Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications

Synthetic Diamond Films

Synthetic Diamond Films Preparation, Electrochemistry, Characterization, and Applications

Edited by

Enric Brillas Carlos Alberto Mart´ınez-Huitle

The Wiley Series on Electrocatalysis and Electrochemistry Series Editor:

Andrzej Wieckowski

A John Wiley & Sons, Inc., Publication

Copyright © 2011 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Synthetic diamond films : preparation, electrochemistry, characterization, and applications / edited by Enric Brillas, Carlos Alberto Mart´ınez-Huitle. p. cm.—(The Wiley series on electrocatalysis and electrochemistry) Includes index. ISBN 978-0-470-48758-7 (cloth) 1. Diamonds—Electric properties. 2. Diamond thin films. I. Brillas, Enric. II. Mart´ınez-Huitle, Carlos Alberto. TK7871.15.D53S96 2011 666 .88—dc22 2010053392 Printed in Singapore oBook ISBN: 978-1-118-06236-4 ePDF ISBN: 978-1-118-06234-0 ePub ISBN: 978-1-118-06235-7 10 9 8 7 6 5 4 3 2 1

Contents

PREFACE

xix

PREFACE TO THE WILEY SERIES ON ELECTROCATALYSIS AND ELECTROCHEMISTRY

xxiii

CONTRIBUTORS

xxv

PART I

SYNTHESIS OF DIAMOND FILMS

1. Electrochemistry on Diamond: History and Current Status

1 3

John C. Angus

1.1

Enabling Technologies / 3 1.1.1 Chemical Vapor Deposition of Diamond / 3 1.1.2 Doping of Diamond / 4 1.1.3 Surface Characterization of Diamond / 5

1.2

First Studies of the Electrochemistry on Diamond / 5 1.2.1 From 1987 to 1996 / 5 1.2.2 From 1996 to Present / 6

1.3

Development of Electrochemical Applications of Diamond / 8 1.3.1 Surface Functionalization / 8 1.3.2 Destruction of Wastes / 9 1.3.3 Sensors and Electroanalysis / 9

1.4

Other Directions / 10 1.4.1 Biolectronic Applications / 10 1.4.2 Anomalous Surface Conductivity of Diamond / 11 v

vi

CONTENTS

1.5

Conclusions / 13 References / 13

2. Synthesis of Diamond Films

21

Vadali V. S. S. Srikanth and Xin Jiang

2.1

Introduction / 21

2.2

Diamond Film CVD Techniques / 23 2.2.1 History / 23 2.2.2 Thermal Decomposition Techniques / 25 2.2.2.1 Hot Filament Chemical Vapor Deposition (HFCVD) / 25 2.2.2.2 Oxy-Acetylene Torch Method / 25 2.2.3 Plasma-Aided Deposition Techniques / 26 2.2.3.1 Microwave Plasma-Enhanced CVD (MWCVD) / 26 2.2.3.2 DC Plasma CVD / 27 2.2.3.3 RF Plasma CVD / 28 2.2.3.4 Electron Cyclotron Resonance Microwave Plasma-Assisted CVD / 28

2.3

Diamond Nucleation and Growth / 28 2.3.1 Nucleation / 28 2.3.1.1 Definition and Types / 28 2.3.1.2 Methods / 30 2.3.2 Growth / 32 2.3.3 Role of Hydrogen and Oxygen / 39

2.4

Diamond Epitaxy / 40

2.5

Nanodiamond Thin Films / 45

2.6

Diamond Nanocomposite Films / 46

2.7

Conclusions / 48 References / 48

3. Types of Conducting Diamond Materials and Their Properties Marco A. Quiroz and Erick R. Bandala

3.1

Introduction / 57

3.2

Conducting Diamond Materials (CDMs) / 62

3.3

CDM Preparation Procedures / 63

3.4

CDM Doping Materials / 63 3.4.1 Characteristics of Boron-Doped CDMs / 63 3.4.2 Electrochemical Properties / 64

57

vii

CONTENTS

3.4.3 3.4.4 3.4.5 3.4.6

Photoelectrochemical Properties / 66 Optical Spectroscopy Properties / 67 Photo- and Cathodoluminescence Properties / 68 Electrical Conductivity and Superconductivity Properties / 69

3.5

Non-Boron-Doped CDMs / 69

3.6

Conclusions / 71 References / 71

PART II ELECTROCHEMISTRY OF DIAMOND FILMS

77

4. Electrochemistry of Diamond

79

Yuri Pleskov

4.1

Introduction / 79

4.2

Principal Electrochemical Properties of Diamond / 80

4.3

The Effect of Semiconductor Nature of Diamond on its Electrochemical Behavior / 83

4.4

The Effect of Crystal Structure on the Electrochemical Behavior of Diamond / 92 4.4.1 The Effect of Crystallographic Orientation of Crystal Faces / 92 4.4.2 The Effect of Surface Morphology / 95 4.4.3 The Effect of the Diamond Grain Size (or the Film Thickness, or the sp 2 -Carbon Impurity) / 98

4.5

Diamond-Based Nanostructures as Electrode Materials: Vacuum-Annealed Undoped Polycrystalline Diamond / 102

4.6

Conclusions / 106

4.7

Acknowledgments / 106 References / 106

5. Applications of Polycrystalline and Modified Functional Diamond Electrodes Yasuaki Einaga and Akira Fujishima

5.1

Introduction / 109

5.2

Preparation of BDD Electrodes / 110

5.3

Electrochemical Properties of BDD as Electrode Materials / 111

5.4

Applications in Electrochemical Analysis Using Polycrystalline BDD electrodes / 111 5.4.1 Detection of Free Chlorine / 111

109

viii

CONTENTS

5.4.2 5.4.3

Detection of Oxalic Acid / 113 Proteins (Including Cancer Markers) / 113

5.5

Modified Functional BDD Electrodes / 116 5.5.1 Production of High-Concentration Ozone-Water Using Free-Standing Perforated Diamond / 116 5.5.2 Modified Functional BDD Electrodes for Electrochemical Analysis / 119 5.5.2.1 Ion-Implanted BDD Electrodes / 119 5.5.2.2 Selective Detection of As(III) and As(V) by Stripping Voltammetry / 124 5.5.2.3 In vivo Dopamine Detection by BDD Microelectrodes / 125 5.5.2.4 BDD Nanograss Array (Whisker BDD) / 126

5.6

Conclusions / 130

5.7

Acknowledgments / 130 References / 131

6. Diamond Ultramicroelectrodes and Nanostructured Electrodes

133

Katherine B. Holt

6.1

Introduction / 133

6.2

Ultramicroelectrodes: Definition and Electrochemical Characteristics / 134

6.3

Boron-Doped Diamond UMEs / 136 6.3.1 Substrate Preparation and Growth of Diamond Films / 136 6.3.2 Insulation Methods and Control of Exposed Electrode Geometry / 140 6.3.3 Electrochemical Performance and Applications / 142

6.4

Boron-Doped Diamond UME Arrays / 143 6.4.1 Fabrication of BDD UME Arrays / 144 6.4.2 Electrochemical Performance and Applications / 146

6.5

Nanostructured BDD Electrodes / 147 6.5.1 Random Array BDD Nanodisk Electrodes / 147 6.5.2 Fabrication of Nanostructured BDD Arrays / 148 6.5.3 Electrochemical Performance and Applications of Nanostructured BDD Electrodes / 149

6.6

Conclusions and Future Directions / 150 References / 151

CONTENTS

ix

PART III ELECTROANALYTICAL APPLICATIONS

153

7. Electroanalytical Applications of Diamond Films

155

Weena Siangproh, Amara Apilux, Pimkwan Chantarateepra, and Orawon Chailapakul

7.1

Introduction / 155

7.2

Pharmaceutical Compounds / 156

7.3

Biomolecules or Biological Compounds / 159

7.4

Pollutant Compounds / 162

7.5

Heavy Metals / 165

7.6

Food and Dietary Contaminants / 166

7.7

Miscellaneous / 168

7.8

Conclusions / 170

7.9

Acknowledgments / 178 References / 178

8. Cathodic Pretreatment of Boron-Doped Diamond Electrodes and Their Use in Electroanalysis Leonardo S. Andrade, Giancarlo R. Salazar-Banda, Romeu C. Rocha-Filho, and Orlando Fatibello-Filho

8.1

Introduction / 181

8.2

Cathodic Pretreatment of Conductive Diamond Films / 182

8.3

Electroanalytical Applications / 192 8.3.1 General Aspects / 192 8.3.2 Determination of Pesticides in Environmental Samples / 193 8.3.2.1 Carbaryl / 193 8.3.2.2 4-Nitrophenol / 193 8.3.2.3 Chlorophenols / 196 8.3.3 Determination of Substances in Food Samples / 198 8.3.3.1 Aspartame / 198 8.3.3.2 Sodium Cyclamate / 199 8.3.3.3 Aspartame and Sodium Cyclamate / 200 8.3.3.4 Total Phenols / 201 8.3.4 Determination of Substances in Pharmaceutical Samples / 201 8.3.4.1 Sulfamethoxazole and Trimethoprim / 201 8.3.4.2 Sulfamethoxazole and Sulfadiazine / 205

181

x

CONTENTS

8.3.4.3 8.3.4.4 8.3.4.5 8.3.4.6

Acetylsalicylic Acid / 205 Paracetamol and Caffeine / 206 Sildenafil Citrate (Viagra®) / 206 Lidocaine / 207

8.4

Gold Deposition and Stripping / 209

8.5

Conclusions / 209 References / 210

PART IV INDUSTRIAL APPLICATIONS 9. Use of Boron-Doped Diamond Electrode in Electrochemical Generation and Applications of Ferrate

213

215

Virender K. Sharma, Enric Brillas, Ignasi Sir´es, and Karel Bouzek

9.1

Introduction / 215

9.2

Electrochemical Generation of the Ferrate Ion with Iron Anodes / 217

9.3

Electrochemical Generation of the Ferrate Ion with Inert Anodes / 222

9.4

Electrochemical Generation of the Ferrate Ion with Boron-Doped Diamond Anode / 223 9.4.1 Acidic Medium / 223 9.4.2 Alkaline Medium / 225

9.5

Applications / 228 9.5.1 Common Inert Anodes / 228 9.5.2 Iron Anodes / 229 9.5.3 BDD Anode / 230

9.6

Conclusions / 233

9.7

Acknowlegments / 233 References / 233

10. Electrochemical Oxidation of Organic Compounds Induced by Electro-Generated Free Hydroxyl Radicals on BDD Electrodes Agnieszka Kapałka, Helmut Baltruschat, and Christos Comninellis

10.1

Introduction / 237

10.2

Influence of Anode Material on the Reactivity of Electrolytic Hydroxyl Radicals / 238

10.3

Electro-Generation and Detection of Quasi-Free Hydroxyl Radicals on BDD Electrode / 240

237

xi

CONTENTS

10.3.1 10.3.2 10.3.3 10.3.4

Hydroxyl Radicals Spin Trapping / 240 Trapping by Salicylic Acid / 240 Competitive Reactions / 242 Formation of Hydrogen Peroxide / 242

10.4

Concentration Profile of Hydroxyl Radicals on BDD Electrode / 244 10.4.1 HO • Concentration Profile during Oxygen Evolution / 244 10.4.2 HO • Concentration Profile during Electro-Oxidation of Organic Compound / 246

10.5

Kinetic Model of Organics Oxidation on BDD Anode / 248 10.5.1 Electrolysis under Current Limited Control (japplied < jlim ) / 249 10.5.2 Electrolysis under Mass Transport Control (japplied > jlim ) / 251

10.6

Electrochemically Induced Mineralization of Organic Compounds by Molecular Oxygen / 253

10.7

Conclusions / 256

10.8

Exercises / 256 10.8.1 Solutions / 257 References / 260

11. Modeling of Electrochemical Process for Water Treatment Using Diamond Films

261

Onofrio Scialdone and Alessandro Galia

11.1

Introduction / 261

11.2

Theoretical Models / 263 11.2.1 General Considerations / 263 11.2.2 Oxidation of Organic Pollutants in Water at BDD by Means of Direct Anodic Oxidation or Reaction with Electro-Generated Hydroxyl Radicals (“Direct Processes”) / 265 11.2.2.1 The Model of Comninellis and Coauthors / 268 11.2.2.2 The Theoretical Works of Polcaro and Coauthors / 272 11.2.2.3 The Approach Proposed by Rodrigo and Coauthors / 273 11.2.3 Oxidation of Organic Pollutants in Water by Means of Electro-Generated Oxidants (“Indirect Processes”) Such as Active Chlorine / 274

11.3

Conclusions / 278

11.4

Acknowledgments / 279 References / 279

xii

CONTENTS

12. Production of Strong Oxidizing Substances with BDD Anodes

281

Ana S´anchez-Carretero, Cristina S´aez, Pablo Ca˜nizares, and Manuel A. Rodrigo

12.1

Electrolyses with Conductive-Diamond Anodes / 281

12.2

Production and Storage of Oxidizing Substances: Experimental Setups / 283

12.3

Production of Hydroxyl Radicals with Conductive-Diamond Anodes / 284

12.4

Synthesis of Peroxoacids and Peroxosalts / 288 12.4.1 Peroxosulphuric Acids / 288 12.4.2 Peroxodiphosphate Salts / 292 12.4.3 Monoperoxophosphoric Acid / 296

12.5

Synthesis of Halogen Oxoanions / 300 12.5.1 Perchlorates / 300 12.5.2 Perbromates / 300

12.6

Synthesis of Ferrates / 301

12.7

Effect of the Type of Diamond on the Efficiency of the Production of Oxidants / 305

12.8

Conclusions / 307

12.9

Acknowledgments / 308 References / 308

13. Ozone Generation Using Boron-Doped Diamond Electrodes

311

Yunny Meas, Luis A. Godinez, and Erika Bustos

13.1

Introduction / 311

13.2

Ozone 13.2.1 13.2.2 13.2.3 13.2.4

13.3

Technologies for Producing Ozone / 317 13.3.1 Corona Discharge Technique / 317 13.3.2 Electrical Discharge Ozone Generators (EDOGs) / 319 13.3.3 Electrolytic Ozone Generators (ELOGs) / 319 13.3.3.1 Anodes for Electrochemically Producing Ozone / 320 13.3.3.2 Boron-Doped Diamond (BDD) / 323

13.4

Reaction Mechanism for the Production of Ozone with Boron-Doped Diamond / 325

/ 311 Physical and Chemical Properties of Ozone / 312 Production of Ozone / 313 Importance of Ozone Applications / 313 Efficiency and Production / 315

CONTENTS

13.5

xiii

Conclusions / 326 References / 327

14. Application of Synthetic Diamond Films to Electro-Oxidation Processes

333

Marco Panizza

14.1

Introduction / 333

14.2

Application in Wastewater Treatment / 335 14.2.1 Oxidation in the Potential Region before Oxygen Evolution / 335 14.2.2 Oxidation in the Potential Region of Oxygen Evolution / 339 14.2.3 Influence of the Nature of Organic Pollutants / 342 14.2.4 Influence of the Concentrations of Organic Compounds / 343 14.2.5 Influence of the Applied Current Density / 343 14.2.6 Influence of the Flow Rate / 345 14.2.7 Influence of the Temperature / 345 14.2.8 Comparison with Other Electrode Materials / 346

14.3

Application in Organic Electrosynthesis / 347

14.4

Conclusions / 348 References / 349

15. Fabrication and Application of Ti/BDD for Wastewater Treatment

353

Xueming Chen and Guohua Chen

15.1

Fabrication of Stable Ti/BBD Electrodes / 353 15.1.1 Introduction / 353 15.1.2 HFCVD Facility / 354 15.1.3 HFCVD Parameter Optimization / 354 15.1.4 Reactive Gas Component Improvement / 357 15.1.5 Methods to Enhance the Service Life of Ti/BDD / 363

15.2

Use of 15.2.1 15.2.2 15.2.3

15.3

Conclusions / 369

Ti/BDD Electrodes for Wastewater Treatment / 365 Oxidation of Acetic and Maleic Acids / 365 Oxidation of Phenol / 365 Oxidation of Dyes / 366

References / 369 16. Application of Diamond Films to Water Disinfection Jessica H. Bezerra Rocha and Carlos A. Mart´ınez-Huitle

16.1

Introduction / 373

373

xiv

CONTENTS

16.2

Disinfection Water / 374

16.3

Science and Technology for Water Purification / 375

16.4

Electrochemical Disinfection/Purification Systems / 376

16.5

Diamond Films for Drinking Water Disinfection / 384

16.6

Production of Inorganic Disinfection by-Products and Inorganic Species Elimination / 388 16.6.1 Chloride, Chlorite, and Chlorate Ions / 389 16.6.2 Perchlorate in Drinking Water / 392 16.6.3 Electrolysis of Nitrates / 394

16.7

Electrochemical Free-Chlorine Systems Using Diamond Films / 396

16.8

Conclusions / 400 References / 400

17. Fenton-Electrochemical Treatment of Wastewaters for the Oxidation of Organic Pollutants Using BDD

405

Enric Brillas

17.1

Introduction / 405

17.2

Fundamentals of Fenton’s Electrochemistry / 406

17.3

Electrogeneration of H2 O2 and Regeneration of Fe2+ / 409

17.4

Degradation of Organics in BDD/O2 Tank Reactors / 413 17.4.1 Herbicides / 414 17.4.2 Dyes / 417 17.4.3 Pharmaceuticals and Amino Acids Precursors / 420

17.5

Degradation of Organics in others Tank Reactors with a BBD Anode / 426

17.6

Degradation of Organics in Batch Recirculation BDD/O2 Flow Cells / 427

17.7

Conclusions / 433 References / 433

18. Electrochemical Energy Storage and Energy Conversion Systems with Diamond Films Juan M. Peralta-Hern´andez, Aracely Hern´andez-Ram´ırez, Jorge L. Guzm´an-Mar, Laura Hinojosa-Reyes, Giancarlo R. Salazar-Banda, and Carlos A. Mart´ınez-Huitle

18.1

Introduction / 437

18.2

Different Techniques Used to Modify BDD Films / 438 18.2.1 Microemulsion Synthesis / 438

437

xv

CONTENTS

18.2.2 18.2.3 18.2.4

Thermal Deposition / 443 Electrodeposition / 446 18.2.3.1 Electrodeposition of Metal Particles on BDD / 449 Sol-Gel Modification / 453

18.3

Application of Modified BDD Films as Electrocatalytic Surfaces for Fuel Cells / 459

18.4

Application of BDD Films in Batteries / 466

18.5

Application of BDD Electrodes as Electrochemical Capacitors / 474

18.6

Conclusions / 477 References / 478

19. Use of Diamond Films in Organic Electrosynthesis

483

Siegfried R. Waldvogel, Axel Kirste, and Stamo Mentizi

19.1

Introduction / 483

19.2

Specific Features of BDD Electrodes / 485

19.3

Stability of BDD Electrodes in Organic Media / 487

19.4

Electrolysis Cells for BDD Electrodes for Organic Transformations / 489

19.5

Anodic 19.5.1 19.5.2 19.5.3 19.5.4 19.5.5 19.5.6

19.6

Cathodic Synthesis on BDD Electrodes / 504 19.6.1 Reduction of Oximes / 504 19.6.2 Reductive Carboxylation / 505

19.7

Conclusions / 506

19.8

Acknowledgment / 506

Transformations on BBD Electrodes / 491 Alkoxylation Reactions / 491 Fluorination Reactions / 493 Cyanation Reactions / 494 Cleavage of C,C-Bonds / 495 Oxidation of Activated Carbon Atoms / 496 Anodic Phenol Coupling Reaction / 496 19.5.6.1 Anodic Homo-Coupling of Phenolic Substrates / 497 19.5.6.2 Nonsymmetrical Phenol Coupling and Phenol-Arene Cross-Coupling Reaction / 499

References / 507

xvi

CONTENTS

PART V BIOELECTROCHEMICAL APPLICATIONS

511

20. Diamond Sensors for Neurochemistry

513

Bhavik Anil Patel

20.1

Introduction / 513

20.2

Central and Peripheral Nervous System / 513

20.3

The Process of Neurotransmission / 514 20.3.1 Neurotransmitters / 516

20.4

Electroanalytical Methods to Study Neurotransmitter Release / 517 20.4.1 Sensors Utilized / 519

20.5

Limitations of Current Techniques for In Vitro and In Vivo Monitoring / 520 20.5.1 Long-Term Recordings / 521 20.5.2 Fouling from Large Biomolecules / 523 20.5.3 Fouling from Redox Reaction By-Products / 525

20.6

Applications of Diamond Sensors and Devices in Neurochemistry / 529 20.6.1 Recording Neuronal Activity / 529 20.6.2 Single Cell Measurements of Vesicular Release / 530 20.6.3 Neurotransmitter Release from Sympathetic Nerves Innervating Mesenteric Arteries / 531 20.6.4 Measuring Transmitter Release from the Gastrointestinal Tract / 533 20.6.4.1 Detection of Histamine Release from Enterochromaffin-Like Cells Located in the Stomach / 534 20.6.4.2 Monitoring Serotonin Release from Enterochromaffin Cells Located in the Mucosa / 535 20.6.4.3 Monitoring Nitric Oxide Release from Myenteric Plexus Neurons / 537 20.6.5 Studying the Neurotransmitter Clearance Process / 538 20.6.5.1 Measurements of Multiple Transmitters from Brain Synaptosomes / 539 20.6.5.2 Investigation of Serotonine Clearance by Transporters Present on Lymphocytes / 540 20.6.6 In vitro and In vivo Measurements from the Central Nervous System / 541 20.6.6.1 In vitro Measurements / 541 20.6.6.2 In vivo Measurements from Anesthetized Animals / 542

20.7

Conclusions and Outlook for the Future / 543

CONTENTS

20.8

xvii

Acknowledgments / 544 References / 544

21. DNA-Modified Diamond Films

551

Nianjun Yang and Christoph E. Nebel

21.1

Introduction / 551

21.2

Diamond Transducer Properties / 558 21.2.1 CVD Diamond Growth / 558 21.2.2 Surface Terminations / 562 21.2.3 Diamond Nanotexture and Wire Formation / 564

21.3

Surface Modification of Diamond / 571 21.3.1 Photochemical Surface Modification of Intrinsic Diamond / 571 21.3.2 Electrochemical Surface Functionalization of Boron-Doped Diamond / 582 21.3.3 Tip Functionalization of Diamond Nanotextures / 589

21.4

DNA Molecules on Diamond / 593 21.4.1 DNA Attachment / 593 21.4.2 Characterizations of DNA Layers / 595

21.5

Sensing 21.5.1 21.5.2 21.5.3

21.6

Summary and Outlook / 612

21.7

Acknowledgments / 614

of DNA Hybridization / 602 DNA Field Effect Transistor / 602 Cyclic Voltammetry and Impedance Spectroscopy / 606 DNA Sensing on Nanotextured Diamond Surfaces / 609

References / 614 INDEX

621

Preface

Diamond is an extremely hard crystalline form of carbon and it is considered an excellent material for many applications due to its unusual physical and chemical properties. For this reason, it has long attracted the attention of scientists and the public. Interest in diamond has been further increased by the discovery of the possibility to produce polycrystalline diamond films with mechanical and electronic properties comparable with natural diamond. Over the last few years, the number of publications has increased considerably regarding the synthesis and/or applications of this new material. Currently, synthetic diamond films have been the subject of applications and fundamental research in several fields of the science. Much effort was spent during the 1960s and 1970s to investigate diamond synthesis until it was successfully achieved by using the chemical vapor deposition (CVD) technique with excellent diamond growth rates, which led to good prospects for films being used for some industrial applications. Since their introduction into electrochemical research in 1987, doped-diamond electrodes have become more and more popular. This is based on their unique properties that distinguish them from conventional electrode materials and make many electrochemical processes more attractive or even possible. These electrodes, then, have been the subject of a large variety of applications and fundamental research in electrochemistry, opening up a novel branch known as the “electrochemistry of synthetic diamond films.” Almost every aspect of electrochemistry has been impacted by the diamond electrode, from instrumental analysis to industrial applications. Electrically conductive films of boron-doped diamond (BDD) have gained increasing popularity in many electrochemical applications, in large part due to the fact that very high quality films possess background currents that can be some orders of magnitude lower than those other types of electrode materials. Other important properties of these electrodes are related to their large potential window, low adsorption, corrosion stability in very aggressive media, high efficiency in oxidation processes, and very low doublelayer capacitance. Therefore, diamond films have become suitable materials for several purposes classified in different areas: synthesis of chemicals, modification of diamond surfaces, electroanalysis, water disinfection, destruction of pollutants in waters, and so xix

xx

PREFACE

on. The versatility of these materials has also been extended to develop sensors, microelectrodes, nanoelectrodes, and biosensors. Recently, synthetic diamond is starting to be commercialized for practical purposes—for example, diamond electrochemical detectors for liquid chromatography and large-scale diamond films for industrial wastewater treatment. The future for the synthetic diamond films is bright. These materials are starting to make important contributions for the measurements and understanding of an extensive range of chemical processes. Several other complementary techniques are emerging and can provide new knowledge for the chemistry of materials. Future interdisciplinary developments based on the close collaboration of chemists, electrochemists, engineers, biologists, and neuroscientists can be envisaged to ensure an effective application and productive use of synthetic diamonds to answer important chemical and medical questions and to resolve vital environmental problems. Several areas of interest have been opened up by these developments, including the measurement of release from single neurons, new sensors that are smaller and yet faster responding, sensor arrays for simultaneously detecting several analytes, novel biosensors for new neurochemical species of interest, and medical and clinical applications to neurochemistry and neuroscience. Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications is a timely book that has gathered together the best international experts of the electrochemical community and who have imagined a large number of approaches to investigate the electrochemical properties of diamond films and their characteristics. All of these contributing experts now focus on aspects that promote efficiency, selectivity, and high performances for a broad variety of laboratory and industrial diamond applications with goals ranging from organic synthesis to environmental problems. The first part of the book (Chapters 1 to 3) deals with diamond history, emphasizing the discovery of synthetic diamond films and types of them. The second part (Chapters 4 to 6) concerns the beginning and new branches of the electrochemistry of diamond films. Chapters 7 and 8, which make up the third part of the book, summarize the studies of the diamond films on electroanalytical applications. The chapters focus on the use of these materials as sensors for detecting organic or inorganic species in order to propose the electroanalysis as an alternative quality control technique as well as for monitoring chemical species on air, soil, and water ecosystems. The fourth part (Chapters 9 to 19) is devoted to industrial applications of diamond films describing specific problems of particular interest for organic chemistry, energy conversion (fuel cells and batteries), and environmental aspects. Since diamond films are the most efficient materials for water treatment and water disinfection, Chapters 9 to 17 are devoted to these industrial applications in order to show the properties of these materials for environmental protection. The electrochemical oxidation with diamond films has been recognized as an electrochemical advanced oxidation process (EAOP), whereas the recent application of synthetic diamond films to other emerging EAOPs, such as electro-Fenton and photoelectro-Fenton, has also opened new prospects for wastewater remediation. The last part of the book (Chapters 20 to 21) focuses on innovative applications of diamond films, ranging from novel biosensors for chemical species of interest, as well as medical and clinical applications to neurochemistry and neuroscience by diamond microelectrodes and nanoelectrodes.

PREFACE

xxi

We strongly believe that this book will greatly promote research in this key field in the forthcoming years for the benefit of our society. Carlos A. Mart´ınez-Huitle Enric Brillas Editors

Preface to the Wiley Series on Electrocatalysis and Electrochemistry

This series covers recent advances in electrocatalysis and electrochemistry and depicts prospects for their contribution into the present and future of the industrial world. It aims to illustrate the transition of electrochemical sciences from its beginnings as a solid chapter of physical chemistry (covering mainly electron transfer reactions, concepts of electrode potentials and structure of electrical double layer) to the field in which electrochemical reactivity is shown as a unique chapter of heterogeneous catalysis, is supported by high-level theory, connects to other areas of science, and includes focus on electrode surface structure, reaction environment and interfacial spectroscopy. The scope of this series ranges from electrocatalysis (practice, theory, relevance to fuel cell science and technology) to electrochemical charge transfer reactions, biocatalysis and photoelectrochemistry. While individual volumes may appear quite diverse, the series promises updated and overall synergistic reports providing insights to help further our understanding of the properties of electrified solid/liquid systems. Readers of the series will also find strong reference to theoretical approaches for predicting electrocatalytic reactivity by such high-level theories as density functional theory. Beyond the theoretical perspective, further vehicles for growth are such significant topics such as energy storage, syntheses of catalytic materials via rational design, nanometer-scale technologies, prospects in electrosynthesis, new instrumentation and surface modifications. In this context, the reader will notice that new methods being developed for one field may be readily adapted for application in another. Electrochemistry and electrocatalysis have both benefited from numerous monographs and review articles due to their depth, complexity, and relevance to the practical world. The Wiley Series on Electrocatalysis and Electrochemistry is dedicated to present the current activity by focusing each volume on a specific topic that is timely and promising xxiii

xxiv

PREFACE TO THE WILEY SERIES ON ELECTROCATALYSIS AND ELECTROCHEMISTRY

in terms of its potential toward useful science and technology. The chapters in these volumes will also demonstrate the connection of electrochemistry to other disciplines beyond chemistry and chemical engineering, such as physics, quantum mechanics, surface science, and biology. The integral goal is to offer a broad-based analysis of the total development of the fields. The progress of the series will provide a global definition of what electrocatalysis and electrochemistry are now, and will contain projections about how these fields will further evolve in time. The purpose is twofold, to provide a modern reference for graduate instruction and for active researchers in the two disciplines, as well as to document that electrocatalysis and electrochemistry are dynamic fields that are expanding rapidly, and are likewise rapidly changing in their scientific profiles and potential. Creation of each volume required the editors involvement, vision, enthusiasm and time. The Series Editor thanks each Volume Editor who graciously accepted his invitation. Special thanks go to Ms. Anita Lekhwani, the Series Acquisitions Editor, who extended the invitation to edit this series to me and has been a wonderful help in its assembling process. Andrzej Wieckowski Series Editor

Contributors

Leonardo Santos Andrade, Departamento de Qu´ımica, Universidade Federal de Goi´as, Campus de Catal˜ao, Avenida Lamartine P. Avelar 1120, 75704-020 Catal˜ao, GO, Brazil John C. Angus, Department of Chemical Engineering, Case Western Reserve University, Cleveland, OH 44106-7217, USA Amara Apilux, Sensor Research Unit, Department of Chemistry, Faculty of Science, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand Helmut Baltruschat, Institute for Physical and Theoretical Chemistry, Universit¨at Bonn, D 53117 Bonn, Germany ´ Erick Roberto Bandala Gonzalez, Universidad de las Am´ericas-Puebla, Departamento de Ingenier´ıa Civil y Ambiental Grupo de Investigaci´on en Energ´ıa y Ambiente, Sta. Catarina M´artir, Cholula-Puebla, M´exico J´essica Horacina Bezerra Rocha, Centro de Ciˆencias Exatas e da Terra, Departamento de Qu´ımica, Universidade Federal do Rio Grande do Norte, Campus Universit´ario-Lagoa Nova, CEP 59.072-970, Natal/RN, Brazil Karel Bouzek, Department of Inorganic Technology, Institute of Chemical Technology Prague, Prague, Czech Republic Enric Brillas, Laboratori d’Electroqu´ımica dels Materials i del Medi Ambient, Facultat de Qu´ımica, Departament de Qu´ımica F´ısica, Universitat de Barcelona, Mart´ı i Franqu`es 1-11 08028 Barcelona, Spain Erika Bustos, Centro de Investigaci´on y Desarrollo Tecnol´ogico en Electroqu´ımica, Parque Tecnol´ogico Quer´etaro, Sanfandila, C.P. 76703, Pedro Escobedo, Edo. de Quer´etaro, M´exico ˜ Pablo Canizares, Department of Chemical Engineering, Faculty of Chemical Sciences, Enrique Costa Building, Universidad de Castilla La Mancha, Campus Universitario s/n 13071 Ciudad Real, Spain xxv

xxvi

CONTRIBUTORS

Orawon Chailapakul, Department of Chemistry, Sensor Research Unit and Center for Petroleum, Petrochemicals, and Advanced Materials, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand Pimkwan Chantarateepra, Program in Biotechnology, Faculty of Science, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand Guohua Chen, Department of Chemical and Biomolecular Engineering, Hong Kong University Science & Technology, Clean Water Bay, Kowloon, Hong Kong, China Xueming Chen, Department of Environmental Engineering, Zhejiang University, 388 Yuhangtang Road, Hangzhou 310058, China Christos Comninellis, Ecole Polytechnique F´ed´erale, Lausanne, Group of Electrochemical Engineering, EPFL, 1015 Lausanne, Switzerland Yasuaki Einaga, Department of Chemistry, Keio University, 3-14-1 Hiyoshi, Yokohama 223-8522, Japan Orlando Fatibello-Filho, Departamento de Qu´ımica, Universidade Federal de S˜ao Carlos, C.P. 676, 13560-970 S˜ao Carlos-SP, Brazil Akira Fujishima,

President, Tokyo University of Science, Kagurazaka, Tokyo, Japan

Lu´ıs A. God´ınez, Centro de Investigaci´on y Desarrollo Tecnol´ogico en Electroqu´ımica, Parque Tecnol´ogico Quer´etaro, Sanfandila, C.P. 76703, Pedro Escobedo, Edo. de Quer´etaro, M´exico ´ Jos´e L. Guzman-Mar, Facultad de Ciencias Qu´ımicas, Centro de Laboratorios Especializados, Universidad de Nuevo Le´on, Pedro de Alba s/n, Cd. Universitaria, San Nicol´as de los Garza, NL, M´exico ´ırez, Facultad de Ciencias Qu´ımicas, Centro de Laborato´ Aracely Hernandez-Ram rios Especializados, Universidad de Nuevo Le´on, Pedro de Alba s/n, Cd. Universitaria, San Nicol´as de los Garza, NL, M´exico Lu´ıs Hinojosa-Reyes, Facultad de Ciencias Qu´ımicas, Centro de Laboratorios Especializados, Universidad de Nuevo Le´on, Pedro de Alba s/n, Cd. Universitaria, San Nicol´as de los Garza, NL, M´exico Katherine Holt, Department of Chemistry, University College London, Christopher Ingold Building, 20, Gordon St., London, WC1H 0AJ, UK Xin Jiang, Institute of Materials Engineering, University of Siegen, Paul-Bonatz-Str. 9-11, 57076 Siegen, Germany Agnieszka Kapalka, Ecole Polytechnique F´ed´erale, Lausanne, Group of Electrochemical Engineering, EPFL, 1015 Lausanne, Switzerland Axel Kirste, Kekul´e-Institut f¨ur Organische Chemie und Biochemie, Rheinische Friedrich-Wilhelms Universit¨at Bonn, Gerhard-Domagk-Str. 1, 53121 Bonn, Germany Carlos Alberto Mart´ınez-Huitle, Centro de Ciˆencias Exatas e da Terra, Departamento de Qu´ımica, Universidade Federal do Rio Grande do Norte, Campus Universit´ario-Lagoa Nova, CEP 59.072-970, Natal/RN, Brazil Yunny Meas, Centro de Investigaci´on y Desarrollo Tecnol´ogico en Electroqu´ımica, Parque Tecnol´ogico Quer´etaro, Sanfandila, C.P. 76703, Pedro Escobedo, Edo. de Quer´etaro, M´exico

CONTRIBUTORS

xxvii

Stamo Mentizi, Johannes Gutenberg-Universit¨at Mainz, Institut f¨ur Organische Chemie, Duesbergweg 10-14, 55128 Mainz, Germany Christoph E. Nebel, Fraunhofer-Institute for Applied Solid State Physics (IAF), Department Micro- and Nano-Sensors (GF5), Tullastrasse 72, 79108 Freiburg, Germany Marco Panizza, Department of Chemical and Process Engineering, Universit`a di Genova, P.le J.F. Kennedy 1, 16129 Genoa, Italy Bhavik A. Patel, Centre for Biomedical and Health Sciences Research, School of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton, BN2 4GJ ´ Juan Manuel Peralta-Hernandez, Facultad de Ciencias Qu´ımicas, Centro de Laboratorios Especializados, Universidad de Nuevo Le´on, Pedro de Alba s/n, Cd. Universitaria, San Nicol´as de los Garza, NL, M´exico Yurnny V. Pleskov, Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Leninsky prospekt 31, 119991 Moscow, Russia Marco Antonio Quiroz-Alfaro, Universidad de las Am´ericas Puebla. Departamento de Ciencias Qu´ımico Biol´ogicas, Grupo de Investigaci´on en Energ´ıa y Ambiente, Sta. Catarina M´artir, Cholula-Puebla, M´exico Romeu Cardozo Rocha–Filho, Departamento de Qu´ımica, Universidade Federal de S˜ao Carlos, C.P. 676, 13560-970 S˜ao Carlos-SP, Brazil Manuel A. Rodrigo, Department of Chemical Engineering, Faculty of Chemical Sciences, Enrique Costa Building, Universidad de Castilla La Mancha, Campus Universitario s/n 13071 Ciudad Real, Spain ´ Cristina Saez, Department of Chemical Engineering, Faculty of Chemical Sciences, Enrique Costa Building, Universidad de Castilla La Mancha, Campus Universitario s/n 13071 Ciudad Real, Spain Giancarlo R. Salazar-Banda, Laborat´orio de Eletroqu´ımica e Nanotecnologia, Instituto de Tecnologia e Pesquisa, Universidade Tiradentes, Av. Murilo Dantas 300, Farolˆandia, 49032-490 Aracaju-SE, Brazil ´ Ana Sanchez-Carretero, Department of Chemical Engineering, Faculty of Chemical Sciences, Enrique Costa Building, Universidad de Castilla La Mancha, Campus Universitario s/n 13071 Ciudad Real, Spain Onofrio Scialdone, Dipartimento di Ingegneria Chimica, Gestionale, Meccanica e Informatica, Universit`a di Palermo, Viale delle Scienze, 90100, Palermo, Italy Virender K. Sharma, Florida Institute of Technology, 150 West University Boulevard Melbourne, Fl 32901, USA Weena Siangproh, Department of Chemistry, Faculty of Science, Srinakharinwirot University, Sukhumvit 23, Wattanna, Bangkok 10110, Thailand Ignasi Sir´es, Laboratori d’Electroqu´ımica dels Materials i del Medi Ambient, Facultat de Qu´ımica, Departament de Qu´ımica F´ısica, Universitat de Barcelona, Mart´ı i Franqu`es 1-11 08028 Barcelona, Spain Vadali V.S.S. Srikanth, School of Engineering Sciences and Technology, University of Hyderabad, Central University (P.O.), Hyderabad 500046, India

xxviii

CONTRIBUTORS

Siegfried R. Waldvogel, Institut f¨ur Organische Chemie, Johannes GutenbergUniversit¨at Mainz, Duesbergweg 10-14, 55128 Mainz, Germany Nianjun Yang, Fraunhofer-Institute for Applied Solid State Physics (IAF), Department Micro- and Nano-Sensors (GF5), Tullastrasse 72, 79108 Freiburg, Germany

Current density (mA cm–2)

0.20 (a) 0.15 0.10 (b) 0.05 (c) 0.00 0.2

0.0

0.4

0.6

0.8

Potential (V) vs. Ag/AgCl Figure 5.9 For caption see page 122.

counter electrode

BDD

stimulating electrode Reference electrode

anesthesia induction

Counter electrode

Stimulating electrode reference electrode

BDD

Brain Cortex Corpus Striatum Substantia Nigra Dopaminergic Neuron

Figure 5.14 For caption see page 127.

(c)

(b)

(d) 75 nA

4.0 3.0 2.0 1.0

0 nA

0.0

10 0 –10 0

20 40 60 Distance (μm)

80

Figure 6.3 For caption see page 145.

Current density(A m–2)

Anodic potential (V) vs. SCE

20

2.0

65

2.2

90

2.5

155

3.5

195

4.1

0.04 Thermodynamics

0.03

Pt

0.02 j (A cm–2)

Height (nm)

it/i(∞)

DDB

0.01 0

–0.01 –0.02

–2.0

–1.0

0.0

1.0

2.0

3.0

E (V) vs. NHE

–0.03 Figure 12.1 For caption see page 282.

Power supply V

A

catholyte

anolyte +

–

Electrochemical cell

Membrane Anode (BDD)

Cathode (AISI 304)

Out

Out

In

In

Figure 12.7 For caption see page 286.

(a)

(b)

(c)

(d)

μm 5

μm 5 4

4 3

3 2

2 1

1

Figure 16.16 For caption see page 398.

Figure 19.5 For caption see page 488.

Platelets Figure 20.6 For caption see page 523.

sensor

sensor

sensor

sensor

sensor BLOOD

Red blood cells

Activated platelets

Protein

Fibrin

Before stimulation

(a)

(b)

After stimulation at 10 Hz

Stimulator 210 µm

210 µm

Diamond Microelectrode (c) Imax = 10.4 pA

Stimulation Current (pA)

Diameter (µm) 30 µm

Figure 20.13 For caption see page 532.

Ileum–cross section

Submucosal plexus neurons EC Cells

Myenteric plexus neurons

Circular Muscle

Longitudinal Muscle

Enterochromaffin cells

Mucosal layer

SERT transporters present in enterocytes

Figure 20.14 For caption see page 533.

Myenteric plexus

(a)

Gastric Pit

(b) 8

Gastric glands Parietal cell

Current / nA

6

4

2 ECL-cell 0 0

100

200

300

500

400

Time (s)

Figure 20.15 For caption see page 534.

Figure 20.20 For caption see page 541. Au

Current (mA cm−2)

Pt Glassy Carbon

(H)SCD B:PCD (USU)

B:PCD (NRL)

B:(H)SCD –3

–2

–1

0

1

Potential (V) vs. SCE

Figure 21.1 For caption see page 552.

2

3

600

3

–1 –2

0

2

–3

CdS

–χ

–4.2

pH2 = 1 bar 1 mbar 1 μbar

0

2

–4.0

μe (eV)

1

μ

EVAC

–1

–4.6 6 8 10 12 14 pH-Value

4

–2 –3

CdSe

μ

–5 –6 GaP

SiC

–7

GaAs Si

–4 –5

H-terminated Diamond

Ge

–6

Diamond

–7

Semiconductors Figure 21.3 For caption see page 554.

1500 Diamond 1000

Gold Silicon

500 Glassy Carbon

0

0

5

1 0

–4.4

–4

Fluorescence intensity (a.u.)

Energy rel. to vacuum level (eV)

2

3

–3.8

Hydrogen Term Diamond EVBM

10

15

20

Hybridization cycles Figure 21.5 For caption see page 555.

25

30

(I)

(II)

Nanocrystalline Diamond

(a)

1 μm

1 μm

Silicon Polycrystalline Diamond

(b)

200 μm 200 μm Substrate Side

(c)

CL Intensity (a.u.)

4 x 106

FE(TO)

T = 16 K E = 13 kV I = 2 μA

3 x 106

2 x 106

1 x 106

0 5.0

1.00

0.5 nm

0.75

0.3 nm

0.50

0.0 nm

0.25 FE(LO) FE(TA) FE(TO + OΓ) 5.1 5.2 5.3 5.4 Photon Energy (eV)

5.5

0

0.25

0.50

Figure 21.6 For caption see page 556.

0.75

0 1.00 μm

(a) Electrochemical nitrophenyl grafting

Nanotextures from Boron-Doped Diamond

(c) Probe DNA immobilization

Nanotextures from Boron-Doped Diamond

(b) Animation & cross-linker attachment

Nanotextures from Boron-Doped Diamond

(d) DNA hybridization & detection

Nanotextures from Boron-Doped Diamond

Figure 21.7 For caption see page 559.

500

400

[nm]

300

200

100

0 0

100

200

300

400

500

[nm] 0.00

0.50

[nm] Figure 21.8 For caption see page 560.

(a)

Current density (mA cm–2)

0.5 (b) (c) 0

–0.5

(d)

–0.5

0

0.5

Potential (V) vs. Ag/AgCl Figure 21.12 For caption see page 564.

1

(c)

(b)

(a)

(d)

Figure 21.13 For caption see page 565.

5

Sensor Area (norm.)

[×1013] 6 (a)

C –2 (F–2)

4

3 (c) 2

1

0 –0.5

2.2 2 1.8 1.6 1.4 1.2 1 0

(d)

5 10 15 Etching time (s)

20

(b)

0

0.5

1 1.5 2 E (V) vs. Ag/AgCl

2.5

Figure 21.16 For caption see page 568.

3

3.5

40

Height / A

32

24

16 0

1000

2000 Distance (nm)

Figure 21.21 For caption see page 573.

Figure 21.24 For caption see page 576.

3000

(

(a)

(

(b)

Figure 21.27 For caption see page 578.

Figure 21.28 For caption see page 579.

100 nN 120 nN >120 nN

6 Height (nm)

5 4 3

26 Å

8Å

2 1 0

0

2000 1000 Distance (nm)

3000

Figure 21.38 For caption see page 588.

Figure 21.41 For caption see page 590.

Figure 21.46 For caption see page 596.

Figure 21.47 For caption see page 597.

(a)

(b)

Figure 21.48 For caption see page 597.

(a) PEEK-Sample Holder (part 1)

(b) PEEK-Sample Holder (part 2, backside)

Figure 21.58 For caption see page 604.

(c) Mounted Sample (part 1 and 2)

NA ds -D

ds -D

ds -D

NA

NA

Pt

Debye Length λD DRAIN

Source

Figure 21.59 For caption see page 604.

(a) Single-stranded (ss) DNA

(b) Double-stranded (ds) DNA

Redox-Molecules diffuse into the ss-DNA film.

ss

Redox-Molecules are repelled by Coulomb Force

ds

Diamond

Diamond

Generation of Redox-Current

No Redox-Current

ss

4

5.0 x 10

0.0 4

–5.0 x 10

–1.0 x 104 –0.4 –0.2

0.0

0.2

0.4

Potential (V vs. Ag/AgCl)

0.6

Current density (A/cm2)

2

Current density (A/cm )

Ferrocyanide (5mM) redox reaction 1.0 x 104

1.0 x 10

4

ds

5.0 x 104 0.0 –5.0 x 10

4

–1.0 x 104 –0.4

–0.2 0.0 0.2 0.4 0.6 Potential (V vs. Ag/AgCl)

Figure 21.62 For caption see page 607.

Part I

Synthesis of Diamond Films

1 Electrochemistry on Diamond: History and Current Status John C. Angus

1.1

ENABLING TECHNOLOGIES

The use of diamond in electrochemistry was not anticipated during the initial explosion of interest in low-pressure diamond growth during the 1980s. Most attention was focused on electronic, optical, and mechanical applications (e.g., for high temperature electronic devices, radiation detectors, high voltage switches, x-ray windows, audio speaker diaphragms, and protective and tool coatings). Little attention was paid to electrochemical applications. In this chapter, we describe the enabling technologies underlying diamond electrochemistry, the early work on diamond electrochemistry itself, and two major threads of current research. The present status of the field is covered in the other chapters of this volume. 1.1.1

Chemical Vapor Deposition of Diamond

Diamond electrochemistry has been made possible by development of methods for the chemical vapor deposition of diamond. The first reported growth of diamond at sub-atmospheric pressure, where diamond is metastable with respect to graphite, was by Eversole [1] at Union Carbide Corporation. He grew diamond on high-surface area powder from methane and carbon monoxide. This effort was ultimately abandoned after Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

3

4

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

the announcement of the successful growth of diamond at high pressures by General Electric Corporation [2]. Eversole’s work was later extended by Angus et al. at Case Western Reserve University [3–7] and by Deryagin and a large group at the Physical Chemistry Institute in Moscow [8–13]. The Angus group showed the beneficial effect of added hydrogen on diamond yield, developed methods for removing unwanted graphitic carbon using atomic hydrogen, and grew boron-doped diamond by chemical vapor deposition [5]. In their first studies, the Soviet workers reported the growth of filamentary diamond whiskers using a vapor-liquid-solid (VLS) technique with molten iron and nickel [8,9]. Surprisingly, this process has received little attention in subsequent years. From the same group at the Physical Chemistry Institute, Varnin et al. [11] grew diamond thin films from the gas phase; Spitsyn, Bouilov, and Deryagin [13] reported both diamond films and isolated diamond crystals grown on copper substrates by chemical vapor transport from a graphite source. These very early studies were the first to show that diamond could be grown at conditions in which it is not the stable phase. At the time, diamond synthesis at low pressures was widely believed to be impossible and in violation of the second law of thermodynamics. Although the early studies showed the feasibility of diamond synthesis by chemical vapor deposition, the growth rates were low. This situation changed dramatically in the mid-1980s. A group at the National Institute for Research in Materials in Tsukuba, Japan, under the leadership of Nobuo Setaka achieved large increases in growth rates by activating the gas phase during growth with a hot filament [14,15], a microwave discharge [16], and an RF discharge [17,18]. Matsui, Matsumoto, and Setaka [19] also performed careful characterization of their product diamond. Since the mid-1980s, the interest in chemical vapor deposition of diamond has expanded enormously, and it is now used to grow extremely high-quality diamond in many shapes and sizes. 1.1.2

Doping of Diamond

For most electrochemical applications, it is necessary to use conducting diamond obtained by doping with the group III element, boron. The effect of boron on the electrical conductivity of diamond has been well studied and is covered in several classic review volumes [20–24]. Poferl, Gardner, and Angus [5] in 1973 were the first to grow boron-doped conducting diamond by chemical vapor deposition. This study was done to provide further confirmation of metastable diamond growth. Since then, numerous studies of diamond doping have been made. The boron-doping agent is usually added through small amounts of diborane, trimethyl boron, or organic borates in the source gases, but solid-state boron sources have also been successfully used. Reviews by Pan and Kania [25] and Werner and Locher [26] describe the state of the art in the mid-1990s; more recent discussions [27–29] are available. At low concentrations, boron promotes p-type semiconductivity in diamond with an acceptor level 0.37 eV above the valence band [30]. At boron concentrations higher than 1017 to 1018 cm−3 , an impurity band forms and the acceptor gap is reduced [31–34]. At boron concentrations from 1020 to 1021 cm−3 , diamond is a semimetal, with resistivities as low as 0.001 -cm. For electrosynthesis and electrodestruction applications, high conductivities are desirable; for photoelectrochemistry and semiconductor electrochemistry, low conductivity electrodes are used. The uniformity of boron concentration is an important variable in electrochemical applications. Spitsyn et al. in 1981 [13] and Janssen et al. in 1990 [35] showed that more boron was incorporated in (111) growth sectors than in (100) sectors.

1.2 FIRST STUDIES OF THE ELECTROCHEMISTRY ON DIAMOND

1.1.3

5

Surface Characterization of Diamond

Diamond, graphene and carbon nanotubes are unique among the common semiconductors in having no stable solid surface oxide. Furthermore, the chemistry of the diamond surface is closely related to well-known organic chemical processes. As a consequence, it is possible to tailor the properties of the diamond surface for sensor and other applications by changing the surface functionalization. The most significant early paper on diamond surfaces was the LEED study by Lander and Morrison at Bell Laboratories in 1966 [36]. They showed that hydrogen-termination of (111) diamond surfaces eliminated the reconstruction of the clean surface and gave an (essentially) bulk-terminated surface. They also found significant atom mobility on the diamond surface between 900◦ C and 1400◦ C. Lander and Morrison explicitly stated that this behavior could make epitaxial extension of diamond feasible. Another important early study of diamond surfaces and their interaction with gases was by Lurie and Wilson in 1977 [37].

1.2

FIRST STUDIES OF THE ELECTROCHEMISTRY ON DIAMOND

Even after the chemical vapor deposition of diamond became more or less routine, there was little availability of boron-doped conducting diamond. Therefore, some of the earliest electrochemical studies done on diamond were performed on samples in which conductivity was induced by damage from ion implantation. In 1983, Iwaki et al. [38] of the Institute of Physical and Chemical Research in Hirosawa, Japan, reported the use of diamond electrodes made conductive by ion implantation of nitrogen, argon, and zinc. However, the ion implantation converted the diamond surface into nondiamond carbon [38,39] so the measurements did not reflect the electrochemical characteristics of diamond itself. 1.2.1

From 1987 to 1996

The first measurements of the inherent electrochemical properties of diamond were by Yuri Pleskov and co-workers at the Frumkin Institute of Electrochemistry in Moscow. In 1987, they reported the photoelectrochemical properties of diamond electrodes [40]. They found a photoresponse of diamond at sub-bandgap wavelengths that they attributed to excitation of electrons from mid-gap defect states to the conduction band. They also observed sluggish evolution of hydrogen at cathodic potentials. Subsequently, in 1989 Natishan and Morrish at the Naval Research Laboratory in Washington, D.C., described the electrochemical properties of composite diamond/molybdenum electrodes in which the diamond acted as an inert insulating material [41]. During the early 1990s, Pleskov and co-workers continued and expanded their earlier work on diamond electrochemistry. They analyzed the impedance of conducting polycrystalline diamond thin films between two ohmic contacts and showed that intercrystalline boundaries are capacitive/resistive barriers in the bulk of diamond [42,43]. This same group continued its photoelectrochemical studies [44,45] and measured the minority carrier diffusion length and the acceptor concentrations in diamond [46]. A large group at the University of Tokyo led by Akira Fujishima made extensive progress on diamond electrochemistry during the early 1990s. Patel, Hashimoto, and

6

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

Fujishima examined the photoresponse of diamond [47,48]. They showed that the diamond electrodes acted as a p-type semiconductor; from the flat band potential, they deduced that the bottom of the conduction band of diamond was near the vacuum level. Tenne and others of this same group exploited the large overpotentials for hydrogen evolution for the reduction of nitrate to ammonia on boron-doped diamond electrodes [49,50]. In 1993 Marchywka et al. [51] described an interesting process in which they used low-energy carbon ion implantation, followed by electrochemical etching of the damaged region, to separate thin diamond layers from bulk diamond crystals. Also in 1993, Ramesham and co-workers [52] deposited diamond films on both glassy carbon and graphite to make a composite electrode that they said “may have some use in electroanalysis”. Miller et al. in 1994 used ion implantation of 140 keV Co ions to form a patterned, nondiamond conductive region on a planar diamond electrode surface [53]. In the early and mid-1990s, Swain and his colleagues published a series of papers [54–58] that demonstrated many of the essential properties of diamond electrodes: their low capacitance, featureless background current, large signal to noise ratio, and chemical stability. They suggested that these properties made diamond electrodes suitable for electroanalysis and sensors. In 1995 a group of researchers at the Shanghai Institute of Metals also reported on the stability and reproducibility of diamond electrodes and noted their potential for sensor applications [59–61]. The large overpotential for hydrogen evolution on diamond was observed in the earliest work [49,50], but the large overpotential for oxygen evolution did not become apparent until high-quality diamond electrodes, without significant nondiamond carbon, were used. In 1995 and 1996, Martin et al. [62], Argoitia et al. [63], and Martin et al. [64] demonstrated that high-quality diamond electrodes did not evolve hydrogen until −1.25 V and oxygen until +2.3 V (versus the standard hydrogen electrode). Bouamrane et al. [65] reported a similar observation in 1996. The high overpotentials for both oxygen and hydrogen evolution on diamond give rise to an extremely wide potential window of water stability that is found on no other electrode material. Figure 1.1 shows the current-voltage characteristics for water electrolysis on high- and low-quality diamond, glassy carbon, and platinum [63]. The wide potential window and the lack of background current within the potential window for high-quality diamond are apparent. 1.2.2

From 1996 to Present

In the latter part of the 1990s, several intensive efforts were made to understand the nature of charge transfer at diamond electrodes and how it differed from conventional carbon electrodes. In 1996, Vinokur et al. [66] performed a careful analysis of electron transfer to both lightly and heavily doped diamond electrodes. Highly reversible electrode behavior was found for heavily boron-doped electrodes, especially at potentials more positive than −0.50 V versus the saturated calomel electrode. Lightly doped, semiconducting electrodes showed very slow kinetics at more negative potentials, which was attributed to a lack of available states in the bandgap of diamond. During this same time period, the Swain [67] group measured the boron distribution on diamond electrodes and Granger et al. [68] presented a detailed study of the electrode kinetics of diamond from two different sources. They examined the difference in electrode kinetics for processes involving outer sphere

1.2 FIRST STUDIES OF THE ELECTROCHEMISTRY ON DIAMOND

(a) Platinum

(b) High-Quality Diamond 30

15 0 −1

−15

1

2

3

i (mA / cm−2)

i (mA / cm−2)

30

−2

15 0 −2

−1

V vs. SHE(V)

1

2

3

i (mA / cm−2)

i (mA / cm−2)

0

V vs. SHE(V)

3

1

2

3

30

15

−30

2

(d) Glassy Carbon

30

−15

1

V vs. SHE(V)

(c) Low-Quality Diamond

−1

−15 −30

−30

−2

7

15 0 −2

−1

−15 −30 V vs. SHE(V)

Figure 1.1 Early comparison of voltammograms of water electrolysis on (a) platinum, (b) high-quality diamond, (c) low-quality diamond, and (d) glassy carbon. The electrolyte was 0.50 M H2 SO4 ; potentials are measured versus the standard hydrogen electrode. The wide potential window between hydrogen evolution at ∼ − 1.5 V and oxygen evolution at ∼ + 2.5 V and lack of background current for the high-quality diamond electrode are evident. Reprinted from [63].

electron transfer and for reactions that go through adsorbed intermediates; they also studied the effect of hydrogen-termination and surface oxidation on the electrode kinetics. In 1999 Martin and associates [69] showed that increasing amounts of sp2 , nondiamond carbon on the diamond surface reduced the potential window of water stability and made the electrodes behave more like glassy carbon or pyrolytic graphite. The results of these and other studies showed the intrinsic differences between diamond and conventional carbon electrodes. Many of these differences arise from the very different types of surface structure and functionalities that are found on diamond electrode surfaces. One of the most striking features is the relatively inert, hydrogen-terminated diamond surface. The high overpotentials for hydrogen and oxygen evolution (and other reactions) are attributed, at least in part, to the limited adsorption of intermediate species on diamond. High overpotentials are not observed for rapid outer-sphere reactions that do not rely on an adsorbed intermediate. However, it should be emphasized that the diamond surface is not completely inert. Anodic polarization and liberation of oxygen changes the surface properties of a previously hydrogen-terminated surface [40,64,69–71]. These changes have been attributed to the replacement of hydrogen with oxygen functionalities to the surface [72]. Rao et al. [72] also showed that changing the surface functionalization changed the absolute position of the band edges and hence the flat band potential. Furthermore, there is evidence that in some cases oxidized diamond surfaces have a greater potential window of water stability than do hydrogenated surfaces [73–76].

8

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

Of great interest is the possibility that the chemically bound hydrogen participates in the hydrogen evolution reaction generating a free radical site, for example, CH + H+ + e− → C • + H2

(1.1)

Anderson and Kang [77] showed that the energetics of this reaction are favorable. Furthermore, Yagi et al. [78] demonstrated deuterium exchange between the solution and a hydrogenated diamond surface during hydrogen evolution. Argoitia et al. [63] and Martin et al. [69] reported a rather surprising result that oxygen was added to hydrogenterminated diamond surfaces during liberation of hydrogen. The preceding results show that the electrochemical processes on diamond are significantly more complex than if the diamond were a completely inert source or sink of electrons. The early work on the electrochemistry of diamond was summarized in reviews by Tenne and Levy-Clement [79], Swain, Anderson, and Angus [80], Angus et al. [81], Kobashi [82], and Pleskov [83].

1.3 DEVELOPMENT OF ELECTROCHEMICAL APPLICATIONS OF DIAMOND By 2000 boron-doped diamond electrodes were receiving much attention. In addition to the very low background currents, low capacitance, and extremely wide potential window for water stability, diamond electrodes were shown to be extremely stable in many corrosive environments. Furthermore, the electrodes resisted fouling, especially when hydrogen-terminated. These properties began to be exploited in a number of laboratories around the world for a variety of applications. 1.3.1

Surface Functionalization

A natural extension of the early surface studies has been the extensive efforts on adding specific functional groups to the diamond surface. One of the first reported studies was that of Freedman and Stinespring [84] who in 1990 fluorinated diamond surfaces with an atomic beam of fluorine atoms. Smetkowski and Yates [85] in 1996 added fluorine to diamond surfaces by x-ray irradiation of perfluoroalkyl iodide layers on the surface. Ohtani et al. [86] in 1998 treated diamond with chlorine under ultraviolet irradiation; subsequent heating in boiling pyridine added quaternary pyridinium salts to the diamond. In 1999 Kuo, McCreery, and Swain [87] electrochemically coupled phenyl groups to diamond surfaces using phenyl diazonium. Polarization-induced modifications of diamond surfaces were described by Goeting et al. [88] in 1999. Throughout the early 2000s, there were many studies of addition of hydrocarbon functional groups to diamond. In 2001 Mathieu [89] reported functionalizing diamondcovered silicon wafers to improve biocompatibility. Wang et al. [90] used Diels-Alder chemistry to functionalize diamond with hydrocarbons. The theory of Diels-Alder reactions on diamond surfaces was discussed by Fitzgerald and Doren [91] and Lu et al. [92]. The reactivity of the 2 × 1 reconstructed (100) surface of diamond to 1, 3 butadiene and cyclopentene was reported by Russell et al. [93]. Hovis et al. [94] and Strother et al. [95] photochemically functionalized polycrystalline diamond surfaces with several

1.3 DEVELOPMENT OF ELECTROCHEMICAL APPLICATIONS OF DIAMOND

9

organic molecules. Another approach was reported by Ida et al. [96,97], who treated hydrogenated diamond surfaces with organic peroxides to add carboxylic acids and other functionalities. In 2002 Bent [98] reviewed the status of chemical functionalization of group IV semiconductor surfaces, including diamond. 1.3.2

Destruction of Wastes

One early application of conducting diamond was the anodic destruction of organic wastes. This process takes advantage of the very high anodic overpotentials that can be achieved on diamond electrodes as well as their chemical and dimensional stability under extreme conditions. Carey, Christ, and Lowery [99] used this method to oxidize photo solutions, phenol, hydroquinone, sodium salts of EDTA, and formic, oxalic, and malonic acids. From 1999 to 2001, Comninellis and co-workers [100–110] published an extensive series of papers on the use of boron-doped diamond electrodes for destruction of wastes. They showed that highly oxidizing species, for example, the persulfate ion, could be generated on diamond electrodes at anodic potentials well above the reversible potential for oxygen evolution [104]. They also demonstrated that hydroxyl groups generated at high anodic potentials were efficient oxidizing agents that kept organic films from forming on the diamond anodes [102,103,107,110]. Other workers used diamond anodes to both detect and destroy organic materials in waste streams. Davila-Jimenez et al. [111] reported in 2000 that textile dyes could be anodically treated using diamond electrodes. In 2002, Terashima et al. [112] used diamond electrodes to detect chlorophenols in drain waters. Bellagamba et al. [113] employed diamond anodes to oxidize dissolved polyacrylates in water, and Van Hege, Verhaege, and Verstraete [114] at the State University of Ghent oxidized organic components and ammonia in the concentrated filtrate from reverse osmosis membranes on diamond anodes. Lawrence et al. [115] detected and oxidized sulfide ion on boron-doped diamond electrodes; Van Andel and Janssen [116] described a process for the regeneration of chrome solutions using diamond electrodes, and Canizares et al. [117,118] described the destruction of phenol wastes using diamond anodes. Another interesting application is the use of boron-doped diamond anodes for the highly oxidizing electro-Fenton reaction [119,120]. Electrochemical oxidation on diamond electrodes is an active and ongoing field; recent work is described in later chapters of this book. 1.3.3

Sensors and Electroanalysis

The potential of diamond electrodes for sensor and analytical applications was recognized early by a number of workers: Ramesham et al. [52], Swain and Ramesham [54], Awada, Strojek, and Swain [58], and Zhu et al. [59]. The first applications followed shortly after this recognition. In 1996, Argoitia et al. [63] reported that the cerium/cerous redox couple could be detected on diamond. Strojek et al. [121] found enhanced signal to noise ratios at diamond electrodes. Also in 1996, Davidson et al. at Vanderbilt University developed diamond-based hydrogen [122] and oxygen [123] gas sensors. Use of diamond electrodes for analysis of biological molecules was an early research direction. In 1998 a group at Heriot Watt University in Edinburgh described the use of heavily boron-doped diamond electrodes for the determination of glucose [124]. In 1999 Rao et al. [125] from the University of Tokyo reported the electrochemical oxidation of nicotinamide adenine dinucleotide (NADH) and sensors based on enzyme-catalyzed

10

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

reactions involving NADH as a cofactor. Methods for determining sulfadiazine [126] as well as histamine and serotonin [127] were developed by Sarada et al., and Popa, Tryk, and Fujishima [128] developed methods for purines. Fujishima, Rao, and Tryk [129] demonstrated that diamond electrodes could be used for the detection of trace amounts of lead, without the use of mercury. In 2001 two reports described the use of transparent diamond electrodes for making in situ spectroscopic observations. In one, the addition of hydroxyl groups to diamond was observed [130] and in the other the electro-oxidation of ferrocyanide and electroreduction of methyl viologen was monitored [131]. Also in 2001 Hayashi et al. [132] used a double layer of undoped- and boron-doped diamond between two platinum electrodes to detect phosphine, diborane, and arsine. Kawarada et al. [133] fabricated electrolytesolution-gate field effect transistors (FETs) as biocompatible ion sensors. Kohn et al. [134] discussed the prospects of diamond micro devices and reviewed the use of diamond in electromechanical, electrothermal, and electrochemical sensors [135]. Saterlay, Foord, and Compton [136] developed a diamond sensor to be used with ultrasound for the detection of 4-chlorophenol in aqueous solution. Fujishima and Rao [137] described diamond detectors for flow-injection analysis of chlorophenols and theophylline. Development of diamond-based sensors is continuing in many laboratories around the world. Carbon-based sensors, including diamond, have been the subjects of an extensive review by McCreery [138]; other recent reviews on diamond-based sensors are by Chailapakul, Siangproh, and Tryk [139] and Zhou and Zhi [140].

1.4

OTHER DIRECTIONS

Two current directions of research on diamond electrochemistry deserve mention: bioelectronic applications of diamond and the anomalous surface conductivity of diamond. Both areas are the subject of much current research and both appear to have significant long-range implications. 1.4.1

Biolectronic Applications

Diamond as a platform for bioelectronic and bioelectrochemical applications is becoming a major research tool. These applications exploit the biocompatibility of diamond, its chemical inertness, and the ability to tailor its surface chemistry to permit attachment of biologically active molecules (e.g., DNA) for bioelectronics and biosensing. Diamond has significant advantages over silicon and the other common semiconductors because it does not have a solid surface oxide interposed between it and the external environment. Biosensor work using diamond has been a continuing focus. These studies include not only early sensor studies mentioned in the previous section but also the electrochemical analysis of nucleic acids [141] and the determination of single and double-stranded DNA [142] at boron-doped diamond electrodes. Another direction was explored by Tachiki et al. [143] who measured the inter-relationship between the surface charging of diamond and the physisorption of DNA. In 2001 Adamschik et al. described a diamond lab on a chip for biochemical applications including DNA synthesis [144]. In 2002 Ushizawa et al. immobilized DNA on diamond powder through an ester linkage [145]. The covalent bonding of the DNA to the diamond was verified by diffuse reflectance infrared spectroscopy. In 2002 and 2003, Hamers and co-workers

1.4 OTHER DIRECTIONS

11

used a photochemical scheme to obtain DNA functionalization of nanocrystalline and polycrystalline diamond films through amine groups [146,147]. In 2003 Takahashi et al. [148] sequentially hydrogenated, chlorinated, aminated, and finally carboxylated a diamond substrate to provide a platform for chemically binding DNA. Also in 2003, Wenmackers et al. [149] used thymidine as a linker molecule to immobilize DNA on diamond; the bound DNA was confirmed by confocal fluorescence microscopy. Functionalization of diamond with DNA for sensor and other applications remains of major current research interest. A significant body of work has been achieved by Hamers and co-workers at the University of Wisconsin. These studies include measuring changes in surface electrical properties of DNA modified diamond films as a function of exposure to complementary and noncomplementary DNA molecules [150], the fabrication of a biologically sensitive field effect transistor on a nanocrystalline diamond film [151], and an electrically addressable functionalized film [152]. This same group also compared the thermal stability of DNA-modified diamond and silicon surfaces [153] and developed methods for functionalizing diamond surfaces to resist protein adsorption, while still permitting specific binding for biorecognition [154,155]. Nebel and co-workers [156] at the Diamond Research Center in Tsukuba, Japan, studied the geometric orientation of DNA covalently bonded to single crystal diamond and the photochemical attachment of oriented amine linkers to hydrogen-terminated single crystal diamond [157]. This same group used periodically arranged benzene linker molecules, also on single crystal diamond, to obtain dense DNA films [158]. Recent reviews on bioelectronics and bioelectrochemistry have been written by Linares, Doering, and Linares [159], Nebel et al. [160] and Vermeeren et al. [161]. 1.4.2

Anomalous Surface Conductivity of Diamond

Most of the interaction between diamond science and electrochemistry exploits the unusual properties of diamond for electrochemical applications. There is one important example of the reverse in which electrochemical principles have lead to a deeper understanding of the properties of diamond. In 1989 Landstrass and Ravi [162] reported a curious observation that hydrogenterminated diamond developed a pronounced p-type surface conductivity upon exposure to air. This observation was extremely unusual because surface conductivity had not been reported previously despite more than a century of study of the electronic properties of diamond. Subsequent research, initially by Gi et al. [163–165], started to unravel the cause of the conductivity. Gi showed that the conductivity increased with exposure to acidic gases such as NO2 , decreased in the presence of basic gases such as NH3 [163,164], and that the sheet concentration of holes was as high as 7 × 1013 cm−2 [164]. In order to explain the effect, Gi proposed that the holes arose from the oxidation of the terminal hydrogen atoms by the hydronium ion, H3 O+ [165]. Gi’s mechanism was one of several put forward to explain the surface conductivity, including the presence of acceptor sites arising from subsurface hydrogen. Maier et al. [166,167] framed the surface oxidation process in terms of an electrochemical redox reaction. They proposed that the hydrogen redox couple, 2H+ + 2e− = H2 , in an adsorbed water film acted as an external electrochemical acceptor to the electrons in the diamond. The electron chemical potential (Fermi energy) of the redox couple lay below the Fermi energy of hydrogen-terminated diamond, which permitted the couple to act as an external electron acceptor, for example, to oxidize the diamond. In this model, the

12

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

role of the hydrogen-termination was to provide a surface dipole that raised the electronic band structure of diamond so that the electron transfer could take place. This proposal also received rather limited support, in part because the partial pressure of H2 in the atmosphere is so low that it was difficult to understand how it could support a reversible redox couple. Subsequently, Foord et al. [168] and Chakrapani et al. [169] gave evidence that the electrochemical mechanism was correct, but that the oxygen redox couple, 4H+ + 4e− + O2 = 2H2 O, was responsible for the effect. Calculations of Petrini and Larson [170] showed that the oxygen and ozone redox couples were both more favorable for electron transfer from diamond than the hydrogen couple. Chakrapani et al. [171] used macroscopic size samples of diamond and water to demonstrate that the changes in pH and dissolved oxygen concentration were consistent with electron transfer to the oxygen redox couple. It is of considerable interest that electron transfer to an external redox couple was proposed by Shapoval et al. earlier (in 1995) to explain conductivity changes in diamond exposed to molten salts [172]. The electrochemical transfer doping of diamond is still a subject of active research. Much more remains to be done to achieve a full understanding of the effect and to explore potential applications. It should be noted in this context that the surface conductivity has been used as the basis of interesting field effect transistors, most notably by Kawarada et al. [133] and Denisenko, Aleksov, and Kohn [173]. The understanding that an electrochemical mechanism played a role in mediating the electrical properties of diamond led to attempts to find the effect in other systems. Ristein [174] discussed the general nature of transfer doping as an alternative to conventional impurity doping. Strobel et al. [175] demonstrated transfer doping of diamond to an adsorbed surface layer of fullerene molecules, C60 . Chakrapani et al. reported electrochemical transfer doping of carbon nanotubes [176] and that the photoluminescence of GaN and ZnO could be mediated by electrochemical transfer doping by changing the

300

Number of publications

250 200 150 100 50 0 1990

1995

2000 Publication year

2005

2010

Figure 1.2 Number of publications on diamond electrochemistry versus year of publication. Three publications appeared prior to 1990: Iwaki et al. [38] in 1983 and Pleskov et al. [40] and Natishan and Morrish [41] in 1989. Source: ISI Web of Knowledge; the count was limited to papers identified by the keyword combinations (diamond electrochemistry), (diamond electrochemical) etc.

REFERENCES

13

acidity or basicity of the ambient humid air [177]. A very interesting development is the prediction that electrochemical charge transfer to the oxygen redox couple can be used to alter the conductivity of graphene [178]. Chen et al. [179] have recently published a comprehensive review of surface transfer doping.

1.5

CONCLUSIONS

The current status of diamond electrochemistry can be found in a recent review volume [180] and in the following chapters of this volume. Major areas of expanding interest include waste treatment using diamond anodes, electroanalysis on diamond electrodes, diamond-based electrochemical sensors, diamond electrodes for functional stimulation of muscle, diamond as a substrate for affixing DNA and other molecules of biological interest, and electronic devices based on diamond surface conductivity. The interest in these and other types of electrochemistry involving diamond is continuing to increase. The number of published papers concerned with diamond electrochemistry as a function of time is shown in Figure 1.2.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

W.G. Eversole, US patents 3,030,187 and 3,030,188, 1962. F.P. Bundy, H.T. Hall, H.M. Strong, R.H. Wentorf, Nature 1955, 176 , 51–55. J.C. Angus, H.A. Will, W.S. Stanko, J. Appl. Phys. 1968, 39 , 2915–2922. J.C. Angus, D.J. Poferl, N.C. Gardner, S. Chauhan, T.J. Dyble, P. Sung, Sint. Almazy 1971, 3 , 30–31 (in Russian). D.J. Poferl, N.C. Gardner, J.C. Angus, J. Appl. Phys. 1973, 44 , 1428–1434. S.P. Chauhan, J.C. Angus, N.C. Gardner, J. Vac. Sci. Tech. 1974, 11 , 423. S.P. Chauhan, J.C. Angus, N.C. Gardner, J. Appl. Phys. 1976, 47 , 4746–4754. B.V. Deryagin, D.V. Fedoseev, V.M. Lukyanovich, B.V. Spitsyn, V.A. Ryabov, A.V. Lavrentiev, J. Cryst. Growth 1968, 2 , 380–384. B.V. Deryagin, D.V. Fedoseev, Russ. Chem Rev . 1970, 39 , 783–788. D.V. Fedoseev, K.S. Uspenskaya, V.P. Varnin, S.P. Vnukov, Izv. Akad. Nauk SSSR, Seriya Khimicheskaya 1978, 6 , 1252–1256. V.P. Varnin, B.V. Deryagin, D.V. Fedoseev, I.G. Teremetskaya, A.N. Khodan, Kristallografiya 1977, 22 , 893–896. D.V. Fedoseev, B.V. Deryagin, V.P. Varnin, G.K. Sokolina, Doklady Akad. Nauk SSSR 1979, 247 , 1201–1204. B.V. Spitsyn, L.L. Bouilov, B.V. Deryagin, J. Cryst. Growth 1981, 52 , 219–226. S. Matsumoto, Y. Sato, M. Kamo, N. Setaka, Jpn. J. Appl. Phys. 1982, 21 , part 2 , L183–L185. S. Matsumoto, Y. Sato, M. Tsutsumi, N. Setaka, J. Mater. Sci . 1982, 17 , 3106–3112. M. Kamo, Y. Sato, S. Matsumoto, N. Setaka, J. Cryst. Growth 1983, 62 , 642–644. S. Matsumoto, J. Mater. Sci. Lett . 1985, 4 , 600–602. S. Matsumoto, M. Hino, T. Kobayashi, Appl. Phys. Lett . 1987, 51 , 737–739. Y. Matsui, S. Matsumoto, N. Setaka, J. Mater. Sci. Lett . 1983, 2 , 532–534. R. Berman, Physical Properties of Diamond , Clarendon Press, Oxford, 1965.

14

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

21. J. E. Field, ed., The Properties of Diamond , Academic Press, London, 1979. 22. G. Davies, Diamond , Adam Hilger Ltd., Bristol, 1984. 23. J. Wilks, E. Wilks, Properties and Applications of Diamond , Butterworth Heinemann, Oxford, 1991. 24. J.E. Field, ed., Properties of Natural and Synthetic Diamond , Academic Press, London, 1992. 25. L.S. Pan, D.R. Kania, eds, Diamond: Electronic Properties and Applications, Kluwer Academic Publishers, Boston, 1995. 26. M. Werner, R. Locher, Rep. Prog. Phys. 1998, 61 , 1665–1710. 27. C.E. Nebel, J. Ristein, eds., Thin Film Diamond I, Semiconductors and Semimetals, Volume 76 , Elsevier Academic Press, Amsterdam, 2003. 28. C.E. Nebel, J. Ristein, Eds., Thin Film Diamond II, Semiconductors and Semimetals, Volume 77 , Elsevier Academic Press, Amsterdam, 2004. 29. R. S. Sussman, ed., CVD Diamond for Electronic Devices and Sensors, John Wiley & Sons, Ltd, Chichester, 2009. 30. A.T. Collins, Phil. Trans. R. Soc. A 1993, 342 , 233–234. 31. D. Narducci, C.R. Guarnieri, J.J. Cuomo, J. Electrochem. Soc. 1991, 138 , 2446–2451. 32. Y.F. Tsai, K.P. Ananthanarayanan, P.J. Gielisse, S.S. Mitra, J. Appl. Phys. 1972, 43 , 3677–3682. 33. J. Mort, K. Okumura, M.A. Machonkin, Phil. Mag. B 1991, 63 , 1031–1036. 34. K. Nishimura, K. Das, J.T. Glass, J. Appl. Phys. 1991, 69 , 3142–3148. 35. G. Janssen, W.J.P. van Enckevort, W. Vollenberg, L.J. Giling, Diamond Relat. Mat . 1990, 1 , 789–800. 36. J. J. Lander, J. Morrison, Surf. Sci . 1966, 4 , 241–246. 37. P. G. Lurie, J. M. Wilson, Surf. Sci . 1977, 65 , 453–475. 38. M. Iwaki, S. Sato, K. Takahashi, H. Sakairi, Nucl. Instrum. Methods Phys. Res. 1983, 209–210 , 1129–1133. 39. M. Iwaki, K. Takahashi, H. Watanabe, S. Sato, RIKEN Review 1996, 12 , 27–28. 40. Y. V. Pleskov, A. Y. Sakharova, M. D. Krotova, L. L. Bouilov, B. V. Spitsyn, J. Electroanal. Chem. 1987, 228 , 19–27. 41. P.M. Natishan, A. Morrish, Mat. Lett . 1989, 8 , 269–272. 42. A.Ya. Sakharova, A.E. Sevastyanov, Y.V. Pleskov, G.L. Teplitskaya, V.V. Surikov, A.A. Voloshin, Soviet Electrochemistry 1991, 27 , 239–244. 43. A. Sakharova, L. Nykos, Y.V. Pleskov, Electrochim. Acta 1992, 37 , 973–978. 44. Y.V. Pleskov, A.Y. Sakharova, E.V. Kasatkin, VA. Shepelin, J. Electroanal. Chem. 1993, 344 , 401–404. 45. A.Y. Sakharova, Y.V. Pleskov, F. Diquarto, S. Piazza, C. Sunseri, I.G. Teremetskaya, V.P. Varnin, Russ. J. Electrochem. 1995, 31 , 169–174. 46. Y.V. Pleskov, V.Y. Mishuk, M.A. Abaturov, V.V. Elkin, M.D. Krotova, V.P. Varnin, I.G. Teremetskaya, J. Electroanal. Chem. 1995, 396 , 227–232. 47. K. Patel, K. Hashimoto, A. Fujishima, J. Photochem. Photobiol. A: Chem. 1992, 65 , 419–429. 48. K. Patel, K. Hashimoto, A. Fujishima, Denki Kagaku 1992, 60 , 659. 49. R. Tenne, K. Patel, K. Hashimoto, A. Fujishima, J. Electroanal. Chem. 1993, 347 , 409–415. 50. C. Reuben, E. Galun, H. Cohen, R. Tenne, R. Kalish, Y. Muraki, K. Hashimoto, A. Fujishima, J.E. Butler, C. Levy-Clement, J. Electroanal. Chem. 1995, 396 , 233–239.

REFERENCES

15

51. M. Marchywka, P.E. Pehrsson, D.J. Vestyck, D. Moses, Appl. Phys. Lett . 1993, 63 , 3521–3523. 52. R. Ramesham, R.F. Askew, M.F. Rose, B.H. Loo, J. Electrochem. Soc. 1993, 140 , 3018–3020. 53. B. Miller, R. Kalish, L. C. Feldman, A. Katz, N. Moriya, K. Short, A. E. White, J. Electrochem. Soc. 1994, 141 , L41–L43. 54. G. M. Swain, R. Ramesham, Anal. Chem. 1993, 65 , 345–351. 55. G. M. Swain, Adv. Mater. 1994, 6 , 388–392. 56. G. M. Swain, J. Electrochem. Soc. 1994, 141 , 3382–3293. 57. S. Alehashem, F. Chambers, J. W. Strojek, G. M. Swain, R. Ramesham, Anal. Chem. 1995, 67 , 2812–2821. 58. M. Awada, J. W. Strojek, G. M. Swain, J. Electrochem. Soc. 1995, 142 , L42–L45. 59. J.Z. Zhu, S.Z. Yang, P.L. Zhu, X.K. Zhang, G.X. Zhang, C.F. Xu, H.Z. Fan, Fresenius J. of Anal. Chem. 1995, 352 , 389–392. 60. P.L. Zhu, J.Z. Zhu, S.H. Yang, X.K. Zhang, G.X. Zhang, Fresenius J. of Anal. Chem. 1995, 353 , 171–173. 61. J.L. Wu, J.Z. Zhu, L. Shan, N.Y. Cheng, Anal. Chim. Acta 1996, 333 , 125–130. 62. H. B. Martin, A. Argoitia, J. C. Angus, A. B. Anderson, U. Landau, in Applications of Diamond Films and Related Materials: Third International Conference ( A. Feldman et al., eds.), NIST Special Publication 885, U.S. Dept. of Commerce, 1995, pp. 91–94. 63. A. Argoitia, H. B, Martin, E. J. Rozak, U. Landau, J. C. Angus, in “Diamond for Electronic Applications” ( D.L. Dreifus et al., eds.), Mater. Res. Soc. Proc. 1996, 416 , 349–354. 64. H. B. Martin, A. Argoitia, U. Landau, A. B. Anderson, J. C. Angus, J. Electrochem. Soc. 1996, 143 , L133–L136. 65. P. Bouamrane, A. Tadjeddine, J.E. Butler, R. Tenne, C. Levy-Clement, J. Electroanal. Chem. 1996, 405 , 95–99. 66. N. Vinokur, B. Miller, Y. Avyigal, R. Kalish, J. Electrochem. Soc. 1996, 143 , L238–L240. 67. J. Wang, G. M. Swain, M. Mermoux, G. Lucazeau, J. Zak, J. W. Strojek, New Diamond Frontier Carbon Tech. 1999, 9 , 317–343. 68. M.C. Granger, M. Witek, J.S. Xu, J. Wang, M. Hupert, A. Hanks, M.D. Koppang, J.E. Butler, G. Lucazeau, M. Mermoux, J.W. Strojek, G.M. Swain, Anal. Chem. 2000, 72 , 3793–3804. 69. H.B. Martin, A. Argoitia, J.C. Angus, U. Landau, J. Electrochem. Soc. 1999, 146 , 2959–2964. 70. C.H. Goeting, J.S. Foord, F. Marken, R.G. Compton, Diamond Relat. Mat . 1999, 8 , 824–829. 71. C.H. Goeting, F. Marken, A. Gutierrez-Sosa, R.G. Compton, DiamondRelat. Mat . 2000, 9 , 390–396. 72. T.N. Rao, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 680–684. 73. T. Yano, E. Popa, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1998, 145 , 1870–1876. 74. I. Yagi, H. Notsu, T. Kondo, D.A. Tryk, A. Fujishima, J. Electroanal. Chem. 1999, 473 , 173–178. 75. T. Yano, E. Popa, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 1081–1087. 76. D.A. Tryk, K. Tsunozaki, T.N. Rao, A. Fujishima, Diamond and Related Mat . 2001, 10 , 1804–1809. 77. A.B. Anderson, D.B. Kang, J. Phys. Chem. A 1998, 102 , 5993–5996.

16

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

78. I. Yagi, K. Ogai, T. Kondo, A. Fujishima, K. Ueda, K. Uosaki, Chem. Lett . 2003, 32 , 1050–1051. 79. R. Tenne, C. Levy-Clement, Israel J. Chem. 1998, 38 , 57–73. 80. G. M. Swain, A. B. Anderson, J. C. Angus, MRS Bulletin 1998, 23, 56–60. 81. J. C. Angus, H. B. Martin, U. Landau, Y. E. Evstefeeva, B. Miller, N. Vinokur, New Diamond Frontier Carbon Tech. 1999, 9, 175–187. 82. K. Kobashi, ed., Collection of articles from New Diamond and Frontier Carbon Tech. 1999, 9 , Nos. 3 & 5. 83. Y. V. Pleskov, Russ. Chem. Rev . 1999, 68 , 381–392. 84. A. Freedman, C.D. Stinespring, Appl. Phys. Lett . 1990, 57 , 1194–1196. 85. V.S. Smentkowski, J.T. Yates, Science 1996, 271 , 193–195. 86. B. Ohtani, Y.H. Kim, T. Yano, K. Hashimoto, A. Fujishima, K. Uosaki, Chem. Lett . 1998, 9 , 953–954. 87. T.C. Kuo, R.L. McCreery, G.M. Swain, Electrochem. Solid State Lett . 1999, 2 , 288–290. 88. C.H. Goeting, F. Marken, A. Gutierrez-Sosa, R.G. Compton, J.S. Foord, Diamond Relat Mat . 2000, 9 , 390–396. 89. H.J. Mathieu, Surf. Interf. Anal . 2001, 32 , 3–9. 90. G.T. Wang, S.F. Bent, J.N. Russell, J.E. Butler, M.P. D’Evelyn, J. Am. Chem. Soc. 2000, 122 , 744–745. 91. D.R. Fitzgerald, D.J. Doren, J. Am. Chem. Soc. 2000, 122 , 12334–12339. 92. X. Lu, G. Fu, N.Q. Wang, Q.N. Zhang, M.C. Lin, Chem. Phys. Lett . 2001, 343 , 212–218. 93. J.N. Russell, J.E. Butler, G.T. Wang, S.F. Bent, J.S. Hovis, R.J. Hamers, M.P. D’Evelyn, Mat. Chem. and Phys. 2001, 72 , 147–151. 94. J.S. Hovis, R.J. Hamers, M.P. D’Evelyn, J.N. Russell, J.E. Butler, J. Am. Chem. Soc. Comm. 2000, 122 , 732–733. 95. T. Strother, T. Knickerbocker, J.N. Russell, J.E. Butler, L.M. Smith, R.J. Hamers, Langmuir 2002, 18 , 968–971. 96. S. Ida, T. Tsubota, O. Hirabayashi, M. Nagata, Y. Matsumoto, A. Fujishima, Diamond Relat. Mat . 2003, 12 , 601–605. 97. S. Ida, T. Tsubota, S. Tanii, M. Nagata, Y. Matsumoto, Langmuir 2003, 19 , 9693–9698. 98. S.F. Bent, Surface Sci . 2002, 500 , 879–903. 99. J.J. Carey, C.S. Christ, Jr., S.N. Lowery, “Method of Electrolysis Employing a Doped Diamond Anode to Oxidize Solutes in Waste Water,” US patent, 5,399,247, 1995. 100. G. Foti, D. Gandin, C. Comninellis, A. Perret, W. Haenni, Electrochem. Solid State Lett . 1999, 2 , 228–230. 101. M. Fryda, D. Herrmann, L. Shafer, C.P. Klages, A. Perret, W. Haenni, C. Comninellis, D. Gandini, New Diamond Frontier Carbon Tech. 1999, 9 , 229–240. 102. D. Gandini, P.A. Michaud, I. Duo, E. Mahe, W. Haenni, A. Perret, C. Comninellis, New Diamond Frontier Carbon Tech. 1999, 9 , 303–316. 103. D. Gandini, C. Comninellis, A. Perret, W. Haenni, Fifth Eur. Symp. Electrochem. Engng. 1999, 145 , 181–190. 104. P.A. Michaud, E. Mahe, W. Haenni, A. Perret, C. Comninellis, Electrochem. Solid State Lett . 2000, 3 , 77–79. 105. S. Ferro, A. DeBattisti, I. Duo, C. Comninellis, W. Haenni, A. Perret, J. Electrochem. Soc. 2000, 147 , 2614–2619. 106. D. Gandini, E. Mahe, P.A. Michaud, W. Haenni, A. Perret, C. Comninellis, J. Appl. Electrochem. 2000, 30 , 1345–1350.

REFERENCES

17

107. M. Panizza, P.A. Michaud, G. Cerisola, C. Comninellis, J. Electroanal. Chem. 2001, 507 , 206–214. 108. M. Panizza, P.A. Michaud, G. Cerisola, C. Comninellis, Electrochem. Comm. 2001, 3 , 336–339. 109. W. Haenni, J. Gobet, A. Perret, L. Pupunat, P. Rychen, C. Comninellis, B. Correa, Diamond Materials VII, Proceedings (Electrochemical Society Series) 2002, pp. 16–23. 110. J. Iniesta, P.A. Michaud, M. Panizza, G. Cerisola, A. Aldaz, C. Comninellis, Electrochim. Acta 2001, 46 , 3573–3578. 111. M.M. Davila-Jimenez, M.P. Elizalde-Gonzalez, A. Gutierrez-Gonzalez, A.A. Pelaez-Cid, J. Chromatography 2000, 889 , 253–259. 112. C. Terashima, T.N. Rao, B.V. Sarada, D.A. Tryk, A. Fujishima, Anal. Chem. 2002, 74 , 895–902. 113. R. Bellagamba, P.A. Michaud, C. Comninellis, N. Vatistas, Electrochem. Comm. 2002, 4 , 171–176. 114. K. Van Hege, M. Verhaege, W. Verstraete, Electrochem. Comm. 2002, 4 , 296–300. 115. N.S. Lawrence, M. Thompson, C. Prado, L. Jiang, T.G.J. Jones, R.G. Compton, Electroanalysis 2002, 14 , 499–504. 116. Y. Van Andel, L.J.J. Janssen, Appl. Electrochem. 2002, 32 , 481–486. 117. P. Canizares, M. Diaz, J.A. Dominguez, J. Garcia-Gomez, M.A. Rodrigo, Ind. & Eng. Chem. Res. 2002, 41 , 4187–4194. 118. P. Canizares, F. Martinez, M. Diaz, J. Garcia-Gomez, M.A. Rodrigo, J. Electrochem. Soc. 2002, 149 , D118–D124. 119. K.J. McKenzie, D. Asogan, F. Marken, Electrochem. Comm. 2002, 4 , 820–824. 120. E. Brillas, B. Boye, I. Sires, J.A. Garrido, R.M. Rodriguez, C. Arias, P.L. Cabot, C. Comninellis, Electrochim. Acta 2004, 49 , 4487–4496. 121. J.W. Strojek, M.C. Granger, G.M. Swain, T. Dallas, M.W. Holtz, Anal. Chem. 1996, 68 , 2031–2037. 122. Y. Gorbuz, W.P. Kang, J.L. Davidson, D.L. Kinser, D.V. Kerns, Sensors and Actuators B-Chemical 1996, 33 , 100–104. 123. Y. Gorbuz, W.P. Kang, J.L. Davidson, D.V. Kerns, J. Electrochem. Soc. 1999, 146 , 387–391. 124. C.E. Troupe, I.C. Drummond, C. Graham, J. Grice, P. John, J.I.B. Wilson, M.G. Jubber, N.A. Morrison, Diamond Relat. Mat . 1998, 7 , 575–580. 125. T.N. Rao, I. Yagi, T. Miwa, D.A. Tryk, A. Fujishima, Anal. Chem. 1999, 71 , 2506–2511. 126. B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, New Diamond Frontier Carbon Tech. 1999, 9 , 365–377. 127. B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, Anal. Chem. 2000, 72 , 1632–1638. 128. E. Popa, D.A. Tryk, A. Fujishima, Diamond Materials VI, Proceedings (Electrochemical Society Series) 2000, 99 , 512–517. 129. A. Fujishima, T.N. Rao, D.A. Tryk, Electrochim. Acta 2000, 45 , 4683–4690. 130. H.B. Martin, P.W. Morrison, Electrochem. Solid State Lett . 2001, 4 , E17–E20. 131. J.K. Zak, J.E. Butler, G.M. Swain, Anal. Chem. 2001, 73 , 908–914. 132. K. Hayashi, Y. Yokota, T. Tachibana, K. Kobashi, T. Fukunaga, T. Takada, J. Electrochem. Soc. 2001, 148 , H17–H20. 133. H. Kawarada, Y. Araki, T. Sakai, T. Ogawa, H. Umezawa, Phys. Stat. Solid. A-Appl. Res. 2001, 185 , 79–83. 134. E. Kohn, M. Adamschik, P. Schmid, A. Denisenko, A. Aleksov, W. Ebert, J. Phys. D-Appl. Phys. 2001, 34 , R77–R85.

18

ELECTROCHEMISTRY ON DIAMOND: HISTORY AND CURRENT STATUS

135. E. Kohn, W. Ebert, M. Adamschik, P. Schmid, A. Denisenko, New Diamond Frontier Carbon Tech. 2001, 11 , 81–100. 136. A.J. Saterlay, J.S. Foord, R.G. Compton, Electroanalysis 2001, 13 , 1065–1070. 137. A. Fujishima, T.N. Rao, Diamond Related Mat . 2001, 10 , 1799–1803. 138. R.L. McCreery, Chem. Rev . 2008, 108 , 2646–2687. 139. O. Chailapakul, W. Siangproh, D.A. Tryk, Sens. Lett . 2006, 4 , 99–109. 140. Y.L. Zhou, J.F. Zhi, Talanta 2009, 79 , 1189–1196. 141. C. Prado, G.U. Flechsig, P. Grundler, J.S. Foord, F. Marken, R.G. Compton, Analyst 2002, 127 , 329–332. 142. T.A. Ivandini, B.V. Sarada, T.N. Rao, A. Fujishima, Analyst 2003, 128 , 924–929. 143. M. Tachiki, Y. Kaibara, Y. Sumikawa, M. Shigeno, T. Banno, K.S. Song, H. Umezawa, H. Kawarada, Phys. Status Solidi A-Appl. Res. 2003, 199 , 39–43. 144. M. Adamschik, M. Hinz, C. Maier, P. Schmid, H. Seliger, E.P. Hofer, E. Kohn, Diamond Relat. Mat . 2001, 10 , 722–730. 145. K. Ushizawa, Y. Sato, T. Mitsumori, T. Machinami, T. Ueda, T. Ando, Chem. Phys. Lett . 2002, 351 , 105–108. 146. W.S. Yang, O. Auciello, J.E. Butler, W. Cai, J.A. Carlisle, J. Gerbi, D.M. Gruen, T. Knickerbocker, T.L. Lasseter, J.N. Russell, L.M. Smith, R.J. Hamers, Nature Materials 2002, 1 , 253–257. 147. T. Knickerbocker, T. Strother, M.P. Schwartz, J.N. Russell, J.E. Butler, L.M. Smith, R.J. Hamers, Langmuir 2003, 19 , 1938–1942. 148. K. Takahashi, M. Tanga, O. Takai, H. Okamura, Diamond Relat. Mat . 2003, 12 , 572–576. 149. S. Wenmackers, K. Haenen, M. Nesladek, P. Wagner, L. Michiels, M. vandeVen, M. Ameloot, Phys Status Solidi A-Appl. Res. 2003, 199 , 4–48. 150. W. Yang, J.E. Butler, J.N. Russell, R.J. Hamers, Langmuir 2004, 20 , 6778–6784. 151. T.L. Lasseter, W. Cai, R.J. Hamers, Analyst 2004, 129 , 3–8. 152. W. Yang and R.J. Hamers, Appl. Phys. Lett . 2004, 85 , 3626–3628. 153. M. Lu, T. Knickerbocker, W. Cai, W. Yang. R.J. Hamers, L.M. Smith, Biopolymers 2004, 73 , 606–613. 154. T. Lasseter, B.H. Clare, B.M. Nichols, N.L. Abbott, R.J. Hamers, Langmuir 2005, 21 , 6344–6355. 155. T.L. Lasseter, B.H. Clare, N.L. Abbott, R.J. Hamers, J. Am. Chem. Soc. 2004, 126 , 10220–10221. 156. B. Rezek, D. Shin, T. Nakamura, C.E. Nebel, J. Am. Chem. Soc. 2006, 128 , 3884–3885. 157. C.E. Nebel, D. Shin, D. Takeuchi, T. Yamamoto, H. Watanabe, T. Nakamura, Diamond Relat. Mat . 2006, 15 , 4–8. 158. D.C. Shin, N. Tokuda, B. Rezek, C.E. Nebel, Electrochem. Comm. 2006, 8 , 844–850. 159. R. Linares, P. Doering, B. Linares, Stud. Health Technol. Inform. 2009, 149 , 284–296. 160. C.E. Nebel, D.C. Shin, B. Rezek, N. Tokuda, H. Uetsuka, H. Watanabe, J. R. Soc. Interf . 2007, 4 , 439–461. 161. V. Vermeeren, S. Wenmackers, P. Wagner, L. Michiels, Sensors 2009, 9 , 5600–5636. 162. M. I. Landstrass, K. V. Ravi, Appl. Phys. Lett . 1989, 55 , 975–977. 163. R.S. Gi, T. Ishikawa, S. Tanaka, T. Kimura, Y. Akiba, M. Iida, Jpn. J. Appl. Phys. 1997, 36 , 2057–2060. 164. R.S. Gi, K. Tashiro, S. Tanaka, T. Fujisawa, H. Kimura, T. Kurosu, M. Iida, Jpn. J. Appl. Phys. 1999, 38 , 3492–3496.

REFERENCES

19

165. R.S. Gi, T. Mizumasa, Y. Akiba, Y. Hirose, T. Kurosu, M. Iida, Jpn. J. Appl. Phys. 1995, 34 , 5550–5555. 166. F. Maier, M. Reidel, B. Mantel, J. Ristein, L. Ley, Phys. Rev. Lett . 2000, 85 , 3472–3475. 167. J. Ristein, M. Riedel, F. Maier, B.F. Mantel, M. Stammler, L. Ley, J. Phys.: Condensed Mat . 2001, 13 , 8979–8987. 168. J.S. Foord, C.H. Lau, M. Hiramatsu, R.B. Jackman, C.E. Nebel, P. Bergonzo, Diamond Relat. Mat . 2002, 11 , 856–860. 169. V. Chakrapani, S.C. Eaton, A.B. Anderson, M. Tabib-Azar, J.C. Angus, Electrochem. Solid State Lett . 2005, 8 , E4–E8. 170. D. Petrini, K. Larsson, J. Phys. Chem. C 2007, 111 , 13804–13812. 171. V. Chakrapani, J.C. Angus, A.B. Anderson, S.D. Wolter, B.R. Stoner, G. U. Sumanasekera, Science 2007, 318 , 1424–1430. 172. V.I. Shapoval, I.A. Novosyolova, V.V. Malyshev, H.B. Kushkhov, Electrochim. Acta 1995, 40 , 1031–1035. 173. A. Denisenko, A. Aleksov, E. Kohn, Diamond Relat. Mat . 2001, 10 , 667–672. 174. J. Ristein, Science 2006, 313 , 1057–1058. 175. P. Strobel, M. Riedel, J. Ristein, L. Ley, Nature 2004, 430 , 439–441. 176. V. Chakrapani, J.C. Angus, A.B. Anderson, G. Sumanasekera, Diamond Electronics Symposium, Mat. Res. Soc. Symp. Proc. 2007, 956 , paper J15-01. 177. V. Chakrapani, C. Pendyala, K. Kash, A.B. Anderson, M.K. Sunkara, J.C. Angus, J. Am. Chem. Soc. 2008, 130 , 12944–12952. 178. S.J. Sque, R. Jones, P.R. Briddon, Phys. Stat. Sol. (a) 2007, 204 , 3078–3084. 179. W. Chen, D. Qi, X. Gao, A.T.S. Wee, Prog. Surf. Sci . 2009, 84 , 279–321. 180. A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, eds, Diamond Electrochemistry, Elsevier/BKS Inc., Tokyo, 2005.

2 Synthesis of Diamond Films Vadali V. S. S. Srikanth and Xin Jiang

2.1

INTRODUCTION

Diamond is a well-known gemstone and considered as a precious possession. About fourth century BC, diamonds were first mined in India, and in the second century BC, Chinese realized diamond’s first industrial use originating from its ultimate hardness. Based on the optical and electrical properties, and the nature and percentage of impurities present, natural diamonds are classified into four types of monocrystalline (types Ia, Ib, IIa, and IIb) and two types of polycrystalline (carbonados and ballas) diamonds. However, type Ia diamonds constitute 98% of the total natural diamonds. The space group for the diamond lattice is O7h (F4,/d 32/m) with two carbon atoms per Bravais cell. Diamond structure can be visualized as two interpenetrating face-centered cubic (FCC) lattices (see Figure 2.1) displaced along the body diagonal by (1/4, 1/4, 1/4)a, where a is the dimension of the cubic unit cell. Each carbon atom of the lattice has a tetrahedral configuration consisting of sp 3 hybrid atomic orbitals. The {111} crystallographic planes are constituted by 6-atom hexagonal rings that are arranged in such a way that the adjacent atoms are alternatively displaced upward and downward from the plane. The stacking sequence along the <111> direction is ABC ABC ABC . . . . In the ˚ same crystallographic direction, the lattice constant and bond length are 3.56 and 1.54 A, respectively. Diamond has two isomers—namely, graphite and lonsdaleite. In graphite, each carbon atom has sp 2 atomic configuration and therefore has three inplane σ bonds; the remaining valence electron forms π bonds using a pz atomic orbital. Consequently, the trigonally bonded 6-carbon rings are situated in a flat plane, contrary to that in the diamond structure. The stacking sequence of the planes in graphite is ABABAB. . . . The lattice constant in ˚ and the in-plane, nearest neighbor the basal plane between repeating layers is 6.707 A, ˚ spacing is 1.42 A. On the other hand, the structure of lonsdaleite is derived from diamond Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

21

22

SYNTHESIS OF DIAMOND FILMS

Figure 2.1

FCC structure of the diamond crystal. (Reprinted with permission from Ref. 23.)

A

B

C

A

B

B

A

A

Figure 2.2 Crystal structure of cubic (left) and hexagonal (right) diamond. The difference in stacking sequence of (111) layer pairs in the two structures has been illustrated. (Reprinted with permission from Ref. 23.)

(see Figure 2.2) in that the positioning of atoms in each plane is the same as that in the cubic diamond. However, the stacking sequence of planes is AB AB AB. . . . In lonsdaleite the atoms thus bond closely; the lattice constants in the a and c directions are 2.52 and ˚ respectively. The distance between adjacent atoms is 1.52 A. ˚ 4.12 A, Apart from being a precious gemstone, the diamond has always attracted a special attention due its unique and unsurpassed properties [1–3]. Consequently, research interests have risen to obtain synthetic diamond. Synthetic diamond was obtained for the first time by employing the high-pressure high-temperature (HPHT) technique [4]. In this

2.2 DIAMOND FILM CVD TECHNIQUES

23

technique, graphite and transition-metal (Fe, Ni, Co, etc.,) catalyst powders are mixed and the mixture is subjected to high pressures (∼80–300 kbar) and high temperatures (∼1900–3000◦ C) such that the catalyst component melts. The reaction is then equilibrated for a certain time; the temperature is then gradually decreased to reduce the carbon solubility in the catalyst, resulting in the precipitation of excess carbon as diamond. The role of catalyst in the HPHT technique is not only to lower the activation energy required for the conversion of graphite into diamond but also to dissolve the carbon atoms in the graphite, thus enabling them to reconstitute as diamond. Due to the availability of only expensive natural diamonds and synthetic diamonds only in the form of abrasive grit (via the HPHT technique), it was understood that most properties of diamond could be realized for technological applications only when it is synthesized as a thin film often chemically bonded to an underlying nondiamond substrate material. For almost five decades, chemical vapor deposition (CVD) has been the most commonly used technique to synthesize diamond films. The CVD of diamond involves growth of a tetrahedrally bonded carbon atom network affected by addition of one carbon atom at a time from a suitable gas phase to an initial template; it is accomplished at much lower pressures when compared to the HPHT technique [5–8]. The CVD of diamond thin films has been an interesting topic of research ever since its accomplishment [9–11] and subsequent modifications [12–23]. A typical diamond film growth process via any CVD method includes three main stages that lead to the incorporation of suitable gas species into the growing diamond solid phase: (1) activation of the gas mixtures; (2) gas-phase reactions; and (3) diffusion of gas species onto the substrate surface. Further chemistry occurs on the reaction surface and finally diamond growth takes place. Diamond films’ microstructure can range from oriented columnar grains (including epitaxial grains) to a nanocrystalline structure, depending on the deposition conditions such as surface pretreatment, temperature, pressure, and gas composition. When gas species responsible for diamond growth reach the reaction surface, adhere to it, and settle quickly into possible equilibrium positions before any structural defects form on the growth front, a single crystalline diamond film can be obtained. On the contrary, when the same atoms do not quickly settle into stable equilibrium positions upon their arrival at the reaction surface, nanocrystalline diamond films can be obtained. At present, diamond thin films are serving in a variety of applications related to machining, field emission, electromechanical systems, electrochemical systems, biomedical, biosensing, and others [24–34]. By making a review of diamond film synthesis, the objective of this chapter is to cover the most important scientific and engineering aspects of diamond CVD. This chapter starts with a review of typical diamond thin film synthesis techniques. In the subsequent sections, various aspects pertaining to diamond film (1) nucleation and growth and (2) epitaxy will be discussed. This chapter concludes with a brief discussion of some future perspectives.

2.2 2.2.1

DIAMOND FILM CVD TECHNIQUES History

In the early stages of research concerning CVD of diamond films, the most exciting feature was the formation of the diamond film itself. Although diamond film could be formed, it had been of little technological or commercial importance due to drawbacks

24

SYNTHESIS OF DIAMOND FILMS

such as extremely low growth rate, lack of substrate selectivity, formation of graphite, and other nondiamond forms of carbon. However, with time these drawbacks have been considerably overcome. Early CVD experiments employed mainly thermal decomposition of carbon containing gases at a pressure less than 1 atm to grow a diamond homoepitaxially on natural diamond crystals maintained at 900◦ C. The growth rate of diamonds in such experiments was a meager 1 nm/h; also, graphite was co-deposited alongside diamond. Later it was shown that atomic hydrogen could preferentially etch graphite during the diamond deposition. Subsequent research works extended the possibilities of diamond CVD by showing that diamond could be grown on nondiamond surfaces as well. In the early 1950s, Eversole and Kenmore [9] developed the cyclic pyrolysis process and demonstrated, for the first time, diamond synthesis under low pressures. In the early 1960s, Angus, Will, and Stanko [10] extended Eversole’s work and synthesized a boron-doped diamond film on diamond grit. Eversole’s work was further extended by Derjaguin et al. [11]. Interestingly, in 1953, Schmellenmeier [35] (of Germany) reported synthesis of crystalline diamond containing hard carbon films from acetylene in an electrical discharge. In 1982, Matsumoto et al. [15] made a major breakthrough in diamond CVD technology. A hot filament was used, and hydrocarbon and hydrogen gas mixtures were passed over it to directly activate the gas phase. The diamond film was then deposited onto a nondiamond substrate located about 1 cm away from the filament. Due to the presence of atomic hydrogen during the deposition, graphite was etched away simultaneously; this rendered the previously used deposition and etching cycles unnecessary, which in turn led to higher diamond growth rates. Since then, various gas phase activation methods such as direct current (DC)-plasma, radio frequency (RF)-plasma, microwave plasma, electron cyclotron resonance (ECR)-microwave plasma, and so on, were developed. Mainly, the role of atomic hydrogen in diamond growth was recognized, and the diamond growth rates approached the rates acceptable by industrial standards. In a further development, Rudder, Posthill, and Markunas [36] suggested that the pyrolysis of fluorocarbons such as tetrafluoromethane (CF4 ) could produce epitaxial diamond growth; this suggestion was based on the understanding that OH, O, O2 , F, and F2 species are better graphite etchants than atomic hydrogen. In this method, CF4 and F2 gas mixture diluted in He was blown onto a diamond substrate maintained at 875◦ C. The deposited diamond film was devoid of any graphite. Although the fluorocarbon pyrolysis process took place nearly at thermodynamic equilibrium, a growth rate of only ∼0.6 μm h−1 was achieved. In another development, hydrogen-deficient carbon-containing noble gas plasma was used to deposit phase-pure nanocrystalline diamond films on Si substrates [37]. On the other hand, other non-CVD methods based on low- [38] and high- [39] energy carbon ion beams have also been developed. Methods based on electron bombardment of the substrate have also been developed that resulted in an increase of diamond growth rates almost up to 20 μm h−1 [40,41]. In large areas, as large as 19 cm2 , low pressure diamond film growth has also been reported [42]. Besides CVD, physical vapor deposition methods were also employed [43,44]. A method wherein diamond growth at low temperatures (as low as 75◦ C) was also reported [45]. At present, various developments in plasma-type CVD processes are not only allowing the growth of polycrystalline diamond films with fewer defects but also the growth of single crystalline CVD diamond films with required and highly consistent properties. A good report on the synthesis of diamond in laboratory explaining several of the above discussed progresses is available [46].

2.2 DIAMOND FILM CVD TECHNIQUES

2.2.2

25

Thermal Decomposition Techniques

2.2.2.1 Hot Filament Chemical Vapor Deposition (HFCVD) The hot-filament CVD method is the earliest [15,16] and the most popular method used for diamond film deposition under low pressures (∼20–40 Torr). In this method, a refractory metal (such as tungsten, tantalum, and rhenium) filament is heated to a temperature above 2000◦ C; a methane (CH4 ) and hydrogen (H2 ) gas mixture, typically about 1:99 in volume, is then passed over the hot filament. Hydrocarbon pyrolysis takes place and diamond is deposited on a substrate (at ∼700–900◦ C) kept at a distance of about 8–10 mm from the hot filament. In the HFCVD process, atomic hydrogen is produced, which etches away graphite much faster than diamond. Thus, the deposition rate of diamond is typically about 1 μm h−1 , which is suitable for industrial purposes. For a schematic diagram of typical HFCVD reactor, please refer to [47, p. 155]. In the HFCVD method, the refractory metal filaments are first carburized prior to the deposition of diamond films. The vapor pressure of metal carbides is less than the respective metal. Therefore, at the working filament temperatures, carburization reduces the metallic impurity incorporation in the diamond films. The HFCVD method possesses the ability to adjust to a wide variety of carbon gas sources such as methane, propane, ethane, and other hydrocarbons. Even oxygen-containing hydrocarbons—including acetone, ethanol, and methanol—can be used. The addition of oxygen-containing species widens the temperature range within which diamond deposition can take place. With the aim of improving the phase purity and growth rate of diamond, some modifications have been incorporated into the typical design of HFCVD. In one such modification, a DCplasma is used in combination with the typical HFCVD setup; a bias voltage is applied to the substrate and filament [41,42,48]. A moderate positive voltage is applied to the substrate and a negative voltage to the filament (or an accessory electrode), resulting in electron bombardment of the substrate, which induces desorption of the surface hydrogen which in turn increases the growth rate (up to about 10 μm h−1 ). This technique is called electron-assisted HFCVD. When the bias voltage is high enough to ignite a stable plasma discharge, the decomposition of H2 and hydrocarbon is greatly enhanced, leading to a remarkable increase in the diamond growth rate (up to about 20 μm h−1 ). When the polarity of the bias is reversed—that is, when the substrate is negatively biased—substrates’ surface undergoes ion bombardment, which results in the enhanced nucleation of diamond on nondiamond substrates. Another modification is to replace the single hot filament with multiple filaments or a filament net for uniform diamond film deposition over large areas. 2.2.2.2 Oxy-Acetylene Torch Method Hirose and Kondo [49] were the first to introduce the combustion flame-assisted diamond CVD. In this method, the smoldering tip of a welding torch is used to oxidize a mixture of acetylene (C2 H2 ) and oxygen (O2 ) gases (ratio 1:1). Diamond film gets deposited on the substrate where the tip of the bright interior region of the flame touches the substrate, which is maintained at a temperature of 800–1050◦ C. Under appropriate experimental conditions, the diamond growth rate in this method can go up to 50–100 μm h−1 . Using this method, homoepitaxial films could be deposited at high rates [50]. The aerosol-doping technique, in combination with the oxyacetylene torch method, was used to dope boron into diamond films [51]. When the torch is used in its transversing mode, large-area diamond coatings could be obtained [52]. The advantages of this method over the conventional CVD methods include simplicity and cost-effectiveness of the equipment, lack of power supply, high growth rate, and the

26

SYNTHESIS OF DIAMOND FILMS

ability to deposit diamond over large areas and on curved substrate surfaces. On the other hand, in this method the deposition is difficult to control and consequently the deposited diamond films are inhomogeneous in both microstructure and composition. Also, the torch produces thermal gradients over the substrates’ surface, thereby causing the substrate to warp or fracture during the process of coating large substrate areas. Murakawa, Takeuchi, and Hirose [52] and Cappelli and Paul [53] have achieved great progress in enhancing the diamond film quality and area by using the oxy-acetylene torch method. 2.2.3

Plasma-Aided Deposition Techniques

2.2.3.1 Microwave Plasma-Enhanced CVD (MWCVD) It was found that the atomic hydrogen concentration under CVD conditions could be increased by using a DC plasma ignited by an electrical discharge [14,54]. Thus, plasma became another pathway to dissociate molecular hydrogen into atomic hydrogen and to simultaneously activate hydrocarbon species, thereby leading to the diamond deposition. Besides DC plasma, two other kinds of plasma with different excitation frequencies are possible. The excitation frequency for microwave plasma CVD is typically 2.45 GHz, whereas for radio frequency (RF) plasma CVD it is 13.56 MHz. A schematic of ASTeX microwave plasma CVD reactor is shown in Figure 2.3. Also, Davis [47, p. 159] provides a typical microwave reactor setup.

Microwave Gas inlet

Water cooling Plasma Substrate

Heating

Pump Bias power supply 0 ± 300 VDC

Figure 2.3 Schematic of a microwave plasma-enhanced CVD reactor manufactured by ASTeX. (Reprinted with permission from Ref. 23.)

2.2 DIAMOND FILM CVD TECHNIQUES

27

The microwave frequency used to ignite the plasma oscillates the electrons in the gas mixtures used to deposit diamond. Consequently, high ionization fractions are generated as electrons collide with gas atoms and molecules. Microwave plasma has hot electrons as well as cold ions and neutrals. A substrate typically about 2–3 cm in diameter is placed on a holder in a tube reactor (quartz or steel) compatible with a waveguide that can guide the microwaves generated by a magnetron. Microwaves enter into the reaction chamber from a proprietary antenna that converts a rectangular microwave signal into a circular one. The microwave proceeds through a quartz window into the reaction chamber filled with gas at a particular pressure. The microwave energy is transmitted to the gas mixtures, thereby igniting a luminous ball-shaped plasma, the size of which can be enlarged by increasing the microwave power. The edge of the luminous plasma will be located typically about 2 cm above the substrate’s surface. In this setup, the substrate can be independently heated. Under suitable conditions of gas pressure, microwave power, gas mixture ratio, gas flow rates of different gases, substrate temperature, and so on, diamond film grows on the substrate. 2.2.3.2 DC Plasma CVD Direct current plasma CVD is another deposition technique wherein DC plasma is used to activate the gas source (typically hydrocarbon and hydrogen) for diamond growth. Suzuki et al . [21] developed the DC-plasma-assisted CVD (DC PACVD). In this method, a glow discharge is generated in the reactor between the substrate and the other electrode when the bias voltage is sufficiently high. The activation of the gas phase is thus achieved via the collisions of high-energy electrons with the neutral gas species, resulting in the dissociation of the gas species and the generation of diamond-forming reactive gas species. In all the DC PACVD setups, substrate is mounted on the anode of the reactor. The DC PACVD possesses the ability to coat large areas, which are limited only by the size of the electrodes and the DC power supply. However, this method has a drawback of depositing diamond at very low rates (<0.1 μm h−1 ). With the increase of gas pressure (∼150 Torr), DC voltage (∼1 kV), and discharge current, growth rates of up to 250 μm h−1 can be approached but with a reduction in the deposition area. Davis [47, p. 166] provides the schematic of a typical DC plasma CVD reactor. Another advance in DC plasma CVD is the DC plasma jet method. Kurihara et al. [55] have designed a DC plasma jet facility termed DIA-JET, which employs a gas- injection nozzle consisting of a cathode rod surrounded by an anode tube [56]. In this method, the substrate is placed on a separate water-cooled holder and the plasma is ignited by generating a DC arc discharge between the electrodes (typically a water-cooled copper anode and a tungsten cathode). Due to the use of high pressures and powers, a DC arc discharge can be generated. Typically, methane will be mixed with H2 /Ar plasma jet at high pressures (∼188 Torr). Due to the high temperatures (∼5000 K) in the plasma, a large fraction of H2 molecules dissociate and a sufficient supply of active diamond-forming carbon gas species are generated. The gas species in the plasma jet are transported to the substrate surface within a few hundred microseconds, thereby reducing hydrogen combination reactions. Typical growth rates up to 80 μm h−1 can be achieved. Because various DC arc methods can synthesize high-quality diamond on nondiamond substrates at fast growth rates, they provide a marketable means for diamond film synthesis. Ohtake and Yoshikawa [57] have achieved a growth rate of 930 μm h−1 using the DC plasma jet CVD. For many years, by using DC plasma jet CVD, thick diamond films are being obtained on a routine basis. Boron- or

28

SYNTHESIS OF DIAMOND FILMS

phosphorous-doped diamond film can also be obtained using DC plasma jet CVD [58]. As previously mentioned, in another DC-plasma deposition variation, Fujimori et al. [48] synthesized diamond films at high rates using a hybrid HFCVD plus DC plasma method. 2.2.3.3 RF Plasma CVD In this technique, radio frequency is used to generate a plasma using two electrode configurations—namely, capacitive-coupled parallel plates and induction at atmospheric pressure. Davis [47, p. 164] provides schematics of the two. RF plasma-assisted CVD utilizes a frequency of 13.56 MHz. RF plasma can be used to deposit diamond on areas larger than are possible using microwave plasmas. However, RF capacitive plasma is limited in that the frequency of the plasma is optimal for sputtering, especially if the plasma contains argon. On the other hand, capacitive coupled RF plasma is not suitable for the growth of high-quality diamond due to the ion bombardment from the plasma, which results in severe defects in the growing diamond film. Polycrystalline diamond films have been synthesized by the induction method using MWCVD deposition conditions [59]. Homoepitaxial diamond films have also been deposited by RF induction plasma enhanced chemical vapor deposition (PECVD) [48]. 2.2.3.4 Electron Cyclotron Resonance Microwave Plasma-Assisted CVD In this method, electron cyclotron resonance (ECR) is used to generate the microwave plasma. The cyclotron resonance is obtained when the frequency of an alternating electric field is made to match the natural frequency of the electrons orbiting the magnetic field force lines. This occurs in a magnetic field of 875 gauss at the standard microwave frequency of 2.45 GHz. A typical ECR-MP-CVD setup is shown in Figure 2.4. As discussed in the previous sections, direct current, radio frequency, and microwave plasmas ionize and decompose hydrogen and hydrocarbon species into hydrogen atoms and hydrocarbon radicals, and thus promote the diamond growth. It can thus be expected that ECR-MP-CVD can be a method capable of synthesizing diamond films since ECRMP generates high-density plasma (>1 × 1011 /cm3 ) that is favorable for diamond growth. In fact, Hiraki et al. [60] used ECR-MP-CVD to synthesize diamond in 1990. The growth temperature could be reduced to 500◦ C. Later, Yara et al. [61] and Mantei et al. [62] succeeded in diamond deposition using the ECR-MP-CVD technique. They obtained uniform films at substrate temperatures as low as 300◦ C. However, due to the extremely low pressure of the ECR process (10−4 –10−2 Torr), diamond growth proceeds at a very low rate. Therefore, this method is mainly used for laboratory experiments.

2.3 2.3.1

DIAMOND NUCLEATION AND GROWTH Nucleation

2.3.1.1 Definition and Types Diamond nucleation is the first and foremost step in diamond growth via CVD routes. The nucleation process affects film thickness, grain sizes, homogeneity, morphology, defects, surface roughness, and adhesion pertaining to the deposited diamond films. Nuclei are the smallest diamond units that have to be formed for the subsequent diamond growth to start. Nucleation is broadly classified into two types: homogeneous (gas phase) and heterogeneous (on the substrate). There are only few experimental evidences regarding gas phase nucleation [63] of diamond resulting from hydrocarbon molecules such as adamantane, tetracyclododecane, and hexacyclopentadecane acting as the possible diamond embryos. However, it was later shown

2.3 DIAMOND NUCLEATION AND GROWTH

H2 + CH4

Magnetic coil A

2.45 GHz microwave

Plasma

Sample loadlock Magnetic coil B

29

Windows

Diamond film Sample holder

Turbo pump

Figure 2.4 Ref. 23.)

Experimental setup for ECR plasma CVD of diamond. (Reprinted with permission from

that the suggested hydrocarbons are very unstable at CVD temperatures [64]. Additionally, diamond particulates if at all formed in the gas phase were in very low concentrations (when compared to the surface nucleation densities) and would have almost no effect on nucleation [65]. This revelation also led to the idea that small molecular species (for example, CH3 and C2 H2 ), which are capable of surviving the CVD atmosphere and reaching (to be incorporated into the growing diamond nuclei) the seeded substrate, are the ones more suitable for diamond nucleation and growth. On the other hand, heterogeneous nucleation (i.e., surface nucleation) of CVD diamond follows heteroepitaxial growth mechanism from a diamond-seeded (pretreated) substrate surface. First, gas atoms impinge on the reaction surface from the gas phase and get adsorbed onto the surface. Subsequently, the adatoms may either desorb or diffuse over the reaction surface; they may also diffuse into the substrate or bond to other surface atoms. With the deposition time, the surface concentration of the adatoms increases, resulting in the formation of nanoclusters. A statistical fluctuation in the local adatom concentration results in the growth or decay of the clusters. When the clusters attain a size just greater than a critical size, they become stable. Critical size is defined as the size above which growth will be greater than decay of the clusters during the concentration fluctuation. The time taken to form a stable nucleus is called the incubation period . The stabilized

30

SYNTHESIS OF DIAMOND FILMS

clusters then act as suitable sites for further growth either from the continued migration of single adatoms or from direct impingement of atoms from the gas phase. 2.3.1.2 Methods Scratching substrate surface with an abrasive powder has been the most common and powerful method for achieving diamond nucleation densities enough for the formation of continuous diamond films with uniform grain sizes. Mitsuda et al. [66] demonstrated for the first time that scratching the substrate surface with diamond powder could enhance the diamond nucleation density. Besides diamond powder, other abrasive powders such as SiC [67], cubic-BN [68], Cu or stainless steel [69], ZrB2 [70], Al2 O3 [71], and others, have also been used to scratch the substrate surface to enhance the diamond nucleation. Among all the powders, diamond powder has been the most effective one. The scratching technique can be applied to most of the substrates used for diamond growth. For Si substrates, a nucleation density of 107 –108 cm−2 can be routinely obtained after thoroughly scratching it with diamond powder. In contrast, the nucleation density in the case of nonscratched Si substrates can reach only 104 cm−2 . The nucleation density is proportional to the scratching time; the morphology changes from large isolated crystals for short scratching times to smaller, high-density crystals for long scratching times [71,72]. The grit size of the diamond powder used for scratching also influences the nucleation density; 0.25 μm grit is the most effective one for manual scratching [73], and a 40–50 μm powder is best for scratching in an ultrasonic bath using a grit suspension [74]. During scratching with diamond, cubic-BN, or a-SiC powder, the residual powder or fragments are unavoidably left in the scratched grooves and act as seeds for diamond growth. Crystal structures of cubic-BN and a-SiC are close to that of diamond and thereby diamond grows easily on them. In fact, Iijim, Aikawa, and Baba [75] observed the presence of diamond fragments in the scratched grooves of Si, upon which the diamond growth occurred. Scratching with abrasive powder changes the surface morphology; it creates edges, steps, dislocations, and other surface defects. These defects are considered as chemically active sites, which prefer to adsorb diamondforming gas species due to enhanced bonding at high-energy intersecting surfaces with a high density of unsaturated bonds and low coordination numbers [71,76,77]. In this context, using electron microscopy, Denning and Stevenson [70] experimentally observed diamond nucleation on the ridges that have been lithographically etched on nonscratched Si. Singh et al. [71] also reported enhanced nucleation on Si etched by HF/HNO3 . Si+ implantation on nonscratched Si has also been found to enhance the nucleation density [78]. In this method, neither diamond seeds nor carbon-rich layers existed; only the surface morphology or surface structure was changed. However, other attempts to enhance nucleation by creating etch pits on nonscratched Si using acid, H+ [79], or other reactive gas treatments [69] proved unsuccessful. Similarly, attempts to generate large numbers of defect sites via implantation with surface saturation levels of C+ [80] or Ar+ [79] on the unscratched substrates did not result in nucleation enhancement. An array of holes of 0.1–0.3 μm in diameter created by a focused Ga+ ion beam on a Si substrate resulted only in the deposition of nondiamond carbon in the grooves [81]. The reason for the discrepancy among these experiments is unclear. In order to avoid any such problems of understanding, the substrate surface must be thoroughly characterized (with nucleation point of view) prior to any diamond deposition since nucleation is a surface phenomenon and the most critical one. Another method to enhance diamond nucleation is to coat the substrate surface with graphite [82,83], amorphous carbon [82,84], diamond-like carbon [85–87], C60 , and

2.3 DIAMOND NUCLEATION AND GROWTH

31

mechanical oil [84]. This method and the previously mentioned scratching method cannot, however, result in oriented nucleation or epitaxial growth on nondiamond substrates. Bias-enhanced nucleation (BEN) is the method that not only results in higher nucleation densities, but under suitable experimental conditions, results in oriented or epitaxial nucleation. Yugo et al. [88] used the BEN method for the first time and obtained a high nucleation density (109 –1010 cm−2 ) on a mirror-polished substrate (without scratching). In this method a negative bias voltage is generally applied on the substrate in a typical MWCVD setup. BEN can lead to nucleation densities as high as 1010 –1011 cm−2 on mirror-polished Si substrates [89]. Using this method, Jiang and associates [90,91] and Stoner and Glass [92] realized the first ever heteroepitaxial growth of diamond on silicon and silicon carbide substrates, respectively. Yugo, Kimura, and Kanai [93] and Gerber et al. [94] explained the nucleation mechanism by suggesting a shallow ion implantation model in which the sp3 -bonded carbon clusters, formed by low-energy ion implantation, act as the active nucleation sites. The negative bias causes the positively charged ions in the gas phase (plasma in the case of MWCVD) to accelerate toward and bombard the substrate surface, thereby removing any contamination and facilitating the sp3 -bonded carbon cluster formation on the surface. These events in turn enhance the diamond nucleation. On the other hand, Stoner et al. [95] indicated that the critical process for nucleation should be the formation of a surface carbide layer and change in plasma chemistry, such as increase in the concentration of atomic hydrogen caused by substrate biasing. Jiang and colleagues [96,97] found that the overall temporal evolution of the nucleation density corresponds well with a surface kinetic model involving immobile active nucleation sites, germs, and nuclei. They also suggested that, in addition to surface defects (point defects, steps and sp3 -bonded carbon clusters) serving as the nucleation sites, the enhanced surface diffusion and sticking probability of carbon on silicon due to ion bombardment should be the decisive factors. The enhancement of surface diffusion of carbon species has been clearly observed by investigating the distribution of the first nearest-neighbor distances [96]. Stubhan et al. [98] and Lin and associates [40,41] showed that BEN also works in the HFCVD system. In the HFCVD method, the thermally activated gas-phase species (atomic hydrogen and hydrocarbon radicals) are neutral, and as a result, a negative substrate bias as in MWCVD cannot induce enhanced nucleation. However, when plasma is generated by proper choice of bias, enhancement of diamond nucleation similar to that in MWCVD can be achieved. Using BEN in HFCVD, a nucleation density of 109 –1010 cm−2 on mirror-polished Si was obtained, which is similar to the result obtained in the MWCVD system. There are also methods of obtaining high diamond nucleation densities without any surface scratching or substrate biasing. One method is by employing very low gas pressures; the other method is by Si+ implantation [78] into mirror-polished Si substrate prior to the introduction of methane into the deposition chamber. High-density diamond nucleation (109 –1011 cm−2 ) has been obtained on mirror-polished Si substrates using either HFCVD or ECR microwave plasma CVD setup working at very low pressures (0.1–1 Torr) [99]. Similar results have also been obtained on Ti substrates. Using this method, diamond grain density greater than 1010 cm−2 can be obtained, which is two orders of magnitude higher than the highest density (107 –108 cm−2 ) that can be obtained on scratched substrates under conventional CVD process pressure (10–50 Torr) values. In order to obtain high diamond nucleation densities by CVD methods, the substrate surface must be treated in such a way that (1) surface adsorption sites can be created for the incoming hydrocarbon radicals and (2) the distributed region of the adsorption sites

32

SYNTHESIS OF DIAMOND FILMS

is large enough (larger than the critical size of the nucleus) for continuing the growth of diamond nuclei. Si+ ions implantation into the mirror-polished Si substrate meets these criteria exactly. It changes only the surface structure without affecting the composition of Si. After a treatment with Si+ energy of 25 keV (lower energy would be better) and an implantation dose of 2 × 1017 cm−2 , diamond can easily nucleate and grow on a Si wafer. Si+ implantation-enhanced nucleation is assumed to create nanoscale surface defects on the Si substrate. These defects serve as active sites for the adsorption of hydrocarbon radicals necessary for initial diamond nucleation. A similar effect can also be found in the case of diamond growth on porous silicon [100] because, its surface has a dense nanoscale microstructure. 2.3.2

Growth

In order to grow diamond thin films more efficiently and economically, minimize defects in the films, know the most effective gas precursors, and so on, it is critical to understand the diamond film growth mechanism. Tsuda, Nakajima, and Oikawa [101,102] gave the first account of atomic scale {111} diamond growth mechanism [101,102]; it was proposed that diamond growth involved C+ 3 cations or a positively charged surface. However, this mechanism could not be generalized since, under HFCVD conditions, C3+ cations are scarce and the substrate surface is uncharged. Later, Chu et al. [103] proposed that the methyl radicals are the dominant species for the growth of all {111}, {100}, and {110} diamond facets under HFCVD conditions. However, at high temperatures, CH4 and C2 H2 will decompose into various products and it is difficult to distinguish from which original source the products originate. To this end, Martin and Hill [104] and Harris and Martin [105] showed that methyl or methane is more effective than acetylene for diamond film growth. Harris [106] used a nine-carbon model compound [bicyclononane (BCN)] and proposed a growth mechanism involving only neutral CH3 and hydrogen atoms. Frenklach and colleagues [107–110] proposed that acetylene, which is present in greater quantities than CH3 under HFCVD conditions [111–113], is the primary growth specie for {111} diamond and other low-index surfaces. On the other hand, the primary growth specie for nanocrystalline diamond growth has been found to be C2 dimer [114,115] but not CH3 or C2 H2 . In the following (111) and (100) diamond growth mechanisms, and the roles of hydrogen and oxygen gases during diamond film growth will be discussed in detail. Figure 2.5 shows high-resolution electron energy loss (HREEL) spectra obtained in situ during the growth of homoepitaxial (111) diamond surfaces and highly oriented (111) diamond films on Si (0 0 1) for different growth temperatures [116]. The loss peaks at 365, 155, 110, 310, and 460 meV correspond to CH stretching, CH bending, C–C stretching, and the first and second overtones of CH bending vibrations, respectively. The vibration modes and characteristic frequencies of CHx molecular subgroups [117,118] clearly indicate that the surface is not CH3 terminated, but CH terminated. Thus, diamond growth on the (111) surface proceeds via two-by-two layers mode. It can be noted that the CH stretching vibration is perpendicular to the (111) diamond surface, while the CH bending vibration is parallel to the (111) diamond surface. The electron-energy-loss spectral selection rule [119] does not allow the vibration parallel to the (111) surface and therefore should be absent in the spectra. However, in Figure 2.5, CH bending vibration loss peak at 155 meV is present. In order to understand this contradictory presence, the H on the (111) diamond surface was replaced by its isotope D [116]. By doing so, two types of adsorption sites were found to exist; one on the (111) surface, and the other

Intensity (arb. unit)

2.3 DIAMOND NUCLEATION AND GROWTH

33

(a)

(b)

14

(c) –100

100

300 500 Energy loss (meV)

700 ◦

◦

◦

Figure 2.5 HREEL spectra of (1 1 1) diamond facets grown at (a) 800 C, (b) 900 C, and (c) 1000 C. (Reprinted with permission from Ref. 116.)

on another facet. If it is speculated that the growing (111) diamond surface consists of (111) faces and (110) steps, it is not difficult to reconcile the contradiction. The bending vibration of CH on the (110) surface steps is perpendicular to the (111) surface and therefore the CH bending vibration is active. As Frenklach pointed out, growth on the (111) surface takes place in two stages [109,110]. In the first, kernel forms; in the second, it propagates. In the first stage, the appearance of island and the (110) steps are possible. The existence of the CH bending vibration can thus be understood. The second stage of the Frenklach model, which details the connection of C2 H2 in a two-layers by two-layers manner, is experimentally supported [116]. In addition, it has been shown experimentally that during the kernel formation, CH3 is the responsible diamond growth specie [120]. In summary, growth on the (111) diamond surface has been proposed to be completed in two stages. In the first stage, a surface carbon (active site) is activated by H abstraction, adsorption, and catenation of CH3 , which results in kernel formation. In the second stage, the (111) surface grows along the (011) direction aided by acetylene alone. As for the growth mechanism of (100) diamond surface, Harris’ predictions [106] (BCN model) agreed well with experiments. However, the steric repulsion for the H–H site on BCN is different from that on the (100) diamond surface. On the (100) diamond surface, which is terminated by CH2 radicals, a very strong steric repulsion should exist

34

SYNTHESIS OF DIAMOND FILMS

between the neighboring hydrogen atoms because the intermolecular H–H spacing is only ˚ Such a repulsion will ˚ about 0.77 A—nearly the same as that in the H2 molecule (0.74 A). greatly affect the growth of the {100} diamond surfaces. Similar reasons hold true for some low-index surfaces as well. Hamza, Kubiak, and Sulen [112] obtained a 1 × 1 lowenergy electron diffraction (LEED) pattern on the (100) diamond surface at temperatures between 500 and 750 K [121]. This was attributed to the saturation of the dangling bonds of a surface carbon atom by two hydrogen atoms, and thus surface reconstruction did not occur. This surface was a nominally di-hydrogenated surface. Above 1300 K, the surface structure showed a 2 × 1 reconstruction due to desorption of one H atom from a surface carbon atom. This surface was a nominally mono-hydrogenated surface with an elongated C–C dimer bond [122]. A 2 × 1 reconstruction of the (100) diamond surface grown at 1000◦ C was also observed by both scanning tunneling microscopy (STM) and atomic force microscopy (AFM) [123,124]. Figure 2.6 shows HREEL spectra of the (1 0 0) diamond surface grown at a temperature of 800◦ C [125]. Intensity losses occur at 156, 180, and 372 meV, with three smaller losses at 110, 310, and 530 meV. Compared with the characteristic frequencies of the molecular subgroups CHx [115,116], the spectrum is consistent with that of the CH2 radical, which has its stretching, scissors, wagging, twisting, and rocking vibrations at 370, 179, 157, 150, and 108 meV, respectively. In the HREEL spectra, 372, 180, and 110 meV losses are assigned to CH2 stretching, scissors, and rocking vibrations, respectively; and 156 meV loss is assigned to the overlapping of wagging and twisting vibrations. The 310 meV is the overtone of the loss at 156 meV, and the loss at 530 meV is the combination of CH2 wagging and stretching vibrations. The film deposited at 800◦ C exhibits a good crystallinity. At 1000◦ C, the film shows a “cauliflower”-like morphology, and the Raman spectrum exhibits a broad peak at 1580 cm−1 , which is characteristic of graphite, whereas its HREELS is similar to the one at 800◦ C. However, a prominent peak appears at 140 meV. This peak corresponds to the bending vibration of the mono-hydrogenated dimer. Due to the appearance of the mono-hydrogenated surface, hydrogen atoms that are bonded to the (100) surface become further separated. As a result, it releases the strong steric repulsion between the H atoms. Nevertheless, the CH bond of mono-hydrogenated dimer possesses to some extent π bond character. As the hydrocarbon radicals attach to the CH bond, they also take in some π bond character. Their tendency to form graphite is the reason why we observe a broad peak at 1580 cm−1 in the Raman spectrum. Noteworthy is that if the (100) surface of the as-grown sample is exposed to atomic hydrogen (no hydrocarbon involved), the 140 meV loss peak also appears in the HREEL spectrum. This can be explained by the abstraction of one of the two hydrogen atoms bonded to a surface carbon by gas-phase atomic hydrogen. If the abstraction is strong enough, mono-hydrogenated dimers appear in some local regions where the surface is similar to the growth surface at 1000◦ C. If the amount of atomic hydrogen is small, or if the abstraction proceeds at a moderate temperature (∼800◦ C), or if hydrocarbon is involved in the gas source, the abstraction of H will not be strong. Thus, di-hydrogenated carbon atoms remain in the neighborhood of C with one abstracted H, and mono-hydrogenated dimers cannot be formed. The carbon atom then keeps only one H and one vacant site, which may bond to the hydrocarbon radical. As a result, diamond growth proceeds steadily. The quantity of carbon with one hydrogen is determined by the growth temperature, the amount of atomic hydrogen near the surface, and the concentration of activated hydrocarbon. At growth temperatures around 1000◦ C, the (100) diamond surface consists of mono-hydrogenated H–C–C–H dimers as well. For even higher temperatures, a large amount of hydrogen

2.3 DIAMOND NUCLEATION AND GROWTH

35

Intensity (arb. unit)

(a)

(b)

(c) 14

(d) –100

100

300 500 Energy loss (meV)

700 ◦

Figure 2.6 HREEL spectra of (a) (1 0 0) diamond facets grown at 800 C (b) (1 0 0) facets grown at ◦ 1000 C, (c) the sample of (a) dosed with atomic hydrogen, and (d) dosed with oxygen. (Reprinted with permission from Ref. 116.)

desorbs from the surface and the surface carbons become CC dimers. The growth rate of the CH2 -terminated surface is determined by the amount of CH3 radicals in the gas phase. It has been shown that atomic hydrogen can easily abstract one hydrogen atom from CH3 ; therefore, CH2 radicals become the precursor of (100) diamond growth [126]. The control of the appearance of diamond grain facets becomes significant not only in practical usage but also in the testing of established diamond growth mechanisms. The morphology of CVD diamond films is correlated with the growth parameters. Systematic studies on the relationship between the appearances of the (111) and (100) facets and the growth parameters such as the ratio of CH4 /H2 , O2 content, and distance from the hot filament to the substrate have been reported [127,128]. It is well known that, for a kinetically controlled growth system, the crystal morphology is determined by the appearance of facets that have the slowest growth rate in their normal direction and by the corresponding relative growth rates [129]. Because the (110) surface is an S (stepped) face and encounters no repulsion between adjacent hydrogen atoms, it should have the highest growth rate and appear as diamond facets on the film surface. In fact, the growth rate of the (110) surface via the CVD method is the highest among the low-index surfaces, and either CH3 radical- or acetylene-speciesbased mechanisms can be postulated. Thus, in most cases, the surface of CVD diamond crystals appear with {100} faces and {111} faces. According to the established growth

36

SYNTHESIS OF DIAMOND FILMS

model, the growth rate of the (100) face depends on the concentration of CH2 or CH3, whereas the growth rate of the (1 1 1) facet relies on both CH3 and C2 H2 for kernel formation and subsequent growth. If the CH3 concentration near the substrate is much higher than the C2 H2 concentration, the {100} growth rate will be high, and hence the {100} faces will be absent. The crystal appears to have just {111} diamond faces. In contrast, if the C2 H2 concentration near the growth surface is dominant, the {111} diamond face grows faster [130], which consequently results in the appearance of {100} diamond facets. In most cases, the concentrations of CH3 and C2 H2 are comparable, so both {100} and {111} facets are present. Harris and Weiner [131] and Harris and colleagues [132] have measured the dependence of the CH3 and C2 H2 concentrations on the CH4 /H2 and O2 /H2 ratios, as well as on the spacing between the hot filament and substrate. Based on their experimental results, Sun, Zhang, and Lin [125] explained the regularity of the appearance of the diamond facets. In general, CVD diamond growth involves site activation by a surface hydrogen abstraction reaction, (Equation 2.1), followed by addition (Equation 2.2) of a hydrocarbon radical like CH3 [133,134]. Cd represents a surface radical. The competition between surface activation (Equation 2.1) and H-atom recombination with the surface radical site (Equation 2.3) determines the number of active nucleation sites available for a particular set of experimental parameters [133]: Cd H + H∗ ↔ Cd∗ + H2

(2.1)

Cd∗ + CH∗3 ↔ Cd − CH3

(2.2)

Cd∗

(2.3)

∗

+ H ↔ Cd − H

The stable nano-sized crystals formed during the nucleation stage typically exhibit spherical shapes. With time, nucleation density increases to a certain value on which it terminates or ceases to occur at a measurable rate. The isolated crystallites now grow and develop facets due to the relatively high rate of surface carbon diffusion from the surrounding surface sites. Once the crystals grow large enough to coalesce with each another, they form grain boundaries and then continue growing as a continuous film. The morphology of a growing diamond surface depends on the rates at which different diamond planes grow. Under typical growth conditions, the morphology assessment of a diamond film can be made by the growth parameter (Equation 2.4): α=

√ ν100 3 ν111

(2.4)

where ν100 and ν111 are the normal growth velocities of (100) and (111) diamond planes respectively [129,135]. The grains exhibiting fastest growth in the direction perpendicular to the substrate overshadow other slower-growing grains to form a continuous film with a columnar structure [136]. This mode of film growth is known as the evolutionary selection principle [137]. On the contrary to evolutionary selection principle, Jiang et al. [138] showed an oriented diamond growth dependent on facet reactivity and selectivity. Due to the presence of a very little amount of tetramethylsilane gas during a typical MWCVD of diamond, nanocrystalline β-SiC phase selectivity forms on different growing diamond

2.3 DIAMOND NUCLEATION AND GROWTH

37

faces—namely, (001) and (111)—appearing to the gas phase. At the considered experimental conditions, (001) diamond faces avoid any solid deposition resulting from siliconcontaining gas species. In addition, the defective nature of {111} diamond faces will decrease the nucleation barrier and provide high density of active nucleation sites for nano–β-SiC deposition. Figure 2.7 shows the scanning electron microscope (SEM) crosssectional image of the film. There is a single grain growth (marked by arrows) in [001] with its size increasing toward the surface, indicating that tetramethylsilane (TMS) addi¯ direction that tion allows only few suitable nuclei to grow. Typical edges along [110] are formed due to the intersection of {111} planes can also be clearly observed. To explain theoretically this novel growth mechanism, the energy levels of the frontier molecular orbitals (FMOs) and other nearby orbitals of the reactants, a CH3 , a diamond nanoparticle, and a SiH3 , determined at HF/6-31G** level of theory, see Figure 2.8. The growth mechanism was explained by analyzing the reaction occurrence by judging the energy difference between the FMOs of the reactants diamond and CH3 /SiH3 , using frontier orbital theory. The theoretical study revealed that {111} diamond facets are more reactive with SiH3 than the {100} diamond facets, readily forming SiC phases on the {111} facets and leaving the {100} facet clean and with only diamond growth. The selection of diamond (001) plane growth is therefore due to the distortion and/or termination of non-{001} planes growth. Here, the evolutionary selection theory [129,137] is no longer valid because the growth of all the non-{001} faces is interrupted due to nano–β-SiC deposition, even though their growth rates could be higher than that of (001) face. The formation of [001] texture in the present case is therefore due to an angular selection. The {001} facets growth without any tilt will be the fastest and occupy the top-most position, and as a result the tilted facets have no space to develop due to the geometric limitations. Diamond growth models are developed by considering the effects of C1 hydrocarbon radicals—namely, CH3 , CH2 , CH, and C atoms—on both monoradical and biradical reaction surface sites [139]. At typical CVD diamond deposition conditions, diamond growth takes place mainly via the addition of CH3 radicals to monoradical sites. Growth can also take place via the addition of CH3 radicals to the biradical sites; here, the

Figure 2.7 Backscattered field emission scanning electron microscope (FE-SEM) cross-sectional morphology of a film deposited with a continuous variation of TMS gas from a value of 0.0506 to 0% during the deposition. The brighter spot-like regions represent β –SiC phase. (Reprinted with permission from Ref. 138.)

38

SYNTHESIS OF DIAMOND FILMS

0.4 0.3

FMO eigenvalues of reactants (a.u.)

0.2

LUMO {111}

LUMO

{111}

{111}

0.1

LUMO

{100}

0.0 –0.1 {111}

–0.2

{111} {111}

–0.3 –0.4

HOMO {100}

HOMO

HOMO

–0.5 –0.6 –0.7 CH3

Diamond

SiH3

Figure 2.8 The energy levels of FMOs (with isosurfaces as inset) and other nearby orbitals of the reactants of CH3 , diamond nanoparticle (C54 H56 ), and SiH3 . The two energy levels with equal value next to the HOMO for CH3 or SiH3 are degenerate. (Reprinted with permission from Ref. 138.)

unused surface dangling bonds can be rapidly hydrogenated during the conversion of CH3 adducts into a CH2 surface groups [140]. At some deposition conditions, the biradical mechanism is dominant; in this case, the probability of species like C2 , C2 H, and so on, to be the biradical site adducts is enhanced. Such reactive species then have the opportunity to cross-link on the surface, creating a strongly bonded (maybe even nonetchable) defect. This surface defect could act as either a renucleation point for a new epitaxial layer, or, if it is misaligned with the existing lattice, a new crystallite growing in a different direction to that of the main bulk. This last possibility, often termed renucleation, leads to a decrease in the average crystal size. If renucleation occurs frequently, the crystallite size can drop from mm to μm, and eventually to nm, and the films are referred to single crystalline diamond, microcrystalline diamond , and (Ultra) nanocrystalline diamond , accordingly. In addition to CH3 , addition of the other less abundant but highly reactive C1 species, particularly atomic C, to either type of radical site at high H-atom concentration, can also be a route to growth, since the dangling bonds on the adduct would be readily hydrogenated converting the adduct into CH2 . However, at low H-atom concentration, the dangling bonds on the adduct can cross-link to lattice sites, again leading to renucleation and subsequent smaller crystal sizes. Models based on the surface migration of reactive species could predict even diamond growth rates and average crystallite sizes [141,142]. May et al. [143] have critically reviewed the diamond growth mechanism to simulate diamond growth on the (100) surface and presented a unified model for diamond growth based on Monte Carlo simulations [143]. Surface migration, nucleation processes, and

2.3 DIAMOND NUCLEATION AND GROWTH

39

the effects of gas impurities and gas surface reactions have been carefully incorporated into the model to generate diamond growth on the (100) surface. The growth rate is determined mainly by a balance between the flux of CH3 species to the surface and the desorption/etching rate. Using values for the various rates from the literature, a growth rate for standard diamond growth conditions of ∼1 μm h−1 has been estimated, which is consistent with experimental values. Migration of species along and across dimer rows is essential to obtain step-edge growth and smooth surfaces. Without migration, the surfaces become rough and spiky, and the growth rate drops typically by a factor of ∼2, although the exact amount depends on the choice of conditions, and can be several times this value if, for example, a much faster migration rate is used. The growth rate enhancement caused by migration is roughly proportional to the surface diffusion length. It follows that the growth rate enhancement is due to the increased probability of the migrating species meeting a step-edge due to sampling a larger number of surface sites than stationary species. By using a parameter to represent random surface defect formation, growth morphologies resembling “wedding cake” structures can be produced, which again, are consistent with atomic-scale morphologies seen experimentally, and may, in principle, be scaled up to rationalize the cauliflower or ballas diamond morphologies seen at higher C:H ratios. β-scission has been shown to be a minor process, removing only 1-in-800 of the adsorbing carbons from the lattice. Thus, its importance in maintaining a smooth growing surface and in removing longer-chained polymeric species from the surface may have been previously overestimated. 2.3.3

Role of Hydrogen and Oxygen

The most crucial aspect in a typical diamond CVD process is that the hydrocarbon gas must be diluted in hydrogen to as low as 1% and that the hydrogen must be dissociated into atomic hydrogen, which has several specific roles. During the deposition process, atomic hydrogen etches graphite about 20–30 times faster than diamond, resulting in rapid removal of graphite and other nondiamond phases from the substrate and thereby allowing only clusters with diamond structure to remain and grow [125]. The process stabilizes the diamond surface and maintains the sp3 hybridization configuration [144]. Atomic hydrogen not only converts hydrocarbons into the respective radicals that are the necessary species for diamond formation but it also abstracts hydrogen from the hydrocarbons attached on the surface [145], thereby creating active sites for adsorption of the diamond-forming species. However, excess atomic hydrogen causes unnecessarily strong abstraction. The formation of mono-hydrogenated dimers will increase and the graphite phase will readily appear, which results in the deterioration of diamond film quality. As already mentioned, diamond growth is a combined process of deposition and carbon etching that take place concurrently. The growth of diamond occurs if the deposition rate is larger than the etching rate. During CVD growth of diamond films, both atomic H and H+ ions in the plasma cause etching, and H+ ions etch even faster than atomic H [146,147]. It was found that a [001]-textured top layer can be prepared on polycrystalline or [111]-textured diamond films by the application of a negative substrate bias potential during diamond growth [147–149]. An etching effect of hydrogen ions on the growth of diamond films was observed and confirmed to play a dominant role for the [001]-textured growth. The H+ ion bombardment was performed by applying a negative substrate bias during a microwave plasma CVD process, using only hydrogen as the reactant gas. It was discovered that the etching efficiency of H+ ions on non-[001]- oriented grains is

40

SYNTHESIS OF DIAMOND FILMS

higher than that on grains with their (001) faces parallel to the substrate. Lateral growth of the (001) faces can occur during the bombardment process. As a result, the size of the (001) faces increases after H+ etching while grains oriented in other directions are etched off. This effect provides a way to improve the orientation degree of [001]-oriented diamond films and may be helpful for obtaining very thin [001]-oriented diamond films. As for the role of oxygen, its addition (in the form of CO, O2 , or alcohol) to the reaction gases not only has a beneficial effect on the growth rate and quality of the diamond films but it also allows low-temperature diamond growth [150]. Although hydrocarbon species are somewhat reduced by oxygen addition, oxygen has only a relatively small effect on the mole fractions of radical species such as H and CH3 [151]. Also, OH is formed at concentrations sufficient for the removal of nondiamond carbon at a rate comparable to that of diamond growth. At low temperatures (<800◦ C), the addition of oxygen not only enhances the growth rate but it also extends the region of diamond formation [152]. At temperatures lower than 500◦ C, oxygen was explained to have stronger preferential etching behavior than hydrogen. Another possible explanation about the role of atomic oxygen is its aid in the effective abstraction of surface hydrogen [153]. These occurrences are helpful for diamond growth. On the other hand, the addition of too much oxygen will cause strong hydrogen abstraction and even oxidation of the surface, which in turn will deteriorate the diamond film.

2.4

DIAMOND EPITAXY

Fabrication of diamond-based electronic devices requires diamond single crystals or epitaxial films. Even though epitaxial diamond films can be grown on natural or HPHT diamond [127,154–160], their industrial viability is quite low. This is because natural diamonds are rare and expensive, whereas HPHT-grown diamond crystals are small for practical purposes. To counter such problems, it is necessary to obtain heteroepitaxially grown diamonds. In this context, cubic boron nitride (cubic-BN) is the best candidate for diamond heteroepitaxy [161–166]. Cubic-BNs’ lattice parameter (1.3% mismatch) and surface energy are similar to that of diamond [167,168]. In spite of its large lattice mismatch (22%) with diamond, β-SiC is the second best candidate for diamond heteroepitaxy. Stoner and Gloass [92] obtained highly oriented diamond crystallites on β-SiC; diamond crystallites with diamond (100)//β-SiC (100), diamond [011}]//β-SiC [011] crystallographic relationships were obtained. Jiang and Klages [90] and their colleagues [91] obtained [001]-oriented diamond films grown epitaxially on Si (001) substrate. High-resolution transmission electron microscopy analysis showed that diamond (001) grew directly on silicon (001) [169,170] (see Figure 2.9). It was therefore understood that under certain experimental conditions when a negative electrical potential is applied to the substrate, there is no necessity for an intermediate β-SiC layer for epitaxial diamond nucleation on Si. Various researchers [171–174] and Stubhan et al. [98] could also obtain heteroepitaxial growth of diamond (0 0 1) on Si (0 0 1) employing the HFCVD method. It was also possible to improve the structural quality of diamond films by reducing the orientation deviation on Si (1 0 0) substrates from a best value of about 9◦ [175], as determined from the full width at half maximum (FWHM) of x-ray rocking curves, to a best value of about 2◦ [176]. The misorientation of diamond grown on β-SiC (001) has also been greatly reduced. An orientation deviation of smaller than 1◦ could be achieved in the case of diamond epitaxy on β-SiC (001) [177].

2.4 DIAMOND EPITAXY

41

Figure 2.9 HRTEM cross-sectional lattice image showing epitaxially grown diamond on Si; the image is taken along the [1 1 0] direction of the diamond-silicon interface. (Reprinted with permission from Ref. 169.)

Diamond was heteroepitaxially grown on other types of substrate materials—namely, BeO [178], Ni [179], Pt [180], Co [181], Ir [182–184], and SrTiO3 [185]. However, practical applications of diamond epitaxy on these substrates are still unclear. As for practical applications, diamond epitaxy on Si is highly desirable not only because of the ease of Si wafer availability but also because of its extensive usage in the electronics industry. Diamond epitaxy on Si will facilitate the integration of diamond-thin film electronics with that of already established Si technology and consequently could give a big leap to the diamond-based electronic device fabrication. Diamond heteroepitaxy on Si, then, has immense technological and scientific importance. In the following, diamond heteroepitaxy on Si will be elucidated. Typically, diamond heteroepitaxy on silicon involves two stages: nucleation and growth. It is well known that diamond nucleation on mirror-polished silicon substrates is difficult. Under conventional nucleation procedures (without substrate pretreatment), a nucleation density of only 104 cm−2 can be obtained, which can lead to the growth of only isolated individual diamond crystals. To avoid the substrate scratching with diamond powder, which severely damages the arrangement and periodic structures of the surface atoms, BEN was employed in a MWCVD system to epitaxially grow diamond on a mirror-polished Si substrate. Figure 2.10 shows a schematic diagram of the process parameters. Prior to BEN, in situ hydrogen plasma etching was performed in order to remove the native surface oxide layer. By carefully controlling the nucleation process, diamond nucleation with more than 90% [001]-oriented nuclei was achieved [97]. Later, high-density nucleation and oriented diamond films were also achieved by applying a negative bias to the substrate in a HFCVD system [96,171]. The investigation of the nucleation process reveals a narrow parameter window for epitaxial nucleation. The crucial parameters are the substrate temperature, methane concentration, applied bias voltage to the substrate during nucleation, and the nucleation time. A critical bias

42

SYNTHESIS OF DIAMOND FILMS

nucleation cooling growth

Magnitude (arb. units)

etching

0.4–4.0 0.4–2.0

CH4 (%)

700–800

Ts (°C)

0

0 ≈ 850

800–900

0 Vb (V) –50––150 –150 0

20

60 40 Time (min)

80

1000

Figure 2.10 Schematic diagram of the MWCVD process for preparing heteroepitaxial diamond films on Si. (Reprinted with permission from Ref. 97.)

Figure 2.11 Diamond nucleation density as a function of substrate bias voltage. (Reprinted with permission from Ref. 186.)

voltage exists, although experimental results show different values for various reaction chambers. Typically, the critical bias voltage for a process pressure of 20–40 mbar is approximately −80 to −120 V (see Figure 2.11) [186]. The early stages of diamond nucleation on Si have been systematically studied by using atomic force microscopy (AFM) and reflection-high-energy-electron diffraction (RHEED) [187,96]. The RHEED patterns (see Figure 2.12) obtained for different nucleation times

2.4 DIAMOND EPITAXY

(a)

(b)

(c)

(d)

43

Figure 2.12 RHEED patterns obtained for different nucleation time (a) 7.5 min, (b) 10 min, (c) 12.5 min and (d) 15 min. (Reprinted with permission from Ref. 23.)

contained diffraction spots corresponding to diamond after several minutes of deposition, which suggested that a certain period of time is necessary for the initiation of nucleation. It is demonstrated that the diamond nuclei formed at the start of the nucleation are oriented. As the nucleation time increases, the nucleation density increases dramatically to 5 × 1010 cm−2 (see Figure 2.13) [96] and the nuclei become randomly oriented. It is therefore important to control the nucleation process for epitaxial diamond growth [97]. Schreck and associates [188,189] studied this processing space for heteroepitaxial nucleation of diamond on Si (001) using the bias process by x-ray diffraction texture measurements. It was found that the bias time lies within a distinct time interval and that the width of the time window and the bias time for optimal crystal alignment decrease sharply as the absolute value of the bias voltage decreases. A possible mechanism of the loss of diamond epitaxy was suggested by Jiang et al. [190] and Schreck et al. [188] after investigating diamond growth under bias conditions. According to this mechanism, slightly misoriented crystallites grow homoepitaxially on {100} facets and the crystal misorientation can be traced back to either the formation of defects during the homoepitaxial growth of the crystallites induced by ion bombardment or the increased renucleation of strongly misoriented grains.

44

SYNTHESIS OF DIAMOND FILMS

Figure 2.13 Nucleation density (islands/cm2 ) versus deposition time. In the abscissa, an induction time of 6.5 min is subtracted. The curve is obtained by computer modeling. The open and full circles in the plots represent data calculated from crystal size distribution and obtained by direct particle counting, respectively.

Textured diamond growth follows Van der Drift’s growth mode [137] if the growth process proceeds to longer times (larger thickness), and permits only crystallites that are oriented in the fastest growth direction (approximately parallel to the general growth direction) to grow further. As previously mentioned, the fastest growth direction can be varied by changing the process parameters, the most important of which are the substrate temperature and methane concentration [137]. Application of a proper bias potential to the substrate is also another parameter that leads to a strong selection of the growth direction [190]. The selection effect is attributed to the direction-dependent etching of the diamond crystal by H+ ions [191]. Therefore, film texture can be achieved for thin films. Both of the aforementioned approaches have been used to prepare epitaxially oriented films [135,149]. Heteroepitaxial [001]-oriented diamond films with considerably increased lateral grain size and strongly improved orientational perfection could be prepared by a microwave plasma-assisted chemical vapor deposition using a [001]-textured growth process on Si (001) substrates followed by a [110]-step-flow growth process [176,192]. The diamond films were characterized by atomic force microscopy, scanning electron microscopy, and transmission electron microscopy. The results indicate that diamond crystals increase their lateral dimensions at the (001) film surface either by coalescence of grains combined with a termination of the propagation of grain boundaries or by changing the grain boundary plane orientations from preferentially vertical to preferentially parallel directions with respect to the (001) growth faces. In the second case, the grains with relatively large angle deviation from the ideal epitaxial orientation are overgrown by those with relatively small angle deviation. As a result, the degree of orientational perfection of the films improves considerably in comparison to that of films prepared by the established process of [001]-textured growth. The presence of boron in the gas phase was found to strongly enhance the step-flow lateral grain growth. It was possible to achieve deposition of a thin boron-doped diamond films characterized by a full width at half maximum value of the measured tilt angle distribution of only 2.1◦ .

2.5 NANODIAMOND THIN FILMS

45

Early studies revealed that the growth of diamond on Si occurred through a β-SiC interlayer [193,194]. This perhaps is expected, considering that the lattice misfit between diamond and β-SiC (22%) is much smaller than that between diamond and silicon (52%). However, a direct lattice observation of the interface between silicon and epitaxially grown diamond crystals has been successfully performed [169,170]. It clearly demonstrated that the diamond nuclei are grown in direct contact with the silicon substrate (see Figure 2.9). Because the ratio of the lattice constants of silicon and diamond is close to 1.5, a nearly perfect 3 to 2 correspondence (1.5% mismatch) of lattice spacing is seen at the interface (e.g., every three Si (111) fringes are matched well with four diamond (111) fringe; the mismatch is about 1.5%). Individual 60◦ interface misfit dislocations for every ˚ can third (111) atomic plane, with a spacing between two such dislocations of about 7 A, be clearly identified (see Figure 2.9). The presented results are reproducible for many of the epitaxially oriented diamond grains observed and they provide strong evidence that diamond crystal can be epitaxially grown directly on Si. From the experimental results, it can be concluded that if the growth temperature is not too high and the incubation time is sufficiently short, diamond can grow directly on the Si substrate, avoiding the formation of the β-SiC interlayer. A thin epitaxial SiC intermediate layer is unnecessary for diamond heteroepitaxy. Theoretical investigations of the interface structure between diamond and silicon have also provided support for the direct growth of diamond on silicon [195,196]. Although great progress has been made in the heteroepitaxy of diamond on silicon, efforts are still needed to obtain large-area, single crystalline diamond films. For the heteroepitaxy of diamond directly on silicon, the following approaches may be taken into consideration. This is the first and essential requirement: The substrate surface must be clean, without surface contamination and oxidation, and the dangling bonds of the Si surface atoms must be saturated by hydrogen. During the BEN process, surface damage due to ion bombardment and the formation of amorphous carbon must be avoided. The bias voltage, biasing time, and pressure strongly influence crystal orientations. The bombardment of energetic ions above a critical energy is necessary for the formation of nuclei. A negative influence by ion bombardment on the alignment of the diamond grains has also been demonstrated. An improved epitaxy requires a compromise of the positive and negative influences. On the other hand, at present the vacuum for CVD diamond nucleation and growth is low and the residual gas in the chamber would contaminate or even oxidize the Si surface and change the surface status. Thus, improvement of the base vacuum in the growth chamber and the purity of gas source would be helpful in achieving better epitaxy.

2.5

NANODIAMOND THIN FILMS

With their unusual structure, properties, and applications, nanocrystalline (NC) films have been attracting rapidly increasing research interests [197]. Nanocrystalline diamond-thin film intrinsically possesses high uniformity, extremely high grain boundary density, and pin-hole-free ultra-smooth surface morphology even for film thicknesses well below 1 μm. Such films serve in a variety of applications [198–204]. Nanocrystalline diamond films were first deposited on Si substrates from a hydrogen-deficient carbon-containing noble gas CVD plasma [37]; phase-pure diamond film with nanometer-sized grains was also grown at very low hydrogen concentrations via a chemical pathway involving the insertion of C2 dimer into carbon-carbon and carbon-hydrogen bonds of the growing

46

SYNTHESIS OF DIAMOND FILMS

film [205]. One may obtain NC diamond films by adding nitrogen [206] to a high methane-containing CVD plasma or a very little amount of tetramethylsilane [207] to a very low methane-containing CVD plasma. Another special route developed to synthesize NC diamond films was based on the effect of ion-bombardment-induced, high-frequency secondary diamond nucleation [208,209]. By means of the ion bombardment as induced by the negative bias applied on the substrate at a voltage above 80 V, grain size in the films can be decreased and consequently the film morphology can be changed. At a bias voltage exceeding −140 V, very smooth NC diamond films can be obtained. Owing to the shallow ion penetration into the subsurface region (ion subplantation) [210], the sp3 –bonded carbon clusters are formed and act as the nucleation precursors. The negative bias applied to the substrate causes the positively charged ions in the growth chamber to accelerate toward the substrate’s surface and bombard it, thereby removing the contamination and facilitating cluster formation at the surface. These events in turn advance the diamond nucleation. Since early 1990s, the substrate bias-induced ion bombardment route has been an established one for the heterogeneous CVD diamond nucleation and made the heteroepitaxial growth of diamond on silicon a great success. On the other hand, only recently, research has been underway regarding the usage of continuous or intermittent ion bombardment effects during diamond film growth. Detailed investigations indicated that the combination of CVD diamond growth from neutral species (H and CH3 ) with the simultaneous bombardment of the growth surface by energetic particles of variable energies can provide a wide range of diamond phase modifications and thereby improved film properties. It was demonstrated [147,190] that, via H+ ion bombardment during the diamond CVD, largely variable diamond film growth can be obtained; from microcrystalline (MC), through a fine crystalline and fully (001) textured film morphology, to a nanocrystalline diamond growth can be obtained, independent of the substrate type. Due to the small grain size and ion bombardment, the grown nanocrystals show much higher density of crystal defects and larger lattice distortions in comparison to microcrystals. Due to the high compressive film stress introduced by ion subplantation, thicker films could hardly be prepared. The boron incorporation can, however, lower the intrinsic compressive stress of the NC diamond films, which is due to an interesting cancellation effect between the boron doping-induced tensile stress and the ion subplantation-induced compressive stress.

2.6

DIAMOND NANOCOMPOSITE FILMS

As for the research concerning diamond-related composite films, most of the work has been aimed at preparing diamond composites [211–218] with (1) carbides of molybdenum, tungsten, and titanium; (2) nitrides of silicon, aluminium, and titanium; (3) nickel alloy; and (4) copper–titanium–gold alloy by employing techniques such as modified MWCVD, metal powder plasma spraying, modified HFCVD, sputtering, electroplating, and laser ablation. Works reporting a-C/nanodiamond composite films [219–221] also exist in the literature. There are also a number of reports that are available regarding the preparation of (graphitic/amorphous) carbon-SiC nanocomposites [222–231]. The previously mentioned composites are not phase mixture diamond nanocomposite films. On the other hand, there are phase mixture diamond nanocomposite systems such as diamond/TiC [232,233], diamond/WC [234], and diamond/β-SiC [235–237].

2.6 DIAMOND NANOCOMPOSITE FILMS

47

The idea behind synthesis of nanocrystalline phase mixture composite films is to obtain films that possess a whole range of combined properties of diamond and the other phase (TiC, WC, β-SiC etc.). An example of a diamond/β-SiC nanocomposite film system will be briefly discussed. This film system is designed in such a way that the resultant material contains a required volume of nanometer-sized grains of both the components, such that the availability of a large volume of grain boundaries can be controlled based on the application requirement. The MWCVD technique was used to carry out the nanocomposite film depositions with the aid of hydrogen-methane-tetramethylsilane (TMS) gas mixtures. The presence of TMS in the gas phase along with hydrogen and methane not only results in a decrease of diamond crystallite size but also helps in the incorporation β-SiC as a second phase in the resulting film. The growth process of the nanocomposite films was attributed to the competitive nature of diamond- and β-SiC-forming species to occupy the reaction sites on the substrate [238]. Based on the microstructural analyses, these nanocomposite films are classified as granular-type composite films that contain diamond and β-SiC components as nanocrystalline grains distributed contiguously and laterally throughout the thickness of the film in a desired volume fraction combinatorial form [207]. Field emission (FE) SEM surface morphology image of a diamond/β-SiC nanocomposite thin film is shown in Figure 2.14. The content of β-SiC in the films could be increased by increasing TMS concentration in the gas phase. This factor was used in depositing films with gradient nature in a single process step, resulting in stress-free diamond top layers [239]. Indentation modulus and hardness values of the nanocomposite films revealed their linear relationship with βSiC content in the films. Very low friction coefficient values were also measured for the nanocomposite films [240]. It was also shown that the nanocomposite films are tougher than typical CVD polycrystalline diamond films and can be used as interlayers to obtain low-stress adherent diamond top layers on metallic substrates and cutting tool inserts [241].

Figure 2.14 Ultra-high magnification FE-SEM surface morphology image of nanocrystalline diamond/β-SiC composite film deposited on a BEN step pretreated (100) Si substrate. Bright regions represent diamond; darker regions represent β-SiC.

48

2.7

SYNTHESIS OF DIAMOND FILMS

CONCLUSIONS

Diamonds possess unmatchable mechanical, thermal, optical, electronic, electrochemical, electromechanical, biosensing, biomedical properties—and many others. Chemical vapor deposition of diamond-thin films was a major breakthrough in synthetic diamond research and has made a variety of diamond-based applications possible. With the aim of increasing the growth rate and improving the quality of diamond films, CVD methods—namely, HFCVD, MWCVD, RF-PCVD, DC-PCVD, ECR-MWCVD, and combustion flame-CVD—have been continuously developed. Standard diamond growth models based on both experimental and theoretical research works have been developed. At present, with the help of these models, CVD diamond growth process can be controlled to obtain diamond films according to the application requirement. A variety of film types—microcrystalline (poly), highly oriented, epitaxial, nano, and nanocomposite diamond—can now be readily synthesized. As a consequence, CVD diamond has opened the door to various applications. Nitrogen-doped diamond films can now be used as electron-emitting cathodes in flat-panel displays. Free-standing diamond windows that are transparent to infrared (IR), x-rays, high-power lasers, and high-power microwaves can be operated under severe temperature, pressure-differential, and reactivity conditions. Water treatment for organic effluents and production of strong oxidants employing conductive diamond electrodes are now possible. Also, CVD diamond has shown great promise as a potential radiation-detection material. It has also been successfully used as a substrate for surface acoustic-wave devices. Additionally, NC diamond films have been used as substrates for the preservation of DNA. The most sorted out application of diamond-thin films that has been already industrialized is using them as protective and/or wear-resistance coatings on cutting tools.

REFERENCES 1. J.T. Field, ed., Properties of Natural and Synthetic Diamond , Academic Press, London, 1992. 2. J.C. Angus, Thin Solid Films 1992, 216 , 126. 3. M.N. Yoder, in Synthetic Diamond: Emerging CVD Science and Technology, 1st ed. (K.E. Spear, J.P. Dismukes eds.), John Wiley & Sons, 1994. 4. F.P. Bundy, H.T. Hall, H.M. Strong, R.H. Wentorf, Nature 1955, 176 , 51. 5. J.E. Field, The Properties of Diamonds, Academic Press, New York, 1979. 6. K. Nassau, J. Nassau, J. Cryst. Growth 1979, 46 , 157. 7. R.C. De Vries, Ann. Rev. Mater. Sci . 1987, 17 , 161. 8. P.K. Bachmann, R. Messier, Chem. Eng. News 1989, 67 , 24. 9. W.G. Eversole, N. Y. Kenmore, US Patents 3030187 and 3030188, 1962. 10. J.C. Angus, H.C. Will, W.S. Stanko, J. Appl. Phys. 1968, 39 , 2915. 11. B.V. Deryagin, D.V. Fedoseev, V.M. Lukyanovich, B.V. Spitsyn, A.V. Ryanov, A.V. Lavrentyev, J. Cryst. Growth 1968, 2 , 380. 12. D.J. Poferl, N.C. Gardner, J.C. Angus, J. Appl. Phys. 1973, 44 , 1418. 13. B.V. Derjaguin, D.V. Fedoseev, Sci. Am. 1975, 223 , 102. 14. B.V. Spitsyn, L.L. Builov, B.V. Derjaguin, J. Cryst. Growth 1981, 52 , 219. 15. S. Matsumoto, Y. Sato, M. Kamo, N. Setaka, Jpn. J. Appl. Phys. 1982, 21 , part 2 , 183. 16. S. Matsumoto, Y. Sato, M. Tsutsumi, N. Setaka, J. Mater. Sci . 1982, 17 , 3106.

REFERENCES

17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

43. 44. 45. 46. 47. 48. 49. 50. 51.

49

M. Kamo, U. Sato, S. Matsumoto, N. Setaka, J. Cryst. Growth 1983, 62 , 642. S. Matsumoto, Y. Matsui, J. Mater. Sci . 1983, 18 , 1785. S. Matsumoto, J. Mater. Sci. Lett . 1985, 4 , 600. Y. Saito, S. Matsuda, S. Nogita, J. Mater. Sci. Lett . 1986, 5 , 565. K. Suzuki, A. Sawabe, H. Yasuda, T. Inuzuka, Appl. Phys. Lett . 1987, 50 , 728. Y. Hirose, M. Mitsuizumi, New Diam. 1988, 4 , 34. S.-T. Lee, Z. Lin, X. Jiang, Mater. Sci. Eng. R 1999, 25 , 123. P.W. May, Endeavour 1995, 19 , 101. K. Okano, S. Koizumi, S.R.P. Silva, G.A.J. Amaratunga, Nature 1996, 381 , 140. M. Seal, Interdiscipl. Sci. Rev ., 1989, 14 , 64. M. Panizza, G. Cerisola, Electrochem. Acta 2005, 51 , 191. D. Gandini, E. Mah´e, P.A. Michaud, W. Haenni, A. Perret, C. Comninellis, J. Appl. Electrochem. 2000, 20 , 45. P.A. Michaud, E. Mah´e, W. Haenni, A. Perret, C. Comninellis, Electrochem. Solid-State Lett . 2000, 3 , 77. M. Bucciolini, Nucl. Instr. Meth. Phys. Res. A 2005, 552 , 189. B. Dishchler, C. Wild, Low-Pressure Synthetic Diamond Manufacturing and Applications, Springer, Heidelberg, 1998. B. Bi, W.-S. Huang, J. Asmussen, B. Golding, Diam. Rel. Mat . 2002, 11 , 677. W. Yang, O. Auciello, J.E. Butler, W. Cai, J.A. Carlisle, J. Gerbi, D.M. Gruen, T. Knickerbocker, T.L. Lasseter, J.N. Russel, Jr., L.M. Smith, R.J. Hamers, Nature Mater. 2002, 1 , 253. W. Ahmed, H. Sein, N. Ali, J. Gracio, R. Woodwards, Diamond Rel. Mat ., 2003, 12 , 1300. H. Schmellenmeier, Exp. Tech. Phys. 1953, 1 , 49. R.A. Rudder, J.B Posthill, R.J. Markunas, Electron. Lett . 1989, 25 , 1220. D.M. Gruen, S. Liu, A.R. Krauss, J. Luo, X. Pan, Appl. Phys. Lett . 1994, 64 , 1502. P.H. Gaskell, A. Saeed, P. Chieux, D.R. McKenzie, Phys. Rev. Lett . 1991, 67 , 1286. J.F. Prins, Mat. Sci. Rep. 1992, 7 , 271. Q. Chen, Z. Lin, Appl. Phys. Lett . 1996, 68 , 2450. Q. Chen, J. Yang, Z. Lin, Appl. Phys. Lett . 1995, 67 , 1853. H.S. Shin, N.G. Glumac, D.G. Goodwin, Fourth Int. Conf. On New Diamond Science & Technology (S. Sato, N. Fujimori, O. Fukunaga, M. Kamo, K. Kobashi, M. Yoshikawa, eds.) Myu, Tokyo, 1994, pp. 27–30. W.M. Lau, S.-T. Lee, F. Qin, Surface Modification Technologies V , 1992, p. 263. X.S. Sun, N. Wang, H.K. Woo, W.J. Zhang, G. Yu, I. Bello, C.S. Lee, S.T. Lee, Diam. Rel. Mat . 1999, 8 , 1414. P. Aublane, V.P. Novikov, L.V. Kuznetsova, M. Mermoux, Diam. Rel. Mat . 2001, 10 , 942. S. Ferro, J. Mater. Chem. 2002, 12 , 2843. R.F. Davis, ed., Diamond Films and Coatings, Noyes Publications, New Jersey, 1992. N. Fujimori, A. Ikegaya, T. Imai, K. Fukushima, N. Ota, in “Diamond and Diamond-like Films” (J. Dismukes et al,. eds.), Electrochem. Soc. Proc. 1989, PV 89-12 , 465. Y. Hirose, N. Kondo, Program and Book of Abstracts, Japan Applied Physics 1988 spring meeting, 1988, p. 434. G. Janssen, W.J.P. van Enckefort, J.J.D. Schaminee, W. Vollenberg, L.J. Giling, M. Seal, J. Cryst. Growth 1990, 104 , 752. K. Doverspike, J.E. Butler in Diamond 1992 (P.K. Bachmann, A.T. Collins, M. Seal, eds.), Elsevier Sequoia, Lausanne, 1993, p. 1078.

50 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.

83.

SYNTHESIS OF DIAMOND FILMS

M. Murakawa, S. Takeuchi, Y. Hirose, Surf. Coat. Technol . 1990, 43/44 , 22. M.A. Cappelli, P.H. Paul, J. Appl. Phys. 1990, 675 , 2596. A.R. Badzian, R.C. De Vries, Mater. Res. Bull . 1988, 23 , 385. K. Kurihara, K. Sasaki, M. Kawarada, N. Koshino, Appl. Phys. Lett . 1988, 526 , 437. M. Kawarada, K. Kurihara, K. Sasaki, A. Teshima, K. Koshino in Diamond Optics II (A. Feldman, S. Holly, eds.) SPIE Proc., 1146, Bellingham, WA, 1990, p. 28. N. Ohtake, M. Yoshikawa, J. Electrochem. Soc. 1990, 137 , 717. N. Fujimori, New Diam. 1988, 2 , 10. D.E. Meyer, T.B. Kustka, R.O. Dillon, J.A. Woollam in Diamond Optics II (A. Feldman, S. Holly, eds.), SPIE Proc., 1146, Bellingham, WA, 1990, p. 21. A. Hiraki, H. Kawarada, J. Wei, J.I. Suzuki, Surf. Coat. Technol . 1990, 43/44 , 10. T. Yara, M. Yuasa, M. Shimizu, H. Makita, A. Hatta, J.I. Suzuki, T. Ito, A. Hiraki, Jpn J. Appl. Phys. 1994, 33 , 4405. D. Mantei, Z. Ring, M. Schweizer, S. Tlali, H. E. Jackson, Jpn. J. Appl. Phys. 1996, 35 , 2516. M. Frenklach, R. Kematick, D. Huang, W. Howard, K.E. Spear, J. Appl. Phys. 1989, 66 , 395. S.E. Stein, Nature 1990, 346 , 517. H. Liu, D.S. Dandy, Diam. Rel. Mat . 1995, 4 , 1173. K. Mitsuda, Y. Kojima, T. Yoshida, K. Akashi, J. Mater. Sci . 1987, 22 , 1557. A. Sawabe, T. Inuzuka, Thin Solid Films 1986, 137 , 89. N. Fujimori, T. Imai, A. Doi, Vacuum 1986, 36 , 99. C.P. Chang, D.L. Flamm, D.E. Ibbotson, J.A. Mucha, J. Appl. Phys. 1988, 63 , 1744. P.A. Denning, D.A. Stevenson, Proceedings of the 2nd International Conference on New Diamond Science and Technology, Pittsburg, PA, MRS, 1991, pp. 403–408. B. Singh, O. Mesker, A.W. Levine, Y. Arie, Growth of Polycrystalline Diamond Particles and Films by Hot-Filament Chemical Vapor Deposition, Report CN-5300, David Sarnoff Research Center, SRI International, Princeton, NJ, 1988. B. Singh, Y. Arie, A.W. Levine, O.R. Mesker, Appl. Phys. Lett . 1988, 526 , 451. C.P. Beetz Jr, T.A. Perry, Hot and Cold Wall Filament-Assisted Growth of Diamond Films, General Motors Research Publication GMR-6093, November 24, 1987, pp. 1–21. S. Yugo, T. Kimura, H. Kanai in Science and Technology of New Diamond (S. Saito, O. Fukunaga, M. Yoshihara, eds.), KTK Scientific Pub., Tokyo, 1990, pp. 119–128. S. Iijima, Y. Aikawa, K. Baba, J. Mater. Res. 1991, 6 , 1491. S.P. Murarka, M.C. Peckerar, Electronic Materials Science and Technology, Academic Press, San Diego, CA, 1989. S. Yugo, A. Izumi, T. Kanai, T. Muto, T. Kimura, Proceedings of the 2nd International Conference on New Diamond Science and Technology, Pittsburg, PA, MRS, 1991, pp. 385–390. J. Yang, X. Su, Q. Chen, Z. Lin, Appl. Phys. Lett . 1995, 66 , 3284. P.K. Bachmann, R. Weimer, W. Drawl, Y. Liou, R. Messier, Diamond Technology Initiative Symposium, Paper T20, SDIC/IST ONR, Crystal City, VA, July 12–14, 1988. M. Marchywka, P.E. Pehrsson, D.J. Vestyck, D. Moses, Appl. Phys Lett . 1993, 63 , 3521. A.R. Kirkpatrick, B.W. Ward, N.P. Sconomou, J. Vac. Sci. Technol. B 1989, 7 , 1947. J.C. Angus, Z. Li, M. Sunkara, R. Gat, A.B. Anderson, S.P. Mehandru, M.W. Geis in A.J. Purdes, J.C. Angus, R.F. Davis, B.M. Meyerson, K.E. Spear, M. Yoder (Eds.), Proceedings of the 2nd International Symposium on Diamond Materials, The Electrochemical Society, Pennington, New Jersey, 1991, p. 125. Z. Feng, K. Komvopoulos, I.G. Brown, D.C. Bogy, J. Appl. Phys. 1993, 744 , 2841.

REFERENCES

84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123.

51

A.A. Morrish, P.E. Pehesson, Appl. Phys. Lett . 1991, 59 , 417. J. Singh, M. Vellaikal, J. Appl. Phys. 1993, 73 , 2831. K.V. Ravi, C.A. Koch, Appl. Phys. Lett . 1992, 57 , 348. K.V. Ravi, C.A. Koch, H.S. Hu, A. Joshi, J. Mater. Res. 1990, 5 , 2356. S. Yugo, T. Kanai, T. Kimura, T. Muto, Appl. Phys. Lett . 1991, 58 , 1036. B.R. Stoner, G.-H.M. Ma, S.D. Wolter, J.T. Glass, Phys. Rev. B 1992, 45 , 11067. X. Jiang, C.-P Klages, Diam. Rel. Mat . 1993, 2 , 1112. ` X. Jiang, C.-P. Klages, R. Zachai, M. Hartweg, H.-J. FuEsser, Appl. Phys. Lett . 1993, 62 , 3438. B.R. Stoner, J.T. Glass, Appl. Phys. Lett . 1992, 60 , 698. S. Yugo, T. Kimura, T. Kanai, Diam. Rel. Mat . 1993, 2 , 328. J. Gerber, S. Sattel, K. Jung, H. Ehrhard, J. Robertson, Diam. Rel. Mat . 1995, 4 , 559. B.R. Stoner, G.H. Ma, S.D. Wolter, W. Zhou, Y.-C. Wang, R.F. Davis, J.T. Glass, Diam. Rel. Mat . 1993, 2 , 142. X. Jiang, K. Schiffmann, C.-P. Klages, Phys. Rev. B 1994, 50 , 8402. X. Jiang, C.-P. Klages, Phys. Stat. Sol . 1996, 154 , 175. ` F. Stubhan, M. Ferguson, H.-J. FuEsser, R.J. Behm, Appl. Phys. Lett . 1995, 66 , 1900. S.T. Lee, Y.W. Lam, Z. Lin, Y. Chen, Q. Chen, Phys. Rev. B 1997, 5524 , 15937. Z. Liu, B.Q. Zong, Z. Lin, Thin Solid Films 1995, 254 , 3. M. Tsuda, M. Nakajima, S. Oikawa, J. Am. Chem. Soc. 1986, 108 , 5780. M. Tsuda, M. Nakajima, S. Oikawa, Jpn. J. Appl. Phys. 1987, 26 , L527. C.J. Chu, M.P. D’Evelyn, R.H. Hauge, J.L. Margrave, J. Appl. Phys. 1991, 70 , 1659. L.R. Martin, M.W. Hill, J. Mater. Sci. Lett . 1990, 9 , 621. S.J. Harris, L.R. Martin, J. Mater. Res. 1990, 5 , 2313. S.J. Harris, Appl. Phys. Lett . 1990, 56 , 2298. M. Frenklach, K.E. Spear, J. Mater. Res. 1988, 3 , 133. D. Huang, M. Frenklach, M. Maroncelli, J. Phys. Chem. 1988, 92 , 6379. M. Frenklach, H. Wang, Phys. Rev. B 1991, 43 , 1520. D. Huang, M. Frenklach, J. Phys. Chem. 1991, 95 , 3692. F.G. Celii, J.E. Butler, Appl. Phys. Lett . 1989, 54 , 1031. M. Frenklach, J. Appl. Phys. 1989, 65 , 5142. D.G. Goodwin, G.G. Gavillet, J. Appl. Phys. 1990, 68 , 6393. D.M. Gruen, MRS Bull . 1998, 239 , 32. P.C. Redfern, D.A. Horner, L.A. Curtiss, D.M. Gruen, J. Phys. Chem. 1996, 100 , 11654. B. Sun, X.P. Zhang, Q.Z. Zhang, Z.D. Lin, Appl. Phys. Lett . 1993, 62 , 31. T. Shimanouchi, Table of Vibrational Frequencies, Natl. Bur. Stand. Ref. Data Ser., Natl. Bur. Stand. (U.S.) Circ. No. 39, U.S. GPO, Washington, DC, 1968, Consolidated Vols. I and II. T. Shimanouchi, J. Phys. Chem. Ref. Data 1977, 6 , 993. H. Ibach, D.L. Mills, Electron Energy Loss Spectroscopy and Surface Vibrations, Academic Press, New York, 1982. L.B. Zhao, K.A. Feng, Z.D. Lin, Diamond Rel. Mater. 1993, 3 , 155. A.V. Hamza, G.D. Kubiak, R.H. Stulen, Surf. Sci . 1990, 237 , 35. S.P. Mehandru, A.B. Anderson, Surf. Sci . 1991, 248 , 369. T. Tsuno, T. Imai, Y. Nishibayasi, K. Hamada, N. Fujimori, Jpn. J. Appl. Phys. 1991, 30 , 1063.

52 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161.

SYNTHESIS OF DIAMOND FILMS

L.F. Sutcu, Appl. Phys. Lett . 1992, 60 , 1685. B. Sun, X. Zhang, Z.D. Lin, Phys. Rev. B 1993, 47 , 9816. K.A. Feng, Z.D. Lin, J. Phys Chem. Solids 1994, 55 (6), 525. K. Kobashi, K. Nishimura, Y. Kawate, T. Horiuchi, Phys. Rev. B 1988, 38 , 4067. Y.J. Baik, K.Y. Eun, Applications of Diamond Films and Related Materials, Elsevier, New York, 1991, p. 521. C. Wild, N. Herres, P. Koidl, J. Appl. Phys. 1990, 68 , 973. Z.D. Lin, B. Sun, J. Vac. Soc. Jpn. 1994, 37 , 579. S.J. Harris, A.M. Weiner, J. Appl. Phys. 1990, 67 , 6520. S.J. Harris, D.N. Betton, A.M. Weiner, S.J. Schmieg, J. Appl. Phys. 1989, 66 , 5353. J.E. Butler, R L. Woodin, Phil. Trans. R. Soc. Lond. A 1993, 342 , 209. D.G. Goodwin, J.E. Butler, in Handbook of Industrial Diamonds and Diamond Films (M.A. Prelas, G. Popovici, L.G. Bigelow, eds.), Marcel Dekker Inc., 1998. C. Wild, P. Koidl, W. M¨uller-Sebert, H. Walcher, R. Kohl, N. Herres, R. Locher, R. Samlenski, R. Brenn, Diam. Rel. Mat . 1993, 3 , 158. R.A. Bauer, N.M. Sbrockey, W.E. Brower Jr., J. Mater. Res. 1993, 8 , 2858. A. van der Drift, Philips Res. Reports 1967, 22 , 267. X. Jiang, V.V.S.S. Srikanth, Y.L. Zhao, R.Q. Zhang, Appl. Phys. Lett . 2008, 92 , 243107. P.W. May, Y.A. Mankelevich, J. Phys. Chem. C 2008, 112 , 12432. S. Skokov, B. Weiner, M. Frenklach, J. Phys. Chem. 1994, 98 , 7073. P.W. May, M.N.R. Ashfold, Y.A. Mankelevich, J. Appl. Phys. 2007, 101 , 053115. Y.A. Mankelevich, P.W. May, Diam. Rel. Mat . 2008, 17 , 1021. P.W. May, N.L. Allan, M.N.R. Ashfold, J.C. Richley, Y.A. Mankelevich, J. Phys.: Condens. Matter 2009, 21 , 364203. B.J. Waclawski, D.T. Pierce, N. Swanson, R.J. Celotta, J. Vac. Sci. Technol . 1982, 21 , 368. T.R. Anthony, Vacuum 1990, 41 , 1356. K. Ishibashi, S. Furukawa, Jpn. J. Appl. Phys. 1985, 24 , 912. X. Jiang, W.J. Zhang, M. Paul, C.-P. Klages, Appl. Phys. Lett . 1996, 68 , 1927. W.J. Zhang, X. Jiang, Appl. Phys. Lett . 1996, 68 , 2195. W.J. Zhang, X. Jiang, Y. B. Xia, J. Appl. Phys. 1997, 82 , 1896. Y. Muranaka, H. Yamashita, H. Miyadera, J. Vac. Sci. Technol. A. 1991, 9 , 76. S.J. Harris, A. M. Weiner, Appl. Phys. Lett . 1989, 55 , 2179. Y. Liou, A. Inspektor, R. Weimer, D. Knight, R. Messier, J. Mater. Res. 1990, 5 , 2305. B. Sun, X. Zhang, Q. Zhang, Z. Lin, J. Appl. Phys. 1993, 73 , 4614. S.-T. Lee, Y.W. Lam, Z.D. Lin, Y. Chen, Z.Q. Gao, Phys. Rev. B. 1996, 54 , 14185. A. Badzian, T. Badzian, X.H. Wang, Carbon 1990, 28 , 804. M. Kamo, H. Yurimoto, Y. Sato, Appl. Surf. Sci . 1988, 33/34 , 553. B.V. Derjaguin, B.V. Spitsyn, A.E. Gorodetsky, A.P. Zakharov, L.I. Bouilov, A.E. Aleksenko, J. Crystal. Growth 1975, 31 , 44. H. Nakazawa, Y. Kanayama, M. Kamo, K. Osumi, Thin Solid Films 1987, 151 , 199. N. Fujimori, T. Imai, H. Nakahata, H. Shiomi, Y. Nishibayashi, Mater. Res. Soc. Symp. Proc. 1990, 162 , 109. M. Yoshikawa, H. Ishida, A. Ishitani, T. Murakami, S. Koizumi, T. Inuzuka, Appl. Phys. Lett . 1990, 57 , 428. S. Koizumi, T. Murakami, K. Suzuki, T. Inuzuka, Appl. Phys. Lett . 1990, 57 , 563.

REFERENCES

53

162. S. Koizumi, T. Inuzuka, Jpn. J. Appl. Phys. 1993, 32 , 3920. 163. A. Argoitia, J.S. Ma, J.C. Angus, Appl. Phys. Lett . 1993, 63 , 1336. 164. A. Argoitia, J.C. Angus, J.S. Ma, L. Wang, P. Pirouz, W. Lambrecht, J. Mater. Res. 1994, 9 , 1849. 165. R.O. Jones, O. Gunnarsson, Rev. Mod. Phys. 1989, 61, 689. 166. T. Tomizuka, S. Shikata, Jpn. J. Appl. Phys. 1993, 32 , 3938. 167. R.C. De Vries, Cubic Boron Nitride: Handbook of Properties, G.E. Representative. No. 72 CRD, 1972, p. 178. 168. T. Suzuki, A. Argoitia, Phys. Stat. Sol . 1996, 154 , 239. 169. X. Jiang, C.L. Jia, Appl. Phys. Lett . 1995, 67 , 1197. 170. C.L. Jia, K. Urban, X. Jiang, Phys. Rev. B 1995, 52 , 5164. 171. J. Yang, Z.D. Lin, L.X. Wang, S. Jin, Z. Zhang, Appl. Phys. Lett . 1994, 65 , 3203. 172. K. Feng, J. Yang, Z.D. Lin, Phys. Rev. B. 1995, 51 , 2264. 173. J. Yang, Z.D. Lin, L.X. Wang, S. Jin, Z. Zhang, J. Phys. D: Appl. Phys. 1995, 28 , 1153. 174. Q.J. Chen, L.X. Wang, Z. Zhang, J. Yang, Z.D. Lin, Appl. Phys. Lett . 1996, 68 , 176. ` ` 175. X. Jiang, C.-P. Klages, M. RoEsler, R. Zachai, M. Hartweg, H.-J. FuEsser, Appl. Phys. A 1993, 57 , 483. 176. X. Jiang, K. Schiffmann, C.-P. Klages, D. Wittorf, C.L. Jia, K. Urban, W. Jaeger, J. Appl. Phys. 1998, 83 , 2511. 177. H. Kawarada, C. Wild, N. Herres, R. Locher, P. Koidl, J. Appl. Phys. 1997, 81 , 3490. 178. A. Argoitia, J.C. Angus, L. Wang, X.I. Ning, P. Pirouz, J. Appl. Phys. 1993, 73 , 4305. 179. W. Zhu, P.C. Yang, J.T. Glass, Appl. Phys. Lett . 1993, 63 , 1640. 180. T. Tachibana, Y. Yokota, K. Miyata, K. Kobashi, Y. Shintani, Diam. Rel. Mat . 1996, 5 , 197. 181. W. Liu, D.A. Tucker, P. Yang, J.T. Glass, J. Appl. Phys. 1995, 78 , 1291. 182. K. Ohtsuka, K. Suzuki, A. Sawabe, T. Inuzuka, Jpn. J. Appl. Phys. Part 2 . 1996, 35 , L1072. 183. K. Ohtsuka, H. Fukuda, K. Suzuki, A. Sawabe, Jpn. J. Appl. Phys. Part 2 . 1997, 36 , L1214. 184. T. Saito, S. Tsuruga, N. Ohya, K. Kusakabe, S. Morooka, H. Maeda, A. Sawabe, K. Suzuki, Diam. Rel. Mat . 1998, 7 , 1381. 185. M. Schreck, H. Roll, B. Stritzker, Appl. Phys. Lett . 1999, 74 , 650. ` 186. X. Jiang, C.-P. Klages, R. Zachai, M. Hartweg, H.-J. FuEsser, Diam. Rel. Mat . 1993, 2 , 407. 187. X. Jiang, K. Schiffimann, A. Westphal, C.-P. Klages, Appl. Phys. Lett . 1993, 63 , 1203. 188. M. Schreck, K.H. Thuerer, R. Klarmann, B. Stritzker, J. Appl. Phys. 1997, 81 , 3096. 189. M. Schreck, K.H. Thuerer, B. Stritzker, J. Appl. Phys. 1997, 81 , 3092. 190. X. Jiang, W.J. Zhang, C.-P. Klages, Phys. Rev. B 1998, 58 , 7064. 191. W. J. Zhang, X. Jiang, C.-P. Klages, J. Cryst. Growth 1997, 171 , 485. 192. X. Jiang, C.L. Jia, Appl. Phys. Lett . 1996, 69 , 3902. 193. S.D. Wolter, B.R. Stoner, J.T. Glass, P.J. Ellis, D.S. Buhaenko, C.E. Jenkins, P. Southworth, Appl. Phys. Lett . 1993, 6211 , 1215. 194. W. Zhu, X.H. Wang, B.R. Stoner, G.H.M. Ma, H.S. Kong, M.W.H. Braun, J.T. Glass, Phys. Rev. B 1993, 47 , 6529. 195. R.Q. Zhang, W.L. Wang, J. Esteva, E. Bertran, Thin Solid Films 1998, 317 , 6. 196. X. Jiang, R.Q. Zhang, G. Yu, S.T. Lee, Phys. Rev. B 1998, 58 , 15351. 197. C.C. Koch, I.A. Ovid’ko, S. Seal, S. Veprek, Structural Nanocrystalline Materials: Fundamentals and Applications, Cambridge University Press, Cambridge, UK, 2007. 198. D.M. Gruen, Ann. Rev. Mater. Sci . 1999, 29 , 211.

54

SYNTHESIS OF DIAMOND FILMS

199. A.R. Krauss, O. Auciello, M.Q. Ding, D.M. Gruen, Y. Huang, V.V. Zhirnov, E.I. Givargizov, A. Breskin, R. Chechen, E. Shefer, V. Konov, S. Pimenov, A. Karabutov, A. Rakhimov, N. Suetin, J. Appl. Phys. 2001, 89 , 2958. 200. O.A. Williams, M. Nesladek, Phys. Stat. Sol. A 2006, 203 , 3375. 201. Y.P. Ma, F.H. Sun, H.G. Xue, Z.M. Zhang, M. Chen, Diam. Rel. Mat . 2007, 16 , 481. 202. O. Auciello, S. Pacheco, A.V. Sumant, C. Gudeman, S. Sampath, A. Datta, R.W. Carpick, V.P. Adiga, P. Zurcher, Z. Ma, H.C. Yuan, J.A. Carlisle, B. Kabius, J. Hiller, S. Srinivasan, IEEE Microw. Mag. 2007, 8 , 61. 203. C.E. Nebel, B. Rezek, D.C. Shin, H. Uetsuka, N. Yang, J. Phys. D: Appl. Phys. 2007, 40 , 6443. 204. T.-H. Kim, W.M. Choi, D.-H. Kim, M.A. Meitl, E. Menard, H.Q. Jiang, J.A. Carlisle, J.A. Rogers, Adv. Mater. 2008, 20 , 2171. 205. D.M. Gruen, S. Liu, A.R. Krauss, J. Luo, X. Pan, Appl. Phys. Lett . 1994, 64 , 1502. 206. D.M. Gruen, MRS Bull . 1998, 23 , 32. 207. V.V.S.S. Srikanth, T. Staedler, X. Jiang, Appl. Phys. A 2008, 91, 149. 208. X. Jiang, C.Z. Gu, L. Sch¨afer, C.-P. Klages, Proceedings of ADC/FCT’99 (M. Yoshikawa, Y. Koga, Y. Tzeng, C.-P. Klages, K. Miyoshi, eds.), September 3, 1999, Tsukuba, Japan, p. 28. 209. C.Z. Gu, X. Jiang, J. Appl. Phys. 2000, 88 , 1788. 210. Y. Lifshitz, S.R. Kasi, J.W. Rabalais, W. Eckstein, Phys. Rev. B 1990, 41 , 10468. 211. M. Kawadara, K. Kurihara, K. Sasaki, J. Appl. Phys. 1992, 71 , 1442. 212. C. Tsai, J.C. Nelson, W.W. Gerberich, J. Heberlein, E. Pfender, Diam. Rel. Mat . 1993, 2 , 617. 213. W.D. Fan, X. Chen, K. Jagannadham, J. Narayan, J. Mater. Res. 1994, 9 , 2850. 214. M. Nagano, T. Urasaki, K. Wakiyama, M. Komori, Diam. Rel. Mat . 1997, 6 , 501. 215. M.L. Terranova, S. Piccirillo, V. Sessa, M. Rossi, G. Cappuccio, Chem. Vap. Deposition 1999, 5 , 101. 216. M.S. Raghuveer, S.N. Yoganand, K. Jagannadham, J. Bailey, R.L. Lemaster, Wear 2002, 253 , 1194. 217. K. Jagannadham, H. Wang, J. Appl. Phys. 2002, 91 , 1224. 218. F.A. Khalid, O. Beffort, U.E. Klotz, B.A. Keller, P. Gasser, Diam. Rel. Mat . 2004, 13 , 393. 219. X.T. Zhou, Q. Li, F.Y. Meng, I. Bello, C.S. Lee, S.T. Lee, Y. Lifshitz, Appl. Phys. Lett . 2002, 80 , 3307. 220. X.T. Zhou, X.M. Meng, F.Y. Meng, Q. Li, I. Bello, W. J. Zhang, C.S. Lee, S.T. Lee, Y. Lifshitz, Diam. Rel. Mat . 2003, 12 , 1640. 221. W.J. Hsieh, P.S. Shih, J.H. Lin, C.C. Lin, U.S. Chen, S.C. Huang, Y.S. Chang, H.C. Shih, Thin Solid Films 2004, 469 , 120. 222. S. Yajima, T. Hirai, J. Mater. Sci . 1969, 4 , 416. ˇ Suˇznjevi´c, I. Deˇzarov, A. Mihajlovi´c, D. Cerovi´c, Carbon 1970, 8 , 283. 223. S. Marinkovi´c, C. 224. J.L. Kaae, T.D. Gulden, J. Amer. Ceram. Soc. 1971, 54 , 605. 225. J.J. Nickl, C.V. Braunm¨uhl, J. Less-Common Metals 1971, 25 , 303. 226. Y.C. Wang, M. Sasaki, T. Hirai, J. Mater. Sci . 1991, 26 , 5495. 227. W.C. Liu, S.J. Sun, M.J. Li, Y.L. Wei, J. Mater. Sci. Letters 1993, 12 , 886. 228. M.R. Richards, A.C. Richards, M. Taya, F.S. Ohuchi, J. Vac. Sci. Technol. A 1995, 13 , 1196. 229. Y. Kim, J.T. Choi, J.K. Choi, K.H. Auh, Mater. Lett . 1996, 26 , 249. 230. I.M. Kostjuhin, S.V. Sotirchos, Ind. Eng. Chem. Res. 2001, 40 , 2586. 231. J.I. Kim, W.J. Kim, D.J. Choi, J.Y. Park, W.-S. Ryu, Carbon 2005, 43 , 1749.

REFERENCES

232. 233. 234. 235. 236. 237. 238. 239. 240.

55

L. Xiang, Ph.D. Thesis, Technical University Braunschweig 2002. F.Z. Ding, Y.L. Shi, Surf. Coat. Technol . 2007, 9-11 , 5050. H.A. Samra, R.J. Hong, X. Jiang, Chem. Vap. Deposition 2007, 13 , 17. X. Jiang, C.-P. Klages, Appl. Phys. Lett . 1992, 61 , 1629. X. Jiang, C.-P. Klages, Diam. Rel. Mat . 1993, 2 , 523. Y.L. Shi, M.H. Tan, X. Jiang, J. Mater. Res. 2002, 17 , 1241. V.V.S.S. Srikanth, M.H. Tan, X. Jiang, Appl. Phys. Lett . 2006, 88 , 073109. V.V.S.S. Srikanth, T. Staedler, X. Jiang, Diam. Rel. Mat . 2009, 18 , 1326. T. Staedler, V.V.S.S. Srikanth, X. Jiang, Mater. Res. Soc. Symp. Proc. 2006, 890 , 0890Y01-04.1. 241. V.V.S.S. Srikanth, Deposition and Characterization of Nanocrystalline Diamond/ β-SiC Composite Film System, Shaker Verlag GmbH, Aachen, Germany, 2008.

3 Types of Conducting Diamond Materials and Their Properties Marco A. Quiroz and Erick R. Bandala

3.1

INTRODUCTION

Historically, carbon has played, an important role in electrochemistry development. However, its use until the last 20 years has been restricted mostly to graphite, one of its natural allotropic forms. As traditionally taught in material science courses, natural carbon possesses two allotropic, structurally different forms: graphite and diamond. Since the now classical works by Curl and Smalley in 1991 [1], fullerenes have been also recognized as a third allotropic form of carbon. Graphite exhibits a structure of layers of carbon atoms arranged in hexagonal rings. ˚ In diamond structure, each carbon The distance between successive layers is 3.35 A. 3 atom is sp -hybridized, giving a tetrahedral geometrical arrangement. In both cases, carbon atoms are linked by covalent bonds. Fullerenes constitute a wide family (from C20 to C720), with closed structures and variable stability; angle strain in the sp2 carbon atoms is a consequence of bond bending to adopt a curve structure. Their stability results from a mix of topological and chemical properties, beyond the subject of this chapter. The diverse electronic structure in carbon atoms at the different structural arrangements determine, in most cases, whether or not the different carbon allotropic forms show electrical conductivity, considered as an unavoidable requirement for their use in

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

57

58

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

electrochemical applications. In this way, it is noteworthy that electrical conductivity requires nonbonding electrons (valence band electrons) to be able to move through the solid (into the conduction band). Separation between both valence and conduction band (the bandgap) determines the electrical conduction properties of the different carbon allotropes. The graphite structure, as mentioned earlier, consists of layers of carbon atoms arranged in hexagonal rings that are stacked in a sequence ABAB . . . (see Figure 3.1a) ˚ The atoms are all sp 2 -hybridized in the with a separation between them of 3.35 A. hexagonal arrangement on every layer, which allows every carbon atom to get three ˚ length covalent bonds oriented 120◦ on the hexagonal plane. The remaining 1.42 A unhybridized 2p orbital is perpendicular to the hexagonal plane and it is used in π bonding. The three σ orbitals contribute six electrons to the valence band. The 2p atomic orbital contributes 2 electrons to the partially filled conduction band (see Figure 3.1b). Superimposition of parallel layers generates an interlayer region with an electronic density that allows an electrical conductivity of 3 × 104 S·cm−1 , very different from the value determined for electrical conductivity in the perpendicular direction to the planes (5 S·cm−1 ). This laminar structure is “glued” by p − p and van der Waals interaction providing graphite with a high anisotropy degree.

(a)

(b)

Energy, E

3σ*sp2 2p atomic orbitals

π2p

Conduction band partially filled

sp2 hybrid orbitals 2s atomic orbital

Figure 3.1

2p atomic orbitals

3σsp2

Valence band completely filled

(a) Structural representation and (b) Electronic diagram of hexagonal graphite.

3.1 INTRODUCTION

Figure 3.2

59

Different structure and size of fullerenes (C60 , C540 , and simple wall nanotube).

Fullerenes are carbon-based structures that vary in size and stability: buckminsterfullerene (C60 ) has 60 carbon atoms and measures around 1 nm of diameter; higher-order fullerenes can have numbers from 540 to 720 carbon atoms (C540 –C720 ) in quasi-spherical geometries [1,2], as shown in Figure 3.2. Every single structure is, by itself, spherically shaped, highly symmetric, and electronegative [3]. It has been observed that fullerene electronic properties may be affected by structure variations and/or doping with atoms other than carbon. Thus, for instance, the C60 fullerene is not an electric conductor but if potassium (K) is added as an impurity a conducting material is obtained. Nevertheless, at high potassium concentrations an isolating compound is generated. In other compounds, including alkaline metals (X3 C60 fullerenes with X = Na, K, Rb, Cs), the origin of the electrical behavior is not completely clear [4]. However, it is noteworthy that fullerenes are capable of forming nanotubes [5,6], just another allotropic form of carbon, which are considered as the best electrical conductors at the nano-sized scale. Fullerene and carbon nanotubes chemistry and electrochemical properties are widely discussed in scientific literature and its application perspectives are currently analyzed in research facilities around the world. The tetrahedral atomic arrangement of diamond is generated by sp 3 -hybridization of the carbon atomic orbitals. This sp 3 -hybridization leads to forming four covalent bonds symmetrically oriented to the neighbor carbon atoms (see Figure 3.3a) and forming a ˚ The superimposicompact tridimensional structure where internuclear distance is 1.54 A. tion of the sp 3 -hybrid orbitals in diamond carbon atoms is so effective that it has several different consequences, such as a large energy gap (5.5 eV) between the valence and conduction bands (see Figure 3.3b). This extreme electron distribution defines two of the main characteristics of diamond: (1) a very strong bonding action, which is extended over the whole crystalline structure, with the consequent diamond being extremely hard and virtually absent of reactivity; (2) the bandgap creates an energetic barrier that makes the electron transference from the valence to conduction band practically impossible, thus generating a material with a high thermal stability and good dielectric behavior (≈1016  cm−1 ). To make use of these properties it is necessary to develop methods of synthesis capable to make synthetic diamond an accessible material, both in quantity and in cost. However, Because many chemically prepared diamonds are now available, it is important to distinguish between diamond and diamond-like materials such as synthetic diamond (SD), diamond-like carbon (DLC), and imitation diamonds (ID), to mention some. Diamond-like carbon is a carbon-based material obtained usually by vapor-phase chemical deposition. To obtain DLC, a mixture of hydrogen and gaseous hydrocarbons is used.

60

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

(a)

(b)

Energy, E

* 3 4σsp

2p atomic orbitals sp3 hybrid orbitals 2s atomic orbital

Figure 3.3

Conduction band empty

Energy gap 5.5 eV 4σsp3

Valence band filled

(a) Structural representation and (b) Electronic diagram of diamond.

Synthetic diamond may be obtained by different chemical-physics methods of synthesis; however, the high pressure high temperature (HPHT) and the chemical vapor deposition (CVD) are the two most commonly published techniques. In the case of CVD, monocrystalline or polycrystalline layers of diamond with grain size ranging from nanometers to some micrometers (see Figure 3.4a) are commonly generated. In addition, the diamond layers’ hardnesses vary depending on the specific preparation method, and the chemical and structural heterogeneity is determined by the arrangement of regions with diamond and nondiamond character (Figure 3.4b). Technological and commercial interests of SD obtained through CVD techniques are diverse and depend on their specific application. In some applications, such as mechanical grinding, the thickness of the films as well as their chemical and structural heterogeneity are important characteristics. For other uses, electrical conductivity will be the most important property. An interesting example on actual SD research trends is the Micromachined Diamond Device Initiative (MIDDI) project in the UK, where high advanced micro- and nanoscaled manufacturing technologies for preparation of synthetic single

3.1 INTRODUCTION

61

(a)

(b)

Figure 3.4 (a) Microcrystalline CVD diamond film on Si growth onto a lower methane atmosphere. (b) Nanocrystalline CVD diamond film on Si growth onto a rich methane atmosphere.

crystal sp3 -carbon diamond is being currently developed for different applications [7]. This concept is recent and it is associated to modern CVD methods. The actual quality of thin films obtained depends on the degree of sp2 -carbon impurity (i.e., graphite carbon), which is proposed to be controlled by regulating the methane-hydrogen ratio: The higher the methane content, the greater the amount of graphite carbon produced. Synthetic diamond, as the natural one, is not a conducting material. Nevertheless, not all kinds of natural diamond are electrical insulators. It is well known that blue diamond, containing boron as impurity, is able to behave as extrinsic p-type semiconductor. This fact is transcendental not only because of the aesthetics point of view but mainly because of the relevance to obtain conducting SD. Semiconducting properties of doped synthetic diamond (DSD) are basically similar to those observed in other wide bandgap semiconductors such as silicon (1.12 eV) or germanium (0.67 eV). The type and conducting level depends on the dopant agent and

62

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

Conduction band (CB)

Band gap 5.5 eV

P (0.6 eV below CB) donors

N (1.6 eV below CB)

B (0.37 eV above VB)

acceptor

Valence band (VB) Figure 3.5 Energy diagrams of selected states in the bandgap of diamond. (Reprinted with permission from Ref. 12.)

its concentration. When DSD is doped with an acceptor material (i.e., with electron deficiency or hole excess), such as boron, a p-type semiconductor is produced [8] (socalled boron-doped diamond, BDD), whereas when doped with donating materials (i.e., electron excess), such as phosphorous or nitrogen, a n-type semiconductor will be produced [9–11] (see Figure 3.5). Conducting capacity in the DSD changes with dopant concentration up to reach a quasi-metallic behavior for boron concentration in the order of 1019 –1021 atoms cm−3 [10,11].

3.2

CONDUCTING DIAMOND MATERIALS (CDMs)

As a consequence of the large energy gap between the filled and empty bands (over 5 eV), there is no a real chance for the highest-energy electrons to move under the influence of an applied electrical potential. Consequently, undoped diamond is not considered a good electrical conductor, so no practical use as electrochemically active materials has been reported for it [12,13]. In summary, pure diamond, just as other large bandgap materials, can be used as conducting material only when doped appropriately. The inclusion of light dopants during diamond crystal growth expands the possibility of creating new and versatile wide bandgap semiconductors [13]. The transformation from insulator to semiconductor by doping is achieved by the creation of discrete acceptor states just above the valence band (i.e., 0.35 eV for boron in diamond) which form a pseudo-conduction band [14]. Several different factors may have influence on the properties of these new conducting diamond materials generated after doping: (1) the potential dependent density of electronic states closely related to the doping type, distribution, and level; (2) the surface

3.4 CDM DOPING MATERIALS

63

chemistry, microstructure, and morphology of the CDMs; and (3) the defect density and nondiamond carbon impurity content. In the case of application of CDMs to electrochemical processes, the effect of these factors on the electrode properties will depend on the reaction mechanism for the particular redox system [15]. Moreover, all the properties listed earlier can be controlled using the appropriate laboratory conditions for deposition.

3.3

CDM PREPARATION PROCEDURES

So far, two main procedures for creating CDM have been reported [12]: chemical vapor deposition (CVD) of thin CDM films (CVD–CDMs) and high-pressure hightemperature (HPTP) CDM particle production. For electrode applications, vacuum annealing of undoped diamond and surface transfer doping of undoped diamond have been also described for preparing conducting diamond electrodes [12]. The detailed description of these major CMD preparation methodologies are provided in Chapter 2. For this reason, in this chapter we will only briefly refer to them in order to correlate them with their influence on the properties of the CDMs. All the aforementioned properties are subject of manipulation providing the appropriate selection of deposition conditions. These conditions may vary depending on the source and affect the resulting CDM on its quality and response [15]. For example, in the case of CVD–CDM, pretreatment of substrate before diamond coating and surface activation with nanoscale diamond particles are important treatment steps [16–18]. Other important conditions are source gas composition, cool-down procedures as well as postgrowth chemical treatment. All these variables can lead to differences in the properties and performance of CDMs [15].

3.4 3.4.1

CDM DOPING MATERIALS Characteristics of Boron-Doped CDMs

Boron is the doping element most used in carbon doping [19]. This element is the only atom able to enter into carbon atom lattice. Substituting carbon atoms at the trigonal sites, boron alters the electronic properties of the material without significantly changing the lattice parameters [20,21]. Application of boron for doping graphite [22–25], C–C composites [16,20,26], and related carbon derivatives [27–33] have been a common practice in the last 30 years to improve the resistance of these materials to oxidizing conditions. In the last decade, boron has also been widely used in carbon nanotubes preparation [34–39]. As occurs with doping of other carbon lattices, boron is also the most common doping element for diamond [12,19,40–44]. By the replacement of carbon atoms by boron, formation of an impurity level with activation energy at 0.37 eV above the valence band in the diamond occurs [46]. Figure 3.5 shows the energy diagram of selected states in the diamond bandgap as one dimensional representation. Because trivalent boron atoms are electron deficient, they will accept electrons from the diamond valence band generating electron-deficient carbon atoms (so-called holes) and they are able to carry a positive charge. Generation of p-type semiconductors is aimed at producing as many holes as possible. In the diamond atomic structure, every hole is associated with a negatively charged boron ion, which maintains the electric neutral

64

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

charge in the semiconductor. Under these conditions, each hole moving through the lattice behaves as a positive charge equilibrated with one electron. As long as the boron concentration reaches high enough values to generate a material with a resistivity between 5 and 100 m cm−1 (usually considered ≥1019 atoms cm−3 ), the hole concentration will exceed the electron thermal excitation and it becomes the major charge carrier. As previously mentioned, p-type conductivity has been reported occurring naturally in diamond only under rare conditions where boron is the main impurity at concentrations under 1 mg/kg. Boron loadings at upper concentrations can only be achieved synthetically in diamond using mainly HPHT or CVD technologies. As reported by Ristein et al. [47], surface conductivity in B-doped diamond (BDD) seems to be related to the following experimental conditions: (1) surface conductivity is enhanced by positive charge carriers moving at 30 to 70 cm2 V−1 s−1 at room temperature, indicating holes transportation within the crystalline solid [48]; (2) the concentration of charge carriers, as well as the mobility, is weakly temperature dependent with values among 1012 and 1013 cm−2 at room temperature [49]; and (3) surface conductivity may vary in orders of magnitude depending on the environment to which the surface is exposed [50]. 3.4.2

Electrochemical Properties

Conducting diamond materials possess very interesting properties as anodes useful in several electrochemical processes. Currently, one of the most attractive of applications for BDD anodes is water treatment—probably the most intensively investigated use. For water treatment, BDD properties such as their large potential window, low species adsorption, corrosion stability, high efficiency, and low double-layer capacitance and background current, among many others, have made the use of CDMs especially interesting for destruction of organic pollutants and waterborne pathogens in water. These exceptional BDD properties are clearly evident when comparing their potentiodynamic profile with that exhibited by a typical platinum electrode, such as illustrated in Figure 3.6a. The potentiodynamic profiles for the BDD electrode obtained after purged with nitrogen gas are featureless, and the background current is very low (less than 1 μA) in both cases. In acidic media the oxygen evolution peak appears about +2.0 V vs. SCE (Figure 3.6a). On the other hand, the curve B exhibits a broad oxidation peak at about +0.9 V vs. SCE. Thus, it can be assumed that this peak is due to water decomposition. These results are in agreement with Yano et al. [11]. Finally, a comparison between Pt and BDD electrodes is shown in Figure 3.6a and b, where BDD electrodes exhibit a most important characteristic: a potential window about 4 V. In alkaline media a similar broad potential window is observed, as shown in Figure 3.6b. As an important difference from conventional anodes—that is, Pt, PbO2 , SnO2 (doped and undoped), IrO2 , and RuO2 —the oxidation mechanism on CDMs does not proceed through preliminary introduction of oxygen into the oxide lattice generating a change in the oxidation state. Rather, the oxidation mechanism proceeds by generation of reactive oxygen species (ROS) and weak oxidants such peroxodisulfate, peroxodicarbonate, and peroxodiphosphate [51–54]. The wide range of applications of CDMs in organic compound oxidation includes degradation of ammonia, phenols, cyanides, dyes and colorants, pharmaceuticals, and an extended list of other pollutants [55–59] (see Chapters 10, 14, and 17) as well as its use in electrochemical disinfection processes for different microorganisms [60–62].

3.4 CDM DOPING MATERIALS

65

(a)

(b)

Figure 3.6 Potentiodynamic profiles (a) for the BDD and Pt electrodes in acidic media (0.5 M H2 SO4 ) and (b) for the BDD electrode in alkaline media (0.25 M NaOH + 0.5 M Na2 SO4 ).

This last case is important because microbiological contamination of fresh water with pathogens constitutes a major sanitary concern. Thus, for this instance, Furuta et al. [60] have shown that use of BDD as CDM anodes for water disinfection generates a high inactivation rate of pathogenic bacteria and viruses, such as is observed in Figure 3.7. More examples and results are commented on in Chapter 16. In addition to water treatment, CDM electrodes have also been used in a wide variety of other interesting industrial processes. Examples are chemical synthesis (see Chapter 19) and electro-analysis (see Chapters 7 and 8), where they have been used as sensors and biosensors [54,63], as well as in the in situ production of common chloride-based disinfection agents [62], fuel cells (see Chapter 18), and metal finishing processes such as electrodeposition of noble metals [64]. The most recent trends for the use of boron-doped CDM are its use as ultramicroelectrodes (UME) and biosensors (see Chapters 5 and 6). A UME is defined as an electrode with an effective diameter smaller than the scale of the diffusion layer, typically a critical diameter <25 μm. As a consequence, a rapid radial diffusion is performed dominating at the electrode surface, and faradic currents can be detected with improved signal:noise

66

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

Virus activity (%) 100.0 80.0 60.0 40.0 20.0 0.0 0

10 Time (min)

15

20

Figure 3.7 Inactivation of Adenovirus (H 40/1; 105 virus/mL) in tap water by chemical chlorine dosing (dashed line) and electrochemical disinfection (solid line). (Reprinted with permission from Ref. 60.)

ratio [65]. BDD-UME have been reported to be suitable for use in scanning electrochemical microscopy (SECM) probes as confirmed by Holt, Hu, and Foord [65], who fabricated and used these BDD-UMEs to obtain approach curves and image electrochemical activity of E. coli cells, as shown in Figure 3.8 and 3.8b. Boron-doped diamond electrodes have been recently investigated as potential substrates for biosensor devices [66]. In particular, H-terminated epitaxial diamond has been used as a convenient substrate to grow neuronal cultures for application in the fabrication of electrodes for recording electrical signals of excitable cells [66]. Recently, Carlisle [63] proposed that the use of electro-active enzymes chemically immobilized on conducting nanocrystalline diamond-conducting materials could be the basis of diamond-based electrochemical biosensors and bio-interfaces. The main advantages identified for CDMs for the generation of biosensors (1) no preparation is required to have electrodes ready to use, (2) the electrode surface is relatively free of oxygen providing pH-independent behavior of voltammetric background, and (3) there is an absence of problems related to surface functional groups and high stability level during exposure to biological environments [67]. Diamond microelectrodes have shown improvement performance for in vivo measurements with respect to other conventional microelectrodes made with Pt, Au, or C, among others [68–71].

3.4.3

Photoelectrochemical Properties

Boron doped-conducting diamond electrodes possess also some attractive photoelectrochemical properties [72–74]. Due to the large energy gap in undoped diamond, low energetic ultraviolet (UV) and visible radiation are not able to promote electrons from the valence to the conducting band. For example, in high-quality diamond electrodes with low nondiamond carbon content, only radiation with energy over the bandgap is able to promote electrons in the conducting band [74,75]. However, by the inclusion of boron as a dopant, measurable photocurrents have been determined probably associated with surface states produced by impurity within the bandgap. Boonma et al. [74] tested the photoelectrochemical behavior of boron-doped diamond electrodes with excimer lasers under different wavelengths of 193, 248 and 351 nm (6.4, 5.0, and 3.53 eV, respectively). They found that photocurrent observed using the highest energy (193 nm; 6.4 eV) was higher than that observed with the other two lasers and that photoelectrochemical properties of the CDM were sensitive to surface conditions [73].

3.4 CDM DOPING MATERIALS

67

(a)

(b) 12 pA 9 pA

350 μm

6 pA 3 pA 0 pA

2000 μm

Figure 3.8 (a) Sharp electrode tip of radius ≈3 μm achieved by bias-enhanced growth of nanocrystalline diamond. (b) SECM image of immobilized E. coli cells obtained by detecting ferrocyanide produced from their respiration. (Reprinted with permission from Ref. 65.)

3.4.4

Optical Spectroscopy Properties

Boron doping in diamond allows several very interesting spectral properties in the infrared band—for example, the absorption spectra in the range from 160 to 830 meV at room temperature due, at 160 meV, to one-phonon-absorption [76] and to transitions to boronbound hole excited states at 304, 347, and 363 meV [77]. Under low temperature conditions, the absorption spectra have been solved over 25 different absorption lines [78,79]. Under these conditions, the ground state splits by an energy of about 2.1 meV into degenerated 8 lower and 7 higher levels. The electronic transitions between these

68

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

energy levels can couple to the defect-induced first-order (DIFO), one phonon density of states spectrum leading to a repetition on absorption bands at higher energies [80]. On this spectrum type, lines at 462 and 508 meV are dominant and are usually shifted by the O Raman phonon. At about 370 meV, photoionization continuum as phononrepeat and extend to the visible region producing a bluish color of IIb-type diamond until reaching 10 μm (<0.1 eV) where free carrier absorption have been observed [78]. Perfect crystalline diamond is not able to absorb into infrared spectra but the presence of impurities or defects are able to produce slightly absorption in the visible region [81]. As described earlier, diamond doping with boron produce the diminishing on its transparency [82]; nevertheless, boron-doped CDMs are currently used in UV-Vis and infrared spectroscopy [83,84]. Boron-doped CDMs are also widely used as optically transparent electrodes (OTE) in spectroelectrochemical determinations in an extensive wavelength range [83–85,137]. The type of CDMs used for the different OTE applications depends on the use requirements. Thin-doped diamond films on a nondiamond substrate such as Si are currently used for IR, whereas quartz substrate is used in UV-Vis spectroelectrochemical measurements [83,12]. More recently [137], other ways for OTE production have been described—such as mechanically polished free-standing diamond electrodes that are prepared by growing thick films of CDM on a metallic substrate. The film is then separated from the substrate by cooling rapidly from the growth temperature, using the differences in thermal expansion coefficients of the CDM and the supporting material. The resulting film is mechanically polished to smooth the surface and to eliminate roughness and avoid light scattering. Generally, CDM-OTEs possess a short-wavelength cutoff about 225 nm, related with the bandgap of the material [85], and its optical window goes well out into the infrared, except for the boron acceptor band and the intrinsic multiphonon absorptions described early. 3.4.5

Photo- and Cathodoluminescence Properties

Doped diamond good-quality photo and cathodoluminescence spectra at 5.355 eV shows a sharp no-phonon (NP) recombination line of the boron-bound exciton (BE) [86]. Some authors have found a linear relationship between conductivity and boron content [87,88] in boron-doped polycrystalline films. They are derived from a phenomenological relationship that allowed the boron-doping level in the range between 1017 cm−3 to 1019 cm−3 from cathodoluminescence spectra. In the same way, doping level has an important effect on the spectral position and line width of the boron-bound exciton; for boron loads over 1019 cm−3 the boron-bound exciton shifts to lower energies. By increasing boron concentration, sudden shifts of the BE (up to 5.05 eV) have been reported [86]. BE luminescence have also been reported in boron-implanted HPHT-CDM. For these materials, high temperatures (about 1000–1200◦ C) are required to heal lattice damage and activate boron on optically active substitutional lattice sites even at, comparatively, low implantation doses (about 2 × 1013 cm−2 ). In addition to this bandgap BE, other boron emission bands have been identified in cathodoluminscence measurements for HPHT and CVD–CDMs. For example, a band centered in the UV region at 4.6 eV (mainly from high boron-loaded areas) has been reported for CVD–CDMs [89] at high temperature and disappears when samples are at room temperature. Other bands in the visible region, centered at about 2.3 eV and related to the donor-acceptor pair between boron and nitrogen or between unidentified donors with the same energy, have been detected [86]. The band at 2.3 eV is

3.5 NON-BORON-DOPED CDMs

69

characteristic of high-boron-loaded regions and it is dependent on boron concentration; nevertheless, nitrogen concentration does not affect it [86]. Another band at 1.8 eV has been reported by Klein et al. [90], which may compete with that at 2.3 eV and associate with transitions from other deep donors to boron acceptors. 3.4.6

Electrical Conductivity and Superconductivity Properties

A recent report on type-II superconductivity in boron-doped CDMs [91] found a revolutionary discovery that may change the actual basic science and technology concept on the field [43,92]. So far, the report of superconducting diamond material synthesized at about 8–9 GPa and 2500–2800 K with Tc = 4 K have brought back the controversy on the mechanisms in covalent materials. Several theoretical proposals on the superconductivity mechanisms of CDMs have been suggested [93,94]. These deal with the theory that superconductivity is an impurity band phenomenon driven by strong electron correlation effects and that the disordered superconductivity state is located in the vicinity of the Mott transition band, Further work is required to evidence any of these theories. As mentioned earlier, when acting as a dopant, boron atoms replace carbon atoms in the diamond lattice by bonding to the neighbor carbon atoms with a sp 3 hybridization. Because boron has one fewer electron than carbon, an electron deficiency (a hole) is generated in the otherwise filled valence band. It has been determined that the generated hole is associated in the ground state to the boron with a binding energy about 0.37 eV. By increasing of doping with boron, an Anderson-Mott transition from insulator to conductor is proposed via the overlapping of impurity wave functions after reaching a critical boron concentration, nc . Some authors have proposed superconductivity in CDMs as a result of the impurity band phenomenon dominated by strong physical correlation within the impurity band subsystem [86]. However, other recent theoretical studies have used band structure calculations and suggested that the superconductivity in CDMs results in strong coupling between phonons and holes at the  point [93–95]. Further research is necessary to generate accurate knowledge to understand the mechanisms of superconductivity in CDMs and for future development of new diamond superconducting devices.

3.5

NON-BORON-DOPED CDMs

Diamond dopage using a wide variety of elements has been an interesting scientific issue in the last few years [96]. Doping agents tested include an extensive number of metals and semiconductors [97]. Besides boron, other elements such as phosphorous [98–100], nitrogen [99,101–103], silicon [104–106], and palladium [107] have been used as dopants. In most cases, the objective of doping diamond is to achieve a p-type or n-type semiconductor [96]. In the case of p-type dopants, despite boron being the most widely reported dopant, other doping diamond elements such as potassium, sodium, and aluminum have been recently tested. In the case of potassium and sodium, they show low acceptor levels; aluminium shows a deeper acceptor level [108]. Research on viable n-type dopants has become increasingly more complex with time [109]. Theoretically, in analog to silicon doping, elements from groups V and VI are expected to become donor impurities. Several different elements such as lithium, sodium [110–111], arsenic, antimony [112–113], as well as previously mentioned nitrogen [114],

70

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

phosphorous, and sulfur [115–117] have been tested as doping agents. Among them, nitrogen (activation energy 1.6–1.7 eV below the CB), phosphorous (activation energy 0.6 eV below the CB), and sulfur are the most useful [46,100,118] as well as co-doped diamond (e.g., nitrogen-boron or boron-sulfur) [99,100,103,118]. Phosphorous or nitrogen doping, known as pentavalent impurities or donor elements, increase the number of negative charge carriers (i.e., electrons) in diamond [119]. The purpose of n-type doping is to produce abundant negative charge carriers in diamond. By adding a pentavalent atom to the crystallite lattice in diamond, the covalent tetrahedral structure in carbon atoms is maintained, allowing one nonbonded electron. In this way, the generation of free electrons will exceed holes and become the main charge carriers in n-type semiconductors. The use of sulfur as a doping agent has usually been restricted to the presence of boron, generating an n-type semiconductor at low boron concentrations [100,118]. Nevertheless, other works [138] have reported that diamond layers obtained by CVD in the presence of H2 S exhibit p-type conduction with activation energy and carrier concentration very similar to those reported for boron-doped diamond. In the case of nitrogen, its use as a dopant depends on growing conditions as well as on crystallographic orientation of the diamond films [120]. It can be incorporated in many different bonding configurations, the most common being as a single substitution nitrogen and as complex as lattice defects. It has been reported that electrochemical parameters obtained by nitrogen doping are only slightly lower to conventional borondoped diamond with a wide potential window and low background current [121]. The inclusion of nitrogen increases the thermoluminescence activity in diamond dosimeters [122]. More studies are needed to determine what optimal nitrogen impurity concentration should be performed to identify the best operation conditions. Recent work dealing with the generation of nitrogen-doped nanocrystalline diamond suggests that high conductivity could not be the result of conventional doping with nitrogen atoms but the increase of sp 2 bonding at grain boundaries and the introduction of π and π * states within the bandgap [123]. These increases may generate that conduction occurs along grain boundaries [98,121], however the exact transport mechanism remains unknown [124]. Phosphorous atoms, which are larger than carbon atoms, are able to form defects in diamond sp 3 structure by substitution [125], forming P-containing carbon tetragons. These structures are stable in CPx sp 2 cage-like structures [126,127]. P-doped diamond has been widely used as light emitting diodes [128] and UV detectors [129], among other uses [130]. An interesting donor level has been reported for phosphorous [111,116,131,132]; however, electron mobility values determined for P-containing diamond are low (23 cm2 V−1 s−1 ) [96,133]. The conductivity dependence on temperature for P-doped CDMs and its low mobility lead some authors to suggest that the observed conduction may be related more to crystal imperfections than to n-type conductivity. Sodium, antimony, arsenic, and oxygen have been tested as n-type donors. In none of these cases, however has n-type doping been definitively demonstrated with experimental evidence [96,134]. Controversy remains on the donor activity for n-type doping. In contrast to p-type doping, which has been demonstrated in diamond using many different techniques [135], n-type doping preparation is considered by some authors to be associated with several

REFERENCES

71

different problems [135]. They believe that these difficulties have limited its application to few special cases [136], mainly because several attempts to generate n-type semiconducting diamonds have resulted in low-quality CDMs. 3.6

CONCLUSIONS

Nowadays, the synthetic diamond is an advanced material with a wide spectrum of applications in fundamental research as well as in technological development. The conductive limitations of diamond in the past are no longer a problem today. It is clear that insertion of foreign atoms into the crystalline structure of diamond can decrease the large energy gap to an acceptable level to allow the electrical conduction but keep their thermal and chemical stability properties. This is a relevant aspect because some of the most important applications of synthetic diamond could be dependent on these latter properties. Boron, nitrogen, and phosphorus are the most common foreign atoms to add to electronic acceptors or charge donors in doping processes. Although the electronic states of nitrogen and phosphorus in diamond are well known, BDD has become the more popular material of synthetic diamond since its introduction in 1987. This fact is due in part to BBD’s strong impact in the field of application and fundamental research of the electrochemistry. Good quality, oxygen, or hydrogen-terminated diamond thin-film electrodes exhibit suitable properties for electrochemical studies such as (1) a low and stable background current over a wide potential range and (2) a good resistance to fouling due to weak adsorption of polar molecules on the nonpolar surface. Knowledge branches such as organic electrosynthesis, electroanalysis, and water treatment have found in these conducting diamond films a new perspective of development. In addition, the versatility of this material has also been exploited in developing sensors and biosensors, also being useful for several bioelectrochemical applications. Finally, doped synthetic diamonds will make scientific development in possible applications such as microsensors and instrumentation, thermal devices, biomedical devices, electronic and microelectronic systems, optic and photonics, microfluids, and nanomaterials, as well as in the fuel cell technology. However, the most relevant application in the future will be associated with the elaboration of nanoparticles of synthetic diamond able to be used in environmental problems that remain unresolved and in medical diagnostic activities. REFERENCES 1. 2. 3. 4. 5. 6. 7.

R.F. Curl, R. Smalley, Sci. Am. 1991, 265 , 54–63. F. Diederich, R. Ettl, Y. Rubin, Science 1991, 252 , 548–551. O. Vasilievna, U. Ortiz, Ciencia UANL 2002, 5 , 475–479. O. Gunnarsson, Rev. Modern Phys. 1997, 69 , 575–606. P. Byszewsri, Z. Klusek, Optoelect. Rev . 2001, 9 , 203–2010. S. Iijima, Springer Proc. Phys. 2001, 87 , 24–34. Micromachined Diamond Device Initiative (MIDDI). Element Six Ltd. and Institute of Photonics, University of Strathclyde, Glasgow, UK, 2008. 8. V.S. Vavilov, Usp. Fiz. Nauk 1994, 164 , 287–293. 9. R. Kalish, Carbon 1999, 37 , 781–785.

72

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

10. M. Nesladek, Semicond. Sci. Technol. 2005, 20 , R19–R27. 11. T. Yano, E. Popa, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 1081–1087. 12. A. Kraft, Int. J. Electrochem. Sci . 2007, 2 , 355–370. 13. M. Pagels, C. Hall, N.S. Lawrence, A. Meredith, T. Jones, H.P. Godfried, J. Pickless, J. Wilman, C. Banks, R. Compton, L. Jiang, Anal. Chem. 2005, 77 , 3705–3708. 14. C. Saguy, Defect Diff. Forum 2004, 31 , 226–228. 15. A.E. Fisher, Y. Show, G.M. Swain, Anal. Chem. 2004, 76 , 2553–2558. 16. S. Ragan, G.T. Emmerson, Carbon 1992, 30 , 339–345. 17. M. Fryda, T. Matthee, S. Mulcahy, A. Hampel, L. Schafer, I. Troster, Diam. Rel. Mat . 2003, 12 , 1950–1955. 18. X. Jiang, M. Fryda, C.L. Jia, Diam. Rel. Mat . 2000, 9 , 1640–1648. 19. A. Marchand in Chemistry and Physics of Carbon ( P.L. Walker, ed.), Marcel Dekker, New York, 1971. 20. X. Wu, L.R. Radovic, J. Phys. Chem. A 2004, 108 , 9180–1987. 21. V. Serin, R. Brydson, A. Scott, Y. Kihn, O. Abidate, B. Maquin, A. Derre, Carbon 2000, 38 , 547–555. 22. D.J. Allardice, P.L. Walker, Carbon 1970, 8 , 375–385. 23. T. Hagio, M. Nakamizo, K. Kobayashi, Carbon 1989, 27 , 259–263. 24. N.M. Rodriguez, R.T.K. Baker, J. Catal . 1993, 140 , 287–293. 25. T. Sogabe, T. Matsuda, K. Kuroda, Y. Hiroata, T. Hino, T. Yamashina, Carbon 1995, 33 , 1783–1795. 26. W. Kowbel, Y. Huang, H. Tsou, Carbon 1993, 31 , 355–363. 27. D.H. Zhong, H. Sano, Y. Uchiyama, K. Kobayashi, Carbon 2000, 38 , 1199–1118. 28. L.E. Jones, P.A. Thrower, Carbon 1991, 29 , 251–258. 29. L.R. Radovic, M. Karra, K. Skokova, P.A. Thrower, Carbon 1998, 36 , 1841–1864. 30. W. Cermignani, T E. Paulson, C. Onneby, C.G. Pantano, Carbon 1995, 33 , 367–374. 31. T. Durkic, A. Peric, M. Lusevic, O. Neskovic, M. Vejkovic, Z. Lausevic, Carbon 1997, 35 , 1567–1575. 32. M. Chesneau, F. Beguin, J. Conard, R. Erre, J. Thebault, Carbon 1992, 30 , 714–716. 33. S. Jacques, A. Guette, X. Burrat, F. Langlais, C. Guimon, C. Labrugere, Carbon 1996, 34 , 1135–1143. 34. B.Q. Wei, R. Spolenak, P. Kohler, M. Ruhle, E. Arzt, Appl. Phys. Lett . 1999, 74 , 3149–3156. 35. P.C.P. Watts, W.K. Hsu, G.Z. Chen, J. Mater. Chem. 2001, 11 , 2482–2490. 36. W.Q. Han, Y. Bando, K. Kurashima, T. Sato, Chem. Phys. Lett . 1999, 299 , 368–375. 37. W.K. Hsu, S. Firth, P. Redlich, M. Terrones, H. Terrones, Y.Q. Zhu, N. Grobert, A. Schilder, R.J. Clark, D. Walton, J. Mater. Chem. 2000, 10 , 1425–1432. 38. A.S. Kotosonov, D.V. Shilo, Mol. Mater. 2000, 13 , 113–121. 39. P.L. Gai, O. Stephan, K. McGuire, A.M. Rao, M.S. Dresselhaus, G. Dresselhaus, C.J. Colliex, J. Mater. Chem. 2004, 14 , 669–675. 40. N. Vinokur, B. Miller, Y. Avygal, R. Kalish, J. Electrochem. Soc. 1996, 143 , L238–L243. 41. C.H. Goeting, F. Jones, J.S. Foord, J.C. Eklund, F. Marken, R.G. Compton, P.R. Chalker, C.J. Johnston, J. Electroanal. Chem. 1998, 442 , 207–213. 42. R.G. Compton, J.S. Foord, F. Marken, Electroanalysis 2003, 15 , 1349–1356. 43. E. Bustarret, P. Achatz, B. Sac´ep´e, C. Chapelier, C. Marcenat, L. Ortega, T. Klein, Phil. Trans. R. Soc. A 2008, 366 , 267–275.

REFERENCES

73

44. B. Sac´ep´e, C. Chapelier, C. Marcenat, J. Kacmarcik, T. Klein, M. Bernard, E. Bustarret, Phys. Rev. Lett . 2008, 96 , 097006.1. 45. J. R. Chelikowsky, S.G. Louie, Phys. Rev. B 1984, 29 , 3470–3477. 46. M. Nesdalek, Semicond. Sci. Technol. 2005, 20 , R19–R28. 47. J. Ristein, M. Riedel, M. Stammler, B.F. Mantel, L. Ley, Diam. Rel. Mat . 2002, 11 , 359–370. 48. H.J. Looi, R.B. Jackman, J.S. Foord, Appl. Phys. Lett . 1998, 72 , 353–370. 49. K. Hayashi, S. Yamanaka, H. Okushi, K. Kajimura, Appl. Phys. Lett . 1996, 68 , 376–378. 50. M. Szameitat, X. Jiang, W. Beyer, Appl. Phys. Lett . 2000, 77 , 1554–1560. 51. P.A. Michaud, E. Mahe, W. Haenni, A. Perret, C. Comninellis, Electrochem. Solid-State Lett . 2000, 3 , 77–85. 52. M.S. Saha, T. Furuta, Y. Nishiki, Electrochem. Solid-State Lett . 2003, 6 , D5–D11. 53. P. Ca˜nizares, F. Larrondo, J. Lobato, M.A. Rodrigo, C. S´aez, J. Electrochem Soc. 2005, 152 , D191–D198. 54. C.A. Martinez-Huitle, E. Brillas, Angew. Chem. Int. Ed . 2008, 47 , 1998–2005. 55. M. Panizza, G. Cerisola, Electrochem. Acta 2005, 51 , 191–198. 56. M.A. Quiroz-Alfaro, S. Ferro, C.A. Martinez-Huitle, Y. Meas-Vong, J. Braz. Chem. Soc. 2006, 17 , 227–236. 57. I. Sir´es, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, E. Brillas, Electrochem. Acta 2006, 52 , 75–82. 58. M. Villanueva, A. Hernandez, J.M. Peralta-Hern´andez, E.R. Bandala, M.A. Quiroz, ECS Transactions 2008, 15 , 411–417. 59. M. Villanueva, A. Hernandez, J.M. Peralta, E.R. Bandala, M.A. Quiroz, J. Hazard. Mater. 2008, 167 , 1226–1230. 60. T. Furuta, H. Tanaka, Y. Nishiki, L. Pupunat, W. H¨anni, P. Rychen, Diam. Rel. Mat . 2004, 13 , 2016–2023. 61. I. Tr¨oster, L. Sch¨afer, M. Fryda, T. Matthe´e, Water Sci. Technol . 2004, 49 , 207–215. 62. W. Haenni, J. Gobet, A. Perret, L. Pupunat, P. Rycehn, C. Comninellis, B. Correa, New Diam. Front. Carbon Technol . 2002, 12 , 83–98. 63. J.A. Carlisle, Nat. Mat . 2004, 3 , 668–675. 64. J. Wurm, New Diam. Front. Carbon Technol . 2002, 12 , 107–115. 65. K. B. Holt, J. Hu, J.S. Foord, Anal. Chem. 2007, 79 , 2556–2563. 66. P. Ariano, A. Lo Giudice, A. Mercantoni, E. Vittone., E. Carbone, D. Lovisole, Biosens. Bioelect . 2009, 24 , 204–2072. 67. S.H. Park, U. Krull, Anal. Chem. Acta 2006, 564 , 133–141. 68. C.A. Martinez-Huitle, Small 2007, 3 , 1474–1476. 69. Y. Zhou, J. Zhi, Talanta 2009, 75 , 1189–1193. 70. Y. Pleskov, M.D. Krotova, V.V. Elkin, V.G. Ralchenko, A.V. Saveliev, S.M. Pimenov, P.Y. Lim, Electrochem. Acta 2007, 52 , 5470–5478. 71. T. Watanabe, Y. Einaga, Biosens. Bioelect . 2009, 24 , 2684–2693. 72. A.Y. Sakharova, Y.V. Pleskov, F. Di Quarto, S. Piazza, C. Sunseri, I.G. Teremetskaya, V.P. Varnin, Russ. J. Electrochem. 1995, 32 , 168–175. 73. T.N. Rao, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 680–688. 74. L. Boonma, T. Yano, K. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1997, 144 , L142–L150. 75. Y. Pleskov, V.M. Mazin, Y.E. Evstefeeva, V.P. Varnin, G. Teremetskaya, V.A. Laptev, Electrochem. Solid State Lett . 2000, 3 , 141–153.

74

TYPES OF CONDUCTING DIAMOND MATERIALS AND THEIR PROPERTIES

76. G. Davies, in The Chemistry and Physics of Carbon ( P.L. Walker, P.A. Thrower, eds.), Marcel Dekker, New York, 1977. 77. A.T. Collins, A.W. Willams, J. Phys. C: Solid State Phys. 1971, 4 , 1789–1795. 78. H. Kim, A.K. Ramdas, S. Rodriguez, T.R. Anthony, Solid State Commun. 1997, 102 , 861–873. 79. H. Kim, R. Vogelgesang, A.K. Ramdas, S. Rodriguez, M. Grimsditch, T.R. Anthony, Phys. Rev. B 1997, 57 , 15315–15323. 80. E. Gheeraert, J. Deneuville, J. Mambou, Diam. Rel. Mat . 1998, 7 , 1509–1514. 81. A.N. Obraztsov, I.Y. Pavlovski, T. Izumi, H. Okushi, H. Watanabe, Appl. Phys. A: Mater. Sci. Process. 1997, 65 , 505–511. 82. R.L. McCreery, Chem. Rev . 2008, 108 , 2646–2653. 83. H.B. Martin, P.W. Morrison, Electrochem. Solid-State Lett . 2001, 4 , E17–E23. 84. J. Stotter, J. Zack, Z. Behler, Y. Show, G.M. Swain, Anal. Chem. 2002, 74 , 5924–5932. 85. J.K. Zak, J.E. Butler, G.M. Swain, Anal. Chem. 2001, 73 , 908–915. 86. K. Thonke, Semicond. Sci. Technol. 2003, 18 , S20–S31. 87. H. Kawarada, Y. Yokota, Y. Mori, K. Nishimura, A. Hiraki, J. Appl. Phys. 1990, 67 , 983–989. 88. H. Kawarada, H. Matsuyama, Y. Yokota, T. Sogi, A. Yamaguchi, A. Hiraki, Phys. Rev. Lett . 1993, 47 , 3633–3645. 89. S.C. Lawson, H. Kanda, T. Tsutsumi, H. Kawarada, J. Appl. Lett . 1995, 77 , 1729–1734. 90. P.B. Klein, M.D. Crossfield, J.A. Freitas, A.T. Collins, Phys. Rev. B 1995, 51 , 9634–9645. 91. E.A. Ekimov, V.A. Sidorov, E.D. Bauer, N.N. Mel’nick, N.J. Curro, J.D. Thompson, S.M. Stishov, Nature 2004, 428 , 542–549. 92. Y. Takano, M. Nagao, I. Sakaguchi, M. Tachiki, T. Hatano, K. Kobayashi, H. Umezawa, H. Kawarada, Appl. Phys. Lett . 2004, 85 , 2851–2863. 93. L. Boeri, J. Kortus, O.K. Andersen, Phys. Rev. Lett . 2004, 93 , 237–242. 94. K.W. Lee, W.E. Pickett, Phys. Rev. Lett . 2004, 93 , 1–13. 95. X. Blas´e, C. Adessi, D. Connetable, Phys. Rev. Lett . 2004, 93 , 237004. 96. A.S. Barnard, S.P. Russo, I.K. Snook, J. Phys. Chem. 2005, 109 , 11991–11998. 97. J.P. Goss, R. Jones, M.I. Heggie, C.P. Ewels, P.R. Briddon, S. Oberg, Phys. Rev. B 2002, 65 , 115207. 98. M. Nesdalek, D. Tromson, P. Bergonzo, P. Hubik, J.J. Mares, J. Kristofik, D. Kindl, O.D. Williams, D. Gruen, Diam. Rel. Mat . 2006, 15 , 607–616. 99. S.C. Eaton, Y.E. Evstefeeva, J.C. Angus, A.B. Anderson, Y. Pleskov, Russ. J. Electrochem. 2003, 39 , 154–159. 100. S. Vaddiraju, S. Eaton-Magana, J.A. Chaney, M.K. Sunkara, Electrochem. Solid-State Lett . 2004, 7 , G331–G343. 101. S. Ferro, A. De Battisti, I. Duo, C. Comninellis, W. Haenni, A. Perret, J. Electrochem. Soc. 2000, 147 , 2614–2621. 102. Q. Chen, D.M. Gruen, A.R. Krauss, T.D. Corrigan, M. Witek, G.M. Swain, J. Electrochem. Soc. 2001, 148 , E44–E53. 103. W. Haenni, P. Rychen, M. Fryda, C. Comninellis, in Semiconductors and Semimetals Series (C, Nebel, ed.), Academic Press, New York, 2004. 104. G. Sittas, H. Kanda, I. Kiflawa, P. Spear, Diam. Rel. Mat . 1996, 5 , 866–875. 105. W.J. Wu, M.H. Hon, Surf. Coat. Technol . 1999, 111 , 134–142. 106. I. Kiflawi, G. Sittas, H. Kanda, D. Fisher, Diam. Rel. Mat . 1997, 6 , 146–154. 107. A.V. Simakin, E.N. Loubnin, G.A. Shafeev, P. Doppelt, Appl. Surf. Sci . 2002, 186 , 227–232.

REFERENCES

75

108. G. Popovici, T. Sung, M.A. Prelas, J. Chem. Vap. Deposition 1994, 3 , 115–123. 109. L. Ley, J. Ristein, F. Meier, M. Riedel, P. Strobel, Physica B: Cond. Mater. 2006, 376-377 , 262–267. 110. S.A. Kajihara, A. Antonelli, J. Bernholc, R. Car, Phys. Rev. Lett . 1991, 66 , 2010–2023. 111. S.A. Kajihara, A. Antonelli, J. Bernholc, Physica B: Cond. Mater 1993, 185 , 144–156. 112. V.S. Vavilov, Phys-Usp. 1994, 37 , 269–275. 113. V.S. Vavilov, Phys-Usp. 1997, 40 , 387–393. 114. J.P. Goss, P.R. Briddon, R. Jones, S. Sque, Diam. Rel. Mat . 2004, 13 , 684–695. 115. D. Saada, J. Adler, R. Kalish, Appl. Phys. Lett . 2000, 77 , 878–894. 116. L.G. Wang, A. Zunger, Phys. Rev. B 2002, 66 , 161202. 117. T. Miyazaki, H. Okushi, Diam. Rel. Mat . 2002, 11 , 323–332. 118. S.C. Eaton, A.B. Anderson, J.C. Angus, Y.E. Evstefeeva, Y.V. Pleskov, Electrochem. SolidState Lett . 2002, 5 , G65–G78. 119. K. Haenen, M. Nesdalek, L. De Schepper, R. Kravets, M. Vanecek, S. Koizumi, Diam. Rel. Mat . 2004, 13 , 2041–2054. 120. F.K. De Theije, J.J. Schermer, J.P. van Enckevort, Diam. Rel. Mat . 2000, 9 , 1439–1445. 121. Y. Pleskov, M.D. Krotova, A.V. Saveliev, V.G. Ralchenko, Diam. Relat. Mat . 2007, 16 , 2114–2123. 122. C. Descamps, D. Tromson, M.J. Guerrero, C. Mer, E. Rzepka, M. Nesdalek, P. Bergonzo, Diam. Relat. Mat . 2006, 15 , 833–845. 123. E. Kohn, A. Denisenko, M. Kubovic, T. Zimmermann, A.O. Williams, D.M. Gruen, Semicond. Sci. Technol . 2006, 21 , L32–L43. 124. Y. Pleskov, S.M. Pismenov, P. Lim, V.G. Ralchenko, Russ. J. Electrochem. 2008, 44 , 861–865. 125. H. Sternschulte, K. Thonke, R. Sauer, S. Koisomi, Phys. Rev. B 1999, 59 , 12924–12935. 126. G.K. Gueorguiev, A. Furlan, H. Hogberg, S. Stafstrom, L. Hultman, Chem. Phys. Lett . 2006, 426 , 374–86. 127. E. Cruz, D.A. Cullen, L. Gu, J.M. Romo, E. Mu˜noz, F. Lopez, B. Sumpter, V. Meunier, J. Charlier, D.J. Smith, H. Terrones, M. Terrones, ACS Nano 2008, 2 , 441–453. 128. S. Koizumi, K. Watanabe, M. Hasegawa, H. Kanda, Science 2001, 292 , 1899–1901. 129. A. BenMoussa, U. Schule, K. Haenen, M. Nesdalek, S. Koizumi, J.F. Hochedez, Phys.Stat. Sol. A 2004, 201 , 2536–2546. 130. A. Lazea, V. Mortet, J. D’Haen, P. Geithner, J. Ristein, M. D’Olieslaeger, K. Haenen, Chem. Phys. Lett . 2008, 454 , 310–318. 131. E. Gheeraert, S. Koizumi, T. Teraji, H. Kanda, M. Nesdalek, Diam. Rel. Mat . 2000, 9 , 948–957. 132. E. Gheeraert, N. Cassanova, S. Koizumi, T. Teraji, H. Kanda, Diam. Rel. Mat . 2001, 10 , 444–453. 133. S. Koizumi, M. Kamo, Y. Sato, H. Okazi, T. Inuzuka, Appl. Phys. Lett . 1997, 71 , 1065–1074. 134. J.F. Prins, Diam. Rel. Mat . 2000, 9 , 1275–1284. 135. S. Koizumi, T. Teraji, H. Kanda, Diam. Rel. Mat . 2000, 9 , 935–943. 136. I. Sakaguchi, M. Gamo, Y. Kikuchi, E. Yasu, H. Haneda, T. Suzuki, T. Ando, Phys. Rev. B 1999, 60 , R2139. 137. S. Haymond, J.K. Zak, Y. Show, J.E. Butler, G.T. Babcock, G.M. Swain, Anal. Chem. Acta 2003, 500 , 137–143. 138. R. Kalish, A. Reznik, C. Uzan, C. Cytermann, Appl. Phys. Lett . 2000, 76 , 757–765.

Part II

Electrochemistry of Diamond Films

4 Electrochemistry of Diamond Yuri Pleskov

4.1

INTRODUCTION

Diamond undoubtedly is a significant electrode material. However, unlike other carbonaceous materials that have long found their practical applications (e.g., graphite, glassy carbon, pyrolytic graphite, carbon fibers, etc.), the electrochemical study of diamond was started relatively recently, in 1987, when the first paper on the (photo)electrochemistry of CVD-diamond electrode was published [1]. These studies were triggered by the fact that during the preceding three decades, highly effective methods of polycrystalline diamond film growth at conducting substrates at subatmospheric, rather than ultra-high, pressure were developed. By doping with acceptor-type impurity (boron), well-conducting films were prepared, which were semiconducting (p-type) or, at a higher doping level, even quasi-metallic in their nature. Thus, ultra-high corrosion resistance and sufficient conductivity were combined in the same material. With these first studies on diamond electrodes, a new field of the semiconductor electrochemistry [2] was established—namely, the electrochemistry of diamond [3]. The electrochemistry of diamond has several specific features. First, diamond is crystalline; thus, it is not unexpected that it demonstrates semiconductor behavior. Second, the charge carrier (hole) mobility in polycrystalline CVD-films is rather low (∼10 cm2 V−1 s1 ) [4]; therefore, to demonstrate substantial conductivity, diamond must be rather heavily doped, so that the semiconductor becomes degenerate and its semiconductor properties appear somewhat blurred. Third, the semiconductor and structural properties of diamond are interrelated rather quaintly. Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

79

80

ELECTROCHEMISTRY OF DIAMOND

In this chapter, we present an overview of the principal electrochemical properties of CVD-diamond, with emphasis on the semiconductor and structural effects.

4.2

PRINCIPAL ELECTROCHEMICAL PROPERTIES OF DIAMOND

We first discuss current–voltage curves taken in supporting electrolyte solutions. In Figure 4.1, we show a typical background current curve taken potentiodynamically at a polycrystalline diamond thin-film electrode in the supporting electrolyte (0.5 M H2 SO4 ) solution [5]. The potential window of diamond electrodes is incomparably wider than that of any other material: It may be as wide as three volts (sometimes even more). The advantage of diamond electrodes in the width of the potential window, as well as in the lowest background current in the mid-window, is most clearly seen when compared with other carbonaceous materials (glassy carbon, graphite, etc.) or platinum, which have long had applications as electrodes. The wide potential window arises from the high overvoltage for solvent (water) decomposition. Indeed, the hydrogen electrode reaction (HER) or oxygen electrode reaction (OER) are inner-sphere reactions; their kinetics are very sensitive to the adsorption of their intermediates (e.g., H • ) at the electrode surface. Diamond is a poor adsorbent; hence, its catalytic effect is rather weak for these reactions. In contrast, outer-sphere reactions that are not limited by the adsorption of intermediates proceed very fast, often in a reversible mode [6]. Thus, within the potential window the solvent is stable, and the diamond electrode is suitable for the studying of reactions that involve neither the solvent nor the electrode substance; the reactions can be studied bearing in mind their electro-analytical, electrosynthetic, other applications. Because the growth process occurs in a hydrogen atmosphere, as-grown CVD-diamond surfaces are hydrogen-terminated; when oxidized (e.g., anodically or under the action of O-plasma), they are oxygen-terminated. The surface termination affects the electrochemical kinetics drastically [7,8] (see Chapter 5 of this book).

1 0.8 0.6 j (μA cm–2)

0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 –1 –1

0

1

2

3

E (V) vs. Ag/AgCI Figure 4.1 Voltammetric curve of boron-doped diamond electrode, recorded in supporting electrolyte (0.5 M H2 SO4 ) at a potential scan rate of 50 mV/s. (Reprinted from Ref. 5.)

4.2 PRINCIPAL ELECTROCHEMICAL PROPERTIES OF DIAMOND

81

Diamond is extremely corrosion-resistant. For example, its long-term potential cycling in 1 M HNO3 + 0.1 M HF solution between the potentials of cathodic hydrogen evolution and anodic oxygen evolution has no effect on the diamond surface morphology. In contrast, glassy carbon and pyrolytic graphite electrodes are strongly damaged under the same conditions [9]. Similarly, diamond demonstrates strong corrosion resistance in 1 M HNO3 + 2 M NaCl solution at a potential of anodic chlorine evolution [10]. In Figures 4.2a and 4.2b, we give anodic and cathodic potentiodynamic curves in solutions containing a single (either oxidized or reduced) form of the [Fe(CN)6 ]3−/4− redox couple, taken at different polarization rates [11]. Contrary to expectations, these curves appeared mutually symmetrical, as with metal electrodes. Meanwhile, an ideal semiconductor electrode should demonstrate current rectification [2]: Usually, its polarization curves distinctly show the forward and blocking directions of the electrode current. Such polarization curves have indeed been experimentally observed time and again for lightly or moderately doped diamond electrodes (which will be discussed later). Yet, for rather heavily doped diamond, the anodic and cathodic curves are mirror images of each other. (In all probability, this is due to the fact that the heavily doped diamond is a degenerate semiconductor (as discussed earlier), or even demonstrates a metal-like behavior.) This characteristic shape of the curves, with the anodic and cathodic current maxima, is evidence that the interfacial charge transfer stage proper is rather fast (quasi-reversible), so that the reaction is limited by diffusion difficulties caused by slow mass transfer of the reacting species from the solution bulk to the electrode surface. This is often the case with outer-sphere reactions. The dependence of the anodic and cathodic current maximum potentials Ep in Figures 4.2a and b on the logarithm of the polarization rate dE /dt = v is given in Figure 4.2c. From the slope of the lines, using the theory of the kinetics of quasi-reversible reactions (as applied to the potentiodynamic curve method) [12,13], the transfer coefficients for the anodic (α) and cathodic (β) reactions can be calculated by the following formula: Ep = const − (RT/2αnF) ln v

(4.1)

where n is the number of electrons participating in the reaction (n = 1 for the [Fe (CN)6 ]3−/4− redox couple). Another approach to measuring the transfer coefficients is based on measuring the electrode’s electrochemical impedance spectra in redox-couple-containing solutions, such as in the “model” redox solution K3 [Fe(CN)6 ] + K4 [Fe(CN)6 ], in which the electrode acquires the equilibrium potential of the redox couple. Figure 4.3 shows a complex-plane plot of impedance spectra for a diamond electrode, measured in such a solution [14]. The low-frequency cutoff of the semicircle gives the Faradaic (or charge-transfer) resistance RF . To calculate the transfer coefficient of cathodic reaction α (or anodic, β), the following expression can be used: j0 = nFk0

1−α α cox cred

(4.2)

where k 0 is the rate constant. The exchange current j0 is interrelated with the Faradaic resistance measured in a redox-couple-containing solution at its equilibrium potential as RF = RT/(nFj0 )

(4.3)

82

ELECTROCHEMISTRY OF DIAMOND

1.2

i (mA/cm2)

(a)

0.8

5 4 3

0.4 2 1 0 0.4

0.6

0.8 E, V (Ag/AgCI)

1.0

1.2

E, V (Ag/AgCI) 0

0.2

0.4

0.6

(b) 1 2

i (mA/cm2)

–0.2 3 –0.4 4 –0.6 5 –0.8 –1.0 1.0

I

0.9 (c) 0.8 E P, V

0.7 0.6 0.5 0.4 0.3 2

0.2 –2.4

–2.0

–1.6 log ν (V/s)

–1.2

–0.8

Figure 4.2 (a) Anodic and (b) cathodic potentiodynamic curves for diamond electrode in 0.5 M H2 SO4 + 0.01 M K4 Fe(CN)6 (resp., K3 Fe(CN)6 ) solution. Potential scan rate (mV/s): (1) 5; (2) 10; (3) 20; (4) 50; and (5) 100. (c) Determination of transfer coefficients: (1) the anodic reaction, ( 2) the cathodic reaction. (Reprinted from Ref. 11.)

Substituting equation (4.3) into (4.2), we reveal the dependence of RF on the variable concentration of one of the forms in the redox couple (e.g., the oxidant, cox ) while the concentration of the other form (the educing agent, cred ) is kept constant, and vice versa. The transfer coefficients of the cathodic and anodic reactions α and β are found by the slope of the corresponding linear dependence of RF on cox (at cred = const) or cred (at cox = const).

4.3 THE EFFECT OF SEMICONDUCTOR NATURE OF DIAMOND

83

–Im Z (kOhm cm2)

420 0.1

12

8000 0

0.1

0.2 0.3 Re Z (kOhm cm2)

Figure 4.3 Complex-plane plot of impedance spectrum, measured at a diamond electrode in solution 1M KCl + 5 × 10−2 M K3 [Fe(CN)6 ] + 5 × 10−2 M K4 [Fe(CN)6 ]). AC frequency (in Hz) is shown at the corresponding points. (Reprinted from Ref. 14.)

4.3 THE EFFECT OF SEMICONDUCTOR NATURE OF DIAMOND ON ITS ELECTROCHEMICAL BEHAVIOR For semiconductor electrodes (diamond, in particular), the transfer coefficients α and β are less than ∼0.5 (the value characteristic of metals); hence, their sum α + β is less than 1. The reason of this is still an open question. (We come back to this point later.) We now turn to the dependence of the electrochemical reaction rate on the diamond electrode doping level. In Figure 4.4, we correlate the charge transfer (Faradaic) resistance RF and the diamond resistivity ρ for two redox systems: [Fe(CN)6 ]3−/4− and benzoquinone/hydroquinone [14,15]. (The RF values are measured experimentally, either from the slope of polarization curves taken in solution containing equal concentrations of [Fe(CN)6 ]3− and [Fe(CN)6 ]4− at the equilibrium potential, or as a “low-frequency cutoff” at the real resistance axis for a complex-plane plot of electrode impedance spectra measured in the above solution, as shown in Figure 4.3.) Recall that the resistance RF is related to the exchange current of the electrode reaction j0 by Equation (4.3). From the relative position of the plots in Figure 4.4, one may conclude that the reaction rate in the [Fe(CN)6 ]3−/4− system is three orders of magnitude higher (RF is three orders of magnitude lower at a fixed ρ) than that in the benzoquinone/hydroquinone system. Most importantly, the exchange current (j0 ∼ 1/RF ) appears to be nearly proportional to the diamond conductivity (1/ρ) over a wide range of ρ values for both reactions. Indeed, the slope of the lines in Figure 4.4 approaches 1. The resistivity is known to be approximately inversely proportional to the charge carrier concentration in semiconductors (provided their mobility may be thought of as ρ-independent). Thus, the heavier the diamond electrode is doped, the faster are electrochemical reactions thereon. (The possible reason for this dependence will be discussed at the end of this section.) Anyhow, various practical applications (e.g., water purification, electroanalysis, etc.) require rather heavily doped electrodes whose resistivity comes to mOhm cm (which corresponds to a boron concentration of approximately 1021 cm−3 ). Now, let’s compare electrochemical kinetics for n- and p-type diamond electrodes. Since 1987, and over the entire time of studies of CVD-diamond electrochemistry, only ptype samples were dealt with because boron remained the only dopant with low ionization

84

ELECTROCHEMISTRY OF DIAMOND

7 2 1 logRF (Ohm cm2)

6

5

4

3

2

4

5

6 log ρ (Ohm

7

8

cm2)

Figure 4.4 Dependence of the Faradaic resistance RF at the equilibrium potential in solutions (a) [Fe(CN)6 ]3−/4− and (b) benzoquinone/hydroquinone on the diamond film resistivity ρ. (Reprinted from Ref. 14.)

energy that imparted diamond reasonable conductivity at room temperature. Only recently [16–18] n-type diamond thin-film electrodes well conducting at room temperature were prepared by co-doping diamond with boron and sulfur. (It appeared rather difficult to find an appropriate donor for diamond; the well-known donors, nitrogen and phosphorus, lie too “deeply” in the bandgap of diamond—1.6 and 0.6 eV below the conduction band—and therefore practically are not ionized at room temperature and hence do not contribute to the diamond conductivity.1 ) Obtaining sulfur-doped n-type diamond for the first time is quite a detective story. From 1999 to 2001, sulfur was reported to give n-type conductivity in synthetic diamond [19,20]. The supposed n-type samples were sent from Japan, where they were first prepared, to Israel for closer examination. However, the measurements indicated that the samples were contaminated with boron and therefore were p-type [21]. Post factum we can say that in works referenced in [19] and [20], the reactor was evidently used earlier for the deposition of boron-doped diamond; hence, it contained residual boron even after careful cleaning. In the samples prepared, the boron overcompensated the introduced sulfur. Later, n-type diamond films [16,17] were CVD-grown by deliberate co-doping of diamond with sulfur and boron, using H2 S and trimethylboron as sources of the S and B dopants, respectively. When the films were grown without adding trimethylboron to the feed gas (no traces of boron were present in the reactor), particle-induced x-ray emission (PIXE) analysis showed no boron in the grown films. Sulfur was not detected using secondary ion mass spectroscopy (SIMS) either, despite the presence of H2 S in 1

Nonetheless, irrespective of the relative position of the energy levels in the bandgap of diamond, nitrogen (a donor) can compensate boron (the adopted acceptor in diamond). Because nitrogen is always present (e.g., in methane used as raw material in the CVD-process), preparation of lightly doped diamond is quite a problem, since the required low acceptor concentration is inevitably obtained as a poorly controlled small difference of two large quantities (the boron and nitrogen concentrations).

4.3 THE EFFECT OF SEMICONDUCTOR NATURE OF DIAMOND

TABLE 4.1 Ref. 23.) Ratio in feed gas (C/H)feed

85

Growth conditions (co-doping with B and S) and the film properties. (Adapted from Ratio in feed gas (S/C)feed (ppm)

Ratio in feed gas (B/S)feed

Crystal orientation of diamond substrate

Seebeck coefficient α ∗ (mV/ ◦C)

Resistivity (Ohm cm)

S concentration (PIXE)/ Detection limit (ppm)

Below detection limit Below detection limit Below detection limit

39000 350000 59000

0/53 6/24 35/215

9.8 1.6 7.5 2050 63 9450

627/550 30/39 270/37 240/29 39/45 84/21

Prior to the boron introducing to reactor 0.40% 0.20% 0.10%

375 500 2000

0 0 0

{110} {111} {111}

After the boron introducing to reactor 0.50% 0.10% 0.10% 0.20% 0.10% 0.05%

1000 5000 2500 1250 100 29000

4 0.5 0.4 0.2 0 0.4

{111} {110} {111} {100} {110} {111}

0.32 103 −25 −78 −0.11 −150

the feed gas. The films appeared to be too resistive (see Table 4.1) for electrochemical or thermoelectric measurements. However, when the film growth was performed with both H2 S and trimethylboron to the feed gas, the grown films were either n- or p-type depending on the relative amount of boron (both residual and added) and sulfur (Table 4.1). The X-ray Photoelectron Spectroscopy (XPS), PIXE, and SIMS analyses showed that sulfur was incorporated into these films at any substrate crystal orientation. The samples were analyzed for the sulfur concentration profile by XPS and SIMS. Both methods showed that the sulfur concentration was high at the surface and decreased several orders of magnitude, to a relatively constant value, in the bulk diamond (Figure 4.5). For example, the SIMS analysis of the samples with {111} and {110} orientations showed the sulfur concentration of 2.5 1019 and 7.0 1016 cm−3 at the surface and in the bulk, respectively. The boron concentration there was 1.0 1018 and 3.0 1017 cm−3 , respectively.2 The sign of the thermoelectric effect (a sign of the Seebeck coefficient) gives a determination of the sign of the charge carriers in the solid. The thermoelectric power measurements were done using a two-point probe. The sign of the thermal electro motive force (EMF) was determined by measuring the voltage between the probes when the heater was turned on. For the n-type silicon samples, the cool probe was charged negatively, relative to the warm probe; for p-type samples, the cool probe was positive, relative to the warm probe. Determination was made of the sign of the thermal EMF, with two series of samples grown on {111} and {110} substrates from a gas mixture in which the sulfur-to-carbon (S/C) atomic ratio was held constant and the boron-to-sulfur (B/S) ratio increased. In these experiments, the residual concentration of boron in the 2 Here, we draw special attention to the enrichment of just the film-subsurface region with B and S. As the film growth goes on, the gas/diamond film interface (the depth = 0 in Figure 4.5) moves to the left, and the dopants’ concentrations in what currently was the subsurface region would take on their steady-state values. One may speculate that the heat effect of the doping process is so high that the system acts as if it “jumps through” its equilibrium state (one may remember the famous Belousov–Zhabotinsky reaction); as time goes, it comes back to the steady state.

86

ELECTROCHEMISTRY OF DIAMOND

Atomic concentration (cm–3)

1020

1019

Sulfur

Boron 1018

1017

1016 0

20

40

60 Depth (nm)

80

100

120

Figure 4.5 SIMS depth profile of sulfur and boron in a film grown on a (111) diamond substrate. Feed conditions were an S/C atomic ratio of 2000 ppm, a B/S ratio of 2000 ppm, and a 0.12% methane concentration. The film thickness was approximately 2.25 μm. (Reprinted from Ref. 16.)

reactor prior to each run was minimal. Figure 4.6 shows a plot of the Seebeck coefficient versus the B/S atomic ratio in the feed gas for a series of diamond films grown on the {111} face at a constant methane concentration (0.12%) and constant S/C atomic ratio (0.001). For these conditions, the change-over from n-type to p-type occurs at a B/S atomic ratio of approximately 0.23. Similar results were observed with a series of samples deposited on {110}-oriented substrates; however, here the change-over from n-type to p-type is observed at a B/S atomic ratio of approximately 0.12. Relatively lightly doped electrodes clearly demonstrated their semiconductor nature. A cyclic voltammogram taken at a moderately-boron-doped p-type diamond electrode (Figure 4.7 [22]) clearly shows that the cathodic reaction is suppressed, whereas the anodic reaction is fast and diffusion-controlled. And vice versa, at the sulfur-doped ntype diamond the forward direction of the current is the cathodic one, whereas the anodic current is small and has kinetic nature (Figure 4.8). Thus, as expected [2], the semiconductor diamond electrodes demonstrated electrical current rectification, even if poorly pronounced, so that the n- and p-type electrodes are as if they are mirror images of each other in their electrochemical properties. Equally, the n- and p-type diamond electrodes differ in their capacitive behavior. To estimate the acceptor or donor concentration in the semiconductor diamond, measurements of differential capacitance were used. The capacitance values are used in plotting the Mott-Schottky graphs showing the relationships between the inverse reciprocal capacitance (C −2 ) and the electrode potential E . Physically, the Mott-Schottky plot reflects the potential dependence of the space charge layer (more precisely, the depleted layer) thickness Lsc in the semiconductor:  Lsc =

2εε0 |E | eNA

1/2 (4.4)

4.3 THE EFFECT OF SEMICONDUCTOR NATURE OF DIAMOND

87

α*(MV/°C) 1 0.5

p-type

0 0.00 0.05 –0.5 n-type

0.10

0.15

0.20

0.25

0.30

0.35

0.40

–1 –1.5 –2 –2.5 B/S ratio in gas phase

Figure 4.6 Plot of the Seebeck coefficient versus the B/S atomic ratio in the feed gas for a series of diamond films grown on the {111} face at a constant methane concentration (0.12%), increasing trimethylboron flow, and constant S/C atomic ratio (0.001). The vertical dashed line shows the approximate B/S atomic ratio at which the passing from n-type to p-type occurs. (Reprinted from Ref. 17.)

I (μA)

40

20

0 0.5

–0.5

1.0 E (V)

–20

Figure 4.7 Cyclic voltammogram at polycrystalline p-type diamond electrode, taken in 1 M KCl + 0.008 M K3 Fe(CN)6 + 0.008 M K4 Fe(CN)6 . Potential scan rate is 20 mV/s. (Reprinted from Ref. 22.)

where ε and ε0 are the diamond and vacuum permittivities, respectively; e is the electron charge. The intercept at the potential axis is the electrode flat-band potential Efb ; the slope of the line allows determination of the noncompensated (NA ) acceptor or donor (ND ) concentration in the semiconductor by using the following formula: N =±

2 [d(C −2 )/dE ]−1 εε0 e

(4.5)

where the “+” or “–” sign relates to the donors and acceptors. [In what follows, when speaking of the acceptor or donor concentration, we shall mean the noncompensated

88

ELECTROCHEMISTRY OF DIAMOND

ja (mA cm–2) 0,15

0,00 –1,4

–0,7

0,0

0,7

1,4

2,8 2,1 E (V)

–0,15

–0,30

jc (mA cm–2)

1013 C2(μF2 CM4)

Figure 4.8 Potentiodynamic curves at single-crystal n-type diamond electrode, taken in 0.5 M H2 SO4 + 0.01 M K3 Fe(CN)6 (cathodic) and 0.5M H2 SO4 + 0.01 M K4 Fe(CN)6 (anodic). Potential scan rate is 200 mV/s. Film growth conditions: CH4 concentration 0.1%, S/C ratio in gas phase 60 ppm. (Reprinted from Ref. 17.)

1.0

5.0

0.0 –0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

E (V) Figure 4.9 Mott-Schottky plot for a sample grown at single crystal (111)-substrate. Film growth conditions: CH4 concentration 0.25%, S/C and B/C ratios in gas phase 0.0025 and 0.4, respectively. The negative slope of the line points to the presence of acceptors; the acceptor concentration calculated by the slope is 2.7 × 1018 cm−3 . (Reprinted from Ref. 17.)

acceptor concentration (NA − ND ), or, correspondingly, the noncompensated donor concentration.] When sulfur is not present, or present in relatively small amounts, the Mott-Schottky plots have a negative slope characteristic of p-type semiconductor electrodes. A representative example for a sample grown at a B/S atomic ratio of 0.40 is shown in Figure 4.9. Despite some scatter of experimental points, we can conclude from the intercept that the flat-band potential is 1.0 V.

4.3 THE EFFECT OF SEMICONDUCTOR NATURE OF DIAMOND

89

1011C–2(μF–2cm4)

2,0

1,5

1,0

0,5

0,0 0.0

0.5 E (V)

1.0

Figure 4.10 Mott-Schottky plot for a sample grown at single crystal (111)-substrate. Growth conditions: CH4 concentration 0.12%, S/C and B/C ratios in gas phase 0.0024 and 0.1, respectively. The positive slope of the line points to the presence of donors; the donor concentration calculated by the slope is 1.4 × 1020 cm−3 . (Reprinted from Ref. 17.)

At still lower B/S ratios, the as-grown samples were n-type. The Mott-Schottky plots show a positive slope, which indicates the prevalence of donors. A representative MottSchottky plot for one of the samples grown from a feed gas with a B/S atomic ratio of 0.1 is shown in Figure 4.10. Its flat-band potential is close to 0.1 V. Qualitatively, the difference in the flat-band potentials corresponds to the difference in the conduction type [2]. It is important that all samples whose thermal EMF sign showed n-type conductance showed Mott-Schottky plots characteristic of an n-type semiconductor. And vice versa, all films with p-type conductance, as evidenced by the thermal EMF measurements, have Mott-Schottky plots characteristic of a p-type semiconductor. By and large, qualitatively, the electrochemical behavior of sulfur-containing diamond electrodes is typical of n-type semiconductors. In particular, this is true for some potentiodynamic curves showing the current rectification. In accord with the type of conductance, the cathodic current flows in forward direction, whereas the anodic current is blocking. It should be noted that the pronounced rectifying properties of semiconductor/electrolyte interfaces have been rarely observed with semiconductors, diamond in particular [23]. The very idea of co-doping diamond with boron and sulfur is based on the suggestion that a small boron atom can facilitate incorporation of a large sulfur atom into diamond [24,25]. Indeed, the quantum-chemical calculations of dopant energies in diamond allowed one to conclude that the BS pair in a substitutional site is energetically more stable than isolated substitutional B and S atoms in diamond [26]. In addition, boron can affect the concentration of sulfur adsorbed on diamond surfaces. An increase in this concentration during the diamond film growth should, in turn, facilitate sulfur incorporation into the diamond. Gas-phase equilibria calculations performed for the feed gas compositions used for the film deposition, showed [26,27] that the hydrogen-boron-sulfur compound (HBS) and BS2 concentrations are greater than the CH3 concentration; and the B/S concentration, although less than the CH3 concentration, is still greater than the concentration of monatomic sulfur vapor.

90

ELECTROCHEMISTRY OF DIAMOND

Calculations of the adsorption energy for these species at the diamond surface were performed. The results show that boron- and sulfur-containing species adsorb on diamond more strongly than those containing only sulfur. For example, the (H3 C)3 C–S and (H3 C)3 C–BS bond energies were estimated to be 280 and 480 kJ/mole, respectively. Therefore, the calculated sulfur concentration on the growing diamond surface increased in the presence of boron by up to two orders of magnitude. Thus, the very process of diamond co-doping with boron and sulfur can be thought of as a two-stage (adsorption of a dopant’s compound and the further diamond overgrowing thereover) process. The detailed microscopic nature of the donors in boron/sulfur-containing diamond still is not understood. Because both the thermoelectric power and impedance (Mott-Schottky plots) measurements sample the near-surface region of the semiconductor, the donor centers are spread over the diamond bulk rather than restricted to its surface. In addition, these centers can form narrow bands in the diamond bandgap [18]. These mid-gap bands are likely the source of the n-type conductivity. The knowledge of the flat-band potential allows one to draw an energy diagram of the diamond/solution interface (Figure 4.11 [27]). The diagram shows the relationship between the electrode potential scale and the physical scale of electron energies. The connection between these scales has been revealed by Gurevich and Pleskov [28] (see also [29]) as: eE = 4.44 + ε

(4.6)

where the electrode potential E is in volts (versus the standard hydrogen electrode, SHE), the electrochemical potential/electron energy ε is in electron volts, and e = −1 is the charge of an electron. The electrochemical potential ε refers to the electron at rest in a vacuum near the electrode surface (but outside the field of purely surface forces). Equation (4.6) permits one to relate the electrode potentials E used by electrochemists to the electron energies measured by other means. For example, shown in Figure 4.11 are the electrochemical potentials (“Fermi levels in solution”) of several common electrochemical redox couples in aqueous solution. Also shown in Figure 4.11 are the estimated positions of the band edges of diamond in contact with an aqueous solution determined by measuring the flat-band potential. The flat-band potential Efb gives the position of the Fermi level EF on the electrode potential scale. Hence, by knowing the potential EF one can obtain the energy εF of the Fermi level from Equation (4.6). For moderately doped p-type diamond (NA ∼ 1019 cm−3 ), the difference between the Fermi level and the valence band maximum εF − εVBM can be estimated as ∼0.1 eV. Thus, knowing the bandgap of diamond εG = εCBM − εVBM = 5.5 eV, one can fix the energy of the conduction band minimum εCBM . In Figure 4.11, the band edges of diamond in aqueous solution were placed using EF = Efb ≈ 1.0 eV(SHE)—see earlier Figure 4.9; this is a typical value for slightly oxidized diamond electrodes. From Equation (4.6), the energy of the Fermi level is given by εF = eEF − 4.44εV = −1.0 − 4.44 = −5.44 eV. The energy of the valence band maximum is approximately 0.1 eV lower than εF ; hence, εVBM ≈ −5.54 eV. Because the bandgap of diamond is 5.5 eV, this puts the conduction band minimum at εCBM ≈ −0.04 eV. Because, according to the Franck–Condon principle, one may conclude that the electron transitions occur between levels with equal energy, and better overlapping of the electrochemical potential levels in solution and energy levels in the valence band of diamond (hence, facilitation of electrode reactions) must take place for redox couples

4.3 THE EFFECT OF SEMICONDUCTOR NATURE OF DIAMOND

91

with more positive equilibrium potentials. Conversely, because the connection has been made between electrode potentials in aqueous solution and the electron energies in diamond, one can use well-defined electrochemical couples to probe the band structure of diamond [6]. Indeed, it was shown (for relatively lightly doped diamond electrodes) that the electrode irreversibility increased as the equilibrium potential of the couple became more negative (moved higher in the bandgap of diamond) for the following series: Fe(o-phenanthroline) Cl3 (1.08 V); ferrocene(1,1 )dimethanol (0.42 V); Ru(NH3 )6 Cl3 (–0.01 V); methyl viologen (–0.45 V). This effect was attributed [6] to a decreased number of available charge carriers in the diamond for couples with electron energies higher in the gap (more negative electrode potentials). In this connection, we now come back to the dependence of the reaction rate on the charge carrier concentration in the semiconductor diamond bulk (Figure 4.4). At an “ideal” semiconductor electrode, the potential drop in the Helmholtz layer at interfaces is small compared to the potential drop in the space-charge layer in the solid [2]. Therefore, energy band edges εCBM and εVBM (Figure 4.11) must act as if “pinned” at the interface, irrespectively of the doping level, hence, on the bulk-free carrier concentration (which is approximately inversely proportional to the diamond resistivity). Therefore, the surfacefree carrier concentration in a semiconductor electrode also does not depend on the doping. Equally, the electrochemical reaction rate on the semiconductor electrode, which is proportional to the surface concentration of the charge carriers participating in the reaction, must be doping-independent [2]. This is not the case with diamond electrodes (see Figure 4.4). From the experimental RF vs. ρ dependence we thus draw the conclusion that the potential drop in the Helmholtz layer at diamond electrodes is not small. More particularly, the potential distribution at the diamond/electrolyte solution interface in

Hydrogen Terminated Diamond

E ε [V][eV]

Solution –6

–4.44

Diamond in Aqueous Solution

εCEM = 1.3 eV 0

εCEM = –0.04eV ε = –1.39eV

Li+ + e– = Li

5.5 eV

–2 –2

MV2+ + e– = MV+ 2H+ + 2e– = H2 O

5.5 eV -4

O2 + 4H+ + 4e– = 2H2O -6

ε = –1.39eV ε = –4.44eV ε = –4.83 eV (pH=14) ε = –5.66 eV (pH=0)

εCEM = –4.2 eV εCEM= -5.54 eV

Figure 4.11 Energy diagram of the diamond/solution interface. The electrode potential (E) scale is referred to the standard hydrogen electrode, and the electrochemical potential ε is referred to as the electron at rest in a vacuum. The electrochemical potentials ε for the couple O2 + 4H+ + 4e− = 2H2 O at pH = 0 and pH = 14 are shown along with the band edges for diamond in contact with an aqueous solution. Also shown are the positions of the two couples: Li+ /Li (E ◦ = −3.045 V (SHE); ε = −1.39 eV) and methyl viologen (E ◦ = −0.45 V (SHE); ε = −3.99 eV). (Reprinted from Ref. 27.)

92

ELECTROCHEMISTRY OF DIAMOND

redox solutions depends on the diamond doping level—even more so for more heavily doped samples. And this is the reason why the semiconductor diamond behavior is far from “ideal.” (For example, no current rectification is observed on more heavily doped diamond electrodes; see Figure 4.2). Thus, we see that moderately doped diamond demonstrates almost ideal semiconductor behavior in inert background electrolytes, in particular, linear Mott-Schottky plots, which indicates band edge “pinning” at the semiconductor surface. In redox electrolytes, however, a metal-like behavior is observed, with the band edges “unpinned” at the surface. This phenomenon, although not yet fully understood, has been observed with numerous semiconductor electrodes (e.g., silicon, gallium arsenide, and others) [30]. It must be associated with chemical interactions between the semiconductor and the redox system, which results in a large and variable Helmholtz potential drop.

4.4 THE EFFECT OF CRYSTAL STRUCTURE ON THE ELECTROCHEMICAL BEHAVIOR OF DIAMOND 4.4.1

The Effect of Crystallographic Orientation of Crystal Faces

It is of interest to compare the electrochemical reaction kinetics at differently oriented faces of single crystal diamond electrodes. For this purpose, single-crystal films deposited onto appropriately oriented faces of undoped (that is, dielectric) single crystal diamond substrates should be used, rather than the polycrystalline films deposited onto nondiamond (metal or silicon) substrates. Here, the growing film replicates the substrate crystal structure (epitaxial growth), yielding a single-crystal film. Electrodes with (111), (110), and (100) faces were prepared in this manner [31,32]. All three crystals were grown simultaneously (at the same trimethyl borate concentration in the gas phase). The acceptor concentration was determined using differential capacitance measurements. Given the Mott-Schottky plots for the films in Figure 4.12, the acceptor concentration values calculated from slopes of the lines are given here (in cm−3 ): (100) (2–3) × 1019

(110) 1.3 × 1020

(111) (5–7) × 1020

Polycrystalline 5 × 1020

The polycrystalline thin-film electrode approaches the (111)-oriented epitaxial one in its electrode behavior; probably the crystallites constituting the polycrystalline film are mainly edged by (111) faces. In Figure 4.13a, b, we give cyclic voltammograms taken in [Ru(NH3 )6 ]2+/3+ and [Fe(CN)6 ]3−/4− redox solutions at differently oriented single-crystal thin-film electrodes. At the electrodes with (111) and (110) faces (as well as polycrystalline film studied for comparison) the curves have the shape characteristic of irreversible, yet reasonably fast electrode reactions. Particularly, the reaction is more reversible at a (111) than (110) face as follows from the difference of the anodic and cathodic current peak potentials Ep —that is, 450 and 555 mV, respectively. (Recall that this difference is a measure of the reaction reversibility [12].) By contrast, at a (100)-oriented electrode the curve has no current peaks at all; the process must be under kinetic, rather than diffusion, control. These results obtained at the (111) and (100) faces agree well with those of Kondo et al. [33] in that the reactions at the former face proceed more quickly than at the latter one.

4.4 THE EFFECT OF CRYSTAL STRUCTURE ON THE ELECTROCHEMICAL

93

C –2 x 10–12(F –2 cm4)

2.0 (100)

1.0

(111)

(110)

0.0 0.0

0.5

1.0

1.5

E (V) Figure 4.12 Mott-Schottky plots for diamond films orientated as (110), (100), and (111), in 2.5 M H2 SO4 solution. (Reprinted from Ref. 31.)

The difference in the electrochemical activity of different crystal faces manifests itself also in the transfer coefficients of the anodic and cathodic reactions α and β. At the (111) face, these quantities are 0.40 and 0.44, respectively; for the (110) face, the quantities are 0.2 and 0.23, respectively. The electrode behavior of the (111) face approaches metallic behavior; that of the (110) face is characteristic of a semiconductor. Thus, the electrode reactions in the [Fe(CN)6 ]3−/4− and [Ru(NH3 )6 ]2+/3+ redox couples experience retardation in the following sequence: (polycrystalline) ≈ (111) >(110) >(100). This is associated with the decrease, in the same sequence, of the doping level in diamond bounded by these crystal faces. Earlier we showed that the rate of electrochemical reactions at polycrystalline diamond electrodes is approximately proportional to the acceptor concentration in a boron-doped diamond (Figure 4.3). The difference in the boron concentration in single-crystal films is explained by well-known different intensities of boron incorporation from gas phase into different diamond crystal faces during the film growth. (In particular, the (111) face “absorbs” boron most intensely; the (100) one is less intense.) If, however, the (111)- and (100)-oriented films are compared at equal acceptor concentration, practically no difference in their electrochemical activity is observed; compare the corresponding cyclic voltammograms in Figures 4.13a and 4.13c. (To this purpose, another (111)-oriented film was deposited separately, the trimethyl borate concentration in the feed gas being chosen much lower than in the first growth process, in order to obtain as lightly doped film as the previously grown (100)-oriented film. The acceptor concentration in this film is 6 × 1018 cm−3 , which is not far from that in the (100) film.) Similar differences in the electrochemical reaction rates were observed [34,35] between different faces of a diamond single crystal grown by the more traditional method—namely, at high temperature and high pressure from a carbon solution in molten metal. The differential capacitance measurements revealed differences in the doping level of the so-called growth sectors (pyramids) inside the crystal (the sectors begin with seed-crystal and end as the faces of different hkl ).

94

ELECTROCHEMISTRY OF DIAMOND

(a) 0.008

(111)

j (Acm–2)

0.004

(110) (100)

0.000

–0.004

–0.008 –0.5

–1.0

0.5

–0.0

1.0

1.5

E (V) (b) 0.004

(111)

0.002 j (Acm–2)

(110) (100)

0.000

–0.002

–0.004

–0.8

–0.6

–0.4

0.2 E (V)

0.0

0.2

0.4

(c) j (mA cm–2)

0.16

0.08

–1

0

1

2 E (V)

–0.08

Figure 4.13 Cyclic voltammograms taken in (a) 0.01 M K3 Fe(CN)6 + 0.01 M K4 Fe(CN)6 + 2.5 M H2 SO4 and (b) 0.1 M NaCl + 0.005 M Ru(NH3 )6 Cl2 + 0.005 M Ru(NH3 )6 Cl3 solutions at singlecrystal (epitaxial) thin-film diamond electrodes with different crystallographic orientation of the faces (shown in the figure). All films were deposited under identical conditions (the same trimethyl borate concentration in gas phase). (c) Cyclic voltammogram taken at lightly doped (111)-electrode in 2.5 M H2 SO4 + 0.01 M K3 Fe(CN)6 + 0.01 M K4 Fe(CN)6 solution. Potential scan rate is 10 mV s−1 . (Reprinted from Ref. 31.)

4.4 THE EFFECT OF CRYSTAL STRUCTURE ON THE ELECTROCHEMICAL

111

95

111

111 100

100

111 Figure 4.14 Schematic presentation of the studied diamond single crystal grown at high temperature and high pressure.

For illustration, Figure 4.14 gives a scheme of a diamond single crystal bounded by (111) and (100) faces. By studying its differential capacitance and kinetics face by face (by insulating the rest of the surface), the electrode behavior of individual faces was revealed, as described earlier; however, here the data were collected from a unique single crystal, rather than from three different single-crystal (epitaxial) films. The observed differences in the electrochemical properties (the impedance and kinetic parameters) of differently oriented crystal faces can be explained mainly by the difference in the acceptor (boron) concentration in the growth sectors terminated by these faces, which in turn is caused by the different ability of the faces to incorporate boron during the synthesis. Evidently, finer effects, if any (e.g., the dependence of electrochemical properties of individual faces on their atomic density, etc.), can be elucidated by comparing faces with different hkl at a constant boron concentration in diamond; for this purpose, a set of electrodes should be appropriately prepared and examined. To summarize this section, at first glance, the “structural” aspect of the diamond electrode behavior actually reduces to a purely “semiconductor” one—that is, the dependence of the electrochemical activity on the doping level of diamond. 4.4.2

The Effect of Surface Morphology

Varying the micro- or nano-sized roughness is another method of structurally diversifying surfaces. We analyzed the effect of the “chaotic” microroughness on the electrochemical kinetics for the previously discussed case of comparable rates of interfacial charge transfer and mass transfer in solution. In Lim et al. [5] and in Pleskov et al. [36], the diamond surface roughness was controlled by varying the roughness of titanium substrates, onto which the diamond film was deposited. In particular, prior to the deposition, the substrates were annealed in air in order to change the crystal structure of the titanium. On removing the scale from titanium by sand blasting, the annealed plates were subjected to etching in hot 6 M HCl for different times. Then the surface topography, average roughness Ra and roughness factor Sef (the true-to-geometrical surface ratio) were measured using a White Light Interferometric Microscope. Shown in Figure 4.15a is an SEM photomicrograph of a roughened titanium substrate prepared in this manner.

96

ELECTROCHEMISTRY OF DIAMOND

(a)

(b)

Figure 4.15 SEM images of (a) roughened Ti substrate surface and (b) diamond film deposited onto the roughened substrate. Note the difference in scale bars: 5 μm and 1 μm, respectively. (Reprinted from Ref. 36.)

4.4 THE EFFECT OF CRYSTAL STRUCTURE ON THE ELECTROCHEMICAL

97

Upon depositing diamond films on the rough substrates (Figure 4.15b), the average roughness Ra and the roughness factor Sef changed but insignificantly. The measured electrochemical characteristics of the rough diamond electrodes are summarized in Table 4.2. We begin their analysis with the differential capacitance measured in background electrolyte. This procedure reveals the general trend in the electrode capacitance changes with varying substrate pretreatment. As follows from Table 4.2, in the sequence of the etched substrates the capacitance by and large follows the value of surface roughness. The same is true for the background current in supporting electrolyte. Concerning the kinetic data, we can conclude, first, that the transfer coefficients α and β are less than 0.5 and their sum α + β is less than 1 (which is characteristic of semiconductors [2], see earlier text). Second, the kinetic characteristics of diamond electrodes are affected by the microroughness of the diamond surface. With increasing roughness, the irreversible character of the electrochemical reaction is less pronounced, as we see from the quantity Ep . The electrochemical activity increases monotonically with increasing roughness, which follows from an increase in α + β. However formal this criterion may be, it still points to the apparent increase in the electrochemical activity, which may be explained by a decrease in the true current density in the series and, hence, gradually approaching the reversible mode of charge transfer. It is noteworthy that the difference in the true surface area did not affect the potentiodynamic curves of the oxidation and reduction reactions in the [Fe(CN)6 ]3−/4− redox system, because the measured quasi-steady-state current is controlled by diffusion, rather than kinetics. Indeed, under the quasi-steady-state conditions of the experiment, the diffusion front has already propagated far away from the electrode surface (to a distance well exceeding the microroughness size). Thus, it is the geometrical, rather than the true, electrode surface area that is the crucial issue in the evaluating of the reaction rates, unlike the kinetic parameter Ep (the anodic and cathodic current peak separation), as well as the background current, that are determined by the true current density; hence, it depends on the true surface area. Therefore, the roughening of the diamond surface gives no gain in the measured current; however, the roughening improves the charge transfer and facilitates the passing of the reaction from kinetic to diffusion control. This may be beneficial for electroanalytical applications, for example.

TABLE 4.2 The electrochemical parameters for roughened polycrystalline diamond electrodes. (Adapted from Ref. 5.) Roughness Seff

2.29 2.44 2.64

Background current in 2.5 M H2 SO4 measured at the potential scan rate of 50 mV s−1 (μA cm−2 )

Potential window width (V)

4.46 5.40 8.45

2.8 2.9 2.95

Transfer coefficients for [Fe(CN)6 ]3−/4− system α β α+β

0.34 0.37 0.38

0.37 0.39 0.39

0.71 0.76 0.77

Differential capacitance C in 2.5 M H2 SO4 per 1cm2 of geometrical surface (μF cm−2 )

Ep (V)

6.9 7.5 8.3

0.36 0.30 0.29

98

ELECTROCHEMISTRY OF DIAMOND

4.4.3 The Effect of the Diamond Grain Size (or the Film Thickness, or the sp2 -Carbon Impurity) We now discuss the effects introduced to the diamond electrode behavior by some peculiarities in the electrode structure. In particular, the electrochemical properties of polycrystalline films depend somewhat on the diamond grain size [37]. As any crystalline solid, diamond is not free from crystal defects that often take the form of nondiamond (sp 2 -) carbon. In particular, intercrystalline boundaries (containing disordered carbon) can be thought of as defects. The effect of the nondiamond carbon on the electrochemical behavior of diamond has been discussed in the literature; different methods of introducing sp 2 -carbon into diamond have been used (e.g., diamond graphitization under the action of high-energy ions [38–40]). Nondiamond carbon was deliberately imposed on a diamond film surface by mechanically rubbing it with a sample of highly oriented pyrolytic graphite. Next, the kinetics of electrochemical reactions in the quinone/hydroquinone and [Fe(CN)6 ]3−/4− redox solutions were studied [41,42]. In both cases the nondiamond carbon made the reactions more reversible, which suggested that the sp 2 -carbon forms active sites at diamond surfaces. Varying the film thickness is another method of the controlling the sp 2 -carbon impurity in polycrystalline CVD-diamond films. Here, the dependence of diamond crystallite size on the film thickness is used. In Figure 4.16, the cross-section of such a film is schematically presented. The crystallites start growing at the substrate surface that is preliminarily sowed with seed-nanocrystallites. Columnar crystallites are oriented at right angle to the substrate surface. Because their size increases gradually as the film grows (“stronger” crystallites suppress the growth of “weaker” ones), the intercrystalline boundary network is denser in the fine-grained layer attached to the substrate than near the film’s growth side. The growth side is faceted; the size of its constituent diamond crystallites l is assumed to be approximately proportional to the film full thickness d , which is demonstrated in Figure 4.17, where SEM photomicrographs of the surfaces of diamond films with different thickness are given [37]. The outcroppings of the intercrystalline boundaries at the surface may be thought of as “active sites.” It is clear that one may control the intercrystalline boundary network density, and hence the disordered carbon content near and at the electrode surfaces, by merely specifying the film growth time. Qualitatively, the amount of the disordered carbon content near and at the electrode surface can be estimated by Raman spectroscopy. Shown in Figure 4.18 are the Raman spectra for two films of different thickness. We see that in the spectrum of the thinner film, the “diamond” peak (at 1332 cm−1 ) is lower, whereas the sp 2 -carbon maximum is more pronounced than those of the thicker film. For the two films, the thickness differs by a factor of 8, while the ratio of the “diamond” maximum is only twice as high as the “nondiamond” maximum. Figure 4.18 clearly shows that the amount of nondiamond carbon in films increases with the thickness of the film. The crystal size at the films’ surfaces was estimated semiquantitatively by counting the number of crystals per unit length (see Table 4.3, second column). Cyclic voltammograms measured in an indifferent electrolyte solution at diamond film electrodes of different thickness are given in Figure 4.19. They are typical for borondoped diamond electrodes of rather high quality: The potential window is wide; even the thinnest (0.5 μm-thick) film evidently has no through-holes; and the background current in mid-window is as low as ∼10 μA cm−2 . We see that with decreasing the film

4.4 THE EFFECT OF CRYSTAL STRUCTURE ON THE ELECTROCHEMICAL

99

l

d

Substrate

Figure 4.16

Schematic presentation of cross-section of polycrystalline diamond film.

TABLE 4.3 Dependence of electrochemical properties of diamond film electrodes on their thickness. (Adapted from Ref. 37.) Thickness d (μm) 0.5 1.4 1.83 4.0

Crystallite average size (μm)

Transfer coefficients α

0.57 ± 0.07 0.75 ± 0.10 1.1 ± 0.03 2.02 ± 0.7

(0.8) 0.15 0.13 0.12

β

Background current JBG (μA cm−2 )

Differential capacitance C (μF cm−2 )

0.36 0.17 0.17 0.1

10 5 5 1

4.94 4.95 3.17 2.27

thickness, the potential of the onset of anodic chlorine evolution (the anodic boundary of the potential window) becomes ever less positive; both the background current and the differential capacitance in the mid-window region increase (Table 4.3). In other words, the overvoltage for the Cl2 (and O2 ) evolution is less for thinner films (and finer grains at their surfaces). It was mentioned earlier that the kinetics of gas evolution reactions is sensitive to the adsorption of their intermediates at the electrode surface. Evidently, the sp 2 -carbon forms additional adsorption sites on the surface, while crystalline diamond per se is a poor adsorbent. The effect of diamond film thickness on the electrochemical kinetics was further explored by measuring the transfer coefficients in the [Fe(CN)6 ]3−/4− model redox system; of the two previously described methods, the second one (impedance spectra taken at the equilibrium potential in K3 [Fe(CN)6 ] + K4 [Fe(CN)6 ] solutions) was applied. Equally, in the [Fe(CN)6 ]3−/4− redox system, with the decreasing of the film thickness d , the values of the transfer coefficients α and β increased from ∼0.15 (which is characteristic of semiconductor electrodes [28] and “poor conductors” in general) up to ∼0.35—that is, approaching the values characteristic of metal-like electrodes (Table 4.3). Another method of exploring the effect of grain size on the electrochemical properties is to compare, on a free-standing diamond film (that is, after the etching-off its substrate), the film’s growth side and its “nucleation” side. The former consists of rather perfect

100

ELECTROCHEMISTRY OF DIAMOND

(a)

(b)

Figure 4.17 SEM photomicrography of surfaces of diamond films with different thickness: (a) 2.6 μm, (b) 3.6 μm. (Reprinted from Ref. 37.)

4.4 THE EFFECT OF CRYSTAL STRUCTURE ON THE ELECTROCHEMICAL

101

160 140 Intensity (arb. units)

d = 0.5 μm 120 100 80 60 d = 4 μm 40 20 1000

1200

1400

1800

1600

Raman shift (cm−1) Figure 4.18 Raman spectra for two films of different thickness. (Reprinted from Ref. 37.)

2.5 μm j (A / cm–2) 3.7 μm d = 0.5 μm 0.5

–1.0

0.0

1.0

2.0 E (V)

Figure 4.19 Cyclic voltammograms measured in indifferent electrolyte solution (1 M KCl) at diamond film electrodes of different thickness (determination of the potential window). Potential scan rate is 50 mV/s. (Reprinted from Ref. 37.)

large crystallites; the latter has defective submicrocrystallites (compare Figure 4.16) and approaches, in its behavior, on the surfaces of very thin diamond films. Such a comparison showed that for thicker films, the difference in electrochemical behavior is quite striking: The electrochemical activity of the coarse-grained growth surface is rather poor, whereas the “nucleation” surface is noticeably more active. In particular, it demonstrates higher

102

ELECTROCHEMISTRY OF DIAMOND

differential capacitance and higher acceptor concentration (the crystal lattice defects may play the role of acceptors) [43,44]. Thus, leaving the question of the corrosion stability open (in all probability, sp 2 -carbon “impurity” impairs the corrosion resistance), we may conclude, even if qualitatively, that the nondiamond carbon “impurity” in crystalline diamond electrodes favors their electroactivity. Very close results were obtained by Bennett et al. [45] who introduced the sp 2 -bonded nondiamond carbon impurity through adjustment of the C/H source gas ratio used during the deposition of polycrystalline diamond films. To conclude this section, we suggest that the “crystallographic” (at first glance) effect of grain size on the CVD-diamond films’ electrochemical properties arises mainly from variations in the amount of sp 2 -carbon therein.

4.5 DIAMOND-BASED NANOSTRUCTURES AS ELECTRODE MATERIALS: VACUUM-ANNEALED UNDOPED POLYCRYSTALLINE DIAMOND By no means is doping the only way of imparting electrical conductance to dielectric diamond. For example, time and again it was attempted to convert dielectric single crystal diamond to conducting electrodes by implanting it with ions. This resulted in the formation of a conducting layer on the diamond surface; however, it appeared to be due to diamond graphitization under the action of the high-energy ions [38–40]. Strictly speaking, such an electrode with the graphite layer on top of diamond cannot be called “diamond electrode.” One more way to convert insulating diamond to a room-temperature-conducting material is the high-temperature annealing of undoped (dielectric) diamond films in a vacuum. The changes in the structure of these films, after annealing at temperatures up to 1900 K, were studied by optical and spectroscopic methods (see, e.g., Khomich et al. [46]). By using high-resolution transmission electron microscopy (HRTEM) [47], the formation of amorphous carbon and even crystalline graphite layers up to ∼20 nm thick along the intercrystalline boundaries was detected (see Figure 4.20, [48]); islets of graphitelike material in defective areas inside the crystallites also emerged. The amount of the newly formed nondiamond phase depends on the annealing temperature and time, as well as the “quality” of the initial films: The more defective the film, the easier to partially graphitize it. The nondiamond phase is formed from disordered carbon constituting the intercrystalline boundaries, as well as the adjacent periphery of diamond crystallites proper. Noteworthy is that the annealing at temperatures exceeding 1700 K forms dislocations in the diamond crystallites near the intercrystalline boundaries. (These dislocations cause specific photoelectrochemical effects at the electrodes made of vacuum-annealed undoped polycrystalline diamond [48,49].) The formation of defects is due to mechanical stresses caused by the diamond-to-graphite transformation. The stresses emerge because of the large difference in the densities of diamond (3.51 g cm−3 ) and nondiamond carbon (∼2 g cm−3 ); hence, the nondiamond carbon requires nearly twice as much volume as the initial diamond. Because diamond is incompressible, high inner pressure arises. (After the annealing at ∼1970 K, the stress is so strong that the polycrystalline diamond disintegrates into separate crystallites [50]). This result is specific to the annealing of polycrystalline diamond. By contrast, when single crystal diamond is subjected to annealing, the defect concentration therein decreases, as is usual with single crystal semiconductors.

4.5 VACUUM-ANNEALED UNDOPED POLYCRYSTALLINE DIAMOND

103

Figure 4.20 High-resolution transmission electron photomicrograph showing intergranular boundary in an undoped polycrystalline diamond film annealed in vacuum at 1670 K. (Reprinted from Ref. 48.)

It is noteworthy that the nondiamond carbon of the anneal-transformed intercrystalline boundaries cannot be thought of as being identically equal to ordinary graphite or amorphous carbon because this material exists at high pressure in the diamond matrix. The pressure may alter its properties, in particular, the electrochemical properties. The amorphous carbon or graphite-like phase emerging at the intercrystalline boundaries form a continuous conducting network permeating the entire diamond film from its nucleation side to the growth side (compare Figure 4.16); the undoped dielectric diamond crystallites appear within the nondiamond matrix. Hence, the initial dielectric film is apparently converted to a (two-phase) conducting material. The outcroppings of the intercrystalline boundaries at the diamond film surfaces may play the role of active spots where the charge transfer across the electrode/electrolyte solution takes place. The electrochemical properties of this novel material were studied [51–53] using a 400 μm-thick diamond film grown on silicon substrate. The substrate was etched off in a nitric acid–hydrofluoric acid mixture, leaving a free-standing film. The specimens were annealed in a vacuum (10−5 Torr) furnace at temperatures from 1725 to 1925 K for one hour. The electrochemical studies were carried out at both sides of each sample—that is, the faceted growth surface (grain size 20–60 μm) and the smooth nucleation surface formed by submicron crystallites. After annealing at temperatures above 1825 K, the initially dielectric films became conducting, which allowed them to be used as electrodes. With a further increase in temperature to 1915 K, the effective resistivity of this two-phase material decreased from the initial value of 1011 –1012 Ohm cm to less than 0.1 Ohm cm; the differential capacitance increased from ∼10−3 up to ∼50 μF per 1 cm2 of geometrical surface. As to the potential window, this material is only slightly inferior to the boron-doped diamond. With an increase in the annealing temperature, all kinetic properties also change. For

104

ELECTROCHEMISTRY OF DIAMOND

(a)

β

0.4

0.2

0 1840

1860

1880

1900

1920

Tann (K) 0.5 (b)

i (mA / cm–2)

0.4

0.3

0.2

0.1

0 1820

1840

1860

1880

1900

1920

Tann (K) Figure 4.21 Kinetic characteristics of vacuum-annealed polycrystalline films as functions of the annealing temperature: (a) transfer coefficient for [Fe(CN)6 ]4− anodic oxidation reaction; (b) cathodic peak height for [Fe(CN6 ]3− cathodic reduction reaction (◦ growth side, • nucleation side). (Reprinted from Ref. 52.)

example, the transfer coefficients in the [Fe(CN)6 ]3−/4− redox couple increased from ∼0.2 to 0.4–0.5 (see Figure 4.21a); the current per se also grows (Figure 4.21b). By and large, both sides of the film (the growth side and the nucleation side) behave similarly to each other. Their possible kinetic difference (caused by the difference in the density of the intercrystalline boundary network) must be leveled off because the reactions in the model redox couple [Fe(CN)6 ]3−/4− are fast and therefore are mainly diffusion-controlled. With increasing annealing temperature, the anodic and cathodic peak current potentials approach each other in the cyclic voltammograms (Figure 4.22a), eventually reaching the “theoretical” value of 59 mV for a one-electron reaction (Figure 4.22b) [12,13]; this points out that the electrode reaction becomes more reversible with an increase in the annealing temperature. The observed annealing-induced changes in the electrode characteristics are associated with the formation of a nondiamond phase along the intercrystalline boundaries. With

4.5 VACUUM-ANNEALED UNDOPED POLYCRYSTALLINE DIAMOND

(a)

105

I (μA) 100

50

0 0.3

0.5

0.7 E (V)

50

100 (b) 1.0

ΔEp (V)

0.8 0.6 0.4 0.2 0 1820

1840

1860

1880

1900

1920

Tann (K) Figure 4.22 (a) Cyclic voltammogram taken in 0.01 M K4 Fe(CN)6 + 0.01 M K3 Fe(CN)6 + 2.5 M H2 SO4 solution at growth side of a film annealed at 1910 K. Potential scan rate is 5 mV/s; (b) dependence of difference of the anodic and cathodic current peak potentials Ep on the annealing temperature (◦ growth side, • nucleation side). Dotted line shows the theoretical Ep value for a reversible reaction. (Reprinted from Ref. 52.)

increasing annealing temperature, the amount of this newly formed phase increases, which is the principal reason for the increase in the differential capacitance because the total area of the “active spots” at the insulating film surface increases. The conductivity of the nondiamond phase also grows. Indeed, the growth of the capacitance has always kept ahead of the growth of the ohmic resistance, as the impedance data showed [54]. These two factors in aggregate underlie the transition of the material from the category of “poor conductor” to that of a “metal-like” conductor. One can see that by and large the behavior of the vacuum-annealed undoped polycrystalline diamond resembles that of boron-doped diamond deliberately enriched with the sp 2 -carbon, described in the preceding section.

106

4.6

ELECTROCHEMISTRY OF DIAMOND

CONCLUSIONS

To conclude, we emphasize the common feature of the variety of diamond-based materials, in particular, microcrystalline diamond (either boron-doped or vacuum-annealed), nitrogenated ultrananocrystalline diamond [55,56], as well as sp 3 -diamond-like carbon doped with metals or nitrogen [57–59]. With increase of conductivity (whatever is its reason), the material properties (electrochemical, in particular) change from those of a “poor conductor” (semiconductor) to metal-like properties.

4.7

ACKNOWLEDGMENTS

The author is grateful to the colleagues who invaluably contributed to the studies overviewed in this chapter: Marina Krotova, Yulia Evstefeeva, Valentin Varnin, Irina Teremetskaya, John Angus, Sally Eaton, Viktor Ralchenko, and Roman Khmelnitskiy. I express my thanks to John Angus for valuable discussions and his stylistic improvement of the text. Financial support from Russian Foundation for Basic Research (project no. 10-03-00011) is acknowledged.

REFERENCES 1. Yu.V. Pleskov, A.Ya. Sakharova, M.D. Krotova, L.L. Bouilov, B.V. Spitsyn, J. Electroanal. Chem., 1987, 228 , 19. 2. Yu.V. Pleskov, Yu.Ya. Gurevich, Semiconductor Photoelectrochemistry, Consultants Bureau, New York, 1986. 3. Yu.V. Pleskov, Elektrokhimiya almaza (Electrochemistry of Diamond), Editorial URSS, Moscow, 2003. 4. D.M. Malta, J.A. von Windheim, H.A. Wynands, B.A. Fox, J. Appl. Phys. 1995, 77 , 1536. 5. P.Y. Lim, F.Y. Lin, H.C. Shih, V.G. Ralchenko, V.P. Varnin, Yu.V. Pleskov, S.F. Hsu, S.S. Chou, P.L. Hsu, Thin Solid Films 2008, 516 , 6125. 6. N. Vinokur, B. Miller, Y. Avyigal, R. Kalish, Electrochem. Solid State Lett . 1998, 1 , 265. 7. I. Yagi, H. Notsu, T. Kondo, D.A. Tryk, A. Fujishima, J. Electroanal. Chem. 1999, 473 , 173. 8. H. Notsu, T. Fukazawa, T. Tatsumo, D.A. Tryk, A. Fujishima, Electrochem. Solid State Lett . 2001, 4 , H1. 9. G.M. Swain, Adv. Mater. 1994, 6 , 388. 10. Q. Chen, M. Granger, T.E. Lister, G.M. Swain, J. Electrochem. Soc. 1997, 144 , 3806. 11. Yu.V. Pleskov, Yu.E. Evstefeeva, M.D. Krotova, P.Y. Lim, S.S. Chu, I.I. Vlasov, V.V. Kononenko, V.P. Varnin, I.G. Teremetskaya, H.C. Shih, Elektrokhimiya 2005, 41 , 387. 12. Z. Galus, Fundamentals of Electrochemical Analysis, Ellis Horwood, Chichester, 1976. 13. P. Delahay, New Instrumental Methods in Electrochemistry, Interscience Publishers, New York, 1954. 14. A.D. Modestov, Yu.E. Evstefeeva, Yu.V. Pleskov, V.M. Mazin, V.P. Varnin, I.G. Teremetskaya, J. Electroanal. Chem. 1997, 431 , 211. 15. A.D. Modestov, Yu.V. Pleskov, V.P. Varnin, I.G. Teremetskaya, Elektrokhimiya 1997, 33 , 60. 16. S.C. Eaton, A.B. Anderson, J.C. Angus, Yu.V. Pleskov, Yu.E. Evstefeeva, Electrochem. Solid State Lett . 2002, 5 , G65.

REFERENCES

107

17. S.C. Eaton, J.C. Angus, A.B. Anderson, Yu.E. Evstefeeva, Yu.V. Pleskov, Elektrokhimiya 2003, 39 , 170. 18. S.C. Eaton, A.B. Anderson, J.C. Angus, Y.E. Evstefeeva, Y.V. Pleskov, Diam. Rel. Mat . 2003, 12 , 1627. 19. I. Sakaguchi, M. Nishitani-Gamo, Y. Kukuchi, E. Yasu, H. Haneda, T. Suzuki, T. Ando, Phys. Rev. B 1999, 60 , 2139. 20. M. Nishitani-Gamo, C. Xiao, Y. Zhang, E. Yasu, Y. Kukuchi, I. Sakaguchi, T. Ando, Thin Solid Films 2001, 382 , 113. 21. R. Kalish, A. Resnick, C. Uzan-Saguy, C. Cytermann, App. Phys. Lett . 2000, 76 , 757. 22. Yu.V. Pleskov, Elektrokhimiya 2003, 39 , 1411. 23. Yu.V. Pleskov, in Advances in Electrochemical Science and Engineering ( R.C. Alkire, D.M. Kolb, eds.), vol. 8, Wiley—VCH, Weinheim, 2003, p. 209. 24. S.C. Eaton, A.B. Anderson, J.C. Angus, Y.E. Evstefeeva, Yu.V. Pleskov, Proceedings of the 6th Applied Diamond Conference/Second Frontier Carbon Technology Joint Conference, NASA/CP-2001-210948, 2001, p. 164. 25. S.C. Eaton, A.B. Anderson, J.C. Angus, Y.E. Evstefeeva, Y.V. Pleskov, in Diamond Materials VII ( G.M. Swain, T. Ando, J.C. Angus, W.D. Brown, J.L. Davidson, A. Gicquel, W.P. Kang, B.V. Spitsyn, eds.), PV 2001-25, The Electrochemical Society, Pennington, New Jersey, 2001. 26. T. Albu, A.B. Anderson, J.C. Angus, Diamond Materials VII ( G.M. Swain, T. Ando, J.C. Angus, W.D. Brown, J.L. Davidson, A. Gicquel, W.P. Kang, B.V. Spitsyn, eds.), PV 2001-25, The Electrochemical Society, Pennington, New Jersey, 2001. 27. J.C. Angus, S.C. Eaton, Yu.V. Pleskov, in Thin-Film Diamond II ( C.E. Nebel, J. Ristein, eds.), Academic Press, Amsterdam, 2004, p. 97. 28. Yu.Ya. Gurevich, Yu.V. Pleskov, Elektrokhimiya 1982, 18 , 1477. 29. A.J. Bard, R. Memming, B. Miller, Pure Appl. Chem. 1991, 63 , 569. 30. Yu.V. Pleskov, Yu.Ya. Gurevich, in Modern Aspects of Electrochemistry ( B.E. Conway, R.E. White, J. O’M. Bockris, eds.), vol. 16, Plenum Press, New York, 1985, p. 189. 31. Yu.V. Pleskov, Yu.E. Evstefeeva, V.P. Varnin, I.G. Teremetskaya, Elektrokhimiya, 2005, 41 , 1023. 32. Yu.V. Pleskov, Yu.E. Evstefeeva, M.D. Krotova, V.P. Varnin, I.G. Teremetskaya, J. Electroanal. Chem. 2006, 595 , 168. 33. T. Kondo, Y. Einaga, B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, J. Electrochem. Soc. 2002, 149 , E179. 34. Yu.V. Pleskov, Yu.E. Evstefeeva, M.D. Krotova, V.Ya. Mishuk, V.A. Laptev, Yu.N. Palyanov, Yu.M. Borzdov, J. Electrochem. Soc. 2002, 149 , E260. 35. Yu.V. Pleskov, Yu.E. Evstefeeva, M.D. Krotova, V.Ya. Mishuk, V.A. Laptev, Yu.N. Palyanov, Yu.M. Borzdov, Elektrokhimiya 2002, 38 , 698. 36. Yu.V. Pleskov, Yu.E. Evstefeeva, M.D. Krotova, P.Y. Lim, H.C. Shih, V.P. Varnin, I.G. Teremetskaya, I.I. Vlasov, V.G. Ralchenko, J. Appl. Electrochem. 2005, 35 , 857. 37. Yu.V. Pleskov, M.D. Krotova, V.P. Varnin, I.G. Teremetskaya, Elektrokhimiya 2010, 46 , 340. 38. M. Iwaki, S. Sato, K. Takahashi, H. Sakairi, Nucl. Instrum. Meth. Phys. Res. 1982, 209–210 , 1129. 39. B. Miller, R. Kalish, L.C. Feldman, A. Katz, N. Moriya, K. Short, A.E. White, J. Electrochem. Soc. 1994, 141 , L41. 40. C. Reuben, E. Galun, H. Cohen, R. Tenne, R. Kalish, Y. Muraki, K. Hashimoto, A. Fujishima, J.F. Butler, C. Levy-Clement, J. Electroanal. Chem. 1995, 396 , 233. 41. I. Duo, A. Fujishima, C. Comninellis, Electrochim. Commun. 2003, 5 695. 42. E. Mah´e, D. Devilliers, C. Comninellis, Electrochim. Acta 2005, 50 , 2263.

108

ELECTROCHEMISTRY OF DIAMOND

43. V.G. Ralchenko, Yu.V. Pleskov, V.I. Polyakov, A.V. Khomich, Yu.E. Evstefeeva, M.D. Krotova, E.N. Loubnin, I.I. Vlasov, Diam. Rel. Mat . 2003, 12 , 531. 44. Yu.V. Pleskov, Yu.E. Evstefeeva, M.D. Krotova, V.G. Ralchenko, I.I. Vlasov, E.N. Loubnin, A.V. Khomich, J. Appl. Electrochim. 2003, 33 , 909. 45. J.A. Bennett, J. Wang, Y. Show, G.M. Swain, J. Electrochim. Soc., 2004, 151 , E306. 46. A.V. Khomich, V.G. Ralchenko, A.V. Vlasov, R.A. Khmelnitskiy, I.I. Vlasov, V.I. Konov, Diam. Rel. Mat . 2001, 10 , 546. 47. L. Nistor, V. Ralchenko, I. Vlasov, A. Khomich, R. Khmelnitskii, P. Potapov, J. Van Landuyt, Phys. Stat. Sol. A 2001, 186 , 207. 48. Yu.V. Pleskov, M.D. Krotova, V. Ralchenko, A. Khomich, R. Khmelnitskii, Elektrokhimiya, 2005, 41 , 343. 49. Yu.V. Pleskov, M.D. Krotova, V.V. Elkin, V.G. Ralchenko, A.V. Khomich, R.A. Khmelnitskiy, Electrochim. Acta 2005, 50 , 1149. 50. V. Ralchenko, L. Nistor, E. Pleuler, A. Khomich, I. Vlasov, R. Khmelnitskii, Diam. Rel. Mat . 2003, 12 , 1964. 51. Yu.V. Pleskov, M.D. Krotova, V. Ralchenko, A. Khomich, R. Khmelnitskii, Elektrokhimiya 2003, 39 , 886. 52. Yu.V. Pleskov, M.D. Krotova, V.G. Ralchenko, A.V. Khomich, R.A. Khmelnitskiy, Electrochim. Acta 2003, 49 , 41. 53. Yu.V. Pleskov, M.D. Krotova, V.G. Ralchenko, A.V. Khomich, R.A. Khmelnitskiy, Diam. Rel. Mat . 2003, 12 , 1957. 54. V.V. Elkin, M.D. Krotova, Yu.V. Pleskov, Elektrokhimiya 2003, 39 , 1053. 55. Yu.V. Pleskov, M.D. Krotova, V.V. Elkin, V.G. Ralchenko, A.V. Saveliev, S.M. Pimenov, P.-Y. Lim, Electrochim. Acta 2007, 52 , 5470. 56. Yu.V. Pleskov, M.D. Krotova, V.G. Ralchenko, A.V. Saveliev, Elektrokhimiya 2007, 43 , 868. 57. Yu.V. Pleskov, Yu.E. Evstefeeva, A.M. Baranov, Diam. Rel. Mat . 2002, 11 , 1518. 58. Yu.E. Evstefeeva, Yu.V. Pleskov, A.M. Kutsay, I. Bello, Elektrokhimiya 2005, 41 , 866. 59. Yu.V. Pleskov, M.D. Krotova, M.L. Shupegin, A.D. Bozhko, V.G. Ralchenko, Elektrokhimiya 2006, 42 , 1002.

5 Applications of Polycrystalline and Modified Functional Diamond Electrodes Yasuaki Einaga and Akira Fujishima

5.1

INTRODUCTION

Conductive boron-doped diamond (BDD) is an alternative to traditional carbon electrodes that provides superior chemical and dimensional stability, low background currents, and a very wide potential window of water stability. Recently, electrochemical applications using BDD electrodes are attracting much attention in many fields, not only in electrochemistry but also in fields such as functional materials science, analytical chemistry, environmental science, biomedical or biological science, and so on [1–3]. In fact, waste water treatment systems and ozone or fluorine generation systems that use BDD electrodes have already become commercially available, and the number of publications involving BDD electrochemistry research is drastically increasing year by year (see Figure 5.1). In this chapter, several examples of electro-analytical applications using polycrystalline BDD electrodes will be shown. Furthermore, some examples using modified functional BDD electrodes such as self-standing perforated BDD with regular patterns of holes, ion-implanted BDD, BDD microelectrodes, and BDD nanograss arrays (whisker BDD),

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

109

110

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

180

160

140

Publications

120

100

80

60

40

0

1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

20

Year Figure 5.1 Yearly research publications on diamond electrochemistry.

which have achieved ozone generation with high efficiency, and the highly sensitive detection of arsenic, dopamine, and glucose, will be introduced.

5.2

PREPARATION OF BDD ELECTRODES

The BDD electrodes were deposited on Si (100) wafers in a microwave plasma-assisted chemical vapor deposition system (ASTeX Corp.) [4]. The vapor of liquid mixtures of acetone and trimethoxyborane (B(OCH3 )3 ) as the source gases were introduced into the reactor by bubbling with hydrogen gas. The liquid mixtures were prepared with appropriate mixing ratios based on Raoult’s law so that the boron/carbon (B/C) ratios in the reactor were controlled. The typical grain size of the resulting BDD thin films was up to ∼5 μm, with a thickness of ∼20 μm for a deposition time of 7 h using 5 kW of plasma power.

5.4 APPLICATIONS IN ELECTROCHEMICAL ANALYSIS USING POLYCRYSTALLINE BDD ELECTRODES

111

5.3 ELECTROCHEMICAL PROPERTIES OF BDD AS ELECTRODE MATERIALS Boron-douped diamond electrodes have an extremely wide potential window of water stability, low background currents (see Figure 5.2), chemical and mechanical stability, resistance to fouling, lack of a surface oxide film, and controllable surface termination. These characteristics have led to the application of BDD electrodes in electrochemical sensing, in electroanalysis, in electrochemical synthesis, and for the anodic destruction of organic wastes. BDD also can be used as a transparent, conducting medium for analytical chemistry and photoelectrochemical applications.

5.4 APPLICATIONS IN ELECTROCHEMICAL ANALYSIS USING POLYCRYSTALLINE BDD ELECTRODES 5.4.1

Detection of Free Chlorine

Chlorine is a strong oxidizing agent and is the conventional chemical used for the continuous disinfection of drinking water, water in swimming pools, and wastewater [6]. The ability of chlorine to disinfect depends on its concentration. If the concentration is too small, the disinfectant effect is insufficient; however, too much chlorine is wasteful and creates other dangerous side products such as trihalomethanes. The World Health Organization drinking water standard states that 2–3 mg L−1 chlorine gives satisfactory disinfection and residual concentration, where the maximum amount of chlorine allowed is 5 mg L−1 . Therefore, the exact determination and continuous on-line monitoring of residual water disinfectant is a strict requirement. The general detection methods for free chlorine include the colorimetric method, the amperometric titration method, and iodometry. However, these methods are unsuitable

Figure 5.2 Electrochemical properties of various electrodes.

112

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

for continuous online monitoring because they each have a number of disadvantages, such as the requirement for many types of reagents that may produce greater toxicity, high detection limits, difficulty of operation, and so on. Electroanalytical methods require fewer reagents, promote easy handling, and can provide high sensitivity as well as long-term response stability. The reduction of HClO and ClO− has mainly been reported for the electrochemical determination of free chlorine. However, there are several problems that should be considered when using this method, such as interference, due to a similar reduction potential to that of dissolved oxygen, metal deposition from the sample solution at the electrode, and the effect of trace metal ions. On the other hand, there have been few investigations on quantitative determination based on the anodic reaction of free chlorine, which should be superior, because it is not subject to the preceding problems. In the present study, we focused on the quantitative determination and the possibility of continuous online monitoring of free chlorine oxidation using BDD electrodes [5]. Cyclic voltammograms (CVs) for various concentrations of NaClO in a 0.1 M NaClO4 obtained at a scan rate of 100 mV s−1 using an as-deposited (ad) BDD electrode are shown in Figure 5.3. A well-defined irreversible oxidation peak was observed at a potential of ca. 1.4 V (vs. Ag/AgCl). This is the typical advantage of wide potential window, because the observation of oxidation current at such high potential is impossible in the case of other conventional electrodes due to the narrow potential window. The inset in Figure 5.3 shows a plot of the oxidation peak current versus the free chlorine concentration. A linear calibration curve (R 2 = 0.999) obtained in the concentration range of 20–100 mg Cl L−1 , indicating that determination of free chlorine can be performed at ad-BDD electrodes. A slope of 0.744 μA cm−2 mg−1 L shows the sensitivity. This sensitivity was three to four times higher in comparison with those at conventional electrodes, suggesting the superiority of using ad-BDD for chlorine oxidation. A background current of 3 μA cm−2 was obtained, which is very small in comparison with that at a Pt electrode under the same conditions (115 μA cm−2 ). It is known that BDD electrodes have extremely small background currents due to the inert surface. The small background current can give rise to a very low detection limit due to

Current density (µA cm–2)

90 80 70

y = 0.744x – 2.767 R 2 = 0.9981

60 40

50

20 30

0

20

40

60

80

100

CIO– Concentration(mg Cl l–1)

10

–10

0

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Potential (V) vs. Ag/AgCl Figure 5.3 Cyclic voltammograms of a 0.1 M NaClO4 solution in the presence of various concentrations of free chlorine (2–100 mg Cl L−1 ) at as-deposited diamond electrodes. Linear calibration plots are shown in the inset. (Reprinted with permission from Ref. 5.)

5.4 APPLICATIONS IN ELECTROCHEMICAL ANALYSIS USING POLYCRYSTALLINE BDD ELECTRODES

113

the decrease in the noise. Furthermore, excellent stability was also shown for repetitive voltammograms. 5.4.2

Detection of Oxalic Acid

Oxalic acid, which exists naturally in many plants (spinach, ginger, chocolate, etc.), combines with Ca, Fe, Na, Mg, or K to form poorly soluble oxalate salts. High levels in the diet lead to irritation of the digestive system, and particularly of the stomach and kidneys. It is also known to contribute to the formation of kidney stones. The urinary level of oxalic acid has long been recognized as an important indicator for the diagnosis of renal stone formation. Cyclic voltammetry and flow injection analysis with amperometric detection were used to study the electrochemical reaction [7]. In this case also, the oxidation current was observed at a high potential (about 1.35 V vs. Ag/AgCl), which cannot be observed with conventional electrodes. A good linear response was observed for the concentration range from 50 nM to 10 μM, with an estimated detection limit of about 0.5 nM (S/N = 3). 5.4.3

Proteins (Including Cancer Markers)

The detection of proteins, including cancer markers, has been attracting increasing attention. The direct, unmediated electron transfer between proteins and electrodes has been studied for many different combinations of proteins and electrode surfaces in recent years. These studies have received much attention because of the requirement in understanding the fundamental reactions of biomolecules, from the viewpoint of developing a direct detection method of the protein at the electrode, or in order to develop new materials by combinations between the protein and the electrode surface. In general, direct electrochemical oxidation of proteins is based on the electro-oxidation of metal ions or electro-active amino acids in the protein structure [8]. The proteins studied in this manner range from small, water-soluble redox proteins, to large, sometimes multi-redox enzyme-centered proteins. However, it can be said that the cytochromes, the blue-copper proteins, and the iron-sulfur proteins are as familiar to electrochemical analysts as they are to biochemists. In contrast, the electrochemical detection of the larger nonmetalated proteins (e.g., albumin) is reported much less frequently. The limited number of reports is not only due to the complexity of the protein structure but also to the strong adsorption of the proteins on the electrode surface, which can lead to signal depression and result in lack of predictability and reproducibility. On the other hand, other superior properties of BDD hydrogen-terminated diamond films are that they are well faceted, hydrophobic, and have a low surface energy. Therefore, it is expected that even proteins can be directly detected electrochemically using BDD because of the surface inertness. Figure 5.4a (upper curve) shows the CV of 300 mg dl−1 BSA in 0.1 M PBS pH 10 at the BDD electrode. Interestingly, the CV included three peaks. It is known that BSA contains 20 types of amino acids, including cystein, tryptophan, and tyrosine, as well as 17 disulfide bonds that are known to be electro-active at BDD electrodes. In order to confirm the active sites of BSA at the BDD electrode, comparison was made between the CVs of BSA and several amino acids (Figure 5.4a [lower curves]). The oxidation peaks can be observed separately at different potentials at pH 10. Thus, it is suggested

114

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

(a) BSA

Current density

2 mA/cm2

tyrosine

tryptophan cysteine 20 mA/cm2

0.2

0.0

0.4

0.6 0.8 Potential (V) vs Ag/AgCl

1.0

1.2

(b) 40

50

3000

30 20

2100

10

40 Current (nA)

0 0

500

30

1000 1500 2000 2500 3000 Concentration/mg dl

1500 1200

900

20

150

600

300

75 mg/dl 10

0 0

10

20

30

40

50

Time (min) Figure 5.4 (a) Cyclic voltammograms of 300 mg dl−1 BSA (up) and 1 mM amino acids (bottom) in 0.1 M phosphate buffer solution at diamond electrodes. (b) Signal responses of BSA in various concentrations by using flow injection analysis with a diamond electrode as the detector in the concentration range 75–3000 mg dl−1 . The applied potential was 0.8 V versus Ag/AgCl. The mobile phase was a 0.1 M phosphate buffer solution at pH 7.4. The flow rate was 1 mL/min. Inset shows linear dynamic concentration of BSA. (Reprinted with permission from Ref. 9.)

that the oxidation peaks of BSA correspond to the oxidation peaks of cystein, tyrosine, and tryptophan. Further, flow injection analysis (FIA) was used to minimize the adsorbing effects of the protein and to obtain calibration curves (see Figure 5.4b). The peak current shows good linearity in the concentration range of 5–3000 mg dl−1 of BSA, with an experimental detection limit of 5 mg dl−1 , as shown in the linear dynamic calibration

115

5.4 APPLICATIONS IN ELECTROCHEMICAL ANALYSIS USING POLYCRYSTALLINE BDD ELECTRODES

in the inset of Figure 5.4b. The stability of the current response is also shown for 5 injections of each concentration in the concentration range of 5–3000 mg dl−1 . Reproducibility was also confirmed. As an example of cancer-maker detection, the possibility of using H-terminated BDD for the direct electrochemical detection of other proteins was also investigated for immunosuppressive acidic protein (IAP). The amperometric FIA responses of IAP showed the possibility of the direct electrochemical detection. A linear detection range of 200–800 μg ml−1 was observed. Since the normal human serum IAP level is ∼355 μg ml−1 , the detectable range should be suitable for cancer screening [9]. Then, as one of the applications of protein detection, the detection of conformational changes in large, nonmetallo proteins such as bovine serum albumin, using FIA coupled with hydrogen-terminated BDD electrodes, was performed [10]. Briefly, the oxidation current was used as a signal reporter in the monitoring of urea-induced BSA denaturation. In the denatured state at high urea concentrations, the electrochemical signal increased, and the amperometric responses for the oxidation potential at 1300 mV were consistent with the results of conventional methods of denaturation monitoring using fluorescence spectroscopy. The oxidation involved at least five redox-active species (cysteine, tryptophan, tyrosine, methionine, and disulfide bonds). Furthermore, the method also showed high sensitivity for the quantitative analysis of proteins. A linear dynamic range for concentrations 50–400 μg ml−1 (r 2 = 0.977), with a lower detection limit (LOD) of 190 ng ml−1 , was achieved for BSA. Furthermore, in order to realize the selectivity of the protein detection, surface modification can be useful. As one example, a poly-o-ABA-modified BDD was developed for a protein immunosensor (see Figure 5.5) [11]. The amperometric sensing of mouse IgG (MIgG) was selected as the model at the poly-o-ABA-modified BDD to compare to the poly-o-ABA-modified glassy carbon (GC) at the same condition. An anti-mouse IgG from goat (GaMIgG) was covalently immobilized at a poly-o-ABA-modified BDD electrode, which used a sandwich-type alkaline phosphatase (ALP) catalyzing amperometric

AAP ALP ALP

ALP ALP

AA

AARe

O

NH

C

O

C

NH O

NH

C

O

NH

C

AAP

C

NH O

AAOx

NH

C

2e–

BDD

BDD

ALP

ALP conjugated

GaMIgG

GaMIgG

MIgG

Figure 5.5 Schematic of the amperometric enzyme immunosensor based on the poly-o-ABA modified BDD electrode. (Reprinted with permission from Ref. 11.)

116

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

immunoassay with 2-phospho-L-ascorbic acid (AAP) as a substrate. The ALP enzyme conjugated at the immunosensor can generate the electro-active ascorbic acid (AA), which can be determined by amperometric detection. The signal was found to be proportional to the quantity of MIgG. The limit of detection (LOD) of 0.30 ng mL−1 (3SD) and 3.50 ng mL−1 (3SD) for MIgG at BDD and GC electrodes was obtained. It also was found that the dynamic range of three orders of magnitude (1–1000 ng mL−1 ) was obtained at BDD, while at GC, the dynamic range was narrower (10–500 ng mL−1 ). The method was applied to a real mouse serum sample that contained MIgG.

5.5

MODIFIED FUNCTIONAL BDD ELECTRODES

In order to improve the electrochemical properties, several modified types of BDD electrodes were also fabricated. 5.5.1 Production of High-Concentration Ozone-Water Using Free-Standing Perforated Diamond Ozone dissolved in water is one of the most powerful chemicals used for disinfection and sterilization. The effectiveness is based on the generation of reactive oxygen species by reactions between ozone and water molecules as well as the strong oxidative ability of ozone. Furthermore, ozone-water is very environmentally friendly because the residual ozone spontaneously decomposes to oxygen, rendering postwashing unnecessary. Due to these advantages, ozone-water has been employed for hand washing in hospitals and welfare institutions, vegetable washing in food factories and kitchens, and so on. Water electrolysis with a solid polymer electrolyte (SPE) cell, or a zero-gap (ZG) electrolytic cell, is a unique technique suitable for ozone-water production. The cell is divided into two compartments by a proton exchange membrane, to which a porous anode and a porous cathode are firmly attached. The system is user-friendly, in that an electrolyte is unnecessary for electrochemical ozone production (EOP). The anode and cathode compartments are filled with pure water; water electrolysis at the anode evolves oxygen and ozone; protons electrogenerated at the porous anode move thorough the membrane toward the porous cathode; and the hydrogen evolution reaction occurs at the cathode surface. When pure water, or tap water, is continuously supplied to the electrolytic cell, electrolyte-free ozone-water can be continuously produced. EOP systems available in Japan have employed a porous lead dioxide (PbO2 ) plate or a platinum (Pt) mesh as an anode. Although they are effective in producing ozone gas or ozone-water, they have the drawbacks of being low in durability, costly, and environmentally harmful. Conductive diamond has attracted much attention as a promising electrode material for EOP because of the large overpotential for the oxygen evolution reaction (OER) as well as the superior chemical and dimensional stability. Here, direct electrochemical ozone-water production (EOWP) was performed with a ZG cell made of an acrylic frame, which included a porous anode, a porous cathode, and a proton exchange membrane, as shown in Figure 5.6a [12]. As the anode, free-standing BDD electrodes with various patterns of perforated holes were used (summarized in the table above Figure 5.7), a photograph of a representative free-standing perforated BDD electrode, D10HN410, on which 410 holes 1 mm in diameter were perforated, is shown in Figure 5.6b). The perfectly circular holes with sheer edges were formed by means

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

Hydrogen water

(a)

117

Ozone water

Porous cathode

Porous anode

Membrane

Water

Water

(b)

Figure 5.6 (a) Schematic of a ZG electrolytic cell. (b) Photograph of the representative free-standing perforated BDD electrode (D10HN410). (Reprinted with permission from Ref. 12.)

of laser beam machining. A Pt mesh was used as the cathode, and they were firmly pressed onto the membrane (Nafion film). A Pt mesh was also adopted as the anode for comparison purposes. At the porous anode, the oxygen evolution and ozone generation reactions occurred, as follows: Hydrogen ions generated at the anode are reduced at the cathode to produce hydrogenated water. Pure water of about 12◦ C with electrical conductivity of less than 1 (μS cm−1 ) was continuously supplied into the anodic compartment at a flow rate of 2.0 L min−1 (120 L h−1 ). Cathodic water was pure water cycled at a flow

118

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

ID2 D10HN166 D10HN288 D10HN312 D10HN410 D10HN144 D10HN196

Thickness (mm)

Total edge length (cm)b

0.77 0.92 0.3,0.54,0.90 0.54,0.94 0.82 0.90

65.78 103.48 111.01 141.80 80.86 105.56

Diameter Hole number (mm) 1.0 1.0 1.0 1.0 1.5 1.5

166 288 312 410 144 196

Ucell (V)

30

20

10

D10HN196 : D10HN288 :

0 50

D10HN312 : D10HN410 : Pt mesh :

ε03(%)

40 30 20 10 0 10

c02(mg L–1)

8 6 4 2 0

0

2

4

6

8

10

/app (A) Figure 5.7 Plots of (a) ozone concentration, (b) current efficiency, and (c) cell voltage with respect to applied current. Ozone-water was produced by electrolysis of pure water with free-standing perforated BDD electrodes and a Pt mesh electrode. (Reprinted with permission from Ref. 13.)

rate of 0.1 L min−1 . Electrogenerated hydrogen was spontaneously released as gas. Water electrolysis was performed under galvanostatic conditions. A ZG cell newly constructed was pre-electrolyzed at 10 A for 2–5 h to normalize the cell conditions. Then, ozone concentration and cell voltage were recorded at a steady state attained after the application of a constant current to the cell. Ozone-water produced at the anodic compartment drained into a reservoir and was then introduced into an ozone meter by a pump. This prevented gas interfusion, thereby providing more reliable data compared with that reported previously [13]. Ozone-water concentration was measured with a dissolved ozone meter based on the UV absorption method.

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

119

Figure 5.7 shows (a) ozone concentration, (b) current efficiency, and (c) cell voltage for EOWP as a function of applied current. The perforated free-standing BDD electrodes with holes 1 mm in diameter exhibited superior performance for EOWP compared to the Pt mesh electrode. Figure 5.7a demonstrates that, accompanied with an increase in the applied current at the BDD electrodes, ozone-water is gradually concentrated to a maximum of about 9 mg L−1 at 10 A. Although applying higher current was not performed due to limitations of the power source, we believe that more concentrated ozone-water could be available. In contrast, at the Pt mesh electrode, the ozone concentration was much lower, less than 0.5 mg L−1 maximum. The low catalytic activity is attributed to the lower overpotential toward OER and the absence of anions in pure water, which could otherwise adsorb on the Pt surface and inhibit the OER process [13]. The current efficiency reached a maximum at 3–4 A, as shown in Figure 5.7b. This tendency is rather different from that reported in the previous research, in which the current efficiency monotonically dropped depending on applied current [13]. The difference is partly attributed to the employment of a flow system in which electrolyzed ozone-water was directly introduced into the ozone meter. The direct injection was accompanied by small bubbles of ozone and oxygen gases, which caused UV light scattering, resulting in an overestimation of the ozone-water concentration. In the present system, however, ozone-water was drained into a reservoir and then introduced into the ozone meter by a pump, which inhibited gas interfusion. Because excess gaseous ozone was dispersed at the reservoir, the current efficiency was probably underestimated. However, in view of the estimation of the current efficiency for ozone-water production, we believe that the present results are more realistic. In summary, EOWP with free-standing perforated BDD electrodes was proven to be dependent on the number of holes, hole diameter, and electrode thickness. In particular, increasing the number of holes per unit area is the most effective method for improving the current efficiency. The electrode configuration optimal for EOWP was D10HN410, which was 0.54 mm in thickness and offered a current efficiency of 47% under moderate conditions, the highest thus far. The advantage of using free-standing diamond plates is obvious. The entire electrode surface retains an ideal diamond crystalline structure, and thus durability is considerably superior under high current conditions. The application of a diamond electrode formed on a mesh or a porous substrate is another option and has some merits, including low cost and high mechanical strength. However, diamond deposition onto complex substrates is a quite difficult task. Thinner areas of the diamond film tend to contain pinholes, which lead to the erosion of the substrate after long-term usage under high current conditions. Accordingly, we believe that the present system is the most appropriate for EOWP in terms of high current efficiency as well as high durability. 5.5.2 5.5.2.1

Modified Functional BDD Electrodes for Electrochemical Analysis Ion-Implanted BDD Electrodes

Detection of Arsenic Using Ir-implanted BDD Electrodes Many arsenic compounds are known to be toxic. Their direct exposure to especially humans and animals and the side effects to the ecosystem remain international problems. Arsenic exists in many different chemical forms in nature; particularly, in groundwater it is found almost exclusively as arsenite (AsO2 − , As3+ ) and arsenate (HAsO4 2− , As5+ ). Arsenite can

120

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

be converted to arsenate under oxidizing conditions. However, the conversion in either direction is difficult. The reduced species can be found in both oxidized environments and vice versa. Typical targets of arsenic toxicity are the respiratory system, the circulatory system, and the reproductive system. The toxicity of arsenic is greatly dependent on the arsenite level, since arsenite is ∼50 times more toxic than arsenate due to its reactions with enzymes in human metabolism. The higher mobility of arsenite in groundwater also accentuates the potential danger compared to arsenate. In addition, arsenite and arsenate are more toxic than organic arsenic. According to the U.S. Environmental Protection Agency and the World Health Organziation arsenic guidelines, the maximum contaminant level of arsenic in community water systems is 10 μg/L−1 . However, arsenic compounds cannot be detected directly by BDD, because BDD does not have sufficient electrochemical catalytic activity due to the surface inertness. On the other hand, metal electrodes such as iridium have electrochemical catalytic properties, even though the sensitivities are not good due to the large background currents [14]. Modification of BDD electrodes with redox-active particle/compounds offers significant advantages in the design and development of electrochemical sensors. The redox-active sites facilitate electron transfer between the substrate electrode and analytes, with a significant reduction in activation overpotential. A wide variety of particles or compounds have been used as electron transfer mediators via modification of BDD surfaces. Preparation of modified diamond by using chemical precipitation and the electrochemical deposition method have been reported. Although electrochemical deposition is a convenient method to prepare metal deposits on substrates, it is not suitable for metal deposition at BDD, because the nonuniform doping of boron in diamond crystals causes the surface conductivity to be inhomogeneous. Furthermore, the stability of the deposited metal is not good. Since its inception, ion implantation has been considered as the most feasible method to change the electrical properties of a diamond substrate. The method modifies the nearsurface structure of the target due to the heavy ion bombardment. Here, BDD electrodes modified with implanted iridium ions were investigated for arsenic (III) detection by using cyclic voltammetry (CV) and flow injection analysis (FIA) [15]. BDD electrodes were implanted with 800 keV Ir+ by using iridium metal powder as the targets. Cyclic voltammetry and flow injection analysis with amperometric detection were used to study the electrochemical reaction (see Figure 5.8). The electrodes exhibited high catalytic activity toward As (III) oxidation, with the detection limit (S/N = 3), sensitivity, and linearity being 20 nM (1.5 ppb), 93 nA μM−1 cm−2 , and 0.999, respectively. The precision for 10 replicate determinations of 50 μM As (III) was 4.56% relative standard deviation. The advantageous properties of the electrodes were its inherent stability with a very low background current. The electrode was applicable for the analysis of spiked arsenic in tap water containing a significant amount of various elements in ionic form. The results indicate that metal implantation could be a promising method for controlling the electrochemical properties of diamond electrodes. Selective Detection of Glucose Using Cu-implanted BDD Electrodes Recently, highly sensitive and simple glucose detection has been considered to be very desirable for the diagnosis and management of diabetes mellitus. However, glucose is normally undetectable using bare BDD electrodes, similar to the previous example of arsenic detection. Glucose oxidation is a complex process that requires a catalytic reaction using an enzyme or active metal surfaces. Although Au, Pt, Ni, and Cu metal electrodes are known for

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

121

(a)

Current density (mA cm2)

0.6 0.4 0.2 0 –0.2 Phosphate Buffer 1 mM As

–0.4 –0.6 –0.8

–0.6

–0.4

–0.2

0

0.2

0.4

0.6

0.8

1

Potential / V (vs. Ag/AgCl) (b) 1000

1200

800

600

100µM

y = 13.615x + 0.502 R 2 = 0.999

900

50µM

600

300 0

20 40 60 80 Concentration (µM)

100

400

0

10µM

200

Current (nA)

1500

0.1µM

0

2

4

0.5µM

6

1µM

8 Time (min)

5µM

10

12

14

16

Figure 5.8 (a) Cyclic voltammograms of 0.1 M phosphate buffer solution in the absence and presence of 1 mM As(III) at an Ir-implanted BDD electrode. (b) Amperometric response of flow injection analysis using Ir-implanted BDD as a detector. Applied potential was 0.6 V versus Ag/AgCl. Mobile phase was 0.1 M phosphate buffer solution. Flow rate was 1 mL min−1 . Inset shows the variation of current vs. As(III) concentration. (Reprinted with permission from Ref. 15.)

showing electrocatalysis for glucose oxidation, the BDD electrode does not have catalytic properties. Here, we have studied the electrochemical detection of glucose using Cu-modified BDD electrodes (Cu-BDD). We report the simple technique of selective glucose detection using the Cu-BDD. The selectivity was derived from the differences in the diffusion processes for interfering species such as ascorbic acid (AA) and uric acid (UA) (linear diffusion to the BDD surface) and for glucose (spherical diffusion to implanted copper particles). Each dispersed copper particle of the Cu-BDD acts as an ultramicroelectrode (UME), because glucose reacts only at the copper surface but not at the BDD surface. On the other hand, interfering substances react at both electrode surfaces. Eventually, the difference of diffusion leads to a dependence or independence

122

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

of the Faradic current on time, and the steady-state component of the current reflects only glucose concentration. That is, the time dependence of the observed current should be followed by Equations (5.1) and (5.2), respectively. √ Interfering species (linear diffusion) : I = nFACD (1/ π Dt) √ Glucose (spherical diffusion) : I = nFACD (1/ π Dt + 4/π r)

(5.1) (5.2)

Current density (mA cm–2)

Figure 5.9 (curve a) shows a CV recorded for the oxidation of glucose at Cu-BDD. Figure 5.9 (curves b and c) show CVs in the absence of glucose at Cu-BDD and in the presence of glucose at a bare BDD electrode respectively. Because no peak was observed in curves (b) and (c), we conclude that glucose was oxidized at the copper surface but not the BDD surface. Furthermore, the copper particles of Cu-BDD act as UMEs for glucose oxidation in the case of low-density modification with Cu, as evidenced by the fact that the curve of the negative scan was superimposed on the positive scan for the Cu-BDD electrodes, whereas this did not occur for pure metallic Cu electrodes. Electrochemical quantitative analyses of glucose at Cu-BDD electrodes were carried out by chronoamperometry. A linear calibration curve for the intercepts of Cottrell plots (current vs. inverse square root of time) depending on the glucose concentration was obtained (shown later), indicating that Cu-BDD electrodes can be used as a glucose sensor. On the other hand, CV plots for both AA and UA showed oxidation peaks at ca. +0.2 V vs. Ag/AgCl, and Cottrell plots of the chronoamperometric response at +0.6 V versus Ag/AgCl exhibited linear curves passing through the origin. These results indicate that the intercepts of the Cottrell plots reflect the glucose concentration at Cu-BDD, even in the presence of interfering substances. As indicated above, at Cu-BDD, glucose differs from interfering substances in the dimensions of diffusion. Figure 5.10 shows a schematic illustration of these differences. Glucose could only be oxidized at modified copper with spherical diffusion, whereas UA and AA could be oxidized at a BDD surface following a linear diffusion process. The

0.20 (a) 0.15 0.10 (b) 0.05 (c) 0.00 0.0

0.2

0.4

0.6

0.8

Potential (V) vs. Ag/AgCl Figure 5.9 Cyclic voltammograms of 3 mM glucose in 0.2 M NaOH at the (c) diamond electrode, (a) Cu-implanted diamond electrode, and (b) Cu-implanted diamond electrode in the absence of glucose. (Reprinted with permission from Ref. 16a.) See color insert.

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

123

(a)

(b)

Figure 5.10 Illustration of the diffusion profiles at Cu-implanted diamond electrodes for (a) ascorbic acid and uric acid and (b) glucose. (Reprinted with permission from Ref. 16a.)

Cottrell plots of 6 mM glucose solutions and a mixed solution containing 6 mM glucose and interfering species (0.5 mM AA and 0.5 mM UA) showed similar values for both intercepts. Electrochemical quantitative analyses of glucose at Cu-BDD electrodes were carried out by chronoamperometry at 0.6 V vs. Ag/AgCl. Figure 5.11 shows chronoamperograms of 0.2 M NaOH aqueous solutions containing 1–5 mM (n = 5) glucose, respectively.

(a)

(b)

Figure 5.11 Chronoamperograms of 1–5 mM (a–e) glucose in 0.2 M NaOH at Cu-implanted diamond electrodes, respectively. The insets show (A) related Cottrell plots and (B) dependence of current on glucose concentration, extracted from the intercepts of the Cottrell plots. (Reprinted with permission from Ref. 16a.)

124

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

Cottrell plots in inset A of Figure 5.11 shows that the intercepts were not zero and the slopes were small, indicating that the amperometric response to glucose was at a steady state. A linear calibration curve for the concentration dependence of the intercept value was obtained (Figure 5.11, inset B), indicating that Cu-BDD electrodes can be used as a glucose sensor by this method. Furthermore, Figure 5.12 shows the Cottrell plots of 6 mM glucose solution (Figure 12a) and a mixed solution containing 6 mM glucose and interfering species (0.5 mM AA and 0.5 mM UA) (Figure 5.12b). According to the previous discussions, similar values for both intercepts reflect the concentration of glucose (6 mM) only. These results suggest that the simple methodology using Cu-BDD could be promising for selective glucose sensors [16]. 5.5.2.2 Selective Detection of As(III) and As(V) by Stripping Voltammetry The electrochemical detection of mixed solutions of As3+ and As5+ has been investigated by stripping voltammetry at gold-modified diamond electrodes [17]. The method was implemented based on the oxidative stripping of As0 deposited at the electrode surface. Whereas As3+ can be deposited by simple electrochemical reduction of As3+ to As0 at −0.4 V (vs. Ag/AgCl), a much more negative potential is required to overcome the activation energy of As5+ reduction. However, in such a negative potential region, hydrogen evolution also occurs. Consequently, one more step should be added to release the hydrogen gas adsorbed at the electrode surface during the reduction step. During the deposition of As5+ , the As3+ species was also simultaneously deposited. Therefore, to differentiate As3+ and As5+ quantitatively in a mixed solution, both stripping voltammetry methods should be performed and compared mathematically. A comparison of stripping voltammograms for both methods for As3+ solution in the absence of As5+ demonstrated similar peak shapes and current intensities, confirming that errors in the calculation of As5+ concentration in the mixed solution with As3+ can be avoided. Good linear responses were observed for each standard solution of As3+ and As5+ . A linear calibration curve could also be achieved for a series of concentrations of 100–1000 ppb As5+ in mixed solutions with 100 ppb As3+ (r 2 = 0.99) and for a series of concentrations of

Figure 5.12 Chronoamperograms of (a) 6 mM glucose and (b) a mixture of 0.5 mM ascorbic acid, 0.5 mM uric acid, and 6 mM glucose in 0.5 M NaOH with Cu-implanted diamond electrodes. The inset shows their related Cottrell plots. (Reprinted with permission from Ref. 16a.)

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

125

5–30 ppb As3+ in mixed solutions with 100 ppb As5+ . Detection limits of 5 and 100 ppb can be achieved for As3+ and As5+ in mixed solution, respectively. Good reproducibility was shown for stripping voltammetry of As3+ and As5+ with RSD values (n = 8) of 7.5 and 8.4%, respectively. Good stability of gold-modified diamond electrodes before and after arsenic detection was also evaluated by SEM images. Application of the method for real sample analysis was performed for arsenic detection in Yokohama (Japan) tap water. 5.5.2.3 In vivo Dopamine Detection by BDD Microelectrodes The borondoped diamond (BDD) microelectrode has received much attention as an electrochemical sensor, especially as an in vivo sensor [18–20]. By using this promising electrode, in vivo electrochemical detection of dopamine (DA) was investigated. Dopamine is one of the neurotransmitters, and thought to act as a trigger of many vital activities. Therefore, in vivo DA monitoring is important. However, in the case of the carbon fiber, which is a conventional in vivo sensor, it is a long-standing problem to separate the response from that of some interfering substances, particularly ascorbic acid (AA). To overcome this issue, we used the BDD microelectrode. Anodically oxidized BDD (ao-BDD) has the property to recognize DA and AA separately by investigating their specificity of oxidation potential [21]. So, we applied this useful property for in vivo analysis with microelectrodes. Finally, a good separation between DA and AA could be obtained not only in vitro but also in vivo. Moreover, some advantages over the conventional carbon fiber electrode were demonstrated. Therefore, we concluded that BDD microelectrodes are highly promising for future in vivo analysis [22]. BDD thin films were grown on chemically etched tungsten wires (50 μm dia.) by using a microwave plasma-assisted chemical vapor deposition system at a hydrogen pressure of 60 Torr and microwave power of 2.5 kW for 3 h. Boron was doped with a concentration of 104 ppm (B/C). A single compartment cell was employed for in vitro electrochemical experiments. Ao-BDD was formed by the electrochemical oxidation of as-deposited, hydrogen terminated, BDD (ad-BDD) at an applied potential of +2.5V (vs. Ag/AgCl) in 0.1 M HCl for 20 min. In vivo measurements were conducted in a mouse brain. Electric stimulations for the medial forebrain bundle (MFB) were applied to release DA. Square wave pulse voltammetry (SWPV) was used for electrochemical characterization. All experiments were conducted by using an Ag/AgCl electrode as reference and a Pt wire as counter electrode. Polycrystalline BDD on tungsten wire with the tip diameter of about 5 μm was obtained. This was the smallest among published reports of BDD microelectrodes, and its small size was expected to be workable for in vivo studies without any invasive damage for cells or tissues (see Figure 5.13 shows a schematic drawing of the diamond microelectrode used for in vivo analysis). At first, the basic electrochemical properties of BDD microelectrodes were studied in vitro. It was already reported that, at an ao-BDD electrode, the oxidation potential of AA was shifted to a higher potential than that of DA, whereas at ad-BDD, AA was oxidized at almost the same potential as that of DA (i.e., about +0.6V vs. Ag/AgCl). This is because AA is charged negatively in the aqueous solution, so that electrostatic repulsion is exhibited between the AA and the surface of the ao-BDD macroelectrode, which is terminated with oxygen. Therefore, for the oxidation of AA, much excess potential is needed to approach to the electrode. This behavior was also observed at the

126

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

about 1cm BDD

resin about 5cm

pre-pulled glass capillary silver epoxy

coated metalwire Figure 5.13 Ref. 22.)

Schematic drawing of a diamond microelectrode. (Reprinted with permission from

ao-BDD microelectrode. Two steps of steady-state current responses appeared for the measurement for the mixture of DA and AA. Moreover, high sensitivity was obtained, with an experimental detection limit of 50 nM in the presence of AA. This value was adequate for the application for in vivo monitoring of DA. Based on these observations, an ao-BDD microelectrode was utilized for the in vivo measurements (see Figure 5.14). The BDD electrode was inserted into the striatum in the mouse brain, and MFB stimulations were applied to induce DA release. Figure 5.15 shows signal responses following the stimulations. Clear signals were recorded for each stimulation and indicated that in vivo monitoring of DA was successfully achieved at the ao-BDD microelectrode. Moreover, a correlation diagram for in vivo analysis showed very similar features to that for in vitro DA. These results confirmed that these signals certainly came from DA oxidation. 5.5.2.4 BDD Nanograss Array (Whisker BDD) In order to expand the electrochemical application field of BDD electrodes, a BDD nanograss array was created [23]. A simple method was applied to prepare the BDD nanograss array on a heavily doped BDD film by reactive ion etching. The boron dopant atoms in the diamond act as the mask during plasma etching, thus avoiding the complicated pre-preparation processes involved in using an intentional mask or removal of the template by using additional processes. The diameter of the prepared nanograss gradually increased from 5 to 20 nm with increasing etching time from 30 s to 10 min, while the morphology of the nanograss was disturbed by use of an etching time of 60 min. Figure 5.16 shows SEM images of the microstructure of a BDD nanograss array. It can be seen that the nanograss array on

127

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

counter electrode

stimulating electrode

BDD

Reference electrode

anesthesia induction

Counter electrode

Stimulating electrode reference electrode

BDD

Brain Cortex Corpus Striatum Substantia Nigra Dopaminergic Neuron

Figure 5.14 Photograph and schematic drawing of in vivo mouse brain experimental setup. (Reprinted with permission from Ref. 22.) See color insert.

Current (nA)

5.4 5.2

0.20 0.18 0.16 0.14 0.12

36

0.8

5.0

0.9 1.0 Time (s)

1.1

4.8

Current (nA)

(a)

Current (nA)

5.6

(b)

34 32

4.6 4.4 200

400

600 Time (s)

800

1000

30 200

400

600 Time (s)

800

1000

Figure 5.15 Differential pulse voltammetry (DPV) monitoring of current response following MFB stimulation (50 Hz, 100 pulses for 2 s) measured at an applied potential of 0.9 V (vs. Ag/AgCl) with (a) a diamond microelectrode and (b) a carbon fiber electrode. The inset shows the dependence of the signal current on the applied potential. DPV settings: frequency, 50 Hz; potential step, 100 mV; pulse amplitude, 150 mV; starting potential, 0.65 V versus Ag/AgCl. (Reprinted with permission from Ref. 22.)

the BDD surface has the following dimensions: ∼20 nm diameter and ∼200 nm length, and the distance between each nanograss structure is about 50 nm. The electron transfer on the electrode surface was investigated by electrochemical impedance spectroscopy (EIS). Figure 5.17 shows the results of EIS for 10 mM [Fe(CN)6 ]3-/4- in 0.1 M KCl on the electrode surface. The lower semicircle diameter obtained on the BDD nanograss array electrode (b), compared to that on the oxidized BDD film electrode (a), indicated that the presence of the nanograss array on the electrode

128

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

Figure 5.16 SEM images of a BDD nanograss array at (A) low resolution and (B) high resolution. (Reprinted with permission from Ref. 23.)

500 (b) oxidized BDD film BDD nanograss array

–Z ’’(ohm)

400

300

200 (a) 100

0 200

400

600 Z ’’(ohm)

800

Figure 5.17 Nyquist plots of 10 mM [Fe(CN)6 ]3−/4− in 0.1 M KCl from 105 Hz to 0.1 Hz at an ac amplitude of 10 mV under open circuit potential conditions, obtained on (a) the oxidized BDD film and (b) the BDD nanograss array. (Reprinted with permission from Ref. 23.)

surface improved the reactive site, reduced the interfacial resistance, and made the electron transfer easier. The effect of the nanograss array structure on the enhancement of electrocatalytic activity of electrodes was demonstrated by detecting dopamine (DA) and uric acid (UA). It is essential to develop simple and rapid methods for their determination in routine analyses. Here, the electrochemical detection of DA and UA was studied on the BDD nanograss array electrode and the oxidized BDD film electrode. In the CVs of 5 × 104 M DA (Figure 5.18a), the oxidation peak potential shifted from 0.585 V on the oxidized BDD film electrode to 0.395 V on the BDD nanograss array electrode, and the oxidation peak current increased from 16.35 to 22.44 mA. That is, on the BDD nanograss array electrode, the oxidation peak potential shifted negatively by ∼0.19 V

5.5 MODIFIED FUNCTIONAL BDD ELECTRODES

129

(a) 5x10–4M DA

Current (µA)

20

15

10

5 oxidized BDD film BDD nanograss array

0

–0.2

0.0

0.2 0.4 Potential (V) vs. Ag/AgCl

0.6

0.8

(b) 16 5x10–4M UA

Current(µA)

12

8

4

oxidised BDD film BDD nanograss array

0

0.0

0.3

0.6

0.9

1.2

Potential (V) vs. Ag/AgCl (c) Current (µA)

6.0

Current (µA)

4.5

6

xxx xxx xxx

4

120µM (b) 100µM

2 0 0

3.0

80µM

(a)

40 80 120 Concentration/µM 60µM 40µM

(c) 1.5

20µM 10µM

0.0

(a) oxidised BDD film (b) BDD nanograss array

blank 0

50

100

150

200

250

300

350

400

Time (s)

Figure 5.18 CVs for (A) 5 × 10−4 MDA, and (B) 5 × 10−4 MUA obtained on the oxidized BDD film (solid line) and the BDD nanograss array (dashed line) electrodes in 0.07 M phosphate buffer solution (PBS) (pH 7.0). The scan rate was 50 mV s−1 . (C) Amperometric response of the oxidized BDD film electrode (a) and the BDD nanograss array electrode (b) through the successive addition of DA in 0.07 M PBS (pH 7.0) under continuously stirred conditions. The applied potential was kept at 0.58 V (vs. Ag/AgCl) for (a) and 0.38 V (vs. Ag/AgCl) for (b), respectively. (Reprinted with permission from Ref. 23.)

130

APPLICATIONS OF POLYCRYSTALLINE AND MODIFIED FUNCTIONAL DIAMOND ELECTRODES

and the oxidation peak current increased by 37.25%, compared to those on the oxidized BDD film electrode. In the case of the oxidation of UA (Figure 5.18b), the oxidation peak current was 11.17 mA at 0.832 V on the oxidized BDD film electrode whereas on the BDD nanograss array electrode, the oxidation peak current of UA at 0.76 V increased by 33.36%. In comparison with the oxidized BDD film electrode, on the BDD nanograss array electrode, the negative shift of peak potential and the increase in peak current on the oxidation of DA and UA were ascribed to the following: The structure of nanograss array could increase the surface area, provide better electric linkage between electrode active sites, promote the electrocatalytic ability, and accelerate the electron transfer. The amperometric response of DA was also investigated through the successive addition of DA into 0.07 M PBS (pH 7.0) under continuously stirring conditions on the BDD nanograss array electrode and the oxidized BDD film electrode, as shown in Figure 5.18c. It was observed that the current response increased with increasing concentration of DA for the two electrodes. The current response of the BDD nanograss array electrode was higher than that of the oxidized BDD film electrode for the same concentration of DA. The inset of Figure 5.18c shows the dependence of the steady-state current on the concentration of DA for the two electrodes. Linearity was observed in the range of 5–120 mM for the two electrodes (r = 0.999). Sensitivity corresponding to the linear range for DA was 465.57 mAM−1 cm−2 on the oxidized BDD film electrode and 642.14 mAM−1 cm−2 on the nanograss array BDD electrode. The detection limits were calculated to be about 1.5 and 0.8 mM, respectively, according to the 3 sb /m criteria. These results also demonstrated the better electrocatalytic activity of the BDD nanograss array toward the detection of DA, with higher sensitivity and a lower detection limit.

5.6

CONCLUSIONS

We have reported not only the preceding examples but also several additional examples of sensitive detection by BDD electrodes, such as uric acid [24], NADH [25], acids [26], toxic gases [27], and others. In fact, some of the examples are now in progress for the development of practical electrochemical sensor applications. Furthermore, in order to add other functions and improve the electrochemical properties, BDD electrodes with surfaces chemically modified by functional organic molecules are also actively being developed [28,29]. Thus, it appears likely that, in the near future, electrochemical sensors using BDD can be applied for environmental analysis and medical analysis.

5.7

ACKNOWLEDGMENTS

The authors would like to express thanks to Prof. Donald A. Tryk for carefully reading the manuscript. We also express thanks to all of co-workers, especially Dr. C. Terashima (Central Japan Railway Company), Dr. T. A. Ivandini, Dr. M. Murata, Dr. M. Chiku, Dr. T. Watanabe, Ms. A. Suzuki (Keio University), Prof. K. Yoshimi, and Prof. S. Kitazawa (Juntendo University).

REFERENCES

131

REFERENCES 1. A. Fujishima, Y. Einaga, T N. Rao, D.A. Tryk, eds., Diamond Electrochemistry, BKC Inc. and Elsevier, 2005. 2. A. Kraft, Int. J. Electrochem. Sci . 2007, 2 , 355. 3. J.H. Luong, K.B. Male, J.D. Glennon, Analyst 2009, 134 , 1965. 4. T. Yano, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1998, 145 , 1870. 5. M. Murata, T.A. Ivandini, M. Shibata, S. Nomura, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2008, 612 , 29. 6. F. Kodera, M. Umeda, Y. Yamada, Anal. Chim. Acta. 2005, 537 , 293. 7. T.A. Ivandini, T.N. Rao, A. Fujishima, Y. Einaga, Anal. Chem. 2006, 78 , 3467. 8. F.A. Armstrong, Encyclop. Electrochem. 2002, 9 , 11. 9. M. Chiku, T.A. Ivandini, A. Kamiya, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2008, 612 , 201. 10. M. Chiku, J. Nakamura, A. Fujishima, Y. Einaga, Anal. Chem. 2008, 80 , 5783. 11. A. Preechaworapun, T.A. Ivandini, A. Suzuki, A. Fujishima, O. Chailapakul, Y. Einaga, Anal. Chem. 2008, 80 , 2077. 12. K. Arihara, C. Terashima, A. Fujishima, J. Electrochem. Soc. 2007, 154 , E71. 13. K. Arihara, C. Terashima, A. Fujishima, Electrochem. Solid-State Lett . 2006, 9 , D17. 14. A. Salimi, M.E. Hyde, C.E. Banks, R.G. Compton, Analyst 2004, 29 , 9. 15. T.A. Ivandini, R. Sato, Y. Makide, A. Fujishima, Y. Einaga, Anal. Chem. 2006, 78 , 6291. 16. (a) T. Watanabe, T.A. Ivandini, Y. Makide, A. Fujishima, Y. Einaga, Anal. Chem. 2006, 78 , 7857. (b) T. Watanabe, Y. Einaga, Biosensors and Bioelectronics 2009, 24 , 2684. 17. D. Yamada, T.A. Ivandini, M. Komatsu, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2008, 615 , 145. 18. B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 1469. 19. A.M. Bond, Analyst 1994, 119 , R1. 20. P.S. Cahill, Q.D. Walker, J.M. Finnegan, G.E. Mickelson, E.R. Travis, R.M. Wightman, Anal. Chem. 1996, 68 , 3180. 21. E. Popa, H. Notsu, T. Miwa, D.A. Tryk, A. Fujishima, Electrochem. Solid State Chem. 1999, 2 , 49. 22. A. Suzuki, T.A. Ivandini, K. Yoshimi, A. Fujishima, G. Oyama, T. Nakazato, N. Hattori, S. Kitazawa, Y. Einaga, Anal. Chem. 2007, 79 , 8608. 23. M. Wei, C. Terashima, M. Lv, A. Fujishima, Z.-Z. Gu, Chem. Commun. 2009, 3624. 24. E. Popa, Y. Kubota, D.A. Tryk, A. Fujishima, Anal. Chem. 2000, 72 , 1724. 25. T.N. Rao, I. Yagi, T. Miwa, D.A. Tryk, A. Fujishima, Anal. Chem. 1999, 71 , 2506. 26. N. Mitani, Y. Einaga, J. Electroanal. Chem. 2009, 626 , 156. 27. T.A. Ivandini, D. Yamada, T. Watanabe, H. Matsuura, N. Nakano, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2010, 645 , 58. 28. T. Kondo, H. Hoshi, K. Honda, Y. Einaga, A. Fujishima, T. Kawai, J. Phys. Chem. C . 2008, 112 , 11887. 29. T. Kondo, S. Aoshima, K. Hirata, K. Honda, Y. Einaga, A. Fujishima, T. Kawai, Langmuir 2008, 24 , 7545.

6 Diamond Ultramicroelectrodes and Nanostructured Electrodes Katherine B. Holt

6.1

INTRODUCTION

Miniaturization of boron-doped diamond (BDD) electrodes is desirable for incorporation into devices and for their effective use in vivo. However, the ability to carry out electrochemical analysis within a confined volume is not the only advantage to using small electrodes. At ultramicroelectrode (UME) dimensions, diffusion of reactant to the electrode surface is very fast, resulting in steady-state (time-independent) limiting currents. This is advantageous for applications requiring continuous monitoring of an analyte concentration by amperometry or for use in imaging applications requiring fast response times and spatial resolution, such as scanning electrochemical microscopy (SECM). UMEs can also be used in highly resistive media—for example with low levels of added electrolyte, or in nonpolar solvents, as they draw small Faradaic currents that result in very minor iR drops in solution. Due to the small electrode area, very low- charging (capacitive) currents result, meaning that higher scan rates can be employed than for conventionalsized electrodes (as charging current increases proportionally with scan rate). This allows short-lived reaction intermediates to be detected and the kinetics and mechanisms of fast time scale electrochemical events to be investigated using fast scan cyclic voltammetry (CV).

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

133

134

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

When coupled with the highly desirable electrochemical characteristics of BDD—a wide potential window, low background currents, robustness, and the ability to resist fouling in biological/organic media—it is clear that the development of diamond UMEs should result in very impressive and versatile sensors. Indeed, very small < 100 μm dimensioned BDD electrodes have successfully been incorporated into capillary electrophoresis systems for the end-column electrochemical detection of dopamine, catechol, ascorbic acid, naphthol, norepinephrine, and epinephrine [1–2]. The analytical performance of BDD microelectrodes was impressive when compared to carbon fiber microelectrodes, achieving lower detection limits and stable, reproducible responses with no evidence of electrode deactivation. BDD microelectrodes have also been used particularly successfully in vivo for the detection of dopamine in mouse brain [3] and for in vitro neurochemistry and neurodynamic studies [4–7]. In many in vivo applications, the exact geometry of the electrode is not important, as long as the performance of the electrode is stable and reproducible. Likewise, although a small size and sharp tip is desirable for use within tissues to minimize damage, the exact exposed electrode area is less important and steady-state electrochemical behavior is not essential. In contrast, in order to be suitable for use in quantitative electrochemical investigations, for example, fast-scan CV for the measurement of heterogeneous and homogeneous rate constants, or in SECM investigations, development of diamond UMEs with a well-defined geometry, is desirable. Importantly, in order to obtain steady-state measurements, we require that the electrode dimensions are of similar magnitude to the diffusion layer thickness. In practice this means that electrodes with at least one dimension of less than 25 μm are required. In this chapter, the electrochemical characteristics of UMEs are briefly summarized (see section 6.2), along with the progress made in fabrication and application of single BDD UMEs of different geometry and sizes < 25 μm (see section 6.3). Most progress has been made with the fabrication and characterization of UME arrays of BDD, where multiple electrodes are embedded in an insulating matrix. The availability of microfabrication techniques, such as photolithography, allows arrays of different geometries to be fabricated with high precision, as discussed in section 6.4. These arrays are highly sensitive, with excellent signal-to-noise characteristics that allow very low detection limits. Some remaining challenges will also be discussed, including issues of inhomogeneity of dopant distribution and crystallite size, and morphology and how this affects the electrochemical response. Finally, the fabrication and application of nanostructured BDD array electrodes will be discussed. This is a very new field where nanoscale structure is introduced onto the surface of BDD electrodes resulting in “nanograss,” “nanowire,” or “nanorod forest” morphologies, with posited applications including selective sensing of dopamine and DNA hybridization.

6.2 ULTRAMICROELECTRODES: DEFINITION AND ELECTROCHEMICAL CHARACTERISTICS An ultramicroelectrode is defined as an electrode with at least one dimension smaller than the diffusion layer developed during a typical electrochemical experiment—in practice this requires one “critical” dimension of 25 μm or less [8]. When a potential is applied to an UME, such that species O in solution is reduced to R, a constant steady-state current is achieved very quickly. This is because diffusion of reactant O to the electrode

6.2 ULTRAMICROELECTRODES: DEFINITION AND ELECTROCHEMICAL CHARACTERISTICS

135

(a)

Macroelectrode linear diffusion (b) Disk UME

(c)

0.5

R i (nA)

insulation

Microelectrode radial diffusion

0.3

0.1

–0.1 0

RG = R / r

0.1

0.2

0.3

0.4

E (V) vs. Ag / Agcl (i)

(ii)

(d) Insulating surround Electrically connected UMEs

Figure 6.1 (a) Schematic of diffusion processes at macroelectrode (linear diffusion predominates) and microelectrode (radial/hemispherical diffusion predominates). (b) Schematic of disk UME with critical dimension, r, and an approximate RG of 3. (c) Typical CV of BDD disk UME of radius 1.5 μm in 1 mM FcOH in 0.2 M KCl, scan rate 20 mV s−1 . (d) Schematic of diffusion profiles formed around UMEs in an array: (i) UMEs widely spaced so that diffusion profiles do not overlap and diffusion remains hemispherical and (ii) for closely spaced UMEs the diffusion profiles overlap and diffusion becomes linear.

is radial rather than linear, as shown in Figure 6.1a, and a thick (relative to the size of the electrode) diffusion field is rapidly established. Mass transport to the electrode is therefore very fast and reactant reaches the surface at a constant flux, and time-independent steady state currents are achieved. Ultramicroelectrodes may be fabricated with a range of geometries—for example disk, hemispherical, spherical, cylinder, cone, and microband—as long as one dimension is < 25 μm. The most common UMEs are of disk geometry (e.g., formed by sealing a metal wire in glass) which gives a steady-state current, i , described by Equation (6.1): i =4 nFDCr

(6.1)

where n is the stoichiometric number of electrons involved in the electrode reaction, F is Faraday’s constant (96485 C mol−1 ), D is the diffusion coefficient of the solution species, C is the bulk concentration of species, and r is the radius of the electrode. Therefore for a 25 μm diameter disk electrode, a typical aqueous redox couple of concentration 1 mM should result in a steady-state current of ∼3–4 nA. The thickness of the insulation surrounding the electrode can also influence the currents achieved; for a thin insulating layer, more diffusion of species around the edges of the electrode is possible,

136

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

leading to larger currents. The ratio of the radius of the whole electrode assembly, including insulation (R) to the radius of the electroactive electrode (r), is termed the RG (illustrated in Figure 6.1b) and the steady state current is found to be dependent on this value for RG of 10 and below. This becomes important for very sharp electrodes with thin insulation, such as those fabricated for scanning probe applications such as SECM, where close approach to a surface are required and a small RG is advantageous [9]. The effect of RG on limiting current may be calculated by changing the suffix “4” in Equation (6.1) (which is valid for RG ≥ 10) to the value most appropriate for the RG in question, which may be found in the literature [10]. Similar equations allow one to calculate the steady-state current for other electrode geometries [8,9]. The cyclic voltammetric (CV) response of an UME is shown in Figure 6.1c, where, in comparison to the conventional “duck-shaped” CV of a macroelectrode, a sigmoidal response is obtained with limiting currents observed rather than peaks. The backward scan follows the forward scan exactly, as products formed at the electrode rapidly diffuse away from the interface and are therefore not detected on the reverse scan. As the CV is mass transport limited, its shape and limiting currents will be independent of scan rate over most routinely used values (< 0.5 V s−1 ). Some UMEs approaching the upper size limit may exhibit quasi-steady-state behavior, where a limiting current is observed at slow scan rates but a peaked CV response is obtained as scan rates are increased. This is particularly the case for electrode geometries where some of the dimensions (other than the critical dimension) are greater than 25 μm, such as cylinder or microband electrodes. Ultramicroelectrode arrays consist of many individual but electrically connected electrodes of UME dimensions distributed within an insulating matrix. Each electrode has the electrochemical behavior of an UME if individually addressed, but the response of the array depends on the spacing between them, as shown in Figure 6.1d. If the electrodes are spaced far enough apart, then their diffusion profiles will not overlap and the CV response will be similar to the sigmoidal steady-state response shown in Figure 6.1c. However, if the electrodes are too close together, the diffusion profiles will overlap and merge, resulting in linear diffusion and a peaked CV response consistent with a planar macroelectrode of the same geometric area. An advantage of a well-spaced UME array is that the steady-state response of a single UME is obtained but with a much larger limiting current, which is proportional to the number of individual electrodes making up the array. An array therefore has the advantages of a single UME but with a much enhanced current response and thus greater sensitivity. 6.3

BORON-DOPED DIAMOND UMEs

The most common strategy for the preparation of single BDD UMEs is to coat a suitable substrate, of the correct geometry and dimensions, with a BDD film using conventional CVD techniques. The entire assembly is then sheathed with an insulating coating, except for a very small electroactive area at the very tip, which is exposed to form the UME. The conditions and methods employed for each of these steps has an influence on the final size and geometry of the resulting UMEs; they are discussed in detail in the sections. 6.3.1

Substrate Preparation and Growth of Diamond Films

The most common substrates for the deposition of single BDD UMEs are sharpened tungsten [3,11–15] or platinum [1,16] wires. Sharpening of 50–250 μm tungsten wires

6.3 BORON-DOPED DIAMOND UMEs

137

is carried out electrochemically in a 2 M NaOH solution with an applied DC voltage. The sharpness of the W tips obtained using this method is found to be independent of initial wire diameter, instead depending on the depth of immersion into the NaOH solution and the threshold voltage at which the power supply is switched off [14]. As the highest etching rate is obtained at the solution/air interface, the lower, immersed part of the wire tends to fracture away when the force of gravity exceeds the yield strength of the etched neck of wire. When there is a greater depth of wire in the solution, the part of wire below the necking point at the solution/air interface is heavier and so it will fracture sooner, leading to larger cross-section tips. The threshold voltage is important because if it is too high, the power supply will be cut off before the lower part necks off; however, if it is too low, etching will continue after necking off, leading to a much blunter tip. It was found that for a 250 μm W wire at 2 mm immersion depth, 2–3 V initial potential and 0.75 V threshold voltages, the sharpest tips of 100 nm diameters can be obtained. In some situations, less sharp tips may be required (e.g., for increased robustness) in which case the threshold voltage and immersion depth may be varied appropriately. Platinum wires (of 10, 25, or 76 μm) can also be sharpened electrochemically using an etching solution of CaCl2 in water and acetone and an applied AC voltage [1]. Both Pt and W wires make excellent substrates for diamond growth due to the ease of their etching and their ability to withstand the high temperatures for CVD deposition. An advantage of using platinum is that it has highly characteristic voltammetric features, which would be discernible if cracks were to form in the diamond film, and hence can be used as an indicator of loss of film integrity. If pinholes form in films deposited on tungsten, leakage of solution through to the underlying tungsten can lead to the formation of an oxide layer and loss of electrical connectivity between the BDD films and the tungsten substrate [14]. Once sharpened, the metal wires must be seeded by sonicating in a suspension of diamond powder, which embeds diamond nanoparticles into the wire and also roughens the surface to provide nucleation sites for growth. Without this nucleation step, very little diamond growth will occur on the smooth etched tips even after 12 h or more of growth. Diamond growth can be carried out under the same conditions as for regular BDD electrodes, by either HF or MW CVD, with the wires typically aligned vertically with the sharpened tips upward, taking some care that the wires are secured mounted to prevent their disturbance during pumping down. As the BDD film coats the sharpened wires, it is inevitable that the diameter of the tip will increase as a function of diamond film thickness and hence growth time. Figure 6.2a shows a typical SEM image of a BDD-coated W wire after 10 h of HFCVD growth, where microcrystalline diamond is found to coat the entire wire. The thickness of the coating is about 5–10 μm, as determined from a cross-section image of a broken tip, as shown in Figure 6.2b. The size of the individual crystallites is dependent on the methane concentrations used in the feed gas mixture during growth. Low methane concentrations (0.3–0.4% CH4 in H2 ) result in larger grains, with grain size increasing toward the sharpened tip. For HFCVD growth, it is not uncommon for single crystals to form at the very tip, as diamond grains tend to coalesce in regions that are higher in temperature. As microcrystalline films are very rough and so more difficult to insulate effectively (see Section 6.3.2), the effect of growth conditions on crystal size and the smoothness of coating has been explored [13,14]. Increasing the methane concentration to 3% increases nucleation and growth rates and results in the deposition of nanocrystalline BDD films, which are much smoother, as shown in Figure 6.2c. However, films grown with such high

138

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

(a)

(b)

Figure 6.2 (a) SEM image of microcrystalline BDD coating a sharpened W wire. (b) SEM image of cross-section through BDD-coated W wire. (c) SEM image of smooth nanocrystalline BDD coating on W wire. (d) SEM image of BDD-coated W wire insulated with electrophoretic paint, with 3 μm diameter disk of BDD exposed at the tip using focused ion beam (FIB) in the ‘‘head-on’’ configuration. (e) SEM image from side of BDD disk UME formed from BDD tip coated in electrophoretic paint and then exposed by FIB operating in ‘‘side-on’’ geometry. [Figures a, b and c: (Reprinted with permission from Ref. 13. Figures d and e: Reprinted with permission from Ref. 15.)]

6.3 BORON-DOPED DIAMOND UMEs

139

(c)

(d)

Figure 6.2

(Continued)

methane content may contain poorer quality diamond and the electrochemical response may be compromised as a result. An alternative method to producing smooth films, even at lower methane concentrations (0.72%), is to apply a negative bias between the hot filament and the W wire substrates throughout the growth procedure. Negative bias-enhanced nucleation (BEN) increases nucleation density and produces continuous smooth film coatings with small grain size (∼1 μm) in less than half the growth time

140

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

(e)

2 μm

Figure 6.2 (Continued)

required for nonbiased growth. As low methane concentrations are used, the films are of higher quality (containing less nondiamond carbon) than the nanocrystalline films grown under methane rich conditions. The growth conditions and resulting film morphology are important, as microcrystalline films typically give superior electrochemical response, but generally provide a poorer coverage and greater incidence of pinholes, as well as being more difficult to insulate. Nanocrystalline films are smoother, are easier to insulate, produce sharper tips, and are more continuous (pinhole free), but being of lesser quality diamond can have a shortened potential window and increased background current. The best compromise, then, appears to be small grain size films grown with BEN under low methane concentration conditions [14]. An alternative approach to producing BDD microelectrodes is employed by Halpern et al. [6] and Xie et al. [7], where a 25 μm diameter tungsten wire is sealed in a quartz glass capillary and the entire assembly is then placed in the HFCVD reactor. Growth of diamond occurs both on the exposed W disk and on the surrounding quartz insulation, resulting in an unusual disk-ring structure due to the different diamond growth rates on the two materials. The electrodes produced in this way thus tend to have larger dimensions than the initial substrate (typically 35 μm) and the geometry and morphology is quite complicated. However, this method could potentially be modified to produce BDD UMEs of better defined geometry. The authors have successfully used these electrodes for a range of neurodynamic investigations. 6.3.2

Insulation Methods and Control of Exposed Electrode Geometry

As-grown BDD-coated wires are typically attached to a copper wire (current collector) using a silver conducting epoxy to provide a suitable electrical contact. The whole

6.3 BORON-DOPED DIAMOND UMEs

141

assembly must then be insulated; various methods have been reported, including sealing in glass, epoxy, polyimide or other polymers, nail polish, polypropylene tubing, and electrophoretic paint. Most of these methods result in the complete coverage of the electrode surface, requiring that the insulation be later removed from a controlled area at the tip of the electrode by physical or chemical means. Although heat-sealing in glass is the most common approach to fabricating disk UMEs from platinum and gold wires or carbon fibers, diamond UMEs are more difficult to seal in glass due to the typical roughness of the polycrystalline film and the difference in thermal expansion coefficients of the two materials. However, Cooper et al. [11] reported the successful fabrication of diamond UMEs of exposed BDD diameter < 7 μm by completely sealing BDD-coated W tips in a glass capillary (by heating in a ceramic furnace) and then partially exposing the diamond tip by careful polishing or by chemically etching the glass in hydrogen fluoride (HF). The authors noted that although glass insulation was advantageous for use in corrosive or organic environments and is very chemically and physically robust, it is difficult to routinely obtain well-sealed electrodes with a small and defined exposed geometry using this method. The most common and simple insulation methods are to coat the electrode with a suitable epoxy, a nail varnish, or a polymer coating (polyimide films have been reported as being particularly successful for this application [1]) and allow it to dry or cure. The tip of the electrode can then be exposed by touching gently to filter or tissue paper soaked in a suitable solvent (e.g., acetone for nail varnish, sodium hydroxide for polyimide). UMEs may be fabricated surprisingly reproducibly using this method—the size of the resulting exposed electrode depends mainly on the geometry of the as-grown BDD-coated wire tip [13]. As well as its ease, another advantage to this technique is that if too much insulating material is accidentally removed, the whole electrode can be recoated very quickly and the procedure repeated. A disadvantage is that success relies on the skill of the person carrying out the selective exposure, but with practice the technique is very reliable. Another problem is the relative stability of the different coatings toward electrochemical cycling, especially at extreme anodic or cathodic polarizations. The chemical composition of the different epoxy and nail varnish brands may vary considerably, so it is advisable to carefully assess the stability of the chosen insulation under the desired experimental conditions. Another common method for the insulation of sharp etched Pt or C fiber UMEs is the use of electrophoretic paint, which forms a thin insulating coating suitable for fabricating very sharp electrodes (i.e., electrodes with a very small RG). Different types of paints are commercially available and they work on the principle that when a negative (cathodic paint) or positive (anodic paint) potential is applied to the sharpened wire, a local pH change around the electrode, caused by water hydrolysis, results in precipitation of the paint in the form of a thin, uniform polymer film. After curing in an oven, the film becomes electrically insulating and inert. If a sharp needle-shaped electrode is used—for example, an electrochemically etched Pt wire—the paint coating shrinks and retracts from the sharpened point during curing, resulting in a small exposed electrode area at the very tip. This particular property of electrophoretic paint works only rarely with BDD UMEs because the paint tends to retract also from the sharp edges of diamond crystallites down the entire length of the electrode, leading to pinholes and incomplete insulation. Repeating the paint deposition/curing cycle many times results in the formation of a thicker, continuous insulating coating; however, a method is then required to selectively remove the paint from the tip of the electrode to expose the underlying BDD.

142

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

As these paints are mostly impervious to common chemical treatments, the removal of the paint by exposure to a suitable solvent, as discussed previously, is not possible. The paint can be removed by gentle polishing but this can lead to damage to the fragile, fine tips, so a more sophisticated method is to etch away the paint using a focused ion beam (FIB) [15]. The apexes of insulated BDD electrodes can be exposed by FIB milling in different geometries, ranging from “head on” where the ion beam is directly incident in a perpendicular geometry onto the diamond surface, to “side on” where the ion beam is parallel to the exposed electrode surface. Figure 6.2d shows a SEM image of the top of an insulated BDD tip, where a 3 μm diameter circle can be seen in the center of the image, corresponding to the area milled by the FIB in a “head on” configuration. The depth of the milled area is around 0.3 μm, which is thicker than the electrophoretic paint coating, to ensure that all of the paint is removed from that area and that the underlying diamond is exposed. It was found, however, that less than one-third of UMEs fabricated using “head on” FIB milling give satisfactory electrochemical performance. This is believed to be due to damage to the exposed diamond surface from the high-energy impact of the ion beam in the “head on” geometry. In contrast, when the FIB was employed in a “side on” geometry, almost all of the resulting UMEs gave a good electrochemical response consistent with small dimensions, small RG, and defined geometry. To fabricate UMEs using this route requires that the tip radius of the BDD-coated W is roughly similar in size to that required for the UME, as shown in Figure 6.2e for a side on SEM image of a BDD UME of diameter 2 μm. The result is a flat, disk-shaped electrode that is quite suitable for applications such as SECM where the tip needs to approach closely to a substrate. 6.3.3

Electrochemical Performance and Applications

Figure 6.1c shows the CV response for a typical UME in 1 mM ferrocenemethanol (FcOH), fabricated using the “side on” FIB method just described, which exhibits the predicted sigmoidal response for a UME. The limiting current of ∼0.5 nA is consistent with an exposed electrode radius of 1.5 μm (assuming a disk geometry and an RG of 1.1) and electrodes of these dimensions or even smaller are routinely achievable using this method. The backward trace follows the forward scan almost exactly, indicative of very low capacitive charging currents. Some small difference in the forward and backward currents is inevitable and acceptable; the degree of hysteresis observed depends to some extent on the porosity and thickness of the insulating medium, with some materials prone to greater charging than others. If large charging currents are observed, this usually indicates a poor seal between the BDD electrode and the insulating surround, and the electrode should be re-insulated or discarded. The CV response of BDD UMEs of ≤ 25 μm in diameter is largely independent of the scan rates used in conventional experiments (≤ 0.5 V s−1 ) due to the steady-state diffusion of reactant to the electrode surface. Electrodes that are recessed or “lagooned” in their insulating coating may give the expected steady state CV response at slow scan rates, but their behavior at higher scan rates may deviate. Thus, it is important that newly fabricated BDD UMEs should be tested over a range of scan rates to ensure that the geometry of the electrode is as expected. For small electrodes, it is often difficult to assess their size and geometry and the integrity of insulation under a microscope, so using simple electrochemical investigations to diagnose potential issues and obtain accurate electrode dimensions is essential.

6.4 BORON-DOPED DIAMOND UME ARRAYS

143

Boron-doped diamond electrodes have much lower background currents than similar carbon fiber and platinum UMEs and so, as expected, perform particularly well for the detection of low concentrations of analytes. BDD UMEs are particularly valuable in applications where other electrode materials become fouled or require activation in order to be effective. The performance of BDD UMEs in biological media is exemplary, having been used within animal tissue to investigate the role of adenosine in the modulation of breathing [6], serotonin as a neuromodulator [7], norepinephrine release in mesenteric artery [16], and dopamine in mouse brain [3]. Under similar experimental conditions, carbon fiber microelectrodes can become irreversibly deactivated and require polishing and pretreatment steps. The sharper BDD UMEs can also be used in SECM experiments and have been used to image the respiratory activity of immobilized living E. coli cells [13], showing the stability of the electrode during the time taken to obtain an image, its ability to withstand fouling in a biological environment, and its sensitivity to very low concentrations of analyte. The advantage to using BDD UMEs in electroanalytical and biological applications is without question, but their use in quantitative or mechanistic studies is less clear. When BDD UMEs from the same batch, fabricated using identical methods, are tested with a range of common outer-sphere redox couples—FcOH, hexaammineruthenium (Ru(NH3 )6 3+/2+ ), hexachloroiridate (IrCl6 2−/3− ), and ferrocyanide (Fe(CN)6 3−/4− ) —the electrodes often vary in their relative responses to the different couples [15]. For example, an electrode may exhibit perfect steady-state behavior, indicating fast electrode kinetics, when a CV is carried out in FcOH (as in Figure 6.1c) but perform poorly in IrCl6 2−/3− . Another electrode may give ill-defined and nonsteady-state CV responses in both of these redox solutions but perform very well in Ru(NH3 )6 3+/2+ . This variation in redox response is little understood at present; however, it has long been known that some couples usually considered to be outer sphere, in particular Fe(CN)6 3−/4− , are very sensitive to the surface termination of BDD and probably require the presence of specific surface functionalities for optimum electron transfer. Additionally, electrochemical and scanning probe investigations of larger BDD electrodes have revealed that their activity is inhomogeneous, with some areas being highly conductive and hence electrochemically active while other regions show much lower activity [17–19]. The heterogeneity in conductivity has been attributed to differences in boron accumulation of the different diamond crystal faces during growth. However, as similar heterogeneity has been noted for nanocrystalline BDD films, it is clear that several factors must play a role, including dopant distribution and surface chemistry. Regions of high and low electrochemical activity are found to cover areas of several tens of microns on the surface of planar macroelectrodes, with the same area showing a different electrochemical response to different redox couples [19]. As the exposed electroactive area of an UME is of similar (or smaller) dimensions to these regions of heterogeneity, it is not surprising that the electrochemical response of the UMEs varies considerably even within the same batch. 6.4

BORON-DOPED DIAMOND UME ARRAYS

The production of BDD microelectrode arrays is desirable in order to exploit the excellent sensitivities obtainable in electroanalytical applications. The variety of microfabrication facilities now routinely available has meant that several approaches to the production of

144

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

these arrays has been reported [20–23] as well as their use in analytical applications, such as for the amperometric detection of dopamine [24], sulphate and peroxysulfate [23], and 4-nitrophenol and manganese (II) [25]. 6.4.1

Fabrication of BDD UME Arrays

Several of the reported fabrication procedures make use of photolithographic macrofabrication methods in order to produce either masks or moulds of the required geometry for the array. Tsunozaki et al. [21] patterned a Si (100) surface with photoresist, which acted as a mask while the surrounding Si was chemically etched. The resulting microstructured Si surface was then coated with BDD using conventional CVD methods. The entire surface of the deposited BDD layer was then spin-coated with an insulating polyimide layer. The BDD microelectrode array was revealed by mechanically polishing the assembly until the diamond tips were exposed. The distance between each UME was controlled by the photoresist mask and was about 250 μm; the size of each electrode was controlled by the etching conditions used to form the structured Si substrate and was about 25–30 μm in diameter. Soh et al. [22,24] have produced UME arrays of several different geometries, again using photolithography techniques to produce a mould for diamond deposition. A conducting Si substrate was coated with a 0.5 μm thick insulating SiO2 layer by thermal treatment and further coated with a 0.5 μm layer of sputtered Mo. The surface was then patterned using a standard photoresist mask and the unmasked areas of Mo and SiO2 were wet-etched away. The resulting mould was seeded with diamond nanoparticles by sonicating in a suspension, and excess diamond powder on the surface was removed by lifting off the photoresist layer. BDD was then deposited under standard CVD conditions, with the diamond preferentially nucleating and growing in the seeded recessed Si regions. Finally, the Mo layer was removed to leave the ultramicroelectrode array, with the electrodes being reported as very slightly recessed from the insulating SiO2 surround. Arrays of 10 μm × 75 μm or 2 μm × 80 μm microbands were produced, as well as square electrodes of 10 μm × 10 μm with spacings of 20 μm or 100 μm. Provent et al. [23] took a similar fabrication approach, producing an array of 106 slightly recessed 5 μm diameter BDD disk UMEs separated by 150 μm by covering a BDD film with a 0.5 μm layer of insulating silicon nitrite and exposing the microdisks by dry etching of the silicon nitrite. A very successful approach to producing an all-diamond electrochemical device has been to coat a patterned BDD array with an insulating diamond layer [20]. A 500 μm thick BDD film was polished to a mirror finish on one side, masked to define an array pattern, and then ablated in the unmasked regions using a UV laser. The result was an arrangement of BDD columnar structures of 10–25 μm in diameter, with a height of up to 50 μm and separated by about 250 μm. A layer of undoped, insulating diamond was then grown over this array to completely coat the surface. Finally, the nonconducting diamond layer was polished back to expose the tips of the BDD columns, creating a flat surface with an array of BDD UMEs embedded coplanar with the insulating diamond surround, as shown in the SEM image in Figure 6.3a. An advantage of this arrangement is that the electrode can be repeatedly cleaned and polished without structural damage to the device.

6.4 BORON-DOPED DIAMOND UME ARRAYS

145

(a)

(c)

(b)

75 nA

Height (nm)

0 nA

10 0 –10 0

20 40 60 Distance (μm)

80

(d) 4.0 3.0 2.0 1.0 0.0

it/i(∞) Figure 6.3 (a) SEM image of all-diamond BDD UME array showing ‘‘side-on’’ cross-section of conducting BDD disk embedded in insulating diamond matrix (Reprinted with permission from Ref. 20). (b) Topographic AFM image of 80 × 80 μm area of all-diamond BDD UME array showing the conductive BDD disks are coplanar with the surrounding insulating diamond. (c) Conductive AFM image taken simultaneously with image in (b) showing heterogeneity in conductivity of the BDD disk. (d) ECM image of reduction of Ru(NH3 )6 3+ at one of the BDD disks showing inhomogeneity in electrochemical activity. Figures (b), (c), and (d): (Adapted from Ref. 17.) See color insert.

146

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

6.4.2

Electrochemical Performance and Applications

The ideal behavior of an array of UMEs is discussed in Section 6.2; namely, if the diffusion profiles of the individual electrodes do not overlap, the steady-state, sigmoidal CV response of an UME should be obtained, but with a limiting current proportional to the number of individual electrodes in the array. If the packing density of electrodes in the array is too high, then the individual diffusion profiles will overlap and the CV behavior of the array will approach that of a planar macroelectrode. Optimum electrode spacing was investigated by Soh and coworkers [22] for 10 μm × 75 μm microband UMEs separated by 20 μm or 100 μm, where the smaller spacing was found to result in macroelectrode behavior at most scan rates and the larger spacing indicated the expected microelectrode behavior. Simms et al. [26] investigated in some detail the electrochemical response of an all-diamond microarray device described earlier, where although calculation suggested that the diffusion profiles of the individual electrodes did not overlap, true steady-state UME behavior was not observed experimentally. For the simple one electron oxidation of Fe(CN)6 4− some scan rate dependence in the resulting CVs was observed, as well as a gentle peak on both the forward and backward scans. This may be because the individual disks were about 40 μm in diameter, which is toward the upper limit of UME dimensions and a relatively larger degree of linear diffusion to the electrode surface would be expected. An advantage of UME arrays is an increased sensitivity (better signal/noise ratio), meaning that lower detection limits are achievable than at a standard planar BDD macroelectrode, as demonstrated by Lawrence et al. [25] for direct detection of 4-nitrophenol (seven-fold increase in sensitivity observed) and the electrochemical determination of Mn(II) concentration (ten-fold lower limit of detection). All-diamond microdisk arrays were also modified with gold, silver, and copper by electrodeposition for the selective detection of arsenic, hydrogen peroxide, and nitrate respectively, with significant (six to seven times) increase in sensitivity per unit area [26]. These BDD microelectrode arrays were found to be excellent and stable substrates for modification by these different metals. Moreover, the deposits could be simply removed after use, either electrochemically or by polishing, allowing the same BDD sensor to be used for multiple analytical tasks without degradation. Such arrays have also been modified by electrodeposition of palladium for the detection of hydrazine, again achieving much higher sensitivity and lower detection limits than previously available [27]. Boron-doped diamond ultramicroelectrode arrays also have advantages for use in specific environments—for example in media of low ionic strength. Lawrence and colleagues [25] showed that a BDD UME array exhibited a defined, approximately steady-state reduction wave for Ru(NH3 )6 3+ in water in the absence of any supporting electrolyte. In contrast, a planar BDD macroelectrode did not produce a well-defined redox response under the same conditions. Provent et al. [23] showed that their BDD UME array performed well for the industrially relevant detection of sulphate and peroxodisulfate. The small size of the array, which was formed along with a crook-shaped BDD counterelectrode on a single 7 mm × 2.8 mm silicon chip, suggests its use as a maintenance free and cost-effective sensor in real industrial applications [23]. Although BDD UME arrays are very valuable for diffusion-limited amperometric detection applications, caution must be exercised in the analysis of voltammetric wave shapes. Colley et al. [17] investigated the spatial heterogeneity in electroactivity of an all-diamond BDD microarray (such as that shown in Figure 6.3a) using conducting-AFM

6.5 NANOSTRUCTURED BDD ELECTRODES

147

(C-AFM) and SECM. Each conducting microdisk within the array is made up of polycrystalline BDD with different grain orientations exposed. Boron uptake is known to vary with grain orientation and so it is not surprising that conductivity of individual grains within the same microdisk was found to vary. Figures 6.3b and 6.3c show simultaneously recorded C-AFM topographic and conductivity images for a single microdisk BDD electrode embedded in an insulating diamond matrix. The mirror finish of the alldiamond microarray is demonstrated clearly in this image, where a surface roughness of only about 10 nm was observed; the BDD disk, however, could readily be distinguished from the nonconducting diamond surround and individual grains easily discriminated. The conductivity image shows clear heterogeneity both from grain to grain and within a single grain. The most conductive grains exhibited metallic conductivity when their current-voltage characteristics were investigated; in contrast, the less conductive regions indicated p-type semiconductivity. The influence of this heterogeneity on electroactivity was investigated using SECM, by imaging the production of Ru(NH3 )6 2+ by the microelectrode disks (electrogenerated by reduction of Ru(NH3 )6 3+ ). Most disks showed a similar (high) level of activity, but it was clear that some electrodes were much less electrochemically active. High-resolution imaging of the electroactivity of one of the less active disks showed clear variations in electroactivity over the electrode, as shown in Figure 6.3d. Only part of the electrode exhibited the currents expected at a metallic electrode, with the variation in electron transfer kinetics most likely linked to the differences in conductivity of the individual grains. For this reason, care should be take in kinetic analysis of voltammetric measurements using these arrays, as the variation in electroactivity will be even more significant at lower electrode overvoltages. 6.5

NANOSTRUCTURED BDD ELECTRODES

Random arrays of BDD nanodisk electrodes may be prepared by selective insulation of a planar BDD macroelectrode, using a similar approach to UME array fabrication [28]. However, the mass transport and hence voltammetric characteristics of nanoarrays vary significantly from UME arrays, as will be discussed in Section 6.5.1. An alternative reported approach is to nanostructure existing BDD films by formation of vertically aligned BDD nanowires on the surface. Production of diamond nanowires was realized some time ago with suggested applications in advanced composites, thermal management, field emission, and microelectronics. However, only very recently have nanostructured BDD films been fabricated to allow their use in electrochemical and biosensing applications [29–31]. In contrast to UME arrays, where conducting BDD electrodes are embedded within an insulating matrix, the entire surface of the nanostructured electrode is exposed and electrochemically active. The electrochemical characteristics, and in particular the mass transport behavior, are thus quite complicated, with greater similarity to a porous macroelectrode than to a microelectrode array. Perhaps as a consequence of this, or also due to nanoscale or even atomic scale morphological and chemical modifications to the diamond surface, the electrochemical response toward some analytes is very different to that at a planar macroscopic BDD electrode. 6.5.1

Random Array BDD Nanodisk Electrodes

A random array of BDD nanodisk electrodes has been prepared by a three-step process [28]. First, a conventional BDD electrode was modified by electrodeposition of

148

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

molybdenum dioxide nanoparticles. The entirety of this modified electrode was then coated in an insulating polymer film layer by the electropolymerisation of 4-nitrophenyldiazonium salt. Finally, the molybdenum oxide nanoparticles were dissolved from the surface using dilute HCl. This step also results in the removal of the polymer film directly above the nanoparticles and hence the exposure of nanodisks of BDD of 20 ± 10 nm in diameter. This method is reported as producing 650 ± 25 million BDD nanodisks per cm2 . Due to the high density of nanodisks, the diffusion profiles of each individual electrode overlap, with the result that diffusion to the surface is linear. The CV response is thus that of a macroelectrode with the maximum current achievable being that of the geometric area of the electrode (i.e., the area of the BDD electrode before modification)—despite most of the surface being covered with an insulating layer. In contrast, the background capacitive current of the nanoarray electrode is significantly reduced compared to the bare BDD, as most of the surface is insulated. This provides significant benefits in the electroanalytical detection of small concentrations of analytes. 6.5.2

Fabrication of Nanostructured BDD Arrays

Two different approaches to the fabrication of BDD nanoarrays have been used to date: the “top-down” formation of nanowires using reactive ion etching (RIE) of a conventional BDD film [29,30] and the “bottom-up” growth of nanostructures on a silicon nanowire template [31]. In the first method, smooth highly doped BDD films are exposed to oxygen plasma, and unmasked areas of the diamond film are etched away to form gaseous CO2 and CO, leaving an array of BDD nanostructures under the masked regions. Yang et al. [29] used a layer of 8–10 nm diamond nanoparticles as the etching mask, which they formed by sonicating the diamond film in a suspension of diamond nanoparticles. The seeded nanoparticle layer is dense, and after RIE treatment vertically aligned nanowires arise where the nanoparticles were deposited. The length of the BDD nanowires depends on the nature of the mask and the time of the plasma treatment—the etching rate of the ´˚ s−1 , meaning that a 10 nm particle will act as a diamond nanoparticles is about 10 A mask for 10 s before being etched away. Therefore, the optimum geometric dimensions of the nanostructures formed after 10 s etching time is 10 nm length nanowires with about 10 nm separation between them and diameters of 1–5 nm (determined by tapping mode AFM measurements). If the time of plasma treatment is increased past the lifetime of the masking layer, then the newly formed nanowires also begin to be etched away, eventually resulting in a smooth film. If silicon nanoparticles are used as a mask, the length of the nanowires can be increased to several microns, as the etching rate of silicon ´˚ s−1 . in the oxygen plasma is only 1 A Wei et al. [30] have taken a different approach for the formation of a BDD “nanograss array” using oxygen plasma RIE. No etching mask was used in this case; it was assumed that in the oxygen plasma environment boron oxide species form on the film surface at boron-rich sites and were themselves able to act as a mask. The boron oxides are less volatile than the CO and CO2 species formed from diamond etching, and so in the early stages of treatment remain on the surface while the surrounding area is etched. Eventually, they too are removed, but tend to redeposit near the tops of the nanostructures and continue to serve as a mask. Final dimensions of the nanograss array (determined

149

6.5 NANOSTRUCTURED BDD ELECTRODES

by SEM) after about 10 min etching time were 20 nm diameter, 200 nm length, and a separation of 50 nm between structures. Luo and Zhi [31] used a more conventional CVD growth approach to the formation of BDD “nanorod forest electrodes,” Electroless metal deposition of silicon onto a p-type (100) silicon wafer was carried out to form a silicon nanowire (SiNW) array, which was used as the substrate for the HFCVD growth of a BDD film. SEM investigation showed complete and continuous coverage of the polycrystalline diamond film over the whole length of the SiNW. This resulted in the formation of vertically aligned BDD nanorods of several microns in length and about 100–500 μm in diameter, with similar spacing between structures, as shown in Figures 6.4a and b. 6.5.3 Electrochemical Performance and Applications of Nanostructured BDD Electrodes The interfacial electronic properties of the nanorod arrays produced by Yang et al. [29] using RIE were assessed using capacitance (C)–voltage (V) measurements and analysis using the Mott-Schottky equation. The active sensor area was estimated to be about 1.3–2.0 times that of the smooth electrode before plasma treatment, which was consistent with the estimated geometric area calculated from AFM images. Additionally, the builtin potential of the BDD surface was found to increase significantly from the 1.7 V of the hydrogen-terminated smooth BDD film, to 3.2 V for the oxygen-terminated plasma treated surface where many surface defects were generated. These BDD nanorod arrays were utilized as DNA biosensors in order to detect DNA hybridization, reaching a detection level of single-base mismatch [32]. The formation of the DNA biosensor relied on a unique property of these nanostructured films. Because of the surface morphology of the nanoarrays, current density is greatest at the very tip of the nanorods allowing selective electrochemical modification of the tips of the nanorods, while the remaining electrode area is unchanged. The first step in the formation of the sensor is the electrochemical attachment of aminophenyl linker molecules to the tip of the nanorods, first by electrochemical reduction and grafting of the corresponding nitrophenyl diazonium salt to form a layer of immobilized nitrophenyl groups, followed by their electrochemical reduction to the aminophenyl moieties. Single-stranded DNA molecules can then be cross-linked to

(a)

10

(b)

(c)

Current (µA)

8 500 nm

Nanoforest BDD 6 4 2 planar BDD

10 µm

1 µm

0 0.0

0.2

0.4 0.6 Potential (V)

0.8

Figure 6.4 (a) and (b) SEM images of ‘‘nanoforest’’ BDD array. (c) Linear sweep voltammogram for the oxidation of glucose at a planar BDD electrode and a nanoforest BDD electrode. (Adapted from Ref. 33.)

150

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

the aminophenyl groups. As only the very tips of the nanostructures are modified in this way, the remaining electrode area is available for electrochemical detection of solution redox probes. When the DNA is in its single-stranded form, negatively charged redox probes such as Fe(CN)6 3−/4− can readily diffuse through the strands to the underlying unmodified electrode where it can be detected voltammetrically. However, when the DNA is hybridized with its target DNA strand, the larger size and increased negative charge of the double-stranded DNA blocks the passage of the Fe(CN)6 3−/4− probe to the electrode surface. The resulting decrease in current can be used as an indicator of DNA hybridization. The nanostructure of this electrode is essential to its operation; if a planar BDD electrode is used, the initial deposition of the aminophenyl linker groups occurs indiscriminately over the whole electrode surface, blocking the electrode and preventing any redox response to Fe(CN)6 3−/4− . It is the ability to modify only the tips of the nanorods while leaving the remaining electrode area unblocked and available for electrochemical sensing that makes these electrodes so promising for biosensing applications. The “nanograss array” electrodes developed by Wei et al. [30] were used for the detection of dopamine and uric acid. In comparison to an oxidized planar BDD film, oxidation of both species took place at a less positive applied potential and higher peak currents were obtained. This is indicative of increased electrocatalytic activity and faster rates of electron transfer, perhaps due to the formation of more active sites, as well as increased electrode area. These observations would suggest that nanostructuring of electrodes may also result in activation of the electrode toward some analytes, by providing additional sites for adsorption and catalysis. Similarly, Luo, Wu, and Zhi [33] used their “nanorod forest” electrodes for the oxidation of adenine and direct amperometric (non-enzymatic) detection of glucose, as shown in Figure 6.4c [33], where in comparison to the equivalent planar BDD much greater sensitivity and selectivity toward the analytes was observed. The authors attribute this advantage to the high surface area and porosity of the nanoarray electrode, as the Faradaic currents for the kinetically sluggish oxidation of glucose will depend on the nanoscopic area of the electrode rather than its geometric area.

6.6

CONCLUSIONS AND FUTURE DIRECTIONS

Several different strategies for the fabrication of single BDD UMEs and arrays of microand nanostructured BDD electrodes are available, allowing these electrodes to be used in a wide range of applications. The advantage of BDD for many of these uses is unquestionable—its biocompatibility and sensitivity to low concentrations of analytes lends its use to in vivo applications in particular. Several areas can be identified as particularly promising for future application of BDD UMEs—for example, in the use of surface modified electrodes (e.g., with metals or enzymes) for sensing of specific analytes. In order to optimize the use of BDD UMEs, further work is required in understanding the fundamental behavior of these electrodes, especially the factors that contribute to observed variation in electrochemical activity. Although research into nanostructured diamond electrodes is at a very early stage, the unusual sensitivity of these films to analytes such as glucose that are not readily detected at planar electrodes is very interesting and suggests future applications in biosensing. As micro- and nanofabrication technologies advance and become more available, it is likely that even more sophisticated BDD electrodes on a ultramicro- and nanoscale will be achievable.

REFERENCES

151

REFERENCES 1. J. Cvacka, V. Quaiserova, J.W. Park, Y. Show, A. Muck, G.M. Swain, Anal. Chem. 2003, 75 , 2678–2687. 2. D. Shin, B.V. Sarada, D.A. Tryk, A. Fujishima, Anal. Chem. 2003, 75 , 530–534. 3. A. Suzuki, T.A. Ivandini, K. Yoshimi, A. Fujishima, G. Oyama, T. Nakazato, N. Hattori, S. Kitazawa, Y. Einaga, Anal. Chem. 2007, 79 , 8608–8615. 4. J. Park, V. Quaiserova-Mocko, B.A. Patel, M. Novotny, A.H. Liu, X. C. Bian, J.J. Galligan, G.M. Swain, Analysis 2008, 133 , 17–24. 5. B.A. Patel, X.C. Bian, V. Quaiserova-Mocko, J.J. Galligan, G.M. Swain, Analysis 2007, 132 , 41–47. 6. J.M. Halpern, S. Xie, G.P. Sutton, B.T. Higashikubo, C.A. Chestek, H. Lu, H.J. Chiel, H.B. Martin, Diam. Rel. Mat . 2006, 15 , 183–187. 7. S. Xie, G. Shafer, C.G. Wilson, H.B. Martin, Diam. Rel. Mat . 2006, 15 , 225–228. 8. A.J. Bard, L.R. Faulkner, Electrochemical Methods Fundamentals and Applications, 2nd ed., John Wiley and Sons, 2001, pp. 168–176. 9. A.J. Bard, M.V. Mirkin, eds., Scanning Electrochemical Microscopy, Marcel Dekker, New York, 2001. 10. D. Shoup, A. Szabo, J. Electroanal. Chem, 1984, 160 , 27–31. 11. J.B. Cooper, S. Pang, S. Albin, J. Zheng, R.M. Johnson, Anal. Chem. 1998, 70 , 464–467. 12. B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 1469–1471. 13. K.B. Holt, J.P. Hu, J.S. Foord, Anal. Chem. 2007, 79 , 2556–2561. 14. J.P. Hu, J.S. Foord, K.B. Holt, Phys. Chem. Chem. Phys. 2007, 9 , 5469–5475. 15. J.P. Hu, K.B. Holt, J.S. Foord, Anal. Chem. 2009, 81 , 5663–5670. 16. J.W. Park, Y. Show, V. Quaiserova, J.J. Galligan, G.D. Fink, G.M. Swain, J. Electroanal. Chem. 2005, 583 , 56–68. 17. A.L. Colley, C.G. Williams, U. D’Haenens Johansson, M.E. Newton, P.R. Unwin, N.R. Wilson, J.V. Macpherson, Anal. Chem. 2006, 78 , 2539–2548. 18. (a) K.B. Holt, A.J. Bard, Y. Show, G.M. Swain, J. Phys. Chem. B 2004, 108 , 15117–15127. (b) N.R. Wilson, S.L. Clewes, M.E. Newton, P.R. Unwin, J.V. Macpherson, J. Phys. Chem. B 2006, 110 , 5639–5646. (c) K. Honda, T. Noda, M. Yoshimura, K. Nakagawa, A. Fujishima, J. Phys. Chem. B 2004, 108 , 16117–16127. (d) S. Szunerits, M. Mermoux, A. Crisci, B. Marcus, P. Bouvier, D. Delabouglise, J.-P. Petit, S. Janel, R. Boukherroub, L. Tay, J. Phys. Chem. B 2006, 110 , 23888–23897. 19. S. Wang, G.M. Swain, J. Phys. Chem. C 2007, 111 , 3986–3995. 20. M. Pagels, C.E. Hall, N.S. Lawrence, A. Meredith, T.G.J. Jones, H.P. Godfried, C.S.J. Pickles, J. Wilman, C.E. Banks, R.G. Compton, L. Jiang, Anal. Chem. 2005, 77 , 3705–3708. 21. K. Tsunozaki, Y. Einaga, T.N. Rao, A. Fujishima, Chem. Lett . 2002, 502–503. 22. (a) K.L. Soh, W.P. Kang, J.L. Davidson, S. Basu, Y.M. Wong, D.E. Cliffel, A.B. Bonds, G.M. Swain, Diam. Rel. Mat . 2004, 13 , 2009–2015. (b) K.L. Soh, W.P. Kang, J.L. Davidson, Y.M. Wong, D.E. Cliffel, G.M. Swain, Diam. Rel. Mat . 2008, 17 , 240–246. 23. C. Provent, W. Haenni, E. Santoli, P. Rychen, Electrochim. Acta 2004, 49 , 3737–3744. 24. K.L. Soh, W.P. Kang, J.L. Davidson, Y.M. Wong, D.E. Cliffel, G.M. Swain, Diam. Rel. Mat . 2008, 17 , 900–905. 25. N.S. Lawrence, M. Pagels, A. Meredith, T.G.J. Jones, C.E. Hall, C.S.J. Pickles, H.P. Godfried, C.E. Banks, R.G. Compton, L. Jiang, Talanta 2006, 69 , 829–834. 26. A.O. Simms, C.E. Banks, S. Ward-Jones, T.J. Davies, N.S. Lawrence, T.G.J. Jones, L. Jiang, R.G. Compton, Analyst 2005, 130 , 1303–1311.

152

DIAMOND ULTRAMICROELECTRODES AND NANOSTRUCTURED ELECTRODES

27. C. Batchelor-McAuley, C.E. Banks, A.O. Simm, T.G.J. Jones, R.G. Compton, Analyst 2006, 131 , 106–110. 28. L. Xiao, I. Streeter, G.G. Wildgoose, R.G. Compton, Sens. Actuators B 2008, 133 , 118–127. 29. N. Yang, H. Uetsuka, E. Osawa, C.E. Nebel, Nano Lett . 2008, 8, 3572–3576. 30. M. Wei, C. Terashima, M. Lv, A. Fujishima, Z.-Z. Gu, Chem. Comm. 2009, 3624–3626. 31. D. Luo, J. Zhi, Electrochem. Commun. 2009, 11 , 1093–1096. 32. (a) N. Yang, H. Uetsuka, C.E. Nebel, Adv. Funct. Mater. 2009, 19 , 887–893. (b) C.E. Nebel, N. Yang, H. Uetsuka, E. Osawa, N. Tokuda, O. Williams, Diam. Rel. Mat . 2009, 18 , 910–917. 33. D. Luo, L. Wu, J. Zhi, ACS Nano 2009, 3 , 2121–2128.

Part III

Electroanalytical Applications

7 Electroanalytical Applications of Diamond Films Weena Siangproh, Amara Apilux, Pimkwan Chantarateepra, and Orawon Chailapakul

7.1

INTRODUCTION

The success of an electrochemical-sensing process relies mainly on the careful choice of electrodes. An ideal electrode should exhibit several features: It should be mechanically stable, chemically inert, and have a wide range of potential and an easily reproducible surface. Electrode materials and designs can be classified into two main categories: solid electrodes and liquid electrodes. Metal electrodes offer very fast electron transfer, but the available working potential window is often quite narrow due to the onset of background reactions, such as metal oxidation and oxygen and hydrogen gas evolution. Alternatively, various types of graphite and carbon materials (e.g., glassy carbon), offer significantly wider working potential ranges. Although a number of different types of carbon materials have been applied to electrochemical methods, there are specific characteristics that electrochemists would like to see improved, such as the undesired adsorption of electroactive species from solution, the undesirably large capacitive background current and oxidation current at higher potentials, and the lack of mechanical ruggedness. Therefore, there remains a need for new, selective, sensitive, and mechanically stable materials.

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

155

156

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

Since the early twentieth century, synthetic conductive diamond thin films, a new carbon material, have been increasingly used and reported [1–6]. In addition, there has been a growing interest in the use of boron-doped diamond (BDD) electrodes in various applications. The BDD electrode is one of the best insulators; boron is utilized as a thin metal film in diamond, at levels between 1016 and 1021 cm−3 , with activation energies down to less than one-tenth of an electron volt. Conductivity can be fine-tuned for a variety of purposes. Therefore, boron-doped diamond is one of the most promising new materials for electroanalytical measurements. The BDD electrode possesses electrochemically attractive properties compared to other electrodes, including low and stable voltammetric background currents [7]. The currents for double-layer charging are quite small due to the low number of carriers and the nearly complete lack of porosity. These attributes make the BDD electrode well suited for current-based electrochemical measurements. Other attractive features include wide working potential window in aqueous electrolyte solution, low adsorption of polar molecules from aqueous solution, long-term response stability, and good activity without any pretreatment. Because of the unique properties of BDD electrodes, they have been widely used in various applications. Therefore, the use of diamond as an electrode material in electrochemistry has been summarized up through early 2006 [8]. In this chapter, we provide a detailed, up-to-date, and compact review covering nearly all aspects of electroanalytical application using boron-diamond electrodes from late 2006 until now.

7.2

PHARMACEUTICAL COMPOUNDS

Pharmaceutical compounds are the usual form of prescribed medicines. In addition to the benefits of these medicines, they may have side effects that can be very serious, so quality control of the active ingredient is required. In pharmacokinetic studies, the levels of pharmaceuticals in biological fluids can provide information about treatment to biomedical scientists. However, the successful exploitation of such an approach is dependent on the availability of analytical techniques that can provide fast quantitative measurements of the concentrations of various therapeutic substances. Electroanalytical chemistry has an important place in contemporary chemical analysis because of its large potential for possible applications. In comparison to other instrumental methods, electroanalytical chemistry offers a great variety of advantages, including high sensitivity, simplicity, quickness, low cost, and relatively easy detector miniaturization. Electrochemical methods using boron-doped diamond electrodes have been successfully applied to analyze pharmaceutical compounds, and some examples follow. In 2006, Karuwan et al. [9] reported the construction and application of boron-doped diamond thin film electrodes as sensors for pharmaceutical analysis. A flow injection system with pulsed amperometric detection (PAD) for anti-fouling of the BDD electrode was developed for the determination of three beta-agonists: salbutamol, terbutaline, and clenbuterol. A linear response was obtained in the concentration ranges of 0.5–100 μM, 1.0–100 μM and 0.5–50 μM for salbutamol, terbutaline, and clenbuterol, respectively. The developed PAD-BDD system was applied to determine salbutamol and terbutaline in commercial pharmaceutical products. In 2007, Oliveira et al. [10] reported an electroanalytical method to detect lidocaine using boron-doped diamond electrodes by cyclic voltammetry (CV) and square-wave

7.2 PHARMACEUTICAL COMPOUNDS

157

voltammetry (SWV). Lidocaine is an active ingredient in many commercial local anesthetic products. Because of its widespread use, the development of a new analytical method is required. From the results, an irreversible oxidation peak of lidocaine in 0.1 M Britton-Robinson buffer was observed at 1.68 V versus Ag/AgCl. The detection and quantification limits obtained from pure water were 10.0 μg L−1 and 34.4 μg L−1 , respectively. Furthermore, the electrochemical responses of lidocaine in pharmaceutical preparations were also determined. The results obtained were in agreement with those determined by standard methods. The presence of propyleneglycol as interference in the gels samples had no effect on the voltammetric responses. Lidocaine recoveries ranged from 97.6 to 99.2%. Next, Topal, Uslu, and Ozkan [11] reported the method for electroanalytical investigation of atorvastatin calcium at boron-doped diamond electrodes using voltammetry. The electrooxidation of atorvastatin was irreversible and exhibited a diffusion-controlled behavior. A linear response was obtained in the range of 9.65 × 10−7 M to 3.86 × 10−5 M in 0.1M H2 SO4 with detection limits of 2.27 × 10−7 M and 1.31 × 10−7 M with differential pulse voltammetry (DPV) and square-wave voltammetry (SWV), respectively. The methodology was successfully validated and applied to a high-throughput process for detection of atorvastatin in tablets, human serum, and human urine with good recovery and repeatability. Interestingly, BBD electrodes also displayed excellent behavior for the determination of fluvastatin in pharmaceutical formulations and human serum without any requirements for separation or complex sample preparation [12]. The electrooxidation of fluvastatin sodium at a boron-doped diamond electrode was investigated using cyclic, differential pulse, and square-wave voltammetry. The oxidation of fluvastatin was irreversible and exhibited a diffusion-controlled manner. The results indicated that a linear dynamic range was observed from 1 × 10−6 M to 6 × 10−4 M in a Britton-Robinson buffer solution (pH 10.0), with a detection limit of 1.37 × 10−7 M by DPV and 1.44 × 10−7 M by SWV. The RSD values obtained were 0.66% and 0.15% for the peak currents with DPV and SWV, respectively. In 2008, many researchers reported on the use of boron-doped diamond electrodes for electroanlytical purpose. Uslu, Topal, and Ozkan [13] proposed the electrooxidation for the determination of pefloxacin in pharmaceuticals, and serum at boron-doped diamond and glassy carbon electrodes using cyclic, linear sweep, differential pulse, and squarewave voltammetry. Using cyclic voltammetry, the oxidation peak was irreversible and exhibited a diffusion-controlled behavior. Linearity was obtained in the concentration range of 2 × 10−6 M to 2 × 10−4 M in 0.5 M H2 SO4 at about 1.20 V versus Ag/AgCl for DPV and SWV. The potential applicability of these methods was measured by successful determination of this drug in pharmaceutical dosage forms and human serum samples with good recovery results, reproducibility, precision, and accuracy. There was no interference from the excipients and other substances in the determination of these real samples. Additionally, two sulfonamides—sulfadiazine and sulfamethoxazole—were independently quantified in pharmaceutical products by SWV using a boron-doped diamond electrode by Souza et al. [14]. A well-resolved irreversible oxidation peak was observed at around 1.1 V of sulfadiazine in ethanol + 0.5 M H2 SO4 (50/50 v/v) and of sulfamethoxazole in ethanol + pH 6.0 phosphate buffer (50/50 v/v) solutions. Linearity was observed in concentration ranges from 8.01 × 10−6 M to 1.19 × 10−4 M (r = 0.9995) for sulfadiazine and 6.10 × 10−6 M to 6.01 × 10−5 M (r = 0.9995) for sulfamethoxazole

158

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

with detection limits of 2.19 × 10−6 M and 1.15 × 10−6 M, respectively. For both sulfonamides’ detection, the accuracy of the proposed methodology had no statistical error when compared to those of standard HPLC methods. The recoveries obtained were 95–104%. These findings indicated that the proposed electroanalytical method is attractive and suitable for the determination of sulfa drugs in pharmaceuticals and other products. In addition, Andrade et al. [15] proposed the simultaneous determination of sulfamethoxazole and trimethoprim in pharmaceutical products. The hydrogen-terminated (HT) and oxygen-terminated (OT) BDD electrodes were used in order to investigate the electrooxidation of these compounds. The HT-BDD electrode demonstrated two well-defined oxidation peaks, and the oxidation current peak increased 20-fold when compared to those of the OT-BDD electrode. The calculated limit of detection (LOD) values for sulfamethoxazole and trimethoprim using the HT-BDD electrode were 3.65 μg L−1 and 3.92 μg L−1 , respectively. Wei et al. [16] reported on the investigation of electrochemical oxidation of procaine hydrochloride (PC center dot HCL, 2-diethylaminoethyl 4-aminobenzoate hydrochloride) at an as-deposited boron-doped diamond electrode, anodically oxidized BDD (ao-BDD) electrode, and glassy carbon (GC) electrode by cyclic voltammetry (CV). The cyclic voltammograms showed the oxidation of PC center dot HCL with a high signal-tobackground ratio, low tendency for adsorption, good reproducibility, and long-term stability at the BDD electrode. Good linearity was observed for a concentration range from 5 μM to 200 μM and a detection limit of 0.5 μM was achieved. Radovan, Cofan, and Cinghita [17] used BDD electrodes to simultaneously determine l-ascorbic acid (AA) and acetaminophen (AC) by several techniques, including cyclic voltammetry (CV), chronoamperometry (CA), and differential pulse voltammetry (DPV). Using differential pulse voltammetry, the relationship between the concentration and peak current was linear for both analytes in the 0.01–0.1 mM concentration range with very high correlation coefficients. The relative standard deviation of 2–3% and high sensitivities were obtained from the DPV data in single and dicomponent systems. In the same year, similar studies were conducted using CV and CA for the simultaneous determination of ascorbic acid and acetaminophen [18]. Both techniques provided good linearity for each of the investigated compounds in single and dicomponent solutions within concentrations of 0.01–0.1 mM with very good correlation parameters. High sensitivity values and 2–3% RSDs were obtained. In conclusion of this work, the BDD electrode was identified as an excellent promising material for the simultaneous determination of AA and AC in real sample solutions from pharmaceutical products. The electrochemical oxidation of promethazine hydrochloride at highly boron-doped diamond electrodes using square-wave adsorptive voltammetry was investigated by Ribeiro et al. [19]. This analyte provided two oxidation peaks. The second peak was generated from the formation of the adsorbed product. The parameters were optimized and the highest current intensities were obtained by applying a potential at 0.78 V for 30 s. The best conditions were obtained at pH 4.0 using Britton-Robinson buffer, a frequency of 50 s−1 , a step of 2 mV, and an amplitude of 50 mV. The linear dynamic response was obtained for the concentration range of 5.96 × 10−7 M to 4.76 × 10−6 M, with detection limits of 2.66 × 10−8 M (8.51 μg L−1 ) for peak 1 and of 4.61 × 10−8 M (14.77 μg L−1 ) for peak 2. Altogether these results indicated that the method was stable, sensitive, and reproducible. Then, in 2009, the determination of acetylsalicylic acid (ASA) in pharmaceutical formulations using boron-doped diamond electrodes was reported by Sartori et al. [20].

7.3 BIOMOLECULES OR BIOLOGICAL COMPOUNDS

159

This method was applied to determine ASA in 0.01 M H2 SO4 by SWV without the need of a previous time-consuming alkaline hydrolysis step. An irreversible oxidation peak at 1.97 V versus Ag/AgCl (in 3.0 M KCl) was observed. The concentration curve was linear from 2.50 × 10−6 M to 1.05 × 10−4 M with a detection limit of 2.0 × 10−6 M. The relative standard deviation values were smaller than 1.4% for a 45 μM ASA solution (n = 10). This method was successfully applied for the determination of ASA in several pharmaceutical formulations. The reports on these studies confirmed that BDD electrodes can be used to determine many active ingredients in various pharmaceutical compounds. The proposed methods showed excellent performances and great potential for application in control processes.

7.3

BIOMOLECULES OR BIOLOGICAL COMPOUNDS

The use of boron-doped diamond electrodes to detect biomolecules has also been a frequently published topic because of their important role in biological system. Due to the overwhelming amount of literature available, biomolecules is another growing trend in this field. Beginning in 2006, the unmodified boron-doped diamond electrode was the first to be used successfully for the simultaneous measurement of tryptophan (Trp) and tyrosine (Tyr) by differential pulse voltammetry [21]. The oxidation peaks of Trp and Tyr using a Na2 PO4 /NaOH buffer solution pH 11.2 were completely separated at a BDD electrode. The detection limits of Trp and Tyr were 1 × 10−5 M and 1 × 10−6 M, respectively. This method was also used for the determination of real amino acid samples. In 2007, Zhaom Li, and Li [22] reported on the use of the BDD electrode without further modification for selective detection of dopamine (DA) in the presence of ascorbic acid (AA) by differential pulse voltammetry. The oxidative peaks of DA and AA were separated by 0.44 V at the BDD electrode, whereas these two peaks cannot be separated at the glassy carbon electrode. The results showed high sensitivity and a detection limit of 1.1 × 10−6 M in a 200-fold excess of AA in acidic media. This method was also applied to determine DA in real samples. In a related report, BDD microelectrodes were used for the in vivo analysis of dopamine in mouse brain by differential pulse voltammetry [23]. The BDD microelectrodes with very small tips (5 μm diameter and 250 μm long) were fabricated by depositing BDD on chemically etched micrometer-sized tungsten wires using microwave plasma-assisted chemical vapor deposition. These microelectrodes were inserted into the corpus striatum of the mouse brain. The results showed a clear signal for DA following electrical stimulation of dopaminergic neurons, suggesting that the BDD microelectrode is a promising electrode for future in vivo electroanalysis. Consequently, in the same year, the electrochemical behavior of native and thermally denatured fish DNA was explored using a boron-doped diamond thin film electrode with cyclic voltammetry by Apilux, Tabata, and Chailapakul [24]. The BDD electrode is advantageous for measuring currents less than 1 μA for DNA solutions due to its low background current. The results of the thermally denatured fish DNA showed two clear oxidation peaks due to the oxidation of guanine and adenine at about 1.1 V versus Ag/AgCl and 1.3 V versus Ag/AgCl, respectively. In contrast, native fish DNA showed a poor oxidation peak at 1.1 V versus. Ag/AgCl (see Figure 7.1). In addition, the electrochemical behavior of thermally denatured fish DNA was studied in the presence of cytosine, cytidine, cytidine-5-monophosphate, tetrakis (1-methylpyridinium-4yl) porphyrin (H2 (TMPyP)4+ ), and RuII(TMPyP)4+ .

160

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

Current (μA)

1.8 1.3

Current(μA)

2.3

0.8

5 4 3 2 1 0 –1 –2 –0.7

–0.2

0.3

0.8

1.3

1.8

Potential (V) vs. Ag/AgCl

0.3 –0.2 –0.7 –0.7

–0.2

0.3

0.8

1.3

1.8

Potential (V) vs. Ag/AgCl Figure 7.1 Cyclic voltammograms of 1 mM in base pair of the thermally denatured fish DNA in 0.2 M acetate buffer pH 4.6 at BDD electrode at scan rate 0.01 Vs−1 . Inset: 1 mM in bp the native fish DNA in 0.2 M acetate buffer pH 4.6 at BDD electrode at scan rate 0.01 Vs−1 . The background cyclic voltammograms are also shown (dashed line). (Reprinted with permission from Ref. 24.)

Next, Fortin et al. [25] reported on the use of BDD electrodes to study the electrochemical oxidation of 2 -deoxyguanosine. Interestingly, the BDD electrode showed excellent mechanical properties for the investigation of 2 -deoxyguanosine that is not easily electrochemically detectable on other electrodes. The oxidation peaks of 2 -deoxyguanosine and 2 -deoxyadenosine were detected at potentials of 1.1 and 1.4 V versus Ag/AgCl, respectively. Moreover, small nucleoside concentrations could be detected because of the low background-to-signal ratio. Furthermore, the attractive behavior and advantages of a boron-doped diamond electrode for HPLC amperometric detection were noticed. Ivandini et al. [26] reported the simultaneous detection of purine and pyrimidine bases in mild acidic media by BDD electrodes. As-deposited (AD) and anodically oxidized (AO) BDD electrodes were used to study the electrochemistry property of the analytes by cyclic voltammetry. Under the optimized conditions for HPLC, the linear range of the calibration curves was from 0.1 μM to 10 μM, with limits of detection from 26.3 μM to 162.1 nM. The sensitivity was an order of magnitude higher than when using the conventional electrodes, and this method was successfully applied for the determination of 5-methylcytosine in real DNA samples with high reproducibility. In 2008, the conductive boron-doped diamond electrodes were also successfully demonstrated for the direct electrochemical detection of proteins, including bovine serum albumin (BSA) and immunosuppressive acidic protein (IAP, a cancer marker) [27]. Cyclic voltammetric results suggested that cysteine, tyrosine. and tryptophan contribute to the direct electrochemical oxidation at BDD electrodes. A linear dynamic response in the range of 10 to 100 μg mL−1 (r = 0.9) with a low detection limit of 10 μg mL−1 was achieved for BSA. For IAP detection, the linearity was observed from 200 to 800 μg mL−1 with a detection limit of 100 μg mL−1 . The excellent reproducibility was demonstrated with an RSD value of 5%. In addition, in 2009, a BDD thin film electrode with amperometric detection was used to detect biphenyl amino derivatives (2-aminobiphenyl, 3-aminobiphenyl, and

7.3 BIOMOLECULES OR BIOLOGICAL COMPOUNDS

161

4-aminobiphenyl) in model drinking and river water samples [28]. The separation of studied analytes was performed at a ChiraDex column (Merck, Germany) with chemically bonded P-cyclodextrin in a mobile phase aqueous buffer/acetonitrile/methanol (40/30/30, v/v/v). The separation time of six minutes was achieved. Nanomolar limits of quantitation were obtained. To enhance sensitivity and selectivity for the detection of various species, the modification of BDD electrode surface has been extensively investigated. Biosensers are the example of surface modification by attachment of organic functional groups and biomolecules. In addition, biosensors based on the use of BDD electrodes with enzymatic immobilized for enhancing the sensitivity of electrochemical analytical applications in conjunction with the specificity of biological recognition have been gaining interest recently. A mediator-free glucose biosensor, termed a third-generation biosensor, was fabricated by immobilizing glucose oxidase (GOD) directly onto a BDD electrode [29]. Glucose was determined in the absence of a mediator. The effects of the response of the biosensor in response to pH, applied potential, and cross-linking time were investigated in order to optimize the performance of the sensor. The biosensor responded to glucose within an analysis time of 5 s and provided acceptable results for lifetime, reproducibility, and measurement repeatability. The calibration curve was linear in the glucose concentration range of 6.67 × 10−5 to 2 × 10−3 M with a detection limit of 2.31 × 10−5 M. The alternative method of a protein immunosensor has been developed at boron-doped diamond (BDD) electrode material [30]. To construct the base of the immunosensor, oaminobenzoic acid (o-ABA) was electropolymerized at an electrode by cyclic voltammetry. After polymerization, the poly-o-ABA-modified BDD was characterized by scanning electron microscopy (SEM) and x-ray photoelectron spectroscopy (XPS). The XPS result found that carboxyl groups were formed at the electrode surface. The carboxyl groups were then used to covalently attach protein probes. The amperometric sensing of mouse IgG (MIgG) was selected as the model at the poly-o-ABA-modified BDD to compare to the poly-o-ABA-modified glassy carbon (GC) at the same condition. An antimouse IgG from goat (GaMIgG) was covalently immobilized at a poly-o-ABA-modified BDD electrode that used a sandwich-type alkaline phosphatase (ALP) catalyzing amperometric immunoassay with 2-phospho-L-ascorbic acid (AAP) as substrate. The ALP enzyme conjugated at the immunosensor can generate AAP to the electroactive species of ascorbic acid, which can be determined by amperometric detection. The signal was found to be proportional with the quantity of MIgG. It also was found that the dynamic range of 3 orders of magnitude (1–1000 ng mL−1 ) was obtained as displayed in Figure 7.2. The examples of surface oxidation and the deposition of metal or metal oxide particles, including the selective detection of dopamine (DA) in the presence of ascorbic acid (AA) on the bare and gold nanoparticle/polyelectrolyte-coated polystyrene colloid modified BDD electrode, were studied [31]. The cyclic voltammetric results indicated a high electrocatalytic activity and strong enhancement of the DA oxidation. The detection limit of DA was 8.0 × 10−7 M and the linear range was 5 × 10−6 M to 100 × 10−6 M in the presence of 1 × 10−3 M AA. Additionally, Shang et al. [32] reported a method for the selective detection of dopamine in the presence of 3,4-dihydroxyphenylalanine (L-DOPA), ascorbic acid, uric acid, and other dopamine metabolites using electropolymerized (poly-tyramine and polypyrrole-1-propionic acid) on a BDD electrode. A layer-by-layer film of tyramine and pyrrole-1-propionic acid (PPA) was fabricated subsequent to electropolymerization on a

162

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

2

R = 0.999 Y = 0.12851 + 0.021x

j / μA cm–2

20

10

g j (μA cm–2)

0 0

400 500 [MIgG] / ppb c b a 0.1

a

2.5

100

200 Time (s)

300

Figure 7.2 Amperograms of products (AA) after adding substrates (AAP) of ALP in Tris buffer solution (pH 8.5) at a poly-o-ABA-modified BDD immunosensor by various target MIgG concentrations (0, 1, 5, 100, 200, 500, and 1000 ng mL−1 (a–g). The potential applied was 0.4 V. The inset in the top is a calibration plot showing the correspondence between the changes in anodic peak current after subtracting the control current (0 ng mL−1 ; absence of MIgG) and the concentration of MIgG, and that below is a zoom (a) 0, (b) 1, (c) 5 ng mL−1 . (Reprinted with permission from Ref. 30.)

BDD electrode with an overall thickness of about 33 nm. The modified BDD electrode exhibited a rapid response to dopamine within 6 s and a detection limit of 50 nM with excellent reproducibility. Moreover, Hason et al. [33] described copper-enhanced label-free anodic stripping method for detection of guanine and adenine bases in acid-hydrolyzed DNA at anodically oxidized boron-doped diamond electrodes (AO-BDDE). In the presence of copper, the guanine oxidation signal was increased by about two orders of magnitude. Additionally, the proposed technique was demonstrated to be suitable for the determination of the purine (particularly guanine) content in DNA samples and was applied to magnetic bead-based DNA assays.

7.4

POLLUTANT COMPOUNDS

The application of BDD electrodes was extended to the investigation of the electrochemical behavior of different kinds of environmental pollutants. Starting with Chang and colleagues [34], they studied the electrochemical behavior of formaldehyde (FA) at BDD electrodes by cyclic voltammetry, electrochemical impedance spectroscopy (EIS), and linear scanning voltammetry (LSV). The cyclic voltammetric results indicated that the oxidation reaction of FA is influenced by the hydroxyl concentration in the solution. A linear response was achieved over the range of 10–100 mM. The differential capacitance from the EIS results indicated that the FA molecules were adsorbed at the BDD electrode surface at low potentials. The results of kinetic studies indicated that the adsorption of

7.4 POLLUTANT COMPOUNDS

163

FA molecules at the BDD electrode is the rate-determining step at low potentials. Next, Pedrosa, Machado, and Avaca [35] proposed the use of anodic square-wave voltammetry to investigate behaviors of 4-chlorophenol (4-CP) in aqueous solution at a BDD electrode. The results revealed the oxidation of 4-CP in a Britton-Robinson buffer solution pH 6.0 and a detection limit as low as 9.2 μg L−1 . Subsequently, Ivandini et al. [36] published the use of bare BDD electrodes to study the electrochemical oxidation of oxalic acid (OA) by cyclic voltammetry and flow injection analysis with amperometric detection. Hydrogen-terminated BDD electrodes exhibited well-defined oxidative peaks of oxalic acid over a wide pH range. A good linear response was observed for a concentration range from 50 nM to 10 μM with a detection limit of 0.5 nM. In 2007, the direct and simultaneous determination of phenol (Ph), hydroquinone (HQ), and 4-nitrophenol (4-NP) at unmodified BBD electrodes by voltammetry were investigated by the same group [37,38]. The oxidative peak currents of these three phenolic compounds were completely resolved at the BDD electrodes in acidic media. Each phenolic compound displayed a good linear relationship between the oxidation peak current and concentration over a rather wide range even though one or two of the other phenolic pollutants were appeared. This method is also promising for the determination of phenolic contaminants in real wastewater samples. The electrochemistry of the benzene oxidation process in aqueous solution at BDD electrodes was investigated [39]. The irreversible oxidation peak of benzene in 0.5 M H2 SO4 was found at 2.0 V versus Ag/AgCl at the BDD electrode by cyclic voltammetry. Hydroquinone, resorcinol, p-benzoquinone, catechol, and phenol were the main products generated during the oxidation process. The benzene oxidation was completed using a rotating BDD electrode and the potential was fixed at 2.5 V for 5 h. All products were measured at concentrations below 10−5 M. In addition, the electrochemical behavior of aniline was investigated at BDD electrodes by linear-sweep cathodic stripping voltammetry [40]. The results showed that a dimeric species (p-aminodiphenylamine and benzidine) formed by anodic oxidation of aniline in acidic media during the accumulation period is involved in the electrochemically reversible redox processes, with a detection limit of 1 μM. In the following work, Radovan and Manea [41] reported that the BDD electrode is a feasible alternative for the analytical determination of sodium diethyldithiocarbamate (DEDTC) in protic media, using cyclic voltammetry or chronoamperometry. Linear plots of current against concentration correlated with an anodic stepwise oxidation mechanism in delimited potential ranges with high correlation coefficients. The voltammetric behaviors of 3-nitrofluoranthene and 3-aminofluoranthene in mixed methanol-water solutions were examined at BDD thin-film electrodes by differential pulse voltammetry [42]. The optimal conditions were reported and limits of detection were 3 × 10−8 M of 3-nitrofluoranthene and 2 × 10−7 M of 3-aminofluoranthene. In addition, the boron-doped diamond electrode was the first developed for detecting the chemical oxygen demand (COD) by amperometric methods [43]. A linear range for the COD test was observed from 20 mg L−1 to 9000 mg L−1 COD and the detection limit was 7.5 mg L−1 COD. This BDD electrode was successfully employed to measure the COD of both synthetic and real wastewater samples from chemical manufacturing and pharmaceutical factories. This method is environmental friendly because no toxic substances are used. To increase the performance of analytical method, the BDD electrodes have also been used as detectors in a flow injection analysis (FIA) system for

164

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

the determination of COD [44]. The effect of several important experimental parameters, such as applied potentials, pH, flow rates, and supporting electrolyte concentrations, on the analytical performance was investigated. Under optimized conditions, the proposed method was applied to COD analysis in synthetic samples. Linear dynamic ranges from 2 mg L−1 to 175 mg L−1 COD and a detection limit of 1 mg L−1 were obtained. In 2008, Cinghita, Radovan, and Dascalu [45] investigated the electrochemical behavior of thioacetamide (TAA), a harmful presumptive pollutant in tap and wastewaters. The investigation was performed at a BDD electrode both in unbuffered and buffered 0.1 M Na2 SO4 solutions as the supporting electrolytes. The results showed a well-defined oxidation peak for TAA and good linearity of the amperometric signal versus concentration. Comparison experiments were carried out using glassy carbon electrodes, highlighting the excellent properties of the unmodified BDD electrode and its potential as an amperometric sensor for environmental applications. Next, Martinez-Huitle et al. [46] discussed the electrochemical oxidation of oxalic acid (OA) in the presence of NaCl and NaBr in alkaline media using BDD. Based on the results obtained, bromide was selected as a more suitable mediator of OA oxidation at BDD electrodes in comparison to chlorides. To perform the fully automatic system, a gas diffusion sequential injection system with amperometric detection using a BDD electrode was developed for the determination of sulfite [47]. The sample was mixed with an acidic solution to generate gaseous sulfur dioxide prior to its passage through the donor channel of the gas diffusion unit (GDU). The sulfur dioxide diffused through a PTFE hydrophobic membrane into a carrier solution of 0.1 M phosphate buffer (pH 8) + 0.1% (w/v) sodium dodecyl sulfate in the acceptor channel of the GDU and was converted into sulfite. The sulfite was carried to the electrochemical flow cell and detected directly by amperometry using a BDD electrode at 0.95 V versus Ag/AgCl. Sodium dodecyl sulfate was added to the carrier solution to prevent electrode fouling. This method was applicable in the concentration range of 0.2 to 20 mg L−1 SO3 2− with a detection limit of 0.05 mg L−1 SO3 2− . This method was successfully applied to the determination of sulfite in wines and the analytical results agreed well with those obtained by iodimetric titration. The relative standard deviations for the analysis of sulfite in wines were in the range of 1.0–4.1%. As mentioned previously, to overcome the limitation of unmodified BDD electrodes, surface modifying process is also utilized to prepare BDD electrodes for electroanalysis of variety pollutant compounds. For example, Roustom et al. [48] developed a new method for the synthesis of bimetallic nanoparticles (Au–Pt) on BDD substrates, and these modified electrodes were used to study oxygen reduction. Kondo et al. [49] reported on the use of photochemical modification of alyltriethylammonium bromide (ATAB) onto BDD thin film surfaces for the electrochemical detection of oxalate. The anodic current for oxalate at ATAB-modified BDD with both cyclic voltammetry and flow-injection analysis using amperometry was up to two times larger in comparison to an unmodified BDD electrode. These results may be due to the electrostatic interaction between the oxalate anion and the electrode surface. In addition, the stability of the electrochemical detection of oxalate was improved at the ATAB-BDD electrode compared to those of the unmodified electrode. In 2009, Hutton et al. [50] reported on the fabrication and use of platinum nanoparticle (NP)-modified polycrystalline boron-doped diamond (pBDD) disk electrodes as amperometric sensors for the determination of dissolved oxygen concentrations in aqueous solutions. The electrodes offer high precision over a pH range of 4.0–10.0.

7.5 HEAVY METALS

7.5

165

HEAVY METALS

Trace metals play an important role in many systems and biological processes. A range of metals is essential for the efficient growth and functioning of aquatic organisms. A potential application of boron-doped diamond electrodes for heavy metal determination in various samples was reported. In 2006, McGaw and Swain [51] reported the performance of the BDD electrode for the anodic stripping voltammetric determination of heavy metal ions (Zn2+, Cd2+ , Pb2+ , Cu2+ , Ag+ ) in comparison with Hg-coated glassy carbon. The linear dynamic range obtained from using the BDD electrode was three to four orders of magnitude (r > 0.995) similar to Hg-GC electrode. However, the BDD electrode provides a detection limit three to five times lower than that the Hg-GC electrode because the BDD electrode has a lower background and a higher signal/noise ratio. The detection limits were 50 ppb Zn2+ , 1.0 ppb Cd2+ , 5.0 ppb Pb2+ , 10 ppb Cu2+ , and 1.0 ppb Ag+ . The results demonstrated that BDD is a viable alternative electrode for metal analysis by anodic stripping voltammetry. In the similar concept, the BDD electrodes were also used to investigate the possibility of detecting trace levels of lead by linear-sweep anodic stripping voltammetry [52]. A low limit of detection (2 nM) was obtained in comparison to those by the other electrodes. At low pH values, the Cu2+ concentrations usually present in drinking water do not affect the measurement. Therefore, these methods are attractive and suitable for the determination of trace levels of lead in drinking and tap water. Furthermore, the determination of Zn2+ , Cd2+ , Pb2+ , and Cu2+ by differential pulse anodic stripping voltammetry at BDD electrodes at very low concentrations was reported [53]. Quantification was possible for the simultaneous determination of Zn2+ , Cd2+ , and Pb2+ despite the fact that the Zn2+ and Cd2+ peaks overlapped. Comparison experiments were performed using a glassy carbon electrode, and the BDD electrode provided a lower baseline current, wider potential range, higher sensitivity, and greater longevity than glassy carbon. In 2008, Seehra, Ranganathan, and Manivannan reported the determination of mercury in solutions by BDD electrodes using differential pulse anodic stripping voltammetry [54]. Gold was added to the solution for enhanced sensitivity. Impurity peaks from Cu2+ and Ag+ were identified by choosing the optimal deposition potential. The modified surfaces have been not only used extensively for the analysis of organic and biochemical compounds but some have also been used for inorganic species. Metal and metallic nanoparticles modified boron-doped diamond electrodes and their application to electrochemical analysis have been widely reported. For example, iridium-modified BDD electrodes were used for the electrochemical detection of As3+ by cyclic voltammetry and flow injection analysis with amperometric detection [55]. This method was also applied to the analysis of spiked arsenic in tap water containing a significant amount of various ion elements. Toghill et al. [56] demonstrated a bismuth nanoparticle modified boron-doped diamond (Bi-BDD) electrode for the simultaneous determination of Pb2+ and Cd2+ by square-wave anodic stripping voltammetry (SWASV). In situ plating was achieved using 0.1 mM Bi(NO3 )3 in 0.1 M HClO4 at pH 1.2. The detection limits obtained from this report were 1.9 μg L−1 and 2.3 μg L−1 for Pb2+ and Cd2+ , respectively. Mixtures of As3+ and As5+ at gold-modified diamond electrodes by stripping voltammetry were also electrochemically detected [57]. Good linear responses were observed for standards of As3+ and As5+ . Linear calibration curves were obtained in the concentration range of

166

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

100–1000 ppb As5+ in a mixture with 100 ppb As3+ (r 2 = 0.99) and in the concentration ranged of 5–30 ppb As3+ in a mixture with 100 ppb As5+ . Detection limits of 5 ppb and 100 ppb were achieved for As3+ and As5+ in a mixture with reproducibility of 7.5% and 8.4%, respectively. This method was used to measure arsenic contaminants in Yokohama tap water.

7.6

FOOD AND DIETARY CONTAMINANTS

Due to the frequent appearance of contaminant compounds from food sample/dietary/supplement matrices, the development of a highly sensitive detection method is needed. Therefore, the use of BDD electrodes as sensors for the electrochemical analysis of food contaminants has been widely published. In 2006, Juliao et al. [58] reported on the electrochemical behavior of nitrofurazone (NFZ) in a predominantly aqueous medium, in the absence and presence of glutathione (reduced form) (GSH), L-cysteine (Cys), and oxygen (O2 ) using BDD electrodes by voltammetry. A linear response was observed for NFZ in the concentration range of 9.9 × 10−7 M to 1.1 × 10−5 M at pH 8.0 with a sensitivity of 2.2 × 106 M cm−2 and a detection limit of 3.4 × 10−7 M. Codognoto et al. [59] demonstrated the examination of electrochemical property of pesticide carbaryl at a BDD electrode with square-wave voltammetry without any previous steps. A linear response was obtained for carbaryl detection from 2.5 × 10−6 M to 30.0 × 10−6 M in 0.1 M Na2 SO4 at pH 6.0 with a detection limit of 8.2 ± 0.2 μg L−1 . Flavonoids, which exhibit high antioxidant activities, were also analyzed in tea samples using BDD amperometric detection coupled to a flow injection system [60]. The experiments were performed at a fixed potential of 0.42 V versus Ag/AgCl with a flow rate of 2.5 mL min−1 . A Britton-Robinson buffer solution at pH 5.0 was used as the carrier stream. A good linear response was observed for a rutin concentration range of 0.1 × 10−4 M to 2.5 × 10−4 M with a detection limit of 7.7 × 10−6 M. In addition, this method was used to determine rutin in three different green teas (Camellia sinensis) with a relative standard deviation of 2.1% (n = 20). In 2008, Medeiros et al. [61] reported the use of square-wave voltammetry in conjunction with a cathodically pretreated BDD electrode for the determination of sodium cyclamate. A linear response was observed for sodium cyclamate in 0.5 M H2 SO4 solution over a concentration range of 5.0 × 10−5 M to 4.1 × 10−4 M and with a detection limit of 4.8 × 10−6 M. The RSD was smaller than 1.2%. In addition, this method was used to determine sodium cyclamate in several dietary products. The use of BDD thin film electrodes to study the electrochemical properties of chloramphenicol using cyclic voltammetry and flow injection analysis was demonstrated [62]. A linear curve for chloramphenicol over the range of 0.1 mM to 10 mM was obtained by cyclic voltammetry. Chloramphenicol was then analyzed by flow injection analysis at −0.7 V versus Ag/AgCl. A linear range of 0.1 μM to 50 μM and a limit of detection of 0.03 μM (S/N = 3) were obtained (see Figure 7.3). This method was successfully applied to the determination of chloramphenicol in sterile eye drops and milk samples using the standard addition method. The average recoveries of chloramphenicol were 98.0% and 93.9 to 103% in eye drops and spiked milk, respectively. N-nitrosamines (N-nitrosopyrrolidine, N-nitrosopiperidine and N-nitrosodiethylamine) in aqueous solutions were also examined at BDD electrodes using cyclic voltammetry

7.6 FOOD AND DIETARY CONTAMINANTS

167

–7.0 Peak current (μA)

(b)

–5.0 Current (μA)

–1.5

(a)

–6.0

(c)

–4.0

y = –0.0216x – 0.0437 R 2 = 0.9948

–1.0 –0.5 0.0 0

(d)

–3.0

(e) –2.0

10

20

30

40

50

60

Chloramphenicol concentration (μM)

(f) (g) (h) (i) (j) (k) (l) (m) (n)

–1.0 0.0 0

500

1000

1500

2000

2500

Time (s) Figure 7.3 Flow-injection analysis with amperometric detection results for various concentrations of chloramphenicol in 0.1 M phosphate buffer (pH 6) in 1% ethanol (a) 1000, (b) 500, (c) 250, (d) 100, (e) 50, (f) 25, (g) 10, (h) 5, (i) 2.5, (j) 1, (k) 0.5, (l) 0.25, (m) 0.1, and (n) 0.05 μM. The calibration graph is also shown in the inset. (Reprinted with permission from Ref. 62.)

[63]. An irreversible oxidation peak was observed at approximately 1.8 V versus Ag/AgCl for all N-nitrosamines. The maximal electrochemical response was obtained using the following square-wave voltammetry parameters: f = 250 Hz, Esw = 50 mV, and Es = 2 mV with a Britton-Robinson buffer solution as the electrolyte (pH 2.0). The detection and quantification limits determined for total N-nitrosamines were 6.0 × 10−8 M and 2.0 × 10−7 M, respectively. Furthermore, the determination of aspartame in dietary products samples without pretreatment was achieved using a BDD electrode [64]. A single irreversible oxidation peak at a potential of 1.6 V versus Ag/AgCl was obtained. A linear response over the range of 9.9 × 10−6 M to 5.2 × 10−5 M with a detection limit of 2.3 × 10−7 M was observed in Figure 7.4. The relative standard deviation (n = 5) was less than 0.2% for a 1.0 × 10−4 M aspartame solution. The proposed method yielded similar results to those obtained from the HPLC method at a 95% confidence level. In 2008, the simultaneous determination of aspartame and cyclamate in dietary products at a BDD electrode was developed [65]. Square-wave voltammetry gave separated oxidation peak potentials in binary mixtures of approximately 400 mV (see Figure 7.5). The detection limits for aspartame and cyclamate were 3.5 × 10−7 M and 4.5 × 10−6 M, respectively, while the relative standard deviations were 1.3% for aspartame and 1.1% for cyclamate. This electrochemical method is simple and highly selective, and can be applied to the determination of aspartame in dietary products. An electrochemical biosensor using a BDD electrode was developed and used for detecting o-nitrophenol released from o-nitrophenyl-β-d-galactopyranose, a reaction catalyzed by β-galactosidase, a marker of E . coli contamination in food [66]. Cyclic voltammogram of o-nitrophenol in a 50 mM phosphate buffer at pH 7.0 showed a well-defined oxidation peak at 0.93 V versus Ag/AgCl. This BDD sensor can be used directly without any surface modification. The enzyme was effectively induced by isopropyl-β-d-thiogalacto-pyranoside. The results showed a biphasic calibration plot

168

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

12

20

xxxx

8

10

6

9 8

4 2 0

I (μA)

11

10 I (μA)

25

1 2 3 4 5 6 [Aspartame]/10–5 mol L–1

15

7 6 5 4 3

10

2

1

5

0 1.3

1.4

1.5

1.6

1.7

1.8

1.9

E (V) vs. Ag/AgCl Figure 7.4 Square-wave voltammetric response of the BDD electrode for different aspartame concentrations: (1) 0; (2) 9.9 × 10−6 ; (3) 1.5 × 10−5 ; (4) 2.0 × 10−5 ; (5) 2.4 × 10−5 ; (6) 2.9 × 10−5 ; (7) 3.4 × 10−5 ; (8) 3.8 × 10−5 ; (9) 4.3 × 10−5 ; (10) 4.8 × 10−5 ; (11) 5.2 × 10−5 mol L−1 . Insert: Analytical curve for the oxidation process of aspartame. (Reprinted with permission from Ref. 64.)

with a linear range between 4 × 104 cells mL−1 and 2 × 105 cells mL−1 , and 2 × 105 and 6 × 106 cells mL−1 for the first and second regions, respectively. The detection limit was 4 × 104 cells mL−1 with a total analysis time of less than 1.5 h. This method is a suitable alternative for the highly sensitive detection of multiple pathogens simultaneously and can be adapted for field tests to rapidly detect traces of E . coli and other food-borne pathogens. Last, Andrade et al. [67] demonstrated the use of a multidimensional high-performance liquid chromatography method coupled with amperometric detection using a BDD electrode for the simultaneous determination of sulfamethoxazole and trimethoprim in bovine milk. Results showed good linearity in the concentration from 50 to 800 μg L−1 and from 25 to 400 μg L−1 for sulfamethoxazole and trimethoprim, respectively. The intraand interassay coefficients of variation were less than 10% for both drugs. It was also found that LOD values were 25.0 μg L−1 for sulfamethoxazole and 15.0 μg L−1 for trimethoprim. 7.7

MISCELLANEOUS

Several outstanding properties of BDD electrodes make them very attractive for use in many potential applications, including for pH measurements. The electrooxidation of ethylenediaminetetraacetic acid (EDTA) at a thin-film BDD polycrystalline diamond electrode was studied by cyclic voltammetry and amperometry [68]. Under optimal conditions, the linear responses were obtained for concentration ranges from 1.0 × 10−5 M to

7.7 MISCELLANEOUS

169

(a) 80

12 10

60

Ip (μA)

70

8 6

I (μA)

4

50

2 0

40

0

1

2 3 4 –5 –1 [Aspartame]/ 10 mol L

5

30 20

j

10

a

0 1.2

1.4

1.6

1.8

2.0

2.2

E (V) vs. Ag/AgCl (b) 60 30

50 Ip (μA)

40

l 20

k

I (μA)

10

30

0 0

20

1 2 3 –4 –1 [Cyclamate]/ 10 mol L

4

10 0 1.4

1.6

1.8

2.0

E (V) vs. Ag/AgCl Figure 7.5 (a) SW voltammograms for various concentrations of aspartame at a fixed concentration of cyclamate (3.0 × 10−4 mol L−1 in 0.5 mol L−1 H2 SO4 ). Aspartame concentrations (a–j): 5.0 × 10−6 to 5.0 × 10−5 mol L−1 . (b) SW voltammograms for various concentrations of cyclamate at a fixed concentration of aspartame (1.0 × 10−4 mol L−1 in 0.5 mol L−1 H2 SO4 ). Cyclamate concentrations (k–l): 5.0 × 10−5 to 5.0 × 10−4 mol L−1 . Insets are the corresponding analytical curves for the peak current corresponding to the oxidation process of aspartame or cyclamate. (Reprinted with permission from Ref. 65.)

5 × 10−4 M and with a detection limit of 1 × 10−6 M, showing that the BDD electrodes can be used in the quantitative determination of EDTA with a low background current and detection limit. Mitani and Einaga [69] developed the simplest method using BDD electrodes for the analyses of acid concentrations in acidic solutions. Linear sweep voltammetry using BDD electrodes was used to measure the proton concentration. These methods were also applied in vivo for measurements unable of being conducted by conventional glass electrodes.

170

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

For surface modification, a method for covalent immobilization onto an amineterminated BDD surface was described in 2007 for the determination of phenolic compounds by Zhou and colleagues [70]. The hydrogen-terminated BDD surface was first functionalized by photochemically linking vinyl groups of allylamine to produce a covalently linked amine-terminated active BDD surface. Then the tyrosinase was immobilized onto the active BDD surface using the carbodiimide coupling reaction. The amperometric response was measured as a function of concentration of phenolic compounds in 0.1 M phosphate buffer solution at pH 6.5. The resulted indicated that the BDD enzyme electrode exhibits a good performance in terms of linear dynamic range, sensitivity, and long-term stability to phenolic derivatives as well as the efficient covalent bonding of the enzyme to the substrate and the electrochemical stability of BDD electrode. The tyrosinase-aminophenyl-modified BDD electrode was developed for the detection of phenolic compounds using amperometry. Moreover, Tyrosinase was covalently immobilized on an aminophenyl-modified BDD (AP–BDD) surface via carbodiimide coupling [71]. The effect of the oxygen level, phenolic compound diffusion, and pH of the solution were studied. The Tyr–AP–BDD electrode showed a linear response from 1 to 200 μM, 1 to 200 μM, and 1 to 250 μM and sensitivities of 232.5, 636.7, and 385.8 mA M−1 cm−2 for phenol, p-cresol, and 4-chlorophenol, respectively. Next, the electrochemistry and electrocatalytic activity of cytochrome C (Cyt C) covalently immobilized on a boron-doped nanocrystalline diamond electrode were studied [72]. The linear response to H2 O2 detection was observed over a concentration range of 1 μM to 450 μM, and the detection limit was 0.7 μM. This method showed excellent electrocatalytic performance in terms of fast response, low detection limit, and high stability toward the reduction of H2 O2 . Afterward, Geng et al. [73] described the SiO2 /Cyt C/SiO2 sandwich on the pretreated BDD electrode. Cyt C was immobilized between the SiO2 gel membranes via electrostatic interactions and by carefully controlling the pH of the solution. The SiO2 interlayer was suggested to play a significant role in the sandwich structure of the SiO2 /Cyt C/SiO2 /BDD electrode. The immobilized Cyt C maintained high stability and good electrochemical performance. This electrode was applied to monitor nitrite, and the oxidation current was proportional to the concentration of nitrite in the range of 1.0 × 10−6 M to 1.0 × 10−3 M with a detection limit of 0.5 μM. Finally, the summarization of analytical applications from the use of BDD electrodes for the determination of wide range of analytes is shown in Table 7.1.

7.8

CONCLUSIONS

Since their introduction into the electroanalytical field in the early twentieth century, boron-doped diamond electrodes have become more and more popular. Their unique properties distinguish them from conventional electrode materials, and allow for many electrochemical processes to become more attractive, simple, or even possible. The BDD electrode can be modified with various species such as metals and functional molecules, including biomolecules such as enzymes and nucleic acids. These contribute to their continued progress in new electroanalytical applications. In the future, we expect to see the broad introduction of diamond electrodes in numerous applications.

171

Cyclic and square-wave voltammetry Differential pulse and square-wave voltammetry Cyclic, differential pulse and square-wave voltammetry Cyclic, linear sweep, differential pulse, square-wave voltammetry square-wave voltammetry

Lidocaine

differential pulse voltammetry

Cyclic voltammetry Cyclic and differential pulse voltammetry and chronoamperometry

Sulfonamides (sulfamethoxazole and trimethoprim)

Procaine hydrochloride L-ascorbic acid, Acetaminophen

Sulfonamides (sulfadiazine and sulfamethoxazole)

Pefloxacin

Fluvastatin

Atorvastatin calcium

Flow injection system with pulsed amperometric detection

Three beta-agonists (salbutamol, terbutaline, and clenbuterol)

Technique

Electroanalytical applications of diamond films.

1. Pharmaceutical compound

Analyte

TABLE 7.1

5–200 μM 0.01–0.1 mM

3.65 μg L−1 (sulfamethoxazole) and 3.92 μg L−1 (trimethoprim) 0.5 μM —

2.19 × 10−6 M (sulfadiazine), 1.15 × 10−6 M (sulfamethoxa-zole)

8.01 × 10−6 − 1.19 × 10−4 M (sulfadiazine), 6.10 × 10−6 − 6.01 × 10−5 M (sulfamethoxazole) —

2 × 10−6 − 2 × 10−4 M

2008 2008

2009

2008

2008

2007

2007

2007

10 μg L−1 2.27 × 10−7 1.31 × 10−7 1.37 × 10−7 1.44 × 10−7 4.65 × 10−7 5.78 × 10−7

2006

Year

—

M (DPV) M (SWV) M (DPV), M (SWV) M (DPV) M (SWV)

Limit of detection

9.65 × 10−7 − 3.86 × 10−5 M 1 × 10−6 − 6 × 10−4 M

0.5–100 μM (salbutamol), 1.0–100 μM (terbutaline), 0.5–50 μM (clenbuterol) —

Linear range

AO-BDDE

Comments

(continued overleaf )

16 17

15

14

13

12

11

10

9

Ref.

172 Cyclic and square-wave voltammetry Square-wave voltammetry

Promethazine hydrochloride

Acetylsalicylic acid (ASA)

Biphenyl amino derivatives (2-aminobiphenyl, 3-aminobiphenyl, and 4-aminobiphenyl)

Dopamine in mouse brain Native and thermally denatured fish DNA 2’-deoxyguanosine Purine and pyrimidine bases Proteins

Tryptophan (Trp) and tyrosine (Tyr) Dopamine

Cyclic voltammetry HPLC amperometric detection Cyclic voltammetry and flow injection analysis Amperometric

Differential pulse voltammetry Differential pulse voltammetry Differential pulse voltammetry Cyclic voltammetry

2. Biomolecules or Biological Compounds

Cyclic voltammetry and chronoamperometry

Technique

L-ascorbic acid, Acetaminophen

Analyte

TABLE 7.1 (Continued)

10 μg mL−1 (BSA), 100 μg mL−1 (IAP) 2 × 10−7 M (2-AB), 3.2 × 10−7 M (3-AB), 5.1 × 10−7 M (4-AB)

10–100 μg mL−1 (BSA), 200–800 μg mL−1 (IAP) 4 × 10−7 −100 × 10−7 M (2-AB), 2 × 10−7 −100 × 10−7 M (3-AB and 4-AB)

—

— 26.3–162.1 nM

—

0.5 nM − 100 μM

— 0.1–10 μM

50 nM

5 × 10−6 M − 1 × 10−4 M

0.8 and 0.86 μM (L-ascorbic acid), 0.97 and 1.42 μM (Acetaminophen) 2.66 × 10−8 M (peak 1), 4.61 × 10−8 M (peak 2) 2.0 × 10−6 M

Limit of detection

1 × 10−5 M (Trp), 1 × 10−6 M (Tyr) 1.1 × 10−6 M

—

2.50 × 10−6 − 1.05 × 10−4 M

5.96 × 10−7 − 4.76 × 10−6 M

0.01–0.1 mM

Linear range

2009

2008

2009 2007

2007

2007

2007

2006

2009

2008

2008

Year

28

27

25 26

24

23

22

21

20

19

18

Ref.

Comments

173

Cyclic voltammetry

Cyclic voltammetry

Dopamine

Dopamine

Phenol (Ph), hydroquinone (HQ) and 4-nitrophenol (4-NP)

4-chlorophenol Oxalic acid

Formaldehyde

3. Pollutant Compounds Cyclic voltammetry and electrochemical impedance spectroscopy square-wave voltammetry Cyclic voltammetry and flow injection analysis Differential pulse voltammetry

Amperometry

Protein

Guanine and adenine bases

Amperometry

Glucose

8 × 10−7 M

5 × 10−6 − 100 × 10−6 M

2006 2006

9.2 μg L−1 0.5 nM 1.82 × 10−6 M (Ph), 1.67 × 10−6 M (HQ), 1.44 × 10−6 M (4-NP) for DPV

0.7–4.0 × 10−5 M 50 nM−10 μM 5 × 10−5 − 1.4 × 10−3 M (Ph), 5 × 10−5 − 7 × 10−3 M (HQ), 5 × 10−5 − 3 × 10−3 M (4-NP) for DPV

2007

2006

2008

—

25 fmol

2009

2008

2008

2006

10–100 mM

—

50 nM

0.3 ng mL−1

1–1000 ng mL−1

—

2.31 × 10−5 M

6.67 × 10−5 − 2 × 10−3 M

37

35 36

34

33

32

31

30

29

(continued overleaf )

GOD/BDD electrode poly-o-ABAmodified BDD AuNPs/polyelec trolyte/BDD electrode Tyramine/PPA/BDD electrode AO-BDDE

174

Oxalic acid Sulfite Oxygen reduction

Thioacetamide

Chemical oxygen demand (COD) COD

Sodium diethyldithiocarbamate 3-nitrofluoranthene and 3-aminofluoranthene

Benzene Aniline

Analyte

TABLE 7.1 (Continued)

Cyclic voltammetry Amperometry

Cyclic voltammetry and flow injection analysis Chronoamperometry

Amperometry

Differential pulse voltammetry

Cyclic voltammetry Linear-sweep cathodic stripping voltammetry Cyclic voltammetry and chronoamperometry

Cyclic voltammetry

Technique

1 mg L−1

2–175 mg L−1 0.02–0.06 mM (in supporting electrolyte 0.1 M Na2 SO4 ) 0.005–0.06 mM (in supporting electrolyte BR2 pH 2.16) — 0.2–20 mg L−1 —

— 0.05 mg L−1 —

0.84 μM

—

3 × 10−8 M (3nitrofluoran-thene), 2 × 10−7 M (3-aminofluo ranthene) 7.5 mg L−1

2 × 10−8 − 1 × 10−6 M (3-nitrofluo ranthene), 2 × 10−7 −1 × 10−5 M (3-aminofluorathnene) 20–9000 mg L−1

2007

0.3 μM

—

2008 2008 2008

2008

2009

2007

2007

2007 2007

8.2 × 10−6 M (Ph), 1.2 × 10−5 M (HQ), 1.1 × 10−5 M (4-NP) for CV — 1 μM

5 × 10−5 − 2 × 10−3 M (Ph), 5 × 10−5 − 1 × 10−2 M (HQ and 4-NP) for CV — 1–5 μM

Year

Limit of detection

Linear range

46 47 48

45

44

43

42

41

39 40

38

Ref.

Py/AuNPs/BDD electrode

Highest Sensitive

Comments

175

Mixtures of As3+ and As5+

Cadmium(II), lead(II)

Arsenite

Hg

Differential pulse anodic stripping voltammetry Cyclic voltammetry and flow injection analysis Square-wave anodic stripping voltammetry Stripping voltametry

Linear-sweep anodic stripping and anodic stripping voltammetry Differential pulse anodic stripping voltammetry

Pb

Zn, Cd, Pb, and Cu

Anodic stripping voltammetry

Cyclic voltammetry and flow injection analysis with amperometry Amperometry and cyclic voltammetric

Zn, Cd, Pb, Cu and Ag

4. Heavy metal

Dissolved oxygen concentration

Oxalate

100–1000 ppb (As5+ ), 5–30 ppb (As3+ )

1.9 μg L−1 (Pb2+ ), 2.3 μg L−1 (Cd2+ ) 100 ppb (As5+ ), 5 ppb (As3+ )

20 nM

0.1–100 μM —

1.6 ppb (Zn), 0.36 ppb (Cd), 1.15 ppb (Pb), 0.9 ppb (Cu) 10 ppt

2 nM

50 ppb (Zn), 1 ppb (Cd), 5 ppb (Pb), 10 ppb (Cu), 1 ppb (Ag)

—

32 nM

5–20 ppb (Zn), 1.2–25 ppb (Cd), 3.8–45 ppb (Pb), 3–20 ppb (Cu) —

50–1000 ppb (Zn), 1–1000 ppb (Cd), 5–1000 ppb (Pb), 10–1000 ppb (Cu), 1–1000 ppb (Ag) 2–100 nM

—

0.8–100 mM

2008

2008

2008

2008

2007

2006

2006

2009

2008

57

56

55

54

53

52

51

50

49

(continued overleaf )

BiNPs/BDD electrode Au/BDD electrode

Ir/BDD electrode

Allyltriethylammonium Bromide PtNPs/BDD electrode

176 Flow injection analysis with amperometric detection Square-wave voltammetry Cyclic voltammetry and flow injection analysis with amperometric detection Cyclic and square-wave voltammetry Square-wave voltammetry Square-wave voltammetry

Cyclic voltammetry

HPLC-Amperometric detection

Flavonoid

Aspartame Aspartame, cyclamate

E. Coli

Sulfonamides (sulfamethoxazole and trimethoprim)

N-nitrosamines

Sodium cyclamate Chloramphenicol

Cyclic voltammetry Square-wave voltammetry

Technique

Nitrofurazone Carbaryl

5. Food and dietary contaminants

Analyte

TABLE 7.1 (Continued)

2.3 × 10−7 M 3.5 × 10−7 M (aspartame), 4.5 × 10−6 M (cyclamate) 4 × 104 cells mL−1

9.9 × 10−6 − 5.2 × 10−5 M 5.0 × 10−6 − 5.0 × 10−5 M (aspartame), 5.0 × 10−5 − 5.0 × 10−4 M (cyclamate) 4 × 104 − 2 × 105 cells mL−1 (first regions), 2 × 105 − 6 × 106 cells/mL (second regions) 50–800 μg L−1 (sulfamethoxazole) and 25–400 μg L−1 (trimethoprim)

25 μg L−1 (sulfamethoxazole) and 15 μg L−1 (trimethoprim)

6.0 × 10−8 M

2 × 10−6 − 1.36 × 10−5 M

67

66

2008

2009

64 65

63

61 62

60

58 59

Ref.

2007 2008

2008

2008 2008

2006

7.7 × 10−6 M 4.8 × 10−6 M 0.03 μM

2006 2006

Year

3.4 × 10−7 M 8.2 ± 0.2 μg L−1

Limit of detection

5.0 × 10−5 − 4.1 × 10−4 M 0.1 mM to 10 Mm for CV 0.1 μM to 50 μM for FI-amperometry

9.9 × 10−7 − 1.1 × 10−5 M 2.5 × 10−6 − 30.0 × 10−6 M 0.1–2.5 × 10−4 M

Linear range

Comments

177

Cyclic voltammetry and amperometry Linear sweep voltammetry Cyclic voltammetry

Amperometry

Cyclic voltammetry Cyclic and differential pulse voltammetry

Ethylenediaminetetraacetic acid (EDTA) pH in acidic solutions Phenolic compound

Phenolic compound

H2 O2

Nitrite

6. Miscellaneous

0.7 μM 0.5 μM

1.0 × 10−6 − 1.0 × 10−3 M

— 1.0 μM (phenol), 0.5 μM (p-cresol), 0.8 μM (4-cholorophenol) 0.2 μM (phenol), 0.1 μM (p-cresol and 4-cholorophenol)

— 1–175 μM (phenol), 1–200 μM (p-cresol and 4-cholorophenol) 1–200 μM (phenol and p-cresol) 1–250 μM, (4-chlorophenol) 1–450 μM

1 × 10−6 M

1.0 × 10−5 − 5 × 10−4 M

2008

2008

2006

2009 2006

2008

73

72

71

69 70

68

Cyt c/BDD electrode SiO2 /Cyt c/SiO2 / BDD electrode

Tyrosinase/BDD electrode

Tyrosinase/BDD electrode

178

7.9

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

ACKNOWLEDGMENTS

The authors would like to thank the Thailand Research Fund and Faculty of Science, Chulalongkorn University.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

A. Fujishima, T.N. Rao, D.A. Tryk, Electrochim. Acta 2000, 45 , 4683–4690. T.N. Rao, A. Fujishima, Diam. Rel. Mat . 2000, 9 , 384–389. A. Fujishima, T.N. Rao, Diam. Rel.Mat . 2001, 10 , 1799–1803. R.G. Compton, J.S. Foord, F. Marken, Electroanalysis 2003, 15 , 1349–1363. A. Fujishima, in Diamond Electrochemistry (Y. Einaga, T.N. Rao, D.A. Tryk, eds.), Elsevier, Tokyo, 2005. O. Chailapakul, W. Siangproh, D.A. Tryk, Sens. Lett . 2006, 4 , 99–119. O. Chailapakul, E. Popa, H. Tai, B.V. Sarada, D.A. Tryk, A. Fujishima, Electrochem. Commun. 2000, 2 , 422–426. W. Siangproh, in Recent Advances in Analytical Electrochemistry (D. A. Tryk, O. Chailapakul, eds.), Transworld Research Network, India, 2007, pp. 167–197. C. Karuwan, T. Mantim, P. Chaisuwan, P. Wilairat, K. Grudpan, P. Jittangprasert, Y. Einaga, O. Chailapakul, L. Suntornsuk, O. Anurukvorakun, D. Nacapricha, Sensors 2006, 6 , 1837–1850. R.T.S. Oliveira, G.R. Salazar-Banda, V.S. Ferreira, S.C. Oliveira, L.A. Avaca, Electroanalysis 2007, 19 , 1189–1194. B.D. Topal, B. Uslu, S.A. Ozkan, Comb. Chem. High Throughput Screening 2007, 10 , 571–582. B.D. Topal, S. Tuncel, B. Uslu, S.A. Ozkan, Diam. Rel. Mat . 2007, 16 , 1695–1704. B. Uslu, B.D. Topal, S.A. Ozkan, Talanta 2008, 74 , 1191–1200. C.D. Souza, O.C. Braga, I.C. Vieira, A. Spinelli, Sens. Actuators B 2008, 135 , 66–73. L.S. Andrade, R.C. Rocha-Filho, Q.B. Cass, O. Fatibello-Filho, Electroanalysis 2009, 12 , 1275–1280. M. Wei, Y.L. Zhou, J.F. Zhi, D.G. Fu, Y. Einaga, A. Fujishima, X.M. Wang, Z.Z. Gu, Electroanalysis 2008, 20 , 137–143. C. Radovan, C. Cofan, D. Cinghita, Electroanalysis 2008, 20 , 1346–1353. C. Cofan, C. Radovan, Sensors 2008, 8 , 3952–3969. F.W.P. Ribeiro, A.S. Cardoso, R.R. Portela, J.E.S. Lima, S.A.S. Machado, P. de Lima, D. De Souza, A.N. Correia, Electroanalysis 2008, 20 , 2031–2039. E.R. Sartori, R.A. Medeiros, R.C. Rocha, O. Fatibello, J. Braz. Chem. Soc. 2009, 20 , 360–366. G.H. Zhao, Y. Qi, Y. Tian, Electroanalysis 2006, 18 , 830–834. G.H. Zhao, M.F. Li, M.L. Li, Cent. Eur. J. Chem. 2007, 5 , 1114–1123. A. Suzuki, T.A. Ivandini, K. Yoshimi, A. Fujishima, G. Oyama, T. Nakazato, N. Hattori, S. Kitazawa, Y. Einaga, Anal. Chem. 2007, 79 , 8608–8615. A. Apilux, M. Tabata, O. Chailapakul, Bioelectrochemistry 2007, 70 , 435–439. E. Fortin, E. Vieil, P. Mailley, S. Szunerits, T. Livache, Anal. Chem. 2007, 79 , 3741–3746. T.A. Ivandini, K. Honda, T.N. Rao, A. Fujishima, Y. Einaga, Talanta 2007, 71 , 648–655. M. Chiku, T.A. Ivandini, A. Kamiya, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2008, 612 , 201–207.

REFERENCES

179

28. K. Peckova, K. Jandova, L. Maixnerova, G.M. Swain, J. Barek, Electroanalysis 2009, 21 , 316–324. 29. W. Jing, Q. Yang, Anal. Bioanal. Chem. 2006, 385 , 1330–1335. 30. A. Preechaworapun, A. Suzuki, A. Fujishima, O. Chailapakul, Y. Einaga, Anal. Chem. 2008, 80 , 2077–2083. 31. M. Wei, L.G. Sun, Z.Y. Xie, J.F. Zhi, A. Fujishima, Y. Einaga, D.G. Fu, X.M. Wang, Z.Z. Gu, Adv. Func. Mater. 2008, 18 , 1608–1608. 32. F.J. Shang, Y.L. Liu, S. Hrapovic, J.D. Glennon, J.H.T. Luong, Analyst 2009, 134 , 519–527. 33. S. Hason, H. Pivonkova, V. Vetterl, M. Fojta, Anal. Chem. 2008, 80 , 2391–2399. 34. C.C. Chang, L.C. Chen, S.J. Liu, H.J. Tien, H.C. Chang, J. Chin. Chem. Soc. 2006, 53 , 839–844. 35. V.A. Pedrosa, S.A.S. Machado, L.A. Avaca, Anal. Lett . 2006, 39 , 1955–1965. 36. T.A. Ivandini, T.N. Rao, A. Fujishima, Y. Einaga, Anal. Chem. 2006, 78 , 3467–3471. 37. G.H. Zhao, Y.T. Tang, M.C. Liu, Y.Z. Lei, X.E. Xiao, Chin. J. Chem. 2007, 25 , 1445–1450. 38. Y.Z. Lei, G.H. Zhao, M.C. Liu, X. Xiao, Y.T. Tang, D.M. Li, Electroanalysis 2007, 19 , 1933–1938. 39. R.T.S. Oliveira, G.R. Salazar-Banda, M.C. Santos, M.L. Calegaro, D.W. Miwa, S.A.S. Machado, L.A. Avaca, Chemosphere 2007, 66 , 2152–2158. 40. T. Spataru, N. Spataru, A. Fujishima, Talanta 2007, 73 , 404–406. 41. C. Radovan, F. Manea, Electroanalysis 2007, 19 , 91–95. 42. K. Cizek, J. Barek, J. Fischer, K. Peckova, J. Zima, Electroanalysis 2007, 19 , 1295–1299. 43. H.B. Yu, H. Wang, X. Quan, S. Chen, Y.B. Zhang, Electrochem. Commun. 2007, 9 , 2280–2285. 44. H.B. Yu, C.J. Ma, X. Quan, S. Chen, H.M. Zhao, Environ. Sci. Technol. 2009, 43 , 1935–1939. 45. D. Cinghita, C. Radovan, D. Dascalu, Sensors 2008, 8 , 4560–4581. 46. C.A. Martinez-Huitle, S. Ferro, S. Reyna, M. Cerro-Lopez, A. De Battisti, M.A. Quiroz, J. Braz. Chem. Soc. 2008, 19 , 150–156. 47. C. Chinvongamorn, K. Pinwattana, N. Praphairaksit, T. Imato, O. Chailapakul, Sensors 2008, 8 , 1846–1857. 48. B.E. Roustom, G. Sine, G. Foti, C. Comninellis, J. Appl. Electrochem. 2007, 37 , 1227–1236. 49. T. Kondo, Y. Niwano, A. Tamura, T.A. Ivandini, Y. Einaga, D.A. Tryk, A. Fujishima, T. Kawai, Electroanalysis 2008, 20 , 1556–1564. 50. L. Hutton, M.E. Newton, P.R. Unwin, J.V. Macpherson, Anal. Chem. 2009, 81 , 1023–1032. 51. E.A. McGaw, G.M. Swain, Anal. chim. Acta 2006, 575 , 180–189. 52. D. Dragoe, N. Spataru, R. Kawasaki, A. Manivannan, T. Spataru, D.A. Tryk, A. Fujishima, Electrochem. Acta 2006, 51 , 2437–2441. 53. O. El Tall, N. Jaffrezic-Renault, M. Sigaud, O. Vittori, Electroanalysis 2007, 19 , 1152–1159. 54. M.S. Seehra, S. Ranganathan, A. Manivannan, Anal. Lett . 2008, 41 , 2162–2170. 55. T.A. Ivandini, R. Sato, Y. Makide, A. Fujishima, Y. Einaga, Anal. Chem. 2006, 78 , 6291–6298. 56. K.E. Toghill, G.G. Wildgoose, A. Moshar, C. Mulcahy, R.G. Compton, Electroanalysis 2008, 20 , 1731–1737. 57. D. Yamada, T.A. Ivandini, M. Komatsu, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2008, 615 , 145–153. 58. M.S.D. Juliao, E.I. Ferreira, N.G. Ferreira, S.H.P. Serrano, Environ. Sci. Technol. 2006, 41 , 5080–5086.

180

ELECTROANALYTICAL APPLICATIONS OF DIAMOND FILMS

59. L. Codognoto, S.T. Tanimoto, V.A. Pedrosa, H.B. Suffredini, S.A.S. Machado, L.A. Avaca, Electroanalysis 2006, 18 , 253–258. 60. V.A. Pedrosa, A.R. Malagutti, L.H. Mazo, L.A. Avaca, Anal. Lett . 2006, 39 , 2737–2748. 61. R.A. Medeiros, A.E. de Carvalho, R.C. Rocha, O. Fatibello, Quim. Nova 2008, 31 , 1405–1409. 62. S. Chuanuwatanakul, O. Chailapakul, S. Motomizu, Anal. Sci . 2008, 24 , 493–498. 63. R.T.S. de Oliveira, G.R. Salazar-Banda, S.A.S. Machado, L.A. Avaca, Electroanalysis 2008, 20 , 396–401. 64. R.A. Medeiros, A.E. de Carvalho, R.C. Rocha, O. Fatibello, Anal. Lett . 2007, 40 , 3195–3207. 65. R.A. Medeiros, A.E. de Carvalho, R.C. Rocha-Filho, O. Fatibello-Filho, Talanta 2008, 76 , 685–689. 66. E. Majid, K.B. Male, J.H.T. Luong, J. Agric. Food. Chem. 2008, 56 , 7691–7695. 67. L.S. Andrade, M.C. de Moraes, R.C. Rocha-Filho, O. Fatibello-Filho, Q.B. Cass, Anal. chim. Acta. 2009, 654 , 127–132. 68. A.E. Denisova, Y.V. Pleskov, Russ. J. Electrochem. 2008, 44 , 1083–1085. 69. N. Mitani, Y. Einaga, J. Electroanal. Chem. 2009, 626 , 156–160. 70. Y.L. Zhou, R.H. Tian, J.F. Zhi, Biosens. Bioelect . 2007, 22 , 822–828. 71. Y.L. Zhou, J.F. Zhi, Electrochem. Commun. 2006, 8 , 1811–1816. 72. Y.N. Zhou, J.F. Zhi, Y.S. Zou, W.J. Zhang, S.T. Lee, Anal. Chem. 2008, 80 , 4141–4146. 73. R. Geng, G.H. Zhao, M.C. Liu, M.F. Li, Biomaterials 2008, 29 , 2794–2801.

8 Cathodic Pretreatment of Boron-Doped Diamond Electrodes and Their Use in Electroanalysis Leonardo S. Andrade, Giancarlo R. Salazar-Banda, Romeu C. Rocha-Filho, and Orlando Fatibello-Filho

8.1

INTRODUCTION

Diamond electrodes have been a subject of investigation and application ever since 1987, when Pleskov et al. [1] reported pioneering electrochemical measurements carried out on undoped diamond films. Most of the early work on diamond electrodes, both fundamental and applied, was summarized in the book edited by Fujishima et al. [2], in book chapters by Angus, Pleskov, and Eaton [3] and by Swain [4,5], and in short reviews on boron-doped diamond (BDD) sensors by Chailapakul, Siangproh, and Tryk [6], Chen [7], and Park et al. [8]. More recently, Peckov´a, Musilov´a, and Barek [9] reviewed the voltammetric determination of organic substances using BDD electrodes, whereas Luong, Male, and Glennon [10] reviewed their functionalization and analytical applications. The electrochemical behavior of BDD electrodes depends on their physical, chemical, and electronic properties, which can be significantly affected by the surface termination (hydrogen, oxygen, and others). Microcrystalline BDD films currently prepared by

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

181

182

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

chemical vapor deposition (CVD) methods are hydrogen terminated (HT-BDD). Many redox couples present relatively high electron transfer rates on HT-BDD. In this sense, in 1997, in one of the first efforts to use BDD thin-film electrodes in electroanalysis, Jolley et al. [11] reported reversible to quasi-reversible electron transfer kinetics for the hexacyanoferrate(II)/hexacyanoferrate(III) [Fe(CN)6 4−/3− ] redox couple on a CVD-prepared BDD electrode that did not undergo any surface pretreatment before use. Afterward, Granger and Swain [12] reported on the influence of oxygen and hydrogen surface terminations of BDD electrodes (as deposited and obtained by plasma treatments) on the electron transfer involving Fe(CN)6 4−/3− and other redox couples. Since the electron transfer was much faster on HT-BDD electrodes, the researchers concluded that the Fe(CN)6 4−/3− redox reaction proceeds through a specific surface interaction available on the hydrogen-terminated surface and that such surface interactions appear to be blocked on the oxygen-terminated surface. However, they observed only a quasireversible response (Ep = 70 mV) for the hydrogen-terminated surface. The surface termination on BDD surfaces is usually generated by electrochemical methods (hydrogen evolution to produce hydrogen terminations; oxygen evolution for oxygen terminations) or RF-plasma treatment (H and O terminations) among others [13–28]. Hydrogen-terminated surfaces are hydrophobic with a negative electron affinity and are highly conductive [29–32], whereas oxygen-terminated ones are hydrophilic, with a positive electron affinity, and present a low conductivity [18,33,34]. As a consequence, the charge transfer rate for some redox couples, including Fe(CN)6 4−/3− , can significantly change with the surface termination on the BDD electrode. Thus, many authors have investigated the charge transfer mechanism and kinetics of several redox couples on BDD electrodes, using, for this purpose, physical and electrochemical methods [12,18, 35–45]. Here, we will concentrate on the cathodic pretreatment of BDD electrodes, which produces HT-BDD electrodes. First, investigations on the effect of the cathodic pretreatment of BDD electrodes on their electrochemical properties will be presented. Second, the use of HT-BDD electrodes on the determination of specific analytes in different matrixes will be reviewed. Third, the effect of the electrochemical pretreatments on the deposition of metals on BDD electrodes will be briefly reviewed.

8.2 CATHODIC PRETREATMENT OF CONDUCTIVE DIAMOND FILMS Despite the fact that enhanced electron transfer kinetics for some redox couples and high conductivity for HT-BDD surfaces (as-prepared or obtained by plasma treatments) were reported since the latter 1990s [12], in 2004, Suffredini et al. [18] called attention to the enhanced electrochemical response of BDD electrodes brought on by a cathodic surface pretreatment, for a variety of reactions. This research was inspired by previous observations in the analytical detection of pentachlorophenol (PCP) and 4-chlorophenol (4-CP) using square-wave voltammetry (SWV) at cathodically [−3.0 V versus Ag/AgCl (3.0 mol L−1 KCl) for 30 s] pretreated BDD electrodes [43,45], when excellent analytical performances were obtained. Thus, Suffredini et al. [18] compared the effect of anodic and cathodic electrochemical pretreatments [applying ±3.0 V versus Ag/AgCl (3.0 mol L−1 KCl), for 30 min, in 0.5 mol L−1 H2 SO4 ] on the electrochemical response

8.2 CATHODIC PRETREATMENT OF CONDUCTIVE DIAMOND FILMS

183

(a) 20

I (μA)

15

10

5

0 0.5

0.6

0.7 0.8 E (V) vs Ag/AgCl

0.9

1.0

8 (b)

I (μA)

6

4

2

0

0.2

0.4

0.6 0.8 E (V) vs Ag/AgCl

1.0

1.2

Figure 8.1 Cyclic voltammetry (ν = 0.05 V s−1 ) on BDD (A = 0.62 cm2 ) for 5.0 × 10−5 mol L−1 pentachlorophenol (a) and 4-chlorophenol (b) in a 0.1 mol L−1 Britton-Robson buffer (pH 5.5), after anodic (dotted lines) and cathodic pretreatments (full lines). (Reprinted with permission from Ref. 18.)

for the oxidation of these analytes. As seen in Figure 8.1, in both cases the electroanalytical response was significantly enhanced at the cathodically pretreated BDD electrode. Suffredini et al. [18] also reported on the effect of the electrochemical pretreatments of BDD surfaces on the cyclic voltammetry and electrochemical impedance spectroscopy (EIS) responses for solutions containing K4 [Fe(CN)6 ] in aqueous 0.5 mol L−1 H2 SO4 . The results of the cyclic voltammetry experiments show that the hexacyanoferrate(II)/hexacyanoferrate(III) redox couple behaves in a quasi-reversible manner after the anodic pretreatment, whereas the cathodic pretreatment leads to a reversible behavior with a Ep value of 60 mV. In addition, the EIS data for this redox couple showed that the value of a resistance associated to a high-frequency element presented values of over 300  cm2 and about 4  cm2 after anodic and cathodic pretreatments, respectively (see Figure 8.2). From these studies the authors concluded that the electrochemical response of BDD electrodes is strongly affected by the type of electrochemical pretreatment applied to their

184

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

(a) 400

−Z ′′ (Ω cm2)

320

240

160 0.15 Hz 80 5.0 Hz 0 0

80

160

240 Z ′ (Ω

400

(b)

12

−Z ′′ (Ω cm2)

320

cm2)

8

4 320 Hz

0 0

4

8 Z ′ (Ω

12

cm2)

Figure 8.2 Complex-plane plots for the BDD electrode in a 1.0 mmol L−1 K4 [Fe(CN)6 ] + 0.5 mol L−1 H2 SO4 aqueous solution after anodic (a) or cathodic (b) pretreatment. Frequency intervals: 50 kHz to 30 mHz (a) and 50 kHz to 10 Hz (b). Measurements carried out at 0.60 V versus HESS—hydrogen electrode in the same solution. (Reprinted with permission from Ref. 18.)

surface before measurements. This effect is very noticeable for some analytes, although it was also present for electron transfer reactions of well-known reversible couples, for which a cathodic pretreatment of the BDD electrode surface prior to measurements leads to an enhanced electrochemical activity. After, Mah´e, Devilliers, and Comminellis [19] also reported data that evidenced a significantly improved electrochemical reactivity of the Fe(CN)6 4−/3− redox couple on

8.2 CATHODIC PRETREATMENT OF CONDUCTIVE DIAMOND FILMS

185

cathodically pretreated BDD electrodes. According to these authors, the cathodic polarization gives rise to a very high rate constant with great reproducibility, which is consistent with the existence of a hydrogen participation into the hole-generation process at BDD surfaces. They also concluded that simple electrochemical treatments could be used to achieve reversible chemical surface modification of BDD electrodes, associated to a drastic change in the rate of electron transfer. In 2006, Salazar-Banda et al. [20] reported on studies to elucidate the effect of different pretreatments (anodic, cathodic, and thermal) on 800 ppm BDD electrodes on the electrochemical response of the hexacyanoferrate(II)/hexacyanoferrate(III) redox couple. As-prepared BDD electrodes were subjected to either an anodic (3.0 V versus HESS, for 30 min) or a cathodic (−3.0 V versus HESS, for 3 or 30 min) pretreatment in a 0.5 mol L−1 H2 SO4 aqueous solution; the thermal pretreatment was carried out at 400◦ C during 30 min in an oven with an oxygen atmosphere. The kinetics of the Fe(CN)6 4−/3− redox couple was greatly affected by the different pretreatments, as can be seen in Figure 8.3. Clearly, the cathodic pretreatment significantly facilitated the redox reaction, leading to a reversible behavior with a Ep of 60 mV and an Ipox /Ipred ratio of 1.0. An analysis of baseline for the cathodically pretreated electrode, also shown in Figure 8.3, concludes that HT-BDD presents a clean surface with no faradaic processes except for a small anodic signal at the positive end of the scan; this signal has been attributed to the presence of small amounts of sp2 carbon as surface impurity [19,46]. On the other hand, the anodic and thermal pretreatments yielded electrode surfaces that do not favor the kinetics of this redox reaction, resulting in irreversible behaviors, with Ep values of 1010 mV and 440 mV associated to Ipox /Ipred ratios equal to 1.5 and 1.3, respectively. This clearly confirms that the cathodic pretreatment of the BDD electrode (resulting in a hydrogen-terminated surface) leads to an enhanced electrochemical response, as previously reported [18].

Figure 8.3 Cyclic voltammograms (ν = 0.05 V s−1 ) for BDD electrodes (A = 0.63 cm2 ) in the presence of a 1.0 mmol L−1 K4 [Fe(CN)6 ] + 0.5 mol L−1 H2 SO4 aqueous solution after cathodic, anodic, and thermal pretreatments. Also included is the background response for the cathodically pretreated electrode. (Reprinted with permission from Ref. 20.)

186

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

Salazar-Banda et al. [20] also investigated the effect of time of cathodic pretreatment on the electrochemical response of the pretreated BDD electrode. They showed that the voltammetric behavior of the Fe(CN)6 4−/3− redox reaction is highly dependent on the duration of the cathodic pretreatment. As seen in Figure 8.4, the reversibility of this redox reaction increases significantly even after pretreatment times as short as 3 s. The value of Ep decreases from 420 mV for the as-received electrode to 110 mV, 72 mV, and 60 mV after pretreatment times of 3 s, 3 min, and 30 min, respectively. Thus, a pretreatment time of 30 min (corresponding to a passed charge of about −60 C cm−2 ) led to a totally reversible electrochemical response of the Fe(CN)6 4−/3− redox couple. It is worth highlighting that short pretreatments time and, consequently, low charges passed already had a significant effect on the BDD electrochemical response. Suffredini et al. [18] attributed the large differences in the electrochemical response of BDD electrodes after cathodic and anodic pretreatments to either an internal transformation of the BDD film or, most probably, to the presence of a discontinuous passive layer. However, Raman studies performed on 800 ppm BDD electrodes before and after a cathodic pretreatment showed that neither important bulk structural differences nor significant changes in the sp2 /sp3 carbon content are introduced into the BDD film by the pretreatment [20]. Thus, this indicates that the enhanced electrochemical response brought on by the cathodic pretreatment is due only to superficial changes, most probably in the surface terminations of the BDD film that becomes mainly hydrogen terminated. In this sense, it is important to mention that several studies have reported that the surface of nondoped diamond is either conducting when it is hydrogen terminated or insulating when oxygen terminated [31,47–51]. Furthermore, the hydrogen-terminated surface becomes insulating when heated under vacuum at temperatures higher than 700◦ C [52], a process that leads to the elimination of hydrogen from the diamond surface.

tcp = 3min 60

tcp = 30min

I (μA)

30 tcp = 0 0

–30 tcp = 3 s –60 0.2

0.4

0.6 E (V) vs. HESS

0.8

1.0

Figure 8.4 Effect of cathodic pretreatment time on the electrochemistry of BDD electrodes (A = 0.63 cm2 ): cyclic voltammograms (ν = 0.15 V s−1 ) for an 800 ppm BDD electrode in a 1.0 mmol L−1 K4 [Fe(CN)6 ] + 0.5 mol L−1 H2 SO4 aqueous solution after cathodic pretreatments at −3.0 V versus HESS, for different lengths of time (tcp ), as indicated on the curves. (Reprinted with permission from Ref. 20.)

8.2 CATHODIC PRETREATMENT OF CONDUCTIVE DIAMOND FILMS

187

However, the conductivity is reinstated when the diamond is hydrogenated again by a plasma treatment. According to Hayashi et al. [29], the conductivity of nondoped hydrogen-terminated diamond is in the range of 10−6 to 10−4 S, due to p-type carriers with a lateral concentration in the range of 1012 to 1013 cm−2 and with a mobility between 10 and 100 cm2 V−1 s−1 . The mobility measured for BDD is not too different, there being a general agreement that the carriers are holes residing in a layer near to the surface [50,51]. Thus, on nondoped diamond, a hydrogen-terminated surface is necessary for the existence of this superficial conductivity that allows the onset of bulk conductivity. On the other hand, if the diamond surface is insulating (oxygen terminated), no bulk conductivity arises. Taking this into account, in electrodes with low boron doping levels it could be assumed that there is a mixed conductivity at the BDD electrode surface, due to areas with high boron contents and to hydrogen-terminated areas. These areas must be connected to boron-rich regions in the diamond bulk, leading to conductive pathways [35]. Probably one of the most interesting results reported by Salazar-Banda et al. [20] is the one that cathodically pretreated BDD electrode surfaces present a dynamic electrochemical behavior; that is, the HT-BDD electrochemical response changes with time of exposition to air. Hence, after a cathodic pretreatment (−3.0 V versus HESS, for 30 min), the HT-BDD electrodes present a progressive decrease of the electron transfer rate for the Fe(CN)6 4−/3− redox couple that results in a loss of the reversibility as a function of time exposed to atmospheric conditions. This dynamic behavior was associated to a loss of superficial hydrogen due to oxidation of the surface by oxygen from the air. This assumption was confirmed by XPS analysis, since a cathodically pretreated electrode exposed to air for 30 days clearly presented an increased superficial content of oxygen. For a cathodically pretreated 800 ppm BDD electrode, Ep changes from 60 mV to 85 mV, after the first 48 h of exposition to the atmosphere, and to 293 mV after 100 days (see Figure 8.5) [20]. These changes must be related to loss of superficial hydrogen due to the oxidation of the surface by atmospheric oxygen or by other species (HCO3 − , OH− ) contained in the thin layer of water naturally formed on the surface of solids exposed to air [31,50]. In fact, when these results were correlated with oxygen content from XPS data, it was possible to observe that after one month of exposition to air, an increase of 85 mV in the value of Ep is related to a difference of 5.8% in the value of the O/C ratio on the electrode surface [20]. In this sense, Kulesza, Patyk, and Rozploch [53] reported on the spontaneous oxidation of hydrogenated nondoped diamond surfaces. When hydrogenated nondoped diamond surfaces are in contact with air, a degradation of the hydrogen layer occurs by oxidation, followed by long-term deterioration of the surface electrical conductivity. The increase rate of the surface resistance was calculated to be approximately (10 ± 6) k month−1 . If one assumes that the surface resistance of BDD electrodes exposed to air increases similarly to the one for hydrogenated nondoped diamond surfaces, this may explain the variations in the electrochemical behavior of the hexacyanoferrate(II)/hexacyanoferrate(III) redox couple (see Figure 8.5). Additionally, Salazar-Banda et al. [20] reported that the dynamic behavior of the electrochemical response of cathodically pretreated BDD electrodes also presents an inverse dependence with the doping level. Figure 8.6 shows how the value of Ep changed with time of exposition to air for the different cathodically pretreated (−3.0 V versus HESS, for 30 min) BDD electrodes. From this figure one can see, for instance,

188

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

Figure 8.5 Effect of exposure to air on the electrochemical response of a cathodically pretreated 800 ppm BDD electrode (A = 0.63 cm2 ): cyclic voltammograms (ν = 0.05 V s−1 ) for the Fe(CN)6 4−/3− redox couple as a function of the time after pretreatment (tap ) that the BDD electrode was exposed to atmospheric conditions, recorded in a 1.0 mmol L−1 K4 [Fe(CN)6 ] + 0.5 mol L−1 H2 SO4 aqueous solution. Cathodic pretreatment: −3.0 V versus HESS, for 30 min, in 0.5 mol L−1 H2 SO4 . (Reprinted with permission from Ref. 20.)

Figure 8.6 Effect of exposure to air on the Ep values for the Fe(CN)6 4−/3− redox couple for cathodically pretreated (−3.0 V versus HESS, for 30 min) BDD electrodes having different doping levels, as indicated in the inset. Data from cyclic voltammograms (ν = 0.05 V s−1 ) recorded in a 1.0 mmol L−1 K4 [Fe(CN)6 ] + 0.5 mol L−1 H2 SO4 aqueous solution. (Reprinted with permission from Ref. 20.)

8.2 CATHODIC PRETREATMENT OF CONDUCTIVE DIAMOND FILMS

189

that after 140 h of exposition to the atmosphere the Ep values for the 300 ppm, 800 ppm, 2000 ppm, and 8000 ppm BDD electrodes increased by 35, 32, 8, and 5 mV, respectively. This dynamic behavior of the electrochemical response inversely dependent on the electrode doping level clearly suggests that the boron content has a stabilizing effect on the H-terminated surface. According to the authors, these results can be further understood if one assumes that there is a mixed conductivity at the surface of the BDD electrodes with low boron doping levels, due to the existence of heterogeneity in the boron content on the surface; the boron-poor areas lose hydrogen quicker than the boron-rich areas, due to a stabilizing effect of boron. These areas must be connected to boron-rich regions in the diamond bulk, leading to conductive pathways [35]. However, this mixed conductivity is less important in electrodes with high boron doping levels, where most of the surface is covered with boron-rich areas (hydrogen stabilized). The findings of the study by Salazar-Banda et al. [20] clearly show that hydrogenterminated sites play an important role in the electrochemical response of BDD electrodes. Findings also indicate that when reproducible results are required, the BDD electrode has to be cathodically pretreated just before the electrochemical experiments are carried out, especially when the electrode was not used for a long period of time. Girard et al. [21] studied the influence of anodic and cathodic pretreatments on the behavior of highly doped (1020 cm−3 ) BDD electrodes. The electrochemical pretreatments were performed during 10 s with two significantly different current densities: ±10−4 A cm−2 (±10−3 C cm−2 ) for mild pretreatments and ±10−1 A cm−2 (±1 C cm−2 ) for severe ones. The authors observed that when the mild cathodic pretreatment (−10−4 A cm−2 ) was carried out, the reversibility of the Ce3+/4+ and Fe(CN)6 4−/3− redox couples diminished (see Figure 8.7). It is worth noticing that the mild cathodic pretreatment is indeed very mild, associated to a potential difference between working and reference electrodes that is nearly stable and about −1.0 V versus MSE [mercury sulfate electrode, equal to about −1.4 V versus Ag/AgCl (3.0 mol L−1 KCl)]. This electrode potential is actually not in the potential range of the hydrogen evolution reaction (see, for instance, Figure 5 in Mah´e, Devilliers, and Comninellis [19] or Figure 2 in Hupert et al. [54]); consequently, this mild pretreatment cannot be considered as a truly cathodic pretreatment in the sense used in previous studies (water electroreduction to produce hydrogen terminations) [18–20]. On the other hand, after the severe cathodic pretreatment {corresponding to a nearly stable electrode potential in the range of −3 V versus MSE [∼−3.4 V vs. Ag/AgCl (3.0 mol L−1 KCl)]}, the Ep value (700 mV with a small peak current increase) in the voltammograms registered with the ceric species (see Figure 8.8) is a bit smaller than that obtained for the as-deposited BDD electrode (750 mV), while Ep for Fe(CN)6 4−/3− redox system reaches the quasi-reversible value of 100 mV with a notable increase of the peak current [21]. In a subsequent study, Simon et al. [22] confirmed the enhancement in the conductivity of moderately doped (1019 cm−3 ) BDD surfaces after cathodic pretreatments of as-deposited samples. They also showed that previously annealed samples (1100◦ C in ultrahigh vacuum, for 12 h, in order to outgas the hydrogen introduced into the diamond layer during the deposition process) do not exhibit an enhancement of conductivity after the electrochemical pretreatments. This shows that the presence of hydrogen in the BDD electrodes seems to be crucial to increase the superficial conductivity after the electrochemical pretreatments.

190

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

400 (a) Ce3+/4+ in H2SO4 medium

j (µA cm–2)

200

0

NT –200

A1 C1

–400 –0.3

300

0.0

0.3

0.6 E (V/ESM)

0.9

1.2

1.5

(b) Fe(CN)64–/3– in KOH medium

j (µA cm–2)

200 100 0 –100

NT A1

–200

C1 –1.5

–1.0

–0.5 E (V/ESM)

0.0

0.5

Figure 8.7 Cyclic voltammograms (ν = 0.02 V s−1 ) for (a) 5.0 mmol L−1 Ce3+/4+ in 0.5 mol L−1 H2 SO4 and (b) 5.0 mmol L−1 Fe(CN)6 4−/3− in 0.5 mol L−1 KOH, at BDD electrodes after mild electrochemical pretreatments (±10−3 C cm−2 ): anodically (dashed lines) or cathodically (dotted lines) pretreated and as-deposited (full lines) electrodes. (Reprinted with permission from Ref. 21.)

A recent tentative to elucidate the electron transfer behavior on HT-BDD electrodes in an electrolytic solution was carried out by Wang et al. [55] using scanning probe microscopy and ab initio methods. The HT-BDD electronic structures were investigated in detail. Shallow acceptors were found in the HT-BDD bandgap and were attributed to the interaction between physisorbed active adsorbates and C–H bondings on the diamond surface. The authors also concluded that these shallow acceptors favor electron transfer. Studies of diamond electrodes in nonaqueous electrolytes are more limited. PastorMoreno and Riley [56] studied the reduction of 1,4-benzoquinone in acetonitrile at BDD electrodes with different surface pretreatments. They showed that the mechanism of reduction is dependent on the electrode pretreatment. Whereas the electrochemistry at an oxygenated BDD surface (prepared by immersion in a hot chromic acid solution) resembles that of a glassy-carbon electrode, the electrochemistry at an HT-BDD

8.2 CATHODIC PRETREATMENT OF CONDUCTIVE DIAMOND FILMS

400

191

(a) Ce3+/Ce4+ in H2SO4 medium

300

j (µA cm–2)

200 100 0 –100 NT –200

A2

–300

C2

–400 –0.5 600

0.0

0.5 E (V/ESM)

1.0

1.5

(b) Fe(CN)64–/3– in KOH medium

j (µA cm–2)

400 200 0 –200 NT

–400

A2 C2

–600 –1.0

–0.5 0.0 E (V/ESM)

0.5

Figure 8.8 Cyclic voltammograms (ν = 0.02 V s−1 ): (a) for 5.0 mmol L−1 Ce3+/4+ in 0.5 mol L−1 H2 SO4 and (b) for 5.0 mmol L−1 Fe(CN)6 4−/3− in 0.5 mol L−1 KOH, at BDD electrodes after severe electrochemical pretreatments (±1 C cm−2 ): anodically (dashed lines) or cathodically (dotted lines) pretreated and as-deposited (full lines) electrodes. (Reprinted with permission from Ref. 21.)

surface (as-prepared or obtained by applying a potential in the hydrogen evolution region, for 6 h, in 1 mol L−1 KCl) indicates the presence of a hydrogen source. In the latter case, the authors concluded that hydrogen in the diamond subsurface may participate in electrochemical processes. In view of the small number of reports available in the literature about the effect of cathodic treatments on the electrochemical and electroanalytical behavior and surface properties of BDD films, it is clear that this issue will demand further investigations in the future. For example, studies must be carried out to (1) elucidate the changes (in the bulk and in the surface) in the physical and/or physicochemical properties of diamond materials caused by cathodic pretreatments; (2) evaluate the physical and/or chemical stability of BDD films after repeated and severe cathodic pretreatments, focusing on the use of in situ

192

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

techniques; and (3) quantify and understand the relationship between different properties of diamond materials before and after cathodic pretreatments [properties such as sp2 type carbon content, grain boundary size (micro and nanodiamond), level of doping, conductivity, and surface termination, among others]. The results from these studies would certainly contribute to advance the fundamental knowledge and the technological applications of BDD in electrochemistry and electroanalysis. 8.3 8.3.1

ELECTROANALYTICAL APPLICATIONS General Aspects

One of the great challenges of analytical chemistry has been the fulfillment of the demand for analytical methods for the quantification of biological, environmental, and pharmaceutical analytes in several kinds of matrixes. The concentration of degradation products and/or impurities present in analytes varies widely and the analysis procedures usually require physical and/or chemical separation(s) prior to any meaningful determination. The assay of chemical species of interest may correspond to a very complex process, due to limiting factors that can be represented by the need to determine ever-smaller amounts (often below the LOD value offered by the techniques available), the interferences arising from the complex character of many matrixes, and the need to distinguish and quantify various chemical species associated with the same product (selectivity). As was pointed out previously in this chapter and by others (see, for instance, refs. 1–5), thin BDD films have an important number of electrochemical properties that distinguish them from other carbons bonded by sp2 , commonly used as electrodes in electroanalytical determinations. Because of these properties, BDD electrodes have been widely studied in recent years, both in terms of fundamental electrochemical properties [1–3,18,20,35,57,58] and electroanalytical [1,4–6,59,60] and environmental (wastewater treatment) applications [61–67]. The wide potential window (up to 3.5 V) presented by BDD electrodes in aqueous solutions—as well as their very low and stable voltammetric background current, long-term response stability, and low sensitivity to dissolved oxygen—are important characteristics that allow the detection of many electroactive species, which otherwise would be masked by the water decomposition reactions. However, it is clear that the analytical performance of BDD electrodes greatly depends on their surface termination (i.e., whether they are hydrogen or oxygen terminated [68–74]). Accordingly, surface modifications of BDD have a strong effect on its electroanalytical behavior when detecting several kinds of analytes. The presence of different functional groups on the BDD surface plays an important role when using diamond electrodes in electroanalysis. Initially, in most studies found in the literature, diamond electrodes were subjected mainly to an anodic treatment of the surface in order to make it hydrophilic (standard procedure) [74–76]. However, as pointed out in Section 8.2, more recently Suffredini et al. [18] and Salazar-Banda et al. [20] highlighted that cathodic pretreatments of BDD electrodes lead to a nearly-ideal reversible behavior for the Fe(CN)6 4−/3− redox system, along with improving the electrochemical responses for other analytes in aqueous media. So, this pretreatment has been used in many studies, especially those involving measurements for the detection of pesticides, drugs, and food additives using classical voltammetric techniques [e.g., square-wave voltammetry (SWV) and differential pulse voltammetry (DPV)]. It has been shown that for many analytes the combination

8.3 ELECTROANALYTICAL APPLICATIONS

193

of a cathodically pretreated (hydrogen-terminated) BDD electrode with these techniques becomes a very powerful analytical tool. Hence, a summary of electroanalytical applications of cathodically pretreated BDD electrodes in the determination of several kinds of analytes in different matrixes is presented hereinafter. 8.3.2

Determination of Pesticides in Environmental Samples

Although chemical substances such as herbicides, germicides, and insecticides (pesticides) are necessary to improve agricultural productivity, they can cause prejudicial effects on the environment since they flow into natural or public water systems. Commonly, most of them are carcinogenic; consequently, their presence in the environment can also be harmful to human health. Since most electroanalytical studies to determine pesticides are done using a mercury electrode, the search for new materials becomes increasingly important. The electroanalytical determination of some pesticides using cathodically pretreated BDD electrodes is summarized here. 8.3.2.1 Carbaryl Carbaryl, like any other insecticide, must be toxic to insects to be effective; like all pesticides, it is also toxic to certain nontarget organisms, including humans. Carbaryl acts both by entering the stomach of the pest with food and by being absorbed through body contact. The occupational exposure of humans to this insecticide has been observed to cause cholinesterase inhibition and reduction of the activity of this enzyme in the blood, which may cause neurological effects [77]. Therefore, simple, sensitive, and reliable methods for the analysis of carbaryl in air, food, grains, and natural water are quite welcome. Codognoto et al. [78] reported the development of an analytical method to determine carbaryl, in aqueous solutions as well as in natural waters, using a BDD electrode and SWV. Prior to the determinations, the electrode was first anodically pretreated [+ 3.0 V versus Ag/AgCl (3.0 mol L−1 KCl) for 30 min] to clean its surface, followed by a cathodic pretreatment [−3.0 V versus Ag/AgCl (3.0 mol L−1 KCl) for 30 min]. The developed method of analysis does not involve any previous extraction, cleanup, preconcentration, or derivatization steps. The obtained analytical curve range was 2.5–30.0 μmol L−1 (see Figure 8.9), with a limit of detection (LOD) of 8.2 ± 0.2 μg L−1 in pure water (analytical sensitivity of 3.07 mA mmol−1 L) and a limit of quantification (LOQ) of 27.5 μg L−1 ; the attained reproducibility and repeatability (n = 10) were 3.9% and 3.2%, respectively. This analytical sensitivity was slightly decreased (to 2.80–2.90 mA mmol−1 L) when the experiments were carried out using water samples collected from two different points in a polluted urban creek. The obtained LOD value is quite adequate, considering that according to the local legislation the maximum amount of carbaryl allowed in natural water is 10 μg L−1 . The authors also evaluated the effect of other pesticides (fenthion and 4-nitrophenol) on the carbaryl determination and found an insignificant influence, as shown in Figure 8.10. In other words, the voltammetric procedure reported by the authors leads to clearly separated electrochemical responses for the three analytes, which can thus be determined simultaneously. 8.3.2.2 4-Nitrophenol 4-Nitrophenol (4-NP), a hazardous substance that can have a high environmental impact due to its toxicity and persistence, is a metabolite of several

194

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

Figure 8.9 Calibration curve from SWV responses of a cathodically pretreated BDD electrode (A = 0.62 cm2 ) for different carbaryl concentrations in the range 2.5–30.0 μmol L−1 in 0.1 mol L−1 Na2 SO4 (pH 6.0). SWV conditions: Es = 2 mV, a = 50 mV, f = 300 Hz. Inset: linear dependence of the peak current on the carbaryl concentration. (Reprinted from Ref. 78.)

Figure 8.10 SWV profile on the BDD electrode (A = 0.62 cm2 ) for 10 μmol L−1 4-nitrophenol (1), 10 μmol L−1 fenthion (2), and 25 μmol L−1 carbaryl (3) in a 0.1 mol L−1 Na2 SO4 (pH 6.0) solution. SWV conditions: Es = 2 mV, a = 50 mV, f = 300 Hz. (Reprinted from Ref. 78.)

organophosphorus pesticides. Thus, some electroanalytical methods have been developed to determine its residues in different types of samples. Pedrosa, Codognoto, and Avaca [79] reported the electroanalytical determination of 4-NP in aqueous solutions on a BDD electrode using SWV. Prior to the experiments the electrode was cathodically pretreated at − 3.0 V versus Ag/AgCl (3.0 mol L−1 KCl) for 30 s. The compound presented only one irreversible peak for both its oxidation at 1.0 V versus Ag/AgCl (3.0 mol L−1 KCl) and its reduction at −0.8 V versus

8.3 ELECTROANALYTICAL APPLICATIONS

195

Ag/AgCl (3.0 mol L−1 KCl), as seen in Figure 8.11. The corresponding LOD values are 8.4 μg L−1 and 12.1 μg L−1 , respectively. Garbellini, Salazar-Banda, and Avaca [80] also investigated the determination of 4-NP on a cathodically pretreated BDD electrode using SWV, but associated with ultrasound radiation. Since this radiation can clean the electrode surface, its fouling commonly caused by strong adsorption of the electroactive species could be minimized and, consequently, the LOD values were improved. Prior to the experiments, the BDD electrode was pretreated in a 0.5 mol L−1 H2 SO4 solution by applying + 3.0 V versus Ag/AgCl (3.0 mol L−1 KCl) for 5 s, followed by −3.0 V Ag/AgCl (3.0 mol L−1 KCl) for 30 s. As can be apprehended when Figure 8.12 is compared with Figure 8.11, clearly the application of ultrasound radiation increased the sensibility of the detection method compared to the silent conditions (about 3 times higher). This was explained by the authors as due 20

(a)

9

20

Ip/µA

15

8

10 5

15 I (µA)

0

7 0

3 4 1 2 5 –5 –1 [4–Nitrofenol]/10 mol L

6

10 5 4 3 2 1

5

0 0.8

0.9

1.0

1.1

1.2

E (V) vs Ag/AgCI 15

–15

9

(b)

–12

–Ip/µA

12 9

8

6

I (µA)

3

–9

7

0 0

1 2 3 [4–Nitrofenol]/10–5mol L–1

6 5

–6 4 3

–3

2 1

0 –0.4

–0.5

–0.6

–0.7

–0.8

–0.9

–1.0

–1.1

E (V) vs Ag/AgCI

Figure 8.11 SWV response of the BDD electrode (A = 0.62 cm2 ) for different 4-NP concentrations in a 0.1 mol L−1 Britton-Robson buffer (pH 6.0). (a) oxidation and (b) reduction-based determinations. SWV conditions: a = 60 mV, Es = 2 mV, f = 100 Hz. Insets: linear dependence of the peak current with 4-NP concentration. (Reprinted with permission from Ref. 79.)

196

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

50 40 Ip/µA

30

12

30

11

20

10

A

I (µA)

10

9

0

20

(a)

13

B

40

8 0

10

20 30 40 50 –6 –1 [4–NP]/10 mol L

60 7 6

10

5 4 1

0 0.6

0.8

3 2

1.0

1.2

E (V) vs Ag/AgCI –120

13

−140

B

−120 −100 Ip/µA

–100 –80

11

−80 −60

10

A

−40

9

I (µA)

−20 0

–60

0

10

20

30

40

50

60

[4–NP]/10–6 mol L–1

–40

8 7 6 5 4

–20

3 2 1

0 –0.2

(b)

12

–0.4

–0.6

–0.8

–1.0

E (V) vs Ag/AgCI

Figure 8.12 SWV responses obtained on a BDD electrode (A = 0.25 cm2 ) using different 4-NP concentrations in a 0.1 mol L−1 Britton-Robson buffer (pH = 6.0) in the presence of ultrasound radiation. (a) oxidation and (b) reduction-based determinations. SWV conditions: f = 100 Hz, a = 50 mV, Es = 2 mV. Ultrasound conditions: d: 5 mm and A: 20%. Insets: The respective analytical curves for the oxidation and reduction processes under silent (A) and ultrasound conditions (B). (Reprinted with permission from Ref. 80.)

to the increase in mass transport and the cleaning of the electrode surface brought on by the ultrasound radiation. Furthermore, the sensitivity of the method based on the 4-NP reduction process was increased much more significantly than that based on the oxidation process. The obtained LOD values were 3.87 μg L−1 for the oxidation process and 2.57 μg L−1 for the reduction process; the corresponding LOQ values were 12.9 μg L−1 and 8.58 μg L−1 , respectively. 8.3.2.3 Chlorophenols Pentachlorophenol (PCP) is a very toxic compound and among the chlorophenols it was extensively used for decades as a fungicide in wood preservation. Its analysis has received special attention in order to measure the amounts of this substance that can be transferred to food by direct contact with wood treated with

8.3 ELECTROANALYTICAL APPLICATIONS

197

PCP. Besides, it can be used as model compound for the development of new analytical techniques [43]. The toxicity and persistence of PCP in contaminated places is well known, so that it is considered as a toxicologically relevant pollutant [82]. Codognoto, Machado, and Avaca [43] reported on the optimization of the experimental parameters for the determination of PCP in pure and contaminated water on a BDD electrode also using SWV. Prior to the experiments, the electrode was cathodically pretreated by applying −3.0 V versus Ag/AgCl (3.0 mol L−1 KCl) for 30 s. The obtained calibration curves for the determination of PCP showed an excellent linear response, for solutions both in pure and contaminated water, but with different slopes (see Figure 8.13). Possible explanations for this behavior, which leads to different LOD values, were put forth by the authors: hindrance of a 4-CP adsorption step by contaminant organic molecules, decrease of the concentration of 4-CP or other chlorophenols by interaction with humic and fulvic substances, or presence of other unknown species that undergo oxidation in the same potential region. The optimized method yielded LOD values for PCP of 5.5 mg L−1 and 15.5 mg L−1 in pure and contaminated water, respectively. According to the authors, the LOD value obtained in pure water was low enough to comply with the limits imposed by environmental and/or health authorities for human consumption, whereas that obtained in polluted matrixes was just above that limit. Two years later, Suffredini et al. [18] clearly demonstrated that the electrochemical response of a BDD electrode in the determination of 4-chlorophenol or PCP was strongly affected by the type of pretreatment applied to its surface before measurements (see Section 8.2). Pedrosa, Machado, and Avaca [45] reported on the simultaneous determination of 4chlorophenol (4-CP), 2,4-dichlorophenol (2,4-CP), and 2,4,6-trichlorophenol (2,4,6-CP) in different water samples on a cathodically pretreated BDD electrode through a deconvolution procedure. This procedure was used in order to obtain a separation of the peaks

Figure 8.13 Analytical curves for PCP in solutions prepared with pure water () and with samples collected at point 1 (•), point 2 (), and point 3 () of a contaminated urban creek. SWV conditions: a = 60 mV, Es = 2 mV, f = 100 Hz. BDD electrode area = 0.62 cm2 . (Reprinted with permission from Ref. 43.)

198

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

TABLE 8.1 Analytical parameters for 4-CP determination in pure and contaminated water (from two different points of a polluted creek) samples, using SWV with a cathodically pretreated BDD electrode or an HPLC method. (Reprinted with permission from Ref. 45.) Method BDD HPLC

Sample

R

LOQ (μg L−1 )

Recovery (%)

Milli Q Point 1 Point 2 Milli Q Point 1 Point 2

0.998 0.995 0.996 0.999 0.998 0.998

9.2 37.2 42.2 1.2 36.1 38.4

94.0 ± 3.1 92.2 ± 4.2 91.1 ± 3.5 96.0 ± 1.1 94.2 ± 5.2 95.1 ± 4.8

for each compound and to construct standard curves for each chlorophenol in the mixture. Analytical curves were obtained with the cathodically pretreated BDD electrode by the standard addition method for the mixture of chlorophenols in the different water samples (pure or collected in two different points of a polluted urban creek) and compared with those obtained using an HPLC method. Recovery experiments were also carried out. From the obtained analytical results, summarized in Table 8.1, the authors concluded that excellent recoveries were attainable even in highly polluted water samples, thus indicating that the reported procedure could be used for determinations in environmental matrices. A slight dependence of the LOQ values on the amount of pollution in the water samples was attributed to the same factors put forth by Codognoto, Machado, and Avaca [43] to explain the variation of the slope of the analytical curves for PCP in pure and contaminated water.

8.3.3

Determination of Substances in Food Samples

8.3.3.1 Aspartame Aspartame (N-L-α-aspartyl-L-phenylalanine methyl ester) has the potential to reduce the amount of sugars and calories in food products, and thus may be combined with sugars (such as sucrose, dextrose, and fructose) and/or sweeteners (such as acesulfame-K, sodium cyclamate, and saccharin). However, aspartame is not always suitable for baking because it often breaks down when heated and it loses much of its sweetness. Considering that most methods developed for aspartame determination require a lengthy pretreatment of the sample prior to analyses, Medeiros et al. [68] proposed an electroanalytical method using a BDD electrode and SWV to determine the aspartame concentration directly in the sample without pretreatment or chemical separation. Prior to the experiments, the BDD electrode was cathodically pretreated by applying −1.0 A cm−2 for 60 s in a 0.5 mol L−1 H2 SO4 solution. As shown in Figure 8.14, the oxidation of aspartame on BDD presents a single peak at 1.6 V versus Ag/AgCl (3.0 mol L−1 KCl), with the characteristics of an irreversible reaction. The obtained analytical curve is linear in the aspartame concentration range 9.9–52 μmol L−1 , with an LOD of 0.23 μmol L−1 ; in repeatability studies (n = 5) of a 0.10 mmol L−1 aspartame solution, the obtained relative standard deviation (RSD) was 0.2%. Several dietary products containing aspartame were analyzed using the proposed SWV method and employing the standard addition method. The inset in Figure 8.14b is the obtained analytical curve

8.3 ELECTROANALYTICAL APPLICATIONS

12 R = 0.9998 10 8 6 4 2 0 0 1 2 3 4 5 6 [Aspartame]/10–5 mol L–1

11 10 9 8 7 6 5 4 3 2

I (µA)

25

I (µA)

20

199

15

10

1

5

(a) 0 1.3

1.4

1.5

1.6

1.7

1.8

1.9

E (V) vs. Ag/AgCl 14 12 10 8 6 4 2 0 –3 –2 –1 0 1 2 3 4 [Aspartame]/10–5 mol L–1

5

I (µA)

25

20

4 3 2

15 I (µA)

1 10

5 (b)

0 1.3

1.4

1.5

1.6

1.7

1.8

E (V) vs. Ag/AgCl

Figure 8.14 (a) SWV response of a cathodically pretreated BDD electrode (A = 0.72 cm2 ) for the oxidation of aspartame in increasing concentrations [9.9–52 μmol L−1 , in 0.5 mol L−1 H2 SO4 (curve 1)]; Inset: corresponding analytical curve. (b) Square-wave voltammograms obtained for the determination of aspartame in a sweetener (Zero cal®). SWV conditions: a = 40 mV, Es = 2 mV, f = 10 Hz. (Reprinted with permission from Ref. 68.)

for the determination of aspartame in a sample of a dietary product, the sweetener Zero cal®. 8.3.3.2 Sodium Cyclamate Sodium cyclamate is a white, crystalline, and odorless powder. Although 30 times sweeter than sucrose, it is noncaloric (zero calories) and more stable than other artificial sweeteners, such as aspartame and saccharin, which allows its use at high and low temperatures [82]. Sodium cyclamate is commonly used together with saccharin because cyclamate can mask the bitter flavor left by saccharin, but its

200

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

use has been banned in the US because it was found that it could be metabolized as cyclohexylamine, which can cause several health problems. So, taking into account the interest in the development of analytical methods for the determination of these sweeteners in dietary products, Medeiros et al. [69] proposed the use of SWV with a BDD electrode for the analytical determination of sodium cyclamate. The BDD electrode was cathodically pretreated the same way as it was done for the determination of aspartame [68] (discussed earlier). Figure 8.15 clearly shows the enhancement of the BDD electrochemical activity brought on by this cathodic pretreatment, compared to its untreated (as-received) response. The samples were analyzed as received in a 0.5 mol L−1 H2 SO4 solution in the concentration range 5.0 × 10−5 −4.1 × 10−4 mol L−1 , with an LOD of 4.8 × 10−6 mol L−1 . In the repeatability and reproducibility studies (n = 5) of a 3.0 mmol L−1 cyclamate solution, RSD values of 1.2% and 2.4% were obtained, respectively; furthermore, the proposed method was applied with success in the determination of sodium cyclamate in several dietary products. 8.3.3.3 Aspartame and Sodium Cyclamate Medeiros et al. [70] also reported on the simultaneous determination of aspartame and sodium cyclamate in dietary products using SWV and a cathodically pretreated BDD electrode. The SWV oxidation peak potentials of aspartame and cyclamate present in binary mixtures are about 400 mV apart, as shown in Figure 8.16. For aspartame, LOD was 0.47 μmol L−1 in the presence of 0.30 mmol L−1 cyclamate; for sodium cyclamate, LOD was 4.2 μmol L−1 in the presence of 0.10 mmol L−1 aspartame. When simultaneously changing the concentration of both sweeteners in a 0.5 mol L−1 H2 SO4 solution, the corresponding LOD values were 0.35 and 4.5 μmol L−1 , respectively. In repeatability tests (n = 5), the obtained RSD values were 1.3%, for a 0.10 mmol L−1 aspartame solution, and 1.1% for a 3.0 mmol L−1 cyclamate solution. The proposed voltammetric method

200 a b

I (µA)

150

100

50

0 1.6

1.7

1.8

1.9

2.0

2.1

E (V) vs. Ag/AgCl (KCl 3.0 mol L–1)

Figure 8.15 SWV responses for 3.0 mmol L−1 sodium cyclamate in 0.5 mol L−1 H2 SO4 using an as-received (a) or a cathodically pretreated (b) BDD electrode. SWV conditions: a = 20 mV, Es = 2 mV, f = 10 Hz. BDD electrode area = 0.72 cm2 . (Reprinted with permission from Ref. 69.)

8.3 ELECTROANALYTICAL APPLICATIONS

201

(b) (a)

(c)

Figure 8.16 (a) SW voltammograms obtained for the oxidation of aspartame and cyclamate in 0.5 mol L−1 H2 SO4 . The concentrations of both aspartame (5.0–50 μmol L−1 ) and cyclamate (0.050–0.50 mmol L−1 ) were changed simultaneously. (b) Analytical curves for aspartame. (c) Analytical curves for cyclamate. SWV conditions: a = 40 mV, Es = 2 mV, f = 10 Hz. BDD electrode area = 0.72 cm2 . (Reprinted with permission from Ref. 70.)

was successfully applied in the simultaneous determination of aspartame and sodium cyclamate in several dietary products, with results in very good agreement with those results obtained using an HPLC method. 8.3.3.4 Total Phenols Total phenol concentration is an essential indicator of the state of quality of pharmaceutical products and food, since its concentration could be related to nutritional, processing, and health aspects. Thus, Dejmkova et al. [83] report on a method to determine total phenols in foods such as teas, juices, and wines. The electrode was electrochemically pretreated in 1 mol L−1 HNO3 by a procedure similar to the one proposed by Suffredini et al. [18]: +3.0 V versus Ag/AgCl (3 mol L−1 KCl) for 20 min, followed by −3.0 V versus Ag/AgCl (3 mol L−1 KCl) also for 20 min. When the applied potential on the cathodically pretreated BDD is positive enough to oxidize the phenols present in the sample, the charge consumed during the oxidation could be related to the total phenol content in the sample. The proposed method is robust, allowing a rapid evaluation of total phenols directly in real samples after a simple dilution with the supporting electrolyte. The electrode does not present the drawback of fouling (an electrochemical cleaning procedure was successfully optimized) and the results were validated using the standard Folin-Ciocalteau assay. 8.3.4

Determination of Substances in Pharmaceutical Samples

8.3.4.1 Sulfamethoxazole and Trimethoprim Sulfonamides, indicated primarily to treat urinary infections, are used in combination with trimethoprim (TMP) for the treatment of ear infections, bronchitis, sinusitis, and pneumocystis pneumonia. The corresponding pharmaceutical products usually consist of a sulfonamide mixed with another

202

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

μ

drug that increases its power, such as the sulfamethoxazole (SMX) and TMP mixture, an often used pharmaceutical product. In the specific case of analysis of drugs, quality control can be considered as one of the most important aspects because it contributes to ensure the efficacy, safety, and fundamentally the quality standards of medicines. Thus, the search for techniques that are sensitive, accurate, and easily accessible for the determination of compounds present in several commercial drugs has received much attention. In this context, Andrade et al. [73] reported on the simultaneous electrochemical detection by differential pulse voltammetry (DPV) of SMX and TMP using cathodically (HT-) or anodically (OT-) pretreated BDD electrodes. The cathodic or anodic pretreatment was carried out by applying −0.5 A cm−2 or 0.5 A cm−2 , respectively, during 60 s, in a 0.5 mol L−1 H2 SO4 solution. Cyclic voltammetric studies (see Figure 8.17) show that on an HT-BDD electrode both the SMX and TMP voltammograms exhibit well-defined irreversible oxidation peaks at 1080 mV and 1100 mV versus Ag/AgCl (3.0 mol L−1 KCl), respectively. Conversely, within the investigated potential range (0.5 V to 1.3 V versus Ag/AgCl), no reduction peaks are observed on the reverse scan. However, since the anodic peak potentials for SMX and TMP are separated by only 20 mV, their simultaneous analyses would not be possible under these conditions. Thus, by changing the solution pH, a greatly increased difference in the oxidation peak potential values (180 mV) was found at pH 7 (see Figure 8.18a), associated to two well-defined oxidation waves [Ep = 920 and 1100 mV versus Ag/AgCl (3.0 mol L−1 KCl)] that correspond to the oxidation of SMX and TMP, respectively. When an OT-BDD electrode was used, the magnitude of these oxidation waves decreased, but this decrease was more significant for the SMX oxidation. As can be seen in Figure 8.18b, independently of whether an HT-BDD or an OT-BDD electrode is used, the SMX oxidation current peak decreases as the pH increases. Besides, the magnitude of the current peak obtained for the SMX and TMP oxidation was always higher when the HT-BDD electrode was used.

Figure 8.17 Cyclic voltammograms (ν = 50 mV s−1 ) for (a) blank solution (0.2 mol L−1 Britton-Robison buffer, pH 2.0) and solutions containing (b) 1.0 mg L−1 SMX or (c) 1.0 mg L−1 TMP in this buffer, on a cathodically pretreated BDD electrode. BDD electrode area = 0.63 cm2 . (Reprinted from Ref. 73.)

8.3 ELECTROANALYTICAL APPLICATIONS

203

(a)

(b)

Figure 8.18 (a) Differential pulse voltammetric responses (oxidation) obtained at an anodically (solid line) or cathodically (dashed line) pretreated BDD electrode (A = 0.63 cm2 ) using a mixture of 1.0 mg L−1 SMX and 1.0 mg L−1 TMP in a 0.2 mol L−1 Britton-Robson buffer (pH 7.0). (b) Effect of pH on the SMX oxidation peak current at an anodically (solid line) or cathodically (dashed line) pretreated BDD electrode. DPV conditions: scan rate, 50 mV s−1 ; pulse amplitude, 60 mV; pulse width, 10 ms. (Reprinted from Ref. 73.)

In an effort to explain the different electrochemical behaviors obtained when using an HT- or an OT-BDD electrode, Andrade et al. [73] analyzed the acid-base chemistry of SMX and TMP. The former is a weak acid (pKa = 5.6), whereas the latter is a weak base (pKa = 7.3). Thus, if pH > pKa , the conjugate base predominates in the solution bulk, the analyte then being negatively charged. Consequently, an electrostatic repulsion between sulfa anions and the OT-BDD electrode could be expected for the measurements carried out at pH > 5.6. In fact, for pH > 5.6, the current peaks obtained for the SMX oxidation on the OT-BDD electrode (see Figure 8.18b) were always lower than those obtained on the HT-BDD electrode. However, when the measurements were

204

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

done at pH < 5.6, the SMX oxidation current peaks were not higher on the OT-BDD electrode when compared with those obtained on the HT-BDD electrode, as it could be expected. Furthermore, in the pH range investigated (2 to 7) the oxidation current peaks always decreased linearly with pH, independently of whether an HT-BDD or an OT-BDD electrode was used. Conversely, for pH < 7.3, the magnitude of the TMP oxidation current peak could be expected to be higher on the OT-BDD electrode than on the HT-BDD electrode, but actually it was about 40% lower. Taking this into account, the authors concluded that the electroanalytical performance of the electrodes is not significantly influenced by electrostatic interaction between the analytes and the BDD surface; hence, the obtained results should be explained mainly taking into account the different surface conductivity presented by the HT-BDD or OT-BDD electrodes. However, Ivandini et al., [84] have pointed out that while HT-BDD acts as an excellent electrode for the detection of negatively charged DNA, OT-BDD works exceptionally well for the detection of the positively charged oxidized form of glutathione [85], due to the operation of strong electrostatic interactions. Clearly, questions involving the influence of surface chemistry of the BDD films on the electrochemical processes occurring on these electrodes are still controversial, needing to be further investigated. A series of DPV voltammograms obtained for the simultaneous determination of SMX and TMP at different concentrations (1.0−10 mg L−1 and 0.2−2.0 mg L−1 for SMX and TMP, respectively) in a 0.2 mol L−1 Britton-Robinson buffer (pH 7) by DPV at an HT-BDD electrode are shown in Figure 8.19. Notice in the figure insets that the respective analytical curves presented a good linearity in the investigated concentration range (r 2 = 0.9993 for both SMX and TMP). The calculated values for LOD and LOQ were 3.65 μg L−1 (14.4 nmol L−1 ) and 12.2 μg L−1 (48.2 nmol L−1 ) for SMX; these values were 3.92 μg L−1 (13.5 nmol L−1 ) and 13.1 μg L−1 (45.1 nmol L−1 ) for TMP.

Figure 8.19 Differential pulse voltammetric responses (simultaneous oxidation) obtained on an HTBDD electrode (A = 0.63 cm2 ) at different concentrations of SMX/TMP in a 0.2 mol L−1 Britton-Robson buffer solution (pH 7.0). SMX/TMP concentrations: (1) 1.0/0.2; (2) 2.0/0.4; (3) 3.0/0.6; (4) 4.0/0.8; (5) 5.0/1.0; (6) 6.0/1.2; (7) 7.0/1.4; (8) 8.0/1.6; (9) 9.0/1.8, and (10) 10/2.0 mg L−1 . Inset: respective analytical curves for SMX and TMP. DPV conditions: scan rate, 50 mV s−1 ; pulse amplitude, 60 mV; pulse width, 10 ms. (Reprinted from Ref. 73.)

8.3 ELECTROANALYTICAL APPLICATIONS

205

Besides, repeatability tests carried out by successive measurements (n = 10) in the same solution (10 mg L−1 SMX and 2.0 mg L−1 TMP) showed RSD values of 0.3% and 0.1%, respectively. Despite the fact that LOD and LOQ are not considered so relevant in drugs determination, because of their high concentration in commercial formulations, the obtained values clearly indicate that quite low concentrations of SMX and TMP can be detected using the HT-BDD electrode. So, the proposed method was applied successfully to determine SMX and TMP by the standard addition method in three different commercial formulations. 8.3.4.2 Sulfamethoxazole and Sulfadiazine Souza et al. [86] also reported on the electrochemical determination of sulfonamides (sulfadiazine and sulfamethoxazole, independently) in pharmaceutical formulations employing a cathodically pretreated BDD electrode and SWV. The BDD electrode was pretreated in 0.5 mol L−1 H2 SO4 similarly to what was done by Salazar-Banda et al. [20]. The as-received electrode was first anodically pretreated (+3.0 V versus SCE for 30 min) to clean its surface, followed by a cathodic pretreatment [−3.0 V versus SCE for 30 min). Then, before each measurement, this cathodic pretreatment was carried out for 30 s (only for the very first pretreatment each day, the applied potential was −2.0 V versus SCE). Good linear analytical curves were achieved for both sulfonamides and the obtained LOD values were 2.19 and 1.15 μmol L−1 for sulfadiazine and SMX, respectively. In the repeatability studies (n = 7) of 90 μmol L−1 sulfadiazine and 0.25 mmol L−1 SMX solutions, RSD values of 0.56% and 0.58% were obtained, respectively. In the corresponding reproducibility study (n = 5), RSD values of 0.78% and 0.71% were obtained for sulfadiazine and SMX, respectively. The recovery values for both sulfonamides were in the range 95–104% and the methodology was successfully compared with the standard HPLC method, with relative errors of −4.31% and −0.79% for sulfadiazine and sulfamethoxazole, respectively. 8.3.4.3 Acetylsalicylic Acid Acetylsalicylic acid (ASA), also known by the trade name aspirin, is one of the oldest medicines that still play an important role in modern therapeutics, being widely employed in pharmaceutical formulations for the relief of headaches, fever, muscular pain, and inflammations caused by arthritis or injury. Commonly, ASA is determined by titration after its conversion to salicylic acid and acetic acid by alkaline hydrolysis. Aiming at obtaining a simpler procedure, Sartori et al. [87] investigated the determination of ASA in pharmaceutical formulations using SWV and a cathodically pretreated BDD electrode. In this proposed electroanalytical method, ASA can be directly determined in a 0.01 mol L−1 H2 SO4 solution (see Figure 8.20). The obtained analytical curve was linear in the ASA concentration range 2.50 × 10−6 −1.05 × 10− mol L−1 , with a LOD of 2.0 μmol L−1 ; in the repeatability study (n = 10) of 45 μmol L−1 ASA solutions, a RSD value of 1.4% was obtained. The proposed method was applied with success in the determination of ASA in several pharmaceutical formulations (commercial adult and children tablets) and the obtained results were in close agreement, at a 95% confidence level, with those obtained using an official method of the British Pharmacopoeia. The reported results demonstrate that the combination of SWV with an HT-BDD electrode is a feasible alternative for the analytical determination of ASA in commercial adult and children tablets without previous hydrolysis of the analyte.

206

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

50

50 I /µA

40

40

30 20 10

I /µA

0 0 20 40 60 80 100 120 –6 –1 [AAS]/10 mol L

30 20

17 10 1 0 1.7

1.8

1.9 2.0 E (V) vs Ag/AgCl

2.1

2.2

Figure 8.20 SWV response (direct current) of an HT-BDD electrode (A = 0.33 cm2 ) for different ASA concentrations in 0.01 mol L−1 H2 SO4 : (1) 0; (2) 2.50 × 10−6 ; (3) 5.00 × 10−6 ; (4) 7.50 × 10−6 ; (5) 1.50 × 10−5 ; (6) 2.25 × 10−5 ; (7) 3.00 × 10−5 ; (8) 3.75 × 10−5 ; (9) 4.50 × 10−5 ; (10) 5.25 × 10−5 ; (11) 6.00 × 10−5 ; (12) 6.75 × 10−5 ; (13) 7.50 × 10−5 ; (14) 8.25 × 10−5 ; (15) 9.00 × 10−5 ; (16) 9.75 × 10−5 ; (17) 1.05 × 10−4 mol L−1 . Inset: analytical curve for the ASA oxidation process. SWV conditions: a = 40 mV, Es = 3 mV, f = 50 Hz. (Reprinted with permission from Ref. 87.)

8.3.4.4 Paracetamol and Caffeine Paracetamol (N-acetyl-p-aminophenol, acetaminophen), a substance derived from p-aminophenol, is considered an excellent analgesic in cases of mild to moderate pain, besides not causing relevant gastrointestinal effects. It is noncarcinogenic and an effective substitute for aspirin for patients with sensitivity to it. Caffeine is used therapeutically in combination with ergotamine in the treatment of migraines or with nonsteroidal anti-inflamatories in analgesic formulations. Lourenc¸a˜ o et al. [88] investigated the use of a cathodically pretreated BDD electrode to develop simple, selective, and sensitive methods for the determination of paracetamol and caffeine, simultaneously and individually. This was achieved using (1) SWV, for paracetamol; and (2) DPV, for caffeine individually and for both drugs simultaneously. The HT-BDD electrode was obtained by applying −1.0 A cm−2 for 180 s in a 0.5 mol L−1 H2 SO4 solution. In the binary mixtures, a separation of about 550 mV between the peak oxidation potentials of paracetamol and caffeine was obtained. Figure 8.21 shows the DPV voltammograms obtained for the simultaneous determination of the drugs. The corresponding calibration curves showed an excellent linear response, in the range 0.50–83 μmol L−1 for both compounds. The LOD values for the simultaneous determination of paracetamol and caffeine were 0.49 and 0.035 μmol L−1 , respectively. In the repeatability study (n = 5) of a 50 μmol L−1 paracetamol and caffeine solution, RSD values of 0.3% and 1.8% were obtained, respectively. The proposed method was successfully applied in the simultaneous determination of paracetamol and caffeine in several pharmaceutical formulations (tablets), with results similar to those obtained using a reference HPLC method. 8.3.4.5 Sildenafil Citrate (Viagra®) Sildenafil citrate (1-[[3-(6,7-dihydro-1methyl-7-oxo-3-propyl-1-H-pyrazolo[4,3-d]pirydin-5-yl)-4-ethoxyphenyl]sulfonyl]-4methylpiperazine citrate), commonly known as Viagra®, is a drug widely used as oral

8.3 ELECTROANALYTICAL APPLICATIONS

207

200

I (μA)

150

100 15 50 1 0 0.4

0.6

0.8 1.0 1.2 E (V) vs. Ag/AgCl

1.4

1.6

Figure 8.21 Differential pulse voltammetric curves obtained for the oxidation of paracetamol and caffeine at equal concentrations in a 0.2 mol L−1 acetate buffer solution (pH 4.5): (1) 0.50, (2) 2.0, (3) 4.0, (4) 5.9, (5) 7.9, (6) 9.8, (7) 19, (8) 28, (9) 37, (10) 45, (11) 54, (12) 61, (13) 69, (14) 76, and (15) 83 μmol L−1 . DPV conditions: scan rate, 70 mV s−1 ; modulation amplitude, 100 mV; modulation time, 7 ms. BDD electrode area = 0.72 cm2 . (Reprinted with permission from Ref. 88.)

therapy for erectile dysfunction. Due to the high consumption of sildenafil, selective, sensitive, and easily used methods of analysis of its pharmaceutical formulations are welcome, to assess not only their quality but also their authenticity. Accordingly, Batista et al. [89] proposed the use of DPV in conjunction with a cathodically (−1.0 A cm−2 for 240 s in 0.5 mol L−1 H2 SO4 ) pretreated BDD electrode for the analytical determination of commercial pharmaceutical samples of Viagra®. According to the authors, the HT-BDD electrode presented a better peak definition and a higher current magnitude, indicating that the cathodic pretreatment of the electrode led to a larger electrochemical activity for sildenafil oxidation. The cyclic voltammetric response of sidenafil presents two electrochemically irreversible anodic peaks, at ∼1.5 and ∼2.0 V versus Ag/AgCl (3.0 mol L−1 KCl). In order to avoid interference from the oxygen evolution reaction, only the first peak was considered for the development of the electroanalytical method. After optimization of the DPV parameters, which included the maximum peak current and the minimum half-peak width values, the authors validated the method taking into account selectivity (possible interferents), linearity, and recovery studies. All these validation criteria were satisfied and the obtained LOD value was 0.64 μmol L−1 . In the repeatability (n = 10) and reproducibility (n = 5) studies of 4.0 μmol L−1 sildenafil solutions, RSD values of 1.1% and 1.9% were obtained, respectively. Table 8.2 summarizes the results obtained for determinations of sildenafil citrate in Viagra® commercial tablets of different dosages employing the proposed DPV method compared with those obtained using a reference HPLC method. 8.3.4.6 Lidocaine Lidocaine (2-(diethylamino)-N -(2,6-dimethylphenyl)acetamide) is a local anesthetic commonly used to relieve pain related to surgical, dental, and gynecological procedures. The therapeutic and toxic effects of lidocaine are directly related to its concentration and metabolites. The toxicity of lidocaine affects primarily the cardiovascular and central nervous systems, and overdoses can result in ventricular

208

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

TABLE 8.2 Sildenafil citrate content in Viagra ® pharmaceutical formulations determined by a proposed differential pulse voltammetric (DPV) method using an HT-BDD electrode and by a comparative HPLC method. (Reprinted from Ref. 89.) Determined value (mg) DPV a

HPLC a

Relative errorb (%)

25.7 ± 0.9 51.5 ± 0.5 105.0 ± 0.9

26.1 ± 0.5 52.4 ± 0.8 102.0 ± 0.6

–1.5 –1.7 2.9

Label value (mg) 25 50 100

a Average of 3 measurements. b 100 × [(DPV value—HPLC value)/HPLC value].

arrhythmia. Taking into account the lack of a simple and direct electroanalytical method for lidocaine determination, Oliveira et al. [90] investigated the use of SWV with a BDD electrode for this purpose. According to the authors, a cathodic pretreatment was necessary for conditioning the BDD surface prior to the electroanalytical determinations. Thus, before each analysis the BDD electrode was pretreated in a 0.1 mol L−1 HClO4 solution by applying + 3.2 V versus Ag/AgCl for 30 s (to clean the electrode surface), followed by −2.8 V versus Ag/AgCl, for 30 s. Figure 8.22 shows the SWV responses obtained for different concentrations of lidocaine as well as the corresponding analytical curve. The obtained LOD and LOQ values were 10.0 and 34.4 mg L−1 , respectively. The proposed method was successfully applied in the determination of lidocaine in three different commercial gel formulations, in which lidocaine is mixed with propyleneglycol; the presence of propyleneglycol had no influence on the voltammetric responses, as it can be verified in Table 8.3.

60 Ip (µA)

50

70 60

I (µA)

40 30

50 40 30 20 10 0

20

background curve

(b) 6 8 10 12 2 4 Concentration (10–5 M)

0

10 0

(a) 0.0

0.4

0.8 1.2 E (V) Ag/AgCl

1.6

2.0

Figure 8.22 (a) SWV response obtained for the oxidation of lidocaine at different concentrations in a 0.1 mol L−1 Britton-Robson buffer (pH 2.0); (b) Linear dependence of the oxidation peak current with the lidocaine concentration. SWV conditions: a = 50 mV, Es = 2 mV, f = 150 Hz. BDD electrode area = 0.25 cm2 . (Reprinted from Ref. 90.)

8.5 CONCLUSIONS

209

TABLE 8.3 Lidocaine recovery percentage from samples of commercial pharmaceutical preparations (gels) analyzed using square-wave voltammetry and a cathodically pretreated BDD. (Reprinted from Ref. 90.)

Cream A Cream B Cream C

Label value (mg g−1 )

Found value (mg g−1 )

Recoverya (%)

RSDb (%)

50.0 50.0 50.0

49.8 49.2 48.8

99.6 98.4 97.6

2.1 2.3 2.6

a 100 × (found value/label value); b Relative standard deviation of measurements done in triplicate.

8.4

GOLD DEPOSITION AND STRIPPING

Finally, it should be noted that BDD films have also been used as electrodes to investigate the deposition of different metals, when the electrochemical pretreatment of their surfaces was also found to play a role. Specifically, Holt et al. [91] reported on the deposition of gold on BDD by the reduction of tetrachloroaurate(III). Considering that the adhesion and the electrochemical reactivity of gold particles on diamond likely depended on the physical and chemical nature of the electrode surface, Holt and colleagues investigated the deposition and stripping characteristics of gold metal on a cathodically or anodically pretreated BDD electrode. The cathodic or anodic pretreatment was carried out in dilute aqua regia by applying − 1.2 V (versus SCE) for 10 s or + 2.0 V (versus SCE) for 30 s, respectively. Significant increases in both the gold deposition current and the stripping efficiency were found when the cathodically pretreated BDD electrode was used. The authors reported that the cathodic pretreatment had a dramatic effect on the reduction process of the tetrachloroaurate(III) ion: decrease of the reduction overpotential by about 0.1 V and enhancement of the reduction current by nearly a factor of 2. Additionally and even more significant, the size of the stripping peak increased, enhancing the stripping efficiency. XPS quantitative analyses revealed that the amount of gold deposited on the cathodically pretreated BDD electrode was about 10 times higher than that on an untreated BDD electrode.

8.5

CONCLUSIONS

The cathodic pretreatment of BDD electrodes, which increases the fraction of hydrogen termination on their surfaces, has been shown to enhance the electrochemical activity of the electrodes toward many redox couples and analytes. Consequently, the use of hydrogen-terminated BDD in the electroanalytical determination of several analytes has been found to be quite convenient, since boosted sensitivities and selectivity could be attained. Such effects have been reported in the determination of different pesticides in environmental samples, sweeteners, and total phenols in food samples, and drugs in pharmaceutical samples. Nevertheless, a better characterization of the superficial state of BDD brought on by different types of cathodic pretreatments is still necessary and might contribute to even more significant advancements in the application of this electrode in electroanalysis.

210

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

REFERENCES 1. Y.V. Pleskov, A.Y. Sakharova, M.D. Krotova, L.L. Bouilov, B.V. Spitsyn, J. Electroanal. Chem. 1987, 228 , 19–27. 2. A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, Diamond Electrochemistry, Elsevier, Tokyo, 2005. 3. J.C. Angus, Y.V. Pleskov, S.C. Eaton, in Semiconductors and Semimetals: Thin Film Diamond II (C.E. Nebel, J. Ristein, eds.), Elsevier, San Diego, 2004, 77 , pp. 97–120. 4. G.M. Swain, in Semiconductors and Semimetals: Thin Film Diamond II (C.E. Nebel, J. Ristein, eds.), Elsevier, San Diego, 2004, 77 , pp. 121–148. 5. G.M. Swain, in Electroanalytical Chemistry (A.J. Bard, I. Rubinstein, eds.), Marcel Dekker, New York, 2004, 22 , pp. 182–278. 6. O. Chailapakul, W. Siangproh, D.A. Tryk, Sens. Lett . 2006, 4 , 99–119. 7. G. Chen, Talanta 2007, 74 , 326–332. 8. J. Park, V. Quaiserov´a-Mocko, B.A. Patel, M. Novotny, A. Liu, X. Bian, J.J. Galliganb, G.M. Swain, Analyst 2008, 133 , 17–24. 9. K. Peckov´a, J. Musilov´a, J. Barek, Crit. Rev. Anal. Chem. 2009, 39 , 148–172. 10. J.H.T. Luong, K.B. Male, J.D. Glennon, Analyst 2009, 134 , 1965–1979. 11. S. Jolley, M. Koppang, T. Jackson, G.M. Swain, Anal. Chem. 1997, 69 , 4099–4107. 12. M.C. Granger, G.M. Swain, J. Electrochem. Soc. 1999, 146 , 4551–4558. 13. E. Popa, H. Notsu, T. Miwa, D.A. Tryk, A. Fujishima, Electrochem. Solid-State Lett . 1999, 2 , 49–51. 14. H. Notsu, I. Yagi, T. Tatsuma, D.A. Tryk, A. Fujishima, J. Electroanal. Chem. 2000, 492 , 31–37. 15. D.A. Tryk, H. Tachibana, H. Inoue, A. Fujishima, Diam. Rel. Mat . 2007, 16 , 881–887. 16. M. Panizza, I. Duo, P.A. Michaud, G. Cerisola, C. Comninellis, Electrochem. Solid-State Lett . 1999, 3 , 429–430. 17. I. Duo, C. L´evy-Cl´ement, A. Fujishima, C. Comninellis, J. Appl. Electrochem. 2004, 34 , 935–943. 18. H.B. Suffredini, V.A. Pedrosa, L. Codognoto, S.A.S. Machado, R.C. Rocha-Filho, L.A. Avaca, Electrochim. Acta 2004, 49 , 4021–4026. 19. E. Mah´e, D. Devilliers, C. Comminellis, Electrochim. Acta 2005, 50 , 2263–2277. 20. G.R. Salazar-Banda, L.S. Andrade, P.A.P. Nascente, P.S. Pizani, R.C. Rocha-Filho, L.A. Avaca, Electrochim. Acta 2006, 51 , 4612–4619. 21. H. Girard, N. Simon, D. Ballutaud, M. Herlem, A. Etcheberry, Diam. Rel. Mat . 2007, 16 , 316–325. 22. N. Simon, H. Girard, M. Manesse, D. Ballutaud, A. Etcheberry, Diam. Rel. Mat . 2008, 17 , 1371–1375. 23. A. Denisenko, C. Pietzka, A. Romanyuk, H. El-Hajj, E. Kohn, J. Appl. Phys. 2008, 103 , 1–8. 24. M. Wang, N. Simon, C. Decorse-Pascanut, M. Bouttemy, A. Etcheberry, M. Li, R. Boukherroub, S. Szunerits, Electrochim. Acta 2009, 54 , 5818–5824. 25. C.H. Goeting, F. Marken, A. Guti´errez-Sosa, R.G. Compton, J.S. Foord, Diam. Rel. Mat . 2000, 9 , 390–396. 26. F. Marken, R.G. Compton, C.H. Goeting, J.S. Foord, S.D. Bull, S.G. Davies, J. Solid State Electrochem. 2001, 5 , 88–93. 27. I. Yagi, H. Notsu, T. Kondo, D.A. Tryk, A. Fujishima, J. Electroanal. Chem. 1999, 473 , 173–178.

REFERENCES

211

28. S. Ferro, A. De Battisti , Electrochim. Acta 2002, 47 , 1641–1649. 29. K. Hayashi, S. Yamanaka, H. Watanabe, T. Sekiguchi, H. Okushi, K. Kajimura, J. Appl. Phys. 1997, 81 , 744–753. 30. J.B. Cui, J. Ristein, L. Ley, Appl. Phys. Lett . 1998, 81 , 429–432. 31. F. Maier, M. Riedel, B. Mantel, J. Ristein, L. Ley, Phys. Rev. Lett . 2000, 85 , 3472–3475. 32. D. Takeuchi, H. Kato, G.S. Ri, T. Yamada, P.R. Vinod, D. Hwang, C.E. Nebel, H. Okushi, S. Yamasaki, Appl. Phys. Lett . 2005, 86 , 1–3. 33. P.K. Bachmann, W. Eberhardt, B. Kessler, H. Lade, K. Radermacher, D.U. Wiecher, H. Wilson, Diam. Rel. Mat . 1996, 5 , 1378–1383. 34. M. Tachiki, T. Fukuda, K. Sugata, H. Seo, H. Umezawa, H. Kawarada, Jpn. J. Appl. Phys. 2000, 39 , 4631–4632. 35. K.B. Holt, A.J. Bard, Y. Show, G.M. Swain. J. Phys. Chem. B 2004, 108 , 15117–15127. 36. D.A. Tryk, K. Tsunozaki, T.N. Rao, A. Fujishima, Diam. Rel. Mat . 2001, 10 , 1804–1809. 37. M.N. Latto, D.J. Riley, P.W. May, Diam. Rel. Mat . 2000, 9 , 1181–1183. 38. M.N. Latto, G. Pastor-Moreno, D.J. Riley, Electroanalysis 2004, 16 , 434–441. 39. I. Duo, A. Fujishima, C. Comninellis, Electrochem. Comm. 2003, 5 , 695–700. 40. D. Becker, K. J¨uttner, J. Appl. Electrochem. 2003, 33 , 959–967. 41. D. Becker, K. J¨uttner, New Diam. Front. Carbon Technol . 2003, 13 , 67–78. 42. D. Becker, K. J¨uttner, Electrochim. Acta 2003, 49 , 29–39. 43. L. Codognoto, S.A.S. Machado, L.A. Avaca, Diam. Rel. Mat . 2002, 11 , 1670–1675. 44. V.A. Pedrosa, L. Codognoto, L.A. Avaca, Quim. Nova 2003, 26 , 844–849. 45. V.A. Pedrosa, S.A.S. Machado, L.A. Avaca, Anal. Lett . 2006, 39 , 1955–1965. 46. S. Ferro, M. Dal Colle, A. De Battisti, Carbon 2005, 43 , 1191–1203. 47. M. Szameitat, X. Jiang, W. Beyer, Appl. Phys. Lett . 2000, 77 , 1554–1556. 48. J. Ristein, F. Maier, M. Riedel, J.B. Cui, L. Ley, Phys. Stat. Sol. (A) 2000, 181 , 65–76. 49. M.C. Rossi, F. Spaziani, S. Salvatori, G. Conte. Phys. Stat. Sol. A 2003, 199 , 71–76. 50. J. Ristein, M. Riedel, L. Ley, J. Electrochem. Soc. 2004, 151 , E315–E321. 51. J. Ristein, W. Zhang, L. Ley, Phys. Rev. E 2008, 78 , 041602–1–041602-8. 52. J.B. Cui, J. Ristein, L. Ley, Phys. Rev. B 1999, 59 , 5847–5856. 53. S. Kulesza, J. Patyk, F. Rozploch, Chem. Phys. Lett . 2004, 391 , 56–59. 54. M. Hupert, A. Muck, J. Wang, J. Stotter, Z. Cvackova, S. Haymond, Y. Show, G.M. Swain, Diam. Rel. Mat . 2003, 12 , 1940–1949. 55. J. Wang, F. Liu, H. Chen, D. Chen, Mater. Chem. Phys. 2009, 115 , 590–598. 56. G. Pastor-Moreno, D.J. Riley, Electrochem. Comm. 2002, 4 , 218–221. 57. Y.V. Pleskov, Russ. J. Electrochem. 2002, 38 , 1275–1291. 58. S. Wang, G.M. Swain, J. Phys. Chem. C 2007, 108 , 3986–3995. 59. Y.V. Pleskov, J. Anal. Chem., 2000, 55 , 1045–1050. 60. V.A. Pedrosa, H.B. Suffredini, L. Codognoto, S.T. Tanimoto, S.A.S. Machado, L.A. Avaca, Anal. Letters 2005, 38 , 1115–1125. 61. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191–199. 62. M.A.Q. Alfaro, S. Ferro, C.A. Martinez-Huitle, Y.M. Vong, J. Braz. Chem. Soc. 2006, 17 , 227–236. 63. C.A. Martinez-Huitle, S. Ferro, Chem. Soc. Rev . 2006, 35 , 1324–1340. 64. L.S. Andrade, L.A.M. Ruotolo, R.C. Rocha-Filho, N. Bocchi, S.R. Biaggio, J. Iniesta, V. Garc´ıa-Garc´ıa, V. Montiel, Chemosphere 2007, 66 , 2035–2043.

212

CATHODIC TREATMENT OF BDD ELECTRODES AND THEIR USE IN ELECTROANALYSIS

65. C.A. Mart´ınez-Huitle, M.A. Quiroz, C. Comninellis, S. Ferro, A. De Battisti, Electrochim. Acta 2004, 50 , 949–956. 66. L.S. Andrade, R.C. Rocha-Filho, N. Bocchi, S.R. Biaggio, J. Iniesta, V. Garc´ıa-Garc´ıa, V. Montiel, J. Hazard. Mater. 2008, 153 , 252–260. 67. A. Kapalka, G. F´oti, C. Comninellis, Electrochim. Acta 2009, 54 , 2018–2023. 68. R.A. Medeiros, A.E. Carvalho, R.C. Rocha-Filho, O. Fatibello-Filho, Anal. Lett . 2007, 40 , 3195–3207. 69. R.A. Medeiros, A.E. Carvalho, R.C. Rocha-Filho, O. Fatibello-Filho, Quim. Nova 2008, 31 , 1405–1409. 70. R.A. Medeiros, A.E. Carvalho, R.C. Rocha-Filho, O. Fatibello-Filho, Talanta 2008, 76 , 685–689. 71. T.A. Ivandini, T.N. Rao, A. Fujishima, Y. Einaga, Anal. Chem. 2006, 78 , 3467–3471. 72. T.A. Ivandini, K. Honda, T.N. Rao, A. Fujishima, Y. Einaga, Talanta 2007, 71 , 648–655. 73. L.S. Andrade, R.C. Rocha-Filho, Q.B. Cass, O. Fatibello-Filho, Electroanalysis 2009, 21 , 1475–1480. 74. D.A. Tryk, T. Kondo, A. Fujishima, in Diamond Electrochemistry (A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, eds.), Tokyo, Elsevier, 2005, pp. 174–217. 75. V. Fisher, D. Gandini, S. Laufer, E. Blank, C. Comninellis, Electrochim. Acta 1998, 44 , 521–524. 76. A. Perret, W. Haenni, N. Skinner, X.M. Tang, D. Gandini, C. Comninellis, B. Correa, G. Foti, Diam. Rel. Mat . 1999, 8 , 820–823. 77. J.M.P. Carrazon, A.D. Recio, L.M.P. Diez, Talanta 1992, 39 , 631–635. 78. L. Codognoto, S.T. Tanimoto, V.A. Pedrosa, H.B. Suffredini, S.A.S. Machado, L.A. Avaca, Electroanalysis 2006, 18 , 253–258. 79. V.A. Pedrosa, L. Codognoto, L. A. Avaca, J. Braz. Chem. Soc. 2003, 14 , 530–535. 80. G.S. Garbellini, G.R. Salazar-Banda, L.A. Avaca, J. Braz. Chem. Soc. 2007, 18 , 1095–1099. 81. D.G. Crosby, Pure Appl. Chem. 1981, 53 , 1051–1080. 82. M.V.S Pinheiro, M.N. Oliveira, A.L.B. Penna, A.Y. Tamine, Int. J. Dairy Technol . 2005, 4 , 193–199. 83. H. Dejmkova, M. Scampicchio, J. Zima, J. Barek, S. Mannino, Electroanalysis 2009, 21 , 1014–1018. 84. T.A. Ivandini, B.V. Sarada, T.N. Rao, A. Fujishima, Analyst 2003, 128 , 924–929. 85. C. Terashima, T.N. Rao, B.V. Sarada, A. Fujishima, Anal. Chem. 2002, 74 , 895–902. 86. C.D. Souza, O.C. Braga, I.C. Vieira, A. Spinelli, Sens. Actuators B 2008, 135 , 66–73. 87. E.R. Sartori, R.A. Medeiros, R.C. Rocha-Filho, O. Fatibello-Filho, J. Braz. Chem. Soc. 2009, 20 , 360–366. 88. B.C. Lourenc¸a˜ o, R.A. Medeiros, R.C. Rocha-Filho, L.H. Mazo, O. Fatibello-Filho, Talanta 2009, 78 , 748–752. 89. E.F. Batista, E.R. Sartori, R.A. Medeiros, R.C. Rocha-Filho, O. Fatibello-Filho, Anal. Letters 2010, 43 , 1046–1054. 90. R.T.S. Oliveira, G.R. Salazar-Banda, V.S. Ferreira, S.C. Oliveira, L.A. Avaca, Electroanalysis 2007, 19 , 1189–1194. 91. K.B. Holt, G. Sabin, R.G. Compton, J.S. Foord, F. Marken, Electroanalysis 2002, 14 , 797–803.

Part IV

Industrial Applications

9 Use of Boron-Doped Diamond Electrode in Electrochemical Generation and Applications of Ferrate Virender K. Sharma, Enric Brillas, Ignasi Sir´es, and Karel Bouzek

9.1

INTRODUCTION

Iron commonly exists in the zero, +2, and +3 oxidation states; however, in certain environments, higher oxidation states of iron such as +4, +5, and +6 can also be obtained [1–4]. Fe(IV) and Fe(V) species have been proposed as reactive intermediates in selective oxygenation of hydrocarbons by iron-induced activation of hydrogen peroxide in organic solvents [5,6]; the ferryl (FeIV = O) and perferryl (FeV = O) species may play an important role in the oxygen activation and transfer reactions mediated by heme and nonheme iron proteins [7–11]. In recent years, iron in the +6 oxidation state, commonly called ferrate (FeVI O4 2− ), has been of great interest because of its role as an oxidant and hydroxylating agent in industrial and water treatment processes, such as the development of a “super iron” battery, the green chemistry synthesis, and the nonchlorine oxidation/disinfection of aqueous effluents for pollutant remediation [12–14]; Fe(VI) provides an environmentally benign, high-energy density battery cathode [15–17], whereas selective oxidations by Fe(VI) can be utilized for synthesizing organic compounds without the release of toxic by-products [18].

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

215

216

USE OF BORON-DOPED DIAMOND ELECTRODE

The aqueous ferrate solutions have a distinctive violet color similar to that of solutions of the permanganate ion. The alkaline solutions of the ferrate ion show a maximum at 510 nm (ε = 1150 ± 25 M−1 cm−1 ) and a shoulder between 275 and 320 nm [19]. The ferrate ion is a powerful oxidizing agent in aqueous media. Under acidic conditions, the reduction potential of the ferrate ion is relatively high compared to that of other oxidants used as disinfectants in water treatment processes, as shown in Table 9.1 [19,20]; however, the reduction potential strongly decreases under more alkaline conditions (Table 9.1). The spontaneous decomposition of ferrate in water leads to the formation of molecular oxygen according to Reaction (9.1) [21]: 2FeO4 2− + 5H2 O → 2Fe(OH)3 + 3/2 O2 + 4OH−

(9.1)

The decomposition rate of ferrate is strongly dependent on the initial ferrate concentration, temperature, pH, and even on the surface properties of the hydrous iron oxide formed upon decomposition. The ferrate ion is an emerging water treatment oxidant, disinfectant, and coagulant, which can address the concerns on disinfection by-products (DBPs) associated with currently used chemicals, such as free chlorine, chloramines, and ozone [22]. Like ozone, Fe(VI) does not react with the bromide ion, and so, the carcinogenic bromate ion is not produced during the treatment of bromide-containing water [22]. Several studies on the reaction of ferrate with a series of pollutants including sulfide, thiourea, cyanides, amines, phenols, and anilines have shown that the destruction of pollutants by ferrate is achieved in seconds to minutes; moreover, it leads to the formation of relatively nontoxic by-products [13]. The reaction rates are pH-dependent; thus, the half-lives determined for the pollutant removal are also pH-dependent [23,24]. Ferrate can also degrade effectively emerging contaminants such as estrogens, bisphenol-A, and sulfonamide antimicrobials present in water [25–30]. Furthermore, the nontoxic byproduct generated from ferrate reaction (i.e., ferric oxide/hydroxide) acts as a powerful TABLE 9.1 Redox potentials for the oxidants/disinfectants used in water treatment [19,20]. Reaction

E ◦ (V/SHE)

OH + H+ + e− ⇔ H2 O OH + e− ⇔ OH− FeO4 2− + 8H+ + 3e− ⇔ Fe3+ + 4H2 O FeO4 2− + 4H2 O + 3e− ⇔ Fe(OH)3 + 5OH− O3 + 2H+ + 2e− ⇔ O2 + H2 O O3 + H2 O + 2e− ⇔ O2 + 2OH− H2 O2 + 2H+ + 2e− ⇔ 2H2 O H2 O2 + 2e− ⇔ 2OH− MnO4 − + 4H+ + 3e− ⇔ MnO2 + 2H2 O MnO4 − + 8H+ + 5e− ⇔ Mn2+ + 4H2 O MnO4 − + 2H2 O + 3e− ⇔ MnO2 + 4OH− HClO + H+ + 2e− ⇔ Cl− + H2 O ClO− + H2 O + 2e− ⇔ Cl− + 2OH− ClO4 − + 8H+ + 8e− ⇔ Cl− + 4H2 O Cl2 + 2e− ⇔ 2Cl− O2 + 4H+ + 4e− ⇔ 2H2 O O2 + 2H2 O + 4e− ⇔ 4OH− ClO2 + e− ⇔ ClO2 −

2.80 1.89 2.20 0.70 2.08 1.24 1.78 0.88 1.68 1.51 0.59 1.48 0.84 1.39 1.36 1.23 0.40 0.95

Oxidant Hydroxyl Radical

• •

Ferrate (VI) Ozone Hydrogen peroxide Permanganate

Hypochlorite Perchlorate Chlorine Dissolved Oxygen Chlorine Dioxide

9.2 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH IRON ANODES

217

coagulant that is suitable for the removal of metals, nutrients, radionuclides, and humic acids [31–33]. Because of the wide range of applications of ferrate, the most economic production ways for individual applications have been sought. Basically, there exist three major approaches for the synthesis of ferrate: (1) dry thermal synthesis, (2) wet chemical synthesis, and (3) electrochemical synthesis. In the thermal synthesis, ferric oxides are heated together with an alkali metal oxide or peroxide to obtain solid ferrate. Different dry syntheses have recently been reviewed [34]. In the wet chemical synthesis, the ferric ion is converted into the ferrate ion by oxidation with hypochlorite in a highly alkaline environment [35]. Ozone and oxone™ (a mixture of K2 SO4 , KHSO4 , and KHSO5 ) can be used instead of hypochlorite to synthesize ferrate [36,37]. The electrochemical synthetic procedure generally occurs in a concentrated solution of alkali metal hydroxides. Traditionally, two kinds of electrochemical setups have been employed, based on the use of (1) inert anodes (e.g., Pt) able to oxidize soluble Fe3+ ions and (2) iron-based anodes (e.g., grey cast iron, white cast iron, steel, mild steel, etc.) that act as the iron source. More recently, boron-doped diamond (BDD) electrode has been applied to the synthesis of ferrate. The present contribution summarizes different electrochemical synthetic approaches for the generation of the ferrate ion. Several descriptions have been recently reviewed [38]. Hence, this chapter briefly reviews the electrochemical synthetic procedures that are available but mainly focuses on the use of the BDD anode for the electrogeneration of the ferrate ion, as well as on the recent applications of this methodology.

9.2 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH IRON ANODES Several studies have been addressed to overcome some of the difficulties associated with the electrochemical synthesis of ferrate [16,17,38–41]. Problems include the formation of a residual passive film on the electrode surface and the extent to which the competitive oxygen evolution reaction (OER) is given at the potential at which ferrate is formed. The current efficiency for the ferrate(VI) generation is very sensitive to the type of electrode pretreatment applied, as well as to the reaction conditions. Therefore, main attention has been paid to the chemical composition, geometry, and mode of activation of the anode, as well as to the electrolyte composition in order to optimize the synthesis of ferrate [38,41]. The influence of other experimental parameters including the temperature, applied current, and electrolysis time has been assessed as well. Various studies have focused on establishing the suitable conditions for electrosynthesizing ferrate when using an iron anode. It has been demonstrated that the presence of high silicon content in grey cast iron allows enhancing the ferrate production, reaching current yields in the range of 20–40% [42]. As can be seen in Figure 9.1, alloys produced by centrifugation led to better results compared to molded alloys, and a content of 2.80% Si produced a current yield of 33%. The degree of porosity of the iron anode also has great influence on the process [43]. For example, Figure 9.2 illustrates the comparative production of ferrate using pellet electrodes. After 1 h of electrolysis, the current yield was much higher using the foil electrode when working at the lowest current density, whereas the pressed iron powder gave the highest yield working over 5.5 mA cm−2 current density [43]. In the latter case, the porous structure of the anode favors the dissolution process, thus accelerating the transformation of the dissolved iron

218

USE OF BORON-DOPED DIAMOND ELECTRODE

40

Current yield (%)

30

20

10

0 2.0

2.4

2.2

2.6

3.2

3.0

2.8

Si content (wt %) Figure 9.1 Dependence of the current yield with the silicon content in grey cast iron for twohour ferrate production using a U-cell with a glass diaphragm. Volume of anodic solution: 50 cm3 ; electrode area: 30 cm2 ; current density: 17–25 mA cm−2 . () Centrifugation, () molded. (Reprinted with permission from Ref. 42.)

100 100 80 Current yield (%)

Current yield (%)

80

60

Pellet

60 40 20

Foil 40

0 0

10 20 30 40 50 60 70 80 j (mA cm–2)

20 Pellet Foil 0

0

100

200 j (mA

300

400

500

cm–2)

Figure 9.2 Current yield as a function of current density after 1 h of electrolysis for (•) iron foil and () pellet electrodes. The hollow circles (◦) correspond to the current yields computed by taking the zero time at the instant at which the electrode potential reached 0.6 V for the pellet electrode. The inset shows the plot for the low current density region. (Reprinted with permission from Ref. 43.)

9.2 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH IRON ANODES

219

into ferrate. It is then evident that the maximum current yield varies depending on the physical appearance of the anode used in the electrolysis. Recently, rapid electrochemical preparation of Na2 FeO4 was studied in detail by using an iron wire gauze anode [44]. The ferrate yield (η) was determined upon variation of the NaOH concentration, anodic current density, and electrolysis time. The increased ratio of effective anode surface area to anolyte volume using a thinner anodic chamber enhanced the production rate of ferrate in the anolyte. The dependence of the ferrate concentration, apparent ferrate yield (ηapp ), NaOH concentration, and Fe(III) concentration versus the electrolysis time is presented in Figure 9.3. The curve for the ηapp shows a vertical step because it represents the 1h-average apparent yield. The highest ηapp occurred during the first hour, which corresponded to 0.148 M ferrate. The maximum ferrate concentration (=0.48 M) was obtained at 6 h. But one of the major causes of relatively low ferrate production was the formation of a passive layer of iron oxide on the iron anode surface. Hence, the study was made using anisomeric square pulse with fluctuating frequency to favor the continuous activation of the anode, aiming at obtaining a high ferrate yield [45]. The results in Figure 9.4, which were obtained in a cylindrical electrolytic bath with bipolar membranes, show that the ferrate concentration increased over time in all cases, whereas the corresponding current efficiency always exhibited a progressive decay. This suggests that it was though possible to destroy the passive iron oxide film, but the formation of the passive layer still occurred. Therefore, a low current efficiency and a gradual decrease of the ferrate generation rate were observed [45]. The optimum fluctuating frequency was 2 Hz, as depicted in the inset of Figure 9.4b. Table 9.2 summarizes the current efficiency values reported for the ferrate production under various conditions, using different iron-containing anodes [46–55]. It seems clear that the nature of the anode and the carbon content have a large influence on the current efficiency. Based on the cyclic voltammograms for the formation of ferrate from metallic iron, the following mechanism involving Reactions (9.2) through (9.7) has been proposed 1.0 16

0.8

SFe-[FeO42–]

0.6

[OH–] happ

15

14 0.4

[OH–] (M)

Content(mM) or happ

[FeO42–]

13 0.2 12

0.0 0

1

2

3

4 5 Time (h)

6

7

8

Figure 9.3 Time course of some species and ηapp during a continuous electrolysis using 14-layer iron gauzes for 8 h at 35.0 ± 0.8◦ C. Initial anolyte: 16 M NaOH; volume: 75 cm3 ; estimated total anode surface area: 690 cm2 . The ηapp is the average apparent yield for 1-h intervals. (Reprinted with permission from Ref. 44.)

USE OF BORON-DOPED DIAMOND ELECTRODE

60

30

(a)

25

55

50

20 h (%)

Oxidant(mM FeO42– )

60

(b)

15

2Hz

h (%)

220

50 45

40

0.2Hz

20Hz

0Hz 40 Frequency/Hz

30

10 2Hz 20Hz 0.2Hz 0Hz

5

2Hz 20Hz 0.2Hz 0Hz

20

0 0

1

2

4 3 Time (h)

5

0

6

1

2

3 4 Time (h)

5

6

7

Figure 9.4 (a) Concentration of electrogenerated FeO4 2− and (b) current efficiency with time at 5 mA cm−2 by varying the applied frequency between 0 and 20 Hz for a CS–CMC bipolar membrane electrolysis cell. (Reprinted with permission from Ref. 45.)

[39–41,48,56–60]. Reactions (9.2) and (9.3) represent both the active dissolution of the anode material and the surface layer restructuration, respectively. The reaction of the passive layer of FeOOH with OH− ions results in the breakage of the iron oxide surface and allows the continuous dissolution of the anode material to form FeO2 − by Reaction (9.4). This can be given simultaneously to the oxidation to the FeO3 2− ion from Reaction (9.5a), followed by disproportionation to the ferrate ion from Reaction (9.6). However, the formed FeO2 − ions dissolved in the anolyte may be further anodically oxidized to yield FeO3 2− via Reaction (9.5b). The concerned reaction of oxygen evolution occurs in the vicinity of ferrate formation zone from Reaction (9.7). Fe + 2OH− → Fe(OH)2 + 2e−

(9.2)

−

Fe(OH)2 + OH → FeOOH + H2 O + e

−

FeOOH + OH− → FeO2 − + H2 O

(9.3) (9.4)

FeOOH + 3OH− → FeO3 2− + 2H2 O + e−

(9.5a)

FeO2 − + 2OH− → FeO3 2− + H2 O + e−

(9.5b)

3FeO3 2− + H2 O → 2FeO2 − + FeO4 2− + 2OH−

(9.6)

−

2OH → H2 O + 1/2O2 + 2e

−

(9.7)

This mechanism is generally in good agreement with the oxidation steps undergone by metallic iron to yield Fe(III), but some controversy exists concerning the formation of the ferrate ion from Fe(III). The electrochemical impedance spectroscopy (EIS) study of the system suggests an alternative four-step mechanism; after the oxidation of the zero-valent iron to FeOOH (Reactions 9.2 and 9.3), the reaction can proceed according to Reactions (9.8) and (9.9) [61]: FeOOH + 3OH− → (FeO3 − )ads + 2H2 O + 2e− −

−

−

3(FeO3 )ads + 2OH → FeO2 + 2FeO4

2−

+ H2 O

(9.8) (9.9)

9.2 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH IRON ANODES

221

TABLE 9.2 Effect of the operation conditions on the current efficiency for the ferrate production with iron-based anodes. Current efficiency (%)

Operation conditions

Reference

12

Anode: Silver steel with 0.08% C j = 1 mA cm−2 [NaOH] = 9.0 M T = 25◦ C Electrolysis time = 15 min

[46,47]

21

Anode: Grey cast iron j = 2 mA cm−2 [NaOH] = 14 M T = 20◦ C Electrolysis time = 60 min Anode: Mild steel j = 4 mA cm−2 [NaOH] = 14 M T = 20◦ C Electrolysis time = 180 min Anode: Porous magnetite j = 3.3 mA cm−2 [NaOH] = 16 M T = 30◦ C Electrolysis time = 300 minutes Anode: White cast iron j = 5 mA cm−2 [NaOH] = 14 M T = 20◦ C Electrolysis time = 60 min Anode: Pure iron j = 4.4 mA cm−2 [NaOH] = 14 M T = 50◦ C Electrolysis time = 60 min Anode: Grey cast iron j = 4.54 mA cm−2 [NaOH] = 14 M T = 20◦ C Electrolysis time = 60 minutes Anode: Silver steel with 0.9% C j = 1 mA cm−2 [NaOH] = 10 M T = 25◦ C Electrolysis time = 5 min

[48]

42

52.3

64

64

68.5

>70

[49]

[54]

[50]

[51]

[55]

[46,47]

More recently, spectroscopic approaches have also been used to gain information on the processes taking place at the electrode surface, which eventually will help to produce ferrate in a more efficient manner [62]. One of the main drawbacks of the electrochemical synthesis is the instability of ferrate in aqueous medium. Therefore, the molten hydroxides approach, which excludes altogether water from the environment, has been applied to perform the direct anodic oxidation process with an iron anode or with soluble iron species using a platinum anode [63–65]. Under molten hydroxides conditions, the ferrate produced is expected to be in dry stable solid form in the cooled-down reaction mixture. Furthermore, the previously

222

USE OF BORON-DOPED DIAMOND ELECTRODE

proposed mechanism involves a chemical reaction; hence, the increased temperature in the molten state can accelerate the overall generation of ferrate. The electrochemical synthesis of ferrate using this approach was therefore performed at 170, 180, and 200◦ C; the anode passivation was less pronounced during the ferrate(VI) generation in comparison with the experiments in an aqueous environment [63–65], which is important for the continuous production of ferrate. However, the problem related to the competing OER found in aqueous solution was noticed in the molten environment as well. Furthermore, electrolysis operation temperature became the key factor in the process. An excessive temperature increase was detrimental regarding the ferrate yield because of the limited thermal stability of ferrate [66,67]. Conversely, only a few studies have discussed the use of an inert electrode for synthesizing ferrate, although it is advantageous because a bulk Fe3+ solution can be used, which allows avoiding the impact of the anode dissolution kinetics on the overall process. Moreover, only three electrons are needed to generate the ferrate ion, instead of the six electrons required when using zero-valent (i.e., metallic) iron, which reduces the electrical consumption to half the value. The reported synthesis of ferrate using an inert electrode is reviewed hereafter.

9.3 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH INERT ANODES In a previous study, cyclic voltammetric (CV) curves were collected using a rotating ringdisk electrode (RRDE), as well as platinum electrodes under stationary conditions [38]. Solutions of 0.03 M Fe(II), Fe(III), and Fe(VI) in a 14 M NaOH electrolyte solution were investigated. CV curves were also recorded in the 14 M NaOH solution using a platinum electrode with the surface covered by cathodically deposited iron. During the potential scan to more positive values, a current plateau in the potential region from +0.060 to +0.675 V versus Hg/HgO was observed in all the studied solutions. This plateau was ascribed to the oxidation of FeO2 − to FeO4 2− . Importantly, the current density of the plateau strongly depended on the electrolyte temperature, whereas it was independent of the rotation speed. This suggests that the mechanism of ferrate formation involves at least one chemical step. Recently, CV curves for the Fe(VI)/Fe(III) system were investigated using a SnO2 -Sb2 O3 /Ti anode in strong basic solution [68,69]. Figure 9.5 shows steady-state CVs obtained with solutions of FeO2 − in 14 M NaOH. They suggest the formation of an intermediate during the oxidation of FeO2 − to FeO4 2− . The slope of the linear relationship in the inset of Figure 9.5 is about 0.028; accordingly, two electrons are transferred in the oxidation step, which is associated with the formation of Fe(V) from Reaction (9.10) and its disproportionation to give the ferrate ion via Reaction (9.11): FeO2 − + 4OH− → FeO4 3− + 2H2 O + 2e− 3FeO4

3−

+ 2H2 O → 2FeO4

2−

−

+ FeO2 + 4OH

(9.10) −

(9.11)

This system was also examined using a powder microelectrode as the anode, which yielded similar conclusions [69]. Not only FeO2 − , but also Fe2 O3 and Fe(OH)3 were used as iron sources, yielding no ferrate ion instead.

9.4 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH BORON-DOPED DIAMOND ANODE

223

(a) 0.15 scan rate: 1 m V s–1

j (mA cm–2)

0.10

0.05

0.00

0.0

0.2

0.4

0.6

0.8

j (V) vs Hg/HgO

j (mA cm–2)

0.08

j(V)vs Hg/HgO

(b) 0.12 0.65

0.60

0.55 –2

–1

0

1

2

log (il(i1–i))

0.04

scan rate: 0.2 m V s–1 0.00

0.1

0.2

0.3 0.4 0.5 j (V) vs Hg/HgO

0.6

0.7

Figure 9.5 Steady-state cyclic voltammogram at (a) 1 mV s−1 and (b) 0.2 mV s−1 at a SnO2 –Sb2 O3 electrode in 14 M NaOH containing 0.02 M FeO2 − . The arrows indicate the scan direction. The inset gives the linear relationship of ϕ vs. log(i/(il − i)). (Reprinted from Ref. 68.)

9.4 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH BORON-DOPED DIAMOND ANODE 9.4.1

Acidic Medium

Original work on the use of the BDD anode for the electrochemical generation of ferrate was carried out in 0.1 M HClO4 [70]. A 0.006 M FeSO4 solution was used to collect the CVs given in Figure 9.6a at different scan rates. The voltammograms exhibited three peaks: two anodic (AI and AII) and one cathodic (CI). The redox couple Fe3+ /Fe2+ is responsible for peaks AI (∼+1.1 V versus Ag/AgCl) and CI (∼+0.4 V versus Ag/AgCl). The peak potential corresponding to AII varied from +2.3 to +2.75 V versus Ag/AgCl

224

USE OF BORON-DOPED DIAMOND ELECTRODE

(a) 4m

CI

0 AI

–4m j (A cm–2)

(a)

(b)

–8m (c) –12m (d) –16m

(e)

–20m

(f) (g)

–24m 3.0

AII 2.5

2.0 1.5 1.0 E (V) vs Ag/AgCl

0.5

0.0

(b) CI 0.0 (a) (b)

AI

j (A cm–2)

–1.0m (c) –2.0m

–3.0m AII (d)

–4.0m 2.5

2.0

1.5 1.0 E (V) vs Ag wire

0.5

0.0

Figure 9.6 Cyclic voltammograms at a boron-doped diamond (BDD) electrode. Plot (A): (a) 0.1 M HClO4 alone, or in the presence of 6 mM FeSO4 at scan rates of (b) 10, (c) 50, (d) 100, (e) 250, (f) 500, and (g) 1000 mV s−1 . The electrochemical cell was a single compartment cell with the surface of the BDD electrode exposed at the bottom of the cell through an O-ring supported opening with a Pt mesh counter electrode, and a Ag/AgCl reference electrode (in saturated KCl). Plot (B): (a) acetonitrile with 0.1 M LiClO4 alone, (b) as in (a) but with addition of 6 mM FeSO4 , (c) as in (b) but with addition of 1.13 M of water, and (d) as in (b) but with addition of 2.83 M of water. Scan rate: 100 mV s−1 . A Pt mesh counter electrode and Ag wire pseudo-reference electrode were used. (Reprinted with permission from Ref. 70.)

depending on the scan rate, and it was assigned to the generation of the ferrate ion from the oxidation of Fe3+ in accordance with Reaction (9.12): Fe3+ + 4H2 O → FeO4 2− + 8H+ + 3e−

(9.12)

9.4 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH BORON-DOPED DIAMOND ANODE

225

This is supported by the known thermodynamic potential in acidic medium (see Table 9.1) [71]. Interestingly, no cathodic peak was seen for the reduction of Fe(VI) to Fe(III) and the amount of oxygen in the system did not affect the peak. Furthermore, the AII peak was much larger than what expected from Reaction (9.12). This was accounted for by the fact that the generated ferrate ion rapidly decomposed in acidic medium to Fe3+ and oxygen by Reaction (9.13): 2FeO4 2− + 5H2 O → 2Fe3+ + 13/2O2 + 10H+

(9.13)

A similar set of CV experiments was performed by using acetonitrile instead of water as the solvent in order to provide evidence for Reaction (9.12). The CVs of Figure 9.6b show a decrease in the overpotential required by the Fe3+ /Fe2+ redox couple, along with a new anodic peak, AII, at a less positive potential compared to that required for the ferrate formation. This demonstrates the important role of the water content in the system, as deduced from Reaction (9.12). The authors concluded that the initial apparent number of electrons transferred resulted in 3, which further confirmed the occurrence of Reaction (9.12). 9.4.2

Alkaline Medium

The work on the electrochemical generation of ferrate ion using the BDD electrode in alkaline medium has been performed only recently [72,73]. The time course of generated ferrate was monitored in experiments using both, stainless steel and BDD electrodes, as depicted in Figure 9.7. The concentration of ferrate was determined by using the chromite method [35]. When the stainless steel electrode was initially used for 2 h, the concentration of the generated ferrate increased progressively and then the production slowed down to reach a constant value. This can be related to the existence of Reaction (9.14): Fe + 8OH− → FeO4 2− + 4H2 O + 6e−

(9.14)

Oxidant (mM FeO42–)

0.25 0.2 0.15 Anode: SS 0.1

Anode: BDD

0.05 0 0

50

100 150 Time (min)

200

250

Figure 9.7 Variation of the ferrate concentration with time during the electrolysis of hydroxide solutions with () stainless steel and () BDD electrodes. Experimental conditions: 14 M NaOH, at current density of 13 mA cm−2 and 30◦ C. (Reprinted with the permission from Ref. 72.)

226

USE OF BORON-DOPED DIAMOND ELECTRODE

The replacement of the stainless steel anode by a BDD one in the solution treated for 2 h resulted in an increase in the concentration of the ferrate ion; this was possibly due to the oxidation of the iron species accumulated in the solution during the preliminary electro-oxidation process. This finding is interesting because it suggests that BDD can increase the efficiency of the ferrate electrogeneration process. In such studies, the cell voltage did not change, which indicates that the usual phenomenon of passivation and deterioration of the electrode did not occur. The efficiency was found to be dependent on the electrical charge consumed [72]. Experiments were also carried out using the BDD anode, with iron powder and Fe(III) hydroxide as the iron source, at different current densities. Figure 9.8 illustrates that iron powder turned out to be the preferred primary iron material. This suggests that the process to generate ferrate using BDD is determined by the availability of the iron species. Figure 9.8 also shows that an increase in the current density led to a larger production of ferrate. This phenomenon is similar to that observed for the synthesis of other oxidants using BDD (i.e., H2 O2 , S2 O8 2− , etc.) [74–77]. Hydroxyl radicals formed when using BDD electrodes may also be coresponsible for the formation of ferrate in the BDD system [72]. The effect of other parameters, such as the concentration of hydroxide anions and the operation temperature, on the efficiency of ferrate production was assessed as well [73]. The effect of the content of hydroxide anions is shown in Figures 9.9a and 9.9b. The production of ferrate at ≤5 M NaOH was not significant, whereas continuous ferrate yield was obtained as the concentration of hydroxide salt increased up to 14 M NaOH (Figure 9.9a). This may be related to the stability of ferrate after its generation in strong alkaline solution. Figure 9.9b demonstrates that the use of KOH instead of NaOH as the electrolyte markedly increased the amount of ferrate produced. The different stability of the generated ferrate salts, K2 FeO4 and Na2 FeO4 , in the system under alkaline conditions may explain these results. 1.6

Oxidant(mM FeO42−)

1.4 1.2 1 0.8 0.6 0.4 0.2 0 0

500

1000

1500

2000

Current density (A m-2) Figure 9.8 Variation of the ferrate concentration with the current density in the electrolysis with BDD electrodes of hydroxide solutions () saturated with 9.4 mM Fe(OH)3 and () with iron-powder bed. Experimental conditions: 10 M KOH and temperature of 30◦ C. (Reprinted with permission from Ref. 72.)

9.4 ELECTROCHEMICAL GENERATION OF THE FERRATE ION WITH BORON-DOPED DIAMOND ANODE

Oxidant (mM) FeO42−)

0.1

227

(a)

0.08 0.06 0.04 0.02 0 0

Oxidant (mM FeO42−)

0.08

1

2 Q (Ah L–1)

3

4

1

2 Q (Ah L–1)

3

4

(b)

0.06

0.04

0.02

0 0

Figure 9.9 Variation of the ferrate concentration with the electric charged passed during the electrolysis of Fe(OH)3 solutions with BDD electrodes at 13 mA cm−2 at 30◦ C. Plot (a): () 5, () 10, and () 14 M NaOH. Plot (b): () 10 M NaOH and () 10 M KOH. (Reprinted with permission from Ref. 73.)

Figures 9.10a and 9.10b show the effect of the electric charge consumed in the electrolysis of iron hydroxide at several temperatures. The former figure presents the initial dependence of the ferrate concentration on the electric charge passed at all temperatures. This fact is very clear at electric charge ≤5 Ah L−1 . Figure 9.10b depicts the change of the ferrate generation yield with the temperature, reaching a maximum at 25◦ C. Two different processes—the solubility of Fe(III) and the stability of ferrate—may be occurring simultaneously to cause such dependence of the ferrate yield on the temperature. At low temperatures, the solubility of Fe(III) is low and a small amount of Fe(III) species is then available to be oxidized, causing a small production of ferrate; conversely, the stability of ferrate decreases with increasing temperature to give low yields of ferrate at too high temperatures (Figure 9.10b). In summary, the current density, electric charged passed (which is related to the electrolysis time), concentration of electrolyte, and temperature have a great influence on the efficiency of the electrosynthesis of ferrate with the BDD anode.

228

USE OF BORON-DOPED DIAMOND ELECTRODE

Oxidant (mM) FeO42−)

0.24

(a)

0.2 0.16 0.12 0.08 0.04 0 0

Oxidant (mM) FeO42−)

0.15

5

10

15 Q (Ah L–1)

20

25

30

(b)

0.1

0.05

0 10

20

30

40

50

60

Temperature (°C) Figure 9.10 Variation of the ferrate concentration (a) with the electric charge passed and (b) with the temperature, during the electrolysis of Fe(OH)3 solutions with BDD electrodes in 10 M KOH at 100 mA cm−2 . Plot (a), temperature: () 17, () 23, () 30, (*) 40, and () 57◦ C. Plot (b), ferrate concentrations obtained at an electric charge passed of 1.5 Ah L−1 . (Reprinted with permission from Ref. 73.)

9.5

APPLICATIONS

In contrast to the wide use of chemically produced ferrate, a very limited amount of work has been conducted on the applications of electrochemically generated ferrate. Examples reported in the field of fuel cells as well as in the remediation of water pollutants are described briefly in subsections below on the basis of the different types of electrodes used as the anode. 9.5.1

Common Inert Anodes

In recent years, direct methanol fuel cells (DMFCs) have shown applications in electronic equipments [78]. In alkaline solution, the anodic reaction of the DMFCs is described as follows: CH3 OH + 6OH− → CO2 + 5H2 O + 6e−

(9.15)

9.5 APPLICATIONS

229

Pt or Pt-based alloy anodes are commonly used for methanol oxidation, but these anodes become easily poisoned by the CO-like intermediates. Recent work demonstrated that addition of ferrate ion and Fe(III) in 1 M methanol and 12 M NaOH promoted both the catalytic activity and poison tolerance of Pt [78]. The assessment of the long-term stability of the methanol electro-oxidation in the presence of ferrate and Fe(III) ions under alkaline conditions was carried out through current-time curves at a steady potential. As can be seen in Figure 9.11, there is a fast current decay during the early stages; then, after ∼500 seconds the decay becomes slower in the four electrolytes, and the current progressively tends to zero during the rest of the electro-oxidation of methanol. The initial trends of the curves suggest poisoning of the electrocatalyst. The lowest current was obtained in the absence of ferrate or Fe(III) ions. Interestingly, the oxidation current in the electrolyte containing the decomposed ferrate ion was the largest compared to solutions with a stoichiometric amount of Fe(NO3 )3 and K2 FeO4 . These results were also in agreement with the CVs of the systems, indicating that the reduction product of ferrate ion greatly promoted the electrocatalytic activity of Pt for the oxidation of methanol in alkaline media. 9.5.2

Iron Anodes

The online preparation and use of ferrate in the wastewater treatment at a pilot scale has been only tested recently [79]. Steel was used as the anode in the pilot plant setup for the electrochemical production of ferrate. The steel had an iron content >99% and carbon content in the range of 0.10–0.12%. The thickness of each steel plate was 2 mm and a bipolar configuration of the electrodes was used. The optimum current density was 3.6 mA cm−2 . In the laboratory scale, the possible risk arising from the hydrogen

Normalized current (mA cm–2)

1.0

0.8

0.6

0.4

0.2 d a

0.0 0

b

c 5000

10000 Time (s)

15000

20000

Figure 9.11 Current–time curves for methanol electro-oxidation in 1 M CH3 OH + 12 M NaOH: (a) at −0.041 V versus Hg/HgO, (b) with a Fe(NO3 )3 loading of 203 mg L−1 , (c) with a K2 FeO4 loading of 164 mg L−1 , and (d) with the decomposed K2 FeO4 at −0.080 V versus Hg/HgO. (Reprinted with permission from Ref. 78.)

230

USE OF BORON-DOPED DIAMOND ELECTRODE

production in the electrochemical process was not observed [80]. However, such risks need to be examined for the electrochemical generation of ferrate at the large scale. In the treatment process, different parameters of water quality such as suspended solids (SS), chemical oxygen demand (COD), biological oxygen demand (BOD), and phosphate content were tested. The concentrations in tests were 242–730 mg L−1 for the SS, 523–1125 mg L−1 for the COD, 235–441 mg L−1 for the BOD, and 11.3–18.5 mg L−1 for the phosphate as total P. The results are presented in Figure 9.12. At a maximum ferrate dose of 0.04 mg L−1 , the average removal percentage achieved was 70% for the SS (Figure 9.12a), 40% for the phosphate (Figure 9.12b), 40% for the COD (Figure 9.12c), and 30% for the BOD (Figuer 9.12d). Interestingly, Table 9.3 shows that the performance of a low ferrate dose (0.03 mg Fe L−1 ) could be similar or even better than that obtained with a high amount of ferric sulfate (37 mg Fe L−1 ). The superior performance of ferrate is not surprising, taking into account that it is a much more powerful oxidant compared to the ferric ion. Another significant advantage of ferrate is related to the residual iron concentration in the final effluent. Table 9.3 shows that the use of ferrate leads to only 0.25 mg Fe L−1 in the effluent, whereas the iron concentration in the case of ferric sulfate was much higher (17.2 mg Fe L−1 ). Therefore, the use of a very low dose of ferrate(VI) would generate a much smaller volume and mass of the sludge than that from high doses of ferric salts. This would then be beneficial concerning the reduction of the cost of sludge handling. However, before implementation of the ferrate technology to a full-scale treatment of water and wastewater, the full assessment of the operation cost should be performed. 9.5.3

BDD Anode

Recently, the ferrate ion has been suggested to enhance the electrochemical oxidation of Acid Yellow 36 azo dye (AY36) using the BDD electrode in a system containing Fe(II) ions under acidic conditions [81]. The decolorization of the dye solution using different

100 (a)

80

P removal (%)

SS removal (%)

100

60 40 20 0

60 40 20 0

0

0.01

0.02

0.03

0.04

0.05

0

0.01

0.02

0.03

0.04

0.05

0.01

0.02

0.03

0.04

0.05

100

100 BOD removal (%)

COD removal (%)

(b)

80

(c)

80 60 40 20 0 0

0.01

0.02

0.03

0.04

0.05

(d)

80 60 40 20 0 0

Ferrate dose(mg Fe(VI) L–1)

Figure 9.12 Average removal with ferrate(VI) in the treatment of crude wastewater. (a) Suspended solids, (b) phosphate ion, (c) COD, and (d) BOD. (Reprinted with permission from Ref. 79.)

231

9.5 APPLICATIONS

TABLE 9.3 Comparative performance of crude sewage treatment with ferric sulfate and ferrate(VI)a [79]. Average percentage removal (%)

Residual iron and pH

Chemical and dose

SS

P

COD

BOD

Fe (mg L−1 )

pH

Ferrate(VI) (0.03 mg Fe L−1 ) Ferric sulfate (37 mg Fe L−1 )

79 ± 5

56 ± 1

50 ± 3

30 ± 5

0.25 ± 0.08

9.8 ± 0.1

78 ± 4

59 ± 2

54 ± 5

43 ± 6

17.2

–

a Crude sewage properties: [SS] = 730 mg L−1 , [P] = 18.5 mg L−1 , [COD] = 1125 mg L−1 , [BOD] = 388 mg L−1 , [Fe] = 1.0 mg L−1 , pH = 8.0.

1.0

AY36 (C1/Co)

0.8 (a) 0.6 (b) (c)

0.4

(d) 0.2 (e) 0.0 0

20

40

60

80

100

120

140

160

180

Time (min) Figure 9.13 Effect of the FeSO4 concentration over the decolorisation of 40 mg L−1 of Acid Yellow 36 under an anodic potential of 2.5 V. (a) Electrochemical oxidation process (EOP), and with FeSO4 content of (b) 12, (c) 8, (d) 6, and (e) 1 mM FeSO4 . (Reprinted with permission from Ref. 81.)

amounts of Fe(II) ions is presented in Figure 9.13. The electrochemical oxidation process (EOP) achieved only ∼41% of the AY36 degradation in 180 min. Comparatively, adding Fe(II) ions improved the degradation markedly under the same conditions. The added amount of the Fe(II) ions had a great influence on the degradation of the dye. It is noteworthy that an increase in the concentrations of Fe(II) ions decreased the degradation efficiency. The results obtained in Figure 9.13 were explained by performing CV measurements over the BDD electrode under the same conditions, as illustrated in Figure 9.14. The absence of Fe(II) ions does not yield any redox signal (curve a); similarly, the addition of 12 mM Fe(II) ions does not yield any redox signal (curve b), whereas a signal appeared from Fe(II) contents lower or equal to 8 mM. Three distinctive peaks could be observed: two anodic (AI and AII) and one cathodic (CI) (curve c). The Fe3+ /Fe2+ redox couple was associated with the AI and CI signals, while AII corresponded to the generation of the ferrate ion at ∼+2.3 V from Reaction (9.12). This findings are similar to those reported by Lee et al. (see Section 9.4.1) [70]. The CI peak increased with Fe(II) concentration decreasing from 6 to 1 mM (curves d and e).

232

USE OF BORON-DOPED DIAMOND ELECTRODE

Cl 0.000 (a) (b) (c) (d)

0.002

j (A cm–2)

0.004

AI

0.006 0.008

(e) AII

0.010 0.012 0.014 0.016 3.0

2.5

2.0

1.5 1.0 E (V) vs Ag/AgCl

0.5

0.0

Figure 9.14 Cyclic voltammograms at BDD in 0.1 M HClO4 for ferrate ion generation in the following conditions: (a) without FeSO4 , and with FeSO4 loading of: (b) 12, (c) 8, (d) 6, and (e) 1 mM. Scan rate: 100 mV s−1 . (Reprinted with permission from Ref. 81.)

H2O BDD Oxidation products

a) Oxidation products

H++e– Fe2+ b) BDD(•OH) f)

Pollutants

[Fe(VI)] d)

Fe2+

Pollutants

e) BDD c)

Fe3+

Fe3+

Fe3+

Figure 9.15 Proposed mechanism for the electrochemical oxidation of organic compounds with the simultaneous ferrate ion formation in the system. (Reprinted with permission from Ref. 81.)

Results of Figures 9.13 and 9.14 clearly demonstrate the role of Fe(II) ions in the enhancement of the electrochemical oxidation of AY36 using a BDD electrode. The mechanism given in Figure 9.15 suggests that the ferrate ion is the species responsible for such an enhancement. In the proposed mechanism, step (a) produces hydroxyl radical, which oxidizes the AY36 (step (b)). The generation of the ferrate ion occurs in step (c), so that AY36 can also be oxidized by this species (step (d)), which shows the enhancement effect of the Fe(II) ions. The Fe(II) ions are regenerated by the reduction of the Fe(III) ions in step (e). A certain amount of Fe(II) is oxidized by hydroxyl radicals (step (f)). It seems that steps (b) and (d) give rise to a synergistic effect for degrading AY36 at low concentrations of the Fe(II) ions by avoiding minimizing steps (e) and (f).

REFERENCES

9.6

233

CONCLUSIONS

Significant advances in terms of the optimization of parameters such as the anode material and operating conditions have been made regarding the electrochemical synthesis of ferrate. However, the mechanism of the Fe(III) oxidation to yield ferrate still needs to be fully understood. The use of classical electrochemical techniques may not be sufficient to explore the mechanism because the reaction of Fe(III) to Fe(VI) is overlapped by oxygen evolution. Hence, spectroscopic approaches are currently being used to gain insight on the processes occurring at the electrode surface. Two relatively new synthesis approaches—the employment of molten hydroxides as the electrolysis environment and the utilization of inert anodes—are currently being pursued to minimize the influence of the anode material and the oxygen evolution in synthesizing ferrate. Studies reporting the use of BDD anode to synthesize ferrate are forthcoming. The efficiency of the synthesis is greatly influenced by the availability of oxidizable iron species in the reaction solution. The potential role of ferrate in the electrochemical oxidation of methanol and dyes using inert anodes have been suggested, but mechanistic understanding of the system is still missing. The in situ electrochemical ferrate production method is promising for treating pollutants in water and wastewater treatment facilities. 9.7

ACKNOWLEGMENTS

V.K. Sharma would like to acknowledge the support of United States National Science Foundation (CHE 0706834). V.K. Sharma and K. Bouzek also acknowledge the partial support of NATO collaborative linkage grant (CBP.EAP.CLG.983119). K. Bouzek acknowledges the support of the Ministry of Education, Youth and Sports of the Czech Republic (project No. ME 890).

REFERENCES 1. J.D. Rush, B.H.J. Bielski, J. Am. Chem. Soc. 1986, 108 , 523–525. 2. C. Jeannot, B. Malaman, R. Gerardin, B. Oulladiaf, J. Solid State Chem. 2002, 165 , 266–277. 3. J.F. Berry, E. Bill, E. Bothe, S.D. George, B. Mienert, F. Neese, K. Wieghardt, Science 2006, 312 , 1937–1941. 4. F.T. De Oliveira, A. Chanda, D. Banerjee, X. Shan, S. Mondal, L. Que, Jr., E.L. Bominaar, E. M¨unck, T.J. Collins, Science 2007, 315 , 835–838. 5. H.C. Tung, C. Kang, D.T. Sawyer, J. Am. Chem. Soc. 1992, 114 , 3445–3455. 6. J.T. Groves, J. Inorg. Biochem. 2006, 100 , 434–447. 7. J. Kaizer, E.J. Klinker, N.Y. Oh, J.-U. Rohde, W.J. Song, A. Stubna, J. Kim, E. M¨unck, W. Nam, L. Que, Jr., J. Am. Chem. Soc. 2004, 126 , 472–473. 8. E.Y. Tshuva, S.J. Lippard, Chem. Rev . 2004, 104 , 987–1011. 9. S.V. Kryatov, E.V. Rybak-Akimova, S. Schindler, Chem. Rev . 2005, 105 , 2175–2226. 10. N.Y. Oh, Y. Suh, M.J. Park, M.S. Seo, J. Kim, W. Nam, Angew. Chem. Int. Ed . 2005, 44 , 4235–4239. 11. X. Shan, L. Que, Jr., J. Inorg. Biochem. 2006, 100 , 421–433. 12. J.D. Carr, P.B. Kelter, A. Tabatabai, D. Splichal, J. Erickson, C.W. McLaughlin, in Proceedings of the Fifth Conference on Water Chlorination: Chemistry, Environmental Impact and Health

234

13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

31. 32.

33. 34.

35. 36. 37. 38. 39. 40. 41.

USE OF BORON-DOPED DIAMOND ELECTRODE

Effects (R.L. Jolley, R.J. Bull, W.P. Davis, S. Katz, M.H. Roberts, V.A. Jacobs, eds.), Lewis Publishers, Chelsea, MI, 1985, Vol. 5, pp. 1285–1298. V.K. Sharma, Adv. Environ. Res. 2002, 6 , 143–156. V.K. Sharma, F. Kazama, J. Hu, A.K. Ray, J. Water Health 2005, 3 , 45–58. S. Licht, B. Wang, S. Ghosh, Science 1999, 285 , 1039–1042. S. Licht, X. Yu, in Ferrates: Synthesis, Properties, and Applications in Water and Wastewater Treatment (V.K. Sharma, ed.), Vol. 985, ACS Symposium Series, 2008, pp. 2–51. S. Licht, X. Yu, in Ferrates: Synthesis, Properties, and Applications in Water and Wastewater Treatment (V.K. Sharma, ed.), Vol. 985, ACS Symposium Series, 2008, pp. 197–256. L. Delaude, P. Laszlo, J. Org. Chem. 1996, 61 , 6360–6370. V.K. Sharma, in Advances in Control of Disinfection By-Products in Drinking Water Systems (A. Nikolau, L. Rizzo, H. Selcuk, eds.), Nova Science Publishers, New York, 2007, pp. 370–380. E. Brillas, I. Sir´es, M.A. Oturan, Chem. Rev . (doi. 9.1021/cr900136g). H. Goff, R.K. Murmann, J. Am. Chem. Soc. 1971, 93 , 6058–6065. V.K. Sharma, S.K. Mishra, N. Nesnas, Environ. Sci. Technol. 2006, 40 , 7222–7227. V.K. Sharma, Water Sci. Technol . 2004, 49 , 69–74. R.A. Yngard, V.K. Sharma, J. Filip, R. Zboˇril, Environ. Sci. Technol. 2008, 42 , 3005–309. J.Q. Jiang, J. Hazard. Mater. 2007, 146 , 617–623. V.K. Sharma, Chemosphere 2008, 73 , 1379–1386. V.K. Sharma, X.Z. Li, N. Graham, R.-A. Doong, J. Water Supply: Res. Technol.- AQUA 2008, 57 , 419–426. G.A.K. Anquandah, V.K. Sharma, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2009, 44 , 62–66. V.K. Sharma, G.A.K. Anquandah, N. Nesnas, Environ. Chem. Lett . 2009, 7 , 115–119. V.K. Sharma, G.A.K. Anquandah, R.A. Yngard, H. Kim, J. Fekete, K. Bouzek, A.K. Ray, D. Golovko, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2009, 44 , 423–442. M.E. Potts, D.R. Churchwell, Water Environ. Res. 1994, 66 , 107–109. U.M. Joshi, R. Balasubramanian, V.K. Sharma, in Ferrates: Synthesis, Properties, and Applications in Water and Wastewater Treatment (V.K. Sharma, ed.), Vol. 985, ACS Symposium Series, 2008, pp. 466–476. Y. Lee, S.G. Zimmermann, A.T. Kieu, U. Von Gunten, Environ. Sci. Technol. 2009, 43 , 3831–3838. Y.D. Perfiliev, V.K. Sharma, in Ferrates: Synthesis, Properties, and Applications in Water and Wastewater Treatment (V.K. Sharma, ed.), Vol. 985, ACS Symposium Series, 2008, pp. 112–123. G.W. Thompson, L.T. Ockerman, J.M. Schreyer, J. Am. Chem. Soc. 1951, 73 , 1379–1381. M.D. Johnson, US patent No. 5746994, 1998. Y.D. Perfiliev, E.M. Benko, D.A. Pankratov, V.K. Sharma, S.K. Dedushenko, Inorg. Chim. Acta 2007, 360 , 2789–2791. Z. M´acov´a, K. Bouzek, J. H´ıveˇs, V.K. Sharma, R.J. Terryn, J.C. Baum, Electrochim. Acta 2009, 54 , 2673–2683. K. Bouzek, M.J. Schmidt, A.A. Wragg, Collect. Czech. Chem. Commun. 2000, 65 , 133–140. F. Lapicque, G. Valentine, Electrochem. Commun. 2002, 4 , 764–766. X. Yu, S. Licht, J. Appl. Electrochem. 2008, 38 , 731–742.

REFERENCES

42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.

63. 64. 65. 66. 67. 68. 69.

70. 71. 72. 73. 74. 75. 76.

235

V. Lescuras-Darrou, F. Lapicque, G. Valentin, J. Appl. Electrochem. 2002, 32 , 57–63. M. De Koninck, D. B´elanger, Electrochim. Acta 2003, 48 , 1435–1442. W. He, J. Wang, C. Yang, J. Zhang, Electrochim. Acta 2006, 51 , 1967–1973. Y. Ren, Z. Chen, Y. Geng, R. Chen, X. Zheng, Chem. Eng. Process. 2008, 47 , 708–715. A. Denver, D. Plecher, J. Appl. Electrochem. 1996, 26 , 815–821. A. Denver, D. Plecher, J. Appl. Electrochem. 1996, 26 , 823–827. K. Bouzek, I. Rouˇsar, J. Appl. Electrochem. 1996, 26 , 919–923. K. Bouzek, I. Rouˇsar, J. Appl. Electrochem. 1993, 23 , 1317–1322. K. Bouzek, I. Rouˇsar, M.A. Taylor. J. Appl. Electrochem. 1996, 26 , 925–931. K. Bouzek, I. Rouˇsar. J. Appl. Electrochem. 1997, 27 , 679–684. K. Bouzek, M.J. Schmidt, A.A. Wragg, Electrochem. Commun. 1999, 1 , 370–374. K. Bouzek, M.J. Schmidt, A.A. Wragg, J. Chem. Technol. Biotechnol . 1999, 74 , 1188–1194. Z. Ding, C. Yang, Q. Wu, Electrochim. Acta 2004, 49 , 3155–3159. J. Xu, J. Wang, W. Yang, C. Chen, J. Zhang, Dianhuaxue 2004, 10 , 87–93. M. Alsheyab, J.-Q. Jiang, C. Stanford, J. Environ. Manage. 2009, 90 , 1350–1356. F. Beck, R. Kaus, M. Oberst, Electrochim. Acta 1985, 30 , 173–183. J.-Y. Zou, D.-T. Chin, Electrochim. Acta 1988, 33 , 477–485. K. Bouzek, I. Rouˇsar, H. Bergmann, K. Hertwig, J. Electroanal. Chem. 1997, 425 , 125–137. K. Bouzek, M. Nejezchleba, Collect. Czech. Chem. Commun. 1999, 64 , 2044–2060. H.B. Shao, J.M. Wang, W.C. He, J.Q. Zhang, C.N. Cao, Electrochem. Commun. 2005, 7 , 1429–1433. Z. M´acov´a, K. Bouzek, V.K. Sharma, in Ferrates: Synthesis, Properties, and Applications in Water and Wastewater Treatment (V.K. Sharma, ed.), Vol. 985, ACS Symposium Series, 2008, pp. 52–67. J. H´ıveˇs, M. Benov´a, K. Bouzek, V.K. Sharma, Electrochem. Commun. 2006, 8 , 1737–1740. J. H´ıveˇs, Z. M´acov´a, M. Benov´a, K. Bouzek, J. Electrochem. Soc. 2008, 155 , E113–E119. J. H´ıveˇs, M. Benov´a, K. Bouzek, J. Sitek, V.K. Sharma, Electrochem. Acta 2008, 54 , 203–208. J. Madar´asz, R. Zboˇril, Z. Homonnay, V.K. Sharma, G. Pokol, J. Solid State Chem. 2006, 179 , 1426–1433. L. Machala, R. Zboˇril, V.K. Sharma, J. Filip, O. Schneeweiss, Z. Homonnay, J. Phys. Chem. B 2007, 111 , 4280–4286. C.Z. Zhang, Z. Liu, F. Wu, L.-J. Lin, F. Qi, Electrochem. Commun. 2004, 6 , 1104–1109. C.Z. Zhang, H.B. Deng, T. Zhao, F. Wu, W. Liu, S. Cai, K. Yang, V.K. Sharma, in Ferrates: Synthesis, Properties, and Applications in Water and Wastewater Treatment (V.K. Sharma, ed.), Vol. 985, ACS Symposium Series, 2008, pp. 81–93. J. Lee, D.A. Tryk, A. Fujishima, S.-M. Park, Chem. Commun. 2002, 5 , 486–487. R.H. Wood, J. Am. Chem. Soc. 1958, 80 , 2038–2041. P. Ca˜nizares, M. Arc´ıs, C. S´aez, M.A. Rodrigo, Electrochem. Commun. 2007, 9 , 2286–2290. C. S´aez, M.A. Rodrigo, P. Ca˜nizares, AIChE J . 2008, 54 , 1600–1607. K. Serrano, P.A. Michaud, C. Comninellis, A. Savall, Electrochim. Acta 2002, 48 , 431–436. M.S. Saha, T. Furuta, Y. Nishiki, Electrochem. Commun. 2004, 6 , 201–204. P. Ca˜nizares, F. Larrondo, J. Lobato, M.A. Rodrigo, C. S´aez, J. Electrochem. Soc. 2005, 152 , D191–D196.

236

USE OF BORON-DOPED DIAMOND ELECTRODE

77. P. Ca˜nizares, C. S´aez, A. S´anchez-Carretero, M.A. Rodrigo, J. Appl. Electrochem. 2009, 39 , 2143–2149. 78. Z. Xu, J. Hu, Z. Yan, S. Yang, J. Zhou, W. Lu, Electrochim. Acta 2009, 54 , 3548–3552. 79. J.-Q. Jiang, C. Stanford, M. Alsheyab, Sep. Purif. Technol. 2009, 68 , 227–231. 80. M. Alsheyab, J.-Q. Jiang, C. Stanford, J. Chem. Health Safety 2008, 15 , 16–20. 81. M. Villanueva-Rodr´ıguez, A. Hern´andez-Ram´ırez, J.M. Peralta-Hern´andez, E.R. Bandala, M.A. Quiroz-Alfaro, J. Hazard. Mater. 2009, 167 , 1226–1230.

10 Electrochemical Oxidation of Organic Compounds Induced by Electro-Generated Free Hydroxyl Radicals on BDD Electrodes Agnieszka Kapałka, Helmut Baltruschat, and Christos Comninellis

10.1

INTRODUCTION

In the last decade, the advanced oxidation processes (AOPs) have been demonstrated as an effective means of wastewater treatment [1]. The main principle of AOPs is based on oxidation of organic pollutants using hydroxyl radicals that are highly reactive and destroy organics with a very high rate and low selectivity. AOPs are classified depending on the reaction phase or HO • generation method, generally classified as chemical (Fenton processes), electrochemical, sonochemical (ultrasound processes), and photochemical methods (UV/H2 O2 , UV/O3 photolysis, photo-Fenton, TiO2 photocatalysis) [1]. The electrochemical method for the oxidation of target organic pollutants is a technology for treatment of dilute wastewater (COD < 5 gl−1 ). In this method, organic pollutants are oxidized by intermediacy of hydroxyl radicals electro-generated from water on high Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

237

238

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

oxidation power anodes. An ideal anode for electrochemical oxidation of organic compounds is a boron-doped diamond electrode (BDD). The aim of this work is to elucidate the principles of electrochemical oxidation of organic compounds induced by electro-generated free hydroxyl radicals on BDD electrode. The following points will be treated: — Influence of anode material on the reactivity of electrolytic hydroxyl radicals — Electro-generation and detection of quasi-free hydroxyl radicals on BDD anode — Concentration profile of quasi-free hydroxyl radicals during the oxygen evolution reaction and electro-oxidation of organic compounds on BDD — Kinetic model of organics mineralization on BDD — Electrochemically induced mineralization of organics by molecular oxygen on BDD 10.2 INFLUENCE OF ANODE MATERIAL ON THE REACTIVITY OF ELECTROLYTIC HYDROXYL RADICALS In the electrochemical mineralization (EM) reactions oxygen is transferred from water to the organic pollutant using electrical energy. According to the generally accepted mechanism of the EM in acid media [2], water is first discharged at the anode active sites M, producing hydroxyl species (either adsorbed species, if the potential is low or, if the potential is high enough, even free radicals), as seen in Equation 10.1: M + H2 O → M(HO • ) + H+ + e−

(10.1)

These electrogenerated hydroxyl species are involved in the mineralization of organic pollutants R (see Equation 10.2): R + M(HO • ) → M + mineralization products + H+ + e−

(10.2)

This reaction (Equation (10.2) is in competition with the side reaction of the anodic discharge of hydroxyl species to dioxygen (see Equation 10.3): 1 M(HO • ) → M + O2 + H+ + e− 2

(10.3)

The activity of these electrolytic hydroxyl species is strongly linked to their interaction with the electrode surface M [3]. As a general rule, the weaker the interaction, the lower the electrochemical activity is toward oxygen evolution (high O2 overvoltage anodes) and the higher the chemical reactivity is toward organics oxidation. Based on this approach, we can classify the different anode materials according to their oxidation power in acid media as it is shown in Table 10.1 [4]. This table shows that the attainable oxidation potential of the anode is directly related to the overpotential for oxygen evolution and to the adsorption enthalpy of hydroxyl species on the anode surface (i.e., for a given anode material, the higher the O2 overvoltage, the higher its oxidation power is). A low oxidation power anode is characterized by a strong electrode-hydroxyl species interaction resulting in a high electrochemical activity for the oxygen evolution reaction

10.2 INFLUENCE OF ANODE MATERIAL

TABLE 10.1

239

Oxidation power of the anode material in acid media. (Adapted from Ref. 4.)

Electrode RuO2 − TiO2 (DSA-Cl2 ) IrO2 − Ta2 O5 (DSA-O2 ) Ti/PtOx Ti/PbO2 Ti/SnO2 -Sb2 O5 p–Si/BDD

Oxidation potential of organics (V)∗

Overpotential of O2 evolution (V)∗∗

1.4–1.7 1.5–1.8 1.7–1.9 1.8–2.0 1.9–2.2 2.2–2.6

0.18 0.25 0.3 0.5 0.7 1.3

Oxidation power of the anode

∗ Depending on the applied current densities ∗∗ Considering an overpotential for O evolution of 0.1 mA cm−2 2

(low overvoltage anode) and to a low chemical reactivity for organics oxidation (low current efficiency for organics oxidation). A typical low oxidation power anode is the IrO2 -based electrode. Concerning this anode, it has been demonstrated using differential electrochemical mass spectrometry (DEMS) [5] that the interaction between IrO2 and hydroxyl species is so strong that a higher oxidation state oxide is formed as an intermediate. This higher oxide can act as mediator for both organics oxidation and oxygen evolution. On low oxidation power anodes, oxidation of organics occurs with a low current efficiency but with a high selectivity. Typically, in electrosynthesis, a high selectivity is achieved under potentiostatic conditions. However, on low oxidation power anodes, a high selectivity can be achieved also under galvanostatic conditions [6]. This can be explained by the fact that on these anodes, under galvanostatic conditions, the potential is “buffered” by the oxygen evolution reaction, which is the competing side reaction. Under these so-called pseudo-potentiostatic conditions, the working potential is fixed by the nature (oxidation power) of electrode material, which is related to the oxygen evolution overpotential (i.e., the higher the oxygen evolution overpotential, the higher the oxidation power of the anode (Table 10.1)). The principle of the anode potential “buffering” induced by the oxygen evolution reaction is shown in Figure 10.1 This figure shows that during 11 IrO2

10 9 j (mA cm–2)

8 7 6 5

Dj

4 3 2 1 0 0.0

DE 0.5

1.0 1.5 E (V) vs SHE

2.0

Figure 10.1 Principle of buffering the anode potential induced by the oxygen evolution reaction using steady-state current potential curve recorded on Ti/IrO2 in 1 M HClO4 + 100 mM i-propanol. (Adapted from Ref. 6.)

240

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

oxidation of isopropanol on IrO2 , a large increase in current results in a small change of anode potential due to potential buffering by the side reaction of oxygen evolution. In fact, a change of one decade in current induces a shift of only 40 mV of potential. On the contrary to low oxidation power anode, the high oxidation power anode is characterized by a week electrode-hydroxyl species interaction resulting in a low electrochemical activity for the oxygen evolution reaction (high overvoltage anode) and to a high chemical reactivity for organics oxidation (high current efficiency for organics oxidation). Boron-doped diamond based anode is a typical high oxidation power anode. Here, the electrode-hydroxyl interaction is so weak that hydroxide radicals are formed. The evidence for the formation of hydroxyl radicals on BDD was found by means of spin trapping, hydroxylation of salicylic acid, and formation of hydrogen peroxide [7], as discussed next.

10.3 ELECTRO-GENERATION AND DETECTION OF QUASI-FREE HYDROXYL RADICALS ON BDD ELECTRODE 10.3.1

Hydroxyl Radicals Spin Trapping

The principle of the spin trapping method is to produce a stable adduct by allowing a specific scavenger to react with a less stable radical. As shown in Equation 10.4, the spin trap 5.5-dimethyl-1-pyrroline-N-oxide (DMPO) reacts with hydroxyl radicals to produce a stable adduct, detectable by electron spin resonance (ESR). H3C H3C

N+

+

HO•

O−

DMPO spin trap

H 3C H 3C

N

OH

O•

DMPO hydroxyl radical spin adduct

(10.4)

The main advantage of using DMPO is that it exhibits different ESR spectra with hydroxyl radical and singlet oxygen. The rate constant between DMPO and hydroxyl radicals is equal to 4.3 × 109 mol−1 s−1 [8]. Figure 10.2 shows ESR spectrum obtained after 2 hours of electrolysis of 10 mM DMPO in 1 M HClO4 at 0.1 mA cm−2 on BDD. This spectrum reveals the existence of the DMPO hydroxyl radical spin adduct (see Equation 10.4). The ESR hyperfine couplings of the DMPO adduct produced by electrolysis (obtained by simulation) are aN = aH = 14.95 G. These hyperfine coupling constants are typical of the spin adduct DMPO-OH [9,10], indicating that hydroxyl radicals are indeed produced during electrolysis on BDD electrode.

10.3.2

Trapping by Salicylic Acid

Salicylic acid (SA) is usually used for trapping free hydroxyl radicals in biological system. In fact, SA reacts with hydroxyl radicals to form dihydroxylated aromatic compounds [7]. The main products of the hydroxylation of salicylic acid with hydroxyl radicals are

10.3 ELECTRO-GENERATION AND DETECTION OF QUASI-FREE HYDROXYL RADICALS

241

Figure 10.2 Electron spin resonance of DMPO adduct obtained after electrolysis of 8.8 mM DMPO solution in 1 M HClO4 for 2 h on BDD electrode at 0.1 mA cm−2 . (Adapted from Ref. 7.)

2,3- and 2,5-dihydroxybenzoic acids (DHBA) (see Equation 10.5). COOH

COOH HO•

OH

COOH

OH

OH +

OH

HO

2,3-Dihydroxybenzoic acid

2,5-Dihydroxybenzoic acid

(10.5)

Figure 10.3 shows the concentration profile of salicylic acid and its conversion during electrolysis on BDD anode. As expected, the hydroxylation of salicylic acid results in formation of dihydroxylated intermediates. High 2,5- and 2,3-DHBA formation yields cannot be achieved because of its further oxidation to catechol and aliphatic acids. On BDD, the current density has almost no influence on the salicylic conversion as well as

100

6

4

60

X (%)

Conc. (mol m–3)

80

40 2 20 0

0 0

0.4

0.8

1.2

1.6

2

Q (Ah L–1) Figure 10.3 Oxidation of salicylic acid in 1 M HClO4 on BDD anode at 20 A m−2 ; (•) salicylic acid, () 2,5-DHBA, () 2,3-DHBA, () conversion of salicylic acid. (Reprinted from Ref. 7.)

242

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

the DHBA selectivity indicating that the chemical reaction of salicylic acid hydroxylation takes place. 10.3.3

Competitive Reactions

When a mixture of organic compounds is present in solution, one can consider that there is a competition for hydroxyl radicals between the organics. Figure 10.4 shows the results of electrolysis performed on BDD anode in an equimolar solution of formic acid and oxalic acid. It can be seen that the concentration of formic acid decreases much faster than the concentration of oxalic acid. In fact, only when the concentration of formic acid is very low, oxalic acid starts to be oxidized with a significant rate. These results indicate that formic acid competes with oxalic acid for hydroxyl radicals. Indeed, the rate constant between hydroxyl radicals and formic acid (k = 108 M−1 s−1 ) is higher by two orders of magnitude than the rate constant between hydroxyl radicals and oxalic acid (k = 106 M−1 s−1 ) [11]. Such a difference in k between these acids explains why the oxidation of formic acid occurs first on BDD. 10.3.4

Formation of Hydrogen Peroxide

If the interaction between the electrode surface and hydroxyl radicals is weak, hydrogen peroxide is expected to be formed by combination of two hydroxyl radicals, as seen in Equation 10.6: HO • + HO • → H2 O2

(10.6)

Figure 10.5 shows generation of hydrogen peroxide during electrolysis of 1 M HClO4 on BDD anode at the current densities from 23 to 160 mA cm−2 . Depending on the applied current density, the concentration of hydrogen peroxide, after 1 h, reaches a plateau ranging from 0.3 to 1 mM. The obtained steady state H2 O2 concentration (plateau) is

Figure 10.4 Evolution of concentration of formic and oxalic acids during electrolysis of a mixture of 0.5 M formic acid and 0.5 M oxalic acid in 1 M HClO4 on BDD at 238 A m−2 . (Reprinted from Ref. 7.)

10.3 ELECTRO-GENERATION AND DETECTION OF QUASI-FREE HYDROXYL RADICALS

243

1.0

[H2O2] (mol m–3)

0.8

0.6

0.4

0.2

0.0 0

5

10

15

Time (h) Figure 10.5 Production of H2 O2 at different current densities; (♦) 23 mA cm−2 , () 47 mA cm−2 , () 95 mA cm−2 , (x) 160 mA cm−2 during electrolysis of 1 M HClO4 on BDD electrode. (Reprinted from Ref. 7.)

due to the further anodic oxidation of hydrogen peroxide to oxygen. The initial current efficiency of H2 O2 formation reaches a value up to 0.5% [12]. Therefore, these results show that upon anodic polarization, a significant amount of hydroxyl radicals is formed at BDD electrode. Considering that free hydroxyl radicals are formed in acid aqueous solution according to Reaction (10.7), it is enlightening to calculate the thermodynamic standard potential of its formation using Equation 10.8 [13]. • − + H+ H2 O(l)  HO(aq) (aq) + e

E◦ = −

Gr◦ zF

(10.7) (10.8)

where E ◦ (V) is the standard thermodynamic potential of reaction, Gr ◦ (kJ mol−1 ) is the standard Gibbs free energy of the reaction, and z is the number of transferred electrons (z = 1). Taking the standard Gibbs free energy of liquid water formation as −237.178 kJ mol−1 and that of hydroxyl radicals in the aqueous state as −7.74 kJ mol−1 [13], one obtains the thermodynamic standard potential for HO • formation equal to 2.38 V.1 On BDD electrodes, the onset potential of oxygen evolution reaction is about 2.3 V (see Figure 10.6). The fact that O2 is evolved close to the thermodynamic potential of HO • formation (and at potentials positive of it) may indicate that electro-generated hydroxyl radicals on BDD surface are quasi-free. This assumption is justifiable, taking into account the inert nature of diamond surface, which contains closely packed sp3 carbon atoms, and a lack of adsorption sites [14]. As shown in Figure 10.6, the oxygen evolution reaction on BDD occurs with a high overpotential with respect to 1 A somewhat higher value of 2.73 V has also been suggested in literature. [P. Wardman, J. Phys. Chem. Ref. Data, 1989, 18, 1637-1755] For a concentration of 10 μM, which is the steady state concentration obtained only of at high current densities (see below), the equilibrium concentration would be 2.4 V, similar to the value given above.

244

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

BDD IrO2

8

j (mA cm–2)

7 6 5 4

•OH/H

Ir(V)/Ir(IV)

3 2

2O

O2/H2O

1 0 0.0

1.0

2.0

3.0

E (V) vs. SHE

Figure 10.6 Oxygen evolution reaction on IrO2 and BDD electrode in acid media. ◦ thermodynamic potential for O2 formation (EOER = 1.23 V versus SHE), but is very ◦ • = 2.38 V versus SHE). close to the thermodynamic potential of HO formation (EHO Therefore, it might be concluded that formation of this active intermediate (H2 O/HO • ) determines the overpotential for OER on BDD. Similarly, on IrO2 , OER proceeds close to the thermodynamic potential of Ir(V)/Ir(IV) redox couple (Figure 10.6), as it was demonstrated by DEMS measurements [5].

10.4 CONCENTRATION PROFILE OF HYDROXYL RADICALS ON BDD ELECTRODE It is interesting to determine the concentration profile of hydroxyl radicals during oxygen evolution and organics electro-oxidation on BDD electrode. To do that, one should consider that anodically polarized BDD electrode generates quasi-free hydroxyl radicals, which mediate the oxidation processes in the vicinity of the electrode surface. 10.4.1

HO • Concentration Profile during Oxygen Evolution

In the absence of organic compounds, electro-generated free hydroxyl radicals can react with each other to form hydrogen peroxide, as given by Equation 10.6. Hydrogen peroxide can be further oxidized to oxygen either by its direct discharge on the electrode surface (Equation 10.9) or via assistance of hydroxyl radicals (Equation 10.10): H2 O2 → O2 + 2H+ + 2e− H2 O2 + 2HO → O2 + 2H2 O •

(10.9) (10.10)

Assuming that the first step (Equation 10.6) is the rate determining step and that four hydroxyl radicals (four electrons) are needed for oxygen evolution, by applying onedimensional Fick law referred to the molecular diffusion in a stagnant layer, the mass balance for hydroxyl radicals in an element of width x can be expressed as [15]:     dcHO • dcHO • 2 • • • − 4kHO cHO • x = −DHO (10.11) −DHO dx x dx x +x

10.4 CONCENTRATION PROFILE OF HYDROXYL RADICALS ON BDD ELECTRODE

245

where DHO • (m2 s−1 ) is the diffusion coefficient of HO • , kHO • (m3 mol−1 s−1 ) is the rate constant of Reaction 10.6, and cHO • (mol m−3 ) is the concentration of HO • . If x goes to 0, we can write: DHO •

d2 cHO • 2 = 4kHO • cHO • dx 2

(10.12)

The boundary conditions are obtained by assuming that far from the electrode (x = ∞) concentration of hydroxyl radicals is 0, whereas at the electrode (x = 0) there is a surface s concentration of hydroxyl radicals (cHO • ): cHO • = 0 at x = ∞

(10.13)

s cHO • = cHO at x = 0 •

(10.14)

Considering these boundary conditions, the solution of differential Equation 10.12 gives the concentration profile of hydroxyl radicals, as a function of the distance x from the electrode surface, during oxygen evolution. cHO • =

 2kHO •

3DHO • 2  • x + 2k3DHO • cs • HO

(10.15)

HO

The gradient of the concentration can be expressed as: dcHO • = dx

 kHO

•

−3DHO • 3  • x + 2k3DHO • cs • HO

(10.16)

HO

Using Equation 10.16, it is possible to calculate the flux of hydroxyl radicals at the electrode surface, as given by Equation 10.17:  [JHO ]x =0 = −DHO •

•

dcHO • dx

 x =0

 s 3 = 1.63 (cHO • ) kHO • DHO •

(10.17)

The flux of hydroxyl radicals at the electrode surface can also be expressed in terms of the current density j : [JHO • ]x =0 =

j F

(10.18)

Thus, comparing Equation 10.17 with Equation 10.18, the surface concentration of hydroxyl radicals during oxygen evolution (OER) can be obtained (see Equation 10.19):  s cHO •

OER

=

3

j2 2.67F 2 kHO • DHO •

(10.19)

246

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

Figure 10.7 Simulated concentration profile of hydroxyl radicals during oxygen evolution according to Equations 10.15 and 10.19; j = 300 A m−2 ; DHO = 2.2 × 10−9 m2 s−1 ; kHO = 5.5 × 109 M−1 s−1 [10]. (Reprinted from Ref. 15.)

Figure 10.7 shows the concentration profile of hydroxyl radicals during oxygen evolution as a function of the distance from the electrode surface. It can be seen that for current density of j = 300 A m−2 , the reaction layer thickness is about one micrometer, whereas the maximum (surface) concentration of hydroxyl radicals reaches the value of several tenths of μM.

10.4.2 HO • Concentration Profile during Electro-Oxidation of Organic Compound By analogy to Equations 10.11 through 10.19, it is possible to determine the concentration profile of hydroxyl radicals during oxidation of organic compounds under these hypotheses: •

Oxidation of organic compounds R proceeds only via assistance of hydroxyl radicals in the vicinity of the electrode surface. R + zHO • → oxidation products

•

(10.20)

Concentration of organic compound is high enough to be considered as a constant in the reaction layer. • Oxygen evolution, via H2 O2 oxidation, is negligible. The two latter hypotheses apply to the charge-transfer controlled region.

10.4 CONCENTRATION PROFILE OF HYDROXYL RADICALS ON BDD ELECTRODE

247

Under these hypotheses, the mass balance for hydroxyl radicals leads to Equations 10.21 and 10.22:     dcHO • dcHO • • • • − zkR cHO cR x = −DHO (10.21) −DHO dx x dx x +x DHO •

d2 cHO • = zkR cHO • cR dx 2

(10.22)

where z is the number of electrons involved in oxidation of organic compound R, kR (m3 mol−1 s−1 ) is the rate constant of Reaction 10.20, and cR (mol m−3 ) is the concentration of organic compound. Thus, solving Equation 10.22 with the boundary conditions given in Equations 10.13 and 10.14, one can obtain the concentration profile of hydroxyl radicals during oxidation of organic compounds (Equation 10.23):  

cHO • =

s cHO •

zkR cR exp − x DHO •

(10.23)

The gradient of the concentration can be expressed as:  dcHO • s = −cHO • dx

  zkR cR zkR cR exp − x DHO • DHO •

(10.24)

Using Equation 10.24, the flux of hydroxyl radicals at the electrode surface is given by Equation 10.25:

s zkR cR DHO • [JHO • ]x =0 = cHO •

(10.25)

Comparing Equation 10.25 with Equation 10.18, the surface concentration of hydroxyl radicals during oxidation of organic compounds R is obtained (Equation 10.26): s cHO • = R

j √ F zkR cR DHO •

(10.26)

In the presence of organic compound, concentration of hydroxyl radicals decreases exponentially (Equation 10.23) with the distance from the electrode surface. The thickness of the reaction layer (reaction cage) depends on the root of the concentration of organic compound, the rate constant of organics oxidation (via hydroxyl radicals), and the applied current density. As a typical example, Figure 10.8 shows the simulated HO • concentration profile during oxidation of formic acid (0.25–1 M) at 300A m−2 . It can be seen that the higher is the formic acid concentration, the lower is the surface concentration of HO • , and the smaller is the thickness of the reaction layer. For investigated concentrations of formic acid, the thickness of the reaction layer (reaction cage) drops down to barely tenths of nanometers, which is significantly lower as compared with that in the absence of organics (Figure 10.7). Figure 10.9 shows the comparison of HO • concentration profile for 1 M formic acid, methanol, and ethanol. Depending on the rate constant, the thickness of the reaction layer varies between few nanometers and tenths of nanometers.

248

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

Figure 10.8 Simulated concentration profile of hydroxyl radicals during oxidation of (1) 1 M, (2) 0.75 M, (3) 0.5 M, and (4) 0.25 M formic acid, according to Equations 10.23 and 10.26; j = 300 A m−2 ; DHO = 2.2 × 10−9 m2 s−1 ; kHCOOH = 1.3 × 108 M−1 s−1 [10]; z = 2. (Reprinted from Ref. 15.)

Figure 10.9 Simulated concentration profile of hydroxyl radicals during oxidation of 1 M (1) formic acid, (2) methanol, and (3) ethanol according to Equations 10.23 and 10.26; j = 300 A m−2 ; DHO = 2.2 × 10−9 m2 s−1 ; kHCOOH = 1.3 × 108 M−1 s−1 ; kCH3OH = 9.7 × 108 M−1 s−1 ; kC2H5OH = 1.9 × 109 M−1 s−1 [10]; z = 2, 6, and 12 for formic acid, methanol, and ethanol, respectively. (Reprinted from Ref. 15.)

10.5

KINETIC MODEL OF ORGANICS OXIDATION ON BDD ANODE

Under electrolysis regime, electrogenerated hydroxyl radicals (Equation 10.1) are the intermediates for both the main reaction of organics oxidation (Equation 10.2) and the side reaction of oxygen evolution (Equation 10.3). Considering this simplified reaction scheme (Equations 10.1, 10.2, and 10.3), a kinetic model is proposed based on following

249

10.5 KINETIC MODEL OF ORGANICS OXIDATION ON BDD ANODE

suppositions: (1) adsorption of the organic compounds at the electrode surface is negligible; (2) all organics have the same diffusion coefficient D; and (3) the global rate of the electrochemical mineralization of organics is a fast reaction and it is controlled by mass transport of organics to the anode surface. The consequence of this last assumption is that the rate of the mineralization reaction is independent of the chemical nature of the organic compound present in the electrolyte. Under these conditions, the limiting current density for the electrochemical mineralization of an organic compound (or a mixture of organics) under given hydrodynamic conditions can be written as Equation 10.27 [3]: jlim (t) = 4Fkm COD(t)

(10.27)

where jlim is the limiting current density for organics mineralization (A m−2 ), F is the Faraday constant (C mol−1 ), km is the mass transport coefficient (m s−1 ) and COD is the chemical oxygen demand (mol O2 m−3 ). At the beginning of electrolysis, at time t = 0, the initial limiting current density (jlim ) is given by: 0 jlim = 4Fkm COD0

(10.28)

where COD0 is the initial chemical oxygen demand. Working under galvanostatic conditions, two different operating regimes are defined: at japplied < jlim the electrolysis is controlled by the applied current, whereas at japplied > jlim it is controlled by the mass transport control. 10.5.1

Electrolysis under Current Limited Control ( japplied < jlim )

In this operating regime, the current efficiency is 100% and the rate of COD removal is constant and can be written as: r =α

0 jlim 4F

(10.29)

where α is the dimensionless current density defined as: α=

japplied with 0 < α < 1 0 jlim

(10.30)

Using Relation 10.28, the rate of COD removal (Equation 10.30) can be given by: r = αkm COD0

(10.31)

It is necessary to consider the mass-balances over the electrochemical cell and the reservoir to describe the temporal evolution of COD in the batch recirculation reactor system given in Figure 10.10. Considering that the volume of the electrochemical reactor VE (m3 ) is much smaller than the reservoir volume VR (m3 ), we can obtain from the mass-balance on COD for the electrochemical cell the following relation: QCODout = QCODin − αkm ACOD0

(10.32)

250

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

CO2 O2 W1

W2 F2

F1

R3

R4

P3

P4 _

+

R1

E

R2

R5

R6 P1

P2

Figure 10.10 Scheme of the two-compartment electrochemical flow cell; R = reservoirs; P = pumps; E = electrochemical cell with membrane; W = heat exchangers, F = gas flow controllers. (Reprinted from Ref. 4.)

where Q is the flow-rate (m3 s−1 ) through the electrochemical cell, CODin and CODout are the chemical oxygen demands (mol O2 m−3 ) at the inlet and at the outlet of the electrochemical cell, respectively, and A is the anode area (m2 ). For the well-mixed reservoir (Figure 10.10), the mass balance on COD can be expressed as: Q(CODout − CODin ) = VR

dCODin dt

(10.33)

Combining Equations 10.32 and 10.33 and replacing CODin by the temporal evolution of chemical oxygen demand COD, we obtain: COD0 Akm dCODin = −α dt VR

(10.34)

Integrating this equation subject to the initial condition COD = COD0 at t = 0 gives the evolution of COD(t) with time in this operating regime (japplied < jlim ):  Akm COD(t) = COD 1 − α t VR 0

(10.35)

This behavior persists until a critical time (tcr ), at which the applied current density is equal to the limiting current density, which corresponds to: CODcr = αCOD0

(10.36)

10.5 KINETIC MODEL OF ORGANICS OXIDATION ON BDD ANODE

251

Substituting Equation 10.36 in Equation 10.35 it is possible to calculate the critical time (Equation 10.37): tcr =

10.5.2

1 − α VR α Akm

(10.37)

Electrolysis under Mass Transport Control ( japplied > jlim )

In this operating regime, the current efficiency is 100% and the rate of COD removal is constant and can be written as: When the applied current exceeds the limiting one (japplied > jlim ), secondary reactions (such as oxygen evolution) start to proceed, resulting in a decrease of the instantaneous current efficiency (see Equation 10.38): ICE =

COD(t) jlim = japplied αCOD0

(10.38)

This regime is realized either at: 0 1. japplied < jlim when the electrolysis is continued over the critical time (COD < 0 αCOD at t > tcr ), or 0 2. japplied > jlim when applied current exceeds its initial limiting value (COD < αCOD0 at any finite time)

In these cases, the COD mass balances on the anodic compartment of the electrochemical cell E and the reservoir R2 (Figure 10.10) can be expressed as: dCOD Akm COD =− dt VR

(10.39)

Integration of this equation from t = tcr to t, and COD = αCOD0 to COD(t) leads to:  Akm 1−α at t > tcr t+ 1. COD(t) = αCOD0 exp − VR α  Akm 0 2. COD(t) = COD exp − t VR

(10.40) (10.41)

From Equations 10.38, 10.40 and 10.41 the instantaneous current efficiency, ICE, is now given by:  Akm 1−α at t > tcr 1. ICE = exp − t+ VR α  Akm 1 2. ICE = exp − t α VR

(10.42) (10.43)

A graphical representation of the proposed kinetic model is given in Figure 10.11. In order to verify it, an anodic oxidation of various organic compounds has been performed on BDD. Figure 10.12 shows both the experimental and predicted values (continuous

252

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

(a)

COD0

A

COD(t ) = COD0(1–

B αAkm t) VR

COD(t) = αCOD0exp ( – CODcr =

i appl

Akm 1–α ) VR t + α

4Fk m

0 (b) ICE(%)

t

100

ICE = exp ( –

o (1–α) Qcr = i cr km

Ak m 1–α ) t+ VR α

0 t

Figure 10.11 Evolution of (a) COD and (b) ICE in function of time (or specific charge); (A) represents the charge transport control; (B) represents the mass transport control. (Reprinted from Ref. 4.)

1.0

2500 ICE/–

0.8

COD (ppm)

2000

0.6 0.4 0.2

1500

0.0 0

5

10

15

Specific charge (Ah L–1)

1000

500

0 0

5

10

15

Specific charge (Ah L–1)

Figure 10.12 Evolution of COD and ICE (inset) in function of specific charge for different organic compounds: (x) acetic acid, () isopropanol, (o) phenol, () 4-chlorophenol, (♦) 2-naphtol; i = 238 A m−2 ; T = 25◦ C; Electrolyte: 1 M H2 SO4 ; the solid line represents model prediction. (Reprinted from Ref. 4.)

line) of both ICE and COD evolution with the specific electrical charge passed during the anodic oxidation of different classes of organic compounds (acetic acid, isopropanol, phenol, 4-chlorophenol, 2-naphtol). This figure demonstrates that the electrochemical treatment is independent on the chemical nature of the organic compound. Furthermore, there is an excellent agreement between the experimental data and predicted values from proposed model. Figure 10.13 shows the influence of current density on both ICE

10.6 ELECTROCHEMICALLY INDUCED MINERALIZATION OF ORGANIC COMPOUNDS

253

1.0

2000 ICE /–

0.8

COD (ppm)

1500

0.6 0.4 0.2 0.0

1000

0

5

10

15

Specific charge (Ah L–1) 500

0 0

5

10

15

–1)

Specific charge (Ah L

Figure 10.13 Influence of the applied current density, (x) 119 A m−2 , (o) 238 A m−2 , (♦) 476 A m−2 on the evolution of COD and ICE (inset) during electrolysis of 5 mM 2-naphtol in 1 M H2 SO4 on BDD; T = 25◦ C; the solid line represents model prediction. (Reprinted from Ref. 4.)

and COD evolution with the specific electrical charge passed during the galvanostatic oxidation of a 5 mM 2-naphtol in 1 M H2 SO4 at different current densities (119–476 A m−2 ). As previously, an excellent agreement between the experimental and predicted values is observed.

10.6 ELECTROCHEMICALLY INDUCED MINERALIZATION OF ORGANIC COMPOUNDS BY MOLECULAR OXYGEN Electrochemical mineralization of organic compounds on BDD anode not only involves hydroxyl radicals but also molecular oxygen present in air/oxygen-saturated aqueous organic solutions [16]. The direct evidence for this process was found using differential electrochemical mass spectrometry (DEMS). Figure 10.14 shows a DEMS measurement, in which the cyclic voltammogram (CV) and the mass spectrometric cyclic voltammograms (MSCV) for C16 O2 , 18 O2 , 16 O18 O, C18 O2 , and C16 O18 O were simultaneously recorded during oxidation of acetic acid solution (1) deaerated and (2) saturated with isotopically labeled 18 O2 . Figures 10.14e and 10.14f show that isotopically labeled C18 O2 and C16 O18 O are evolved in parallel to 18 O2 diminution (Figure 10.14c) indicating that oxygen dissolved in the solution is involved in mineralization of acetic acid. The rate of C18 O16 O is higher than that of C18 O2 due to the higher probability of C18 O16 O formation, because one atom of oxygen (16 O) comes from water. It is interesting to note that ionic current of C18 O2 reaches a limitation at higher potentials (Figure 10.14e). The limiting current may indicate that oxidation of acetic acid by 18 O2 (similarly, 16 O18 O formation in Figure 10.14d) is limited by diffusion of 18 O2 to the electrode surface. Under investigated conditions, 7% of additional, non-Faradaic CO2 was evolved. This non-Faradaic enhancement of the acetic acid electro-oxidation proceeds most likely through the sequence of reactions that are initiated by HO • formed on the electrode

254

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

CV

I f (mA)

4

1–2

2

(a) 0 MSCV, m/z = 44 (C16O2)

I i (nA)

6

1–2

3 (b) 0 1.0

2

16

I i (nA)

MSCV, m/z = 36 (18O2)

I j (pA) 0.8

12 1 (c) 8

0.6 75

MSCV, m/z = 34 (16O18O)

I i (pA)

2 50

25 27

(d)

1

MSCV, m/z = 48 (C18O2)

I i (pA)

2 18

9 (e)

1

0

I i (nA)

0.4

MSCV, m/z = 46 (C16O18O) 2

0.2 1

(f) 0.0 1.8

2.0

2.2 2.4 E (V) vs RHE

2.6

2.8

Figure 10.14 Simultaneously recorded cyclic voltammogram (CV) (a) and mass spectrometric CV (MSCV) for (b) C16 O2 (m/z = 44), (c) 18 O2 (m/z = 36), (d) C16 O18 O (m/z = 34), (e) C18 O2 (m/z = 48), and (f) C16 O18 O (m/z = 46) of (1) deaerated and (2) 18 O2 -saturated solution of 50 mM acetic acid; scan rate 10 mV s−1 , flow rate 5 μls−1 , electrolyte 1 M HClO4 , T = 25◦ C. (Adapted from Ref. 16.)

10.6 ELECTROCHEMICALLY INDUCED MINERALIZATION OF ORGANIC COMPOUNDS

255

surface. Such a reaction scheme was proposed in radiolysis of hydrocarbons (RH). It has been reported that during the ionizing irradiation of aqueous solution of organic compounds (upon x-ray or γ -ray), molecular oxygen enhances the oxidation processes [17–19]. This occurs via addition of molecular oxygen to an organic free radical (R • ) resulting in formation of an organic peroxy radical (RO2 • ), which can participate in subsequent reactions. The organic free radical (R • ) is formed via dehydrogenation of hydrocarbons (RH) initiated by HO • formed during radiolysis of water. Therefore, by analogy between radiation-induced oxidation of organic compounds and our system, in which HO • are generated electrochemically, a general model of electrochemically induced oxidation of organic compounds via molecular oxygen on BDD electrode can be proposed, as shown in Figure 10.15. Such a non-Faradaic enhancement of electrolysis opens the possibilities for designing a less energy consuming electrochemical mineralization process in which current is applied mainly to initiate formation of HO • , whereas complete degradation of pollutant proceeds via reaction with molecular oxygen in aerated solutions (Figure 10.15, reactions 1–4). To do so, it is necessary to ensure an efficient transport of molecular O2 to the interface and thus the organic radicals (via porous electrodes). The key factors of the efficient mineralization process are the stability of organic radicals (reaction 5 in Figure 10.15, corresponding to formation of 18 O16 O in Figure 10.14d) and its reactivity toward molecular oxygen dissolved in solution (reaction 3 in Figure 10.15). It can be further concluded that such electrochemically induced activation process may occur also on other electrode materials on which organic radicals are formed during a direct discharge of organic compounds at the electrode surface.

RO•2 (4) (5)

(3) O2 (aq)

O2 R•

RO

1/2O2 (2)

RH (6)

H2O

(1)

H+

HO• e–

H+

e–

BDD electrode surface Figure 10.15 Simplified diagram for electrochemically induced oxidation of organic compounds via molecular oxygen dissolved in aerated solution on boron-doped diamond electrode; (1) water discharge to hydroxyl radical HO • ; (2) dehydrogenation of organic compound RH via HO • and formation of free organic radical R • ; (3) addition of molecular oxygen to R • resulting in formation of an organic peroxy radical RO2 • ; (4) decomposition of RO2 • leading to regeneration of HO • and formation of RO; (5) decomposition of RO2 • to R • ; (6) oxygen evolution, a side reaction. (Adapted from Ref. 16.)

256

10.7

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

CONCLUSIONS

On boron-doped diamond electrodes, organic compounds are oxidized by intermediacy of quasi-free hydroxyl radicals electro-generated from water at high anodic potentials. These hydroxyl radicals initiate chain reactions in which molecular oxygen dissolved in aerated aqueous solution contributes to mineralization of organic compounds at ambient temperature. In general, oxidation of organic compounds on BDD proceeds with very high current efficiency, which makes BDD an ideal anode material for electrochemical mineralization of organic wastes. 10.8

EXERCISES

1. Calculate the concentration of hydroxide radicals (DHO • = 2.2 × 10−9 m2 s−1 ) at the surface of BDD electrode (x = 0), at a distance of 0.1 μm and at 0.5 cm from the electrode surface during oxygen evolution in 1 M HClO4 at 200 A m−2 , T = 25◦ C. Consider that the rate determining step of the process is the following reaction: HO • + HO • → H2 O2

kHO • = 5.5 × 106 m3 mol−1 s−1

2. Calculate the concentration of hydroxide radicals (DHO • = 2.2 × 10−9 m2 s−1 ) at the surface of the BDD electrode (x = 0) during oxygen evolution under galvanostatic conditions in 1 M HClO4 at 25◦ C at current densities of: a) 300 A m−2 b) 100 A m−2 c) 10 A m−2 Consider that the rate determining step of the process is this reaction: HO • + HO • → H2 O2

kHO • = 5.5 × 106 m3 mol−1 s−1

3. The anodic oxidation of 1 M HCOOH solution in 2 M HClO4 has been carried out at 25◦ C under galvanostatic conditions (300 A m−2 ) using BDD electrodes. Calculate the concentration of hydroxide radicals (DHO • = 2.2 × 10−9 m2 s−1 ) at the surface of the BDD electrode (x = 0) after: a) Conversion of 30% of HCOOH to CO2 b) Conversion of 60% of HCOOH to CO2 c) Conversion of 90% of HCOOH to CO2 Consider that the process is controlled by the reactions: HO • + HO • → H2 O2

kHO • = 5.5 × 106 m3 mol−1 s−1

HCOOH + 2HO • → CO2 + 2H2 O

kHCOOH = 1.3 × 105 m3 mol−1 s−1

4. Calculate the initial concentration of hydroxide radicals (DHO • = 2.2 × 10−9 m2 s−1 ) at the surface of the BDD electrode (x = 0) during the anodic oxidation of the following organic compounds under galvanostatic conditions (100 A m−2 ) in 2 M HClO4 at T = 25◦ C:

10.8 EXERCISES

257

a) 1 M HCOOH (kHCOOH = 1.3 × 105 m3 mol−1 s−1 ) b) 1 M CH3 OH (kCH3OH = 9.7 × 105 m3 mol−1 s−1 ) c) 1 M C2 H5 OH (kC2H5OH = 1.9 × 106 m3 mol−1 s−1 ) 5. The mass transfer coefficient of an electrolytic cell, under given hydrodynamic conditions, has been estimated to be 1.2 × 10−5 m s−1 using the ferri-ferrocyanide (FFC) redox couple. Calculate: a) The mass transfer coefficient of formic acid, oxidized at the same conditions b) The limiting current density for the oxidation of 0.1 M formic acid aqueous solution given that DFFC = 1.1 × 10−9 m2 s−1 and DHCOOH = 1.7 × 10−9 m2 s−1 . 10.8.1

Solutions

• 1. .a) HO concentration at the surface of the electrode (x = 0):  j2 3 s cHO • = 2 OER 2.67F kHO • DHO •  2002 3 = = 5.1 × 10−2 mol m−3 2.67 · 964852 · 5.5 · 106 · 2.2 · 10−9

b) HO • concentration at a distance of 0.1 μm from the electrode (x = 0.1 μm = 10−7 m): cHO • =

3DHO •  2  • 2kHO • x + 2k3DHO s •c • HO

HO

3 · 2.2 · 10−9 −2 −3 =  2 = 1.4 × 10 mol m  3·2.2·10−9 2 · 5.5 · 106 10−7 + 2·5.5·10 6 ·5.1·10−2 c) HO • concentration at a distance of 0.5 cm from the electrode (x = 0.5 cm = 5 × 10−3 m): cHO • =

3 · 2.2 · 10−9 −11 mol m−3  2 = 2.4 × 10  −9 3·2.2·10 2 · 5.5 · 106 5 · 10−3 + 2·5.5·10 6 ·5.1·10−2

• 2. a . ) HO concentration at the surface of the electrode (x = 0) for galvanostatic conditions at 300 A m−2 :  j2 3 s cHO • = 2 OER 2.67F kHO • DHO •  3002 3 = = 6.7 × 10−2 mol m−3 2.67 · 964852 · 5.5 · 106 · 2.2 · 10−9

258

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

b) HO • concentration at the surface of the electrode (x = 0) for galvanostatic conditions at 100 A m−2 :  s cHO •

OER

=

3

2.67 ·

964852

1002 = 3.2 × 10−2 mol m−3 · 5.5 · 106 · 2.2 · 10−9

c) HO • concentration at the surface of the electrode (x = 0) for galvanostatic conditions at 10 A m−2 :  s cHO •

OER

=

3

2.67 ·

964852

102 = 6.9 × 10−3 mol m−3 · 5.5 · 106 · 2.2 · 10−9

3. .a) During oxidation of an organic compound: s cHO • = R

j √ F zkR cR DHO •

with

cR = cR0 (1 − X )

where X is the conversion rate; cR0 = 1M = 103 mol m−3 HCOOH + 2OH • → CO2 + 2H2 O z = 2 b) HO • concentration at the surface of the electrode (x = 0) after conversion of 30% of HCOOH to CO2 : s cHO • = R

j 300 =

√ 5 F zkR cR DHO • 96485 2 · 1.3 · 10 · 103 (1 − 0.3) · 2.2 · 10−9

= 4.9 × 10−3 mol m−3 c) HO • concentration at the surface of the electrode (x = 0) after conversion of 60% of HCOOH to CO2 : s cHO • = R

300 = 6.5 × 10−3 mol m−3

5 96485 2 · 1.3 · 10 · 103 (1 − 0.6) · 2.2 · 10−9

d) HO • concentration at the surface of the electrode (x = 0) after conversion of 90% of HCOOH to CO2 : s cHO • = R

300 = 1.3 × 10−2 mol m−3

5 96485 2 · 1.3 · 10 · 103 (1 − 0.9) · 2.2 · 10−9

10.8 EXERCISES

259

4. .a) HO • concentration at the surface of the electrode (x = 0) during anodic oxidation of 1 M (103 mol m−3 ) HCOOH (kR = 1.3 × 105 m3 mol−1 s−1 ) at 100 A m−2 : HCOOH + 2OH • → CO2 + 2H2 O j s cHO • = √ R F zkR cR DHO • =

z =2

100 √ 96485 2 · 1.3 · 105 · 103 · 2.2 · 10−9

= 1.4 × 10−3 mol m−3 b) HO • concentration at the surface of the electrode (x = 0) during anodic oxidation of 1 M (103 mol m−3 ) CH3 OH (kR = 9.7 × 105 m3 mol−1 s−1 ) at 100 A m−2 : z =6 CH3 OH + 6OH • → CO2 + 5H2 O 100 s = 2.9 × 10−4 mol m−3 cHO • = √ R 96485 6 · 9.7 · 105 · 103 · 2.2 · 10−9 c) HO • concentration at the surface of the electrode (x = 0) during anodic oxidation of 1 M (103 mol m−3 ) C2 H5 OH (kR = 1.9 × 106 m3 mol−1 s−1 ) at 100 A m−2 : C2 H5 OH + 12OH • → 2CO2 + 9H2 O z = 12 100 s = 1.5 × 10−4 mol m−3 cHO • = √ R 96485 12 · 1.9 · 106 · 103 · 2.2 · 10−9 5. .a) Estimation of the mass transfer coefficient of HCOOH: kd =

D at invariant δ : δ

kd,HCOOH = kd,FFC

DHCOOH DFFC = kd,HCOOH kd,FFC

DHCOOH 1.7 · 10−9 = 1.2 · 10−5 = 1.9 × 10−5 m s−1 DFFC 1.1 · 10−9

b) The limiting current density for the oxidation of 0.1 M (102 mol m−3 ) HCOOH: jlim = z · F · kd,HCOOH · cHCOOH where z = 2 (HCOOH + 2OH • → CO2 + 2H2 O) jlim = z · F · kd · c = 2 · 96485 · 1.9 · 10−5 · 102 = 367 A m−2 = 0.37 kA m−2

260

ELECTROCHEMICAL OXIDATION OF ORGANIC COMPOUNDS

REFERENCES 1. S. Parsons, ed., Advanced Oxidation Processes for Water and Wastewater Treatment , IWA Publishing, London, 2004. 2. C. Comninellis, Electrochim. Acta 1994, 39 , 1857–1862. 3. G. F´oti, C. Comninellis, in Modern Aspects of Electrochemistry (R.E. White, B.E. Conway, C.G. Vayenas, M.E. Gamboa-Adelco, eds.), Modern Aspects of Electrochemistry, No. 37, Kluwer Academic / Plenum Publishers, New York, 2004, pp. 87–130. 4. A. Kapałka, G. F´oti, C. Comninellis, J. Appl. Electrochem. 2008, 38 , 7–16. 5. S. Fierro, T. Nagel, H. Baltruschat, C. Comninellis, Electrochem. Commun. 2007, 9 , 1969–1974, Electrochem. Solid State Lett ., 2008, 11 , E20–E23. 6. S. Fierro, E. Passas-Lagos, E. Chatzisymeon, D. Mantzavinos, C. Comninellis, Electrochem. Commun. 2009, 11 , 1358–1361. 7. B. Marselli, J. Garcia-Gomez, P.A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 150 , D79–D83. 8. G. Liu, J. Zhao, H. Hidaka, J. Photochem. Photobiol. A: Chem. 2000, 133 , 83–88. 9. G.R. Buettner, Free Rad. Biol. Med . 1987, 3 , 259–303. 10. M. Hermes-Lima, N.C. Santos, J. Yan, M. Andrews, H.M. Schulman, P. Ponka, Biochim. Biophys. Acta 1999, 1426 , 475–482. 11. NDRL Radiation Chemistry Data Center, http://allen.rad.nd.edu. 12. P.A. Michaud, Comportement anodique du diamant synth´etique dope au bore, PhD thesis, EPFL, No 2595, 2000. 13. J.P. Hoare, in Standard Potentials in Aqueous Solution (A.J. Bard, R. Parsons, J. Jordan, eds.), Marcel Dekker Inc., New York, 1985. 14. A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, eds., Diamond Electrochemistry, BKC Inc., Tokyo & Elsevier B.V., Amsterdam, 2005. 15. A. Kapałka, G. F´oti, C. Comninellis, Electrochim. Acta. 2009, 54 , 2018–2023. 16. A. Kapałka, B. Lanova, H. Baltruschat, G. F´oti, C. Comninellis, Electrochem. Commun. 2008, 10 , 1215–1218. 17. R. Scholes, J. Weiss, Radiat. Res. 1959, 1 , 177–189. 18. G.R.A. Johnson, J. Weiss, Chem. & Ind . 1955, 358–359. 19. W.M. Garrison, H.R. Haymond, W. Bennett, S. Cole, Radiat. Res. 1959, 10 , 273–282.

11 Modeling of Electrochemical Process for Water Treatment Using Diamond Films Onofrio Scialdone and Alessandro Galia

11.1

INTRODUCTION

In electrochemistry, as in other chemical methods, situations in which a single reactive path is obtained are seldom encountered. In most experimental cases the target reactive path is accompanied by side reactions leading to unwanted side products and/or to a higher energetic consume. Therefore, the optimization of the process becomes obviously a relevant issue. Quite often the high number of operative parameters that may be adjusted makes an empirical investigation exceedingly onerous in order to individuate the conditions that allow the optimization of an electrochemical process. In this perspective, theoretical models can offer useful strategies for both the individuation of the parameters that affect the selectivity of the process, and the design of new materials/apparatuses that can favor the selective occurrence of the desired route [1]. Furthermore, the experimental validation of mathematical models can furnish precious indications for scale-up stages and confirm the assumptions on which the model is based, thus allowing a proper description of the process. In the field of the electrochemical abatement of organic pollutants in wastewater, unwanted reactions should be severely minimized in order to avoid the formation of secondary pollutants and/or an increase of energetic costs. Hence, the Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

261

262

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

modeling of these processes has attracted the attention of numerous researchers in the last several years [2–10]. Both oxidation and reduction routes were used for the treatment of organic compounds in water [11,12]. Oxidation processes allow the conversion of organic pollutants in carbon dioxide (electrochemical “incineration” or “combustion”) (see Equation 11.1), to nontoxic compounds or to biocompatible organics that can be treated in a conventional biological process. Combined processes are more difficult to optimize. Therefore, the development of an oxidation process that allows treating the wastewater without the necessity of a posttreatment stage is more often pursued by researchers. Reduction processes are usually used for the conversion of some organic pollutants to nontoxic compounds. For example, halogenated aliphatic compounds can be electrochemically reduced to corresponding dehalogenated aliphatic hydrocarbons (see Equation 11.2 for the conversion of chloroalkanes to corresponding alkanes) [12], whereas nitrate ions can be converted in nitrogenous gas [13]. Cm Hn + 2mH2 O → mCO2 + (n + 4m)H+ + (n + 4m)e−

(11.1)

Cm H2m+2−p Clp + 2pe− + p H+ → Cm H2m+2 + p Cl−

(11.2)

Focusing on the oxidation route, it is relevant to observe that many organic pollutants can be anodically oxidized at a lower potential with respect to that involved for oxygen evolution, but in these conditions a decrease of the anode activity is often observed as a consequence of poisons formation [2]. These poisoning species can usually be oxidized only at high anodic potentials in the region of oxygen evolution. Therefore, the anodic abatement of organic pollutants is generally carried out at high potentials with simultaneous oxygen evolution. At these potentials the oxidation of organics can take place both by a direct anodic oxidation or by means of hydroxyl radicals generated by water oxidation. Depending on the nature of the electrode material, hydroxyl radicals can be present as free radicals in a reaction layer adjacent to the electrodic surface, with a thickness drastically lower than that of the diffusion layer due to the fast chemical evolution of these very reactive species, or adsorbed to the anodic surface. Both a weak physical and a chemical adsorption are furthermore possible depending on the electrodic material. A simplified mechanism for the electrochemical selective oxidation or combustion of organics with simultaneous oxygen evolution was proposed in 1994 by Comninellis [2]. According to this mechanism, selective oxidation to stable compounds occurs with oxide anodes (MOx ) forming a so-called higher oxide MOx +1 (chemisorbed hydroxyl radicals) and combustion occurs with phisisorbed hydroxyl radicals. On the basis of this assumption, Comninellis proposed two theoretical expressions for the instantaneous current efficiency (ICE) for the organics oxidation achieved by means of physical adsorbed hydroxyl radicals at “nonactive anodes” (electrodes which do not participate in the oxidation) or chemical adsorbed oxygen at “active anodes” (electrodes that participate in the oxidation) [2]. First, the author focused on processes that occurred in the absence of mass transfer limitations. Later, Comninellis and, co-authors (Simond, Schaller, and Comnindellis [3] and Simond and Comninellis [4]) developed a theoretical model for active electrodes that included the effect of the substrate concentration ([RH ]b ) polarization. More recently, an extension of the above mentioned theoretical models was proposed, which takes in account the occurrence of both direct anodic oxidation and oxidation mediated by physical or chemical adsorbed hydroxyl radicals [5]. Mass transfer control,

11.2 THEORETICAL MODELS

263

oxidation control, and mixed kinetic regimes were considered. Theoretical predictions were in good agreement with the experimental results obtained for oxalic and formic acids oxidation at Ti/IrO2 –Ta2 O5 anode [5,6]. In general, in the literature it has been observed more times that the weaker the interaction between hydroxyl radical with the electrodic surface, the lower is the electrochemical activity toward oxygen evolution (high oxygen overvoltage anodes) and the higher is the chemical reactivity toward organics oxidation [14]. In this frame, the high activity of Boron-doped diamond (BDD) toward the electrochemical incineration of a wide class of organics can be rationalized on the bases of the weak interactions between the diamond surface and hydroxyl radicals. These interactions are described to be so weak that the HO • can be considered as quasi-free, thus leading to high overpotential for oxygen evolution and to high reactivity toward organics oxidation [14]. The oxidation of organic pollutants can also take place in the homogeneous phase by electro-generated agents such as active chlorine, ozone, S2 O8 2− and CeIV [11]. Active chlorine or peroxidisulfuric acid, for example, can be generated by the oxidation of chlorides or sulfate ions, respectively, at the anodic surface. Particular interest has been focused on chlorine mediation, due to the ubiquitous character of Cl− species in wastewaters and its effective action. The homogeneous oxidation performed by means of active chlorine could coexist with the direct oxidation of the organics at the electrode surface, the oxidation mediated by electro-generated hydroxyl or oxychloro radicals or with both these paths. As a consequence, in this case it is not easy to evaluate a priori the effect of operative parameters on the performances of the process. In this context it highlights again the potential paramount importance of theoretical models to select the key operative parameters and design suitable materials/apparatus. So, many authors have focused their attention on the modeling of electrochemical oxidation of organic pollutants in water. Particular attention has been devoted to processes performed at BDD, due to the extreme efficacy of this electrode toward the oxidation of numerous pollutants. The modeling of both direct and indirect oxidation processes was attempted by various authors. A brief review of these studies is reported here.

11.2 11.2.1

THEORETICAL MODELS General Considerations

The electrochemical oxidation of water at BDD has been extensively investigated by various authors [11,15]. As prepared boron-doped diamond shows hydrogenated terminations that are oxidized before the oxygen evolution to oxygen-containing functionalities [16,17]. It has been suggested that hydroxyl groups are the result of strong electrochemical oxidation. Oxygen evolution at BDD starts at about 2.3 V versus SCE by oxidation of water and formation of hydroxyl radicals (see Equation 11.3) [15]. The experimental evidence for the formation of hydroxyl radicals on BDD electrodes was reported by Marselli et al. [18] by spin trapping, using electron spin resonance. Recently, Enache and co-authors have shown by DP voltammetry an oxidation peak at 2.1 V versus Ag/AgCl associated with the water discharge process and the electrochemical generation of hydroxyl radicals [19]. Many authors have supposed, on the bases of the poor adsorption ability of the BDD surface, that hydroxyl radicals generated by water oxidation at BDD are quasi-free or very weakly adsorbed on the electrode surface [14]. The

264

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

oxygen evolution can involve different mechanisms. In particular, two probable mechanisms involve the oxidation of the hydroxyl radical (Equations 11.4 and 11.6) or the coupling of two HO • (Equations 11.5 and 11.6) [20–22]. Note that coupling of HO • can proceed toward the formation of hydrogen peroxide (Equation 11.7), which, at its turn can be oxidized to oxygen (Equation 11.8) [21,22]. Reactions 11.4 and 11.5 can also coexist, but it is not easy, up to present knowledge, to assert what route should predominate. In this context one can consider that recombination of hydroxyl radicals is likely to be more favored on surfaces that present, in contrast from what is expected for BDD, a high active surface sites concentration [22]. On the other hand, the occurring of the recombination reaction at BDD is supported by the fact that Michaud et al. observed low H2 O2 concentrations during HClO4 electrolyses at BDD [21]. H2 O → HO • + H+ + e− HO

+

→O +H +e

•

•

−

2HO → O + H2 O •

2O

•

→ O2

•

(11.3) (11.4) (11.5) (11.6)

2HO → H2 O2

(11.7)

H2 O2 → O2 + 2H+ + 2e−

(11.8)

•

In the presence of organics, hydroxyl radicals are involved in their oxidation as proposed by Feng and Johnson [23], that of course takes place in competition with the oxygen evolution. BDD( • OH)n + RH → BDD + CO2 + H2 O

(11.9)

Furthermore, organics can be directly oxidized at the anodic surface or they can be oxidized by means of other electro-generated reagents such as O3 , H2 O2 , H2 S2 O8 , and so on [11]. Direct anodic oxidation and oxidation by means of hydroxyl radicals, whose main stages occur at the electrode surface, were often classified in literature as “direct processes”, whereas oxidation by means of other electro-generated oxidants, that usually take place in homogeneous phase, are called “indirect processes” [11]. Please consider that, in the case of free hydroxyl radicals, this classification is not so stringent. On the other hand, for the sake of simplicity, all the oxidation routes involving reactions with hydroxyl radicals were often grouped among the “direct processes” [11]. Concerning the “direct processes”, it is useful to observe that anodic oxidation could be potentially performed at BDD for some organics, such as phenol derivatives, at lower potentials with respect to that necessary for oxygen evolution, but the formation of a passive layer at these potentials usually prevents this possibility [7,8,24,25]. Hence, abatement of organics is usually performed in the range of potential of oxygen evolution. In these conditions, no formation of the passivation layer occurs, oxidation of organics can coexist with oxygen evolution [8–10], and both anodic oxidation and oxidation by means of hydroxyl radicals can potentially concur to the abatement of the organics.

11.2 THEORETICAL MODELS

265

Please note that in the following “direct” and “indirect” processes will be studied separately because a very different effect of operative parameters on the performances of the wastewater treatment can occur with these different groups of reactions. Amperostatic electrolyses will be considered, since the galvanostatic mode is general preferred from an applicative point of view.

11.2.2 Oxidation of Organic Pollutants in Water at BDD by Means of Direct Anodic Oxidation or Reaction with Electro-Generated Hydroxyl Radicals (‘‘Direct Processes’’) When the abatement of organic pollutants takes place by means of direct anodic oxidation or reaction with electro-generated hydroxyl radicals (“direct processes”), the oxidation processes arise on the anode surface or in a thin reaction layer adjacent to the electrode surface with a thickness dramatically lower with respect to that of the diffusion layer. Thus, hydroxyl radicals generated at BDD are expected to exist as weakly physical adsorbed species on the anodic surface or as free molecules that can diffuse through a very thin portion of the diffusion layer due to their very high reactivity. It follows that the oxidation process can be considered in these conditions as a surface or a pseudo-surface process [26] that can take place under oxidation reaction control, mass transfer control or mixed kinetic regimes depending on the rate of mass transfer of the pollutant toward the anodic surface in comparison with the oxidation rate. Please note that the process could also be under the kinetical control of adsorption or desorption stages. On the other hand, these stages are usually neglected for the poor adsorption ability of the BDD surface. When the rate of the mass transfer of the pollutant is dramatically lower than that of its anodic oxidation, the concentration of the pollutant at the anodic surface/reaction layer C0 is close to zero and the oxidation process is under mass transfer control. This case arises, for an oxidation process that proceeds up to the total oxidation of the organic pollutant, when the limiting current density ilim = nFkm [RH]b  iapp ICEOC (where n is the number of electrons exchanged for the anodic oxidation of RH to carbon dioxide; F is the Faraday constant (96487 C mol−1 ); [RH]b and km are the bulk concentration and the mass transfer coefficient of the organic RH, respectively; iapp is the applied current density; and ICEOC is the instantaneous current efficiency for the oxidation of RH under oxidation reaction control under adopted operative conditions). Conversely, when ilim  iapp ICEOC , mass transfer is significantly faster with respect to oxidation rate, C0 is very close to the concentration in the bulk Cb and the process is under reaction oxidation control. When more pollutants are present in the bulk during the electrolysis, one can focus on the chemical oxygen demand (COD) of the solution. In particular, the limiting current density, for a process that proceeds up to the total oxidation of organics, is given by ilim = 4Fkm COD and a mass transfer control will arise if ilim  iapp ICEOC (where ICEOC is the average current efficiency for a process under oxidation reaction control and COD is the chemical oxygen demand computed on a molar base). The effect of operative parameters on the performances of the process was recently investigated [5,6,30,31], by considering, for the sake of simplicity, the complete oxidation of an organic pollutant that proceeds in competition with the oxygen evolution with no formation of intermediates with significant concentrations in the bulk. In this case, the instantaneous current efficiency for a process under galvanostatic mode can be rapidly estimated for both mass transfer and oxidation reaction control regimes.

266 •

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

Under mass transfer control, the current efficiency ICEMT is given by the ratio between the limiting current ilim and the applied current density iapp independently by the oxidation mechanism [5,7,8,30,31]. For [RH]b  C ∗ ICEOC , ICE = ICE MT ≈ [RH]b /C ∗ = nFkm [RH]b /iapp (11.10) where C ∗ = iapp /nFkm . It follows that, under mass transfer control, the ICE should depend on the fluidodynamics of the system (trough its effect on km ), on the applied current density and on [RH]b (with a linear dependence on this parameter). It is important to observe that, under these conditions, the performances of the process are readily predictable. Thus, the mass transfer coefficient is readily given by the ratio km = D/δ where the diffusion coefficient D can be often found in literature or can be estimated by electroanalytical experiments or using the Wilke-Chang expression, whereas the thickness of the diffusion layer δ is easily estimated by typical limiting current essays using, as an example, the couple hexacyanoferrate (II)/hexacyanoferrate (III). Hence, it is also possible to predict the organics concentration as a function of the charge or of the time. Indeed, the instantaneous current efficiency is given, by definition, by: ICE = −nFV d[RH]b /dQ

(11.11)

and eliminating the term ICE by Equations 11.10 and 11.11, one easily obtains the relationships reported in Equations 11.12 and 11.13 (where V is the volume, A is the anodic surface, and t and Q are the time and the charge passed, respectively). [RH]b,Q = [RH]b,Q = 0 exp[−Q/(nFVC ∗ )] [RH]b,t = [RH]b,t = 0 exp[−(Akm /V )t] •

(11.12) (11.13)

For a process under the kinetic control of the oxidation reaction, the instantaneous current efficiency ICEOC is determined by the competition between the oxidation of the organic and the evolution of the oxygen [5,30,31,40]: For [RH]b  C∗ ICEOC

iRH iRH = iapp iRH + iO2 1 1 = = iO2 [RH]∗ 1+ 1+ iRH [RH]b

ICEOC =

(11.14)

where iRH and iO2 are the current densities involved in the oxidation of the organic and in the oxygen evolution process, respectively. The term [RH]∗ is the value of [RH]b , which gives a current density for the RH oxidation iRH equal to the current density involved for the oxygen evolution reaction iO2 (e.g. the value of [RH]b that gives ICE = 50%) [5]. Please consider that the physical meaning of [RH]* depends on the oxidation route. Thus, if the oxidation of the organic takes place by direct anodic oxidation [RH]* is given by [RH]∗dir = 2r(E )/nk (E ), where k (E ) is the heterogeneous rate constant for the oxidation of RH and r(E ) is the rate of

11.2 THEORETICAL MODELS

267

the solvent oxidation. Otherwise, if the oxidation takes place by means of hydroxyl radicals, [RH]∗HO is given (as an example, considering for the sake of simplicity physically adsorbed hydroxyl radicals) by: [RH]∗OH = 2kor /nkO2

(11.15)

where kO2 (s−1 ) and kor (M−1 s−1 ) are the rate constants of Reactions 11.16 and 11.17, respectively, and n is the number of adsorbed hydroxyl radicals necessary to convert the substrate to carbon dioxide [5,7,30]. BDD( • OH) → BDD + 0.5 O2 + H+ + e−

(11.16)

BDD( • OH) + RH → BDD + H2 O + R •

(11.17)

It is interestingly to observe that [RH]∗dir assumes a constant value with the potential (and applied current density), given by the following relationship, if the transfer coefficients α of the two competitive oxidation routes assume similar values: [RH]∗dir = (2/n)(k  /k

◦

RH )exp[(1

− α)F (E

◦

RH

−E

◦

W )/(RT)]

(11.18)

where k is given by the product of the standard rate constant for the oxidation of the solvent and the solvent concentration and E ◦ W and E ◦ RH are the standard potentials for the oxidation of the water and RH, respectively [5,7,30]. Hence, in the case of a direct anodic oxidation process, expression reported in Equation 11.14 can be readily used to predict the effect of some operative parameters such as current density and organic concentration on the performances of the process. A more complex scenario is expected for an indirect process mediated by hydroxyl radicals. In this case, in fact, [RH]∗ is expected to depend on the working potential (e.g., on the applied current density) so that a less easy prediction of the effect of operative parameters on ICE is achievable [5,30]. Anyway, this aspect is of less practical relevance if the kinetic constant of mediated reaction is so high that in any case [RH]∗ is close to zero and ICE under oxidation control is close to 1. In general, it is interesting to observe that according to Equations 11.10 and 11.14, the instantaneous current efficiency of the process is expected to depend on the concentration of the organic pollutant in the bulk [RH]b for both mass transfer and oxidation reaction control. On the other hand, in the case of a mass transfer control, ICE should depend linearly on [RH]b , whereas in the case of an oxidation reaction control, a linear relationship is expected between 1/ICE and 1/[RH]b . Furthermore, if the kinetic constant of the reaction between the organic pollutant and hydroxyl radical is very high, as usually expected for diamond anodes, for a process under oxidation reaction control, the dependence of the ICE by [RH]b can be practically observed only for very low values of both [RH]b and the applied current density (e.g., by working with electrodes characterized by a very large surface area). Please note also that, under oxidation reaction control, the process is expected to be dramatically affected by the nature of both the organic pollutant and of the electrode material through their effect on [RH]* but not by the flow dynamic regime. However, in the case of a mass transfer control, the ICE depends strongly on the flow dynamic regime (trough its effect on km ), slightly on the nature of the organic pollutant (trough the value of the diffusion coefficient) but not significantly on the nature of the anode. It follows

268

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

that the utilization of boron-doped diamond anodes is expected to affect positively the abatement of organic pollutants in water if the process is under oxidation reaction control but not if a mass transfer control arises. On the other hand, please consider that a mass transfer control is expected when [RH]b  C∗ ICEOC . Therefore, if the same pollutant is treated at diamond anode with a high value of ICEOC (e.g., a low value of [RH]*) or at an electrode characterized by a very low value of ICEOC , for the same operative conditions BDD can experience a mass transfer control and the other electrode an oxidation reaction control with a slower rate with respect to that of the mass transfer at BDD. This case, as an example, was observed for the oxidation of oxalic acid in basic conditions at BDD and Ti/IrO2 -Ta2 O5 [6,42]. Thus, the oxidation reaction was so slow at the iridium-based anode that a mass transfer control never took place at this electrode under adopted operative conditions [42]. Let us now discuss the general case of a process whose rate determining step changes during the electrolysis from oxidation reaction control in the first stages to mass transfer control in the last part with a mixed regime between them. In this case, a general expression for the instantaneous current efficiency has to be adopted. In particular, under the previously mentioned conditions, it is easy to show that the following expression applies [5,30,31,40]: ICE =

1 2[RH]∗ 1+ [RH] + ([RH] 2 + 4[RH]∗ [RH]b )0.5

(11.19)

where [RH] = [RH]b − [RH]*−C * and C ∗ = i /nFkm . Theoretical predictions based on Equations 11.10, 11.14, and 11.19 were used, up to now, to predict the effect of operative parameters on the abatement of two quite simple molecules such as oxalic and formic acid. A very good agreement between experimental data and theoretical predictions was observed changing severely the current density, the flow rate, and the initial concentration of the acid (e.g., see Figures 11.1 and 11.2 for the oxidation of oxalic acid). This result appeared rather interesting also considering that the kinetic constant of the reaction between hydroxyl radicals and the organic presents very different values for these two acids. Please note that numerous pollutants can often be present in the water system from the beginning of the electrolysis or as a result of the formation of some intermediates with appreciable concentrations in the bulk. In this case, the modeling of the process has to take in account the contemporaneous presence of various organic pollutants in the solution. This general problem was studied by various groups, including those of Comninellis, Polcaro, Rodrigo, and Scialdone with different approaches. Some of these studies are briefly summarized in the following. 11.2.2.1 The Model of Comninellis and Coauthors Comninellis and coauthors observed that the oxidation of many organics, such as phenol, proceeds at BDD toward the complete incineration with very high current efficiency if no mass transfer limitations arise [7,8,24,25]. On the bases of these results, the authors developed a very simple kinetic model for a batch recirculation system with galvanostatic alimentation based on the assumption that when the oxidation of organics is performed at BDD at high anodic potentials, close to oxygen evolution, the electrochemical incineration of the organic

11.2 THEORETICAL MODELS

269

(a) 0.12

[OA]b(M)

0.10 0.08 0.06 0.04 0.02 0.00 0

2000

4000 Q(C)

6000

8000

0

2000

4000 Q(C)

6000

8000

0

2000

4000 Q(C)

6000

8000

(b) 0.12 0.10

[OA]b(M)

0.08 0.06 0.04 0.02 0.00

(c) 0.12 0.10

[OA]b(M)

0.08 0.06 0.04 0.02 0.00

Figure 11.1 Profile of oxalic acid concentration [OA]b versus charge passed for electrolysis performed at BDD with (a) 17, (b) 39, and (c) 56 mA/cm2 , at different flow rates. Flow rate: 1.7 (), 3.4, (-) 4.9 l/min (•). Amperostatic electrolyses. Theoretical curves (—) obtained by Equation 11.19 with [RH]* = 0.013 M. System solvent supporting electrolyte (SSE): Water, Na2 SO4 , H2 SO4 . Initial oxalic acid concentration: 100 mM. T = 25◦ C. (Reprinted with permission from Ref. 30.)

270

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

(a) 0.20 0.04 0.15 [OA]b(M)

0.02 0.00

0.10

0

5000

10000

15000

0.05

0.00 0

5000 10000 Charge passed (C)

15000

0

0.05

0.15

(b) 1.2 1

ICE

0.8 0.6 0.4 0.2 0

0.1 [OA]b(M)

(c) 5

1/(ICE)

4

3

2

1

0 0

40 ([OA]b)–1(M–1)

80

Figure 11.2 Oxalic acid concentration [OA]b versus (a) charge passed, (b) ICE versus [OA]b , and (c), 1/ICE versus 1/[OA]b . Initial substrate concentration: 0.01 (-), 0.05 (•), 0.1 () and 0.2 M (◦). Flow rate: 4.9 l/min. Current density: 17 mA/cm2 . Amperostatic electrolyses at BDD anode. System solvent supporting electrolyte (SSE): Water, Na2 SO4 , H2 SO4 . T = 25◦ C. Theoretical curves (—) in Figure 2a obtained by Equation 11.19 with [RH]* = 0.013 M. Theoretical curves in Figures 2b and 2c (–) obtained by Equations 11.14 and 11.10 with [RH]* = 0.013 M. (Reprinted with permission from Ref. 30.)

11.2 THEORETICAL MODELS

271

compound is a fast reaction and it is controlled by mass transport toward the anode. Particularly, authors assumed that: •

The reservoir volume (V ) is much greater than that of the electrochemical reactor. The electrochemical reactor and the reservoir are perfectly mixed. • The oxidation in the bulk by electro-generated oxidants is not considered. • If iapp < ilim (e.g., if the electrolysis is under current limit contro)l, the current efficiency for the abatement of the organic is 100%. As a consequence, COD decreases linearly with time, as shown by the relationship reported in the following equation (where COD◦ is the initial COD, A is the anodic surface, V is the solution volume, and t is the time). •

  iapp A α  km A ◦ t = COD 1 − t COD = COD − nFV V ◦

(11.20)

◦ where α  = iapp /ilim . • If iapp > ilim (e.g., if the electrolysis is under mass transfer control), secondary reactions (such as oxygen evolution) commence, resulting in a decrease of COD. The current efficiency can be simply estimated as

ICE = ICEMT ≈ ilim /iapp

(11.21)

and, as a consequence, the COD removal follows an exponential trend described by the following equation (where CODcr and tcr are the initial values of COD and time when the process is under mass transfer control from the beginning of the electrolysis or the values of COD and time achieved when iapp = ilim ):     1 − α Akm Akm ◦  (t − tcr ) = α COD exp − t+ COD = CODcr exp − V V α (11.22) It is important to underline the fact that this very simple model does not present any adjustable parameter. Thus, the mass transfer coefficient is readily given by the ratio km = D/δ, where the diffusion coefficient D can be often found in literature or can be estimated by electroanalytical experiments or using the Wilke-Chang expression, whereas the thickness of the diffusion layer δ is easily estimated by typical limiting current essays using, as an example, the couple hexacyanoferrate (II)/hexacyanoferrate (III). Of course, this model is expected to fit with accuracy experimental data when the current efficiency in the absence of mass transfer limitations is close to 1. Furthermore, the model does not describe mixed kinetic regimes or the evolution of the concentrations of different species present in the system. On the other hand, this simple model presents a very good agreement with experimental data for several organic compounds and various operative conditions for the evolution of COD with the time or the charge passed [7,8,14,15,24,25]. (see, as an example, Figures 11.3 and 11.4 for the case of the oxidation of 2-naphtol at acidic pH.) Therefore, it represents a very interesting tool for the prediction of the effect of operative parameters, such as current density, flow rate, and organic concentration, on the “direct” abatement of COD at a BDD anode in a batch recirculation system.

272

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

Figure 11.3 Influence of the initial 2-naphthol concentration () 9 mM; ( × ) 5 mM; (•) 2 mM on the trends of COD and ICE (inset) during electrolysis, using a BDD anode. Supporting electrolyte: 1 M H2 SO4 ; T = 30◦ C; j = 30 mA cm−2 . The solid lines represent the model prediction. (Reprinted with permission from Ref. 8.)

60

1.2 1 0.8 ICE

COD (mol O2m–3)

50 40

0.6 0.4 0.2

30

0 0

2

4

6

8

10

14

16

Q (Ah dm–3)

20 10 0 0

2

4

6

8 10 Q (Ah dm–3)

12

Figure 11.4 Influence of applied current density (•) j = 15 mA cm−2 ; (×) j = 30 mA cm−2 ; () j = 60 mA cm−2 on the trends of COD and ICE (inset) during electrolysis in 1 M H2 SO4 + 5 mM 2naphthol, using a BDD anode. T = 30◦ C. The solid lines represent the model prediction. (Reprinted with permission from Ref. 8.)

11.2.2.2 The Theoretical Works of Polcaro and Coauthors Polcaro and coauthors reported a more complex model with the aim of evaluating the trends of the concentrations of the starting pollutant and intermediate products in the stagnant layer and in the bulk [26]. Thus, it has been shown that the incineration of some organics at BDD proceeds trough the formation of intermediates. As an example, the oxidation of

11.2 THEORETICAL MODELS

273

phenol can be represented by the following schematic reaction path: Phenol → Cyclic intermediates → Aliphatic acids → CO2 Furthermore, this model applies also to the case of an oxidation process that gives rise to a current efficiency lower than 100% in the absence of mass transfer limitations. The model was based on the following assumptions: •

Oxidation of pollutants and intermediate compounds takes place by homogeneous chemical reactions with hydroxyl radicals (with first-order kinetic reaction with respect to both organics and hydroxyl radicals concentrations) in the diffusion layer in competition with the chemical deactivation of hydroxyl radicals, which is described by a first-order kinetic reaction not depending on the working potential. • A diffusion-reaction model is used to model the diffusion layer, whereas the bulk of the solution is represented by a stirred-tank reactor. The model required a numerical solution of pertaining equations (mass balance equations in the bulk and in the stagnant layer for the starting pollutant and intermediates and in the diffusion layer for hydroxyl radicals). As a result, authors were able to predict the trend with time of the concentration of different compounds present in the bulk of the solution during the electrolyses, as well as the evolution of the space profile of the species. As an example, the authors compared the theoretical predictions of the model with experimental data for the oxidation of Phenol at BDD. The model was able to predict with a good accuracy the trend with time of the concentrations of phenol and main intermediates (aromatic and aliphatic acids) by changing significantly the current density and the flow dynamic regime, and using one adjustable parameter (the average kinetic constant between hydroxyl radicals and aromatic compounds) [26]. A good agreement between theoretical predictions and experimental data was reported by the authors also in the cases of cyanuric acid and atrazine by using, as an adjustable parameter, the kinetic constant between hydroxyl radicals and the adopted pollutant [26]. As mentioned, the model involved also the calculation of the concentration profiles in the diffusion layer of hydroxyl radicals, pollutants, and intermediates with the time. Quite interestingly, according to this model, hydroxyl radicals diffuse in the stagnant layer for few tens of nanometers. As an example, a thickness of the reaction layer of about 50 nm was computed for a current density of 25 mA/cm2. 11.2.2.3 The Approach Proposed by Rodrigo and Coauthors In order to reduce the mathematic complexity of theoretical models aimed to determine the concentrations of compounds involved in the oxidation process, Rodrigo and coauthors considered some simplifying assumptions. Their model divides the electrochemical reactor into three zones: two are close to the electrodes (electrochemical zones) and a third zone corresponds to the bulk of the solution (chemical zone) that is considered as consecutive stirred-tank reactors [27–29]. Hence, in each zone, the concentration is assumed to depend on the time passed but not on the position. The concentration of each compound in the chemical zone is taken as the value measured experimentally. Mass transport processes between the electrochemical and chemical zones are quantified by assuming that the local exchange rate is proportional to the difference of concentration between the two zones. The authors observe that a number of processes can occur at the electrode

274

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

surface, and therefore the total current applied is shared among all these processes. The rate of each process is given by ri = (iapp A/F)αielectrode

(11.23)

where αielectrode , which gives the ratio of current density for a particular electrochemical process with respect to the applied current density, is a measure of the relative oxidizability of a compound and αielectrode = 1. In a first version of the model, αielectrode was a function of the oxidizability factors of the compounds present in the system that were computed as adjustable parameters [27]. Later, the authors assumed that the fraction of iapp used in each process depends on the cell potential (Vwork ) and the oxidation (or reduction) potential (Vi ) of each process by the relationship reported below, so that no adjustable parameters were used in the model [28]. αielectrode = (Vwork − Vi )/



(Vwork − Vi )

(11.24)

The model accounted also for the presence of homogeneous oxidation reactions between electro-generated oxidants and organics assuming second-order rate expressions depending on the concentrations of both the oxidant and the organic compound. In the case of phenol and carboxylic acids, no presence in the bulk of oxidants was evaluated with a I2 /I− test [28]. It was therefore concluded that no mediated oxidation processes occurred in the bulk zone. Hence, mediated process, involving oxidation by means of hydroxyl radicals, might occur only in the electrochemical zone and were considered as direct reactions. The model showed a good agreement with a very large set of experimental data obtained for the incineration of various organic compounds, including phenol and different carboxylic acids at BDD in a water solution of Na2 SO4 [28]. In the case of phenol, for example, the model considered the direct oxidation of phenol at the anode to carbon dioxide, through the formation of maleic and oxalic acids as intermediates, and the hydrogen evolution at the cathode (see Figure 11.5). As shown in Figure 11.6, theoretical predictions were in a good agreement with experimental data obtained under a very large range of operative conditions.

11.2.3 Oxidation of Organic Pollutants in Water by Means of Electro-Generated Oxidants (‘‘Indirect Processes’’) Such as Active Chlorine As previously mentioned, electrochemical oxidation of organics can occur also by the action of electro-generated oxidants such as active chlorine, ozone, peroxidisulfuric acid, CeIV , and more. Hence, various researchers included in their model the homogeneous oxidation of organics by means of electro-generated oxidants [10,28,29,40,44]. It is interesting to observe that in indirect processes the oxidation of the organics take place mainly by a homogeneous reaction and a very different effect of operative parameters on the performances of the process can arise with respect to direct ones. Thus, the current efficiency of the process is expected to be determined mainly by the current efficiency due to the electro-chemical formation of oxidants and by the competition between the homogeneous chemical oxidation of organic pollutants and competitive reaction paths

11.2 THEORETICAL MODELS

Anodic reaction zone (a)

anode

1

Phenol*

e–

Cathodic reaction zone (c)

cathode

Phenol

2 3

Maleic*

e–

Chemical reaction zone (b)

275

Maleic

9

Oxalic

Oxalic*

5

H2 e–

4 H2O

6

CO2

CO2* 7 H2O

e–

8 O2

(1) r1 = k·A·([Phenol]b–[Phenol]a) (2)

anode r2 = I a2 F

(3) r3 = k·A·([Maleic]b–[Maleic]a)

(4) r4 = k·A·([Oxalic]b–[Oxalic]a)

(7) r7 = k·A·([CO2]b–[CO2]a)

(5)

anode r5 = I a5 F

I anode (8) r8 = a8 F

(6)

anode r6 = I a 6 F

I cathode (9) r9 = a9 F

Figure 11.5 Sketch representing the processes considered in the modeling of the electro-oxidation of phenol-polluted wastewater using BDD anodes. (Reprinted with permission from Ref. 28.)

involving the electro-generated oxidants. In the following, this aspect will be studied in the frame of the oxidation of organics by means of active chlorine. This process was considered as a model case for numerous reasons. First, the effect of chloride ions on the performances of the process has been the object of numerous researches [11] due to the ubiquitous character of Cl− species in waste waters and to the fact that chloride ions can cause an increase in the removal efficiency of organic pollutants for the involvement of active chlorine in the oxidation process. Furthermore, it has been shown that the addition of chloride ions can give rise in some cases to the formation of halogenated intermediates more toxic than the starting compounds [32]. Second, in the presence of Cl− , a quite intriguing change of the effect of some operative parameters on the performances of the process can occur with respect to “direct processes.” When chloride ions are added to a water solution, their oxidation can lead to the formation of chlorine, hypochlorous acid, and/or hypochlorite, depending on the pH (Equations 11.25–11.27) that can oxidize the organics near to the anode or/and in the bulk of the solution (Equation 11.28) in alkaline medium [33–36]. 2Cl− → Cl2 + 2e− Cl2 + H2 O → HOCl + H+ + Cl−

(11.25) (11.26)

35

Concentration (mmol C dm–3)

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

50

Concentration (mmol C dm–3)

Concentration (mmol C dm–3)

276

(a)

60 50 40 30 20 10 0 0

5

10

15

20

25

30

60

(b)

50 40 30 20 10 0 0

5

Concentration (mmol C dm–3)

Charge (A h dm–3) (c)

80 70 60 50 40 30 20 10 0 0

10

20

30

40

Charge (A h dm )

25

(d)

70 60 50 40 30 20 10 0

–3

10 15 20 Charge (A h dm–3)

0

10

20 30 40 50 Charge (A h dm–3)

60

70

Figure 11.6 Results obtained [simulation (lines) versus experimental data (points)] for four experimental runs for the abatent of phenol: () phenol, () maleic acid, () oxalic acid, (-) carbon dioxide. (a) j = 30 mA cm−2, T = 25◦ C, pH = 2. (b) j = 30 mA cm−2 , T = 25◦ C, pH = 12. (c) j = 30 mA cm−2 , T = 25◦ C, pH = 9. (d) j = 60 mA cm−2 , T = 25◦ C, pH = 12. (Reprinted with permission from Ref. 28.)

HOCl ↔ H+ + OCl− Organics + OCl− → intermediates → CO2 + Cl− + H2 O

(11.27) (11.28)

These reactions take place in competition with oxygen evolution, chlorate chemical and electrochemical formation, and cathodic reduction of oxidants in the presence of undivided cells. Amperostatic electrolyses of chlorides gives rise at DSA anodes mainly to hypochlorite and chlorate, whereas at BDD, a more complex mixture—including hypochlorite, chlorine dioxide, and reactive oxygen species—was found [32,37]. The authors supposed that these species could be formed by reaction paths involving the reaction between hydroxyl radicals and active chlorine. It has been proposed [33,38] that adsorbed chloro- and oxychloro-radicals could be also involved in the oxidation mechanism. Furthermore, the possibility that some role could be played by the anodic shift of the oxygen evolution, caused by Cl− ions in the solution, has also been taken in consideration for platinum electrodes [38,39]. It follows that the oxidation of organics performed by means of active chlorine could coexist with the direct oxidation at the electrode surface, the reaction with hydroxyl or oxychloro radicals, or with both these paths. As a consequence, it is not easy for these complex systems to rationalize the effect of operative parameters, on the performances of the process. Furthermore, the coexistence of these oxidative routes makes this process completely different from the chemical oxidation with hypochlorite. This has been confirmed by many experimental works that usually report higher abatement of organic pollutants for the electrochemical oxidation with chlorides with respect to the chemical one with hypochlorite [33–38].

11.2 THEORETICAL MODELS

277

A very simplified theoretical approach was recently presented with the aim to examine how the more important operative parameters are expected to affect the performances of the process both in the absence and in the presence of chlorides [40]. In particular, it was assumed, for the sake of simplicity, that the reaction between the organics and active chlorine takes place mainly in the bulk of the solution. In the absence of chlorides or in the presence of a very high ratio between the concentration of organics and NaCl, main processes should be “direct” ones, such as a direct anodic oxidation or a mediated oxidation by hydroxyl or oxychloro radicals. Hence, the current efficiency for the abatement of the organics should be favored, as previously explained in detail, by (1) high values of the organic concentration; (2) the use of an electrodic material such as BDD, which provides an high oxygen over potential (e.g., a low value of k  (E ) and [RH]*); (3) higher flow rates and lower current densities (when the process is not under oxidation reaction control). Let us focus our attention on the opposite case of a sufficiently high ratio between the concentrations of chlorides and organics so that the contribution of the direct processes can be neglected. In this case, the current efficiency should depend mainly on the formation in the bulk of active chlorine ICE AC and on the competition between homogeneous oxidation of organics and chlorate formation [40]. Interestingly, ICEAC has been shown to increase for both BDD and DSA anodes in the presence of higher chloride concentrations and high current density, whereas a more complicated effect of the flow rate is reported. Higher and lower current efficiencies are reported by increasing the flow rate at BDD and DSA, respectively [37,40]. Hence, an opposite effect of current density is expected if the oxidation of organics takes place by a direct anodic process or by an oxidation mediated by active chlorine. Furthermore, BDD presents quite low current efficiency for the active chlorine formation with respect to low oxygen overpotential electrodes such as iridium and ruthenium oxides. Thus, the addition of chlorides in the case of BDD can result, depending on the adopted operative conditions, in a lower abatement of organics. Please note that when the direct process is under mass transfer control, the addition of chlorides is expected to result, if sufficiently high values of current density are imposed, in higher current efficiency for the mediated process. This is because in this case the process is no longer affected negatively by the kinetic limitations imposed by the mass transfer of the pollutant toward the anodic surface. Another important consequence is that BDD, which presents generally drastic higher abatement of organics with respect to DSA, can present, in the presence of suitable amounts of NaCl, a lower abatement with respect to these anodes. Real systems are complicated by the fact that direct and indirect processes could coexist under the same operative conditions. In this case, flow rate, current density, and anodic material should have an opposite effect on homogeneous and heterogeneous oxidation processes so that their overall effect on the performances of the process depends on the relative rates of heterogeneous and homogeneous oxidation reactions. Interestingly, the previously mentioned considerations on the effect of various operative parameters were confirmed by various studies [40,43,47], including one on the electrochemical incineration of oxalic acid at BDD and iridium-based anodes performed in the absence and in the presence of various amounts of NaCl [40]. Thus, a very different influence of the nature of the anodic material, the flow rate, and the current density on the performances of the process was observed in the absence and in the presence of chlorides so that optimization of the two processes required very different operative conditions. In the absence of chlorides, high current efficiency (CE) were obtained at BDD

278

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

when most of the process was under oxidation reaction kinetic control (i.e., when low current densities and high flow rates were imposed). On the other hand, in the presence of high concentrations of NaCl, higher CE was usually obtained at DSA with respect to BDD and at higher values of current densities. A simple theoretical model for the oxidation of organics in the presence of Cl− was proposed by Szpyrkowicz and Radaelli [44]. These authors described the kinetics of decolorization of simulated textile wastewaters in the presence of Cl− by means of a second-order rate constant: d(abs)/dt = −k (abs)[Cl2 ]

(11.29)

where (abs) is the measured absorbency and [Cl2 ] is the concentration of dissolved chlorine. The model was used by the authors, assuming that the Faradaic efficiency for the chlorine evolution is constant trough the electrolysis, with good results in the frame of the scale up of the oxidation of a textile wastewater at Ti/Pt-Ir anode from a batch to a continuous flow unit [44]. Recently, Anglada, Urtiaga, and Ortiz [45] studied in detail the electro-oxidation of landfill leachate at BDD on a pilot scale. Organic matter and ammonia oxidation were highly influenced by the applied current density value. Interestingly, the abatement of the organic matter was successfully modeled by using the approach proposed by Comninellis and coauthors (see Section 11.2.2) when a current density of 300 and 450 A m−2 was applied [45]. At higher current densities, the oxidation rate was higher than that predicted by the model, thus suggesting the occurrence of a mediated oxidation process by means of active chlorine. The authors also used the model proposed by Szpyrkowicz and Radelli [44] to predict the evolution of ammonium based on the hypothesis that the degradation takes place by a second-order reaction between ammonium and active chlorine. The kinetic model was able to predict the evolution of the concentration of the pollutant as a function of the current density in the range 300–600 A m−2 . At higher current densities a worse correlation between predicted and experimental results occurred. In these conditions a rapid decrease in chloride concentration occurred, thus suggesting that one of the model’s assumptions, the constancy of the Faradaic efficiency for the chlorine evolution during the electrolysis, was not complied [45]. It is relevant to observe that, as stated by Anglada et al. [45,46], many of the theoretical models proposed for the description of the electrochemical oxidation of organics in water still have to be validated for real wastewaters with complex and unknown detailed composition. In this case, in fact, the simultaneous occurrence of many competitive reactions at the anodic surface and in the homogeneous phase can create very complex scenarios that are very difficult to model in a simple and accurate way.

11.3

CONCLUSIONS

In the last years, several studies have proven that conducting diamond thin films can be used as very effective electrodes for the complete oxidation of organic pollutants in water and for water disinfection. The kinetic modeling of the electrochemical abatement of organics in water at diamond anodes has been attempted by many authors with the main objective of predicting the trend of the concentrations of pollutants and intermediates, of chemical oxygen demand, and of current efficiency by changing severely the operative

REFERENCES

279

conditions. The electrochemical oxidation of organic pollutants in water at diamond electrodes can take place by means of hydroxyl radicals generated by water oxidation and by direct anodic oxidation (“direct processes”) or in homogeneous phase by means of electro-generated reagents such as O3 , H2 O2 , H2 S2 O8 , active chlorine, and so on (“indirect processes”). “Direct processes” can be considered as surface or pseudo-surface processes and are strongly affected by the mass transfer rate of the organic pollutants toward the anodic surface. The modeling of these processes has been performed by various research groups, and a very good agreement between experimental data and theoretical predictions was generally reported. In “indirect processes”, the oxidation of the organics involves a homogeneous reaction with electro-generated oxidants, and a quite different effect of operative parameters on the performances of the process can take place with respect to the case of “direct processes”. Thus, in the case of “indirect processes”, the abatement of the organics is expected to depend mainly on the current efficiency due to the electrochemical formation of oxidants and on the competition between homogeneous chemical oxidation of organic pollutants and competitive reaction paths involving the electro-generated oxidants. The oxidation of organics in the presence of active chlorine was, in particular, briefly discussed as a model case.

11.4

ACKNOWLEDGMENTS

Universit`a di Palermo and Ministero dell’Istruzione, dell’Universit`a e della Ricerca (MIUR) are acknowledged for their financial support.

REFERENCES 1. C. Amatore, “Micro and Macropenomena,” in Organic Electrochemistry (H. Lund, M. Baizer, eds.), Marcel Dekker, Inc., New York, 1991, p. 220. 2. C. Comninellis, Electrochim. Acta 1994, 39 , 1857–1862. 3. O. Simond, V. Schaller, Ch. Comninellis, Electrochim. Acta 1997, 42 , 2009–2012. 4. O. Simond, Ch. Comninellis, Electrochim. Acta 1997, 42 , 2013–2018. 5. O. Scialdone, Electrochim. Acta 2009, 54 , 6140–6147. 6. O. Scialdone, S. Randazzo, A. Galia, G. Filardo, Electrochim. Acta 2009, 54 , 1210–1217. 7. M.A. Rodrigo, P.A. Michaud, I. Duo, M. Pamizza, G. Cerisola, Ch. Comninellis, J. Electrochem. Soc. 2001, 148 , D60–64. 8. M. Panizza, P.A. Michaud, G. Cerisola, Ch. Comninellis, J. Electroanal. Chem. 2001, 507 , 206–214. 9. A.M. Polcaro, S. Palmas, Ind. Eng. Chem. Res. 1997, 36 , 1791–1798. 10. A.M. Polcaro, M. Mascia, S. Palmas, A. Vacca, Ind. Eng. Chem. Res. 2002, 41 , 2874–2881. 11. C.A. Martinez-Huitle, S. Ferro. Chem. Soc. Rev . 2006, 35 , 1324–1340. 12. O. Scialdone, A. Galia, G. Filardo, Electrochim. Acta 2008, 53 , 7220–7225. 13. B.P. Dash, S. Chaudhari, Water Res. 2005, 39 , 4065–4072. 14. A. Kapalka, G. Foti, Ch. Comninellis, J. Appl. Electrochem. 2008, 38 , 7–16. 15. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191–199. 16. R. DeClements, G.M. Swain, T. Dallas, M.W. Holtz, R.D. Herric, J.L. Stickney, Langmuir 1996, 12 , 6578–6586.

280

MODELING OF ELECTROCHEMICAL PROCESS FOR WATER TREATMENT USING DIAMOND FILMS

17. I. Yagi, H. Notsu, T. Kondo, D.A. Tryk, A. Fujishima, J. Electroanal. Chem. 1999, 473 , 173–178. 18. B. Marselli, J. Garcia-Gomez, P.A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 150 , D79–D83. 19. T.A. Enache, A.M. Chiorces-Paquim, O. Fatibello-Filho, A.M. Oliveira-Brett, Electrochem. Comm. 2009, 11 , 1342–1345. 20. K. Kinoshita, in Electrochemical Oxygen Technology, The Electrochemical Society, New York,Wiley, 1992, chap. 2. 21. P.A. Michaud, M. Panizza, L. Quattara, T. Diaco, G. Foti, Ch. Comninellis, J. Appl. Electrochem. 2003, 33 , 151–154. 22. M.H.P. Santana, L.A.D. Faria, J.F.C. Boodts, Electrochim. Acta 2005, 50 , 2017–2027. 23. J. Feng, D.C. Johnson, J. Electrochem. Soc. 1990, 137 , 507–510. 24. J. Iniesta, P.A. Michaud, M. Panizza, G. Cerisola, A. Aldaz, Ch. Comninellis, Electrochim. Acta 2001, 46 , 3573–3578. 25. F. Montilla, P.A. Michaud, E. Morallon, J.L. Vazquez, Ch. Comninellis, Electrochim. Acta 2002, 47 , 3509–3513. 26. A.M. Polcaro, A. Vacca, S. Palmas, M. Mascia, J. Appl. Electrochem. 2003, 33 , 885–893. 27. P. Canizares, M. Diaz, J.A. Dominguez, J. Garcia-Gomez, M.A. Rodrigo, Ind. Eng. Chem. Res. 2002, 41 , 4187–4194. 28. P. Canizares, J. Garcia-Gomez, J. Lobato, M.A. Rodrigo, Ind. Eng. Chem. Res. 2004, 43 , 1915–1922. 29. P. Canizares, J. Garcia-Gomez, J. Lobato, M.A. Rodrigo, Ind. Eng. Chem. Res. 2004, 43 , 1923–1931. 30. O. Scialdone, A. Galia, C. Guarisco, S. Randazzo, G. Filardo, Electrochim. Acta 2008, 53 , 2095–2108. 31. O. Scialdone, A. Galia, S. Randazzo, in preparation. 32. M. Bergmann, J. Rollin, Catalysis Today 2007, 124 , 198–203. 33. L. Szpyrkowicz, J. Naumaczky, F. Zilio-Grandi, Toxicolog. and Environ. Chem. 1994, 44 , 189–202. 34. C. Comninellis, A. Nerini, J. Appl. Electrochem. 1995, 25 , 23–28. 35. L.C. Chiang, J.E. Chang, T.C. Wen, Water Res. 1995, 29 , 671–678. 36. C.H. Yang, C.C. Lee, T.C. Wen, J. Appl. Electrochem. 2000, 30 , 1043–1051. 37. A.M. Polcaro, A. Vacca, M. Mascia, F. Ferrara, J. Appl. Electrochem. 2008, 38 , 979–984; A.M. Polcaro, A. Vacca, M. Mascia, S. Palmas, J.R. Ruiz, J. Appl. Electrochem. 2009, 39 , 2083–2092. 38. F. Bonfatti, S. Ferro, F. Lavezzo, M. Malacarne, G. Lodi, A. De Battisti, J. Electrochem. Soc. 2000, 147 , 592–596. 39. C.A. Martinez-Huitle, S. Ferro, A. De Battisti, Electrochem. Solid State Lett . 2005, 11 , D35–D39. 40. O. Scialdone, S. Randazzo, A. Galia, G. Silvestri, Water Res. 2009, 43 , 2260–2272. 41. K. Serrano, P.A. Michaud, C. Comninellis, A. Savall, Electrochim. Acta 2002, 48 , 431–436. 42. O. Scialdone, S. Randazzo, A. Galia, G. Filardo, Electrochim. Acta 2009, 54 , 1210–1217. 43. M. Wu, G. Zhao, M. Li, L. Liu, D. Li, J. Hazardous Mat . 2009, 163 , 26–31. 44. L. Szpyrkowicz, M. Radaelli, J. Appl. Electrochem. 2006, 36 , 1151–1156. 45. A. Anglada, A. Urtiaga, I. Ortiz, Environ. Sci. Technol. 2009, 43 , 2035–2040. 46. A. Anglada, A. Urtiaga, I. Ortiz, J. Chem. Technol. Biotechnol . 2009, 84 , 1747–1755. 47. M. Zhou, H. S¨arkk¨a, M. Sillanp¨aa¨ , Separ. Purif. Techn. 2011, 78 , 290–297.

12 Production of Strong Oxidizing Substances with BDD Anodes Ana S´anchez-Carretero, Cristina S´aez, Pablo Canizares, and ˜ Manuel A. Rodrigo

12.1

ELECTROLYSES WITH CONDUCTIVE-DIAMOND ANODES

From the late nineties of the twentieth century to present, applications of conductive diamond surfaces in electrochemistry have grown significantly. One of these applications is their use as anodes in electrochemical wastewater treatment processes, where these anodes have led to very powerful oxidation processes, with high efficiencies in the removal of organic pollution. During the characterization of these processes, there have been reports about the significant action of mediated electro-oxidation in the achievement of good results, and the role of many particular oxidants has been described. In particular, the role of hydroxyl radical has been established to be very relevant. Its occurrence during diamond electrolyses of aqueous wastes was demonstrated by Marselli et al. [1], and later works have confirmed this occurrence with very different observations. This allowed classifying conductivediamond electrochemical oxidation (CDEO) of wastewaters as an advanced oxidation process (AOP). In this context, one of the more illustrative observations [2,3] about the role of hydroxyl radicals in wastewater treatment compares results of the electrolyses of phenol working at anodic potentials below and over the necessary amounts for water oxidation (around 2.5–2.7 V versus NHE). In this way, Figure 12.1 shows a comparison of the electrochemical windows of Pt and BDD electrodes and also summarizes the anodic Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

281

282

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

Current density(A m–2)

Anodic potential (V) vs. SCE

20

2.0

65

2.2

90

2.5

155

3.5

195

4.1

0.04 Thermodynamics

0.03

Pt

j (A cm–2)

0.02

DDB

0.01 0

–0.01 –0.02

–2.0

–1.0

0.0

1.0

2.0

3.0

E (V) vs. NHE

–0.03 Figure 12.1 Experimental methodology to demonstrate the role of hydroxyl radicals in CDEO process. (Adapted from Ref. 3.) See color insert.

potential measurement during the conductive-diamond phenol oxidation using different current densities. Figure 12.2 shows how the electrolyses of phenol solutions at potentials below water electrolyses (and hence without formation of hydroxyl radicals) lead to the formation of significant amount of intermediates (aromatic and aliphatic). Electrolyses at potentials higher than this value lead to the formation of negligible intermediates and to high efficient processes, suggesting that hydroxyl radicals improve the oxidation conditions and lead to very strong oxidation conditions. However, the better performance of conductive-diamond electrolyses over other advanced oxidation processes (those based on the production and use of hydroxyl radicals), and also over other electrolytic processes, indicates that in addition to direct oxidation and to hydroxyl radicals-mediated oxidation, there should be many other oxidants with a significant role in the destruction of organics during electrochemical wastewater oxidation with diamond anodes. In this context, the roles of chlorine [4–7], sulphates [8–10], phosphates [11,12], and many other types of salts on the electrochemical destruction of organics have been extensively studied in the literature. According to this fact, Figure 12.3 summarizes the main mechanism proposed to explain the good results of CDEO in the oxidation of organic pollutants during wastewater treatment. From this figure, it seems clear that the electrolyses of water solutions of different salts can lead to the formation of oxidants, and if proper conditions are found, oxidants produced can be separated and stored for later use (see Figure 12.4). This presence of oxidant species has encouraged many research groups to study the synthesis of particulate oxidants and to isolate them as valuable products. This work is intended to review some of these processes as examples of the significant applications of CDEO in the production of high-value oxidants. To do this, the

350

1200

300

1000

250

COD/g m–3

TOC(g m–3)

12.2 PRODUCTION AND STORAGE OF OXIDIZING SUBSTANCES: EXPERIMENTAL SETUPS

200 150 100

800 600 400 200

50

0

0 0

500

1000

1500

0

2000

Time(min)

500

1000 1500 Time(min)

2000

350 Carboxylic acids/g Cm–3

70 Aromatics(g Cm–3)

283

60 50 40 30 20 10

300 250 200 150 100 50 0

0 0

10

20 30 Q (kAh m–3)

40

50

0

20 40 Q (kAh m–3)

60

Figure 12.2 Results of the electrolyses of phenol solutions at different anodic potentials (C0 : 6.4 mM phenol; pH 7; T:25◦ C). (+) 2.0 V versus NHE, (◦) 2.2 V versus NHE, () 2.5 V versus NHE, () 3.5 versus NHE, () 4.1 V versus NHE. (Adapted from Ref. 2.)

production of hydroxyl radicals and more stable oxidants (peroxosalts and peroxoacids, ferrates, and halogen oxoanions) are going to be described. 12.2 PRODUCTION AND STORAGE OF OXIDIZING SUBSTANCES: EXPERIMENTAL SETUPS Prior to the description of processes that produce oxidants, it is necessary to take into account that contrary to the electro-oxidation of wastewaters, the synthesis of oxidants should be carried out in a double-compartment electrochemical flow cell in which anodic and cathodic compartments are separated by means of an ionic exchange membrane. This avoids the cathodic reduction of oxidants, and hence increases significantly the efficiency of the process. Usually, the ionic membrane is a cathionic one, because many oxidants are anionic species, and this allows storing the produced oxidants in the anodic compartment. Figure 12.5 illustrates with an example (obtained during the electrolytic production of perphosphates with boron-doped diamond anodes) the improvement obtained in the production of oxidants with the use of double compartment cells. At this point, it is important to be reminded that the use of double compartment cells in the destruction of the organic pollutants contained in wastewaters is not necessary because most reactions are irreversible. The separation of the anionic and cationic compartments only leads to an increase in the cell potential, and consequently in the operation cost, without any clear improvement in the treatment results.

284

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

HYDROXYL RADICALS MEDIATED ELECTROLYSES

Anode DIRECT ELECTROLYSES

Direct

Si

Si e–

Si+1

With oxidants produced from saltsas mediators

Si+1

Mred-Mox

Si

Mred

2SO42– → S2O82–

H2O e–

Si+1

Mox

2PO43– → P2O84– Cl– → ClO–

OH· With ozone as mediator

O2

Si

2H+ + O2 Oxygen evolution

O3

Mred

H2O

Si

e–

Mox

Si+1

H2O2

Si+1 With hydrogen Si peroxide as mediator

Si+1

Mediated electrolyses with oxidants produced from salts contained in the wastewater Figure 12.3 Main mechanisms for the electrolyses with conductive-diamond anodes of wastewaters polluted with organics.

Another important point to be considered in the production of oxidants is their stability. Light, temperature, and pH use have a very significant influence on this stability, which is usually favored at low temperature, dark, and extreme pHs. An example of the huge influence of these parameters in the stability of oxidants is shown in Figure 12.6, in which the decomposition rate of peroxophosphates is shown as a function of the pH and temperature. For this reason, a typical setup to produce oxidants with diamond anodes should include a cryostat device to control temperature, a pH control device, and dark glass in tubes and storage tanks. With such requirements, an example of a complete setup to produce oxidants with conductive-diamond anodes electrolyses is shown in Figure 12.7. 12.3 PRODUCTION OF HYDROXYL RADICALS WITH CONDUCTIVE-DIAMOND ANODES Production of hydroxyl radicals during conductive-diamond electrolysis of aqueous wastes was first demonstrated by Marselli and co-workers [1] using selective reactions of hydroxyl radicals. This was a very significant work because it was the first

12.3 PRODUCTION OF HYDROXYL RADICALS WITH CONDUCTIVE-DIAMOND ANODES

285

Anode

Mred H2O

Mred-Mox

e–

Mox 2SO42– → S2O82–

OH·

2PO43– → P2O84–

O2

Cl– → ClO–

2H+ + O2 Oxygen evolution

O3 H2O

Mred e–

H2O2

Mox

Figure 12.4 anodes.

Mechanisms for the production of oxidants with electrolyses with conductive-diamond

Oxidants(mmol) P2O84–

300

Double-compartment electrochemical flow cell

250 200 150 100

Single-compartment electrochemical flow cell

50 0 0

50

100

150

200

250

300

Time(min)

Figure 12.5 Effect of cell compartments on the efficiency of the production of oxidants by electrolysis with conductive diamond anodes. (C0 : 1 M K3 PO4 ; pH 12.5; T: 25◦ C; j: 1250 A m−2 ).

286

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

pH Decomposition rate(mol (1 min)–1)

1.00E–05 0

1

2

3

4

5

6

7

1.00E–06

1.00E–07 Figure 12.6 Effect of temperature and pH on the decomposition rate of peroxophosphates. (C0 : 1 M HK3 PO4 ; j: 64 A m−2 ). (+) 35◦ C, () 22◦ C, () 10◦ C. (Reprinted from Ref. 36.)

Power supply V

A

catholyte

anolyte +

–

Electrochemical cell

Membrane Anode (BDD)

Out

In

Cathode (AISI 304)

Out

In

Figure 12.7 Typical setup to produce oxidants with CDEO technology. See color insert.

12.3 PRODUCTION OF HYDROXYL RADICALS WITH CONDUCTIVE-DIAMOND ANODES

287

demonstration that conductive-diamond electrochemical oxidation belongs to the group of the advanced oxidation technologies. Figure 12.8 summarizes graphically the two main findings of the work: the selective formation of DMPO and the selective hydroxylation of salicylic acid. The selective oxidation of 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) involves trapping the hydroxyl radicals by an addition reaction (spin trapping) to produce a more stable radical spin adduct. The anodic oxidation of salicylic acid at the BDD anode leads to the formation of two particular dihydroxylated products, 2,3- and 2,5-dihydroxybenzoic acids, just the same products formed by Fenton reaction. After that, many other works have supported this achievement with different observations that indicate the significance of hydroxyl radicals in the electrolyses with conductive diamond. The main drawback of hydroxyl radical is their small stability, which makes its storage impossible and even makes their detection very difficult (average lifetime lower than 200 μs [13]). However, the occurrence of hydroxyl radicals explains the production of many more stable oxidants, particularly two relevant oxidants: ozone and hydrogen peroxide. Thus, although many other technologies can provide for ozone or hydrogen peroxide in a more efficient way, there are many works in literature in which the production of both oxidants is obtained with conductive-diamond electrolyses. As it has been describe previously, hydroxyl radicals are produced during conductivediamond electrolyses of aqueous solutions. These radicals decompose following a complex system of reactions in which ozone [14] and hydrogen peroxide [14] can be intermediates or final products. In literature, it is reported that in absence of organic compounds, hydroxyl radicals can react with each other to form hydrogen peroxide from Reaction (12.1) [15]. In fact, Michaud et al. [16] observed the generation of hydrogen peroxide during the electrolysis with BDD anodes of HClO4 solutions. In addition, it is reported that electro-generated hydrogen peroxide can be further oxidized to oxygen either by its direct discharge on the

Figure 12.8 Demonstration of the occurrence of hydroxyl radicals during CDEO.

288

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

electrode surface by Reaction (12.2) or by hydroxyl radical-mediated Reaction (12.3). •

OH · + • OH· → H2 O2 +

H2 O2 → O2 + 2H + 2e

(12.1) −

H2 O2 + • OH· → O2 + 2H2 O

12.4

(12.2) (12.3)

SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

Peroxoacids and peroxosalts are in a group of oxidants characterized by the presence of a group –O–O– in the molecules. Typically, these species come by the substitution of a group –OH by a –O–OH in oxoacids or oxosalts of the groups IV, V, and VI of the periodic table. Typical examples include peroxocarbonates, peroxonitrates, peroxosulphates, and peroxophosphates, although the last two are more important because of their actual and potential applications. The efficiency in the production of these products is high enough to study commercial applications; consequently, many applications have been proposed for these species. The main use of peroxosulphuric acids is as an initiator in polymerization processes (olefins, vinyl chloride, styrene-butadiene, vinyl acetate, acrylic ester). Its use as a reagent in the etching of printed circuit boards and in the removal of photo resists are also important. Other applications concern dyes oxidation, whitening fibers, promotion for radical polymerization, total organic compound measurements, and so on. Peroxophosphates have a wide variety of applications in areas such as oxidizing agents in organic synthesis [17,18] cosmetics [19,20], agriculture [21], wastewater treatment [22], and also as bleaching agents in the detergent industry [23,24]. Due to its similar properties, peroxodiphosphate can also be used in other processes as a substitute of peroxodisulphate. In these later uses, peroxodiphosphate has an important advantage: It is a more environmental-friendly reagent because its reduction product (phosphate) can be easily and economically removed from aqueous wastes that the reduction product of persulphates (sulphates). Electrolyses with conductive-diamond anodes have shown to be effective in the production of these species. In the following subsections, some details about the electrolytic production with diamond anodes of peroxodisulphuric acid, monoperoxophosphoric acid, and peroxodiphosphate are going to be described. 12.4.1

Peroxosulphuric Acids

Peroxosulphuric acids can be produced by electrolytic oxidation of sulphuric acid solutions. Two different species are included in this group: peroxomonosulphuric and peroxodisulphuric acid (Figure 12.9). They both have a very high reduction potential (1.81 and 2.08 V, respectively). Peroxodisulphuric acid is more stable than peroxomonosulphuric acid, but it can also suffer thermal decomposition giving to the generation peroxomonosulphuric acid and also hydrogen peroxide. The process is more significant at temperatures higher than 50◦ C, but it can be observed at any range of temperatures, and hence mixtures of both acids are typically encountered in commercial products in which the dimmer is the primary species.

12.4 SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

O HO

S O OH

O Monoperoxosulphuric acid

O

289

O

HO S O O S OH O O Peroxodisulphuric acid

Figure 12.9 Peroxosulphuric acids.

Traditionally, the synthesis of peroxosulphuric acids have been carried out by electrolyses of sulphuric acid by Reaction (12.4) with noble metal-coated anodes such as platinized, titanium, tantalum, or niobium [25]. 2H2 SO4 → H2 S2 O8 + 2e−

(12.4)

Efficiency of the process was found to depend strongly on the electrode material selected, and particularly on the high overvoltage value for oxygen evolution reaction provided by the anode material. In this context, the use of conductive-diamond electrodes was first proposed by Gandini et al. [26] and by Michaud et al. [27] in 2000, being one of the first electrochemical applications proposed for the anodes of conductive diamond. Figure 12.10 shows the results of the electrolyses with diamond electrodes of sulphuric acid solutions. As it can be observed, the production of significant concentrations of peroxosulphuric acids is obtained, and the efficiency is high enough to be used commercially. During the batch electrosynthesis, the oxidant concentration increases with the specific charge until a constant value is achieved, and current efficiency decreases continuously. This behavior could be explained in terms of mass transfer limitations (in the batch system studied, the concentration of reactant decreases continuously) or, most likely, some sort of chemical or electrochemical destruction of the oxidants formed, which promotes the concentration of a formed product. Electrolytic production of peroxosulphuric acids was found to depend significantly on the concentration of raw sulphuric acid, temperature, and current density. The effect of these parameters is shown in Figures 12.11 and 12.12, respectively. The production of peroxosulphuric acids and the efficiency of the processes increase with the concentration of raw H2 SO4 , with low temperatures and with high current densities. The influence of concentration is difficult to explain although it is as expected: The higher the concentration of raw materials, the higher the concentration of product. Preliminarily, it could be explained by mass transfer controlling mechanisms. However, concentration of sulphuric acid does not become small in any case, and hence some sort of electrochemical decomposition of peroxosulphuric acid should be considered to explain the effect of concentration. This also has to be related to the stabilization of the concentration of peroxosulphuric acids during batch electrolysis, and the plateau range of current charge could indicate the zone in which the production rate is equal to the decomposition rate. The influence of temperature is easier to explain. It is related with the thermal stability of peroxosulphuric acids, which are known to decompose with temperature to yield

290

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

Oxidants(mmol H2S2O8)

40

30

20

10

0 0

20

40

60

80

100

Q (Ah dm–3)

Current efficiency(%)

40

30

20

10

0 0

20

40 60 Q (Ah dm–3)

80

100

Figure 12.10 Variation of the oxidants () and of current efficicency () with the specific electrical charge passed in the electrosynthesis of peroxosulphuric acids (C0 : 2 M H2 SO4 ; T: 10◦ C; j: 1250 A m−2 ).

sulphuric acid and hydrogen peroxide according to the mechanisms proposed in Reactions (12.5) to (12.7) [28]. S2 O8 2− + H2 O → 2 SO4 2− + 2H+ + 1/2O2

(12.5)

2− + S2 O8 2− + H2 O → SO2− 5 + SO4 + 2H

(12.6)

SO5 2− + H2 O → H2 O2 + SO4 2−

(12.7)

The influence of current density is the more interesting fact in the production of peroxosulphuric acids: There are two different behaviors as a function of the current density. Current densities higher than 1100 A m−2 lead to a more efficient process. Initially, this observation could be related to the massive formation of hydroxyl radicals in the reaction media, which complements the direct electrochemical formation of peroxosulphuric acid

12.4 SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

291

Oxidants(mmol H2S2O8 dm–3)

30 25 20 15 10 5 0 0

0.5 1 1.5 Concentration raw H2SO4(M)

2

Figure 12.11 Effect of the concentration of raw sulphuric acid on the production of peroxosulphuric acids for a specific current charge of 40 Ah dm−3 . ( j: 1250 A m−2 , T: 10◦ C).

Oxidants(mmol H2S2O8 dm–3)

30 25 20 15 10 5 0 0

10

20 30 Temperature(°C)

40

50

Oxidants(mmol H2S2O8 dm–3)

30 25 20 15 10 5 0 0

500

1000 1500 Current density(A m–2)

2000

Figure 12.12 Effect of the operation conditions on the production of peroxosulphuric acids for a current charge of 40 A h dm−3 . (C0 : 1 M H2 SO4 , pH 2).

292

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

from Reactions (12.8) to (12.10). As it is going to state with other oxidants, current densities around 1100 A m−2 correspond to anodic potentials for water electrolysis in the experimental setup used. HSO4 − + • OH → SO4 − • + H2 O SO4

2−

+ OH → SO4 •

−•

+ H3 O

(12.8)

+

(12.9)

SO4 − • + SO4 − • → S2 O8 2− 12.4.2

(12.10)

Peroxodiphosphate Salts

Peroxophosphates chemistry is similar than that of persulphates, although peroxophosphates are less known because their synthesis methods were not very efficient, and they usually led to unpure species, but contaminated with some of the additives dosed to increase the efficiency of the production process. Similar to peroxosulphuric acid chemistry, two different species can be found: peroxomonophosphate and peroxodiphosphate salts (see Figure 12.13). Peroxomonophosphate is stable at acid pH, whereas peroxodiphosphate stability is higher at basic pH. Both are powerful oxidants, being their oxidation potentials are similar to those of persulphates (2.07 V peroxodiphosphate versus 2.08 V persulphates). Typically, potassium peroxodiphosphate, K4 P2 O8 , is obtained by electrolysis on platinum electrodes at alkaline conditions of potassium phosphate solutions containing some reagents [29], mainly fluoride or thiocianate. These reagents are primarily added to promote the blockage of oxygen evolution sites [30] of the anode in the synthesis process. Consequently, the direct oxidation of phosphate to peroxodiphosphate is favored over the water oxidation process and higher current efficiencies are obtained. However, some of the uses of peroxodiphosphates can be affected by the impurities. Likewise, some of the reagents can be highly corrosive to platinum and others can be toxic. Thus, in order to obtain impurities-free peroxodiphosphate, extensive purification is required [31]. This greatly increases the manufacturing costs. The synthesis of K4 P2 O8 can also be carried out with platinum without using any additives, but in these cases low current efficiencies are obtained [32]. Figure 12.14 shows the production of peroxodiphosphate by electrolysis with conductive-diamond of K3 PO4 in alkaline conditions (pH 12.5; T: 25◦ C; j: 1250 A m−2 ) [33]. As it can be observed, the process results are surprising. Huge conversions are obtained with very good efficiencies. In addition, peroxodiphosphate can be precipitated and separated from the reaction media simply by the addition of methanol to the electrolyzed solution. It can also be observed that the rate of generation of peroxodiphosphate decreases progressively during a batch process. This behavior (just the same that was observed in the electrochemical production of peroxosulphuric acids OH O P O OH OH Monoperoxophosphoric acid

OK

OK

O P O O P O OK

OK

Potassium peroxodiphosphate

Figure 12.13 Peroxophosphoric acids and salts.

12.4 SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

293

100

Conversion(%)

80 60 40 20 0 0

30

60

90

120

150

180

210

Q (Ah dm–3)

Current efficiencies(%)

100

80

60

40

20

0 0

30

60

90

120

150

180

210

Q (Ah dm–3) Figure 12.14 Variation of the conversion and of the current efficiency with the specific electrical charge passed in the electrosynthesis of peroxodiphosphate (C0 : 1 M K3 PO4 ; pH 12.5; T: 25◦ C; j: 1250 A m−2 ). (Reprinted from Ref. 33.)

with diamond anodes) is justified in terms of a progressive decrease in the efficiency of the process. This continuous decrease can be due to mass transfer limitations or some sort of electrochemical destruction of the oxidants formed. In this case, the high conversions obtained and the high stability of peroxodiphosphate suggests that mass transfer limits the process. Figure 12.15 shows the influence of the raw material used to produce peroxodiphosphate, particularly the concentration of potassium phosphate and the pH. As expected, process efficiencies increase with the concentration of raw phosphate in the solution to be electrolyzed and hence for the same current charge passed the production of peroxodiphosphate increases with the initial concentration of phosphates. However, on a few occasions it was observed that electrolyses of solutions with very high concentrations of K3 PO4 seemed to damage to the diamond surface and small corrosion circles appeared on the surface of the diamond after the treatment. This problem never appeared

294

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

90 80

Conversion(%)

70 60 50 40 30 20 10 0 0

1 Concentration of raw potassium phosphate/M

2

90 80 Conversion(%)

70 60 50 40 30 20 10 0 8

9

10

11 pH

12

13

14

Figure 12.15 Effect of the characteristic of the raw materials on the electrosynthesis of peroxodiphosphate at an electrical charge passed of 130 Ah dm−3 (T: 25◦ C; j: 1250 A m−2 ). (Reprinted from Ref. 33.)

when working with initial concentration of 1 M K3 PO4 ; hence, this concentration should be recommended to warrant good results, and to protect the anodic surface. Likewise, pH influences greatly in the process performance and only good yields are obtained in strongly alkaline solutions. Optimum operation pH was 12.5—the same that literature indicates is the maximum stability of peroxodiphosphate. Figure 12.16 shows the effect of the main operation parameters (temperature and current density) on the efficiency of the process. Temperatures higher than 25◦ C lead to low conversions, and thus, to low current efficiencies. The thermal decomposition of peroxodiphosphate to give pyrophosphate and oxygen from Reaction (12.11) might justify the observed decrease, although the complex chemistry of the system needs more chemical studies to clarify this point. P2 O8 4− → P2 O7 4− + 1/2O2

(12.11)

12.4 SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

295

80 70

Conversion(%)

60 50 40 30 20 10 0 10

15

20

25 30 Temperature(°C)

35

40

45

100

Conversion(%)

80 60

40 20

0 0

50

100

150

200

250

j(mA cm–2) Figure 12.16 Effect of the operation conditions on the electrosynthesis of peroxodiphosphate with diamond anodes at an electrical charge passed of 80 Ah dm−3 (C0 : 1 M K3 PO4 ; pH 12.5). (Reprinted from Refs. 33,41.)

The influence of the current density is the same as that observed for peroxosulphuric acids. It can be seen that the conversion increases continuously with the specific charge passed, and that for a given current charge passed not a continuous change is obtained as a function of the current density, but only two limit behaviors can be discerned. For current densities below 1000 A m−2 maximum conversions do not exceed 30%, whereas for higher current densities, conversions over 70% are obtained. The abrupt change in the efficiency may be related to the mechanisms of peroxodiphosphate formation on BDD surfaces (see Figure 12.17). Thus, it is proposed [30] that peroxodiphosphate can be formed by direct electrooxidation from Reaction (12.12) on the surface of electrodes such as platinum. 2PO4 3− → P2 O8 4− + 2e−

(12.12)

296

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

e–

PO4–3

anode

(PO4–2)• (PO4–2)• e–

H2O OH

Direct oxidation P2O8

–4

PO4–3 (PO4–2)•

(PO4–2)•

P2O8–4

OH mediated oxidation

Figure 12.17 Mechanism for the production of peroxodiphosphate with diamond anodes.

However, the higher efficiencies obtained with BDD anodes can be due to the presence of hydroxyl radicals. It is reported [1] that large quantities of these radicals are formed during electrolyses of aqueous solutions. These radicals can combine with phosphate ions and form PO4 2− • radicals [30]. These later radicals can oxidize other compounds in a region close to the anode surface (e.g., water peroxide to oxygen), or they can combine between them to form peroxodiphosphate or with hydroxyl radicals to form peroxomonophosphate. This can justify the higher efficiencies obtained in the electrosynthesis of these compounds when using BDD as anode materials, as these anodes combine both direct and hydroxyl-mediated oxidation processes. Conversely, this complex mechanism might justify the abrupt change in the efficiencies by the promotion of one or the two mechanisms in the electrochemical oxidation. Nevertheless, further research should be done in order to clarify this point. 12.4.3

Monoperoxophosphoric Acid

The synthesis of peroxomonophosphoric acid (H3 PO5 ) is usually based on highly exothermic chemical reactions, which, as a major drawback, led to the rapid decomposition of the peroxo-compounds generated and consequently to a low efficiency in the production of monoperoxophosphoric acids. Several improvements, such as the addition of inert diluents, have been proposed [34], but the low reproducibility of the results limits their use. For this reason, the hydrolysis of peroxodiphosphate salts in very acidic conditions appeared as one of the best synthesis method of peroxomonophosphoric acid [35]. However, the requirement of powerful reagents (peroxodiphosphate and perchloric acid or hydrogen peroxide), and its large manufacturing costs have limited its commercial application. In this context, the direct electrochemical production of monoperoxophosphoric acid using conductive-diamond electrodes can become in an important way of manufacturing this product. Figure 12.18 shows the results of batch electrolyses with conductive diamond of phosphoric acid [36]. As shown, monoperoxophosphoric acid is formed. As in the synthesis of other compounds, oxidant concentration increases to achieve a plateau. This plateau value results from the compensation of the rate of formation of

12.4 SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

297

4.5

Oxidants(mmol PO53–)

4 3.5 3 2.5 2 1.5 1 0.5 0 0

20

0

20

40 Q (Ah dm–3)

60

80

4.5 4

Efficiency(%)

3.5 3 2.5 2 1.5 1 0.5 0 40 Q (Ah dm–3)

60

80

Figure 12.18 Variation of the monoperoxophosphoric acid concentration and current efficiencies with electrical charge passed in the electrolysis of phosphoric acid (C0 : 1 M H3 PO4 ; pH 1.2; T: 13◦ C; j: 64 A m−2 ). (Reprinted from Ref. 36.)

oxidant and its decomposition rate. The results obtained exceed those obtained with other methods. However, the efficiency and the amount of oxidant generated are significantly smaller than the production of peroxodiphosphates salts using electrolyses of alkaline potassium phosphate. In spite of that, the efficiencies are high enough to suggest that this method could be a promising alternative to acidification of peroxodiphosphate solutions. Figure 12.19 shows the influence of the raw phosphoric acid (concentration and pH of the solution) on the production of monoperoxophosphoric acid for a particular electrical current charge passed. With respect to the H3 PO4 concentration, the amount of oxidant obtained is largely influenced by the concentration of the solute. The quantity of peroxomonophosphate increases significantly with the amount of H2 PO4 − ions available to be oxidized until achieving a given value (around 1 M H3 PO4 ). and then a decrease in the amount of peroxomonophosphate produced is observed. The optimum concentration for the peroxomonophosphate synthesis with BDD anodes in the operating conditions used in this work is 1 M H3 PO4 .

298

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

Oxidants(mmol PO53–)

1 0.8 0.6 0.4 0.2 0 0

0.5

1

1.5

2

2.5

3

5

6

Initial concentration(mol dm–3)

Oxidants(mmol PO53–)

5 4 3 2 1 0 0

1

2

3

4

pH Figure 12.19 Amount of oxidant generated in the electrosynthesis of peroxomonophosphate as a function of the initial concentration of H3 PO4 at an electrical charge passed of 1.5 Ah dm−3 (pH 5; T: 13◦ C; j: 20 A m−2 ). (Reprinted from Ref. 36.)

This unexpected behavior has to be explained in terms of the hydroxyl radical’s role in the formation of monoperoxophosphoric acid. Thus, hydroxyl radicals can contribute to the generation of H2 PO4• . This radical can be produced by direct electrolyses from Reaction (12.13) on the electrode surface or by the action of hydroxyl radicals from Reaction (12.14) in the nearest of the electrode. Moreover, the presence of hydroxyl radicals is also needed to promote the generation of peroxomonophosphoric acid by Reaction (12.15). As a result, an increment of the initial concentration of phosphate is not enough to ensure the generation of higher amount of oxidant generated, but the coexistence of both radical reagents is required. H2 PO4 − → (H2 PO4 ) • + e−

(12.13)

H2 PO4 − + • OH· → (H2 PO4 ) • + OH−

(12.14) −1 −1

(H2 PO4 ) + OH → H3 PO5 k = 4 · 10 L mol s •

•

9

(12.15)

12.4 SYNTHESIS OF PEROXOACIDS AND PEROXOSALTS

299

Conversely, the pH seems not to have a strong influence on the results, and within the range of 1 to 5 almost no changes are observed. This is important because due to the acid-base equilibrium given by Reactions (12.16) to (12.18), the hydrogenphosphoric anion can also be precursor of the formation of monoperoxophosphoric acid by Reaction (12.19). H3 PO4 ↔ H2 PO4 − + H+

pKa = 2.14

(12.16)

H2 PO4 − ↔ HPO4 2− + H+

pKa = 6.86

(12.17)

HPO4 2− ↔ PO4 3− + H+

pKa = 12.4

(12.18)

HPO4 2− → (HPO4 − ) • + e−

(12.19)

Figure 12.20 shows the influence of the operation conditions (temperature and current density) on the process at particular current charge passed. As it can be observed, the

Oxidants(mmol PO53–)

4

3

2

1

0 0

5

10

15 20 Temperature(°C)

25

30

35

Oxidants(mmol PO53–)

5

4 3

2 1 0 0

300

600

900

1200

1500

Current density(Am−2)

Figure 12.20 Amount of oxidant generated in the electrosynthesis of peroxomonophosphate as a function of the temperature at an electrical charge passed of 10 Ah dm−3 (1 M H3 PO4 ; pH 5; j: 64 A m−2 ). (Reprinted from Refs. 36,41.)

300

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

conversion of phosphoric acid to monoperoxophosphoric acid shows a marked dependence with this operating parameter. The higher the temperature, the lower the oxidant concentrations obtained at the steady state. The influence of the current density is just as reported for the oxidants previously described in this chapter. Two significantly different types of behavior can be discerned. At low current densities, a smaller efficiency suggests that the direct mechanism is the main responsible for peroxoacid generation. At higher intensities, the mechanism may involve also the hydroxyl radicals as it was explained previously.

12.5 12.5.1

SYNTHESIS OF HALOGEN OXOANIONS Perchlorates

The good properties in the production of oxidants have encouraged the interest of comparing its performance with the performance of dimensionally stable anodes (DSA) in the production of hypochlorite. However, the properties of conductive diamond are completely different from that of DSA. Thus, Figure 12.21 shows the variation of the concentration of the main oxidant species generated during the electrolysis of 0.1 M NaCl solutions with DSA and BDD anodes. As it can be observed, hypochlorite, chlorite, and chlorate are the species produced in higher concentrations for low current charges in the case of conductive-diamond anodes [37–39]. Likewise, from a given electrical charge passed, significant amounts of perchlorate begin to be detected in the reaction system and the conversion to this product is quantitative for large current charges. These results are opposite to those obtained with other electrodes such as DSA in which chloride ions are almost only transformed into hypochlorite ions. It is well documented [40] that DSA electrodes can be successfully used for the electrochemical generation of hypochlorite, although they are not able to attain significant concentrations of perchlorate salts. 12.5.2

Perbromates

Perbromates are oxoanions of bromine that cannot be presently found commercially due to the nonexistence of a good synthesis method. Figure 12.22 shows results about the production of perbromate during the oxidation of bromate solutions at strong alkaline media. It can be easily verified that conductive diamond allows the further oxidation of bromate ions, giving to the formation of Br(VII) species [41]. This observation is a very important insight because, to the authors’ knowledge, the electrochemical generation of perbromate has not been previously reported by electrochemical methods in aqueous media. However, at the present moment, the low concentration attained and the low current efficiency serve as warnings about the use of this electrochemical technique as a promising synthesis method. So, further studies must be done to establish the optimum operating conditions that allow improving the current efficiency and the generation of large amount of perbromate.

12.6 SYNTHESIS OF FERRATES

301

90 DSA 80

% Conversion

70 60 50 40 30 20 10 0 0

50

100

150

Q (Ah dm–3) 40

BDD

35

% Conversion

30 25 20 15 10 5 0 0

5

10

15

20

25

30

Q (Ah dm–3) Figure 12.21 Production of perchlorates with dimensionally stable anodes (DSA) and conductive− diamond anodes (DDB) (C0 : 0.1 M NaCl, pH 10, T: 35◦ C, j: 300 Am−2 ). () ClO− , (•) ClO− 3 , () ClO2 , . () ClO− 4

12.6

SYNTHESIS OF FERRATES

Typical oxidation states for iron species are +2 and +3; however, unusual oxidation states are observed to be stable in particular conditions, such as +6 in the form of ferrate (VI), FeO4 2− . Ferrate ions are very powerful oxidizing agents with standard halfcell reduction potentials ranging from 2.20 V at acidic pHs to 0.72 V versus NHE at alkaline condition [42,43]. Moreover, during the oxidation process, ferrate (VI) ions will be reduced to Fe (III) ions or ferric hydroxide, making them suitable to be used in a wide

302

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

12

Oxidants(mmol BrO4–)

10 8 6 4 2 0 0

100

200

300

400

Q (Ah dm–3) Figure 12.22 Production of perbromates with conductive diamond anodes. (C0 : 0.1 M NaBrO3 , pH 10, T: 17◦ C, j: 1000 Am−2 ).

range of applications: “green” organic synthesis [44,45], wastewater treatment (oxidation and coagulation) [46,47], and water treatment (persistent disinfection, coagulation, and oxidation of nondesirable compounds) [42,48–50]. Due to their highly oxidized iron basis, multiple electron transfer, and high intrinsic energy, ferrate (VI) can be used as “super-iron” catode [45]. The ferrate synthesis can be divided into the following three categories: (1) thermal chemical synthesis, by heating/melting various iron oxide containing minerals under conditions of strong alkaline and oxygen flow [51]; (2) wet chemical synthesis, by oxidizing Fe(III) salt at strong alkaline condition and using hypochlorite or chlorine as the oxidant [52] or electrochemical techniques; and (3) anodic oxidation using iron or alloy as anode and NaOH or KOH as electrolyte [53]. However, these generation techniques have shown important drawbacks. Thus, it is reported that thermal techniques require the use of very high temperatures that unfortunately also favor the decomposition rate of the ferrate generated. Likewise, wet or dry oxidation technologies also lead to important drawbacks such as low efficiencies and in some cases the use of hazardous compounds as reagents. In this context, electrochemical synthesis of ferrate using iron electrodes in alkaline solutions has shown better yields, but it has also shown significant problems such as the formation of passivation layers on the electrode surfaces. Figure 12.23 shows the variation of the concentration of ferrate and the current efficiencies during the electrolysis with conductive-diamond anodes of an alkaline solution (14 M NaOH saturated with iron (III) hydroxide). As can be seen, there is a rapid increase in the ferrates concentration during the first stages of the electrolyses. Then, the oxidation rate decreases markedly to a constant value, and the ferrate concentration starts to increase in a slower way. Consequently, the current efficiency of the electrosynthesis decreases during the electrolyses being very low even at the initial stages of the electrolyses. In this context, the maximum concentration of soluble iron species (around 0.2 mM) clearly suggests that the small amounts of available iron can limit markedly the efficiency of the electrosynthesis [54,55].

12.6 SYNTHESIS OF FERRATES

303

Oxidants / mmol FeO42–dm–3

0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 0

1

2

3

4

Q (Ah dm–3)

Current efficiency/%

1 0.8 0.6 0.4 0.2 0 0

1

2

3

4

Q (Ah dm–3) Figure 12.23 Batch production of ferrates with diamond electrodes (C0 :14 M NaOH, j: 1300 A m−2 , T: 30◦ C). (Reprinted from Ref. 54.)

Figure 12.24 shows the effect of the raw matter on the efficiency of the process. Notice that the process depends strongly on the concentration of hydroxyl ions, and it is only efficient for a very high content of this anion. According to literature, the stability of ferrate is greatly influenced by the pH of the system [45,56,57], and ferrate salts seem to be more stable in strongly alkaline conditions. In this context, hydroxide anion concentration was also found to be a key parameter in the electrosynthesis of ferrates with iron electrodes [45]. To increase the availability of oxidizable-iron species, and therefore improve the efficiency of the electrosynthesis with BDD, an iron-powder bed placed near to the anode surface (separated of the anode surface by means of a very thin plastic mesh) was used as raw material. During the electrolysis of aqueous solutions, the oxygen evolution that takes place on the anodic surface, consumes large amounts of hydroxyl anions. This fact leads to changes in the pH in a region very close to the anode surface that can favor the chemical dissolution of the iron particles [58]. This would increase the amount of iron ionic species in the reaction system (coming from the dissolution of iron particles) and, consequently, the efficiencies of the electrosynthesis of ferrate. Part b of Figure 12.24 shows the variation of the ferrate concentration with the electrical charge passed during

304

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

Oxidants/mmol FeO42–dm–3

0.1

0.08

0.06

0.04

0.02

0 0

5 10 NaOH(mol dm–3)

15

Oxidants/mmol FeO42–dm–3

1.2 1 0.8 0.6 0.4 0.2 0 0

50

100 150 200 250 300 350 400 450 500 550 Q (Ah dm–3)

Figure 12.24 Effect of the raw materials on the production of ferrates with diamond anodes (C0 : 10 M KOH, j: 1000 A m−2 , T: 10◦ C). (Reprinted from Ref. 55.)

the electrolysis of hydroxide solution (10 M KOH) using iron-powder as raw iron. As the figure shows, high ferrates concentrations are obtained (onefold higher than those obtained with Fe(OH)3 ). This also means that efficiencies increase by onefold and the process could become interesting from the commercial viewpoint. Figure 12.25 shows the effect of the operation conditions in electrolysis with the ironpowder bed as raw material for two particular current charge passed. The concentration of obtained ferrates is strongly dependent on the temperature of the electrosynthesis, and the generation process seems to be favored at temperatures around 25◦ C. The observed maximum may be explained in terms of two opposite processes: the solubility of iron (III) (raw material) and the stability of ferrates. Thus, low temperature can lead to lower concentration of iron species available to be oxidized and thus to a more significant masstransfer control. This explains the lower efficiency obtained at this operation condition.

12.7 EFFECT OF THE TYPE OF DIAMOND ON THE EFFICIENCY OF THE PRODUCTION OF OXIDANTS

305

Oxidants / mmol FeO42–dm–3

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0

1500 500 1000 Current density(A m–2)

2000

Oxidants / mmol FeO42–dm–3

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0

10

20 30 40 Temperature(°C)

50

60

Figure 12.25 Effect of the operation conditions on the results of the production of ferrates (C0 : 10 M KOH). (Reprinted from Ref. 54.)

Conversely, higher temperatures can disfavor the stability of the oxidant electrogenerated and lead to higher decomposition rate. The effect of current density is as expected according to the results shown for other oxidants produced with conductive diamond. The higher the current density, the better the results obtained. This behavior can be explained in terms [27] of the hydroxyl radical contribution, and indicates that the role of hydroxyl radicals is not only important in the production of peroxo compounds but also on the production of other oxidants. 12.7 EFFECT OF THE TYPE OF DIAMOND ON THE EFFICIENCY OF THE PRODUCTION OF OXIDANTS The rapid development of the diamond technology has focused attention on the search for applications and not on the fundamental aspects of diamond technology. Electrocatalysis of conductive-diamond anodes is still unclear, and the effects of the characteristics of diamond on the process efficiencies have to be further studied in the near future. An example of this is shown in Figure 12.26 for the production of peroxophosphates salts during the discontinuous electrolyses of alkaline solutions containing 1 M K3 PO4 .

306

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

2.2 Oxidants/mmoles P2O84–

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0

0.5

1

2

1.5

Q (Ah dm–3) Figure 12.26 Electrolyses of potassium phosphate solution at pH12.5 and current density of 1250 Am−2 . (♦) 1049-4, () 1038-4, () 1050-4, (×) 983-5, ( ) 802-3, (◦) 60–61, (+) 908-5, () 858-5, (-) 792-1, ♦ 805-1, (•) 805-3, () 825-6.

Table 12.1 summarizes the main characteristics of the lots of conductive-diamond electrodes used in this work [59,60]. In every case, the pH was controlled at 12.5 ± 0.1, and the current density was fixed at 1250 A m−2 . As it can be observed, the use of conductive diamond with different characteristics leads to a variety of responses in which difference is even more than onefold. However, almost no differences are observed if the experiments are repeated with the same electrodes or with other electrodes of the same lot. This observation confirms that the characteristics of electrodes influence strongly on the electrosynthesis results; hence, they should be taken into account in the study of the applications of conductive-diamond electrolyses. Figure 12.27 shows the results of simple statistical analyses of these results in which the effect of different parameters of the diamond surfaces on the efficiency of production of peroxophosphates is compared. As seen, boron content and thickness of the diamond anode seem to have a significant effect on the production of peroxodiphosphates, ˜ TABLE 12.1 Characteristics of the conductive-diamond electrodes lots used by Canizares et al. [54].

Reference of the BDD 1049-4 1050-4 60-61-G1 1038-4 825-6 908-5 983-5 792-1 805-1/3 858-5 802-3

Conductive-diamond layer Boron Ratio Thickness BDD layer (μm) contents (ppm) sp3 /sp2 100 200 500 1300 1300 2500 1300 100 1300 2500 8000

65 75 93 66 45 68 77 89 105 43 80

1.09 1.14 2.4 1.33 2.33 1.15 2.27 1.03 2.25 1.13 1.05

p-Si substrate Si-Resistivity Roughness, (m cm) Si-Surfinra (μm) 100 100 100 100 100 100 10 10 10 10 10

0.3-0.5 0.3-0.5 0.3-0.5 0.3-0.5 <0.1 <0.1 0.3-0.5 <0.1 <0.1 <0.1 <0.1

12.8 CONCLUSIONS

307

0.6 0.4

99.5 % significance

0.2

95.0 % significance

0

Effect =

–0.2

95.0 % significance

–0.4

99.5 % significance

Positive y+ – n+

y− n−

Negative

–0.6 –0.8

sp

re si

st iv i ty r ou 3/ sp gh 2 ne sp (fr ss 3/ om th sp ic lo 2 kn w (fr es to om s m sp m ed 3/ ed i sp u iu m 2 m ) bo (fr to ro o m hi n gh lo bo (fro w ) m ro t o n lo hi w (fr gh om to ) m m e ed di bo um iu ro m ) n to (fr om hi gh lo ) w to hi gh )

Effect/mmol peroxophosphate (Ah)–1

0.8

Parameter

Figure 12.27 Statistical analyses of the influence of the main properties of diamond anodes on the production of peroxodiphosphates.

appearing around the line of 99.5% of significance. This means that these are the most significant parameters from a statistical viewpoint. According to these observations, an increase in the boron content or a decrease in the thickness of the diamond layer improves the efficiencies of the electrolyses. A graphical analysis of the influence of the boron content does not result in any significant conclusion, except for a very large value of the efficiency in the case of the highly boron-doped diamond (8000 ppm). However, the graphical analysis of the influence of the thickness shows some important observations. Points corresponding to the thinner layer are significantly over the points corresponding to the thicker layer and the same points are randomly distributed as a function of the characteristic of the p-Si substrate. In addition, there is no significant influence of the diamond to graphite ratio (except for points corresponding to thicker layer are below in this graph) and the boron content does not influence except for the 8000 ppm boron-doped diamond. At the light of the actual knowledge, it is difficult to explain the results obtained. However, they would be related to the nature of the mechanisms involved in the oxidation processes.

12.8

CONCLUSIONS

From this work, the following conclusions can be drawn: — Strong oxidants can be produced by electrolysis with conductive-diamond anodes. The process is particularly efficient in the production of peroxoacids and peroxosalts.

308

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

— Current density has important influences on the results obtained. Values higher than 1000 A m−2 promotes the formation of hydroxyl radicals and increases the efficiency of the processes. — The oxidation of chlorine and bromine species leads to the formation of the perhalogen acid. This is particularly significant in the case of bromine because it allows a new route to produce this species efficiently. — Production of ferrates is highly efficient and is limited only by mass transport. Use of iron-beds allows significant increases in the efficiency of the process because of the higher availability of oxidizable-iron species in the reaction system. — Properties of diamond surfaces strongly influence the process efficiency. Thin layer seems to increase the yield of the processes.

12.9

ACKNOWLEDGMENTS

Financial support from Consejer´ıa de Educaci´on y Ciencia of the Junta de Comunidades de Castilla-La Mancha through the project PCI-08-0068-9073 is gratefully acknowledged.

REFERENCES 1. B. Marselli, J. Garc´ıa-G´omez, P.A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 150 , 79–83. 2. P. Ca˜nizares, R. Paz, C. S´aez, M.A. Rodrigo, J. Electrochem. Soc. 2007, 154 , 37–44. 3. P. Ca˜nizares, C. S´aez, J. Lobato, M.A. Rodrigo, Afinidad 2006, 63 , 19–25. 4. M. Panizza, G. Cerisola, Electrochim. Acta 2003, 48 , 1515–1519. 5. J. Hastie, D. Bejan, M. Teutli-Leon, N.J. Bunce, Ind. Eng. Chem. Res. 2006, 45 , 4898–4904. 6. A. Cabeza, A.M. Urtiaga, I. Ortiz, Ind. Eng. Chem. 2007, 46 , 1439–1446. 7. M. Mascia, A. Vacca, A.M. Polcaro, S. Palmas, J. Rodriguez Ruiz, A. Da Pozzo, J. Hazard. Mater. 2010, 174 , 314–322. 8. E. Brillas, I. Sir´es, C. Arias, P.L. Cabot, F. Centellas, R.M. Rodr´ıguez, J.A. Garrido, Chemosphere 2005, 58 , 399–406. 9. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191–199. 10. E. Guinea, E. Brillas, F. Centellas, P. Ca˜nizares, M.A. Rodrigo, C. S´aez, Water Res. 2009, 43 , 2131–2138. 11. P. Ca˜nizares, B. Louhichi, A. Gadri, B. Nasr, R. Paz, M.A. Rodrigo, C. S´aez, J. Hazard. Mat . 2007, 146 , 552–557. 12. P. Ca˜nizares, M. Hernandez, M.A. Rodrigo, C. Saez, C.E. Barrera, G. Roa, Ind. Eng. Chem. Res. 2009, 48 , 1298–1301. 13. E. Kilinc, Talanta 2005, 65 , 876–881. 14. A. Kraft, M. Stadelmann, M. W¨unsche, M. Clanschke, Electrochem. Commun. 2006, 8 , 883–886. 15. A. Kapałka, G. F´oti, C. Comninellis, Electrochim. Acta 2009, 54 , 2018–2023. 16. P.A. Michaud, M. Panizza, L. Ouattara, T. Diaco, G. F´oti, C. Comninellis, J. Appl. Electrochem. 2003, 33 , 151–154. 17. J.A. Rosso, F.J. Rodr´ıguez, M.C. Gonz´alez, D.O. Martire, J. Photochem. Photobiol. A. Chem. 1998, 116 , 21–25.

REFERENCES

309

18. J.A. Rosso, P. Caregnato, V.C. Mora, M.C. Gonz´alez, D.O. Martire, Helvetica Chim. Acta 2003, 86 , 2509–2524. 19. B. Cohen, P.R. Mucenieks, FMC Corp. 1971 US Patent 3 555 147. 20. B. Puetzer, A.V. Finn, L. Mackles, Tintex Corp. 1967 US Patent 3 337 466. 21. Hunan Agricultural College, T’uJang (Nanking) 1978 (1), 9–11. 22. A.S. Groisman, A.G. Belyaeva, O.B. Khachaturayn, I.Y. Sugurova, 1991 US Patent 1 699 953 A1. 23. A. Gaffar, D.B. Calvin, 1983 DE Patent 3 246 946 A1. 24. A. Gaffar, C.J. Edwards, 1983 DE Patent 3 627 759 A1. 25. G.B. Balazs, P.R. Lewis 1999 U.S. Patent 5919350. 26. D. Gandini, E. Mah´e, P.A. Michaud, W. Haenni, A. Perret, C. Comninellis, J. Appl. Electrochem. 2000, 30 , 1345–1350. 27. P.A. Michaud, E. Mah´e, W. Haenni, A. Perret, C. Comninellis, Electrochem. Solid-State Lett . 2000, 3 , 77–79. 28. K. Serrano, P.A. Michaud, C. Comninellis, A. Savall, Electrochim. Acta 2002, 48 , 431–436. 29. C.J. Battaglia, J.O. Edwards, Inorg. Chem. 1965, 4 , 552–558. 30. A. El Ghazali, S. Beldaci, J.J. Rameua, F. Dalard, J. Appl. Electrochem. 1998, 28 , 725–735. 31. J.S.C. Chiang, 1986 US Patent 4 626 326. 32. O.G. Tyurikava, N.B. Miller, 1984 US Patent 1 089 174 A. 33. P. Ca˜nizares, F. Larrondo, J. Lobato, M.A. Rodrigo, C. S´aez, J. Electrochem. Soc. 2005, 152 , D191–D196. 34. G. Toennies, J. Am. Chem. Soc. 1937, 59 , 555–557. 35. T. Zhu, H. Chang, J.F. Kadla, Can. J. Chem. 2003, 81 , 156–160. 36. E. Weiss, C. S´aez, K. Groenen-Serrano, P. Ca˜nizares, A. Savall, M.A. Rodrigo, J. Appl. Electrochem. 2007, 38 , 93–100. 37. H. Bergmann, J. Rollin, Catalysis Today 2007, 124 , 198–203. 38. H. Bergmann, J. Rollin, L. Lourtchouk, Electrochim. Acta 2008, 54 , 2102–2107. 39. P. Ca˜nizares, C. S´aez, A. S´anchez-Carretero, M.A. Rodrigo, J. Appl. Electrochem. 2009, DOI: 10.1007/s10800-009-9792-7. 40. C.R. Costa, C. Botta, E. Espindola, P. Olivi, J. Hazard. Mater. 2008, 153 , 616–627. 41. C. S´aez, P. Ca˜nizares, A. S´anchez-Carretero, M.A. Rodrigo, J. Appl. Electrochem. 2010, DOI 10.1007/s10800-010-0108-8. 42. V.K. Sharma, Adv. Environ. Res. 2002, 6 , 143–156. 43. C. Li, X.Z. Li, N. Graham, Chemosphere 2005, 61 , 537–543. 44. L. Delaude, P. Laszlo, J. Org. Chem. 1996, 61 , 6360–6370. 45. W. He, J. Wang, C. Yang, J. Zhang, Electrochim. Acta 2006, 51 , 1967–1973. 46. D.S.J. Luca, M. Cantelli, D.M.S. Luca, Water Sci. Technol . 1992, 26 , 2077–2080. 47. J.Q. Jiang, B. Lloyd, Water Res. 2002, 36 , 1397–1408. 48. M. Gilbert, T.D. Waite, C. Hare, J. Am. Water Works Assoc. 1976, 56 , 466–474. 49. A. Basu, K.R. Williams, J. Biol. Chem. 1987, 262 , 9601–9607. 50. S. Licht, B. Wang, S. Ghosh, Science 1999, 285 , 1039–1042. 51. Y.M. Kiselev, N.S. Kopelev, N.A. Zavyalova, Y.D. Perfiliev, P.E. Kazin, Russ. J. Inorg. Chem. 1989, 34 , 1250–1253. 52. G.W. Thompson, L.T. Ockerman, J.M. Schreyer, Chem. Anal . 1951, 73 , 1379–1381. 53. K. Bouzek, I. Rousar, J. Appl. Electrochem. 1996, 26 , 919–924. 54. C. S´aez, P. Ca˜nizares, M.A. Rodrigo, AIChE Journal 2008, 54 , 1600–1607.

310 55. 56. 57. 58. 59.

PRODUCTION OF STRONG OXIDIZING SUBSTANCES WITH BDD ANODES

P. Ca˜nizares, M. Arc´ıs, C. S´aez, M.A. Rodrigo, Electrochem. Commun. 2007, 9 , 2286–2290. M. De Koninck, T. Brousse, D. Belanger, Electrochim. Acta 2003, 48 , 1425–1433. V. Lescuras-Darrou, F. Lapicque, G. Valentin, J. Appl. Electrochem. 2002, 32 , 57–63. J.J. Pignatello, D. Liu, P. Huston, Environ. Sci. Technol. 1999, 33 , 1832–1839. P. Ca˜nizares, C. S´aez, A. S´anchez-Carretero, M.A. Rodrigo, Electrochem. Commun. 2008, 10 , 602–606. 60. E. Guinea, F. Centellas, E. Brillas, P. Ca˜nizares, C. S´aez, M.A. Rodrigo, Appl. Catal. B. Environ. 2009, 89 , 645–650.

13 Ozone Generation Using Boron-Doped Diamond Electrodes Yunny Meas, Luis A. Godinez, and Erika Bustos

13.1

INTRODUCTION

Ozone is an attractive gas for wastewater treatment due to its high oxidizing power; it can oxidize a large number of organic and inorganic compounds. This chapter reviews the various ozone technologies and focuses on the production of ozone using boron-doped diamond (BDD) electrodes by electrochemical methods. BDD is a good candidate for use in electrochemical production of O3 due to its high overpotential for O2 production, high chemical stability to corrosion, and low adsorption properties. Although the efficiency of BDD ozone production could be improved by modifying the BDD composition and by increasing the rate of adsorption of oxygenated species in the reaction mechanism, BDD is a good candidate material for electrodes in ozone generators.

13.2

OZONE

The discovery of ozone was officially announced by Sch¨onbein [1], at the Academy of Munich, in 1840. The name ozone was derived from the Greek word ozein, meaning ‘to smell’ [2]. In 1845, ozone was obtained by submitting pure dry oxygen to an electric spark [3], and later it was described as existing in four electronic resonance structures (see Figure 13.1) [4,5]. Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

311

312

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

O O

O O

O

O O

O

O O

O

O

Figure 13.1 Structure of the ozone molecule as four canonical forms.

13.2.1

Physical and Chemical Properties of Ozone

Ozone (O3 ) has a molecular weight of 48 g/mol, a gas density of 2.144 g L−1 at 0◦ C, and a boiling point of −112◦ C at atmospheric pressure. Ozone follows Henry’s Law, which states that at equilibrium, the concentration of a gas in water is directly proportional to its partial pressure in the vapor phase above the liquid (i.e., its gas phase concentration). The proportionality constant (Henry’s constant) varies with temperature. Ozone solubility decreases with decreasing temperature, in order of the Henry’s Law (Equation 13.1) considering the adequated conversi´on for other conditions (Equation 13.2), where Cliquid is the dissolved concentration in liquid, Cgas is the gas concentration, β is the Bunsen coefficient (solubility) with temperature dependent, Pgas is the gas pressure, T1 in K and P1 in Pa [6]. Cliquid = Cgas βtemperature Pgas

(13.1)

[O3 ](T1,P1) = [O3 ](T0,P0) (273.15/T1 )(P1 /101325)

(13.2)

Ozone is a reactive substance. The level of measurable dissolved ozone at any instant in time is affected by the specific aqueous conditions present, including pH, alkalinity, and ozone demand from any oxidizable substances present [7]. Ozone is much more soluble in acetic acid, acetic anhydride, dichloroacetic acid, chloroform, and carbon tetrachloride than it is in water. Ozone is an unstable molecule that quickly reacts to form oxygen. The half-life of ozone in environmental air is short, depending on the temperature and humidity. The half-life in water ranges between seconds, hours, or days, depending on the temperature, pressure, velocity, pH, and oxidizable substance concentrations (see Table 13.1) [6,7].

TABLE 13.1 Typical ozone half-life in water at normal conditions: 21◦ C and 1 atm. Phase of ozone

Gas Dissolved at pH 6.0 Dissolved at pH 7.0 Dissolved at pH 8.0

Half-life time ≈3 months ≈18 days ≈8 days ≈3 days ≈1.5 h ≈1.5 s ≈20 min ≈30 min ≈20 min ≈15 min ≈12 min ≈8 min 5 min

Temperature (◦ C) −50 −35 −25 20 120 250 21 15 20 25 30 35 21

Stability

Stable Unstable Stable Stable Unstable

13.2 OZONE

313

Molecular ozone (O3 ) is more stable in high purity water when the pH is less than 6. As pH increases above 7.0, hydroxyl free radicals are formed as ozone decomposes: O3 + H2 O → O2 + • OH + HO− [2,7]. The rate of ozone decomposition to form • OH increases with increasing pH, with the process becoming essentially instantaneous at pH 10.0. The hydroxyl free radical, with available hydrogen (H+ ) ions, is a stronger oxidizing agent than molecular ozone (oxidation potential of 2.76 V versus 2.076 V for O3 ). However, hydroxyl free radicals are so unstable and reactive that they are consumed within microseconds by most oxidizable constituents (e.g., bicarbonate or carbonate ions and organic compounds dissolved in the water) [8]. The principal physic-chemical characteristics of ozone are summarized in Table 13.2. 13.2.2

Production of Ozone

Usually, ozone (O3 ) is generated by the exposure of air or another gas containing normal oxygen (O2 ) to a high-energy source, which, in commercial production, is a high electrical voltage discharge or ultraviolet radiation. Typical oxygen sources are air, bottled “dry” oxygen, liquid oxygen (LOX), and oxygen concentrators [7–9]. Ozone must be manufactured on site for immediate use because it is unstable and quickly decomposes to form oxygen [10]. When generated from air, ozone has a bluish color in the generator cell, but ozone/air mixtures are invisible even at the high concentrations that exit ozone generators. The distinctive, pungent odor of ozone is readily detected at concentrations above 0.02 ppm in air. Although the gas is only partially soluble in water, concentrations up to 10 ppm can be achieved in water, under normal conditions [7,11]. To produce ozone it is necessary to deliver 10–20 kW h per kg of O3 at 1–5 g m−3 concentration. For this reason, the ionization is 2 or 3 times more expensive than chlorination [10]. O3 is known to be generated simultaneously with O2 during anodic discharge if particular electrode materials are used in a variety of aqueous media [12]. 13.2.3

Importance of Ozone Applications

O3 is a clean and strong oxidant, and its application in so-called green chemical processes (e.g. pollutant combustion, removal of contaminants, etc.) has received special attention

TABLE 13.2 Physico–chemical properties of ozone [6–9,11]. Properties Color

Molecular Weight Density (0◦ C and 101.3 KPa) Boiling Point (101.3 KPa) Fusion Point of Solid O3 Redox Potential Bunsen coefficient (solubility) in water at 0◦ C

Values Gas: blue colored Dissolved in water: purple blue in concentration >20 ppm 48 g mol−1 2.154 g L−1 –111.9◦ C –192.5◦ C 2.07 V (Hydroxyl radical • OH 2.80 V) 20 mg L−1

314

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

in recent years [13–23]. It is used to minimize the environmental impact of industrial activity, because it is a very powerful oxidant and does not produce toxic products on decomposition [14,16,20,24–30], in contrast with other highly oxidizing compounds. Ozone, therefore, is chosen for industrial processes as the best available technology with minimum cost and without pollution [2,11,31]. The ozone used in both potable water and wastewater treatment may be classified as both an oxidant and a germicidal compound. It therefore has the same properties exhibited by aqueous chlorine, and there is tendency to view the two substances as competitors. Therefore, the distinct properties of ozone are (1) as a bactericide, (2) as a vermicide, and (3) as a powerful chemical oxidant [11,17,23,32,33]. The solubility of ozone in water is a limiting factor that greatly affects the process of ozonation. Although ozone is more soluble than oxygen, chlorine is 12 times more soluble than ozone. In pure aqueous solutions, ozone is thought to decompose as follows [11]: O3 + H2 O → HO3 + + OH− +

−

(13.3)

HO3 + OH → 2HO2

(13.4)

O3 + HO2 → • OH + 2O2

(13.5)

OH + HO2 → H2 O + O2

(13.6)

•

The free radicals (HO2 + and • OH) that form when ozone decomposes in aqueous solutions have great oxidizing power; they may react with impurities present (e.g., metal salts, organic matter, hydrogen, and hydroxide ions present in solution). Ozone, while it exists, does not loose its oxidizing capacity in an aqueous solution. When ozone reacts with hydrogen peroxide, the process is called peroxone, and a hydroxyl radical is formed [10,11,34,35]. Ozone is a powerful oxidizing agent. Its oxidation potential is −2.07 V versus the hydrogen electrode at 25◦ C and at unity H-ion activity. Only fluorine has a more electronegative oxidation potential [11]. The oxidation-reduction potentials provided in Table 13.3 show that ozone is the most powerful of the commonly used drinking water TABLE 13.3 [7,36].

Standard potentials of different oxidants

Reaction OH + H+ + e− → H2 O O3 + 2H+ + 2e− → O2 + H2 O • OH + e− → HO− H2 O2 + 2H+ + 2e− → 2H2 O + − MnO− 4 + 4H + 3e → MnO2 + 2H2 O + HOCl + H + 2e− ↔ Cl− + H2 O HOBr + H+ + 2e− ↔ Br− + H2 O O2 + 4H+ + 4e− ↔ 2H2 O HOI + H+ + e− ↔ I− + H2 O CIO2(aq) + e− ↔ CIO− 2 OCl− + H2 O + 2e− → Cl− + 2OH− OBr− + H2 O + 2e− → Br− + 2OH− OI− + H2 O + 2e− → I− + 2OH− •

Oxidation-reduction potential (V) 2.760 2.076 2.020 1.776 1.679 1.482 1.331 1.229 0.987 0.954 0.810 0.761 0.485

13.2 OZONE

315

oxidants [7]. Ozonation is one of the most widely used advanced oxidation technologies with strongly increased process efficiencies [10,34,35]. Frequently identified by-products from the ozonation of organic substances contained in drinking or wastewaters are aldehydes, carboxylic acids, and other aliphatic aromatic or mixed oxidized compounds. Such substances are often quite easily biodegradable, and it is not surprising to observe the decrease of toxic effects [36,37]. Certain halogenated and other heat-resistant or so-called refractory organics (e.g., benzene, tetrachloroethylene, or other volatile organic compounds (VOCs) in drinking water) can be oxidized only by hydroxyl free radicals ( • OH) that are produced when ozonation is coupled with ultraviolet (UV) radiation and/or hydrogen peroxide (H2 O2 ) treatment [7,10]. These coupled ozonation processes produce hydroxyl free radicals ( • OH) rapidly over a wide pH range and destroy most refractory VOC organic compounds [10]. This coupled ozone technology, in which • OH is deliberately produced from ozone decomposition, is called an advanced oxidation process (AOP) [7,10,38]. AOPs are identified as chemical processes that accelerate the decomposition of ozone to yield sufficient quantities of the potent hydroxyl free radical to cause specific improvements in the water treatment process. These processes produce peroxone, which is the result of the reaction between hydrogen peroxide and ozone [39–42], ultraviolet radiation (UV) with ozone, elevated pH with ozone, and UV with hydrogen peroxide. The AOPs enable water producers to achieve definitive water quality goals at lower costs and greater reliability than conventional oxidation methods [11]. As a highly reactive gas, ozone quickly corrodes most metals (e.g., iron, copper, and mild steel) and damages most plastics if used without consideration for ozone compatibility. Natural rubber exposed to ozone will harden quickly and crack. Gaskets, sealing compounds, and piping should be chosen with attention to ozone compatibility [7]. Generally, the main areas in which ozone is used are: •

Disinfection: pathogenic pollution (excluding parasitic organisms), bacterial disinfection, viral and cyst inactivation, and befouling control [7,37] • Oxidation of inorganic compounds: Fe2+ , Mn2+ , As3+ , organically bound heavy metals and cyanides [7,37] • Oxidation of organic compounds: alkenes, chlorinated compounds, chloroaromatics, aldehydes, alcohols, carbonic acids, aliphatics, some pesticides, polyaromatic hydrocarbons, some detergents, phenols, and trihalomethane (THM), including taste, odor, color removal, algae control, and particle removal (particles of 0.5–2 mg L−1 ), micro flocculation (of soluble organics), pretreatment of organics for biological oxidation and precursor control [7,37] • Treatment of wastewater. Pulp and paper mill (reduction of color, COD, absorbable organic halogens), municipal (disinfection, odor control, sludge reduction, and biodegradability), pesticides (elimination of recalcitrant compounds) [2,43–48] 13.2.4

Efficiency and Production

Ozone has been used to treat groundwater and surface water in preparation for the drinking supply or for use in swimming pools in Europe since 1906. By 1982, there were well over 2000 municipal water treatment plants and many more public swimming pools using ozone throughout the world. It is estimated that more than 3000 water

316

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

treatment plants use ozone today; more than 200 of these are in the US. The EPA’s water treatment regulations for primary disinfection and control of chlorination by-products encourage municipalities to use ozone for both purposes. Actually, a larger proportion of US municipal water treatment plants are expected to be using ozone for both disinfection by-product controls [7]. In the past, the attitudes in the United States and Mexico were that ozone was simply an alternative disinfectant to chlorine. Although ozone is the most effective and most rapidly acting primary disinfectant available, its residual disinfection property can be more stable in distribution systems. The most critical consideration for water quality is persistent contamination. Using ozone as a primary oxidizing agent (or as the primary disinfectant) does not require the addition of chemicals to the water, whereas chlorine, potassium permanganate, chlorine dioxide, and so on, do [7]. Recently, Americans have begun to realize that even though some European plants use ozone as the primary disinfectant after filtration in municipal water treatment, it is generally recommended that ozone be accompanied by small amounts of chlorine, chlorine dioxide, or monochloramine to produce a disinfectant residue throughout the water distribution system. Much of the disinfectant demands of the water are satisfied by ozonation steps. For this reason, only small amounts of chlorine-containing compounds are necessary to produce the required stable residue [7]. Ozone is increasingly recognized as being more of a chemical oxidant, in the early stages of water treatment (prefiltration), than simply a disinfectant. Ozone performs oxidation better than other available oxidants with the formation of fewer problematic by-products [7]. Most large-scale applications of ozone treatment in municipal water treatment plants can be transferred to small systems markets. This is because many of the same contaminants are present in both types of aqueous environments. All of the treatment techniques used in professional water treatment (i.e., filtration, ion exchange, activated carbon adsorption, or reverse osmosis) can be coupled beneficially with ozone to solve specific water treatment problems. In fact, incorporation of ozone can solve some of the problems resulting from the use of these other devices (e.g., befouling control of ion exchange resins and membranes, and disinfection following activated carbon adsorption) [7]. Within the last 15 years, the cost-effectiveness of ozone production systems has improved. Several technological advances have contributed to cost reduction [37]: •

Higher ozone yield per unit electrode area, due to medium frequency technologies Increased ozone concentration with modern ozone generators, from 6% to 14 wt % in oxygen • Increase in the unitary ozone production capacity by a factor of two or three • 40% reduction in specific energy consumption (since 1990) • Improved operational reliability •

Considering the investment and operation costs, ozonation still is not a cheap technology. Although safe operation is no longer a problem, ozonation systems require considerable safety precautions, increasing the investment costs. Especially in smaller applications, the investment and capital costs cannot be neglected, because they can considerably lengthen the time required for investment return [37].

317

13.3 TECHNOLOGIES FOR PRODUCING OZONE

An ozonation system, from a customer’s perspective, must meet the following technical and economical requirements [37]: •

Achieve the treatment goal (e.g., reduce contamination below the legal limits). Use minimum oxygen. • Keep the energy costs low. • Safe to operate. •

13.3

TECHNOLOGIES FOR PRODUCING OZONE

Ozone is an unstable oxidizing gas with a maximum half-life in very clean water (double distilled) on the order of only a few hours, which is reduced when pollutants are present [7]. The methods of ozone generation and their differences are summarized in Table 13.4 [37]. The first two methods of ozone production, electrical discharge and electrolysis, are the only methods of practical importance, both in the laboratory and in full industrial-scale applications. Ozone production from ambient air or pure oxygen, in an electrical discharge chamber, is the most widespread technology for ozone generation. Considering that ozone is a threeatom molecule modification of molecular oxygen, its production directly from oxygen seems reasonable. In recent years, electrochemical ozone production by electrolysis of water has gained some attention in certain applications [37]. Ozone is made by rupturing the stable oxygen molecule, forming two oxygen fragments that can combine with oxygen molecules to form ozone: O2 ↔ 2[O]

(13.7)

2[O] + 2O2 ↔ 2O3

(13.8)

Nature generates ozone continuously by means of sunlight acting on oxygen in the atmosphere, or intermittently by lightning passing though the air. Humans simulate this natural process of generating ozone by passing high-voltage electrical discharges, high or low electrical frequencies, or high-energy radiation through air or oxygen. Ozone is also generated unintentionally by humans as a by-product during electrical power generation, electrostatic precipitators, welding equipment, electrostatic copying machines, UV lights, and a variety of other electrical devices. Generation of ozone is energy intensive, with some 90% of the power supplied to the generator being utilized to produce light, sound, and primarily heat, rather than the desired ozone. Thus, minimizing electrical power requirements in future technological advances for ozone production is a prime target of ozone generator manufacturers. 13.3.1

Corona Discharge Technique

For the generation of ozone by the corona discharge technique, the type, thinness, and surface area of the dielectric medium used, width of the discharge gap between electrodes, degree of flawlessness of the dielectric medium (no pin-holes), pressure, temperature, and

318

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

TABLE 13.4 Overview of the types of ozone generation, working principles, and fields of application [7–8,37]. Method of ozone generation

Working principle

Ozone source

Electrical

Electrical discharge (ED)

Electrochemical

Electrolysis (EL)

Photochemical (λ = 185 nm)

Irradiation (abstraction of electrons) X-rays, radioactive γ -rays

Radiation chemistry

Thermal

Light arc ionization

Field of application

Air or O2

Standard method, from laboratory to large-scale. Uses a 10–20 kV potential with larger performance ranges, >100 g O3 /h Predominately for pure water Water without organic applications, laboratory to compounds or small industrial scale, with an suspended particles, ozone performance of with a conductivity of ∼15 mg O3 /h with an energy <20 μS cm−1 of ∼12 Wh g−1 O2 (air), water (drinking Generation of [O3 ] < 1 g m−3 , water quality or highly with high energy use, around purified) 3 kW h g−1 Water (highly purified) Very seldom, solely experimental. Use of radioactive isotopes, such as 137 Cs, 60 Co, and 90 Sr. Ozone yields of ∼82 kW h g−1 Water Very seldom, solely experimental. Ozone yield ∼5 kW h g−1

rate of flow of feed gas through the ozone generator, composition (air versus oxygen), and moisture content of the feed gas are among the most important factors [2,49]. Because ozone is only slightly soluble in aqueous media, contact between ozone and water requires the bubbling of ozone mixed with air or oxygen. Mass transfer of ozone from the gaseous bubbles occurs across the gas/liquid interface into the water [23]. Factors that affect the mass transfer of ozone into liquids, and that themselves are affected by the design and operation of the contactor systems include miscibility with water, ozone demand of the substances to be ozonized, concentration of ozone in the gas, whether the carrier gas is air or oxygen, method and duration of contact, bubble size, pressure, and temperature. In designing an ozone contacting system, it is important to minimize the amount of ozone required for the specific purpose for which the ozone is to be used. There are only four different types of gas/liquid contacting systems: spray towers (liquid sprayed into gas), packed beds, bubble plate or sieve towers (an intermediate situation between the last), and units for dispersing gas bubbles into liquid [49,50]. The main drawback associated with this technology is that the limited current efficiency leads to low O3 concentrations in the gaseous phase (O2 + O3 ), thus restricting O3 application [46,51]. Such a process is represented as follows: O • + O2 + M → O3 + M ∗

(13.9)

where M can be the reactor wall, a nitrogen molecule, or a molecule of carbon dioxide. M acts by removing the energy excess acquired during ozone molecule formation, resulting in an excited inert body, M* [2].

13.3 TECHNOLOGIES FOR PRODUCING OZONE

13.3.2

319

Electrical Discharge Ozone Generators (EDOGs)

Electrical discharge, sometimes also called silent discharge ozone generators, ionizes molecular oxygen by applying high-power alternating current to the gas. Air or pure oxygen can be used as a feed gas, at either ambient or elevated pressures (Pabs = 100–600 katm). Ozone is formed by recombination of ionized oxygen atoms and ionized molecular oxygen. In this process, only 4–12% of the energy supplied is used for the formation of ozone, and the rest is transformed into heat. An efficient cooling system must be installed, because ozone decays quickly at elevated temperatures, with T = 50◦ C regarded as the critical value. In full-scale systems, the gas is cooled to T = 5–10◦ C. Newly developed techniques, such as double-sided cooling of the electrodes, help increase energy efficiency [37]. Various types of discharge chambers are available with plate or tubular geometries. The classical and most frequently used type is the tubular, often also called van der Made-type or Welsbach-type. The central rod electrode technology is a more recent development, especially designed for applications with pure oxygen as the feed gas [37]. 13.3.3

Electrolytic Ozone Generators (ELOGs)

In an electrolytic ozone generator (ELOG), ozone is produced from the electrolysis of high purity water. In the electrolytic cell, water is broken down into molecular hydrogen (H2 ), oxygen (O2 ), and ozone (O3 ) by the action of electrons supplied by the catalytic properties of the electrode material. Water, O2 , and O3 leave the cell on the anode side, whereas H2 is produced at the cathode side of the specialized electrolytic cell [2,23,37,46,52–54]. The cell anode and cathode spaces are separated by a solid electrolyte membrane. The anode is made of a porous, water-penetrable, and current-conductive carrier material coated with an active layer. However, the active catalytic layer and the electrolyte membrane are present, ozone is produced when a direct current of 3 to 6 V at currents of up to 50 A (corresponding to a current intensity of 0.2 to 3.0 A cm−2 ) are applied [37]. Because all materials of the cell must be electrochemically stable and must provide high conductivities, the cell is constructed with refined metals or metal oxides at their highest oxidation levels. The feed water must be of high purity, because it must pass through the porous anode and cathode materials without clogging or causing chemical damage. Therefore, ions and other impurities, present in normal drinking water, must be removed by ion-exchange, ultra or nanofiltration, reverse osmosis, or distillation [37]. Electrolytic ozone generators are supplied by several producers. The ozone production capacity of one cell is 1–4 g O3 h−1 , but several cells can be combined in one generator. Cell temperature strongly influences ozone production, and efficient cooling is necessary [37]. Relevant reduction half-reactions and their corresponding standard potentials are [12,29]: Anode reaction: ◦

3H2 O ↔ O3 + 6H+ + 6e−

E = 1.51V vs. NHE

2H2 O ↔ O2 + 4H+ + 4e−

E = 1.23V vs. NHE

◦

(13.10) (13.11)

320

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

Cathode reaction: 2H+ + 2e− → H2

(13.12)

It is apparent from these E ◦ values that anodic generation of O2 is thermodynamically favored over generation of O3 . Strategies for increasing the current efficiency for O3 generation generally include one or more of the following: 1. The electrode material should present good conductance and resistance to anodic corrosion and a high anodic overpotential for the oxygen evolution reaction, such as. PbO2 and glassy carbon (GC) [12,15,16,29,46,52–56,37,52]. 2. The morphological characteristics (roughness, porosity) of the electrode material significantly affect the current efficiency and require further investigation of the electrode preparation in order to optimize ELOG, thus reducing energy demands [2,16,23]. 3. Anions and cations from the electrolyte should not engage in competitive reactions with the oxygen and hydrogen evolution reactions or electrochemical ozone production. This can be possible using acidic electrolytes (e.g. H2 SO4 , HClO4 ) [15,57], which must be select in order of protect the electrode material [2,53,54]. 4. Electrochemical ozone production efficiencies can also be increased if special electrolytes are used, such as solid polymer electrolytes (SPEs) [14,26,29,40] or aqueous electrolytes, wh conducting salts containing fluoride-containing anions [12,15,23,52]. Addition of an adsorbate (e.g., F− , BF4 − , or PF6 − ) can help block the O2 evolution mechanism [12,15,16,46,52,55,56]. 5. Apply a large anodic current density to achieve a large positive electrode potential [2,53,54]. 6. Lower the temperature [2,16,53,54]. 7. Use a solid polymer electrolyte-based cell [13,16,40,58]. 8. Allow for diverse configurations [52]. Relative to the corona process, the ELOG process produces higher O3 concentrations, making the ELOG process more appropriate for situations in which a high O3 concentration is essential. However, the ELOG process is associated with higher costs of electrical energy consumption [16,26,52,59]. In addition to UV ozone generation, electrochemical production of ozone is a better alternative to corona discharge techniques [60]. Electrochemical ozone production has the advantage that it is a low DC voltage technology, and that ozone can be produced directly in water with a support electrolyte, thereby eliminating or minimizing technical problems associated with dissolving ozone in water. The widespread use of electrochemical ozone production requires an increase in current efficiency, simplified production systems, stable electrode materials, and/or easy handling of the electrolytes [61]. 13.3.3.1 Anodes for Electrochemically Producing Ozone The drastic operation conditions required for ELOG (very low interfacial pH and high anodic potentials) impose stringent specifications on the electrode materials [59]. In addition, the ELOG

13.3 TECHNOLOGIES FOR PRODUCING OZONE

321

technology’s water electrolysis produces oxygen as the main competitive reaction to ozone production. Thermodynamically, oxygen evolution is strongly favored over ozone production [61]. Therefore, high current efficiencies for electrochemical ozone production are possible only if high oxygen overvoltage anode materials are used. Examples of such materials that have been investigated for ozone production are Pt [2,19,46,48,52,62,63]; Ti/Pt [5]; Ti/IrO2 [13]; PbO2 [2,12–15,20,23,26–29,40,46–48,57,64–69]; PbO2 /Ebonex (Ebonex is a commercial material based on titanium suboxides) [2,46];-Fe-F-PbO2 [70]; Ti4 O7 /PbO2 [71]; Ti5 O9 /PbO2 [71]; Ti/RuO2 [52,72]; SnO2 [52,72–74]; Sb(SnO2 ) [74]; Ni(SnO2 ) [74]; PbO2 /Ti [12,20]; Fe-PbO2 /Ti [12]; Ti/[IrO2 -Nb2 O5 ] [46,75–77]; TaO2 [77]; Ti(Pt-TaO2 /Ti) [77,78]; Ti/(IrO2 + Ta2 O5 ) [46]; or glassy carbon [2,20,46,52]. Electrochemical ozone production efficiencies can also be increased if special electrolytes are used—namely SPEs or aqueous electrolytes with fluoride-containing conducting salts [16,75]. Using PbO2 at the high anodic potentials involved, the same oxygen species may be involved in the formation of O3 in addition to O2 , as illustrated in the pathway described below [2,16,20,23,48,79–85]: Electrochemical Steps: Kinetic Control: H2 O → H+ + ( • OH)ads + e−

(13.13)

( • OH)ads → (O • )ads + H+ + e−

(13.14)

Chemical Steps: Efficiency Control: (O • )ads → [1 − θ ](O • )ads + θ (O • )∗ads

(13.15)

(0 < θ <1)

[1 − θ ](2O )ads → [1 − θ ](O2 )ads •

[1 − θ ](O2 )ads → [1 − β][1 − θ ](O2 )ads + β[1 − θ ](O2 )∗ads

(13.16) (0 < β<1) (13.17)

Oxygen Evolution: 2( • OH)ads → O2 + 2H+ + 2e−

(13.18)

2(O • )ads → (O2 )ads → O2

(13.19)

[1 − β][1 − θ ](O2 )ads → O2 ↑

(13.20)

Ozone Formation: (O • )ads + (O2 )ads → (O3 )ads → O3

(13.21)

θ (O • ) ∗ads +β[1 − θ ](O2 )∗ads → [θ + β[1 − θ ]](O3 )ads

(13.22)

[θ + β[1 − θ ]](O3 )ads → O3 ↑

(13.23)

where θ and β indicate the partial surface coverages describing what characterizes the competition between oxygen evolution reaction and electrochemical ozone production processes. respectively; whereas “*” represents the surface coverage by oxygenated species leading that yield to O3 —formation.

322

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

Current efficiencies for O3 generation are significantly larger at Fe (III)-doped PbO2 film electrodes as compared to the current efficiencies of pure β-PbO2 , due to the apparent electrocatalytic effect of the Fe (III) sites in Fe-PbO2 anodes. The oxygen from the hydroxyl species on Pb (IV) sites is transferred to the O2 molecule adsorbed at an adjacent Fe (IIII) sites to produce O3 , as [12]: H2 O → ( • OH)ads + H+ + e−

(13.24)

(O2 )ads + ( • OH)ads → O3 + H+ + e−

(13.25)

or [20]: H2 O → • OH + H+ + e−

(13.26)

( • OH) → (O)ads + H+ + e−

(13.27)

2(OH)ads → (O2 )ads → (O2 )

(13.28)

(O)ads + (O2 )ads → O3

(13.29)

In the specific case of Ti/[IrO2 − Nb2 O5 ] or Ti/(IrO2 + Ta2 O5 ), Santana, De Faria, and Boodts [77] and Da Silva et al. [46] propose that the ozone is formed in the high overpotential region (j ≥ 0.4 A cm−2 ). In this case, O3 formation strongly depends on an effective encounter between O2 and O • , which, in turn, depends on the partial surface concentration of the oxygen intermediates, as: Kinetic Control Steps: ≡ S + H2 O →≡ S −∗ OH + H+ + e−

(13.30)

≡ S −∗ OH →≡ S − OH

(13.31) +

≡ S − OH →≡ S − O + H + e

−

2 ≡ S − OH → 2 ≡ S + O2 + 2H+ + 2e−

(13.32) (13.33)

Efficiency Control Steps: ≡ S − O → [1 − θ ] ≡ S − O + θ ≡ S∗ − O

(13.34)

[1 − θ ]2 ≡ S − O → [1 − θ ]2 ≡ S − O2

(13.35)

[1 − θ ]2 ≡ S − O2 → [1 − β][1 − θ ]2 ≡ S − O2 + β[1 − θ ]2 ≡ S∗ − O2 (13.36) Oxygen Evolution: [1 − β][1 − θ ]2 ≡ S − O2 → O2 ↑

(13.37)

Ozone Formation: ≡ S − O + O2 →≡ S + O3

(13.38)

θ ≡ S ∗ −O + β[1 − θ ]2 ≡ S ∗ −O2 → [θ + β(1 − θ )]3 ≡ S − O3

(13.39)

[θ + β(1 − θ )]3 ≡ S − O3 → O3 ↑

(13.40)

13.3 TECHNOLOGIES FOR PRODUCING OZONE

323

where S is an active surface site, θ and β indicate the partial surface coverages describing the competition between oxygen evolution and electrochemical ozone production, and the asterisk represents the surface coverage by oxygenated species that ultimately form O3 . One of the major technical limitations for the electrochemical generation of ozone is the anode material. For example, •

PbO2 anodes are currently used because the material is cheap and it is relatively stable under the high positive potentials required [20], with relatively high current efficiencies [86]. It must be protected from chemical reduction by applying an external current [87]. • Glassy carbon is used in a very limited range of materials suitable for ozone evolution [20]. It is easily corroded in acid solution [87]. • Pt is too costly for practical applications and produces high current efficiencies only at very low temperatures [52,88]. It is consumed due to a high rate of wear [87]. • Dimensional stable anode (DSA) materials have not proven to be stable for ozone evolution and have shown very poor current efficiencies [52,72]. On the basis of these comparisons, BDD is expected to be useful for ozone generation in which a high product purity would be required [87]. In the last several years, new doped diamond electrodes have been developed and thoroughly investigated. Several studies have attempted to use diamond as an electrochemical electrode for wastewater treatment and electroanalysis. It was discovered that diamond possesses a wider electrochemical window and a lower background current than other electrodes [46,87–94]. Aside from these interesting properties, diamond is distinguished by its exceptionally high overvoltage for oxygen evolution in aqueous electrolytes, which makes it an even more highly efficient • OH radical producer. 13.3.3.2 Boron-Doped Diamond (BDD) Diamond is a wide bandgap p-type semiconductor [95,96] with a unique and extraordinary combination of electronic properties (high breakdown voltage [97,98], electron/hole mobility [99], and carrier saturation velocity [100]); mechanical properties (high hardness and strength); chemical inertness; stability; and high thermal conductivity [101–104]. In contrast to natural diamonds, synthetic diamonds contain inclusions of a metalcarbon solid solution that originate from the growth environment. The number of inclusions can be considerably reduced by decreasing the crystal growth rate, but this yields several technical problems [95]. During synthesis, the structure of diamond crystals can be doped with boron in the C–H–B system. The results demonstrate that the addition of boron to the binary system C–H (naphthalene) notably reduces the pressure and temperature required for diamond synthesis. One possible reason for this observation is that boron-doping forms structurally imperfect graphite as an intermediate step to diamond synthesis [96]. Microwave plasma chemical vapor deposition (MPCVD) [105–107,113] and hot filament chemical vapor deposition (HFCVD) [96,108,109,112] are two of the most extensively used techniques for depositing diamond films [102]. High-pressure high-temperature (HPHT) technologies are used as well [93,104,110–113]. Boron-doped diamond presents a very high overpotential for both oxygen and hydrogen evolution in acidic conditions. This electrode leads to a wide potential

324

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

window (approximately 3.5 V) that can be used for other electrochemical reactions in the presence of aqueous electrolytes [93]. Diamond electrodes, indeed, have the largest potential window measured thus far in aqueous electrolytes. This differentiates them from common electrode materials, such as gold, platinum, or mixed metal oxide DSA-type electrodes [114]. For example, the oxygen evolution potential on BDD is 2.4 V in 0.2 M H2 S O4 [114,115]. This is much higher than 1.5 V for Pt, which demonstrates that BDD is more beneficial for the direct and indirect oxidation of organic compounds than Pt, and the reaction efficiency on BDD is higher than for Pt [115]. The width of the window decreases with the quality of the film and the incorporation of nondiamond sp2 carbon impurities, and its response resembles that obtained from glassy carbon and highly oriented pyrolytic graphite [93,96,116–126]. The large oxygen overpotential of BDD in water is due to the fact that BDD is a nonactive electrode material (it not oxidizable, similar to SnO2 or PbO2 ). Formation of hydroxyl radicals on the surface of some electrode materials has been demonstrated by in situ studies involving spin-trapping reagents. One obvious difference between BDD and these other nonactive electrode materials is that BDD is not a highly oxidized form of carbon [127]. Boron-doped diamond electrodes are considerably more effective at oxidizing organic compounds [114] into smaller fragments and CO2 than conventional material substrates such as Pt, Au, or graphite. BDD can also oxidize inorganic compounds, such as carbonates or sulfates into percarbonate and persulfate [126,128]. Hydrodynamic modulation experiments support the effectiveness of BDD over other materials. In the presence of noble metals, oxygen evolution is in strong competition with oxidation of organic compounds, resulting in the observed poor current efficiencies for the oxidation of the organics. Graphite is itself attacked under these conditions. These studies complement the larger-scale demonstrations of organic incineration at BDD and offer a more controlled study of the reaction stages [93,122–128]. Diamond electrodes are distinguished from conventional electrode materials in several other ways as well. They show very low double-layer capacitance and background current. Surface oxide formation or reduction reactions are also absent when using diamond electrodes, whereas these reactions are prevalent among conventional metal or metal oxide electrode materials. Diamond is an inert surface with low adsorption properties [93]. The diamond–electrolyte interface is ideally polarizable, and the current between—1000 and +1000 mV versus SCE is <50(A cm−2 ). The double-layer capacitance is up to one order of magnitude lower than that of glassy carbon [93,129]. Boron-doped diamond is an inert substrate due to its weak adsorption properties. Consequently, • OH only slightly adsorbs onto the surface, and BDD is very reactive toward organic oxidation. In general, organic compounds can be completely mineralized, (i.e., converted into CO2 , water, and inorganic ions) by this anode with a high current efficiency [130]. The BDD shows remarkable corrosion stability in very aggressive media. The morphology of diamond electrodes is stable during long-term cycling between hydrogen and oxygen evolution, even in acidic fluoride media [93,130–132]. BDD has been widely investigated because it is an inert electrode and presents weak interactions between its surface and hydroxyl radicals. The mechanism by which the organic substrates are oxidized at this electrode surface has been examined. BDD has low adsorption properties and a strong tendency to resist deactivation [93,133]. Hydroxyl radicals are generated by water oxidation and can competitively evolve toward oxygen

13.4 REACTION MECHANISM FOR THE PRODUCTION OF OZONE

325

evolution. At high current densities, the oxidation process shifts from charge transfer control to mass transfer control [134]. The BDD anode is included in advanced oxidation processes (AOPs), which are defined as processes that “involve the generation of hydroxyl radicals in quantities sufficient to achieve water purification” [37,115,135–137].

13.4 REACTION MECHANISM FOR THE PRODUCTION OF OZONE WITH BORON-DOPED DIAMOND Comparing the hydrogen-adsorption BDD film and the oxygen-adsorption BDD film electrodes, discussed in the previous paragraph, BDD electrodes possess a wide electrochemical window, a larger diamond film resistance and capacitance, and a larger polarization resistance than BDD electrodes. Due to BDD’s wide electrochemical window, exceptionally high overpotential for oxygen evolution, and high efficiency for HO • production, much research has been devoted to elucidating the mechanism for generation of ozone at BDD electrodes. The rate of O3 generation is highly dependent on the applied current density and increases with increasing current densities. The applied current is lower than the limiting current density at high mass transfer rates, which maintains the generation of O3 molecules on the anode [93]. The mechanism for O3 production, proposed by several authors [75,87,93,137–141] is the following: H2 O → • OH + H+ + e− •

+

(13.41)

−

OH → O + H + e •

(13.42)

2O → O2 •

(13.43)

2 • OH → H2 O2

(13.44) +

H2 O2 → O2 + 2H + 2e

−

O2 + O • → O3

(13.45) (13.46)

where the general reaction can be expressed as [87,93,94,137]: 3H2 O → O3 + 6e− + 6H+

(13.47)

This reaction proceeds in competition with the following reaction: 2H2 O → H2 O2 + 2e− + 2H+ −

2H2 O → O2 + 4e + 4H

+

(13.48) (13.49)

At the cathode, hydrogen is generated simultaneously: 2H+ + 2e− → H2

(13.50)

A rate comparison of ozone gas production at BDD and Pt electrodes, by electrolysis of a 1 M H2 SO4 solution, demonstrates that BDD exceeds Pt in efficiency by more than

326

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

BDD H2O 1

O2

2 O2

+ H+ + e–

O + H+ + e–

H+ + e–

O3

BDD(OH•)

1

2 CO2

BDD(H2O2)el

(H2O2)sol

+ H+ + e–

Figure 13.2 Scheme of the proposed mechanism for water oxidation on BDD electrodes in an acidic solution containing a non-electroactive supporting electrolyte (1 M HClO4 ). (Reprinted with permission from Ref. 137.)

a factor of two at each applied density current condition down to 2 A cm−2 , and the electrolytic water decomposition reactions proceed as follows [137]: ◦

2H2 O → O2 + 4H+ + 4e−

E = +1.23V ◦

3H2 O → O3 + 6H+ + 6e− +

H2 O + O2 → O3 + 2H + 2e

E = +1.51V −

◦

E = +2.07V

(13.51) (13.52) (13.53)

Ozone concentrations in the gas phase increase almost linearly from 110 to 800 ppm, for applied current densities in the range of 23–150 mA cm−2 in 1 M HClO4 . Higher O3 concentrations, up to 10,000 ppm, were also obtained during electrolysis in 1 M H2SO4 with a much higher current density. The efficiency limit arises from the inert surface of BDD, which does not favor adsorption of the intermediate oxygenated species produced during ozone evolution [96]. The generation of ozone is better at temperatures between −5 and 0◦ C, but when the temperature is increased beyond 30◦ C, production decreases [96]. The mechanism proposed in the literature is shown in Figure 13.2 [137].

13.5

CONCLUSIONS

O3 is a reactive substance. Its concentration levels at any instant in time are affected by the aqueous conditions, including pH, alkalinity, and demand for ozone by oxidizable substances present. For these reasons, ozone must be manufactured on site for immediate use because it is unstable and quickly decomposes to molecular oxygen. An ozonation system, from a customer’s viewpoint, must meet the following technical and economic requirements: achieve the treatment goal, use a minimum quantity of oxygen, keep the energy cost low, and be safe to operate.

REFERENCES

327

The most commonly used technologies for the production of ozone are the corona discharge technique, the electrical discharge or double-discharge ozone generators, and the electrolytic ozone generators, in which the boron-doped diamond represents a new electrode material with better characteristics for generating O3 . Boron-doped diamond is a good electrode material for the production of O3 by electrochemical processes because it has remarkable corrosion stability in very aggressive media; stability during long-term cycling between hydrogen and oxygen evolution, even in acidic fluoride media; low adsorption properties; a strong tendency to resist deactivation; and it allows control by charge and mass transfer. Although BDD is a promising candidate material for ozone production electrodes, it can be improved by optimizing the adsorption of oxygenated species to increase the production of ozone.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

C.F. Sch¨onbein, Comptes Rendus Hebd. Seances Acad. Sci . 1840, 10 , 706. L.M. Da Silva, M.H.P. Santana, Qu´ım. Nova, 2003, 6 , 880. M.P. De la Rive, Comptes Rendus Hebd. Seances Acad. Sci . 1845, 20 , 591. P.S. Bailey, Ozonation in Organic Chemistry, New York, Academic Press, 1978, p. 8. E.R. Kotz, S. Tucki, J. Electroanal. Chem. 1987, 228 , 407. Technology for Productivity: Ozone Data & Conversion Tables. http://www.mksinst.com. P. Lukes, M. Clupek, V. Babicky, P. Sunka, Czech. J. Phys. 2004, 54 , Suppl. C . C908. F. J. Rodr´ıguez Vidal. Processes de Potabilizaci´on del Agua e Influencia del Tratamiento de Ozonizaci´on. Ediciones D´ıaz de Santos, S. A. Madrid, 2003. S.J. Masten, S.H.R. Davies, Environ. Sci. Technol. 1994, 4 , 180A. G. Kiely, Ingenier´ıa Ambiental: Fundamentos, Entornos, Tecnolog´ıas y Sistemas de Gesti´on. McGraw–Hill/Interamericana de Espa˜na, S. A. Madrid, 1999. C. White, Handbook of Chlorination and Alternative Disinfectants, 4th ed. New York, John Wiley & Sons, 1999. J. Feng, D.C. Johnson, S.N. Lowery, J.J. Carey. J. Electrochem. Soc. 1994, 10 , 2708. Y. Beaufils, C. Comninellis, P. Bowen, Electrochem. Eng. Icheme Symp. Ser. 1999, 145 , 191. S. Stucki, G. Theis, R. K¨otz, H. Devantay, H. Chrsiten, J. Electrochem. Soc. 1985, 132 , 367. P.C. Foller, W. Tobias, J. Electrochem. Soc. 1982, 129 , 506. L.M. Da Silva, L.A. De Faria, J.F.C. Boodts, Electrochim. Acta 2003, 48 , 699. R. Amadelli, A. De Battisti, D.V. Girenko, S.V. Kovalyov, A.B. Velinchenko, Electrochim. Acta 2000, 46 , 341. A. Chernik, V.B. Drozdovich, I.M. Zharskii, Russ. J. Electrochem. 1997, 33 , 259. A. Babak, R. Amadelli, V.N. Fateev, Russ. J. Electrochem. 1998, 34 , 166. A. Babak, R. Amadelli, A. De Battisti, V.N. Fateev, Electrochim. Acta 1994, 39, 1597. R. Amadelli, L. Armelao, A.B. Velinchenko, N.V. Nikolenko, D.V. Girenko, S.V. Kovalyov, F.I. Danilov, Electrochim. Acta 1999, 45 , 713. L.M. Da Silva, D.V. Franco, L.A. De Faria, J.F.C. Boodts, Electrochim. Acta, 2004, 49 , 3977–3988. L.M. Da Silva, L.A. De Faria, J.F.C. Boodts, Pure Appl. Chem. 2001, 12 , 1871. M.X. Meng, J.S. Hsieh, Tappi J . 2000, 83 , 67. P.C. Foller, M.L. Goodwin, Ozone Sci. Eng. 1984, 6 , 29.

328

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

26. S. Stucki, H. Baumann, H.J. Christen, R. K¨otz, J. Appl. Electrochem. 1987, 17 , 773. 27. A. Chernik, V.B. Drozdovich, I.M. Zharskii, Russ. J. Electrochem. 1997, 33 , 259. 28. R.G. Rice, A. Netzer, eds., Handbook of Ozone Technology and Applications, Vol. 1, Ann Arbor Science, UK, 1982. 29. K. Onda, T. Ohba, H. Kusunoki, S. Takezawa, D. Sunakawa, T. Araki, J. Electrochem. Soc. 2005, 10 , D177. 30. K. Arihara, C. Terashima, A. Fujishima, J. Electrochem. Soc. 2007, 4 , E71. 31. J.N. Armor, Appl. Catal. A: Gen. 1999, 189 , 153. 32. U.V. Gunten, Water Res. 2003, 37 , 1469. 33. M. Peleg, Water Res. 1976, 10 , 361. 34. N. Kisimoto, Y. Morita, H. Tsuno, T. Oomura, H. Mizutani, Water Res. 2005, 39 , 4661. 35. P. Ca˜nizares, R. Paz, C. S´aez, M.A. Rodrigo, Electrochim. Acta 2008, 53 , 2144. 36. K. Rajeshwar, J. Iba˜nez, Environmental Electrochemistry: Fundamentals and Applications in Pollution Abatement ,. New York,Academic Press, 1997. 37. C. Gottschalk, J.A. Libra, A. Saupe, Ozonation of Water and Waste Water: A Practical Guide to Understanding Ozone and Its Application. Wiley–VCH, Weinheim 2000. 38. S.J. Masten, S.H.R. Davies, Environ. Sci. Technol. 1994, 4 , 180A. 39. C.D. Adams, P.A. Scanlan, N.D. Secrist, Environ. Sci. Technol. 1994, 28 , 1812. 40. P. Tatapudi, J.M. Fenton, J. Electrochem. Soc., 1994, 141 , 1174. 41. J. Murphy, G.D. Hitchens, US005972196A, 1999. 42. M.E.H. Bergmann, J. Rollin, T. Iourtchouk, Electrochim. Acta, 2009, 54 , 2102. 43. M.B. Rubin, Bull. Hist. Chem. 2001, 26 , 40. 44. T. Bechtold, A. Turcanu, W. Schrott, Diam. Rel. Mat . 2006, 15 , 1513. 45. S.J. Masten, S.H.R. Davies, Environ. Sci. Technol. 1994, 4 , 180A. 46. L.M. Da Silva, D.V. Franco, L.A. De Faria, J.F.C. Boodts, Electrochim. Acta, 2004, 49 , 3977. 47. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191. 48. K. Arihara, C. Terashima, A. Fujishima, J. Electrochem. Soc. 2007, 4 , E71. 49. R.G. Rice, M.E. Browning, Ozone Treatment of Industrial Wastewater. New Jersey, Noyes Data Corporation, 1981. 50. L. Chia-Chang, C. Cheng-Yu, L. Mei-Yun, L. Ya-Ling, J. Hazard. Mater. 2009, 167 , 1014. 51. F.W. Pontius, Water Quality and Treatment , 4th ed., New York, Mc Graw-Hill, 1993. 52. P.C. Foller, G.H. Kelsall, J. Appl. Electrochem. 1993, 23 , 996. 53. P. Tatapudi, J.M. Fenton, in Environmental Oriented Electrochemistry ( C.A.C. Sequeira, ed.), Amsterdam, 1994. 54. W.T. Shin, A. Mirmiran, S. Yiacoumi, C. Tsouris, Sep. Purif. Technol. 1999, 15 , 271. 55. J.C.G. Thanos, H.P. Fritz, D. Wabner, J. Appl. Electrochem. 1984, 14 , 389. 56. J.E. Graves, D. Pletcher, R.L. Clarke, F.C. Walsh, J. Appl. Electrochem. 1992, 22 , 200. 57. P.C. Foller, C.W. Tobias, J. Electrochem. Soc. 1982, 3 , 506. 58. A. Babak, V.N. Fateev, R. Amadelli, G.F. Potapova, Elecktrokhimiya, 1994, 6 , 814. 59. M.H.P. Santana, L.A. De Faria, J.F.C. Boodts, Electrochim. Acta 2004, 49 , 1925. 60. M.B. Rubin, Bull. Hist. Chem. 2001, 26 , 40. 61. A. Kraft, M. Stadelmann, M. W¨unsche, M. Blaschke, Electrochem. Commun. 2006, 8 , 883. 62. J.D. Seader, C. Tobias, Ind. Eng. Chem. 1952, 9 , 2207. 63. T.R. Beck, R.W. Moulton, J. Electrochem. Soc. 1956, 4 , 247. 64. P. Tatapudi, J.M. Fenton, J. Electrochem. Soc. 1993, 140 , 3527.

REFERENCES

329

65. R. Amadelli, L. Armelao, A.B. Velichenko, N.V. Nikolenko, D.V. Girenko, S.V. Kovalyov, F.I. Danilov, Electrochim. Acta 1999, 7 , 13. 66. I. Sir´es, E. Brillas, G. Cerisola, M. Panizza, J. Electroanal. Chem. 2008, 613 , 151. 67. N. Kisimoto, Y. Morita, H. Tsuno, T. Oomura, H. Mizutani, Water Res. 2005, 39 , 4661. 68. C. Flox, C. Arias, E. Brillas, A. Savall, K. Groenen–Serrano, Chemosphere 2009, 74 , 1340. 69. A. Baak, V.N. Fateev, R. Amadelli, G.F. Potapova, Elektrokhimiya, 1994, 6 , 814. 70. A.B. Velichenko, D.V. Girenko, S.V. Kovalyov, A.N. Gnatenko, R. Amadelli, F.I. Danilov, J. Electroanal. Chem. 1998, 454 , 203. 71. J.E. Graves, D. Pletcher, R.L. Clarke, F.C. Walsh, J. Appl. Electrochem. 1992, 22 , 200. 72. J.C. Hwang, Electrolytic Generation of Ozone, paper 592, 166th Meeting of the Electrochemical Society, New Orleans, 7–12 October, 1984. 73. S.–A. Cheng, K.–Y. Chan, Electrochem. Solid State Lett . 2004, 7 , D4. 74. Y.–H. Wang, S. Cheng, K.–Y. Chan, X. Y. Li, J. Electrochem. Soc. 2005, 152 , D197. 75. A. Kraft, M. Stadelmann, M. W¨unsche, Electrochem. Commun. 2006, 8 , 883. 76. C. Flox, C. Arias, E. Brillas, A. Savall, K. Groenen-Serrano, Chemosphere 2009, 74 , 1340. 77. M.H.P. Santana, L.A. De Faria, J.F.C. Boodts, Electrochim. Acta 2004, 49 , 1925. 78. M.I. Award, S. Sata, K. Kaneda, M. Ikematsu, T. Okajima, T. Ohsaka, Electrochem. Commun. 2006, 8 , 1263. 79. J. Ge, D.C. Johnson, J. Electrochem. Soc. 1995, 142 , 1525. 80. J. Feng, L.L. Houk, D.C. Johnson, S.N. Lowery, J.J. Carey, J. Electrochem. Soc. 1995, 142 , 3626. 81. J.E. Vitt, D.C. Johnson, J. Electrochem. Soc. 1992, 139 , 774. 82. J. Feng, D.C. Johnson, J. Electrochem. Soc. 1991, 138 , 3328. 83. L.A. Larew, J.S. Gordon, Y.L. Hsiao, D.C. Johnson, J. Electrochem. Soc. 1990, 137 , 3071. 84. H. Gerischer, A. Heller, J. Electrochem. Soc. 1992, 139 , 113. 85. R. Amadelli, A. De Battisti, D.V. Girenko, S.V. Kovalyov, A.B. Velichenko, Electrochim. Acta 2000, 46 , 341. 86. D.P. Semchenko, E.T. Lyubushkina, V. Lyubushkin, Electrokhimiya 1973, 11 , 1744. 87. N. Katsuki, E. Takahashi, M. Toyoda, T. Kurosu, M. Lida, S. Wakita, Y. Nishiki, T. Shimamune, J. Electrochem. Soc. 1998, 7 , 2358. 88. J.D. Seader, C.W. Tobias, Ind. Eng. Chem. 1952, 9 , 2207. 89. T. Kondo, Y. Einaga, B.V. Sarada, J. Electrochem. Soc. 2002, 149 , E179. 90. T.A. Ivandini, B.V. Sarada, C. Terashima, J. Electroanal. Chem. 2002, 1–2 , 117. 91. S. Ferro, A.D. Battisti, I. Dao, J. Electrochem. Soc. 2000, 7 , 2614. 92. W. Meifen, Z. Guohua, L. Mingfang, L. Lei, L. Dongming, J. Hazard. Mater. 2009, 163 , 26. 93. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191. 94. S.–G. Park, Russ. J. Electrochem. 2003, 3 , 321. 95. A.I. Chepurov, A.P. Yelisseyev, E.I. Zhimulev, V.M. Sonin, I.I. Fedorov, A.A. Chepurov, Inorg. Mater. 2008, 4 , 377. 96. E.A. Ekimov, V.A. Sidorov, A.V. Rakhmanina, N.N. Mel’nik, M.A. Timofeev, R.A. Sadykov, Inorg. Mater. 2006, 11 , 1198. 97. A.V. Bogdanov, I.M. Vikulin, Sov. Phys. Semicond . 1982, 16 , 720. 98. E.A. Konoirova, Sov. Phys. Semicond . 1983, 17 , 146. 99. V.K. Bazhenov, I.M. Vikulin, Sov. Phys. Semicond . 1985, 19 , 829. 100. L. Reggiani, S. Bosi, Solid State Commun. 1979, 30 , 333. 101. Y.S. Touloukian, Thermophysical Properties of Matter, New York, Plenum, 12, 1970.

330

OZONE GENERATION USING BORON-DOPED DIAMOND ELECTRODES

102. X. Jisheng, M. Yaowu, X. Yiben, J. Jianhua, W. Hong. J. Shanghai University, 1997, 2 , 159. 103. E.Y. Zarechnaya, E.I. Isaev, S.I. Simak, Y.K. Vekilov, L.S. Dubrovinsky, N.A. Dubrovinskaia, I.A. Abrikosov, J. Experim. Theor. Phys. 2008, 4 , 781. 104. Y. Haitao, Y. Haisue, R.B. Jackman, Semicond. Sci. Technol. 2005, 20 , 296. 105. Y. Takano, M. Nagao, I. Sakaguchi, Appl. Phys. Lett . 2004, 85 , 2851. 106. V.A. Sidorov, E.A. Ekimov, E.D. Bauer, Diam. Rel. Mat . 2005, 14 , 335. 107. S.G. Buga, V.D. Blank, S.A. Terent’ev, M.S. Kuznetsov, S.A. Nosukhin, V.A. Kulbachinskii, A.V. Krechetov, V.G. Kytin, G.A. Kytin, J. Experim. Theor. Phys. 2007, 4 , 586. 108. N.G. Ferreira, L.L. G. Silva, E.J. Corat, Diam. Rel. Mat . 2002, 11 , 1523. 109. S. Ferro, A.D. Battisti, Electrochim. Acta 2002, 47 , 1641. 110. J.B. Zang, Y.H. Wang, H. Huang, W.Q. Liu, J. Appl. Electrochem. 2009. DOI 10.1007/ s10800-009-9838-x. 111. I. Tr¨oster, M. Fryda, D. Hermann, L. Scha¨afer, W. H¨anni, A. Perret, M. Blaschke, A. Kraft, M. Stadelmann. Diam. Rel. Mat . 2002, 11 , 640. 112. O. Scialdone, A. Galia, G. Filardo, Electrochim. Acta 2008, 53 , 7220. 113. A. Cieciwa, R. W¨uthrich, C. Comninellis, Electrochem. Commun. 2006, 8 , 375. 114. A. Kraft, Int. J. Electrochem. Sci . 2007, 2 , 355. 115. G. Zhao, S. Shen, M. Li, M. Wu, T. Cao, D. Li, Chemosphere 2008, 73 , 1407. 116. G.M. Swain, A.B. Anderson, J.C. Angus, MRS Bull . 1998, 23 , 56. 117. H.B. Martin, A. Argoitia, U. Landau, A.B. Anderson, J.C. Angus, J. Electrochem. Soc. 1996, 143 , 133. 118. G.L. Lin, L.I. Tzeng, C.L. Chen, Thin Solid Films 1997, 301 , 126. 119. M. Nishitani-Gamo, T. Tachibana, K. Kobashi, Diam. Rel. Mat . 1998, 7 , 783. 120. I. Pocsik, M. Hundhausen, M. Koos, L. Ley, J. Non-Cryst. Solids 1998, 227 , 1083. 121. V.A. Krivchenko, D.V. Lopaev, P.V. Minakov, V.G. Piragov, A.T. Rakhimov, N.V. Suetin, Technical Physics 2007, 52 , 1471. 122. B. Marselli, J. Garc´ıa–G´omez, P.–A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 3 , D79. 123. G. Sin´e, I. Duo, B. El Roustom, G. F´oti, C. Comninellis, J. Appl. Electrochem. 2006, 36 , 847. 124. B. El Rouston, G. Sin´e, G. F´oti, C. Comninellis, J. Appl. Electrochem. 2007, 37 , 1227. 125. I. Duo, C. Levy-Clement, A. Fujishima, C. Comninellis, J. Appl. Electrochem. 2004, 34 , 935. 126. M.A. Quiroz Alfaro, S. Ferro, C.A. Mart´ınez–Huitle, Y. Meas Vong, J. Braz. Chem. Soc. 2006, 2 , 227. 127. D. Sopchak, B. Miller, Y. Avyigal, R. Kalish, J. Electroanal. Chem. 2002, 538–539 , 39. 128. E.J. Ruiz, R. Ortega–Borges, J.L. Jurado, T.W. Chapman, Y. Meas, Electrochem. Solid-State Lett . 2009, 1 , E1. 129. G.M. Swain, R. Ramesham, Anal. Chem. 1993, 65 , 345. 130. I. Sir´es, E. Brillas, G. Cerisola, M. Panizza, J. Electroanal. Chem. 2008, 613 , 151. 131. G.M. Swain, J. Electrochem. Soc. 1994, 141 , 3382. 132. R. Ramesham, M.F. Rose, Diam. Rel. Mat . 1997, 6 , 17. 133. G.M. Swain, Adv. Mater. 1994, 5 , 388. 134. O. Scialdone, A. Galia, G. Filardo, Electrochim. Acta 2008, 53 , 7220. 135. G. Junxia, Z. Guohua, S. Wei, L. Dongming, Chemosphere, 2009, 75 , 519. 136. N. Kishimoto, T. Nakagawa, M. Asano, M. Abe, M. Yamada, Y. Ono, Water Res. 2008, 42 , 379.

REFERENCES

331

137. P.A. Michaud, M. Panizza, L. Ouattara, T. Diaco, G. Foti, C. Comninellis, J. Appl. Electrochem. 2003, 33 , 151. 138. J. Iniesta, P.A. Michaud, M. Panizza, G. Cerisola, A. Aldaz, C. Comninellis, Electrochim. Acta 2001, 46 , 3573. 139. G. F´oti, D. Gandini, C. Comninellis, A. Perret, W. Haenni, Electrochem. Solid State Lett . 1999, 2 , 228. 140. M. Polcaro, A. Vacca, M. Mascia, S. Palmas, R. Pompei, S. Laconi, Electrochim. Acta 2007, 52 , 2595. 141. M.E.H. Bergmann, J. Rollin, T. Iourtchouk, Electrochim. Acta, 2009 , 54 , 2102.

14 Application of Synthetic Diamond Films to Electro-Oxidation Processes Marco Panizza

14.1

INTRODUCTION

Electrochemical oxidation providing versatility, high-energy efficiency, and amenability to automation, environmental compatibility, and cost-effectiveness has reached a promising stage of development and can now be effectively used both in wastewater treatment for the destruction of toxic or biorefractory organics and in selective organic electrosynthesis [1–3]. Electro-oxidation of organic pollutants can be performed in different ways, including direct and indirect processes. In the direct electrolysis, the pollutants are oxidized after adsorption on the anode surface without involving any substances other than the electron. Direct electro-oxidation is theoretically possible at low potentials, before oxygen evolution, but the reaction rate usually has low kinetics. Moreover, the main problem of electro-oxidation at anodic potentials before oxygen evolution is a decrease in the catalytic activity, commonly called the poisoning effect, due to the formation of a polymer layer on the anode surface [4–8].

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

333

334

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

In the indirect oxidation, organic pollutants do not exchange electrons directly with the anode surface, but through the mediation of some electroactive species regenerated there, which act as an intermediary for shuttling electrons between the electrode and the organics. The oxidation mediators may be a reversible redox reagent turned over several times and recycled, such as some metallic or inorganic ions (e.g., Ag+ /Ag2+ , Co2+ /Co3+ , Ce3+ /Ce4+ , Cl− /ClO− , Br− /BrO− ) [9–12], or strong oxidizing chemicals (e.g., ozone, hydrogen peroxide) [13–18] generated in situ. The main drawbacks of the indirect electrolysis are the need to subsequently separate the oxidation products from the mediator or the possible formation of undesirable chlorinated or brominated by-products. Another mechanism for the electrochemical oxidation of organics at a high potential is based on intermediates of the oxygen evolution reaction [19–24]. This process involves the transfer of oxygen from H2 O to the organics via adsorbed hydroxyl radicals generated by the water discharge. It has been generally observed that the overall performances of the electro-oxidation processes are influenced by the choice of appropriate electrolysis conditions and, above all, by the selection of the electrode materials. Synthetic thin film boron-doped diamond (BDD) is a relatively new electrode material that has recently received great attention, thanks to its unique and superior properties that distinguish it from conventional electrodes, such as: •

An extremely wide potential window in aqueous and nonaqueous electrolytes: In the case of high-quality diamond, hydrogen evolution starts at about −1.25 V versus SHE and the oxygen evolution at +2.3 V versus SHE; therefore, the potential window may exceed 3 V [25]. • Corrosion stability in very aggressive media: The diamond electrode morphology is stable during long-term cycling from hydrogen to oxygen evolution, even in acidic fluoride media [26]. • Inert surface with low adsorption properties and a strong tendency to resist deactivation: The voltammetric response toward ferri/ferrocyanide is remarkably stable for up to two weeks of continuous potential cycling [27]. • Very low double-layer capacitance and background current: The diamond-electrolyte interface is ideally polarizable and the current between −1000 and +1000 mV versus SCE is < 50 μA cm−2 . Double-layer capacitance is one order of magnitude lower than that of glassy carbon [28]. Thanks to these properties, the BDD anodes during electrolysis in the region of water discharge promote the production of weakly adsorbed hydroxyl radicals [29]: BDD + H2 O → BDD( • OH) + H+ + e−

(14.1)

Depending on the experimental conditions, these electrogenerated hydroxyl radicals can be used in wastewater treatment for the nonselective and complete mineralization of pollutants or in electrosynthesis for the selective oxidation of organic compounds [30,31]: BDD( • OH) + R → BDD + CO2 + H2 O +

BDD( OH) + R → BDD + RO + H + e •

(14.2) −

(14.3)

14.2 APPLICATION IN WASTEWATER TREATMENT

335

The aim of this chapter is to elucidate the basic fundamentals of electro-oxidation of organics with a BDD anode and to discuss the recent progress dealing with the application of diamond electrodes in the electrochemical treatment of wastewater and in the organic electrosynthesis.

14.2

APPLICATION IN WASTEWATER TREATMENT

As the environment preservation gradually becomes a matter of major social concern and more strict legislation is being imposed on effluent discharge, more effective processes are required to deal with nonreadily biodegradable and toxic pollutants. There are different methods for the treatment of wastewater containing organic pollutants, including physicochemical processes (filtration, coagulation, adsorption, and flocculation), chemical oxidation (use of chlorine, ozone, hydrogen peroxide, wet air oxidation), and advanced oxidation processes (Fenton’s reaction, O3 /UV, photocatalysis). The choice of the treatment depends on economics as well as ease of control, reliability, and efficiency of the treatment. In this context, the electrochemical technologies have recently attracted a great deal of attention thanks to intensive investigations that have improved the catalytic activity and stability of electrode materials and optimized reactor geometry, and offer an alternative solution to many environmental problems in the process industry [3,32]. In the literature, several anode materials have been tested for the direct electrochemical oxidation of organic compounds, but the complete mineralization to CO2 and a good Faradic efficiency was only obtained using high oxygen overpotential anodes, such as SnO2 , PbO2 , and BDD. In fact, during the electrolysis, these electrodes involve the production of oxygen evolution intermediates, mainly hydroxyl radicals that oxidize the pollutants. Despite their notable ability to remove organics, the doped SnO2 anodes have the major drawback of a short service life, whereas the PbO2 anodes can release toxic lead ions during the electrolysis [33]. On the contrary, BDD shows a high anodic stability and wide potential window for water discharge and hence, it is undoubtedly a promising material for the complete combustion of organics during wastewater treatment. 14.2.1

Oxidation in the Potential Region before Oxygen Evolution

The activity of the BDD anode in the potential region before oxygen evolution has been studied by cyclic voltammetry and chronoamperometry in the presence of many model compounds such as carboxylic acids (e.g., acetic, formic, and oxalic acid) and aromatic and heterocyclic compounds (e.g., phenol, 4-chlorophenol, naphthol, 3-methylpyridine, naphthalene, and anthraquinone sulfonic acids). In the presence of the investigated carboxylic acids, the cyclic voltammograms display no significant differences from the voltammogram of the supporting electrolyte. The only difference is a slight decrease in the starting potential of oxygen evolution. These results indicate that the BDD electrode, which does not provide an active site for the adsorption of reactants, has no electrocatalytic activity in oxidation of the carboxylic acids; therefore, the oxidation of these compounds involves some intermediates formed during the oxygen evolution. The behavior of the BDD electrodes is quite different during the oxidation of aromatic and heterocyclic compounds. In fact, at least one anodic peak, corresponding to the

336

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

1.2

i (mA cm–2)

1

ip (mA cm–2)

1.4

0.8

0.6 0.5 0.4 0.3 0.2 0.1 0

Peak E = 1.1 V Peak E = 2.0 V

(f) y = 0.50 x R2 = 0.999

0

0.25

0.6

0.5 0.75 concentr./mM

(e) 1

(d) (c)

0.4 0.2

(b) (a)

A

0 0.6

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

E (V) vs. SHE Figure 14.1 Cyclic voltammetric (first scan) behavior of BDD at a scan rate of 100 mV s−1 in 1 M H2 SO4 with different 2-naphthol concentrations (mM): (a) 0; (b) 0.125; (c) 0.25; (d) 0.5; (e) 0.75; (f) 1. The dependence of the peak current density on the 2-naphthol concentration is shown in the inset. A: anodic start of the cyclic voltammograms. (Reprinted from Ref. 8.)

oxidation of the organic compound by direct electron transfer, is observed in the potential region of water stability. For example, Figure 14.1 shows cyclic voltammograms (first cycle) obtained with a scan rate of 100 mV s−1 in the potential region between +0.65 and +2.25 V (vs. SHE) for different 2-naphthol concentrations in 1 M H2 SO4 [8]. Three oxidative peaks corresponding to the oxidation of naphthol are observed at approximately +1.12 V, +1.65 V, and +2.0 V versus SHE. These peaks increase linearly with the naphthol concentration in the studied range of 0.125 mM −1 mM naphthol (Figure 14.1, inset: slope = 0.5 mA cm−2 mM−1 , R 2 = 0.999). In order to obtain more information about the oxidation mechanisms, several voltammetric measurements in the presence of 0.5 mM naphthol have been carried out in a wider potential range; these results are shown in Figure 14.2 [8]. During the first scan (curve a), a cathodic peak is obtained at about −0.2 V versus SHE besides the anodic peaks. During the second scan (curve b), an anodic peak is obtained at about +0.8 V along with the other peaks. The addition of 1,2-naphthoquinone into the electrolyte results in an increase of both the anodic (at +0.8 V) and cathodic (at −0.2 V) peaks current (Figure 14.2, curve c). This fact indicates that the two peaks at −0.2 and +0.8 V correspond to the 1,2-naphtoquinone/1,2-dihydroxynaphthalene couple. Figure 14.2 also shows that the separation of anodic and cathodic peak potential (Ep ) for the couple naphthoquinone/dihydroxynaphthalene is higher than the predicted value of about 29.5 mV for a two-electron transfer reaction, indicating an irreversible behavior and a slow electron transfer kinetics at the diamond surface. On the basis of these results, the following reaction mechanism is proposed for the anodic oxidation of 2-naphthol in the potential region of supporting electrolyte stability:

14.2 APPLICATION IN WASTEWATER TREATMENT

337

0.7 0.6 i(mA cm–2)

(a) 0.5 0.4 0.3 (c)

0.2 A

0.1

(b)

0 –0.5 –0.1

0.5

0

1 E(V) vs. SHE

1.5

2

–0.2 Figure 14.2 Cyclic voltammograms for the oxidation of 0.5 mM 2-naphthol in 1 M H2 SO4 : (a) first cycle, 0 mM 1,2-naphthoquinone; (b) second cycle, 0 mM 1,2-naphthoquinone; (c) first cycle, 5 mM 1,2-naphthoquinone. Scan rate 100 mV s−1 , T = 25◦ C. A: anodic start of the cyclic voltammograms. (Reprinted from Ref. 8.)

•

Two-step, one-electron oxidation of 2-naphthol to naphthyloxy radical and naphthyloxy cation: OH

O•

−e−

O+

−e−

−H+ •

(14.4)

Reaction of the naphthyloxy cation with water in a two-step reaction giving 1,2naphthoquinone: O O+

+

O

O +H O 2 −3H+ −2e−

(14.5)

The formation of 1,2-naphthoquinone according to Reaction (14.5) can explain the appearance of the 1,2-naphtoquinone/1,2-dihydroxynaphthalene peak during the second scan (Figure 14.2, curve c) In a parallel reaction pathway the naphthyloxy cations (or radicals) can react together, producing polymeric materials that decrease electrode activity: OH

−e− −H+

O• Polymerization

(14.6)

This assumption is confirmed by the consecutive cyclic voltammograms in a solution containing 1 mM 2-naphthol in 1 M H2 SO4 at a scan rate of 100 mV s−1 (Figure 14.3)

338

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

1.2 1.2 ipeak / i °peak

1

1 0.6

(2)

0.4

Peak 1.1 V vs. SHE Peak 2.0 V vs. SHE

0.2

0.8 i(mA cm–2)

(1 = 6)

0.8

(3) (4) (5)

0 0 0.6

500 1000 1500 Charge passed(mC cm–2)

0.4

0.2

A

0 0.6

0.8

1

1.2

1.4 1.6 E (V) vs. SHE

1.8

2

2.2

2.4

Figure 14.3 Consecutive cyclic voltammograms on BDD for 5 mM 2-naphthol in 1 M H2 SO4 . (1–5) consecutive cycles; (6) after reactivation at +2.6 V versus SHE for 10 minutes. Scan rate 100 mV s−1 . The inset shows the dependence of the normalized current peak on charge passed during the reactivation at +2.6 V versus SHE in 1 mM 2-naphthol in 1 M H2 SO4 . (Adapted from Ref. 8.)

[8,34]. As the number of cycles increases, the anodic peak currents for naphthol oxidation rapidly decrease until a steady state is reached (5 cycles). This decrease in electrode activity is originated by deposition of polymeric adhesive products on the electrode surface. The loss of electrode activity depends strongly on the naphthol concentrations, and in particular at low naphthol concentrations (<0.5 mM) the electrode deactivation seems to be less pronounced. Washing with organic solvents (isopropanol) cannot reactivate the electrode. However, the electrode surface can restore its initial activity by an anodic polarization in the same solution at +2.64 V versus SHE, as demonstrated by the fact that the naphthol oxidation peak returned to its initial position (Figure 14.3, curve 6). In fact, this potential is in the region of water discharge and on BDD it involves the production of hydroxyl radicals (Equation 14.1) that oxidize the polymeric film on the surface: •

OH

poly-naphthol film −−→ CO2

(14.7)

0 0 , where ipeak is The inset in Figure 14.3 shows the normalized peak currents (ipeak /ipeak the peak current during the first scan) during the reactivation at the fixed anode potential of +2.64 V versus SHE. When the charge passed exceeds 750 mC cm−2 (about 4 min), the naphthol oxidation peaks come back to their initial positions, meaning the complete reactivation of electrode surface. The results of the cyclic voltammetries have been also confirmed by the chronoamperometry measurements. Figure 14.4 shows potentiostatic i-t curves recorded in stirred solutions in 5 mM 2-naphthol at 2.0 V versus SHE [34]. Curve 1 shows that the current density decreases to very low values, confirming the blocking of the electrode surface by

14.2 APPLICATION IN WASTEWATER TREATMENT

339

8

i (mA cm–2)

6 4 (1 = 3) 2 (2) 0 0

10

20

30

40

50

t (s) Figure 14.4 Potentiostatic i-t curves recorded in a solution of 5 mM naphthol in 1 M H2 SO4 with a BDD anode at +2.0 V versus. SHE. Curve (1) new electrode; curve (2) after washing with isopropanol; curve (3) after anodic polarization at +2.64 V versus SHE for 10 min. (Reprinted from Ref. 34.)

the depositing polymeric adhesive products. Simple washing with organic solvents (isopropanol) can not reactivate the electrode (Figure 14.4, curve 2), but the initial activity can be restored after anodic polarization at +2.64 V versus SHE for 10 min (Figure 14.4, curve 3). A similar behavior has been also observed for the oxidation of other aromatic compounds such as phenol [35], chlorophenols [7,36] nitrophenols [37,38], hydroxybenzene [39], and synthetic dyes [40]. The results indicate that in the potential region before oxygen evolution, only reactions involving simple electron transfer are active; for example, the oxidation of the phenolic compounds to the phenoxy radicals and, subsequently, to the corresponding phenoxy cation: Ar − OH → Ar − O • → Ar − O+

(14.8)

These two electrochemically formed intermediates are very reactive and can couple to form polymers that deactivate the BDD surface. 14.2.2

Oxidation in the Potential Region of Oxygen Evolution

The cyclic voltammetry and chronoamperometry measurements have shown that the polarization in the potential region of oxygen evolution inhibits the electrode fouling and moreover restores the initial activity of the deactivated electrodes due to the electrogeneration of hydroxyl radicals (Equation 14.7)). For this reason, the electrochemical treatment of a wide range of organic compounds has been performed in a one-compartment flow cell with parallel plate electrodes (Figure 14.5) under galvanostatic conditions applying current values corresponding to anode potentials in the region of oxygen evolution. BDD was used as the anode and stainless steel AISI 304 as the cathode. Both electrodes were circular with a geometric area of 50 cm2 each and an interelectrode gap of 1 cm. The electrolyte was stored in a 0.5 dm3 thermoregulated glass reservoir and circulated through the electrochemical cell by a centrifugal pump with different recirculation flow rates. Some examples are shown in Table 14.1. At high potentials, all the organic compounds are completely mineralized by the reaction with electrogenerated • OH, according

340

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

1

T

2 3

1

3

2 4

+ –

6

5

4 7 (b)

(a)

Figure 14.5 Electrochemical cell for bulk oxidation of organics on diamond electrode. (a) Setup used: 1: thermoregulated reservoir, 2: electrochemical cell, 3: power supply, 4: pump. (b) Electrochemical cell, 1: outlet 2: anode, 3: cathode, 4: electrolysis compartment, 5: and 6: electrical contacts, 7: inlet.

TABLE 14.1 Some examples of organic compounds oxidized on BDD anodes. Pollutants Phenol 4-chlorophenol Naphthol Anionic surfactants Landfill leachate Benzensulfonic acids Mecoprop herbicide 3,4,5-trihydroxybenzoic acid 3-methylpyridine Synthetic dyes

Experimental conditions −2

i = 5–60 mA cm ; concentration: 20 mM i = 15–60 mA cm−2 ; concentration: 3.9–15.6 mM; T = 25–70◦ C i = 15–60 mA cm−2 ; concentration: 2–9 mM; T = 30–60◦ C i = 25–75 mA cm−2 ; flow rate: 60–180 dm−3 h−1 ; concentration: 750 mg dm−3 Initial COD = 350 mg dm−3 ; i = 10–60 mA cm−2 i = 20–60 mA cm−2 ; flow rate: 60–180 dm−3 h−1 ; initial COD = 1370 mg dm−3 i = 6–40 mA cm−2 ; flow rate: 75–300 dm−3 h−1 ; concentration: 180–700 mg dm−3 i = 10–60 mA cm−2 ; concentration: 1 g dm−3 i = 2.5–60 mA cm−2 ; concentration: 5 mM Methylene Blue, Alizarin Red, Eriochrome Black T, Methyl Red, Acid Yellow 1, Acid Blue 22

Ref. [35] [7,44] [8] [45] [46] [47,48] [49] [43] [50] [40,51–55]

to Equation 14.1. The current efficiency and the amount of intermediates are strongly affected by the experimental conditions. In particular, for high concentrations of organic compounds or low current densities, chemical oxygen demand (COD) decreases linearly, forming a large amount of intermediates, whereas instantaneous current efficiency (ICE, calculated with Equation 14.9) remains at about 100%, indicating a kinetically controlled process. Conversely, for low concentrations of organic compounds or high current densities, pollutants are directly mineralized to CO2 but ICE is below 100%, due to mass transport limitation and side reactions of oxygen evolution. The ICE values for the anodic oxidation of organic compounds can be calculated from the COD values using

14.2 APPLICATION IN WASTEWATER TREATMENT

341

the following reaction [41]: ICE = 4FV

[(COD)t − (COD)t+t ] I t

(14.9)

where (COD)t and (COD)t+t are the chemical oxygen demands at times t and t + t (in molO2 dm−3 ) respectively, I is the current (A), F is the Faraday constant (96487 C mol−1 ) and V is the volume of electrolyte (dm3 ). In order to describe these results, a comprehensive kinetic model to predict COD trends and current efficiency for the electrochemical combustion of the organic with BDD electrodes has been proposed [7,8,42,43]. The model, developed for an electrochemical reactor operating in a batch recirculation mode under galvanostatic conditions, is based on the following assumptions: (1) adsorption of the organic compounds at the electrode surface is negligible; (2) all the organics have the same diffusion coefficient (D); and (3) the global rate of the electrochemical mineralization of organics is a fast reaction and it is controlled by mass transport of organics to the anode surface. The consequence of this last assumption is that the mineralization reaction rate is independent on the chemical nature of the organic compound present in the electrolyte. The formulation of the model is based on the estimation of the limiting current density for the electrochemical oxidation of an organic compound (or a mixture of organics) from the value of the COD: ilim (t) = 4 · F · km · COD(t)

(14.10)

where ilim (t) is the limiting current (A m−2 ) at a given time t, 4 is the number of exchanged electrons, F is the Faraday’s constant (C mol−1 ), km is the average mass transport coefficient in the electrochemical reactor (m s−1 ) and COD(t) is the chemical oxygen demand (molO2 m−3 ) at a given time t. Depending on the relation between the current density (iappl ) and the limiting current density (ilim ), which decreases during treatment, two different operating regimes are identified: • iappl

< ilim , when the applied current is low or the concentration of the organics is sufficiently high, the electrolysis is under current limited control, the current efficiency is 100%, and the COD decreases linearly over time. • iappl > ilim , when the applied current is high or the concentration of the organics is low, the electrolysis is under mass-transport control, secondary reactions (such as oxygen evolution) commence, resulting in current efficiency decrease. In this regime, the COD removal follows an exponential trend due to mass-transport limitation. A graphic representation of the proposed kinetic model and the equations that describe the trends over time of COD and ICE in both regimes are given in Figure 14.6. The main advantage of this model is that it does not include any adjustable parameters. Hence, the system behavior can be predicted if the experimental conditions (applied current intensity, concentration of organic compounds, and mass transport coefficient) are known. The model has been tested for anodic oxidation of several organic compounds (Table 14.1) in different experimental conditions, and the good agreement between the experimental and modeling results obtained in all cases validates the assumptions on which the model is based.

342

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

(a) COD0 Akm t VR

COD

COD(t ) = COD0 1–α

iappl CODcr = 4Fk m

COD(t ) = αCOD0 exp –

Akm t + 1–α α VR

0 t

ICE(%)

(b) 100

0 Qcr = i lim

ICE = exp –

1–α km

Akm t +1–α α VR

0 t

Figure 14.6 Evolution of (a) COD and (b) ICE as a function of time (or specific charge); (A) represents the charge transfer control; (B) represents the mass transport control.

14.2.3

Influence of the Nature of Organic Pollutants

Figure 14.7 shows both the experimental (symbols) and predicted values (continuous line) of both ICE and COD evolution with the specific electrical charge passed during the anodic oxidation of different classes of organic compounds (acetic acid, isopropanol, phenol, 4-chlorophenol, and 2-naphtol) [31]. For all the compounds the electrochemical treatment is independent on the chemical nature of the organic pollutants and the complete

ICE / -

–3

COD(mg dm )

2500 2000 1500

1.0 0.8 0.6 0.4 0.2 0.0 0

1000

5 10 15 Specific charge(A dm–3)

500 0 0

5

10

15

Specific charge(A dm–3)

Figure 14.7 Evolution of COD and ICE (inset) as a function of the specific charge for different organic compounds: (×) acetic acid, () isopropanol, (◦) phenol, () 4-chlorophenol, () 2-naphtol; i = 238 A m−2 ; T = 25◦ C; Electrolyte: 1 M H2 SO4 ; the solid line represents model prediction. (Reprinted from Ref. 31.)

14.2 APPLICATION IN WASTEWATER TREATMENT

343

mineralization is obtained after the passage of 15 Ah dm−3 . Furthermore, these results demonstrate that there is an excellent agreement between the experimental data and the values predicted from proposed model. 14.2.4

Influence of the Concentrations of Organic Compounds

The effect of the initial concentration of the pollutants on the COD and ICE evolution during electrolysis of 4-chlorophenol (4-CP) when using a current density of 30 mA cm−2 is presented in Figure 14.8 [7]. Overall, COD removal is achieved in all cases and the time for total mineralization increased with the 4-CP concentration as expected from the presence of a greater amount of organic matter in the solution. However, at a high initial concentration and at the beginning of the electrolysis the COD decreases linearly with specific charge and ICE remains at about 100% (Figure 14.8 inset). This indicates that the electrolysis is performed at a current below the limiting one and under these conditions the oxidation of 4-CP is controlled by the rate at which electrons are delivered at the anode. On the contrary, at a low concentration, the ICE decreases linearly to zero, meaning that the oxidation is carried out at a current density higher than the limiting one and the process is under mass-transport control. Under the latter conditions, which are characteristic of electrolyses with low COD values, the process is controlled by the rate at which organic molecules are transported from the bulk liquid to the electrode surface [7,56]. Again, there is an excellent agreement between the experimental and predicted values. 14.2.5

Influence of the Applied Current Density

Electrolyses of 1 g dm−3 3,4,5-trihydroxybenzoic acid have been performed at 10 and 60 mA cm−2 (i.e., below and above the limiting current density), which is 46.4 mA cm−2 3500

1.2 1

3000 ICE/–

0.8 COD(mg dm–3)

2500

0.6 0.4

2000

0.2 0

1500

0

5

10 15 Q (Ah dm–3)

20

1000 500 0 0

5

10

15

20

25

Q (Ah dm–3)

Figure 14.8 Influence of 4-CP concentration on the evolution of COD and ICE (inset) with the specific electrical charge passed during the electrolyses on BDD anode. Electrolyte: 1 M H2 SO4 ; T = 25◦ C; i = 30 mA cm−2 ; initial 4-CP concentration: () 3.9 mM; (×) 7.8 mM; (•) 15.6 mM. The solid lines represent model prediction. (Reprinted from Ref. 7.)

344

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

ICE /–

1200

COD (mg dm–3)

1000 800 600

1 0.8 0.6 0.4 0.2 0 0

2

4 t (h)

400 200 0 0

1

2

3

4

5

t (h)

Figure 14.9 Influence of current density on the time evolution of COD and ICE (inset) during the electrolyses of 1 g dm−3 3,4,5-trihydroxybenzoic acid in HClO4 0.5 M. Current density: () 10 mA cm−2 ; () 60 mA cm−2 . The solid lines represent model prediction. (Reprinted from Ref. 43.) 1200

dm–3)

800

COD (mg

1000

600 400 200 0 0

5

10

15

20

Q (Ah dm–3)

Figure 14.10 Influence of current density on the evolution of COD with specific charge during the electrolyses of 1 g dm−3 3,4,5-trihydroxybenzoic acid in HClO4 0.5 M. Current density: () 10 mA cm−2 ; () 60 mA cm−2 . The solid lines represent model prediction.

in our experimental conditions (calculated with Equation 14.13). Figure 14.9 shows the time evolution of COD and current efficiency, whereas Figure 14.10 shows the trend of COD with electrical charge consumed [43]. At a low current density, the current efficiency remains at 100% during almost all the oxidation, and the mineralization requires a low electrical charge but a long electrolysis time because some of the reactor capacity is underused. On the contrary, when operating current exceeds the limiting one, the electrolysis is fast but the current efficiency decreases and therefore the specific charge consumed increases because a portion of the current is wasted on the secondary reaction of oxygen evolution: 2 • OH → O2 + 2H+ + 2e−

(14.11)

As can be seen, the model can satisfactorily predict the experimental data for all the current densities.

14.2 APPLICATION IN WASTEWATER TREATMENT

180 Color removal (%)

100

160 140 COD(mg dm–3)

345

120 100

80 60 Q = 60 dm3 h–1

40

Q = 100 dm3 h–1 Q = 180 dm3 h–1

20

80

0 0

60

50

100

150

t (min)

40 20 0 0

200

400

600

800

1000

1200

1400

t (min)

Figure 14.11 The influence of the electrolyte flow rate on the COD evolution and color removal (inset) during the direct electrolyses of 80 mg dm−3 Methylene Blue. Electrolyte: Na2 SO4 0.5 M; T = 20◦ C; i = 20 mA cm−2 . (Reprinted from Ref. 51.)

14.2.6

Influence of the Flow Rate

The effect of the flow rate during the oxidation of the synthetic dye Methylene Blue at 20 mA cm−2 is shown in Figure 14.11 [51]. The COD and dye removals are faster at higher flow rates, meaning that the oxidation is a mass-controlled process. In fact, the increase in the flow rate produces a higher concentration of organics that can react with electro-generated hydroxyl radicals, near the electrode surface, avoiding their decomposition to oxygen (Equation 14.11). As previously noted, an excellent agreement between the experimental and predicted values is observed. 14.2.7

Influence of the Temperature

Figure 14.12 reports the effect of temperature on the oxidation of 1 g dm−3 of the nitro dye Acid Yellow 1 in HClO4 1 M [55]. In a medium that cannot generate oxidizing species, such as HClO4 , higher temperatures do not yield a significant increase of the oxidation rate during the electrochemical incineration process (see Figure 14.12). The small difference between the results obtained at the two temperatures is due only to an increase of the diffusion rate with rising temperature due to the decrease of the medium viscosity. However, several studies report that an increase of temperature favors organic oxidation [51,54,57–59]; this behavior is not caused by the increase of the activity of the BDD anode, but by the increase of the indirect reaction of organics with electro-generated oxidizing agents from the electrolyte oxidation. In fact, electrolysis with BDD anodes in media containing chloride, sulphate, or phosphate ions generates chlorine (Equation 14.15), peroxodisulfate (Equation 14.13), and peroxodiphosphate (Equation 14.14). 2Cl− → Cl2 + 2e−

(14.12)

2SO4 2− → S2 O8 2− + 2e− 2PO4

3−

→ P2 O8

4−

+ 2e

−

(14.13) (14.14)

346

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

[Acid Yellow 1] (g L–1)

1 0.8 T = 25°C 0.6

T = 40°C

0.4 0.2 0 0

0.5

1

1.5

2

t (h) Figure 14.12 Effect of the temperature on the evolution of 1 g dm−3 Acid Yellow 1 concentration during the electrolysis with BDD anode. Conditions: i = 30 mA cm−2 ; Flow rate = 300 L h−1 ; T = () 20◦ C; () 40◦ C. (Reprinted from Ref. 55.)

These powerful oxidizing agents can oxidize the organic matter by a chemical reaction whose rate increases with temperature, following the Arrhenius Law. 14.2.8

Comparison with Other Electrode Materials

The degradation ability of BDD has been also compared with traditional electrodes, such as Pt, PbO2 , and Ti–Ru–Sn ternary oxide for the oxidation of Methyl Red [52]. The results are presented in Figure 14.13 evidence that BDD enables the highest oxidation rate and current efficiency. To explain its best performance, it is speculated that on BDD—which has an inert surface with weak adsorption properties—electro-generated hydroxyl radicals are weakly adsorbed and consequently more reactive toward organic oxidation. Furthermore, it is assumed that • OH action is extended to a “reaction cage” in the vicinity of the electrode surface, rather than limited to the surface itself. 250

TiRuSnO2 Pt PbO2 BDD

COD(mg dm–3)

200 150 100 50 0 0

2

4

6

8

10

12

t (h) Figure 14.13 Comparison of the trend of COD during the oxidation of 200 mg dm−3 Methyl Red in 0.5 M Na2 SO4 at different anodes. Conditions: I = 500 mA; flow rate 180 dm3 h−1 . (Reprinted from Ref. 52.)

14.3 APPLICATION IN ORGANIC ELECTROSYNTHESIS

347

Other papers have also demonstrated that BDD electrodes are able to achieve faster oxidation and better incineration efficiency than other electrode materials in the treatment of naphthol [34,60], 4-chlorophenol [44], chloranilic acid [61], surfactants [62], and herbicides [49,63]. 14.3

APPLICATION IN ORGANIC ELECTROSYNTHESIS

The electrosynthetic processes for the production of organic compounds possess several advantages over chemical reactions [64], such as: (1) no need for oxidizing agents, thus eliminating the hazardous involved in handling these materials and reducing the costs associated with buying and disposing hazardous reagents applied at an industrial scale; (2) precise control of reaction conditions, reducing the amount of toxic effluents that must be treated prior to discharge; and (3) selective transformation can take place by either controlling the applied voltage or by taking advantage of differences in the affinity for metal electrodes of similarly electron-rich functional groups. Despite these advantages, the availability and stability of the performance of the electrodes are the main limitation of the electrosynthesis; therefore, only few processes have been applied on an industrial scale. As mentioned earlier, BDD electrodes have a high corrosion stability in very aggressive media, and in the potential region of oxygen evolution electrode deactivation is avoided due to the production of hydroxyl radicals. Therefore, BDD electrodes open new possibilities for the electrosynthesis of organic compounds. However, with BDD the reaction path and the nature of products are not influenced by controlling the electrode potential, as in conventional electrosynthesis, but by controlling the γ ratio between the production rate of hydroxyl radicals (r • OH ) and the flux of reactants to the electrode surface (rR ) [65]: γ =

iapp r • OH = rR F · km · [R]

(14.15)

where iapp is the applied current density (in A m−2 ), F is the Faraday’s constant, km is the mass-transport coefficient in the electrochemical reactor (m s−1 ), and [R] is the concentration of organics in the electrolyte (mol m−3 ). In particular, for a given reaction, if the value of γ is lower than the number of electrons (or hydroxyl radicals) involved in the reaction, there is only an accumulation of products, without appreciable mineralization of the reactants. For example, for a γ value lower than 6, nicotinic acid, an important pharmaceutical intermediate, can be produced by the selective oxidation of 3-methylpyridine (3-MP) [50]: OH

CH3

O

+ 6 •OH N 3-methylpyridine

+ 4H2O

N nicotinic acid

(14.16)

The results of a high-performance liquid chromatography (HPLC) of the solution during the oxidation of 5 mM 3-MP to nicotinic acid carried out at 2.5 mA cm−2 is given in

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

Concentration (mmol dm–3)

6

400 350

5

300 4

250

3

200 150

2

100 1

TOC(mg dm–3)

348

50

0

0 0

0.2

0.4

0.6

0.8

1

1.2

Q (Ah dm–3) Figure 14.14 Concentration trends of: () 3-MP, () nicotinic acid, (◦) oxidation intermediates, and (×) TOC during 3-MP electrolysis on BDD anode. Electrolyte: 0.5 M HClO4 ; initial 3-MP concentration = 5 mM; i = 2.5 mA cm−2 . (Reprinted from Ref. 50.)

Figure 14.14. Under these conditions, corresponding to γ = 2.6–5.2, the concentration of 3-MP decreases linearly with the specific charge, forming nicotinic acid as the main product, for 3-MP conversion up to 50%. The total organic carbon (TOC) in the solution remains almost constant, indicating that the mineralization of 3-MP to CO2 does not occur under these conditions. During the electrolysis, the electrode potential remains E = +2.7±0.1 V versus SHE, meaning that there is no deactivation under these conditions. In a similar way, benzoquinone can be produced from the selective oxidation of phenol working with γ lower than 4 [35]: O

OH + 4 •OH

+ 3H2O O

(14.17)

The results of bulk electrolysis of 20 mM phenol in a one-compartment cell at 5 mA cm−2 , corresponding to γ = 1.3–1.6, are presented in Figure 14.15. For low phenol conversion (<20%), the concentration of phenol decreases linearly with the specific charge, forming benzoquinone as the main product and small amounts of hydroquinone and catechol. The TOC of the electrolyte remains almost constant during the electrolysis, confirming the partial oxidation of the phenol. 14.4

CONCLUSIONS

Conducting diamond thin film is a new electrode material that has received great attention recently because it possesses several technologically important characteristics such as an inert surface with low adsorption properties; remarkable corrosion stability, even in strong acidic media; and an extremely wide potential window for water discharge. Thanks to these properties, diamond electrodes offer significant advantages over other electrode materials for the oxidation of organic compounds in wastewater treatment. In the potential

REFERENCES

349

1400

18

1200

16 14

1000

12 800

10

600

8 6

400

TOC(mg dm–3)

Concentration(mmol dm–3)

20

4 200

2 0

0 0

0.5

1

1.5

2

Q (Ah dm–3) Figure 14.15 Variation of the concentration of: () phenol, () benzoquinone, () hydroquinone, (◦) catechol, and (×) TOC during phenol electrolysis on BDD anode. Electrolyte: 1 M HClO4 ; initial phenol concentration = 20 mM; i = 5 mA. (Reprinted from Ref. 35.)

region before oxygen evolution, only reactions involving simple electron transfer are active on BDD but they cause electrode fouling. On the contrary, in the potential region of oxygen evolution, BDD produces a great amount of hydroxyl radicals that avoid electrode deactivation and oxidize organic compounds with high current efficiency, even close to 100%. The capacity of the BDD anode to oxidize the organics is independent of the nature of the pollutants, and the current efficiency is limited only by mass transfer of the organic molecules to the anode surface, particularly when the concentration of the pollutants reaches low values during the treatment or when high current density is applied. Moreover, controlling the production rate of the hydroxyl radicals and the mass transfer of the organic species to the anode surface, it is possible to control the nature of the reaction products and thus the BDD anode opens new possibilities in the field of organic electrosynthesis.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

G. Chen, Sep. Purif. Technol. 2004, 38 , 11–41. M. Panizza, G. Cerisola, Chem. Rev . 2009, 109 , 6541–6569. C. Comninellis, G. Chen, Electrochemistry for the Environment ; New York, Springer, 2009. M. Gattrell, D. Kirk, J. Electrochem. Soc. 1993, 1534–1540. G. Foti, D. Gandini, C. Comninellis, Current Topics Electrochem. 1997, 5 , 71–91. J.D. Rodgers, W. Jedral, N.J. Bunce, Environ. Sci. Technol. 1999, 33 , 1453–1457. M.A. Rodrigo, P.A. Michaud, I. Duo, M. Panizza, G. Cerisola, C. Comninellis, J. Electrochem. Soc. 2001, 148 , D60–D64. M. Panizza, P.A. Michaud, G. Cerisola, C. Comninellis, J. Electroanal. Chem. 2001, 507 , 206. D.F. Steele, Platinum Metals Rev . 1990, 34 , 10–14. J. Bringmann, K. Ebert, U. Galla, H. Schmieder, J. Appl. Electrochem. 1995, 25 , 846–851. F. Bonfatti, S. Ferro, F. Lavezzo, M. Malacarne, G. Lodi, A. De Battisti, J. Electrochem. Soc. 2000, 147 , 592–596.

350

APPLICATION OF SYNTHETIC DIAMOND FILMS TO ELECTRO-OXIDATION PROCESSES

12. M. Panizza, G. Cerisola, Electrochim. Acta 2003, 48 , 1515–1519. 13. M. Panizza, G. Cerisola, Water Res. 2001, 35 , 3987–3992. 14. S.-G. Park, G.-S. Kim, J.-E. Park, Y. Einaga, A. Fujishima, J. New Mat. Electr. Sys. 2005, 8 , 65–68. 15. L.M. Da Silva, D.V. Franco, J.C. Forti, W.F. Jardim, J.F.C. Boodts, J. Appl. Electrochem. 2006, 36 , 523–530. 16. I. Sir´es, N. Oturan, M.A. Oturan, R.M. Rodr´ıguez, J.A. Garrido, E. Brillas, Electrochim. Acta 2007, 52 , 5493–5503. 17. I. Sir´es, J.A. Garrido, R.M. Rodr´ıguez, E. Brillas, N. Oturan, M.A. Oturan, Appl. Catal. B-Environ. 2007, 72 , 382–394. 18. M. Panizza, G. Cerisola, Water Res. 2009, 43 , 339–344. 19. H. Chang, D.C. Johnson, J. Electrochem. Soc. 1990, 137 , 2452–2457. 20. C. Comninellis, Electrochim. Acta 1994, 39 , 1857–1862. 21. K.T. Kawagoe, D.C. Johnson, J. Electrochem. Soc. 1994, 141 , 3404–3409. 22. J. Feng, D.C. Johnson, S.N. Lowery, J.J. Carey, J. Electrochem. Soc. 1994, 141 , 2708–2711. 23. J. Feng, L.L. Houk, D.C. Johnson, S.N. Lowery, J.J. Carey, J. Electrochem. Soc. 1995, 142 , 3626–3632. 24. O. Simond, V. Schaller, C. Comninellis, Electrochim. Acta 1997, 42 , 2009–2012. 25. H.B. Martin, A. Argoitia, U. Landau, A.B. Anderson, J.C. Angus, J. Electrochem. Soc. 1996, 143 , L133–L136. 26. R. Ramesham, M.F. Rose, Diam. Rel. Mat . 1997, 6 , 17–27. 27. G.M. Swain, Adv. Mater. 1994, 5 , 388–392. 28. Y.Y. Pleskov, Russ. J. Electrochem. 2002, 38 , 1275–1291. 29. B. Marselli, J. Garcia-Gomez, P.A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 150 , D79–D83. 30. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191–199. 31. M. Panizza, E. Brillas, C. Comninellis, J. Environ. Eng. Manage. 2008, 18 , 139–153. 32. C.A. Mart´ınez-Huitle, S. Ferro, Chem. Soc. Rev . 2006, 12 , 1324–1340. 33. B. Correa-Lozano, C. Comninellis, A. De Battisti, J. Appl. Electrochem. 1997, 27 , 970–974. 34. M. Panizza, G. Cerisola, Electrochim. Acta 2003, 48 , 3491–3497. 35. J. Iniesta, P.A. Michaud, M. Panizza, G. Cerisola, A. Aldaz, C. Comninellis, Electrochim. Acta 2001, 46 , 3573–3578. 36. P. Ca˜nizares, J. Garc´ıa-G´omez, C. S´aez, M.A. Rodrigo, J. Appl. Electrochem. 2003, 33 , 917–927. 37. P. Ca˜nizares, C. S´aez, J. Lobato, M.A. Rodrigo, Electrochim. Acta 2004, 49 , 4641–4650. 38. P. Ca˜nizares, C. S´aez, J. Lobato, M.A. Rodrigo, Ind. Eng. Chem. Res. 2004, 43 , 1944–1951. 39. P. Ca˜nizares, C. S´aez, J. Lobato, M.A. Rodrigo, Ind. Eng. Chem. Res. 2004, 43 , 6629–6637. 40. C. S´aez, M. Panizza, M.A. Rodrigo, G. Cerisola, J. Chem. Technol. Biotechnol . 2007, 82 , 575–581. 41. C. Comninellis, C. Pulgarin, J. Appl. Electrochem. 1991, 21 , 703–708. 42. M. Panizza, P.A. Michaud, G. Cerisola, C. Comninellis, Electrochem. Commun. 2001, 3 , 336. 43. M. Panizza, A. Kapalka, C. Comninellis, Electrochim. Acta 2008, 53 , 2289–2295. 44. L. Gherardini, P.A. Michaud, M. Panizza, C. Comninellis, N. Vatistas, J. Electrochem. Soc. 2001, 148 , D78. 45. M. Panizza, M. Delucchi, G. Cerisola, J. Appl. Electrochem. 2005, 35 , 357–361. 46. M. Panizza, M. Zolezzi, C. Nicolella, J. Chem. Technol. Biotechnol . 2006, 81 , 225–232.

REFERENCES

47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.

351

M. Panizza, G. Cerisola, Int. J. Environ. Pollut . 2006, 27 , 64–74. M. Panizza, G. Cerisola, Ann. Chim. Rome 2003, 93 , 977–984. I. Sir´es, E. Brillas, G. Cerisola, M. Panizza, J. Electroanal. Chem. 2008, 613 , 151–159. J. Iniesta, P.A. Michaud, M. Panizza, C. Comninellis, Electrochem. Commun. 2001, 3 , 346–351. M. Panizza, A. Barbucci, R. Ricotti, G. Cerisola, Sep. Purif. Technol. 2007, 54 , 382–387. M. Panizza, G. Cerisola, Appl. Catal. B-Environ. 2007, 75 , 95–101. M. Panizza, G. Cerisola, Ind. Eng. Chem. Res. 2008, 47 , 6816–6820. M. Panizza, G. Cerisola, J. Hazard. Mater. 2008, 153 , 83–88. J. Rodriguez, M.A. Rodrigo, M. Panizza, G. Cerisola, J. Appl. Electrochem. 2009, 39 , 2285–2289. A.M. Polcaro, A. Vacca, S. Palmas, M. Mascia, J. Appl. Electrochem. 2003, 33 , 885–892. P. Ca˜nizares, J. Garc´ıa-G´omez, C. S´aez, M.A. Rodrigo, J. Appl. Electrochem. 2004, 34 , 87–94. X. Chen, G. Chen, Sep. Sci. Technol. 2006, 48 , 45–49. P. Ca˜nizares, R. Paz, J. Lobato, C. S´aez, M.A. Rodrigo, J. Hazard. Mater. 2006, 138 , 173–181. M. Panizza, G. Cerisola, Electrochim. Acta 2004, 49 , 3221–3226. C.A. Mart´ınez-Huitle, M.A. Quiroz, C. Comninellis, S. Ferro, A. De Battisti, Electrochim. Acta 2004, 50 , 949–956. E. Weiss, K. Groenen-Serrano, A. Savall, J. New Mat. Electr. Sys. 2006, 9 , 249–256. C.A. Mart´ınez-Huitle, A. De Battisti, S. Ferro, S. Reyna, M. Cerro-Lopez, M.A. Quiroz, Environ. Sci. Technol. 2008, 42 , 6929–6935. D. Pletcher, F.C. Walsh, Industrial Electrochemistry, London, Chapman and Hall, 1982. M. Panizza, P.A. Michaud, J. Iniesta, C. Comninellis, G. Cerisola, Ann. Chim. Rome 2002, 92 , 995–1006.

15 Fabrication and Application of Ti/BDD for Wastewater Treatment Xueming Chen and Guohua Chen

15.1 15.1.1

FABRICATION OF STABLE Ti/BBD ELECTRODES Introduction

Diamond is known for its inertness, hardness, and wide window of water electrolysis. Boron-doped diamond (BDD), p-type semi-conductor, has become an interesting and a promising alternative electrode material for environmental-related applications such as electro-oxidation, electroreduction, as well as electrochemical green processing of chemicals [1,2]. The history of diamond electrode utilization has been surveyed briefly by Rao, Fujishima, and Angus [3]. This material can now be produced by a low-pressure chemical vapor deposition (CVD) method converting a gas phase carbon into a solid cubic crystalline form [4,5]. Following the release of the methods of growing diamond on a nondiamond substrate in 1976, various reports have been made for the production of this composite material. There are usually three types of CVD methods available: plasma-assisted CVD, combustion flames, and hot filament CVD (HFCVD). The HFCVD method is probably the simplest, with the least requirement of the production facility and the most versatile technique. It is therefore the primary technique for the preparation of large-area BDD on different substrates such as silicon, niobium, tungsten, molybdenum, and titanium. Since titanium is an ideal material for the substrates of

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

353

354

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

dimensionally stable electrodes [6], the preparation of Ti/BDD electrodes using HFCVD is therefore of more industrial relevance. For the growth of diamond using plasmaassisted CVD, a combined process of a homogeneous plasma bulk and a heterogeneous process at the plasma-substrate interface, readers are referred to a recent published book Diamond Electrochemistry edited by Fujishima and colleagues [7]. There are many types of plasma-generating techniques such as direct current plasma, radio frequency plasma, microwave plasma, electron cyclotron resonance microwave plasma, as well as high pressure plasma. The plasma contains the dissociated hydrogen and hydrocarbon species. The diamond is grown on the substrate surface following the stabilization of the sp3 carbon phase [8]. For the electrode fabrication in general, readers are referred to Guo and associates in the current book edited by Comninellis and Chen [1], in which the thermal decomposition method and the surface modification method are discussed, along with the chemical vapor deposition method [9]. 15.1.2

HFCVD Facility

The major facilities required for the HFCVD growth of diamond are as follows: a system for stable, quantified, and uniform supply of precursor gases (carbon containing gas, hydrogen, boron source, and additives); a vacuum pump to achieve the desired low pressure; a system to heat the filament (W or Ta) to the desired temperature and to maintain the spacing between the hot filament and the substrate; a system to regulate the substrate temperature; a reaction chamber to house the substrate and the filament; a ventilation system for the nonreacted gases; and a control system for the whole process. The BDD films on different substrates are now produced in batches with a maximum area of 0.5m2 available industrially. A HFCVD experimental setup used for deposition of Ti/BDD electrodes is schematically shown in Figure 15.1 [10]. 15.1.3

HFCVD Parameter Optimization

Diamond films are generally obtained at low pressures by CVD techniques from the mixture of H2 + CH4 with a low CH4 concentration. It is understandable that the quality of diamond is highly dependent on the CVD conditions. The typical conditions for depositing diamond on Si using the HFCVD technique are substrate temperature 700–1000◦ C, filament temperature ∼2000◦ C, CH4 concentration 1%, and total gas pressure 13–133 mbar [11,12]. Nevertheless, it should be pointed out that the performance of Ti as a substrate material of diamond films is rather different from that of Si because of the much stronger affinity of titanium for carbon. The strong affinity for carbon may hinder diamond nucleation and form a thick TiC layer at the interface [13]. Moreover, when BDD-coated titanium is used for electrochemical purpose, the situation is complicated by electrochemical corrosion. Therefore, it is expected that the proper HFCVD conditions for Ti/BDD electrode fabrication are different from the typical conditions mentioned above. Chen and Chen [14] have investigated the stability dependence on the major HFCVD parameters including substrate temperature (Tsub ), filament temperature (Tfil ), CH4 concentration, filament-substrate distance (dfil−sub ), and deposition time (tdep ). Accelerated life tests were used to assess electrode stability. Table 15.1 summarizes the experimental results for electrodes prepared under different conditions. As Fan, Jagannadham, and Narayan [15] and Buccioni et al. [16] reported, diamond films are easy to detach from titanium substrates. It is found that the occurrence of detachment depends strongly on

15.1 FABRICATION OF STABLE Ti/BBD ELECTRODES

Methane

355

2

1

3

1

6 Hydrogen 7 4

8

9

10 11

13

5

Vacuum pump 12

Figure 15.1 Diagram of HFCVD experimental setup. 1: gauges; 2: CH4 mass flow controller; 3: H2 mass flow controller; 4: organic additive container; 5: B(OCH3 )3 supply; 6: copper pipe; 7: tantalum filament; 8: titanium substrate; 9: niobium support; 10: tantalum heating element; 11: ceramic support; 12: thermal couple; 13: quartz bell. (Reprinted from Ref. 9.) TABLE 15.1 Ti/BDD stability dependence on HFCVD conditions. Pressure 30 mbar; accelerated life test conditions: 10,000 A m−2 , 3 M H2 SO4 , 50◦ C [13]. Tsub (◦ C) 720 770 770 770 820 820 850 850 850 850 880 880

Tfil (◦ C)

CH4 (%)

dfil-sub (mm)

tdep (h)

Initial adhesion status

∼2000 ∼2000 ∼2000 ∼2000 ∼2050 ∼2050 ∼2050 2120–2150 2160–2180 2180–2210 2120–2150 2200–2220

0.5 0.5 0.5 1.0 0.5 0.5 0.8 0.8 0.8 0.8 0.8 0.8

5 5 5 5 5 5 8 8 8 8 8 8

5 5 20 20 5 20 15 15 15 12 10 10

good adhesion good adhesion good adhesion good adhesion good adhesion severely detached good adhesion good adhesion slightly detached severely detached slightly detached severely detached

Service life (h) <0.5 7.6 59 36 11 — 56 95 — — — —

the HFCVD conditions. Using high substrate and filament temperatures, a short filamentsubstrate distance, and a long deposition time tends to result in severe film detachment. The mechanisms to cause diamond film detachment are complex. The easy film detachment at a high substrate temperature is primarily attributed to the formation of a thick titanium carbide (TiC) layer. As mentioned earlier, a very important feature of titanium is its strong affinity for carbon. The growth of TiC can compete with diamond formation for

356

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

the available carbon. Since TiC formation is more temperature-sensitive than the growth of a diamond film, use of a high substrate temperature results in formation of a thick TiC layer. This thick TiC layer may crack because of volume change and can induce diamond film to crack, promoting interfacial debonding and spallation of the diamond film [17]. The easy film detachment at a high substrate temperature is also associated with the large residual stress from the thermal expansion mismatch between the diamond film and Ti substrate. The thermal expansion coefficients of diamond and Ti are 0.8 × 10−6 K−1 and 8.4 × 10−6 K−1 , respectively. The over one magnitude difference in thermalexpansion coefficients will cause a large residual stress when a Ti/BDD sample is cooled from a high deposition temperature to an ambient temperature. This stress inevitably increases with the substrate surface temperature used. Ager and Drory [18] have calculated that the limiting stress is 17 GPa, beyond which diamond film detachment will occur. The reason why using a high filament temperature and a short filament-substrate distance caused film detachment is not clear. It is probably related to uneven heating. When the filament temperature is high and the filament-substrate distance is short, part of the substrate surface located under the filament will be overheated due to the spiral geometry of the filament. The local overheating may facilitate formation of a thick TiC layer there and cause the substrate to deform, leading to BDD film detachment. Although BDD films could be deposited on titanium substrates with good adhesion at low substrate and filament temperatures, and a short deposition time, the durability is still poor. This is attributed to the poor diamond quality and uncompleted coverage. As Tsub , Tfil , and tdep increased properly, the service life increased significantly. In the investigated ranges, the optimal conditions that gave a service life of 95 h in the accelerated test are: Tsub 850◦ C, Tfil 2,120–2,150◦ C, CH4 0.8%, dfil−sub 8 mm, tdep 15 h, and pressure 30 mbar. Figure 15.2 displays the Raman spectra of the diamond film deposited at optimized conditions [14]. The film has a peak slightly away from 1332 cm−1 , which is the characteristic signature of the diamond structure [19]. The peak shift was caused by the compression stress that was produced due to the large difference of the thermal expansion coefficients between a diamond film and a titanium substrate [20]. Apart from the diamond peak, there is a broad band, which is from amorphous or graphitic sp2 carbon impurities [21]. The ratio of the sp2 carbon to the diamond scattering intensities counts to be about 1.5. It was reported that the cross-sectional scattering coefficients for diamond and graphite are 9 × 10−7 cm−1 /sr and 500 × 10−7 cm−1 /sr, respectively [19]. Normalizing the band intensities with these coefficients indicates that the sp2 carbon content is <0.2%. Figure 15.3 shows the XRD pattern of a Ti/BDD sample [14]. Distinct diamond peaks were detected. In addition, intensive TiC peaks can be clearly observed, revealing the existence of an internal TiC layer between the titanium substrate and diamond film. Figure 15.4 displays the typical SEM images of the BDD film deposited on Ti [14]. Well-defined diamond crystals can be observed clearly. The diamond crystals are highly faceted and have an average particle size of about 2 μm. The continuous film was formed, with few cracks detected.

15.1 FABRICATION OF STABLE Ti/BBD ELECTRODES

357

8000 1336

Counts

7000

6000

5000 1490 4000

3000

1200

1300

1400

1500

1600

Shift (cm–1) Figure 15.2 Raman spectrum of the film deposited at Tsub 850◦ C, Tfil 2,120–2,150◦ C, CH4 0.8%, dfil-sub 8 mm, tdep 15 h, and pressure 30 mbar. (Reprinted from Ref. 13.)

300

TiC TiC (111) (200)

250 Diam (111)

150

50

TiC (220)

Ti (110)

Ti (103)

Diam(220) TiC(311)

100

TiC(222)

Counts

200

Ti (201)

Diam (311)

0 30 35 40 45 50 55 60 65 70 75 80 85 90 95 2θ (degrees) Figure 15.3 XRD pattern of Ti/BDD prepared at Tsub 850◦ C, Tfil 2,120–2,150◦ C, CH4 0.8%, dfil-sub 8 mm, tdep 15 h, and pressure 30 mbar. (Reprinted from Ref. 13.)

15.1.4

Reactive Gas Component Improvement

It has been established that the Ti/BDD electrode failure is a result of the BDD film delamination [14,22]. The BDD film delamination is essentially associated with the electrochemical corrosion of the TiC layer. Therefore, the key to enhance the Ti/BDD stability is to prevent the formation of a thick TiC layer. In addition, increasing the thickness of the protective BDD layer is also helpful.

358

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

Figure 15.4 SEM image of BDD film deposited on Ti using H2 + CH4 under conditions of Tsub 850◦ C, Tfil 2,120–2,150◦ C, CH4 0.8%, dfil-sub 8 mm, td 15 h, and pressure 30 mbar. (Reprinted from Ref. 13.)

Technically, an effective method to enhance the Ti/BDD stability is to add an interlayer between the Ti substrate and the BDD film before deposition. The material of the interlayer can be selected from Si, Nb, Ta, and W. Since Si, Ta, Nb, and W have much lower affinity for carbon than Ti, addition of an interlayer of Si, Nb, Ta, or W can hinder formation of a thick TiC layer, leading to an increase in the electrode stability. Diamond films can be deposited on the surfaces of Si, Nb, Ta, and W well. Actually, Fryda et al. [22] have reported that Si/BDD, Nb/BDD, Ta/BDD, and W/BDD electrodes are very stable for electrochemical attack even under high loads. Another effective method to enhance the Ti/BDD stability is to promote diamond deposition. The fast diamond nucleation and growth not only increases the BDD film thickness at a given deposition time but it also suppresses formation of a thick TiC layer, leading to an increase in the electrode stability. Diamond deposition can be promoted by varying the reactive gas component. It was reported that addition of a small amount of H2 O [23] or O2 [24,25] to the conventional H2 + CH4 mixture or use of oxygen-containing compounds such as CH3 OH, C2 H5 OH, (CH3 )2 CHOH, (CH3 )3 COH, CH3 COCH3 , [(CH3 )2 CH]2 O, C2 H5 OC2 H5 , CH3 COOCH3 , CH3 CHO [26], and CO [27] as carbon sources could increase the diamond growth rate significantly. HFCVD of diamond on Si using the mixture of H2 + CH3 COCH3 , for instance, had a growth rate of 8–10 μm h−1 , over 10 times faster than that using the conventional H2 + CH4 [26]. Additionally, it was reported that use of chlorinated hydrocarbons (CH2 Cl2 , CHCl3 , CCl4 , C2 H5 Cl) or fluorinated hydrocarbons (CF4 and CHF3 ) as carbon sources could also promote diamond growth significantly [21,28–30]. Chen and colleagues [10,31,32] have investigated three different organic compounds including CH3 COCH3 , CH2 Cl2 , and CH2 (OCH3 )2 . Due to the severe corrosion to the

359

15.1 FABRICATION OF STABLE Ti/BBD ELECTRODES

tantalum filament, these compounds were not used as sole carbon sources but as additives to the H2 + CH4 mixture. It was found that addition of CH3 COCH3 to H2 + CH4 did increase the film growth rate significantly. As H2 and CH4 were used as reactive gases, the increased weight of a sample after 15 h of deposition was about 8 mg only. When CH3 COCH3 was added, the increased sample weight was as high as 16.8 mg despite the deposition time being reduced to 5 h. However, Raman analysis revealed that the films formed were sp2 carbons. Accelerated life tests showed that the electrodes thus prepared were very unstable. The films could be dissolved quickly. A total of five samples were prepared, but none of them had a service life over 2 hours [31]. Therefore, CH3 COCH3 is not a good additive for deposition of BDD films on Ti substrates, although it was reported that a good-quality diamond film could be obtained using Si as a substrate and CH3 COCH3 as a carbon source [26]. Unlike CH3 COCH3 , CH2 Cl2 could enhance Ti/BDD electrode stability dramatically. On the average, addition of CH2 Cl2 to H2 + CH4 increased the service life by 80%. However, it was found that CH2 Cl2 deteriorated the diamond quality [31]. CH2 (OCH3 )2 is found to be the best additive investigated. Table 15.2 compares the service lives of electrodes prepared with and without addition of CH2 (OCH3 )2 under accelerated life test conditions [32]. The addition of CH2 (OCH3 )2 to the reactive gases increased the electrode service life by a factor of 2.35–2.97 depending on the HFCVD conditions. It is generally believed that • CH3 and • H radicals play very important roles in diamond growth. For the conventional H2 + CH4 reactive gas mixture, • H and • CH3 radicals are generated through the following reactions: H2 → 2 • H

(15.1)

CH4 + • H → • CH3 + H2

(15.2)

The mechanism of addition of • CH3 to the diamond structure is complex and has not been clearly explained. The GDSB mechanism [33], as shown in Figure 15.5, is now widely believed to be a principal route for dimer opening and carbon insertion. Obviously, the rate of the diamond growth depends on the concentrations of • CH3 and • H radicals available at the depositing surface. The higher the concentrations of • CH3 and • H are, the faster the diamond growth rate is. For the H2 + CH4 + CH2 (OCH3 )2 system, in addition to Reactions (15.1) and (15.2), there are the following possible reactions to generate • CH3 and • OH radicals: TABLE 15.2 Service life comparison of Ti/BDD electrodes prepared with and without adding CH2 (OCH3 )2 . H2 100 sccm; CH2 (OCH3 )2 13.8 mg h−1 ; accelerated life test conditions: 3 M H2 SO4 , 10,000 A m−2 , 50◦ C [31]. HFCVD conditions Tsub 770◦ C, Tfil 2000◦ C, CH4 0.5%, dfil-sub 5 mm, tdep 20 h Tsub 850◦ C, Tfil 2,120-2,150◦ C, CH4 0.8%, dfil-sub 8 mm, tdep 15 h

H2 + CH4

H2 + CH4 + CH2 (OCH3 )2

59 h

175 h

95 h

223 h, 264 h

360

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

H

H

H H Cd

CH3

H CH3

Cd

Cd

Cd

CH2

H H

Cd

H

Cd

H Cd

Cd

Cd

H

H C

Cd

C

H Cd

Cd

Figure 15.5 GDSB mechanism.

CH2 (OCH3 )2 → • CH2 OCH3 + • OCH3 • •

(15.3)

CH2 OCH3 → • CH2 O + • CH3 +

(15.4)

OCH3 + H → CH3 + OH •

•

(15.5)

As shown in Table 15.3, the bond dissociation energy of C–O is only 351 kJ mol−1 , much lower than that of CH3 − H, 435 kJ mol−1 , indicating easier generation of • CH3 by thermal decomposition of CH2 (OCH3 )2 . Moreover, the bond dissociation energy of HO–H is 498 kJ mol−1 , larger than that of H–H, 435 kJ mol−1 , suggesting • OH is more powerful than • H in abstraction of H from the diamond surface. Therefore, addition of CH2 (OCH3 )2 could promote diamond synthesis significantly through the modified mechanism, as shown in Figure 15.6. TABLE 15.3 Bond dissociation energies. Bond

Bond dissociation energy (kJ mol−1 )

References

435 435 498 351 (average)

[34] [34] [34] [35]

CH3 –H H–H HO–H C–O

H

H

H H

Cd

Cd

CH3

H CH3

Cd

C

Cd

Cd

CH2 Cd

OH

H

H

Cd

H H

Cd

H

Cd

OH

Cd

H

H C

H Cd

Cd

Figure 15.6 Modified mechanism of addition of CH3 to the diamond.

15.1 FABRICATION OF STABLE Ti/BBD ELECTRODES

361

2800 2600

1334

2400

Counts

2200 2000 1800 1600 1520 1400 1200 1000 800 1200

1300

1400

1500

1600

Shift (cm–1) Figure 15.7 Raman spectrum of the BDD film deposited using H2 + CH4 + CH2 (OCH3 )2 . (Reprinted from Ref. 31.)

Figure 15.7 shows the Raman spectrum of the BDD film deposited with H2 + CH4 + CH2 (OCH3 )2 [32]. The film obtained has an intensive diamond peak around 1334 cm−1 , whereas the broad amorphous or graphitic sp2 carbon band signals are quite weak. The sp2 carbon content in the film is <0.5%, indicating good quality of the diamond film formed. Figure 15.8 displays the XRD pattern of the Ti/BDD sample prepared with CH2 (OCH3 )2 added [32]. Compared with Figure 15.2, the diamond crystal peak intensities increased, while the TiC peak intensities decreased, indicating that CH2 (OCH3 )2 could really enhance the diamond deposition and inhibit the formation of TiC. The BDD film thickening and the TiC thinning help to improve the Ti/BDD stability. 300

Diam (111)

TiC (111)

250

TiC (200)

150

Ti (101)

50

Ti (100)

100

TiC (220) Ti (103)

TiC (311) Diam (220) TiC (222)

Counts

200

Ti (201)

Diam (311)

0 30 35 40 45 50 55 60 65 70 75 80 85 90 95 2θ (degrees) Figure 15.8 XRD pattern of the BDD film deposited using H2 + CH4 + CH2 (OCH3 )2 . (Reprinted from Ref. 31.)

362

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

Figure 15.9 Surface morphology of Ti/BDD electrode prepared using H2 + CH4 + CH2 (OCH3 )2 . (Reprinted from Ref. 31.)

Figure 15.9 shows the morphology of the BDD film deposited on Ti using H2 + CH4 + CH2 (OCH3 )2 [32]. It was found that the diamond crystallites obtained was smaller than those synthesized using H2 + CH4 . This is attributed to an increase in the nucleus density in the initial period. The reduction of crystalline size can help reduce the porosity of the diamond film. As a result, the penetration of the electrolyte through the BDD film is suppressed, and the electrochemical corrosion rate can be reduced. Figure 15.10 shows the cyclic voltammograms obtained on Ti/BDD electrodes prepared using H2 + CH4 + CH2 (OCH3 )2 [10]. A relatively large voltammetric current was 600

Current density (A m2)

500 400

1 cycle 2 cycles 5 cycles 20 cycles

300 200 100 0 –100 .5

1.0

1.5 2.0 Potential (V) vs NHE

2.5

3.0

Figure 15.10 Cyclic voltammogram development of Ti/BDD electrode prepared using H2 + CH4 + CH2 (OCH3 )2 at a scan rate of 0.1 V s−1 in 0.5 M H2 SO4 solution. (Reprinted from Ref. 9.)

15.1 FABRICATION OF STABLE Ti/BBD ELECTRODES

363

detected in the first few cycles. Then, the voltammetric current decreased quickly. After about 20 cycles, the shape of CV became identical. The large voltammetric current in the first few cycles resulted from an electrochemical “etching” or surface cleaning effect [36]. As shown in Raman spectroscopic studies, the diamond-thin film is composed principally of diamond and a low concentration of sp2 carbon impurities. The nondiamond carbon impurities on the surface may react during voltammetric investigation to form CO and/or CO2 according to the following reactions [36]: ◦

C + H2 O = CO + 2H+ + 2e−

E = 0.548 V vs. NHE

C + 2H2 O = CO2 + 4H+ + 4e−

E = 0.207 V vs. NHE

◦

(15.6) (15.7)

The quick decrease in the voltammetric current with the scan cycles reveals the ease of removing the sp2 carbon impurities electrochemically. This feature is very important for electrooxidation of pollutants. Otherwise, CE will decrease remarkably due to the parasitical current produced. The rapid increase in the anodic current in the high potential region is attributed to the O2 evolution. In the first few scan cycles, the O2 evolution currents detected were relatively large. This was also associated with the sp2 carbon impurities that could serve as active sites for O2 evolution. The voltammograms obtained on the Ti/BDD electrodes prepared using H2 + CH4 + CH2 (OCH3 )2 and H2 + CH4 gas mixtures after reaching steady states are almost identical, with an onset potential of 2.7 V versus NHE for O2 evolution [31]. This indicates that addition of CH2 (OCH3 )2 to the conventional H2 + CH4 gas mixture will not affect the eventual electrochemical properties of the electrodes prepared and good activity for pollutant oxidation is expected. Figure 15.11 shows the voltammograms for the [Fe(CN)6 ]3−/[Fe(CN)6 ]4− redox couple on the Ti/BDD electrode prepared using H2 + CH4 + CH2 (OCH3 )2 [31]. The separation of anodic and cathodic peak potentials was 0.08 V at a scan rate of 0.02 V s−1 and increased to 0.11 V when the scan rate increased to 0.2 V s−1 , demonstrating that the [Fe(CN)6 ]3− /[Fe(CN)6 ]4− redox couple behaved in a quasi-reversible manner on the Ti/BDD electrode. 15.1.5

Methods to Enhance the Service Life of Ti/BDD

A long service life is a desirable property of an electrode. For the Ti/BDD electrode fabricated by HFCVD, it is possible to act as a stable anode when the operating parameters are optimized [37]. In fact, even longer service life may be obtained when the process is modified slightly. One approach tested was to deposit a silicon interlayer between the titanium substrate and the diamond crystals. This Si interlayer will minimize the formation of TiC and facilitate better growth of diamond crystals [38]. By evaporation deposition of a thin layer of Si (200–300 nm) on the pretreated Ti substrate, the accelerated service life of the so formed Ti/Si/BDD was found to increase by over 20% compared with Ti/BDD. The quality of the diamond crystals, Figure 15.12, was found to be better than that directly grown from Ti (as seen in Figure 15.7) with the graphite content almost not detectable. Another approach was to adopt a staged substrate temperature. Specifically, the temperature of the substrate was started at a relatively low value, 650◦ C, to grow a layer

364

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

60 50 d

40 c

Current density (A m2)

30 b

20

a

10 0 –10 –20 –30 –40 –50 –60 0.2

0.3

0.4 0.5 Potential (V) vs. NHE

0.6

0.7

Figure 15.11 Voltammograms obtained on Ti/BDD in 10 mM [Fe(CN)6 ]3- + 10 mM [Fe(CN)6 ]4- + 0.5 M Na2 SO4 solution at different scan rates. (a) 0.02, (b) 0.05, (c) 0.1, and (d) 0.2 V s−1 . 16000 14000

Counts

12000 10000 8000 6000 4000 2000 800 900 1000 1100 1200 1300 1400 1500 1600 1700 shift (cm–1)

Figure 15.12 Raman spectrum of the Ti/Si/BDD film deposited using H2 + CH4 + CH2 (OCH3 )2 . (Reprinted from Ref. 36.)

of stable diamond-like material that has better adhesion to the substrate than diamond crystals. This layer would serves as a cushion to absorb the thermal stress during the cooling of the material so as to avoid the chance of diamond film delaminating from the substrate. The diamond film can grow on top of this cushion layer at a higher substrate temperature, 750◦ C, so as to have a high quality of diamond polycrystalline film [39]. The accelerated service life of the Ti/BDD so obtained was found to be nearly four

15.2 USE OF Ti/BDD ELECTRODES FOR WASTEWATER TREATMENT

365

times of that obtained at a fixed substrate temperature of 750◦ C. The staged substrate temperature is easy to achieve. Thus, it should be a viable option for industrial operation.

15.2 15.2.1

USE OF Ti/BDD ELECTRODES FOR WASTEWATER TREATMENT Oxidation of Acetic and Maleic Acids

Usually, acetic and maleic acids are important intermediates produced in electro-oxidation of complex organic compounds. They are very difficult to destruct on common electrodes such as Pt [40]. Figure 15.13 shows the chemical oxygen demand (COD) variation with the charge loading for oxidation of acetic and maleic acids on the Ti/BDD electrode [31]. It can be seen that that the activity of the Ti/BDD electrode was very good in oxidizing acetic and maleic acids. COD decreased linearly with the charge loading up to about 600 mg L−1 , with the current efficiency (CE) being close to 100%, suggesting that pollutant degradation was under current control in the early stage. This is consistent with the observations of oxidizing formic, acetic, and oxalic acids [41], 4-chlorophenol [42] and 2-naphthol [43] on Si/BDD electrodes. After COD was lowered to a level < 600 mg L−1 , the oxidation rate decreased gradually. The final COD values were only 33 mg L−1 and 46 mg L−1 for the acetic acid and maleic acid, respectively. 15.2.2

Oxidation of Phenol

Phenol is one of the most common toxic organic pollutants present in many industrial effluents. It has been widely used as a model pollutant for different oxidation processes. The result for phenol oxidation on the Ti/BDD electrode is shown in Figure 15.14 [31]. It was found that the initial CE for oxidation of phenol was only about 80%, lower than that obtained in oxidizing acetic and maleic acids, indicating that phenol is more resistant than acetic and maleic acids to oxidation on the Ti/BDD electrode. This could 1200

COD (mg L–1)

1000 800 600 Maleic acid

400 Acetic acid 200 0

0

1

5 2 3 4 Charge loading (Ah L–1)

6

7

Figure 15.13 Oxidation of 1060 mg L−1 acetic acid and 1500 mg L−1 maleic acid at current density 200 A m−2 , reaction temperature 30◦ C, and Na2 SO4 1500 mg L−1 .

366

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

1400 1200

COD (mg L–1)

1000 800 600 400 200 0

0

1

2 3 Charge loading (Ah L–1)

4

5

Figure 15.14 Oxidation of 500 mg L−1 phenol at current density 100 A m−2 reaction temperature 30◦ C, and Na2 SO4 1500 mg L−1 .

be attributed to formation of polymeric intermediate products. It was observed that a brown polymeric film was formed on the Ti/BDD electrode surface after electrolysis for about half an hour. This polymeric film modified the electrochemical properties of the Ti/BDD electrode, leading to a decrease in initial CE. Despite the formation of the polymeric intermediate products, the Ti/BDD electrode could still reduce COD from initial 1,175 mg L−1 to 40 mg L−1 at the end of the reaction, indicating its good activity for phenol oxidation. It should be noted that formation of polymeric products in electro-oxidation process is not a new finding, but a common phenomenon. Many researchers have observed polymeric film formation in oxidizing phenol on different electrodes [44–46]. A reaction mechanism via phenate anions with the formation of a polyoxyphenylene film was proposed [44]. Comninellis and Pulgarin [45] reported that the polymeric film formed had good electrical conductivity. They also pointed out that the formation of the polymeric film depended strongly on the experimental conditions. Alkaline media (pH >9), low current densities (< 300 A m−2 ), high temperatures (≥ 50◦ C ), and high phenol concentrations (≥ 50 mM) favored film formation. The decay of electrode activity is believed to be one of the major problems of electrooxidation. Figure 15.15 demonstrates the COD removal efficiency variation in oxidizing phenol on Ti/BDD [47]. No remarkable drop tendency to the COD removal efficiency was observed after nine experimental runs, indicating that the Ti/BDD electrode had good reproducibility. 15.2.3

Oxidation of Dyes

It is well known that dyes are very resistant to oxidation. Most dyes are biologically refractory organic compounds. In order to know if the Ti/BDD electrodes are effective for dye degradation, Chen and associates [10,48] investigated the electrochemical degradation of different dyes. Except for Orange II, all others are reactive dyes. Figure 15.16 shows

15.2 USE OF Ti/BDD ELECTRODES FOR WASTEWATER TREATMENT

367

COD removal efficiency (%)

100

95

90

85

80

0

1

2

3

4

5

6

7

8

9

10

Cycling number

Figure 15.15 Reproducibility of oxidation of 500 mg L−1 phenol on Ti/BDD at Initial COD 1175 mg L−1 , current density 100 A m−2 , temperature 30◦ C, and charge loading 4.85 Ah L−1 . (Reprinted from Ref. 46.)

1200

COD (mg L–1)

1000 800 Orange II 600 400 Reactive red HE-3B 200 0

0

1

2

3

4

5

6

7

Charge loading (Ah L–1)

Figure 15.16 Oxidation of 750 mg L−1 Orange II and 1500 mg L−1 reactive red HE-3B at current density 200 A m−2 , reaction temperature 30◦ C, and Na2 SO4 1500 mg L−1 . (Reprinted from Ref. 9.)

the results of oxidizing two typical dyes (i.e., Orange II and reactive red HE-3B) on the Ti/BDD electrode. Aqueous Orange II solutions are alkali, whereas aqueous reactive red HE-3B solutions are acidic. Both Orange II and reactive red HE-3B contain sulphonic and hydroxyl groups. Their chemical structures are shown in Figure 15.17. It can be seen that the Ti/BDD electrode also had good activity in oxidizing Orange II and reactive red HE-3B. At a charge loading of 6.25 Ah L−1 , COD was reduced from 1130 mg L−1 to lower than 100 mg L−1 for Orange II, and from 920 mg L−1 to

368

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

OH

N

NaO3S

N

Orange II

N NaO3S

SO3H

N Cl

OH NH

N

C

N

C

C

SO3Na

N

NH

NH

N

C

N

C

N

C

NH

SO3H

OH N

N

Cl NaO3S

SO3Na

Reactive red HE-3B Figure 15.17 Chemical structures of Orange II and reactive red HE-3B.

about 50 mg L−1 for reactive red HE-3B. In addition, it was found that Orange II was easier to destruct than reactive red HE-3B on Ti/BDD. The initial CE for Orange II was close to 100% as obtained for acetic and maleic acids. In contrast, the initial CE for reactive red HE-3B was about 90%. Polymeric products were also observed in anodic oxidation of Orange II and reactive red HE-3B on Ti/BDD electrodes. However, no films were adhered onto the electrodes. The amount of the polymeric products depended on operational conditions. High pH, high temperature, low current density, and low initial concentration tended to hinder formation of polymeric products. The results for oxidizing other dyes, all of which are reactive dyes, are summarized in Table 15.4. Clearly, the Ti/BDD electrode is very effective in degrading various dyes. COD could be reduced from initial 402–980 mg L−1 to 8–93 mg L−1 at a current density of 100 A m−2 . CE ranged from 51 to 90%. The energy consumption was 8.9–17.9 kWh m−3 . After treatment, solutions turned colorless except the one containing Monozol T-blue HFG.

369

REFERENCES

TABLE 15.4 Results of anodic oxidation of other reactive dyes on Ti/BDD. Experimental conditions: 100A m−2 , 30◦ C, 2,000 mg L−1 Na2 SO4 , 25 mL, initial pH 4.70–5.53 [9]. Dyes Cibacron yellow HW200 Cycafiw yellow FLN250 Cycafix navy-blue F2B Monozol black SGRN Monozol blue BRF-150 Monozol red F3B150 Monozol T-blue HFG Monozol yellow F3R150 Procion blue HE-RD Reactive blue R Reactive red HE-7B Samafix red S-3B Samafix yellow S-3R Unicion green S6B Unicion red S-3BF80

15.3

Charge (Ah L−1 )

Initial COD (mg L−1 )

Residual COD (mg L−1 )

CE (%)

Energy (kWh m−3 )

3.52 3.02 2.77 2.52 2.52 2.39 4.03 2.52 3.02 2.90 2.52 2.52 2.52 4.03 2.27

737 610 659 710 634 654 980 667 902 803 402 607 440 711 589

28 31 64 78 71 72 93 74 89 61 19 38 50 8 66

67.5 64.2 72.0 84.0 74.8 81.6 73.7 78.8 90.2 85.7 51.0 75.6 51.8 58.4 77.2

14.4 9.7 8.9 10.3 10.3 10.2 17.9 10.9 13.9 9.3 11.5 11.1 10.6 16.7 9.7

CONCLUSIONS

The low-pressure conversion of carbon to diamond crystals has made it possible to grow a thin layer of diamond film on suitable substrates. The doped diamonds, especially boron-doped diamond (BDD), have been tested extensively as promising electrodes for various applications. Hot filament chemical vapor deposition (HFCVD) is an industrially viable technique to fabricate active and stable diamond-based electrodes using titanium as a substrate material. The damage of the diamond film is usually caused by the residual stress resulting from the different thermal expansion coefficients between the diamond and the substrate. When the process is optimized, the Ti/BDD service life can be sufficiently increased. Ti/BDD has proven activities in the application as an anode for electro-oxidation of pollutants in wastewater such as phenol, dyes, maleic acid, and so on.

REFERENCES 1. C. Comninellis, G. Chen, Electrochemistry for the Environment , Springer, New York, 2009. 2. G. Chen, Sep. Purif. Technol. 2004, 38 , 11–41. 3. T.N. Rao, A. Fujishima, J.C. Angus, “History Survey of Diamond Electrodes,” in Diamond Electrochemistry (A. Fujishima, Y. Einaga, T.N. Rao, D. A. Tryk, eds.), Elsevier, New York, 2005, pp. 1–10. 4. W.G. Eversole, US Patent 3,030,187 and 3,030,188, 1962. 5. J.C. Angus, H.A. Will, W.S. Stanko, J. Appl. Phys. 1968, 39 , 2915–2922. 6. H.B. Beer, US Patent 3,632,498, 1972. 7. Fujishima et al. is Diamond Electrochemistry (A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, eds.), Elsevier, New York, 2005. 8. T.A. Ivandini, Y. Einaga, K. Honda, A. Fujishima, “Preparation and Characterization of Polycrystalline Chemical Vapor Deposited Boron-Doped Diamond Thin Films,” in Diamond Electrochemistry (A. Fujishima, Y. Einaga, T.N. Rao, D. A. Tryk, eds.), Elsevier, New York, 2005, pp. 11–25.

370

FABRICATION AND APPLICATION OF TI/BDD FOR WASTEWATER TREATMENT

9. L. Guo, X.Y. Li, G. Chen, “Techniques of Electrode Fabrication,” in Electrochemistry for the Environment (C. Comninellis, G. Chen, eds.), Springer, New York, 2009, pp. 55–98. 10. X.M. Chen, G. Chen, P.L. Yue, Chem. Eng. Sci . 2003, 58 , 995–1001. 11. S. Matsumoto, Y. Sato, M. Kakmo, N. Setaka, J. Mater. Sci . 1982, 17 , 3106–3112. 12. S. Matsumoto, Y. Sato, M. Kakmo, N. Setaka, Jpn. J. Appl. Phys. 1982, 2 , L183–L185. 13. P.O. Joffreau, R. Haubner, B. Lux, J. Refract. Hard Mater. 1988, 7 , 186–194. 14. X.M. Chen, G. Chen, J. Electrochem. Soc. 2004, 151 , B214–B219. 15. W.D. Fan, K. Jagannadham, J. Narayan, Surf. Coat. Technol . 1997, 91 , 32–36. 16. E. Buccioni, E. Braca, J.M. Kenny, M.L. Terranova, Diam. Rel. Mat . 1999, 8 , 17–24. 17. X.L. Peng, T.W. Clyne, Thin Solid Films 1997, 93 , 261–269. 18. J.W. Ager III, M.D. Drory, Phys. Rev. B. 1993, 48 , 2601–2607. 19. D.S. Knight, W.B. White, J. Mater. Res. 1989, 4 , 385–393. 20. V. Fisher, D. Gandini, S. Laufer, E. Blank, C. Comninellis, Electrochim. Acta 1998, 44 , 521–524. 21. F. Hentschel, I. Schmidt, C. Benndorf, Thin Solid Films 1996, 290–291 , 196–199. 22. M. Fryda, D. Herrmann, L. Schafer, C.P. Klages, A. Perret, W. Haenni, C. Comninellis, D. Gandini, New Diamond Front. Carbon Technol . 1999, 9 , 229–240. 23. Y. Saito, K. Sato, K. Tanaka, K. Fujita, S. Matuda, J. Mater. Sci . 1988, 23 , 824–846. 24. J.A. Mucha, D.L. Flamm, D.E. Ibbotson, J. Appl. Phys. 1989, 65 , 3448–3452. 25. C.F. Chen, Y.C. Huang, S. Hosomi, I. Yoshida, Mater. Res. Bull . 1989, 24 , 87–94. 26. Y. Hirose, Y. Terasawa, Jpn. J. Appl. Phys. 1986, 25 , L519–L521. 27. Y. Saito, K. Sato, K. Gomi, H. Miyadera, J. Mater. Sci . 1990, 25 , 1246–1250. 28. F.C.N. Hong, J.C. Hsieh, J.J. Wu, G.T. Liang, J.H. Hwang, Diam. Rel. Mat . 1993, 2 , 365–372. 29. I. Schmidt, F. Hentschel, C. Benndorf, Solid State Ionics 1997, 101–103 , 97–101. 30. I. Schmidt, C. Benndorf, Diam. Rel. Mat . 1998, 7 , 266–271. 31. X.M. Chen, High-Performance Electrodes for Wastewater Treatments, Ph.D. dissertation, the Hong Kong University of Science and Technology, 2002. 32. X.M. Chen, G. Chen, F.R. Gao, P.L. Yue, Environ. Sci. Technol. 2003, 37 , 5021–5026. 33. B.J. Garrison, E.J. Dawnkaski, D. Srivastava, D.W. Brenner, Science 1992, 255 , 835–838. 34. L.G. Wade, Organic Chemistry, Prentice-Hall, Inc., New Jersey, 1999, p. 142. 35. M. Lazar, J. Rychly, V. Klimo, P. Pelikan, L. Valko, in Free Radicals in Chemistry and Biology, CRC Press, Inc., Florida, 1989, p. 9. 36. G.M. Swain, J. Electrochem. Soc. 1994, 141 , 3382–3393. 37. L. Guo, G. Chen, Diam. Rel. Mat . 2007, 16 , 1530–1540. 38. Y. Tian, X.M. Chen, C. Shang, G. Chen, J. Electrochem. Soc. 2006, 153 , J80–J85. 39. L. Guo, G. Chen, J. Electrochem. Soc. 2007, 154 , D657–D661. 40. S. Stucki, R. Kotz, B. Carcer, W. Suter, J. Appl. Electrochem. 1991, 21 , 99–104. 41. D. Gandini, E. Mah´e, P.A. Michaud, W. Haenni, A. Perret, C. Comninellis, J. Appl. Electrochem. 2000, 30 , 1345–1350. 42. M.A. Rodrigo, P.A. Michaud, I. Duo, M. Panizza, G. Cerisola, C. Comninellis, J. Electrochem. Soc. 2001, 148 , D60–D64. 43. M. Panizza, P.A. Michaud, G. Cerisola, Ch. Comninellis, J. Electroanal. Chem. 2001, 507 , 206–214.

REFERENCES

44. 45. 46. 47. 48.

G. Mengoli, S. Daolio, M.M. Musiani, J. Appl. Electrochem. 1980, 10 , 459–471. C. Comninellis, C. Pulgarin, J. Appl. Electrochem. 1991, 21 , 703–708. N.B. Tahar, A. Savall, J. Electrochem. Soc. 1998, 145 , 3427–3434. X.M. Chen, F.R. Gao, G. Chen, J. Appl. Electrochem. 2005, 35 , 185–191. X.M. Chen, G. Chen, Sep. Purif. Technol. 2006, 48 , 45–49.

371

16 Application of Diamond Films to Water Disinfection Jessica H. Bezerra Rocha and Carlos A. Mart´ınez-Huitle

16.1

INTRODUCTION

The many problems worldwide associated with the lack of clean, fresh water are well known: 1.2 billion people lack access to safe drinking water, 2.6 billion have little or no sanitation, millions of people die annually—3900 children a day—from diseases transmitted through unsafe water or human excreta1. Countless more are sickened from disease and contamination. Intestinal parasitic infections and diarrheal diseases caused by waterborne bacteria and enteric viruses have become a leading cause of malnutrition owing to poor digestion of the food eaten by people sickened by water [1–3]. In both developing and industrialized nations, a growing number of contaminants are entering water supplies from human activity: from traditional compounds such as heavy metals and distillates to emerging micro-pollutants such as endocrine disrupters and nitrosoamines. Increasingly, public health and environmental concerns drive efforts to decontaminate waters previously considered clean. More effective, lower-cost, robust methods to disinfect and decontaminate waters from source to point-of-use are needed, without further stressing the environment or endangering human health by the treatment itself. Water also strongly affects energy and food production, industrial output, and the quality of our environment, affecting the economies of both developing and industrialized nations [1]. In the coming decades, water scarcity may be a watchword that prompts action ranging from wholesale population migration to war, unless new ways to supply clean water

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

373

374

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

are found. Fortunately, a recent flurry of activity in water treatment research offers hope in mitigating the impact of impaired waters around the world. Conventional methods of water disinfection, decontamination, and desalination can address many of these problems with quality and supply [1]. However, intensive chemical treatments (such as those involving ammonia, chlorine compounds, hydrochloric acid, sodium hydroxide, ozone, permanganate, alum and ferric salts, coagulation and filtration aids, anti-scalants, corrosion control chemicals, and ion exchange resins and regenerants) and residuals resulting from treatment (sludge, brines, toxic waste) can add to the problems of contamination and salting of freshwater sources. Fortunately, there is much more that science and technology can do to mitigate environmental impact and increase efficiency because current treatment methods are still far from natural-law limits in their ability to separate compounds, deactivate or remove deleterious pathogens and chemical agents, transport water molecules, and move ions against concentration gradients [1].

16.2

DISINFECTION WATER

Overarching goals for providing safe water is affordably and robustly to disinfect water from traditional and emerging pathogens, without creating more problems due to the disinfection process itself. Waterborne pathogens have a devastating effect on public health, especially in the developing countries. Waterborne infectious agents responsible for these diseases include a variety of helminthes, protozoa, fungi, bacteria, rickettsiae, viruses, and prions [4]. Although some infectious agents have been eradicated or diminished, new ones continue to emerge, and so disinfecting water has become increasingly more challenging. Viruses are of particular concern, accounting, together with prions, for nearly half of all emerging pathogens in the last two to three decades [1,5]. Enteric viruses received less attention in the past compared with bacterial pathogens (e.g., Vibrio cholerae) and protozoan parasites (e.g., Cryptosporidium parvum), partly because they were difficult to detect and partly because free chlorine was very effective in inactivating them. This method is the most popular for drinking water disinfection by the addition of chlorine and/or chlorine by-products that are able to eliminate all harmful microorganisms in several areas. Despite the great effectiveness of chlorine as a water disinfection method, there are disadvantages such as unfavorable taste and odor, its ineffectiveness when used alone against some resistant microorganisms, and the generation of products potentially toxic such as trihalomethanes (chlorinated compounds that are mutagenic) [6,7] and chloroform (the most common chemical by-product of water disinfection, considered harmful for its cancer risks) [7–11]. Therefore, the effective control of waterborne pathogens in drinking water calls for the development of new disinfection strategies, including multiple-barrier approaches that provide reliable physicochemical removal (e.g., coagulation, flocculation, sedimentation, electrochemical systems, and media or membrane filtration) along with effective photon-based and/or chemical inactivation. A number of these unwanted disinfection by-products are a major health concern because of their carcinogenic properties. The World Health Organization has set guidelines for these compounds based on an excess cancer risk. In addition, it should be noted that several epidemiological studies have been carried out to investigate the possible carcinogenic properties of chlorinated drinking water, but the International Agency for the Research on Cancer considered that the degree of evidence for an association between chlorination and the occurrence of cancer from these studies is inadequate [12,13]. In the

16.3 SCIENCE AND TECHNOLOGY FOR WATER PURIFICATION

375

case of the chloroform, the controversy began in March 1998, when the Environmental Protection Agency released new data on disinfection by-products, and announced that it was considering changing the goal for chloroform contamination in drinking water from zero to 300 mg L−1 , supposing the first acknowledgment of a threshold dose for a regulated carcinogen [14,15]. But, at the end of 1999, the agency retreated from this position, and set the goal at zero (final ruling). As a result of all these disadvantages, a high number of alternatives to chlorine for drinking water disinfection have been proposed.

16.3

SCIENCE AND TECHNOLOGY FOR WATER PURIFICATION

The most interesting alternatives to chlorine include [1]: (1) chemical systems such as ozone, silver, copper, ferrate, iodine, bromine, hydrogen peroxide and potassium permanganate [16–18]; (2) physico-chemical systems such as titanium photocatalysis and photodynamic disinfection [19–21]; (3) electrochemical disinfection; and (4) physical systems such as ultraviolet irradiation, ultrasonication, pulsed electric fields, irradiation, magnetic enhanced disinfection, and microwave systems [1,22,23]. Whereas ozone and ultraviolet irradiation have gained acceptance within the water treatment process, most of the other alternatives at present do not fulfill the requirements for primary and residual drinking water disinfection [1,7]. Sequential disinfection schemes such as UV/combined chlorine and ozone/combined chlorine are being considered by many drinking-water utilities as the inactivation component of their multi-barrier treatment plants because, compared with free chlorine, both UV and ozone are very effective in controlling C. parvum oocysts. In addition, combined chlorine can provide a residual in distribution systems without forming high levels of regulated DBPs [1]. However, changing disinfection technologies has raised new concerns because viruses, although effectively controlled by ozone, are resistant to both UV and combined chlorine disinfection. Moreover, ozone can form the DBP carcinogen bromate ion in water containing bromide ions, and combined chlorine can form other unregulated disinfection by-products (DBPs)—for example, haloacetonitriles and iodoacetic acid [24,25]—that may be more toxic and carcinogenic than those associated with free chlorine. In contrast, electrochemical disinfection has emerged as one of the most promising alternatives to chlorine, providing both primary and residual disinfection [7]. In recent years, effective electrochemical disinfection systems for conventional water treatment have been developed. The advantages of these procedures make them more attractive than other methods. The electrochemical technology is environmentally friendly, low in cost, easily operated, and known to inactivate a wide variety of micro-organisms from bacteria to viruses and algae [6,7,26]. The most useful systems for electrochemical disinfection of drinking water are based on the electrogeneration of disinfecting agents. Other technologies such as the electrosorption of bacteria on the electrode surface [27], electrocution [28], and electrophoresis [29,30] have also been explored. The potential use of electrochemical methods for disinfection has been discussed since the 1950s, but systems other than that electro-generating chlorine [6,7,31–33] have yet to gain widespread acceptance within the water industry. Although the mechanism of electrochemical disinfection using chloride-containing solutions, so-called electrochlorination, has been mainly attributed to the action of electro-generated active chlorine, conflicting research concerning the generation of other

376

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

disinfecting agents in water treated with these systems has been considered. However, the debate as to whether electrochemical systems can replace chlorine is still open.

16.4

ELECTROCHEMICAL DISINFECTION/PURIFICATION SYSTEMS

Numerous electrochemical systems and electrode materials have been proved against a variety of microorganisms and their effectiveness to the abatement of bacteria, viruses, and protozoa is largely dependent on the electrochemical reactor, anode material, electrolyte composition, and electrolysis conditions. Table 16.1 summarizes these parameters for the most relevant research in the frame of the inactivation of microorganisms in chloride-containing waters by electro-chlorination [34–42]. In addition, Table 16.1 also indicates the estimation of energy consumption based on the data collected from each research developed in the frame of the inactivation of micro-organisms. Similar data obtained for tap waters with a very low chloride content (<4 mg L−1 ) or free-chlorine waters [7,43–52] are collected in Table 16.2. Application of alternating current or potential, pulse voltage, and mainly direct current to a large variety of undivided electrochemical cells can be observed. As anode material, which is the most important parameter in the disinfection process, Ti, TiN, activated carbon fiber, carbon cloth, Pt, Pt–Nb, mixed metal oxides of Ir and/or Ru (IrO2 , Ti/RuO2 , Ti/IrO2 –TiO2 , Ti/RuO2 –TiO2 , Ti/IrO2 –Sb2 O5 –SnO2 ), and conductive boron-doped diamond films (Nb/BDD, Si/BDD) have been utilized. Tables 16.1 and 16.2 also show the electrolyte and concentration of cells treated, the applied anode potential, cell voltage and/or current density, and the kind of bacteria, viruses, and algae inactivated. The most popular method of electrochemical disinfection is electrochlorination. Its main advantage is the onsite generation of disinfectants, thus avoiding the problems of common chlorination such as transport and storage of dangerous chlorine [26]. There are two types of electro-chlorination procedures involving either the synthesis of free chlorine from brine in an electrolytic generator or the direct production of oxidants from the water to be treated through the electrolyser (see Table 16.1). Active chlorine species such as Cl2 , HOCl, OCl− , and ClO2 have been widely recognized as key oxidants responsible for inactivating cells in electrochlorination. These species can be produced at the anode via the following total reactions [6]: H2 O + Cl− → ClOH • + H+ + 2e− H2 O + ClOH • + Cl− → Cl2 + O2 + 3H+ + 4e− −

−

−

(16.1) (16.2)

Cl2 + 2OH → H2 O + OCl + Cl

(16.3)

ClOH • + Cl2 → ClO2 + 3H+ + 2Cl− + e−

(16.4)

Some researchers have pointed out that the disinfecting efficacy of this method is much higher than chlorination due to the competitive electrogeneration of other oxidants. Thus, Venczel et al. [54] found more rapid inactivation kinetics for Escherichia coli , the rugose strain of V. cholerae, Clostridium perfringens spores, and bacteriophage MS2 in pH 6–10 with onsite electro-generated oxidants from brine than with free chlorine. Similarly, Son et al. [53] reported the better disinfecting efficacy of electrochemically generated oxidants

377

Pt wire

Direct current Stirred tank reactor

87 cm2 Ti/RuO2 -TiO2 rod

Ti/RuO2 rod (260 mm length, 5 mm diameter)

1.35 cm2 Ti/TiO2 foil irradiated with a 150 W Xe lamp

Electrochemical assisted photocatalysis Stirred tank reactor with a quartz window for anode illumination

Tubular reactor

30 cm2 Ti/RuO2 plate

Pulse voltage Flow-through cell

Anode

600 mL of 3 × 106 CFU mL−1 algal suspension in water with Cl− (pH ≈ 7) flowing in batch. 1–10 mA cm−2 (cell voltage of 3.5–9.2 V). 265 mL of deionized water with bacterial suspension and up to 0.1 M NaCl. 11 mA cm−2

10 mL of EE buffer (pH = 8.3)c with a suspension of 103 CFU mL−1 . 25–350 mA (cell voltage of 25–350 V).

10 mL of 5 × 105 CFU mL−1 bacterial suspension in a Ringers solutionb. 1 V vs. SCE.a

500 mL of water with germinated brown rice and NaCl (pH = 5.5) containing 102 −107 CFU mL−1 treated in batch at 87 mL min−1 . Cell voltage of 1.0 or 1.5 kV at 5 Hz.

Experimental Conditions

Microorganisms inactivated in chloride-containing waters by electrochlorination.

Undivided electrochemical cell

TABLE 16.1

0.08

0.05

0.332

0.015

100

Energy consumption (kWh m−3 )

[38]

[37]

[36]

[35]

[34]

Ref.

(continued overleaf )

Escherichia coli

Escherichia coli, Pseudonomas aeruginosa, bacteriophage MS2 alga Mycrocystis aerunigosa

Escherichia coli

Legionella

Inactivated microorganisms

378 919.6 mm2 Ti/IrO2 -Sb2 O5 SnO2 pellets

522 cm2 Pt-Nb mesh

65 cm2 Si/BDD plate

Typical dual-electrode cell

Zappi™ cell

DiaCell® reactor

Potable water with 50 mg L−1 Cl− , 240 mg L−1 SO4 2− and a suspension of 105 –107 CFU mL−1 in continuous at 3 L min−1 . 0.5–4 A (cell voltage of 4–14 V). Synthetic solutions with a suspension of 107 −108 CFU mL−1 and 0.016-0.032% or 0.5–1.0% in weight of NaCl. 0–2 A (cell voltage of 0 to 18 V). 10 L of contaminated 0.010 M NaCl in batch at 6 L min−1 . 4 mA cm−2 (cell voltage of 5 V). Tap water with 75 mg L−1 Cl− or deionized water with 330 mg L−1 NaCl in continuous at 160 L h−1 and further used for disinfection. 25–150 mA cm−2.

Experimental Conditions

a Applied anode potential in a three-electrode cell. b Solution composed of 2.25 g L−1 NaCl, 0.105 g L−1 KCl, 0.120 g L−1 CaCl and 50 mg L−1 NaHCO . 2 3 c Buffer composed of 30 mM Tris and 150 mM KCl.

30 cm2 Ti/IrO2 -TiO2 plate

Anode

Flow-through cell

Undivided electrochemical cell

TABLE 16.1 (Continued)

0.8

1

0.48

0.3

Energy consumption (kWh m−3 )

Escherichia coli, bacteriophage MS2 Legionella pneumophila

bacteriophage MS2

Bacillus subtilis, Escherichia coli, Saccharomyces cerevisiae

Inactivated microorganisms

[41] [42]

[7]

[40]

[39]

Ref.

379

Direct current Flow-through cell

Flow-through cell

Alternating current or potential Stirred tank reactor

Activated carbon fiber (18 mm diameter, 100 mm length, 5 mm thickness)

260 or 1170 cm2 carbon-cloth sheet

2 cm2 TiN mesh

Activated carbon fiber (34 mm diameter, 100 mm length, 9 mm thickness)

25 cm2 Ti plate

Anode 350 mL of tap water (pH = 8.1) with a suspension up to 26, 800 cell mL−1 . 2.5–5.0 mA cm−2 (cell voltage of 45–100 V) at 0.5 Hz. Continuous tap water with a suspension of 2.3 × 103 cell mL−1 flowing at 300 mL min−1 . 1.0 V vs. SCE for 20 min and cycling from 0.2 to −0.8 V vs. SCE for 10 min.a Continuous tap water with a suspension of 73 cell mL−1 flowing at 15 mL min−1 . 1.2 V vs. Ag/AgCl for 60 min and -0.6 V vs. Ag/AgCl for 30 min.a Continuous tap water with a suspension of 102 cell mL−1 . 0.5–0.7 V vs. SCE.a Continuous drinking water with 22 cell mL−1 flowing at 2 mL min−1 for 12 h. After stopping for 24 h, it started again at 1 mL min−1 for 6 h. 0.8 V vs. SCE.a

Experimental Conditions

Escherichia coli

(continued overleaf )

[47]

[46]

[45]

[44]

Escherichia coli

Aenomas hydrophila, Bacillus subtilis, Escherichia coli, Saccharomyces cerevisiae, Klebsiella pneumoniae, Pseudonomas cepacia, Pseudonomas fluorescens Escherichia coli

[43]

Ref.

Coliforms

Inactivated microorganisms

Inactivated microorganisms in tap waters with very low chloride content or in free-chlorine waters by electrochemical disinfection.

Undivided electrochemical cell

TABLE 16.2

380

Tubular reactor

Stirred tank reactor

Undivided electrochemical cell

TABLE 16.2 (Continued)

87 cm2 Ti/RuO2 -TiO2 rod

30 cm2 Si/BDD plate

6 cm2 Nb/BDD plate or 5 cm2 Pt sheet

4.6 cm2 Pt sheet

Anode 50 mL of 0.1 M phosphate buffer (pH = 7.1) with a suspension of 2 × 106 CFU mL−1 . 0.1–1 A. 80 mL of 0.2 M phosphate buffer (pH = 7.1) with a suspension of 105 CFU mL−1 . 0.1–100 mA cm−2 1 mM Na2 SO4 with a suspension of about 102 CFU mL−1 flowing in batch or continuous up to 100 mL min−1 . 1.5–13.3 mA cm−2 (2.8–3.1 V vs. SCE).a 265 mL of deionized water with bacterial suspension and 0.01 M NaNO3 or 0.1 M Na2 SO4 . 11 mA cm−2

Experimental Conditions

Escherichia coli

Escherichia coli, Enterococcus faecalis, coliform Enterobacter, coliform Acinectobacter

Escherichia coli

Saccharomyces cerevisiae

Inactivated microorganisms

[38]

[51]

[49] [50]

[48]

Ref.

381

(Continued)

IrO2 , Pt, BDD-Diamond®

Not specified

a Applied anode potential in a three-electrode cell.

65 cm2 Si/BDD plate

DiaCell® reactor

Pt-Nb mesh

Anode 522

cm2

Zappi™ cell

Undivided electrochemical cell

TABLE 16.2

10 L of contaminated 0.030 M Na2 SO4 or 0.036 M NaH2 PO4 flowing in batch at 6 L min−1 . 24–27 mA cm−2 (cell voltage of 5 V). Tap water or deionized water with 476 mg L−1 NaHCO3 or 440 mg L−1 Na2 SO4 in continuous at 160 L h−1 and further used for disinfection. 25–150 mA cm−2 Tap water with 1.4 × 108 CFU of bacterial suspension and glucose (9 g O2 L−1 of COD).

Experimental Conditions

Escherichia coli

Legionella pneumophila

Escherichia coli, bacteriophage MS2

Inactivated microorganisms

[52]

[41] [42]

[7]

Ref.

382

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

than free chlorine for E. coli and Bacillus subtilis spores at pH 8.2 considering the same content of total oxidants. Recent studies have attributed the higher disinfecting power of electrochlorination to the oxidant role of reactive oxygen species (ROS) such as hydroxyl radical ( • OH), atomic oxygen ( • O), hydrogen peroxide, and ozone, which can be generated from water discharge at the anode as follows [51,55,56]: H2 O → • OH + H+ + e−

(16.5)

OH → • O + H+ + e−

(16.6)

•

2 O → O2

(16.7)

2 OH → H2 O2

(16.8)

•

•

O2 + O → O 3 •

(16.9)

•

OH is the second-most oxidizing species known after fluorine, with a high standard potential (E ◦ = 2.8 V versus NHE) that ensures its fast reaction with most organics. However, the short lifetime of • OH and other ROS in solution only makes it possible to underline their possible role in disinfection using direct current. Liang et al. [37] utilized a tubular electrochemical cell with a Ti/RuO2 anode (see Figure 16.1) for the batch treatment of 600 mL of a chloride aqueous solution containing a suspension of the alga Microcystis aeruginosa at pH about 7. Under these conditions, the cell content of the algal suspension decreased rapidly and proportionally to the current density and electrolysis time. After 52 min of treatment at 10 mA cm−2 , the population of M. aeruginosa was reduced from 3 × 106 to 0.6 × 106 colony-forming units (CFU) mL−1 . As can be seen in Figure 16.2, scanning electron microscopy revealed surface damage and apparent leakage of intracellular contents after electrochemical disinfection. As a result, chlorophyll-a was released from the cells and degraded up to 96% by electrochemically generated oxidants, similarly to that obtained from ozone oxidation [37]. The kinetics of the different series/parallel steps involved in Reactions (16.1) through (16.9) depends on the anode material; and this determines the predominant oxidant species

Tubular reactor DC power source

Stirring apparatus

Iron cathode Insulated layer

+

Flowmeter

Ti/RuO2 anode

–

Reservoir

A Ammeter

Pump

Figure 16.1 Scheme of the setup utilized for the inactivation of the alga Microcystis aeurinogosa using a tubular electrochemical cell with a Ti/RuO2 anode. (Reprinted with permission from Ref. 37.)

16.4 ELECTROCHEMICAL DISINFECTION/PURIFICATION SYSTEMS

(a)

383

(b)

Figure 16.2 Scanning electron micrographs of Microcystis aeurinogosa obtained after (a) and before (b) disinfection using the electrochemical tubular cell of Figure 16.1. (Reprinted with permission from Ref. 37.)

produced. Based on this assertion, Kraft [57] showed that the electrochemical production of free chlorine at IrO2 and IrO2 /RuO2 electrodes is higher than BDD and Pt anodes, under comparable conditions. A similar finding was recently reported by Scialdone et al. [58]. Then, mixed metal oxides of Ir and/or Ru are preferable as anodes for electrochlorination to enhance the generation of active chlorine species as the main disinfectants [6]. New electrode materials were developed in order to avoid the production of disinfection by-products and disinfecting hazardous agents. They could be engineered into flow-through reactors for high-throughput systems. The configuration and associated cost of such systems could make them economically viable for applications ranging from large water treatment plants supplying potable and nonpotable water to point-of-use systems with segregated lines dedicated to human consumption and hygiene. Recently, BDD thin films electrodes were found to be particularly attractive as anodes due to their outstanding properties, including large potential window, low species adsorption, corrosion stability in very aggressive media, high efficiency in oxidation processes, and very low double-layer capacitance and background current, which are significantly different from those of other conventional anodes such as Pt, PbO2 , doped and undoped SnO2 , IrO2 , RuO2 , and so on [42,55]. Diamond films are suitable materials for some industrial applications such as chemical synthesis, electroanalysis, and sensors and biosensors, although they have been mainly applied in anodic oxidation to destroy refractory organic pollutants or toxic substances for wastewater treatment [52,57–63]. BDD anode is also able to produce much more appreciable amounts of ROS and other oxidizing species such as peroxodisulfate, peroxodicarbonate, and peroxodiphosphate coming from the oxidation of ions present in the solution, also allowing a fast and permanent disinfection (see Reactions (16.10) through (16.12)) [6,64–67]. 2HSO4 − → S2 O8 2− + 2H+ + 2e−

(16.10)

2HCO3 − → C2 O6 2− + 2H+ + 2e−

(16.11)

2PO4

3−

→ P2 O8 + 2e 4

−

(16.12)

384

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

Unlike PbO2 , SnO2 , and TiO2 , BDD thin films deposited on Si, Ta, Nb, and W by chemical vapor deposition have shown excellent electrochemical stability [57–61]. The application of BDD electrodes for wastewater treatment has been mostly studied with Si-supported devices, in spite of the difficulties related to their fragility and relatively low conductivity of the Si substrate. Although BDD films synthesized on Nb, Ta, and W are promising, their large-scale preparation is impossible due to the unacceptably high costs of these metal substrates. The industrial use of diamond films for wastewater treatment then seems to be too long a time until a high-quality support for industrial scale to be obtained. A possible alternative is titanium, which possesses all required features to be a good substrate material, and Ti/BDD anode has been already used for the destruction of some pollutants. However, diamond deposition on Ti needs to be strongly improved because cracks appear and cause the detachment of the diamond film during long-term electrolysis [55,62]. For these reasons, Si/BDD electrodes have been proposed in the last years, as material to drinking water disinfection, where minor anode dimensions are required in comparison to a wastewater treatment plant.

16.5

DIAMOND FILMS FOR DRINKING WATER DISINFECTION

Electrochemical disinfecting methods with generation of oxidants at diamond films are still under investigation. Nevertheless, the efficient direct in situ production of common chlorine based disinfection agents [41,42,57,68], along with the high generation of ROS [57,69], via Reactions (16.1) through (16.9) may achieve more accurate dosage and simplifies the handling of chemicals. Electrochemical production of oxidants at the diamond surface can thus be exploited for the disinfection of drinking water and removal of color and odor to prevent waterborne diseases. Tables 16.1 and 16.2 illustrate the most relevant studies performed in this way under different conditions. A DiaCell® reactor is a typical example of electrochemical disinfection reactor (Figure 4 in Ref. 6), which, with a Si/BDD anode, has been applied to prepare electrolyzed water with residual oxidants for disinfection of Legionella pneumophila at 104− 106 CFU mL−1 [41,42]. The cell operated in continuous by circulating either tap water without or with addition of 75 mg L−1 Cl− or deionized water with 330 mg L−1 NaCl, 476 mg L−1 NaHCO3 or 440 mg L−1 Na2 SO4 at 160 L h−1 . Total inactivation of Legionella cells (>90%) was reached when the tap water was electrolyzed at more than 150 mA cm−2 and the contact time was longer than 1 hour. The bacteria abatement in tap water (with 3.5 mg L−1 Cl− ) was at least three times faster with the electrochemical disinfection from the diamond cell than with conventional chlorine dosing. A low level of electro-generated oxidant (<1 mg L−1 ) was sufficient for a rapid disinfection. The inactivation efficacy increased gradually as the electrolyzed water contained more chloride, even at low current densities. Using 80 mg L−1 Cl− , for example, Legionella cells were completely inactivated by applying a current density as small as 50 mA cm−2 with contact times of 1 min. Bicarbonate solutions electrolyzed in the diamond cell also inactivated the bacteria due to the formation of a low content of oxidant peroxodicarbonate from Reaction (16.11). The generation of this oxidant can then explain the rapid Legionella abatement attained with electrolyzed tap water, which contains a high HCO3 − concentration of 324 mg L−1 . In contrast, water electrolyzed with sulfate had not impact on Legionella cells because of the low oxidizing power of peroxodisulfate produced via Reaction (16.10). Tr¨oster et al. [52] reported the better performance of a diamond anode in comparison to common

16.5 DIAMOND FILMS FOR DRINKING WATER DISINFECTION

385

electrode materials like Pt and IrO2 for the treatment of a solution containing 1.4 × 108 CFU of E. coli and glucose with a chemical oxygen demand (COD) of 9 g O2 L−1 . The use of diamond anode not only yielded a considerable reduction in bacteria population but it also caused a simultaneous removal of COD by combustion of the sugar. Both effects can be related to the action of large amounts of strong oxidants such as ROS formed from Reactions (16.5) through (16.9), indicating that the generation of these species at diamond electrodes can largely improve disinfection and simultaneous decontamination of waters. Conversely, Haenni et al. [68] showed that the DiaCell® reactor can be efficiently utilized for the disinfection of chloride-containing swimming pool water. The Si/BDD anode exhibits continuous chlorine productivity and higher disinfection performance against bacteria in comparison to directly addition of NaOCl in water. Other interesting electrochemical applications with diamond films involve the disinfection of water circuits and process water in industries and energy supply, humidifiers in air-conditioning systems, cooling towers (inactivation of algae, Legionella and germs), warm water systems in hotels and hospitals (Legionella removal), biologically cleaned wastewater (sewage), ballast water, and medical instruments [52,70,71]. As can be observed in Tables 16.1 and 16.2, many studies have examined the electrochemical disinfection-generating oxidants (with or without chlorine in solution) limiting to a single electrode material without comparison to other electrodes and were therefore unable to demonstrate the effect of electrode material on the generation of the oxidants comparatively. For this reason, Jeong, Kim, and Yoon [72] examined the role of electrode material on the generation of oxidants, and they elucidated the different reaction pathways for generating individual oxidants by employing BDD, Ti/RuO2 , Ti/IrO2 , Ti/Pt–IrO2 , and Pt as anode materials. The efficiency of • OH production, as determined by para-chlorobenzoic acid (pCBA) degradation, was in the order of BDD  Ti/RuO2 ≈ Pt, while no significant production of • OH was observed at Ti/IrO2 and Ti/Pt–IrO2 . The large difference in the • OH production shown in Figure 16.3 was attributed to the different nature of the electrode materials employed (active and nonactive—see Chapter 15). Thus, it has been suggested that the reaction pathway of • OH, after it is produced at the electrode surface (M), strongly depends on the chemical interaction between the electrode surface and • OH. This radical on a nonactive electrode preferentially diffuses to the bulk solution to react with any oxidizable species adsorbed on or in the vicinity of electrode surface. For this reason, BDD, PbO2 , and SnO2 -based DSA anodes have been reported as the nonactive electrodes. The finding that the pCBA degradation was much faster at BDD than that at the other electrodes, as shown in Figure 16.3, may be explained by the nonactive characteristic of BDD, which allows • OH to participate in degradation of pCBA in the bulk solution. Jeong and colleagues [49,50,72] also demonstrated that the • OH plays a key role in O3 generation at BDD surface, contrary to the other electrode materials. In the case of BDD, the O3 concentration gradually increased to a maximum of 1.1 mgL−1 for 20 min of electrolysis, compared to below 0.1 mg L−1 in the other electrodes. Then, the researchers confirmed that BDD exhibits a much higher efficiency in O3 generation than the other electrodes, adding an excess of t-BuOH (0.05 M) in solution during O3 production. And it remarkably inhibited the formation of O3 at BDD, which confirmed the key role played by • OH in the generation of O3 at BDD. In the case of the production of active chlorine, Jeong, Kim, and Yoon [72] showed that it was in the order of Ti/IrO2 > Ti/RuO2 > Ti/Pt–IrO2 > BDD > Pt, which agreed

386

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

1.0

0.6

0

10

Time (min) 20 30

40

0.0 –0.5

0.4

Ln(C/C0)

|pCBA| / |pCBA|0

0.8

BDD Ti/RuO2 Pt Ti/Pt-IrO2

0.2

–1.0 –1.5 –2.0 –2.5

Ti/IrO2 0.0 0

5

10

15 20 Time (min)

25

30

35

Figure 16.3 Effect of electrode material on the production of • OH as accessed by the degradation of pCBA (i = 100 mA cm−2 , [pCBA]0 = 1.9310 L−6 M, [KH2 PO4 ]0 = 0.2 M, pH = 7.1, T = 25◦ C). (Reprinted with permission from Ref. 72.)

with a previous study achieved by Kraft (10 years before, using few electrode materials [57]). The large difference in this order from that of ROS was attributed to the difference in the electrocatalytic activity of each electrode material toward the production of active chlorine, as evidenced by linear sweep voltammetry (LSV) measurements (see Figure 16.4). Five electrode materials were employed in this study applying different current densities values, using 0.2 M KH2 PO4 as supporting electrolyte and 0.1 M NaCl to understand the behavior of active species. In addition, they examined the characteristics of microbial inactivation as a function of electrode material under the presence of an inert electrolyte, using E. coli as an indicator micro-organism, however, these results will be discussed later in Section 16.7. Polcaro et al. [73] have recently reported the identification of products of the oxidation at BDD anode of chloride ions in aqueous solutions during galvanostatic electrolyse performed in a filter-press reactor operating both in batch and continuous mode. The results obtained in this work showed that at low chloride concentrations, electrolysis with BDD anode produced a mixture of powerful oxidants. Low current density, high mass transfer conditions and low residence time were determined by them as optimal conditions to maximize the concentration of oxidants and minimize the concentration of chlorates. It is portent to remark that the chloride effect has been investigated by other authors to remove organic pollutants from wastewater or drinking water; however, a small number of these papers have explained the real mechanistic or electrochemical phenomena. However, this scientific contribution about the role of the electrode material (BDD) in the production of the oxidant species process has allowed researchers to determine that the strong influence of material in the selectivity and efficiency is an important parameter.

16.5 DIAMOND FILMS FOR DRINKING WATER DISINFECTION

387

Oxidants Conc. (mg/L as total ( l )

100 Ti/IrO2 Ti/RuO2 Ti/Pt-IrO2 BDD Pt

80

60

40

20

0 0

2

4

6

8

10

12

Time (min) Figure 16.4 Effect of electrode material on the evolution of chlorine (i = 17 mA cm−2 , [NaCl]0 = 0.017 M, T = 25◦ C). (Reprinted with permission from Ref. 72.)

To interpret this behavior, a new comprehensive reaction mechanism has been proposed that may also justify the controversial effect of chloride ions in wastewater treatments: The electrolysis carried out with BDD anodes and electrolyte-containing chloride concentration higher 1 g L−1 could meet the target of the process only if the active chlorine is effective in oxidation of the pollutant that must be removed. Microcystins (MCs), produced by blue-green algae, are one of the most common naturally occurring toxins found in natural environment. The presence of MCs in drinking water sources poses a great threat to people’s health. There are more than 76 different analogues of MCs. One of them is microcystin-RR (MC-RR, C49 H75 N13 O12 )), which has been isolated from natural blooms or laboratory cultures of cyanobacteria [73,74]. It is far more common than other MCs in Asia due to the higher temperature [75]. MCs are known to be chemically stable compounds due to their cyclic structure [75,78]. The toxicity of MCs, coupled with their stability, is the major problem in water supply. Conventional methods in drinking water treatment such as coagulation and sedimentation have only limited efficacy in removing MC-RR from potable waters; then, more research efforts are needed to develop some powerful alternatives. Therefore, the degradation behavior of microcystin-RR on BDD electrode has been recently studied by Zhang, Fu, and Gu [78]. Electrochemical conditions such as reaction time, supporting electrolyte, and applied current density were varied in order to determine their effects on disinfection process. The degradation rate of MC-RR as a function of electrolysis time operating at pH of 3.23 (adjusted with 1.0 M H2 SO4 ) in presence of lower NaCl concentrations (CRR,0 : 0.5 μg ml−1 , current density: 18.2 mA cm−2 ) was studied by Zhang, Fu, and Gu. They concluded that the removal rate increases linearly with time at the beginning of the electrolysis. At these conditions, 86.8% of MC-RR was removed after 30 min, 98.5% after 60 min, and full removal was achieved after 2 h. However, in order to clarify

388

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

1.0

MC-RR removal, (c0–c)/c0

0.8

0.6

0.4

0.2

0.0 0

10

20

30 40 Time (min)

50

60

Figure 16.5 MC-RR removal rate versus different supporting electrolyte (, RR-02, 20 mM NaCl); (•, RR-03, 20 mM KCl); (, RR-04, 20 mM K2 SO4 ); (, RR-05, 20 mM Na2 CO3 ); (, RR-06, 20 mM KNO3 ); CRR,0 : 0.5 μg ml−1 , current density: 18.2 mA cm−2 , no pH adjustments. (Reprinted with permission from Ref. 78.)

the roles of different electrolytes in electrochemical degradation of MC-RR, a series of inorganic salts with the concentration of 20 mM were employed into the comparative study. Then, the results clearly demonstrated the impacts of different electrolytes on the degradation of MC-RR (see Figure 16.5), confirming that the influence of cations during the electrolysis. As it can be seen in Figure 16.5, the use of NaCl leads to a fast promotion on the removal rate of MC-RR. Therefore, Zhang, Fu, and Gu [78] studied the effect of NaCl concentration on the removal rate of MC-RR, over a range of 10–35 mM. At NaCl concentration of 35 mM, the destruction was immediate that MC-RR was undetectable by the first sampling point (5 min; see Figure 16.6). This proves that the effect of NaCl concentration is highly positive for the destruction of MC-RR molecules due to a rise in active chlorine generation, which may be considered as a factor for the full removal of MC-RR by electrochlorination with BDD electrode. However, from the HPLC results reported by Zhang, Fu, and Gu [78] it was possible to discover the formation of disinfection by-products. Based on these results, although no hazardous by-products in MC systems were detected so far, it is worthy to note that further studies need to be extensively conducted to confirm these results for the application of BDD technology.

16.6 PRODUCTION OF INORGANIC DISINFECTION BY-PRODUCTS AND INORGANIC SPECIES ELIMINATION Research into new and effective methods for water disinfection has prompted interest in the direct electrochemical treatment of natural waters. The aim of electrochemical

16.6 PRODUCTION OF INORGANIC DISINFECTION BY-PRODUCTS

389

1.0

MC-RR removal, (c0–c)/c0

0.8

0.6

0.4

0.2

0.0 0

10

20

30 40 Time (min)

50

60

Figure 16.6 MC-RR removal rate versus concentrations of sodium chloride (, RR-07, 35 mM NaCl); (•, RR-02, 20 mM NaCl); (, RR-08, 10 mM NaCl); CRR,0 : 0.5 μg ml−1 , Iappl: 18.2 mA/cm2 . (Reprinted with permission from Ref. 78.)

treatment is to generate oxidizing species with high bactericidal effect; these species are generally obtained by the oxidation of chloride ions that are usually present, even at low concentrations, in natural waters. As commented in the last section, the use of BDD electrodes for effective electrochlorination disinfection has been proposed in recent years. However, there is a lack of information about the formation of undesired inorganic byproducts during the treatment of water with low chloride ion content, using BDD anodes. Recently, Bergmann and Rollin [79] highlighted the problems due to the formation of disinfection by-products such as ClO2 , ClO3 − , and ClO4 − during electrolysis by using BDD anodes. Although the effect of these compounds on human health is still under study, it is suspected that they carry health risks; so the WHO recommends very low concentrations of ClO3 − and ClO4 − in drinking water [80]. Therefore, the understanding about the formation of products and by-products during the electrolysis of water with low concentration of chloride ions at BDD anodes is a crucial point, in order to individuate the operating conditions that minimize the formation of undesired products to extend the application of BDD anodes to environmentally real oriented purposes.

16.6.1

Chloride, Chlorite, and Chlorate Ions

Little is known about the formation of inorganic disinfection by-products using BDD. Nevertheless, Polcaro et al. [51] concluded that several intermediates were formed when chloride-free waters containing sulphate were electrolyzed. Similar findings were also recently obtained by Bergmann and Rollin [79] and Polcaro et al. [81], demonstrating that chlorate and perchlorate were found in remarkable concentration after electrolysis treatment. Borutzky, Bergmann, and Junghannss [82] found disinfection effects in

390

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

sulphate containing waters treated by electrolysis using BDD anodes, concluding that chloride-based disinfection by-products have to be expected when drinking waters are subjected to electrolysis. It is known that BDD anodes produce oxidants in a series of reactions on and near the electrode, as shown earlier in Reactions (16.5) to (16.9), together with the following simplified reactions (16.13) and (16.14): OH− → • OH + e− •

OH + H2 O2 → HO2 • + H2 O

(16.13) (16.14)

However, the formation of oxidants such as ozone and OH radicals is also known or assumed for other electrode materials [7,83]. Several researchers have studied the anodic chloride oxidation and chlorine formation on BDD anodes [84,85]. Then, from processes such as ozonation and combined disinfection processes, different disinfection by-products can be expected [86]. This is especially true for BDD because it tends to produce oxidants electrochemically with further chemical reactions in a quasi-adsorbed state. The production of free or active chlorine components (free available chlorine, dissolved Cl2 , HOCl, and OCl− ) is usually the aim of electrolysis process due to the high disinfecting ability of these species. First studies were carried out by Ferro et al. [84], discussing and demonstrating the mechanism for forming Cl2 from 2 Cl− . This mechanism occurs by way of dissolved chlorine that quickly reacts with water and OH− ions to form hypochlorous acid and hypochlorite ions. However, no kinetic studies were presented in the literature for solutions with chloride concentration in the ppm range. Then, it motivated the studies performed by Bergmann and Rollin [87] using mixed oxide electrodes. They showed that the chlorine formation exists even at very low chloride concentrations of some ppm. Based on these results, Bergmann and Rollin have assumed that chlorine formation is possible at extremely low chloride concentrations, but side reactions may lower the chlorine concentration again. One side reaction was obviously based on hydrogen peroxide that is formed anodically or cathodically. This means, for each system with defined temperature, electrode materials, distance between anode and cathode, current density, and so on, one minimum chloride concentration value existed for clear measurement of produced active chlorine. For example, the researchers found that for one BDD system containing only chloride and 50 ppm nitrate, a minimum concentration of 10 ppm of Cl− was necessary for chlorine detection. Then, active chlorine is an electrolysis product both at mixed oxide and doped diamond anodes. However, mechanism of formation, current efficiency, and reaction behavior differ significantly in the two cases. Bergmann and Rollin demonstrated that, while the measured chlorine values in experiments using mixed oxide anodes usually increased continuously until reaching a limit, in experiments with diamond anodes a maximum of active chlorine was reached soon after starting the experiment (see Figure 16.7). In addition, Bergmann and Rollin [87] discussed that the destruction of active chlorine by ROS to give chloride could be attained, being that chlorite and chlorate are the principal products. However, chlorite cannot be measured by ion chromatography because chlorite is a short-lived component and it may react with many other species or directly at the anode. The very good reactivity during electrolysis of sodium chlorite solutions is shown in Figure 16.8. Nevertheless, the researchers have observed the formation of chlorine dioxide and chlorate, indicating a stepwise mechanism of oxidation by • OH

AC/[Cl(0)]

16.6 PRODUCTION OF INORGANIC DISINFECTION BY-PRODUCTS

391

0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0

10

20

30

40

50

60

Time of electrolysis (min) Figure 16.7 Active chlorine concentration (AC) related to starting chloride concentration (Cl(0)) versus. time (rotating BDD anode, IrO2 cathode, water containing 50 ppm [Cl− ] as sodium salt, 100 ml, 200 mA, 20◦ C, 300 rpm, and DPD method of active chlorine analysis). (Reprinted with permission from Ref. 79.)

350 61.4 ppm chlorite 311 ppm chlorite

Chlorite concentration (ppm)

300

250

200

150

100

50

0 0

10

20

30

40

50

Time of electrolysis (min) Figure 16.8 Chlorite depletion in two discontinuous experiments varying starting concentration as indicated in the legend (rotating BDD anode, IrO2 cathode, sodium salt, 100 ml, 200 mA, 20◦ C, and 300 rpm). (Reprinted with permission from Ref. 79.)

radicals. It is interesting to note that other authors have proposed many assumptions for chlorate formation at other electrodes (Pt, IrO2 /RuO2 ), including chlorite as an intermediate of chlorate formation on electrocatalytically active surfaces [88,89], and it is in contradiction with BDD anode considered a nonactive electrode. In addition, according to Figure 16.8, chlorate formation is present during the experiments starting with water containing chloride and hypochlorite; contrasting to

392

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

the observed on mixed oxide anodes where active chlorine concentration continuously decrease. Chlorate was formed in both experiments, but slowly in the mixed oxide anodes experiment and faster (going though a relatively high maximum) in the case of BDD anode. It is important to indicate that chlorate is usually a stable product in disinfection processes based on chlorine or chlorine dioxide addition, but the disappearance of chlorate in the BDD experiment indicates possible chemical or electrochemical reactions. Therefore, the formation of perchlorate could be assumed because this process is known in the electrolysis at high concentrated chlorate electrolytes [90]. However, the occurrence of perchlorate in drinking water is a serious problem due to its carcinogenic potential, and it will be separately discussed in next section. 16.6.2

Perchlorate in Drinking Water

The presence of perchlorate is one of the most typical discussion points concerning drinking water in the US [91]. After the contamination of large areas in the United States, two states set a limiting concentration of 2 and 6 ppb for perchlorate in drinking water. Perchlorate formation was also reported by Ferro et al. [84], suggesting that the reactions between reactive oxygen species and active chlorine may be the reason for the influence of the applied current density and stirring rate on the system behavior observed during electrolysis at BDD. Higher current density may cause not only the observed initial faster production of oxidizing species but also their faster decrease since all the subsequent steps are accelerated by the high reactant concentration. Moreover, low stirring rate may lead to poor mixing, thus giving rise to dead zones in the reactor with local accumulation of reactants enhancing the further oxidation of active chlorine to ClO3 − and ClO4 − . Recently, Bergmann, Rollin, and Iourtchouk [92] performed electrochemical studies to estimate the risks of perchlorate formation in drinking water disinfected by direct electrolysis. They tested BDD anodes in laboratory and commercially available cells at 20◦ C, and, for comparison, other anode materials such as platinum and mixed oxide were also tested. It was found that BDD anodes have a 1000-fold higher perchlorate formation potential compared with the other electrode materials that were tested. Perchlorate is thought to be nonadsorptive and nonreactive in electrochemical systems. In most situations this is true but under special conditions exceptions exist. So, in previous studies it was reported that perchlorate may be reacted to Cl2 O7 on Pt anodes [93]. When a perchlorate solution is electrolyzed, the limiting current typical for electrode processes with • OH radical formation prior to the oxygen evolution is reduced by a factor of approximately 2. In terms of adsorption, perchlorate ions may block active sites and thus reduce the efficiency of • OH radical formation. This discussion appears logical because on BDD surfaces the concentration of active sites is radically reduced compared with conventional mixed oxide electrodes. Under these conditions, the competing reactivity and/or adsorption of ions in the same ppm concentration range near the electrode can have a higher influence with respect to penetration of adsorbing water molecules from the rare active sites. Figure 16.9 depicts results showing the results obtained by Bergmann et al. [92] at preliminary long-term electrolysis of drinking water. It was surprising that all chloride ions reacted to perchlorate (i.e., no gases containing Cl atoms were stripped and all intermediates were converted to higher oxidized species and finally to perchlorate). Perchlorate in critically high concentration was found even for small specific charges of

16.6 PRODUCTION OF INORGANIC DISINFECTION BY-PRODUCTS

393

Perchlorate (ppm)

250 hypochlorite chlorite chlorate

200 150 100 50 0 0

20

10

30

40

50

60

70

Time of electrolysis (min) Figure 16.9 Perchlorate formation during electrolysis using rotating BDD anode in electrolytes containing 102 ppm hypochlorite (Ca(OCl)2 ) + 3 ppm chloride (NaCl), 111.2 ppm chlorite + 5.5 ppm chloride (NaClO2 + NaCl) and 179.4 ppm chlorate (NaClO3 ), respectively (IrO2 cathode, 100 ml, 200 mA, 20◦ C, and 300 rpm). (Reprinted with permission from Ref. 79.)

Yield of chloride to perchlorate coversion (mmol mol–1)

1200 1000 800 600 400 200 0 0

0.5

1

1.5

Specific charge (Ah

2

2.5

L–1)

Figure 16.10 Conversion yield of chloride to perchlorate in a discontinuous experiment using rotating BDD anode and real drinking water (water composition: 42 ppm [Cl− ] + 162 ppm [SO4 2− ] + 12.9 ppm [NO3 − ], 200 Am−2 , IrO2 cathode, 300 rpm, 100 mL, 20◦ C, initial pH 7.9). (Reprinted with permission from Ref. 92.)

about 0.01 Ah L−1 , which are adequate to single-pass operation in technical cells (see also Figure 16.10). Bergmann and associates [92] found that, in long-term discontinuous experiments, all the chloride finally reacted to form perchlorate. The same result was obtained when probable oxychlorine intermediates (OCl− , ClO2 − , ClO3 − ) were electrolyzed in synthetic waters in the ppm range of concentrations. The tendency to form perchlorate was confirmed by them when the flow rate of drinking water was varied between 100 and 300 L h−1 and the temperature increased to 30◦ C. In a continuous flow mode of operation, a higher chloride concentration in the water resulted in a lower perchlorate

394

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

6.4 mmol/L - Chloride 1.94 mmol/L - Hypochlorite 1.65 mmol/L - Chlorite 2.1 mmol/L - Chlorate

Initial concentration: 500

Perchlorate (ppm)

400 350

ppm

450

300 250

50 45 40 35 30 25 20 15 10 5 0 0 0.05 0.1 0.15 0.2 0.25 Ah L–1

200 150 100 50 0 0

0.5

1

1.5

2

2.5

Specific charge (Ah L–1) Figure 16.11 Perchlorate formation in discontinuous experiments using rotating BDD anode and single chloride, hypochlorite, chlorite and chlorate solutions in ppm range of concentration (200 Am−2 , IrO2 cathode, 300 rpm, 100 mL, 20◦ C). (Reprinted with permission from Ref. 92.)

formation. This assumption was explained by reaction competition of species near and on the anode surface for experiments both with synthetic and local drinking waters. According to these results reported by Bergmann in Figure 16.11, all products are approximately of the same initial concentration except for the chloride. At a higher specific charge, all chlorite, hypochlorite, and chlorate are totally converted to perchlorate. Surprisingly, hypochlorite seems to react faster than chlorite and it is able to form ClO2 , as reported by Bergmann and colleagues [92]. This research work showed that electrochemical, ozone, and • OH radical-based chemical steps participate in the formation of perchlorate. Chloride was the only starting species that exhibited an exponential growth of perchlorate formation. The reason was the better adsorption/reactivity of chloride ions on the surface converted to chlorate (Figure 4 in Ref. 92). Only when most of the chloride was converted to other species might chlorate replace chloride on the active sites of the electrode. In most of the practical applications, no perchlorate will occur in mixed oxide anodes cells if a low current density is chosen. Both producers and users must be aware of the risks associated with the use of mixed oxide anodes and take appropriate action. The probability of perchlorate formation (in ppm) for different electrode materials is given as follows: IrO2 = 0; 30% IrO2 + 70% RuO2 = 1.3; 70% IrO2 + 30% RuO2 = 1.2; RuO2 = 0; Pt = 2.35 and BDD = 123. These results were obtained in a cell with a rotating disk anode, 220–240 ppm [Cl− ], 100 mL, 200 Am−2 , 1.0 AhL−1 operated at 300 rpm at 20◦ C. As can be seen, for the IrO2 and RuO2 anodes no perchlorate was found. The measurements confirm the hypothesis that the generation of ozone and radicals using these anodes differs from that when using BDD. 16.6.3

Electrolysis of Nitrates

Nitrate is a drinking water component with a limiting concentration of 50 ppm according to the rules of many countries, but others have a limit about 15 ppm. Mostly,

16.6 PRODUCTION OF INORGANIC DISINFECTION BY-PRODUCTS

395

waterworks try to adjust to lower concentration values because nitrate is suspected to cause nitrosoamines in the human body. It is also known from the literature that nitrate can be reduced at the cathode in acidic, neutral, and alkaline media. Doped-diamond cathodes were studied in neutral media by several authors [94–97], and nitrate could be reduced, but reproducible results were obtained only in experiments using alkaline electrolytes. Levy-Clement et al. [96] found that at applied potentials between −1.5 and −1.7 V, the amount of NO3 − reduced (10%) and that it is mainly transformed into gaseous products, then it increases to 29% at −2 V with almost equal parts of nitrite and nitrogenous gas formed, without the production of ammonium (see Figure 16.12). On the whole, the mechanisms of nitrate reduction are not yet clear and usually sum reactions are used to describe the processes of nitrite and ammonia formation from Reactions (16.15) to (16.17): NO3 − + H2 O + 2e− → NO2 − + 2OH−

(16.15)

NO3 − + 6H2 O + 8e− → NH3 + 9OH− −

−

(16.16)

−

NO2 + 5H2 O + 6e → NH3 + 7OH

(16.17)

Ammonia may react to ammonium. In fact, nitrite and ammonium ions were clearly detected by them when performing nitrate electrolysis in the ppm concentration range using BDD. Figure 16.12 shows the results with respect to ammonium ions from corresponding experiments applying different cell currents, but no significant differences were observed in these experiments. This behavior probably was achieved because the electrocatalytic process of ammonia oxidation is not characteristic for the BDD anode and ammonium may accumulate, as suggested by the authors [95]. However, more investigations are being developed to determine the strong influence of BDD as cathodes material in the selectivity and efficiency of nitrates removal.

30

30

–

NO3

Formed products (%)

20

Nitrogenous gas

20 15

15 10

–

NO2

10

5

Removed nitrates (%)

25

25

5 +

NH4 0 –2.1

0 –2.0

–1.9

–1.8

–1.7

–1.6

–1.5

–1.4

Applied potential (V/SCE) Figure 16.12 Variation of the concentration of reduced nitrate () and formed nitrogenous compounds [nitrite (◦), ammonium (•) and gas ()] during 16 h of electrolyses in 1 M KNO3 as a function of the applied potential. (Adapted from Ref. 96.)

396

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

16.7 ELECTROCHEMICAL FREE-CHLORINE SYSTEMS USING DIAMOND FILMS As discussed in the last sections, the aim of electrochemical treatment is to generate oxidizing species with high bactericidal effect; these species are generally obtained by the oxidation of chloride ions that are usually present in natural waters. However, the studies developed for some electrochemists have demonstrated that more investigation is needed to assess possible problems involved with the formation of disinfection byproducts having a high health risk, such as ClO2 , ClO3 − , and ClO4 − . An alternative to these problems is the development of electrochemical systems avoiding the use of chlorine in solution. Therefore, new evidences on the oxidant action of ROS ( • OH, • O, H2 O2 , O3 ) in the electrochemical disinfection with diamond films have been recently obtained by electrolyzing free-chlorine waters. Polcaro et al. [51] have reported the treatment of bacterial suspensions of E. coli, Enterococcus faecalis, and coliforms in 1 mM Na2 SO4 using the system with a stirred tank reactor containing a Si/BDD anode. Figure 16.13 shows that the concentration of oxidants accumulated in electrolyzed solution (H2 O2 , O3 , and peroxodisulfate formed from Reactions (16.8) through (16.10)) under continuous treatment depends on effluent flow rate. A decrease in oxidant species with raising the effluent flow can be observed. However, higher concentrations of oxidants are produced when high current density is applied and low stirring velocity (characterized by its Reynolds number) is used during the experiments. On the basis of these results, several disinfection trials were performed and a fast inactivation of all bacteria was found by applying 10 mA cm−2 under batch conditions, as illustrated in Figure 16.14. Good performances regarding the reduction of micro-organism populations were achieved, from 1 × 103 CFU mL−1 to the detection limit in 60, 100, and 300 s for E. coli, coliforms, and enterococcci , respectively. The high quality of these results is evident when these are compared with similar electrochemical processes with other anode materials taken from Table 16.2. For example, Kerwich

[Oxidants] (meq mL–1)

0.03

0.02

0.01

0 0

20

40

60

80

100

120

Flow rate (mL min–1) Figure 16.13 Trend of the concentration of oxidants produced in the system of Figure 16.5 as a function of the effluent flow rate under continuous treatment. Applied current density: 13.3 mA cm−2 (full symbols), 6.6 mA cm−2 (empty symbols). Reynolds number: 1.5 × 103 (circles), 1.0 × 104 (squares). (Adapted from Ref. 51.)

16.7 ELECTROCHEMICAL FREE-CHLORINE SYSTEMS USING DIAMOND FILMS

397

0

In N/N0

–2

–4

–6

–8 0

20

40

60 80 Electrolysis time (s)

100

120

140

Figure 16.14 Survival ratio with electrolysis time for Escherichia coli (), coliforms (), and enterococci (•) under batch treatment at 10 mA cm−2 and a Reynolds number of 1.0 × 104 . [E.coli]0 = 6.4 × 102 CFU mL−1 , [coliforms]0 = 2.3 × 103 CFU mL−1 , [enterococci]0 = 4.4 × 103 CFU mL−1 . Full lines represent the least squares regression lines for the data and dotted lines represent the standard error of the regression. (Adapted from Ref. 51.)

Generated Gas

Generated Gas Contaminated water into cell Inlet

Electrode connection to power source

Electrode connection to power source Pt-Nb wire mesh anode

Steel cathode

Water moves freely through the open cell without gaskets and out the sides into the collection vessel Figure 16.15 Scheme of the Zappi™ cell used to disinfect waters contaminated with E. coli and the bacteriophage MS2. (Reprinted with permission from Ref. 7.)

et al. [7] treated in batch 10 L of 0.030 M Na2 SO4 or 0.036 M NaH2 PO4 at 6 L min−1 through the Zappi™ cell sketched in Figure 16.15, showing a 4 log inactivation of E. coli and bacteriophage MS2 cells in both media after long electrolysis times of 60–75 min at 24–27 mA cm−2 due to the lower production of ROS at the Pt-Nb anode than at diamond films. Patermarakis and Fountoukidis [43] exposed total coliforms at population density of 200–26, 800 cell mL−1 in tap water to alternating current of 2.5 mA cm−2 using Ti electrodes, but the culturable counts were reduced only by an order of magnitude in 15.7 min.

398

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

Matsunaga et al. [46] reduced E. coli from a population density of 102 cell mL−1 in tap water to less than 2% of the initial number after 10 min of electrolysis with a carbon-cloth electrode at 0.7 V. The good efficiency of the direct electrochemical disinfection with a Si/BDD anode in diluted Na2 SO4 solutions corroborates the important role that the anode material plays to produce ROS. A detailed study made by Jeong et al. [49] about the inactivation of E. coli cells with an electrochemical free-chlorine system containing 0.2 M phosphate buffer and using Nb/BDD as anode should also be mentioned. The morphological changes of cells after 5 min of electrolysis at 100 mA cm−2 were simultaneously followed by transmission electron microscopy (TEM) and atomic force microscopy (AFM). Comparison of TEM images of Figures 16.16A and 16.16B for untreated and treated bacteria, respectively, evidences the existence of drastic changes in the nature of the contents of cells, as well as in the structure of their walls, after electrolysis. The cells become mostly empty and their membranes appear to be no uniform. Figures 16.16C and 16.16D present the AFM images of the same cells before and after electrolysis. While the surface of the untreated cells appears to be smooth and flat, the treated cells have a rough and sunken surface, as if they had shrunk when the inner contents escaped from them. These morphological changes can be interpreted by the attack of ROS disrupting the integrity of the cell membrane and leading to the lyses of

(a)

(b)

(c)

(d)

µm 5

µm 5 4

4 3

3 2

2 1

1

Figure 16.16 Morphological change of E. coli cells resulting from the electrolysis at 100 mA cm−2 for 5 minutes using a Nb/BDD anode. [E. coli]0 = 108 CFU mL−1 , [KH2 PO4 ]0 = 0.2 M, pH = 7.1, 25◦ C, (A) before electrolysis (TEM); (B) after electrolysis (TEM); (C) before electrolysis (AFM); and (D) after electrolysis (AFM). (Reprinted with permission from Ref. 49.) See color insert.

16.7 ELECTROCHEMICAL FREE-CHLORINE SYSTEMS USING DIAMOND FILMS

399

2.5 Without t-BuOH With 0.05 M t-BuOH

Log inactivation

2.0

1.5

1.0

0.5

0.0 BDD

Pt

Ti/RuO2 Ti/Pt-IrO2

Ti/IrO2

Figure 16.17 Effect of electrode material on the inactivation of E. coli without and with t-BuOH in the electrolyte of KH2 PO4 (i = 100 mAcm−2 , electrolysis time = 3 min, [KH2 PO4 ]0 = 0.2 M, [t-BuOH]0 = 0 or 0.05 M, pH = 7.1, T = 25◦ C). (Reprinted with permission from Ref. 72.)

the cells. This study clearly shows that strong oxidants as ROS formed by electrolyzing water at diamond films can cause a significant inactivation of micro-organisms, as much as chlorine in electrochlorination. In 2009, once again, Jeong et al. [72] performed new experiments in order to study the effect of electrode materials on the inactivation of E. coli during the electrolysis for 3 min in 0.2 M of KH2 PO4 electrolyte in the absence and presence of t-BuOH (0.05 M). These experiments were done in order to assess the role of • OH on the inactivation of microorganisms with five different electrode materials. The preliminary tests confirmed that electrolyzed water resulting from the electrolysis of KH2 PO4 for 3 min had no residual disinfecting activity toward E. coli . The highest reduction in the inactivation of E. coli was found at BDD, as an approximately 2.4 log inactivation of E. coli , accomplished after 3 min without t-BuOH, but this process was inhibited by the addition of 0.05 M t-BuOH. This result indicated that • OH was mainly responsible for the inactivation of E. coli at BDD, which agreed well with the result of previous report by the same authors [49]. Similar findings were obtained at Ti/RuO2 , indicating the main effect of • OH in the inactivation but much lower than that at BDD. The large difference in the level of inactivation between at BDD and Ti/RuO2 was ascribed to the different production of • OH. In the case of Pt, an approximately 1.3 log inactivation of E. coli was achieved for 3 min without t-BuOH, which was much higher than that observed at Ti/RuO2 as the production of • OH. In the presence of t- BuOH, the inactivation of E. coli at Pt was partially reduced but not completely inhibited, as can be seen in Figure 16.17. The potential role of these strong oxidant species, which possess higher oxidizing power than chlorine, deserves to be underlined in the treatment of spore-forming micro-organisms that are difficulty inactivated by only chlorine.

400

16.8

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

CONCLUSIONS

There is an increasing incidence in health problems related to environmental issues that originate from inadequate treatment of potable waters. This has compelled scientists and engineers to engage in innovative technologies to achieve a maximum disinfection at affordable costs. Therefore, the development of new approaches to disinfecting waters by using diamond films may lead to an entirely new class of electrochemical free-chlorine systems. In that case, the recent advances obtained with diamond electrodes suggest that their application to the disinfection water technology should be rapidly developed because of their better performance respect to other anode materials. The fast bacterial abatement and total oxidation of organic substances achieved due to the great amounts of ROS produced during water electrolysis are important profit determinants with the adoption of diamond-coated electrodes. As compared with other chemical disinfection methods, the advantages of electrochemical disinfection are obvious: no transport, storage, or dosage of disinfectants are required. The disinfecting effect can be adjusted according to the on-site demand. Electrochemical disinfection shows a reservoir effect and is often more cost effective and requires less maintenance than other disinfection methods. Photovoltaic power supply makes it possible to use electrochemical disinfection far from the electrical supply grid. This may be important for its application to drinking water in developing countries. Electrochemical disinfection can also be used in conjunction with other disinfection methods. Imagine, for example, the use of such practical commercial technology for disinfection, washing, sterilization of medical articles, drinking water disinfection, and treatment of purulent and septic diseases of humans and animals, also allowing application to waste and/or sewage treatment plants, swimming pools, poultry factories, livestock farms, and extremely epidemic danger areas. This opens new perspectives for an easy, effective, and free-chemical water treatment by means of the electrochemical technology with diamond films.

REFERENCES 1. M.A. Shannon, P.W. Bohn, M. Elimelech, J.G. Georgiadis, B.J. Marinas, A.M. Mayes, Nature 2008, 452 , 301–310. 2. A.A.M. Lima, et al. J. Infect. Dis. 2000, 181 , 1643–1651. 3. J.R. Behrman, H. Alderman, J. Hoddinott, “Hunger and malnutrition,” in Copenhagen Consensus—Challenges and Opportunities, OCLC57489365 (London, School of Hygiene and Tropical Medicine, 2004); http://www.copenhagenconsensus.com/Files/Filer/CC/Papers/ Hunger%5Fand%5FMalnutrition%5F070504.pdf. 4. World Health Organization, Emerging Issues in Water and Infectious Disease, World Health Organization, Geneva, 2003, pp. 1–22. 5. G.K. Pitman, Bridging Troubled Waters—Assessing The World Bank Water Resources Strategy, World Bank Publications, Washington, DC, 2002. 6. C.A. Mart´ınez-Huitle, E. Brillas, Angew. Chem. Int. Ed . 2008, 47 , 1998–2005. 7. M.I. Kerwick, S.M. Reddy, A.H.L. Chamberlain, D.M. Holt, Electrochim. Acta 2005, 50 , 5270–5277. 8. Water Chlorination: Chemistry, Environmental Impact and Health Effects, Lewis Publishers, Inc., Chelsea, MI, 1985.

REFERENCES

401

9. H. Komulainen, V.-M. Kosma, S.-L. Vaittinen, T. Vartianinen, E. Kaliste-Korhonen, S. Lotjonen, R.K. Tuominen, J. Tuomisto, J. Nat. Cancer Inst . 1997, 89 , 848–856. 10. A.M. Driedger, J.L. Rennecker, B.J. Marinas, Water Res. 2000, 34 , 3591–3597. 11. R.J. Bull, S.W. Krasner, P. Daniel, R.D. Bull, ”Health Effects and Occurrence of Disinfection By-Products,” AWWA Research Foundation and American Water Works Association, 2001. 12. F.X.R. Van Leeuwen, Food Chem. Toxicol . 2000, 38 , S51–S58. 13. Guidelines for Drinking-Water Quality, Second Edition: Recommendations, Vol. 1, World Health Organization, Geneva, 1993. 14. T. Reichhardt, Nature 1999, 399 , 718. 15. T. Collins, Nature 1999, 399 , 17–18. 16. B. Langlais, D.A. Reckhow, D.R. Brink, Ozone in Water Treatment-Application and Engineering, Lewis Publishers, New York, 1991. 17. N. Facile, B. Barbeau, M. Prevost, B. Koudjonou, Water Res. 2000, 34 , 3238–3246. 18. P. Xu, M.L. Janex, P. Savoye, A. Cockx, V. Lazarova, Water Res. 2002, 36 , 1043–1055. 19. D.M. Blake, P.C. Maness, Z. Huang, E.J. Wolfrum, J. Huang, W.. Jacoby, Sep. Purif. Meth. 1999, 28 , 1–50. 20. D. Blake, NREL Report, Golden, CO, 1999. 21. A. Mills, S. LeHunte, J. Photochem. Photobiol. A: Chem. 1997, 108 , 1–35. 22. Y. Kikuchi, K. Sunada, T. Iyoda, K. Hashimoto, A. Fujishima, J. Photochem. Photobiol. A: Chem. 1997, 106 , 51–56. 23. R. J. Watts, S. Kong, M.P. Orr, G.C. Miller, B.E. Henry, Water Res. 1995, 29 , 95–100. 24. S.W. Krasner et al. Environ. Sci. Technol. 2006, 40 , 7175–7185. 25. M.G. Muellner et al. Environ. Sci. Technol. 2007, 41 , 645–651. 26. K. Rajeshwar, J.G. Ibanez, Environmental Electrochemistry: Fundamentals and Applications in Pollution Abatement , Academic Press, San Diego, 1997. 27. D. Golub, E. Ben-hur, Y. Oren, A. Soffer, Bioelectrochem. Bioenerg. 1987, 17 , 175–182. 28. J.-C. Park, M.S. Lee, D.-W. Han, D.H. Lee, B.J. Park, I.-S. Lee, M. Uzawa, M. Aihara, K. Takatori, Appl. Environ. Microbiol . 2004, 70 , 1833–1835. 29. N.J. Rowan, S.J. Macgregor, J.G. Anderson, D. Cameron, O. Farish, Appl. Environ. Microbiol . 2001, 67 , 2833–2836. 30. P. Wouters, A.P. Bos, J. Ueckert, Appl. Environ. Microbiol . 2001, 67 , 3092–3101. 31. Y. Kondo, H. Umezawa, T. Koizumi, “Sterilizing Method and Electrolysis Water Purification Apparatus,” US patent 2003024828, 2003. 32. R.F. Kenyon, N. Meakin, “Disinfectant Solutions Containing Electrochemically Activated Water,” PCT WO 2005094904, 2005. 33. B. Hakansson, L. Nyman, F. Herlitz, J. Echardt, T. Shimamune, “Electrolytic Cell for Treating Contaminated Water,” PCT WO 2005058761, 2005. 34. C. Feng, K. Suzuki, S. Zhao, N. Sugiura, S. Shimada, T. Maekawa, Bioresour. Technol . 2004, 94 , 21–25. 35. P.S.M. Dunlop, J.A. Byrne, N. Manga, B.R. Eggins, J. Photochem. Photobiol. A: Chem. 2002, 148 , 355–363. 36. K.P. Drees, M. Abbaszadegan, R.M. Maier, Water Res. 2003, 37 , 2291–2300. 37. W. Liang, J. Qu, L. Chen, H. Liu, P. Lei, Environ. Sci. Technol. 2005, 39 , 4633–4639. 38. X.Y. Li, H.F. Diao, F.X.J. Fan, J.D. Gu, F. Ding, A.S.F. Tong, J. Environ. Eng. 2004, 130 , 1217–1221. 39. H. Bergmann, T. Iourtchouk, K. Schops, K. Bouzek, Chem. Eng. J . 2002, 85 , 111–117. 40. Q. Fang, C. Shang, G. Chen, J. Environ. Eng. 2006, 132 , 13–22.

402

APPLICATION OF DIAMOND FILMS TO WATER DISINFECTION

41. T. Furuta, H. Tanaka, Y. Nishiki, L. Pupunat, W. H¨anni, P. Rychen, Diam. Rel. Mat . 2004, 13 , 2016–2019. 42. T. Furuta, P. Rychen, H. Tanaka, L. Pupunat, W. H¨anni, Y. Nishiki, ”Application of Diamond Electrodes for Water Disinfection, in Diamond Electrochemistry (A. Fujishima, ed.), Elsevier, Amsterdam, 2005, Chap. 23. 43. G. Patermarakis, E. Fountoukidis, Water Res. 1990, 24 , 149–1496. 44. M. Okochi, T.K. Lim, M. Nakamura, T. Matsunaga, Appl. Microbiol. Biotechnol . 1997, 47 , 18–22. 45. T. Matsunaga, M. Okochi, M. Takahashi, T. Nakayama, H. Wake, N. Nakamura, Water Res. 2000, 34 , 3117–3122. 46. T. Matsunaga, S. Nakasono, T. Takamuku, J.G. Burgess, N. Nakamura, K. Sode, Appl. Environ. Microbiol . 1992, 58 , 686–689. 47. T. Matsunaga, S. Nakasono, Y. Kitajima, K. Horiguchi, Biotechnol. Bioeng. 1994, 43 , 429–433. 48. S. Guillou, N. El Murr, J. Appl. Microbiol . 2002, 92 , 860–865. 49. J. Jeong, J.Y. Kim, J. Yoon, Environ. Sci. Technol. 2006, 40 , 6117–6122. 50. J. Jeong, J.Y. Kim, M. Cho, W. Choi, J. Yoon, Chemosphere 2007, 67 , 652–659. 51. A.M. Polcaro, A. Vacca, M. Mascia, S. Palmas, R. Pompei, S. Laconi, Electrochim. Acta 2007, 52 , 2595–2602. 52. I. Tr¨oster, L. Sch¨afer, M. Fryda, T. Matth´ee, Water Sci. Technol . 2004, 49 , 207–212. 53. H. Son, M. Cho, H. Chung, S. Choi, J. Yoon, J. Ind. Eng. Chem. 2004, 10 , 705–709. 54. L.V. Venczel, C.A. Likirdopulos, C.E. Robinson, M.D. Sobsey, Water Sci. Technol . 2004, 50 , 141–146. 55. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191–199. 56. M.H.P. Santana, L.A.D. Faria, J.F.C. Boodts, Electrochim. Acta 2005, 50 , 2017–2027. 57. A. Kraft, Int. J. Electrochem. Sci . 2007, 2 , 355–385. 58. O. Scialdone, S. Randazzo, A. Galia, G. Silvestri, Water Res., 2009, 43 , 2260–2272. 59. P. Rychen, L. Pupunat, W. H¨anni, E. Santoli, New Diam. Front. Carbon Technol . 2003, 13 , 109–117. 60. P. Rychen, W. H¨anni, L. Pupunat, Chimia 2003, 57 , 655–658. 61. T. Matth´ee, M. Fryda, S. Mulcahy, W. Stalizus, L. Nyman, Ultrapure Water 2006, 23 , 53–56. 62. C.A. Mart´ınez-Huitle, S. Ferro, Chem. Soc. Rev . 2006, 35 , 1324–1340. 63. M.A. Quiroz Alfaro, S. Ferro, C.A. Mart´ınez-Huitle, Y. Meas Vong, J. Braz. Chem. Soc. 2006, 17 , 227–236. 64. I. Sir´es, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, E. Brillas, Electrochim. Acta 2006, 52 , 75–85. 65. P.-A. Michaud, E. Mah´e, W. Haenni, A. Perret, C. Comninellis, Electrochem. Solid State Lett . 2000, 3 , 77–79. 66. M.S. Saha, T. Furuta, Y. Nishiki, Electrochem. Solid State Lett . 2003, 6 , D5–D7. 67. P. Canizares, F. Larrondo, J. Lobato, M.A. Rodrigo, C. Saez, J. Electrochem. Soc. 2005, 152 , D191–D196. 68. W. Haenni, J. Gobet, A. Perret, L. Pupunat, P. Rychen, C. Comninellis, B. Correa, New Diam. Front. Carbon Technol . 2002, 12 , 83–88. 69. B. Marselli, J. Garcia-Gomez, P.–A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 150 , D79–D83. 70. P. Rychen, L. Pupunat, T. Furuta, M. Sekimoto, H. Tanaka, Y. Nishiki, S. Wakita, “Electrochemical Sterilizing and Bacteriostatic Method,” US Patent 2005023227, 2005.

REFERENCES

403

71. J. Horn, A. Stepan, Method for cleaning, sterilizing and disinfecting dishes and other kitchen utensils and cleaning device for such processes, PCT WO 2006117201, 2006. 72. J. Jeong, C. Kim, J. Yoon, Water Res. 2009, 43 , 895–901. 73. A.M. Polcaro, A. Vacca, M. Mascia, S. Palmas, J. Rodiguez Ruiz, J. Appl. Electroche. 2009, 39 , 2083–2092. 74. G.A. Codd, Ecol. Eng. 2000, 16 , 51–60. 75. S. Haider, V. Naithain, P.N. Viswanathan, P. Kakkar, Chemosphere 2003, 52 , 1–21. 76. H. Yan, G. Pan, M.M. Zhang, H.L. Chen, H.L. Chen, H. Zou, Acta Scientiae Circumstantiae 2004, 24 , 355–359. 77. K. Lahti, J.M. Rapala, M. Niemela, K. Sivonen, Water Res. 1997, 31 , 1005–1012. 78. C. Zhang, D. Fu, Z. Gu, J. Hazard. Mater. 2009, 172 , 847–853. 79. M.E.H. Bergmann, J. Rollin, Catal. Today 2007, 124 , 198–203. 80. Guidelines for Drinking-Water Quality, Vol. 1, World Health Organization, Geneva, 2006. 81. A.M. Polcaro, A. Vacca, M. Mascia, F. Ferrara, J. Appl. Electrochem. 2008, 38 , 979–984. 82. U. Borutzky, H. Bergmann, U. Junghannss, Proceedings of 1st European Conference on Environmental Application of Advanced Oxidation Processes (EAAOP) Chania/Crete 2006, Book of Abstracts, 229 and CD full text version 2006, P207 , 1–6. 83. L.M. da Silva, M.H.P. Santana, J.F.C. Boodts, Quim. Nova 2003, 26 , 880–888. 84. S. Ferro, A. De Battisti, I. Duo, C. Comninellis, W. Haenni, A. Perret, J. Electrochem. Soc. 2000, 147 , 2614–2619. 85. M. Murata, T.A. Ivandini, M. Shibata, S. Nomura, A. Fujishima, Y. Einaga, J. Electroanal. Chem. 2008, 612 , 29–36. 86. V.U. Gunten, Water Res. 2003, 37 , 1469–1487. 87. M.E.H. Bergmann, A.S. Koparal, J. Appl. Electrochem. 2005, 35 , 1321–1329, and Erratum 2005, 36 , 845–846. 88. A. Tasaka, T. Tojo, J. Electrochem. Soc. 1985, 132 , 1855–1859. 89. L.R. Czarnetzki, L.J.J. Janssen, J. Appl. Electrochem. 1992, 22 , 315–324. 90. T. Lehmann, P. Stenner, Germ. Patent no. DE OS 102 58 652 A1 (January, 2002), Deutsches Patent- und Markenamt Muenchen. 91. E.T. Urbansky, M.R. Schock, J. Environ. Manage. 1999, 56 , 79–95. 92. M.E.H. Bergmann, J. Rollin, T. Iourtchouk, Electrochim. Acta 2009, 54 , 2102–2107. 93. Kirk-Othmer, Encyclopedia of Chemical Technology, Vol.5, 3rd ed., John Wiley & Sons, New York, 1979, p. 647. 94. F. Bouamrane, A. Tadjeddine, J.E. Butler, R. Tenne, C. Levy-Clement, J. Electroanal. Chem. 1996, 405 , 95–99. 95. A.N. Ndao, F. Zenia, A. Deneuville, M. Bernard, C. Levy-Clement, Diam. Rel. Mat . 2000, 9 , 1175–1180. 96. C. Levy-Clement, N.A. Ndao, A. Katty, M. Bernard, A. Deneuville, C. Comninellis, A. Fujishima, Diam. Rel. Mat . 2003, 12 , 606–612. 97. G. Sine, L. Ouattara, M. Panizza, C. Comninellis, Electrochem. Solid State 2003, 6 , D9–D11.

17 Fenton-Electrochemical Treatment of Wastewaters for the Oxidation of Organic Pollutants Using BDD Enric Brillas

17.1

INTRODUCTION

Over the past decade, emerging indirect electro-oxidation methods based on Fenton’s chemistry are being developed for the treatment of acidic wastewaters containing persistent organic pollutants. These procedures are environmentally friendly techniques and considered as electrochemical advanced oxidation processes (EAOPs) because their main oxidant is the in situ electrogenerated hydroxyl radical ( • OH) [1–4]. This species is the second strongest oxidant known after fluorine with E ◦ ( • OH/H2 O) = 2.8 V/SHE; hence, it is able to nonselectively destroy most organics via abstraction of a hydrogen atom (dehydrogenation) or addition to a nonsaturated bond (hydroxylation) until total mineralization or conversion into CO2 , water, and inorganic ions [5–7]. The most common of these EAOPs is electro-Fenton (EF) that can be applied using divided or undivided electrolytic cells. In the first case, homogeneous • OH is produced by the catalytic Fenton’s reaction between Fe2+ and electrogenerated H2 O2 ; in the second case, this radical is also formed from water oxidation at the anode surface. The EF Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

405

406

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

process in undivided cells is then influenced by the nature of heterogeneous • OH formed at the anode. It has been found that the use of boron-doped diamond (BDD) films is much more efficient for organic degradation than classical electrodes such as Pt [4]. This phenomenon has also been observed using the photoelectro-Fenton (PEF) method, where the efficiency of EF is enhanced due to the catalytic action of UVA irradiation over the treated solution. This chapter presents the main applications of these emerging EAOPs based on Fenton’s chemistry using a BDD anode. Fundamentals of these methods are initially presented and discussed to analyze their characteristics and oxidation power for organic pollutants removal. 17.2

FUNDAMENTALS OF FENTON’S ELECTROCHEMISTRY

EAOPs based on Fenton’s chemistry are designed to destroy organics in waters using hydrogen peroxide generated at the cathode of an electrolytic cell. Note that H2 O2 is a ‘green’ chemical since it is converted into oxygen gas and water as by-products. It is also a hazardous product since it can be oxidized and easily decomposed by several metallic ions, UVC light, and high temperature. This compound has very low oxidation power and can only attack reduced sulfur compounds, cyanides, chlorine, and certain organics such as aldehydes, formic acid, and some nitro-organic and sulfo-organic compounds [8,9]. Activation of H2 O2 in acidic effluents with Fe2+ ion as catalyst (Fenton’s reagent) is then commonly used for the treatment of organic pollutants in waters by the traditional chemical Fenton method [5–7]. It is well known since 1882 that H2 O2 can be continuously supplied to a solution contained in an electrolytic cell from the two-electron reduction of dissolved O2 gas at a carbonaceous cathode with high surface area [10]. In acidic solution, this reduction process takes place according to Reaction (17.1) with standard potential E ◦ = 0.68 V/SHE: O2(g) + 2H+ + 2e− → H2 O2

(17.1)

This reaction is energetically easier than the four-electron reduction of O2 to water with E ◦ = 1.23 V/SHE. It has been found that H2 O2 production and stability depend on factors such as cell configuration, cathode properties, and operation conditions. For example, its electrochemical reduction at the cathode by Reaction (17.2) and, in much lesser extent, its disproportion in the bulk by Reaction (17.3) are general parasitic reactions that result in the loss of oxidant with a drop in current efficiency [11]: H2 O2 + 2e− → 2OH−

(17.2)

2H2 O2 → O2(g) + 2H2 O

(17.3)

When an undivided cell is used, H2 O2 is also oxidized to oxygen at the anode by Reactions (17.4) and (17.5) with formation of hydroperoxyl radical (HO2 • ) as intermediate [12]: H2 O2 → HO2 • + H+ + e− +

−

HO2 → O2(g) + H + e •

(17.4) (17.5)

17.2 FUNDAMENTALS OF FENTON’S ELECTROCHEMISTRY

407

Reaction (17.4) competes with the anodic oxidation of other products generating other reactive oxygen species (ROS) that can be used to destroy the organics contained in the electrolyzed solution. The strongest ROS is the radical • OH, which is formed at a high O2 -overvoltage anode (M) from water oxidation by Reaction (17.6) [3,13]: M + H2 O → M( • OH) + H+ + e−

(17.6)

The mineralization action of this radical is largely ineffective for classical electrodes such as Pt, but it is much more efficient when a boron-doped diamond thin layer is used as an anode. Operating at the same high current, within the water discharge region, much higher quantity of reactive BDD( • OH) than Pt( • OH) is produced, so that aromatics and unsaturated compounds such as carboxylic acids can even be completely converted into CO2 [13]. The low adsorption capability of • OH on BDD favors its dimerization to H2 O2 by Reaction (17.7), whereas the high oxidation power of this anode facilitates the generation of other weaker oxidants such as ozone by Reaction (17.8) and S2 O8 2− from oxidation of SO4 2− and/or HSO4 − present in the electrolyte by Reactions (17.9) and (17.10), respectively, if sulphuric acid is used to set the solution pH [13–16]. 2BDD( • OH) → BDD + H2 O2 +

3H2 O → O3(g) + 6H + 6e

(17.7)

−

(17.8)

2SO4 2− → S2 O8 2− + 2e− −

2HSO4 → S2 O8

2−

(17.9) +

+ 2H + 2e

−

(17.10)

The removal of organic pollutants by anodic oxidation (AO) with a BDD anode is then based on their direct reaction at the anode and/or their mediated oxidation with ROS such as BDD( • OH) and in less extension O3 , as well as with other weaker oxidizing species formed from the anion of the electrolyte such as S2 O8 2− . In an undivided BDD/O2 cell, organics can be additionally oxidized by other ROS such as the weak oxidants H2 O2 and HO2 • generated from Reactions (17.1), (17.4), and (17.7), leading to the anodic oxidation with electrogenerated H2 O2 (AO-H2 O2 ). This indirect electro-oxidation method has been used in our laboratory to test the superiority of EAOPs based on Fenton’s chemistry such as EF and PEF with Pt/O2 or BDD/O2 cells, as discussed below. The EF treatment of contaminated aqueous solutions involves the continuous generation of H2 O2 from O2 directly injected as pure gas or compressed air, which is efficiently reduced at carbonaceous cathodes such as carbon-polytetrafluoroethylene (PTFE) [17–35], carbon felt [26,27,36–44], and carbon sponge [45] by Reaction (17.1). A small catalytic quantity of Fe2+ is added to the solution to react with electrogenerated H2 O2 to form Fe3+ and homogeneous • OH by the classical Fenton’s Reaction (17.11) [46]: Fe2+ + H2 O2 → Fe3+ + • OH + OH−

(17.11)

This reaction has an optimum pH of 2.8 and is propagated by the catalytic behavior of the Fe3+ /Fe2+ pair [26,46]. Table 17.1 summarizes the main reactions for the Fenton system, along with the corresponding absolute second-order rate constant (k2 ) value [47]. As can be seen, three kinds of reactions can take place after the initiation Reaction (17.11). Thus, Fe2+ can be regenerated by H2 O2 from Fenton-like Reaction (17.12), by HO2 • from

408

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

TABLE 17.1 Main reactions and the corresponding absolute second-order rate constant for a Fenton chemistry system at pH ca. 3 [47]. k2 (M−1 s−1 )

Number

55

(17.11)

Catalysis: regeneration Fe3+ + H2 O2 → Fe2+ + HO2 • + H+ Fe3+ + HO2 • → Fe2+ + O2 + H+ Fe3+ + O2 •− → Fe2+ + O2 Fe3+ + O2 •− + 2 H2 O → Fe2+ + 2 H2 O2

3.1 × 10−3 2 × 104 5 × 107 1.0 × 107

(17.12) (17.13) (17.14) (17.15)

Propagation H2 O2 + • OH → H2 O + HO2 • HO2 • ↔ H+ + O2 •− RH + • OH → R • + H2 O ArH + • OH → ArHOH •

3.3 × 107 4.8a 107 –109 108 –1010

(17.16) (17.17) (17.18) (17.19)

Inhibition Fe2+ + • OH → Fe3+ + OH− Fe2+ + HO2 • + H+ → Fe3+ + H2 O2 O2 •− + HO2 • + H+ → H2 O2 + O2 HO2 • + HO2 • → H2 O2 + O2 HO2 • + • OH → H2 O + O2 O2 •− + • OH → OH− + O2 O2 •− + • OH + H2 O → H2 O2 + OH− + 1/2 O2 • OH + • OH → H2 O2

4.3 × 108 1.2 × 106 9.7 × 107 8.3 × 105 7.1 × 109 1.0 × 1010 9.7 × 107 5.2 × 109

(17.20) (17.21) (17.22) (17.23) (17.24) (17.25) (17.26) (17.27)

Reaction Initiation Fe2+ + H2 O2 → Fe3+ + • OH + OH− Fe2+

a Equilibrium constant.

Reaction (17.13), and/or by superoxide ion (O2 •− ) from Reactions (17.14) and (17.15). The propagation reactions include the production of HO2 • by Reaction (17.16) and O2 •− by Reaction (17.17), along with the attack of • OH to saturated or aromatic organics giving dehydrogenated or hydroxylated derivatives via the general Reactions (17.18) or (17.19), respectively. The inhibition Reactions (17.20) through (17.27) promote the removal of reactive radicals and can restrict the range of several experimental parameters. For example, the existence of the parasitic Reaction (17.20) prevents the use of high concentrations of Fe2+ , being usually lower than 1.0 mM. An interesting differentiation of the EF process in relation to the chemical Fenton method is that Fe2+ can be regenerated from the electroreduction of Fe3+ at the cathode from Reaction (17.28) with E ◦ = 0.77 V/SHE [4]: Fe3+ + e− → Fe2+

(17.28)

The continuous regeneration of Fe2+ by Reaction (17.28) accelerates the production of • OH from Fenton’s Reaction (17.11). As a result, Fe3+ can also be alternatively added to the aqueous medium yielding similar decontamination to that reached with Fe2+ as initial catalyst [36,42]. However, in an undivided electrolytic cell, the slow oxidation of Fe2+ to Fe3+ at the anode is feasible by Reaction (17.29) [26]: Fe2+ → Fe3+ + e−

(17.29)

17.3 ELECTROGENERATION OF H2 O2 AND REGENERATION OF Fe2+

409

The EF method using an undivided BDD/O2 cell, for example, involves the simultaneous oxidation of pollutants with heterogeneous BDD( • OH) produced from Reaction (17.6) and by homogeneous • OH formed from Fenton’s Reaction (17.11), although slower degradation with weaker oxidants (H2 O2 , HO2 • , O3 , S2 O8 2− , etc.) is feasible. The following major advantages for this indirect electro-oxidation method compared with the chemical Fenton process have been claimed [1,2,4]: 1. The on-site production of H2 O2 that avoids the risks related to its transport, storage, and handling; 2. The possibility of the control of the degradation kinetics to allow mechanistic studies; 3. The higher degradation rate of organic pollutants because of the continuous regeneration of Fe2+ at the cathode, which also minimizes sludge production; and 4. The feasibility of overall mineralization at a relatively low cost if the operation parameters are optimized. A related EAOP is PEF in which the solution treated under EF conditions is illuminated with artificial UVA light to enhance the degradation process of organics. The action of this radiation is complex and can be accounted for by (1) a greater regeneration of Fe2+ and production of additional • OH from photoreduction of Fe(OH)2+ , the predominant Fe3+ species in acidic medium [46], by the photolytic Reaction (17.30) and/or (2) the fast photodecarboxylation of complexes of Fe(III) with generated carboxylic acids such as oxalic. Reaction (17.31) shows the photolytic processes for Fe(III)-oxalate complexes (Fe(C2 O4 )+ , Fe(C2 O4 )2 − , and Fe(C2 O4 )3 3− ) [48]. Fe(OH)2+ + hν → Fe2+ + • OH 2Fe(C2 O4 )n

(3−2n)

+ hν → 2Fe

2+

(17.30) + (2n−1)C2 O4

2−

+ 2CO2

(17.31)

A disadvantage for the industrial application of PEF is the high electrical cost of lamps supplying UVA light. An interesting alternative possibility is the use of sunlight as an inexpensive and renewable energy source of wavelength >300 nm [29]. This method, called solar photoelectro-Fenton (SPEF), enhances photolytic processes that are extended in the visible region, as in the case of Reaction (17.31) occurring from 300 to 480 nm.

17.3

ELECTROGENERATION OF H2 O2 AND REGENERATION OF Fe2+

Two-electrode undivided tank reactors and flow cells are typically employed to apply the AO-H2 O2 , EF, PEF, and SPEF treatments with a BDD anode to wastewater remediation. Figure 17.1 shows the sketch of a two-electrode undivided tank reactor with a carbonPTFE O2 -diffusion cathode [12], whereas Figure 17.2 presents a similar cell with a carbon-felt cathode [42]. In both cases, a Pt anode was also alternatively used. The scheme of a flow plant with an undivided BDD/O2 filter-press reactor coupled to a solar photoreactor for the SPEF degradation of organics wastewaters is depicted in Figure 17.3 [24,29]. The electrochemical systems with a carbon-PTFE O2 -diffusion cathode are fed with an excess of pure gas to avoid percolation of the liquid through the carbon cloth. When the

410

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

6 W UVA light or direct sunlight

O2 flow

(–) Nichrome wire

Holder of polypropylene

(+) Pt or BDD anode

Water to thermostat

Carbon-PTFE O2-fed cathode

Solution Water from thermostat

Magnetic bar Figure 17.1 Scheme of a bench-scale open and stirred two-electrode undivided tank reactor containing a carbon-PTFE gas diffusion cathode directly fed with pure O2 for H2 O2 electrogeneration. This reactor was used for the electro-Fenton (EF), photoelectro-Fenton (PEF) with a 6-W UVA irradiation, and solar photoelectro-Fenton (SPEF) treatment of organics. (Adapted from Ref. 12.)

Power supply Compressed air

Air drying solution

Carbon-felt cathode

Air diffuser

Magnetic bar

Pt or BDD anode

Figure 17.2 Sketch of a bench-scale electrochemical system consisting of a stirred two-electrode one-compartment tank reactor with a carbon-felt cathode for the EF degradation of organics. Compressed air was bubbled through the solution for H2 O2 electrogeneration. (Adapted from Ref. 42.)

17.3 ELECTROGENERATION OF H2 O2 AND REGENERATION OF Fe2+

411

(a) Solar photoreactor

–

Gas-diffusion cathode

+

Power supply

BDD anode

O2

Cell

Sampling Flowmeter Pump

Heat exchangers

Reservoir Purge valves (b) Ni mesh collector Cathode

O2

Liquid compartment

Outlet

O2 chamber

gasket

Inlet BDD anode

End plate

Figure 17.3 Sketches of (a) a batch recirculation solar flow plant and (b) the one-compartment filter-press cell with a BDD anode and a carbon-PTFE O2 -diffusion cathode, both of 20-cm2 geometric area, used for the SPEF degradation of organics in acid medium. (Adapted from Refs. 24 and 29.)

carbon-felt cathode is employed, compressed air is usually bubbled through the solution under vigorous stirring to produce H2 O2 from Reaction (17.1). The initial pH of effluents treated by EF is normally regulated to a value near 3.0, close to the optimum pH of 2.8 for Fenton’s Reaction (17.11) [46], to ensure the fastest generation of homogeneous • OH. Accumulation of electrogenerated H2 O2 , regeneration of Fe2+ to Fe3+ at the cathode, and its inverse process at the anode depend on the applied current and catalyst content. These parameters affect the production of • OH and are optimized in each system. Figure 17.4a shows the change of the concentration of accumulated H2 O2 and iron ions with time during the electrolysis of 200 mL of 0.05 M Na2 SO4 with 4.0 mM Fe3+ at pH 3.0 in the BDD/O2 and Pt/O2 cells of Figure 17.1 with 3-cm2 electrodes operating

412

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

10 (a) 8

6

Concentration (mM)

4

2

0

(b)

0.20

0.15

0.10

0.05

0.00

0

10

20

30

40

50

60

70

Time (min)

Figure 17.4 Variation of Fe3+ , Fe2+ , and H2 O2 concentrations with time during the electrolysis of 200 mL of 0.05 M Na2 SO4 solutions with different Fe3+ contents at pH 3.0, 300 mA, and room temperature in a one-compartment cell. (a) 4.0 mM Fe3+ and a 3-cm2 O2 -diffusion cathode and (b) 0.20 mM Fe3+ , air saturated solution, and a 70-cm2 carbon-felt cathode. Species: (◦) Fe3+ , () Fe2+ , and () H2 O2 using a 3 cm2 -Pt anode; (•) Fe3+ , () Fe2+ , and () H2 O2 using a 3-cm2 BDD anode. (Reprinted from Ref. 26.)

at 300 mA [26]. In these EF conditions, a similar behavior can be observed for both tank reactors, since a steady-state of about 9 mM is attained for H2 O2 , whereas the Fe3+ concentration always remains approximately equal to its initial value. The steady H2 O2 concentration is attained just when the rates for its production from Reaction (17.1) and its destruction, mainly at the anode, from Reaction (17.4) become equal. These results demonstrate such a high ability of H2 O2 generation at the O2 -diffusion cathode that regenerated Fe2+ from Reaction (17.28) is rapidly converted into Fe3+ , without significant influence of using a BDD or Pt anode. In contrast, the use of a carbon-felt cathode with a large specific area only allows H2 O2 accumulation in the bulk in the absence of iron ions, since it is not detected when EF is applied, as shown in Figure 17.4b [26] using the cell of Figure 17.2 equipped with a 70 cm2 carbon-felt electrode after adding 0.2 mM Fe3+ as catalyst. A very fast conversion of all initial Fe3+ into Fe2+ can be observed in Figure 17.4b for the Pt anode operating at 300 mA, reaching almost 0.2 mM Fe2+ after 20 minutes of electrolysis, as expected if the rate of Reaction (17.28) is so high that Fe2+ is rapidly regenerated from Fe3+ reduction at the cathode, thus removing all

17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

413

60 50

[H2O2] (mM)

40 30 20 10 0 0

120

240

360

480

600

Time (min)

Figure 17.5 Variation of H2 O2 concentration with time during the electrolysis of 2.5 L of a 0.05 M Na2 SO4 solution of pH 3.0 in the flow plant shown in Figure 17.3 at: (•) 150, () 100, and () 50mA cm−2 , 25◦ C, and liquid flow rate of 180 L h−1 . Data of curve () correspond to a 100 mg L−1 MCPP solution with 0.5 mM of Fe2+ of pH 3.0 degraded at 50 mA cm−2 by SPEF. (Reprinted from Ref. 29.)

electrogenerated H2 O2 from Fenton’s Reaction (17.11). However, a BDD anode causes a much faster destruction of Fe2+ from Reaction (17.29) and Fe2+ is poorly accumulated, disappearing by prolonging the electrolysis (see Figure 17.4b). This behavior indicates that greater amounts of homogeneous • OH are obtained at the Pt/carbon felt cell under EF conditions, when only a small catalytic concentration of Fe2+ or Fe3+ is required, even lower than 0.2 mM, to obtain the maximum rate for • OH production. For undivided cells equipped with a carbon-PTFE O2 -diffusion cathode, H2 O2 always reaches a steady concentration, which rises linearly with increasing applied current as a consequence of the concomitant acceleration of Faradaic Reactions (17.1) and (17.4). Figure 17.5 exemplifies this trend for the electrolysis of 2.5 L of a 0.05 M Na2 SO4 solution of pH 3.0 in the batch recirculation flow plant of Figure 17.3 containing a filterpress reactor with a 20-cm2 BDD anode and a 20-cm2 O2 -diffusion cathode by applying between 150 and 50 mA cm−2 [29]. A high decay of the H2 O2 concentration takes place when a 100 mg L−1 solution of the herbicide mecoprop (MCPP) with 0.5 mM Fe2+ of pH 3.0 is treated by SPEF at 50 mA cm−2 if the electrochemical reactor is coupled to a solar photoreactor of 600-mL of irradiated volume, owing to the acceleration of Fenton’s Reaction (17.11) by the rapid consumption of • OH to degrade the organic matter.

17.4

DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

We have reported the destruction of aromatic pollutants by Fenton’s electrochemistry using a BDD/O2 cell. Some common herbicides [21,28], dyes [22,23], and pharmaceuticals [25–27,30,31,33,35] and amino acids precursors [34] have been degraded in a small undivided electrolytic tank reactor such as that in Figure 17.1 containing a 3-cm2 BDD thin layer deposited on a conductive Si sheet from CSEM as anode and a 3-cm2 carbon-PTFE O2 -diffusion cathode from E-TEK fed with pure gas at a flow rate of

414

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

12–20 mL min−1 . The contaminated solutions contained 0.05 M Na2 SO4 as supporting electrolyte and their pH was adjusted in the range 2.0–6.0 with H2 SO4 . Electrolyses were performed at constant current between 100 and 450 mA and temperature of 25◦ C or 35◦ C. EF with BDD was usually carried out by adding 0.2–2.0 mM FeSO4 to initial solutions, whereas a 6-W UVA light was used to irradiate the solution in PEF. The characteristics of these degradations will be presented and discussed in following subsections. Apart from these studies, Montanaro, Petrucci, and Merli [32] considered a divided BDD/O2 cell with an anionic membrane to reduce the phosphorous content of a real effluent from the manufacture of phosphorous-based flame retardants. The process was initiated by the treatment of 100 mL of a fresh solution by AO with BDD and the resulting pretreated solution was further oxidized by EF in the catholyte. The anionic membrane allowed the passage of OH− ions from the cathodic to the anodic compartment to maintain the catholyte pH close to 1.5 for avoiding iron precipitation. A sequential running saved charge and time by using both anode and cathode performances in parallel, only requiring 240 min at 10 mA cm−2 to decrease the phosphorous content below the limits needed. 17.4.1

Herbicides

The treatment of 100 mL of solutions with common chlorophenoxy acid herbicides such as 4-CPA (4-chlorophenoxyacetic acid), MCPA (4-chloro-2-methylphenoxyacetic acid), 2,4-D (2,4-dichlorophenoxyacetic acid), 2,4,5-T (2,4,5-trichlorophenoxyacetic acid), and 2,4-DP (2-(2,4-dichlorophenoxy)-propionic acid) has been studied by EAOPs with BDD or Pt anodes under comparable conditions [17–21,28]. These compounds showed similar degradation behavior and the oxidation power of the methods increased in the sequence AO-H2 O2 <EF
17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

120

415

(a)

TOC (mg L–1)

100 80 60 40 20 0

0

4

8

12

16

20

Specific charge (Ah L–1)

300

(b)

3 ln (c0 /c)

Concentration (mg L–1)

250 200

2 1

150 0 0

100

5

10 15 Time (min)

20

50 0

0

10

20

30

40

Time (min) Figure 17.6 (a) TOC abatement with specific charge for the degradation of 100 mL of 230 mg L−1 2,4-D solutions of pH 3.0 using the cell of Figure 17.1 with a 3-cm2 carbon-PTFE O2 -diffusion cathode at 300 mA and 35◦ C. (◦) Anodic oxidation with electrogenerated H2 O2 (AO-H2 O2 ) with a 10-cm2 Pt anode, (•) EF with a 10-cm2 Pt anode and 1.0 mM Fe2+ , () AO-H2 O2 with a 3-cm2 BDD anode, and () EF with a 3-cm2 BDD anode and 1.0 mM Fe2+ . (b) Herbicide concentration decay during the degradation of: (•) 194 mg L−1 4-CPA, () 200 mg L−1 MCPA, () 230 mg L−1 2,4-D, and () 266 mg L−1 2,4,5-T solutions of pH 3.0 at 100 mA and 35◦ C by EF with BDD and 1.0 mM Fe2+ . The inset panel gives the corresponding kinetic analysis assuming a pseudo–first-order reaction for each compound. (Reprinted from Ref. 21.)

The comparative oxidation power of all EAOPs can be better explained from their mineralization current efficiency (MCE, in %), calculated for each treated solution at a given electrolysis time t (h) from Equation (17.32) [24]: MCE =

nFVs (TOC)exp × 100 4.32 × 107 mIt

(17.32)

where n is the number of electrons consumed in the mineralization process, F is the Faraday constant (= 96, 487 C mol−1 ), Vs is the solution volume (L), (TOC)exp is the

416

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

experimental TOC decay (mg L−1 ), 4.32 × 107 is a conversion factor involving the molar mass of carbon (= 3600 s h−1 × 12,000 mg C mol−1 ), m is the number of carbon atoms in the pollutant molecule, and I is the current (A). For the aforementioned herbicides, total mineralization involves their conversion into carbon dioxide and chloride ion with n = 32 for 4-CPA from reaction (17.33), n = 38 for MCPA from reaction (17.34), n = 30 for 2,4-D from Reaction (17.35), and n = 28 for 2,4,5-T from reaction (17.36): C8 H7 ClO3 + 13H2 O → 8CO2 + Cl− + 33H+ + 32e−

(17.33)

C9 H9 ClO3 + 15H2 O → 9CO2 + Cl− + 39H+ + 38e− −

+

C8 H6 Cl2 O3 + 13H2 O → 8CO2 + 2Cl + 32H + 30e

(17.34) −

C8 H5 Cl3 O3 + 13H2 O → 8CO2 + 3Cl− + 31H+ + 28e−

(17.35) (17.36)

Table 17.2 summarizes the percentage of TOC removal and MCE values found after 3 h of electrolysis at 100 mA of solutions with 100 mg L−1 TOC of 4-CPA, MCPA, 2,4-D, and 2,4,5-T of pH 3.0 using Pt/O2 and BDD/O2 cells [17–21]. In the first system, decontamination degrees of 12-19% for AO-H2 O2 , 53–66% for EF, and 90–99% for PEF are obtained, corresponding to efficiencies of 3.1–6.0, 14–20, and 25–29%, respectively. The first two methods with Pt do not allow total mineralization and are much less TABLE 17.2 Percentage of TOC removal and mineralization current efficiency after 3 h of treatment of 100 mL solutions of 100 mg L−1 TOC of chlorophenoxyacetic acid herbicides at pH 3.0 by EAOPs at 100 mA. The last column collects the pseudo–first-order rate constant found for their reaction with hydroxyl radical [17–21]. Methoda

T (◦ C)

% TOC removal

MCE (%)

k1 (s−1 )

4-CPA

AO-H2 O2 with Pt AO-H2 O2 with BDD EF with Pt EF with BDD PEF with Pt

35

19 54 66 75 96

5.6 16 20 22 29

8.2 × 10−5 9.0 × 10−5 2.0 × 10−3 4.5 × 10−3 2.7 × 10−3

MCPA

AO-H2 O2 with Pt AO-H2 O2 with BDD EF with Pt EF with BDD PEF with Pt

19 58 65 76 91

6.0 18 20 24 29

9.8 × 10−5 1.2 × 10−4 2.0 × 10−3 2.3 × 10−3 2.3 × 10−3

2,4-D

AO-H2 O2 with Pt AO-H2 O2 with BDD EF with Pt EF with BDD PEF with Pt

25 35 25 35 25

13 57 57 78 90

3.6 16 16 22 25

1.3 × 10−4 1.2 × 10−4 3.0 × 10−3 5.2 × 10−3 3.8 × 10−3

2,4,5-T

AO-H2 O2 with Pt AO-H2 O2 with BDD EF with Pt EF with BDD PEF with Pt

35

12 59 53 80 99

3.1 15 14 21 26

2.0 × 10−4 1.1 × 10−4 1.8 × 10−3 4.0 × 10−3 2.0 × 10−3

Herbicide

a AO-H O with Pt: anodic oxidation with electrogenerated H O in a Pt/O cell; AO-H O with BDD: anodic 2 2 2 2 2 2 2 oxidation with electrogenerated H2 O2 in a BDD/O2 cell; EF with Pt: electro-Fenton with 1 mM Fe2+ in a Pt/O2 cell; EF with BDD: electro-Fenton with 1 mM Fe2+ in a BDD/O2 cell; PEF with Pt: photoelectro-Fenton with 1 mM Fe2+ in a Pt/O2 cell under 6-W UVA irradiation.

17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

417

efficient than AO-H2 O2 and EF with BDD, respectively, which mineralize completely all solutions practically at 6 h (see, for example, the case of 2,4-D in Figure 17.6a). This confirms the much greater oxidation power of BDD than Pt, producing enough amount of reactive BDD( • OH) to destroy the initial contaminants and their by-products. Using a Pt anode, complete mineralization can only be attained using PEF owing to the efficient photodecarboxylation of Fe(III)-oxalate complexes as ultimate by-products under UVA irradiation [17–20]. The decay kinetics for the above herbicides was monitored by reversed-phase chromatography. Figure 17.6b exemplifies their quick disappearance by EF with BDD, varying from 12 min for 2,4-D to 30 min for MCPA. The inset of Figure 17.6b illustrates that they follow a pseudo–first-order reaction with the different hydroxyl radicals. This behavior was also observed for all procedures. The last column of Table 17.2 shows a similar pseudo–first-order rate constant (k1 ) for both AO-H2 O2 methods, regardless of the anode used, indicating that the initial chloroaromatics are removed at similar rate by BDD( • OH) and Pt( • OH). Similar and two-order magnitude greater k1 -values can be observed for EF with Pt, EF with BDD, and PEF with Pt, thereby confirming the quicker reaction of herbicides with • OH generated from Fenton’s Reaction (17.11) without significant participation of the photolytic Reaction (17.30). The greater oxidation ability of AO-H2 O2 and EF using a BDD/O2 cell compared with a Pt/O2 one can then be related to the faster mineralization of final by-products with BDD( • OH) than with Pt( • OH). Gas chromatography-mass spectrometry (GC-MS) and high-performance liquid chromatography (HPLC) analysis of solutions electrolyzed in a BDD/O2 cell revealed the formation of primary phenol intermediates such as 4-chlorophenol for 4-CPA, 4-chloroo-cresol for MCPA, 2,4-dichlorophenol for 2,4-D, and 2,4,5-trichlorophenol for 2,4,5-T [21]. These species react rapidly with the same oxidant since they were only detected while starting herbicides were destroyed. Moreover, HPLC analysis of generated carboxylic acids revealed the large persistence of the ultimate oxalic acid in AO-H2 O2 or its Fe(III) complexes in EF, since these species can only be converted to CO2 by BDD( • OH). This explains the long time for total mineralization (tTM ) of pollutants in EF with BDD. The much greater mineralization found for PEF with Pt is related to the fast photolysis of Fe(III)-oxalate complexes [17]. The Cl− ion was also released during all the indirect electro-oxidation processes of chlorophenoxy acid herbicides. This ion remained stable in all EAOPs tested with the Pt/O2 cell, but it was slowly removed from the medium in the analogous processes with the BDD/O2 cell, since it was converted into Cl2 by reaction with BDD( • OH) [49]. The great oxidation power of EAOPs with BDD was confirmed from the degradation of 2,4-DP [28]. It was observed again that overall decontamination was reached in EF with 1 mM Fe2+ because the great production of • OH in the bulk from Fenton’s Reaction (17.11) destroyed rapidly the aromatic compounds, whereas the high generation of reactive BDD( • OH) favored the removal of final carboxylic acids. UVA irradiation in PEF also had little effect on the degradation rate of aromatics. Chlorohydroquinone and chloro-p-benzoquinone were detected as primary aromatic intermediates, further on being oxidized to maleic, fumaric, malic, lactic, pyruvic, acetic, formic, and oxalic acids. 17.4.2

Dyes

The higher performance of using a BDD/O2 tank reactor instead of a Pt/O2 cell has been corroborated from the comparative treatments of acidic aqueous solutions containing up

418

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

to 0.9 g L−1 of the dye Indigo Carmine (C. I. Acid Blue 64) by AO-H2 O2 [22] and by EF and PEF [23]. As shown in Figure 17.7a for 220 mg L−1 of the dye at pH 3.0 and 33 mA cm−2 , the EF treatment with Pt and 1.0 mM Fe2+ promotes a fast degradation to 46% mineralization at Q = 3Ah L−1 (3 h), but at longer time decontamination becomes so slow that TOC is only reduced by 49% at Q = 9 Ah L−1 (9 h).The PEF method with Pt only leads to partial mineralization of 84% at the end of electrolysis, as expected if other Fe(III) complexes than those of oxalic acid cannot be photolyzed by UVA light. A quicker TOC removal occurs by EF with BDD and 1.0 mM Fe2+ , with 91% of TOC reduction at 9 h and complete mineralization at tTM = 13 h, indicating that all complexes of Fe(III) with final carboxylic acids are efficiently oxidized with BDD( • OH). Figure 17.7a also

120

(a)

TOC (mg L–1)

100 80 60 40 20 0

120

0

2

4 6 Specific charge (Ah L–1)

8

10

(b)

TOC (mg L–1)

100 80 60 40 20 0

0

3

6

9

12

15

Specific charge (Ah

18

21

24

L–1)

Figure 17.7 (a) TOC decay versus specific charge for the mineralization of 100 mL of 220 mg L−1 Indigo Carmine solutions in 0.05 M Na2 SO4 of pH 3.0 at 33 mA cm−2 and 35.0◦ C using the cell of Figure 17.1 with 3-cm2 electrodes. (•) EF with Pt and 1.0 mM Fe2+ , () PEF with 1.0 mM Fe2+ , () EF with BDD and 1.0 mM Fe2+ , and () PEF with Pt and 1.0 mM Fe2+ + 0.25 mM Cu2+ . (b) TOC removal with specific charge for the treatment of the above solution at: () 33, () 100, and ( ) 150 mA cm−2 by EF with BDD and 1.0 mM Fe2+ . (Reprinted from Ref. 23.)

17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

419

shows that overall destruction of all by-products at Q = 7 Ah L−1 (7 h) is possible by PEF with Pt if 1.0 mM Fe2+ + 0.25 mM Cu2+ are combined as co-catalysts. For all the EAOPs tested, a quicker mineralization was found with increasing current density by the concomitant generation of more BDD( • OH) or Pt( • OH), and/or homogeneously formed • OH because more H2 O2 is accumulated in the bulk (see Figure 17.5). Nevertheless, Q increases due to the higher loss of charge associated to parasitic reactions, as explained in the case of herbicides (see Subsection 17.4.1). This phenomenon is exemplified in Figure 17.7b for the EF treatment with BDD, since the specific charge for total mineralization (QTM ) raises from 12 Ah L−1 at 33 mA cm−2 to 22.5 Ah L−1 at 150 mA cm−2 , while the corresponding tTM drops from 12 to 5 h. The degradation rate also rises with increasing dye concentration, as can be seen in Figure 17.8a for EF with BDD at 100 mA cm−2 . This trend is confirmed by the increasing efficiency depicted in Figure 17.8b as a result of the larger extent of reactions involving the destruction of organics with hydroxyl radicals and the concomitant decay in the rate of nonoxidizing parasitic reactions. Note that the efficiency decreases at long electrolysis times owing to the progressive decay of organic content that, moreover, is more difficult to be oxidized with homogeneous and heterogeneous • OH. 500 (a)

TOC (mg L–1)

400 300 200 100 0

(b)

MCE (%)

50 40 30 20 10 0

0

5

10

15

20

25

30

Specific charge (Ah L–1) Figure 17.8 Dependence of (a) TOC and (b) mineralization current efficiency on specific charge for 100 mL of solutions with Indigo Carmine concentration of: (•) 881, () 440, () 220, and () 112 mg L−1 at pH 3.0, 100 mA cm−2 , and 35.0◦ C treated by EF with BDD and 1.0 mM Fe2+ . (Reprinted from Ref. 23.)

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

2.0

15

1.5

10

1.0

5

0.5

+

20

0 0

60

[NO3– ] (mg L–1)

[NH4 ] (mg L–1)

420

0.0 120 180 240 300 360 420 480 540 600 Time (min)

Figure 17.9 Concentration of (•,) ammonium and (◦,) nitrate ions released during the treatment of 100 mL of 220 mg L−1 Indigo Carmine solutions of pH 3.0 at 33 mA cm−2 and 35.0◦ C by (•,◦) EF with BDD and 1.0 mM Fe2+ and (,) PEF with Pt and 1.0 mM Fe2+ + 0.25 mM Cu2+ . (Reprinted from Ref. 23.)

The degradation process of Indigo Carmine is accompanied by the complete loss of the initial nitrogen in the form of NH+ 4 as major inorganic ion, along with a much smaller fraction of NO− , as can be observed in Figure 17.9 for the two more potent EAOPs 3 tested, EF with BDD and PEF with Pt and 1.0 mM Fe2+ + 0.25 mM Cu2+ . Reversed-phase HPLC analyses of treated solutions showed that Indigo Carmine followed a pseudo–zero-order reaction kinetics, disappearing at the same time as its aromatic derivatives isatin 5-sulfonic acid, indigo, and isatin, mainly by reaction with the • OH radicals produced from Fenton’s Reaction (17.11). Ion-exclusion HPLC revealed the formation of oxalic and oxamic acids as ultimate by-products, and the evolution of their Fe(III) and/or Cu(II) complexes shown in Figures 17.10a and 17.10b allowed the explanation of the different oxidation power of EAOPs. So, Fe(III)-oxalate and Fe(III)-oxamate remain stable in EF with Pt because they cannot be oxidized either with Pt( • OH) or with • OH, but the former kinds of complexes are efficiently photolyzed under UVA irradiation in PEF with Pt. The opposite behavior can be observed for EF with BDD where both complexes are completely destroyed with BDD( • OH). When Cu2+ is used as cocatalyst in PEF with Pt, Cu(II)-oxalate and Cu(II)-oxamate are competitively produced and completely destroyed by Pt( • OH) and/or • OH. This evidences the synergistic action of combining Fe2+ , Cu2+ , and UVA light for the treatment of wastewaters with aromatic pollutants containing nitrogen if oxamic acid is formed as by-product. Based on these results, the reaction pathways of Figures 17.11a and 17.11b were proposed to describe the mineralization of oxalic and oxamic acids, respectively, in the situations tested. 17.4.3

Pharmaceuticals and Amino Acids Precursors

The application of Fenton’s electrochemistry was also extended to the removal of common pharmaceuticals from wastewaters. Thus, the AO-H2 O2 , EF, PEF, and/or SPEF treatments of the blood lipid regulator metabolite clofibric acid (2-(4-chlorophenoxy)2-methylpropionic acid) [25], the NSAID salicylic acid (2-hydroxybenzoic acid) [30],

17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

421

120 (a) [Oxalic acid] (mg L–1)

100 80 60 40 20

[Oxamic acid] (mg L–1)

0

(b)

50 40 30 20 10 0 0

120

240

360

480

600

720

840

Time (min)

Figure 17.10 Evolution of the concentration of (a) oxalic and (b) oxamic acids during the degradation of a 220 mg L−1 Indigo Carmine solution under the conditions given in Figure 17.7. (Reprinted from Ref. 23.)

the biocide chloroxylenol (4-chloro-3,5-dimethylphenol) [31], the anti-inflammatory ibuprofen (2-(4-isobutylphenyl)propionic acid) [33], and the fluoroquinolone antibiotic enrofloxacin [35] were comparatively tested in undivided Pt/O2 and BDD/O2 cells. Similarly to previously studied herbicides and dyes, it was found that the oxidation power of these methods also increased in the order AO-H2 O2 <EF
422

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

(a)

COOH COOH Oxalic acid Fe

3+

Cu2+ •

OH Cu(II)-oxalate complexes

Fe(III)-oxalate complexes hν •

BDD( OH)

–Fe2+

•

OH

CO2

(b)

COOH CONH2 Oxamic acid Cu2+

Fe3+

•

Fe(III)-oxamate complexes

BDD(•OH)

OH

CO2

Cu(II)-oxamate complexes

•

OH

+ NH4 –

NO3

Figure 17.11 Proposed pathways for the mineralization of (a) oxalic and (b) oxamic acids by the EF and PEF processes. BDD( • OH) denotes the hydroxyl radical generated from water oxidation at the BDD anode.

only feasible for the more potent method of SPEF with BDD owing to the formation of small amount of very stable by-products. As expected, the MCE values of the earlier treatments depicted in Figure 17.12b follow the same tendency as the oxidation power of the applied method. For the most efficient process of SPEF with BDD, a maximum MCE of 34% is obtained at Q = 0.66 Ah L−1 (40 min) when aromatics are totally removed and not as easily oxidizable organics such as carboxylic acids are accumulated, since they are more slowly transformed into CO2 . The ibuprofen decay followed pseudo–first-order kinetics with very similar k1 -values of 1.86 × 10−3 s−1 for EF with Pt, 2.00 × 10−3 s−1 for PEF with Pt, 2.06 × 10−3 s−1 for EF with BDD, and 2.08 × 10−3 s−1 for PEF with BDD. This means that the drug

17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

35

423

(a)

TOC (mg L–1)

30 25 20 15 10 5 0

(b)

30

MCE (%)

25 20 15 10 5 0

0

1

2

3

4

5

6

7

Specific charge (Ah L–1) Figure 17.12 (a) TOC decay and (b) mineralization current efficiency versus consumed specific charge for the degradation of 100 mL of a solution with 41 mg L−1 ibuprofen (near saturation), 0.5 mM Fe2+ , and 0.05 M Na2 SO4 of pH 3.0 using the cell of Figure 17.1 with 3-cm2 electrodes at 33.3 mA cm−2 and 25.0◦ C. (◦) EF with Pt, (•) EF with BDD, () PEF with Pt, () PEF with BDD, () SPEF with Pt, and () SPEF with BDD. (Reprinted from Ref. 33.)

and its aromatic by-products react primordially with homogeneous • OH formed in the bulk from Fenton’s Reaction (17.11). When the most potent SPEF method with Pt or BDD was tested, ibuprofen disappeared much more rapidly as a result of the generation of larger amounts of • OH from the additional participation of the photolytic Reaction (17.30) induced by the high intensity of solar irradiation. Similar trends for TOC and pollutant concentration decays were found for clofibric acid, salicylic acid, chloroxylenol, and enrofloxacin by the same EAOPs. For all procedures, increasing MCE was obtained with increasing initial concentration or decreasing current density due to the minimization of parasitic reactions, as explained earlier in the case of Indigo Carmine (see Subsection 17.4.2). Analysis of treated ibuprofen solutions by GC-MS and HPLC allowed the identification of aromatic intermediates such as 1-(1-hydroxyethyl)-4-isobutylbenzene, 4isobutylphenol, 4-isobutylacetophenone, and 4-ethylbenzaldehyde, and carboxylic acids such as pyruvic, acetic, formic, and oxalic. As exemplified in Figure 17.13a for 4isobutylacetophenone and Figure 17.13b for the ultimate oxalic acid, all by-products

424

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

[4-isobutylacetophenone] (mg L–1)

1.5 (a)

1.0

0.5

0.0 0

10

20

30 40 Time (min)

50

60

70

60

120

180

300

360

420

20

[Oxalic acid] (mg L–1)

(b) 15

10

5

0 0

240

Time (min) Figure 17.13 Evolution of the concentration of (a) 4-isobutylacetophenone detected as the main aromatic intermediate and (b) oxalic acid detected as the final carboxylic acid during the degradation of ibuprofen for the same trials as Figure 17.12. (Reprinted from Ref. 33.)

were more rapidly destroyed in SPEF, practically independent of the anode used. This EAOP is the most potent since the high intensity of solar irradiation enhances the photolysis of some aromatics and their degradation from • OH production by photocatalytic Reaction (17.30), as well as the photodecomposition of Fe(III)-carboxylate complexes such as Fe(III)-oxalate. On the basis of these results, the reaction sequence of Figure 17.14 involving the detected aromatics is proposed to explain the initial degradation of ibuprofen by EF, PEF, and SPEF under the action of BDD( • OH), Pt( • OH), • OH, and/or photons. The intermediates detected in the electrolyzed solutions of the other pharmaceuticals were (1) 4-chlorophenol, 4-chlorocatechol, hydroquinone, p-benzoquinone, and 2hydroxyisobutyric, tartronic, maleic, fumaric, formic, and oxalic acids for clofibric acid; (2) 2,3-dihydroxybenzoic, 2,5-dihydroxybenzoic, 2,6-dihydroxybenzoic, α-ketoglutaric, glycolic, glyoxylic, maleic, fumaric, malic, tartronic, and oxalic acids for salicylic acid; (3) 2,6-dimethylhydroquinone, 2,6-dimethyl-p-benzoquinone, 3,5-dimethyl-2-hydroxy-pbenzoquinone, and maleic, malonic, pyruvic, acetic, and oxalic acids for chloroxylenol;

425

17.4 DEGRADATION OF ORGANICS IN BDD/O2 TANK REACTORS

O H3C

O H3C

OH •OH,

H3C

OH •OH, Pt(•OH)

Pt(•OH)

BDD(•OH) H3C

H3C

BDD(•OH),hν –CO2 –CH3-CHOH-CH3

O 4-ethylbenzaldehyde

OH CH3

CH3 Ibuprofen

2-[4-(1-hydroxyisobutyl)phenyl] propionic acid

Pt(•OH) BDD(•OH) • OH

O OH H3C

H3C

OH

H3C

O

OH

OH •OH,

BDD( –CO2 H3C

•OH,

Pt(•OH) •OH),

hν

2-(4-isobutylphenyl)2-hydroxypropionic acid

OH, Pt(•OH)

BDD(•OH) –CH3-COOH

•OH)

BDD(

H3C CH3

•

Pt(•OH)

H3C CH3

1-(1-hydroxyethyl)4-isobutylbenzene

CH3 4-isobutylacetophenone

H3C CH3

4-isobutylphenol

Figure 17.14 Proposed reaction scheme for the initial degradation of ibuprofen by EF, PEF, and SPEF. The sequence includes aromatics detected and hypothetical intermediates in brackets. Pt( • OH) and BDD( • OH) represent the hydroxyl radical electrogenerated from water oxidation at the Pt and BDD anode, respectively, and • OH denotes the hydroxyl radical produced in the medium. (Reprinted from Ref. 33.)

and (4) hydroxylated aromatics, diethylamine, ethylformamide, and succinic, tartaric, maleic, fumaric, malonic, oxalic, oxamic, carbamic, and acetyl carbamic acids for enrofloxacin. Hydroxylation followed by the generation of short carboxylic acids is then the main degradative route of all these aromatic drugs. More recently, a common precursor of drugs, the amino acid α-methylphenylglycine (α-MPG, S-2-amino-2-phenylpropionic acid), was degraded by EF and SPEF using the BDD/O2 reactor of Figure 17.1 [34]. Solutions (250 mL) of 500 mg L−1 α-MPG in 0.05 M Na2 SO4 with a small Fe2+ content of 10 mg L−1 at pH 2.9 were comparatively treated by both EAOPs at several constant currents and by the analogous chemical Fenton and solar photo-Fenton processes with H2 O2 addition. From the αMPG concentration and TOC removals with time, it was established that the electrochemical methods were faster than the chemical ones. Furthermore, the presence of sunlight was beneficial for both configurations. The SPEF treatment with BDD was then found to be the most viable process for the degradation of this amino acid. GC-MS analysis of degraded solutions revealed the presence of aromatic by-products such as α–amino–hydroxy–α–methylbenzeneacetic acid, α–amino–α–methylbenzenemethanol, α-amino–hydroxy–α–methylbenzenemethanol, phenol, o-, m-, and p-acetophenol, o-,

426

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

m-, and p-hydroxyphenol, acetophenone, and benzoic and hydroxybenzoic acids, along with carboxylic acids such as succinic, malonic, maleic, fumaric, tartaric, oxalic, and oxamic acids. The superiority of solar-driven processes was related to the rapid photolysis of Fe(III)-oxalate complexes from Reaction (17.31), along with the slower, but efficient, removal of Fe(III)-oxamate ones. Although the initial N content was converted − (in smaller proportion) in all cases, their mass balance was not into NH+ 4 and NO3 closed, probably due to the generation of NOx or other volatile nitrogen species.

17.5 DEGRADATION OF ORGANICS IN OTHERS TANK REACTORS WITH A BBD ANODE In a collaboration between our laboratory and Oturan’s group, 200 mL solutions of the antimicrobials chlorophene (o-benzyl-p-chlorophenol) [26], and triclocarban (N -(4-chlorophenyl)-N  -(3,4-dichlorophenyl)urea) and triclosan (2,4,4 -trichloro-2 hydroxydiphenyl ether) [27] with Fe3+ as catalyst were comparatively degraded by EF using the undivided cells of Figures 17.1 and 17.2. Since in the Pt/O2 and BDD/O2 cells, H2 O2 was largely accumulated but Fe3+ content remained unchanged (see Figure 17.4a), the pollutant decay was enhanced by increasing the initial Fe3+ content that promoted faster Fe2+ regeneration at the cathode with greater • OH production from Fenton’s Reaction (17.11). In contrast, when the carbon-felt cathode was employed, Fe2+ was largely regenerated and 0.2 mM Fe3+ was only needed to obtain the maximum • OH generation rate (see Figure 17.4b). Under these latter conditions, k2 -values of 1.00 × 1010 and 5.49 × 109 M−1 s−1 were obtained for the corresponding decay reactions of chlorophene and triclosan. Poor mineralization degree was found for the Pt/O2 cell because of the difficult oxidation of final Fe(III)-oxalate with • OH. These complexes were totally destroyed using a BDD anode thanks to the formation of the stronger oxidant BDD( • OH) on its surface. Overall mineralization was also achieved in both Pt/carbon felt and BDD/carbon felt cells owing to the efficient oxidation of Fe(II)-oxalate complexes with • OH in the bulk. The highest oxidation power regarding total mineralization was reached for the BDD/carbon felt cell by the additional destruction of Fe(II)-oxalate complexes with BDD( • OH). The overall release of Cl− was confirmed for all treatments. Primary intermediates such as 2,4-dichlorophenol, 4-chlorocatechol, chlorohydroquinone, and chloro-p-benzoquinone were identified for triclosan, whereas urea, hydroquinone, chlorohydroquinone, 1-chloro-4-nitrobenzene, and 1,2-dichloro-4-nitrobenzene were detected for triclocarban. The efficient degradation of 500 mL of the azo dye Acid Orange 7 in an O2 -saturated 0.05 M Na2 SO4 solution with 0.10 mM Fe2+ at pH 3.0 by EF in BDD/carbon felt and Pt/carbon felt cells such as the one shown in Figure 17.2 was further reported by Oturan’s group [43]. Similar dye decays and TOC removals were found for both systems, although the former gave slightly greater performance owing to the faster reaction of organics with the stronger oxidant BDD( • OH) than with Pt( • OH). Thus, for the destruction of 0.05 mM dye at 60 mA, a k1 -value of 0.67 min−1 or 0.51 min−1 for its reaction with hydroxyl radical was obtained with a 35-cm2 BDD or a 4.4.5-cm2 Pt anode and a 60-cm2 carbon-felt cathode, respectively. A k2 -value of (1.10 ± 0.04) × 1010 M−1 s−1 was found in both cases, as determined from the competition kinetic method. More than 98% TOC reduction was reached for a 0.53 mM dye solution after 9 h of electrolysis in both systems operating at 250 mA, indicating that all by-products

17.6 DEGRADATION OF ORGANICS IN BATCH RECIRCULATION BDD/O2 FLOW CELLS

427

and their Fe(III) or Fe(II) complexes can be efficiently removed under the action of homogeneous and heterogeneous hydroxyl radicals. Sulfanilic and phthalic acids, 2-naphthol, 1-amino-2-naphthol, 1,2-naphthoquinone, and 4-hydroxybenzensulfonate were identified as aromatic intermediates coming from the cleavage of the –N=N– bond of Acid Orange 7, followed by consecutive hydroxylations. Maleic, acetic, formic, and oxalic acids were detected as generated carboxylic acids. Ammonium, in much major proportion than nitrate, and sulphate were released as inorganic ions to the medium. The superiority of the EF process using a BDD/carbon felt cell compared with a Pt/carbon felt cell has been corroborated again for the removal of the herbicide atrazine (2-chloro-4-(ethylamine)-6-(isopropylamine)-s-triazine) [44]. After 10 h of electrolysis of 150 mL of an O2 -saturated 0.20 mM herbicide solution with 0.1 mM Fe3+ of pH 3.0 at 250 mA, 82% mineralization was obtained using a 15-cm2 BDD anode and a 60-cm2 carbon-felt cathode, a value higher than 72% found for an analogous system with a 4.5-cm2 Pt anode. This behavior was associated with the partial oxidation of the end aromatic by-product, the cyanuric acid, with BDD( • OH) to yield short linear carboxylic acids that are converted into CO2 . Note that this is feasible due to the higher oxidation power of the BDD anode, since this acid is not oxidized using a Pt anode, as well as nonelectrochemical advanced oxidation processes. Recently, a novel three-dimensional carbon sponge electrode has been tested as cathode for the EF treatment of the herbicide propham (isopropyl phenylcarbamate) with a BDD or Pt anode [45]. Again, the use of BDD favored the degradation rate, although almost overall mineralization (99% TOC removal) was achieved in all cells after 6 h of EF treatment of 150 mL of an O2 -saturated 0.5 mM propham solution with 0.2 mM Fe3+ at pH 3.0, 100 mA, and 35◦ C using a cell similar to that of Figure 17.2 with a 15-cm2 anode and a 1.0-cm × 1.0-cm × 4.0-cm cathode. At a given time, a gradual decrease in MCE with increasing current from 100 to 450 mA was observed, in the same way as reported above for tank reactors with an O2 -diffusion cathode [23]. The slower TOC destruction for electrolysis times longer than 2 h was related to the difficult oxidation of Fe(III) and Fe(II) complexes of the ultimate oxalic and oxamic acids. The initial N content of the herbicide was released in the form of NH4 + , along with NO3 − to a lesser extent. 17.6 DEGRADATION OF ORGANICS IN BATCH RECIRCULATION BDD/O2 FLOW CELLS Taking into account the large oxidation power and performance of SPEF with BDD, this method was tested in our laboratory using the batch recirculation flow plant schematized in Figure 17.3 to degrade cresols [24], the herbicide MCPP (2-(4-chloro2-methylphenoxy)-propionic acid) [29], and the fluoroquinolone enrofloxacin [35], as a first step of its scaling-up to investigate its viability at industrial scale. These trials were carried out at constant current (I , in A) and the energy cost per unit volume (EC, in kWh m−3 ) and per unit TOC mass (EC, in kWh (g TOC)−1 ) at time t (h), and the energy cost for total mineralization (ETM , in kWh m−3 ) at tTM (h) were calculated from Equations (17.37) through (17.39): EC (kWh m−3 ) =

VIt Vs

(17.37)

428

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

EC (kWh(g TOC)−1 ) = ETM (kWh m−3 ) =

VIt (TOC)exp Vs

(17.38)

VItTM Vs

(17.39)

where V is the average cell voltage (V), Vs is the solution volume (L), and (TOC)exp is the experimental TOC removal (mg L−1 ). Figure 17.15a shows that the degradation of 2.5 L of 100 mg L−1 of MCPP with 0.5 mM Fe2+ of pH 3.0 at 50 mA cm−2 and 25◦ C is much faster for SPEF than for EF or PEF with UVA irradiation [29]. The latter process was applied by replacing the solar

60

(a)

TOC (mg L–1)

50 40 30 20 10 0

0

1

2

3

Specific charge (Ah 400

4

L–1)

(b)

350 TOC (mg L–1)

300 250 200 150 100 50 0

0

120

240

360

480

600

Time (min) Figure 17.15 (a) Variation of TOC with specific charge for the treatment of 2.5 L of 100 mg L−1 mecoprop solutions in 0.05 M Na2 SO4 of pH 3.0 using the recirculation flow plant of Figure 17.3 at 50 mA cm−2 , 25◦ C, and liquid flow rate of 180 L h−1 . (•) AO-H2 O2 with BDD, () EF with BDD, () PEF with BDD and a 160 W UVA lamp, and () SPEF with BDD. In the three latter methods 0.5 mM Fe2+ was added as catalyst. (b) TOC decay with time for solutions with a MCPP concentration of (•) 634, () 375, () 200, and () 100 mg L−1 , by SPEF with BDD. (Reprinted from Ref. 29.)

17.6 DEGRADATION OF ORGANICS IN BATCH RECIRCULATION BDD/O2 FLOW CELLS

120

429

(a)

[MCPP] (mg L–1)

100 80 60 40 20 0

0

5

10

15

20

Time (min) 700

(b)

4 ln (C0 /C )

[MCPP] (mg L–1)

600 500 400

3 2 1

300 0 0

200

10

20

100 0

30

40

50

Time (min)

0

10

20

30

40

50

60

Time (min) Figure 17.16 Mecroprop decay with time during the treatment of 2.5 L of herbicide solutions in the same conditions as Figure 17.15. In plot (a), degradation of 100 mg L−1 MCPP by: (•) UVA light (without current), () EF with BDD, () PEF with BDD, and () SPEF with BDD. In plot (b), treatment of: (•) 634, () 375, () 200, and () 100 mg L−1 of herbicide using SPEF with BDD. The inset panel presents the corresponding kinetic analysis assuming a pseudo–first-order reaction for MCCP. (Reprinted from Ref. 29.)

photoreactor with an annular photoreactor containing a 160-W UVA light. As expected, AO-H2 O2 leads to the slowest TOC abatement. In the SPEF process, however, TOC is very slowly removed after being reduced by 82% at 100 min (0.67 Ah L−1 ), attaining 96% mineralization at 9 h (3.6 Ah L−1 ) with 46-kWh m−3 energy cost, practically under the same conditions as the PEF process. Similar mineralization degrees were obtained after 9 h of SPEF by increasing the herbicide concentration up to 0.64 g L−1 , as can be seen in Figure 17.15b. This caused the progressive increase in efficiency calculated from Equation (17.32), with values as high as 337% in the early stages of the most concentrated solution, as expected from the quick photolysis of Fe(III) complexes under solar irradiation. This was confirmed by determining the MCCP decay in the different EAOPs. Figure 17.16a illustrates that the herbicide is not directly photolyzed by sunlight, while it decays at a similar rate by EF, PEF, and SPEF, but much more slowly by AO-H2 O2

430

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

(data not shown). This demonstrates that the main oxidant of MCPP is homogeneous • OH formed from Fenton’s Reaction (17.11), without significant participation of photolytic Reaction (17.30). The production of this oxidant remains constant under given operation conditions and when the herbicide concentration increases, its attack is gradually concentrated on by-products strongly accelerating TOC removal (see Figure 17.15b) but causing a deceleration of the MCPP decay that always follows a pseudo–first-order reaction (see Figure 17.16b and inset). Figures 17.17a through -17.17c present the time-course for several aromatics such as 4-chloro-o-cresol, 2-methylhydroquinone, and 2-methyl-p-benzoquinone [29]. As can be seen, all these intermediates are rapidly accumulated and removed in less than 30 min, following the same evolution for EF, PEF, and SPEF because homogeneous • OH is their main oxidizing species. In contrast, Fe(III)-pyruvate complexes are removed in 150 min (see Figure 17.17d), while Fe(III)-oxalate complexes are quickly photodecarboxylated under UVA light and more rapidly with sunlight (see Figure 17.17e), but Fe(III)-acetate complexes are only removed at a slightly higher rate under light exposition (see Figure 17.17f). Based on these results, the reaction pathways shown in Figures 17.18a and 17.18b are proposed for the evolution of aromatics and carboxylic acids, respectively. The former path involves the hydroxylation of MCPP to give 4-chloro-o-cresol 6

8

(a)

4-chloro-o-cresol

5

(d)

Pyruvic acid

(e)

Oxalic acid

(f)

Acetic acid

6 4 4

3 2

2 1 0

concentration (mg L–1)

0 (b)

2-methylhydroquinone

50

2

40 30

1

20 10

0

0 (c)

2-methyl-p-benzoquinone

5

25

4

20

3

15

2

10

1

5

0

0

5

10

15

20

25

30

0

0

120

240

360

480

600

Time (min)

Figure 17.17 Evolution of the concentration of intermediates detected during the degradation of mecoprop under the same conditions of Figure 17.16a. (Reprinted from Ref. 29.)

17.6 DEGRADATION OF ORGANICS IN BATCH RECIRCULATION BDD/O2 FLOW CELLS

431

(a) CH3 O

CH3

Cl MCPP

•

CH3

OH

CH3 COOH Lactic acid

CH3

•

OH −–Cl

Cl

COOH

O

OH

OH

COOH

4-chloro-ocresol

•

CH3

OH

OH

O

2-methylhydroquinone

2-methyl-pbenzoquinone

(b) CH3 COOH COOH Lactic acid

CH3

•OH

BDD(•

OH)

CO

•OH •

BDD( OH)

COOH Pyruvic acid

Fe3+

CH3–COOH

Fe(III)-acetate complexes

Acetic acid BDD(•OH)

COOH–COOH Oxalic acid BDD(•OH)

BDD(•OH) Fe3+

Fe(III)-oxalate complexes hν

–Fe2+

BDD(•OH)

CO2

Figure 17.18 Proposed reaction scheme for (a) aromatics and (b) carboxylic acids formed during mecoprop mineralization by EF, PEF, and SPEF. (Adapted from Ref. 29.)

with loss of lactic acid. This by-product is hydroxylated releasing a chloride ion to yield 2-methylhydroquinone, which is subsequently oxidized to 2-methyl-p-benzoquinone. The second path explains the progressive transformation of the initially generated lactic acid into pyruvic, acetic, and oxalic acids. This latter acid also comes from the oxidation of aromatics. The fact that TOC is very slowly removed at long electrolysis time of SPEF (see Figure 17.15a) can then be ascribed to the large persistence of Fe(III)-acetate complexes, which can only be slowly but progressively oxidized by BDD( • OH). A different behavior was found when 2.5 L of o-, m-, and p-cresol solutions were treated in the flow plant by EF and SPEF with BDD [24]. Whereas overall decontamination was quickly obtained by SPEF, about 50–60% mineralization was only reached at the same tTM value by EF because of the very slow oxidation of final Fe(III)-oxalate complexes with BDD( • OH) and/or • OH compared with their efficient photolysis by the incident solar irradiation. Table 17.3 collects the results obtained in the case of p-cresol at 2 h of SPEF treatment. A higher percent of TOC removal with increasing current density from 25 to 100 mA cm−2 , along with the concomitant fall in efficiency, can be observed, confirming a greater production of BDD( • OH) and • OH, but with a faster acceleration of their nonoxidizing reactions. Table 17.3 also reveals that the rise in p-cresol concentration up to 1 g L−1 enhances its mineralization, as expected if the waste reactions of the above oxidants are gradually less significant. A MCE value as high as about 400% is found for the most concentrated solution, indicating a very high efficiency of sunlight to photolyze Fe(III)-oxalate complexes. In addition, p-cresol solutions are more rapidly destroyed with 1.0 mM Fe2+ than with 0.25 mM Fe2+ since greater amounts of such

432

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

TABLE 17.3 Percentage of TOC removal, mineralization current efficiency after 2 h of treatment, time for total mineralization (tTM ), and energy cost per unit volume for total mineralization (ETM ), obtained for 2.5 L solutions of several initial p-cresol contents of pH 3.0 by SPEF using a batch recirculation flow BDD/O2 reactor with electrodes of 20-cm2 area at different current densities, 30◦ C, and liquid flow rate of 180 L h−1 [24]. [p-cresol]0 (mg L−1 ) 128

256 512 1024

[Fe2+ ]0 (mM)

Current density (mA cm−2 )

% TOC removal

MCE (%)

tTM (min)

0.25 1.0 0.25 1.0 0.25 1.0 1.0 1.0

25 25 50 50 100 50 50 50

42 56 75 87 89 55 36 34

119 158 104 118 60 148 195 373

300 240 240 180 180 330 450 —a

ETM (kWh m−3 ) 8.3 6.6 20 15 51 27 37 —a

a Not determined.

TABLE 17.4 Pseudo–first-order rate constant and percentage of TOC removal and energy consumptions (EC) per unit volume and per unit TOC mass after 5 h of electrolysis of 2.5 L solutions of the fluoroquinolone antibiotic enrofloxacin at pH 3.0 using the same flow system as in Table 17.3 with a Pt/O2 or BDD/O2 cell operating in batch mode at 50 mA cm−2 , 35◦ C, and liquid flow rate of 200 L−1 [35].

Method AO-H2 O2 with Pt AO-H2 O2 with BDD EF with Ptc EF with BDDc SPEF with Ptc SPEF with BDDc

k1 (s−1 )

% TOC removal

ECa (kWh m−3 )

ECb (kWh (g TOC)−1 )

6.33 × 10−5 9.77 × 10−5 1.12 × 10−3 1.16 × 10−3 1.23 × 10−3 1.31 × 10−3 8.64 × 10−4 3.68 × 10−4

8 28 29 45 69 86 89 93

17.00 20.80 17.80 21.05 15.60 21.20 21.10 21.00

2.125 0.743 0.613 0.467 0.226 0.246 0.118 0.045

C0 (mg L−1 ) 158

316 790

a Calculated from Equation (17.37). b Calculated from Equation (17.38). c Experiments performed with 0.2 mM Fe2+ .

complexes are photodecomposed. Table 17.3 shows a strong drop in the value of ETM as less current density is applied due to the drop in cell voltage and the gradual increase in MCE. The value of this parameter decreases with increasing Fe2+ concentration owing to the faster photolysis of Fe(III)-oxalate complexes, but increases with greater initial pollutant concentration because longer tTM is needed for higher contents of p-cresol. The lowest ETM value of 6.6 kWh m−3 is found when treating a 128 mg L−1 p-cresol solution with 1.0 mM Fe2+ at 25 mA cm−2 . The above results evidence the high efficiency and low energy cost required for the degradation of aromatics by SPEF with BDD, making this technique viable for industrial application. More recently, this behavior has been confirmed again by studying the degradation in the flow plant of a complex molecule like the fluoroquinolone enrofloxacin [35]. Relevant results obtained comparatively for 2.5 L solutions after 5 h of electrolysis by different EAOPs using BDD/O2 and Pt/O2 reactors at 50 mA cm−2 and 35◦ C are summarized in Table 17.4. These results show the same trends as those of other aromatics,

REFERENCES

433

thus the k1 -value for enrofloxacin decay is quite similar by EF and SPEF, both with Pt and BDD, being at least two orders of magnitude higher than for both AO-H2 O2 methods, as expected if it is predominantly oxidized by homogeneous • OH produced from Fenton’s Reaction (17.11). The highest TOC removal of 93% is achieved for the most concentrated solution treated by SPEF with BDD, corresponding to the smallest energy cost of 0.045 kWh (g TOC)−1 . This means that this method could be used for relatively high organic concentrations promoting larger and more inexpensive decontamination.

17.7

CONCLUSIONS

Simple, ecological, and inexpensive treatments for the oxidation of organics in wastewaters have been proposed by Fenton’s electrochemistry. These methods involve H2 O2 electrogeneration from O2 reduction at gas diffusion, carbon-felt, and carbon sponge cathodes. This species reacts with catalytic Fe2+ added at the medium to produce the homogeneous strong oxidant • OH via Fenton’s Reaction (17.11). In undivided cells, a BDD anode is always more efficient than a Pt one since pollutants are more rapidly removed with BDD( • OH) than with Pt( • OH), formed at the corresponding anode surface from Reaction (17.6). In AO with electrogenerated H2 O2 , organics are effectively destroyed by BDD( • OH), while the EF treatment removes more rapidly the aromatic pollutants because they mainly react with • OH formed from Fenton’s reaction. However, complexes of Fe(III) with final carboxylic acids such as oxalic and oxamic acids can only be slowly oxidized with BDD(OH). The PEF method is more efficient due to the parallel and quicker photodecarboxylation of final Fe(III)-oxalate complexes by UVA irradiation. When oxamic acid is produced, its Fe(III) complexes cannot be photolyzed and the use of Cu2+ as co-catalyst enhances the mineralization process because Cu(II)-oxamate complexes are quickly removed with hydroxyl radicals. Solar photoelectro-Fenton can be a viable technique for the treatment of industrial wastewaters due to its higher efficiency and lower operational cost, as proved with a small solar batch BDD/O2 flow plant. All organics are removed much more rapidly in it as a result of the synergistic action of BDD( • OH), • OH, and sunlight. Total mineralization of organics wastewaters is easily reached if only oxalic acid is formed as persistent carboxylic acid, because its Fe(III) complexes are rapidly photolyzed by solar irradiation. However, the process becomes much more difficult when acetic and oxamic acids are generated because their Fe(III) complexes can only be slowly oxidized with hydroxyl radicals.

REFERENCES 1. E. Brillas, M.A. Oturan, in Pesticides: Impacts Environmentaux, Gestion et Traitements ´ (M.A. Oturan, J.M. Mouchel, eds.), Presses de l’Ecole Nationale des Ponts et Chauss´ees, Paris, France, 2007; pp. 61–73. 2. M.A. Oturan, E. Brillas, Port. Electrochim. Acta 2007, 25 , 1–17. 3. M. Panizza, E. Brillas, Ch. Comninellis, J. Environ. Eng. Manage. 2008, 18 , 139–153. 4. C.A. Mart´ınez-Huitle, E. Brillas, Appl. Catal. B-Environ. 2009, 87 , 105–145. 5. R. Andreozzi, V. Caprio, A. Insola, R. Marotta, Catal. Today 1999, 53 , 51–59.

434

FENTON-ELECTROCHEMICAL TREATMENT OF WASTEWATERS FOR THE OXIDATION

6. M. Tarr, ed., Chemical Degradation Methods for Wastes and Pollutants. Environmental and Industrial Applications, Marcel Dekker, Inc., New York, 2003. 7. M. Pera-Titus, V. Garc´ıa-Molina, M.A. Ba˜nos, J. Gim´enez, S. Esplugas, Appl. Catal. BEnviron. 2004, 47 , 219–256. 8. L. Plant, M. Jeff, Chem. Eng. 1994, 101 , EE16–EE20. 9. D. Pletcher, Acta Chem. Scand . 1999, 53 , 745–750. 10. P.C. Foller, R.T. Bombard, J. Appl. Electrochem. 1995, 25 , 613–627. 11. A. Alvarez Gallegos, Y. Vergara Garc´ıa, A. Zamudio, Sol. Energy Mater. Sol. Cells 2005, 88 , 157–167. 12. E. Brillas, R.M. Bastida, E. Llosa, J. Casado, J. Electrochem. Soc. 1995, 142 , 1733–1744. 13. M. Panizza, G. Cerisola, Electrochim. Acta 2005, 51 , 191–199. 14. M. Panizza, P.A. Michaud, G. Cerisola, Ch. Comninellis, J. Electroanal. Chem. 2001, 507 , 206–214. 15. C. Flox, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, E. Brillas, Chemosphere 2006, 64 , 892–902. 16. E. Weiss, K. Groenen-Serrano, A. Savall, J. Appl. Electrochem. 2008, 38 , 329–337. 17. E. Brillas, J.C. Calpe, J. Casado, Water Res. 2000, 34 , 2253–2262. 18. B. Boye, M.M. Dieng, E. Brillas, Environ. Sci. Technol. 2002, 36 , 3030–3035. 19. B. Boye, M.M. Dieng, E. Brillas, J. Electroanal. Chem. 2003, 557 , 135–146. 20. E. Brillas, B. Boye, M.M. Dieng, J. Electrochem. Soc. 2003, 150 , E583–E589. 21. E. Brillas, B. Boye, I. Sir´es, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, P.L. Cabot, Ch. Comninellis, Electrochim. Acta 2004, 49 , 4487–4496. 22. S. Ammar, R. Abdelhedi, C. Flox, C. Arias, E. Brillas, Environ. Chem. Lett . 2006, 4 , 229–233. 23. C. Flox, S. Ammar, C. Arias, E. Brillas, A.V. Vargas-Zavala, R. Abdelhedi, Appl. Catal. B-Environ. 2006, 67 , 93–104. 24. C. Flox, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, E. Brillas, Appl. Catal. B-Environ. 2007, 75 , 17–28. 25. I. Sir´es, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, P.L. Cabot, E. Brillas, Appl. Catal. B-Environ. 2007, 72 , 373–381. 26. I. Sir´es, J.A. Garrido, R.M. Rodr´ıguez, E. Brillas, N. Oturan, M.A. Oturan, Appl. Catal. B-Environ. 2007, 72 , 382–394. 27. I. Sir´es, N. Oturan, M.A. Oturan, R.M. Rodr´ıguez, J.A. Garrido, E. Brillas, Electrochim. Acta 2007, 2 , 5493–5503. 28. E. Brillas, M.A. Ba˜nos, M. Skoumal, P.L. Cabot, J.A. Garrido, R.M. Rodr´ıguez, Chemosphere 2007, 68 , 199–209. 29. C. Flox, J.A. Garrido, R.M. Rodr´ıguez, P.L. Cabot, F. Centellas, C. Arias, E. Brillas, Catal. Today 2007, 129 , 29–36. 30. E. Guinea, C. Arias, P.L. Cabot, J.A. Garrido, R.M. Rodr´ıguez, F. Centellas, E. Brillas, Water Res. 2008, 42 , 499–511. 31. M. Skoumal, C. Arias, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, E. Brillas, Chemosphere 2008, 71 , 1718–1729. 32. D. Montanaro, E. Petrucci, C. Merli, J. Appl. Electrochem. 2008, 38 , 947–954. 33. M. Skoumal, R.M. Rodr´ıguez, P.L. Cabot, F. Centellas, J.A. Garrido, C. Arias, E. Brillas, Electrochim. Acta 2009, 54 , 2077–2085. 34. A. Serra, X. Dom`enech, C. Arias, E. Brillas, J. Peral, Appl. Catal. B-Environ. 2009, 89 , 12–21. 35. E. Guinea, J.A. Garrido, R.M. Rodr´ıguez, P.L. Cabot, C. Arias, F. Centellas, E. Brillas, Electrochim. Acta 2010, 55 , 2101–2115.

REFERENCES

36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.

435

M.A. Oturan, J. Appl. Electrochem. 2000, 30 , 475–482. E. Guivarch, N. Oturan, M.A. Oturan, Environ. Chem. Lett . 2003, 1 , 165–168. K. Hanna, S. Chiron, M.A. Oturan, Water Res. 2005, 39 , 2763–2773. S. Irmak, H.I. Yavuz, O. Erbatur, Appl. Catal. B-Environ. 2006, 63 , 243–248. M. Diagne, N. Oturan, M.A. Oturan, Chemosphere 2007, 66 , 841–848. G. Kaichouh, N. Oturan, M.A. Oturan, A. El Hourch, K. El Kacemi, Environ. Technol . 2008, 29 , 489–496. M.A. Oturan, E. Guivarch, N. Oturan, I. Sir´es, Appl. Catal. B- Environ. 2008, 82 , 244–254. S. Hammami, N. Bellakhal, N. Oturan, M.A. Oturan, M. Dachraoui, Chemosphere 2008, 73 , 678–684. B. Balci, N. Oturan, R. Cherrier, M.A. Oturan, Water Res. 2009, 43 , 1924–1934. A. Ozcan, Y. S¸ahin, A. Savas¸ Koparal, M.A. Oturan, Appl. Catal. B-Environ. 2009, 89 , 620–626. Y. Sun, J.J. Pignatello, Environ. Sci. Technol. 1993, 27 , 304–310. M.A. Oturan, N. Oturan, J.J. Aaron, Actual. Chimique 2004, 277 - 278 , 57–63. Y. Zuo, J. Hoign´e, Environ. Sci. Technol. 1992, 26 , 1014–1022. S. Ferro, A. De Battisti, I. Duo, Ch. Comninellis, W. Haenni, A. Perret, J. Electrochem. Soc. 2000, 147 , 2614–2619.

18 Electrochemical Energy Storage and Energy Conversion Systems with Diamond Films Juan M. Peralta-Hern´andez, Aracely Hern´andez-Ram´ırez, Jorge L. Guzm´an-Mar, Laura Hinojosa-Reyes, Giancarlo R. Salazar-Banda, and Carlos A. Mart´ınez-Huitle

18.1

INTRODUCTION

Using petroleum as a main source of energy during the twentieth century has caused a considerable amount of damages, among which are atmospheric pollution and global warming. At the same time, many geologists and economists expect that the discovery rate of new petroleum sources will not follow the market’s demand, causing a serious shortage of petroleum in the near future. These alarming perspectives have motivated scientists around the world to develop new clean and renewable energy sources. In that manner, electrochemists intensified their research in order to develop the fuel cell technology. This consists in directly converting chemical energy into electricity. The main advantage of this technology over traditional energy production is that the fuel cell energy efficiency is Carnot cycle independent. The theoretical efficiency of such an energy source is nearly 90%. Direct alcohol fuel cell (DAFC), fed with ethanol and air (or oxygen), is a very interesting option for this kind of energy production. Ethanol is a

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

437

438

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

renewable energy source of vegetal origin whose CO2 production is consumed by plants during the cycle. In recent years, remarkable research efforts have been carried out with the objective to develop catalysts-supported materials for energy storage and conversion with improvement of the kinetic of the anode and cathode reactions, enhancing the activity for oxygen reduction and methanol and ethanol oxidation, with the goal of avoiding mainly the CO poisoning and intermediates on the electrode surfaces [1]. In the case of anode materials, the use of Pt–Ru catalyst is considered the most promising approach. In recent years, conductive films of boron-doped diamond (BDD) have been used by many researchers as an outstanding electrode material for electrosynthesis [2], a conductive support in electrocatalysis and mainly in environmental applications [3–10]. BDD exhibits attractive properties such as wide potential window, low background current, and a high chemical and dimensional stability, making it feasible for many electrochemical processes. Recently, Shao et al. [11] discussed in a critical review that BDD materials can be used as a potential support in the polymer electrolyte membrane (PEM) fuel cells. In this context, the major advantage of conductive BDD is the mechanical and chemical stability that it offers to modify this substrate. The aim of this chapter is to summarize the basic surface modification of BDD materials, with emphasis in different techniques to improve the catalytic efficiency of supported catalysts for PEM fuel cells (methanol and ethanol oxidation) using BDD materials.

18.2

DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

The search of new catalysts formulations for fuel cell applications has been conducted for a few decades. The study has been extended to hybrid systems containing BDD. The deposition of metal or metal oxide clusters onto the BDD film electrodes as nanoparticles are used to exploit the much higher catalytic activity of such nanoparticles using only very small amounts compared to the conventional bulk material [12]. Many deposition techniques have been tested in an effort to improve particle adherence and dispersion. A wide range of methods for nanoparticles synthesis has been explored. These methods are well known to be efficient ways to prepare particles and nanoparticles; however, the shape and size distribution of the obtained particles are strongly dependent on the synthesis technique. The choice of the material being deposited depends on the intended application of the electrode (sensitivity and selectivity in electroanalysis, electrocatalytic activity toward redox reactions such as chlorine, hydrogen, or oxygen evolution), and its required stability under harsh working conditions. The deposition technique should be simple and yield good dispersion of the particles on the substrate surface. In this section, we present a general review of techniques used for modified BDD surfaces related to the most relevant applications of all these electrodes for fuel cells. Fundamentals of each technology are also briefly discussed to better understanding its advantages and limitations for the modification of BDD surfaces. 18.2.1

Microemulsion Synthesis

A microemulsion is defined as a thermodynamically stable isotropic dispersion of two immiscible liquids consisting of microdomains of one or both liquids stabilized by an interfacial film of surface active molecules. The microemulsion system is characterized

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

439

by transparency (optical isotropic), droplet size (from 6 to 80 mm) and stability (thermodynamic) [13,14]. The synthesis of inorganic nanoparticles is usually carried out in water-in-oil (w/o) microemulsions. The microemulsion method has been used as microreactors to produce nanoparticles with narrow size distribution, since the first work described by Boutonnet et al. [15]. Water-in-oil microemulsion is the coexistence of an excess water phase and the surfactant molecules that aggregate in the oil phase in the form of reverse micelle. The water core of these aggregates is surrounded by surfactant molecules that have the nonpolar part of their molecule toward the oil phase. In the water core of this aggregate, metal salts could be solubilized. These metals will be then transformed into inorganic precipitates by using an appropriate reducing or precipitating agent. There are two main ways of preparation in order to obtain nanoparticles from microemulsions (see Figure 18.1):

(a)

(b) (1)

(1)

A A

B

+

NaBH4 +

(2) (2) A

B

(3) (3)

(4)

3 nm

Figure 18.1 Modes of particle preparation from microemulsion technique: (a) direct addition of the reducing agent to the microemulsion and (b) mixing of two microemulsions. The steps involved in the procedure are in (a): 1. Mixing; 2. Nucleation; and 3. Growing of the part side of the inverse micelles; and in (b): 1. Mixing; 2. Collision of droplets, exchange of the reactants, and metal reduction; 3. Nucleation; and 4. Growing of the part side of the inverse micelles. ((a) Reprinted with permission from Ref. 16; (b) reprinted with permission from Ref. 19.)

440

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

1. By mixing two microemulsions, one containing the precursor and the other the precipitating agent. 2. By adding the precipitating agent directly into the microemulsion containing the metal precursor. The former way of preparation is the most common technique used to prepare precipitates of nanoparticles employing microemulsions. It is also more favorable for achieving a homogeneous distribution of the precipitating/reducing agent since it rapidly reacts with the metal precursors through the collision and coalescence between two aggregates containing the different reactants. The microemulsion droplets in these systems provide an environment for controlled nucleation and growth. Particle growth is limited by the rate of intermicellar exchange and the steric stabilization provided by the surfactant prevents aggregation. The final size and shape of the nanoparticles can be controlled by varying the water-to-surfactant molar ratio or by varying the micro emulsion itself. Monodispersity of particles and stabilization of particles are very important criteria in controlled synthesis. The stabilizer (emulsifier) provides sites for the particle nucleation and stabilization of growing particles. They not only act as microreactors for processing reactions but they also exhibit the process aggregation of particles because the surfactants could adsorb on the particle surface when the particle size approaches to that of the water pool. As a result, the particles obtained in such a medium are generally very fine and monodisperse. Since the development of the microemulsion technique [15], a few publications have been presented in which the technique has been conducted for the synthesis of metallic nanoparticles where the catalyst has been supported in BDDs (see Table 18.1). BDD has been used as substrate of Pt [16], Pt–Ru [17], Pt–Sn [18], and Pt–Ru–Sn [19]. The choice of Pt and Pt–based particles was motivated by its useful potential application in alcohol (methanol or ethanol) electro-oxidation. Sin´e and Comninellis [16] obtained platinum nanoparticles by reduction of chloroplatinic acid (H2 PtCl6 ) with hydrazine at room temperature in a water-in-oil microemulsion of tetraethylene glycol monododecyl ether (BRIJ-30)/n-heptane using a two-microemulsion-step method. After the microemulsion step, transmission electron

TABLE 18.1 Survey of recent publications on metal/metaloxide particles from microemulsions deposited onto BDD thin-film electrodes with applications in fuel cells.

Catalyst Pt Pt/Sn Pt/Ru

Pt/Ru/Sn

Microemulsion system

Metal precursor

Reducing/ precipitating agent

Particle diameter (nm)

Brij-30a/nheptane Brij-301 /nheptane Brij-301 /nheptane

H2 PtCl6

NH2 NH2

2–3

H2 PtCl6 , SnCl2 H2 PtCl6 , RuCl3 ,

NaBH4

2–5

NaBH4

2–5

Brij-301 /nheptane

H2 PtCl6 , RuCl3 , SnCl2

NaBH4

2–5

a Brij-30: Polyoxiethylene (4) lauryl ether (non ionic surfactant).

Catalytic reaction Methanol oxidation Ethanol oxidation Methanol and ethanol oxidation Methanol and ethanol oxidation

Reference [16] [17] [18]

[19]

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

441

microscopy (TEM) was used to characterize nanoparticle sizes. The catalyst displayed similar particle size 2–5 nm. TEM micrograph of platinum nanoparticles synthesized by the microemulsion method is presented in Figure 18.2. Platinum nanoparticles were deposited onto the BDD substrate putting of the suspension on the diamond substrate and the excess water was dried under nitrogen atmosphere. In order to avoid detachment of Pt nanoparticles on BDD, Nafion® films were used to mechanically stabilize the electrode in order to avoid the detachment of Pt nanoparticles from the BDD surfaces. The BDD–Pt–Nafion® electrode was prepared adding Nafion® solution. BDD–Pt–Nafion® was more active than BDD–Pt for methanol oxidation. Subsequently, Sin´e and coworkers prepared bimetallic binary Pt–Ru [17], Pt–Sn [18], and ternary Pt–Ru–Sn [19] nanoparticles supported on BDD substrates by mixing the microemulsion with solid sodium borohydride as a reducing agent. They also used TEM and x-ray photoelectron spectroscopy (XPS) techniques to characterize particle sizes and morphology, and to determine the effective particle compositions and the identification of oxidation states of metals in the different samples, respectively. Conversely, the morphology and microstructure of metal-BDD electrodes were characterized by XRD, and the specific electrochemical surface activity and stability were analyzed by cyclic voltammetry. Pt/Ru nanoparticles of different compositions were synthesized by mixing appropriate ratios of Pt and Ru precursors in the aqueous phase of the microemulsion. The size distributions obtained by them were similar for all the samples and the size domain of the particles was approximately 2–3 nm of diameter. In addition, cyclic voltammetry (CV) was used as a surface analysis tool to provide information on the surface state of

Figure 18.2 TEM micrograph of platinum nanoparticles synthesized by the microemulsion method. (Reprinted with permission from Ref. 16.)

442

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

BDD-supported nanoparticles. As showed in Figure 18.3, CV of Pt, Pt50 Ru50 , and Ru nanoparticles deposited on BDD in pure supporting electrolyte (1 M HClO4 ) achieved substantial decrements in the quality of the H adsorption–desorption characteristic and increase in the background current in the double-layer region when the Ru content in the particles was increased. This information could improve the synthesis of more efficient nanoparticles for the electrocatalysis of methanol oxidation. In fact, other syntheses were carried out by the same authors in order to understand the effect of particle size and morphology on the efficiency. Therefore, (a)

I (mA)

0.4

0.0

–0.4

–0.8 0.0

0.2

0.4

0.6

0.8

1.0

0.8

1.0

E (V) vs. SHE

(b)

I (mA)

0.4

0.0

–0.4

–0.8 0.0

0.2

0.4 0.6 E (V) vs. SHE

0.2

0.4

(c)

I (mA)

0.4

0.0

–0.4

–0.8 0.0

0.6

0.8

1.0

E (V) vs. SHE Figure 18.3 CV of microemulsion-synthesized Pt (a), Pt50 Ru50 , (b) and Ru nanoparticles, (c) deposited on BDD. Recorded in N2 -saturated 1 M HClO4 solution at 50 mV s−1 and 25◦ C. (Reprinted with permission from Ref. 17.)

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

443

bimetallic Pt/Sn particles of several compositions with theoretical atomic contents Pt80 Sn20 , Pt60 Sn40 , Pt50 Sn50 , Pt40 Sn60 , and Pt20 Sn80 were synthesized via the microemulsion method [18]. TEM micrographs of Pt (A), Pt60 Sn40 (B), and Pt20 Sn80 nanoparticles (C) are presented in Figure 18.4. As can be seen, small isolated and well-spherical units of diameter in the 2–5 nm range were obtained by this method. In the case of ternary Pt–Ru–Sn nanoparticles, specific nominal compositions (80:10:10) were synthesized by Sin´e et al. [19]. The particle size measured by TEM was in the 2–5 nm range. Also, using XPS analyses of Pt80 Ru10 Sn10 nanoparticles produced by microemulsion technique, the relative atomic amounts of Pt, Ru, and Sn in the nanoparticles were 90%, 3%, and 7%, respectively. XPS Pt4f spectra for Pt/Ru, PtSn and Pt/Ru/Sn alloy nanoparticles are shown in Figure 18.5. As can be seen, Pt4f binding energies for the Pt/Ru and Pt/Sn alloy nanoparticles were lower than those for clean Pt nanoparticles. This phenomenon was more marked in the case of Pt80 Sn20 particles, in which the Pt4f contribution appeared at lower binding energy than in Pt80 Ru20 . The change in the electronic structure of the Pt component in the alloys (Pt/Sn and Pt/Ru) could modify the Pt work function and thus weaken bonding of adsorbed intermediates Pt–CO that could produce an enhancement in rates of methanol oxidation. 18.2.2

Thermal Deposition

The thermal decomposition of appropriate precursors that have been dissolved in suitable solvents and spread on a metallic support [20] has been applied to deposit iridium oxide [17,21], gold [17,22], and platinum nanoparticles [23] onto boron doped-diamond film. The nature of the precursor and the decomposition temperature must be controlled during the procedure because it affects the particle size, nonstoichiometry, and morphology of the oxide layer. The goal of the modification of doped-diamond with IrO2 , Au, or Pt nanoparticles was to produce electrodes with the well-known properties of iridium oxide, gold, or platinum electrodes by using only very low amounts of these precious metals. However, long-term stability is not sufficient at the current state of development. Duo et al. [21] studied the deposition of IrO2 particles onto BDD mild hydrophilic surface by the thermal decomposition technique. They observed that the oxygen evolution reaction (OER) was dramatically enhanced with the surface modification. IrO2 has a remarkable stability toward chemical and electrochemical attack, a metal-like conductivity, and appreciable catalytic (an electrocatalytic) activity. In this case, solutions of H2 IrCl6 in 2-propanol was applied to the BDD surface (1 cm2 ). After solvent evaporation at 80◦ C, calcination was performed at 350◦ C to oxidize the precursor acid to IrO2 . The calculated amount of IrO2 deposited was 2.5 μg cm−2 . Subsequently, Sin´e and associates [17] also deposited IrO2 onto BDD film electrode using as precursor H2 IrCl6 at different precursor concentrations, in order to vary the amount of deposited IrO2 . The calcination was performed at 450◦ C. At low IrO2 loading, isolated IrO2 particles had a size of about 2–3 nm and were concentrated at the grain boundaries of the diamond crystals. At higher IrO2 loading, the particles were larger (10 nm) and their concentration at the grain boundaries of the diamond crystals was significantly higher. The voltammetric curves obtained with BDD–IrO2 electrodes provided a fingerprint of electrode surface transitions occurring during the potential scan. Figure 18.6a shows the response of a BDD–IrO2 (G = 6.4) electrode, recorded by CV at 2 V s−1 (curve 2), in comparison with the background current of a bare BDD electrode (curve 1). The high capacitive current is related to changes in the oxidation state of the

444

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

(a)

(b)

(c)

Figure 18.4 TEM micrographs of (a) Pt, (b), Pt60 Sn40 , and (c) Pt20 Sn80 nanoparticles synthesized via the microemulsion method. (Reprinted with permission from Ref. 17.)

Intensity (a.u.)

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

445

Pt60Ru10Sn10 Pt60Sn20 Pt60Ru20

72.5

71.5

70.5

69.5

Binding energy (eV) Figure 18.5 XPS chemical shift of the Pt 4f7/2 line of Pt0 in Pt-rich ternary and binary nanoparticles supported on Au. The dashed vertical line represents the position of the signal in pure Pt nanoparticles. (Reprinted with permission from Ref. 19.)

IrO2 surface during the potential scan. The two pairs of peaks, seen at 0.40 and 0.95 V, can be related to the surface redox couples Ir(III)/Ir(II) and Ir(IV)/Ir(III), respectively. The very low currents recorded at a BDD electrode were certainly related to the absence of electroactive surface functionalities on the electrode surfaces. The capacitance was also calculated by them from the total anodic and cathodic currents, i , as a function of the scan rate, m, for each IrO2 loading. Figure 18.6b shows that the voltametric charge increased linearly with the IrO2 loading. Thus, the behavior of the composite electrode can be totally attributed to the electrochemical properties of the particles, even at low deposited IrO2 loading. Gold nanoparticles onto BDD electrode were deposited using a sputter deposition method of a thin gold layer with a sputtering time of 6 to 12 s followed by a heat treatment at 400–600◦ C [17,22]. Gold exhibits an extraordinary electrocatalytic activity higher than that of bulk gold and other transition metals toward many reactions like CO oxidation, catalytic hydrogenation of unsaturated alcohols and aldehydes, and O2 reduction [22]. Gold nanoparticles with an average size of 5–35 nm were prepared by this method on polycrystalline BDD film electrode, the particle size being dependent on the amount of deposited gold. Figure 18.7 shows typical SEM images of gold nanoparticles deposited on BDD (Au/BDD electrodes) using this technique. This figure shows clearly that the higher the deposited amount of gold (gold loading), the larger is the average size of the obtained Au nanoparticles. This figure also depicts that the gold nanoparticles grow essentially on specific surface sites such as grain boundaries of the polycrystalline BDD surface. Conversely, platinum particles were deposited on p-Si/BDD substrate by thermal decomposition procedure [23]. The reaction of methanol oxidation in acid medium was used as reaction test of the prepared p-Si/BDD/Pt electrode. This method consisted in the application of 5 μL of a platinum precursor solution (0.2–3 mM H2 PtCl6 in 2-propanol) on the diamond surface (1 cm2 ), evaporation of the solvent at 60◦ C during 5 min, and finally, thermal decomposition of the precursor by treatment in an oven at 350◦ C during 1 h. Figure 18.8 shows the SEM images of platinum deposited on p-Si/BDD by thermal decomposition with 15 nmol cm−2 of Pt. The agglomeration of these platinum particles was related to the inhomogeneity of the interfacial surface tension of the BDD support.

446

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

1.0

(a) 2

j (mAc m–2)

0.6 1

0.2

–0.2

–0.6

–1.0 0.0

0.5 E (V) vs. SHE

1.0

(b)

C (mF cm–2)

80

60

40

20

0 0

2

4

6

Γ Figure 18.6 (A) Cyclic voltammograms: (1) BDD mild electrode and (2) BDD–IrO2 electrode with G = 6.4 prepared at 450◦ C. Electrolyte: 0.5 M H2 SO4 . Scan rate 2000 mV s−1 , T = 25◦ C. G = 1 corresponds to 1015 molecules IrO2 cm−2 . (B) Capacitance values ( E = 0.64, scan rate 20 mV s−1 ) as a function of IrO2 loading on BDD electrodes. (Reprinted with permission from Ref. 18.)

However, after CV and SEM analysis of the electrode surfaces, the authors concluded that thermal decomposition procedure was not a suitable method to obtain a well dispersed and electrochemically stable catalyst. 18.2.3

Electrodeposition

The electrodeposition is one of the most widely used methods for the preparation and deposition of particles on BDD. As an electroanalytical tool, BDD has been used in the detection of numerous analytes, but it has also been successfully employed as an inert substrate for catalytically active metals and metal oxides [24,25]. The modification of a BDD electrode surface has been reported for a limited range of metal nanoparticles, including Ag, Au, Pt, Pd, Cu, Bi, Ni, Hg, Pb, Co, Ir, Ru, Te, Ti, and Fe [26–28] (see Table 18.2 for Pt and Ru, as an example).

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

447

(a)

(b)

Figure 18.7 SEM pictures of gold deposits on polycrystalline BDD substrate heated at 600◦ C as a function of the deposited amount of gold: (a) 9.6 × 1015 and (b) 1.9 × 1016 gold atom per cm2 geometrical surface area, obtained with sputtering times of 6 and 12 s, respectively. (Reprinted with permission from Ref. 18.)

One important feature of the BDD electrode is the nonuniform electroactivity across its surface. This is due to the boron-doped nature of the polycrystalline diamond, leading to areas of increased reactivity depending on the concentration of B atoms, which are present in an approximate ratio of 1 boron atom to 1000 carbon atoms [29]. Increased activity at the diamond grain boundaries is a feature suggested by the SEM images of polished polycrystalline high-quality BDD surfaces [30,31], and varied local reactivity has also been reported via confocal Raman imaging, photoluminescence [32], and AC impedance experiments [33,34]. As such, BDD is an excellent electrode material to promote the formation of nanoparticles as opposed to films. Nanoparticles are well known to be ideal for use in catalysis owing to their high surface area to volume ratio and often improved catalytic behavior due to their changed properties from the corresponding bulk material [35]. The advantages of electrodeposition include the fact that most compound semiconductor is obtained at or near room temperature, which is considered low temperature deposition. Electrodeposition also promotes controlled growth and it is generally

448

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

(a)

(b)

Figure 18.8 SEM images of p-Si/BDD/Pt electrode at two magnifications: (a) 2000 and (b) 20,000 prepared by the thermal decomposition technique, Pt loading: 15 nmol cm−2 . (Reprinted with permission from Ref. 23.)

TABLE 18.2 Electrodeposition of some metals onto BDD thin-film electrodes with applications in fuel cells. Metals Pt Pt–Ru

Electrodeposition Conditions CA: +1 to +0.02 V in 1 M of HClO4 CV: 50 mV s−1 between −0.2 and 1.0 V in methanol 0.1 M/H2 SO4 0.5 M

Metal precursor

Particle diameter (nm)

K2 PtCl6

40–700 nm

K2 PtCl6 and RuCl3

46±13 nm of Pt and 105±57 nm Pt–Ru.

Application Methanol electrooxidation. Electrocatalyst for methanol oxidation.

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

449

a low-cost methodology when compared to the dry methods. The deposition and codeposition of the different metals on diamond films have been the most studied systems due to their high interest in electrocatalysis. Deposits in BDD received great attention due to their applicability in fuel cell systems, being the methanol oxidation the preferred test reaction [36]. 18.2.3.1 Electrodeposition of Metal Particles on BDD As Pt is one of the more used metals in the electrodeposition technique, the electrodeposition of Pt particles on a BDD electrode is generally performed by applying a potential step to a deaerated 2 mM H2 PtCl6 solution in 1 M HClO4 . The potential is shifted from an equilibrium potential (1 V, where no reduction of platinic ions takes place) to a potential at which the reduction of Pt4+ to metallic Pt occurs (0.02–0.15 V). The electrodeposition mechanism is studied by means of chronoamperometry. Figure 18.9 shows a typical current versus time profile for a potentiostatic step experiment, with a potential step from +1 to +0.02 V. At the beginning of the first interval (zone I), a sharp increase, followed by a current decay, is observed, corresponding to the double-layer charging current and the initial nucleation process. Free growth of independent nuclei and formation of new nucleation sites without overlapping effects explain how the current increase of the second interval (zone II) are obtained. During the third interval (zone III), the growth of independent nuclei and their overlap can occur simultaneously; the current increases up to a maximum (jm ) at which nuclei overlapping occurs. Finally, during the fourth interval (zone IV), the current decreases due to the overlapping of the diffusion zones of different nuclei and to their coalescence. From the time tm corresponding to jm , the current decreases due to the decrease in the surface area of deposited platinum particles, and it can be considered that a change from hemispherical to linear mass transfer diffusion occurs because of the increase in particle size. In the domain of free growth of nuclei (zone II), the I 1/2 vst plot is linear. This indicates a 3D-nucleation process of hemispherical nucleus with diffusion control [37].

I II

III

IV

5.0

j (mA cm–2)

jm

2.5

0.0 0

1 tm

2

3

4

5

t (s) Figure 18.9 Chronoamperometric curve of the electrodeposition of platinum on a BDD electrode. Potential step from +1 to +0.02 V in 2 mM H2 PtCl6 + 1 M HClO4 solution, T = 25◦ C. The different deposition time intervals are marked (zones I–IV) as well as the maximum current density (jm ) and the corresponding maximum time (tm ). (Reprinted with permission from Ref. 18.)

450

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

Based on this behavior, Scharifker and Hills [38] have developed a model that allows distinguishing between instantaneous and progressive mechanism of nucleation. In the instantaneous nucleation mechanism, all nuclei are rapidly created during the first stages of the process and their number remains constant throughout growth. In the progressive nucleation mechanism, new nuclei are continuously formed during the whole deposition process because the nucleation rate process is low. Figure 18.10 compares the experimental I–t curve (curve a) with those relative to the two limiting mechanisms (curve b and c for progressive and instantaneous nucleation, respectively). The experimental curve of Figure 18.10 agrees better with that of the progressive nucleation mechanism. Similar results have also been obtained for the electrodeposition of platinum on graphite [39]. The most important techniques for characterization of electrodeposits on BDD are scanning electron microscopy – energy dispersive X-ray spectroscopy (SEM/EDS), xray diffraction (XRD), atomic force microscopy (AFM), energy dispersive x-ray (EDX), Raman spectroscopy, scanning electrochemical microscopy (SECM), x-ray photoelectron spectroscopy (XPS), and CV. Figure 18.11 shows SEM micrographs of a BDD–Pt electrode prepared by performing a potential step from 1 V to 0.02 V in a 2 mM H2 PtCl6 + 1 M HClO4 solution for 5 s. Spherical and isolated particles are observed with a quite large size variation that covers the 40–700 nm range. This is indicative of continuous formation of new nuclei during deposition and it is in agreement with the progressive mechanism of nucleation of Pt on BDD. A typical cyclic voltammogram for electrodeposited Pt particles on BDD is shown in Figure 18.12. This voltammogram exhibits the characteristic feature of Pt (i.e., two distinctive H adsorption–desorption peaks between 0.05 and 0.35 V, followed by a fine double-layer region corresponding to metallic Pt). Potential cycling at more anodic potentials, which induces Pt oxide formation (broad plateau), was not performed in order to avoid surface rearrangement and aggregation of particles. The electrochemical response of this BDD–Pt composite electrode can be attributed solely to the deposited 1.0

0.8

( j /j m)2

0.6 c 0.4 a

0.2

b

0.0 0

1

2

3

4

t /t m Figure 18.10 I–t response plotted with reduced variables; jm and tm as shown in Figure 18.1. (a) Experimental values, (b) theoretical data for progressive nucleation, and (c) theoretical data for instantaneous nucleation. (Reprinted with permission from Ref. 18.)

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

451

Figure 18.11 SEM pictures at two different magnification scales of a BDD–Pt electrode prepared by electrodeposition with a single potential step (60 s) from 1 to 0.02 V in a N2 -saturated 2 mM H2 PtCl6 + 1 M HClO4 solution. (Reprinted with permission from Ref. 23.)

Pt particles—even at very low Pt loadings, due to the chemical inertness and low background current of the diamond substrate. This justifies the choice of BDD for the electrochemical study of supported catalytic nanoparticles. Although electrodeposited Pt particles on BDD are efficient for methanol electrooxidation, their size domain is so broad that they cannot be strictly classified as nanoparticles. The literature attributes this heterodispersity to the inhomogeneous nature of the BDD substrates [40], mainly to the presence of nondiamond sp2 impurities that act as preferential deposition sites. Therefore, a “size-effect” cannot be reasonably expected in this case, and some alternative synthesis techniques have to be employed to deposit real Pt nanoparticles on BDD. The Pt–Ru binary metallic catalyst is most commonly accepted as the best electrocatalyst for methanol oxidation. The fundamental mechanism studies for Pt–Ru catalysts

452

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

0.04

I (mA)

0.02 0.00 –0.02 –0.04 –0.06 0.0

0.2

0.4

0.6

0.8

1.0

E (V) vs. SHE Figure 18.12 Cyclic voltammograms of electrodeposited Pt particles on BDD electrode. Recorded in a N2 -saturated 1 M HClO4 solution at 50 mV s−1 and 25◦ C. Conditions of electrodeposition: single potential step (5 s) from 1 to 0.02 V in a N2 -saturated 2 mM H2 PtCl6 + 1 M HClO4 solution. (Reprinted with permission from Ref. 18.)

indicate that methanol is oxidized according to a bifunctional mechanism [41]. Surfacesited Pt atoms oxidatively dehydrogenate the chemisorbed methyl moiety in consecutive steps to yield a residual Pt–CO fragment that cannot be oxidized to CO2 at DMFC potentials, and Pt adsorbed CO is removed via an oxygen-transfer step from electrogenerated Ru–OH. Pt + CH3 OH → Pt − CO − 4H+ + 4e−

(18.1)

Ru + H2O → Ru − OH + H+ + e− Ru − OH + Pt − CO → Ru + Pt + CO2 + e

(18.2) −

(18.3)

Ru transfers oxygen more effectively than Pt due to its ability to oxidatively absorb water at less positive potentials [42]. In this case, the metal depositions were carried out with solutions of potassium hexachloroplatinate (IV) and ruthenium (III) chloride. The platinum and ruthenium were electrochemically deposited sequentially by means of CV, with a potential sweep between −0.2 and 1.0 V at 50 mV s−1 in 2 mM K2 PtCl6 /0.5 M H2 SO4 and 2 mM RuCl3 /0.5 M H2 SO4 solutions. After that, the particles were analyzed by SEM technique (see Figure 18.13). The average particle size was approximately 105±57 nm for Pt–Ru with coverage of 1.4 × 109 cm−2 . Here again, the deposition conditions require further optimization in order to reach practical dimensions. There has been a recent report of a pulsed galvanostatic electrodeposition technique being used for practical Pt–Ru electrocatalysts on high area carbon, in which a range of particle sizes from 20 to 50 nm was achieved [43]. The beneficial effect of Ru addition to Pt can clearly be observed in Figure 18.14, which shows the linear sweep voltammograms and the onset of methanol oxidation on Pt and Pt–Ru. The onset of methanol oxidation is significantly shifted to lower potentials on the bimetallic surface. For instance, in the case of methanol oxidation, a potential

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

(a)

453

(b)

Figure 18.13 SEM images obtained at two different magnifications (18,000 × and 10,000 × ) of (a) Pt and (b) Pt–Ru particles sequentially deposited on boron-doped diamond films by CV, doing a potential sweep between −0.2 and 1.0 V in 2 mM K2 PtCl6 /0.5 M H2 SO4 and 2 mM RuCl3 /0.5 M H2 SO4 solution, respectively. (Reprinted with permission from Ref. 57.)

j(A mol–1met)

400

200

0 0.35

0.40

0.45 E (V) vs. SHE

0.50

0.55

Figure 18.14 Linear sweep voltammograms for methanol electro-oxidation on Pt (thin line) and Pt–Ru (thick line) deposited on BDD. Recorded in N2 -saturated 1 M HClO4 + 0.1 M CH3 OH solution at 1 mV s−1 and 25◦ C. (Reprinted with permission from Ref. 18.)

of 0.51 V is needed to reach a specific molar current of 50 Amol−1 at Pt, whereas a value of 0.43 V is sufficient at Pt–Ru. This indicates that continuous alcohol adsorption and dehydrogenation occurs on Pt surface sites that are available. It is concluded that on the bimetallic surface, the surface reaction between adsorbed CO and oxygen-containing species occurs at lower potentials, and that the bifunctional mechanism is valid at Pt–Ru. 18.2.4

Sol-Gel Modification

Surface modifications of BDD electrodes have been carried out with several metal oxides and some mixed composites using the sol-gel method [44,45]. It is well known that the sol-gel technique is a suitable process for coating substrates for energy storage materials

454

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

and electrochemical devices; in either case, there are many interfaces between components and many components that have to perform reliably and safely. For instance, there are interfaces between electrodes and current collectors, between electrodes and electrolytes, and between electrodes and interconnects. In some cases, the interfaces are the location of failure in an operating fuel cell, and problems due to chemical reactions and increased contact resistance can occur. Also, elevated temperatures lead to microstructure changes, crystallization, thermal expansion mismatch, and delamination. In these complicated materials systems, the use of sol-gel processing is well suited to the need for accurate placement of critical materials [46]. The sol-gel method (see Figure 18.15) [47] starts with a solution consisting of metal compounds, such as metal alkoxides, and acetylacetonates as sources of oxides, water as the hydrolysis agent, alcohol as the solvent and acid or base catalyst. Metal compounds undergo hydrolysis and polycondensation at room temperature, giving rise to sol, in which polymers or fine particles are dispersed. Further reaction connects the particles, solidifying the sol into wet gel, which still contains water and solvents. Vaporization of water and solvents produces a dry gel (xerogel); an aerogel results from a supercritical drying process. Heating gels to several hundred degrees produces dense oxides as products. Coating films can be made by dip coating or spin coating of the sol. Unsupported films can be made by synthesizing the film at the interface between alkoxide solution and water. Membranes are prepared by pouring the sol on the porous oxide with coarse pores. Particles with sharp size distribution can be precipitated and grown in the sol [48]. The advantages of sol-gel technique in the preparation of ceramics include better homogeneity, lower temperature processing, and more uniform phase distribution in multicomponent systems, easy preparation of thin films and coatings, better size and

Sc . Drying Water catalyst Precursors solution

Spin or dip coating

Sol

Gel

Aerogel

Drying Drawing Xerogel Xerogel film

Thermal treatment

Thermal treatment Fibers

Ceramics, glass

Dense film

Figure 18.15 Sol-gel method showing the different products that can be prepared. (adapted from reference 47.)

455

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

morphological control in powder synthesis, and opportunities for the preparation of new crystalline and noncrystalline solids [49]. The main factors that are important in development of thin films are the uniformity and thickness of film, its adhesion to the substrate, and its resistance to cracking. In this context, some research groups are working on the BDD surface modification by the sol-gel technique for different catalytic coatings such as metallic oxides (MO2 , M = Pb, Ru, and Ir) [44] or metal catalyst as platinum. Suffredini et al. [44] reported for the first time the preparation of Pt–RuO2 deposits on a carbon black substrate using the sol-gel method; their activity toward methanol electrooxidation was investigated and they found superior activity of the Pt–RuO2 /C anodes prepared by sol-gel than those of similar composition but prepared by alternative methods [50]. Lately, Salazar-Banda, Suffredini, and Avaca [51] carried out the deposition of platinum oxide particles PtOx on BDD electrodes by this method and testing several pre- and post-treatments of the surface for electrochemical experiments. They studied the electrochemical stability of the catalytic coatings indicating that the electrodes retained 91.6% of the coated material after 1000 voltammetric cycles conducted in the water decomposition. Their results demonstrate that the sol-gel method produces more stable PtOx deposits on BDD surfaces than other reported techniques [51]. In other cases [52], the BDD electrode was modified with Pt, Pt–RuO2 , and Pt–RuO2 –RhO2 by the sol-gel process aiming to carry out the oxidation of methanol and ethanol. Each catalyst was deposited as coating film on BDD electrodes using as precursors the corresponding metallic acetylacetonates. The sol-gel solutions were prepared with Pt(II), Ru(II), and Rh(III) acetylacetonates in a mixture of isopropyl alcohol and acetic acid (3:2, v/v) obtaining a 0.01 M as a final concentration of each solution. These solutions were applied onto the BDD surfaces after previous thermal pretreatment at 400◦ C for 30 min in air; after that, the solutions were evaporated at 80◦ C. This procedure was repeated 15 times and finally the electrodes were annealed at 400◦ C for 1 h in an argon atmosphere. The XRD patterns registered for each prepared electrode indicated that only pure Pt was deposited instead platinum oxide, whereas Ru and Rh were deposited as RuO2 and RhO2. Estimates from XRD diffractograms showed the mean crystallite size of the Pt, Pt–RuO2 , and Pt–RuO2 –RhO2 coatings indicated that the sol-gel method was an efficient technique to produce nanometric catalytic deposits with the desired composition (see Table 18.3). Additionally, EDX analysis showed that the catalyst particles were randomly and homogeneously deposited over the entire electrode with a Pt>Ru>Rh increasing content on the surface. The AFM topological images for the BDD surface and for Pt/BDD, Pt–RuO2 /BDD and Pt–RuO2 –RhO2 /BDD modified electrodes were recorded in a 50 μm × 50 μm area. Figure 18.16 illustrates the image of BDD topology showing the regular pyramidal structures without holes or cracks; however, the modified BDD TABLE 18.3

Electrodes:Composition and characterization.

Coating Pt Pt–RuO2 Pt–RuO2 –RhO2

Atomic ratio in solution

Atomic ratio on the surfacea (EDX)

Particle sizeb from XRD (nm)

100 50:50 50:25:25

100 54:46 56:23:21

6.5 4.4 4.3

a Mean value after four measurements in different places of the surface. b Mean crystallite size calculated using the Scherrer equation.

456

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

134 μm 50 μm 50 μm

329 nm 0 nm 50 μm

50 μm

25 μm

25 μm

25 μm

25 μm

0 μm 0 μm

0 μm 0 μm

Pt/BDD

BDD

340 nm 0 nm 50 μm 50 μm

50 μm 25 μm

25 μm

0 μm 0 μm

Pt-RuO2/BDD

312 nm 156 nm 0 nm 50 μm 25 μm

25 μm

0 μm 0 μm

Pt-RuO2-RhO2/BDD

Figure 18.16 AFM images for BDD, Pt/BDD, Pt–RuO2 /BDD, and Pt–RuO2 –RhO2 /BDD electrode surfaces. (Reprinted with permission from Ref. 52.)

electrodes exhibit as well small islands corresponding to the catalytic coatings (clear areas in the images). For all BDD-modified samples some agglomerations of the catalytic coating were observed, which indicate the presence of heterogeneous sites containing some clusters with small size (1–5 μm) [52]. And after that, the behavior of sol-gel Pt, Pt–RuO2 , and Pt–RuO2 –RhO2 on BDD electrodes was evaluated by means of CV when the modified electrodes were used on methanol and ethanol electro-oxidation reaction in acid media (explained in the next section). The modification of the BDD electrode with other Pt-metal oxide catalysts prepared by sol-gel has been investigated with the aim to improve its electrocatalytic response to be used as a fuel cell anode. In this context, BDD electrodes with IrO2 , PbO2 , SnO2 , Ta2 O5 , and some mixed composites prepared by sol-gel have been investigated by electrochemical techniques to establish their catalytic activity toward methanol and/or ethanol oxidation reactions [44,53–56]. Table 18.4 presented different coating catalysts synthesized by sol-gel technique on BDD support electrode, indicating the precursors used in the synthesis, the crystallite size obtained, and the electro-oxidation reaction that was studied. In all cases a Nafion® film was incorporated onto the modified BDD to improve the stability of the coating on the diamond surface. Table 18.4 summarizes the most important examples of coating catalyst on BDD support synthesized by sol-gel technique. The surface modification of BDD with IrO2 and PbO2 was studied by AFM technique and the results indicated the existence of sites with heterogeneous deposition; both catalysts showed good electrocatalytic activity; however, IrO2 /BDD electrode exhibits better performance for the oxygen evolution reaction (OER) respect to diamond (unmodified) and PbO2 /BDD electrodes, as illustrated by Suffredini et al. [44]. Platinum-ruthenium oxide carbon powder composite (Pt–RuO2 /C) deposited on BDD surface was studied on the oxidation of methanol and ethanol in H2 SO4 solutions. The composite was prepared by the sol-gel process and the fixed BDD electrode showed

18.2 DIFFERENT TECHNIQUES USED TO MODIFY BDD FILMS

457

TABLE 18.4 The most important examples of coating catalyst on BDD support synthesized by sol-gel technique. Fuel cell system

Catalyst deposited on BDD electrode

Characterization technique

Average size crystallite (nm)

XRD, EDX, AFM, SEM,

4.3–6.5

AFM

—

Methanol and Ethanol

Pt, Pt–RuO2 , Pt–RuO2 –RhO2

Ethanol

PtOx , PtOx –RuO2 , RuO2 , IrO2 , PbO2

Methanol and Ethanol

Pt–RuO2 /C

XRD, EDX

7.2

Methanol

Pt–RuOx

SEM, XRD

500

Ethanol

Pt–RuO2 /C Pt–PbOx /C Pt–IrO2 /C

XRD, EDX

5.0

XRD, EDX, SEM, AFM

4.6–9.1

Methanol and Ethanol

Pt–(RuO2 –IrO2 )/C Pt–(RuO2 –PbOx )/C Pt–(IrO2 –PbOx )/C Pt, Pt–SnO2 Pt–Ta2 O5

Precursors Acetylacetonates of Pt(II), Ru (III), and Rh; in a mixture of isopropyl alcohol in acetic acid. Acetylacetonates of Pt(II), Ru (III), Pb(II) and Ir(III); in a mixture of isopropyl alcohol in acetic acid. Acetylacetonates of Pt(II), Ru (III); in a mixture of isopropyl alcohol in acetic acid/carbon black powder (Vulcan ® XC72R). Acetylacetonates of Pt(II), and Ru (III); in a mixture of isopropyl alcohol in acetic acid. Acetylacetonates of Pt(II), Ru (III), Pb(II) and Ir(III); in a mixture of ethanol in acetic acid/carbon black powder (Vulcan ® XC72R).

Acetylacetonate of Pt(II), Sn(IV) bisacetylacetonatedibromide, Ta (V) ethoxide; in a mixture of isopropyl alcohol in acetic acid

enhanced performance respect to commercial Pt/C composite, under the same experimental conditions. From the XRD pattern of the Pt–RuO2 /C composite, an average size crystallite of 7.2 nm was estimated using the Scherrer equation, confirming the efficiency of the sol-gel technique in the deposition of nanostructured catalysts [45]. Alternatively, the composite catalysts Pt–PbOx /C, Pt–IrO2 /C, Pt–(RuO2 –IrO2 )/C, Pt–(RuO2 –PbOx )/C, and Pt–(IrO2 –PbOx )/C were also fixed on BDD substrate to be used in direct ethanol fuel cells (DEFC). The sol-gel method gives as a result the formation of the nanometric crystallite dimensions of the composites that can be responsible for the enhanced catalytic activity toward ethanol oxidation. The XRD analysis revealed that Pb was deposited as a mixture of PbO and PbO2 , and the EDX measurements indicated that

458

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

TABLE 18.5 EDX compositions and crystallite dimensions for different composites. Composite Pt–PbOx /C Pt–IrO2 /C Pt–(RuO2 –IrO2 )/C Pt–(RuO2 –PbOx )/C Pt–(IrO2 –PbOx )/C

Composition (EDX)

Particle sizea from XRD (nm)

56%Pt–44%Ir 40%Pt–60% 52%Pt–22%Ru–26%Ir 38%Pt–40%Pb–22%Ru 33%Pt–40%Pb–27% Ir

5.7 3.4 6.9 6.8 3.3

a Mean crystallite size calculated using the Scherrer equation.

Pb is preferentially deposited as compared with Pt [55]. Some examples are presented in Table 18.5 about the compositions and the crystallite dimensions for the different composites obtained by sol-gel. Hence, the lead oxide-based catalysts deposited by the sol-gel technique on carbon powder exhibited enhanced catalytic activity for the ethanol oxidation compared with Pt/C commercial powder. On the other hand, Pt–SnO2 and Pt–Ta2 O5 catalysts have been incorporated on BDD surface in order to study the methanol and ethanol electro-oxidation [56]. The characterization of the BDD-modified electrodes was accomplished by XRD, AFM, SEM, and EDX studies. XRD diffractograms for each sample are shown in Figure 18.17; in the Pt/BDD pattern (curve a) can be observed the corresponding reflections for polycrystalline Pt and two characteristic peaks of diamond substrate. The Pt–SnO2 and Pt–Ta2 O5 XRD patterns showed the main peaks corresponding to SnO2 and Ta2 O5 crystalline structure. In this study, the authors estimated the mean crystallite size for Pt, Pt–SnO2 , and Pt–Ta2 O5 coatings, achieving values of 4.6, 5.0, and 9.1 nm, respectively. Although all these catalysts showed high catalytic activities toward ethanol and methanol oxidation, Pt–SnO2 catalysts exhibited better performance than Pt–Ta2 O5 for the oxidation

SnO2 Pt Cdiam.

Intensity (a.u.)

(c)

Ta2O5

(b)

(a)

10

20

30

40

50 60 2θ (degress)

70

80

90

Figure 18.17 XRD diffractograms of (a) Pt/BDD, (b) Pt–SnO2 /BDD, and (c) Pt–Ta2 O5 /BDD electrodes. (Reprinted with permission from Ref. 56.)

18.3 APPLICATION OF MODIFIED BDD FILMS AS ELECTROCATALYTIC

459

of ethanol; whereas, Ta2 O5 decreases the poisoning effect caused by the adsorbed CO generated in the methanol oxidation [56]. 18.3 APPLICATION OF MODIFIED BDD FILMS AS ELECTROCATALYTIC SURFACES FOR FUEL CELLS Salazar-Banda et al. [52] proposed the use of sol-gel method to deposited different materials over BDD surface such as Pt, and alloys Pt–RuO2 and Pt–RuO2 –RhO2 with the objective to generate electrodes to carry out the oxidation of methanol and ethanol. According to the results obtained, voltammetric assays showed that the BDD-modified surface improve the electroactive area in approximately five times with respect to original BDD surface. An important result of this study is shown in Figure 18.18, where the voltammetric response for bare BDD and BDD after modification with Pt, Pt–RuO2 , and Pt–RuO2 –RhO2 is evaluated. The study shows evidence that the oxygen reduction (OER) and the hydrogen evolution (HER) reactions are shifted toward low potentials due to the catalytic effect of the deposited metals after modification. The curve in Figure 18.18c displays the typical electrochemical behavior of a polycrystalline Pt surface, such as hydrogen absorption/desorption and the oxide formation, evidencing that the Ptparticles have a adequate contact with the BDD surface, high purity and homogeneous distribution. 3

0.8

(a)

(b) 0.4

1

i (mA cm–2)

i (mA cm–2)

2

0 –1

–1.2

–3 –1

0 1 2 E (V) vs. HESS

3

0.0

1.6

0.3

0.6 0.9 1.2 E (V) vs. HESS

1.5

1.2 (c)

(d) 0.8 i (mA cm–2)

i (mA cm–2)

–0.4 –0.8

–2

1.2

0.0

0.8 0.4 0.0

0.4 0.0 –0.4

–0.4 –0.8 0.0

0.3

0.6

0.9

E (V) vs. HESS

1.2

1.5

–0.8

0.0

0.3

0.6

0.9

1.2

1.5

E (V) vs. HESS

Figure 18.18 Cyclic voltammetric studies in a 0.5 M of H2 SO4 aqueous medium for (a) BDD, (b) Pt/BDD, (c) Pt–RuO2 /BDD, and (d) Pt–RuO2 –RhO2 /BDD. (Reprinted with permission from Ref. 52.)

460

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

Cyclic voltammetric assays were carried out for methanol and ethanol oxidation at a scan rate of 0.005 V s−1 , in acidic medium (H2 SO4 ) adding 0.5 M alcohol concentration; their results are shown in Figure 18.19. These studies revealed that the CO poisoning effect for both alcohols oxidation reaction was mainly inhibited on the ternary alloy Pt–RuO2 –RhO2 /BDD electrode (solid lines in Figure 18.19) due to the Rh presence, which promotes a better catalytic effect for these reactions by either prompting the oxidation of the adsorbed intermediate species to CO2 or diminishing the absorption of CO and the others intermediates over Pt surface. Another important result was presented by Suffredini et al. [45], who reported the electro-oxidation of methanol and ethanol using a Pt–RuO2 /C composite prepared by 5

(a)

4

i (mA cm–2)

3

2

1

0 0.0

5

0.2

0.4

0.6 0.8 E (V) vs. HESS

1.0

1.2

0.2

0.4

0.6 0.8 E (V) vs. HESS

1.0

1.2

(b)

i (mA cm–2)

4 3 2 1 0 0.0

Figure 18.19 Cyclic voltammetric study (second cycle) for electrochemistry oxidation of 0.5 M of (a) methanol and (b) ethanol. Dotted line corresponds to Pt/BDD, dashed line to Pt–RuO2 /BDD, and solid line to Pt–RuO2 –RhO2 /BDD electrodes materials (v = 0.005 V s−1 ). (Reprinted with permission from Ref. 52.)

18.3 APPLICATION OF MODIFIED BDD FILMS AS ELECTROCATALYTIC

461

sol-gel method supported on BDD, which were evaluated by CV. Catalytic properties of Pt–RuO2 /C supported over BDD were also evaluated by means of cyclic voltammetric technique. In this context, electrochemical assays were conducted using a glassy carbon electrode as the substrate for the composite. The results are presented in Figure 18.20. In this figure is important to remark that the authors evaluated the capacity in the potential window 100–400 mV versus HESS using glassy carbon (GC) electrode (1.2610−5 C), and it was considerably larger than that calculated at BDD (2.69 × 10−6 C), and therefore was partially justified as a better performance when BDD surface are used. The potential region of methanol oxidation (inset in Figure 18.20) for forward and reverse scans as well as for the peak showed that the Pt–RuO2 /C composite on the BDD substrate presented a higher current density than the composite supported on GC electrode. In this frame, the authors emphasized an interesting difference between the two voltammograms of Figure 18.20, indicating that the forward and backward lines of BDD substrate were almost coincident, whereas a large difference was observed for GC electrode. However, other tests using the same Pt–RuO2 /C material, indicating that the differences should be attributed to the substrate and that probably reflect the great capacitive effect of the GC. In light of this discussion, the most important contribution to the larger oxidation currents was that the use of BDD surfaces practically avoids the substrate contribution; thus, the response of electrode was only dependent on the catalyst. Results presented in Figures 18.21a and b correspond to methanol and ethanol oxidation responses, respectively, for Pt–RuO2 /C catalyst over BDD studied by CV at 10 mV s−1 . As well, the authors included in this figure the responses of a commercial 10% Pt/C catalyst on BDD

300 150 250

j /A (g Pt)–1

150

j/A (g Pt)–1

200

100 50 0 –50

–100 100

–150 0

50

100 200 300 400 500 E/mV vs. HESS

0 –50

GLASSY CARBON DIAMOND (BDD)

–100 0

200

400 E (mV) vs. HESS

600

800

Figure 18.20 Cyclic voltammetric response of the Pt–RuO2 /C composite fixed on BDD (full line) and on glassy carbon (dotted line) for the oxidation of 1.0 M methanol in 0.5 M H2 SO4 . The insert shows the region of capacitive responses for both electrode configurations, ν = 10 mV s−1 . (Reprinted with permission from Ref. 54.)

462

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

250

(a)

j /A (g Pt)–1

200 150 100 50 0 –50 0

200

400

600

800

1000

800

1000

E (mV) vs. HESS 120

(b)

j /A (g Pt)–1

80

40

0

–40 0

200

400 600 E (mV) vs. HESS

Figure 18.21 Voltammetric oxidation of 1.0 M of methanol (a) and ethanol (b) in 0.5 M H2 SO4 on Pt–RuO2 /C (full lines) and Pt/C (traced lines) composites fixed on BDD surfaces. Baselines (dotted lines) were included as a comparison. Scan rate 10 mV s−1 . (Reprinted with permission from Ref. 54.)

as a comparison. As observed from Figure 18.21a, the oxidation of methanol started at 380 mV versus HESS on both substrates and these results were agreement with the data reported by He et al. [42] where Pt–Ru nanoparticles were electrodeposited on carbon nanotubes. For the case of ethanol oxidation (Figure 18.21b), the electrochemical responses were extremely different for both cases, showing the presence of a reactivation process on the catalyst surface, but in the case of Pt–RuO2 /C material, the oxidation ethanol potential was much lower than for the Pt/C. However, the response in current density for Pt–RuO2 /C material was fairly large for an extended potential window, indicating a multistep process during methanol oxidation. Polycrystalline BDD films was proposed by Gonz´alez-Gonz´alez, Tryk, and Cabrera [57] as the alternative to obtain high-area carbon supports using electrodepositation, with potential application for direct methanol fuel cell electrocatalysts. The electrocatalytical behavior of Pt/BDD, Pt–Ru/BDD, and BDD electrodes toward the oxidation of methanol in acid media was evaluated by CV in Figure 18.22. Thus, the maximum currents densities

18.3 APPLICATION OF MODIFIED BDD FILMS AS ELECTROCATALYTIC

0.00100

Current density (Ac m–2)

0.00080

463

A B C

0.00060

0.00040

0.00020

0.00000

–0.00020 –0.40 –0.20

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

Potential vs. Ag/AgCl, V Figure 18.22 Study of CV of diamond films with electrocatalyst deposition on 0.1 MeOH/0.5 m H2 SO4 at 50 mV s−1 . (a) Pt/BDD, (b) Pt–Ru/BDD, and (c) BDD. (Reprinted with permission from Ref. 57.)

for methanol oxidation were obtained about 0.73 mA cm−2 for Pt and 0.94 mA cm−2 for Pt–Ru deposited on BDD. However, as indicated by the authors, the fact that Pt–Ru exhibited lower potentials than Pt may be expected in the basis of previous studies [40] and hence, more investigation is necessary to completely understand the composition and particle size effects. In 2007, Salazar-Banda, Eguiluz, and Avaca [53] reported another interesting and innovative study where they carried out the modification of BDD powder with metallic oxides (Pt–RuOx ) using the sol-gel technique to prepare high area and stable surface electrodes to the methanol oxidation. Its comparison with a commercial catalyst (Pt–Ru/C). Pt–RuOx /BDD powder electrode was electrochemically evaluated by means of CV, discovering that the incorporation of ruthenium presents the inhibition of the hydrogen adsorption/desorption signals. Additionally, good performances on currents were observed in the double-layer region due to an increase of the capacitive currents and to the ruthenium redox processes, in accordance with Figure 18.23. As can be seen in the Figure 18.23, methanol oxidation and onset potentials (i = 0.04 mA cm−2 ) displayed close values on both electrodes (∼0.40 V versus HESS). Furthermore, the magnitude of the current densities in the common fuel cell operation was approximately from 0.4 to 0.8 V versus HESS. As a consequence, the BDD powder modification presented an important enhancement of the catalytic activity to methanol oxidation with respect to other materials such as carbo-modified composites. At the same time, an analogous idea was proposed by Swope et al. [58], where they prepared conductive diamond powders as a new catalyst for fuel cells. The researchers reported the development of higher surface area, approximately 100 m2 g−1 , and good corrosion resistance by conductive diamond powders for application as the electrocatalyst support, using electrodeposition. For this investigation, they carried out the electrochemical measurements using a glassy carbon rotating disk electrode (GC RDE)

464

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

0.48 200

i (mA cm–2)

0.40

i (mA cm–2)

0.32

100 0 –100 –200

0.24

0.0

0.2

0.4

0.6

0.8

1.0

E (V) vs. HESS

0.16

0.08 Pt-RuOx/BDD powder Pt-Ru/C (E-Tek)

0.00

0.0

0.2

0.4

0.6

0.8

1.0

1.2

E (V) vs. HESS Figure 18.23 Cyclic voltammograms for the electrochemical oxidation of 0.5 M of methanol in 0.5 M H2 SO4 aqueous solution carry out on the Pt–RuOx /BDD powder/BDD (solid line) and on the Pt–Ru/C/BDD (dashed line) electrodes at ν = 10 mV s−1 . Insert corresponds to the cyclic voltammogram recorded on the Pt–RuOx /BDD powder/BDD electrode in 0.5 M H2 SO4 aqueous solution at ν = 50 mV s−1 . (Reprinted with permission from Ref. 53.)

Current density (A cm–2)

as the substrate. As illustrated in Figure 18.24, the larger background current for the 500 nm diamond powder electrode was due to higher specific area. However, no reduction and oxidation signal was observed between −500 and 700 V, suggesting that the electrode surface is largely free of sp2 carbon impurities. Moreover, the researchers

0.0008

500 nm Diamond Powder 8–12 μm Diamond Powder

0.0000

–0.0008 0.0 0.8 Potential (V) vs. Ag/AgCl Figure 18.24 Cyclic voltammogram for an 8–12μm diamond powder electrode, grown under microcrystalline conditions, and a 500 nm diamond powder electrode, grown under nanocrystalline conditions. Experiments were carried out in 1 M KCl at 100 mV s−1 . (Reprinted with permission from Ref. 58.)

18.3 APPLICATION OF MODIFIED BDD FILMS AS ELECTROCATALYTIC

465

affirmed that the featureless backgrounds for the voltammograms are evidence of good particle conductive. Recently, Salazar-Banda et al. [56] presented a similar study for the preparation BDD film surfaces modified with Pt, Pt–SnO2 , and Pt–Ta2 O5 nanocrystalline deposits, by means of the sol-gel method, to evaluate the methanol and ethanol oxidation. Analyzing Figure 18.25, the results for BDD before (thin solid line) and after modification with Pt (thick-solid line), Pt–SnO2 (dashed line), and Pt–Ta2 O5 (dotted line) showed the increase in the current for each material. However, the most important result for this study is the evidence that the presence of tantalum oxides produced an increase in defective surface sites, which enhances the interfacial capacitance and also raises the ability of charge accumulation—as already observed for the addition of tantalum oxide to the SnO2 –IrO2 system in the same study. Figure 18.26 shows the voltammograms carried out on BDD surface electrode without (solid line curve) and in the presence of 0.5 M of methanol and ethanol (dashed and dotted lines, respectively). In this figure it is possible to observe that methanol and ethanol are not electroactive in the potential region commonly used to evaluate the fuel cell systems (from 0.4 to 0.8 V versus HESS). Based on these results, is possible to observe that this electrode showed onset potential (i = 0.5 mA cm−1 ), respectively, to 1.49 and 1.54 V for the methanol and ethanol oxidation process, in accordance with the insert in the Figure 18.26. The possible explanation for this behavior maybe due to the low adsorption of species characteristics of diamond surfaces. In light of these results, the substrate, when modified, presents a small capacitive current, and the study of the alcohols oxidation processes is facilitated, because they do not compete with these reactions. Nevertheless, this report showed that the oxidation of ethanol starts at approximately 0.39, 0.35, and 0.61V on the same materials, clearly indicating the enhancement of the catalytic activity of the Pt coatings in the presence of Sn or Ta oxide, mainly for the Sn-containing coating. In addition, the cyclic

0.6 0.4

i (μA cm–2)

0.2 0.0 –0.2 –0.4

BDD Pt/BDD

–0.6

Pt-SnO2/BDD

–0.8 –1.0 0.0

Pt-Ta2O5/BDD 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

E (V) vs HESS Figure 18.25 Cyclic voltammograms recorded in 0.5 M H2 SO4 for the BDD (thin solid line), Pt/BDD (thick solid line), Pt–SnO2 /BDD (dashed line), and Pt–Ta2 O5 /BDD (dotted line) surfaces, ν = 50 mV s−1 . (Reprinted with permission from Ref. 56.)

466

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

20

i (mA cm–2)

12

i (mA cm–2)

16

2 back ground methanol ethanol

1

0

8

1.2

1.6

1.4

1.8

2.0

E (V) vs HESS 4

0 0.0

0.3

0.6

0.9

1.2

1.5

1.8

2.1

E (V) vs HESS Figure 18.26 Comparative voltammograms (first cycle) for the oxidation of methanol (dashed line) and ethanol (dotted line) in 0.5 M H2 SO4 and background responses in the supporting electrolyte (solid line), recorded used a nonmodified BDD electrode, ν = 5 mV s−1 . (Reprinted with permission from Ref. 56.)

voltammograms showed that the Pt–SnO2 /BDD electrode exhibited higher current densities at elevated potentials (from 0.7 to 0.85 V) and lower reactivation currents, indicating a faster kinetic for this reaction as compared to the other two electrodes. Or perhaps it could be interpreted as a more efficient complete oxidation that leads a decrement on the production of unwanted intermediates on the surface. Then, the successful deposition of the Pt–SnO2 and Pt–Ta2 O5 catalysts on BDD films was demonstrated. Since no loss or diminution of the catalytic activities of the electrodes was observed during the whole experiments, these deposits also showed a high stability on the diamond surfaces as already demonstrated for sol-gel-made composites deposited on BDD surfaces. In view of the fact that these catalysts showed high catalytic activities toward both the ethanol and methanol oxidation reactions, these authors suggested the Pt–SnO2 and Pt–Ta2 O5 deposition on high-area BDD material (powder or felt) to further test as anodes in fuel cells applications. However, emerging methods and other complementary techniques are necessary to provide measurements or preparation to include other metals on BDD surface. Future developments will rely on the close collaboration of analytical chemists, engineers, and electrochemists to ensure effective application and exploitation of new catalysts to increase the efficiency of fuel cells using BDD anodes.

18.4

APPLICATION OF BDD FILMS IN BATTERIES

Both primary and rechargeable batteries, which can generate clean electric energy from the stored chemical energy through the desired electrochemical reactions, are essential to the convenience and sustainability of human development in the modern mobile society [59].

18.4 APPLICATION OF BDD FILMS IN BATTERIES

467

Rechargeable batteries are currently prevailing portable power sources because they are material saving due to repeated charge and discharge. Today, environmental awareness and high energy density demand lead to the popularity of Li ion and Ni–MH batteries, which have gradually become an alternative power source to traditional lead-acid and Ni–Cd batteries [59]. Rechargeable Li ion battery systems have become a prominent technology in the global battery market, since they offer the highest energy density available to date for rechargeable batteries. Current research and development of Li ion batteries can hardly keep up with the growing demand of the ever-increasing 3C (computer, communication, and consumer electronics) market. New-generation wireless communication technologies require batteries with lighter weight, higher energy/power density, and longer cycle life. Commercial Li ion batteries generally utilize classic Li+ intercalation compounds (LiCoO2 ) and carbons as active materials, which fall in short with limited inherent capacity [60]. Although the lithium anode has superior theoretical capacity (3862 mAh g−1 ) and a high redox potential, there are several problems, such as dendrite and poor cyclability, to be resolved before it can have practical applications [61–63]. For more than two decades, numerous researchers have endeavored to find solutions to this problem by introducing different solvent mixtures [64,65], novel electrolyte salts [66], and additives to the electrolytes [67,68]. Carbonaceous anodes are the most utilized anodic material due to their low cost and availability. However, the theoretical capacity (372 mAh g−1 ) is poor compared with the charge density of lithium (3862 mAh g−1 ). Some efforts with novel graphite varieties and carbon nanotubes have tried to increase this reversible capacity. Reported measurements to date of the lithium ion capacity for single-walled carbon nanotubes (SWCNTs) are generally between 400 and 460 mAh g−1 [69–71]. However, there is a large first-cycle hysteresis that leads to high irreversible capacity loss for SWCNTs. This effect has been attributed to the high surface area of SWCNTs, which affects the extent of solvent decomposition leading to the solid-electrolyte-interface formation [70]. Large research efforts have also been made in the development of lithium metal alloys (LiM) that possess a very high specific capacity and are expected to replace the conventional graphite in advanced lithium-ion high-energy batteries. However, these materials suffer from morphological changes during the charge-discharge cycling, which in turn results in a very poor cycle life. As an alternative material, doped diamond electrode materials have wide potential window in aqueous, nonaqueous, and ionic liquids media, and they are chemically inert (see Chapter 4). Also of importance is that BDD is hydrogen-terminated, which in turn makes it a very stable surface and exhibits excellent electrochemical properties when compared with other carbon forms like glassy carbon and highly oriented pyrolytic graphite. The use of diamond materials in studies for further battery applications is rather recent. The direct insertion of lithium into as-prepared H-terminated BDD electrodes with different levels of boron doping (1018 –1021 B cm−3 ) and grown on cloth of graphite fibers (see Figure 18.27) was demonstrated in 2003 by Ferreira and coworkers [72]. The effect of boron concentration was evident. Electrodes with lower boron content displayed higher capacity for reversible lithium insertion, although they present a smaller electronic conductivity that increases the ohmic drop of the electrode. The electrode with 1021 part cm−3 reached a specific capacity, during the first insertions, of 95.7 mAh g−1 , whereas the sample with 1018 part cm3 reached 234.9 mAh g−1 (see Figure 18.28).

468

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

Figure 18.27 SEM image taken on a BDD/carbon composite, for a doping level of 1021 part cm−3 . (Reprinted with permission from Ref. 72.)

3.5

1018 part. cm–3 1021 part. cm–3

3.0

0.12 Voltage (V)

0.10

Voltage (V)

2.5 2.0

0.08 0.06 0.04 0.02

1.5

0.00 500 1000 1500 0 Specific capacity (mAh. g–1)

1.0 0.5 0.0 0

100

200

300

Specific capacity (mAh g–1) Figure 18.28 Voltage versus specific capacity of the BDD/carbon electrodes as a function of the boron-doping level. The discharge/charge current for both electrodes is 0.020 mA. The inset figure corresponds to the discharge of carbon cloth electrode. (Reprinted with permission from Ref. 72.)

18.4 APPLICATION OF BDD FILMS IN BATTERIES

469

Figure 18.29 SEM image of BDD/CF electrode with a doping level of 1018 part cm−3 . (Reprinted with permission from Ref. 73.)

A continuation of this research was reported in 2005 when the authors investigated the lithium electrochemical intercalation into BDD films grown on carbon felt (BDD/CF electrodes, see Figure 18.29) also with different boron-doping levels (1018 –1021 B cm−3 ) [73]. The grain sizes and conductivity of the BDD layers had great influence in the lithium intercalation process. In contrast to the first study (BDD grown on graphite fibers) [72], higher electronic conductivity (higher boron-doping level) increased the reversible electrode capacity of the BDD grown on carbon felt. Composite electrodes containing diamond layers with higher boron concentration (1021 part cm−3 , curve D in Figure 18.30) also have smaller grain sizes. Consequently, they are rich in grain boundaries or sp2 sites and display the highest reversible capacity for lithium storage. On the contrary, the low-doped diamond layer with a boron concentration of 1018 part cm−3 (curve B in Figure 18.30), which has large grain sizes and low electronic conductivity, was not efficient for lithium storage and intercalation. Nevertheless, the reason for these incongruous results from the two studies is unclear and was not explored in the study. According to the authors, this new class of electrodes can be very useful because they are free of the binder polymers traditionally used in the preparation of lithium batteries. Hypothetically, BDD composite electrodes can become very competitive if a boron–diamond layer provides an elevated sp2 /sp3 sites ratio. In this sense, nanodiamond layers, with a large quantity of grain boundaries, grown on felt substrates deserve further investigation. The study of the cycling performance of BDD powder prepared by the chemical vapor deposition method by assembling Li/BDD cells at ambient temperature was recently reported [74]. BDD powders with a doping level of 3500 ppm of B were prepared on single crystal p-type Si (100) wafers (see Figure 18.31). The as-grown BDD contained some graphitic (sp2 ) phase and was hydrogen terminated. Activation of BDD by anodic polarization (in 1 M H2 SO4 at 25◦ C for 30 min) was carried out to eliminate most of the sp2 -type carbon and absorbed hydrogen from the surface. According to Christy et al. [74], the diamond grains of the BDD layer have effective participation in the lithium storage, and the electron reaches the diamond through the

470

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

A

3.0

B

C

D

Voltage (V vs. Li/Li+)

2.5 2.0 1.5 1.0 0.5 0.0 0

200

400

600

800

Specific capacity (mAh

1000

1200

g–1)

Figure 18.30 First charge/discharge for carbon felt (a) and BDD/CF electrodes doped from 1018 up to 1021 part cm−3 (b, c, d). (Reprinted with permission from Ref. 73.)

sp2 carbons located in the grain boundaries. Both graphitic and nongraphitic carbons should provide sites for lithium insertion. Graphitic sp2 -type carbons accommodate Li between grapheme layers, whereas the sp3 -type carbons can accommodate only in defect sites caused by the presence of the trivalent boron, although it was reported that the lithium insertion in the interstitial sites of the sp3 -bonded carbon structure is energetically favorable and that the mobility of Li in the diamond lattice seems to be elevated [75]. It is worthwhile mentioning that the defect structure and, therefore, the sp2 character can be enhanced by incorporation of more boron in the carbons. Thus, the results of Christy and coworkers suggested that BDD anode materials could be very promising if BDD provides an elevated number of both sp2 carbon sites and also sp3 sites with good intercalation kinetics. This supposition is reinforced if one considers that a high fraction of sp2 carbon is preferred for high lithium storage capacity when as-deposited diamond-like carbon (DLC) films with different sp2 /sp3 ratios were characterized as anode materials for Li–ion batteries [76]. DLC is a metastable form of amorphous carbon containing sp2 -bonded clusters interconnected by a random network of sp3 -bonded atomic sites [77]. In a different approach, functional sp2 –sp3 carbon composite materials (carbon nanotube/nanohoneycomb BDD, CNT-NANO) were fabricated by introducing multiwalled carbon nanotubes into the pores of nanohoneycomb diamond of 400 nm diameter using the CVD method [78]. Highly BDD films were deposited by microwave-assisted plasma CVD. Nanohoneycomb structures were prepared by oxygen plasma etching through anodic alumina masks with 400 nm pore diameter on polished diamond films, as can be observed in Figure 18.32, and the carbon nanotubes were prepared by pyrolysis of phthalocyanine with a Fe catalyst using CVD. The electrochemical behavior of these electrodes was examined using CV, electrochemical impedance spectroscopy, and galvanostatic measurements in LiClO4 /propylene carbonate electrolyte. In contrast to

18.4 APPLICATION OF BDD FILMS IN BATTERIES

Figure 18.31

471

SEM images of BDD powder. (Reprinted with permission from Ref. 74.)

the previously mentioned studies [72–74], neither Li+ intercalation nor deintercalation was observed on the cyclic voltammograms for the as-deposited BDD (curve (a) in Figure 18.33). On the other hand, the behavior of Li+ insertion into CNTs was observed in the cathodic sweep at −3.3 V (vs. Ag/AgCl) in CV (curves (b through d) in Figure 18.33). The current density for Li+ intercalation at HD CNT-NANO was 2343 μA cm−2 (geometric area), and this at LD CNT-NANO was 2173 μA cm−2 (geometric) at −3.3 V (vs. Ag/Ag+ ). Alternating current (AC) impedance measurements have indicated that at the nanohoneycomb diamond densely deposited CNTs (HD CNT-NANO), only the Li+ intercalation process is observed. In contrast, the nanohoneycomb diamond modified with CNTs in low-density (LD CNT-NANO) exhibited the combination behavior of Li+ intercalation at CNTs and the electrochemical double-layer discharging on the diamond surface.

472

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

Figure 18.32 Top view of SEM images for a carbon nanotubes/nanoporous diamond composite electrode. (a) Low magnification and (b) high magnification images for (A) HD CNT-NANO and (B) LD CNT-NANO. (Reprinted with permission from Ref. 78.)

In galvanostatic measurements, HD CNT-NANO behaved as a pure Li+ ion battery anode, and the specific capacity (per 1 g of activated material) was found to be 894 mAh g−1 , which is higher than that obtained for mesophase carbon materials. For LD CNT-NANO, in the initial time following the start of discharging, the behavior of the double-layer discharging was observed in addition to Li+ deintercalation. Suppression of the potential drops associated with Li+ deintercalation by rapid discharging from the electrical double-layer could increase the specific power for LD CNT-NANO. The combination function of the super-capacitor and the Li+ -ion battery that works simultaneously supporting each other in one electrochemical cell suggests the possible realization of a hybrid electrode material with high-energy density and high-specific power. In summary, two different functionalities were simultaneously realized by combining two different materials with totally different electrochemical characteristics. In this case, the increase in the performance of the one functionality results in the trade-off of the other functionality. Therefore, in the case of the actual use of this hybrid electrode, the ratio of the combination of sp2 and sp3 carbon must be selected according to the requirement from the application.

18.4 APPLICATION OF BDD FILMS IN BATTERIES

473

0.2 (c)

Current density (mA cm–2)

0.1

(b)

0.0

–0.1

(d) (a)

–0.2

–0.3

–0.4

–3.0

–2.0

–1.0

0

Potential (V) vs. Ag / Ag+ Figure 18.33 Cyclic voltammograms for (a) as-deposited diamond (AD), (b) CNT-AD, (c) HD CNTNANO, and (d) LD CNT-NANO electrodes in 1M LiClO4 /propylene carbonate. Sweep rate: 0.3 mV s−1 ; geometric area: 0.03 cm2 . (Reprinted with permission from Ref. 78.)

BDD films were also used early as substrates for the deposition of Al thin films for the study of the underpotential deposition (UPD) of lithium as an indirect application of diamond for battery systems [79]. Thus, the electrochemical properties of clean aluminum in LiClO4 (poly(ethylene oxide)) solutions have been investigated in ultrahigh vacuum using as electrodes both foils and thin films vapor deposited on BDD layers supported on Si substrates. Voltammetric scans recorded at temperatures of about 55◦ C yielded a set of deposition/stripping peaks at potentials more positive to the onset of Li/Al alloy formation, attributed to Li UPD on Al. The amount of stored Li was found to increase with the thickness of the Al film; however, uncertainties in the real amount of Al did not allow more quantitative conclusions to be drawn. In view of the scarce quantity of reports available in the literature and due to the controversial results discussed in this section for the intercalation of lithium ion on BDD materials, it is clear that this issue is in the beginning of development. Several studies must be carried out for the application of BDD materials on rechargeable battery systems. Studies with emphasis on the quantification and the relationship within the properties of the diamond materials, such as sp2 -type carbon content, grain boundary size (micro, and nanodiamond), level of doping, diamond conductivity, surface termination, among others, on the lithium intercalation behavior are still needed. Future outlooks are apparently related to the use of boron-doped nanodiamond materials with high levels of doping, joining both a high sp2 /sp3 carbon ratio and a high quantity of atomic defect

474

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

sites. This direction would be a worthwhile area to pursue for research and technological applications for anode materials for battery systems.

18.5 APPLICATION OF BDD ELECTRODES AS ELECTROCHEMICAL CAPACITORS Electrochemical capacitors (ECs)—often called super-capacitors, electrical double-layer capacitors (EDLCs), pseudocapacitances, ultracapacitors, power capacitors, gold capacitors, or power caches—have attracted worldwide research interest because of their potential applications as energy storage devices in many fields [80,81]. However, due to the fact that there are, in general, additional contributions to the capacitance other than double-layer effects, we will call these capacitors as electrochemical capacitors (ECs) throughout this chapter. Like batteries, ECs charge by undergoing electrochemical reactions. However, they store the charge like a capacitor, with physical separation of charges, rather than as the chemical energy in a battery. The result is a unit that can charge and discharge much faster than a battery, although the energy density (as measured in kilowatt-hours per kilogram) is not as high. ECs can have potentially much longer life than batteries. Typically, they exhibit 20 to 200 times larger capacitance per unit volume or mass than conventional capacitors [82]. Therefore, a number of applications now use ECs or are strongly considering them for use, including electric vehicles, digital communication devices, digital cameras, mobile phones, electric hybrid vehicles, electric tools, pulse laser technique, uninterruptible power supplies for computers, and storage of the energy generated by solar cells [83,84]. One of the major features of such vehicles is their ability to give the necessary energy for the vehicle’s range during which the vehicle is stopped or in startup. The goal is to store that energy as efficiently as possible, so that it can be used in accelerating the vehicle at its next move [85]. New types of carbon materials, such as carbon nanotubes and nanofibers, have been studied as possible EC electrode materials. They have larger surface areas than conventional activated carbon and thus offer higher capacitance. Recent studies have suggested that carbon materials with nanopore structures can exhibit even higher capacitance, ostensibly because ions in confined geometries are stripped of their solvating molecules, which decreases their effective size [86]. Although these materials are attractive due to their high surface area and good matrix conductivity, it is desirable to have an electrode material with high capacitance and a wide working potential range in highly conductive aqueous electrolytes. Considering purely material properties, synthetic high-area BDD materials are very attractive candidates. In this context, Honda and coworkers demonstrated the suitability of BDD nanoporous honeycomb electrodes for the development of aqueous electrochemical capacitors [87]. Nanoporous honeycomb diamond films were fabricated from microwave plasma chemical vapor deposited diamond films by oxygen plasma etching through an alumina mask. XPS measurements indicated the presence of a large amount of oxygen on these films. These films exhibited a wide working potential range (about 2.5 V) in the aqueous electrolytes, just as in the case of unetched as-deposited diamond electrodes. The capacitance of the honeycomb diamond electrode was found to be 1.97 × 10−3 F cm−2 (geometric area), which was 200 times greater than the unetched counterpart (as-deposited surface). The capacitance values obtained from galvanostatic measurements, although somewhat higher,

18.5 APPLICATION OF BDD ELECTRODES AS ELECTROCHEMICAL CAPACITORS

475

were consistent with this value, thus indicating that honeycomb diamond electrodes are some promise for electrochemical capacitor applications. Afterward, the oxygen plasma-etched nanohoneycomb diamond-thin film electrodes were examined for electrochemical capacitor applications in nonaqueous electrolytes [88] since BDD films used in nonaqueous electrolytes exhibited 1.5–2.5 times wider potential windows (7.3 V) than those in aqueous electrolytes. For pore type 400 nm × 1.8 μm in nonaqueous electrolyte, the power and energy densities could reach only similar values as those in aqueous electrolytes. However, the impedance behavior observed in nonaqueous electrolytes was significantly different from that in aqueous electrolyte. Indications were that the AC signal cannot penetrate to the bottom of the honeycomb pores in the nonaqueous electrolytes due to their low conductivity, and that not all the surface may contribute to the double-layer capacitance. Therefore, the authors concluded that the combination of pore type 400 nm × 1.8 μm and aqueous electrolyte could be best for examined thus far. Later, nanoporous BDD films with various pore diameters (30–400 nm) and pore depths (50 nm to 3 μm) were fabricated by etching polished polycrystalline diamond films through porous alumina masks with oxygen plasma [89]. The capacitance values increased with increasing roughness factor, based on the pore dimensions. The honeycomb diamond electrode with pore dimensions 400 nm × 3 μm exhibited a 400-fold increase in the capacitance (3.91 × 10−3 μF cm−2 , geometric area) in comparison to the as-deposited surface, and this value was 80 and 500 times greater than that for GC and HOPG, respectively. For the porous film with 30 nm diameter pores, there was only a very small effect of the pore structure on the capacitance due to the high pore impedance. In 2007, Almeida et al. [90] developed nanocrystalline diamond (NCD) grown on carbon fibers (CF) substrate to be used as electric double-layer capacitor. A high specific capacitance (2.6 mF cm−2 ) and rectangular-shaped CV curves were obtained up to a high potential scan rate (100 mV s−1 ) in 0.5 mol L−1 H2 SO4 aqueous solution for NCD/CF-1300 (CF-1300 consists of felt disks with 0.15-cm thickness diameter). These results showed that the NCD/CF electrodes could be an excellent candidate for electrochemical double-layer capacitors by controlling deposition parameters and CF substrate microstructures. Later, these authors carried out the morphologic and electrochemical characterization of carbon fibers (CF) and their hybrid material formed by BDD films grown on CF [91]. These substrates were produced from a polyacrylonitrile precursor at two different heat treatment temperature (HTT) of 1000 (CF-1000) and 2000◦ C (CF-2000). The influence of CF carbonization temperature on diamond surface morphology was analyzed from SEM images. It was also explained taking into account the CF chemical surface contribution using XPS measurements. Figure 18.34 shows the composition of CF surfaces, treated at 1000 and 2000◦ C, associated with their respective SEM micrographs. The two major peaks observed at 200 and 600 eV corresponding to carbon (C1s) and oxygen (O1s) species. The oxygen content on the surface decreases from CF-1000 to CF-2000, as expected. Small amounts of silicon and nitrogen were also found for CF-1000 (A), which can come from the HTT process. SEM images obtained for the BDD/CF-1000 (A, A1, and A2) and BDD/CF-2000 electrodes (B, B1, and B2) exhibited in Figure 18.35 demonstrate that the CF substrates were completely covered by a polycrystalline diamond coating. These micrographs also showed that the grain size of BDD/CF-2000 electrode was larger than that for BDD/CF1000 electrode, supporting the strong influence of CF structural parameters on diamond growth. BET measurements (see Table 18.6) also showed that BDD/CF-2000 electrode

476

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

(a)

Intensity (cps)

(a)

C1s – 284 eV

O1s – 531.5eV

Si in SiO2 N 1s – Si 2s Si 2p 154ev 103ev 400eV

O (Auger) 978eV

Na (Auger) 497eV 1000

800

600

400

0

200

Binding energy (eV) 3µm

(b)

C1s – 284 eV

Intensity (cps)

(b)

O (Auger) 978eV

1000

800

O1s – 531.5eV

600

400

200

0

Binding energy (eV) 3µm

Figure 18.34 Scanning electron microscopy images and x-ray photoelectron spectroscopy spectra of (a) CF-1000 and (b) CF-2000. (Reprinted with permission from Ref. 91.)

(a)

7 µm (b)

7 µm

(a1)

4 µm (b1)

4 µm

(a2)

5 µm (b2)

5 µm

Figure 18.35 Scanning electron microscopy images of the BDD films growth on carbon fibers: BDD/CF-1000 (A, A1, and A2) and BDD/CF-2000 (B, B1, and B2). A, A1, B, and B1 images show the surface film morphology; images A2 and B2 show the film thickness around each fiber. (Reprinted with permission from Ref. 91.)

18.6 CONCLUSIONS

477

TABLE 18.6 BET area, capacitance, and parameters used for fitting the impedance results in both the Nyquist and Bode plots. Electrode CF-1000 CF-2000 BDD/CF-1000 BDD/CF-2000

BET (m2 g−1 )

C (μF cm−2 )

R1 ()

R2 (103 )

CPE1 (10−4 −1 sn )

n1

1.51 0.33 4.05 14.00

266 245 459 1940

142.6 16.26 305.4 38.6

0.474 1.338 4.41 0.820

0.06639 0.1717 0.1936 1.213

0.8427 0.8493 0.9004 0.9086

displays BET surface area almost 4 times higher than that for the BDD/CF-1000 composite and 40 times higher than that for CF-2000 itself, without BDD film. Cyclic voltammetric curves demonstrated that BDD films grown on CF carbonized at 2000◦ C presented the highest capacitance value when compared with that for BDD/CF1000 or those for CF electrodes, without diamond films. In addition, in electrochemical impedance experiments the BDD/CF-2000 electrode displayed almost an ideal capacitive behavior. The capacitance value for the BDD/CF-2000 was 1940 μ F cm−2 (geometric area) that was approximately nine times larger than that for CF-2000 (Table 18.6). The best BDD/CF-2000 electrode capacitive behavior was attributed to its increased surface area as a result of the singular diamond film morphology formed on such carbon fiber. Similar to batteries there are few reports available in the literature about the use of BDD materials as ECs. Thus, it is clear the necessity of studies with emphasis on the development of high-area BDD materials with both wide electrochemical windows and high capacitance. As a consequence, this is an open research area and an important challenge for future investigations.

18.6

CONCLUSIONS

The modification of BDD surfaces with micro- and nanometric metallic and/or metallic oxide deposits using different methods such as electrodeposition, sol-gel, thermal deposition, microemulsion methods, among others, was broadly investigated in the last two decades [11–56]. The deposition of metal or metal oxide clusters onto the BDD film electrodes have been used to exploit the much higher catalytic activity of such nanoparticles using only very small catalyst amounts compared to the conventional bulk material. The use of these hybrid systems containing BDD as new anode catalysts formulations for future fuel cell applications has been widely studied. However, further developments should be carried out upon the close collaboration of analytical chemists, engineers, and electrochemists to ensure effective application and exploitation of new catalysts to increase the efficiency of fuel cells using BDD anodes tested in real fuel cell operating conditions. On the other hand, several efforts should be carried out for the application of BDD materials on rechargeable battery or electrochemical capacitors systems, due to the limited amount of reports available in the literature, which demonstrate that these subjects are in the beginning of development. Studies emphasizing the quantification and the understanding of the relationship within the properties of the diamond materials—such as sp2 -type carbon content, grain boundary size (micro and nanodiamond), level of doping, diamond conductivity, surface termination, among others—on its electrochemical behavior are

478

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

still needed and might contribute to more significant advancements in the application of this electrode in electrochemical energy storage and energy conversion systems.

REFERENCES 1. E. Antolini, Appl. Cat. B: Environ. 2004, 74, 337–350. 2. U. Griesbach, D. Zollinger, H. P¨utter, C. Comninellis, J. Appl. Electrochem. 2005, 35 , 1265–1270. 3. M. Panizza, E. Brillas, C. Comninellis, J. Environ. Eng. Manage. 2008, 18 , 139–153. 4. P. Canizares, A. Gadri, J. Lobato, B. Nasr, R. Paz, M.A. Rodrigo, C. Saez, Ind. Eng. Chem. Res. 2006, 45 , 3468–3473. 5. J. Iniesta, P.A. Michaud, M. Panizza, G. Cerisola, A. Aldaz, C. Comninellis, Electrochim. Acta. 2001, 46 , 3573–3578. 6. M.A. Rodrigo, P.A. Michaud, I. Duo, M. Panizza, G. Cerisola, C. Comninellis, J. Electrochem. Soc. 2001, 148 , D60–D64. 7. B. Boye, P.A. Michaud, B. Marselli, M.M. Dieng, E. Brillas, C. Comninellis, New Diam. Front. C. Technol . 2002, 12 , 63–72. 8. I. Sir´es, N. Oturan, M.A. Oturan, R.M. Rodr´ıguez, J.A. Garrido, E. Brillas, Electrochim. Acta. 2007, 52 , 5493–5503. 9. C. Flox, J.A. Garrido, R.M. Rodr´ıguez, P.L. Cabot, F. Centellas, C. Arias, E. Brillas, Catal. Today. 2007, 129 , 29–36. 10. I. Sir´es, F. Centellas, J.A. Garrido, R.M. Rodr´ıguez, C. Arias, P.L. Cabot, E. Brillas, Appl. Catal. B-Environ. 2007, 72 , 373–381. 11. Y. Shao, J. Liu, Y. Wang, Y. Lin, J. Mater. Chem. 2009, 19 , 46–59. 12. A. Kraft, Int. J. Electrochem. Sci . 2007, 2 , 355–385. 13. M. Boutonnet, S. L¨ogdberg, E.E. Svensson, Curr. Opin. Colloid Interface Sci . 2008, 13 , 270–286. 14. S. Eriksson, U. Nyl´en, S. Rojas, M. Boutonnet, Appl. Catal. A: Gen. 2004, 265 , 207–219. 15. M. Boutonnet, J. Kizling, P. Stenius, G. Maire, Colloids Surf . 1982, 5 , 209–225. 16. G. Sin´e, C. Comninellis, Electrochim. Acta. 2005, 50 , 2249–2254. 17. G. Sin´e, G. F´oti, C. Comninellis, J. Electroanal. Chem. 2006, 595 , 115–124. 18. G. Sin´e, I. Duo, B. El Roustom, G. F´oti, C. Comninellis, J. Appl. Electrochem. 2006, 36 , 847–862. 19. G. Sin´e, D. Smida, M. Limat, G. F´oti, C. Comninellis, J. Electrochem. Soc. 2007, 154 , B170–B174. 20. F. Cardarelli, P. Taxil, A. Savall, C. Comninellis, G. Manoli, O. Leclerc, J. Appl. Electrochem. 1998, 28 , 245–250. 21. I. Duo, P.A. Michaud, W. Haenni, A. Perret, C. Comninellis, Electrochem. Solid-State Lett . 2000, 3 , 325–334. 22. B.E. Roustom, G. F´oti, C. Comninellis, Electrochem. Commun. 2005, 7 , 398–405. 23. F. Montilla, E. Morall´on, I. Duo, C. Comninellis, J.L. V´azquez, Electrochim. Acta 2003, 48 , 3891–3897. 24. A. De Battisti, S. Ferro, M. Dal Colle, J. Phys. Chem. B 2001, 105 , 1679–1689. 25. J.S. Gao, T. Arunagiri, J.J. Chen, P. Goodwill, O. Chyan, J. Perez, D. Golden, Chem. Mater. 2000, 12 , 3495–3504.

REFERENCES

479

26. M. Li, G. Zhao, R. Geng, H. Hu, Bioelectrochemistry 2008, 74 , 217–224. 27. V. Saez, J. Gonzalez-Garcia, M.A. Kulandainathan, F. Marken, Electrochem. Commun. 2007, 9 , 1127–1133. 28. K.E. Toghill, G.G. Wildgoose, A. Moshar, C. Mulcahy, R.G. Compton, Electroanalysis 2008, 20 , 1731–1739. 29. G.M. Swain, R. Ramesham, Anal. Chem. 1993, 65 , 345–352. 30. K.B. Holt, A.J. Bard, Y. Show, G.M. Swain, J. Phys. Chem. B 2004, 108 , 15117–15123. 31. C.H. Goeting, F. Jones, J.S. Foord, J.C. Eklund, F. Marken, R.G. Compton, P.R. Chalker, C. Johnston, J. Electroanal. Chem. 1998, 442 , 207–211. 32. M. Mermoux, B. Marcus, G.M. Swain, J.E. Butler, J. Phys. Chem. B 2002, 106 , 10816–10821. 33. D. Becker, K. J¨uttner, J. Appl. Electrochem. 2003, 33 , 959–968. 34. D. Becker, K. J¨uttner, Electrochim. Acta 2003, 49 , 29–38. 35. N.R. Stradiotto, K.E. Toghill, L. Xiao, A. Moshar, R.G. Compton, Electroanalysis 2009, 21 , 2627–2637. 36. J. Wang, G.M. Swain, J. Electrochem. Soc. 2003, 150 , E24–E32. 37. M. Paunovic, M. Schlesinger, Fundamentals of Electrochemical Deposition, John Wiley & Sons, New York, 1998. 38. B. Scharifker, G. Hills, Electrochim. Acta 1983, 28 , 879–886. 39. F. Gloaguen, J.M. L´eger, C. Lamy, A. Marmann, U. Stimming, R. Vogel, Electrochim. Acta 1999, 44 , 1805–1812. 40. O. Enea, B. Riedo, G. Dietler, Nano. Lett . 2002, 2 , 241–249. 41. H.A. Gasteiger, N. Markovic, P.N. Ross, E.J. Cairns, J. Phys. Chem. 1993, 97 , 12020–12029. 42. Z. He, J. Chen, D. Lui, H. Zhou, Y. Kuang, Diam. Rel. Mat . 2004, 13 , 1764–1770. 43. C. Coutanceau, A.F. Rakotondrainib´e, A. Lima, E. Garnier, S. Pronier, J.M. L´eger, C. Lamy, J. Appl. Electrochem. 2004, 34 , 61–68. 44. H.B. Suffredini, G.R. Salazar-Banda, S.T. Tanimoto, M.L. Calegaro, S.A.S. Machado, L.A. Avaca, J. Braz. Chem. Soc. 2006, 17 , 257–264. 45. H.B. Suffredini, V. Tricoli, N. Vatistas, L.A. Avaca, J. Power Sources, 2006, 158 , 124–128. 46. L.C. Klein, M. Aparicio, F. Damay, “Sol-Gel Processing for Battery and Fuel Cell Applications” in Applications of Sol-Gel Technology, Vol. III ( S. Sakka, ed.), Kluwer Academic Publisher, New York, 2004, p. 311. 47. C.J. Brinker and G.W. Scherer, Sol-Gel Science: The Physiscs and Chemistry of Sol-Gel Processing Academic Press, New York, N.Y. 1990. 48. S. Sakka, ed., Applications of Sol-Gel Technology, Kluwer Academic Publisher, New York, 2004, p. 3. 49. C.P. Ballard, A.J. Fanelli, “Sol-Gel Route for Materials Synthesis” in Chemistry of Advanced Materials ( C.N.R. Rao, ed.), Blackwell Scientific Publications, 1993. 50. H.B. Suffredini, V. Tricoli, L.A. Avaca, N. Vatistas, Electrochem. Commun. 2004, 6 , 1025–1028. 51. G.R. Salazar-Banda, H.B. Suffredini, L.A. Avaca, J. Braz. Chem. Soc. 2005, 16 , 903–906. 52. G.R. Salazar-Banda, H.B. Suffredini, M.L. Calegaro, S.T. Tanimoto, L.A. Avaca, J. Power Sources 2006, 162 , 9–20. 53. G.R. Salazar-Banda, K.I.B. Eguiluz, L.A. Avaca, Electrochem. Commun. 2007, 9 , 59–64. 54. H.B. Suffredini, V. Tricoli, N. Vatistas, L.A. Avaca, J. Power Sources 2006, 15 , 124–128. 55. H.B. Suffredini, G.R. Salazar-Banda, L.A. Avaca, J. Power Sources 2007, 171 , 355–362.

480

ELECTROCHEMICAL ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS

56. G.R. Salazar-Banda, H.B. Suffredini, L.A. Avaca, S.A.S. Machado, Mater. Chem. Phys. 2009, 117 , 434–442. 57. I. Gonz´alez-Gonz´alez, D.A. Tryk, C.R. Cabrera, Diam. Rel. Mat . 2006, 15 , 275–278. 58. V.M. Swope, I. Sasaki, A. Ay. G.M. Swain, ECS Trans. 2007, 3 , 27–36. 59. M. Winter, J.B. Brodd, Chem. Rev . 2004, 104 , 4245–4269. 60. M.S. Whittingham, Chem. Rev . 2004, 104 , 4271–4301. 61. J.J. Zhang, Y.Y. Xia, J. Electrochem. Soc. 2006, 153 , A1466–A1471. 62. J. Hassoun, S. Panero, B. Scrosati, J. Power Sources 2006, 160 , 1336–1341. 63. H. Jusef, S. Panero, P. Simon, P.L. Taberna, B. Scrosati, Adv. Mater. 2007, 38 , 1632–1635. 64. T. Hirai, I. Yoshimatsu, J.I. Yamaki, J. Electrochem. Soc. 1994, 141 , 611–614. 65. X. Wang, E. Yasukawa, S. Kasuya, Electrochim. Acta 2001, 46 , 813–819. 66. M. Ishikawa, S.I. Machino, M. Morita, J. Electroanal. Chem. 1999, 473 , 279–284. 67. K. Kanamura, H. Takezawa, S. Shiraishi, Z. Takehara, J. Electrochem. Soc. 1997, 144 , 1990–1996. 68. R. Mogi, M. Inaba, S.K. Jeong, Y. Iriyama, T. Abe, Z. Ogumi J. Electrochem. Soc. 2002, 149 , A1578–A1583. 69. A.S. Claye, J.E. Fischer, C.B. Huffman, A.G. Rinzler, R.E. Smalley, J. Electrochem. Soc. 2000, 147 , 2845–2852. 70. E. Frackowiak, F. Beguin, Carbon 2002, 40 , 1775–1787. 71. I. Mukhopadhyay, S. Kawasaki, F. Okino, A. Govindaraj, C.N.R. Rao, H. Touhara, Phys. B 2002, 323 , 130–132. 72. N.G. Ferreira, L.L. Mendonc¸a, V.J.T. Airoldi, J.M. Rosolen, Diam. Rel. Mat . 2003, 12 , 596–600. 73. E.C. Almeida, V.J. Trava-Airoldi, N.G. Ferreira, J.M. Rosolen, Diam. Rel. Mat . 2005, 14 , 1673–1677. 74. A.Y.M. Christy, K.S. Nahm, Y.J. Hwang, E.K. Suh, M.A. Kulandainathan, T. Premkumar, A.M. Stephan, Ionics 2008, 14 , 157–161. 75. S.A. Kajihara, A. Antonelli, J. Bernholc, Phys. Rev. Lett . 1991, 66 , 2010–2013. 76. Z.G. Lu, C.Y. Chung, Diam. Rel. Mat . 2008, 17 , 1871–1876. 77. J. Robertson, Mater. Sci. Eng. R 2002, 37 , 129–281. 78. K. Honda, M. Yoshimura, K. Kawakita, A. Fujishima, Y. Sakamoto, K. Yasui, N. Nishio, H. Masuda, J. Electrochem. Soc. 2004, 151 , A532–A541. 79. L.-F. Li, Y. Luo, G.G. Totir, D.A. Totir, G.S. Chottiner, D.A. Scherson, J. Phys. Chem. B 1999, 103 , 164–168. 80. A. Lewandowski, M. Galinski, J. Power Sources 2007, 173 , 822–828. 81. H.D. Abru˜na, Y. Kiya, J.C. Henderson, Phys. Today 2008, 61 , 43–47. 82. Y. Zhang, Y. Gui, X. Wu, H. Feng, A. Zhang, L. Wang, T.C. Xia, Int. J. Hydrogen Energ. 2009, 34 , 2467–2470. 83. J.R. Miller, Electrochim. Acta 2006, 52 , 1703–1708. 84. C-Z. Yuan, B. Gao, X.-G. Zhang, J. Power Sources 2007, 173 , 606–612. 85. A.K. Cuentas Gallegos, M.E. Rincon, J. Power Sources 2006, 162 , 743–747. 86. J. Chmiola, G. Yushin, Y. Gogotsi, C. Portet, P. Simon, P.L. Taberna, Science 2006, 313 , 1760–1763. 87. K. Honda, T.N. Rao, D.A. Tryk, A. Fujishima, M. Watanabe, K. Yasui, H. Masuda, J. Electrochem. Soc. 2000, 147 , 659–644.

REFERENCES

481

88. M. Yoshimura, K. Honda, R. Uchikado, T. Kondo, T.N. Rao, D.A. Tryk, A. Fujishima, Y. Sakamoto, K. Yasui, H. Masuda, Diam. Rel. Mat . 2001, 10 , 620–626. 89. K. Honda, T.N. Rao, D.A. Tryk, A. Fujishima, M. Watanabe, K. Yasui, H. Masuda, J. Electrochem. Soc. 2001, 148 , A668–A679. 90. E.C. Almeida, A.F. Azevedo, M. R. Baldan, N.A. Braga, J.M. Rosolen, N.G. Ferreira, Chem. Phys. Lett . 2007, 438 , 47–52. 91. E.C. Almeida, M.R. Baldan, J.M. Rosolen, N.. Ferreira, Diam. Rel. Mat . 2008, 17 , 1529–1533.

19 Use of Diamond Films in Organic Electrosynthesis Siegfried R. Waldvogel, Axel Kirste, and Stamo Mentizi

19.1

INTRODUCTION

Since in electrochemical transformations only electrons are used as reagents, almost no reagent waste is produced. Because of the outstanding atom economy and often pronounced energy efficiency, most electrochemical processes are considered as “green” [1]. Almost 7% of the total industrial electricity consumption is used for electrochemical processes [2]. Consequently, electrosynthesis has a significant technical impact. Furthermore, electric current is considered to be the prime energy of the future since bulk renewable energy resources will consist of photovoltaics, wind energy, water power, and so on. Therefore, novel electrosynthetic transformations will be of particular interest for technical innovations and future applications. Based on coupled nature of anodic and cathodic processes in electrochemical transformations, both reactions can be employed for synthetic purposes [3]. Usually a divided electrolysis cell is applied, which separates the anodic and catholytic compartment [4]. In a few cases the transformation on the anode and cathode can be simultaneously exploited for preparative purposes. These so-called 200%-cells fulfill all aspects of sustainable electrochemical processes and represent the state-of-the-art [5]. The technical realization is the “paired electrolysis” that has been applied by the BASF SE in Germany [6]. An alternative strategy for selective electrochemical transformations employs mediators. Mediators represent electrochemically regenerated reagents. Therefore, the specific reactivity of a given reagent can be applied with using it catalytically. In addition to the electrochemically fueled transformation, these mediators can reach hidden places such

Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

483

484

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

as active sites of enzymes [7] or solid supports [8]. The application of electrochemically generated radicals in organic and biochemical processes is rather limited, since appropriate techniques and in particular suitable electrode materials are still missing. Because aqueous media are commonly used in inorganic technical processes, the electrolysis of water is the significant side-reaction. Electrode materials with a high offset potential for molecular hydrogen and oxygen in cathodic and anodic processes, respectively, can suppress this unwanted drain of electric energy. The investigation and application of boron-doped diamond (BDD) as novel and innovative electrode material is a vivid area of research. The nondestructive conversion of organic compounds on BDD electrodes is a novel and emerging field. In the recent past years, BDD electrodes became available in a variety of sizes, with different support materials and an enhanced stability in organic media. High hopes are placed in this new BDD technology, since tremendous ameliorations in academic and industrial aspects seem to be possible. This is mainly due to the extremely large potential window for boron-doped diamond in aqueous solutions. In particular, anodic processes allow the formation of • OH radicals at potentials well below the offset of oxygen evolution. Boron-doped diamond electrodes thus can be used for new oxidation reactions that otherwise are not possible in water [9]. In methanol containing solutions, also methoxyl radicals were discussed as reactive intermediates [10]. The formation of OH spin centers has been demonstrated using spin traps [11]. On the other hand, these intermediates are employed for disinfection, detoxification, and wastewater treatment: At intermediate potentials a direct, simple electron transfer occurs, whereas the formation of hydroxyl radicals at highly positive potentials leads to a complete incineration (e.g., of 2-naphthol, 4-chlorophenol, and other compounds) via complex oxidation sequences [12]. Partial oxidation (e.g., of phenol to benzoquinone) is also observed [13]. All advantages described in the previous chapters (14, 16, and 17) for the degradation of organic compounds turn here into potential disadvantages that have to be handled (see Figure 19.1).

Figure 19.1 Mode of action for the transformation of organic substrates at BDD electrodes.

19.2 SPECIFIC FEATURES OF BDD ELECTRODES

485

The desired nondestructive pathway toward the product competes with the electrochemical incineration. High current densities will strongly promote the mineralization. Consequently, low current densities should be beneficial for a synthetic and nondestructive transformations. The compartment of electrochemical reactions caused by hydroxyl or methoxyl radicals can be estimated in the range of a few micrometers close to the BDD electrode. Migration of products out of the electrochemical scene into bulk is crucial, since this will prevent the overoxidation. Control of these two competing and critical processes will end either in failure (mineralization) or a selective electroorganic synthesis.

19.2

SPECIFIC FEATURES OF BDD ELECTRODES

The most intriguing properties of boron-doped diamond are the unusual large offset potentials in aqueous media for both electrodes anode and cathode for the evolution of molecular hydrogen and oxygen, respectively. The overpotential for the discharging of protons in a neutral aqueous electrolyte is accounted to be 1.1 V. This is a similar value found for mercury as cathode [14]. Because of the toxicity of mercury and eventually appearing mercury compounds, there are severe environmental concerns connected with the application of mercury as cathode in process chemistry. Therefore, boron-doped diamond may find future applications in electrochemically fueled reductions wherein aqueous media cannot be circumvented (e.g., enzyme catalysis). In contrast, on the anodic side an overpotential for the generation of molecular oxygen from water of about 2.3 V can be determined. This offset potential is significantly larger than common noble electrodes like gold could provide [15]. In aqueous media the offset potentials for the evolution of hydrogen and oxygen create an electrochemical window that should enable a variety of chemical transformations (see Table 19.1) [14–16]. Recently, we found when using fluorinated alcohols instead of water this window is tremendously opened up. Not only on the anodic part is the potential window as anticipated enlarged to 3.2 V but also in the reductive region almost 2 V were gained for performing chemistry. Therefore, the 1,1,1,3,3,3-hexafluoroisopropanol–BDD system currently represents the protic electrolyte with the largest electrochemical window of about 5 V (see Figure 19.2). The supporting electrolyte used in the system is methyltriethylammonium methylsulfate [17,18]. The limitation in the anodic part might be caused by the supporting TABLE 19.1 Offset potential (η) for the evolution of molecular hydrogen or oxygen, respectively. Electrode material Ag Au Pt (solid) Pd Hg Pb Graphite BDD

η for O2 evolution (V)

η for H2 evolution (V)

0.94 [16] 1.53 [16] 1.50 [16] 1.12[16] — 1.02 [16] 1.12 [16] 2.30 [14]

−0.76 [16] −0.32 [16] −0.40 [16] — −1.21 [16] −1.26 [16] −0.99 [16] −1.10 [15]

486

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

Figure 19.2 Unusual broad electrochemical window for the system BDD/1,1,1,3,3,3-hexafluoroisopropanol (HFIP) with 0.1 M Et3 NCH3 O3 SOCH3 . (Reproduced from Ref. 18.)

Figure 19.3 Simplified mode of action BDD anodes in electroorganic synthesis.

electrolyte. The depicted foot potential shows slight conversions that can be attributed to impurities in the 1,1,1,3,3,3-hexafluoroisopropanol (HFIP). In the course of synthetic studies, BDD anodes show a unique reactivity that is based on the selective formation of oxyl radicals. As long as these intermediates are formed preferentially, a simplified mode of action for BDD anodes can be used (see Figure 19.3). The high reactivity of hydroxyl radicals is outstanding and their estimated oxidation potential is right between fluorine and ozone [19]. Control of these species requires innovative strategies in order to accumulate the desired product.

19.3 STABILITY OF BDD ELECTRODES IN ORGANIC MEDIA

487

In particular, for anodic processes BDD electrodes have very promising characteristics. Due to the highly reactive intermediates on the diamond surface, no electrode fouling by coating of by-products or carbonization is anticipated. Eventually, formed coatings on the surface will be mineralized. The resulting products consist of small molecules and will be easily removed from the electrolyte and the electrolysis cell. Since no deactivation of the anode surface is expected and a dramatically reduced effort for the maintenance of electrodes might be encountered, the technical interest for these electrodes exists. Commonly used graphite electrode stacks have to disassembled and mechanically treated to reinstall a useful anode surface. Consequently, such self-cleaning electrodes based on BDD material may cause a small but important cost advantage for chemical processes.

19.3

STABILITY OF BDD ELECTRODES IN ORGANIC MEDIA

Boron-doped diamond coatings are currently available in a variety of different support materials. The most prominent are silicon, titanium, and niobium. However, the conductive diamond surfaces were originally developed for the electrolysis in aqueous media. Under such conditions the electrodes stay intact even when the electrochemically very stable diamond coating is perforated. As soon as a pin hole is formed and the support comes in contact with the aqueous electrolyte, insoluble oxides will be formed by the anodic action of water (see Figure 19.4, I and II). In particular, when silicon is used

Figure 19.4 Corrosion of BDD electrodes in aqueous or organic media.

488

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

Figure 19.5 Macroscopic hole in BDD on niobium upon corrosion with methanolic electrolyte: diameter of perforation: 2 mm, diameter of crater: 8 mm. See color insert.

as support material this is the case and a compact and insulating protection of silicon dioxide is formed. This efficiently deactivates the pin hole. That area will not be anymore electrochemically active and further corrosion is prevented. Such particular passivation pathway of eventually formed pin holes does not anymore apply if working in alcohols, e.g. methanol or phenol, because no protective oxide layer could be formed. In these organic media well soluble tetraalkoxy silicon species are anodically formed. Upon generation they can diffuse into bulk and further anodic dissolution of the support is allowed (see Figure 19.4, III). The boron-doped diamond layer will be still stay intact, whereas the support is continuously eroded. Finally, the diamond surface at that area will collapse mechanically and opens up the door for a quick and dramatic corrosion. The microscopic event turns to a macroscopic hole. The anodic instability of the exposed support material is not limited to silicon. Titanium as well as niobium show quick degradation under anodic treatment in organic media. Figure 19.5 displays a corrosion event due to a pin hole in the diamond layer. Anodic conditions in the presence of methanol perforated a 2 mm thick niobium metal plate within about 30 min. Therefore, good-quality diamond coatings are required when performing electrochemistry in organic media. Currently, 20 μm thick diamond layers as a result of multiple diamond coatings are available and should work for these electrochemical transformations. However, developments of boron-doped diamond coatings on anodically inert support materials are currently under way. Such electrodes will remove current concerns about the employment of BDD electrodes and will provide appropriate BDD electrode material for a broad scope of applications. More innovative supports might be based on conductive ceramics or carbon materials. Self-curing support materials could an alternative strategy providing reasonable stable BDD electrodes.

19.4 ELECTROLYSIS CELLS FOR BDD ELECTRODES FOR ORGANIC TRANSFORMATIONS

489

19.4 ELECTROLYSIS CELLS FOR BDD ELECTRODES FOR ORGANIC TRANSFORMATIONS Several commercial providers for boron-doped diamond electrode materials offer electrolysis cells in a broad variety. Some modular electrochemical cells using BDD electrodes are described in detail [20]. Since these cells were designed for wastewater treatment, they were generally not useful for electroorganic synthesis in nonaqueous media. For synthetic purposes, electrolysis cells with a small volume made of organic-compatible materials are required. Additionally, any contact of the support with the organic electrolyte has to be strictly eliminated in order to avoid the corrosion. Most BDD electrodes are on silicon support, which causes eventual loss of the BDD electrode by brittle nature of crystalline silicon. Consequently, the material used for sealing has to be inert but soft enough to avoid friction of the silicon support. The available BDD electrodes usually exhibit a planar geometry, with BDD coating on a sole side. For experiments with electrolyte volumes in the range 40 to 120 mL, an electrolysis cell equipped with flange fittings was developed (see Figure 19.6). The depicted cell is undivided and the BDD electrode is employed as anode. A net of nickel serves as cathode. The BDD electrode surface is approximately 5 cm2 . The flange arrangement is fixed by a standard flange clamp [21]. This assembly allows the use of every planar type of BDD electrode, which is easily contacted by a metal foil on the back. If the applied voltage on such an electrolysis cell is tremendously too high, usually the oxide layer on the support causes the electric resistance. Slight treatment of the support face with sandpaper prior contacting turns out to be beneficial. Particular attention has to be given to the sealing between the BDD electrode and the flange of glass. Several materials, including Teflon™ and a variety of silicones, have been tested. Best results were obtained by a sealing made of EPDM, which is a highly

Figure 19.6 Preparative electrolysis for organic transformations on BDD electrodes. Electrolyte volume: 40–120 mL.

490

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

resistant polymer. EPDM foils are easily available since it is widely used as foil for garden ponds and chemical sealings. Thermal stability was found up to 120◦ C. Hot organic and aggressive media as well as highly anodic potentials seem not to affect the sealing over a long process time. A swelling with hot phenols is observed. When switching to another phenolic substrate, replacement is recommended since contamination of the electrolyte may occur. The displayed cell geometry offers the possibility to heat the electrolyte. Usually, a sand bath or silicon oil heating are applied. The cooling jacket in the upper part keeps the more volatile components in the electrolysis compartment [21]. Recent electrochemical investigations indicate that these transformations are best performed at high concentrations of substrate up to 90% of the electrolyte composition. Furthermore, good mixing of the electrolyte seems to be crucial, since slow exchange close to the anode promotes electrochemical incineration and consequently diminishes the synthetic utility of such conversions. For such preparative work, a conceptual novel architecture is required. The two major issues are an electrolysis cell on microscale and avoiding nonemployed volumes by efficient stirring of the electrolyte. A microelectrolysis cell with mutual volumes of 0.7–4 mL was designed (see Figure 19.7). All components are made of Teflon™ or coated by this inert material. A cylindrical Teflon™ tube that is closed on one side was routed off in a parallel fashion. Onto these planes the sealing and electrode materials are placed. A clamp will tighten the operating cell by gentle pressing. The round open part for the electrode materials is very close to the magnetic stirrer (part b [dark grey] of Figure 19.7). Suitable magnetic stirrers with a flat surface on one side are commercially available. Good exchange of the electrolyte from the electrode area is found.

Figure 19.7 Design of the microelectrolysis cell, schematically, and without electrodes. (a) Teflon™ body; (b) magnetic stirrer; (c) routed surface for arrangement of electrodes; (d) access to electrolyte.

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

19.5

491

ANODIC TRANSFORMATIONS ON BBD ELECTRODES

The most important features of polycrystalline boron-doped diamond material are its very wide electrochemical window in aqueous and protic media and the capability for specific formation of oxyl radicals. These very reactive spin centers can be created in high concentration in aqueous or alcoholic solutions [22]. This opens unique reaction pathways that have previously been considered as inaccessible by electrochemical means. Therefore, anodic transformations are the most obvious on BDD electrodes. The generated hydroxyl or alkoxyl species may act either as radicals being implemented into the product or serve as strong oxidants that transform the substrate. In the latter reaction sequence the oxyl spin centers play the role of a mediator. 19.5.1

Alkoxylation Reactions

Because of the outstanding oxidative power of the oxyl radicals and their successful employment in wastewater treatment, the use of these intermediates seems to be selfcontradictorily in terms of synthetic use. However, alkoxylations of hydrocarbons is a very important technical field because it allows the installation of functionalities without using the halogenation pathway. The specific introduction of functional groups on nonactivated hydrocarbons lacks in selectivity and was not yet realized by BDD technology. In contrast, the direct anodic methoxylations of activated carbons exhibiting benzylic or allylic moieties can be performed at BDD anodes. The results obtained on BDD electrode are quite similar to the ones when graphite serves as anode material [23]. The anodic synthesis of benzaldehyde dimethylketals is industrialy relevant. A detailed study of the anodic methoxylation of p-tert-butyltoluene (1) at BDD was carried out [24]. Usually, the first methoxylation product (2) and the twofold functionalized derivative (3) are found upon electrochemical treatment (see Scheme 19.1). When using BDD, the accumulation of the dibenzyl intermediate (4) is observed. Most interestingly, this intermediate is not detected when graphite electrodes are applied. This difference has been attributed to a mechanism based on the noncatalytic character of BDD electrodes and to the presence of active functionalities on graphite electrodes [25]. The occurrence of (4) might be also the consequence of hydrogen atom abstraction by

O

O + MeOH –2e–, –2H+

+ MeOH –2e–, –2H+

1

O

2 Scheme 19.1 Anodic methoxylation of p-tert-butyltoluene.

3

492

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

2

1

BDD anode + MeOH BDD anode 4

3 Scheme 19.2

Potential pathway of BDD-mediated methoxylation reaction of 4-tert-butyltoluene.

BDD anode O 5

MeOH

O

O

O

6 Scheme 19.3 Electrochemical methoxylation of furan.

intermediate methoxyl radicals. However, in the course of the electrochemical conversion, (4) is then further converted into the desired products (2) and (3) by cleavage of the central C,C -bond (see Scheme 19.2). The C,C -bond cleavage of stilbenes or bibenzyl derivatives are found at graphite electrodes only at enhanced current densities of approximately 10 A dm−2 . The different reaction pattern is further supported by CV studies, which reveal that BDD electrodes exhibit a 400 mV larger electrochemical window in methanol compared to graphite. This promotes a series of alkoxylation reactions that are not selective on conventional electrodes. Another technically relevant alkoxylation process is the direct methoxylation of furan derivatives. Anodic treatment of furan (5) in an undivided cell provides 2,5-dimethoxy2,5-dihydrofuran (6) (see Scheme 19.3). This particular product represents a double protected 1,4-dialdehyde and is frequently used as C4 building block for N -heterocycle synthesis and material sciences. Industrial electroorganic processes employ graphite electrodes and sodium bromide which act both as supporting electrolyte and mediator [26]. The same electrolysis of (5) can be carried out on BDD electrodes, but no mediator is required. The conversion is performed with 8% furan in MeOH, 3% Bu4 N+ BF− 4, 15 ◦ C and 10 A dm−2 . When 1.5 F mol−1 were applied (6) is obtained in 75% yield with almost quantitative current efficiency. Treatment with 2.3 F mol−1 is rendered by 84% chemical yield for (6) and a current efficiency of 84% [27]. Most probably furan is anodically oxidized in the initial step. Trimethyl orthoformate (8) is a highly activated C1 building block and a formic acid equivalent that is commonly used in organic condensation reactions. (8) is produced on a industrial scale by two major processes: The first commences with chloroform and sodium methanolate, but produces three equivalents of sodium chloride as waste. The second route is based on the methanolysis of cyanhydric acid, which also causes stoichiometric amounts of ammonium chloride. Additionally, the use and handling of

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

O

O

O

O

BDD electrode

493

O 8

7

Scheme 19.4 Electroorganic synthesis of trimethylorthoformate.

O

O O O TBDPS

O

MeOH, KOH

TBDPS

BDD anode O 9

O O 10

Scheme 19.5 Electroorganic formation of benzoquinone ketal.

cyanhydric acid causes high safety costs. Therefore, anodic methoxylation of formaldehyde dimethylacetal (7) to trimethyl orthoformate (8) is an interesting alternative. The conversion is performed on BDD electrodes in undivided cells (see Scheme 19.4). Product (8) is obtained in 75% selectivity with a partial conversion of 27% (7). The consumed charge accounts to 0.4 F mol−1 (7) leading to a current efficiency of 41%. The electrolysis is carried out at ambient conditions in an undivided cell equipped with BDD anode and steel cathode (9 A dm−2 ). The electrolyte consists of 24% MeOH, 70% (7) and LiN(SO2 CF3 )2 /NaOMe [28]. It is noteworthy that this transformation is not possible employing graphite electrodes. The clear advantages for the electrochemical synthesis of (8) on BDD electrodes are based on the attractive raw materials and the reduced costs because of no generation of waste salts and safer operating conditions. The anodic transformation of 1,2- or 1,4-dihydroxysubstituted benzenes to the corresponding quinones or it masked derivatives is well known, since they represent valuable synthetic intermediates [29]. Benzoquinone ketals of electron rich arenes like (9) can be challenging because the oxidative aryl-aryl coupling reaction might compete. When using BDD anodes, the benzoquinone ketal (10) is obtained in almost quantitative manner demonstrating the superior properties of this electrode material (see Scheme 19.5). Despite the basic conditions, no deblocking of the silyl-protected phenol moiety is observed [30]. 19.5.2

Fluorination Reactions

The outstanding electrochemical and chemical stability of diamond seems to be the ideal basis for electrochemical fluorination reactions. The installation of fluorine α to heteroatom substituted positions can be anodically achieved by hydrogen fluoride/triethylamine mixtures. The Fuchigami group tested several electrode materials for the fluorination of oxindole derivative (11). In this conversion to (12), BDD shows only a slight superior behavior (see Scheme 19.6). The results are comparable with platinum as anodic material. Glassy carbon turned out to be less efficient [31].

494

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

SPh

PhS F

–2e–, –2H+ O

O

N

0.1 M Et4NF• 4HF/CH3CN

N

Ph

Ph 12

11

Entry

Anode

1

BDD

66

2

Pt

67

3

GC

31

Yield 12 (%)

Scheme 19.6 Anodic fluorination using Et4 NF · 4HF/CH3 CN.

F

F

F

F

F

Et4NF• 4HF

13 Scheme 19.7

14 Fluorination of 1,4-difluorobenzene on BDD.

The electrochemical fluorination of 1,4-difluorobenzene (13) using BDD electrodes was reported more than a decade ago and represents the first preparative use of BDD anodes for synthetic purposes. As electrolyte NEt4 F·4HF was employed (see Scheme 19.7). The electrolysis was monitored by gas chromatography and the identity of the product found by comparison of the same and known electrochemical conversion at platinum electrodes. The electrolysis was conducted at constant voltage conditions with 2.75 V (vs. Ag/Ag+ ) at the BDD anode. Further preparative details (yield, current efficiency, etc.) for the formation of 3,3,6,6-tetrafluoro-1,4-cyclohexadiene (14) were not given [32].

19.5.3

Cyanation Reactions

In the electrochemical cyanation reaction, the cyanide acts similar to fluoride. After oxidation of the organic substrate, the nucleophilic cyanide enters the reaction scene and forms a less electron-rich product that is deactivated for further anodic conversions. Therefore, the electrochemical cyanation reaction has some significance for aromatic substrates [33].

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

495

CN –2e–, –2H+ N

N

NaCN/MeOH CF3

CF3 15

16

Entry

Electrode

1

BDD

77

2

Pt

75

3

GC

0

Scheme 19.8

N

NaCN, MeOH

17

N

Current efficiency (%)

Cyanation of tertiary amines.

CN

NaCN, MeOH

NC

18 Scheme 19.9

N

CN

19

Anodic cyanation reaction of pyrrole.

The anodic cyanation reaction at BDD electrodes was studied by the Fuchigami group using aliphatic and heteroaromatic amine substrates. Anodic treatment of N,N dibutyl-N -2,2,2-trifluoroethylamine (15) affected the installation of a cyano moiety on a nonfluorinated alkyl portion forming the amino nitrile (16). Best results were obtained for BDD and platinum as anode materials (see Scheme 19.8), wherein BDD turned out to be slightly superior [34]. The cyanation reaction of N -methylpyrrole (17) was carried out at similar conditions providing the mono (18) and dicyanated product (19) (see Scheme 19.9). Platinum and BDD anodes gave both compounds in excellent to total current efficiency. In contrast, glassy carbon gave to rather poor results [34]. In conclusion, the BDD electrodes are very useful in the cyanation reaction and behave very similar to platinum anodes. Since the performance is several examples the best, BDD might be the material of choice. 19.5.4

Cleavage of C,C-Bonds

Angular arenes with exposed multiple bonds are specifically prone to oxidative conversions. The synthetic value of those double bonds arises from their high reactivity toward simple electrophiles, enophiles, and radicals. Heavy metal catalysis, periodate oxidation, and ozonolysis are the standard tools for oxidative bond cleavage in such molecules. For economic and safety reasons, technically applicable alternatives are of great interest. Phenanthrene (20) represents a typical molecule with such activated multiple bond. Electrogenerated oxyl radicals on BDD selectively attack the double bond in positions 9 and

496

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

O

O O

O 20

O

O O

O

21

22

Scheme 19.10 Degradation of phenanthrene.

10 of (20) (see Scheme 19.10). It is noteworthy that aliphatic olefins proved to be stable toward electrochemical oxidation on BDD anodes in aqueous media [35]. However, (20) was successfully converted using acidic methanolic electrolytes [36]. The electrolysis was performed with 2% (20) and 0.5% H2 SO4 in methanol. It was operated with a current density of 3.4 A dm−2 at 54◦ C and an applied charge of 10 F mol−1 (20). With these conditions (21) and (22) were isolated in 15% and 39% yield, respectively. Despite the C,C bond cleavage described in Scheme 19.2 of the dibenzyl derivative (4), the diester (22) represents in this case the major product and not the masked dialdehyde (21). 19.5.5

Oxidation of Activated Carbon Atoms

A variety of organic transformations in aqueous media using BDD anodes have been tested. The pronounced stability of BDD electrodes in the presence of water makes it obvious. However, the yields are usually low and therefore less attractive for synthetic purposes. At the BASF company the anodic oxidation of butin-1,4-diol (23) was investigated. The anodic treatment in an electrolyte of dilute sulphuric acid gave small amounts of the monoacid (25) and the acetylene dicarboxylic acid (24) (see Scheme 19.11). The relative low product efficiency might be caused by electrochemical incineration processes. Anodic treatment of 3,5-lutidine (26) on BDD electrodes turned out to be challenging as well. Only traces of the targeted pyridine-3,5-dicarboxylic acid (27) could be detected (see Scheme 19.12). As electrolyte a dilute NaOH solution was employed. Most probably, the mineralization is also, in this case, the dominant reaction pathway.

O HO

OH

BDD anode

HO

OH

H2SO4

HO

O 23

24

OH O 25

Scheme 19.11 Oxidation of butindiol in aqueous electrolyte.

19.5.6

Anodic Phenol Coupling Reaction

Biphenols are very common motifs in natural products [37] as well as in technical applications [38]. In the past decades, several methodologies were elaborated for the oxidative

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

O

O

BDD anode

497

HO

OH

NaOH N

N

26

27

Scheme 19.12 Unsuccessful synthesis of pyridine-3,5-dicarboxylic acid.

coupling process of electron rich arenes. In particular, the selective ortho-coupling process of phenols has been addressed by numerous catalytic and stoichiometric approaches, wherein sterically hindered tert-butylated phenols as well as naphthols represent preferred substrates [39], since simple phenols and derivatives with methyl substituents on the aromatic core tend to side-reactions. Most oxidative aryl-aryl coupling reactions do not rely on leaving functionalities in the substrates. Commonly, only hydrogen is sacrificed in the course of the reaction. Therefore, such transformations are of outstanding atom economy and consequently of particular interest. Using BDD electrodes in the anodic coupling of such phenols was not very promising from the very beginning since several previous studies have demonstrated the mineralisation of phenols, naphthols and chlorinated derivatives [40]. In particular, in aqueous media a sequence degradation products is observed, which is consistent with the transformations described earlier. The most important challenge will be the control of the intermediate oxyl species in order to avoid the mineralization reaction. 19.5.6.1 Anodic Homo-Coupling of Phenolic Substrates Recently, biphenols based on simple methyl substituted phenol (e.g., 2,4-dimethylphenol) (28) have drawn significant attention as components of highly potent ligand systems [41]. The sustainable synthesis of such biphenols, despite their rather simple scaffolds, is challenging. Methylsubstituted phenols are prone to side reactions and in particular 2,4-dimethylphenol (28) results on anodic treatment in predominantly polycyclic architectures [42]. Direct electrolysis in basic media led only in traces to the desired biphenol 29 and the major components of product mixture consisted of Pummerer’s ketone 30 and consecutive pentacyclic product (31) (see Scheme 19.13) [43]. For an efficient electrochemical access to 3,3 ,5,5 -tetramethyl-2,2 -biphenol (29), we developed a boron-based template strategy [44]. A protocol for the conversion of neat phenol (28) on BDD electrodes was developed in our laboratory. The conversion in solvent containing electrolyte ended up mainly in mineralization reactions. To circumvent this undesired overoxidation, the electrolysis was carried out in almost neat phenol as electrolyte. Gratifyingly, the chemoselectivity toward the desired biphenol was approximately 90 to 95%. In order to improve the yield and reduce electrochemical incineration, only partial conversion of about 30% is performed. In that case the electrochemical transformation is very clean. By using these particular conditions, only 2,4-dimethylphenol (28) is efficiently converted to the biphenol 29 (see Scheme 19.14). The scope of the method was limited because other phenols gave either no detectable products or ended up in complete tar formation—for example, as sesamol [21]. The free path length of oxyl radicals that are generated on BDD electrodes might be only in the range of some nanometers.

498

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

O OH

O O

OH

H O

O

HO 28

29

30

Entry

Anode

1

Pt

31

Electrolyte

Product ratio 29:30:31

MeOH, Ba(OH)2·8H2O neat + 11% H2O, BDD Et3NCH3 O3SOCH3

2

1:11:6 18:1:0

Scheme 19.13 Product distribution for the anodic coupling of 2,4-dimethylphenol.

OH

OH BDD anode

2

additive, MeNEt3 O3SOMe HO 28

Scheme 19.14

29 Phenol coupling on BDD anodes with redox-stable additives.

F3C

OH

33

32

CF3

CF2 OH

HF2C

F2 C

34 OH

OH

36

HF2C

OH

37

C F2

OH

35 O

O CF3

F3C

F2 C

OH

F3C

F3C

38

NH2 39

Scheme 19.15 Survey of tested additives.

Detailed studies of the electrochemically formed products reveal that also the solvents got anodically degraded [21]. Consequently, more redox-stable additives were encountered, leading to fluorinated alcohols (see Scheme 19.15). Previous studies excluded simple acids for this transformation. tert-Butanol (32) is commonly employed as inert alcohol for oxidative transformations. When (32) was used in the conversion of (28) unattractive yields were found (Scheme 19.13; Table 19.2, entry 1). A variety of differently

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

TABLE 19.2 Entry 1 2 3 4 5 6 7 8

499

Anodic treatment of (28) using different additives.a Additive

T (◦ C)

j (mA cm−2 )

29 (%)b

CE (%)c

32 33 34 35 36 37 38 39

45 45 45 45 45 70 20 100

4.7 4.7 9.5 4.7 4.7 4.7 9.5 9.5

10 26 17 18 47 43 15 5

10 26 17 18 47 43 15 5

a Reaction conditions: 2.44 g (28), 30 mL additive, 0.68 g supporting electrolyte, BDD anode, nickel cathode, Q =

1.0 F per mol (28). b Determined from crude product by GC using an internal standard. c Current efficiency.

fluorinated and acidic additives (33–39) were tested [17]. Primary alcohols with fluorous moieties in β position like 2,2,2-trifluoroethanol (33), 2,2,3,3-tetrafluoropraponol (34), or 2,2,3,3,4,4,5,5-octafluoropentanol (35) could be used but give rise to only slight improvement (Table 19.2). Significantly better results were obtained when 1,1,1,3,3,3hexafluoroisopropanol (37) was employed as an additive. In a very clean conversion, 2,4-dimethylphenol (28) is transformed to (29) in 47% (entry 5). Application of more electric current than 1 F per mol substrate did not improve the yield and quality of the product (29). A trifluoromethyl group can be replaced by a phenyl moiety in order to stabilize the oxyl spin center. The stabilizing effect is clearly not based on the acidic nature of these additives. The yield and the current efficiency were studied over the course of electrolysis under optimized conditions for (29), wherein the chemical yield reached a maximum of about 50% when 1 to 1.3 F current was applied per mol (28) and represents a reasonable compromise between yield and current efficiency [17]. The scope of the elaborated protocol is useful for a variety of differently substituted phenols (see Scheme 19.16). Crucial for the conversion is the solubility of the phenolic substrate in the fluorous electrolyte. Sterically demanding phenol (42) could be converted in moderate yield (Table 19.3, entry 1). Better results were obtained with 2-naphthol (41) providing binol as sole product (entry 2). Most remarkably, sesamole (40) was anodically coupled in very good yield (entry 3). Moreover, halogenated phenolic substrates (43–46) were selectively coupled to the corresponding biphenols (entries 4–7). The low yield is based on the electron deficient nature of the substituents. Despite the low yield for the fluorinated biphenol (43) it represents the first direct example for oxidative coupling of such fluorinated substrate. The protocol is easy to perform and practical. When only 1 F electric current is applied, a clean reaction mixture is obtained consisting of products, starting material, and electrolyte [17]. 19.5.6.2 Nonsymmetrical Phenol Coupling and Phenol-Arene CrossCoupling Reaction The cross-coupling reaction to nonsymmetric biaryls is a very versatile and synthetically useful transformation [45]. In most examples leaving functionalities on both reaction partners are required. Furthermore, toxic transition metal catalysts based on palladium are necessary for the arylation reaction [46]. The most prominent methods utilize arylboronic acids, arylstannanes, benzoic acid derivatives, arylzinc, or arylmagnesium reagents creating waste by the employed

500

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

OH OH BDD anode 2

R

37, MeNEt3 O3SOMe

R

R HO OH

OH

OH

O O 40 OH

41 OH

F

42 OH

OH

Cl

Br

Cl 43 Scheme 19.16

44

45

46

Selection of successfully converted phenolic substrates.

leaving groups [47]. The direct oxidative cross-coupling of arenes is a cutting edge concept that only sacrifices hydrogen substituents and is consequently very attractive in terms of atom economy. This approach requires a specific reactivity of one reaction partner toward the employed oxidant that induces the reaction sequence. The oxidized intermediate then attacks the other partner and the transformation can be accomplished. TABLE 19.3 Electrochemical synthesis of biphenols.a Entry

Substrate

j (mA cm−2 )

Yield (Product) (%) b

CE (%)c

42d 41e 40f 43 44 45g 46

4.7 4.7 2.8 4.7 4.7 4.7 4.7

22 41 74 13 30 24 30

22 41 74 13 30 12 30

1 2 3 4 5 6 7

a Reaction conditions: 0.02 mol phenol, 0.1 M supporting electrolyte in 30 mL HFIP (37), BDD anode, nickel cathode,

Q = 1.0 F per mol phenol.

b Isolated biphenol. c Current efficiency. d 0.01 mol phenol. e 0.005 mol phenol. f 0.012 mol phenol.

g 0.04 mol phenol, Q = 2.0 F per mol phenol.

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

501

O OH

OH

O BDD anode

2

OH

37, MeNEt3 O3SOMe

O 47

48

Scheme 19.17 Nonsymmetrical dehydrodimer of 4-methyl guaiacol (47).

TABLE 19.4 Entry 1 2 3

Variation of the current density. j (mA cm−2 )

Current (F mol−1 )

Yield (%)

CE (%)

2.8 4.7 9.5

1.0 1.0 1.0

27 33 14

27 33 14

j = current density; CE = current efficiency; current refers to (47).

This concept was demonstrated by Dohi et al. using stoichiometric amounts of phenyliodine(III)-bis(trifluoroacetate) [48]. Anodic treatment of arenes results usually in the formation of the homo-coupling product because the oxidation potential is the key property. In a few examples the reactive radical cation can be trapped by an abundant reaction partner that is not affected by the electrode in the applied potential range [49]. When treating 4-methylguaiacol (47) by the elaborated protocol using 1,1,1,3,3,3hexfluoroisopropanol on BDD anodes, a selective and symmetric coupling ortho to the phenolic hydroxyl group was anticipated. However, the reactions exclusively provided the ortho-meta coupled product (48) (see Scheme 19.17). The yield of (48) is highly dependent on the current density, indicating that most probably more than one electrode reaction is involved in the sequence. If the current density is in the range of 2.8–4.7 mA cm−2 (48) is easily isolated in about 30% yield (Table 19.4, entries 1 and 2). A rationale for the formation of (24) includes the direct or indirect generation of phenoxyl radicals on the BDD electrode. Because of the applied conditions, lead to concentration recombination of oxyl spin centers can be excluded. The anodic treatment will cause an Umpolung effect because the electron-rich phenol is oxidized [50]. Despite the liberation of a proton, the phenoxyl species still represents an electrophile [51]. The electrophilic attack occurs on the most electron-rich position, which provides in the connectivity of (48) (Scheme 19.17). The specific role is not yet clear but without 1,1,1,3,3,3-hexafluoroisopropanol (32) the conversion does not proceed. However, it is known that (32) enhances the stability of radical intermediates by several magnitudes [52]. Inspired by this transformation and some more mechanistic insights, a novel concept for the first phenol-arene cross-coupling reaction was realized (see Scheme 19.18). First, the phenoxyl species I is electrogenerated and represents a highly reactive intermediate. Excess of arene B will efficiently quench this species forming the intermediate II. That radical or its tautomer III will undergo directly or indirectly the final oxidation to accomplish product AB. A variety of different electron rich arenes

502

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

R2

R3

OH

OH R3

H

BDD anode

+

+

HORf

H R1 A

R3

R2

R3

R1 AB

B

BB

component A e– R1

H+ B D D

R2

R1

R2

H

H

O•

O I

A N O D E

R1 H

H

R2

R2 OH

e–

+ component B

R1

H+

III

R3

O

H R3 II

product AB Scheme 19.18 Ref. 18.)

Mechanistic picture for the anodic cross-coupling process. (Reproduced from

(component B) were used for the cross-coupling with the 4-methyl guaiacol system A. Remarkably, under these conditions no dehydrodimer (48) was found. First, 4-methyl guaiacol was electrolysed in presence of 1,2,4-trimethoxybenzene (see Table 19.5, entry 1). Due to the very electron rich nature of this arene, some oxidative homo-coupling is found. If the amount of electric current is doubled and the current density lowered, the yield as well as the selectivity for the cross-coupling product (49) is tremendously increased. Employing 1,3,5-trimethoxy benzene affords the mixed biaryl (50) in good selectivity, wherein alteration of current density has only little influence (Table 19.5, entry 2). The cross-coupling reaction is compatible with bromo substituents, as product (51) reveals. The high chemoselectivity is remarkable. Anodic treatment on BDD of 3,4,5trimethoxy toluene results in the exclusive formation of the mixed biaryl (52) (Table 19.5, entry 4). The cross-coupling can successfully performed with benzo[1,3]dioxole containing reaction partners resulting the biaryl (53) and (54) in acceptable yields (see Scheme 19.19). Furthermore, naphthalene moieties can be installed onto 4-methyl guaiacol as the examples (55) and (56) reveal (entries 7 and 8 in Table 19.5). The cross-coupling is not limited to 4-methylguaiacol as phenolic moiety. Mixed biaryls based on 2,4-dimethylphenol or 2-bromocresol as component A are also formed by this protocol [18]. All depicted cross-coupling products (49–58) are novel compounds and

19.5 ANODIC TRANSFORMATIONS ON BBD ELECTRODES

O OH

O

Br O

O

O

OH

HO

O

O

O 51

50

52

O O

O

OH

O

OH

O

OH

O

O

O

O

O

53 O

O 56

55

54 O O

OH

HO

O

O

O

49

OH

O

O

O

O O

503

O

OH Br

O

O

57

Scheme 19.19

58

Selection of mixed biaryls that have been made by anodic cross-coupling at BDD.

TABLE 19.5 Converted substrates with electrolysis conditions: BDD anode, nickel cathode, 30 mL HFIP (37), 50◦ C, A:B = 1:10. Entry 1

2 3 4 5 6 7 8 9 10

a b c d a b a b c a b a b a b

Coupling product

j (mA cm−2 )

Current (F mol A−1 )

AB:BBa (GC)

Yield of AB (%)

CE (%)

49

4.7 4.7 2.8 2.8 4.7 4.7 4.7 4.7 2.8 4.7 4.7 4.7 2.8 4.7 4.7 4.7 4.7 2.8 2.8

1.0 2.0 2.0 3.0 1.0 2.0 1.0 2.0 2.0 1.0 2.0 2.0 2.0 1.0 2.0 2.0 1.0 2.0 2.0

1:1 1.5:1 5:1 1.5:1 11:1 7:1 >50:1 15:1 >50:1 >50:1 13:1 9:1 23:1 >50:1 >50:1 12:1 2.5:1 1:1 1:13

17 39 47 34 12 16 18 14 8 11 18 30 25 10 18 33 11 15 5

34 39 47 22 23 16 37 15 8 23 15 30 25 19 15 33 22 15 5

51 50 52 53 54 55 56 57 58

a AB: cross-coupling product; BB: homo-coupling product of the arene component.

504

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

accessible in a single step. In the workup procedure 1,1,1,3,3,3-hexafluoroisopropanol is almost completely recovered as it represents the most volatile component in the reaction mixture. Subsequently, the nonconverted starting materials can be recycled by short-path distillation at about 80% efficiency. This demonstrates that such electrodes cannot only be used for destructive purposes. BDD electrodes represent a novel and innovative material to generate oxyl spin centers for selective transformations. The mineralization of substrates can be depressed by adding fluorinated alcohols and opened the door to a novel electrochemical concept for a sustainable approach to biaryls. Taming of the intermediate oxyl spin centers by the development of suitable additives will be the key to making this methodology attractive to several other anodic cross-coupling reactions.

19.6

CATHODIC SYNTHESIS ON BDD ELECTRODES

Electric current is by far the less expensive reduction reagent that could be applied. Compared to the anodic conversions, the electrochemical reduction is less elaborate. A significant reason is based in the very few electrode materials that can be applied. In particular, protic media allow only cathode materials with high offset potentials for the evolution of hydrogen. Therefore, lead, mercury, and cadmium usually show the best results. The high toxicity of these electrode materials and eventually formed organometallic species thereof lead to a reluctant application of such reductions. Consequently, an environmental-benign substitute for these heavy metal electrodes could have significant academic and technical impact. Due to the protective cathodic polarization of the BDD electrodes, no problems in stability and corrosion of BDD were observed.

19.6.1

Reduction of Oximes

The synthesis of amines starting from the corresponding oximes is a very common method because the full carbon skeleton is provided. The reduction can be performed in a Bouveault-Blanc-type reaction by treatment with alkali metals in the presence of a proton source [53]. In order to avoid reagent waste, the conversion can be conducted electrochemically, wherein mainly reductions at a mercury pool are used [54]. The high overpotential for the H2 evolution by electrolysis of protic electrolytes should be beneficial for cathodic transformations [55]. The electrochemical and chemoselective reduction of cyclopropyl phenylketone oxime (59) to rac-α-cyclopropyl benzylamine (60) was achieved with 96% yield on a BDD cathode with a niobium support (see Scheme 19.20). As electrolyte a MeOH/NaOMe mixture in a divided cell was applied. The results are similar to the conversion at the same conditions (1% NaOMe in anhydrous MeOH, 40◦ C, 3.4 A dm−2 ) using lead cathodes [56]. The cathodic reduction is superior to a catalytic hydrogenation of oxime (59) to the desired compound (60). The product of the conventional reduction by hydrogenation on noble metal catalyst is concomitant with significant amounts of ring-opened compounds, making purification by distillation very challenging. Inspired by this work, the electroorganic reduction was tested on the sterically more demanding menthone oxime (61). The substrate (61), which is derived from optically pure L-menthone, could provide upon reduction both the epimeric products, (+)neomenthylamine (62) and (−)-menthylamine (63), respectively (see Scheme 19.21). For both diastereomers there exist unique applications [57]. The reduction was previously

19.6 CATHODIC SYNTHESIS ON BDD ELECTRODES

505

HO N

NH2 BDD anode divided cell

59

60

Scheme 19.20 Electrochemical reduction of oximes.

BDD cathode OH N

divided cell

61

NH2

NH2

62

63

Scheme 19.21 Cathodic synthesis of optically pure methylamines.

conducted by treatment with 30 equivalents of sodium metal leading safety concerns for a scale-up [58]. However, detailed studies with BDD cathodes revealed that (61) is not a useful substrate. Almost no conversion is observed and only traces of both epimeric amines could be detected. In that case, cathodic treatment using lead results in almost quantitative reduction and a splendid current efficiency (6 F mol−1 , 66%) [59]. 19.6.2

Reductive Carboxylation

Carbon dioxide is good electron acceptor that can be employed for cathodic transformations [60]. Because of the carbon dioxide balance and due to ecological considerations the electrochemical fixation of this particular C1 building block into chemicals seems to be very attractive. 2-Hydroxy-4-methylsulfanylbutyric acid (65), often named as methionine hydroxy analogue (MHA), is an important technical product that is required for animal feeding. (65) exhibits an improved bioavailability compared to the essential amino acid methionine. MHA (65) is made on a several thousand ton scale by treatment of methylsufanylpropionaldehyde (64) with cyanhydric acid to form the corresponding cyanohydrins. Subsequent nitrile hydrolysis provides (65). An electrochemical approach to (65) exploits a reductive carboxylation of (64) with magnesium as sacrificial anode (see Scheme 19.22). The transformation requires a CO2 atmosphere using for both electrodes magnesium. − −1 a Electrolysis in a DMF/NBu+ 4 BF4 electrolyte resulted upon application of 5 F mol conversion of 90% with a selectivity of 75% for (65) [61]. The conversion can also be carried out on BDD electrodes. If a divided cell is used, less magnesium has to be applied. Unfortunately, the direct reduction of the aldehyde (64) providing alcohol (66) is a significant side reaction. The yield for (65) in this process seems to be significantly

506

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

O H

S

OH

BDD cathode + CO2 sacrificial anode: Mg

64

CO2H

S

S

65

OH 66

Scheme 19.22 Cathodic carboxylation to methionine analogue.

O 2 CO2

undivided cell sacrificial anode: Zn

OH

HO O

67

68

Scheme 19.23 Cathodic reduction of carbon dioxide.

lower than on magnesium electrodes (conversion 66%, current efficiency 22%, divided flow cell, Nafion membrane, BDD on silicon support, 20–25 ◦ C, 0.6 A dm−2 ) [62]. However, this is an impressive example for the versatility of BDD electrodes in preparative electro-organic synthesis. The hydrodimerization of carbon dioxide to form oxalic acid seems to be possible in an undivided cell using zinc as sacrificial electrode material (see Scheme 19.23). The − electrolysis is performed at 6 mA cm−2 in 12 mM NBu+ 4 BF4 in DMF. The current efficiency for the formation of oxalic acid is accounted to be 60%. Currently, no further details are available about this particular process [63].

19.7

CONCLUSIONS

The synthetic use of boron-doped diamond electrodes in preparative organic chemistry has just started. Consequently, only a small flavor of the synthetic possibilities was explored. The unique reactivity of the diamond surface promises a fruitful field for future developments. Since BDD stands extremely strong oxidizing species, the fluorination and even fluorine generation of BDD seems to be an interesting topic. A major issue for anodic transformations will be the reliable control of the reactive intermediates generated on BDD. Potential strategies in order to reduce mineralization involve ultrastable electrolytes. Alternatively, the exchange between electrode surface and bulk should be enhanced for diluting the reactive intermediates. Such approaches can include electrolysis in microreactors or exploiting ultrasonic methods.

19.8

ACKNOWLEDGMENT

The authors are very thankful for the financial support by the SFB 813 Chemistry at Spin Centers (DFG). Special tribute goes to Malte Brutschy for the picture of the corroded BDD electrode.

REFERENCES

507

REFERENCES 1. E. Steckhan, T. Arns, W.R. Heinemann, G. Hilt, D. Hoormann, L. Kr¨oner, B. Lewall, H. P¨utter, Chemosphere 2001, 43 , 63. 2. K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, 3rd ed., VCH, Weinheim, 1997. 3. A.J. Fry, Synthetic Organic Electrochemistry, 2nd ed., Wiley, New York, 1989. 4. (a) H.-J. Sch¨afer, in Encyclopedia of Electrochemistry, Vol. 8 (A.J. Bard, M. Stratmann, H.-J. Sch¨afer, eds.), Wiley-VCH, Weinheim, 2004, pp. 125–170; (b) H.-J. Sch¨afer, in Organic Electrochemistry (H. Lund, O. Hammerich, eds., Marcel Dekker, New York, 2001, pp. 207–222. 5. (a) G. Hilt, Angew. Chem. Int. Ed . 2003, 42 , 1720; (b) M. Ishifune, H. Yamashita, M. Matsuda, H. Ishida, N. Yamashita, Y. Kera, S. Kashimura, H. Masuda, H. Murase, Electrochim. Acta 2001, 46 , 3259; (c) Y.-L. Chen, T.-C. Chou, J. Appl. Electrochem. 1996, 26 , 543. 6. H. P¨utter, in Organic Electrochemistry (H. Lund, O. Hammerich, eds.), Marcel Dekker, New York, 2001, p. 1259. 7. (a) F. Hildebrand, C. Kohlmann, A. Franz, S. Luetz, Adv. Synth. Catal . 2008, 350 , 909; (b) C. Kohlmann, W. Maerkle, S. Luetz, J. Mol. Cat. B: Enzym. 2008, 51 , 57; (c) F. Hildebrand, S. Luetz, Tetrahedron: Asymmetry 2007, 18 , 1187; (d) W. Maerkle, S. Luetz, Electrochim. Acta 2008, 53 , 3175; (e) N. Rao, S. L¨utz, K. Seelbach, A. Liese, in Industrial Biotransformations, 2nd ed., Wiley-VCH, Weinheim, 2006, pp. 115–145. 8. (a) S. Nad, R. Breinbauer, Angew. Chem. Int. Ed . 2004, 43 , 2297; (b) S. Nad, R. Breinbauer, Synthesis 2005, 3654. 9. J. Iniesta, P.A. Michaud, M. Panizza, C. Comninellis, Electrochem. Commun. 2001, 3 , 346. 10. D. Zollinger, U. Griesbach, H. P¨utter, C. Comninellis, Electrochem. Commun. 2004, 6 , 600. 11. B. Marselli, J. Garcia-Gomez, P.A. Michaud, M.A. Rodrigo, C. Comninellis, J. Electrochem. Soc. 2003, 150 , D79. 12. (a) M.A. Rodrigo, P.A. Michaud, I. Duo, M. Panizza, G. Cerisola, C. Comninellis, J. Electrochem. Soc. 2001, 148 , D60; (b) M. Panizza, P.A. Michaud, G. Cerisola, C. Comninellis, J. Electroanal. Chem. 2001, 507 , 206. 13. J. Iniesta, P.A. Michaud, M. Panizza, G. Cerisola, A. Aldaz, C. Comninellis, Electrochim. Acta 2001, 46 , 3573. 14. C. Comninellis, G. Chen, Electrochemistry for the Environment , Springer, Berlin, 2009. 15. A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, Diamond Electrochemistry, Elsevier, Tokyo, 2005. 16. (a) A. Hickling, F.W. Salt, Trans. Faraday Soc. 1940, 36 , 1226; (b) A. Hickling, S. Hill, Discuss. Faraday Soc. 1947, 1 , 225. 17. A. Kirste, M. Nieger, I.M. Malkowsky, F. Stecker, A. Fischer, S.R. Waldvogel, Chem. Eur. J . 2009, 15 , 2273. 18. A. Kirste, G. Schnakenburg, F. Stecker, A. Fischer, S.R. Waldvogel, Angew. Chem. Int. Ed . 2010, 49 , 971. 19. (a) W.H. Koppenol, J.F. Liebman, J. Phys. Chem. 1984, 88 , 99; (b) K. Honda, Y. Yamaguchi, Y. Yamanaka, M. Yoshimatsu, Y. Fukada, A. Fujishima, Electrochim. Acta 2005, 51 , 588. 20. W. Haenni, A. Perret, P. Richen, 2002, patent WO 2002061181. 21. I.M. Malkowsky, U. Griesbach, H. P¨utter, S.R. Waldvogel, Eur. J. Org. Chem. 2006, 4569. 22. N.R. Wilson, S.L. Clewes, M.E. Newton, P.R. Unwin, J.V. Macpherson, J. Phys. Chem B 2006, 110 , 5639. 23. H. P¨utter, A. Weiper-Idelmann, C. Merk, M. Fryda, C.-P. Klages, L. Sch¨afer, A. Hampel, 2003 EP 1 036 861 B1. 24. C. Bosma, S. Gouws, P. Loyson, B. Zeelie, S. Afr. J. Chem. 1999, 52 , 133.

508

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

25. D. Zollinger, U. Griesbach, H. P¨utter, C. Comninellis, Electrochem. Commun. 2004, 6 , 605. 26. D. Degner, Top. Curr. Chem. 1988, 148 , 1. 27. (a) C. Reufer, K. Moebus, T. Lehmann, C. Weckbecker, patent WO 2004087999; (b) C. Reufer, T. Lehmann, R. Sanzenbacher, C. Weckbecker, patent WO 2004085710. 28. R. Fardel, U. Griesbach, H. P¨utter, C. Comninellis, J. Appl. Electrochem. 2006, 36 , 249. 29. S. Quideau in Modern Arene Chemistry (D. Astruc, ed.), Wiley-VCH, Weinheim 2002, pp. 539–572. 30. T. Saitoh, E. Suzuki, A. Takasugi, R. Obata, Y. Ishikawa, K. Umezawa, S. Nishiyama, 217rd ECS Meeting, Vancouver, 2010 (Abstract 809). 31. T. Fuchigami, personal communication 2009. 32. F. Okino, H. Shibata, S. Kawasaki, H. Touhara, K. Momota, M. Nishitani-Gamo, I. Sakaguchi, T. Ando, Electrochem. Solid-State Lett . 1999, 2 , 382. 33. K. Yoshida, J. Am.Chem. Soc. 1979, 101 , 2116. 34. T. Fuchigami, 203rd ECS Meeting, Paris, 2003 (Abstract 2604). 35. U. B¨aumer, H.-J. Sch¨afer, J. Appl. Electrochem. 2005, 35 , 1283. 36. U. Griesbach, D. Zollinger, H. P¨utter, C. Comninellis, J. Appl. Electrochem. 2005, 35 , 1265. 37. G. Bringmann, C. G¨unther, M. Ochse, O. Schupp, S. Tasler, in Progress in the Chemistry of Organic Natural Products, Vol. 2 (W. Herz, H. Falk, G.W. Kirby, R.E. Moore, eds.), Springer, Wien, 2001. 38. (a) W.S. Knowles, Adv. Synth. Catal . 2003, 345 , 3; (b) R. Noyori, Adv. Synth. Catal . 2003, 345 , 15; H. U. Blaser, E. Schmidt eds., Asymmetric Catalysis on Industrial Scale, Wiley-VCH, Weinheim, 2004. 39. G. Lessene, K.S. Feldman, in Modern Arene Chemistry (D. Astruc, ed.), Wiley-VCH, Weinheim, 2002, pp. 479–538. 40. B. Boye, P.A. Michaud, B. Marselli, M. M. Dieng, E. Brillas, C. Comninellis, New Diamond Front. Carbon Technol . 2002, 12 , 63. 41. Recent examples in monodentate P(III) ligands and catalysis: (a) C. Hawner, K. Li, V. Cirriez, A. Alexakis, Angew. Chem. Int. Ed . 2008, 47 , 8211; (b) L. Palais, I.S. Mikhel, C. Bournaud, L. Micouin, C.A. Falciola, M. Vuagnoux-d’Augustin, S. Rosset, G. Bernardinelli, A. Alexakis, Angew. Chem. Int. Ed . 2007, 46 , 7462; (c) M. Vuagnoux-d’Augustin, S. Kehrli, A. Alexakis, Synlett 2007, 2057; (d) K. Li, A. Alexakis, Chem. Eur. J . 2007, 13 , 3765; (e) C.A. Falciola, A. Alexakis, Angew. Chem. Int. Ed . 2007, 46 , 2619; (f) E. Raluy, M. Dieguez, O. Pamies, J. Org. Chem. 2007, 72 , 2842; (g) Y. Mata, O. Pimies, M. Dieguez, Chem. Eur. J . 2007, 13 , 3296; (h) A. Alexakis, D. Polet, S. Rosset, S. March, J. Org. Chem. 2004, 69 , 5660; (i) K. Li, A. Alexakis, Angew. Chem. Int. Ed . 2006, 45 , 7600. Examples in bidentate P(III) ligands and catalysis: (j) J.M. Garner, C.P. Lenges, R.J. Mc Kinney, PCT Int. Appl ., patent WO 2008008927, 2008; (k) M. Bartsch, R. Baumann, G. Haderlein, T. Aechtner, J. Scheidel, H. Luyken, P. Pfab, W. Siegel, V. Weiskopf, PCT Int. Appl ., patent WO 2006040023, 2006; (l) M. Bartsch, R. Baumann, G. Haderlein, M.A. Flores, T. Jungkamp, H. Luyken, J. Scheidel, W. Siegel, PCT Int. Appl ., patent WO 2005042547, 2005; (m) M. Bartsch, R. Baumann, G. Haderlein, M.A. Flores, T. Jungkamp, H. Luyken, J. Scheidel, W. Siegel, F. Molnar, PCT Int. Appl ., patent WO 2004087314, 2004; (n) M. Bartsch, R. Baumann, D. P. Kunsmann-Keitel, G. Haderlein, T. Jungkamp, M. Altmayer, W. Siegel, F. Molnar, PCT Int. Appl ., patent WO 2003033142, 2003. 42. I.M. Malkowsky, C.E. Rommel, K. Wedeking, R. Fr¨ohlich, K. Bergander, M. Nieger, C. Quaiser, U. Griesbach, H. P¨utter, S.R. Waldvogel, Eur. J. Org. Chem. 2006, 241. 43. (a) J. Barjau, P. K¨onigs, O. Kataeva, S.R. Waldvogel, Synlett 2008, 2309; (b) S.R. Waldvogel, Pure Appl. Chem. 2010, 82 , 1055.

REFERENCES

509

44. (a) I.M. Malkowsky, U. Griesbach, H. P¨utter, S.R. Waldvogel, Chem. Eur. J . 2006, 7482; (b) I.M. Malkowsky, R. Fr¨ohlich, U. Griesbach, H. P¨utter, S.R. Waldvogel, Eur. J. Inorg. Chem. 2006, 1690. 45. (a) A. de Meijere, F. Diederich, eds., Metal-Catalyzed Cross Coupling Reactions, Vol 1 & 2, Wiley-VCH, Weinheim, 2004; (b) M. Beller, C. Bolm, eds., Transition Metals for Organic Synthesis, 2nd ed., Wiley-VCH, Weinheim 2004; (c) J. Hassan, M. S´evignon, C. Gozzi, E. Schulz, M. Lemaire, Chem. Rev . 2002, 102 , 1359; (d) A. Zapf, M. Beller, Top. Catal . 2002, 19 , 101. 46. (a) L. Ackermann, ed., Modern Arylation Methods, Wiley-VCH, Weinheim, 2009; (b) X. Chen, K.M. Engle, D.-H. Wang, J.-Q. Yu, Angew. Chem. Int. Ed . 2009, 48 , 5094; (c) I.V. Seregin, V. Gevorgyan, Chem. Soc. Rev . 2007, 36 , 1173; (d) L. Ackermann, Top. Organomet. Chem. 2008, 24 , 35 (e) G. Dyker, ed., Handbook of C-H Transformations, Wiley-VCH, Weinheim, 2005. 47. (a) N. Miyaura, A. Suzuki, Chem. Rev . 1995, 95 , 2457; (b) G.A. Molander, N. Ellis, Acc. Chem. Res. 2007, 40 , 275; (c) V. Farina, V. Krishnamurthy, W.J. Scott, Org. React . 1997, 50 , 1; (d) L.J. Gooßen, N. Rodriguez, K. Gooßen, Angew. Chem. Int. Ed . 2008, 47 , 3100; (e) E.-I. Negishi, Acc. Chem. Res. 1982, 15 , 340; (f) P. Knochel, ed., Handbook of functionalized Organometallics, Wiley-VCH, Weinheim 2005. 48. T. Dohi, M. Ito, K. Morimoto, M. Iwata, Y. Kita, Angew. Chem. Int. Ed . 2008, 47 , 1301. 49. S. Yamamura, S. Nishiyama, Synlett 2002, 533. 50. Umpolung effect by electrochemical treatment: (a) H.J. Sch¨afer, Angew. Chem. Int. Ed . 1981, 20 , 911; (b) R.D. Little, K.D. Moeller, Electrochem. Soc. Interface 2002, 11 , 36; (c) K.D. Moeller, Tetrahedron 2000, 56 , 9527; (d) K.D. Moeller, Top. Curr. Chem. 1997, 185 , 50. 51. The electrophilic nature of aryl radicals was recently demonstrated (σ -type): (a) M.R. Heinrich, Chem. Eur. J . 2009, 15 , 820; (b) A. Wetzel, V. Ehrhardt, M.R. Heinrich, Angew. Chem. Int. Ed . 2008, 47 , 9130; (c) O. Blank, A. Wetzel, D. Ullrich, M.R. Heinrich, Eur. J. Org. Chem. 2008, 3179. The electrophilic nature of phenoxyl radicals: (d) G. Albarran, R.H. Schuler, J. Phys. Chem. A 2005, 109 , 9363; (e) T. Iida, J. Ohshita, N. Ohta, K. Komaguchi, Y. Itagaki, M. Shiotani, A. Kunai, J. Organomet. Chem. 2003, 688 , 192; (f) E.N. Hristea, I.C. CovaciCimpeanu, G. Ionita, P. Ionita, C. Draghici, M.T. Caproiu, M. Hillebrand, T. Constantinescu, A.T. Balaban, Eur. J. Org. Chem. 2009, 626. 52. (a) L. Eberson, M.P. Hartshorn, O. Persson, Angew. Chem. Int. Ed . 1995, 34 , 2268; (b) L. Eberson, M.P. Hartshorn, O. Persson, J. Chem. Soc., Perkin Trans. II 1995, 1735. 53. (a) L. Bouveault, G. Blanc, Compt. Rend. 1903, 136 , 1676; (b) B.I., Seo, L.K. Wall, H. Lee, J.W. Buttrum, D.E. Lewis, Synth. Commun. 1993, 23 , 15. 54. J. Tafel, E. Pfeffermann, Ber. Dtsch. Chem. Ges. 1902, 35 , 1510. 55. H.B. Martin, A. Argoita, U. Landau, A.B. Anderson, J.C. Angus, J. Electrochem. Soc. 1996 143 , L133. 56. U. Griesbach, D. Zollinger, H. P¨utter, C. Comninellis, J. Appl. Electrochem. 2005, 35 , 1265. 57. (a) M.C. Schopohl, C. Siering, O. Kataeva, S.R. Waldvogel, Angew. Chem. Int. Ed . 2003, 42 , 2620; (b) C. Siering, S. Grimme, S.R. Waldvogel, Chem. Eur. J . 2005, 11 , 1877; (c) M.C. Schopohl, A. Faust, D. Mirk, R. Fr¨ohlich, O. Kataeva, S.R. Waldvogel, Eur. J. Org. Chem. 2005, 2987; (d) M. Bomkamp, C. Siering, K. Landrock, H. Stephan, R. Fr¨ohlich, S.R. Waldvogel, Chem. Eur. J . 2007, 13 , 3724; (e) W. Schade, C. Bohling, K. Hohmann, C. Bauer, R. Orghici, S.R. Waldvogel, D. Scheel, Phot. Int . 2007, 1 , 32; Photonik 2006, 38 , 70; (f) R. Orghici, U. Willer, M. Gierszewska, S.R. Waldvogel, W. Schade, Appl. Phys. B 2008, 90 , 355; (g) S. B¨orner, R. Orghici, S.R. Waldvogel, U. Willer, W. Schade, Appl. Opt . 2009, 48 , B183; (h) U. Schwartz, R. Großer, K.-E. Piejko, B. B¨omer, D. Arlt, DE3532356A1; (i) M. Grose-Bley, B. B¨omer, R. Großer, D. Arlt, W. Lange, DE4120695; (j) B. B¨omer, R. Großer, W. Lange, U. Zweering, B. K¨ohler, W. Sirges, M. Grose-Bley, DE19546136A1; (k)

510

58. 59. 60. 61.

62. 63.

USE OF DIAMOND FILMS IN ORGANIC ELECTROSYNTHESIS

W. Lange, R. Grosser, B. K¨ohler, S. Michel, B. B¨omer, U. Zweering, DE19714343A1; (l) J. Looft, T. V¨ossing, J. Ley, M. Backes, M. Blings, EP1989944A1. M.C. Schopohl, K. Bergander, O. Kataeva, R. Fr¨ohlich, S.R. Waldvogel, Synthesis 2003, 2689. U. Griesbach, S.R. Waldvogel, J. Kulisch, I.M. Malkowsky, PCT Int. Appl ., patent WO 2008003620 A2 20080110, 2008. J. Yoshida, K. Kataoka, R. Horcajada, A. Nagaki, Chem. Rev . 2008, 108 , 2265. (a) T. Lehmann, R. Schneider, C. Weckbecker, E. Dunach, S. Oliviero, 2002, patent WO 02/16671; (b) T. Lehmann, R. Schneider, C. Reufer, R. Sanzenbacher, GDCh-Monographie 2001, 23 , 251. C. Reufer, M. Hateley, T. Lehmann, C. Weckbecker, R. Sanzenbacher, J. Bilz, 2006 EP 1 631702. T. Lehmann, E. Dunach, personal communication, 2009.

Part V

Bioelectrochemical Applications

20 Diamond Sensors for Neurochemistry Bhavik Anil Patel

20.1

INTRODUCTION

Monitoring of signaling chemicals in the human body is essential for understanding the means of communication between two cells and, to a greater extent, the function of a biological process. Analytical tools, specifically electrochemical techniques, have over the past 40 years become the method of choice to study neurochemical processes since their conception by Ralph Adams [1–5]. Electrochemical methods coupled with the application of microelectrodes allow neurochemicals to be recorded with a high temporal and spatial resolution, which make them suitable for in vitro and in vivo recordings. Carbon-based microelectrodes (namely, carbon fiber electrodes) to date have been predominantly used for biological measurements due to the biocompatibility of these electrodes. However, diamond electrodes are showing all the attributes that make them attractive for in vitro and in vivo sensing measurements of neurotransmitters. These electrodes provide a measure of the flux of the neurochemical, which is influenced by the mechanisms and kinetics of the neurotransmission process. This specific point has allowed a means of gaining an insight into how chemical signaling alters during various disease states. 20.2

CENTRAL AND PERIPHERAL NERVOUS SYSTEM

The nervous system is extremely complex and signaling between components of the nervous system controls bodily functions. The nervous system is subdivided into smaller systems, which are shown in Figure 20.1. The nervous system is initially divided into the central nervous system (CNS) and the peripheral nervous system. The majority of Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

513

514

DIAMOND SENSORS FOR NEUROCHEMISTRY

NERVOUS SYSTEM Central nervous system Brain

Spinal Cord

Peripheral nervous system Afferent nervous system

Efferent nervous system

Somatic nervous system Sympathetic nervous system

Autonomic nervous system

Parasympathetic nervous system

Enteric nervous system

Figure 20.1 Organization of the nervous system.

sensing-based measurements of neurotransmitters has been carried out in the CNS. These include measurements of dopamine from various regions of the brain using carbon fiber microelectrodes [6–9]. Limited electrochemical-sensing research has been carried out in the periphery, as this area of the nervous system was felt to pose more challenges for measurement than the CNS and was also regarded to be of slightly less importance than understanding transmission in the brain. However, not only has the importance of the peripheral nervous system increased over recent years but this area has also become an important initial focus of diamond sensors. The peripheral nervous system is further divided to the afferent (sensory) nervous system and efferent (motor) nervous system. The efferent nervous system is further subdivided into the somatic and autonomic systems. Somatic nerves carry impulses to skeletal muscle, whereas autonomic nerves carry impulses to smooth muscle, the heart, and glands. The autonomic system is further subdivided into the sympathetic nervous system, parasympathetic nervous system, and enteric nervous system. However, the enteric nervous system is also believed to be independent from the autonomic nervous system. The transmitters in the sympathetic nervous system are required to regulate the contraction of veins and arteries, which is essential for maintaining blood flow to various organs in the body. The enteric nervous system is utilized to regulate the major functions of the digestive tract such as motility. Diamond electrodes have been already utilized to study the role of neurotransmitters in sympathetic veins and arteries [10,11], and to detect neurotransmitters released from the enteric nervous system [12–14]. These areas have been investigated only recently, as they were regarded as complex matrixes for measurements using conventional carbon electrodes. Thus, diamond electrodes have opened the door for measurements to be carried out in all regions of the nervous system and will help in our understanding of the role and function of chemical signaling in the whole body. 20.3

THE PROCESS OF NEUROTRANSMISSION

Neurons are the cellular unit of the central and peripheral nervous systems and are specialized for relaying signals from one location to another within the body. Figure 20.2 shows the common features of a neuron, which contains a large cell body, dendrites, and

20.3 THE PROCESS OF NEUROTRANSMISSION

Dendrite

Transporter

Transmitter molecule

515

Axon Cell body

Postsynaptic receptor

Terminal

Vesicle

Synapse Presynaptic cell terminal

Presynaptic receptor (autoreceptor)

Postsynaptic neuron

Figure 20.2 Anatomy and function of a neuron.

an axon that conveys messages toward the terminal. It is at the terminal where the signal is relayed to other cells by releasing chemical messengers called neurotransmitters. The site of contact between a presynaptic terminal and a target cell (either another neuron or an effector cell, such as a muscle cell) is called a synapse. Information is passed through a neuron in the form of electrical potentials across the cell membrane. This electrical activity is essential, as it is required to open calcium channels that are potential dependent. A neuron has a negative resting potential (this resting membrane potential varies between different neurons, giving the cell its specific characteristics), which is with respect to the outside environment, due to an ionic concentration gradient across the cell membrane. Stimuli received by the dendrites are processed by the cell body and a resulting response is provided by an action potential. The action potential is a rapid electrical event that propagates down the axon to the terminal, which drives the resting potential more positive. At the terminal the action potential causes an exchange of ions across the neuronal membrane. Initially voltage-gated ion channels are opened, causing an influx of sodium, and the neuron resting potential becomes more positive, causing a depolarization. This depolarization critically is responsible for opening voltage-gated calcium ion channels. The influx of calcium leads to the release of neurotransmitters from the presynaptic membrane into the synaptic cleft from individual packages known as vesicles. It takes longer for voltage-gated potassium channels to open, and thus potassium rushes out of the cell, reversing this depolarization and returning the neuron resting membrane potential back to its initially negative value. The key processes of the presynaptic cellular terminal are shown in Figure 20.3. After release, the neurotransmitter has to diffuse across the synapse (1–50 nm) and interacts with receptors on the postsynaptic membrane. The remainder of the released neurotransmitter molecules can undergo three major fates: (1) reuptake back into the presynaptic membrane via transporters, which is the predominate fate of the majority

516

DIAMOND SENSORS FOR NEUROCHEMISTRY

Action potential Na+ Channel Na+

Ca2+ Channel Ca2+ Presynaptic neuron terminal

Na+/K+ ATPase autoreceptor

K+ Na+ Transporter

vesicle

Neurotransmitter

Postsynaptic cell terminal

Figure 20.3 Presynaptic processes involved in the release of neurotransmitters.

of neurotransmitters released, (2) metabolism either in the synaptic cleft or within the presynaptic cell, or (3) diffuses into a remote location. These fates are limited to classical neurotransmitters such as dopamine and serotonin. The receptors on the postsynaptic membrane are responsible for relaying the communication, but receptors on the presynaptic, also known as autoreceptors, are essential for regulating the release of the neurotransmitter. Reuptake by the transporters on the presynaptic neuron is the predominant fate and occurs on a millisecond time scale so that the cell can reset for the next message. After uptake into the releasing cell, the transmitter may be repackaged into vesicles for re-release, or metabolized. This complete process of transmission is responsible for altering the flux of the transmitter in the extracellular matrix, and thus is reflected during electrochemical measurements. Alterations in the response can therefore be attributed to changes in the amount of neurotransmitter released or it rate of reuptake. 20.3.1

Neurotransmitters

The nervous system contains a huge variety of neurotransmitters and neuromodulators (a chemical that can influence the release of a neurotransmitter or the response of the postsynaptic cell to a neurotransmitter), and more molecules are consistently being discovered over time. These neurotransmitters can be classified into five major groups: (1) small amines molecules such as serotonin (5-HT), dopamine (DA), and norepinephrine

20.4 ELECTROANALYTICAL METHODS TO STUDY NEUROTRANSMITTER RELEASE

517

(NE); (2) neuronal gases such as nitric oxide, hydrogen sulphide, and carbon monoxide; (3) amino acids such as glutamate, glycine, aspartate, and γ -aminobutyric acid (GABA); (4) peptides such as substance P; and (5) other small molecules such as acetylcholine and adenosine tri-phosphate (ATP). The structures of some of these neurotransmitters are shown in Figure 20.4. Although there are large numbers of neurotransmitters present, they are not all found in the same location and they vary in their distribution throughout the central and peripheral nervous systems. This ultimately provides a degree of selectivity that aids detection; however, another degree of selectivity is provided as not all neurotransmitters are easily oxidized within a short potential window (±1 V) at the electrode. The transmitters shown in grey in Figure 20.4, are not currently measured directly using electrochemical sensors, but enzyme incorporated electrodes have been used to measure these molecules [15–20].

20.4 ELECTROANALYTICAL METHODS TO STUDY NEUROTRANSMITTER RELEASE The ideal type of electrode and electroanalytical technique will depend greatly on the particular biological application investigated and the signaling molecules to be detected. Five major parameters that influence the choice of technique utilized for in vivo and in vitro electrochemical measurements are: (1) electrode response time, (2) electrode size, (3) electrode material, (4) sensitivity of the method, and (5) selectivity of the method [21–24]. From the variety of electroanalytical methods utilized, only a few methods have been widely used for biological measurements, as they offer the excellent temporal resolution that is essential for detection of neurotransmitter events, which occur over the millisecond domain [8,25–29]. Other than the rapid duration of such signaling, the alterations in release of such signaling molecules are of great interest and thus real-time long-term measurements are very desirable but at present are extremely limited. Constant potential amperometry (CPA) and fast scan cyclic voltammetry (FSCV) are the most widely used

OH O

HO

Norepinephrine

N

O Acetylcholine

N H

NH2

HO Dopamine NH2

Serotonin

O N H O

O OH

HN

HO

OH

H 2N

GABA HO Figure 20.4

N

NH2

OH

H2N

NH2 Histamine

O

OH Glutamate

Melatonin

O N H

N O Nitric oxide

Structure of small molecule neurotransmitters present within the nervous system.

518

DIAMOND SENSORS FOR NEUROCHEMISTRY

techniques from all other electroanalytical approaches [24,30–35]. When utilizing CPA, the current is continuously monitored at a fixed potential where the oxidation/reduction of the analyte of interest occurs. This technique is ideal for in vitro and in vivo investigations, for it has excellent time resolution. However, this technique offers poor selectivity and therefore applications have been limited to situations where only one analyte is known to change concentration during recordings. CPA has been used widely for the measurement of single vesicular events from a variety of single cells and to study nitric oxide release from a variety of biological preparations [28,36–41]. CPA has been used with diamond electrodes for the measurement of norepinephrine from mesenteric arteries and veins [10–11,42] and has also been successfully applied to conduct measurements of serotonin overflow from ileum and colon tissue sections [12,13]. The waveform used for FSCV consists of a linear potential ramp applied to the working electrode to change the potential (see Figure 20.5). This is then reversed and returns to the initial potential. The resulting current versus the applied potential is typically recorded. The current response, however, contains a high capacitive response due the application of such high scan rates, but background subtraction techniques are applied to enhance the Faradaic response of the measured analyte. FSCV, which is widely used to measure neurotransmitter levels, utilizes potential ramps at scan rates greater than 100 V s−1 , where the oxidized form of the analyte of interest is re-reduced before any following homogenous reactions can occur. Many investigations on neurotransmitter dynamics have carried out using FSCV measurements at scan rates between 300 and 1000 V s−1 [25,26,43–45] for the detection of dopamine and serotonin. FSCV offers a degree of selectivity as different neurochemicals in the extracellular fluid can be distinguished by their oxidation and reduction potentials, but lacks the ability to conduct real-time measurements. Fast scan cyclic voltammetry has been predominately used for in vitro and in vivo measurements. The Wightman group have elegantly utilized these methods for the measurement of dopamine release from nucleus accumbens (NAc) from awake animals to study behavioral traits and alterations during drug abuse [46,47]. The potential sweep utilized by the FSCV method has allowed for the detection of multiple neurotransmitters [44,48,49] when used in tandem with principle component analysis.

(b)

E2 Time (s)

E1 t0

t1 One cycle

E pa

Current (A)

Potential (V)

(a)

E pc Potential (V)

Figure 20.5 Fast scan cyclic voltammetry where (a) shows the waveform applied. During the cycle, the potential goes from E1 to E2 , and then returns to E1 . One cycle is shown between the time ranges t0 and t1 . In (b), a typical cyclic voltammetric i–E curve is shown where the solid line shows the capacitive response, and the dotted line shows the Faradaic peaks observed at Epa and Epc , which become much clearer following background subtraction.

20.4 ELECTROANALYTICAL METHODS TO STUDY NEUROTRANSMITTER RELEASE

519

There have also been other techniques used for the detection of neurotransmitters, such as potential step methods and chronoamperometry [50–54]. Chronoamperometry has been successfully applied to study the clearance (reuptake) of a variety of neurotransmitters from synaptosomes [55,56]. Other novel methods are also being established such as AC voltammetry, which can provide a means of measuring multiple neurotransmitters without the need of subtraction methods. Scanning-based electrochemical methods [57–59] can provide 3D maps of transmitter release from cells or tissue sections, which give excellent spatial information. Such new and novel methods will improve our ability to provide excellent spatial resolution and selectivity, which are greatly lacking in the currently utilized techniques. 20.4.1

Sensors Utilized

A variety of different electrodes have been utilized for measurements of neurotransmitters, but nearly all of these experiments have been conducted using microelectrodes that are 1 to 30 μm in diameter [60–64]. These microelectrodes provide excellent spatial resolution, which is important for biological measurements in order to obtain suitable localized measurements of cellular or tissue activity. The ability of microelectrodes to detect neurotransmitters depends not only on their sensitivity but also on the discrimination of a small analytical signal on top of an often large background consisting of the electrode capacitance and/or interfering species in the background [33,65–67]. The most common microelectrode that has been utilized to date for in vitro and in vivo measurements has been the carbon fiber microelectrode [7,68–73]. This electrode has been widely used, due to its biocompatibility, and provides a suitable surface for electron transfer of nearly all neurotransmitter known to be electroactive. The major difference in the carbon fiber microelectrodes used for previous studies has been the geometry, where cylindrical (protruding tip) [62,67,74] or elliptical (disc) electrodes [66,75–77] have been widely used for biological measurements. For the measurements of dopamine, protruding tip carbon fiber electrodes had a four-fold increase in the signal-to-noise ratio compared with disc electrodes [78]. However, quantification with cylindrical electrodes also poses a problem as the sides of the electrode shaft are exposed, thereby making it difficult to know from which area measurements are conducted. These effects are greater in solution rather than during measurements from cells. Carbon fiber microelectrodes have been utilized to good success for measurements of individual vesicular release from single cells [28,36,79,80], for measurements of transmitter release from tissue slices [81–85] and also have been implanted in vivo for recordings [6,9,21,26]. For reproducible measurements, however, the lifespan of carbon fiber microelectrodes for biological measurements are limited and this has been shown to limit measurements from biological tissue to around 10–50 s. Some enhancements have been offered by the application of film membranes, which not only help prolong the lifespan of the electrode but also improve sensitivity and, in some cases, selectivity of the electrode. Nafion®, a perm-selective film membrane used to coat the surface of the electrode, has been widely used to prevent large biological macromolecules from blocking the electrode surface [86–88]. It has also been known to reduce sensitivity to uric acid, DOPAC, and ascorbic acid [89–92]. Other membranes, such as overoxidized polypyrrole and poly-lysine/polystyrene, have been shown to offer similar enhancements [77,93,94]. Other sensors have been used for detection of neurotransmitters, such as platinum, gold, and copper [95]. Copper electrodes specifically have been shown to provide a means

520

DIAMOND SENSORS FOR NEUROCHEMISTRY

of detecting amino acids and peptides [96–97]. However, none of these electrodes has been extensively used for biological measurements due to the clinical stigma of using metal-based sensors for monitoring biological molecules.

20.5 LIMITATIONS OF CURRENT TECHNIQUES FOR IN VITRO AND IN VIVO MONITORING Measurements of neurotransmitters using microelectrodes has been carried out for over 40 years and there are still some major limitations; however, the difficulty of such measurements and the biological significance of the results obtained from such electrochemical measurements drive researchers forward to enhance methodology, techniques, and sensors. As pointed out before, the majority of research articles published for the measurements of neurotransmitters from biological systems are mainly from regions within the brain or from single cells [6,9,26,28,34,36,98–100] and also predominately for the detection of one transmitter, which is dopamine [7,8,47,101–106]. The former point is not due to a limitation of areas of interest for detection of signaling molecules, but the brain poses the most exciting prospect for neurotransmitter detection. The latter point is the one that accounts for the first of many limitations of current techniques. Dopamine is an easily oxidized neurotransmitter that is well located in specific areas of the brain, which is handy in terms of selectivity, but also from a sensing perspective, once oxidized it forms oxidative by-products that are known to foul the electrode. This fouling is less problematic when oxidizing dopamine compared to other transmitters such as serotonin, histamine, and melatonin. Due to the ease of measuring dopamine and also norepinephrine on carbon fiber microelectrodes and the lack of subsequent fouling, these transmitters have been the primary focus of the majority of currently utilized measurements and these signaling molecules are abundant and present in the central nervous system, with some roles within the periphery. As fouling is a limitation, the time course for measurements also becomes a limitation, as stable measurements can only be carried out in the domain of seconds. This poses less of an issue for single vesicular measurements where vesicles fuse and release their content within a few milliseconds. The Wightman group has measured release of dopamine in vivo using carbon fiber microelectrodes, where the stimulated release occurs between 2 and 10 s [8,47,46,84,107]. These time domains are suitable for fast dynamic transmission mechanisms that occur in the examples mentioned earlier, but these characteristics are not observed in all biological regions. For example, steady-state transmission has been observed when monitoring serotonin release from the gastrointestinal tract. The other important factor is that although the biological signaling mechanism occurs over a short domain, long-term measurements are of upmost interest to all researchers. Recording signals in a domain of minutes to possibly days is of upmost interest because it allows for a means to study the effects of pharmacological agents on the regulation of the signaling mechanism. There is a strong belief that just like motor pattern generation observed from neuronal cells, transmitter release could follow a pattern or oscillation type of behavior, which could only be understood with longer-term repeated measurements. Another limitation is the spatial nature of current measurements. As our need for greater understanding of transmission process increases, we require the means of gaining spatial information over a region of tissue to understand the variations within a tissue section. At present, carbon fiber microelectrodes of diameters 5–30 μm are utilized and

20.5 LIMITATIONS OF CURRENT TECHNIQUES FOR IN VITRO AND IN VIVO MONITORING

521

are felt to be biocompatible and are well known to be less invasive than microdialysis probes [108,109]. However, tissue sections are felt to have regions of altered release and reuptake and therefore measurements in various patches of tissue are of interest. There have been approaches that have proven to be successful when using single microelectrodes on tissue sections, using scanning electrochemical microscopy (SECM), but these methods cannot be used for in vivo measurements and they require specialized setups for biological measurements [57,58]. Other approaches have used microelectrode arrays for detection of transmitters over various regions and provides the most likely means forward [19,20,72,110]. As one limitation is always linked or due to another, the limitation in the time frame for measurements due to fouling from oxidative by-products of neurotransmitters that are just too difficult to measure have not just been ignored. Modifications in the electrodes and the electroanalytical techniques have been approached to combat this problem. Carbon fiber electrodes are still the material of choice, but chemical modification of the electrode surface or application of film membranes have been used for improving stability, sensitivity, and, to some degree, selectivity. In some cases, for example, it has been believed that the application of Nafion® has improved selectivity against ascorbate [89–92] and 4-sulfobenzene modified carbon fiber microelectrodes have improved sensitivity and selectivity of dopamine detection relative to the conventionally used carbon fiber microelectrode [111]. However, hardly any of these modifications to the electrode have had any influence on reducing the level of electrode fouling, except for one modification where carbon nanotubes have been trapped within a film membrane over the electrode [44]. The second approach to overcome these problems has led to the application of FSCV, which was discussed in detail previously. This rapid technique utilizes scan rates high enough to outrun any preceding chemical reaction following oxidation of the molecule of interest. In practice, the concept is a good one and has shown to work very well for the detection of dopamine [33,69,78,112,113], but detection of serotonin over time poses a greater challenge and requires scan rates of 1000 V s−1 [25,81], where fouling is still observed. The other limitation, albeit a small one, is that by conducting FSCV, each scan is completed in a duration of 10 to 100 ms, which is of course not the same as realtime monitoring. However, the best approaches from sensor modifications and technique modifications have enhanced selectivity and sensitivity to the point where zeptomole quantities of neurotransmitter can be measured [114] to date. The major limitation at present for in vitro and in vivo measurements, then, all fall back to the same problem: a limitation in the ability to carry out stable biological measurements over a duration of minutes to hours. This limitation can only be solved by changing the material for sensing, as this is practically all we have got left from the technique to alter, and thus diamond-based electrodes can provide a suitable option. Of course other electrodes have been investigated such as carbon nanotubes and diamond-like carbon (DLC) materials [44,115]. The next sections look deeper into the key problems influencing the ability to conduct long-term stable biological measurements and the current diamond sensor research that has provided promise to solve such issues. 20.5.1

Long-Term Recordings

One of the major limitations of current biological measurements is the ability to conduct measurements over a period of hours if not days. To understand changes in biological

522

DIAMOND SENSORS FOR NEUROCHEMISTRY

processes and, more importantly, to study the onset of disease, measurements of chemical markers over a longer time frame is required. At present, in vitro or in vivo longterm measurements have been limited to techniques utilizing microdialysis sampling. Boutelle and co-workers have used this form of sampling to measure glucose and lactate release from human patients suffering from head injuries. These measurements have been recorded for two important timeframes: (1) during surgical procedures, which can be typically between 1 and 6 h; and (2) in the postsurgery intensive care unit where recordings can be obtained for day to over a week. These two timeframes are critical for understanding changes in cerebral blood flow and the onset of ischemia during and following surgery. These timeframes for some biological scenarios may also be short, which may be the case in neurophysiological disorders such as dementia or understanding the onset of cognitive decline. The other important issue is from the measurement point of view. Monitoring the chemical matrix is of importance for understanding changes, but also sensors and tools can also be used diagnostically. For example, deep brain stimulation has been used to trigger dormant areas of the brain during various neurodegenerative diseases, where microelectrode arrays are implanted and pulse-short electrical trains trigger neurological activity. The lifespan of these devices depends greatly on the biocompatibility of the device, and thus a robust inert material can provide the ability for long-term stability for activity, which is where diamond-coated or diamond-based devices are currently being investigated [120,121]. If you consider the current measurements of neurotranmsitters from all the areas of the body that have been investigated using electrochemical sensors, measurements have been in the time domain from seconds to minutes [34,84,122,123]. Limited measurements have been carried out for the duration of hours and in the case of current challenges, sensors and devices need to break into this time domain. Measurements are carried out only in the short-time domain as sensors are influenced from biocompatibility responses of the electrode to the tissue and vice versa. This balance influences if the electrode is prone to fouling or if the tissue responds from the invasive nature of this electrode. The majority of measurements have been limited due to these factors; however, some insights are present in the current literature. Carbon fiber microelectrodes have effectively dominated measurements of neurotransmitters, as many have shown that these electrodes are biocompatible, but these electrodes are so biocompatible that proteins, cells, and anything else in the matrix is attracted to the sensor, causing major fouling over time. Critically this also influences the biological areas in which the carbon fiber microelectrodes can be utilized, as in the brain and from single cells, measurements have been made for 10 s to less than a minute, but in the gastrointestinal tract, measurements would be limited in the duration of a few seconds. The solution to these limitations has been approached either by using film membranes on the electrode surface or by using biocompatible agents released in the vicinity of the electrode to prolong the activity of the electrode. The first approach is a nonrunner because most film membranes currently fail over time and thus would provide variation in responses over time. The use of biocompatible agents released to prevent the formation of fouling from the biological matrix has already proven to be a good solution to allow for long-term measurements. Typically, application of small doses of nitric oxide have been used to prevent biofouling. By addition of endogenous substances in a biological matrix, however, there may be alterations to the activity and function of the tissue.

20.5 LIMITATIONS OF CURRENT TECHNIQUES FOR IN VITRO AND IN VIVO MONITORING

523

Altering the electrode material has recently started to be considered as a better approach, and diamond- and nanotube-based electrodes are being currently investigated. Diamond-based electrodes have all the attributes that can provide an answer to conducting long-term biological measurements. Due to the robust nature of the electrode, the sp3 hydribized structure and the lack of oxygen-functional groups on the surface, the electrode material not only overcomes the effects of fouling, but the surface of the electrode can be stable for a longer time frame. Boron-doped diamond (BBD) microelectrodes have already shown the ability to conduct stable measurements of neurotransmitters in complex biological matrixes [10–12,42,124,125]. For example, measurements of serotonin overflow have been conducted from the mucosa of isolated ileum tissue sections for over an hour, which is far greater than the capabilities of any other microelectrode that has been used over the past 40 years for biological measurements. Diamond-based sensors have a good future in providing the bridge of the currently used temporal and cumulative techniques that allow for measurements of neurotransmitters during all time scales of importance.

20.5.2

Fouling from Large Biomolecules

For all biological measurement either in vitro or in vivo, the measurements are mainly influenced from protein fouling. For electrochemical measurements from the brain to the blood, sensors are placed in highly complex biological extracellular matrixes that are full of proteins and other large biomolecules, which have a huge affinity to adsorb to the electrode surface. Biomolecule adsorption to the electrode surface alters electrochemical responses by impeding the diffusion of the analyte of interest to the electrode surface, thus decreasing reaction kinetics or modifying electrode geometry and/or area [94]. Furthermore, protein fouling is very likely to trigger a pathological response from the tissue environment. For example, when measurements are carried out in the blood, the sensor is prone to fibrous capsulation, which has been well described in the prosthesis literature [126–128], and its mechanism is similar to that of blood clotting or an immune response. In Figure 20.6, the immune response that occurs during sensing measurements in the blood is shown. Following initial protein fouling, platelets aggregate on the electrode, but only after these platelets become activated is that sensor completely fouled, as a fibrous capsule is formed around the electrode [129]. Proteins fouling to the electrode surface can be expected as a physiological response to the implantation of an

Platelets

sensor

sensor

sensor

sensor

sensor BLOOD

Red blood cells

Activated platelets

Protein

Fibrin

Figure 20.6 Biofouling process that occurs during measurements in blood. (Adapted from Ref. 129.) See color insert.

524

DIAMOND SENSORS FOR NEUROCHEMISTRY

in vivo sensor and thus limit the ability to conduct long-term recordings. The biological response may alter the sensor’s calibration and the presence of the electrode in the biological matrix may also influence concentrations of the analyte released [130]. To overcome these problems, the best approach is to prevent biological macromolecules such as proteins from adsorbing on to the electrode material. This is the initial step in the biofouling process that triggers a biochemical and cellular cascade of events leading to inflammation and encapsulation. However, solutions not only need to offer protection to the electrode surface to conduct measurements but they also must not hinder the natural biological activity of the tissue that the measurements are conducted from. Over the past 30 years, one major solution to this problem has been to use film membranes to cover the electrode surface to offer biocompatibility [94,131,132]. These film coatings, however, have limited stability over time and thus offer a short-term fix. A more successful recent approach has been to utilize bioactive polymers that release small signaling molecules such as nitric oxide [129,133,134] to prevent the formation of a fibrous encapsulation over the electrode and to allow for longer-term measurements. Note, however, that when endogenous molecules are released in the biological sampling matrix, this may alter the biological activity. Diamond electrodes are an alternative means of overcoming the effects of protein fouling; they also allow for a means of carrying out long-term measurements. Figure 20.7 shows response of ruthenium (III) hexaammine (an outer sphere reversible redox couple used for studying electrochemical behavior) on a 3 mm BDD electrode in PBS buffer, and also in the presence of 4% serum albumin (which is the most abundant protein in blood) and a homogenised liver matrix. In the presence of albumin and the liver matrix, there is a decrease in the current; however, this current is stable over a period of 30 min. This decrease may be due to blockage of some active sites on the electrode. In comparison, glassy carbon and gold electrodes, these electrodes are completely fouled in the same duration. Another important feature is that there seems to be minimal alteration in the kinetic behavior of the reduction of ruthenium (III) hexaammine on the electrode in these complex matrixes. However, on predominately glassy carbon electrodes, electron transfer kinetics become more sluggish over time. This demonstrates that the BDD electrode is suitable for stable measurements in protein and biomolecule matrixes and can provide

Current (μA)

10

(b) Control + 4 % albumin + liver matrix

5 0 –5 –10 –15 –0.3 –0.2 –0.1

0.0 0.1 0.2 Potential (V) vs. Ag|AgCl

BDD 0

Cathodic current (μA)

(a) 15

0.3

–5

–10

–15

–20

Control + 4 % albumin + liver matrix

Figure 20.7 Response of BDD electrode in protein environments. In (a), cyclic voltammograms of ruthenium (III) hexaammine are shown in the presence of 4% albumin and in a liver matrix, and in (b), the cathodic current is shown. (Reprinted from Ref. 94.)

20.5 LIMITATIONS OF CURRENT TECHNIQUES FOR IN VITRO AND IN VIVO MONITORING

525

the means of conducting long-term measurement in a variety of complex biological environments such as the blood stream and the digestive tract. There have been other studies of protein fouling on diamond electrodes. Unicellular microalgae Chlorella vulgaris were entrapped within a bovine serum albumin (BSA) membrane and immobilized directly onto the surface of a diamond electrode for heavy metal detection; fouling was considerably lower than with platinum electrodes [135]. In another long-term study, ultrananocrystalline diamond (UNCD) films were used as a coating for retinal microchips and were implanted for 4 to 6 months. The diamond film allowed for the stable long-term performance of the microchip and prevented biofouling [136]. 20.5.3

Fouling from Redox Reaction By-Products

Electrode fouling by redox reaction by-products can result from poor solubility of reaction by-products, in which case they simply fall out of solution and deposit on the electrode, or it can be driven by specific chemical forces, such as dipole-dipole and ion-dipole interactions with a polar, oxygenated electrode surface. The adsorption of redox reaction intermediates and/or products is often irreversible and the concentrations of these species are far greater in the immediate vicinity of the electrode surface compared with the bulk solution. The greatest problem that occurs from this mode of fouling is that these species are usually being generated from the biological analyte of interest for measurement, thus the strategies available for decreasing or preventing electrode fouling are limited. This form of fouling is observed when the neurotransmitter serotonin (5-HT) is oxidized at the electrode surface. There have been three methods used to overcome this complication: (1) surface and/or potential waveform modifications; (2) application of permselective polymer coatings (e.g., Nafion); and (3) use of surfactants in the biological media [137–141]. All are effective at reducing the extent of electrode fouling from biological spectator molecules, but they provide no practical means of prolonging the electrode response for 5-HT recordings during biological measurements. To overcome the effect of redox by-products, a greater understanding of the electrooxidation mechanism of the analyte is required to overcome the fouling issue. Wrona and Dryhurst [142–144] have carried out a comprehensive study of the oxidation mechanism of 5-HT at carbon paste electrodes in phosphate buffer. 5-HT is initially oxidized in a reversible one-electrode reaction to form a radical cation (5-HT •+ ) which, in a ratedetermining step, deprotonates to produce the 5-HT radical (1; 5-HT • ) as shown in Figure 20.8. The fouling will therefore be dependent on flux: For small (∼10 μm) electrodes the threshold for fouling can be as low as 5 μM and is correspondingly less severe for

NH3+

NH3+

NH3+ H

HO

–e N H 5-HT

H

HO

–H+ N H

O N H 5-HT

Figure 20.8 Oxidation mechanism of serotonin.

526

DIAMOND SENSORS FOR NEUROCHEMISTRY

larger electrodes or at faster scan rates. The 5-HT radical then reacts with another 5-HT molecule to form dimers and trimers. These products are all cationic and show a high affinity for the electrode surface [142,143]. Fouling is enhanced due to the sp2 hybridized carbon structure as well as the presence of oxygen functional groups on the surface of the electrode, which are feature of carbon fiber electrode. Diamond electrodes are an ideal sensor for detecting redox species that have fouling by-products, as they overcome the problems associated with carbon fiber microelectrodes as they lack oxygen function groups and are sp3 hybridized. Figure 20.9 shows cyclic voltammetric i-E curves for 50 μM 5-HT in PBS buffer at a BDD and carbon fiber microelectrode. At this concentration, a single oxidation peak (Ia ) is observed on the forward scan at about +500 mV for both electrodes, which corresponds to the reversible one-electron oxidation/deprotonation reaction to form the 5-HT radial (5-HT • ) from 5-HT. After scan reversal, a reversible couple (IIc and IIa ) appears for the carbon fiber microelectrode but is absent for the BDD microelectrode. The reversible couple observed on the carbon fiber is similar to that observed by Wrona and Dryhurst on a carbon paste electrode [142,143]. The appearance of the reversible couple IIc /IIa , with the difference in peak potential being approximately zero (Ep ∼ 0) and the ratio of the anodic to cathodic currents being approximately equal to one (Iap /Icp ∼ 1), suggests that this couple is due to an adsorbed dimer or oligomer produced on the first scan that undergoes fast electron transfer [144]. The absence of this redox couple (IIc and IIa ) on BDD electrode indicates that the dimers or oligomers may be produced but they are not adsorbed on the electrode surface. Therefore, the BDD electrode surface has a lower affinity for adsorption of such oxidation reaction by-products and results in increased resistance to fouling. This decreased affinity is most likely due to the fact that BDD has relatively low coverage of carbon-oxygen functional groups on the surface and does not posses an extended π –π electron system like carbon fiber, making the BDD

300 200

Ia

IIa

Current (nA)

100 0 –100 –200 IIc –300

Carbon fiber BDD

–400 –0.4

0.0 0.4 Potential (V) vs Ag|AgCl

0.8

Figure 20.9 Cyclic voltammetric i–E curve for 5-HT obtained on carbon fiber (black line) and BDD (grey line) microelectrode in phosphate-buffered saline (PBS) buffer pH 7.4 carried out at a scan rate of 50 mV s−1 . Arrows indicate the direction of the scan.

20.5 LIMITATIONS OF CURRENT TECHNIQUES FOR IN VITRO AND IN VIVO MONITORING

527

microelectrode less prone to strong adsorption of polar molecules. Similar responses have been observed before for the detection of 5-HT on BDD electrodes relative to glassy carbon electrodes [145]. Measurements were carried out to compare the long-term stability of BDD microelectrodes in comparison with carbon fiber microelectrodes for the measurements of 5-HT. Repeated cyclic voltametric i-E curves for 10 μM 5-HT in PBS buffer on a BDD and carbon fiber microelectrode is shown in Figure 20.10. As clearly shown, the carbon fiber microelectrode is completely fouled (oxidation peak is completely lost) after the 105th cycle (about 600 s), whereas the diamond electrode is still functional after 330 cycles (about 1900 s). Responses of the normalized oxidative current versus time are shown for 10 and 1 μM 5-HT on both electrodes in Figure 20.11. Concentrations of 5-HT and all other neurotransmitters within the central nervous system and periphery lie within 10 to 1 μM, and thus for accurate and stable measurements of neurochemical fouling from oxidative by-products these concentrations ranges should be investigated prior to in vitro or in vivo measurements. The BDD microelectrode within this concentration range is extremely stable, as shown in Figure 20.11. The BDD microelectrode current decreases for the initial 200 s, which is felt to be due to fouling of favorable sites, which are most likely to be remaining sp2 sites on the electrode surface or possibly grain boundaries in between the diamond crystal structure, which appear to promote molecular adsorption. However, the flux dependent nature of electrode fouling from these by-products is observed for the carbon fiber microelectrode, where the sensor is complete fouled after 600 and 5900 s for 10 and 1 μM 5-HT, respectively. The biggest problem for long-term measurements of 5-HT is due to fouling from oxidative by-products, which has limited the amount of electrochemical investigation of 5-HT release from biological tissue; however, this is not only selectivity to 5-HT, as all classical neurotransmitters can foul electrochemical sensors to varying degrees [42,78,112]. Similar responses for BDD microelectrodes in comparison with carbon fiber microelectrodes for the detection of norepinephrine [10,11,42] and with glassy carbon electrodes for the detection of histamine have also been observed [146]. In addition, diamond films have been shown to be stable for the detection of dopamine [147] as well (a) 300

(b) 300 1st cycle 45th cycle 105th cycle 330th cycle

200

1st cycle 45th cycle 105th cycle

250 Current (nA)

Current (nA)

250

150 100

200 150 100

50

50

0

0 –50

–50 0.2

0.8 0.6 0.4 Potential (V) vs Ag|AgCl

1.0

0.2

0.6 0.8 0.4 Potential (V) vs. Ag|AgCl

1.0

Figure 20.10 Cyclic voltametric i–E curves for 10 μM 5-HT. Where (a) shows the response from the BDD microelectrode and (b) shows response from carbon fiber microelectrode. All cyclic voltammetric i–E curves were carried out at 100 mV s−1 were PBS buffer was the supporting electrolyte.

Normalised current (%)

(a)

DIAMOND SENSORS FOR NEUROCHEMISTRY

100 80 60 40 Carbon fibre BBD

20 0

(b)

100

Normalised current (%)

528

80 60 40 20 0

0

500

1000

1500

2000

0

1000 2000 3000 4000 5000 Time (s)

Time (s)

Figure 20.11 Long-term electrochemical measurements of 5-HT. Where (a) shows normalized current response of carbon fiber and BDD microelectrode in 10 μM 5-HT in PBS buffer and (B) shows responses in 1 μM 5-HT in PBS buffer pH 7.4.

as melatonin [13]. Thus, BDD sensors are opening avenues for the stable detection of signaling molecules that have been known to be a nuisance for currently used electrodes. Although BBD electrodes allow for resistance to fouling and therefore provided a means of long-term measurements, there is one slight limitation. Electrochemical studies of various neurotransmitters at slower scan rates (<100 mV s−1 ) have shown that the diamond electrode requires a greater overpotential for the oxidation of neurotransmitters compared to carbon fiber microelectrodes. This indicates that there is sluggish kinetics for the electron transfer during oxidation of transmitters on the diamond surface. For example, the oxidation of norepinephrine and 5-HT requires a further 200 mV to oxidize the compound compared to carbon fiber microelectrodes [12,42]. When faster scan rates are employed, an additional 500 mV overpotential is required for electron transfer on the diamond electrode compared to the carbon fiber microelectrode for mostly all neurotransmitters, as shown in Table 20.1. The sluggish reaction kinetics on the diamond electrode are due to the fact that most all neurotransmitters undergo electron transfer following an adsorption step on the electrode surface, which is more favorable on the carbon fiber microelectrode compared to the diamond electrode and thus a greater overpotential is required for the oxidation of the neurotransmitters. This may pose a limitation because for some biological applications, the potential window used may be restricted.

TABLE 20.1 Summary of cyclic voltammetric data for several bioanalytes at diamond and carbon fiber microelectrodes at high potential sweep rates. All solutions were 0.01 mM of the redox system, except for ascorbic acid and DOPAC (0.2 mM), in 0.1 M phosphate buffer (pH 7.4). Data are presented for the oxidation of each redox system at 250 V s−1 scan rate. Results are presented as mean ± standard deviation (n ≥ 3). Redox system

Ep a (Diamond) (mV)

Dopamine Norepinephrine Serotonin DOPAC Ascorbic acid

829 ± 50 902 ± 34 635 ± 11 1076 ± 55 1012 ± 53

Ep a (Carbon Fiber) (mV) 370 ± 20 420 ± 20 440 ± 10 580 ± 150 585 ± 80

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

529

For example, neurons have a fixed resting membrane potential that will be influenced if higher potentials are poised for measurement, altering their physiological activity. Overall, BDD microelectrodes have opened avenues for long-term measurements to be conducted of all small molecule classical neurotransmitters. At present, the influence of oxidative by-product fouling has limited measurements on carbon fiber microelectrodes of predominately dopamine for fewer than 30 s and forced the application of fast-scan cyclic voltammetry to outrun the chemical reactions that precede the oxidative reaction for monitoring other transmitters.

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY A wide range of applications have been investigated using diamond sensors in biological environments. In the following sections a summary of some of these applications are discussed in detail with indications on some of the benefits. An interesting feature of the applications that follow is that they are predominately in new areas of biological measurement, such as the sympathetic and enteric nervous systems, where there have been limited analytical measurements. Thus, studies in this area are dramatically advancing our knowledge of the field. Another feature of the current work in using diamond electrodes in neurochemistry is that measurements are carried out in complex matrices for a longer duration. 20.6.1

Recording Neuronal Activity

Recording neuronal activity is important to gain an understanding of the function and behavior of a biological system. Voltage-gated ion channels are responsible for maintaining the cellular membrane potential and generating trains of electrical impulses known as action potentials. These action potentials can be modulated by pharmacological agents to alter their firing rate and pattern, and, importantly, they can be monitored using extracellular microelectrodes. A variety of microelectrodes such as tungsten wires have been used, which are attractive to use for electrophysiological monitoring because they can record neuronal activity without penetrating the cellular membrane [148]. Diamond-based electrodes have only recently been introduced into this area, shortly after the introduction of planar microelectrode arrays as a suitable platform for monitoring electrical activity and simultaneous culturing neurons. However, due to the limitation of the materials that can be used for the fabrication of microelectrode arrays, diamond-based sensing platforms were identified as an alternative approach. Diamond materials are attractive to this area due to its chemical inertness, biocompatibility, optical transparency, and high conductance when the surface is functionalized. It is the optical transparency that is the main attraction as other electrode materials offer limited ability for monitoring the cells using microscopy and conducting electrical measurements in tandem. Diamond surfaces through modulation of the surface topography and fictionalization have allowed for cell adhesion and growth of PC12 neuronal cells [135] and rat hippocampal neurons [120]. A limited number of studies have used diamond electrodes, but there has been successful application of electrical monitoring on diamond films. Hydrogen-terminated conductive diamond films have successfully been used for the culture of neuronal GT1-7 cells and provided the means to measure extracellular electrical activity. Biphasic responses

530

DIAMOND SENSORS FOR NEUROCHEMISTRY

were observed for a duration of 8 and 60 ms, which are comparable to noble metal microelectrode arrays [120,121]. Attempts to use BBD microelectrodes for neuronal stimulation in motor neurons of the pond snail, Aplysia californica, have also been conducted [149]. The application of such electrodes allows of a means of multi-parametric recordings of electrical activity from living cells, which is of importance in understanding the complete picture of cellular signaling. Another benefit of the application of diamond films and electrodes to study electrical signaling is that they have been shown to be biocompatible, as neurons have been successfully cultured on these surfaces. So this suggests that BDD sensors can eventually be used to monitor electrical activity in vivo. The advances of current measurements may also provide the way forward for developing devices for deep brain stimulation, and thus using the electrode as a source of electrical activity to stimulate dormant regions of the brain, as observed in Parkinson’s disease [150–152]. 20.6.2

Single Cell Measurements of Vesicular Release

Measurements of the mechanisms of transmitter release have developed as an important field for helping in our understanding of the communication process between single cells. Electroanalytical tools have provided the means of measuring the release of neurotransmitters from small packages known as vesicles into extracellular space. Carbon-based electrodes have been used extensively for such measurements, but BDD microelectrodes have been successfully applied to measure such events. Disc-shaped BDD microelectrodes with diameter of about 40 μm were used to measure 5-HT release from an identified serotonergic neuron from the pond snail, Lymnaea stagnalis. CPA was used for real-time recordings and individual vesicular release events as observed, as shown in Figure 20.12. Events varied from 10 to 30 ms in duration and are similar to those observed on carbon fiber microelectrodes. The BDD microelectrode seems able to detect such events even if the oxidation of 5-HT on the electrode is more sluggish than the carbon fiber microelectrode. But, due to

a

b

c

50 pA pA a b

c

100 100 ms ms 50 pA pA 20 s

Figure 20.12 Single vesicular events from a serotonergic neuron in the intact CNS of the pond snail, Lymanea stagnalis. Measurements obtained using 40 μm BDD microelectrode at potential of +750 mV versus Ag| AgCl.

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

531

the rapid duration of these events and the ease of mass electrode fabrication, the carbon fiber microelectrode seems to be more suitable. These rapid measurements and low concentrations of release also negate any major fouling issues to the electrode surface. However, the low capacitive current of diamond electrodes allow for the detection of more events that may be hidden in the noise from carbon fiber electrodes and can thus can give a more clearer picture of the complicated field of vesicular release dynamics. Another area that diamond electrodes could be of use is for long-term measurements of vesicular events in the form of microelectrodes or electrode arrays. Various diamondbased microelectrode arrays have been fabricated [153,154] and could provide excellent spatial cellular measurements, as current microelectrode arrays can only be utilized for short-term measurements due to fouling from cells, matrix components, and biological waste products. Other than measuring neurotransmitter released from single cells and the dynamics of this release profile, the measurements of neurotransmitters inside the cell is also desirable. This concept has been conducted using separation methods [102,155] as sensors but to date they fail to have the features required for such measurements. Diamond sensors could provide the breakthrough, as they are robust enough to allow for cell membrane penetration and do not suffer from protein fouling, which other metalbased sensors are prone to. Diamond-based sensors will have a strong future in single cell and cellular measurements as the demand to understand internal and external activity of the cell grows.

20.6.3 Neurotransmitter Release from Sympathetic Nerves Innervating Mesenteric Arteries Through a series of papers, Park et al. [10,11,42] have described the use of diamond electrodes to measure norepinephrine (NE) release from sympathetic nerves innervating smooth muscle cells in rat mesenteric arteries. NE is a vasoconstrictor neurotransmitter, released from sympathetic nerves, that acts on the α1-adrenergic receptors of smooth muscle cells to elicit a contractile response. Gaining a better understanding of the neurogenic control mechanisms of arterial and venous tone, and how these control mechanisms are altered in hypertension, is the main goal of these studies. Two of the key findings from these initial studies were that the bare diamond microelectrode exhibited (1) a pHindependent voltammetric background current and (2) superb resistance to deactivation and fouling during electrically evoked NE release measurements [10,42]. In contrast, a bare carbon fiber microelectrode exhibited (1) a pH-dependent background voltammetric current response with evidence for electroactive surface carbon-oxygen functional groups and (2) deactivated irreversibly during exposure to tissue. For example, in continuous amperometric measurements of the electrically evoked release of NE over a 4-h period, with the adipose and connective tissues removed from the blood vessel, the diamond microelectrode exhibited only a 7% NE oxidation current response attenuation, whereas a carbon fiber lost 30% of the original signal [10]. This work was followed up by a more extensive study that used continuous amperometry with a diamond microelectrode and video microscopy to investigate in vitro endogenous NE release simultaneously with the evoked contractile response of a rat mesenteric artery [11]. Using these two techniques along with several drugs, the NE released at sympathetic neuroeffector junctions in the vicinity of the microelectrode was recorded as an oxidation current. NE release was elicited by electrical stimulation (40–70

532

DIAMOND SENSORS FOR NEUROCHEMISTRY

V, 60 pulses, 0.3 ms pulse width) at frequencies between 1 and 60 Hz, with the maximum oxidation current seen at 20 Hz. Detection of NE was accomplished at +800 mV versus Ag/AgCl. However, for assay of neurotransmitter release, confirmation that the oxidation current is due to the transmitter monitored is often supported through the use of several pharmacological agents that can alter the transmission mechanism. Such studies in the investigation of NE release and contractility were conducted using a variety of pharmacological agents. Tetrodotoxin (TTX, 0.3 μM), a voltage-dependent sodium channel antagonist that blocks nerve conduction, abolished both the oxidation current and the arterial constriction. The addition of cocaine (10 μM), an NE reuptake blocker, caused both the oxidation current and the contractile response to increase. These results, combined with the fact that the hydrodynamic voltammetric E1/2 for endogenous NE was identical to that for a standard solution, confirmed that the oxidation current was due to NE and that this compound caused, at least in part, the contractile response. The responses are impressive as both the activity and function show excellent correlations to the applied pharmacological agent. The results also demonstrate that continuous amperometric monitoring of NE with a diamond microelectrode and video imaging of vascular tone allow real-time local measurement of the temporal relationship between nerve-stimulated NE release and arterial constriction, as shown in Figure 20.13. The Swain group (Park et al. [10,11,42]) has gone further to show a major biological finding using the developed technique, where NE-oxidation currents were larger in DOCA-salt hypertensive rats compared to sham mesenteric arteries, but there were no differences between currents recorded from sham and DOCA-salt mesenteric veins. The work by

Before stimulation

(a)

(b)

After stimulation at 10 Hz

Stimulator 210 µm

210 µm

Diamond Microelectrode (c) Stimulation

Imax = 10.4 pA

Current (pA)

Diameter (µm) 30 µm

Figure 20.13 Simultaneous recordings of nerve-stimulated NE release and arterial constriction from rat mesenteric arteries. In (a),the experimental setup is shown with the diamond electrode placed on the mesenteric artery. The stimulator is applied and in (b) the response post stimulation is observed. The trace in (c) shows the temporal dynamics of NE release following constriction of the artery as observed in (b). Video imaging is used to study the amount of constriction of the artery. See color insert.

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

533

Swain’s group is the first major extensive in vitro study of a small molecule neurotransmitter using a diamond microelectrode. Although there have been many other previous studies using diamond-based electrodes for in vitro or in vivo applications, this study has been sustained and has clearly demonstrated to be an area where only diamond electrodes allow for long-term stable recordings. Following the characterization and novel simultaneous monitoring of function and activity, a physiological significant response has also been delivered from studies using diamond electrodes. 20.6.4

Measuring Transmitter Release from the Gastrointestinal Tract

Neurotransmission plays a pivotal role in the gastrointestinal (GI) tract, where it is responsible for influencing motility [156–159]. Alterations in these neurotransmitters have led to digestive disorders such as ulcerative colitis, Crohn’s disease, and irritable bowel syndrome. Various neurotransmitters are present and the GI tract posseses its own nervous system, the ENS. Some of the key neurotransmitters, such as 5-HT, NO, and histamine are released in high concentrations [12,13,125]. Figure 20.14 shows a basic schematic diagram of the neuronal pathways that innervate the majority of all regions of the GI tract and also an image of a cross-section of the ileum. Immunohistrochemical images in Figure 20.14 show staining for serotonin from EC cells located in the villus of the mucosal layer, which is the inner-most layer of the GI tract. The final image shows staining for the enzyme nitric oxide synthase (NOS) in myenteric neurons, which is responsible for the production of NO. Sensor fouling is a major problem for measurements within the GI tract because within this biological area fouling occurs from oxidative by-products of the analytes of

Ileum–cross section

Submucosal plexus neurons EC Cells

Myenteric plexus neurons

Circular Muscle

Longitudinal Muscle

Enterochromaffin cells

Mucosal layer

SERT transporters present in enterocytes

Myenteric plexus

Figure 20.14 Schematic diagram and immunohistochemical images of neurons and cells responsible for function of the ileum. See color insert.

534

DIAMOND SENSORS FOR NEUROCHEMISTRY

interest and also from the tissue. This is because the GI tract is fully lined in mucosal villus, which is a sticky protective layer to the tissue and allows for high surface area for adsorption. There are also goblet cells located in the villus that generate and release mucus into the lumen. This biological application is one of a few where both forms of fouling occur. Thus, the BDD microelectrode at present is the only electrode that is able to conduct stable measurements of neurotransmitters in all areas of the GI tract. 20.6.4.1 Detection of Histamine Release from Enterochromaffin-Like Cells Located in the Stomach Histamine is located in enterochromaffin-like (ECL) cells and is responsible for activating parietal cells within the stomach wall to release gastric acid, as shown in Figure 20.15a. Limited investigations, mainly using microdialysis with a separation technique, have been used to investigate the release of histamine from the stomach, but these methods have very limited temporal resolution [160,161]. Histamine also possesses a problem for electrochemical detection, as oxidative by-products adsorb to the electrode surface fouling the sensor; thus, electrochemical investigations using sensors have been limited. The biological matrix within the stomach is a complex one as well, as pH levels can be as low as pH 2, but there is also a local buffering system at the mucosal surface that uses carbonate to control pH levels. These alterations in local pH levels can have a major influence on the electroanalytical measurement, as oxidation/reduction potentials are known to shift with pH [66,162,163]. The Swain group (Park et al. [10,11,42]) has shown stable responses for the detection of norepinephrine at various pH-buffered solutions from 3.0 to 7.2 on a BDD microelectrode [42]. This stability is critical for accurate interpretations of histamine release as shifts in the oxidation peak will result in underestimations of the current. Figure 20.15b shows experimental traces of histamine released from tissue. Reproducible responses were observed for multiple measurements over various sites of the

(a)

Gastric Pit

(b) 8

Gastric glands Parietal cell

Current / nA

6

4

2 ECL-cell 0 0

100

200

300

400

500

600

Time (s) Figure 20.15 Measurement of histamine release from the stomach. In (a) a schematic diagram of the localization of the ECL cell and the parietal cell in a gastric pit located in the mucosal wall of the stomach and in (b) response of histamine release detected on the BDD microelectrode at potential of +1.2 V versus Ag| AgCl. The grey boxes indicate when the electrode is placed 0.5 mm over the tissue surface. See color insert.

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

535

mucosal surface, while pH values varied from 4 to 6. This response shows that in the presence of varying pH levels, which have been observed using a pH electrode, the responses of histamine and, more importantly, the baseline do not change during the course of the measurement. Subsequently, levels of histamine were measured in tandem with pH levels using an iridium oxide pH electrode to study the influence of various pharmacological agents [125]. On paper, these measurements of histamine from the stomach look no different from any recordings of biogenic amine from biological tissue. But the stomach is a very difficult biological matrix and the ability to conduct stable recordings is important to study the transmission mechanism. The BDD microelectrodes’ robust response in an environment with a low and constantly changing pH provides an important niche of the electrode. These responses begin to support the claim that BDD electrodes are suitable for complicated biological matrices. This cannot be said for other electrodes, where potential shifts are regularly observed and stability of the sensor is lost over time [66,162]. Due to mechanical strength of diamond materials and the inert nature of these materials, these sensors also have the ability for conducting measurements in stomach tissue for a long period of time. 20.6.4.2 Monitoring Serotonin Release from Enterochromaffin Cells Located in the Mucosa Serotonin (5-HT) plays an important role during chemical transmission in the GI tract [156,164]. In the GI tract, 5-HT is released from neurons present in the myenteric plexus and EC cells located in the mucosa. Ninety percent of the body’s 5-HT content is found in these EC cells, which act as sensory transducers that respond to mechanical or chemical stimulation [165]. 5-HT transmission in the gut contributes to the control of gastrointestinal motility and aids in processing the passage of food along the length of the gut. 5-HT is released from vesicles in a Ca2+ dependent manner, and thus is similar to the exocytosis process observed in the brain [166,167]. The released 5-HT is cleared from the lumen by the serotonin transporter (SERT) present in enterocytes that completely surround the EC cells. Changes in the extracellular levels of 5-HT have been implicated to contribute to motility disturbances and have led to the onset of gastrointestinal disorders such as irritable bowel syndrome (IBS) [168–170]. Boron-doped diamond microelectrodes have been successfully used to measure 5HT levels present in the micromolar range. At these levels, carbon fiber sensors are widely prone to fouling, as shown in Figure 20.16 from tissue recordings in biological buffer. In Figure 20.16a, the BDD electrode was poised ∼1 mm over the tissue, which is referred to as the “over tissue current” (OTC ). This OTC occurs as shearing forces are applied to the villus due to the fluid flow causing constant basal stimulation. The tissue is then additionally stimulated using a glass capillary as a means of mechanical stimulation, which is referred to as the ‘touching tissue current’ (TTC ). On the carbon fiber microelectrode the OTC increases but does not stabilize like the response on the BDD electrode. After touching the mucosa, there is a slight plateau during the TTC , but during this phase the current is already beginning to decline, due to the fouling of the carbon fiber electrode and subsequently no further measurements can be obtained. The BDD microelectrode allows for reproducible measurements of 5-HT from isolated in vitro ileum tissue as seen in Figure 20.16, and is indicated by the stable baseline observed in-between each measurement (Figure 20.16b). Based on these measurements 5-HT levels observed released from the mucosa of the ileum are within 1–2 μM.

536

DIAMOND SENSORS FOR NEUROCHEMISTRY

(a) 1.6

(b) 2.0 Carbon fibre BDD

1.4

1.5 Current (nA)

Current (nA)

1.2 1.0 0.8 0.6 0.4

1.0

0.5

0.2 0.0 0

40

120 80 Time (s)

160

200

0.0 0

200

600 400 Time (s)

800

1000

Figure 20.16 Measurements of 5-HT release from EC cells present in ileum mucosa. Where (a) shows the comparison between BDD and carbon fiber microelectrode and (b) shows repeat measurements on the BDD microelectrode. (Reprinted from Ref. 12.)

Multiple methods exist to stimulate tissue, and when the ileum is stimulated by mechanical force, a current response of 201±4.3 pA (n = 34) was observed, but when electrical stimulation was used, the current response increased to 537±2 pA (n = 21). Although an increase in the current is observed using an electrical stimulation, the variation is lower in comparison to the application of mechanical force. For biological experiments, mechanical stimulation is used even though electrical stimulation is more reproducible as it mimics the natural activity of the tissue. It is the pressure from solid particles of food in the intestine that stimulate 5-HT release, which influences intestinal motility. Measurements using the BDD microelectrode were conducted from various regions of the GI tract to see how the 5-HT release mechanism was altered. In Figure 20.17, responses of 5-HT overflow from duodenum, ileum, and colon tissue are shown. The OTC levels and the levels observed during mechanical stimulation (TTC ) are represented. As clearly demonstrated, the concentration of 5-HT decreases the further you go down the GI tract. This is the case for both the OTC and TTC responses. No significant differences are observed for TTC responses between ileum and colon tissue, due to the high variation in the colon tissue. This is most likely due to varying activity in sections of the tissue over the colon that regulates specific motility patterns. Decreases in 5-HT levels in the colon are also expected as this neurotransmitter plays a key role in motility and lower levels in the colon reflect of slower transit, which allows for maximum water adsorption and waste removal, whereas in the duodenum, rapid varying motility patterns induces further digestion from food leaving the stomach. The BDD microelectrode has proved to be a vital tool to study 5-HT overflow from EC cells. As indicated earlier, carbon fiber sensors are prone to fouling from oxidative by-products of 5-HT at concentrations greater than 1 μM. In all areas of the GI tract, the levels of 5-HT are greater than this concentration and therefore prevent the use of carbon or any noble metal electrode to conduct such measurements. The GI tract is also a harsh biological environment for measurement, but this is one feature that diamond sensors thrive on and thus provide the ability to investigate neurochemical matrixes that have been out of reach at present for biological measurements.

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

537

[Serotonin] (µM)

(b) 5

(a) Duodenum Ileum Colon

4 3 2 1 0 Duodenum

500 pA 1 µM

Ileum

Colon

Duodenum Ileum

Colon

(c) 1.4

20 s

[Serotonin] (µM)

1.2 1.0 0.8 0.6 0.4 0.2 0.0

Figure 20.17 5-HT release from various regions of the GI tract. In (a) responses obtained on the BDD microelectrode are shown where the shaded area indicates the TTC phase. In (b) 5-HT levels during OTC and in (C) responses during TTC phase are shown.

20.6.4.3 Monitoring Nitric Oxide Release from Myenteric Plexus Neurons In the GI tract, NO is an inhibitory nonadrenergic and noncholinergic neurotransmitter to the muscle layers where it causes relaxation [171,172]. It is synthesised by the enzyme nitric oxide synthase (NOS), which is located in neurons present within the myenteric plexus. Figure 20.18 shows immunohistochemical staining for these NOS neurons in the myenteric plexus. Detection of NO release using electrochemical methods has been widely carried out using carbon fiber or platinum microelectrodes from single cells to in vitro tissue slices [37,39,41,71,173]. Electrochemical methods for the detection of NO has provided major breakthroughs on understanding the role of this gaseous signaling molecule, as sensors provide a real-time direct means of measuring NO from the site of release, which other techniques fail to offer. If NO is not detected directly then byproducts such as nitrite can be detected as a marker of NO activity. BDD microelectrodes have successfully been used for reproducible detection of nitrite in comparison to glassy carbon electrodes [174]. Due to the difficulties in measuring NO, many experimental controls are needed to confirm selective detection of NO release. Figure 20.18 shows measurements of NO release stimulated using nicotine, where a variety of experimental controls (Figure 20.18a) and biological controls were carried out to confirm the detection of NO (Figure 20.18b). Based on a differential pulse voltammogram (DPV) of a pure NO solution, the oxidation potential of NO on the BDD electrode was +1 V versus Ag/AgCl. When a lower potential

538

DIAMOND SENSORS FOR NEUROCHEMISTRY

1000 mV vs Ag|AgCl Over longitudinal muscle tissue Response without tissue 750 mV vs Ag|AgCl 100 pA

Krebs buffer + 100 nM TTX + 100 µM L-NNA + 0.05 mg/ml myoglobin

50 s

100 pA 50 s

Figure 20.18 No release from myenteric neurons using a BDD microelectrode. Experimental and biological controls are shown, where the grey boxes indicated the duration of nicotine perfusion. (Reprinted from Ref. 14.)

is applied, no interference of other neurotransmitters were observed, as they may be below the limit of detection. When measurements are made over muscle areas rather than neurons, no responses are observed, indicating that NO release observed is solely from the myenteric neurons. The biological controls also confirm the detection of NO, as in the oxidation current elicited by nicotine is abolished in the presence of the NOS antagonist, L-NNA. In addition, TTX (100 nM) nearly abolished the current, indicating that NO is released almost exclusively by neurons via an action potential-dependent process. In the presence of the NO radical scavenger, myoglobin, the current elicited by nicotine, is completely abolished. The BDD electrode was able to clearly detect NO, and there have been other investigations that have used diamond electrodes for the detection of NO and/or nitrite [174]. NO is an important signaling molecule and is known to play a key role in a variety of important biological areas. NO can be easily detected from the myenteric plexus, which is a relatively simple biological environment, but NO plays a key role in the cardiovascular system and is known to influence the process of angiogenesis. Long-term detection to NO from arteries and veins to observe differences in the flux of NO during normal and diseased tissue can provide information on variety of cardiovascular conditions. The blood is one of the most difficult biological environments to measure from, as the sensor can easily be fouled from proteins that are activated and influence other molecules to rapidly cover the sensor, limiting the ability to conduct precise measurements. Diamond electrodes have shown the ability to measure signaling molecules in the presence of blood proteins for a long durations and thus can provide the ideal means of detecting NO in the cardiovascular system to enhance our knowledge. 20.6.5

Studying the Neurotransmitter Clearance Process

The reuptake process is responsible for the clearance of the neurotransmitters from the extracellular fluid following release from the presynaptic cell through transporters. It is an essential part of the neurotransmission mechanism. A spectrum of disorders has been attributed to either a loss or alteration of the reuptake process. These include depression, primary pulmonary hypertension, anorexia, and irritable bowel syndrome [156,175–177]. Understanding and monitoring the alterations in the reuptake process of a variety of classical neurotransmitters is of importance. There have been studies conducted in vivo [52,53]

20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

539

and also in isolated cells [54]; however, it is the latter area where BDD microelectrodes have been applied. Perez and Andrews [55] and Perez, Bianco, and Andres [56] have previously indicated that electrochemical methods are the most suitable means to study reuptake of neurotransmitters. This is because the conventionally used radiochemical assays require a filtration process for sample preparation, which leads to substantial loss in the amount of transported neurotransmitter, leading to lower uptake rates [55,56]. For the electrochemical measurements carried out to date, chronoamperometry has been used with carbon fiber microelectrodes. During experimental assays to study reuptake function, a fixed addition of the neurotransmitter of interest is added to the buffer containing the cells and the clearance of this transmitter is monitored over time. The exponential decay that is observed due to the reuptake of the neurotransmitter can be monitored to obtain the activity of the uptake transporter [55]. This approach has been successful, but is limited because the carbon fiber electrodes suffered from fouling at concentrations greater than 1 μM of the neurotransmitter. To gain any suitable kinetic information on the activity of the reuptake transporter, concentrations up to 10 μM need to be studied. As uptake decays in a similar fashion to fouling, a stable electrode is required for accurate measurements. BDD microelectrodes have recently been investigated to study their stability over this wider concentration range. Measurements of reuptake have been conducted from synaptosomes isolated from the brain and also from peripheral lymphocytes that are present in the blood. 20.6.5.1 Measurements of Multiple Transmitters from Brain Synaptosomes Synaptosomes are a type of neuronal liposome that are derived from neurons and function as a small anucleate cell that possesses active ion and neurotransmitter transport systems across the plasma membrane. Synaptosomes are an attractive model with which to study the reuptake, as this process can be studied independently of neurotransmitter release and diffusion, which have been shown to contribute to changes in neurotransmitter concentrations detected at microelectrodes during in vivo recordings. Under these experimental conditions, the chemical identity and concentration of the neurotransmitter can be closely controlled and studied while conducting electrochemical measurements. As a result, data generated from chronoamperometry studies with synaptosomes have led to a straightforward determination of changes in the kinetics of uptake as they relate to reductions in either the affinity or density of transporters [55,56]. Figure 20.19 shows the response obtained from a BBD microelectrode usd to measure reuptake from synaptosomes isolated from the frontal cortex. In these experiments, 5 μM of the neurotransmitter is added and the clearance of the transmitter is monitored, which can relate to the activity of the transporter. As clearly shown, 5-HT and NE are the only transmitters cleared from the media and dopamine levels fail to change. This is due to the fact that at this region of the brain only has functional transporters for 5-HT and NE. These results provide a good indication of the function and activity of this brain area and can also be used for investigating the role of pharmacological agents. The application of higher concentrations can provide information on the kinetics of the transporter as the slope from Figure 20.19 can be analyzed. The stability and reproducibly shown by the diamond electrode was not feasible on carbon fiber microelectrode at the same concentration, and thus limited the ability of gaining complete kinetic understanding of the rate of reuptake. Due to the stability of BDD microelectrodes over a wide concentration range, the ability to conduct measurements

540

DIAMOND SENSORS FOR NEUROCHEMISTRY

(b) [Dopamine] (μM)

[Serotonin] (μM)

5 4 3 2 1 0

(c) [Noreadrenaline] (μM)

(a)

5 4 3 2 1 0

0 200 400 600 800 1000 Time (s)

0 200 400 600 800 1000 Time (s)

5 4 3 2 1 0 0 200 400 600 800 1000 Time (s)

Figure 20.19 Investigation of reuptake of transmitters from synaptosomes isolated from the frontal cortex. In (a), (b), and (c) response for 5-HT, dopamine, and NE are shown respectively, where additions of the transmitter were added to make final solution concentration of 5 μM.

over a wider concentration range is plausible. The robust nature of diamond microelectrodes also provides a means of conducting high throughput assays of reuptake, which is of biological and clinical significance. The other benefits of reuptake studies is that they are not limited to only synaptosomes, but may include any cell type that contains a functional transporter. 20.6.5.2 Investigation of Serotonine Clearance by Transporters Present on Lymphocytes Due to advantages of BDD microelectrodes for studies of reuptake from synaptosomes, other cell types have also been used for investigation. Lymphocytes are a type of white blood cell, but most importantly they contain serotonin transporters and thus serve as a good cell type to study reuptake function. The major benefits of studying uptake from lymphocytes compared to synaptosomes is that they can be accessed with more ease and can also be obtained from humans, thus allowing for a means of conducting clinical measurements and screenings of uptake function. The majority of studies on reuptake function have been conducted to investigate the use of pharmacological agents; however, there is also a number of commonly occurring noncoding SERT gene polymorphisms that have been identified to influence reuptake function. One polymorphism, termed the human serotonin transporter gene-linked polymorphic region (h5-HTTLPR), is the most extensively studied SERT gene variant and is shown to influence anxiety-related personality traits, vulnerability to developing neuropsychiatric disorders, and variability in drug responsiveness [178–181]. There is, however, contradictions in the literature on which allele influence differences in reuptake [182,183]. In Figure 20.20, reductions in serotonin uptake rates are observed from lymphocytes with one or two copies of the short allele of the rh5-HTTLPR (s/s<s/l
20.6 APPLICATIONS OF DIAMOND SENSORS AND DEVICES IN NEUROCHEMISTRY

541

Figure 20.20 5-HT reuptake from blood lymphocytes. Serotonin uptake is decreased with the short-allele of the rhesus 5-HTTLPR in the peripheral blood lymphocytes. See color insert.

sensors. The application of diamond electrodes to study pharmacological and genetic alterations of reuptake is one example of where the electrode is showing signs from being used from a laboratory to a clinical setting for diagnosis. 20.6.6 In vitro and In vivo Measurements from the Central Nervous System Although the majority of the examples described thus far have shown the use of diamond sensors in the periphery, which is truly where the diamond electrode has proved to be one step ahead of other electrode materials, there has been some investigations in the central nervous system. The major limitation to date for central nervous system measurements is the insulation and fabrication of diamond microelectrodes. Although diamond electrodes have been made to less than 10 μm in diameter and in some cases down to nanometer size, the electrical connection and insulation of such devices are far behind the fabrication process of carbon fiber microelectrodes. At present, producing a batch of electrodes with the same sensing surface area for in vivo use is difficult to obtain, and once this limitation is overcome, in vivo measurements will be routinely carried out. Carbon fiber microelectrodes also have the edge due to the ease of fabrication at a mass. The major advantage of diamond electrodes for CNS measurements is that they can be used for conducting long-term recordings from the brain and also be coupled to electrical activity measurements. 20.6.6.1 In vitro Measurements A wide amount of in vitro work has been carried out for detection of neurotransmitters from the sympathetic and enteric nervous systems, but limited in vitro measurements of transmitters from the CNS have been conducted. 5-HT release was observed from the CNS of the pond snail, A. Californica, where the BDD microelectrode was easily placed and thus allowed a means to study the role of 5-HT. Real-time detection of 5-HT release was observed following electrical stimulation and responses were obtaining using FSCV and amperometry. These measurements not

542

DIAMOND SENSORS FOR NEUROCHEMISTRY

only allow for a means of measuring the dynamics of neurotransmitter release from the CNS but they also provide the means to studying the role of the transmitter on tissue. As in this case, the modulation of 5-HT on feeding behavior was well demonstrated [149]. There has also been an investigation related to the role of adenosine release from brain slices. A limited number of studies has focused on the dynamics of adenosine release, but this molecule not only serves as a neuromodulator but is also known as metabolite produced from adenosine triphosphate (ATP), which is key marker of metabolism. FSCV using carbon fiber microelectrodes have been previously used for the detection of adenosine, where the limit of detection of the electrode was ∼10 μM. The 25 μm BDD microelectrode was able to detect adenosine to a better standard than the carbon fiber microelectrode, where the limit of detection was 20 nM and the signal-to-noise ratio for the sensor was 20. Measurements made in rodent medullary slice preparations were able to show transient changes in adenosine when stimulated using A2A receptor agonist CGS 21680 [184]. This study shows two real advances of the diamond electrode: (1) the ability to carry out direct detection of adenosine rather than using enzyme biosensors that can provide comparable limits of detection and (ii) the ability of the diamond electrode to clearly detect small concentration changes in transmitter releases from the central nervous system. There is no doubt that the diamond microelectrodes have the ability to detect transmitters from the CNS, as demonstrated from the studies previously mentioned. The lower capacitive current on diamond electrodes helps enhance the signal-to-noise ratio and improves the limit of detection, which help detect such lower concentrations that are observed in the CNS. The major limitation is the electrode size, where, for measurements in specific regions of the brain, sensors smaller than 7 μm are required for CNS measurements in rodents. 20.6.6.2 In vivo Measurements from Anesthetized Animals An in vivo measurement from the CNS or practically in any area of the body is a major enhancement on our knowledge of a biological system, as natural activity can be measured. To date, there have been very limited in vivo studies that have used diamond microelectrodes—however, one good study does exist. Conical electrodes with an exposed length of 250 μm and a tip diameter of 5 μm was used for measurements of dopamine. Measurements of dopamine were carried out in the corpus striatum following stimulation in the medial forebrain bundle from an anesthetized mouse. Dopamine was detected selectively from ascorbic acid with the application of an anodically oxidized BDD electrode. The BDD microelectrode was shown to have a better analytical performance than the carbon fiber microelectrodes for in vivo measurements, as the signal-to-noise ratio was two times better [185]. Better stability was also observed for the BDD microelectrode when measurements were conducted for over 10 minutes, which is an order of magnitude greater than that currently carried out on carbon fiber microelectrodes. Increases in dopamine were observed as expected, during the administration of the dopamine reuptake inhibitor nomifensine. These measurements are encouraging and support the benefits from in vitro work, where the lower capacitive current of the diamond sensors enhances the ability of conducting measurements and thus improves signal-to-noise ratios. A stable electrode is essential for measurements in the brain, as the capacitive current that is generated from the ions within extracellular fluid are constantly evolving. Thus, electrodes like the carbon fiber microelectrodes that have been shown to be extremely sensitive to ionic changes

20.7 CONCLUSIONS AND OUTLOOK FOR THE FUTURE

543

within the media [65,73,186] will provide higher errors during biological measurements. This sole point provides the major advantage for using diamond electrodes for accurate and precise measurements of neurotransmitters in vitro and in vivo in the CNS. However, the diamond electrodes used for all CNS studies described are at least an order of magnitude larger in area than currently used carbon fiber microelectrodes. This limitation prevents neuroscientists from gaining the all-important spatial resolution required to understand specific changes in regions of the brain.

20.7

CONCLUSIONS AND OUTLOOK FOR THE FUTURE

Without doubt, diamond electrodes have made their own niche in the wide field of neurochemical monitoring and measurements. Most importantly, the diamond electrode has enabled the ability to conduct measurements in biological systems where other sensors over the past 40 years have failed. The low capacitive currents, biocompatibility, reduced fouling from proteins and oxidative by-products, and mechanical robust nature of diamond electrodes are the key features that have provided the platform for reliable reproducible measurements. This has led to robust stable measurements of transmission in the peripheral nervous system and thus measurements have been made in the blood as well as in the sympathetic and enteric nervous systems. Although 90% of all the work using diamond sensors have been conducted in the periphery, there have been investigations in the CNS, and these studies will continue to grow as diamond-sensing technology improves. Previous studies and emerging studies using diamond electrodes have also allowed the means to study real-time changes in neurotransmitters and thus provide a novel alternative to currently used imaging and biochemical assays used to study various pathophysiological conditions. There are many promising outlooks for the future of diamond sensors in neurochemistry. Initially, more important biological data will be generated from the application of diamond electrodes in biology and thus will promote the features of the material that are superior to other sensors. BDD microelectrodes so far have been already considered to be the electrode material of choice for any nature of 5-HT measurements [34,187]. New biological areas are also expected to be investigated, such as the kidneys and in urine for transmitters and pharmacological agents, as well as within the heart and lungs to study classical and gaseous signaling molecules. More measurements in neurochemical environments will also provide greater exposure to diamond sensors and will allow biologists, neurochemists, and medics to employ such materials. One area that can prove to be of great promise is for the direct detection of neurochemicals that require the application of enzyme biosensors to mediate the electrochemical reaction at the electrode. Studies have shown the ability to conduct direct detection of adenosine [184] and glucose [188] on BDD electrodes. Other markers such as acetylcholine and adenosine triphosphate may also be detected. Measurements of these neurochemical markers are important, for they play key roles in metabolism throughout the body and these key markers are known to change during the onset of ischemia [189–191]. As the amount of applications increase for the use of diamond sensors, the electrode fabrication will improve and techniques will be enhanced. Reproducible small (∼10 μm diameter) microelectrodes are required for in vivo measurements and enhanced spatial resolution. At present, due to insulation issues, no disc-type electrodes have been fabricated and thus currently in vitro and in vivo measurements of transmitter are not limited

544

DIAMOND SENSORS FOR NEUROCHEMISTRY

to the single cell but to the uniformity in transmission from a tissue section or region. There have, however, been microelectrodes made in the range of 1–25 μm for measurements using scanning electrochemical microscopy (SECM) [192]. Diamond sensors used in tandem with SECM will be an attractive tool for spatial mappings of brain slices and/or neurons and will be a major step forward in our understanding of various biological tissues. Currently, individual microelectrode arrays have been identified as ideal platforms for biological and neurological measurements. These devices are seen as a one-stop shop for all techniques to be carried out to gain full understanding of biological processes. For example, fluorescence imaging, optical microscopy, electrophysiology, and microdialysis can all be used together with electrochemical measurements from a microelectrode array to gain understanding of cellular mechanics. Although there have been limited developments in the fabrication of diamond-based microelectrode arrays [153,154,193], more will be developed and used for biological measurements in coming years.

20.8

ACKNOWLEDGMENTS

I thank EPSRC LSI grant for financial support for this research program. I am especially grateful for Professor Greg Swain for introducing me to diamond electrodes and assisting by all possible means to support my research. Also, I give a wholehearted acknowledgment of the support and contribution from students and fellow colleagues in joint publications and research programs, without whom none of this work would be plausible. My special thanks to Professor Anne Andrews and Yogesh Singh for their study of serotonin reuptake from lymphocytes, and to Raphael Trouillon for his protein fouling studies.

REFERENCES 1. 2. 3. 4. 5. 6.

7. 8. 9. 10. 11. 12.

R.N. Adams, Anal. Chem. 1976, 48 , 1126A–1138A. P.T. Kissinger, J.B. Hart, R.N. Adams, Brain Res. 1973, 55 , 209–213. R.M. Wightman, E. Strope, P.M. Plotsky, R.N. Adams, Nature 1976, 262 , 145–146. C.A. Marsden, M.P. Brazell, N.T. Maidment, in Measurement of Neurotransmitter Release in vivo (C.A. Marsden, ed.), John Wiley & Sons, Chichester, 1984, p. 127. J.A. Stamford, Z.L. Kruk, J. Millar, R.M. Wightman, Neurosci. Lett . 1984, 51 , 133–138. R.M. Wightman, D.S. Brown, W.G. Kuhr, R.L. Wilson, in Voltammetry in the Neurosciences: Principles, Methods & Application (J.J.B. Justice, N.J. Clifton, eds.), Humana Press, Clifton, 1987, p. 103. B.J. Venton, D.J. Michael, R.M. Wightman, J. Neurochem. 2003, 84 , 373–381. M.L.A. V. Heien, A. Khan, J.L. Ariansen, J.F. Cheer, P.E. M. Phillips, K.M. Wassum, R.M. Wightman, PNAS 2005, 102 , 10023–10028. D.L. Robinson, R.M. Wightman, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 17. J. Park, J.J. Galligan, G.D. Fink, G.M. Swain, Anal. Chem. 2006, 78 , 6756–6764. J. Park, J.J. Galligan, G.D. Fink, G.M. Swain, J. Physiol . 2007, 584 , 819–834. B.A. Patel, X. Bian, V. Quaiserova-Mocko, J.J. Galligan, G.M. Swain, Analyst 2007, 132, 41–47.

REFERENCES

545

13. B.A. Patel, Analyst 2008, 133 , 516–524. 14. B.A. Patel, J.J. Galligan, G. Swain, X. Bian, Neurogastroenterol. Motil . 2008, 20 , 1243–1250. 15. F. Tian, A.V. Gourine, R.T.R. Huckstepp, N. Dale, Anal. Chim. Acta, 2009, 645 , 86–91. 16. E. Llaudet, S. Hatz, M. Droniou, N. Dale, Anal. Chem. 2005, 77 , 3267–3273. 17. J.-F. Masson, C. Kranz, B. Mizaikoff, E.B. Gauda, Anal. Chem. 2008, 80 , 3991–3998. 18. B.N. Barsoum, W.M. Watson, I.M. Mahdi, E. Khalid, J. Electroanal. Chem. 2004, 567 , 277–281. 19. K.N. Hascup, E.C. Rutherford, J.E. Quintero, K. Day, J.R. Nichell, F. Pomerleau, P. Huettl, J.J. Burmeister, G.A. Gerhardt, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 407. 20. J. Castillo, S. Isik, A. Blochl, N. Pereira-Rodrigues, F. Bedioui, E. Csoregi, W. Schuhmann, J. Oni, Biosensors Bioelectronics 2005, 20 , 1559–1565. 21. J.A. Stamford, Brain Res. Rev . 1985, 10 , 119–135. 22. P.E.M. Phillips, R.M. Wightman, TrAC Trends Anal. Chem. 2003, 22 , 509–514. 23. K.P. Troyer, M.L.A. V. Heien, B.J. Venton, R.M. Wightman, Curr. Opin. Chem. Biol . 2002, 6 , 696–703. 24. S.E. Hochstetler, R.M. Wightman, Biophysics Textbook Online, 2000. 25. B.P. Jackson, S.M. Dietz, R.M. Wightman, Anal. Chem. 1995, 67 , 1115–1121. 26. D.L. Robinson, B.J. Venton, M.L.A.V. Heien, R.M. Wightman, Clin. Chem. 2003, 49 , 1763–1773. 27. R.M. Wightman, J. Jankowski, R. Kennedy, K. Kawagoe, T. Schroeder, D. Leszczyszyn, J. Near, E. Diliberto, Jr, O. Viveros, PNAS , 1991, 88 , 10754–10758. 28. A.G. Ewing, D.M. Cannon Jr., N. Winograd, Annu. Rev. Biophys. Biomol. Struct . 2000, 29 , 239–263. 29. L.A. Sombers, A.G. Ewing, in Electroanalytical Methods for Biological Materials (A. BrajterToth, J. Q. Chambers, eEds.), Marcel Dekker, New York, 2000, p. 279. 30. J.A. Stamford, J. Neurosci. Meth. 1990, 34 , 67–72. 31. J.A. Stamford, J.J.B. Justice, Anal. Chem. 1996, 68 , 359A–366A. 32. C.A. Marsden, M.H. Joseph, Z.L. Kruk, N.T. Maidment, R.D. O’Neill, J.O. Schenk, J.A. Stamford, Neuroscience 1988, 25 , 389–400. 33. D.J. Michael, R.M. Wightman, J. Pharmaceut. Biomed. Anal . 1999, 19 , 33–46. 34. K.L. Adams, M. Puchades, A.G. Ewing, Annu. Rev. Anal. Chem. 2008, 1 , 329–355. 35. R.A. Clark, S.E. Zerby, A.G. Ewing, in Electroanalytical Chemistry, Vol. 20 (A. Bard, I. Rubinstein, eds.), Marcel Dekker, New York, 1998, p. 227. 36. G.Y. Chen, A.G. Ewing, Crit. Rev. Neurobiol . 1997, 11 , 59–90. 37. C. Amatore, S. Arbault, Y. Bouret, B. Cauli, M. Guille, A. Rancillac, J. Rossier, Chem. Phys. Chem. 2006, 7 , 181–187. 38. C. Amatore, S. Arbault, I. Bonifas, Y. Bouret, M. Erard, M. Guille, Chem. Phys. Chem. 2003, 4 , 147–154. 39. F. Bedioui, N. Villeneuve, Electroanalysis 2003, 15 , 5–18. 40. K. Shibuki, Neurosci. Res. 1990, 9 , 69–76. 41. B.A. Patel, M. Arundell, K.H. Parker, M.S. Yeoman, D. O’Hare, Anal. Chem. 2006, 78 , 7643–7648. 42. J. Park, Y. Show, V. Quaiserova, J.J. Galligan, G.D. Fink, G.M. Swain, J. Electroanal. Chem. 2005, 583 , 56–68.

546

DIAMOND SENSORS FOR NEUROCHEMISTRY

43. C.E. John, S.R. Jones, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 35. 44. B.E.K. Swamy, B.J. Venton, Analyst 2007, 132 , 876–884. 45. D.J. Michael, J.D. Joseph, M.R. Kilpatrick, E. Travis, R.M. Wightman, Anal. Chem. 1999, 71 , 3941–3947. 46. M.F. Roitman, G.D. Stuber, P.E.M. Phillips, R.M. Wightman, R.M. Carelli, J. Neurosci . 2004, 24 , 1265–1271. 47. G.D. Stuber, R.M. Wightman, R.M. Carelli, Neuron 2005, 46 , 661–669. 48. M.L.A.V. Heien, M.A. Johnson, R.M. Wightman, Anal. Chem. 2004, 76 , 5697–5704. 49. F.-M. Zhou, Y. Liang, R. Salas, L. Zhang, M. De Biasi, J.A. Dani, Neuron 2005, 46 , 65–74. 50. J.O. Schenk, R.N. Adams, in Measurement of Neurotransmitter Release in vivo (C.A. Marsden, ed.), John Wiley & Sons, Chichester, 1984, p. 193. 51. J.-F. Pujol, M. Jouvet, F. Gonon, R. Cespuglio, J.-L. Ponchon, Anal. Chem. 1979, 51 , 1483–1486. 52. L.C. Daws, G.M. Toney, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 63. 53. J.M. Gulley, G.A. Larson, N.R. Zahniser, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 83. 54. X.A. Perez, A.J. Bressler, A.M. Andrews, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 103. 55. X.A. Perez, A. Andrews, Anal. Chem. 2005, 77 , 818–826. 56. X. Perez, L.E. Bianco, A.M. Andrews, J. Neurosci. Meth. 2006, 154 , 245–255. 57. J.M. Liebetrau, H.M. Miller, J.E. Baur, S.A. Takacs, V. Anupunpisit, P.A. Garris, D. O. Wipf, Anal. Chem. 2003, 75 , 563–571. 58. L. Stoica, S. Neugebauer, W. Schuhmann, in Biosensing for the 21st Century (R. Renneberg, F. Reinhard, eds.), Springer, Heidelburg, 2008, p. 455. 59. A. Schulte, W. Schuhmann, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 353. 60. R.S. Kelly, R.M. Wightman, Anal. Chim. Acta 1986, 187 , 79–87. 61. K.F. Martin, C.A. Marsden, F. Crespi, TrAC Trends Anal. Chem. 1988, 7 , 334–339. 62. A. Schulte, R.H. Chow, Anal. Chem. 1998, 70 , 985–990. 63. J.O. Besenhard, A. Schulte, K. Schur, P.D. Jannakoudakis, in Microelectrodes: Theory and Applications (M.I. Montenegro, M.A. Queiros, J.L. Daschbach, eds.), Kluwer Academic, Dordrecht, 1991, p. 189. 64. J. Millar, C.W.A. Pelling, J. Neurosci. Meth. 2001, 110 , 1–8. 65. A.J. Gotch, R.S. Kelly, T. Kuwana, J. Phys. Chem. B 2003, 107 , 935–941. 66. P.L. Runnels, J.D. Joseph, M.J. Logman, R.M. Wightman, Anal. Chem. 1999, 71 , 2782–2789. 67. P.M. Kovach, A.G. Ewing, R.L. Wilson, R.M. Wightman, J. Neurosci. Meth. 1984, 10 , 215–227. 68. D. Bruns, Methods 2004, 33 , 312–321. 69. B.J. Venton, K.P. Troyer, R.M. Wightman, Anal. Chem. 2002, 74, 539–546. 70. M.J. Logman, E.A. Budygin, R.R. Gainetdinov, R.M. Wightman, J. Neurosci. Meth. 2000, 95 , 95–102. 71. N.R. Ferreira, A. Ledo, J.G. Frade, G.A. Gerhardt, J. Laranjinha, R.M. Barbosa, Anal. Chim. Acta 2005, 535 , 1–7.

REFERENCES

547

72. B. Zhang, K.L. Adams, S.J. Luber, D.J. Eves, M.L. Heien, A.G. Ewing, Anal. Chem. 2008, 80 , 1394–1400. 73. B.T. Chen, M.E. Rice, Electroanalysis 1999, 11 , 344–348. 74. S. Marinesco, T.J. Carew, J. Neurosci. Meth. 2002, 117 , 87–97. 75. A.G. Ewing, M.A. Dayton, R.M. Wightman, Anal. Chem. 1981, 53 , 1842–1847. 76. C. Amatore, S. Arbault, D. Bruce, P. d’Oliveira, M. Erard, M. Vuillaume, Faraday Discussions 2000, 116 , 319–333. 77. K. Pihel, Q.D. Walker, R.M. Wightman, Anal. Chem. 1996, 68 , 2084–2089. 78. B.D. Bath, D.J. Micheal, B.J. Trafton, J.D. Joseph, P.L. Runnels, R.M. Wightman, Anal. Chem. 2000, 72 , 5994–6002. 79. A.G. Ewing, T.G. Strein, Y.Y. Lau, Accounts Chem. Res. 1992, 25 , 440–447. 80. R.G.W. Staal, S. Rayport, D. Sulzer, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 337. 81. M.A. Bunin, R.M. Wightman, J. Neurosci . 1998, 18 , 4854–4860. 82. J.J. O’Connor, Z.L. Kruk, J. Neurosci. Meth. 1991, 38 , 25–33. 83. J.A. Stamford, P. Palij, C. Davidson, S.J. Trout, Bioelectrochem. Bioenergetics 1995, 38 , 289–296. 84. J.D. Joseph, R.M. Wightman, in Electroanalytical Methods for Biological Materials (A. Brajter-Toth, J.Q. Chambers, eds.), Marcel Dekker, New York, 2002, p. 255. 85. C.E. John, S.R. Jones, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 35. 86. K.A. Mauritz, R.B. Moore, Chem. Rev . 2004, 104 , 4535–4586. 87. E.W. Kristensen, W.G. Kuhr, R.M. Wightman, Anal. Chem. 1987, 59 , 1752–1757. 88. F.O. Brown, J.P. Lowry, Analyst 2003, 128 , 700–705. 89. G.A. Gerhardt, A.F. Oke, G. Nagy, B. Moghaddam, R.N. Adams, Brain Res. 1984, 290 , 390–395. 90. G. Nagy, G.A. Gerhardt, A.F. Oke, M.E. Rice, R.N. Adams, I. Moore, R.B. Moore III, M.N. Szentirmay, C.R. Martin, J. Electroanal. Chem. 1985, 189 , 85–94. 91. M.E. Rice, C. Nicholson, Anal. Chem. 1989, 61 , 1805–1810. 92. M.P. Brazell, R.J. Kasser, K.J. Renner, J. Feng, B. Moghaddam, R.N. Adams, J. Neurosci. Meth. 1987, 22 , 167–172. 93. N. Diab, J. Oni, A. Schulte, I. Radtke, A. Blochl, W. Schuhmann, Talanta 2003, 61 , 43–51. 94. R. Trouillon, C. Cheung, B.A. Patel, D. O’Hare, Analyst 2009, 134 , 784–793. 95. D.P. Manica, Y. Mitsumori, A.G. Ewing, Anal. Chem. 2003, 75 , 4572–4577. 96. S.A. Brazill, P. Singhal, W.G. Kuhr, Anal. Chem. 2000, 72 , 5542–5548. 97. P. Luo, F. Zhang, R.P. Baldwin, Anal. Chem. 1991, 63 , 1702–1707. 98. D.L. Robinson, A. Hermans, A.T. Seipel, R.M. Wightman, Chem. Rev . 2008, 108 , 2554–2584. 99. K. Pihel, E. Travis, R. Borges, R. Wightman, Biophys. J . 1996, 71 , 1633–1640. 100. J.A. Stamford, F. Crespi, C.A. Marsden, in Monitoring Neuronal Activity, A Practical Approach (J.A. Stamford, ed.), Oxford University Press, Oxford, 1992, p. 113. 101. B.B. Anderson, A.G. Ewing, J. Pharmaceut. Biomed. Anal . 1999, 19 , 15–32. 102. J.B. Chien, R.A. Wallingford, A.G. Ewing, J. Neurochem. 1990, 54 , 633–638. 103. M.A. Hickey, G.P. Reynolds, A.J. Morton, J. Neurochem. 2002, 81 , 46–59. 104. E.N. Pothos, V. Davila, D. Sulzer, J. Neurosci . 1998, 18 , 4106–4118. 105. R.G.W. Staal, E.V. Mosharov, D. Sulzer, Nature Neurosci . 2004, 7 , 341–346.

548

DIAMOND SENSORS FOR NEUROCHEMISTRY

106. K.P. Troyer, R.M. Wightman, J. Biol. Chem. 2002, 277 , 29101–29107. 107. B.J. Venton, H. Zhang, P.A. Garris, P.E.M. Phillips, D. Sulzer, R.M. Wightman, J. Neurochem. 2003, 87 , 1284–1295. 108. L.M. Borland, G. Shi, H. Yang, A.C. Michael, J. Neurosci. Meth. 2005, 146 , 149–158 109. H. Yang, A.C. Michael, in Electrochemical Methods for Neuroscience (A.C. Michael, L. Borland, eds.), CRC Press, Boca Raton, 2007, p. 489. 110. H.-F. Cui, J.-S. Ye, Y. Chen, S.-C. Chong, F.-S. Sheu, Anal. Chem. 2006, 78 , 6347–6355. 111. A. Hermans, A.T. Seipel, C.E. Miller, R.M. Wightman, Langmuir 2006, 22 , 1964–1969. 112. B.D. Bath, H.B. Martin, R.M. Wightman, M.R. Anderson, Langmuir 2001, 17 , 7032–7039. 113. K.T. Kawagoe, J.B. Zimmerman, R.M. Wightman, J. Neurosci. Meth. 1993, 48 , 225–240. 114. S.E. Hochstetler, M. Puopolo, S. Gustincich, E. Raviola, R.M. Wightman, Anal. Chem. 2000, 72 , 489–496. 115. R. Hauert, Diam. Rel. Mat . 2003, 12 , 583–589. 116. A.J. Strong, M.G. Boutelle, H. Landolt, M.C. Parkin, M. Lauritzen, M. Fabricius, J. Neurotrauma 2004, 21 , 206. 117. A.J. Strong, M. Fabricius, M.G. Boutelle, S.J. Hibbins, S.E. Hopwood, R. Jones, M.C. Parkin, M. Lauritzen, Stroke 2002, 33 , 2738–2743. 118. M.C. Parkin, S.E. Hopwood, M.G. Boutelle, A.J. Strong, TrAC Trends Anal. Chem. 2003, 22 , 487–497. 119. D.A. Jones, M.C. Parkin, H. Langemann, H. Landolt, S.E. Hopwood, A.J. Strong, M.G. Boutelle, J. Electroanal. Chem. 2002, 538–539 , 243–252. 120. P. Ariano, P. Baldelli, E. Carbone, A. Gilardino, A. Lo Giudice, D. Lovisolo, C. Manfredotti, M. Novara, H. Sternschulte, E. Vittone, Diam. Rel. Mat . 2005, 14 , 669–674. 121. P. Ariano, A. Lo Giudice, A. Marcantoni, E. Vittone, E. Carbone, D. Lovisolo, Biosensors Bioelectronics 2009, 24 , 2046–2050. 122. H.-Y. Cheng, R.H. Schenk, R.N. Adams, J. Electroanal. Chem. 1979, 100 , 23–31. 123. R.M. Wightman, Science 2006, 311 , 1570–1574. 124. E. Fortin, J. Chane-Tune, D. Delabouglise, P. Bouvier, T. Livache, P. Mailley, B. Marcus, M. Mermoux, J.-P. Petit, S. Szunerits, E. Vieil, Electroanalysis 2005, 17 , 517–526. 125. E. Bitziou, D. O’Hare, B.A. Patel, Anal. Chem. 2008, 80 , 8733–8740. 126. J. Anderson, Annu. Rev. Mat. Res. 2001, 31 , 81–110. 127. F. Vacanti, Plastic Reconstructive Surgery 2004, 113 , 949. 128. F. Katou, N. Andoh, K. Motegi, H. Nagura, J. Cranio-Maxillofacial Surgery 1996, 24 , 155–162. 129. M. Frost, M.E. Meyerhoff, Anal. Chem. 2006, 78 , 7370–7377. 130. N. Wisniewski, M. Reichert, Colloids Surf. B 2000, 18 , 197–219. 131. F. Moussy, D.J. Harrison, Anal. Chem. 1994, 66 , 674–679. 132. S. Zhang, G. Wright, Y. Yang, Biosensors Bioelectronics 2000, 15 , 273–282. 133. Y. Wu, M.E. Meyerhoff, Talanta 2008, 75 , 642–650. 134. D.A. Riccio, K.P. Dobmeier, E.M. Hetrick, B.J. Privett, H.S. Paul, M.H. Schoenfisch, Biomaterials 2009, 30 , 4494–4502. 135. K.F. Chong, K.P. Loh, S.R.K. Vedula, C.T. Lim, H. Sternschulte, D. Steinmuller, F.-S. Sheu, Y.L. Zhong, Langmuir 2007, 23 , 5615–5621. 136. X. Xiao, J. Wang, C. Liu, J.A. Carlisle, B. Mech, R. Greenberg, D. Guven, R. Freda, M.S. Humayun, J. Weiland, O. Auciello, J. Biomed. Mat. Res. B 2006, 77B, 273–281. 137. J. Wang, M.S. Lin, Anal. Chem., 1988, 60 , 499–502.

REFERENCES

138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168.

169. 170. 171. 172. 173. 174.

549

J. Wang, P. Tuzhi, Anal. Chem. 1986, 58 , 1787–1790. T.J. Self, F. Crespi, J. Mat. Sci . 1992, 3 , 418–425. B. Hoyer, N. Jensen, Electrochem. Comm. 2006, 8 , 323–328. B. Hoyer, N. Jensen, J. Electroanal. Chem. 2007, 601 , 153–160. M.Z. Wrona, G. Dryhurst, J. Org. Chem. 1987, 52 , 2817–2825. M.Z. Wrona, G. Dryhurst, J. Electroanal. Chem. 1990, 278 , 249–267. G. Dryhurst, Chem. Rev . 1990, 90 , 795–811. B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, Chem. Lett . 1999, 28 , 1213–1214. B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, Anal. Chem. 2000, 72 , 1632–1638. S. Raina, W.P. Kang, J.L. Davidson, Diam. Rel. Mat . 2009, 18 , 574–577. A. Hermans, R.M. Wightman, Langmuir 2006, 22 , 10348–10353. J.M. Halpern, S. Xie, G.P. Sutton, B.T. Higashikubo, C.A. Chestek, H. Lu, H.J. Chiel, H.B. Martin, Diam. Rel. Mat . 2006, 15 , 183–187. L. Timmermann, E. Florin, C. Reck, Expert Rev. Med. Devices 2007, 4, 651–661. A.L. Benabid, L. Vercucil, A. Benazzouz, A. Koudsie, S. Chabardes, L. Minotti, P. Kahane, M. Gentil, D. Lenartz, C. Andressen, P. Krack, P. Pollak, Adv. Neurol . 2003, 91 , 293–302. P.A. Starr, J.L. Vitek, R.A. Bakay, Neurosurgery 1998, 43 , 898–1013. K.L. Soh, W.P. Kang, J.L. Davidson, S. Basu, Y.M. Wong, D.E. Cliffel, A.B. Bonds, G.M. Swain, Diam. Rel. Mat . 2004, 13 , 2009–2015. A.L. Colley, C.G. Williams, U. D’Haenens-Johansson, M.E. Newton, P.R. Unwin, N.R. Wilson, J.V. Macpherson, Anal. Chem. 2006, 78 , 2539–2548. T.M. Olefirowicz, A.G. Ewing, J. Neurosci. Meth. 1990, 34 , 11–15. M.D. Gershon, Aliment. Pharmacol. Ther. 1999, 13 , 15–30. M.D. Gershon, Adv. Exper. Med . 1991, 221–230. J.J. Galligan, Curr. Opin. Pharmacol . 2002, 2 , 623–629. J.J. Galligan, Gastroenterology 2004, 126 , 1897–1899. V. Fykse, E. Solligard, M.O. Bendheim, D. Chen, J.E. Gronbech, A.K. Sandvik, H.L. Waldum, Acta Physiol . 2006, 186 , 37–43. A.K. Sandvik, M.J. Lewin, H.L. Waldum, Br. J. Pharmacol . 1989, 96 , 557–562. R. Kachoosangi, R. Compton, Anal. Bioanal. Chem. 2007, 387 , 2793–2800. A. Bard, L.R. Faulkner, in Electrochemical Methods: Fundamentals and Applications, John Wiley & Sons, New York, 2001. M.D. Gershon, Curr. Opin. Pharmacol . 2003, 3 , 600–607. H.E. Raybould, H.J. Cooke, F.L. Christofi, Neurogastroenterol. Motil . 2004, 16 , 60–63. K. Racke, H. Schworer, Pharmacol. Res. 1991, 23 , 13–25. K. Racke, A. Reimann, H. Schworer, H. Kilbinger, Behavioural Brain Res. 1995, 73 , 83–87. M.D. Coates, C.R. Mahoney, D.R. Linden, J.E. Sampson, J. Chen, H. Blaszyk, M.D. Crowell, K.A. Sharkey, M.D. Gershon, G.M. Mawe, P.L. Moses, Gastroenterology 2004, 126 , 1657–1664. R. Spiller, Neurogastroenterol. Motil . 2007, 19 , 25–31. M.D. Gershon, J. Tack, Gastroenterology 2007, 132 , 397–414. A. Lecci, P. Santicioli, C.A. Maggi, Curr. Opin. Pharmacol . 2002, 2 , 630–641. T. Takahasi, J. Gastroenterol . 2003, 38 , 421–430. B.A. Patel, J.J. Galligan, G. Swain, X. Bian, Neurogastroenterol. Motil . 2008, 20 , 1243–1250. N. Spataru, T.N. Rao, D.A. Tryk, A. Fujishima, J. Electrochem. Soc. 2001, 148 , E112–E117.

550 175. 176. 177. 178. 179. 180. 181. 182. 183.

184. 185. 186. 187. 188. 189. 190. 191. 192. 193.

DIAMOND SENSORS FOR NEUROCHEMISTRY

T.R. Insel, J. Zohar, C. Benkelfat, D.L. Murphy, Ann. NY Acad. Sci . 1990, 574–586. H.Y. Meltzer, Ann. NY Acad. Sci . 1990, 486–499. E.H. Cook Jr, B.L. Leventhal, Curr. Opin. Pediatr. 1996, 8 , 348–354. D.L. Murphy, A. Lerner, G. Rudnick, K.P. Lesch, Mol. Interv . 2004, 4 , 109–123. D.L. Murphy, M.A. Fox, K.R. Timpano, P.R. Moya, R. Ren-Patterson, A.M. Andrews, A. Holmes, K.P. Lesch, J.R. Wendland, Neuropharmacology 2008, 55 , 932–960. M.A. Fox, C.L. Jensen, P.S. Gallagher, D.L. Murphy, Neuropharmacology 2007, 53 , 643–656. R. Uher, P. McGuffin, Mol. Psychiatry 2008, 13 , 131–146. J.E. Lim, A. Papp, J. Pinsonneault, W. Sadee, D. Saffen, Mol. Psychiatry 2006, 11 , 649–662. R.V. Parsey, R.S. Hastings, M.A. Oquendo, X. Hu, D. Goldman, Y.Y. Huang, N. Simpson, J. Arcement, Y. Huang, R.T. Ogden, R.L. Van Heertum, V. Arango, J.J. Mann, Am. J. Psychiatry 2006, 163 , 48–51. S. Xie, G. Shafer, C.G. Wilson, H.B. Martin, Diam. Rel. Mat . 2006, 15 , 225–228. A. Suzuki, T.A. Ivandini, K. Yoshimi, A. Fujishima, G. Oyama, T. Nakazato, N. Hattori, S. Kitazawa, Y. Einaga, Anal. Chem. 2007, 79 , 8608–8615. G.A. Gerhardt, A.F. Hoffman, J. Neurosci. Methods 2001, 109 , 13–21. C.A. Mart´ınez-Huitle, Small 2007, 3 , 1474–1476. J. Lee, S.-M. Park, Anal. Chim. Acta 2005, 545 , 27–32. P. Hashemi, R. Bhatia, H. Nakamura, J.P. Dreier, R. Graf, A.J. Strong, M.G. Boutelle, J. Cereb. Blood Flow Metab. 2008, 29 , 166–175. S. Deeba, E. Corcoles, B. Hanna, P. Pareskevas, O. Aziz, M. Boutelle, A. Darzi, Diseases Col. Rect . 2008, 51 , 1408–1413. L. Hillered, P.M. Vespa, D.A. Hovda, J. Neurotrauma 2005, 22 , 3–41. K.B. Holt, J. Hu, J.S. Foord, Anal. Chem. 2007, 79 , 2556–2561. M. Pagels, C.E. Hall, N.S. Lawrence, A. Meredith, T.G.J. Jones, H.P. Godfried, C.S.J. Pickles, J. Wilman, C.E. Banks, R.G. Compton, L. Jiang, Anal. Chem. 2005, 77 , 3705–3708.

21 DNA-Modified Diamond Films Nianjun Yang and Christoph E. Nebel

21.1

INTRODUCTION

Genomics research has elucidated many new biomarkers that have the potential to greatly improve disease diagnostics [1–3]. The availability of multiple biomarkers is important in diagnoses of complex diseases such as cancer [4–6]. In addition, various markers will be required to identify different stages of disease pathogenesis to facilitate early detection. The use of multiple markers in health care will, however, ultimately depend on the development of detection techniques that will allow rapid detection of many markers with high selectivity and sensitivity. Currently, extensive quests for proper transducer materials, for optimization of detection techniques and sensitivities, for realization of highly integrated sensor arrays, and for bio-interfaces that show high chemical stability, which are required in high through-put systems, are therefore ongoing. Most of established substrate materials (“transducers”) such as latex beads, polystyrene, carbon electrodes, gold, and oxidized silicon or glass do not possess all desired properties like flatness, homogeneity, chemical stability, reproducibility, and biochemical surface modifications [7–11]. In addition, future technologies require integration of bio-functionalized surfaces with microelectronics or micromechanical tools, which adds significant complexity to this topic [11–15], as most microelectronic-compatible materials like silicon, SiOx , and gold show degradation of their bio-interfaces in electrolyte solutions [15]. Diamond can become a promising candidate for bio-electronics because it shows effective electronic [16–18] and chemical properties [19–21]. Figure 21.1 shows voltammograms for water electrolysis of various electrodes. The supporting electrolyte is 0.5 M Synthetic Diamond Films: Preparation, Electrochemistry, Characterization, and Applications, First Edition. Edited by Enric Brillas and Carlos Alberto Mart´ınez-Huitle. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

551

552

DNA-MODIFIED DIAMOND FILMS

Au

Current (mA cm−2)

Pt Glassy Carbon

(H)SCD B:PCD (USU)

B:PCD (NRL)

B:(H)SCD –3

–2

–1

0

1

2

3

Potential (V) vs. SCE Figure 21.1 Voltammograms for water electrolysis on various electrodes. The supporting electrolyte is 0.5 M H2 SO4 . The graphs are shifted vertically for comparison. Two polycrystalline films, B:PCD(NRL) with 5 × 1019 B cm−3 and B:PCD(USU) with 5 × 1020 B cm−3 are compared with a single crystalline boron-doped diamond B:(H)SCD with 3 × 1020 B cm−3 and with an undoped diamond (H)SCD. Also shown are data for Pt, Au, and glassy carbon from Ref. 18. Oxidation reactions (e.g., oxygen evolution) have positive currents and emerge around 1.8 V for all diamond samples. Reduction reactions (e.g., hydrogen evolution) have negative currents and show very different properties. Note that the background current within the regime between hydrogen and oxygen evolution for diamond is very low, and the electrochemical potential window is large, compared to glassy carbon, Pt, and Au. (Reprinted from Refs. 21 and 22.) See color insert.

H2 SO4 . Please note that each current/voltage scan has been shifted vertically for better comparison. Two polycrystalline films, B:PCD(NRL) with 5 × 1019 B cm−3 and B:PCD(USU) with 5 × 1020 B cm−3 [22,23] are compared with a single crystalline boron-doped diamond B:(H)SCD with 3 × 1020 B cm−3 and with an undoped diamond (H)SCD [24]. The electrochemical potential-window of diamond is significantly larger. In addition, by tuning the boron-doping level, the onset of hydrogen evolution (rise of current at negative potentials) can be reduced or switched completely off by decreasing the boron-doping level from extremely high with >1020 cm−3 boron (“metallic”) to “undoped” (=intrinsic diamond). There are some other parameters affecting the electrochemical potential-window such as crystal orientation [25], structural perfection of polycrystalline diamond [26], and surface termination [19]. Their discussion in this context is, however, is beyond the scope of this chapter. Another feature that makes boron-doped diamond electrodes attractive for electroanalysis and biochemical sensing applications is the fact that the background current of diamond is considerably lower than that of conventional materials. Figure 21.2 shows a comparison of background currents of boron-doped diamond (3 × 1020 B cm−3 ) with gold (Au) and glassy carbon (GC). The electrochemical background current in phosphate buffer is about 10 times lower than that of Au and about 100 times lower than that of GC. Surface-induced conductivity of hydrogen-terminated undoped diamond in electrolyte solutions is another unique property that attracted significant attention in recent years [27]. It is generated by transfer doping of hydrogen-terminated diamond immersed into electrolyte solution. The term transfer doping indicates that the surface conductivity of diamond rises from missing valence-band electrons as these electrons “transfer” into

21.1 INTRODUCTION

glassy carbon

10–1 Current density (mA cm–2)

553

10–2 Au 10–3 diamond

10–4 10–5 –1.5

–1

–0.5

0

0.5

1

1.5

2

Potential (V) vs. Ag/AgCl Figure 21.2 Comparison of the background current of boron-doped diamond electrode with that of gold and glassy carbon electrode in 0.1 M Na2 SO4 at a scan rate of 100 mV s−1 . The boronconcentration of diamond electrode is 3 × 1020 B cm−3 .

the electrolyte [28–30]. To achieve transfer, the chemical potential of an electrolyte must be equal or below the energy level of the valence-band maximum, EV . For most semiconductors this is not the case, as can be seen in Figure 21.3. Even for oxidized diamond, chemical potentials are usually well above EV . It changes drastically if the surface of diamond, which consists of about 2 × 1015 cm−2 carbon bonds, is terminated with hydrogen. Hydrogen-carbon bonds are polar covalent bonds (electronegativity of carbon: 2.5 and of hydrogen: 2.1); therefore, a dense surface dipole layer is generated with slightly negative charged carbon (C− ) and slightly positive charged hydrogen (H+ ). From basic electrostatics such a dipole layer causes an electrostatic potential step V perpendicular to the surface over a distance of the order of the C–H bond length of ˚ Simple calculations show that the energy variation over this dipole is in the range 1.1 A. of 1.6 eV (for a detailed discussion, see Ref. 31). This dipole energy increases all energy levels of diamond for about 1.6 eV with respect to the chemical potential of an electrolyte (see Figure 21.3). Conduction-band states of diamond are shifted above the vacuum level. This scenario is called negative electron affinity. (See Figure 21.3: clean diamond, where the vacuum level is about 0.3 eV above the conduction band minimum and H-terminated diamond, and the vacuum level is 1.3 eV below the conduction band minimum [30,32].) As all electronic states are shifted for the same dipole energy, occupied valenceband states emerge above the chemical potential, μ, of electrolytes. Electrons from the diamond valence-band (electronically occupied states) can therefore tunnel into empty electronic states of the electrolyte until thermodynamic equilibrium between the Fermi level of diamond and the electrochemical potential of the electrolyte is established. This is schematically shown in Figure 21.4a). Fermi-level and chemical potential, μ, align and ˚ in width, which is, in effect, a confined form a narrow valence-band bending of 20–30 A hole accumulation layer [33,34]. Such alignment requires defect-free bulk and surface properties. The role of H-termination is to minimize the surface defect density. During recent years, the growth of diamond has been optimized to such a level in combination with an optimized H-termination of the surface (for reviews, see Ref. 35, 36).

554

DNA-MODIFIED DIAMOND FILMS

3

1 0 –1

2

–2 –3

CdS

–χ

–4.2

pH2 = 1 bar 1 mbar 1 μbar 0

2

–4.0

μe (eV)

Energy rel. to vacuum level (eV)

2

3

–3.8

Hydrogen Term Diamond EVBM μ

4

EVAC

0

–4.4

–1

–4.6 6 8 10 12 14 pH-Value

–2 –3

CdSe

μ

–4 –5 –6 SiC

–7

1

GaP

GaAs Si

–4 –5

H-terminated Diamond

Ge

–6

Diamond

–7

Semiconductors

Figure 21.3 Energies of the conduction- and valence-band edges of a number of conventional semiconductors, and of hydrogen-terminated and hydrogen-free diamond relative to the vacuum level EVAC are shown. Please note that the H-terminated diamond shows a negative electron affinity (-χ) as the conduction band edge is above the vacuum level of the electrolyte. The dashed horizontal line marks the chemical potential μ for electrons in an acidic electrolyte under conditions of the standard hydrogen electrode. The insert shows the chemical potential under general nonstandard conditions as a function of pH and for different partial pressures of hydrogen in the atmosphere as given by Nernst’s equation [30]. (Reprinted from Ref. 21.) See color insert.

Energy Hole Channel in Diamond

EVBM(x = 0) EVBM(x)

pH 14 Chemical Potential μ pH 0

Fermi Level Valence Band 0

Diamond Electrolyte C–H

X

(b) 0.6 Gate Potential [V]

(a)

Sensitivity –55 (±6) mV

0.5 0.4 0.3 0.2 0.1 0.0 –0.1

Diamond ISFET 2

4

6

8

10

12

pH

Figure 21.4 (a) Fermi-level and chemical potential alignment at the interface diamond/electrolyte after equilibration. Due to transfer doping, electrons are missing in diamond so that a thin two-dimensional (2D) hole accumulation layer is generated [33,34]. The hole density in this layer depends on the chemical potential as indicated by arrows. Reprinted from Ref. 21. (b) pH-sensitivity of a diamond ionsensitive field effect transistor (ISFET) [35,36,40–42]. The gate potential shift shows a pH-dependence of 55 mV/pH, which is close to the Nernst prediction. (Reprinted from Ref. 21.)

Evidence from theory and experiments suggest that the negative electron affinity of diamond in contact with water is significantly smaller (about (0 to −0.4) eV) than that observed in high vacuum [37–39]. A dominant interaction of diamond energy levels with the H2 /H+ redox states seems to be therefore less likely. But diamond valenceband states will still scale with interactions to the O2 /H2 O couple, giving rise to the discovered phenomena. As the chemical potential of electrolytes is changing with pHvalue, a variation of the surface conductivity can be detected experimentally, following closely the Nernst prediction with 55 mV/pH (see Figure 21.4b) [40–42].

21.1 INTRODUCTION

555

Diamond is known to be biocompatible [43–45] and therefore has a potential for in vivo electronic applications. When Takahashi et al. in 2000 [46,47] first introduced a photochemical chlorination/amination/carboxylation process of the initially H-terminated diamond surface, a giant step toward biofunctionalization of diamond was taken, as the obstacle of “chemical inertness” of diamond had finally been removed. This triggered more activities so that two years later, in 2002, Yang and colleagues introduced a new photochemical method to modify nanocrystalline diamond surfaces using alkenes [15], followed by electrochemical reduction of diazonium salts that have been successfully applied to functionalize boron-doped ultrananocrystalline diamond [48]. Shortly later, a direct amination of diamond was introduced [49]. Such functionalized surfaces have been further modified with DNA, enzymes, and proteins, and characterized using fluorescence microscopy and impedance spectroscopy [15,50,51] voltammetry and gate-potential shifts of ion-sensitive field-effect transistors [52,53]. Perhaps the most influential argument for diamond applications in biotechnology has been given by Yang et al. in 2002 [15]. They characterized the bonding stability of DNA to nanocrystalline diamond and other substrates applying DNA hybridization/denaturation cycles and fluorescence microscopy investigations. The result is shown in Figure 21.5 in comparison with Au, Si, and glassy carbon. It demonstrates that DNA bonding to diamond, or more general to carbon atoms, is significantly better than to other substrates, as no degradation of fluorescence intensity could be detected. The long-term bonding stability is especially important in multiarray sensor applications, which are costly to produce and therefore need long-term stability in high through-put systems. Applications of diamond sensors will ultimately depend on the commercial availability of diamond films. This has improved significantly during recent years as meanwhile, nano- and polycrystalline CVD diamond films can be grown by plasma-enhanced chemical vapor deposition (CVD) heteroepitaxially on silicon and other substrates on large area. Growth parameters are currently optimized to deposit films at low temperature to

Fluorescence intensity (a.u.)

1500 Diamond 1000

Gold Silicon

500 Glassy Carbon

0

0

5

10

15

20

25

30

Hybridization cycles Figure 21.5 Stability of DNA bonding to ultrananocrystalline diamond, Au, Si, and glassy carbon as detected during 30 successive cycles of hybridization and denaturation. In each case the substrates were amine-modified and then linked to thiol-terminated DNA [15]. (Reprinted from Ref. 21.) See color insert.

556

DNA-MODIFIED DIAMOND FILMS

allow integration into established silicon technology [54,55]. Single crystalline diamond produced by high-pressure high-temperature growth is commercially available due to an increasing number of companies producing diamond. The size of these layers is relative small, typically 4 mm × 4 mm, which is, however, large enough to be use as substrate for homoepitaxial growth of high-quality single crystalline CVD diamond (“electronic grade quality”). With respect to electronic applications, a careful selection of “diamond” material is required. Figure 21.6 summarizes the structural properties of nano-, poly and singlecrystalline diamond. Ultranano-, nano- and polycrystalline diamond layers are dominated by grain boundaries that are decorated with sp2 and amorphous carbon [56–58]. The (I)

(II)

Nanocrystalline Diamond

(a)

1 μm

1 μm

Silicon Polycrystalline Diamond

(b)

200 μm 200 μm Substrate Side

(c)

CL Intensity (a.u.)

4 x 106

FE(TO)

T = 16 K E = 13 kV I = 2 μA

3 x 106

2 x 106

0 5.0

0.5 nm

0.75

0.3 nm

0.50

0.0 nm

0.25

6

1 x 10

1.00

FE(LO) FE(TA) Γ FE(TO + O ) 5.1 5.2 5.3 5.4 Photon Energy (eV)

5.5

0

0.25

0.50

0.75

0 1.00 μm

Figure 21.6 Comparison of different diamond films: (a) Ultranano-, nano- and (b) polycrystalline diamond layers are dominated by grain boundaries that are decorated with sp2 and amorphous carbon [54,56–58]. The volume-fraction of sp2 and grain boundaries depends on the growth parameter and varies from layer to layer. Amorphous carbon and sp2 generate a continuous electronic density-ofstates distribution in the gap of diamond. In addition, such diamond films show a significant surface roughness in the range of 30–50 nm for nanocrystalline diamond (see (a), II) and micrometer to tens of micrometers for polycrystalline layers (see (b), II). On the other hand, single crystalline CVD diamond has been optimized over recent years to electronic-grade quality with atomically smooth surfaces (see (c), II) [63–65]. A typical cathodoluminescence spectrum measured at 16 K is shown in (c), I. (Reprinted from Ref. 21.) See color insert.

21.1 INTRODUCTION

557

volume fraction of sp2 and grain boundaries depends on growth parameter and varies from layer to layer. Especially ultrananocrystalline diamond contains a high volume fraction of up to 5% [55]. Amorphous carbon and sp2 generate a continuous electronic density-ofstates distribution in the gap of diamond. These states will affect sensor sensitivity and dynamic properties. Applications of polycrystalline diamond as photo- or high-energy particle detector show memory and priming effects that arise by metastable filling of grain-boundary states [59,60]. In addition, such diamond films show a significant surface roughness in the range of 30–50 nm for nanocrystalline diamond and micrometer to tens of micrometer for polycrystalline layers. Commercially available polycrystalline diamond is therefore often mechanically polished to achieve a smooth surface. This generates, however, a thin highly damaged diamond surface that cannot be tolerated in surfacerelated electronic applications. This is because surface defects, about 2.5 eV above the valence band maximum, will pin the Fermi-level and will deteriorate heterojunction properties [61,62]. On the other hand, single crystalline CVD diamond has been optimized over recent years to electronic grade quality with atomically smooth surfaces (see Figure 21.6c) [63–65]. These films show even at room temperature strong free-exciton emissions at 5.27 eV and 5.12 eV, which are fingerprints of low defect densities, typically below 1015 cm−3 . The bulk resistivity of undoped films at 300 K is larger than 1015 cm [56,57]. Atomic force microscopy (AFM) characterization of such films show surface morphologies that indicate atomically flat properties with step-etch growth, where terraces run parallel to the (110) direction. After H-terminated of such layers, heterojunction properties follow very well predicted properties of defect-free diamond. Diamond can be doped p-type by boron, which results in a doping level 360 meV above the valence band maximum [66]. Phosphorus doping has been introduced for n-type doping with the phosphorus doping level 0.6 eV below the conduction band minimum [67]. Both levels are basically too deep for room temperature electronic applications, which is the typical regime for bio-electronics. One way to overcome this problem is the application of metallic doping levels, which means for boron doping densities >3 × 1020 B cm−3 [68]. Such a high acceptor density gives rise to significant wave-function overlap of acceptor atoms to allow hole propagation in these states, without thermal activation to the valence-band. Highly boron-doped diamond is therefore well established in electrochemistry. Applications of n-type diamond in electro- or biochemical sensors are not favorable. The Fermi-level (0.6 eV below the conduction band) and chemical potential of electrolytes (typically 4.5 eV below the vacuum level (see Figure 21.3) are too different, giving rise to energy barrier limited electronic interactions. This brief introduction of major properties of diamond indicates that it is indeed an interesting transducer material for biosensor applications. To realize biosensors from diamond, the details of biomolecular functionalization needs to be investigated and optimized with respect to sensing requirements [15,46,47,49,50,69]. This includes controlled deposition of linker layers to achieve optimization of biomolecule densities on the transducer [70]. In the following we review our achievements with respect to interface properties of single crystalline CVD diamond to organic linker-molecule layers and DNA films. We describe surface functionalization using amine- and phenyl-layers, which are currently attracting significant attention. There are other photochemical modifications of diamond available (for example, see Refs. 46, 47 or [49]) that are based on direct or indirect surface amination. It is very likely that sooner or later the modification spectrum will

558

DNA-MODIFIED DIAMOND FILMS

become even broader. In the following, however, we want to focus on (1) amine and (2) phenyl-related modifications, as these techniques are established, are used by a growing number of scientists, and are characterized reasonably well. For our experiments we used homoepitaxial grown, atomically smooth CVD diamond, either undoped or metallically boron doped, that are free of grain boundaries, sp2 -carbon, or other defects. We apply (1) photochemical attachment chemistry of alkene molecules to undoped diamond [71–73] and (2) electrochemical reduction of diazonium salts [71,74,75] to form nitrophenyl linker molecules on boron-doped CVD diamond. The bonding mechanisms, kinetics, and molecule arrangements and densities will be introduced using a variety of experiments such as x-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), atomic force microscopy (AFM), cyclic voltammetry, and several electronic characterization techniques. By use of a heterobifunctional cross-linker, thiol-modified single-stranded probe DNA (ss-DNA) is bonded to diamond. Finally, such surfaces are exposed to fluorescence labeled target ss-DNA to investigate hybridization by use of fluorescence microscopy. We applied AFM in electrolyte solution to gain information about geometrical properties of DNA, bonding strength, as well as the degree of surface coverage. Finally, we introduce results with respect to labelfree electronic sensing of DNA hybridization using Fe(CN6 )3−/4− redox molecules as mediator in amperometric experiments and variation of gate-potential threshold shifts in DNA-FET structures. For a detailed summary of these results, see Ref. 71. We will discuss a novel way to generate vertically aligned diamond nanostructures [76,77] with geometrical properties ranging from a few nanometers (“nanotextures”), to micrometers (“wires”). We apply self-aligned etching masks from diamond and Ni nanoparticles to realize nanotextures for controlled DNA bonding [78–82]. Wires give rise to surface enhancements. As shown schematically in Figure 21.7 [78,81], the tips of diamond nanowires can be functionalized by electrochemical bonding of nitrophenyl molecules (Figure 21.7a). These functionalized tips act as anchors for single-stranded marker DNA molecules (Figure 21.7b and c), thereby giving rise to a controlled spacing between DNA molecules. We will describe electrochemical properties of diamond nanotextures before and after electrochemical grafting with nitrophenyl molecules, and subsequent DNA attachment and sensing using redox molecules as mediator for DNA hybridization detection (Figure 21.7d).

21.2 21.2.1

DIAMOND TRANSDUCER PROPERTIES CVD Diamond Growth

High- quality undoped, single crystalline diamond films of 200 nm thickness have been grown homoepitaxially on 3 mm × 3 mm (100) oriented synthetic Ib substrates, using microwave-plasma chemical vapor deposition. Growth parameters were substrate temperature 800◦ C, microwave power 750 W, total gas pressure 25 Torr, and total gas flow 400 sccm with 0.025% CH4 in H2 . Note that the used substrates have been grown commercially by the high pressure high temperature (HPHT) technique and contain typically up to 1019 cm−3 dispersed nitrogen (type Ib diamond). To achieve H-termination after growth, CH4 is switched off and the diamond is exposed to a pure hydrogen plasma for 5 min with otherwise identical parameters. After switching off the hydrogen plasma, the diamond layer is cooled down to room temperature in H2 atmosphere. A detailed

21.2 DIAMOND TRANSDUCER PROPERTIES

559

(a) Electrochemical nitrophenyl grafting

(b) Animation & cross-linker attachment

Nanotextures from Boron-Doped Diamond

Nanotextures from Boron-Doped Diamond

(c) Probe DNA immobilization

Nanotextures from Boron-Doped Diamond

(d) DNA hybridization & detection

Nanotextures from Boron-Doped Diamond

Figure 21.7 Schematic illumination of the biofunctionalization of diamond nanotextures. (a) Electrochemical grafting of diazonium salts, (b) electrochemical reduction of nitrophenyl into aminophenyl and the addition of the cross-linkers, (c) immobilization of probe DNA, and (d) optical and electronic detection of DNA hybridization. (Reprinted from Ref. 81.) See color insert.

560

DNA-MODIFIED DIAMOND FILMS

500

400

[nm]

300

200

100

0 0

100

200

300

400

500

[nm] 0.00

[nm]

0.50

Figure 21.8 Typical AFM surface morphology of a single crystalline CVD diamond surface, as ˚ detected by AFM and used in these studies. The root mean square (RMS) roughness was below 1 A. (Reprinted from Ref. 21.) See color insert.

discussion of sample growth properties can be found in Refs. 63–65. Layers are highly insulating with resistivities larger than 1015 cm. Surfaces are smooth, as characterized by atomic force microscopy (AFM). A typical result is shown in Figure 21.8. The root ˚ mean square (RMS) surface roughness is below 1 A. Boron-doped single crystalline diamond films have been grown homoepitaxially on synthetic (100) Ib diamond substrates with 4 mm × 4 mm × 0.4 mm size, using microwave plasma-assisted chemical vapor deposition (CVD). Growth parameters are microwave power 1200 W, which generates a substrate temperature around 900◦ C; gas pressure 50 Torr; and gas flow 400 sccm with 0.6% CH4 in H2 . B2 H6 , as boron source, is mixed in CH4 , where the boron/carbon atomic ratio (B/C) was 16,000 ppm. Typically, 1 μm thick films have been grown within 7 h. To measure bulk properties, boron-doped diamond is wet-chemically oxidized by boiling in a mixture of H2 SO4 and HNO3 (3:1) at 230◦ C for 60 min. Figure 21.9a shows the conductivity of boron-doped diamond as a function of the doped boron concentration. When the boron concentration is lower than 1.0 × 1020 cm−3 , the conductivity of diamond films is quite low. When the boron concentration is higher than 3 × 1020 cm−3 , metallic boron-doped diamond films can be achieved. Using the experimental conditions we mentioned earlier, the result was metallic conductive diamond film. Figure 21.9b

21.2 DIAMOND TRANSDUCER PROPERTIES

561

(a) 106 105

Resistivity (Ω cm)

104 103 102 101 100 10–1 10–2 10–3 1016

1017

1018

1019

Boron concentration

1020

1021

(cm–3)

(b)

Figure 21.9 (a) Conductivity of boron-doped diamond films as a function of boron-concentration. (b) Typical temperature-dependent conductivity of a metallically boron doped CVD diamond. The doping level is in the range of 5 × 1020 B cm−3 .σ is activated with 2 meV, which indicates hopping propagation of holes in the acceptor band. (Reprinted from Ref. 21.)

shows a typical result of conductivity, σ , which is in the range of 200 (cm)−1 at 300 K, showing a negligible activation energy of 2 meV (“metallic properties”). It is achieved by ultra-high doping of diamond with 3 × 1020 cm−3 boron acceptors as detected by secondary ion mass spectroscopy (SIMS). The crystal quality is not deteriorated by this high-boron incorporation. A series of x-ray diffraction (XRD) and Raman experiments have been applied to investigate the details of crystal quality, which will be discussed elsewhere. The boron concentration not only affects the conductivity of diamond but it also impacts dramatically the electrochemical properties—for example, the background current, peak splitting of redox couples. Figure 21.10a shows cyclic voltammograms of 1.0 mM Fe(CN)6 3−/4− on diamond films with different boron concentration. Different peak potentials and currents were obtained on these samples. Figure 21.10b summarizes the peak splitting as a function of boron concentration. The peak splitting varies from more than 1 V at the boron concentration of 1019 cm−3 to around 60 mV, at the boron concentration of (3–5) ×1020 –1021 cm−3 . The diamond films with boron concentration

562

DNA-MODIFIED DIAMOND FILMS

(a) 0.2 1020 cm–3

1019 cm–3

Current (mA cm–2)

0.1

0

1021 cm–3

–0.1

–0.2 –1.5

–1

–0.5 0 0.5 Potential (V) vs. Ag/AgCl

1

1.5

(b) 2.0

Peak splitting (V)

1.6

1.2

0.8

0.4

0.0 1019

1020

1021

Boron concentration (cm–3) Figure 21.10 (a) Cyclic voltammograms of 1.0 mM Fe(CN)6 3−/4− in 0.1 M KCl on boron-doped diamond films with different boron concentration. The scan rate was 100 mV s−1 and the boron concentration was measured using SIMS. (b) The peak splitting of Fe(CN)6 3−/4− on boron-doped diamond electrodes as a function of boron concentration.

of (3–5) ×1020 cm−3 show lowest background current and highest peak currents but with narrowest peak splitting (around 60 mV). Please note that higher background current was detected on diamond electrodes when the boron concentration is higher than (3–5) ×1020 cm−3 . 21.2.2

Surface Terminations

Surface termination of diamond is important for diamond properites and applications. Diamond surfaces can be terminated with a variety of atoms and molecules, most prominently with hydrogen –H, –OH, –O–, and F. As the electron affinity of diamond can be changed from −1.3 eV to +1.7 eV (over a range of about 3 eV) by changing the

21.2 DIAMOND TRANSDUCER PROPERTIES

(a)

(b)

(c)

(d)

563

Figure 21.11 Schematic plots of surface arrangements diamond shown as top and side view with respect to (a) H-, (b) OH-, (c) reconstructed carbon, and (d) O-termination. (Reprinted from Ref. 83b.)

termination from hydrogen to oxygen, the electronic properties of the solid/electrolyte interface can be tuned with respect to chemical potentials of electrolytes as well as with respect to HOMO/LUMOS energy levels of molecules [83]. Figure 21.11 summarizes typical atomic arrangements of four surface terminations: (1) H temination, (2) OH, (3) reconstructed carbon, and (4) O termination [83]. H-termination can be achieved using the same parameters as described earlier for diamond growth or through cathodic treatment in acidic solution with high negative potential [84]. Electrochemical experiments on boron-doped diamond are performed on typical areas of 3 mm2 size. Ohmic contacts to boron-doped diamond are evaporated outside this area and sealed with silicon rubber. To obtain patterns of H- and O-termination on diamond surfaces, we apply photolithography, using photoresist as a mask to protect H-terminated areas while uncovered surface parts are exposed to a 13.56 MHz RF oxygen plasma. Plasma parameters are oxygen (O2 ) gas pressure 20 Torr, plasma power 300 W, and duration of 2.5 min. To realize OH-terminated diamond surfaces, wetchemical oxidation of the initially hydrogen-terminated diamond surface is applied in a mixture of nitric and sulphuric acid (1:3) at 200◦ C for 2 h. Wetting angle experiments of

564

DNA-MODIFIED DIAMOND FILMS

H-terminated surfaces show angles >94◦ indicating strong hydrophobic properties. After plasma oxidation, the wetting angle approaches 0◦ , as the surface becomes hydrophilic. Different electron transfer rates of redox probes were observed on the boron-doped diamond electrodes with different surface terminations. Redox system of Fe(CN)6 3−/4− was chosen because it is recognized as an out-sphere system and is sensitive to the surface terminations of substrates applied. Figure 21.12 shows cyclic voltammograms of Fe(CN)6 3−/4− on (1) H-, (2) OH-, (3) reconstructed carbon, and (4) O-terminated boron-doped diamond electrodes at a scan rate of 100 mV s−1 . The doping level of diamond electrodes is in the range of 5 × 1020 B cm−3 . On H-terminated surface (a), the splitting of the peak potentials is around 75 mV, whereas on OH-terminated surface (b), it increases to 400 mV, further to 1100 mV on reconstructed (c) and O-terminated surface (d), indicating a slowing down electron transfer process. 21.2.3

Diamond Nanotexture and Wire Formation

Several types of diamond nanostructured surfaces have been fabricated using bottomup or top-down approaches. For the bottom–up approach, diamond nanostructures were produced by overgrowing diamond homoepitaxially [85,86] or coating other materials like Si or carbon nanotubes [87–89] with nanodiamond. The most often applied technique is a top-down approach, where diamond is etched with reactive ions in a plasma. Masking materials such as Al, SiO2 , Au, Ni, and Mo can be deposited with help of lithographic techniques [90,91] on the diamond surface to protect selected areas from etching. Another approach uses unintentionally sputtered material from the substrate holder onto the sample during the etching process [91] or deliberately depositing thin metal layers that will act as masking materials [92,93].

(a)

Current density (mA cm–2)

0.5 (b) (c) 0

(d)

–0.5 –0.5

0

0.5

1

Potential (V) vs. Ag/AgCl Figure 21.12 Cyclic voltammograms of 1.0 mM Fe(CN)6 3−/4− in 0.1 M KCl on (a) H-, (b) OH-, (c) reconstructed carbon, and (d) O-terminated boron-doped diamond electrodes at a scan rate of 100 mV s−1 . The doping level is in the range of 5 × 1020 B cm−3 . See color insert.

21.2 DIAMOND TRANSDUCER PROPERTIES

565

The first realization of diamond nanowires using top-down technology was performed in 1997 by Shiomi [94], who demonstrated the formation of porous diamond films from reactive ion etching (RIE) using O2 . Later, in 2000, nanostructured honeycomb films were prepared by etching diamond through a porous anodic alumina mask [95] triggering some activities that are summarized in an article of Shenderova et al . [96]. Growth-induced formation of nanoscaled tubular structures were reported for the first time in 2003 via application of hydrogen microwave plasmas in combination with DC bias potentials [97]. In 2008, Zou et al. [98] and later Babchenko et al. [99] reported the fabrication of nanopillar arrays using self-aligned Au nanoparticles as the etching mask in a bias-assisted RIE system by application of hydrogen/argon plasmas. Recently, etching diamond without any mask in oxygen plasma also results in the generation of mircometer long diamond nanowires (nanograss) [100,101]. We introduce here a novel way [76,77] to generate diamond nanostructures from metallically boron-doped, single crystalline chemical vapor deposition (CVD) diamond by use of a top-down procedure. Briefly, it contains four steps: (1) growth of borondoped diamond, (2) formation of etching masks from diamond or Ni nanoparticles, (3) etching using inductive coupled plasmas (ICP) or reactive ion etching (RIE), and (4) removal of the etching mask by wet-chemical cleaning. The geometrical properties of these nanostructures are dependent on the etching mask as well as on the ICP/RIE parameters. Figure 21.13 shows schematically the approach that we have developed to generate diamond nanotextures. Initially, metallically boron-doped (p-type) diamonds with atomically smooth surfaces are grown by homoepitaxy on Ib diamond substrates, using a microwave-assisted CVD technique [105]. The root-mean-square (RMS) roughness of ˚ Then, an etching mask from diamond nanoparticles the surface is typically about 0.8 A. is deposited using particles of typical 8–10 nm in size [102] dissolved in water [103,104]. The particles are deposited on the diamond plate by immersion into the suspension and subsequent sonication to seed diamond nanoparticles on the diamond surface. For diamond etching, RIE in an O2 (97%)/CF4 (3%) gas mixture is applied for typical times ˚ s−1 . Vertically aligned between 2 and 60 s. The diamond etching rate is about 10 A diamond nanotextures arise where diamond nanoparticles have been deposited. Figure 21.14 shows four 3-D AFM tapping mode images of diamond surface after RIE etching for 5, 10, 20, and 60 s, using diamond nanoparticles as hard mask. Each of these surfaces has been evaluated using Fourier-analysis to determine the geometrical properties. For Fourier-analysis line scans of AFM topographies were used. The y-axis

(a)

(d)

(c)

(b) Plasma

Plasma

Figure 21.13 Schematics of the generation of the diamond nanotextured surface using diamond nanoparticles 8–10 nm in diameter as the etching mask and reactive ion etching. (Reprinted from Ref. 76.) See color insert.

566

DNA-MODIFIED DIAMOND FILMS

(a)

(b)

(c)

(d)

Figure 21.14 AFM images of diamond surface after RIE-etching with a gas mixture of 3% CF 4 and 97% O2 for an etching time of 5 (a), 10 (b), 20 (c), and 60 (d) s, respectively. (Reprinted from Ref. 76.)

shows the amplitude of features and the x -axis shows the periodicity. To average over a large number of lines the graphs were normalized. An etching for 5 s (Figure 21.15a) ˚ in length. The average distance between wires is generates short wires of less than 20 A in the range of 5 nm, varying from 2 to 9 nm. An etching for 10 s (Figure 21.15b) results in wires with diameters of 1–5 nm. They are distributed regularly on the surface with about 11 nm distance in between wires. Etching for longer times results in deterioration of these properties, as can be seen in Figure 21.15c and d. After 20 s of etching, the wires become undefined, varying in shape, height, and periodicity. Finally, after 60 s etching, the wire structure vanishes and a relative smooth surface is generated. To evaluate the interfacial electronic properties of diamond nanotextures, we have performed capacitance (C )–voltage (V ) measurements at 1 kHz in pH 7.4 phosphate buffer. The results are shown in Figure 21.16 as the Mott-Schottky plot (C −2 as a function of potential). We used the Mott-Schottky equation to calculate the built-in potential and the variation of the sensors’ area, given by [106]. 1 2 = 2 C ε0 εr eA2 NB

  kT V − Vbi − e

(21.1)

where ε0 is the dielectric constant, εr = 5.7 (diamond), A is the active sensor area, NB is the boron acceptor density in diamond, Vbi is the built-in potential, k is the Boltzmann

567

21.2 DIAMOND TRANSDUCER PROPERTIES

(a)

(b)

Amplitude (a.u.)

10

0 0

2

4

6

8

10

0

(c)

4

8

12

16

20

(d)

10

0 0

10

20

30

0

2

4

6

8

Spacing (nm)

Figure 21.15 Surface analysis of nanowires fabricated for an etching time of 5 (a), 10 (b), 20 (c), and 60 (d) s, respectively. (Reprinted from Ref. 76.)

constant, T is the temperature (300 K), and e is the elementary charge. In these experiments, the acceptor concentration was fixed (NB = const.) as parts of the same diamond have been used. If NB is constant, variations of slope arise from variations in the active area. Smooth diamond shows a doping concentration of 7 × 1019 cm−3 and a built-in potential of 1.6 V, as the surface is hydrogen terminated. The depletion layer width at ˚ After etching for 5, 10, and 20 s the Mott-Schottky graphs the interface is about 10 A. change significantly. The active sensor area increases by RIE, showing a maximum of 2.1 × A0 after 10 s RIE, where A0 is the smooth surface area. The built-in potential changes as hydrogen is removed from the surface and surface defects give rise to surface Fermi-level pinning. It is interesting that the highest barrier of 3.1 eV is generated by 10 s RIE, whereas shorter (5 s) or longer (20 s) etching times do not change the built-in potential. Diamond nanotextures with optimized geometrical dimensions of 10 nm length and an average separation of 11 nm were obtained with an etching time of 10 s. As the ˚ s−1 , the particles are etched away etching rate of diamond nanoparticles is about 10 A

568

DNA-MODIFIED DIAMOND FILMS

5

Sensor Area (norm.)

[×1013] 6 (a)

C –2 (F–2)

4

3 (c) 2

1

0 –0.5

2.2 2 1.8 1.6 1.4 1.2 1 0

(d)

5 10 15 Etching time (s)

20

(b)

0

0.5

1 1.5 2 E (V vs. Ag/AgCl)

2.5

3

3.5

Figure 21.16 Mott-Schottky plots of a smooth diamond (a) and diamond nanowires for an etching time of 5 s (b), 10 s (c), and 20 s (d) in 0.1 M pH 7.4 phosphate buffer at a fixed frequency of 1.0 kHz. The insert is the sensor area calculated as a function of etching time. (Reprinted from Ref. 76.) See color insert.

after 10 s. The realized texture resembles approximately the geometrical properties of the nanoparticle etching mask, as shown in Figure 21.13. Figure 21.17 shows the production of micrometer long diamond wires using selfaligned Ni particles. Ni is stable in ICPs using oxygen as etching gas. For the formation of Ni particles, a Ni layer of about 1 nm thickness is evaporated onto diamond. Then the nickel layer is thermally annealed in vacuum at 700◦ C for 5 min. Nickel particles of typically 10 nm to 30 nm in diameter were generated and distributed in a homogeneous pattern on the diamond surface. As schematically shown in Figure 21.17, after the formation of Ni particles an oxygen-inductive coupled plasma (ICP) is applied. Typical etching parameters were oxygen gas flow 20 sccm, pressure 5 × 10−2 mbar, duration 5 min, power 1000 W. After the ICP process, residual Ni particles are removed from the surface by wet-chemical cleaning in aqua-regia. The length of nanowires and the spacing in between can be controlled by the size of Ni particles and the etching time. Wire diameters of 10 nm to 60 nm can be realized. The lengths can be varied from a few nanometers to several microns, depending on the thickness of wires and the etching time. After ICP treatment, the resulting diamond surface shows wires with typical diameters of (23 ± 7) nm (see Figure 21.18a). The density is about 1011 cm−2 , which corresponds to the size and distribution of Ni particles. The height or length of wires is measured to be (1200 ± 200) nm (see Figure 21.18b and c) and the aspect ratio is in the range of 50. Calculations of the surface enlargement, assuming cylindrically shaped wires reveal a surface enlargement of 10–80. This is in agreement with experimentally obtained data of the surface enlargement using capacitance-voltage measurements, which will be discussed later. Note that narrow-standing diamond wires are cylindrical in shape, whereas single, stand-alone wires are conical in shape, as can be seen in Figure 21.18b and c. This effect

569

21.2 DIAMOND TRANSDUCER PROPERTIES

Figure 21.17 The schematics of micrometered long diamond nanowire using thermally self-formed Ni nanoparticles as etching mask and using oxygen ICP plasma. The Ni nanoparticles were formed by evaporating a nanometer thick Ni layer and further annealing the sample at 700◦ C for 5 min. (Reprinted from Ref. 77.)

(a)

(b)

(c)

100 µm

500 µm

Figure 21.18 SEM pictograms of diamond nanowires in a top view (a) and in a view (b) taken at a 45◦ tilt angle with respect to sample surface. (c) Higher magnification of individual nanowires. (Reprinted from Ref. 77.)

may arise from plasma shielding during etching due to charging of the wires and reflected ions at wires [107]. To measure the surface enlargement, we have applied cyclic voltammetry experiments on our structured electrodes. Figure 21.19a compares the samples obtained from the same diamond substrate without (solid line) and with nanowires (dashed line). Dimensions of nanowires were height (1200 ± 200) nm, width (35 ± 5) nm, and density ∼ 1010 cm−2 . In order to verify the surface enhancement of diamond nanowires, the background currents were measured in 0.1 M KCl at a scan rate of 50 mV s−1 . The background current is proportional to the surface area in contact with the electrolyte solution and given by [108]: IC = Cn · AGeom · ν

(21.2)

where Ic is the current related to the charging of the interfacial capacitance Cn given in units of (F cm−2 ) and AGeom is the total surface area. We assume that the capacitance per unit area is constant and not affected by the etching and formation of the wires, as the diamond substrate has not been changed (boron doping is the same). We assume that the depletion layer width at the surface of diamond in buffer solution is therefore the same. The background current on diamond nanowires is more than 10 times larger than

570

DNA-MODIFIED DIAMOND FILMS

Current density (μA cm–2)

(a)

10

0

–10

–0.2

0

0.2

0.4

0.6

Potential (V) vs Ag/AgCl (b)

Current density (μA cm–2)

10

5

0

–5

–10 –0.2

0

0.2

0.4

0.6

Potential (V) vs Ag/AgCl

Figure 21.19 Electrochemical experiments on smooth diamond surface (solid lines) and on diamond nanowires (dashed lines). (a) Background currents measured by cyclic voltammetry in 0.1 M KCl at a scan rate of 50 mV s−1 . (b) Redox currents measured by cyclic voltammetry using 1 mM Fe(CN)6 3−/4− in 0.1 M KCl at a scan rate of 100 mV s−1 . (Reprinted from Ref. 77.)

on the smooth diamond electrode, as shown in Figure 21.19a. This is in agreement with geometrical calculations as previously discussed. The electrochemical activity of diamond nanowires was characterized using redox couples of Fe(CN)6 3−/4− as probes. Figure 21.19b shows cyclic voltammograms of 1.0 mM Fe(CN)6 3−/4− at a scan rate of 100 mV s−1 . The peak splitting of oxidation and reduction currents (potential of the anodic peak minus potential of cathodic peak) is (80 ± 10) mV. It is close to the Nernst limit, which is a result of metallically ultra-high doping. The enhancement in current density is, however, only 1.5. For a diffusion-controlled process, the peak current (IP ) is proportional to the square root of the scan rate (ν), to the diffusion coefficient (D) and to the electrochemically active area (AEl Chem ), as given by Equation (21.3) [108]: IP ∼ AEl Chem · D 1/2 ν 1/2

(21.3)

21.3 SURFACE MODIFICATION OF DIAMOND

571

For comparison of smooth diamond and diamond nanowires we assume again that the only variable parameter is the area AEl Chem . Hence, the ratio of peak current slopes (Ip plotted against ν 1/2 , not shown here) can be used to calculate the surface enlargement, which is only 1.5. This is much lower than the real increase of surface area and attributed to the following: The capacitive background current in KCl is proportional to the total surface area in contact with electrolyte solution. On the other hand, the Faradaic current of ferri/ferrocyanide is dominated by their diffusion length, which is typically in the range of tens of microns. Structures that are significantly smaller than diffusion length are not recognized and the Faradaic current of ferri/ferrocyanide is therefore about the same as on a smooth surface. This diamond nanowire electrode does not behave as a microelectrode ensemble but rather as a macroelectrode. As a result, the electrochemically active area is significantly smaller than the total surface area. To evaluate the surface enlargement truly by this method requires redox molecules, which either drift in an electric field and/or show a decreased electrochemical activity like hydrogen peroxide [109]. 21.3 21.3.1

SURFACE MODIFICATION OF DIAMOND Photochemical Surface Modification of Intrinsic Diamond

Undoped single crystalline diamond surfaces are modified by photochemical reactions with 10-amino-dec-1-ene molecules protected with trifluoroacetic acid group (TFAAD) [71–73,110–113]. The chemical attachment is characterized using an x-ray photoelectron spectroscopy (XPS) system (Tetra Probe, Thermo VG Scientific) with a monochromatized AlKα source (1486.6 eV) at a base pressure of 10−10 Torr. Unless otherwise noted, electrons are ejected between 25◦ and 80◦ with respect to the surface normal (atomic sensitivity factors: C, 0.296; F, 1; N, 0.477; O, 0.711). The mean free path of electrons ˚ for perpendicular excitation. Microscopic morphology and structural properties is 36 A of amine layers have been characterized by atomic force microscopy (AFM) and scanning tunneling microscopy (STM) (Molecular Imaging PicoPlus) [110–113]. The STM experiments are performed in air using Pt0.8 Ir0.2 tips with typical tunneling currents in the range of 50–100 pA to detect TFAAD grafted surface areas on diamond. For these currents the typical potential to STM tip was +0.4 V. Restricted Hartree-Fock calculation of theoretical geometric properties of TFAAD molecules were performed with the Gaussian 98 package with density functional theory (B3LYP/6-31G(d)) [114]. A typical result is shown in Figure 21.20. The length of the ˚ and the diameter is 5.01 A. ˚ It is interesting to note that the protecting molecule is 11.23 A, cap molecule shows a tilted arrangement. In case of an upright arrangement of molecules ˚ on diamond, one therefore expects a monolayer thickness of around 11–15 A. The chemically reactive end of TFAAD is terminated with an olefin (C=C); the other is protected from reactions using a trifluoroacetic cap. Chemical attachment is accomplished by placing 4 μl of TFAAD on the diamond substrate. The TFAAD is then homogeneously distributed by spin-coating with 4000 rounds min−1 in air for 20 s which forms a 5 μm thick liquid TFAAD layer. After accomplishment of spin-coating, samples are sealed into a chamber with a quartz window in nitrogen atmosphere. Then UV illumination is switched on for a given period of time. The ultraviolet light is generated in a high-pressure mercury lamp with emission at 250 nm of 10 mW cm−2 intensity. The H-terminated samples photochemically reacted with TFAAD. A variety of attachment experiments on H-terminated and oxidize diamond surfaces show that this process

572

DNA-MODIFIED DIAMOND FILMS

Figure 21.20 10-amino-dec-1-ene molecule protected with trifluoroacetic acid group (TFAAD), as determined by molecular orbital calculations [114]. (Reprinted from Ref. 21.)

works only on H-terminated diamond. To characterize the molecule arrangement and layer formation of TFAAD on hydrogen-terminated CVD diamond, we applied AFM scratching experiments with the aim to (1) identify the threshold force for mechanical removal of TFAAD molecules and (2) use the step-edge of scratched areas to measure the thickness of the TFAAD layers. Please note that diamond is ultrahard (100 GPa hardness). The AFM cantilever will therefore not penetrate diamond even at the highest applied loading forces. Such contact-mode AFM experiments carve out rectangular trenches in the amine layer. AFM tapping-mode line scans are then used to measure the height profile across scratched areas. Forces below 100 nN remove only a fraction of TFAAD molecules whereas forces equal to or larger than 100 nN give rise to complete removal (for details, see Ref. 110). A typical result of AFM characterization applied to an amine layer that has been photochemically attached for 1.5 h is shown in Figure 21.21. The TFAAD molecules were removed by 200 nN applied loading force from an area of 2 × 2 μm2 to truly remove the amine layer. The upper image shows topography and the lower image shows a typical line profile. Within the cleaned area, the surface roughness of diamond is in the ˚ On the amine film we detect a roughness between 3.8 and 9.3 A. ˚ range of 1.2–2.3 A. ˚ which indicates The TFAAD-layer thickness is determined to be in the range of 5–13 A, dispersed layer properties, as for a closely packed, dense layer the height is expected to ˚ be around 11 A.

21.3 SURFACE MODIFICATION OF DIAMOND

573

40

Height (Å)

32

24

16 0

1000

2000

3000

Distance (nm) Figure 21.21 AFM topography and line scan of a TFAAD film attached by 1.5 h of illumination to diamond. The scratched area is 2 × 2 μm2 and the loading force was 200 nN. (Reprinted from Ref. 21.) See color insert.

To investigate the chemical bonding of TFAAD molecules to diamond, x-ray photoelectron spectroscopic measurements (XPS) have been applied. Figure 21.22a shows XPS survey spectra of a clean hydrogen-terminated single crystalline diamond surface before and after exposed to TFAAD and 10 mW cm−2 UV illumination intensity (254 nm) for 2 h. Before XPS measurements, samples were rinsed in chloroform and methanol (each 5 min in ultrasonic). The overall spectrum shows a strong fluorine peak with a binding energy of 689 eV, an O(1s) peak at 531 eV, a N(1s) peak at 400 eV, and a large C(1s) bulk peak at 284.5 eV. Note that on clean H-terminated diamond, no oxygen peak can be detected. The C(1s) spectrum reveals two additional small peaks at 292.9 eV and 288.5 eV (see Figure 21.22b); they are attributed to carbon atoms in the CF3 cap group and in the C=O group, respectively. From these experiments, we conclude that UV light of about 250 nm (5 eV) initiates the attachment of TFAAD to H-terminated single crystalline diamond. The ratio of the F(1s) XPS signal (peak area) to that of the total C(1s) signal (RFC ) as a function illumination time is shown in Figure 21.22c. The time dependence of RFC follows approximately an exponential law: RFC = A{1-exp(–t/τ )},

574

DNA-MODIFIED DIAMOND FILMS

1.2 C(1s) (284.5 eV)

4 x 104

(a) F(1s) (689 eV)

3 x 104

O(1s) (531 eV) N(1s) (400 eV)

2 x 104 1 x 104

2h illuminated

0 300

400

500 600 Energy (eV)

700

0.8 TFAAD-modified

0.6 Diamond 0.4

C(1s) (284.5 eV)

0.2 0.0 275

800

H-terminated Diamond C(1s) (284.5 eV)

(b)

1.0 Counts (a.u.)

Count (1/s)

5 x 104

280

C=O (288.5 eV) CF3 (292.9 eV)

290 285 Energy (eV)

295

(c)

0.8 0.6 0.4 0.2

Time Constant τ = 1.7 h

0.0 0

2 4 6 Illumination Time (h)

8

F(1s)/C(1s) Ratio

F9(1s)/C(1s) (norm.)

0.6 1.0

(d) 0.5 0.4 0.3 0.2 20

30 40 50 60 70 80 Angle to Surface Normal (°)

Figure 21.22 (a) XPS survey spectrum of a hydrogen-terminated single crystalline diamond surface that was exposed to TFAAD and 20 mW cm−2 UV illumination (250 nm) for 2 h. (b) The C(1s) spectrum reveals two additional small peaks at 292.9 eV and 288.5 eV, both of which are attributed to carbon atoms in the CF 3 cap group and in the C=O group, respectively. (c) The ratio of the F(1s) signal (peak area) to that of the total C(1s) signal as a function illumination time is time dependent, follows an exponential increase (dashed line), with a characteristic time constant τ of 1.7 h. (d) Angle resolved (with respect to the surface normal) XPS experiments show an increase of the F(1s)/C(1s) peak intensities, rising from 48◦ to 78◦ . (Reprinted from Ref. 21.)

with a characteristic time constant τ of 1.7 h. Saturation of the area ratio F(1s)/C(1s) is achieved after about 7 h. Angle-resolved XPS experiments shown in Figure 21.22d are used to calculate the density of bonded TFFAD molecules to diamond in absolute units. As parameters, we used the following data: density of carbon atoms = 1.77 × 1023 atoms cm−3 , atomic sensitivity factors for C (0.296), and for F (1), and a mean free path ˚ Taking into account the area ratio F(1s)/C(1s) for perpendicular illumination = 36.7 A. results in about 2 × 1015 cm−2 TFAAD molecules bonded after 7 h. This corresponds to the formation of a monolayer TFAAD as the surface density of carbon bonds on diamond is 1.5 × 1015 cm−2 . However, the TFFAD layer itself consists of 12 carbon atoms that contribute to the signal, so that the real coverage is smaller than this number, as will be discussed later. The variation of film properties as a function of UV-photochemical attachment times is shown in Figure 21.23. For times shorter than 4 h, the average layer thickness increased ˚ (4 h), which is about the length of upright from zero (0 h) to 9.25 (1.5 h) and 13.3 A standing TFAAD molecules on diamond. Illumination time between 4 and 10 h does not ˚ result in a remarkable enlargement of the layer thickness, which saturates around 14 A, but the average roughness decreases. These films are closed with no structural defects like pinholes in it. Illumination longer than 10 h then gives rise to further enlargement: ˚ after 20 h, which is about three times The film becomes thicker—for example, 31.2 A

21.3 SURFACE MODIFICATION OF DIAMOND

45

575

2

40 1.5

30 25 1 20

I(F)/I(C)

Thickness (Å)

35

15 0.5

10 5 0

0

4

8

12

16

20

0

Illumination Time (h) Figure 21.23 Variation of TFAAD film thickness as detected by AFM (circles) as a function of illumination time and the variation of the integrated peak intensity ratio F(1s)/C(1s) (squares) from XPS experiments as a function of illumination time. Each averaged thickness was measured at four different scratched areas. Note that the indicated bars represent the width of height variations. Triangles show the onset of cross-polymerization of TFAAD molecules. (Reprinted from Ref. 21.)

the length of TFAAD molecules. Notice that the bars in Figure 21.23 indicate an average roughness of films as measured by tapping-mode AFM on the layer. After short-term ˚ but the layers attached for attachment (<10 h) the roughness is large (about ±4.0 A), 10–12 h are relatively smooth. Longer attachment times again give rise to enlarged ˚ after 20 h). The dashed red line is a calculated exponential fit to roughness (±8.0 A the data, which reveals a time constant of attachment of 1.7 h in our experiments. For photochemical attachment times shorter than 8 h, the average layer thickness increased ˚ (8 h), which is about the length of the TFAAD molecule. from zero (0 h) to 15 A For longer times (>12 h) cross-polymerization takes place, which gives rise to further ˚ h−1 (indicated by the black dashed line growth of the layer with a speed of about 1.5 A in Figure 21.23). We attribute this to (1) 2D-formation of a dispersed submolecular layer for times <10 h, (2) 2D-formation of a dense monomolecular layer for times between 10 and 12 h, and (3) cross-polymerization and 3D growth for times longer than 12 h. A perfectly smooth surface of (100) oriented diamond contains about 1.5 × 1015 cm−2 carbon bonds that are terminated by hydrogen, as shown schematically in Figure 21.24 [115]. The diameter of the 10-amino-dec-1-ene molecules (not taking into account the ˚ Thus, in the case of a closely packed TFAAD film trifluoroacetic acid top) is 5.01 A. the upright standing molecule will require an area of about 2 × 10−15 cm−2 (assuming rotational symmetry). This gives a closed layer density of 5 × 1014 cm−2 . Each TFAAD molecule covers the area of six hydrogen atoms. Only one of those hydrogen bonds needs to be broken to bond an amine molecule. Therefore, most of the diamond surface may be still H-terminated after generation of a densely packed TFAAD layer. Because hydrogen cannot be detected by XPS, we applied an additional experiment to investigate if the surface is still terminated with hydrogen. It is known that H-terminated undoped CVD diamond shows a surface conductivity in buffer solution that arises from transfer doping [29,34,35]. We therefore measured the drain-source current variation before and after TFAAD attachment using a typical ISFET geometrical

576

DNA-MODIFIED DIAMOND FILMS

Figure 21.24 Side and top view of a reconstructed diamond (100)(2 × 1):1H surface (data from Ref. ˚ as shown in 115). Also shown are typical areas of TFAAD molecules, assuming a diameter of 5.04 A, Figure 21.20. (Reprinted from Ref. 21.) See color insert.

arrangement in SSPE buffer solution. Figure 21.25 shows the conductivity of a perfect H-terminated single crystal diamond before (open circles) and after (open squares) photochemical attachment of TFAAD for 20 h in SSPE buffer solution. A perfectly (100%) H-terminated diamond gives rise to a drain-source current of approximately 7.5 μA at UG = −0.6 V. After photoattachment of TFAAD molecules for 20 h, the drain-source

Figure 21.25 Comparison of diamond ion-sensitive field effect transistor properties (ISFET) measured in SSPE buffer with perfecty H-termination of the surface (circles) and after photoattachment of amine molecules for 20 h (squares). The drain-source current is decreasing for about 60%, but the surface of diamond remains conductive after photoattachment. (Reprinted from Ref. 21.)

21.3 SURFACE MODIFICATION OF DIAMOND

577

current decreased to 3 μA, which is 40% of the initial drain-source conductivity. In the case of 100% H-removal by photoattachment, the drain-source conductivity would completely disappear. As the wetting properties of TFAAD covered diamond surfaces are also changing, which may affect transfer-doping properties of the surface, we leave a quantitative interpretation of this result to further research activities. The highest theoretical packing density of TFAAD of 5 × 1014 cm−2 would require the removal of only 17% of all hydrogen atoms terminating the surface. Following recently published data in the literature, the packing density is most likely less than this number and in the range of 2 × 1014 cm−2 [73,110,116,117]. Our experimental data as well as these theoretical considerations indicate that the diamond TFAAD interface is still reasonably well terminated with hydrogen. This is promising, with respect to sensor applications, because high sensitivity will require a defect-free interface. Detailed characterization of the attachment process indicates that electron emission by subbandgap light triggers the covalent bonding of TFAAD molecules to diamond [72,73]. In this process, valence-band electrons are optically excited into empty hydrogen-induced states slightly above the vacuum level, as shown schematically in Figure 21.26 [30]. From there, they can reach unoccupied π * states of TFAAD molecules, generating a nucleophilic situation in the C=C bonding structure.

Energy ECBM χ = –1.1 eV

π∗-states

7.2 eV

5.47 eV

EVAC

hv

EVBM

π-states

Diamond

x

Liquid-phase alkene

Figure 21.26 Schematic diagram of the photoexcitation mechanism at the surface of diamond in contact with TFAAD (from Refs. 72,73). Valence-band electrons are photoexcited into empty surface states of diamond and then into empty electronic states of TFAAD molecules, which generate nucleophilic properties. (Reprinted from Ref. 21.)

578

DNA-MODIFIED DIAMOND FILMS

To investigate the initial growth of the molecular layer, we applied a combination of AFM experiments either in tapping mode (topography) or in contact mode (scratching experiments) to remove the layer from the surface (contact mode) and, subsequently, to measure the height of the layer by tapping mode (Figure 21.21). A typical surface topography, as detected after 1.5 h of illumination time, shows nanoscopic islands or a “tip” structure (see details in Ref. 113). The size of islands can be determined by evaluation of the base diameter of each tip that is varying between 15 and 40 nm. To reveal the underlying growth mechanism at the early stage of growth, we applied STM experiments on films grown for 10 and 40 min on diamond. The variation of tunneling currents can be used to reveal the amine molecular layer formation on diamond. Figure 21.27 shows two STM images as detected after 10 and 40 min, respectively. The red areas correspond to amine modified surface parts (I), the blue areas are bare diamond (II), and the green refers to a kind of transition state (III) where STM data cannot clearly been attributed either to state (I) or (II). After 10 min of attachment, small but densely grafted spots appear (Figure 21.27a), which cover about 10% of the surface. After 40 min of light exposure these spots become larger, forming closely grafted surface areas that cover between 25 and 35% of the surface (Figure 21.27b). If we take into account exponential growth with a time constant of 1.7 h, one expects 10% coverage after 10 min and 33% after 40 min. These data are therefore in reasonable agreement with data shown in Figure 21.23. Obviously, islands grow at the periphery and become larger with time. Both AFM and STM experiments applied on short-time photoattached amine layers to diamond clearly reveal island formation. On hydrogen-terminated silicon surfaces, Lopinski, Wayner, and Wolkow [118] and later Cicero et al. [119] showed that the addition of styrene molecules follows a chain reaction mechanism with self-directed growth. The model assumes a three-step process: (1) Hydrogen abstraction from silicon takes place to generate a silicon dangling-bond at the surface. (2) Bonding of styrene molecules to silicon dangling bonds follows, thereby forming a carbon-centered radical. ˚ apart Thus, create a (3) This radical abstracts hydrogen from an adjacent dimer 3.8 A

(

( (a)

(b)

Figure 21.27 Typical STM images of 200 × 200 A˚ areas as detected on diamond (a) after 10 min and (b) after 40 min of attachment. Red areas indicate amine modified diamond surface; blue areas are bare diamond. Green represents a nondefined transition level. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article). (Reprinted from Ref. 113.) See color insert.

21.3 SURFACE MODIFICATION OF DIAMOND

579

new Si dangling-bond in close vicinity that acts as a bonding site for the next molecule attachment. It is a chain reaction that proceeds along the dimer row on (100)(2 × 1):H silicon or randomly on H–Si(111) [118,119]. On diamond we attached long-chain amine molecules that have the same reactive end like styrene molecules (H2 C=CH–R) and that also show island formation. The structure of an alkene molecule, as calculated by molecular orbital calculations (Gausssian98, B3LYP with 6-31G (d) basis set) is shown in Figure 21.28 [110,114]. It shows a cylin˚ and a length of 14.7 A. ˚ The geometric drical structure with a diameter of about 5 A properties of a (100)(2 × 1):H reconstructed diamond surface is also shown [120]. The carbon lattice is significantly denser than the Si equivalent. Hydrogen abstraction reaction, ˚ to which is required for a chain reaction, will need to occur over a distance of about 5 A allow accommodation of a TFAAD molecule in close vicinity (see Figure 21.28). Several possible sites are available, as indicated by blue/red-colored atoms in Figure 21.28, that will allow self-defined nondirectional growth. Surface electronic characterizations after deposition of dense TFAAD monomolecular layers (12 h attachment) indicate that only a small fraction of hydrogen has been removed from the surface during the attachment [8]. Obviously, most of hydrogen atoms bonded to diamond is protected against abstraction by the large size of TFAAD molecule (see Figure 21.28). To initiate chain reactions, carbon dangling bonds on the surface of diamond are required.

Figure 21.28 Geometry of TFAAD (left) and top and side view of (100)(2 × 1):H reconstructed ˚ (Reprinted from Ref. 113.) See color insert. diamond surface (all numbers in A).

580

DNA-MODIFIED DIAMOND FILMS

1. Diamond surfaces may show residual dangling bonds, as perfect reconstruction may not be achievable. This has been recognized with respect to Schottky contact formation of different metals on diamond. The Schottky contact barrier is nearly invariant, indicating surface defects that pin the Fermi level at the interface [121]. 2. There are single atomic step-edge terraces present [122] where growth can be initiated at step-edge defects. 3. UV light illumination itself may give rise to hydrogen abstraction. The binding energy of hydrogen to carbon (C–H) is 4.16 eV, which is slightly smaller than the energy of photons used in these photochemical attachment experiments. These have energies between 4.8 and 5.3 eV (for details, see Ref. 73, 116, 117, 123). However, no experimental evidence for this is currently available (unpublished data). On the contrary, we have applied surface defect-sensitive experiments like total photo-yield spectroscopy (TPYS) that do not reveal any defect formation by UV-light illumination. 4. Nucleophilic amine molecules may cleave hydrogen from the diamond surface. This seems to be the most reasonable model for the initiation of the growth, as electrons are indeed photochemically transferred into amine molecules, which should result in nucleophilic and therefore very reactive molecules. There is, however, a significant difference between transferred photoexcited electrons and amine molecules, which are bonding to the surface of diamond (see Ref. 73, 116, 117, 123). We have shown that photoattachment is a weak function of photoelectron excitation, as only one photo-excited electron (= amine radical) out of 1500 electrons (= amine radicals) gives rise to amine attachment to diamond [11]. This may be attributed to the fact that most of these reactive molecules form cluster in the liquid state as discussed by Nichols et al. [116] (see also Ref. 73, 117, 123) while only a small fraction bonds to the surface. The details of the “nucleophilic” bond-breaking mechanism—its dynamic properties and involved bond rearrangements—are up to now not clear. This is the case for Si but also for diamond. Further experiments are clearly required to elucidate the initiation of the chain reaction on diamond and the related chemical rearrangements. These radical anions may abstract hydrogen from the surface, as shown schematically in Figure 21.29, creating a carbon dangling bond at the surface which itself is very reactive toward olefins. To obtain covalently bonded ordered monomolecular layers on diamond requires, however, some additional features. In case of random self-assembly of molecules on diamond, a chaotically organized layer would be generated as the basic requirement of surface mobility to allow intermolecular forces to play their ordering role is missing. Such a process would resembles a “dart game,” since the grafted moieties would be irreversibly immobilized on the surface due to the formation of strong covalent C–C bonds. In our case, ordered formation of monolayers are detected, which is a strong argument in favor of the model of Cicero et al. [119]. In their investigation of olefin addition to H-terminated silicon, they deduced a model that required the formation of alkene anions that abstract H atoms from the surface, thereby creating carbon surface-danglingbonds that covalently bond to other alkene molecules in the liquid. Surface-dangling bonds are reactive toward alkenes as demonstrated by Cicero et al. [119]. The authors showed that the olefin addition on H-terminated Si (111) surfaces follows a chain reaction, initiated at isolated Si dangling bonds. This surface-dangling bonds further react with olefins

21.3 SURFACE MODIFICATION OF DIAMOND

581

Figure 21.29 Schematic grafting mechanism of the diamond surface by amines (from Refs. 110,119). (i) (a) The generation of radical anions by electron transfer from diamond to the olefin is shown. The nucleophilic properties of the radical cause a hydrogen abstraction (b), which results in a surface carbon-dangling bond. (ii) The dangling bond reacts with an olefin molecule (c) to form a diamondcarbon/olefin-carbon bond. This olefin abstracts a hydrogen atom from the diamond surface (d), which is a new site for olefin addition. (iii) The hydrogen abstraction reaction results in a chain reaction. (iv) In the case of extended illumination (>10 h) a 3D growth sets in due to cross-polymerization of olefin molecules. (Reprinted from Ref. 21.)

to form carbon-centered radicals bonded to diamond. These radicals abstract hydrogen from neighboring H–C bonds, thereby regenerating surface-dangling bonds which then propagate the reaction as depicted schematically in Figure 21.29. Such an ordered and self-organized layer formation seems to be a very reasonable model for our findings. The surface of grafted diamonds after 10 h of attachment appear closed with no pinholes. Some modulation, however, is obvious, which may reflect the fact that the ordered growth will break down and compete with other spots where the same process has been triggered. Thus, one can expect some disorder due to competing domain growth. The number of electrons photo-excited from diamond into the olefin film is huge as only one out of about 1500 triggers chemical bonding to diamond [73]. These electrons create radical anions that give rise to cross-polymerization and particle formation [116]. The cross-polymerization on the monomolecular amine layer bonded to diamond is slow, however, and may arise from the fact that the amine layer on diamond prevents the generation of radical anions as TFAAD molecules cannot approach the surface for a electron transition. The radical anions are generated predominantly on diamond surface areas that are not grafted. These radicals then need to diffuse, driven by statistical properties, to find reaction partners and to cross-polymerize. To bond to immobilize TFAAD molecules seems less likely than to react with other molecules in the near vicinity in liquid phase.

582

DNA-MODIFIED DIAMOND FILMS

21.3.2 Electrochemical Surface Functionalization of Boron-Doped Diamond Electrochemically induced covalent attachment of nitrophenyl molecules has been performed using an Electrochemical Analyzer 900 (CHI instruments), and a three-electrode configuration with a platinum counter electrode and an Ag/Ag+ (0.01 M) reference electrode (BAS, Japan) [74,75]. The active area of the boron-doped diamond working electrode is about 0.03 cm2 . Electrolyte solution for the reduction of 1.0 mM 4-nitrobenzene diazonium tetrafluoroborate is 0.1 M tetrabutylammonium tetrafluoroborate (NBu4 BF4 ) in dehydrated acetonitrile (Wako chemicals, H2 O:<50 ppm). The diazonium salts reduction is performed in a N2 -purged glove box. Nitrophenyl-modified diamond surfaces are then sonicated with acetone and acetonitrile. XPS, AFM, STM, and voltammetric experiments have been applied to characterize the surface bonding properties and to reduce the nitrophenyl groups to aminophenyl groups. The nitrophenyl groups grafted on single crystalline diamond substrate can be considered as covalently bonded-free nitrobenzene to diamond, as shown in Figure 21.30 (Molecular Orbital Calculations). In the following we summarize the electrochemical modification of highly conductive p-type single crystalline CVD diamond [74,75,124]. Figure 21.31 shows cyclic voltammograms of 4-nitrobenzene diazonium salts (1 mM) reactions on highly B-doped single crystalline CVD diamond film with different surface terminations in 0.1 M NBu4 BF4 acetonitrile solution at a scan rate of 0.2 V s−1 . The electrochemical reduction of diazonium salts on the H-terminated diamond surface (curve a) and on OH-terminated surface (curve b) show the same cathodic peak potentials at −0.08 V with respect to Ag/Ag+ . An irreversible cathodic peak of the first sweep at −0.08 V (vs. Ag/AgCl) indicates nitrophenyl group attachment by diazonium salt reduction [74,75,124]. The reduction peak on single crystalline diamond films decreases rapidly with increasing number of scans within +0.5 to −1.0 V (vs. Ag/AgCl), due to increasing surface passivation with nitrophenyl molecules. The amplitude detected on the H-terminated diamond surface is about three times larger than that detected on the OH-terminated diamond surface. Phenyl addition on reconstructed (curve c) carbon and on O-terminated (curve d) surfaces show a peak which is shifted significantly toward more negative potentials of −0.34 and −0.27 V, respectively, and with significantly smaller amplitudes. The low reduction potential and the larger cathodic peak current (curve a) on H-terminated diamond indicate that the reactivity of diazonium salts is strongest here. It is interesting to note the peaks of

(a) Front view

(b) Side view

Figure 21.30 Front (a) and side view (b) of a nitrophenyl molecule, as calculated by molecular orbital calculations [114]. (Reprinted from Ref. 21.)

21.3 SURFACE MODIFICATION OF DIAMOND

583

Current density (mA cm–2)

0

(d) –0.2

(b) (c)

–0.4

–0.6

(a) –1

–0.5 0 Potential (V) vs. Ag/Ag+

0.5

Figure 21.31 Cyclic voltammograms of electrochemical reduction of 1.0 mM 4-nitrobenzene diazonium tetrafluoroborate on diamond with (a) H-, (b) OH-, (c) reconstructed carbon, and (d) O-termination measured at a scan rate of 0.2 V s−1 . The electrolyte solution was 0.1 M mtetrabutylammonium tetrafluoroborate in dehydrated acetonitrile. (Reprinted from Ref. 21.)

H- and OH-terminated diamond are very close as well as peaks of carbon reconstructed and O-terminated diamond. Compared to H-termination, the reaction on OH-terminated diamond is about three times smaller in amplitude with otherwise unchanged potential (curve b). It is known that wet-chemical oxidation replaces hydrogen by OH groups. However, it is not clear how much hydrogen is replaced by OH. As nitrophenyl is attached with the same potential on both samples, we conclude that this is either a result of residual H-termination that is remaining after wet chemical etching, or the difference between binding energies of C–H and C–OH is too small (see Table 21.1) to cause a significant shift of the attachment potential. Attachment experiments on carbon reconstructed (C=C) and on O-terminated surfaces (C=O) show peak shifts toward more negative potentials as well as a significant decrease in current (charge). This is a result of the change of bonding energies as here C–O–C and C=C bonds need to be broken by the nitrophenyl attachment to the surface. The bonding energy of C=C is 145 kcal mol−1 and C–O–C is 187 kcal mol−1 [125]. To measure the amount of nitrophenyl groups bonded to diamond, we evaluated the charge consumed during attachment. Unfortunately, the charge does not accurately represent the number of nitrophenyl molecules attached to the surface. Because the electrochemical reduction of diazonium salt is a fast reaction, a fraction of phenyl molecules TABLE 21.1 Binding energies of O, OH and H to carbon. Structure C—H C=O C=C C—OH

Energy (kcal mol−1 ) 98 187 145 78

584

DNA-MODIFIED DIAMOND FILMS

will not bond but diffuse back to the bulk solution [124]. One reliable way to evaluate the density of nitrophenyl molecules bonded to diamond is the evaluation of redox activity of nitrophenyl molecules [74,124]. Then after electrochemical derivatization, the diamond substrates are sonificated in acetonitrile, acetone, and isopropanol in order to investigate properties of attached nitrophenyl layers in 0.1 M NBu4 BF4 solution. Nitrophenyl groups grafted on single crystalline diamond films show two reversible electron transfer steps that are reproducibly detected in all potential sweeps (see Figure 21.32). The generalized reversible redox reactions of this system are summarized in Figure 21.33. To estimate the surface coverage () of nitrophenyl groups on diamond, we use the transferred charge of the electron transfer reaction at −1.17 V (see shaded area in Figure 21.32). This results in 3.8 × 10−7 C cm−2 or 8 × 1013 molecules cm−2 . This value is smaller than the calculated surface concentration of a close-packed nitrophenyl monolayer [126,127], indicating the formation of a submonolayer on diamond [74,124]. If one assumes that a H-terminated (100) oriented surface of diamond consists of about 2 × 1015 cm−2 carbon bonds, which is the density of a perfectly terminated surface with hydrogen, one may calculate form this charge that only 4% of hydrogen atoms are replaced by nitrophenyl

10

Current density (μA/cm–2)

Q 0

–10

–20

–30 –2.0

–1.5

–1.0

–0.5

–0.0

Potential (V) vs. Ag/AgCl Figure 21.32 Cyclic voltammograms from 4-nitrophenyl modified single crystalline diamond in blank electrolyte solution. Electrolyte solution: 0.1 M NBu4 BF 4 in CH3 CN. For details, see Ref. 75. (Reprinted from Ref. 21.)

Figure 21.33 Schematic reduction/oxidation reactions of nitrophenyl bonded to diamond, giving rise to a two electron transfer reaction mechanism [114]. (Reprinted from Ref. 21.)

21.3 SURFACE MODIFICATION OF DIAMOND

585

˚ and molecules. Please note that the diameter of a nitrophenyl molecules is about 7.8 A ˚ Therefore, the forthe distance between H atoms on (100)-oriented diamond is 1.5 A. mation of a close-packed nitrophenyl layer will replace a maximum of 19% of hydrogen atoms on the surface of diamond. Combining these results with the charges consumed during the attachment shown in Figure 21.31, one can conclude that diluted monolayers of nitrophenyl have been formed on H, OH, and reconstructed carbon-terminated surface. Subsequently, nitrophenyl groups are electrochemically reduced to aminophenyl (–C6 H5 NH2 ) in 0.1 M KCl solution of EtOH-H2 O solvent (see Figure 21.34). As the potential becomes more negative from −0.5 down to −1.5 V (vs. Ag/AgCl), the nitro groups are reduced to amino groups at potentials, leading to a broad negative-going (reduction) peak. The first voltammetry sweep gives rise to an irreversible reduction peak at −0.94 (vs. SCE), which is not detected in the second and higher cyclic voltammetry cycles. Since the grafted nitrophenyl layer can be electrochemically reduced into aminophenyl films with the involvement of six electrons, we used the charge involved in the electrochemical conversion to calculate the true density of nitrophenyl molecules on the diamond surface with different surface terminations [48,74,124,128]. Figure 21.35 shows linear sweep voltammograms for electrochemical reduction of nitrophenyl layers attached on different surfaces. Broad negative cathodic waves with different peak potentials and currents are detected on all four samples, representing the electrochemical reduction of nitroto amino-phenyl. On the H-terminated diamond surface, the charge is largest, decreasing to about 70% of this value on the OH-terminated surface, and further down to 50% on oxygen-terminated surface. Several cycles were applied to make sure all nitrophenyl molecules were changing to aminophenyl. The charges consumed during these reductions were calculated and are

Current density (mA cm–2)

0.1

0.0

–0.1

–0.2 Reduction of NO2 to NH2 –0.3 –1.5

–1.0

–0.5

0.0

0.5

1.0

Potential (V) vs. SCE Figure 21.34 The nitro-groups (NO2 ) are electrochemically changed to amino-groups (NH2 ) by cyclic voltammograms in 0.1 M KCl solution with 10:90 (v/v) EtOH − H2 O. Scan rate: 0.1 V s−1 . During the first sweep, a pronounced peak is detected which reflects the reduction of NO2 to NH2 . The variation in the second and third sweep is minor compared to the first sweep reduction. (Reprinted from Ref. 21.)

586

DNA-MODIFIED DIAMOND FILMS

Current density (mA cm–2)

0

(d) –0.1 (b) (c) (a) –0.2

–1.5

–1

–0.5

0

0.5

Potential (V) vs. Ag/AgCl Figure 21.35 Linear sweep voltammograms of electrochemical reduction of nitro- to amino- phenyl benzene diazonium tetrafluoroborate on diamond with (a) H-, (b) OH-, (c) reconstructed carbon, and (d) O-termination measured in a mixture of water/ethanol (9:1) containing 0.1 M KCl as supporting electrolyte at a scan rate of 0.1 V s−1 . (Reprinted from Ref. 83b.)

summarized in Figure 21.36. This figure combines the charge consumed (open circles) on different diamond surface for the electrochemical conversion of nitro- to aminophenyl, as shown in Figure 21.35 with cathodic peak potentials (full circles) used for electrochemical attachment of nitrophenyl to these different surfaces (shown in Figure 21.31), revealing that only on H-terminated surface is the attachment very effective. To summarize this figure: The cathodic peak potential shows which surface molecules are replaced by phenyl. Using this technique we can identify hydrogen, oxygen or carbon-reconstructed surfaces. The charge consumed for the conversion of nitro- into aminophenyl is used to 0

Charge (mC cm–2)

–0.1

0.4

–0.2

0.2

0

–0.3

H–term. OH–term. C–Term. O–Term.

Peak potential (V vs. Ag/Ag+)

0.6

–0.4

Figure 21.36 Summary of the charges (open circles) consumed for the electrochemical conversion of nitrophenyl to aminophenyl molecules and cathodic peak potentials (full circles) for electrochemical reduction of diazonium salts on diamond with different surface terminations. The bars are deviations of measurements and the dashed lines are guiding-the-eyes lines. (Reprinted from Ref. 83b.)

21.3 SURFACE MODIFICATION OF DIAMOND

587

calculate the density of phenyl molecules attached to the surface, which is either equally to remaining hydrogen or to other atomic configurations. If we use the charge consumed for the electrochemical conversion of nitro- to aminophenyl on these surfaces and compare it with the bonding density of phenymolecules on initially H-terminated diamond, which is 8 × 1013 cm−2 , the remaining hydrogen density on OH-terminated diamond after wet-chemical oxidation is in the range 5 × 1013 cm−2 , which is about 2.5% of the fully hydrogen terminated surface (2 × 1015 cm−2 ). The reactivity of diazonium salts with hydrogen-terminated carbon bonds, therefore, gives quantitative numbers with respect to hydrogen bonding on surfaces. We assume that this is accurate on surfaces that have been wet-chemically or plasma-chemically oxidized, since the density of hydrogen is strongly reduced here. The activity of diazonium salts on O-terminated and reconstructed diamond is calculated to be 4.5 × 1013 and 3.9 × 1013 cm−2 , respectively. Compared to the bonding density on H-terminated diamond of 8 × 1013 cm−2 , this is only about a factor 2 smaller. It demonstrates that diazonium salts are very reactive on a variety of materials. In the future it will be essential to discuss the meaning of potentials involved in this conversion to identify molecular structures on the surface by this fingerprint. Electrochemical grafting of diamond with diazonium salts can be used for the qualitative identification chemical reactions ruling the bonding as well as to calculate quantitatively the density of hydrogen bonded on the surface. Although cyclic voltammetry is a rather invasive technique that changes the surface structure during the potential scan, this method shows new possibilities to characterize hydrogen bonding in absolute values to transducers. It is an alternative and new method to established infra-red (IR) spectroscopy. It will give new insights into surface properties of a variety of transducer materials in the future. On the other hand, detailed investigations of phenyl-layer formation on other electrodes show that this simple interpretation is misleading [126,127]. In most cases, multilayers are deposited. We have therefore applied additional experiments to characterize the growth and thickness of the phenyl-layer on diamond [124]. Generally, all performed contactand oscillatory-mode AFM experiments on deposited nitrophenyl layers reveal layer ˚ Taking into account the length of nitrophenyl thicknesses in the range of 28–68 A. ˚ (Figure 21.30), it clearly indicates multilayer formation by cyclic molecules of about 8 A voltammetry deposition. A better controlled growth of phenyl films on diamond can be achieved by electrochemical means, applying fixed potentials for a given period of time instead of potential cycles [87]. In our case we applied −0.2 V (vs. Ag/AgCl) and measured the transient current during the deposition. The result is shown in Figure 21.37. In case of unlimited electron transfer and diffusion limited attachment, the dynamics follow the Cottrell law (i (t) ∼ t −0.5 ) as introduced by Allongue et al. in 2003 [127]. At the very beginning of the transient current, such a characteristic can be detected. However, for longer times, the current decays faster than predicted by this law. We assume that the growing phenyl-layer limits an effective electron transfer, slowing down the bonding process. The density of electrons involved in this reaction saturates at around 4 × 1015 cm−2 . To verify the layer formation, contact mode AFM has been applied. With increasing force to the tip, the phenyl-layer can be removed from diamond. A typical result is shown in Figure 21.38, where the nitrophenyl layer has been attached by one cyclic voltammetry scan (from +0.5 V to −1.0 V vs. Ag/AgCl at a scan rate of 200 mV s−1 ). Forces below ˚ thickness is 100 nN do not damage to the phenyl film. Above 100 nN, a layer of 26 A

588

DNA-MODIFIED DIAMOND FILMS

(2)

Current Density (A/ cm–2)

(1)

(3)

10–4

NO2 10–5 N2

(1)

N2+

N2+

+

(3)

(2) Diamond

10–6 0.1

t–0.5

NO2

NO2

1

10

100

Time (S)

Figure 21.37 Transient current as detected during nitrophenyl attachment at a constant potential of −0.2 V (vs. Ag/AgCl). Also shown is the theoretical decay following a t−0.5 time dependence. The inset shows a schematic growth model where the phenyl layer starts to grow (1), to become thicker (2), and finally terminates the growth (3), as the electron tunneling transition through the phenyl layer decreases to zero [124]. (Reprinted from Ref. 21.) 100 nN 120 nN >120 nN

6 Height (nm)

5 4 3

26 Å

8Å

2 1 0

0

1000 2000 Distance (nm)

3000

Figure 21.38 AFM scratching experiments on nitrophenyl modified diamond. With forces >100 nN, most of the phenyl-layer can be removed, whereas forces >120 nN are required to remove the linker layer to diamond completely [124]. (Reprinted from Ref. 21.) See color insert.

˚ in height. removed and forces above 120 nN give rise to the removal of a thin layer, 8 A We assume that firstly a random oriented phenyl layer is removed, while forces above 120 nN are required to remove phenyl linker molecules bonded to diamond. Atomic force microscopy characterization on phenyl layers, which have been grown at a constant potential of −0.2 V for different times, indicate 3D growth as shown in

21.3 SURFACE MODIFICATION OF DIAMOND

589

3.0

Layer thickness (nm)

2.5

Diamond

Diamond

2.0 1.5

Diamond

1.0 0.5

U = –0.2V H-terminated Diamond

0.0 1

10 100 Attachment time (s)

1000

Figure 21.39 Nitrophenyl layer growth during constant potential attachment experiments with −0.2 V (vs. Ag/AgCl) is governed by three-dimensional (3D) growth properties. After short-time attachment, the layer thickness varies strongly, whereas after 90 s the variations become much smaller, indicating a dense layer formation of about 25 A˚ thickness (for details, see Ref. 124). (Reprinted from Ref. 21.)

Figure 21.39 (applied tip force: 200 nN, scan rate: 4 μm s−1 ). After short-time attachment ˚ The thickness variation (5 s) the thickness of the layer is already between 8 and 23 A. ˚ Taking into account decreases with increasing attachment time, saturating at around 25 A. the saturated electron density of 4 × 1015 cm−2 and the final thickness of the phenyl-layer ˚ the phenyl molecule density in the layer is about 2 × 1021 cm−3 . of 25 A, The orientation of phenyl molecules has been characterized by angle-resolved XPS experiments. The integrated peak intensities of O(1s) to C(1s) shows a strong angle dependence for attachments at −0.2 V for times up to 40 s (see Figure 21.40). We attribute this to an oriented growth of nitrophenyl, with NO2 molecules preferentially located on the growing top of the layer. This is different in case of much thicker layers ˚ attached by 5 cycles in the range of +0.5 V to −1.0 V versus Ag/AgCl at a (30–65 A) scan rate of 200 mV s−1 . Here, the XPS angle variation is weak; molecules are arranged in a more disordered structure. From these experiments we conclude that the formation of phenyl layers on diamond is governed by 3D growth, with preferential alignment of NO2 cap molecules on the top of growing films, if films are not growing too thick. Growth saturates at a layer thickness of ˚ using constant potential attachment (–0.2 V), whereas significantly thicker about 25 A, ˚ are detected after cyclic attachment (+0.5 V to −1 V). layers in the range 35–65 A Properties of such thick layers are governed by a more random molecule orientation. This is schematically summarized in Figure 21.41. 21.3.3

Tip Functionalization of Diamond Nanotextures

The deposition of nitrophenyl molecules to diamond has been discussed earlier in detail [124]. Different nitrophenyl layers with respect to thickness and density could be obtained by altering the potential for the reduction of diazonium salts and by variation of the attachment time. On nanostructured surfaces, the electric field concentrates at tips of wires,

590

DNA-MODIFIED DIAMOND FILMS

Figure 21.40 Angle resolved XPS experiments show oriented growth of nitrophenyl-layers, grown with constant potential. Cyclic potential attachment method gives rise to significantly thicker layers of typically 30–70 A˚ with less pronounced molecule arrangement. (Reprinted from Ref. 21.)

Figure 21.41 Nitrophenyl groups at an initial stage of attachment grow three-dimensional (3D) as shown here schematically, forming layers of varying heights and densities. Layer thicknesses of up to 80 A˚ are detected for cyclic voltammetry attachment after 5 cycles, whereas the layer becomes denser and only about 25 A˚ thick in case of constant potential attachment. (Reprinted from Ref. 21.) See color insert.

which stimulate current flowing preferentially through wires and their tips. Therefore, this will lead to preferential bonding of nitrophenyl to the tips of wires, giving rise to preferential modifications of nanowire tips [76]. For biosensor applications, a wire separation distance of 10 nm has been selected since it will result in a DNA density of about 1012 cm−2 if these wires are used for anchoring DNA molecules. Thus, in the following sections, the diamond nanowires used were fabricated with an etching time of 10 s. Figure 21.42 shows the current measured as a function of time during nitrophenyl attachment on diamond nanotextures (dashed line) using a constant potential of −0.05 V (vs. Ag/Ag+ ) for 4 s. The attachment potential has been selected from cyclic voltammetric attachment experiments, performed to detect the reduction peak potential of diazonium salts on oxidized diamond surface and on diamond nanowires. The current-time graph follows the Cottrell law with a time dependence: I ∼ t −0.5 . For comparison, attachment of nitrophenyl on smooth and oxidized diamond is shown in Figure 21.42 using

21.3 SURFACE MODIFICATION OF DIAMOND

591

Current density (µA cm–2)

102

flat electrode 101

nanotextures 0

1

2

3

4

Time (s) Figure 21.42 Electrochemical reduction of 1.0 mM 4-nitrobenzene diazonium tetrafluoroborate in 0.1 M NBu4 BF 4 in acetonitrile on diamond nanotextures (dashed line) and on a smooth diamond electrode (solid line) at −0.05 V versus Ag/Ag+ for 4 s in a nitrogen-purged glove box. (Reprinted from Ref. 81.)

the same experimental conditions. Here, on the smooth surface the nitrophenyl density per area is 5 × 1014 cm−2 . The current density on nanostructured diamond is about two times smaller but the area is two times larger. This indicates that the number of phenyl molecules bonded to the tip-structured surface is smaller than on smooth surfaces. In our experiments negatively charged redox molecules will be used later as mediators for DNA hybridization sensing [130–132], so the variation of Fe(CN)6 3−/4− redox currents on diamond nanowires before and after phenyl attachment have been characterized and are shown in Figure 21.43. The voltammogram (curve a) shows a large splitting of peak potentials (868 mV) for the redox reaction Fe(CN)6 3− ↔ Fe(CN)6 4− on diamond nanotextures. This is due to the relative low boron doping concentration (7 × 1019 cm−3 ) of the diamond as well as to surface oxidation of diamond during the RIE fabrication process of the nanowires. On the other side, the redox peak currents on diamond nanowires (curve a in Figure 21.43) are almost two times larger than those on smooth diamond (curve c in Figure 21.43), indicating a two-time larger surface area. Note that for these experiments, the doping of diamond was in the range of 7 × 1019 cm−3 , which is nearly a factor of 10 below the doping density conventionally applied. It gives rise to huge redox current peak splitting; nonetheless, these experiments clearly show the major properties that come by fabrication of vertically aligned diamond nanowires. For applications as gas sensors we expect even low doping levels; for biosensor applications the doping level needs to be significantly enhanced. A reduction of peak splitting potential can be achieved by increasing the boron doping concentration to >3 × 1020 cm−3 , which is

592

DNA-MODIFIED DIAMOND FILMS

5 (a)

(b)

Current (μA)

(c)

0

–5

(d)

–0.5

0

0.5

1

Potential (V) vs. Ag/AgCl Figure 21.43 Cyclic voltammograms of 1.0 mM Fe(CN)6 3−/4− at a scan rate of 0.1 V s−1 in pH 7.4 phosphate buffer on diamond nanotextures before (a) and after (b) nitrophenyl attachment. Curves (c) and (d) are voltammograms on a smooth diamond surface before and after nitrophenyl attachment under the same conditions. (Reprinted from Ref. 81.)

performed in our laboratory. As shown in curve b of Figure 21.43, after phenyl attachment the redox current peaks decreased by about 16% and the peak splitting enlarged to 1160 mV. These changes are a result of phenyl attachment as phenyl is an insulator giving rise to slow down exchange rates between transducer and redox molecules. Note that the redox response on the smooth diamond surface (curve d in Figure 21.43) disappeared totally after nitrophenyl attachment. Smooth diamond surface modified with phenyl molecules are therefore not suitable for application in amperometric biosensors. Despite the imperfection of our diamond layer, the functionalized diamond nanowire surface is still transparent for negatively charged redox molecules, indicating the formation of a thin and dispersed nitrophenyl layer on diamond nanowires. In order to identify locations of nitrophenyl attachment to diamond nanowires, constant-current mode STM has been employed to image the diamond nanowires before and after grafting with nitrophenyl. Sharp tips of 3–10 nm size and 8–10 nm length are detected in the clean diamond nanowires (Figure 21.44a). This result agrees with AFM data shown in Figure 21.14b. STM on phenyl-modified wires shows a drastic change of tip features. A “double peak” structure is detected, which is a result of constant tunneling current mode applied here. As nitrophenyl films are insulating, they give rise to decreased STM tunneling currents. To keep the tunneling current constant, the STM tip will approach toward the surface, which will result in inverted cone-shaped tips as shown in Figure 21.44(b). Please note that the molecular size of nitrophenyl is about 0.67 nm, whereas nanowires are in the range of 10 nm. The short-time electrochemical modification of wires gives rise to preferential phenyl bonding to the very tip of wires. In between, most of the surface remains uncoated. This is promising for the fabrication of bioelectrochemical sensors.

593

21.4 DNA MOLECULES ON DIAMOND

STM Tip

STM Tip Image

Image Phenyl

Phenyl

Diamond

Diamond Surface

Surface

nm 100

nm 14

nm 100

nm 11

80

80

60 0 100

40 80

60

20 40

(a)

20

60 0 100

40 80 60

0 0

(b)

20 40

20

0 0

Figure 21.44 STM images of diamond nanowires (a) before and (b) after the electrochemical grafting of a nitrophenyl film by a constant potential technique at −0.05 V versus Ag/Ag+ for 4 s. (Reprinted from Ref. 76.)

21.4 21.4.1

DNA MOLECULES ON DIAMOND DNA Attachment

Probe DNA attachment on the photochemical or electrochemical modified surface is schematically shown in Figure 21.45. To provide chemically reactive amine groups to the photochemically treated diamond samples, the trifluoroacetamide protecting group was removed by refluxing the TFAAD-modified surface in 0.36 M HCl in methanol at 65◦ C for 24 h. The electrochemically modified surface of boron-doped diamond with nitrophenyl groups (–C6 H5 NO2 ) is electrochemically reduced to aminophenyl (–C6 H5 NH2 ) in 0.1 M KCl solution of EtOH–H2 O (V /V = 1/9) to provide reactive aminophenyl groups, as shown in Figure 21.34. At least three cycles were required. For more cycles, neither the oxidation nor the reduction sweep shows a peak which would indicate further reduction. The disappearance marks the completed reduction of the −NO2 groups to −NH2 groups. Aminophenyl is utilized to bond the hetero bifunctional cross-linker molecules to the film. The cross linker molecules are used to covalently add single-stranded marker DNA molecules to the surface of diamond. To attach DNA, we applied the recipe introduced by Yang et al. [15], and Hamers et al. [133], where the protected amine is firstly deprotected, leaving behind a primary amine. The primary amine is then reacted with the cross-linker and finally reacted with thiol-modified DNA to produce the DNA-modified diamond surface. In our experiments we use a 4 μl droplet that is placed on the diamond layer and that covers a circular area of about 2 mm in diameter, thereby covering oxidized and H-terminated surface parts. To assess whether DNA-modified diamond surfaces have been generated, such surfaces have been exposed to complementary oligonucleotides that were labeled with fluorescence tags FAM. The detailed procedure is the following. To attach DNA, the amine- or the phenyl-layer is then reacted with 14 nM solution of the heterobifunctional cross-linker

594

DNA-MODIFIED DIAMOND FILMS

(a)

(a)

(b)

(b)

(c)

(d)

(c)

(d)

(e)

(e)

Figure 21.45 Schemes of DNA attachment on electrochemical (I) and photochemical (II) modified diamond surface: Scheme I: Nitrophenyl linker molecules are electrochemically bonded to H- or O-terminated diamond. Nitrophenyl is reduced to aminophenyl and reacted with a heterobifunctional cross-linker. Finally, thiol-modified ss-DNA is attached. Scheme II: Amine molecules are photochemically covalently attached on H-terminated diamond. The linker molecules are then unprotected and reacted with hetero-bifunctional cross-linker molecules and thiol-modified ss-DNA. (Reprinted from Ref. 21.)

sulphosuccinimidyl-4-(N-maleimidomethyl) cyclohexane-1-carboxylate (SSMCC) in 0.1 M pH 7 triethanolamine (TEA) buffer for 20 min at room temperature in a humid chamber. The NHS-ester group in this molecule reacts specifically with the –NH2 groups of the linker molecules to form amide bonds. The maleimide moiety was then reacted with 2–4 μl thiol-modified DNA (300 μM thiol DNA in 0.1 M pH 7 TEA buffer) by placing the DNA directly onto the surface in a humid chamber and allowing for reaction for given times between 10 min to 12 h at room temperature. All DNA molecules have been purchased from TOS Tsukuba OligoService (http://www.tos-bio.com, Japan). As probe ss-DNA we used the sequence S1 (= 5 -HS-C6 H12 -T15 -GCTTATCGAGCTTTCG-3 ) and as target ss-DNA the sequence F1 (= 5 FAM-CGAAAGCTCGATAAGC-3 ), where FAM indicates the presence of a fluorescence tag of fluorescein phosphoramidite. To investigate mismatched interactions, a 4-based mismatched target ss-DNA (5 -FAM-CGATTGCTCCTTAAGC-3 ) and one base mismatched target DNA (5 -TGA CCT AAC CAT ACA TAA AAC-3 ) have

21.4 DNA MOLECULES ON DIAMOND

595

been used. As controlled experiments, non complementary DNA was used with the sequence of 5 -Cy5-ACT TCC TTA CTA GTT TAT TCA CG-3 , where Cy5 indicates the presence of a dye color center. For DNA hybridization, 5 μl of SSPE buffer (300 mM NaCl, 20 mM sodium phosphate, 2 mM EDTA, and 7 mM sodium dodecyl sulfate, pH 7.4) containing complementary DNA with different concentrations was dropped on the sensor surface for 1 h at room temperature in a humid cell. After hybridization, the samples are washed in deionized water for 1 h at 37◦ C to remove nonintentional bonded DNA molecules. As for the discrimination of single-based mismatched complementary DNA, 5.0 μL 10 nM solution was used for hybridization under identical conditions. Denaturation of samples has been performed in 8.3 M urea-solution for 30 min at 37◦ C, followed by rinsing in deionized water. Samples are then hybridized again for another DNA cycle. 21.4.2

Characterizations of DNA Layers

For DNA characterization, layers were immersed into SSPE buffer (300 mM NaCl, 20 mM NaH2 PO4 , 2 mM ethylenediaminetetraacetic acid [EDTA], and 6.9 mM sodium dodecyl sulphate [SDS], titrated to pH 7.4 by 2 M NaOH). The buffer solution enables DNA to assume natural conformation and avoids effects of water meniscus around the AFM tip. Surface morphologies are investigated in oscillating-mode AFM (O-AFM), where the tip-surface interaction is controlled by adjusting the tip oscillating amplitude to a defined value (AFM set-point ratio measurements) [75–77]. The set-point ratio is defined as rSP = AO /ASP , where AO is the amplitude of free cantilever oscillations and ASP is the amplitude of the tip, approaching the surface. Measurements are made typically with AO of 6 and 10 nm. In addition, we used also cantilever phase shift detection (phase lag of cantilever oscillation with respect to oscillation of the excitation piezo-element) to enhance the material contrast between diamond and DNA. Molecule bonding properties (mechanical properties) of linker and DNA layers have been characterized by contact mode AFM where we applied different loading forces to the AFM tip (C-AFM) in the range of 6–200 nN. The scan rate was 10–20 μm s−1 . For forces above a critical threshold, linker and DNA molecules are removed. The difference in height is then measured in O-AFM. Doped silicon AFM cantilevers are used in these experiments with a spring constant of 3.5 N m−1 . The cantilever resonance frequency is 75 kHz in air and 30 kHz in buffer solution. Fluorescence microscopy has been applied using a Leica Fluorescence Imaging System DM6000B/FW4000TZ where the fluorescence intensity is evaluated by grey-scale analysis (Leica QWin software). Note that we have characterized all diamond layers before surface modifications to detect fluorescence emission arising form the bulk of diamond, like for example from nitrogen/carbon-vacancy complexes. Those samples have been excluded from our experiments. The shown fluorescence is therefore truly from fluorescence-labeled DNA. For some fluorescence experiments the green label (FAM) has been replace by red fluorescence markers (Cy5). Figure 21.46 shows a fluorescence image of S1 ss-DNA marker molecules bonded to diamond after hybridization with its complementary ss-DNA target molecules F1 labeled with Cy5. The image shows DNA bonding to initially H-terminated diamond and to oxidized diamond. The laid “T” shape pattern in Figure 21.46 arises from surface oxidation. The fluorescence from this area is about 10% darker than from hydrogenterminated diamond. As the light intensity is proportional to the density of fluorescence

596

DNA-MODIFIED DIAMOND FILMS

Figure 21.46 Fluorescence microscopy image of a ds-DNA-functionalized single crystalline diamond electrode after DNA hybridization with complementary oligonucleotides terminated with Cy5 dye molecules. The layer was originally H-terminated but a T-shape has been oxidized. Here, the fluorescence intensity appears weaker. To generate a contrast with respect to clean diamond, a scratched area has been realized. (Reprinted from Ref. 21.) See color insert.

centers, the density of DNA bonded to oxidized diamond is about 10% smaller than on H-terminated diamond. No fluorescence can be detected using a 4-base mismatched ss-DNA target molecule for hybridization. Figure 21.47 shows the result of intense green fluorescence (=100%) from originally H-terminated regions and less intense fluorescence (∼ = 70%) from oxidized surface areas. The weak fluorescence contrast has two reasons. The first is that noncovalently bonded DNA is attached to oxidized diamond. This will be discussed soon, using atomic force microscopy experiments (AFM). The other is that transparent diamond gives rise to light trapping so that the transparent diamond appears green in case of fluorescence emission. DNA-functionalized and hybridized surfaces are characterized by AFM measurements in 2 × SSPE/0.2% SDS (sodium dodecyl sulphate) buffer solution [134,135]. By performing contact mode AFM scratching experiments, DNA can be removed, and at the interface between clean and DNA covered diamond, the height of DNA has to be measured. In addition, the force required to penetrate and remove DNA has to be determined, giving insight into the mechanical stability of the bonding. Scratching experiments were performed with different tip loading forces between 10 and 200 nN. A typical result of such experiments on a diamond surface modified with double-strand (ds) DNA is shown in Figure 21.48a. For each force, an area of 2 μm × 10 μm has been scratched at a scan rate of 20 μm s−1 . Forces around 45 nN (±12 nN) generate a surface tha appears to be clean of DNA, as also detected by fluorescence microscopy shown in Figure 21.48b. C-AFM experiments on the boundary between initially oxidized and H-terminated diamond are shown in Figure 21.49. With C-AFM, we detect noncovalently bonded DNA on oxidized diamond. These molecules can be removed with forces around 5 nN. This is 5 times lower than DNA bonded to the H-terminated surface. The layer is also significantly thinner as on H-terminated diamond.

21.4 DNA MOLECULES ON DIAMOND

597

Figure 21.47 Fluorescence microscopy image of double-stranded (ds) DNA helixes on diamond using green fluorescence tags (Cy5) attached to complementary DNA. The bright areas arise from initially H-terminated diamond; the less intense regions were originally oxidized. Black areas are Au contacts. (Reprinted from Ref. 21.) See color insert.

(a)

(b)

Figure 21.48 Typical surface morphology of a DNA film on diamond after application of contact mode AFM (a), where increasing AFM tip loading forces have been applied. Forces larger than 45 nN give rise to DNA removal on photochemically treated and initially H-terminated diamond. After removal of DNA from the surface, it appears dark in fluorescence microscopy (b). (Reprinted from Ref. 21.) See color insert.

By measuring across the boundary between the DNA-functionalized and the cleaned surface, using O-AFM, the DNA layer thickness can be obtained as shown in Figure 21.50. O-AFM is preferable to C-AFM on soft layers as the tip-surface interaction can be minimized by monitoring the phase shift of the cantilever oscillations [136]. The phase shift was measured as a function of the set-point ratio, rsp = Ao /Asp , where Asp is the set-point amplitude and Ao is the amplitude of free cantilever oscillation, on DNA-functionalized and cleaned diamond surface regions (see squares in Figure 21.50a).

598

DNA-MODIFIED DIAMOND FILMS

Oxidized

C-AFM

H-term

Hight (A)

200

150

100

50 0

5

10

15

x (μm) Figure 21.49 Oscillatory AFM measurement applied at the boundary of cleaned diamond surface to initially oxidized (left) and initially H-terminated (right) diamond show that on both areas DNA molecules are present. The height on O-terminated diamond, however, is lower than on H-terminated diamond. Molecules on oxidized diamond can be removed with forces of about 5 nN. (Reprinted from Ref. 21.)

(a)

(b)

Figure 21.50 (a) Optimized oscillatory AFM measurement at the boundary of a cleaned diamond surface to diamond with attached double-stranded DNA molecules. The squares denote the regions where AFM phase shifts were evaluated. (b) AFM height profile across the boundary reveals a DNA layer thickness of 76 A˚ (see also Ref. 134). (Reprinted from Ref. 21.)

21.4 DNA MOLECULES ON DIAMOND

599

Figure 21.51 AFM set-point ratio dependence of AFM measurements of DNA height and phase contrasts across a DNA-functionalized and cleaned diamond surface for free oscillation amplitudes ˚ (Reprinted (Ao ) of 6 and 10 nm. Extrapolation to set-point ratio 1 results in a DNA height of about 76 A. from Ref. 21.)

Figure 21.51 summarizes results of phase contrast and set-point ratio measurements. The phase contrast between diamond and DNA is positive and approaches zero for set-point ratios approaching 1 (i.e. for increasing tip-surface distance). For a phase contrast near zero, which corresponds to minimized tip-surface interaction, the DNA ˚ Simple O-AFM measurements result in about 70 layer thickness reaches about 75–78 A. ˚ (see Figure 21.50), which is slightly smaller than the real DNA thickness. The height A ˚ is, however, still significantly lower than the expected height of about of 75–78 A ˚ for upright standing DNA (105 A) ˚ on linker (12 A) ˚ and cross-linker molecules 130 A ˚ and fluorescence marker FAM6 (5 A). ˚ We attribute this to a tilted arrangement (6 A) of DNA molecules, as shown in Figure 21.52 (schematic view of molecules). Using triangular geometry, the tilt angle is around 33◦ –36◦ , which is similar to results of DNA bonded to gold [135,136]. There, the detected film thickness of DNA layers has been attributed to orientational properties of DNA duplexes in the monolayer. Due to the large charge density of the DNA backbone (2-/base pair without condensed counter-ions), the orientation of the individual helixes is very sensitive toward neighboring molecules and to surface charges. As a consequence, small changes in applied electrochemical potential can cause drastic changes in helical orientation (for a summary, see Ref. 135). In our case, the tilted average helical packing orientation arrangement of 35◦ certainly reflects the minimization of Coulomb repulsion forces between individual duplexes. Effects from diamond surface charges (C–H dipole) or from externally applied electric fields on DNA arrangements are currently investigated by variations of buffer solution ionicities and by application of externally applied electric fields to the diamond transducer. A topographic surface profile of DNA double helix molecules, bonded on diamond is shown in Figure 21.53. It reveals broad undulations due to collective interaction of several DNA oligomers with the tip. The height is modulated with a periodicity of about ˚ No pinholes can be detected 30–50 nm. The DNA surface roughness is around ±5 A. in the layer. Obviously a closed DNA film has been synthesized on diamond using photochemical attachment.

600

DNA-MODIFIED DIAMOND FILMS

Figure 21.52 From AFM, a compact DNA layer of 76 A˚ height is resolved by optimizing phase and height contrast measurements. The axis of the double-helix DNA is therefore tilted by about 30–36◦ with respect to the diamond surface as shown schematically in this figure. (Reprinted from Ref. 21.)

3.0

Relative height (nm)

2.5 2.0 1.5 1.0 0.5 0.0 0

20

40

60

80

100

Lateral distance (nm) ˚ Figure 21.53 AFM height profile shows a dense DNA layer with a RMS height modulations of ±5 A. (Reprinted from Ref. 21.)

For sensor applications, dilute DNA films in the range 1012 –1013 cm−2 are required [137]. To decrease the bonding density, we reduced the time of marker DNA attachment. A saturated and very dense film is achieved after 12 h of exposure. Following the arguments of Takahashi et al. [45,46], this will result in a DNA density of about 1013 cm−2 . By decreasing the time of marker ss-DNA attachment, the density can be decreased to 1012 cm−2 , as shown in Figure 21.54. Here, we have evaluated the change in fluorescence intensity after hybridization, where the attachment of ss-DNA marker molecules has been varied from 10 min to 12 h. The attachment kinetics is well described empirically by an exponential function with a time constant, τ , of about 2 h. Geometrical properties as well as density and bonding strength of DNA bonded by this electrochemical technique to boron-doped diamond have been characterized by AFM ˚ experiments, as described earlier. The height of DNA layers is detected to be around 90 A. This is slightly higher than in case of photoattachment and arises from the thicker linker

21.4 DNA MOLECULES ON DIAMOND

601

Figure 21.54 Variation of the ss-DNA marker molecule attachment time, between 10 min and 12 h gives rise to an exponentional increasing fluorescence intensity. The fluorescence intensity of hybridized DNA follows an activated property (full line) with a time constant of 2 h. The DNA density on diamond varies between 1012 and 1013 cm−2 . (Reprinted from Ref. 21.)

˚ in the case of phenyl and 12 A ˚ for amine linkers. molecule layer, which is about 25 A Again, a tilted arrangement is deduced, comparable to results from photoattachment (≈35◦ ). The layer is dense with no pinholes. The removal forces are between 60 and 122 nN, the statistical average is about 76 nN. It is interesting to note that on initially oxidized diamond the forces are lower, in the range of 34 nN. A comparison of forces is shown in Figure 21.55. These results indicate strong bonding of DNA to diamond for both photo- and electrochemical surface modifications. Removal forces are about two

DNA removal force (nN)

120

Electrochemical: H-terminated Diamond: Oxidized Diamond:

100 80 60

Photochemical: H-terminated Diamond: Oxidized Diamond:

40 20 0 0

10

20

30

40

50

60

70

80

90

DNA layer thickness (Å) Figure 21.55 Comparison of critical removal forces of electrochemically attached ds-DNA on Hterminated and oxidized diamond and of photochemical attached ds-DNA on H-terminated and oxidized diamond. Attachment to initially H-terminated diamond of both linker molecules systems gives rise to stronger bonding than to initially oxidized diamond surfaces. DNA bonded with phenyl linker molecules show the strongest bonds. (Reprinted from Ref. 21.)

602

DNA-MODIFIED DIAMOND FILMS

140

DNA removal force(nN)

120 100 80 (b)

60

(d)

40 (a) 20 (c)

0

a

d

d

ol

G

ol

G

ic

M

d on ) m e ia in D m A (

d on l) y m ia n D he P (

Figure 21.56 Comparison of DNA removal forces as detected in our experiments on diamond and compared with DNA bonding to gold (a: [136], b: [137], c: [138]), and mica (d: [139]). (Reprinted from Ref. 21.)

times higher than detected on gold and mica as summarized in Figure 21.56 [140]. This is promising with respect to diamond biosensor applications where exceptional chemical stability is required.

21.5

SENSING OF DNA HYBRIDIZATION

Most DNA detection techniques are based on DNA hybridization events. In DNA hybridization, the target ss-DNA is identified by a probe ss-DNA that gives rise to hybridization. This reaction is known to be highly efficient and extremely specific. Commonly used DNA detection techniques (radiochemical, enzymatic, fluorescent) are based on the detection of various labels or reagents and have been proven to be time consuming, expensive, and complex to implement. For fast, simple, and inexpensive detection, direct methods are required. In the following we will introduce recently achieved results with respect to field effect and voltammetric sensing, using single crystalline CVD diamond and diamond nanowires as transducers. 21.5.1

DNA Field Effect Transistor

To realize in-plane gate DNA field-effect transistors (DNA-FET), undoped CVD diamonds with atomically smooth surfaces have been grown by microwave plasma assisted by chemical vapor deposition on Ib substrates. The layer thickness is typically 200 nm. After hydrogen-termination of the diamond surface, an H-terminated sensor area of 2 mm × 0.7 mm size is processed by photolithography and plasma oxidation. This area is surrounded by insulating diamond that has been oxidized. The H-terminated surface is chemically modified (as described later) to covalently bond DNA to it. For experiments in electrolyte solution, drain and source contacts are insulated by silicon rubber. We

21.5 SENSING OF DNA HYBRIDIZATION

Oxidized Diamond Surface

603

Au Contacts Source

Drain

0.7 mm

2 mm H-term. Diamond Surface Figure 21.57 Typical geometry and arrangement of realized CVD diamond DNA ions-sensitive fieldeffect-transistor (FET) structures (DNA-ISFET). The image has been generated by scanning electron microscopy (SEM), where the sensor area (‘‘gate’’) is H-terminated and surrounded by oxidized diamond. Drain and source contacts are from Au. The H-terminated area is photochemically modified to bond ss-DNA marker molecules covalently to diamond. (Reprinted from Ref. 21.)

use Pt as gate electrode (not shown) in buffer solutions. Two Au contacts (0.7 mm × 0.5 mm) evaporated to each end of the H-terminated surface serve as drain and source contacts (see Figure 21.57). For electronic characterization or realization of DNA-fieldeffect transistors (DNA-FET), we deposited Ohmic contacts on H-terminated diamond by thermal evaporation of 200 nm-thick Au onto photoresist patterned diamond, followed ˚ ˚ by a lift-off process. In case of highly boron-doped diamond Ti (100 A)/Pt (100 A)/Au ˚ contacts have been realized using e-beam evaporation. (2000 A) Alkene cross-linker molecules are then attached by photochemical means for 12–20 h, followed by the attachment of probe ss-DNA. The density of ss-DNA has been varied between 1012 and 1013 cm−2 for these experiments. Samples are then transferred into polyetheretherketon (PEEK) sample holders (see Figure 21.58a). Note that during the transfer the ss-DNA layer is covered by sodium chloride buffer solution (1 M with 0.1 M phosphate of pH 7.2). The drain and source contacts are sealed against contact with buffer solution by a silicon rubber (see Figure 21.58b) that is also used to press platinum wires to the drain and source Au contacts (Figure 21.58a). A top view of the setup is shown in Figure 21.58c. The PEEK top part has been designed in a way that 1 mm of the sensor area is exposed to buffer solution. To apply well-defined gate potentials we use a thin Pt wire of 0.2 mm diameter that is immersed into the sodiumchloride buffer. Drain and source currents are measured as function of gate potential that has been varied between 0 and −0.6 V. The gate threshold potential of DNA-FETs has been characterized by applying several cycles where properties of the FET with probe ss-DNA have been determined, followed by determination of FET properties after hybridization with complementary target ss-DNA. Then the sample has been denatured and characterized again. Diamond ion-sensitive field effect transistors show sensitive variation of pH with about 55 mV/pH [38–40]. This is close to the Nernst limit of 60 mV/pH. The effect arises from transfer doping so that no gate-insulator layer is required. The separation of surface channel and electrolyte is therefore very small. The application of such a FET system for DNA hybridization detection is new. We show in the following some typical results, achieved on structures with 1012 cm−2 –1013 cm−2 ss-DNA markers bonded to diamond. A schematic figure of the device with hybridized ds-DNA is shown in Figure 21.59. We used a 1M NaCl solution (containing 0.1 M phosphate with pH 7.2) with a Debye length

604

DNA-MODIFIED DIAMOND FILMS

(a) PEEK-Sample Holder (part 1)

(b) PEEK-Sample Holder (part 2, backside)

(c) Mounted Sample (part 1 and 2)

Figure 21.58 DNA-FET sample holder arrangement. The diamond sensor is mounted in a PEEK plate as shown in (a). The drain and source pads are contacted using Pt wires. A second PEEK part (b) with silicon rubber is mounted on to part (a) and closed to seal the drain source pads from electrolyte buffer (c). A Pt wire is used as gate electrode and the exposed sensor area is 0.7 mm × 1 mm. (Reprinted from Ref. 21.) See color insert.

NA -D ds

ds

ds

-D

-D

NA

NA

Pt

Debye Length λD DRAIN

Source

Figure 21.59 Schematic description of DNA hybridized on a diamond DNA ion-sensitive field-effecttransistor (DNA-ISFET) sensor. The surface conductivity of diamond will be changed by accumulation of compensating cations in the DNA layer, which is caused by the negatively charged back-bone structure of DNA (for details, see Ref. 84). The 1 M NaCl buffer shrinks the Debye length in our ˚ (Reprinted from Ref. 21.) See color insert. experiments to about 3 A.

˚ as calculated by [137]: of 3 A  λD =

εel ε0 kT 2z 2 q 2 l

1/2 (21.4)

21.5 SENSING OF DNA HYBRIDIZATION

605

where k is the Boltzmann constant, T is the absolute temperature, εo is the permittivity of vacuum, εel is the dielectric constant of the electrolyte, z is the valence of ions in the electrolyte, q is the elementary charge, and I represents the ionic strength, for a 1-1 salt, it can be replaced by the electrolyte concentration no . Since the linker molecule is ˚ (amine) and the cross-linker molecule 5 A ˚ long, the DNA is not in touch with 10–15 A the Helmholtz layer in our experiments. Figure 21.60a shows a comparison of drain-source currents (IDS ) measured at a fixed drain-source potential of −0.5 V as function of gate potential for ss-DNA marker molecules attached on the gate, for hybridized marker and target DNA on the gate, and after removal of DNA from diamond. The initial ss-DNA density bonded to diamond is 4 × 1012 cm−2 . The drain-source current increases by hybridization as detected also for a sensor where the initial ss-DNA density is slightly smaller (1012 cm−2 ) or larger (1013 cm−2 ). The gate potential variations from ss-DNA to complementary ds-DNA bonding vary, as shown in Figure 21.60b between 30 and 100 mV. There is a clear trend that with decreasing DNA density, the potential shift becomes larger (as predicted by Poghossian et al. [137]). Taking into account the ion sensitivity of diamond ISFETs of 55 mV/pH, this reflects a decrease of pH of the buffer solution of about 1–1.4 by hybridization. Poghossian et al. [137] calculated that the average ion concentration within the intermolecular spaces after hybridization can be more than three to four times higher for cations than before hybridization. In the case of a Nernstian slope of the sensor, theresearchers predict a gate potential shift of 28–35 mV. Our results indicate a stronger change. The increase in drain-source current with hybridization can be well described by the transfer doping model as the increase in cation density will cause a decrease of pH. Therefore, the chemical potential will increase giving rise to enhanced surface conductivity. A summary of published sensitivities of DNA FETs from silicon is shown in Figure 21.61 from Ref. 137, 141, 142. The initial sensitivities reported before 2004 of more than 100 meV threshold-potential shifts by hybridization have not been reproduced.

–3.0 x

10–6

(b) Without DNA

UDS = –0.5 V

IDS (A)

–2.5 x 10–6 –2.0 x 10–6

ds-DNA

80 mV

–1.5 x 10–6 –1.0 x 10–6 –5.0 x

10–6

ss-DNA

0.0 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0.0 UG (V)

0.1

Gate Potential Shift (mV)

(a) –3.5 x 10–6

140 120 100 80 60 40 20

Diamond FET 1M NaCl (pH 7.2)

0 1011 1012 1013 DNA Molecule Density (cm–2)

Figure 21.60 (a) Drain-source current variations measured as a function of gate potential at a fixed drain-source potential of −0.5 V for ss-DNA (marker-DNA), after hybridization with complementary target ss-DNA to form ds-DNA and after removal of DNA by washing. A gate potential shift of about 80 mV is detected on this DNA-ISFET with about 4 × 1012 cm−2 molecules bonded to the gate. (b) Gate-potential shifts as detected on diamond transistor structures with 1012 cm−2 , 4 × 1012 cm−2 , and 1013 cm−2 ss-DNA marker molecules bonded to the gate area. The threshold potential is increasing toward less dense grafted diamond gates areas. (Reprinted from Ref. 21.)

606

DNA-MODIFIED DIAMOND FILMS

500 Threshold voltage change (mV)

450

DNA-FET Si

400 350 300 250 200 150

Diamond DRC Ishige

100 50 0 1996

Theoritical: (28–35) mV

KL 1998

2000

2002

2004

2006

2008

Year Figure 21.61 Comparison of silicon-based DNA-ISFET sensitivities as a function of data publication (black half-filled squares from Ref. 79, and half-filled triangles from Ref. 87) with diamond sensitivities as shown in this chapter (half-filled diamonds, DRC) and with data deduced on polycrystalline CVD diamond films from Ref. 88 (half-filled circles). The shaded area indicates the theoretically predicted sensitivity, following the model of Poghossian et al. [137]. (Reprinted from Ref. 21.)

On the contrary, sensitivities are decreasing towards experimental reproducible as well as theoretical predictable values in the range 30–80 meV. 21.5.2

Cyclic Voltammetry and Impedance Spectroscopy

For amperometric detection of DNA hybridization we use Fe(CN6 )3−/4− as mediator redox molecule [130–132,143]. The detection principle is shown in Figure 21.62. After immobilization of marker DNA (ss-DNA) the negatively charged redox molecule Fe(CN)6 3−/4− can still diffuse through the ss-DNA layer and can reach the diamond surface. This results in peak-shaped voltammograms, as shown schematically in Figure 21.62a. However, after hybridization with target DNA, the enlarged DNA molecule (ds-DNA) repels negatively charged redox molecules due to electrostatic repulsive Coulomb force arising from the negatively charged sugar phosphate backbone of DNA. The spacing in between ds-DNA molecules is decreased and therefore prevents the diffusion of negatively charged redox molecules toward the electrode. Please also note that the surface resistance will be enhanced after hybridization. These facts result in a wider peak splitting and a decrease of redox current, and in some cases, even in a total disappearance of redox activity, as shown schematically in Figure 21.62b. The difference of redox peak currents and potentials on ss-DNA modified electrodes compared to that of ds-DNA modified electrodes reflects DNA hybridization and is therefore used as quantitative parameter for DNA sensing. The result is shown in Figure 21.63 where we used cyclic voltammetry on H-terminated metallically doped (p-type) single crystalline diamond in 0.5 mM Fe(CN6 )3−/4− , 100 mM KCl, 100 mM KNO3 measured with respect to Ag/AgCl with a scan rate of 100 mV s−1 . The H-terminated diamond shows a well-pronounced oxidation peak at +280 mV and a corresponding reduction wave with a peak at +126 mV (not

21.5 SENSING OF DNA HYBRIDIZATION

(a) Single-stranded (ss) DNA

(b) Double-stranded (ds) DNA

Redox-Molecules diffuse into the ss-DNA film.

ss

607

Redox-Molecules are repelled by Coulomb Force

ds

Diamond

Diamond

Generation of Redox-Current

No Redox-Current Current density (A/cm2)

2

Current density (A/cm )

Ferrocyanide (5mM) redox reaction 1.0 x 104

ss

4

5.0 x 10

0.0 –5.0 x 104 –1.0 x 104 –0.4 –0.2

0.0

0.2

0.4

0.6

Potential (V vs. Ag/AgCl)

1.0 x 10

4

ds

5.0 x 104 0.0 –5.0 x 104 –1.0 x 104 –0.4

–0.2 0.0 0.2 0.4 0.6 Potential (V vs. Ag/AgCl)

Figure 21.62 Schematic picture of DNA sensing on diamond electrodes by using negatively charged redox mediator molecules. (a) Diffusion of redox mediators toward the diamond electrode results in a peak-shaped voltammogram. (b) Repulsion of redox mediators due to DNA hybridization leads to a decrease in peak currents and a broader peak splitting. (Reprinted from Ref. 82b.) See color insert.

0.20

Current density (mA cm−2)

ss-DNA 0.15 0.10 0.05 0.00 –0.05 –0.10

ds-DNA

–0.15 –0.2

–0.1

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Potential (V) vs SCE

Figure 21.63 Cyclic voltammograms on ss- and ds-DNA grafted metallically boron-doped ( p-type) single crystalline diamond in 0.5 mM Fe(CN6 )3−/4− , 100 mM KCl, and 100 mM KNO3 measured with respect to Ag/AgCl with a scan rate of 100 mV s−1 . The oxidation and reduction peaks are decreasing for about 50 μA cm−2 by hybridization. (Reprinted from Ref. 81.)

shown here). These characteristics are well known and have been published in the literature [22–24]. After electrochemical attachment of phenyl linker molecules and ss-DNA marker molecules, the redox amplitude is decreasing to about 30% of the clean diamond surface. The peaks are not significantly shifted in potential or broadened by chemical modifications of the electrode (see Figure 21.62). By hybridization, peaks are

608

DNA-MODIFIED DIAMOND FILMS

slightly shifted toward higher oxidation and lower reduction potentials. The change in amplitude is about 50 μA cm−2 . This change in voltammetric signal is reproducible and can be detected for several hybridization/denaturation cycles. In the future, this technique needs to be characterized in depth to evaluate the sensitivity, durability, and reproducibility of this sensor array. Yang et al. reported in 2004 [49] on cyclic voltammetry experiments using Fe(CN6 )3−/4− on boron-doped nanocrystalline diamond films coated with amines as linker and DNA. After ss-DNA marker attachment, their redox currents decreased drastically and the authors concluded that the application of cyclic voltammetry was inhibited by the highly insulating nature of the molecular amine-layer linking DNA molecules to diamond. Our voltammetric experiments show that a detailed control of phenyl-molecule deposition is required. Insulating properties with respect to Fe(CN6 )3−/4− are detected if the phenyl layer is grown slowly and thick, to form a dense scaffold on diamond. By short-time attachment, using the constant potential attachment technique, dispersed layers are generated so that diffusion of Fe(CN6 )3−/4− is not suppressed. Alternatively, Yang et al. [49], Hamers et al. [133] and Gu, Su, and Loh [50] applied impedance spectroscopy where they detected hybridization-induced variations at low frequency. We have also applied such a detection scheme on our grafted diamond layers. A typical result is shown in Figure 21.64. A clear difference between single-strand DNA bonded to diamond and double-strand DNA is detected. This technique allows discriminating hybridization of matched and mismatched DNA. The interpretation of these data requires, however, to know details about dielectric variations of ss-DNA and ds-DNA layers, about thickness variations by applied external electric fields [135], as well as the effect of redox-molecules like Fe(CN6 )3−/4− on the dielectric and conductivity properties of DNA films on diamond.

80

Zim (kOhm)

60 ss DNA 40 mismatched DNA 20 ds DNA

0

50

100

150

200

250

Zre (kOhm) Figure 21.64 Impedance spectroscopic properties of DNA-modified nanodiamond films. The impedance is shown in the complex plane as detected for ss-DNA, for exposure to 4-base mismatched DNA and after exposure to complementary DNA in pH 7.4 phosphate buffer containing 1 mM Fe(CN6 )3−/4− .

21.5 SENSING OF DNA HYBRIDIZATION

609

(b)

(a) 5

0M

100 µM target DNA

1.5

10 pM

clean wires

0.1 nM

Current (μA)

ss DNA

1.0 nM 1

10 nM

ds DNA 0

0.5

–5 –0.5

0

0.5

1

0

0.3

0.6

0.9

Potential (V) vs. Ag/AgCl

Figure 21.65 (a) Cyclic voltammograms of 1.0 mM Fe(CN)6 3−/4− at a scan rate of 0.1 V s−1 in pH 7.4 phosphate buffer on diamond nanowires before (solid line) ss-DNA attachment, after ssDNA attachment (dashed line), and after hybridizing complementary target DNA using 5 μL 100 μM buffer solution (dotted line). (b) Differential pulse voltammograms of 1.0 mM Fe(CN)6 3−/4− in pH 7.4 phosphate buffer on diamond nanowires after (solid line) ss-DNA attachment and after hybridizing with complementary DNA at concentrations of 10 pM, 0.1 nM, 1.0 nM, and 10 nM, respectively. The scan rate was 0.1 V s−1 and the pulse amplitude was 2.5 mV. (Reprinted from Ref. 81.)

21.5.3

DNA Sensing on Nanotextured Diamond Surfaces

Figure 21.65a shows cyclic voltammograms of Fe(CN)6 3−/4− as mediators on diamond nanotextures after immobilizing ss-DNA (dashed line) and after hybridizing with c-DNA (dotted line). A relative large variation in amplitude and a much wider peak splitting are detected, indicating efficient DNA hybridization on diamond nanowires. The peak potential and peak current did not change even after 30 hybridization/denaturation cycles. As a control experiment, we exposed diamond nanotextures functionalized with marker DNA to solutions containing non–c-DNA. No variation was detected electrochemically before and after exposure, indicating that this sensor does not respond to the presence of base mismatched oligonucleotides. Detection of DNA hybridization on diamond nanotextures using c-DNA, which was varied from 0.1 mM to 10 pM, was done by differential pulse voltammetry at a scan rate of 0.1 V s−1 with an amplitude of 2.5 mV. Differential pulse voltammograms of 1.0 mM Fe(CN)6 3−/4− in pH 7.4 phosphate buffer on diamond nanowires after ss-DNA attachment (solid line) and after hybridizing with c-DNA at concentrations of 10 pM, 0.1 nM, 1.0 nM, and 10 nM are shown in Figure 21.65b. In these tests, 5 μl c-DNA solutions were used. Please note that all hybridization treatments have been performed at a constant temperature of 20◦ C applied for 60 min. The amplitude variation from curve a to curve e represents the maximum variation in signal between ss- and ds-DNA. A maximum amplitude decrease of 33% can be achieved by full hybridization. The positive shift of anodic peak potential was observed, indicating slowed-down interactions of mediators’ molecules.

610

DNA-MODIFIED DIAMOND FILMS

The difference of peak current for the oxidation of 1.0 mM Fe(CN)6 3−/4− in pH 7.4 phosphate buffer on ss-DNA attached diamond nanotextures from that after hybridization of ss-DNA with c-DNA, Ip , as a quantitative parameter, was plotted as a function of the concentration of c-DNA, cc−DNA . Figure 21.66a shows the variation of redox peak current difference (between denaturated and hybridized DNA) as a function of complementary target DNA concentration in buffer solution. The bars represent variations of Ip as detected by repetition of experiments. The standard deviation is in the range of 7%, indicating good reproducibility. To identify the detection limit, differential pulse voltammograms of ss-DNA–modified diamond nanotextures before and after hybridization using 1.0–9.0 pM c-DNA concentrations were recorded. The result, Ip versus cc-DNA, is shown in Figure 21.66b. A detection limit of about 2 pM on 0.03 cm2 sensor area is detected using 100 μl solution but keeping the other parameters constant (20◦ C and 60 min hybridization time). We expect that by miniaturization of the sensor area, a significantly lower detection limit can be achieved, which is currently in progress in our laboratory. The overall performance of diamond-nanotextures-based DNA sensors is about 100 to 1000 times better compared to published data available in the literature where comparable DNA strands have been used with comparable redox mediator molecules on planar gold electrodes [144–147], gold nanowires [148,149], and diamond [49,50,133]. Stability measurements of probe DNA on diamond nanotextures were performed using 5.0 μl 1.0 μM c-DNA solutions over extended cycles of hybridization/denaturation treatments. The cyclic voltammograms (scan rate: 0.1 V s−1 ) and differential pulse voltammograms (scan rate: 0.1 V s−1 , amplitude: 2.5 mV) of 1.0 mM Fe(CN)6 3−/4− in pH 7.4 phosphate buffer on probe DNA-modified diamond nanotextures before and after hybridization were recorded. The anodic peak currents from above voltammograms were adopted as a parameter to describe the stability of probe DNA on diamond nanotextures. Figure 21.67 shows the chemical stability of marker DNA on diamond nanotextures as detected from cyclic voltammetry (a) and differential pulse voltammetry (b). No degradation is detected over 30 cycles of DNA hybridization/denaturation, which is comparable to (b)

(a)

ΔIp (μA)

0.6

0.09

0.4

0.06

0.2

0.03

0

102

101

100 10–1 10–2 10–3 10–4 10–5 10–6 cc-DNA (μM)

0

0

2

4

6

8

10

cc-DNA (pM)

Figure 21.66 (a) Difference of peak currents before and after hybridization with complementary DNA. Ip is plotted as a function of concentration of the target DNA, cc−DNA . (b) The variation of Ip with cc−DNA in the range from 1.0 pM to 9.0 pM. (Reprinted from Ref. 81.)

21.5 SENSING OF DNA HYBRIDIZATION

(a)

101 Iss DNA (μA)

611

(b) 100

10–1

0

5

10

15

20

25

30

Hybridization/denaturation cycles Figure 21.67 Variation of peak currents of 1.0 mM Fe(CN)6 3−/4− in pH 7.4 phosphate buffer on diamond nanowires functionalized with marker DNA as function of hybridization denaturation cycles as detected by (a) cyclic voltammograms and (b) differential pulse voltammograms (b).

optical detected DNA stability on diamond [15]. The higher chemical stability of DNA on diamond compared to substrates like gold results from stronger covalent carbon–carbon bonds, which has recently been demonstrated by AFM scratching experiments [124,134]. Moreover, as diamond is chemically inert and ultrahard, nanotextures from diamond will survive in harsh environments where other materials such as gold and silicon will fail. Figure 21.68 shows differential pulse voltammograms as detected on diamond nanotextures after immobilization of marker ss-DNA (a), after exposure to 10 nM single-base mismatched DNA (b), and after hybridization with complementary target ss-DNA (c). After marker attachment, large anodic redox currents are measured. After exposure to single-base mismatched DNA, an amplitude decrease of 8% is detected with a small shift in peak potential. After hybridization with c-DNA, the peak current decreases by about 40% and the peak potential shifts by 80 mV toward positive potential. It demonstrates single-base mismatch discrimination, which, again, is promising for applications in nucleic acid sensors. In summary, diamond nanotextures with controlled geometrical properties such as length and distance between wires can be tip-functionalized electrochemically as a result of electric field concentrations at the tips of wires. Tip-modified wires were utilized for controlled geometrical bonding of DNA molecules. Nucleic acid molecules are bonded in this way with well-defined spacing to the transducer, giving rise to optimized DNA hybridization kinetics. Compared to self-assembled monolayers (SAMs) or dendritic modifications for DNA bonding, the special advantages of diamond nanotextures are the combination of modified surfaces (tips) and the unmodified surfaces for sensing. Due to control of DNA spacing, the hybridization kinetics are optimized, resulting in high sensitivity as well as in single-base mismatch discrimination. The detected sensitivity is promising for miniaturization of the sensor area that is required for real device applications. Furthermore, this technique will be applied in ultramicro and nanoelectrode arrays

612

DNA-MODIFIED DIAMOND FILMS

(a)

Current (μA)

1.5

(b)

(c) 1

0.5

0 0.3

0.6

0.9

Potential (V) vs. Ag/AgCl

Figure 21.68 Differential pulse voltammograms of 1.0 mM Fe(CN)6 3−/4− in pH 7.4 phosphate buffer as detected on diamond nanowires (a) after marker ss-DNA immobilization, (b) after exposure to single-base mismatched DNA (10 nM), and (c) after exposure to complementary target DNA (10 nM). (Reprinted from Ref. 81.)

to realize oligonucleotide sensors that suit the demands in clinical environments where high through-put is required on a 24 h day−1 basis.

21.6

SUMMARY AND OUTLOOK

In this review, we have summarized results with respect to photo- and electrochemical surface modifications of diamond. Both techniques have been optimized to a level that allowed the realization of first-generation electrochemical and field-effect DNA sensors. By AFM experiments we detected a formidable bonding stability of biomolecular arrangements on diamond. It confirms earlier reports by Yang et al. [15] about exceptional DNA bonding stability in hybridization cycles. Applications of diamond biosensors in high through-put systems where especially high bonding stability is required will therefore be of significant interest in the future. Our experiments also show that basic understanding of growth mechanisms, of electronic and chemical properties of each layer of the composite biorecognition film (for example, amine/crosslinker/ss-DNA) is required to achieve progress with respect to optimization of sensor performance. Finally, the proper selection of diamond transducer materials for biosensor applications is of comparable importance. Electronic detection of bonding events in electrolyte solutions requires high-quality diamond with minimized defect densities, no grain boundaries and no sp2 , and with a defect-free surface termination. Such diamond transducers are and will be more expensive than established semiconducting materials like Si. We

21.6 SUMMARY AND OUTLOOK

613

assume, therefore, that future applications of diamond will be in clinical, high throughput systems that require high bonding stability. Multiarray sensors from single crystalline diamond can be realized using established technologies and chemistry. Since prices for established DNA multiarray optical sensors are rather expensive, the costs for diamond substrates of typically 150–250¤ will not be a strong argument against the use of diamond for such applications. In any case, it is important to know that the device properties and performances of diamond are superior to other transducer layers to be commercially successful. There are several fields where nano- and polycrystalline diamond seems to become a promising leading application. Diamond nanoparticles with a diameter of typically 5 nm are currently investigated as core material for rapid, low-volume solid phase extraction of analytes, including proteins and DNA from a variety of biological samples [150–155]. These particles act as color center and can carry specific functional groups. These colorcenters are resistant against photo-bleaching and blinking [150,155]. The increasing demand for secure, mobile, wireless communication has stimulated interest in technologies capable of reducing the size and power consumption of wireless modules, and enhancing the bandwidth efficiency of communication networks. Nanodiamond micro-electrical-mechanical-systems (MEMS) are at the leading edge of this field as the frequency (f ) quality (Q) product of such nanodiamond oscillators reached f = 1.51 GHz and the quality factor Q = 11,555 [156–158]. In the field of biosensing, such high-quality mechanical oscillation systems are also of significant importance for improved detection and sensitivity [159,160]. Direct manipulation of living cells or transfer of molecules into cells (“cell surgery”) is an emerging and increasingly important technology in biology. It requires new tools with dimensions below the micrometer regime. A typical gene surgery tip should be about 40 μm long with a diameter of around 400 nm (for a review, see Ref. 161). In addition, the material should not poison the cell during manipulation, the surface should be optimized with respect to friction of cell membranes, and it should allow applying given potentials to the tip for electrostatic bond or releasing DNA fragments. Best candidates for these applications are nano- and polycrystalline diamond, becauses they show (1) the required mechanical stability (hardness 50–100 GPa), (2) the surface can be adjusted to optimize friction, and (3) if properly designed, the core of the tip can be conductive by boron doping. Diamond nanotextures are proved to be a new concept toward next generation electrochemical sensor platforms. The presented data show significant improvements with respect to sensitivity and chemical stability required to meet future needs in various fields. The sensitivity of pico-Mole achieved on such macroscopic sensor areas is promising with respect to miniaturization. Decreasing the sensor area from square millimeter to square micrometer will improve the sensitivity by a factor 106 atto-mole. Although detailed electronic characterization of diamond nanowires have not yet been performed, it is very likely that these structures resemble mesoscopic properties as known from other material where single molecular sensing has been achieved [51]. In addition, mesospacing of DNA molecules will improve single-base mismatch discrimination. Based on these arguments, we are convinced that bioapplication of diamond either monocrystalline for electronic sensing or poly- and nanocrystalline for mechanical techniques, are very realistic opportunities for this promising material. Diamond will surely find its place in the rapidly growing field of biology and biotechnology. Moreover, diamond nanowires provide robust interfaces for a variety of applications such as for gas,

614

DNA-MODIFIED DIAMOND FILMS

chemical, and biochemical sensing. In the future, the geometrical, electronical, optical, and electrochemical properties of these wires need to be characterized in detail to establish a better understanding of this novel device structure.

21.7

ACKNOWLEDGMENTS

The authors thank Dr. T. Nakamura for the synthesis of TFFAD molecules, and Dr. H. Watanabe, Dr. Ri, and Dr. Tokuda for the growth of excellent intrinsic and borondoped diamond films. Furthermore, we offer our gratitude to Dr. O.A. Williams for his activities with respect to diamond nanoparticle seeding, Dr. D. Shin and Dr. B. Rezek for characterization activities, as well as W. Smirnov for Ni-etching of diamond surfaces.

REFERENCES 1. R. Etzioni, N. Urban, S. Ramsey, M. McIntosh, S. Schwartz, B. Reid, J. Radich, G. Anderson, L. Hartwell, Nat. Rev. Cancer 2003, 3 , 243–252. 2. C. Sander, Science 2000, 287 , 1977–1978. 3. P.R. Srinivas, B.S. Kramer, S. Srivastava, Lancet Oncol . 2001, 2 , 698–704. 4. M.K. Brawer, Prostate Specific Antigen, Marcel Dekker, New York, 2001. 5. J.M. Buriak, Chem. Rev . 2002, 102 , 1271–1308. 6. J.D. Wulfkuhle L.A. Liotta, E.F. Petricoin, Nat. Rev. Cancer 2003, 3 , 267–275. 7. J.N. Kremsky, J.L. Wooters, J.P. Dougherty, R.E. Meyers, M. Collins, E.L. Brown, Nucl. Acids Res. 1987, 15 , 2891–2909. 8. S.R. Rasmussen, M.R. Larsen, S.E. Rasmussen, Anal. Biochem. 1991, 198 , 138–142. 9. K.M. Millan, A.J. Spurmanis, S.R. Mikkelsen, Electroanalysis 1992, 4 , 929–932. 10. K. Hashimoto, K. Ito, Y. Ishimori, Anal. Chem. 1994, 66 , 3830–3833. 11. T. Strother, W. Cai, X.S. Zhao, R.J. Hamers, L.M. Smith, J. Am. Chem. Soc. 2000, 122 , 1205–1209. 12. C.E. Nebel, B. Rezek, A. Zrenner, Phys. Stat. Sol. A 2004, 11, 2432–2438. 13. L. Bousse, N.F. de Rooji, P. Bergeveld, IEEE Trans. Electron Dev . 1983, 30 , 1263–1270. 14. M.R. Linford, P. Fenter, P.M. Eisenberg, C.E.D. Chidsay, J. Am. Chem. Soc. 1995, 117 , 3145–3155. 15. W. Yang, O. Auciello, J.E. Butler, W. Cai, J.A. Carlisle, J.E. Gerbi, D.M. Gruen, T. Knickerbocker, T.L. Lasseter, J.N. Russell Jr, L.M. Smith, R.J. Hamers, Nat. Mater. 2002, 1 , 253–257. 16. C.E. Nebel, J. Ristein, eds., Thin-Film Diamond I, Semiconductors and Semimetals, Vol. 76, Elsevier, New York, 2003. 17. C.E. Nebel, J. Ristein, eds., Thin-Film Diamond II, Semiconductors and Semimetals, Vol. 77, Elsevier, New York, 2004. 18. (a) R. Jackman, ed., Diamond Electronics, Semiconductor Science and Technology, Vol. 18, 2003. (b) A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk, eds., Diamond Electrochemistry, Elsevier/BKC-Tokyo, 2005. 19. J.C. Angus, Y.V. Pleskov, S.C. Eaton, “Electrochemistry of Diamond,” in Thin Film Diamond II, Semiconductors and Semimetals, Vol. 77 (C.E. Nebel, J. Ristein, eds.), Elsevier, New York, 2004, pp. 97–119.

REFERENCES

615

20. G.M. Swain, “Electroanalytical Applications of Diamond Electrodes,” in Thin Film Diamond II, Semiconductors and Semimetals, Vol. 77 (C.E. Nebel, J. Ristein, eds.), Elsevier, New York, 2004, pp. 121–148. 21. C.E. Nebel, B. Rezek, D. Shin, H. Uetsuka, N. Yang, J. Phys. D: Appl. Phys. 2007, 40 , 6443–6466. 22. M.C. Granger, J. Xu, J.W. Strojek, G.M. Swain, Anal. Chim. Acta 1999, 397 , 145–161. 23. M.C. Granger, M. Witek, J.S. Xu, J. Wang, M. Hupert, A. Hanks, M.D. Koppang, J.E. Butler, G. Lucazeau, M. Mermoux, J.W. Strojek, G.M. Swain, Anal. Chem. 2000, 72 , 3793–3804. 24. D. Shin, H. Watanabe, C.E. Nebel, J. Am. Chem. Soc. 2005, 127 , 11236–11237. 25. T. Kondo, Y. Einaga, B.V. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima, J. Electrochem. Soc. 2002, 149 , E179–E184. 26. Y.V. Pleskov, Y.E. Evstefeeva, M.D. Krotova, V.V. Elkin, V.M. Mazin, V.Y. Mishuk, V.P. Varnin, I.G. Teremetskaya, J. Electroanal. Chem. 1998, 455 , 139–146. 27. M.I. Landstrass, K.V. Ravi, Appl. Phys. Lett . 1998, 55 , 975–977. 28. R.S. Gi, T. Mizumasa, Y. Akiba, Y. Hirose, T. Kurosu, M. Iida, Jpn. J. Appl. Phys. 1995, 34 , 5550–5555. 29. J. Shirafuji, T. Sugino, Diam. Rel. Mat . 1996, 5 , 706–713. 30. F. Maier, M. Riedel, B. Mantel, J. Ristein, L. Ley, Phys. Rev. Lett . 2000, 85 , 3472–3475. 31. S.J. Sque, R. Jones, P.R. Briddon, Phys. Rev. B 2006, 73 , 085313. 32. D. Takeuchi, H. Kato, G.S. Ri, T. Yamada, P.R. Vinod, D. Hwang, C.E. Nebel, H. Okushi, S. Yamasaki, Appl. Phys. Lett . 2005, 86 , 152103. 33. C.E. Nebel, N. Yang, H. Uetsuka, T. Yamada, H. Watanabe, J. Appl. Phys. 2008, 103, 013712. 34. C.E. Nebel, B. Rezek, A. Zrenner, Diam. Rel. Mat . 2004, 13 , 2031–2036. 35. C.E. Nebel, B. Rezek, D. Shin, H. Watanabe, Phys. Stat. Sol. A 2006, 203 , 3273–3298. 36. C.E. Nebel, New Diamond Front. Carbon Technol . 2005, 15 , 247–264. 37. V. Chakrapani, S.C. Eaton, A.B. Anderson, M. Tabib-Azar, J.C. Angus, Electrochem. Solid State Lett . 2005, 8 , E4–E8. 38. G. Piantanida, A. Breskin, R. Chechik, O. Katz, A. Laikhtman, A. Hoffman, C. Coluzza, J. Appl. Phys. 2001, 89 , 8259. 39. T.N. Rao, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Electrochem. Soc. 1999, 146 , 680–684. 40. C.E. Nebel, B. Rezek, D. Shin, H. Watanabe, T. Yamamoto, J. Appl. Phys. 2006, 99 , 33711. 41. C.E. Nebel, H. Kato, B. Rezek, D. Shin, D. Takeuchi, H. Watanabe, T. Yamamoto, Diam. Rel. Mat . 2006, 15 , 264–268. 42. B. Rezek, D. Shin, H. Watanabe, C.E. Nebel, Sens. Actuat. B 2007, 122 , 596–599. 43. L. Tang, C. Tsai, W.W. Gerberich, L. Kruckeberg, D.R. Kania, Biomaterials 1995, 16 , 483–488. 44. F.Z. Cui, D.J. Li, Surf. Coat. Technol . 2000, 131 , 481–487. 45. K. Takahashi, M. Tanga, O. Takai, H. Okamura, Bio Industry 2000, 17 , 44–51. 46. K. Takahashi, M. Tanga, O. Takai, H. Okamura, Diam. Rel. Mat . 2003, 12 , 572–576. 47. J. Wang, M.A. Firestone, O. Auciello, J.A. Carlisle, Langmuir 2004, 20 , 11450–11456. 48. G.-J. Zhang, K.S. Song, Y. Nakamura, T. Ueno, T. Funatsu, I. Ohdomari, H. Kawarada, Langmuir 2006, 22 , 3728–2734. 49. W. Yang, J.E. Butler, J.N. Russel Jr., R.J. Hamers, Langmuir 2004, 20 , 6778–6787. 50. H. Gu, X. Su, K.P. Loh, J. Phys. Chem. B, 2005, 109, 13611–13618. 51. K.S. Song, M. Degawa, Y. Nakamura, H. Kanazawa, H. Umezawa, H. Kawarada, Jap. J. Appl. Phys. Part 2-Lett. Express Lett . 2004, 43 , L814–L817.

616

DNA-MODIFIED DIAMOND FILMS

52. A. H¨artl, E. Schmich, J.A. Garrido, J. Hernando, S.C.R. Catharino, S. Walter, P. Feulner, A. Kromka, D. Steinmueller, M. Stutzmann, Nat. Mater. 2004, 3 , 736–742. 53. M. Hasegawa, private communication, 2006. 54. O.A. Williams, M. Nesladek, Phys. Stat. Sol. A 2006, 203 , 3375–3386. 55. J.A. Carlisle, O. Auciello, The Electrochemical Society, Interface, Springer, New York, 2003, pp. 28–31. 56. C.E. Nebel, “Transport and Defect Properties of Intrinsic and Boron-Doped Diamond,” in Thin-Film Diamond I, Semiconductors and Semimetals, Vol. 76 (C.E. Nebel, J. Ristein, eds.), Elsevier, New York, 2003, pp. 261–324. 57. C.E. Nebel, Semicond. Sci. Technol. 2003, 18 , S1–S11. 58. M. Nesladek, K. Haenen, M. Vanecek, in Thin-Film Diamond I, Semiconductors and Semimetals, Vol 76 (C.E. Nebel, J. Ristein, eds.), Elsevier, New York, 2003, pp. 325–377. 59. P. Bergonzo, D. Tromson, C. Mer, Semicond. Sci. Technol. 2003, 18 , S105–S112. 60. P. Bergonzo, R. Jackman, Diamond-Based Radiation and Photon Detectors, in Thin-Film Diamond I, Semiconductors and Semimetals, Vol. 76 (C.E. Nebel, J. Ristein, eds.), Elsevier, New York, 2003, pp. 197–309. 61. J.B. Cui, J. Ristein, L. Ley, Phys. Rev. Lett . 1998, 81 , 429–432. 62. S. Kono, M. Shiraishi, T. Goto, T. Abukawa, M. Tachiki, H. Kawarada, Diam. Rel. Mat . 2005, 14 , 459–465. 63. H. Watanabe, D. Takeuchi, S. Yamanaka, H. Okushi, K. Kajimura, T. Sekiguchi, Diam. Rel. Mat . 1999, 8 , 1272–1276. 64. D. Takeuchi, S. Yamanaka, H. Watanabe, S. Sawada, H. Ichinose, H. Okushi, K. Kajimura, Diam. Rel. Mat . 1999, 8 , 1046–1049. 65. H.H. Okushi, Diam. Rel. Mat . 2001, 10 , 281–288. 66. A.T. Collins, P.J. Dean, E.C. Lightowler, W.F. Sherman, Phys. Rev . 1965, 140 , A1272–A1274. 67. S. Koizumi, M. Kamo, Y. Sato, H. Ozaki, T. Inuzuka, Appl. Phys. Lett . 1997, 71 , 1065–1067. 68. T.H. Borst, O. Weis, Phys. Stat. Sol. A 1996, 154 , 423–444. 69. S. Kuga, J.H. Yang, H. Takahashi, K. Hirama, T. Iwasaki, H. Kawarada, J. Am. Chem. Soc. 2008, 130 , 13251–13263. 70. G. Zheng, F. Patolsky, Y. Cui, W.U. Wang, C.M. Lieber, Nat. Biotechnol . 2005, 23 , 1294–1301. 71. C.E. Nebel, D. Shin, B. Rezek, N. Tokuda, H. Uetsuka, J. Royal Soc. Interface, 2007, 4 , 439–461. 72. C.E. Nebel, D. Shin, D. Takeuchi, T. Yamamoto, H. Watanabe, T. Nakamura, Diam. Rel. Mat . 2006, 15 , 1107–1112. 73. C.E. Nebel, D. Shin, D. Takeuchi, T. Yamamoto, H. Watanabe, T. Nakamura, Langmuir 2006, 22 , 5645–5653. 74. D.C. Shin, N. Tokuda, B. Rezek, C.E. Nebel, Electrochem. Commun. 2006, 8 , 844–850. 75. D.C. Shin, B. Rezek, N. Tokuda, D. Takeuchi, H. Watanabe, T. Nakamura, T. Yamamoto, C.E. Nebel, Phys. Stat. Sol. A 2006, 203 , 3245–3272. 76. N. Yang, H. Uetsuka, E. Osawa, C.E. Nebel, Nano Lett . 2008, 8 , 3572–3576. 77. W. Smirnov, A. Kriele, N. Yang, C.E. Nebel, Diam. Rel. Mat 2010, 19 , 186–189. 78. N. Yang, H. Uetsuka, E. Osawa, C.E. Nebel, Angew. Chemie Int. Ed. 2008, 47 , 5183–5185. 79. C.E. Nebel, N. Yang, H. Uetsuka, E. Osawa, N. Tokuda, O. William, Jpn. New Diamond Forum, 2008, 24 , 10–18. 80. N. Yang, H. Uetsuka, E. Osawa, C.E. Nebel, Diam. Rel. Mat . 2009, 18 , 592–595.

REFERENCES

617

81. N. Yang, H. Uetsuka, C.E. Nebel, Adv. Funct. Mater. 2009, 19 , 887–893. 82. (a) C.E. Nebel, N. Yang, H. Uetsuka, E. Osawa, N. Tokuda, O. William, Diam. Rel. Mat . 2009, 18 , 910–917; (b) N. Yang, H. Uetsuka, O.A. William, E. Osawa, N. Tokuda, C.E. Nebel, Phys. Stat. Sol. A, 2009, 206 , 2048–2056. 83. (a) C.E. Nebel, Science 2007, 318 , 1391–1392; (b) N. Yang, J. Yu, H. Uetsuka, C.E. Nebel, Electrochem. Commun. 2009, 11 , 2237–2240. 84. R. Hoffmann, N. Yang, C.E. Nebel, unpublished results. 85. Q. Yang, W. Chen, C. Xiao, A. Hirose, R. Sammynaiken, Diam, Rel. Mat . 2005, 14 , 1683–1687. 86. Y.K. Chih, Y.L. Chueh, C.H. Chen, J. Hwang, L.J. Chou, C.S. Kou, Diam. Rel. Mat . 2006, 15 , 1246–1249. 87. M.L. Terranova, S. Orlanducci, A. Fiori, E. Tamburri, V. Sessa, M. Rossi, A.S. Barnard, Chem. Mater. 2005, 17 , 3214–3220. 88. D. Luo, L. Wu, J. Zhi, ACS Nano 2009, 3 , 2121–2128. 89. D. Luo, J. Zhi, Electrochem. Commun. 2009, 11 , 1093–1096. 90. D.S. Hwang, T. Saito, N. Fujimori, Diam. Rel. Mat . 2004, 13 , 2207–2210. 91. Y. Ando, Y. Nishibayashi, A. Sawabe, Diam. Rel. Mat . 2004, 13 , 633–637. 92. C.Y. Li, A. Hatta, Diam. Rel. Mat . 2006, 15 , 357–360. 93. E.-S. Baik, Y.-J. Baik, S.W. Lee, D. Jeon, Thin Solid Films 2000, 377 –378 , 295–298. 94. H. Shiomi, Jpn. J. Appl. Phys. 1997, 36 , 7745–7748. 95. H. Masuda, M. Watanabe, K. Yasui, D. Tryk, T. Rao, A. Fujishima, Adv. Mater. 2000, 12 , 444–447. 96. O.A. Shenderova, C.W. Padgett, Z. Hu, D.W. Brenner, J. Vac. Sci. Technol. B 2005, 23 , 2457–2464. 97. K. Kobashi, T. Tachibana, Y. Yokota, N. Kawakami, K. Hayashi, K. Yamamoto, Y. Koga, S. Fujiwara, Y. Gotoh, H. Nakahara, H. Tsuji, J. Ishikawa, F.A. K¨ock, R.J. Nemanich, J. Mater. Res. 2003, 18 , 305–326. 98. Y.S. Zou, Y. Yang, W.J. Zhang, Y.M. Chong, B. He, I. Bello, S.T. Lee, Appl. Phys. Lett . 2008, 92 , 053105. 99. O. Babchenko, A. Kromka, K. Hruska, M. Michalka, J. Potmesil, M. Vanecek, Cent. Eur. J. Phys. 2009, 7 , 310–314. 100. M. Wei, C. Terashima, M. Lv, A. Fujishima, Z.Z. Gu, Chem. Commun. 2009, 3624–3626. 101. A. Szunerits, Y. Coffiner, E. Galopin, J. Brenner, R. Boukherroub, Electrochem. Commun. 2010, 12 , 438–441. 102. A. Kr¨uger, F. Kataoka, M. Ozawa, T. Fujino, Y. Suzuki, A.E. Aleksenskii, A.Ya, Vul, E. Osawa, Carbon 2005, 43 , 1722–1730. 103. O.A. Williams, O. Douheret, M. Daenen, K. Haenen, E. Osawa, M. Takahashi, Chem. Phys. Lett . 2007, 445 , 255–258. 104. O.A. Williams, M. Daenen, D.J. Haen, K. Haenen, J. Maes, V.V. Moshchalkov, M. Nesladek, D.M. Gruen, Diam. Rel. Mat . 2006, 15 , 654–658. 105. N. Tokuda, H. Umezawa, T. Saito, K. Yamabe, H. Okushi, S. Yamasaki, Diam. Rel. Mat . 2007, 16 , 767–770. 106. H.O. Finklea, ed., Semiconductor Electrodes, Elsevier, Amsterdam, 1988, p. 27. 107. I.W. Rangelow, Deep Etching of Silicon, Oficyna Wydawnicza Politechniki Wroclawskiej, Wroclaw, 1996. 108. J. Wang, Analytical Electrochemistry, 2nd ed., Wiley-VCH, New Jersey, 2000.

618

DNA-MODIFIED DIAMOND FILMS

109. S.M. Macdonald, K. Szot, J. Niedziolka, F. Marken, M. Opallo, J. Solid State Electrochem. 2008, 12 , 287–293. 110. N. Yang, H. Uetsuka, H. Watanabe, T. Nakamura, C.E. Nebel, Chem. Mater. 2007, 19, 2852–2859. 111. N. Yang, H. Uetsuka, H. Watanabe, T. Nakamura, C.E. Nebel, Diam. Rel. Mat . 2008, 17 , 892–895. 112. N. Yang, H. Uetsuka, H. Watanabe, T. Nakamura, C.E. Nebel, Diam. Rel. Mat . 2008, 17 , 1376–1379. 113. C.E. Nebel, N. Yang, H. Uetsuka, Surf. Sci . 2009, 603 , L5–L8. 114. M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery, Jr.; R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, P. Salvador, J.J. Dannenberg, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A Pople, Gaussian 98, Revision A.11; Gaussian, Inc., Pittsburgh, PA, 2001. 115. J. Furthm¨uller, J. Hafner, G. Kresse, Phys. Rev. B 1996, 53, 7334–7350. 116. B.M. Nichols, J.E. Butler, J.N. Russel Jr., R.J. Hamers, J. Phys. Chem. B 2005, 109 , 20938–20947. 117. P.E. Colavita, B. Sun, X. Wang, R.J. Hamers, J. Phys. Chem. C , 2009, 113 , 1526–1535. 118. G.P. Lopinski, D.D.M. Wayner, R.A. Wolkow, Nature 2000, 406 , 48–51. 119. R.L. Cicero, E.D. Chidsey, G.P. Lopinski, D.D.M. Wayner, R.A. Wolkow, Langmuir 2002, 18 , 305–307. 120. J. Furthm¨uller, J. Hafner, G. Kresse, Phys. Rev. B 1996, 53 , 7334–7351. 121. M. Werner, Semicond. Sci. Technol. 2003, 18 , S41–S46. 122. K. Bobrov, A.J. Mayne, G. Dujardin, Nature 2001, 413 , 616–619. 123. X. Wang, P.E. Colavita, K.M. Metz, J.E. Butler, R.J. Hamers, Langmuir 2007, 23 , 11623–11630. 124. H. Uetsuka, D. Shin, N. Tokuda, K. Saeki, C.E. Nebel, Langmuir 2007, 23 , 3466–3472. 125. http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/B/BondEnergy.html. 126. J. Pinson, F. Podvorica, Chem. Soc. Rev . 2005, 34 , 429–439. 127. P. Allongue, C.H. de Villeneuve, G. Cherouvrier, R. Cortes, M.-C. Bernard, J. Electroanal. Chem. 2003, 550/551 , 161–174. 128. W. Yang, S.E. Baker, J.E. Butler, C. Lee, J.N. Russell Jr., L. Shang, B. Sun, R.J. Hamers, Chem. Mater. 2005, 17 , 938–940. 129. S. Xu, S. Miller, P.E. Laibinis, G.Y. Liu, Langmuir 1999, 15 , 7244–7251. 130. A. Sassolas, B.D. Leca-Bouvier, L.J. Blum, Chem. Rev . 2008, 108 , 109–139. 131. F. Lucarelli, G. Marrazza, A.T. Turner, M. Mascini, Biosens. Bioelectron. 2004, 19 , 515–530. 132. J.J. Gooding, Electroanalysis 2002, 17 , 1149–1156. 133. J.R. Hamers, J.E. Butler, T. Lasseter, B.M. Nichols, J.N. Russell Jr, K.-Y. Tse, W. Yang, Diam. Rel. Mat . 2004, 14 , 661–668. 134. B. Rezek, D. Shin, T. Nakamura, C.E. Nebel, J. Am. Chem. Soc. 2006, 128 , 3884–3885. 135. S.O. Kelley, J.K. Barton, N.M. Jackson, L.D. McPherson, A.B. Potter, E.M. Spain, M.J. Allen, M.G. Hill, Langmuir 1998, 14 , 6781–6784.

REFERENCES

619

136. D. Erts, B. Polyakov, H. Olin, E. Tuite, J. Phys. Chem. B 2003, 107 , 3591–3597. 137. A. Poghossian, A. Cherstvy, S. Ingebrandt, A. Offenhaeusser, M.J. Schoening, Sens. Actuat. B 2005, 111–112 , 470–480. 138. D. Takeuchi, S. Yamanaka, H. Watanabe, H. Okushi, Phys. Stat. Sol. A 2001, 186 , 269–280. 139. N. Crampton, W.A. Bonass, J. Kirkham, N.H. Thomson, Langmuir 2005, 21 , 7884–7891. 140. (a) D.J. Zhou, K. Sinniah, C. Abell, T. Rayment, Langmuir 2002, 18 , 8278–8281; (b) P.V. Schwartz, Langmuir 2001, 17 , 5971–5977. 141. Y. Ishige, M. Shimoda, M. Kamahori, Jap. J. Appl. Phys. 2006, 45 , 3776–3783. 142. J.-H. Yang, K.-S. Song, S. Kuga, H. Kawarada, Jap. J. Appl. Phys. 2006, 45 , L1114–L1117. 143. S.O. Kelley, E.M. Boon, J.K. Barton, N.M. Jackson, M.G. Hill, Nucl. Acids Res. 1999, 27 , 4830–4837. 144. A. Numnuam, K.Y. Chumbimuni-Torres, Y. Xiang, R. Bash, P. Thavarungkul, P. Kanatharana, E. Pretsch, J. Wang, E. Bakker, J. Am. Chem. Soc. 2008, 130 , 410–411. 145. O. Panke, A. Kirbs, F. Lisdat, Biosens. Bioelectron. 2007, 22 , 2656–2662. 146. H. Aoki, H. Tao, Analyst 2007, 132 , 784–791. 147. C. Fan, K.W. Plaxco, A.J. Heeger, Proc. Nat. Acde. Sci. USA 2003, 100 , 9134–9137. 148. M.A. Lapierre-Devlin, C.L. Asher, B.J. Taft, R. Gasparac, M.A. Roberts, S.O. Kelley, Nano Lett . 2005, 5 , 1051–1055. 149. R. Gasparac, B.J. Taft, M.A. Lapierre-Devlin, A.D. Lazareck, J.M. Xu, S.O. Kelley, J. Am. Chem. Soc. 2004, 126 , 12270–12271. 150. H.-C. Chang, K.W. Chen, S. Kwok, Astrophys. J . 2006, 639 , L63–L66. 151. J.-K. Lee, M.W. Anderson, F.A. Gray, P. John, J.-Y. Lee, Diam. Rel. Mat . 2005, 14 , 675–678. 152. E. Wu, V. Jacques, F. Treussarta, H. Zeng, P. Grangier, J.-F. Roch, J. Lumin. 2006, 119–120 , 19–23. 153. F. Treussart, V. Jacques, E. Wu, T. Gacoin, P., Grangier, J.-F. Roch, Physica B, 2006, 376–377 , 926–929. 154. P.-K. Wei, P.-H- Tsao, H.-C. Chang, W. Fann, C.-C. Fu, H.-Y. Lee, K. Chen, T.S. Lim, H.-Y. Wu, P.-K. Lin, PNAS , 2007, 104 , 727–732. 155. A.M. Schrand, H. Huang, C. Carlson, J.J. Schlager, E. Ohsawa, S.M. Hussain, L. Dai, J. Phys. Chem. B, 2007, 111 , 2–7. 156. J. Wang, J.E. Butler, D.S.Y. Hsu, C.C.-T. Nguyen, Solid State Sensor, Actuator, and Microsystems Workshop Proceedings, Hilton Head Island, SC, 2002, pp. 61–62. 157. J. Wang, J.E. Butler, T. Feygelson, C.T.-C. Nguyen. Proceedings of the Seventeenth IEEE Micro Electro Mechanical Systems Conference, Maastrich, The Netherlands, Jan. 25–29, 2004, pp. 641–644. 158. L. Sekaric, J.M. Parpia, H.G. Craighead, T. Feygelson, B.H. Houston, J.E. Butler, Appl. Phys. Lett . 2002, 81 , 4455–4457. 159. T. Braun, V. Barwich, M.K. Ghatkesar, A.H. Bredekamp, C. Gerber, M. Hegner, H.P. Lang, Phys. Rev. E 2005, 72 , 31907. 160. B. Ilic, Y. Yang, K. Aubin, R. Reichenbach, S. Krylov, H.G. Craighead, Nano Lett . 2005, 5 , 925–929. 161. S.W. Han, C. Nakamura, I. Obataya, N. Nakamura, J. Miyake, Biochem. Biophys. Res. Commun. 2005, 332 , 633–639.