BIOTEMPLATING Complex Structures from Natural Materials
This page intentionally left blank
Simon R Hall University ...
15 downloads
879 Views
28MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
BIOTEMPLATING Complex Structures from Natural Materials
This page intentionally left blank
Simon R Hall University of Bristol, UK
BIOTEMPLATING Complex Structures from Natural Materials
ICP
Imperial College Press
A-PDF Merger DEMO : Purchase from www.A-PDF.com to remove the watermark
Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
BIOTEMPLATING Complex Structures from Natural Materials Copyright © 2009 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.
ISBN-13 978-1-84816-403-1 ISBN-10 1-84816-403-3
Printed in Singapore.
Shelley - Biotemplating.pmd
1
4/8/2009, 3:44 PM
This book is dedicated to Caroline, Corey, Emily and all from 4W 2.25, W504 and S401.
This page intentionally left blank
Contents
1. Introduction 1.1 History of biotemplating ................................................................................... 2 1.2 Mechanisms and models ................................................................................... 4 1.3 Crystallization in nature/skeletons in the beaker ............................................... 5 1.4 References ......................................................................................................... 12 2. Simple Mono- and Oligosaccharides 2.1 Structure ............................................................................................................ 2.2 Use as a source of carbon .................................................................................. 2.3 Use as a chelating agent .................................................................................... 2.4 Other uses.......................................................................................................... 2.5 References .........................................................................................................
14 16 20 23 26
3. Complex Polysaccharides 3.1 Structure and properties..................................................................................... 3.2 Cellulose............................................................................................................ 3.3 Dextran .............................................................................................................. 3.4 Starch................................................................................................................. 3.5 References .........................................................................................................
28 30 38 52 59
4. Hydrocolloids 4.1 Structure and properties..................................................................................... 4.2 Carrageenan....................................................................................................... 4.3 Alginate ............................................................................................................. 4.4 Gelatin ............................................................................................................... 4.5 Agar, curdlan, gellan, pectin and the gums ....................................................... 4.6 References .........................................................................................................
63 65 74 81 86 89
5. Chitin/Chitosan 5.1 Structure and properties..................................................................................... 92 5.2 Chitin/chitosan .................................................................................................. 93 5.3 References ......................................................................................................... 114
vii
viii
Biotemplating
6. Proteins and Lipids 6.1 Proteins – structure and properties .................................................................... 6.2 Lipids................................................................................................................. 6.3 Structure and properties..................................................................................... 6.4 References .........................................................................................................
117 141 141 149
7. Viruses and Bacteria 7.1 Viruses – structure and properties ..................................................................... 7.2 Spherical/polyhedral viruses ............................................................................. 7.3 Helical/filamentous viruses ............................................................................... 7.4 Bacteria – structure and properties .................................................................... 7.5 DNA .................................................................................................................. 7.6 References .........................................................................................................
154 156 157 162 170 172
8. Complex Biostructures as Templates 8.1 Complex biostructures....................................................................................... 174 8.2 References ......................................................................................................... 194 9. Into the Future – Genetic Engineering and Beyond 9.1 Genetic engineering........................................................................................... 196 9.2 References ......................................................................................................... 202
Index .............................................................................................................................. 205
Chapter 1
Introduction
In terms of structural complexity, the natural world produces examples of stunning beauty and high functionality, usually with the minimum of material and energy expenditure. Scientists can harness these amazing structures as readymade scaffolds on which to grow inorganic phases which replicate them, thereby producing materials with greatly enhanced physical properties. With the recent explosion of research into nanotechnology, biomaterials provide ideal templates as complexity in biopolymers is invariably on the nanoscale. This book highlights the wide range of natural materials that have been used in this way and the inorganic phases which result from them. Covering simple molecules such as cellulose and chitin, to large biological constructs such as bacterial proteins, viruses and pollen, practically every inorganic material has been synthesized using biotemplating methods, from simple oxides and carbonates such as silica and calcite, to complex semi- and superconducting materials. The book also discusses the formation of these materials from a mechanistic point of view, thereby enabling the reader to better understand the processes involved in biotemplated mineralization. Many of these materials can be classified in a number of different ways, for example alginate can be considered as a polysaccharide, a hydrocolloid and potentially owing to its behaviour, as a complex biostructure. Inclusion of a biopolymer in one section does not therefore preclude its consideration as one of the others, although to avoid repetition this multiple inclusion is largely avoided. The classifications in this work are based primarily on the properties of the biotemplate being utilized for a particular product.
1
2
Biotemplating
1.1 History of biotemplating Through the four billion years since the first prokaryotic cells appeared, evolution has worked and re-worked life on earth, continually adapting, amending and improving the survivability of organisms in response to a plethora of stimuli. At around 550 million years ago, organisms began to utilize their simple organic molecules in order to grow mineralic phases. These ‘hard’ materials conferred a significant evolutionary advantage, allowing the organisms to survive harsher environments, grow larger, evade predation and so on. As a result, these simple organic molecules eventually became part of structurally complex organic matrices, specifically tailored to biomineralize inorganic phases which precisely fit form to function. This means that today we have at our fingertips, an entire world full of organic matter, often with built-in nano-scale complexity, ready to be pressed into use as templates for the creation of complex functional materials. The processes involved in creating architectural elegance using the minimal amount of material has long fascinated Man, who has endeavoured to understand how such intricate construction can be accomplished through the simple flow of inorganic ions and strategically placed macromolecules. “When the demands of the environment are the blueprints of the construction, structures are produced with the utmost efficiency”. This quotation from D’Arcy Thompson, in his seminal work ‘On Growth and Form’ (1917) represents the kernel of what first drove Man to attempt synthesis based on naturally occurring materials and methods1. Even as far back as the 16th century, scientist and astronomer Johannes Kepler noted that ‘Nature uses as little as possible of anything.’ Both these and many other luminaries throughout history have held nature in the highest esteem as an engineer par excellence. This is perhaps best exemplified by the extraordinary feats of engineering undertaken throughout the Victorian-era, when engineers turned to nature for inspiration, when stable, complex constructions were required. One of the earliest ‘bioinspired’ architectural projects was the construction of the Crystal Palace for the Great Exhibition of 1851. The architect Joseph Paxton conceived the Crystal Palace largely as a result of his work as Head Gardener to the Duke of Devonshire at Chatsworth House, Derbyshire. Whilst at Chatsworth, Paxton built the largest conservatory in the World at that time, utilizing glass and iron for strength and durability. In 1837, the arrival of a lily from Guyana required a custom-built heated pool which Paxton designed. He was intrigued by the huge leaves of the plant which he dubbed ‘a natural feat of engineering’ and
Introduction
3
tested their strength by floating his daughter on one of them. The secret of their mechanical stability was clear to Paxton; an array of radiating ribs connected with flexible cross-ribs. Experimentation over the following years enabled Paxton to improve on his glass and iron structures, culminating in the incorporation of the waterlily’s structural features in his design for the Crystal Palace. Another striking example of engineering inspired by nature can be seen in Isambard Kingdom Brunel’s Royal Albert Bridge near Plymouth. The bridge is a clever combination of arch and suspension bridge. An arch bridge produces a net outward thrust at the abutments, whereas a suspension bridge pulls the abutments inwards. By combining the two concepts in one bridge, the overall force at the abutments is almost zero. In Brunel’s bridge, the arches consist of iron tubes with an oval cross-section, which produce the outward thrust to balance the inward pull of the draped chains. The minimal force carried by the abutments allows for a lighter and more importantly, cheaper construction. Inspiration for this may have come from the observation that this method of force balancing is one which every four-legged animal adopts. For example, in an elephant, the legs are the abutments, the belly the chains and the spine the arch of the bridge. Although it is not known whether Brunel (in the manner of Paxton) first considered nature before embarking on his design, it is likely that such a natural analogue would not have been far from his mind. In the 20th century, scientists began to take a more active interest in the architectural constructs of the biological world, particularly keen to understand the procedures used by flora and fauna in the production of inorganic structural elements. The father of this approach was R.J.P. (Bob) Williams of Oxford University, who instigated a study of the detailed functional use of inorganic elements in biological systems2. By applying principles from inorganic chemistry such as the complex-ion formation and redox potential, to biological systems, he was able to deduce many hitherto unknown biomineralization mechanisms. Among the discoveries from this time were the elucidation of the special inorganic chemistry of unusual metal binding sites in nature, and the role and mode of action of calcium in the formation of calcified structures3. One of the students of the Williams’s ‘Oxford School’ was Stephen Mann. Mann realised that by understanding the processes of biomineralization in terms of the movement and precipitation of inorganic elements within a ‘biological environment’ it should be possible to replicate or mimic them under laboratory conditions4. Mann supposed that as mineralization usually takes place due to constraint within an organism, then by replicating those constraining factors
4
Biotemplating
synthetically, either physically in the form of for example, vesicles, or chemically by control of localized supersaturation, bio-analogues could be created. These experiments in ‘biomimetics’ yielded complex and often strikingly ‘lifelike’ inorganic materials by following closely (but not exactly) the protocols used in the natural analogue5. Currently, research World-wide into biomimetic control of mineralization is strong, producing many diverse and often industrially valuable materials6. All syntheses however still rely either directly or indirectly on the ‘boundary organized biomineralization’ concepts introduced by Williams, Mann et al. The complexity of biological structures and the complex systems which give rise to them are not easily replicated in the laboratory. Even the most advanced and succinct synthetic protocols can only ever offer a poor imitation of the natural analogue. This has convinced many scientists to ‘cut out the middle man’ and directly utilize naturally occurring materials as part of their synthetic procedures. The advantage of this approach is clear; by using a pre-formed, often hierarchically complex material, the scientist aims to transfer the physical properties of the original, to that of the synthetic analogue. It is this simple and elegant synthetic protocol that is the raison d’être of this book. After two decades of concerted research, the time is right for an overview of the research that has been done and that can still be done in the field of biotemplating. This book gathers together for the first time, research on virtually every biomaterial that has been used as a template for mineralization; from simple monosaccharides and peptides, to macromolecular complex bioconstructs such as pollen, diatoms and cuttlebone. In doing so, it is hoped that the reader will get a feel for the breadth and depth of research in biotemplating and perhaps stimulate further research in this most fascinating of fields. 1.2 Mechanisms and models The fact that biomineralization and therefore biotemplating succeeds so spectacularly is largely due to the complimentary interaction between oppositely charged entities (ions, molecules, etc.). There are however, other effects which play a part and other ways of conceptualizing the process of biomineralization/biotemplating which sometimes better describe the effect being observed. The rest of this first chapter discusses the interaction between organic and inorganic phases and explores what happens when soft meets hard. The mechanisms and models described in this chapter apply to all of the examples of biotemplating which follow in this book. If the mechanism is not explicitly stated
Introduction
5
in the discussion of these materials, the reader is invited to revisit this chapter and deduce the mode of interaction at play in that particular example. 1.3 Crystallization in nature/skeletons in the beaker The process of crystallization is well understood, although in practice, there are many factors which can influence the growth of a crystalline solid from a solution. This means that even after careful determination of the potential for successfully growing a particular crystalline phase, it can be difficult to achieve in practice. Broadly speaking, the entire process can be broken down into a nucleation event followed by subsequent crystal growth. The inducement for nucleation is the formation of a stable cluster of ions in solution. This process is dynamic and many nucleation events occur in solution only for the cluster to dissipate as soon as it has been formed. It is only once a critical size has been reached that the stability conferred on the cluster by aggregation allows it to persist without reverting back to individual ions. Spontaneous nucleation and growth in the absence of any seed is considered to be a rare event. Even supersaturated solutions of some compounds will remain uncrystallized providing there is no contamination or disturbance to the system. Once a suitable seed has provided the impetus for a nuclei to persist, crystal growth can occur. Providing the system is in a state of supersaturation, this process will be repeated continuously, with the number of nucleation sites increasing with time until the concentration of ions in solution dips below supersaturation and the system reaches equilibrium. Crystal growth can then occur as ions in solution add to the nucleated clusters. It is obvious then, that the more seed sites that are present, the more likely (all other things being equal) that crystallization will occur. In these cases of ‘classical’ crystallization, both thermodynamic and kinetic factors play an important role in determining the rate of crystal growth and several models were proposed in order to account for the observed formation of crystals within any given system7-10. The kinetic control of crystallization is a key concept in biomineralization and biotemplating. Crystallization under kinetic control can be thought of conceptually as a progressive modification of the activation-energy barriers of nucleation and growth. This step-wise progression can lead to the appearance of several intermediate species in biomineralizing systems, often with the starting point an amorphous precursor phase (Figure 1.1). The progression along the sequence to the final form of the biomineral is entirely dependent on the
6
Biotemplating
solubilities of the inorganic phases present and on the free energies of their interconversions11.
Figure 1.1 – Crystallization pathways under thermodynamic and kinetic control. Whether a system follows a one-step route to the final mineral phase (pathway A) or proceeds by sequential precipitation (pathway B), depends on the free energy of activation (∆G) associated with nucleation (n), growth (g), and phase transformation (t). Amorphous phases are common under kinetic conditions. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
As these step-wise changes of structure in the inorganic phase proceed by way of dissolution-reprecipitation, this process is further complicated by the hydrodynamic properties of the ions in solution. In nature however, the mechanism of crystallization cannot simply be accounted for by these phenomena, as the presence of organic matter serves as a ready-made substrate for the formation of nuclei, thereby allowing crystallization to occur more favourably. One of the earliest studies on the effect of molecular additives to crystal growth was that of Buckley who proposed an essentially epitaxial mechanism to account for the observed morphology of certain crystals when
Introduction
7
grown under these conditions7. By interacting with specific crystal faces, the molecular additive ‘poisons’ a particular face or faces to further growth by adhering preferentially to that surface. Crystals grown under the influence of molecular additives therefore have the potential to develop into non-classical morphologies, directed by the specificity of the additive to different crystal faces in different degrees. By extension, this concept can be applied to biomineralizing entities, with the organic elements not only inducing crystal growth and directing morphology, but also providing a macroscopic scaffold on which crystal growth occurs. In an early work on how this concept can be applied to biotemplating, Mann and co-workers described that it is possible to draw analogies between the formation of inorganic nuclei on the surface of an organic matrix and the interaction between an enzyme and substrate12. In each case, nuclei can be considered to be kinetically stabilized by specific molecular interactions with the surface layer of the organic material. The overall effect that the organic substrate has is to lower the activation energy of nucleation (∆G#). Furthermore, it is entirely likely that as different sets of symmetry-related crystal faces show different levels of complimentarity for organic substrates, ∆G# may be dependent on the absolute 3D structure of the organic matrix, leading to further complexity of mineralization. From experiments done since these early studies on complimentarity, it appears that the prime factor in organic-inorganic recognition is the charge matching between the inorganic ions and appropriate unlike charges on functional groups of the organic substrate. A good example of this in nature is in the biomineralization of calcitic structures. The majority of organisms which biomineralize calcium carbonate have organic fragments which are rich in carboxylate (COOH) groups. Charge matching is therefore possible between the COO- anions and Ca2+ cations, leading to preferred sites of nucleation for the subsequent growth of calcium carbonate. This complimentarity immediately suggests a model for the long range directed growth of the inorganic phase on the substrate. As the organic matrix in a biomineralizing organism will be (usually) a protein or polysaccharide, the organic fragments which carry the appropriate charge for inducing inorganic ion binding will be disposed in a regular manner across the surface of the substrate. We will see in later chapters how the construction of macro-molecular assemblies with complex morphologies can be used effectively to replicate this complex form as a result of the regularity of reactive organic fragment disposition. This leads to preferred nucleation and
8
Biotemplating
growth of inorganic material at specific locations on the organic fragment. This direct ‘epitaxial matching’ of organic to inorganic entity has been postulated and indeed observed to occur in certain cases. In work by Mann et al., it was found that short-chain α-ω-dicarboxylic acids [(CH2)n(CO2H)2] are particularly effective at stabilizing faces parallel to the {11.0} surface of calcite provided that both carboxylate groups are in their ionized form13. The group postulate that this is due to the fact that these faces contain both Ca2+ and CO32- ions with the latter oriented such that the plane of the triangular anion is perpendicular to the surface. This leads to a direct matching of the carbonate anions into the {11.0} face during growth, through bidentate binding of two of the three oxygen atoms to Ca2+ ions in the surface. By simulating a calcite crystal surface, the group discovered that both carboxylate groups in the additive molecule would bind simultaneously to two different calcium ions if the spacing between the CO2groups is close to 0.4 nm (Figure 1.2). This leads to the conclusion that by altering the chain length of the additive molecule, specificity can be controlled in a very precise manner. For example, both malonate (n=1) and the unsaturated diacid, maleate (cis--O2CCH=CHCO2-) will adopt an epitaxially matched conformation on the calcite surface, but the more rigid conformation adopted by the maleate ion will reduce the binding affinity. The sensitivity of this epitaxial matching model to absolute conformation was confirmed by the group in experiments on the diacid trans-isomer, fumarate, which they found had no effect on the control of calcite morphology owing to the molecule being the “wrong” shape to take part in the co-operative binding. These epitaxial effects are pronounced at lower additive concentrations, as higher concentrations will lead to non-specific binding of the additive over all crystal faces14. Similar work by Mann and Heywood identified the formation of oriented barium sulfate phases by the interaction between long-chained sulfated molecules and barium ions15. Stabilization of the {011} set of faces in BaSO4 were found to be a result not only of the ions possessing the correct stoichiometry to be structurally complimentary to the organic phase, but also to be highly polar, leading to a stronger interaction between the organic and inorganic phases. A study by Weiner et al. found that this model has a natural analogue16. Acidic macromolecules extracted from adult sea urchins were found to interact specifically with calcite prismatic faces lying almost parallel to the {11.0} surface. These acidic molecules have a large number of glutamic and aspartic acid residues, which mimic the coordination environment of ions in the {11.0} face by binding to the growing surface of the calcite. Low concentrations of
Introduction
9
naturally-occurring longer biopolymers such as carrageenans and alginates have also been observed binding specifically to crystals, adopting preferred configurations along edge sites17. In this way, nucleation and growth of the edges of crystals are inhibited and non-classical morphologies begin to dominate.
Figure 1.2 – Perspective drawing of the calcite {1-1.0} face showing a possible binding site for malonate anion. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
This epitaxial model of biotemplated growth is one which enables the researcher to reconcile the formation of inorganic phases with certain crystallographic features, but it only partially explains the growth of crystal phases in the presence of organic macromolecules. Another important consideration is the electrostatic environment which surrounds the nucleation centre. The localization of specific inorganic binding entities will concentrate areas of electrostatic charge, which will further improve the specificity of that part of the organic fragment for the inorganic phase. For example, the presence of glutamic acid residues in the inner cavity of the hollow spherical iron transport protein ferritin increase the electrostatic field in the inner surface of the protein relative to the outer surface. This increases the likelihood that iron sequestration
10
Biotemplating
will take place inside the capsid rather than on the outside18. Work by Yamashita et al. revealed that the rate of growth of hydroxyapatite could be controlled by altering the electric polarization conditions of the synthesis19. Hydroxyapatite is polarizable owing to the ease of reorientation of dipole moments between O2- and H+ of the OH- ions in the crystal lattice. By performing the crystallization in the presence of an electric field of 1,000 V, larger crystals of hydroxyapatite could be formed than when the electric field was absent. In addition, the group found that there was a linear relationship between the field strength (and hence polarity of the hydroxyapatite) and speed of crystal growth. At an optimum polarization, crystals of hydroxyapatite were found to grow at six times the rate of nonpolarized samples. When the same experiments were undertaken with dehydrated hydroxyapatite (an absence of hydroxyl groups), no enhancement of the crystal growth rate was observed. The presence of polarizable hydroxyl groups in this case clearly leads to improved crystallization. The mechanism of enhanced growth is clearly an electrostatically driven one as it is not difficult to imagine that cations (in this case Ca2+) are preferentially adsorbed to the more polarized surface of the nucleation centre. The growth of these nuclei are therefore accelerated by the presence of stronger dipole moments. Once a nucleation/growth event is underway and the inorganic phase is a viable (i.e. stable) entity, the possibility exists for the further growth of the newly formed crystallites following the topology of the organic material. In the aforementioned work by D’Arcy Thompson ‘On Growth and Form’, the scene was set for the conceptualization of inorganic crystal growth that was not of the ‘classical’ morphology of straight edges and fixed angles. At longer length scales, it is the influence of mechanical stress and gravity which determine the macroscopic shape of a scaffold. In order to better understand this process, Mann et al. proposed that conceptually, the production of macroscopic, three dimensional structures by the mineralization of an organic scaffold could be considered as either dependent on the chemical and spatial modification of crystal growth (contingent) or as a consequence of the spatial conformation adopted by the organic structures (prescribed)20. As examples, he cites the formation of calcitic spicules in some corals as contingent growth and the formation of delicate siliceous skeletons of diatoms as being an example of prescribed growth. Coralline spicules are polycrystalline calcitic structures which form in discrete vesicles within the body of the coral. The shape that the spicules adopt is determined by the local environment in which the deposition body finds itself; disturbances of the reaction volume and concentration fluctuations across
Introduction
11
the fluid-solid interface are largely responsible for the final morphology of the calcite. This means that spicules, even within the same organism, are not necessarily morphologically identical, rather that they may loosely resemble one another. The stunning accuracy of reproduction of fine skeletal structure in diatoms therefore cannot be explained as contingent growth. The regularity of morphology produced within a single species of diatoms points towards a much more organized mineralization protocol, one which in addition to relying on the compartmentalization of a reaction volume, has that volume itself spatially preorganized within the organism. With a pre-defined environment in which the mineralizing volume sits, the organism is better able to control the local chemical environment by directed transport of ions and molecules to and from the site. The remainder of this book highlights how these basic models of organicinorganic interaction are pressed into use in the formation of complex functional materials, often with enhanced emergent properties. Early studies on biotemplating involved inorganic materials with a known affinity for bio-organic matter; calcium carbonate, calcium phosphate and iron oxide phases feature heavily, with researchers concentrating on developing alternate crystal morphologies through the use of biotemplates not normally associated with that particular mineral in nature. More recently, groups have developed biotemplating strategies as a means to achieving complex form in more ‘functional’ materials such as those which exhibit semi- and superconducting, non-linear optical, photonic and electronic behaviour. In addition, the use of increasingly complex materials as biotemplates, for example micellar and surfactant-like long chained molecules and wood, wings, chitin and cuttlebone have enabled the construction of complex inorganic materials in 3D. By providing a macromolecular organic scaffold, inorganic phases can be constructed not only with control over crystal growth, but also be ordered into micro- and macrostructures. In 1995, Mann identified three ‘Main Thrusts’ in the new field of biomimetics, of which Biotemplating plays a significant part21. They were; 1) Use of natural materials to control crystallization chemically and structurally, 2) Use of living organisms to deposit inorganic material and 3) Using biomineralization concepts to direct syntheses. He concluded that the success of the fields of biomimetics and bio-inspired materials chemistry would rely on the interdisciplinary nature of research and more critically, on the training at an early level, of students who would come through the academic system with an open and communal view of science, rather than that of the traditional ‘boxes’ of Physical, Organic and Inorganic chemistry. We are beginning to see the
12
Biotemplating
emergence of the first green shoots of this ethos, with young academics now leading research groups who embrace this interdisciplinarity. Nowhere more than in biotemplating research is this evident; scientists are now routinely invoking principles and practices which run the whole gamut of science, from biology to physics. The wide intellectual net that is cast by biotemplating has caught the imaginations of many thousands of research groups worldwide and in doing so has hopefully managed to, as Mann requested, ‘unlock our imagination from the straightjackets of conventional disciplines’. The syntheses of anisotropic nanoparticles discussed in this book have been the subject of a Review published in 200922. Aspects of the history of biotemplating in this book have appeared in a previous Royal Society publication23.
1.4 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
Thompson, D.A.W. On Growth and Form, Edn. 1st. (Cambridge University Press, 1917). Williams, R.J.P. Metal Ions in Biological Systems. Biological Reviews of the Cambridge Philosophical Society 28, 381-415 (1953). Williams, R.J.P. Inorganic Biochemistry and Myself. J. Inorg. Biochem. 28, 81-84 (1986). Mann, S. On the Nature of Boundary-Organized Biomineralization (BOB). J. Inorg. Biochem. 28, 363-371 (1986). Walsh, D. & Mann, S. Fabrication of Hollow Porous Shells of Calcium Carbonate from Self-Organizing Media. Nature 377, 320-323 (1995). Ratner, B.D. & Bryant, S.J. Biomaterials: Where we have been and where we are going. Annu. Rev. Biomed. Eng. 6, 41-75 (2004). Buckley, H.E. Habit Modification in Crystals as a Result of the Introduction of Impurities During Growth. Discussions of the Faraday Society, 243-254 (1949). Burton, W.K. & Cabrera, N. Crystal Growth and Surface Structure. 1. Discussions of the Faraday Society, 33-39 (1949). Cabrera, N. & Burton, W.K. Crystal Growth and Surface Structure. 2. Discussions of the Faraday Society, 40-48 (1949). Whetstone, J. The Adsorption of Dyes by Crystals. Discussions of the Faraday Society, 132-140 (1954). Colfen, H. & Mann, S. Higher-order organization by mesoscale self-assembly and transformation of hybrid nanostructures. Angew. Chem.-Int. Edit. 42, 2350-2365 (2003). Mann, S. et al. Crystallization at Inorganic-Organic Interfaces - Biominerals and Biomimetic Synthesis. Science 261, 1286-1292 (1993). Mann, S. The chemistry of form. Angew. Chem.-Int. Edit. 39, 3393-3406 (2000). Heywood, B.R. & Mann, S. Template-Directed Nucleation and Growth of Inorganic Materials. Adv. Mater. 6, 9-20 (1994). Heywood, B.R. & Mann, S. Crystal Recognition at Inorganic Organic Interfaces - Nucleation and Growth of Oriented BaSO4 under Compressed Langmuir Monolayers. Adv. Mater. 4, 278-282 (1992).
Introduction 16.
17.
18. 19. 20. 21. 22. 23.
13
Berman, A., Addadi, L. & Weiner, S. Interactions of Sea-Urchin Skeleton Macromolecules with Growing Calcite Crystals - A Study of Intracrystalline Proteins. Nature 331, 546-548 (1988). Butler, M.F., Glaser, N., Weaver, A.C., Kirkland, M. & Heppenstall-Butler, M. Calcium carbonate crystallization in the presence of biopolymers. Cryst. Growth Des. 6, 781-794 (2006). Wade, V.J. et al. Influence of Site-Directed Modifications on the Formation of Iron Cores in Ferritin. J. Mol. Biol. 221, 1443-1452 (1991). Yamashita, K., Oikawa, N. & Umegaki, T. Acceleration and deceleration of bone-like crystal growth on ceramic hydroxyapatite by electric poling. Chem. Mat. 8, 2697-2700 (1996). Mann, S. Biomineralization: the form(id)able part of bioinorganic chemistry! J. Chem. Soc.Dalton Trans., 3953-3961 (1997). Mann, S. Biomineralization and Biomimetic Materials Chemistry. J. Mater. Chem. 5, 935-946 (1995). Hall, S.R. Biotemplated syntheses of anisotropic nanoparticles. Proc. R. Soc. A, 465, 335-366 (2009). Davies, A.G. & Thompson, J.M.T. (eds) Advances in Nanoengineering: Electronics, Materials and Assembly. 328pp, World Scientific Press, 2007.
Chapter 2
Simple Mono- and Oligosaccharides
In the previous chapter, we have seen how biotemplates are able to induce the changes in crystallization of inorganic phases, either by chemical or physical constraint. In terms of morphology, the simplest sugars rarely form complex structures as they crystallize with no long-range order. This does not preclude their use as a biotemplate however, as many researchers have ingeniously employed them as methods to produce inorganic materials with complex form.
2.1 Structure Monosaccharides are the simplest carbohydrates. As a group, it is the carbohydrates that are perhaps the most important to life on earth. The process of photosynthesis, the primary means by which life is sustained, is a biosynthetic process concerned solely with the manufacture of carbohydrates. As a store of energy the carbohydrates provide sustenance as food, but also can be used to generate heat and light after a period of geological processing, as oil, coal and peat. Monosaccharides have the generalized chemical formula (CH2O)n + m with the structure H(CHOH)nC=O(CHOH)mH. They are generally colourless, crystalline solids with an uncomplicated structure (Figure 2.1). With a wide range of compositions, there are many possible monosaccharides, each varying from the others in terms of the number of carbon atoms they contain and whether they contain an aldehyde or ketone group. The most commonly encountered monosaccharides are the simplest. They include glucose, fructose and galactose, all found widely in nature. The simplest monosaccharides have found little use in biotemplating owing to their inability to form extended complex structures of their own volition. Of slightly more use are the simple oligosaccharides as they are able to chelate ions more readily and to begin to form complex macromolecular structures of their own. There have been
14
Simple Mono- and Oligosaccharides
15
a few examples where mono- and oligosaccharides have been used and it is on these that this chapter will concentrate.
Figure 2.1 – Molecular representation of the structure of fructose.
Di- and oligosaccharides are simply composed of monosaccharide units (typically 2 to ten) joined together, usually through a condensation reaction forming a glycosidic bond (Figure 2.2).
Figure 2.2 – A molecule of sucrose with the glycosidic bond indicated.
As any of the hydroxyl groups on each component sugar unit can form the glycosidic bond, there exists the potential for a huge range of isomers of even very simple oligosaccharides. Differences in stereochemistry in the disaccharides can result in isomers with very different physical and chemical properties. In addition, as they contain aldehyde or ketone groups, the mono-, di- and
16
Biotemplating
oligosaccharides are able to take part in simple chemical reactions. For example, reduction of glucose gives an alcohol and mild oxidation gives a carboxylic acid, typical behaviour for compounds containing an aldehyde group. This limits their use as templates for inorganic mineralization, as many inorganic phases are subject to reduction when saccharides are present. Conversely, it is just this sort of activity that can be used in a synthesis to arrive at the reduced form of a desired mineral.
2.2 Use as a source of carbon As a biotemplate, the mono- and oligosaccharides are most frequently used as a convenient source of carbon, either through chemical oxidation by acids, or by the action of heat in inert gas environments. Carbon, particularly when finely divided or highly porous, has found numerous applications in electrochemistry, catalysis and as an adsorbent1-5. In terms of a structured template however, many applications of saccharides involve simple filling of a porous material which is itself acting as a template, so no morphology is conferred directly by the sugar. Once filled, the composite materials are then subsequently calcined or acted on chemically to carbonize the sugar, thereby replicating the fine pore structure of the template. By careful consideration of the porous body, it is possible to produce the carbonaceous phase, then to subsequently remove the template, thereby leaving a monolithic carbon material with complex form. The first time this was achieved was by Ryoo et al. In their groundbreaking work, they infiltrated the mesopore structure of a siliceous molecular sieve (MCM-48) with sucrose solution and then carbonized the sucrose using sulfuric acid6. Removal of the silica molecular sieve was able to be achieved by immersion of the composite in sodium hydroxide solution, leaving a microporous 3D-carbon replica. In addition, they found that their material (designated CMK-1) contained mesopores of around 3 nm in diameter. This was postulated to be due to a systematic transformation of the structure, which occurred during the removal of the silica framework. A change in the powder X-ray diffraction (XRD) patterns of materials before and after silica removal revealed that the final carbon material was subtly structurally distinct from the composite. Before framework removal, the XRD pattern was able to be indexed to the cubic Ia3d space group of MCM48 (with a small lattice contraction due to thermal effects on the lattice). On removal of the silica, a new peak appeared in the XRD pattern, which corresponded to the (110) reflection for Ia3d. This reflection is symmetrically
Simple Mono- and Oligosaccharides
17
forbidden for Ia3d and therefore indicates that a new ordered structure has been formed, one in which the (110) reflection is not forbidden. The XRD pattern of CMK-1 had no peaks at 2θ values higher than 10°, which showed that the carbon framework was disordered at the atomic scale. Adsorption-desorption isotherms and pore size distribution measurements for CMK-1 showed the presence of pores 3.0 nm in diameter, which would not be expected had the carbonization exactly replicated the pore structure of MCM-48. Structural transformation of the CMK-1 was therefore attributed to strain in the carbon frameworks induced by silica template removal. The CMK-1 material contained micropores of between 0.5–0.8 nm in diameter in addition to the 3.0 nm mesopores, although the total pore volumes of CMK-1 materials were relatively low at around 1.1 cm3g-1. This particular technique opened the way for the creation of porous carbon with structural periodicity at low dimensions. This was important, as up until that point infiltration of carbon sources into micro- and mesopores was particularly difficult, owing to carbonization in macropores producing thick carbon coatings which remained immutable on removal of the template. Since this work was published, many other similar examples have been published7-11. Porous materials such as mesoporous silicas and zeolites have been filled with mono- and oligosaccharides. One such route was reported by Bohme et al. In this work a zeolite was used as a highly porous material which was filled with sucrose12. After pelletizing under pressure and treating with ethene to encourage the formation of pores, the composite materials were calcined in an inert atmosphere at 800 °C for 3 hours. The resulting carbon/zeolite composites were then treated with HF in order to remove the zeolite matrix. The carbon monolith had complex form, not only replicating that of the underlying zeolite template, but also itself possessing porosity. In the filling of the zeolite matrix, slight imperfections in the packing of the voids by sucrose lead to eventual mesopores in the final carbonized material. One application of these materials is in electrochemical hydrogen storage. Song et al.13 converted sucrose to carbon chemically, by the action of sulfuric acid on the composite materials. As before, the ordered carbon materials exhibited porosity on the mesoscale. This proved to be important in demonstrating hydrogen storage, as the carbon material with higher specific surface areas of 720 m2g-1 and pore volume of 0.86 cm3g-1 naturally had a higher hydrogen storage capacity than carbon materials with lower specific surface areas (61 m2g-1) and pore volumes (0.66 cm3g-1). What was surprising was that both of the two ordered mesoporous carbons exhibited higher hydrogen storage
18
Biotemplating
capacities than single-walled carbon nanotubes. Also, cyclic voltammetry indicated that the ordered mesoporous carbon possessed a higher electrochemical activity than single-walled carbon nanotubes. This opened up the possibility for a low-cost route to better hydrogen storage materials as carbon nanotubes tend to be much more expensive than the sucrose-templated method. Sucrose has also found application in the synthesis of even more complex carbon-containing materials. Spong et al. have discovered that by carbon-coating the olivine phase LiFePO4, a key component of lithium iron batteries, the capacity of the batteries can be improved14. Specifically, they prepared the lithium-ion batteries via a novel, one-step, low-cost synthesis route using aqueous precursor solutions of Fe(NO3)3, LiCH3COO, H3PO4 and sucrose. Sucrose acts in this system to suppress crystallization during the initial stages of the synthesis, which is important in ensuring homogeneity of the reactants and as an aid to achieving the correct olivine phase at lower temperatures. Furthermore, owing to the aldehydic nature of the sucrose molecule it acts as a reducing agent in the synthesis. This is all in addition to its use in the more usual role of carbon source when the temperature of the reactant system is raised. The carbon network in the final material acts as a conductive network which permeates the entire structure of the LiFePO4 battery, leading to a higher capacity than materials in which sucrose was substituted with acetylene black. An additional treatment with sucrose at 700 °C, allowed the material to achieve a capacity of 162 mAh g−1 at the C/14 rate and 158 mAh g−1 at C/3.5 in the voltage range 2.0/4.5 V vs lithium alone. One interesting development in the use of sucrose was in the synthesis of silicon carbide materials with high surface areas. Mokaya et al. showed that SiC was able to be synthesized as whiskers and nanotubes, via a synthetic route using mesoporous silica as sacrificial template15. The SiC materials were obtained via carbothermal reduction of a mesoporous silica which contained carbon (Figure 2.3). Varying the temperature and length of carbothermal reduction, it was found that the morphology of the silicon carbide could be modified. In addition to bulk SiC material, a large proportion of nanotubes and whiskers were able to be obtained, with a high degree of crystallographic control. It was determined that the optimum conditions for whiskers to grow in the (111) direction was when the calcination proceeded at high temperatures, typically in the range 1,250 °C to 1,300 °C and for 14 hours.
Simple Mono- and Oligosaccharides
19
Figure 2.3 – SiC nanotubes synthesized at 1250 °C from mesoporous silica and sucrose. Reprinted with permission from15. Copyright 2004 American Chemical Society.
Whisker diameters were found to be in the region of 50 nm to 90 nm, with lengths well over 20 µm. In terms of surface area it was found that SiC materials with whiskers had a surface area of 120–145 m2/g with a pore volume of 0.42 cm3/g. Increasing the carbothermal reduction to 20 hours at the same calcination temperatures resulted in the formation of solid SiC nanotubes of diameter 60–100 nm and around 10 µm in length. The compaction of the material gave a higher surface area of up to 190 m2/g. Sucrose as a carbon source was also used in the preparation of composite clay materials. Intercalation of sucrose, followed by ‘caramelization’ produced macromolecular intercalation compounds. In this work, Ruiz-Hitzky et al. managed to produce nanocomposite materials using a microwave assisted reaction of sucrose with a phyllosilicate clay16. By microwaving physical mixtures of sucrose and clays, intercalative polymerization of sucrose resulted which formed a hybrid material with enhanced mechanical properties.
20
Biotemplating
Further heating of the sample to 750 °C for three hours caused the caramel to carbonize, producing a carbonaceous composite material which was shown to be porous and electrically conductive. The measured conductivity values for the caramel–clay nanocomposite materials prior to carbonization were typically in the range exhibited by insulators, namely 10-12 S cm-1. This changed on carbonization however, with the electrical conductivity increasing to 10-4 S cm-1 at room temperature, making these composites suitable for use in electrode technologies. This essentially dual role of the sugar, as source of carbonaceous matter and also as a crystal growth mediator/morphological directing agent opens up the way for the crystallization of many other inorganic phases to be controlled by the incorporation of sugars into the synthetic protocol.
2.3 Use as a chelating agent Sucrose has been investigated in the control of mineral phase growth, acting as a chelating agent during synthesis, passivating the growing crystal surfaces and limiting the particle size. Almasi et al. used sucrose in this way to form the mineral forsterite (Mg2SO4) from an aqueous solution of magnesium nitrate and colloidal silica17. Addition of a sucrose solution to a sol of magnesium nitrate and colloidal silica encouraged the chelation of MgII ions once nitric acid had decomposed the glycosidic link in sucrose. The overall effect was to provide – OH and –COOH ions to ensure the sol produced was totally homogeneous, essential for the growth of nanoparticulate phases. Once gelled, calcination of the material resulted in carbonization of the saccharides, leaving a porous solid. Evolution of gasses from the saccharide decomposition prevented the sintering of forsterite nanoparticles and encouraged effective mixing of the precursors during the process, both effects leading to retention of nano-morphology (Figure 2.4). Hydroxyapatite (HAp) has been synthesized using a similar protocol. Of all the biominerals investigated, hydroxyapatite is perhaps the most intensively researched, owing to its use in orthopedic and dental replacement/augmentation technologies18-23. As the principal constituent of human (and other animal) bones, research in this area is driven by the need to model and understand the organic/inorganic interaction between the mineral phase and the templating organic phase. In humans, the templating phase is represented primarily by Type I collagen, a fibrous protein able to form bundles with incredible strength and elasticity. The collagen provides a growth matrix for the HAp crystals, limiting
Simple Mono- and Oligosaccharides
21
their size and imparting a preferred orientation to the crystallites. Once fully formed, the organic/inorganic hybrid material has a high flexibility, imparted by the collagen and a high compressive strength from the HAp24, 25. In order to replicate this, researchers are constantly looking at ways to achieve a symbiotic organic/inorganic strengthening effect in vitro.
Figure 2.4 – SEM micrograph of synthesized forsterite powder. Precursor powders were calcined at 800 °C for 3 h. Reprinted from17. Copyright 2007, with permission from Elsevier.
This is particularly important when the bone replacement is to be used in load bearing applications as mechanical failure in vivo is best not countenanced. One problem with synthesizing HAp in vitro is that when grown in the absence of any additive or template, the crystals have a tendency not to sinter together well and as a result the product will tend to be friable and mechanically very weak. This is manifest particularly in hydrated environments, which makes pure HAp wholly unsuitable for implantation as a bone replacement and even as a filling material for non- or low load bearing implant applications. It has been
22
Biotemplating
determined that with a higher surface area, nano-sized HAp will be more amenable to sintering and thereby produce a more structurally sound product. Saha et al. utilized the same chelation/decomposition of sucrose as in the forsterite work in order to produce nano-sized HAp crystals26. By performing a chelation/gelation with sucrose, HAp powders were obtained with a uniform morphology and a particle size of between 30 nm and 50 nm. An indication of the generality of this approach was shown by Bose et al. and Ram et al. in the synthesis of perovskite lead titanate (PZT) nanocrystals, made by the addition of sucrose to PbII, ZrIV and TiIV ions in solution27, 28. Use as a chelating agent does not preclude the possibility of subsequent removal of the templating saccharide to produce porous bodies. Lal et al. describe how sucrose can be used to control the crystal growth of alumina foams29. By heating together an acidic mixture of aluminum nitrate and sucrose, a viscous resin is produced which can then be subsequently mixed with an alumina slurry and heated. Calcination of the material achieves the correct inorganic phase in addition to removing the sucrose template. It was found that sintering the composite foam mixture with a sucrose to alumina weight ratio in the range 0.69-1.03, produced alumina foam structures with almost 97% porosity. The structure of the foam bodies was intimately linked to the sucrose to alumina weight ratio such that a transition from a reticulated structure to cellular foam structure occurred at a sucrose to alumina weight ratio below 0.89. Average pore sizes were similarly dependent on the sucrose to alumina weight ratio, typically being formed in the range 0.48–2.69 mm. Glucose has been used in a similar fashion to prepare porous alumina, although via a sol-gel method in order to produce mesoporous structures. In this method, Xu et al. took aluminum iso-propoxide and glucose and dissolved them both in water, then a diluted aqueous nitric acid solution was added dropwise30. After quiescent storage for five hours, the mixture was heated to 100 °C in open air to remove water and any other volatile species. The resulting solid was then calcined at 600 °C for six hours to remove the glucose template. The calcined sample displayed a type IV isotherm, which is characteristic of a mesoporous material. Experiments using gas physisorption showed that the material had a high surface area of 422 m2/g and a relatively monodisperse pore size of 5.1 nm pore diameter. Even after subsequent high heat treatment, these materials exhibited a high surface area. The surface area of a sample after calcination at 800 °C for four hours was still over 220 m2/g. The fact that this material has a high thermal stability can be attributed to the relatively thick pore walls (5.2 nm)
Simple Mono- and Oligosaccharides
23
compared with those reported in previous work (at around 1.7–3.7 nm). The researchers found that the presence of glucose in the precursor material was crucial in the formation of mesoporous alumina, as no uniform pores were able to be obtained in a control sample prepared by the same process without the addition of glucose. As with many of the studies highlighted in this book by using a biomaterial, the inorganic phase, in this case alumina, can be imbued with a high surface area, narrow pore size distribution and thermal stability. This capability for gelation/chelation has been exploited by a number of groups, in the direct control over crystal growth. Qian et al. used glucose as a method of producing cubic Cu2O crystals31. They were able to show that a solution of CuSO4 was reduced in the presence of glucose, which then subsequently acted as a capping molecule in the formation of Cu2O cubes. These nanocubes were then able to be used themselves as templates in the synthesis of copper sulfide nanocages. Liu et al. made use of glucose both as a gelating agent and as a viscous medium from which to spin fibres of a zinc nitrate/glucose composite material32. On drying, calcination of the material produced ZnO fibres with diameters between 3 µm and 5 µm. Yan et al. have used a similar method in the synthesis of nanocrystalline zirconia33. In this work, sucrose was investigated as a chelating agent and template material. Nanocrystalline zirconia is useful as a support for the preparation of nickel catalysts and they showed that the nickel catalyst supported on zirconia prepared by this method showed higher activity and stability.
2.4 Other uses Owing to their simple structures, there are few other uses for mono- and oligosaccharides as templating materials. Some researchers have utilized the reducing nature of the simple sugars directly in the directed growth of mineral phases. For example, selenium nanotubes were synthesized by Gao et al. in a facile one-step hydrothermal method, using Na2SeO3 as the selenium source and glucose as reducing agent34. Synthesis of selenium nanotubes usually produce polycrystalline or amorphous materials, but by using glucose in a hydrothermal synthesis at 180 °C for 10 hours, reaction with Na2SeO3 produces a grey precipitate with a cotton-like fibrous appearance. Differences in the contrast of nanostructures on the TEM images indicate that the materials are primarily
24
Biotemplating
hollow tubular structures with solid nanorods present. The nanotubes are between 100–400 nm in diameter, with a 20–40 nm wall thickness. The length of the nanostructures varied, with some nanotubes tens of microns in length. HRTEM images at showed pronounced lattice fringes, allowing the group to calculate an interplanar spacing of 0.49 nm, close to the (001) lattice spacing of trigonal selenium (t-Se). Trigonal selenium has been shown to be an important elemental semiconductor, and has found use in such diverse applications as photocells, pressure sensors and exposure meters35, 36 owing to its high photoconductivity and large piezoelectric and thermoelectric effects. In this work, it is likely that the glucose is acting as a weak reducing agent for Na2SeO3 under hydrothermal conditions, which allows for a dissolutionreprecipitation leading to a spontaneous growth of nanostructures. With elongated reaction times, highly reactive Se reprecipitates and crystallizes on the surface of the t-Se nanoparticles leading to extended nanotube lengths. For the creation of technologically useful porous bodies, the majority of studies have used various surfactants as templating agents, which lead to pore formation in the inorganic product on their removal. Pore sizes can be controlled by varying the length of the alkyl chain of the surfactant, typically in the 1 nm to 10 nm range, or by the addition of space filling bulky organic molecules. Removal of the templating material however is energy intensive and often leads to structural collapse of the pore wall structures and results in a more disordered material. The use of surfactants and organic filler molecules which have to subsequently be disposed of is therefore not the most environmentally or structurally friendly of options. Wei et al. attempted to circumvent these problems by the synthesis of mesoporous silica using glucose as a pore forming agent37. They demonstrated that mesoporous silica could be successfully synthesized by HCl-catalyzed hydrolysis and polycondensation of tetraethyl orthosilicate (TEOS) in the presence of (among others) glucose and maltose as the pore-forming agents. By using sugars as pore forming agents, the synthesis as a whole is much more environmentally friendly and template extraction is a simple matter of solvent extraction. The materials synthesized had a high surface area of approximately 1,000 m2g-1 with a mesoporous pore sizes of between 2 nm and 6 nm in diameter. Currently, the mechanism for how this nonsurfactant templating leads to the formation of mesoporosity is not understood, although studies on the formation of mesoporous materials via a neutral nonsurfactant pathway have suggested that as the rate of polycondensation in the sol-gel
Simple Mono- and Oligosaccharides
25
reactions is very fast at near neutral pH, the nonsurfactant templates direct mesophase formation by co-operative assembly, particularly during polycondensation and gelation stages. As well as offering a low cost and environmentally friendly synthesis, the use of templates such as glucose mean that the material will have good biocompatibility. The group suggest that by synthesizing mesoporous materials using a glucose template under nearly neutral conditions could lead to a material suitable for the entrapment of enzymes and other biologically active substances in the mesoporous framework. As the pore sizes are in the mesoscopic range, leakage of enzyme molecules is prevented whilst enabling the bulk transport of reactants and products to the enzyme reactive sites. Yim et al. have used a similar process with an acyl mono/di-saccharide acting as a porogen in the formation of nanoporous siloxane films38. Porous thin films have found use as low dielectric materials in integrated circuits, where they are used to reduce resistance and capacitance delay. In their work they prepared a porous silsesquioxane (SSQ) based polymer by acid catalyst controlled hydrolytic polycondensation of 2,4,6,8tetramethyl-2,4,6,8-tetra (trimethoxysilylethyl) cyclotetrasiloxane in the presence of either glucose pentaacetate, sucrose octaacetate or sucrose benzoate. The sugar molecules were functionalized in this study to provide better interaction with the siloxane polymer and also to finely tailor the eventual pore size on decomposition. They found that the saccharide compounds assembled into discrete domains in the silsesquioxane polymer via the intermolecular interaction between acyl groups of the porogens. The resulting pore diameter of the porous siloxane film prepared with 30 wt.% of benzoyl sucrose in spin-coating solution was 2.2 nm. Once decomposition of the saccharide phase had occurred, the material was a flexible, porous thin film with a dielectric constant of around 2.0. All of the above applications for the simple sugars as templating materials rely on them as ‘passive’ templates; either by the fact that they are easily removed ‘occupiers of space’ or as convenient sources of carbon. In the next (and subsequent) chapters, with increasing molecular complexity, biomolecules can begin to form complex structures of their own which can then impart their complexity to the material being templated. It is in this more ‘active’ role that we see biotemplating really come into its own as a force for morphological complexity.
26
Biotemplating
2.5 References 1. 2. 3.
4. 5.
6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
19. 20.
Han, S.J. & Hyeon, T. Simple silica particle template synthesis of mesoporous carbons. Chem. Commun., 1955-1956 (1999). Jun, S. et al. Synthesis of new, nanoporous carbon with hexagonally ordered mesostructure. J. Am. Chem. Soc. 122, 10712-10713 (2000). Kruk, M., Jaroniec, M., Ryoo, R. & Joo, S.H. Characterization of ordered mesoporous carbons synthesized using MCM-48 silicas as templates. J. Phys. Chem. B 104, 7960-7968 (2000). Joo, S.H. et al. Ordered nanoporous arrays of carbon supporting high dispersions of platinum nanoparticles. Nature 412, 169-172 (2001). Hu, X.G. & Dong, S.J. Metal nanomaterials and carbon nanotubes - synthesis, functionalization and potential applications towards electrochemistry. J. Mater. Chem. 18, 1279-1295 (2008). Joo, S.H., Jun, S. & Ryoo, R. Synthesis of ordered mesoporous carbon molecular sieves CMK-1. Microporous Mesoporous Mat. 44, 153-158 (2001). Wang, A.P., Kang, F.Y., Huang, Z.H. & Guo, Z.C. Preparation of natural zeolite templated carbon and its nanopore formation mechanism. New Carbon Mater. 22, 141-147 (2007). Su, B.L., Vantomme, A., Surahy, L., Pirard, R. & Pirard, J.P. Hierarchical multimodal mesoporous carbon materials with parallel macrochannels. Chem. Mat. 19, 3325-3333 (2007). Wang, A.P., Kang, F.Y., Huang, Z.H. & Guo, Z.C. Preparation of porous carbons from halloysite-sucrose mixtures. Clay Clay Min. 54, 485-490 (2006). Kim, T.W. & Solovyov, L.A. Synthesis and characterization of large-pore ordered mesoporous carbons using gyroidal silica template. J. Mater. Chem. 16, 1445-1455 (2006). Sakintuna, B. & Yurum, Y. Templated porous carbons: A review article. Ind. Eng. Chem. Res. 44, 2893-2902 (2005). Klepel, O., Strauss, H., Garsuch, A. & Bohme, K. Several ways to produce porous carbon monoliths by template assisted routes. Mater. Lett. 61, 2037-2039 (2007). Liu, Y.L., Li, L.X., Chen, X.H. & Song, H.H. The electrochemical hydrogen storage properties of ordered mesoporous carbons. Acta Phys.-Chim. Sin. 23, 1399-1404 (2007). Spong, A.D., Vitins, G. & Owen, J.R. A solution-precursor synthesis of carbon-coated LiFePO4 for Li-ion cells. J. Electrochem. Soc. 152, A2376-A2382 (2005). Yang, Z.X., Xia, Y.D. & Mokaya, R. High surface area silicon carbide whiskers and nanotubes nanocast using mesoporous silica. Chem. Mat. 16, 3877-3884 (2004). Darder, M. & Ruiz-Hitzky, E.R. Caramel-clay nanocomposites. J. Mater. Chem. 15, 3913-3918 (2005). Saberi, A., Alinejad, B., Negahdari, Z., Kazemi, F. & Almasi, A. A novel method to low temperature synthesis of nanocrystalline forsterite. Mater. Res. Bull. 42, 666-673 (2007). Zurlinden, K., Laub, M. & Jennissen, H.P. Chemical functionalization of a hydroxyapatite based bone replacement material for the immobilization of proteins. Materialwiss. Werkstofftech. 36, 820-827 (2005). Ranito, C.M.S., Oliveira, F.A.C. & Borges, J.P. in Bioceramics 17, Vol. 17 341-344 (Trans Tech Publications Ltd, Zurich-Uetikon; 2005). Cochran, D.L. A comparison of endosseous dental implant surfaces. J. Periodont. 70, 1523-1539 (1999).
Simple Mono- and Oligosaccharides 21. 22. 23. 24.
25.
26. 27. 28. 29. 30. 31.
32. 33. 34. 35. 36. 37. 38.
27
Suchanek, W. & Yoshimura, M. Processing and properties of hydroxyapatite-based biomaterials for use as hard tissue replacement implants. J. Mater. Res. 13, 94-117 (1998). Borzacchiello, A. et al. Isothermal and non-isothermal polymerization of a new bone cement. J. Mater. Sci.-Mater. Med. 9, 317-324 (1998). Selvig, K.A. Crystal Structure of Hydroxyapatite in Dental Enamel as seen with Electron Microscope. Journal of Ultrastructure Research 41, 369-375 (1972). Landis, W.J. The Strength of a Calcified Tissue Depends in part on the Molecular Structure and Organization of its Constituent Mineral Crystals in their Organic Matrix. Bone 16, 533-544 (1995). Glimcher, M.J., Hodge, A.J. & Schmitt, F.O. Macromolecular Aggregation States in Relation to Mineralization - The Collagen-Hydroxyapatite System as Stuied in vitro. Proc. Natl. Acad. Sci. U. S. A. 43, 860-867 (1957). Bose, S. & Saha, S.K. Synthesis of hydroxyapatite nanopowders via sucrose-templated sol-gel method. J. Am. Ceram. Soc. 86, 1055-1057 (2003). Ram, S. & Mandal, T.K. Synthesis and characterization of thin ferroelectric PbZr0.52Ti0.48O3 fibrils. J. Am. Ceram. Soc. 88, 3444-3448 (2005). Bose, S. & Banerjee, A. Novel synthesis route to make nanocrystalline lead zirconate titanate powder. J. Am. Ceram. Soc. 87, 487-489 (2004). Prabhakaran, K., Gokhale, N.M., Sharma, S.C. & Lal, R. A novel process for low-density alumina foams. J. Am. Ceram. Soc. 88, 2600-2603 (2005). Xu, B.J. et al. Synthesis of mesoporous alumina with highly thermal stability using glucose template in aqueous system. Microporous Mesoporous Mat. 91, 293-295 (2006). Wang, D.B., Yu, D.B., Mo, M.S., Liu, X.M. & Qian, Y.T. Seed-mediated growth approach to shape-controlled synthesis of Cu2O particles. J. Colloid Interface Sci. 261, 565-568 (2003). Liu, Y., Song, Y.F., Chen, D.R., Jiao, X.L. & Zhang, W.X. Sol-gel synthesis of polycrystalline ZnO and ZnS fibres. J. Dispersion Sci. Technol. 27, 1191-1195 (2006). Rezaei, M., Alavi, S.M., Sahebdelfar, S. & Yan, Z.F. Nanocrystalline zirconia as support for nickel catalyst in methane reforming with CO2. Energy Fuels 20, 923-929 (2006). Chen, M.H. & Gao, L. Selenium nanotube synthesized via a facile template-free hydrothermal method. Chem. Phys. Lett. 417, 132-136 (2006). Ma, Y.R., Qi, L.M., Ma, J.M. & Cheng, H.M. Micelle-mediated synthesis of singlecrystalline selenium nanotubes. Adv. Mater. 16, 1023-1026 (2004). Zhang, H. et al. Selenium nanotubes synthesized by a novel solution phase approach. J. Phys. Chem. B 108, 1179-1182 (2004). Wei, Y. et al. Preparation and physisorption characterization of D-glucose-templated mesoporous silica sol-gel materials. Chem. Mat. 11, 2023-2029 (1999). Yim, J.H., Jeong, H.D. & Pu, L.S. The preparation of nanoporous siloxane films using saccharide derivatives as new porogen. Thin Solid Films 476, 46-50 (2005).
Chapter 3
Complex Polysaccharides
In the last chapter, we considered the simpler sugars as potential candidates for biotemplating. These molecules are limited however, owing to their relative structural simplicity and act in a more passive role in the templating of materials. With polysaccharides, the potential for more complex templating emerges as they are now able to crystallize with long-range order and thereby impart this complexity to the final material. With increasing chain length, there exists the possibility to synthesize an incredibly large number of polysaccharides, both in the laboratory and in nature. A flavour of the range of these will be given, but the majority of the chapter will be devoted to the most commonly researched complex polysaccharides.
3.1 Structure and properties As the number of sugar residues increases and the glycosidic bonds multiply, longer and longer chains form. With the potential for glycosidic bonds to appear which branch the chain structure, the number of structurally distinct polysaccharides is almost limitless. With increasing chain length, the physical properties alter progressively away from the sweet-tasting, soluble simple monosaccharides to molecules which tend to be insoluble, tasteless and in the case of higher molecular weight polysaccharides, form viscous, sticky masses. Polysaccharides have a general formula of Cn(H2O)n-1 where n is usually a large number between 200 and 2500, although some of the larger polysaccharides have molecular weights in the hundreds of thousands of daltons. Despite the large range of molecular weights and structures, the polysaccharides can be distinguished broadly by the type of glycosidic bond present in the molecule. In making a glycosidic bond, there are two possibilities for linking the monosaccharide units together. The first is the alpha-linkage where (for a Dsugar configuration) the bond is made at the anomeric carbon of one 28
Complex Polysaccharides
29
monosaccharide to another monosaccharide unit so that the hydrogen atoms on the carbon either side of the glycosidic bond are on the same side of the ring plane (Figure 3.1).
Figure 3.1 – An alpha-linkage between two glucose residues forming maltose.
The other possibility is that the hydrogen atoms on the carbon either side of the glycosidic bond are on opposite sides of the ring plane. This is the beta-linkage (Figure 3.2).
Figure 3.2 – A beta-linkage showing beta-glucose forming cellobiose.
Despite looking almost identical, alpha and beta linkages have a drastic effect on the properties of the polysaccharides. In particular, it is the way in which they are processed by animals which mark the two linkages down as being of
30
Biotemplating
particular import. Humans and some animals are only able to process alphalinked polysaccharides, which are hydrolyzed by enzymes called amylases and lack the enzymes which can digest beta-linked polysaccharides. In the above examples, the glycosidic bond is between the ‘1’ carbon on one monosaccharide and the ‘4’ carbon of the other (a 1→4 bond). It can be easily seen that this configuration naturally leads to a linear chain polymer. There exist other possibilities for glycosidic bonds however and it is this that further extends the complexity of polysaccharides. A 1→6 glycosidic link is the most common after the 1→4 and leads to branching of the polymer structure. With a high proportion of 1→6 linkages, the resulting polysaccharide will be highly branched and therefore be thick and viscous. Organisms such as bacteria make use of these types of polysaccharide, often secreting then in order to shield their cells from triggering an immune response in the host organism and to increase adhesion. Polysaccharides are usually considered as groups in terms of their function. Structural polysaccharides contain high proportions of beta-linkages and often provide support for cell walls in plants and bacteria. They feature among their number cellulose, dextran, xanthan gum and pullulan. Storage polysaccharides are those which are exclusively composed of alpha-linkages and comprise a large family of starches.
3.2 Cellulose Cellulose is a polysaccharide found in plants, where it is used as the primary structural component of the cell walls. It is formed exclusively from beta-glucose units which are 1→4 linked, therefore giving rise to a straight-chained polymer. With extended straight chains, the possibility exists for extensive interchain hydrogen bonding, which serves to induce the formation of rod-like microfibrils. With strong interchain interactions, cellulose has considerable structural stability. With the inherent complexity exhibited by cellulose, its use as a template is extensive. Many studies have made use of the fibrous nature of cellulose in order to impart a fibre-like morphology to the inorganic phase. Lau et al. have used the fibrous nature of cellulose to create calcium carbonate (CaCO3) aggregates which replicate the underlying cellulose template1. By sonicating cotton fibres whilst adding sequentially calcium and carbonate ions, they were able to mineralize the fibres with CaCO3. Subsequent calcination of the composite material produces hollow fibres of CaCO3. This method was able to be applied equally well to create CaCO3 replicas of woven cellulose in the form of cotton cloth (Figure 3.3).
Complex Polysaccharides
31
Figure 3.3 – SEM morphologies of the original cloth (a,b) and the “cloth CaCO3” obtained with 0.005 M CO32- as the reaction/adsorption initiator (c,d). SEM images clearly show that each fibre is constructed from discrete thin strips (e,f). TEM images (g,h) are obtained from a single strip of the “cloth CaCO3”. The inset of g shows the SAED pattern of a single strip. Reproduced with permission of the American Chemical Society. Reprinted with permission from2. Copyright 2007 American Chemical Society.
In a similar manner, Kunitake et al. used cellulose to form fibres of electrically conductive indium tin oxide (ITO)2. In this work, indium
32
Biotemplating
methoxyethoxide (In(OCH2CH2OMe)3 and tetraisopropoxytin-isopropanol adduct (Sn(OiPr)4 • iPrOH were dissolved in methoxyethanol and in isopropanolmethanol (1:1, v/v) respectively. Combining the two solutions gave the precursor material which was then heated to 50 °C. The templating of cellulose was simply done by passing this hot solution over a piece of filter paper in a vacuum filtration assembly. After calcination at 450 °C, the resultant ITO replica was formed as a layer of between 90 µm and 220 µm thick. The electrical conductivity of the templated fibres was confirmed, with the best conductivity arising from an ITO sheet with In/Sn ratio of 93.5/6.5 at 0.53 S cm-1. This value is higher than that observed in similarly nanostructured ITO films prepared by other templated routes. They postulate that the nanostructured ITO synthesized using a cellulose template may find applications in the next generation of nano/micro-sized batteries, capacitors and in electrolytic wastewater treatment. This group have also looked at the coating of cellulose fibres in order to impart anisotropic morphology to an organic material. In this work, oxidative polymerization of pyrrole monomer using copper chloride in 2-propanol was carried out in the presence of cellulose, very simply by passing the solution through a piece of commercially available filter paper3. This method is of importance as recently, there has been considerable interest in the production of conductive textiles4-8. The successful coating of fibres with conductive polymers can be hard to achieve however, as most polymers will tend to exhibit colloidal behaviour when interacting with substrates, leading to agglomeration and sedimentation. By finely controlling the deposition time, Kunitake et al. found no such agglomeration when polymerizing pyrrole on cellulose fibres. Further analysis revealed to the group that the composite sheet is composed of continuous interconnected nanofibre networks which retain the original morphology of the cellulose fibres. On closer inspection, they found that the electron beam-induced rupture of the PPy layer revealed a cable-like core-shell structure. TEM imaging showed clearly a flat and homogeneous polypyrrole layer around each individual cellulose fibre. By changing the deposition time, they found that they were able to control the thickness of the polypyrrole layer quite precisely between 20 nm and 25 nm. Although no measure of the conductivities of these composites were made, this technique remains an extremely efficient method of producing conjugated polymer materials with highly anisotropic morphologies.
Complex Polysaccharides
33
Deng et al. have used cellulose to form core-shell and hollow nanoparticles of titania in a layer by layer method using partially digested cellulose as a template9. The aim of the work was to directly replicate the underlying structure of the cellulose and produce highly anisotropic nanoparticles of TiO2. In order to achieve this, the cellulose was subjected to a sulphuric acid digestion to break down the macroscopic fibres and leave nanofibres. In the layer by layer method, poly(diallyldimethylammonium chloride) (PDADMAC) was added to a dilute solution which contained digested nanofibres, followed by titanium (IV) bis(ammonium lactato)dihydroxide (TALH). After calcination, hollow titania nanoparticles were formed (Figure 3.4)
Figure 3.4 – TEM image showing hollow titania nanoparticles after calcination derived from LBL method with 10 TALH/PDADMAC bilayers. Scale bar is 50 nm. Reproduced by permission of IoP Publishing Ltd.
It is obvious from Figure 3.4 however, that the original intention of faithful replication of the template was not achieved. They found that for high aspect ratio sub-micron templates, the eventual nanoparticle shape was further away from the shape of the template with each layer of deposited inorganic material.
34
Biotemplating
This had the effect of decreasing the aspect ratio of the coated particles dramatically as the number of bilayers increased (Figure 3.5).
6
Aspect ratio of coated particle
5
4
3
2
Experimental: LBL Coating Experimental: Sol-gel Coating 1
Calculated: Even Coating
0 0
5
10
15
20
25
30
35
40
45
Aspect ratio of cellulose template
Figure 3.5 – Dependence of aspect ratio of coated particles derived from LBL and sol-gel on aspect ratio of cellulose whisker template for a fixed coating thickness. For the LBL method, 10 bilayers were deposited, and for sol-gel method, the coating conditions were given in the experimental section. Reproduced by permission of IoP Publishing Ltd.
From these data they postulated that the polyelectrolytes displayed preferential adsorption on the tips of the whiskers. This lead to the coating layer becoming disproportionately thicker on the sides than on the tips, with the consequent loss of anisotropy of the particles. They explain this observation by consideration of total surface free energy effects. By reducing the particle size, the total surface free energy of particles is reduced. When undergoing growth, all particles will try to minimize their surface free energy by reducing their aspect ratio. This means that for anything other than the thinnest of coatings, preservation of the template shape will not be possible unless the template is spherical. This is a key finding and one which has repercussions for all potential biotemplates. It would appear that when looking to retain anisotropy in the final
Complex Polysaccharides
35
material, replication of a template with complex form must be at length scales larger than the nanoscopic, particularly if thicker, robust coatings are desired. There have been many other studies on the coating of cellulose fibres with inorganic phases, all based around the simple deposition and growth onto the cellulose, producing either fibres or nanoparticles of inorganic phases such as AgO10, Fe2O311, CdS12 and hydroxyapatite13 with greater or lesser degrees of success as Deng et al. have indicated. One possible solution to improving homogeneity of coating is to functionalize the cellulose in order to increase the specificity of interaction between the inorganic phase and the substrate. Cellulose is amenable to functionalization as the hydroxyl groups on the molecule can be replaced, giving a range of cellulosic materials with different properties. Esterification of cellulose is a simple option, which can create film-forming and fibre-forming derivatives such as nitrocellulose and cellulose acetate. Other cellulose derivatives such as carboxymethyl cellulose (CMC) and hydroxymethyl cellulose (HMC) find use in medical applications such as pharmaceutical excipients and artificial tissue replacement technologies, as they are able to encourage remineralization and revascularization and act as scaffolds for complex, high surface area drug delivery vectors. Zhu et al. have made use of CMC in the synthesis of silver sulphide (Ag2S) nanocubes14. Ag2S is an important semiconductor material, which has found a number of applications as a photoelectric and thermoelectric material15 and has been synthesized with morphological control previously in the form of spheres and chains16, 17. As the most energetically favourable morphology, the isotropic spherical shape is the easiest to achieve synthetically. More challenging therefore is to be able to synthesize nanoparticles in a cubic morphology. Cubic nanoparticles will be able to pack together more easily and with greater ordering than they would in more isotropic morphologies. This may prove useful as a method for generating photonic arrays and denser catalytic supports of nanoparticles. Zhu et al. found that they were able to synthesize hollow and concave Ag2S nanocubes through a simple one-step process based on the reaction between AgNO3 and thioacetamide (TAA) in aqueous solution in the presence of carboxymethyl cellulose (CMC). In a typical synthesis, 0.05g of CMC was dissolved in 10 ml of water, to which was added 5 ml of 4 x 10-3 M aqueous AgNO3. After 3 hours of stirring, 5 ml of 3 x 10-3 M aqueous TAA was added to the solution and stirred in dark for 3 hours. The reaction mixture was then heated at 80 °C for 8 hours, followed by cooling to room temperature. TEM images of the product revealed that there were many cubic structures, with dimensions of
36
Biotemplating
around 200 nm each side. They found that temperature played a key role in the eventual morphology of the nanoparticles. When the solution was stirred at room temperature for 11 hours after the introduction of TAA, solid, concave cubic structures were obtained. If the ageing was conducted at 80 °C, hollow cubic structures resulted. Ultraviolet-visible (UV-Vis) spectra of these materials revealed relatively broad peaks with the maximum absorption of hollow cubes at 261 nm and concave structures at 286 nm. Shifts in the UV have been used previously as an indicator of hollow morphology at the nano-scale18. The mechanism of formation is still unclear, although it is likely that the CMC is acting as a chelator of Ag ions in solution, thereby providing discrete, preferred sites of nucleation and growth of the Ag2S phase. Functionalization of cellulose was also employed by Greil et al. in the synthesis of silica nanotubes19. Their work was based on the observation that some synthetic and biological polymers that contain secondary and tertiary amino groups within the chain, or primary amino functions on the side groups, are capable of catalyzing the formation of silica spheres with diameters ranging from 50 to 400 nm. The proposed mechanism involves the influence of Nmethylpropylamino groups, which provide steric fixation and catalyze the condensation of silica precursors. Taking their cue from this observation, they introduced the oligopropylamino side chain responsible for the control of biosilica deposition in diatoms, into the cellulose biopolymer They observe that the cellulose forms a rigid backbone because of intramolecular hydrogen bridges between the secondary hydroxyl group at the C3 position of an anhydroglucose unit (AGU) and the oxygen of the pyrane ring in the next AGU. This means that the free rotation around the glycosidic bond is hindered. In addition, intermolecular network formation of the cellulose chains is suppressed because of the derivatization with oligopropylamine. Because these oligopropylcellulose functionalized celluloses are soluble in water, the silica formation is able to be performed under mild conditions with water-soluble silica precursors, such as tetrakis(2-hydroxyethyl) orthosilicate (THEOS), or by using more the more common tetraethyl orthosilicate (TEOS). Two nitrogen atoms separated by a propylene group were considered to be the catalytically active site for silica formation on the cellulose chain, which act as the analogue of the controlled silica deposition found in natural systems. In the functionalization experiment, Dipropylenetriamine (DPTA) was reacted with the cellulose tosylate because of the nucleophilic properties of primary amino functions. Silica nanotubes were then able to be successfully synthesized from DPTA-cellulose solutions.
Complex Polysaccharides
37
Calcination at 1000 °C resulted in the removal of the cellulosic template leaving hollow silica nanotubes. The silica nanotubes formed with diameters from 10–30 nm, inner core diameters of approximately 3 nm, and lengths of up to 500 nm. The group postulate that a single or two DPTA cellulose molecules are acting as molecular templates for the formation of silica nanotubes. In addition, the oligopropylamine derivatives act as a catalyst in the polycondensation of the silica precursors during sol-gel synthesis. A change from isotropic morphology was also able to be induced by the use of hydroxyethylcellulose (HEC) in the controlled synthesis of CdS nanorods. In the work by Liu et al. gamma radiation was used on an inverse microemulsion synthesis of CdS in the presence of HEC20. By irradiating an emulsion of Na2S2O3 • 5H2O and CdSO4 • 8H2O in isopropyl alcohol, HEC and emulsifying agents, radiolysis of the water produced active species such as e-aq. This species was then able to reduce sulfur to S2-, which reacted with Cd2+ to generate CdS. Again, the chelation of metal ions by the modified cellulose provides the morphological control required. It has been demonstrated that morphological complexity can be generated in the cellulosic material itself, even before the addition of an inorganic phase. Park et al. have shown that by electrospinning cellulose triacetate (CTA) with methylene chloride (MC) and ethanol, very fine nanofibres of cellulose acetate could be produced with a high degree of porosity21. These porous CTA fibres electrospun with MC had discrete circular pores with a narrow size distribution in the range of 50–100 nm. By increasing the ratio of MC to ethanol to 90/10 v/v, the pores became interconnected and larger in size, between 200–500 nm in diameter. The porosity arises from the rapid evaporation of solvent during the electrospinning process. This was evident from the fact that non-porous corrugated fibres were obtained from MC/EtOH (85/15 v/v) and MC/EtOH (80/20 v/v) due to their lower vapor pressure. They found that the pore sizes in ultrafine CTA fibres electrospun with MC showed a bimodal distribution centered at 17 and 64 nm. CTA fibres electrospun with MC/EtOH (90/10 v/v) showed the greatest porosity due to their larger intra-fibre pores and fibre diameter. Porosity has also been induced in thin films of a modified cellulose by BarNir et al. In this work, the group used water condensation at the surface of a cellulosic solution in order to form honeycomb structures22. Pore size and distribution was able to be controlled in several ways, one of which is the hydrophilicity of the cellulosic phase used. The amphiphillic nature of the
38
Biotemplating
celluloses was modified with varying lengths of ethylene glycol side chains using 2,6-thexyldimethylsilyl cellulose. They found that the side chains did not directly affect the honeycomb pore formation, with longer ethylene glycol chains leading to increased pore uniformity but having no effect on the size of the pores. Such materials could be expected to find applications as photonic crystals, microreactors or microarray chips, or as catalytic supports. As the cellulosic material is biocompatible, it is probably as a scaffold for tissue re-growth or other biomedical applications that these materials will find the most use.
3.3 Dextran Dextran is a complex, branched polysaccharide composed of a large number of glucose units in a wide range of lengths so that the molecular weight can vary between 1,000 and several hundred thousand daltons. The backbone of the dextran molecule consists of α1→6 glycosidic linkages, with branches forming primarily from α1→3 links. Depending on the source, the dextran can also be branched from α1→2 and α1→4 linkages as well. In nature, dextran is a product of bacterial action on sucrose. The most common source of dextrans are from Leuconostoc mesenteroides and Streptococcus mutans bacteria. Dextran is used extensively in medicine as it has a high affinity for platelets and red blood cells, binding strongly to them, and thereby reducing their tendency to aggregate. In this manner, dextran acts an extremely effective anti-thrombosis agent23. Higher molecular weight dextrans are metabolized slowly within the body and therefore remain in vivo for longer, extending the therapeutic effect. Solutions of dextran also find application as osmotically neutral fluids which can be introduced intravenously into the body to provide a source of water and glucose. The structural complexity of dextran, in particular the tangled, multi-branched nature of the molecule has made it an interesting molecule to use as a biotemplate. One of the first to employ dextran in this role was a study by Mann et al. on the use of dextran as a template to form metallic and metal oxide sponges24. They based their work around the observation that as dextran has a high solubility in water at room temperature, concentrated aqueous solutions of metal salts could be incorporated with the dextran, contrasting with the sometimes low loadings of inorganic template required when using less soluble templates such as cellulose. In addition, the metal-ion-loaded dextran gels could
Complex Polysaccharides
39
be shaped to appropriate moulds to produce sponge-like materials of silver, copper oxide, titania and gold with a range of macroscopic morphologies (Figure 3.6). Finally, as dextran is a reducing sugar, the reductive aldehyde groups could be exploited for in situ reduction of metal ions and nucleation of monodisperse metallic clusters. Experimentally, the creation of dextran/metal composite materials was extremely simple. In a typical experiment on the creation of silver sponges, silver nitrate was dissolved in distilled water, to which dextran was added in a 2:1 ratio by weight. Mixing the constituents together formed a thick paste, which was able to be transferred to a ceramic crucible mould and left to harden for one day at room temperature. The resulting paste was then heated to a temperature of 520 °C at 10 °C per minute for 1 hour in a furnace, leaving a porous silver body. TEM and SEM studies on the calcined monolith showed that the porous body consisted of an interconnected framework of silver filaments, approximately 4 µm in width, composed of rows and rings of fused micrometre-sized particles that enclosed pores 1 to 20 µm in size. X-ray diffraction (XRD) data showed reflections at 2.37, 2.04, 1.44,1.23 and 1.18 Å corresponding to the (111), (200), (220), (311) and (222) d spacings, respectively, of metallic silver with a face-centred-cubic unit cell structure (JCPDS—Joint Committee on Powder Diffraction Standards—card #4-783). The group found that thermal gravimetric analysis (TGA) showed a marked weight loss for the Ag(I)-loaded dextran composites at 167 °C associated with melting and decomposition of the dextran matrix. In the absence of AgNO3, pure dextran softened at 160 °C to become a viscous liquid at around 170 °C that darkened to a fluid at 180 °C, followed by decomposition at 200 °C. It was noted that the presence of AgNO3 lowered the dextran decomposition temperature due to the release of oxygen during thermal transformation of the silver nitrite, followed by reduction to silver metal, which occurred at 160 °C. Increasing the temperature above 170 °C resulted in minimal additional weight loss (final weight loss, 65%), indicating that near-complete conversion of the dextran matrix to carbon dioxide and water was coupled with the decomposition of the inorganic salt. The formation of the porous body was rationalized as rapid expansion in the volume of the matrix due to outgassing and trapping of oxygen, nitrogen oxide, steam and carbon dioxide bubbles within the viscous polysaccharide matrix. This suggests that growth and sintering of the silver particles occurred preferentially along the wall structure of the expanded dextran matrix to produce metallic replicas with macroporous architecture.
40
Biotemplating
Figure 3.6 – Metallic and metal oxide sponges of different compositions. (a) SEM image of silver sponge containing evenly distributed copper oxide particles prepared at 800 °C. Scale bar = 5 µm. (b) Corresponding energy-dispersive X-ray spatial map showing localized copper-rich domains. (c) Silver/copper oxide sponge after hydrazine reduction showing localized regions of copper metal and copper oxide in the silver framework. Scale bar = 5 µm. (d) Gold sponge prepared at 800 °C. Scale bar = 5 µm. (e) Copper oxide sponge prepared at 800 °C. Scale bar = 20 µm. (f) Maghemite (γ-Fe2O3) sponge prepared at 600 °C. Scale bar = 50 µm. Reprinted by permission from Macmillan Publishers Ltd.24, copyright 2003.
Complex Polysaccharides
41
In support of this model, the group found that changes in dextran viscosity had a pronounced effect on the stability of the expanded matrix associated with the decomposition reactions. This was demonstrated by preparing silver sponges at 520 °C using dextrans with molecular weights of 17,000 and 2,000,000 instead of the 70,000 used previously. SEM observations showed that the reduced viscosity of the low-Mr dextran produced silver sponges that were fragile due to incomplete connectivity within the metal framework, whereas highly compact silver monoliths with occasional channels and voids due to gas venting were produced when the highly viscous, high molecular weight dextran was used. The use of dextran as a porogenic template in this simple way allowed for the simple addition of additives to the synthesis in order to produce functional composite materials. Addition of a catalytic promoter to the silver sponge material was accomplished simply by mixing copper nitrate with the silver nitrate solution before mixing with dextran and heating to 800 °C for 30 minutes. This resulted in a silver sponge with pores of 2–50 µm and an interconnecting metallic framework that contained discrete copper oxide crystallites, 1–2 µm in size. In addition, the group found that the copper oxide particles could be dissolved with aqueous HCl to form voids in the silver framework (Figure 3.7), or reduced by treatment with hydrazine solution for several hours, to form orange sponges containing localized disks of copper metal along with copper(I) and (II) oxides. Changes in the composition and potential functionality of the wall structure were also able to be achieved by addition of preformed nanoparticles to the silver nitrate solution followed by mixing with dextran and heat treatment at 600 °C. In other metal systems, the group investigated the addition of small quantities of a colloidal titania photocatalyst (anatase, particle size = 100 nm, 0.15 wt%). The presence of colloidal titania resulted in a sponge-like silver/TiO2 framework of reduced width (1–2 µm). It was postulated that the incorporation of the titania nanoparticles disrupted the bonding between the dextran polymer chains and reduced the viscosity of the matrix to give highly expanded metallic sponges that were structurally unstable for titania contents exceeding 0.45 wt%. In the case of gold, sponges with macropores, 2 to 100 µm in size, were prepared by dissolving dextran in aqueous solutions of gold (III) chloride and heating the air-dried pastes. Thermal decomposition of the gold–dextran matrix occurred at a much higher temperature than with silver nitrate due to the absence of released oxygen from salt decomposition. The presence of significant amounts of carbon in the replicas calcined at 600 °C necessitated calcination to 800 °C in order that a purely metal replica could be achieved. The gold monoliths consisted of
42
Biotemplating
continuous open frameworks of metallic gold strands that were composed of fused crystallites, 1–2 µm in width. Magnetic iron oxide sponges were fabricated by addition of a colloidal suspension of magnetite (Fe3O4) nanoparticles to aqueous solutions of dextran followed by air drying and heating. The intact brown monoliths responded strongly to an applied magnetic field and consisted of a loose framework of curved strands of fused particles, about 5 µm in size, with pores between 2 and 80 µm across.
Figure 3.7 – SEM micrograph of silver sponge containing voids due to acid dissolution of localised copper oxide particles. Reprinted by permission from Macmillan Publishers Ltd.24, copyright 2003.
This important and extensive work showed that dextran was particularly useful as a biotemplate, making use of the expansion of the dextran matrix during thermal degradation, giving rise to an open framework structure and associated spatial patterning of interconnecting metal/metal oxide rods or filaments. This
Complex Polysaccharides
43
was extended by Khan et al. who investigated the effect of the addition of a simple surfactant to this system25. In their work, the composite materials were prepared by mixing a solution of AgNO3 and Triton X-114 with a high-Mr dextran (Mr = 2,000,000). The material was dried for 2 days at room temperature then calcined at 550 °C for 8 hours. The porous silver sponges resembled those synthesized by Mann et al. with Mr = 70,000 dextran; the presence of the surfactant however ensured that the final material retained its porosity with none of the densification previously observed. By adopting a lower calcination temperature, Tang et al. showed that dextran/silver nitrate composite system could be used to produce silver nanowires26. By holding the composite paste at 180 °C for 12 hours, reduction by the aldehyde groups on the dextran produced discrete silver nanoparticles, the sintering of which was prevented by the slowly oxidizing biopolymer. They propose that selective binding of dextran to certain crystallographic faces of the silver nanoparticles, coupled with Ostwald ripening, produces outgrowth of some silver nanowires. SEM images show that the nanowires are highly uniform, with lengths of between 20 µm and 80 µm with widths of around 270 nm. Electron diffraction confirmed that these nanowires were cubic single crystals of silver. They note that when a higher concentration of dextran was used, more nanoparticles and fewer nanowires were produced. They postulate that for higher concentrations of dextran, there will be a greater number of Ag nuclei generated from the redox reaction, whilst the greater viscosity prevents nanowire outgrowth. Dextran has also been used as a template to produce materials with specific bio-functionalities. Mann et al. have used dextran to produce hydroxyapatite (Ca10(PO4)6(OH)2, HAp) sponges to provide a scaffold on which to stimulate the growth and proliferation of human bone marrow stromal cells27. Incorporation of pre-formed nanoparticles of amino acid-coated hydroxyapatite into a dextran paste, afforded porous HAp sponges on calcination (Figure 3.8). The sponges comprised an interconnected network of relatively large pores of between 100 µm and 200 µm in size. The group undertook biocompatibility studies by examining the viability, adhesion, spreading and proliferation of human bone marrow (HBM) stromal cells over a period of 14 days (Figure 3.9). It was found that few necrotic cells were observed, which meant that the sponges could be considered as successful substrates for the support of cell growth.
44
Biotemplating
Figure 3.8 – SEM images of sponge-like monoliths prepared from dextran sulphate/HAp pastes in the presence of (a) aspartic acid and (b) alanine. Reprinted from27. Copyright 2005, with permission from Elsevier.
In addition, the cells showed indications of the formation of new bone matrix, so whilst not suitable for load bearing applications, these composite materials could certainly be used for cartilage or soft tissue engineering. Another functional material which has been synthesized using dextran is the high-temperature (high-Tc) superconductor YBa2Cu3O7-δ (Y123). Historically, metal oxide superconductors, in particular the cuprates, have been made by repeatedly grinding together mixed-metal and/or mixed-metal oxide powders, followed by sintering. More recently, advances in the understanding of materials chemistry have led to the use of sol-gel methods to produce finer and more
Complex Polysaccharides
45
homogeneous superconducting particles28-33. By producing homogeneous, nanosized superconducting particles, it is hoped that important fundamental properties such as the critical current density (Jc) can be greatly improved. Despite this advance, researchers are still left with the problem of grinding and pelletizing the resulting oxide, in an attempt to control the bulk morphology, or by producing anisotropic ‘tapes’ of superconducting material. These weakly selforganizing tapes are an attempt to extend the length over which these materials are able to operate. This self-organization is important, as it is believed that mis-aligned grain boundaries act as weak links in limiting the Jc of bulk high-Tc superconductors. The elimination or minimization of large-angle grain boundaries by morphological control is therefore a very desirable target.
Figure 3.9 – Photomicrographs of HBM stromal cells grown for 14 days on HAp/A and HAp/D sponge-like scaffolds showing, respectively, (a,c) expression of alkaline phosphatase (red staining) and (b,d) collagen matrix production (sirius red/alcian blue staining). Reprinted from27. Copyright 2005, with permission from Elsevier.
46
Biotemplating
The field of biotemplating can solve the problems of morphological and size polydispersity in the high-Tc superconductors by allowing the creation not only of nanoparticles which are morphologically homogeneous and more monodisperse, but also which either concurrently or subsequently can be ordered in a very precise manner. Work by Hall et al. demonstrates this with the creation of superconducting sponges through the admixing of dextran with a solution of metal nitrate salts of the correct stoichiometry to form Y12334. In a typical synthesis, sponges were prepared by the addition of dextran (20g, Mr = 70,000) to a superconductor precursor solution (10 ml). The precursor solution was prepared by the dissolution of Y(NO3)3 • 6H2O (1.915g, 0.05M), Ba(NO3)2 (2.613g, 0.1M) and Cu(NO3)2 • 2.5 H2O (3.489g, 0.15M) in 100 ml of H2O. Dextran and precursor solution were mixed together in a crucible to form a light blue, viscous paste and left for one day at room temperature to harden. The resulting paste was then heated to a temperature of 920 °C at a ramp rate of 10 °C/min and held at that temperature for two hours. After this time, the material was allowed to cool back to room temperature in the furnace at a rate of approximately 2 °C/min. In order to improve structural stability, silver doped Y123 sponges were prepared in the same manner as above, with the addition of AgNO3 (0.8 g, 4.7 mM) to the precursor solution. SQUID magnetometry shows that the resulting sponge-like material is superconducting, with an onset critical temperature of 90 K (Figure 3.10). In this sample, the field cooled (FC) data and the zero field cooled (ZFC) data are extremely similar, i.e. the magnetisation is highly reversible, implying weak bulk pinning. Nonetheless, a high critical current density of 1.8 kA cm−2 at 77 K and 1 T field is obtained, rising to 1.1 MA cm−2 at 10 K, 1 T, which was attributed to surface pinning of the porous structure, as well as its small crystallite size. This contrasts markedly with a control sample synthesized via a solid-solid reaction. A commercially obtained Y123 powder was found to have a Tc of 92 K and a critical current density of just 0.04 kA cm−2 at 77 K, 1 T and 0.02 MA cm−2 at 10 K, 1 T. X-ray diffraction (XRD) of the dextran-templated material showed strong peaks corresponding to Y123, with only minor indications of other constituents (Figure 3.11). In contrast to control Y123, which shows the typical polydispersity in terms of crystal size and morphology, a Y123 dextran-templated sponge had a uniform, macroporous, open architecture. Transmission electron microscopy imaging confirmed that the fine structure of the Y123 sponge consisted of a gel-like composition with a typical strut size of around 60 nm (Figure 3.12).
Complex Polysaccharides
47
Figure 3.10 – SQUID magnetometry showing critical current densities (Jc) and critical temperatures (Tc) of various superconducting sponges: in parts a and b, a dextran-templated Y123 sponge; in parts c and d, a silver-doped dextran-templated Y123 sponge; and in parts e and f, a sodium-doped dextran-templated Y123 sponge. Reprinted with permission from34. Copyright 2007 American Chemical Society.
48
Biotemplating
Figure 3.11 – XRD patterns of (a) a dextran-templated Y123 sponge, (b) the control Y123 sample (without dextran), (c) a silver-doped dextran-templated Y123 sponge, and (d) a sodium-doped dextran-templated Y123 sponge. Reprinted with permission from34. Copyright 2007 American Chemical Society.
The inclusion of silver in the Y123 sponges was also investigated. Silver is ubiquitous in YBCO research as a means to improve mechanical strength and grain connectivity. By forming a conductive layer between Y123 crystallites, the presence of silver usually leads to an increase in critical current density. SEM showed that the introduction of silver to the Y123-dextran synthesis resulted in a more pronounced crystalline structure, the sponge with well defined 500 nm sized crystallites now able to be manipulated without structural collapse. XRD confirmed that the Y123 phase was retained in this material. SQUID magnetometry revealed that the Tc for this material was 91 K, with a Jc of 0.6 kA cm−2 at 77 K, 1 T and 0.2 MA cm−2 at 10 K, 1 T. The increased splitting of the FC and ZFC curves with respect to the pure dextran-templated sample implies an improved bulk pinning as a result of the incorporation of silver.
Complex Polysaccharides
49
Figure 3.12 – SEM images of (a) the control Y123 sample and (b) a dextran-templated Y123 sponge and (c) TEM image showing detail of the gel-like nature of the dextran-templated Y123 sponge and (d) a silver-doped dextran templated Y123 sponge. Scale bar in part a is 1 µm, in parts b and d is 10 µm, and in part c is 100 nm. Reprinted with permission from34. Copyright 2007 American Chemical Society.
The efficacy of this synthetic protocol is due to the fact that the dextran acts in situ as an anti-sintering, sparging template during calcination and thereby produces nano-scale crystallites of Y123 whilst retaining a homogeneous dispersion of the flux-pinning agent. By simply controlling the crystal morphology of Y123 during synthesis by biotemplating, Hall et al. have created a highly porous, fine grained superconductor in which the critical current density is improved over that available commercially by over an order of magnitude. Another group of functional materials to have benefitted from the use of dextran as a templating agent are zeolites. Zeolites are naturally occurring minerals that have a micro-porous structure and have been used historically as a means to purify water35. They can do this as they possess a network of fine,
50
Biotemplating
micropores and channels, which can effectively screen out most contaminants, particularly of a biological origin when absorbing water. The purified water can then be released by heating the zeolite. There are over 150 types of man-made zeolites along with the 48 naturally occurring ones36, 37. Consisting of hydrated alumino-silicate, they have the additional ability to sequester small cationic species, thereby allowing for ion-exchange and catalytic applications. The organization of zeolites into hierarchical materials is therefore of considerable interest to scientists, as this would allow the zeolites to be integrated into catalysis and separation applications. Mann et al. used dextran to organize zeolitic nanoparticles to form centimeter-sized macroporous framework monoliths of interconnected filaments38 (Figure 3.13).
Figure 3.13 – SEM micrograph showing open macroporous zeolite/silica framework prepared by dextran templating; scale bar = 100 µm. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
Complex Polysaccharides
51
The synthesis of the monoliths involved a simple dissolution of dextran (Mr = 70 000) in an aqueous suspension which consisted of 20:1 wt ratio of submicron sized NaY zeolite crystallites with sizes ranging from 0.7–1 mm and silica nanoparticles (diameter, 14 nm) producing a viscous paste. Left to dry in air for 48 hours, subsequent calcination of the paste resulted in thermal decomposition of the dextran matrix with associated outgassing of steam and carbon dioxide. This had the effect of producing a carbonaceous foam structure on which the thermally induced aggregation of the silica/zeolite particles could take place. On further heating to 600 °C, the carbonized dextran template was destroyed, leaving the inorganic replica. SEM investigations showed that the replicas consisted of a continuous open framework of interconnected filaments, about 15 mm in width and 100–150 mm in length. The filaments intersected at distinct three-point junctions to produce open cells that were usually five- or six-sided and 100–300 mm across. Highresolution images showed filaments composed of discrete micrometer-sized zeolite crystals embedded within a continuous matrix of condensed silica nanoparticles. Significantly, X-ray diffraction studies showed almost identical reflections for the NaY crystals before and after heating the dextran composites to 600 °C and BET analysis revealed no corresponding change in the type I isotherm, micropore size distribution, or surface area (650 m2g-1), thus indicating that no degradation in the zeolite structure occurred during formation of the sponge like material. The group also investigated the formation of macroscopic fibres of aligned silicalite crystals. These crystals were pre-organized in the form of discrete self-assembled chains owing to the hydrothermal synthesis conditions in the presence of the water-soluble cationic polymer, poly(diallydimethylammonium) hydroxide. Surfactant organization is a wellknown phenomenon and has been used previously in many inorganic systems to promote controlled aggregation39. In this work, an aqueous suspension of the chains was mixed with dextran and long fibres up to 20 cm in length were pulled from the mixture, and air-dried. This produced an intact composite fibre consisting of densely packed networks of silicalite chains. As before, heating to around 600 °C resulted in the removal of the dextran template, leaving behind a superstructure of silicalite crystals. These threads were highly ordered, with chains of silicalite crystals preferentially aligned along the fibre axis. XRD measurements indicated that the silicalite micropore structure was retained after thermal treatment, and that the peak intensity of the (0100) reflection was reduced compared to a non-oriented control sample. This indicated to the group
52
Biotemplating
that the silicalite crystals within the oriented superstructure were preferentially aligned such that the crystallographic b axis was parallel to the fibre direction. With a very low structural stability, it is unlikely that these materials will find an immediate device application, although they may be able to be applied to highresolution separation technologies where the column is disposable and isolated from mechanical stresses. As a route to complex, porous inorganic phases, dextran has been shown to be particularly efficacious.
3.4 Starch Starch is a complex glucose polymer, a mixture of amylose and amylopectin, in a 1:4 ratio. Starch is the most common foodstuff in the human diet owing to the fact that it occurs in a wide range of staple foods such as root vegetables and cereals; rice, potatoes and pasta probably account for the largest consumption of starch worldwide. In addition, starch finds use in the manufacture of textiles, adhesives and as a recyclable template for the moulding of jellied sweets. Owing to the large number of hydroxyl groups, starch is able to be chemically functionalized to produce cationic or anionic species. This was first made use of in the papermaking industry, where cationic starch was able to bind to negatively charged cellulose fibres of the paper, thereby imparting an enhanced structural stability to the final product40. This chelating ability forms the basis for many of the template uses for starch. Chen et al. used starch modified with mercapto groups which have a high metallophilicity, in order to create starch/silver core shell particles41. They employed a step-wise method of modification of starch with mercapto groups, followed by the deposition of silver nanoparticles by reduction and the removal of the starch core with the enzyme α-amylase. In the first step, confirmation of the successful esterification reaction between starch and mercaptoacetic acid was by FTIR, which showed two peaks indicative of the stretching vibration of C-O and S–H groups at 1730 cm-1 and 2560 cm-1 respectively. The second step involved the emulsification of the functionalized starch by a surfactant in order to form micelles. Introduction of silver ions formed the starch/silver composite particle, the silver ions chelating to the starch through the mercapto groups. TEM revealed that the particles possessed a core-shell morphology, with a diameter of between 50 nm and 150 nm, with a silver shell thickness of around 10 nm. Dynamic light scattering indicated that there was a bimodal distribution of particles, with average sizes of 164 nm and 295 nm. The group attributed the
Complex Polysaccharides
53
difference between these data and that seen from the TEM by the difference between the dried and hydrated states of the system. The final stage of the process was the removal of the starch core by enzymatic degradation by the enzyme α-amylase. For this, the composite nanoparticles were degraded with a 1 g/L α-amylase solution at room temperature for two weeks. The silver shell was found to be permeable to the enzyme solution and so starch degradation was possible, albeit rather slowly. TEM images of the nanoparticles post enzyme digestion revealed hollow silver structures of similar diameter to the composite particles. Chelation of metal ions was also used by Ji et al. in the creation of Nanoporous nickel oxide (NiO) and nickel nanoparticles/amorphous carbon (Ni/C) composites42. In this work, they simply used the introduction of Ni ions to effect the cross-linking and gelation of a starch solution. NiO is an important material for gas sensors43 and electrochemical capacitors for power generation applications44 and Ni/C composite materials have been postulated to be useful in catalysis and medical applications45, 46. In each case, activity will be improved as the particle size of the material reduces. In a simple procedure, soluble starch was dissolved in distilled water, then an aqueous solution of nickel acetate was added. Adjusting the pH of the solution to 9, followed by heating to 100 °C induced gelation. The gel was dried, ground and then calcined either at 380 °C in air or at 550 °C under nitrogen to produce the carbon-containing products. TEM shows that the 8 nm-sized Ni nanoparticles are well dispersed in the amorphous carbon matrix, with little aggregation. N2 adsorption–desorption analysis of the composites revealed a surface area and pore volume of 263 m2g-1 and 0.15 cm3g-1 respectively. The high surface area is attributed to the fact that large numbers of nanoporous form spontaneously in the starch during the process of carbonization. In this work, the starch provides a duel role on calcination, both as carbothermal reducing agent and as an anti-sintering agent, preventing extensive growth of the nanoparticles. Finally, as proof of the functionality of the NiO material, the group found that with cyclic voltammetry, their product exhibited a specific capacitance of 329 Fg-1 at a sweep rate of 10 mVs-1. The group attribute this high specific capacitance of the NiO sample to the nanoporous structure and large pore volume. These physical characteristics can both facilitate the transfer of OH- ions in the electrolyte, thereby allowing for an efficient Faradaic reaction. Titania (TiO2) is another material which has been synthesized using starch as a template. For catalysis and in medical implant technologies, TiO2 is one of the most important materials47, 48. Nanosized TiO2 was synthesized in a manner
54
Biotemplating
similar to that seen in the synthesis of NiO. Conducting the synthesis of TiO2 in the presence of starch allowed Kochkar et al. to produce nanoparticles of 23 nm in diameter49. Iwasaki et al. used the foaming of starch on heating as a route to a complex mesoporous/microporous sponges of TiO250. The group successfully synthesized starch sponges with high internal macroporosities by freezing and thawing of starch gels, which were then subsequently infiltrated with colloidal suspensions of titania nanoparticles (Figure 3.14).
Figure 3.14 – SEM images of the TiO2 sponge prepared in 7 mass% TiO2 dispersion. (a) Lowmagnification image showing a fragment of the intact monolith, (b) higher magnification image showing individual pores and wall structures (scale bar = 10 µm). Reproduced by permission of Springer Publishing Co. Ltd.
Complex Polysaccharides
55
The experimental procedure was a simple addition of potato starch to water, which was heated to gelation. The gel was then frozen at -20 °C and subsequently thawed, to create a macroporous sponge structure. This was the critical step in inducing macroporosity, in the absence of this freeze/thaw step, the starch gels did not show a well-formed sponge structure. Infiltration of the sponges was achieved by immersing the dried gels in a colloidal solution of TiO2, which was stirred gently for up to four days. The composite structure was then air dried and if a thicker coating was required, re-immersed. On drying, the starch foam-TiO2 composite materials were found to have pores of up to 200 µm diameter, with the colloidal particles covering the surface of the sponges. Calcination at 600 °C for 2 hours removed the sponge template, leaving a TiO2 replica of the structure. The photocatalytic activity of the TiO2 sponges was measured by the rate of photocatalytic degradation of acetaldehyde. Figure 3.15 shows the concentration of acetaldehyde with photoirradiation time (t) for (a) TiO2:starch composite material and (b) TiO2 sponge under UV irradiation. The concentration of acetaldehyde was found to decrease with t in both samples, indicating that both pre- and post-calcined materials were photocatalytically active (Figure 3.15).
Figure 3.15 – Variations in CH3CHO concentration with irradiation time for the crushed pieces of TiO2-starch sponge composites (TiO2 loading: 12.5 mass%) (a) and TiO2 sponge (b). Reproduced by permission of Springer Publishing Co. Ltd.
56
Biotemplating
The group found that the apparent rate constant (k/h−1), calculated from the slope of the plots of ln(CO/C) vs. t for the two samples, where CO and C denote the concentrations of acetaldehyde at t = 0 and t = t, respectively, was 0.51 and 2.40 for the TiO2-starch composite and TiO2 sponge, respectively. Normalization of the data to account for the weight difference of the samples revealed that the calculated values (k’/h-1 (TiO2)g-1) were 86.6 and 50.5 respectively. These values are significantly larger than those seen in commercially available TiO2 catalysts. AMT-100 and Degussa have values for (k’/h-1 (TiO2)g-1) of 6.87 and 12.6 respectively. Nanowires have also been synthesized using starch as a template. Shi et al. successfully used the starch as a template on which nanowires of polypyrrole (PPy) were grown51. Polypyrrole is a highly conductive polymer constructed from pyrrole heterocycles (Figure 3.16).
Figure 3.16 – Molecular structure of polypyrrole.
The group found that in a starch solution, the addition of pyrrole monomers resulted in adsorption of the pyrrole onto the starch via hydrogen bonding. The adsorbed pyrrole monomers were then polymerized, the polymerization following the chains of starch molecules to form PPy nanowires. They also found that PPy nanowires could be electrochemically generated on various electrodes including tin-doped indium oxide (ITO) film, stainless steel, titanium, gold and graphite using starch as a template for growth. SEM and TEM images of the PPy
Complex Polysaccharides
57
materials showed a uniform wire-like PPy nanostructure with average diameter of approximately 100 nm, with lengths varying from hundreds of nanometers to several microns. Formation of PPy in the absence of starch resulted in the uncontrolled polymerization and growth of PPy aggregates. Concentration effects of both starch and pyrrole were investigated. Varying the concentration of pyrrole from 0.06 to 0.22 M, resulted in the average diameter of PPy nanowires increasing from 60 nm to 150 nm. Further increase of the pyrrole concentration (more than 0.22 M) lead to PPy without an anisotropic morphology. The concentration of soluble starch was increased from 0.004 to 0.020 wt% resulting in the average diameter of PPy nanowires decreasing from 100 nm to 70 nm. At the lower limit of the concentration of soluble starch (less than 0.004 wt%), more PPy resembling the control morphology formed. The group measured the reflectance FTIR spectra of soluble starch, PPy nanowires, and cauliflower-like PPy synthesized without soluble starch over the range 400–4000 cm-1. They found that the spectra of nanowires and control PPy are nearly identical and showed characteristic PPy peaks, such as the asymmetric and symmetric ring stretching at 1555 cm-1 and 1469 cm-1, the C–N stretching vibration at 1188 cm-1, and C–H wagging vibrations at 782 cm-1. However, no characteristic peaks of soluble starch were observed in the IR spectrum of PPy nanowires. These results indicated that nanowires of PPy were prepared without any other complex present and that the soluble starch is not incorporated into the PPy nanowires. Measurements of the electrochemical activity of electrodes made with PPy nanowires was by cyclic voltammetry in saline solutions, which showed that whilst both nanowire and control samples were electrochemically active, PPy in the nanowire morphology contained stronger corresponding cathodic and anodic peaks. Deposition of PPy nanowires directly onto functional substrates may therefore provide an method to enhance conductivity in electrochemical applications. Another conductive material, tellurium has been made in nanowire form by the use of a starch template. Tellurium is useful as a semiconductor, with a narrow band gap of approx. 0.35 eV and a pronounced piezoelectric effect. Tellurium nanowires have been synthesized previously, using hydrazine as a reducing agent and a combination of surfactants, which rely on, or produce harmful reagents and by-products52, 53. It was found that by utilizing starch as a reducing sugar, the compound H2TeO4 • 2H2O could be reduced under hydrothermal conditions to form tellurium nanowires in a high yield54. As well as
58
Biotemplating
being a more benign synthesis, the starch acted to sequester the tellurium salt and provided the reducing capability in the system, thereby ensuring a homogeneous distribution of the desired phase throughout the composite precursor material. A typical synthesis consisted of a simple admixture of H2TeO4 • 2H2O in a starch solution, followed by 15 hours in an autoclave at 160 °C. TEM images show the presence of long nanowires of widths approximately 25 nm and lengths of several tens of microns (Figure 3.17).
Figure 3.17 – TEM image of the obtained Te sample with starch as reducing agent. Scale bar is 2 µm. Reprinted with permission from54. Copyright 2005 American Chemical Society.
Complex Polysaccharides
59
The preferred direction of growth of the nanowires was determined from TEM as being the [001] direction, the mechanism of formation likely to be similar to that of the PPy nanowires above. Finally, the use of starch as a simple carbon source was demonstrated by Shen et al. in the synthesis of mesoporous carbon materials55. The mixing of starch, sulfuric acid and a surfactant with tetraethoxysilane (TEOS) produced a gel which was then heated to various temperature under nitrogen to produce a porous carbon material. As was the case for the simple sugars, the majority of these studies are still using the biopolymer as a simple space filling agent to induce porosity, however we have now seen the emergence of a more active role for the organic phase. The more complex the biopolymer, the greater the potential for the direct templating of that complex form. In the next chapter, we examine a group of polysaccharides which possess an even greater level of molecular complexity. This imbues them with the properties of extremely effective gelation, film formation and an affinity for metal cations; all of which make them ideal as templating agents.
3.5 References 1. 2. 3.
4. 5. 6. 7.
8.
Zheng, Z. et al. Biomimetic growth of biomorphic CaCO3 with hierarchically ordered cellulosic structures. Cryst. Growth Des. 7, 1912-1917 (2007). Aoki, Y., Huang, J.G. & Kunitake, T. Electro-conductive nanotubular sheet of indium tin oxide as fabricated from the cellulose template. J. Mater. Chem. 16, 292-297 (2006). Huang, J.G., Ichinose, I. & Kunitake, T. Nanocoating of natural cellulose fibres with conjugated polymer: hierarchical polypyrrole composite materials. Chem. Commun., 1717-1719 (2005). Huang, C.T., Shen, C.L., Tang, C.F. & Chang, S.H. A wearable yarn-based piezo-resistive sensor. Sens. Actuator A-Phys. 141, 396-403 (2008). Kaynak, A., Najar, S.S. & Foitzik, R.C. Conducting nylon, cotton and wool yams by continuous vapor polymerization of pyrrole. Synth. Met. 158, 1-5 (2008). Locher, I. & Trosler, G. Screen-printed textile transmission lines. Text. Res. J. 77, 837-842 (2007). Rattfalt, L., Linden, M., Hult, P., Berglin, L. & Ask, P. Electrical characteristics of conductive yarns and textile electrodes for medical applications. Med. Biol. Eng. Comput. 45, 1251-1257 (2007). Shaw, R.K., Long, B.R., Werner, D.H. & Gavrin, A. The characterization of conductive textile materials intended for radio frequency applications. IEEE Antennas Propag. Mag. 49, 28-40 (2007).
60 9.
10. 11.
12. 13. 14. 15. 16.
17. 18. 19. 20.
21. 22. 23.
24. 25. 26. 27. 28. 29.
Biotemplating Nelson, K. & Deng, Y.L. The shape dependence of core-shell and hollow titania nanoparticles on coating thickness during layer-by-layer and sol-gel synthesis. Nanotechnology 17, 3219-3225 (2006). Maneerung, T., Tokura, S. & Rujiravanit, R. Impregnation of silver nanoparticles into bacterial cellulose for antimicrobial wound dressing. Carbohydr. Polym. 72, 43-51 (2008). Liu, S.L., Zhang, L., Zhou, J.P. & Wu, R.X. Structure and properties of cellulose/Fe2O3 nanocomposite fibres spun via an effective pathway. J. Phys. Chem. C 112, 4538-4544 (2008). Hwang, S.H. et al. Construction of CdS quantum dots via a regioselective dendritic functionalized cellulose template. Chem. Commun., 3495-3497 (2006). Tsioptsias, C. & Panayiotou, C. Preparation of cellulose-nanohydroxyapatite composite scaffolds from ionic liquid solutions. Carbohydr. Polym. 74, 99-105 (2008). Wu, M.D., Pan, X.X., Qian, X.F., Yin, J. & Zhu, Z.K. Solution-phase synthesis of Ag2S hollow and concave nanocubes. Inorg. Chem. Commun. 7, 359-362 (2004). Wang, C., Zhang, X., Qian, X., Wang, W. & Qian, Y. Ultrafine powder of silver sulfide semiconductor prepared in alcohol solution. Mater. Res. Bull. 33, 1083-1086 (1998). Qian, X.F., Yin, J., Feng, S., Liu, S.H. & Zhu, Z.K. Preparation and characterization of polyvinylpyrrolidone films containing silver sulfide nanoparticles. J. Mater. Chem. 11, 2504-2506 (2001). Liu, S.H. et al. Synthesis and characterization of Ag2S nanocrystals in hyperbranched polyurethane at room temperature. J. Solid State Chem. 168, 259-262 (2002). Zeng, J. et al. Necklace-like noble-metal hollow nanoparticle chains: Synthesis and tunable optical properties. Adv. Mater. 19, 2172-2176 (2007). Zollfrank, C., Scheel, H. & Greil, P. Regioselectively ordered silica nanotubes by molecular templating. Adv. Mater. 19, 984-987 (2007). Liu, W.J. et al. Fabrication of CdS nanorods in inverse microemulsion using HEC as a template by a convenient gamma-irradiation technique. J. Cryst. Growth 290, 592-596 (2006). Han, S.O., Son, W.K., Youk, J.H., Lee, T.S. & Park, W.H. Ultrafine porous fibres electrospun from cellulose triacetate. Mater. Lett. 59, 2998-3001 (2005). Kadla, J.F., Asfour, F.H. & Bar-Nir, B. Micropatterned thin film honeycomb materials from regiospecifically modified cellulose. Biomacromolecules 8, 161-165 (2007). Fadhli, H.A., Fine, D.P. & Mazuji, M.K. Intra-Arterial Infusion of Dextran - A New Concept to Prevent Sludging and Thrombosis in Vascular Surgery. J. Thorac. Cardiovasc. Surg. 53, 496-499 (1967). Walsh, D., Arcelli, L., Ikoma, T., Tanaka, J. & Mann, S. Dextran templating for the synthesis of metallic and metal oxide sponges. Nat. Mater. 2, 386-390 (2003). Khan, F., Eswaramoorthy, M. & Rao, C.N.R. Macroporous silver monoliths using a simple surfactant. Solid State Sci. 9, 27-31 (2007). Kong, R., Yang, Q. & Tang, K.B. A facile route to silver nanowires. Chem. Lett. 35, 402-403 (2006). Gonzalez-McQuire, R. et al. Fabrication of hydroxyapatite sponges by dextran sulphate/amino acid templating. Biomaterials 26, 6652-6656 (2005). Yeoh, L.M., Ahmad, M. & Abd-Shukor, R. Superconductivity in Ru-based cuprate Ru(Sr1.5Ca0.5)PbCu2O8 prepared by sol-gel route. Mod. Phys. Lett. B 22, 1441-1446 (2008). Zalga, A., Grigoraviciute, I. & Kareiva, A. Sol-Gel preparation of nonstoichiometric Bi,Pb2223 superconductors. Chemija 18, 7-10 (2007).
Complex Polysaccharides 30.
31.
32. 33. 34.
35. 36. 37.
38. 39. 40. 41. 42.
43. 44. 45. 46. 47. 48. 49.
50.
61
Nenartaviciene, G., Beganskiene, A., Tautkus, S., Jasaitis, D. & Kareiva, A. Chromium substitution effects in Y-124 superconductor prepared by aqueous sol-gel method. Chem. Phys. 332, 225-231 (2007). Zalga, A., Reklaitis, J., Norkus, E., Beganskiene, A. & Kareiva, A. A comparative study of YBa2Cu4O8 (Y-124) superconductors prepared by a sol-gel method. Chem. Phys. 327, 220-228 (2006). Grigoryan, S. et al. A new way of preparing the Y-Ba-Cu-O high-temperature superconductor using the sol-gel method. Supercond. Sci. Technol. 16, 1202-1206 (2003). Sanjines, R., Thampi, K.R. & Kiwi, J. Preparation of Monodispersed Y-BA-CU-O Superconducting Particles via Sol-Gel Methods. J. Am. Ceram. Soc. 71, C512-C514 (1988). Walsh, D., Wimbush, S.C. & Hall, S.R. Use of the polysaccharide dextran as a morphological directing agent in the synthesis of high-Tc superconducting YBa2Cu3O7-delta sponges with improved critical current densities. Chem. Mat. 19, 647-649 (2007). Apreutesei, R.E., Catrinescu, C. & Teodosiu, C. Surfactant-modified natural zeolites for environmental applications in water purification. Environ. Eng. Manag. J. 7, 149-161 (2008). Cheetham, A.K., Ferey, G. & Loiseau, T. Open-framework inorganic materials. Angew. Chem.-Int. Edit. 38, 3268-3292 (1999). Coombs, D.S. et al. Recommended nomenclature for zeolite minerals: Report of the subcommittee on zeolites of the International Mineralogical Association, Commission on New Minerals and Mineral Names. Can. Mineral. 35, 1571-1606 (1997). Walsh, D. et al. Preparation of higher-order zeolite materials by using dextran templating. Angew. Chem.-Int. Edit. 43, 6691-6695 (2004). Li, M., Schnablegger, H. & Mann, S. Coupled synthesis and self-assembly of nanoparticles to give structures with controlled organization. Nature 402, 393-395 (1999). Vanderburgh, L.F. Use of Cationic Starch in Papermaking. Pulp and Paper Magazine of Canada 71, 81 (1970). Wang, L.M. & Chen, D.J. A facile method for the preparation of hollow silver spheres. Mater. Lett. 61, 2113-2116 (2007). Chen, Y.P. et al. Novel synthesis of nanoporous nickel oxide and nickel nanoparticles/amorphous carbon composites using soluble starch as the template. Chem. Lett. 35, 700-701 (2006). Alcock, C.B., Li, B.Z., Fergus, J.W. & Wang, L. New Electrochemical Sensors for Oxygen Determination. Solid State Ion. 53-6, 39-43 (1992). Conway, B.E. Transition from Supercapacitor to Battery Behaviour in Electrochemical Energy Storage. J. Electrochem. Soc. 138, 1539-1548 (1991). Dravid, V.P. et al. Controlled Size Nanocapsules. Nature 374, 602-602 (1995). Lu, A.H. et al. Highly stable carbon-protected cobalt nanoparticles and graphite shells. Chem. Commun., 98-100 (2005). Kung, H.H. & Ko, E.I. Preparation of oxide catalysts and catalyst supports - A review of recent advances. Chem. Eng. J. 64, 203-214 (1996). Lindberg, F., Heinrichs, J., Ericson, F., Thomsen, P. & Engqvist, H. Hydroxylapatite growth on single-crystal rutile substrates. Biomaterials 29, 3317-3323 (2008). Kochkar, H., Triki, M., Jabou, K., Berhault, G. & Ghorbel, A. Novel synthesis route to titanium oxides nanomaterials using soluble starch. J. Sol-Gel Sci. Technol. 42, 27-33 (2007). Iwasaki, M., Davis, S.A. & Mann, S. Spongelike macroporous TiO2 monoliths prepared from starch gel template. J. Sol-Gel Sci. Technol. 32, 99-105 (2004).
62 51.
52. 53.
54. 55.
Biotemplating Shi, W., Liang, P.F., Ge, D.T., Wang, J.X. & Zhang, Q.Q. Starch-assisted synthesis of polypyrrole nanowires by a simple electrochemical approach. Chem. Commun., 2414-2416 (2007). Mo, M.S. et al. Controlled hydrothermal synthesis of thin single-crystal tellurium nanobelts and nanotubes. Adv. Mater. 14, 1658-1662 (2002). Mayers, B. & Xia, Y.N. One-dimensional nanostructures of trigonal tellurium with various morphologies can be synthesized using a solution-phase approach. J. Mater. Chem. 12, 1875-1881 (2002). Lu, Q.Y., Gao, F. & Komarneni, S. A green chemical approach to the synthesis of tellurium nanowires. Langmuir 21, 6002-6005 (2005). Shen, W.Z. et al. The effect of carbon precursor on the pore size distribution of mesoporous carbon during templating synthesis process. Mater. Lett. 60, 3517-3521 (2006).
Chapter 4
Hydrocolloids
In Chapter 3, we were introduced to some of the more complex polysaccharides. The complexity arose in these molecules through a combination of branched structures and extended chain lengths. Whilst still complex polysaccharides, the so-called hydrocolloids have another layer of complexity, in that they form extended macro-structures by the nature of their molecular configuration. Hydrocolloids are hydrophilic in nature and are found in almost every biosphere on earth; in plants, animals and many bacteria. They contain a large number of hydroxyl groups arranged usually in a fairly regular manner along the backbone of the molecule, which allows for the chelation of mono- and divalent cations, thereby cross-linking the hydrocolloid chains together and forming complex macrostructures. This determines the main function they have in nature; the cross-linking affording the hydrocolloid a good degree of mechanical strength, which allows them to act in a structure forming role. Because of this sequestering ability, hydrocolloids have found a great deal of use in the food industry, as they selectively alter the physical behaviour of foodstuffs, particularly in the aqueous phase. The most utilized properties of the hydrocolloids in this way are to stabilize emulsions, to thicken and gel, to stabilize foams and to prevent large crystals of ice from forming. When the physical nature/behaviour of the foodstuff is important (i.e. the flow of tomato ketchup from the bottle) it is more often than not, a hydrocolloid that is responsible for the behaviour. As a result, the hydrocolloids are some of the most well-known polysaccharides (Table 4.1)
4.1 Structure and properties As with the complex polysaccharides, the large number of residues and range of glycosidic bonding forms the basis for structural complexity. The hydrocolloids are different however, in that the majority of these polysaccharides 63
64
Biotemplating
contain more than one type of residue. As they are all of biogenic origin, there can be quite a wide variation of composition for any given hydrocolloid, but in each case an idealized structure will be given which will represent the major constituent parts of each. The simplest interactions of hydrocolloids are (as the name suggests) with water and it is this interaction that is key to their use in foodstuffs. The hydrocolloids are effective water adsorbers and to a greater and lesser degree will be solubilized by water. Owing to the high number of hydroxyl groups, water is held within the molecular structure by hydrogen bonding and also within the voids created by the complex molecular configuration. Table 4.1 – A list of common hydrocolloids. Hydrocolloid
Source
Agar Alginate Carrageenan Curdlan Gelatin Gellan Guar gum Gum arabic Locust bean gum Pectin Xanthan gum
Red algae/seaweed Brown algae/seaweed Red seaweed Agrobacterium biobar bacteria Animal collagen Sphingomonas elodea bacteria Guar beans Acacia trees Carob tree seeds Plant cell walls Xanthomonas campestris bacteria
When hydrated, the hydrocolloids form viscous solutions and gels of varying strength and flow rates, depending on the degree of entanglement of the polymer. In addition, mixtures of hydrocolloids can act synergistically, forming structures with enhanced properties. Gelation in the hydrocolloids occurs when intra- and inter-molecular hydrogen bonding (and sometimes salt formation) is preferred over simple ionic and hydrogen bonding with water. In the gel state, hydrocolloids form extended networks of complex entangled molecules, often with the resultant formation of helical regions. In forming a regular, complex morphological arrangement, the hydrocolloid provides a higher-order structure amenable to templating. As a further complication, mixtures of hydrocolloids can give rise to even more complexity of structure and physical behaviour. For example, mixtures of
Hydrocolloids
65
concentrated locust bean gum and xanthan gum behave quite differently in solution than they would on their own. Together, they behave as a hyperentangled macromolecular network, resistant to flow, whereas xanthan gum alone will only form weak gel-like networks which are ‘wet’ and able to flow without a great deal of resistance. Many of the hydrocolloids have not been investigated as templating agents for inorganic phases owing to more complex behaviour being able to be generated from other hydrocolloids. Equally, some of the more common ones have a plethora of studies carried out using them, many of which can be grouped as conceptually similar. This chapter will concentrate on those broad concepts which have been investigated using each hydrocolloid and the complex morphologies resulting from their use.
4.2 Carrageenan Carrageenan is a polysaccharide derived from the family of red seaweeds Rhodophycae. The majority of carrageenan is extracted from seaweeds from the genus Chondrus, Eucheuma, Gigartina and Iridaea. This gives rise to several distinct types of carrageenan, classified largely according to the degree of sulfation of the polymer backbone. The three types are kappa-carrageenan, iotacarrageenan and lambda-carrageenan. As with all of the biopolymers, the determined structure of carrageenan can only ever be an ideal, as in practice, seaweed extracts can contain varying degrees of sulfation even at different locations along the polymer backbone (Figure 4.1). Most commonly used for biotemplating studies is the kappa form of carrageenan. Kappa carrageenan consists of repeating disaccharide units of alternating (1→3)-α-D-galactose-4-sulphate and (1→4)-β-3,6-anhydro-Dgalactose residues joined in a linear chain. Addition of counter cations into kappa-carrageen is known to change its gelation mechanism. Chandrasekaran et al.1 used X-ray fibre diffraction to prove that on addition of calcium cations, the carrageenan molecules formed a threefold, right-handed, half-staggered parallel double helix, with a pitch of 26.42°. This method of binding was found to be as a result not of the traditional helix-helix interaction, but of cations binding with the sulfate groups, tying the helices together. The result of this cation-mediated interaction is the formation of a much stronger, contracted gel. Chandrasekaran et al. also studied X-ray diffraction patterns from both polycrystalline and well-ordered fibres of carrageenan, when chelating calcium,
66
Biotemplating
rubidium and potassium salts, and found that the biopolymers formed fibres with larger unit cell distances with trigonal symmetry1. It is this sequestration of discrete metal cations which provides the basis for nucleation and growth of nanoparticulate, biotemplated materials. The morphological regularity of metal-sulfate binding in carrageenan provides spatially-defined sites of nucleation and growth, thereby imbuing the inorganic material (at least in the very first instances of mineralization) with a complex morphology matching that of the underlying matrix.
Figure 4.1 – Molecular structures of the three most common types of carrageenan. (a) kappacarrageenan, (b) iota-carrageenan, (c) lambda-carrageenan.
Hydrocolloids
67
Axelos et al. looked at the effect of different mono- and divalent cations on the gelation and phase separation stabilities of iota- and kappa-carrageenan2. They found that the storage modulus (G’) in kappa-carrageenan was always higher than that seen in iota-carrageenan. This study no doubt directed future research which tends to prefer the kappa- form of carrageenan as being more suitable for studies on mineralization. Jones et al. found that by neutralizing iron salts in aqueous solutions of carrageenan, an iron oxyhydroxide/polysaccharide hybrid colloidal material could be produced which was stable up to pH 13 3. The carrageenan was found to be forming a tight polymer layer surrounding the inorganic particles, thereby stabilizing the material to degradation. They found that in addition, they were able to react the colloids further, producing stabilized magnetite nanoparticles. Developing this synthetic protocol, they found that they were able to use carrageenan gels at high pH to also stabilize the formation of cobalt and nickel oxide nanoparticles. In a simple synthesis, a metal/polymer composite solution was added to a solution of carrageenan at a molar ratio of 1:1 (metal cation:SO4), followed by a raising of the pH to 13. The addition of Fe3+, Ni2+, and Co2+ to kappa-carrageenan followed by a large increase in pH resulted in transparent solutions of polysaccharide-stabilized metal hydroxides which were brown, green and blue respectively. The group found that whereas the iron and nickel colloids were stable over several months, the cobalt system developed a redispersable sediment. As the solutions were largely colloidally stable, a measurement of the metal cation environment in the gels was able to be gleaned from UV-Vis spectroscopy. They found that the iron oxyhydroxide-carrageenan spectrum is similar to that expected for Fe3+ solutions at pH > 3, but the peaks at approximately 250 and 400 nm wavelength seen in the cobalt spectrum do not correspond to tetrahedral cobalt (which has peaks from 600 to 700 nm). Instead the group attribute these peaks to Co located inside the nanoparticles which cannot undergo solvent induced surface transitions (Figure 4.2). Ni(OH)2 composite materials showed a small but significant absorbance in the 600–700 nm region typical of Ni2+ in the octahedrally coordinated configuration and a small absorption at 400 nm that can also be attributed to the nickel aqua ion. Owing to the different electronic structure of the nickel system, it is quite stable with regards to oxidation. Each of the metal/carrageenan composite systems was subjected to ultracentrifugation in order to qualitatively determine the strength of binding of the metal cations. Despite the vigorous
68
Biotemplating
centrifugation conditions imposed on the composites, there was only a minimal amount of metal released from each system under investigation. This was shown by the fact that there was only a very slight decrease in absorption amplitude on the UV-Vis spectra. From TEM and SEM studies, the group found that with cobalt at pH 13, the carrageenan acted effectively to control both the size and shape of the nanoparticle. Crystals formed were hollow shells or hollow polygons of 80 nm diameter which eventually fill and sediment over time.
Figure 4.2 – UV-Vis spectra of Co2+ in carrageenan at pH 13 versus time showing oxidation to Co3+. Reprinted with permission from3. Copyright 2000 American Chemical Society.
Similarly, iron III oxyhydroxide nanoparticles formed initially as hollow cages, with the added level of complexity in that they formed strings of nanoparticles. These iron nanoparticles were stable with regards to transformation into goethite even after 10 months of ageing. Nickel composites showed little or no control over crystal growth however, owing to the fact that the nickel phase is more colloidally stable in the carrageenan. A small reduction in the overall particle size was observed as the carrageenan was effectively acting as an anti-coagulating agent in the growth of the nickel
Hydrocolloids
69
phase. The bulk aggregate morphology of the crystallites was also changed by using carrageenan. The group found that in their cobalt system, spherical 10 µm diameter conglomerates formed, which consisted of 100–200 nm sized platelets. The formation and growth of platelets is analogous to the mineralization in constrained volumes which is a feature of biomineralization. In this case, it is the carrageenan that is providing a cocooning shell around the inorganic phase, preventing uncontrolled growth; the biopolymer ‘shell’ is stabilized by the crosslinking action of the metal cations. The successful functioning of the carrageenan in this system is determined by the survivability of the microgel structure. It was determined that for iron oxyhydroxide and nickel hydroxide, the stability was conferred to the nanoparticles by the fact that the reaction forming the metal hydroxide releases at most 50–75% of the carrageenan from the microgel structure after mineralization. This results in discrete nanoparticles of the inorganic phase being established before the cocooning effect of the biopolymer is lost. In contrast, on oxidation, cobalt mineralization resulted in 65–75% of the carrageenan being released. This had the effect of disrupting the stability of the microgel and therefore producing a weaker interaction of carrageenan with the cobalt (III) hydroxide. This destabilization allowed large aggregates to form. An exploration of the synthesis of magnetite nanoparticles using different carrageenans was undertaken by Gil et al. In their work, they examined the effect of the type of carrageenan on metal uptake and subsequent growth4. They found that as expected, the overall size of the magnetite particles were smaller when carrageenan was used (as compared to the control samples) and that generally, the size of the nanoparticles decreased with increasing carrageenan concentration. More interestingly, the group found that ί-carrageenan nanoparticles showed higher than average sizes, compared to λ- and κcarrageenan. They attribute this to the stabilization of hydrolyzed iron species by the particular sulfate charge distribution of ί-carrageenan. Finally, they note that the addition of ferric ions to ί-carrageenan leads to the formation of a heterogeneous gel, whereas for both λ- and κ-carrageenan/iron gels are homogeneous. This is an indication that for this particular system, ί-carrageenan is not able to bind all of the introduced iron, therefore leading to unconstrained nucleation and growth of larger nanoparticles. Another group investigated the ability of carrageenan to adsorb heavy metal ions from solution. The removal of heavy metal ions from water is a major area of research, as the number of industrial processes which involve the discharge of
70
Biotemplating
heavy metal contaminated waste water into the environment has increased dramatically5-7. In their work, Son et al.8 highlight the concerns that processes such as metal plating facilities, mining and tanneries, generate aqueous waste streams which contain heavy metal pollutants9. Increasingly stringent legislation has ensured that there is a large market for remediative processes, with the World Health Organization announcing a list of heavy metal contaminants of most concern, including aluminum, chromium, manganese, iron, cobalt, nickel, copper, zinc, cadmium, mercury and lead. The usual recovery processes for heavy metal remediation involve chemical precipitation or electrolytic recovery, both of which have drawbacks in the cost and amount of energy required to drive the process. Son et al. investigated several biopolymers, chitosan, alginic acid and carrageenan with particular regard to their ability to effectively chelate the heavy metal ions of mercury, lead and copper8. With the polysaccharides immobilized on a polyvinyl alcohol bead in a column, waste water streams were flowed through and the metal uptake recorded. The group investigated λ-carrageenan as it contains three sulfate groups per monomer unit and therefore should theoretically have the greatest affinity of all the carrageenans for metal cations. The solution of heavy metal cations and immobilized polysaccharide was mechanically agitated at 250 rpm for 24 hrs, and then the suspending biomass was removed by centrifugation at 10,000 rpm for 20 min. The supernatant was collected to measure the final concentration of heavy metal by inductively coupled plasma spectroscopy. It was found that of the heavy metal ions, copper showed the greatest affinity for λ-carrageenan, which had an adsorption capacity of 3.35 mmol/g. The group found that, perhaps as expected, polysaccharides such as cellulose, with only hydroxyl functional groups showed a significantly lower adsorption (between 1% and 10%) than the carrageenan, indicating that for bioremediation, it is the functional group of the polysaccharide which is a key factor to determine its adsorption capacity with metal ions. Carrageenan has also been used to control the crystal growth of calcium carbonate phases. Butler et al. investigated a series of biopolymers, carrageenan included, for the effect on growth of calcite10. They found that in the presence of κ-carrageenan, crystallization of the usual rhombohedral calcite was suppressed in favour of rosette-like polycrystalline aggregates (Figure 4.3).
Hydrocolloids
71
Figure 4.3 – STEM micrograph of calcium carbonate shells grown in the presence of κ-carrageenan. Reprinted with permission from10. Copyright 2006 American Chemical Society.
By increasing the concentration of κ-carrageenan, the group found that the increased calcium binding tended to inhibit calcite crystal formation. This behaviour has been noted before in studies on poly-aspartic acid11, 12 although this effected a polymorphic change when low concentrations of the polymer were introduced into the system. Amorphous materials are also able to have complex form imbued upon them using carrageenan. Shchipunov investigated the condensation of silica in the presence of κ-, λ-, and ί-carrageenans, using the precursor tetrakis(2hydroxyethyl) orthosilicate (THEOS)13. First introduced by Hoffmann et al.14, this siloxane is of particular value to biopolymer-mediated mineralization, as it is soluble in water. The vast majority of silica-based syntheses use either
72
Biotemplating
tetraethoxysilane (TEOS) or tetramethoxysilane (TMOS), both of which are water insoluble and require organic solvents in their use. In addition, the hydrolysis and condensation chemistry of TEOS and TMOS results in the production of ethanol and methanol respectively. This can be detrimental as the presence of organic solvents is contraindicatory to most biological systems and can lead to denaturing of organic matter, thereby negating any templating effect. With THEOS, no organic solvents are used or produced in the condensation of silica and so carrageenan can be used unfettered. It was found, as shown above, that an increase in the concentration of the cationic species produced crosslinking of the biopolymer, leading to a stiffer and more structurally stable composite gel. Composites made with κ-carrageenan showed a marked shrinkage of the material, whereas λ- and ί-carrageenans did not. This is in line with the gelation behaviour of the different carrageenans; as λ-carrageenan does not exhibit gelling behaviour and therefore is not substantively affected by the introduction of silica ions. SEM investigations of the silica/carrageenan composites show a marked influence of the carrageenan on the silica phase. In the absence of biopolymer, THEOS condensation produces solid particles of silica in the region of 10 nm to 40 nm diameter. Absolute determination of the particle size was difficult however, as there was extensive condensation of the nanoparticles, resulting in a solid gel-like network. In the presence of κ-carrageenan however, it was found that a 3D fibrous network of silica was formed, with an absence of gel-like aggregations. With an amorphous material such as silica, the templating effect of the biopolymer is likely to be more pronounced, as there is no impetus to form well-defined crystal faces and the inorganic phase will more readily adopt the underlying biopolymer morphology. In the main, a biopolymer acts largely as an anti-sintering matrix for the mineral phase. In previous chapters, the biopolymers were all relatively chemically benign, in the sense that there was an absence of reactive functional groups in the molecule. This, as we have seen however, is not the case for carrageenan. When applied to the synthesis of superconducting yttrium-bariumcopper-oxide (YBCO) nanoparticles, the carrageenan proved to be too chemically reactive. Hall et al. prepared YBCO in the presence of carrageenan as a gelating agent. After calcination to form the superconducting phase, nanoparticles were indeed produced15, TEM showing that crystal morphology was controlled, producing crystallites of between 100 nm and 200 nm in diameter. In all cases where carrageenan was used as a template, the formation
Hydrocolloids
73
of morphologically distinct nanotapes was also observed. Nanotape dimensions varied considerably, their length ranging from 100–1500+ nm, and their widths from 40–80 nm (Figure 4.4).
Figure 4.4 – (a) TEM image of attempted YBCO growth in the presence of carrageenan and (b) HRTEM image of BaSO4 nanotape. Scale bar in (a) is 500 nm, in (b) 5 nm. Reprinted from15. Copyright 2008, with permission from Elsevier.
High resolution TEM images of these nanotapes revealed that they were single crystals of barium sulfate, BaSO4. The nanotapes possessed pronounced lattice fringes, which ran parallel to the crystallographic axis of elongation. The
74
Biotemplating
lattice spacing perpendicular to the length of the nanotapes was determined to be 0.7 nm, consistent with the crystallographic a axis of BaSO4 (orthorhombic lattice, space group Pbnm, JCPDS pattern #24-1035). Electron diffraction of the nanotapes produced intense spot patterns which were indexed to BaSO4 and powder X-ray diffraction confirmed that the product as a whole consisted of a complex mixture of phases. The persistence of BaSO4 as an impurity in the reaction mixture has an adverse effect on the stoichiometry of the superconducting phase. By abstracting a large amount of barium, the sulfate groups on the carrageenan render the attainment of a superconducting phase unlikely, instead leading to the formation of a large number of phases normally regarded as ‘impurity’ phases in YBCO synthesis. SQUID magnetometry showed that unlike the control, the samples made with carrageenan were nonsuperconducting, even down to 10 K. The appearance of more reactive functional groups in the biopolymer reminds us that care must be taken to ensure that no competing reactions are possible between the templating phase and any of the precursor materials of the synthesis.
4.3 Alginate Alginate is a polysaccharide derived largely from the family of brown seaweeds Phaeophyceae, which includes the kelp Laminaria. Chemically, the polymer is an unbranched, anionic polysaccharide, consisting of blocks of polyguluronate (-G-)n, polymannuronate (-M-)n and alternating (-G-M-)n residues, the proportions of which depend on the species of seaweed from which the alginate is extracted. Within a particular species the amount and position of the poly-G and poly-M blocks can vary, depending on the properties that the alginate must confer to the organism. As it is the poly-guluronate blocks which strongly sequester divalent cations such as Ba2+ and Cu2+, regions of the seaweed which need to be resistant to tidal forces, for example the holdfast and stalk, will be rich in poly-G residues, whereas the fronds which require flexibility to move with the tide, will be richer in poly-M residues (Figure 4.5). The ability of poly-G blocks to bind cations strongly is a feature of the molecular shape conferred on these blocks by the nature of the residue linkages. When two poly-G regions on separate alginate molecules align, a pocket is created in which cross-linking can occur readily (Figure 4.6).
Hydrocolloids
75
Figure 4.5 – Molecular representation of G and M groups in alginate.
As the poly-G regions are regularly spaced in any given alginate, this leads to a regular array of cross-linking junction zones, dubbed the ‘egg-box model’. It should be immediately obvious that this feature can be exploited for the controlled growth of other cationic species. By introducing metal cations to a solution of alginate, there will be preferential uptake of the metallic species into the poly-G egg-box, leading to controlled nucleation and growth of nanoparticulate species when subsequent processing steps are taken. This stratagem was adopted by Hall et al. in the synthesis of single crystal superconductor nanowires16. In this work, use was made of the egg-box binding model to produce ordered arrays of barium cations, which were able to subsequently act as preferential sites of nucleation for barium carbonate nanoparticles. Once again, the cocooning effect of the biopolymer prevented coalescence of the barium nanoparticles, leading to uniform, homogeneously dispersed nanoparticles throughout the matrix. On calcination, once the correct temperature was reached, growth of superconductor needles occurred radially from partially embedded barium carbonate nanoparticles exposed on the surface of the amorphous material. SEM revealed that the material comprised a complex
76
Biotemplating
microstructure of fibrous aggregates, up to several microns in thickness (Figure 4.7a).
Figure 4.6 – Schematic showing the cross linking of a metal cation within a poly-G pocket of alginate.
These fibres had smooth surfaces and a circular cross section of 50–80 nm in diameter. On further investigation, TEM imaging of fractured fibres indicated that they themselves are comprised of bundles of straight-sided filaments, typically 10 nm in diameter (Figure 4.7b). Electron diffraction showed that these fibres were uniaxially aligned, with a preferred direction of growth. SQUID magnetometry showed that these fibres had a critical temperature of 77 K, which is only slightly lower than the 85 K typical for Y124. On a more fundamental level, alginate has been demonstrated to alter the crystallization of calcite on crystallographically oriented substrates of indium tin oxide (ITO). Crystallization of calcium carbonate was studied by Pavez et al. using an electrochemical method in which an ITO substrate as a working electrode was immersed in an alkaline electrolyte solution of CaCl2 (2 mM), NaHCO3 (6 mM) and NaCl (10 mM)17. As the cell works on the electrochemical reduction of O2 by applying a negative potential, there will be localized regions
Hydrocolloids
77
of pH increase, leading to surface bicarbonate ions transforming to carbonate and thus precipitation of CaCO3. Morphologically, the crystal structure of the CaCO3 phase was found to be intimately linked to the growth conditions, in particular the presence of alginate in the system. With no control of mineralization, the usual calcite rhombohedra grew, with an average crystal size of 8 µm.
Figure 4.7 – (a) SEM image and (b) TEM image of Y124 nanowires synthesized in the presence of alginate. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
On addition of Mg2+ to the electrolyte, crystal growth changed dramatically, with the appearance of individual and star-shaped aggregates of CaCO3. This particular change in morphology can be explained by the formation of aragonite as well as an elongation in the c-axis of CaCO3, producing pronounced truncated rhombohedra of Mg-calcite. The group found that on addition of alginate to the system, a drastically different morphology was observed, with the appearance of flat, discoidal crystals which varied greatly in size. Whilst the rhombohedra crystals of the control and Mg-calcite gave peaks in the XRD which were indexed to calcite, once alginate was incorporated, these peaks disappeared and only a very weak signal due to the (012) reflection of aragonite was present. The group postulate that the alginate is acting to prevent incorporation of Mg2+ into the crystal lattice of Mg-calcite by strongly sequestering it, thereby inhibiting the growth of this polymorph in favour of the aragonitic phase. The excellent calcium binding characteristics of alginate have also been used in the controlled growth of calcium phosphate spheres. In their work, Barbosa et al. used powdered inorganic material, either calcium titanium phosphate (CPT) or hydroxyapatite which were admixed with a solution of Na-alginate to form a
78
Biotemplating
homogeneous paste18. This paste was then added dropwise to a 0.1M CaCl2 solution, whereupon spherical particles instantly formed. The size of the spherical particles was able to be controlled by the flow-rate of the droplet using a syringe pump. After ageing for 30 minutes, the particles were isolated, washed and dried. As calcium phosphate phases find applications as drug delivery and bone regeneration materials, these spheres were tested for release properties by the immobilization and release of the therapeutic enzyme glucocerebrosidase, employed in the treatment of Gaucher disease19. Owing to the simple nature of the synthesis, the enzyme was able to be incorporated into the ceramic-alginate matrix in two different ways. Either the enzyme was adsorbed onto the ceramic particles prior to incorporation with the alginate or it was dispersed into the alginate directly. The method of incorporation had a marked effect on the release profiles of the enzyme. The group found that slow enzyme release occurred when adsorption of the enzyme to the ceramic powders, whereas an initial fast release was achieved when the enzyme and the ceramic particles were dispersed in the alginate solution before producing the microspheres. The difference in the release profiles can be explained by differences in behaviour of the enzyme in the two different encapsulation environments. With the enzyme incorporated into the ceramic-alginate mixture prior to gel formation, the release profile suggested that enzyme diffusion from the surface of the beads into solution initially predominated, leading to a rapid increase in the concentration of enzyme in solution. Once surface bound enzyme was depleted, the release profile showed a steady increase, indicating the migration of enzyme trapped within the pores of the spheres. The group surmise that as the overall release profile was not significantly different from that seen using a pure alginate matrix, the enzyme is not interacting significantly with the embedded ceramic particles. When the enzyme was incorporated into the ceramic powders prior to the preparation of the ceramic-alginate mixture however, the group found that the initial rapid release of enzyme was reduced and an overall slower release rate was observed. This suggests that the alginate does not offer any additional resistance to the migration of enzyme through the microsphere structure. Microspheres of alumina have also been produced using alginate as a morphological structure directing agent. In their work, Watanabe et al. describe how once incorporated into an alginate matrix, chemistry can be done to produce an inorganic alumina phase, whose growth is constrained by the morphology of the alginate material20. An alginate gel was prepared with the addition of urease, an enzyme which has been shown previously to induce the precipitation of metal
Hydrocolloids
79
ions from aqueous solutions. The group made a solution of ammonium alginate and urease which was added dropwise to a solution of ammonium sulfate and urea. The alginate droplets very quickly tuned opaque due to the cross-linking of the alginate molecules with Al3+ cations. Leaving the gel spheres immersed in the Al2(SO4)3/urea solution allowed urea to diffuse into the gel, whereupon the urease acted on it to hydrolyse the urea to ammonia. The increasingly basic conditions within the gel spheres allowed the Al3+ to precipitate out as basic aluminum sulfate. After five days of immersion, the spheres were recovered, dried and calcined to 1000 °C to remove the alginate and to convert the basic aluminum sulfate to alumina. The removal of the templating alginate phase resulted in porous spheres of 2 mm in diameter which XRD showed were crystalline alumina. The success of this method relies on the fact that alginate is particularly stable at high pH and can therefore allow chemistry to be performed within alginate gels without structural collapse or chemical reaction with the polymer network. The formation of alginate spheres occurs due to cross-linking on contact with a solution containing metal cations, is a convenient way to prepare spheroidal constrained reaction volumes. The size of the alginate gel droplets can only be controlled in a narrow range however, as these studies use a drop-wise addition of the metal solution to the alginate (or vice versa) through a simple syringe delivery system. One study took this concept a stage further and used a nebulizer to form a mist of droplets in order to broaden the range of possible particle sizes. In their novel system, Mann et al. used a nebulizer to create a reagent aerosol composed of approximately 1 µm diameter droplets, which could then be laid onto a reactive receiver solution in a completely enclosed system21. By nebulizing a solution of chitosan containing CaCl2, the microdroplets could be passed over a receiving trough which contained a solution of Na2HPO4 in alginate. Chitosan and alginate have been noted previously to form robust synergistic gels with each other on contact22, 23, a behaviour which has been used to form capsules in which solutions of drugs can be held for controlled release24-26. In this case, after a few hours of nebulization, a thin, flexible film formed on the surface of the alginate solution, which was robust enough to be able to be lifted from the trough and stored in distilled water (Figure 4.8). SEM studies revealed that initially, for the first 20 minutes of deposition, the surface of the alginate contained discrete microspheres of a composite calcium phosphate/organic matrix, between 1 µm and 2 µm in diameter. Further deposition resulted in an increase of the mineral phase, which eventually merged
80
Biotemplating
to form the thin film. After prolonged deposition, the group found that the films could be grown several millimeters thick. FT-IR spectra of the films revealed that they contained both polysaccharides and also calcium phosphate in the films. Absorption bands were seen at 553, 573, and 1030 cm-1 (PO34-), 1130 cm-1 (C-O-C), 1381 and 1405 cm-1 (COO-), 1600 cm-1 (NH3+), 2859 and 2902 cm-1 (C-H), and 3480 cm-1 (OH). The fine structure of the thin film, arising from the discrete microspheres of calcium phosphate was a testament to the strong affinity of alginate for calcium, being cross-linked rapidly when the chitosan containing CaCl2 made contact with the surface and affording a well defined reaction volume.
Figure 4.8 – SEM micrographs: (a) Underside (solution-facing) of HAp film showing randomly packed hollow hemispheres, scale bar = 10 µm; inset shows broken sphere with hollow interior, scale bar = 1 µm. (b) Top surface of HAp film showing disordered open network of hollow spheres, scale bar = 10 µm; inset shows high-resolution image of microsphere showing highly texture surface, scale bar = 1 µm. (c) Cross section of HAp film formed after 4 h showing sublayering after drying, scale bar = 100 µm. (d) Cross section of carbonated HAp film produced after 2 h showing a continuous inorganic matrix and well-dispersed voids, scale bar = 100 µm. Reprinted with permission from21. Copyright 2007 American Chemical Society.
Hydrocolloids
81
The viscous nature of the alginate receiving solution was also suggested to be beneficial in that it prevented large-scale migration of the chitosan/CaCl2 whilst cross-linking was underway. This synergistic behaviour of alginate was employed in the creation of needlelike crystals of hydroxyapatite. Teng et al. found that by progressively increasing the concentration of alginate in an alginate/gelatin mixture, the morphology of crystals of HAp grown therein could be controlled in a precise manner27. TEM images showed that as the alginate concentration increased, crystals developed from needle-like to long fibres, which at the highest concentrations tended to self-aggregate into clusters. The group surmise the change in morphology to be a result of the crosslinking of alginate to gelatin and the formation of the egg-box zones, allowing the preferred sequestration and growth of calcium phosphate phases throughout composite organic matrix. This could account for the preferred c-axis orientation of the crystals, as growth in the highly viscous alginate/gelatin matrix would be largely directed along the fibre long axis. The group also make note of the fact that in the egg-box model of alginate ion binding, the strong, thermoresistive gel formed when calcium ions chelate between four guluronate residues results in a cation in a tetrahedral environment. This leads to a structural match with the tetrahedral phosphate anion and as has been previously observed, whenever electrostatic, structural or stereo-chemical complimentarity is encountered, a preferred morphology and/or crystallography of the mineral phase is usually produced. Lastly the film forming properties of alginate were used to create microcapsules of calcium alginate which were reinforced by their encapsulation within an alginate/aminopropyl-silicate/alginate membrane. Sakai et al. demonstrated that the microscopic structure of the membrane was determined in part by the microscopic structure of the core alginate material28. This has important implications for the release of entrapped drug molecules; the most common application for these microbeads. Molecular permeability through the outer membrane was related directly to the molecular weight of the core alginate.
4.4 Gelatin Although considered one of the hydrocolloids, gelatin is one of the few that is not a polysaccharide. It is a mixture of polypeptides derived from the breakdown
82
Biotemplating
of animal tissues such as skin and bone collagen. Commonly found as an emulsifying agent in foods, it also finds use in the pharmaceutical29 and cosmetic30 industries and in the photographic industry31. At a molecular level, collagen is quite varied, with a mixture of single and multi-stranded polypeptides, of between 200–4000 polypeptide fragments. These polypeptides behave in a similar manner to the hydrocolloid polysaccharides in that they form complex fibres by the aggregation of helices. Gelatin solutions undergo coil-helix transitions followed by aggregation to form right-handed triple-helical proline/hydroxyproline rich junction zones32, 33. With a triple helical conformation, the gelatin gels are much more resistant to swelling by water and therefore be much stronger. Cross-linking of the gels can be achieved by chemically cross-linking the peptide residues using molecules commonly encountered in biological research, for example gluteraldehyde which will bind lysine fragments together. Perhaps one of the earliest uses of a biotemplate to control the morphology of an inorganic material could be considered to be the use of gelatin as a flexible support for silver halide particles employed by George Eastman in the development of the Kodak Process for manufacturing photo sensitive material. Up to that point, photo-reactive silver halide grains were dispersed on a paper ground. On exposure to light, photographs were able to be taken, however the fibrous grain of the paper was reproduced in the final image. By using gelatin instead of traditional paper as the support for silver halide grains, Eastman found that there was no granularity in the final photograph. Although Eastman did not actively investigate the gelatin support for its control of silver halide particulate growth, it is likely that from what we have seen so far in this chapter, particulate growth was indeed controlled by the gelatin matrix. Deng et al. did in fact demonstrate the efficacy of silver halide in gelatin by creating sheets of metallic nanostructures34. By exposing a silver halide impregnated gelatin sheet to an image of a printed circuit, the resulting pattern of grains (not yet electrically continuous) could be made conducting by the subsequent electroless deposition of silver. The group found that the structures resulting from this work could be used in situ on the gelatin sheet, or carefully freed to find use as passive, structural materials such as wire frames or meshes, or to be used in microfluidic, microanalytical, and microelectromechanical systems. The control over particle growth in gelatin was utilized by Zhou et al. who grew nanoparticulate ZnO in a gelatin matrix35. ZnO is of particular interest as it possesses fast-switching optical and electrical behaviour as it is a semiconductor
Hydrocolloids
83
with a band gap of 3.37 ev36. This band gap makes it particularly efficacious as a light emitting material in the ultra-violet region. Gelatin was dissolved in a hot solution of ZnSO4 and cooled to form a gel. Pieces of the gel were then left to soak in a solution of Na2CO3 for 24 hours, which induced the precipitation of zinc carbonate phases within the gel structure. After washing to remove any residual Na2CO3, the gel was dissolved in hot water to leave a precipitate which was dried and calcined at 950 °C for 2 hours. XRD revealed that the resulting material was phase pure hexagonal ZnO. Initially forming at a temperature of 250 °C, as the calcination temperature increased, so the peaks due to ZnO became sharper and more intense, signifying that at higher temperatures the ZnO was more crystalline with a larger particle size. The group found that whilst control samples of ZnO gave spherical or elliptical particles ranging from 20 nm to 30 nm in diameter, those prepared in the presence of gelatin ranged from 26.5 nm to 20.5 nm with the increase of gelatin concentration from 5% to 17%. The effective limiting of particulate growth by gelatin was demonstrated by Chang et al. in their work on the growth of hydroxyapatite within this biopolymer37. By precipitating HAp within the gelatin, the group found that the number of nucleation events of HAp increased, leading to an overall lower particle diameter. By heating the gel to above 50 °C, the mobility of the gelatin molecules was such that constrained crystallization was no longer possible; the free migration of calcium and phosphate ions leading to larger particle growth. An additional level of complexity was realized using gelatin as a porous support for the growth of HAp materials. In a prelude to the larger bioconstructs that we will encounter in later chapters, Kang et al. used a freeze-thaw method on gelatin gels to produce a rigid, porous structure38. This first necessitated the cross-linking of the gelatin by gluderaldehyde in order to produce a rigid structure which would survive the water removal procedure. Once cross-linked, the gels were frozen and treated to a freeze-drying procedure in order to sublimate the water ice. By varying the freezing parameters, the group found that they were able to affect the pore size and structure. Gels frozen in liquid nitrogen were observed to have a 2D ordered pore structure, whilst those frozen at -20 °C had a more 3D ordered structure. These scaffolds had a range of pore sizes from 45 µm to 200 µm, depending on the temperature at which freezing occurred. This has important ramifications for the eventual application of these porous materials. For example, in a tissue replacement technology, pore sizes greater than 100 µm are required in order to allow the vascularization of the material.
84
Biotemplating
This will enable nutrients for cells colonizing the implant to flow freely throughout the structure, thereby decreasing the time taken for the organism to regenerate the area under repair. Another application for freeze-dried gelatin foams was demonstrated by Nabeta et al. who demonstrated the effect of producing a foam structure on the orientation of embedded carbon nanotubes39. By simply performing the freezedrying steps on a gelatin foam in which carbon nanotubes had been added, the restricted movement afforded by the gelatin prevented the nanotubes from making contact and thereby aggregating. In this study, a final step was to calcine the composite foam structure, so that the gelatin was removed, leaving behind a foam replica entirely consisting of carbon nanotubes. This structure was found to be mechanically stable and had the added bonus of being electrically conductive. Foams of the composite material prior to calcination (gelatin with carbon nanotubes dispersed throughout) were insulating, even at the highest nanotube concentration investigated (5.0 mg/mL). This suggests that the carbon nanotubes are not only well dispersed throughout the foam, but are being kept isolated from each other by the enveloping effect of the gelatin matrix. On calcination, the material shrunk considerably to around 20% of its starting volume, albeit retaining porosity. Electrical resistivity of the calcined foams was found to have dropped to 103 Ωcm. The group ascribe the resistivity to the fact that as the material is essentially a thin film of carbon nanotubes with semiconducting properties, the inherent resistivity will tend to stay high. They determined that higher calcination temperatures would produce a less resistive material, although at the expense of the structural integrity of the foam structure. For use in electronic devices, the group found that the material could be impregnated by a simple low density polyethylene, which would infill the pore structure and increase structural stability. The pore forming properties of gelatin was used in a slightly different way by Lin et al. in the synthesis of mesoporous silica40. Mesoporous silicas have long been a hot topic of research, owing to the fact that they can be synthesized simply using surfactants to produce materials with very high surface areas (> 1000 m2/g). Once the surfactant has been extracted, the mesoporous silica can be used as a catalytic support, molecular sieve or as the static phase in separation technologies. The problem with using surfactants however, is that they can do a great deal of environmental damage as the hydrophobic parts of any surfactant will only be degraded very slowly, which leads to prolonged environmental persistence. Lin et al. investigated gelatin therefore as an alternative to the
Hydrocolloids
85
amphiphillic surfactants. Being readily biodegradable, the synthesis and extraction steps would lead to no environmentally persistent materials. They achieved the templating of silica by the hydrolysis and condensation of an acidic solution of sodium silicate in a solution of gelatin. After adjusting the pH to between 4.0 and 6.0, a mild ageing of the solution at 40 °C, the solution was processed hydrothermally in an autoclave at 100 °C for 24 hours. Once the solid product was dried, calcination at 550 °C removed the gelatin template to leave mesoporous silica. Control experiments performed in the absence of gelatin merely produced a slow-condensation of the silica source, which after 48 hours was still gel-like with no precipitation. TEM investigations of the porous silica showed that unlike surfactant templated mesoporous silica, there was no ordered pore structure in the material, which was to be expected, as gelatin does not exhibit a regular structure on the mesoscale. An estimation of the pore sizes from TEM was that pores were on average a few nanometres in diameter. This observation was supported by N2 adsorption experiments which showed pore sizes between 8 nm and 16 nm. Calculation of the surface area of the material showed that depending on the pH of the silicate solution, the surface area was between 280 m2g-1 and 385 m2g-1. These values are in the region expected for porous silica in which porosity has been introduced randomly. It is clear that a compromise has to be made between the surface area and the method of introducing high surface area; traditional surfactants will produce surface areas in the region of 1000 m2g-1 due to their ability to adopt a complex, more efficiently packed form. As well as control over ZnO and HAp particle growth seen previously, other more ‘exotic’ inorganic materials have been synthesized in the presence of a gelatin matrix. For example, phosphorescent Y3Al5O12 (YAG) materials have been created by Zhou et al.41. By doping YAG with terbium, the material will not lose luminescence as temperature changes, important in applications where heat is likely to be generated (lasers, flat-panel displays). As ever, control over the homogeneity of synthesis is of critical importance. Zhou et al. note that this system is particularly suited to biotemplating, as the traditional heat and beat synthetic approach to YAG materials calls for heating to over 1600 °C and extensive ball-milling. The synthesis of Tb-YAG nanoparticles was simply achieved by the group by dissolving stoichiometric amounts of aluminum, yttrium and terbium nitrates in a solution of gelatin. Once gelled, the precipitation of hydroxide species within the gelatin network was achieved by soaking the gel in an ammonium hydroxide
86
Biotemplating
solution. After drying the gels, the Tb-YAG was formed by a calcination step which involved heating of the gels at temperatures up to 1000 °C for three hours. The Tb-YAG phosphor particles produced were between 40 nm and 53 nm in diameter and spherical or elliptical in morphology. This morphology was of importance as this would allow a greater packing efficiency of the phosphor particles, thereby further augmenting the resolution benefits garnered from the particles being at the nano-scale. In terms of performance, the group measured the particles by luminescence spectroscopy and found that the photoluminescent intensity of the particles decreased with increasing grain size. The overall brightness of nanoparticles synthesized at 800 °C was found to be only around 10% less than a commercially available equivalent material. By synthesizing the material under the influence of an anti-sintering biopolymer, the improved crystallinity and homogeneous particle size of the Tb-YAG allows for a much improved (and cheaper) synthesis of these technologically important materials. Hollow structures have also been synthesized using gelatin, this time as a sacrificial template. Miao et al. used gelatin as a route to cadmium chalcogenide (S, Se, Te) hollow spheres by the addition of Cd(OH)Cl to NaHTe or Na2SeSO3 solutions in an inert atmosphere42. The group found that the synthesis conditions employed allowed a shell of cadmium chalcogenide to form around the gelatin sphere, which itself was gradually degraded away, thereby forming hollow shell structures of the order of 5 nm in diameter. The electrochemiluminescence behaviour of the CdS hollow nanotubes was measured and it was determined that the light emission showed good stability and an enhanced intensity. The group attributed this to the aligned crystallites in the aggregation morphology of the CdS nanotubes. A similar protocol was adopted by Xu et al. in the synthesis of hollow Cu2O spheres43. 4.5 Agar, curdlan, gellan, pectin and the gums The remaining hydrocolloids have all been used as biotemplating agents, albeit to a much lesser degree than the three discussed so far. This may be due to the fact that they are (a) less commonly available than carrageenan, alginate or gelatin and (b) less complex molecularly. When they have found use, it is usually as a method of creating synergistic gels with other hydrocolloids, in order that properties such as viscosity and degree of gelation can be minutely controlled. Agar, is ubiquitous in biological research as a culture medium for the growth of bacteria and fungi. It also finds use as a thickener in foodstuffs where the use
Hydrocolloids
87
of gelatin is not possible (i.e. a vegetarian alternative) and a clarifying agent in the brewing industry. It is a polysaccharide consisting entirely of galactose subunits (agarose and agaropectin) and as such will formed branched structures, albeit not to a great extent. Agar is very similar to carrageenan in two main respects. Firstly, agar has L-3,6-anhydro-α-galactopyranose subunits, similar to the D-3,6-anhydro-α-galactopyranose of carrageenan. Secondly, chemically agar differs from carrageenan only in the absence of sulfate groups. As a biotemplate, agar has not found as wide a use as the other hydrocolloids, possibly due to the slightly less complex structural behavior of agar when compared to e.g. carrageenan. In a gel form, agar does exhibit a slight degree of structural complexity, in the formation of agarose double helices which are stabilized by cross-linking water molecules. These double helices can hydrogen bond with each other to form large suprafibres44. The ability of the hydrocolloids to form synergistic gels has been exploited in one of the few studies of agar as a biotemplate. In this work, Jiu et al. investigated the co-operative interaction of agar with a surfactant, in order to greatly enhance the properties of the final product45. The group found that by performing the synthesis of a silica gel in the presence of a mixed surfactant/agar templating phase, a complex, mesoporous silica was able to be produced. The creation of a hybrid poly(ethyleneoxide)106–poly(propyleneoxide)70– poly(ethyleneoxide)106 (pluronic F127) surfactant with agar produced a synergistic effect which was key in the formation of the complex silica structure. The simple synthesis involved an acidified tetraethoxysilane (TEOS) solution being added to an aqueous solution of F127 with agar. This mixture was then aged at 90 °C, followed by removal of the template by calcination at 600 °C for 6 hours. TGA studies of the system showed that for each component of the organic templating phase, the maximum weight loss occurred at 243 and 303 °C for F127 and agar respectively. The corresponding temperature for the mixed organic phase was found to be 297 °C, which suggested to the group that an interaction between the two organic components was indeed taking place. They suggest that it is the hydroxyl groups of the agar which are hydrogen bonding with the hydrophilic blocks of the F127 molecules. XRD patterns of the post-calcined material showed one intense peak and broad reflections between 1° and 5° 2θ, which were able to be indexed to the (100) and (110) reflections seen in mesoporous silicas. The crystallographic d-spacing of the (100) peak suggests a pore size of around 7 nm, consistent with that observed previously in mesoporous silicas synthesized using pluronic block co-polymer surfactants. TEM studies on
88
Biotemplating
the silica showed a highly ordered pore structure with areas of hexagonal symmetry. The group note that when prepared with either the F127 or agar alone as templating agent, no ordered pore structures could be obtained. They attribute this to the observation that micelles of the F127 are able to form at lower concentrations in the presence of the agar, leading to the creation of an ordered liquid crystalline structure of F127/agar. SEM images of the mesoporous silica before and after calcination showed that there was no significant loss of macroscopic structure, even after template removal at 600 °C for 6 hours. Structural collapse is a significant hurdle to overcome in the synthesis of mesoporous silicas; the preservation of pore structure and shape is of critical importance for the incorporation of mesoporous silicas in applications. The group attribute the structural stability of the calcined materials to the stability of the F127/agar gel in the early stages of calcination, providing structural support whilst densification of the silica structure is occurring. Another hydrocolloid, curdlan has been used by Jiu et al. in an almost identical study to their work with agar/F127 composites46. The replacement of agar with curdlan resulted in the same synergistic creation of mesoporous silicas as shown in their work with agar. Numata et al. however, used a curdlan to create a silica material which was comprised of nano-fibres47. In this work they used the molecular structure of curdlan, β-1,3-glucan as a direct morphological template on which to grow fibres of silica, making use of the hydrophobic cavity of the one-dimensionally aligned hydrocolloid as a constrained reaction volume. The group were interested in this approach as they had previously used fibres of an organo-gel to produce fibrous silica, although had experienced difficulties in removing the organic template without subsequent loss of inorganic structural integrity. Calcination to remove the organic part of the system would invariably result in structural collapse as outgassing occurred. By synthesizing the silica inside the organic template, calcination should leave the inorganic part intact. The synthesis was carried out using Schizophyllan (SPG), which is a natural polysaccharide present in the fungus Schizophyllum commune. Molecularly, SPG can be considered to consist of three β-(1-3) glucoses and one β-(1-6) glucose side chain linked at every third main-chain glucose. Chains of SPG are therefore able to self-interact to produce the one-dimensional hydrophobic cavity desired for this study. A solution of SPG in dimethysulfoxide (DMSO) was mixed with TEOS and a large excess of water to give an H2O/DMSO ratio of 95/5. After ageing for a week at room temperature, the group dialyzed the system, which they found induced
Hydrocolloids
89
precipitation of the silica. TEM results were striking in that the SPG produced highly anisotropic fibres of silica, of around 15 nm in diameter. Systematically, the group invoke a simple precipitation within a constrained volume to explain the behaviour of silica in this system. Of the remaining hydrocolloids, even less has been performed by way of biotemplating. Gellan, pectin and xanthan gum all appear in the study by Butler et al., which is discussed above in the section on carrageenan10. Gum Arabic (commonly called acacia gum) has been used by Yin et al. not in a strictly biotemplating role, but as an effective method for ensuring the dispersity of carbon nanotubes synthesized using anodic aluminum oxide templates. By adsorbing acacia gum onto the surface of carbon nanotubes, well aligned and non-aggregated carbon nanotubes were able to be synthesized48.
4.6 References 1. 2. 3. 4. 5. 6. 7.
8. 9. 10.
11.
12.
Janaswamy, S. & Chandrasekaran, R. Cation-induced polymorphism in iota-carrageenan. Carbohydr. Polym. 60, 499-505 (2005). Michel, A.S., Mestdagh, M.M. & Axelos, M.A.V. Physico-chemical properties of carrageenan gels in presence of various cations. Int. J. Biol. Macromol. 21, 195-200 (1997). Jones, F., Colfen, H. & Antonietti, M. Iron oxyhydroxide colloids stabilized with polysaccharides. Colloid Polym. Sci. 278, 491-501 (2000). Daniel-da-Silva, A.L. et al. In situ synthesis of magnetite nanoparticies in carrageenan gels. Biomacromolecules 8, 2350-2357 (2007). Ahluwalia, S.S. & Goyal, D. Microbial and plant derived biomass for removal of heavy metals from wastewater. Bioresour. Technol. 98, 2243-2257 (2007). Lestan, D., Luo, C.L. & Li, X.D. The use of chelating agents in the remediation of metalcontaminated soils: A review. Environ. Pollut. 153, 3-13 (2008). Sud, D., Mahajan, G. & Kaur, M.P. Agricultural waste material as potential adsorbent for sequestering heavy metal ions from aqueous solutions - A review. Bioresour. Technol. 99, 6017-6027 (2008). Son, B.C., Park, K., Song, S.H. & Yoo, Y.J. Selective biosorption of mixed heavy metal ions using polysaccharides. Korean J. Chem. Eng. 21, 1168-1172 (2004). Lovley, D.R. & Coates, J.D. Bioremediation of metal contamination. Curr. Opin. Biotechnol. 8, 285-289 (1997). Butler, M.F., Glaser, N., Weaver, A.C., Kirkland, M. & Heppenstall-Butler, M. Calcium carbonate crystallization in the presence of biopolymers. Cryst. Growth Des. 6, 781-794 (2006). Didymus, J.M., Mann, S., Benton, W.J. & Collins, I.R. Interaction of Poly(alpha,betaaspartate) with Octadecylamine Monolayers-Adsorption Behaviour and Effects on CaCO3 Crystallization. Langmuir 11, 3130-3136 (1995). Pai, R.K., Hild, S., Ziegler, A. & Marti, O. Water-soluble terpolymer-mediated calcium carbonate crystal modification. Langmuir 20, 3123-3128 (2004).
90 13. 14. 15.
16. 17. 18. 19. 20.
21. 22.
23.
24.
25.
26.
27.
28.
29. 30.
31.
Biotemplating Shchipunov, Y.A. Sol-gel-derived biomaterials of silica and carrageenans. J. Colloid Interface Sci. 268, 68-76 (2003). Meyer, M., Fischer, A. & Hoffmann, H. Novel ringing silica gels that do not shrink. J. Phys. Chem. B 106, 1528-1533 (2002). Smith, E., Schnepp, Z., Wimbush, S.C. & Hall, S.R. On the suppression of superconducting phase formation in YBCO materials by templated synthesis in the presence of a sulfated biopolymer. Physica C: Superconductivity 468, 283-2287 (2008). Schnepp, Z.A.C., Wimbush, S.C., Mann, S. & Hall, S.R. Structural evolution of superconductor nanowires in biopolymer gels. Adv. Mater. 20, 1782-1786 (2008). Pavez, J., Silva, J.F. & Melo, F. Effects of alginic acid from marine algae on calcium carbonate electrodeposited coating. J. Cryst. Growth 282, 438-447 (2005). Ribeiro, C.C., Barrias, C.C. & Barbosa, M.A. Calcium phosphate-alginate microspheres as enzyme delivery matrices. Biomaterials 25, 4363-4373 (2004). Grabowski, G.A., Leslie, N. & Wenstrup, R. Enzyme therapy for Gaucher disease: the first 5 years. Blood Rev. 12, 115-133 (1998). Unuma, H., Hirose, Y., Ito, M. & Watanabe, K. Preparation of the precursor of porous alumina particles using immobilized urease in alginate gel templates. J. Ceram. Soc. Jpn. 112, 409-411 (2004). Walsh, D. et al. Aerosol-mediated fabrication of porous thin films using ultrasonic nebulization. Chem. Mat. 19, 503-508 (2007). Gaserod, O., Jolliffe, I.G., Hampson, F.C., Dettmar, P.W. & Skjak-Braek, G. The enhancement of the bioadhesive properties of calcium alginate gel beads by coating with chitosan. Int. J. Pharm. 175, 237-246 (1998). Murata, Y., Maeda, T., Miyamoto, E. & Kawashima, S. Preparation of Chitosan-Reinforced Alginate Gel Beads - Effects of Chitosan on Gel Matrix Erosion. Int. J. Pharm. 96, 139-145 (1993). Murata, Y., Kontani, Y., Ohmae, H. & Kawashima, S. Behavior of alginate gel beads containing chitosan salt prepared with water-soluble vitamins. Eur. J. Pharm. Biopharm. 53, 249-251 (2002). Murata, Y., Miyamoto, E. & Kawashima, S. Additive effect of chondroitin sulfate and chitosan on drug release from calcium-induced alginate gel beads. J. Control. Release 38, 101-108 (1996). Shu, X.Z. & Zhu, K.J. The release behavior of brilliant blue from calcium-alginate gel beads coated by chitosan: the preparation method effect. Eur. J. Pharm. Biopharm. 53, 193-201 (2002). Teng, S.H., Shi, J.J., Peng, B.X. & Chen, L.J. The effect of alginate addition on the structure and morphology of hydroxyapatite/gelatin nanocomposites. Compos. Sci. Technol. 66, 1532-1538 (2006). Sakai, S., Ono, T., Ijima, H. & Kawakami, K. Permeability of alginate/sol-gel synthesized aminopropyl-silicate/alginate membrane templated by calcium-alginate gel. J. Membr. Sci. 205, 183-189 (2002). Djagny, K.B., Wang, Z. & Xu, S.Y. Gelatin: A valuable protein for food and pharmaceutical industries: Review. Crit. Rev. Food Sci. Nutr. 41, 481-492 (2001). Li, G.Y., Fukunaga, S., Takenouchi, K. & Nakamura, F. Comparative study of the physiological properties of collagen, gelatin and collagen hydrolysate as cosmetic materials. International Journal of Cosmetic Science 27, 101-106 (2005). Jolley, J.E. Microstructure of Photographic Gelatin Binders. Photographic Science and Engineering 14, 169-177 (1970).
Hydrocolloids 32. 33. 34.
35. 36. 37.
38. 39. 40.
41. 42. 43. 44.
45. 46. 47. 48.
91
Bigi, A., Panzavolta, S. & Rubini, K. Relationship between triple-helix content and mechanical properties of gelatin films. Biomaterials 25, 5675-5680 (2004). Dolz, R. & Engel, J. in Fleischmajer, R., B. R. Olsen and K. Kuehn 421-424 (1990). Deng, T., Arias, F., Ismagilov, R.F., Kenis, P.J.A. & Whitesides, G.M. Fabrication of metallic microstructures using exposed, developed silver halide-based photographic film. Anal. Chem. 72, 645-651 (2000). Zhou, J.G., Zhao, F.Y., Wang, Y.L., Zhang, Y. & Yang, L. Size-controlled synthesis of ZnO nanoparticles and their photoluminescence properties. J. Lumines. 122, 195-197 (2007). Aoki, T. et al. p-type ZnO layer formation by excimer laser doping. Phys. Status Solidi BBasic Res. 229, 911-914 (2002). Chang, M.C., Douglas, W.H. & Tanaka, J. Organic-inorganic interaction and the growth mechanism of hydroxyapatite crystals in gelatin matrices between 37 and 80 degrees C. J. Mater. Sci.-Mater. Med. 17, 387-396 (2006). Kang, H.W., Tabata, Y. & Ikada, Y. Fabrication of porous gelatin scaffolds for tissue engineering. Biomaterials 20, 1339-1344 (1999). Nabeta, M. & Sano, M. Nanotube foam prepared by gelatin gel as a template. Langmuir 21, 1706-1708 (2005). Lin, Y.C., Hsu, C.H., Lin, H.P., Tang, C.Y. & Lin, C.Y. Preparation of mesoporous silica and carbon using gelatin or gelatin-phenol-formaldehyde polymer blend as template. Chem. Lett. 36, 1258-1259 (2007). Zhou, J.G. et al. Template synthesis and luminescent properties of nano-sized YAG : Tb phosphors. J. Lumines. 119, 237-241 (2006). Miao, J.J., Ren, T., Dong, L., Zhu, J.J. & Chen, H.Y. Double-template synthesis of CdS nanotubes with strong electrogenerated chemiluminescence. Small 1, 802-805 (2005). Xu, L.S. et al. Solution-phase synthesis of single-crystal hollow Cu2O spheres with nanoholes. Nanotechnology 17, 1501-1505 (2006). Rees, D.A. Conformational Analysis of Polysaccharides. 2. Alternating Copolymers of AgarCarrageenan-Chondroitin Type by Model Building in Computer with Calculation of Helical Parameters. Journal of the Chemical Society B-Physical Organic, 217-226 (1969). Jiu, J.T., Kurumada, K., Pei, L.H. & Tanigaki, M. Syntheses of ordered mesoporous silica by new hybrid template. Colloid Surf. B-Biointerfaces 38, 121-125 (2004). Jiu, J.T., Kurumada, K., Wang, F.M. & Pei, L.H. Syntheses of highly ordered mesoporous silica by new hybrid template. Mater. Chem. Phys. 86, 435-439 (2004). Numata, M. et al. beta-1,3-Glucan polysaccharide can act as a one-dimensional host to create novel silica nanofibre structures. Chem. Commun., 4655-4657 (2005). Yin, A.J., Chik, H. & Xu, J. Postgrowth processing of carbon nanotube arrays - Enabling new functionalities and applications. IEEE Trans. Nanotechnol. 3, 147-151 (2004).
Chapter 5
Chitin/Chitosan
With their complex behaviour, particularly in the hydrated state, polysaccharides are finding extensive use as effective templates to controlled crystal growth. One polysaccharide which deserves consideration separately has been more widely used than all of the others for biotemplating as it has a particular significance in marine biomineralization. Chitin and its derivatives are widely and cheaply available, chemically stable over a wide range of synthesis conditions and exhibit complex molecular and supra-molecular form. In nature, the primary use of chitin is as a structural component. Chitin occurs widely across the animal kingdom, being found in the cuticles of arthropods, the shells of brachiopods and molluscs, in the cuttlebone of cuttlefish and the pen of squid. The exoskeletons of insects are largely composed of chitin as are fungal cell walls. It has been estimated that around ten gigatons (1013 kg) of chitin are synthesized in the biosphere each year. By examining the complex behaviour of chitin and chitosan and given the knowledge of constrained reaction volumes from earlier chapters, clues as to the efficacy of these polysaccharides in biotemplating are readily seen.
5.1 Structure and properties Chitin is a polysaccharide composed of N-acetylglucosamine sub-units. These fragments are β-1,4 linked, which makes it structurally similar to cellulose (Figure 5.1). It is obvious that as the two biomolecules are structurally similar, they are likely to find use in similar roles. This indeed is the case, as both are found predominately as structural components of plant and animal cells. The main difference between the two is that chitin has an acetylamine group on each monomer, where cellulose has an hydroxyl group. This means that inter-chain
92
Chitin/Chitosan
93
hydrogen bonding is stronger in chitin, leading to an increase in strength and structural stability. Owing to its large affinity for meal cations, chitin finds extensive industrial use as a water purification material or ion-exchange resin1, 2. The chitin molecule is chemically resistant to attack, the acetylamino group providing resistivity to solvents and acids. Despite this, chitin is able to be dissolved, although a powerful base such as triethylamine is required. Modification of the acetylamino group by deacetylation (i.e. to leave only an amino residue) results in a molecule which is still structurally complex, albeit less chemically stable than chitin (Figure 5.2). As this deacetylated form of chitin (chitosan) is much easier to work with, a plethora of biotemplating studies have been performed with it.
Figure 5.1 – Molecular structures of (a) chitin and (b) cellulose. Image courtesy of Caroline Hall.
5.2 Chitin/chitosan As it is relatively inert to many chemical processes, particularly the sol-gel chemistry favoured in biotemplating studies, chitin itself has been relatively poorly used. It comes into its own in biotemplating however, when part of a
94
Biotemplating
complex biological construct such as cuttlebone, which will be covered in a later chapter. Studies involving mineralization and chitin tend to use the chitin as more as a bio-substrate from which to grow crystals, rather than the chitin itself acting actively to control morphology. One study on the growth of calcite upon chitin membranes was performed by Manoli et al. In this work, they examined the interaction between saturated calcium carbonate solutions and a substrate of chitin3. Crystals of calcite were grown with no apparent change in morphology, the usual rhombohedral form of calcite being predominant. Whilst not strictly biotemplating, the group note the influence of the polarity of the amino group in the initiation of crystallization. Falini et al. investigated the nucleation and growth of CaCO3 on a composite organic layer of β-chitin and silk fibroin4. With the addition of acidic macromolecules extracted from molluscs, this organic assemblage provided a convincing mimic for the mineralization conditions in the sea shell. The conditions were such that the in vitro study produced the same control over calcium carbonate polymorph expression as in the mollusc; acidic macromolecules extracted from calcitic or aragonitic shells produced the same polymorph on the chitin/silk fibroin layer.
Figure 5.2 – Molecular representations of (a) chitin and (b) chitosan. Image courtesy of Caroline Hall.
Chitin/Chitosan
95
Fibres of chitin have been coated with layers of CaCO3 in the presence of a range of polymers in order to mimic the polymorphism seen in nature, without the need for naturally produced macromolecules. Kato et al. looked at the effect of poly(acrylic acid) (PAA), poly(L-aspartate) (poly-ASP), and poly(Lglutamate) (PLG) macromolecules on the growth of calcium carbonate on chitin5. They found that these polymers induce a polymorphic change from calcite to vaterite, with the latter polymorph dominating over the former with an increase in the concentration of the polymer in the system. Polymorphism was also observed by Falini et al. in a study into the precipitation and growth of calcium carbonate crystals within a chitinous squid pen, depending on the supersaturation of the mineralizing solution trapped within the pore structure of the matrix6. Chitin has been used as an analogue to collagen in the formation of calcium phosphate phases, with a view to using the composite material as a bone replacement material. Wan Andrew et al. used solutions of chitin which had been functionalized with carboxymethyl or phosphoryl groups to inhibit the rate of formation of hydroxyapatite from solution7. To solutions of dicarboxymethyl chitosan, the group added calcium acetate and disodium hydrogen phosphate, to produce precipitates of an amorphous calcium phosphate. With low crystallinity afforded by the dicarboxymethyl chitosan inhibition, the amorphous material was more readily able to be shaped to treat bone lesions and in dentistry. As previously noted, removal of the acetyl group from chitin gives chitosan. Deacetylation of chitin is never 100% and so chitosan can therefore be thought to be comprised of a mixture of β-[1→4]-linked N-acetyl-D-glucosamine and Dglucosamine units. Chitosan is readily soluble in weak acidic solutions, giving a viscous sol to which inorganic precursor materials can be readily incorporated, making it an ideal biotemplating material. Yamaguchi et al. performed a series of studies on chitosan, which demonstrated how resilient this material is in terms of mechanical strength8. Preparation of the chitosan involved boiling crab tendons at 100 °C in a concentrated NaOH solution, in order to remove any proteinaceous material and calcium salts, followed by treatment in ethanol. On observing the resultant chitosan through a polarizing optical microscope, the group found that there was no change to the optical behaviour, implying the retention of the molecular structure of the chitosan. Additionally, the group determined the tensile strength of the chitosan to be quite high at around 68 MPa. On further heating of the as-prepared chitosan to 120 °C, the group measured the tensile strength of the biopolymer to be around 235 MPa, a fact that they attribute to the
96
Biotemplating
formation of intermolecular hydrogen bonds as the complex molecular structure densifies. Yi et al. highlight the most interesting feature of chitosan, namely the presence of a primary amine at the C-2 position of the glucosamine residues throughout the molecule9. This unusually high concentration of primary amines for a biomolecule mean that the properties of the chitosan can be finely controlled using pH (Figure 5.3).
Figure 5.3 – Schematic illustrating chitosan’s versatility for fabrication. At low pH (less than about 6), chitosan’s amines are protonated conferring polycationic behavior to chitosan. At higher pH (above about 6.5), chitosan’s amines are deprontonated and reactive. Also at higher pH, chitosan can undergo interpolymer associations that can lead to fibre and network (i.e., film and gel) formation. Reprinted with permission from9. Copyright 2005 American Chemical Society.
From the figure, it can be seen that a change of pH will radically alter the behaviour of the chitosan. The group note that at a low pH, the amine groups will be protonated and carry a net positive charge, allowing the chitosan to be readily solubilized in water. At higher pH values, the amine groups will become deprotonated, thereby causing the chitosan to lose charge and become insoluble.
Chitin/Chitosan
97
As the pKa of chitosan is near neutrality and the transition from soluble to insoluble takes place between pH 6 and 6.5, applications in biology will be particularly favourable. It is this biocompatibility that has made chitosan particularly intensively investigated in the mineralization of hydroxyapatite for biomedical applications10-13. Chitosan is particularly suited to the controlled growth of HAp, owing to the chelating ability of the hydroxyl group on the chitosan molecule to provide sites of preferential nucleation of Ca2+ cations. Charge stabilization of the composite structure is achieved by the positively charged amino groups associating with anionic species in the synthesis. Ehrlich et al. used diffusion across a membrane of chitosan to induce HAp formation within the biopolymer14. As in the majority of chitosan/HAp studies, precipitation of HAp was from a solution of CaCl2 and NaH2PO4, this time on either side of a chitosan membrane. The group found that mineralization of the chitosan readily occurred, forming crystals which morphologically resembled HAp. Confirmation that HAp was formed was given by FTIR, which showed the characteristic bands at 633 cm-1 due to OH libration, at 1026 cm-1, 1106 cm-1 and 961 cm-1 (phosphate stretching) and 604 cm-1 and 563 cm-1 (phosphate bending). In mineralizing chitosan with HAp, researchers have been concerned to show that the composite system is biocompatible and also osteocompatible, allowing and even encouraging the growth and differentiation of pre-bone cells on and within the matrix. As these materials have been conceptualized to be tissue replacement materials, the viability of cells in association with the materials is key. Chesnutt et al. examined the effect a chitosan/HAp composite film had on the survivability of osteoblasts15. When they immersed chitosan/HAp composite films in cell media, they found that additional calcium phosphate precipitated, indicating that bone reformation would be possible in this material. After culturing the composite film for seven days with osteoblasts, the group found that there were almost three times as many cells present than in control experiments without the film. This indicates that not only are osteoblast cells viable on chitosan/HAp films, but that the film encourages the active differentiation and growth of cells. With the knowledge of the viability of cell lines on chitosan/HAp, other studies have concentrated on improving the mechanical properties of the composite materials for implanting. Wang et al. looked at the effect of complexing the chitosan with silk fibroin prior to incorporation of HAp, in an effort to improve the crystallinity of the composite material16. Silk fibroin was
98
Biotemplating
used as it is a simple protein (17 amino acid residues) which is readily available from silk moth cocoons and has the ability to cross-link with chitosan, forming 3D porous networks. In this way, the crystallization volume was altered, with the intention of controlling the morphology of HAp. The synthesis of chitosan/silk fibroin/HAp materials was achieved by adding an acidic solution of chitosan containing H3PO4 to a suspension of silk fibroin in Ca(OH)2. The pH of the synthesis mixture initially 6.2 was then adjusted to 9.0 by the addition of ammonium hydroxide, in order to ensure the HAp phase of calcium phosphate was formed preferentially. This was confirmed by both XRD and FR-IR which showed a monophasic HAp material was produced. TEM images of the crystals showed that for both control and chitosan/silk fibroin/HAp materials, there was no change in morphology, both sets having the characteristic HAp needle-like morphology. The group found that the crystals were between 20 nm and 50 nm in length and 10 nm in width. Although no morphological change was observed, there appeared to be a degree of aggregation of the needles into discrete bundles. The formation of a copolymer to alter crystal morphology has been employed in other biotemplating work with chitosan. Tsai et al. used polyacrylic acid (PAA) with chitosan in order to form a spherical nanoparticulate assemblage with a net positive surface charge17. Complexation of the PAA to chitosan was achieved by dissolving chitosan in a lactate solution to which was added a small amount of a 60 wt% PAA solution with vigorous shaking. After ageing for 24 hours, the spherical chitosan/PAA nanoparticles were used to form the core of a silica nanoparticle, by the simple addition of colloidal silica to the synthesis. Colloidal silica has a net negative charge and so electrostatic interaction between the silica and the negatively charged organic core resulted in the decoration of the chitosan/PAA by the colloidal silica (Figure 5.4). TEM studies of the product revealed that the chitosan/PAA spheres had a diameter of 200 nm, which was found to have increased to 220 nm after the addition of colloidal silica, suggesting to the group that the thickness of the silica ‘shell’ was around 10 nm. Dissolution of the chitosan/PAA core with HCl produced hollow silica shells and induced a slight dissolution of the silica, which subsequently condensed back onto the surface, producing a smoother siliceous material. The group measured the pore size and surface area of the silica shells and found that the material possessed 5.4 nm pores giving a high surface area of around 350 m2g-1.
Chitin/Chitosan
99
Figure 5.4 – Model of hollow silica sphere formation. Reprinted from17. Copyright 2006, with permission from Elsevier.
Chitosan/PAA has also been used to encourage the growth of calcium phosphate phases. In the work by Beppu et al. films of chitosan/PAA were immersed in simulated body fluid (SBF) for a period of seven days at a pH of 7.4 to 7.818. They found that for chitosan films with PAA as co-polymer, mineralization was greater, when compared to films with no PAA present. The group attribute this to the ability of PAA to strongly bind to calcium ions through the carboxyl groups on the molecule. The phase of calcium phosphate produced was largely indeterminate, with X-ray diffraction indicating the presence of an amorphous phase. EDXA analysis of the films seemed to support this, as for chitosan/PAA films, the Ca/P ratio was 2.46, close to the ratio in the depositing SBF solution. For chitosan films alone however, the group note that the possibility of small crystallites of HAp cannot be ruled out, particularly as the Ca/P ratio on these films was between 1.5–1.7, which is of the order found in HAp grown on organic substrates. A chitosan/gelatin blend was found to be a suitable medium for the growth of HAp however. Li et al. found that small variations in the gel concentration and temperature determined the crystallite size of HAp19. Conceptually, the use of gelatin is similar to that of PAA used by Beppu et al. in that both co-polymers contain reactive carboxyl groups which act as sites of preferred nucleation of calcium cations by electrostatic interaction. In Li’s study, they found nanocrystallites between 17 nm and 19 nm when the synthesis was carried out below 50 °C and of the order of 52 nm when carried out at 70 °C (Figure 5.5). The use of a chitosan/co-polymer blend in the formation of hollow nanostructures was explored more fully in work by Cheng et al. In this work, they were able to successfully synthesize hollow nanostructures of chitosan/polypyrrole (PPy) using crystals of AgBr20. AgBr is well known to form
100
Biotemplating
a range of crystal morphologies depending on the processing conditions21. Cheng et al. used different AgBr morphologies on which to bind chitosan/PPy, either spheres, cubes or hexagons, all at the nanoscale.
Figure 5.5 – TEM micrographs of nHA crystalline formatted on the surface of CG-3 network films in different calcium concentration tris buffer solution at 35 °C, (a) nHCG-3(0.1, 20); (b) nHCG3(0.1, 20); (c) nHCG-4(0.1, 20); (d) nHCG-5(0.1, 20). Reprinted from19. Copyright 2007, with permission from Elsevier.
The synthesis of AgBr in different morphologies has long been a subject of study, particularly as we have seen that AgBr has found extensive use in the photographic industry. Hexagonal nanoplates and nanocubes of AgBr were grown and then subsequently passivated by the biopolymer. After an ageing step, the group were able to dissolve the AgBr core by the addition of ammonium thiocyanate (NH4SCN) which solubilized the silver as [Ag(SCN)2]- ions. These were removed from the system by dialysis, which was then subjected to treatment with acetone in order to precipitate the hollow nanoparticles of chitosan/PPy. TEM images of the nanostructures revealed that in the case of both pre- and post-removal of the AgBr core, the diameter of the particle was
Chitin/Chitosan
101
approximately unchanged. This suggests that the biopolymer shell is unaffected by core removal, thereby allowing for a precise tailoring of the dimensions of the eventual organic nanostructure from the start of the synthesis. The group observed that nanospheres of chitosan/PPy formed around colloidal AgBr prior to morphologically controlled growth were of the order of 70 nm in diameter. After modification/growth of the AgBr into nanoplates and nanocubes, the size of the composite particles were 450 nm and 140 nm in width respectively. For all three morphologies however, the thickness of the hollow nanoparticle was between 15 nm and 20 nm. Chitosan and alginate have a marked affinity for each other; co-polymer blends of these two form synergistic gels, which have enhanced structural stability and chemical resistivity. Green et al. used the synergistic gel-forming affinity of chitosan for alginate to prepare microcapsules, which were mineralized and subsequently used for the incorporation of a variety of materials such as viable cells and cell growth factors22. The mineralized microcapsules were prepared by the addition of droplets of sodium alginate which contained Na2HPO4, to an aqueous solution of chitosan which contained CaCl2 and acetic acid. In this way, both formation of microspheres by the complexation of chitosan and alginate and mineralization by calcium phosphate occurred concurrently. The group could control the degree of mineralization of the capsules by varying the amount to calcium phosphate precursors present in each polysaccharide solution. In doing so, the release of cells and/or growth factors by the degradation of the calcium phosphate shell could be precisely controlled. The range of cells and growth factors successfully incorporated in this composite system was broad, but ultimately concerned with bone-regeneration materials. Successful incorporation of promyoblasts, chondrocytes, adipocytes, adenovirally transduced osteoprogenitors, immunoselected mesenchymal stem cells, and the osteogenic factor, rhBMP-2 were all demonstrated. Tantalizingly, the group demonstrated that it was possible to incorporate ‘beads within beads’ of capsules, which contained different cell populations. The spatial separation of cell-types with different functions is an important first step on the road to the creation of artificial analogues of complex living cells. Chitosan poly(acrylic acid) (PAA) has been employed by Ding et al. in order to create magnetic hollow spheres of Fe3O423. In this work, nanoparticles of Fe3O4 stabilized with poly(vinyl alcohol) (PVA) were pre-made and dispersed in a solution of chitosan/acrylic acid. The group found that PVA and PAA
102
Biotemplating
interacted forming a strong composite gel through intermolecular hydrogen bonding, which encouraged the incorporation of the PVA stabilized magnetite nanoparticles into the chitosan/PAA hollow spheres. Once incorporated, the acrylic acid was polymerized by the addition of potassium persulfate (K2S2O8) and the whole was allowed to react for two hours at 353 K under an inert atmosphere. Subsequently, gluteraldehyde was added in order to cross-link the chitosan. The group measured the magnetic properties of the composite material using vibrating sample magnetometry (VSM). The data show that there was no detectable hysteresis and coercivity in the material, suggesting that the magnetite nanoparticles are superparamagnetic at room temperature (Figure 5.6)
Figure 5.6 – Magnetization versus applied magnetic field: (a) PVA stabilized Fe3O4 nanoparticles at room temperature and (b and c) hybrid hollow spheres at room temperature and 78 K, respectively. Reprinted with permission from23. Copyright 2006 American Chemical Society.
Lowering the temperature of the composite to 78 K produced a hysteresis in the material with a coercivity of 400 Oe, indicating that the magnetite was behaving as a ferromagnetic material at that temperature.
Chitin/Chitosan
103
In addition, the group found that the saturation magnetizations of PVAstabilized magnetite nanoparticles and hybrid CS-PAA hollow spheres at room temperature were 62.3 emu/g and 40.7 emu/g, respectively, values which are lower than that of bulk magnetite at 89 emu/g. This loss of magnetization in the biotemplated material was attributed to both the small particle size of the magnetite (as a noncollinear spin arrangement occurs in small particles near the surface) and also to the presence of organic material, which will reduce overall total magnetization of the sample. By functionalizing the chitosan, the chemical and physical behaviour of the molecule can be altered. One of the most interesting instances of this was by the addition of lauric acid to the chitosan in acidic solution. This was first demonstrated by Wong et al24. In this work, chitosan was functionalized by mixing an acidic solution of chitosan in a high speed blender with a range of fatty acids. For longer chain fatty acids, the mixture was also heated above the melting point of the lipid in order to facilitate complexation to the chitosan. The incorporation of a fatty acid moiety in this way provided a degree of hydrophobicity to the polymer and thereby improved the resistance to water, both in terms of hydration ability and also the transmission of water through cast thin films of the chitosan. They demonstrated that in the case of lauric acid, the moiety was incorporated homogeneously throughout the chitosan, giving rise to stacks of sheet-like layers of the biopolymer. The group attribute the enhanced resistivity to water migration through the biopolymer to this layered structure, as a water molecule diffusing through the composite material will have to pass through channels between the chitosan layers lined with hydrophobic molecules. The resulting functionalized chitosan laurate will consist of a hydrophilic part (chitosan) and a hydrophobic part (laurate) and as such, can be expected to act as a surfactant, forming micelles in solution. Khiew et al. investigated the biotemplating properties of a chitosan laurate solution, utilizing the micelle forming ability to grow ZnS particles25. This work consisted of the simple addition of a micellar solution of chitosan laurate containing 0.1 M Na2S to a similar micellar solution containing 0.1 M Zn(NO3)2. After stirring for 2 hours, the precipitate was collected by centrifugation. Repeated washing with water and ethanol removed any residual reactants and chitosan laurate surfactant. XRD studies on the precipitate indicated to the group that the synthesis produced a phase-pure material, with broad peaks at 28.58, 47.78 and 56.68 2-theta, corresponding to the (111), (220) and (311) Miller
104
Biotemplating
indices of the fcc-phase of ZnS. The broadening of the peaks was indicative of the small crystallite size, which the group estimated using the Debye-Scherrer equation at full width half maximum for each peak (Figure 5.7).
Figure 5.7 – The powder XRD patterns of ZnS nanoparticles prepared in micellar solution containing (a) 0.5 wt.%, (b) 1.0 wt.%, (c) 2.0 wt.% of chitosan laurate. Reprinted from25. Copyright 2005, with permission from Elsevier.
From this, the group estimated that the average crystallite size of the ZnS particles prepared in 0.5 wt.% surfactant was approximately 7.74 nm. Increasing the surfactant concentration to 1.0 wt.% and 2.0 wt.% produced a corresponding decrease in the particle size, to 6.27 nm and 4.04 nm respectively. The estimations of particle size were confirmed by TEM investigations, which showed that the particles were indeed nanosized, monodispersed and mostly spheroidal (Figure 5.8). The ability of chitosan to remain stable and to exhibit complex sequestering behaviour over a range of processing conditions, coupled with the incredible stability at high (for a biopolymer) temperatures allows a great deal of chemistry
Chitin/Chitosan
105
to be performed on and in the presence of chitosan even with the absence of a copolymer providing enhanced structural stability. The relationship that chitosan has for metal cations was explored in detail by Inoue et al. in their seminal 1993 work on the adsorption of a range of metal cations on chitosan26. In this work, the group investigated the behaviour of metal ions adsorbing onto chitosan from aqueous ammonium nitrate and hydrochloric acid solutions. They found that in the case of divalent copper, oxovanadium, zinc, nickel and cobalt cations, the chitosan had an excellent loading capacity and selectivity, surpassing even that of iminodiacetic acid, a common chelating agent (Figure 5.9). The enhanced ability of chitosan to sequester metal cations was suggested to be due to the cooperative action of the protonated amino groups providing added chelation to that of the usual hydroxyl-mediated mechanism. This particularly strong affinity for metal cations has provided the impetus for investigations into a wide range of nanoparticle syntheses where the control over nucleation and growth of metal species provides the driving force for the overall control of crystal growth. Rather surprisingly, there appears to be little work done on the growth of calcium carbonate phases in the presence of chitosan. This may be because there is little inducement for the calcium carbonate to form in anything other than the usual rhombohedral crystal morphology, albeit with an overall reduction in crystallite size. However, once again by functionalizing the chitosan, the morphology of the calcite can be altered. Liang et al. demonstrated that by functionalizing chitosan with carboxymethyl groups, crystallization is controlled to form petunia-shaped calcitic structures27. The group attribute the formation of this unique morphology to the CO2 bubbles produced, acting as a transient template around which calcite will form28. By having a supersaturated solution of Ca(HCO3)2 in an unsealed beaker, the group were able to promote the crystallization of calcium carbonate in the upper regions of the reaction volume, owing to the diffusion of CO2 to the surface. The viscocity of the carboxymethyl chitosan means that diffusion of CO2 through it is reduced, allowing for trapped bubbles to act as sites of calcium carbonate growth. These regions of CO2 concentration produced the head of the petinua-shaped aggregates. This gas-mediated templated mineralization was previously observed by Colfen et al.29 in the production of calcite by the use of double hydrophilic block copolymers as templates. Surfactants such as block-copolymers have played and continue to play an important role in the control of crystallization of inorganic phases.
106
Biotemplating
Figure 5.8 – TEM micrographs of ZnS nanoparticles prepared with different surfactant concentration: (a) 0.5 wt.%, (b) 1.0 wt.%, (c) 2.0 wt.% of chitosan laurate. Reprinted from25. Copyright 2005, with permission from Elsevier.
Chitin/Chitosan
107
Figure 5.9 – Adsorption of n-valent metal ion on chitosan from ammonium nitrate solution. Reproduced by permission of The Chemical Society of Japan.
The functionalization of chitosan was also employed by Huang et al. who demonstrated the functionalization of carboxymethyl chitosan to produce amphiphillic (2-hydroxyl-3-butoxy)propyl-carboxymethyl chitosan (HBPCMCS)30. In doing so, the group rationalize that as calcite crystallization begins at preferred sites of nucleation on the HBP-CMCS, crystallite growth is controlled by the enhanced self-interaction of the biopolymer. The surfactant-like behaviour of the HBP-CMCS produces a different reaction volume than observed with chitosan alone, thereby leading to a peanut-shaped crystallite. XRD studies on the calcium carbonate produced in these experiments revealed that not only was crystallite size controlled, but that a degree of polymorph alteration was achieved. The majority phase produced was determined to be calcite, but approximately 21% of the material was vaterite. In addition, as the concentration of HBP-CMCS was increased, XRD beaks became progressively broadened, indicating that the crystallite size was reduced in solutions of higher biopolymer concentration. At a very low molar ratio of calcium to HBP-CMCS, the group found that inhibition of particle growth was such that precipitation of calcium carbonate was prevented. This was attributed to the amphiphillic behaviour of the
108
Biotemplating
HBP-CMCS acting as a colloidal stabilizer, keeping the calcium carbonate crystallites from flocculating. In a similar fashion, the production of nanoparticulate gold was able to achieved by Huang et al. by using the chitosan as a capping agent in the reduction of the gold salt HAuCl431. By utilizing the reducing ability of chitosan, colloidal gold nanoparticles were able to be produced with sizes between 10 nm and 15 nm, comparable with that seen in the citrate reduction of HAuCl4. As in the case of calcite, colloidal stability was conferred by the chitosan chains ‘cocooning’ the inorganic phase. Although useful, calcite and gold are perhaps the least ‘functional’ of inorganic materials in terms of range of applications. In demonstrating that calcite and calcium phosphate and even gold can have their particle size controlled by growth in the presence of chitosan, studies on functional inorganic phases were the next logical step. For many materials, physical properties are enhanced as particle size decreases owing to the higher proportion of atoms of the particle being at the more reactive surface. One such material synthesized using chitosan as a template is alumina. Alumina is a prime example of the need to control crystal growth, as it is a common catalyst, both intrinsically and as a support for other catalytic materials when synthesized as a high surface area material32, 33. Fajardo et al. used a very simple method of obtaining mesoporous spheres of alumina34. In an acidic solution of chitosan, they dissolved aluminum nitrate, which once homogeneous was added dropwise to an ammonia solution. The change in pH induced precipitation of aluminum hydroxide throughout the chitosan. After retrieval and drying for 72 hours, the spheres were calcined to 700 °C for one hour to remove the chitosan and form alumina. SEM images revealed that the alumina spheres were of the order of 900 microns in diameter. Surface area measurements by BET analysis showed that the spheres were highly porous, with a surface area of 464 m2/g, indicating their suitability as a useful catalytic support. Testing of this material was performed by the incorporation of a small amount of nickel nitrate with the aluminum nitrate during the synthesis, in order to allow the catalytic CO2 reforming of CH4 to be examined. Carried out at 650 °C, the conversion rates of CH4 over the highly porous alumina were between 40% and 60%, depending on how the alumina was calcined. The group attribute this higher catalytic ability to the strong interaction between Ni and the alumina at higher calcination temperatures. With a closer metal-support interaction, the group note that there will be less coke deposition at the nickel sites during the
Chitin/Chitosan
109
catalysis, a common cause of catalyst deactivation in this reaction. The group observed that the stability of the catalyst and support was also good, essentially invariant in catalytic turnover from 5 hours up to 24 hours. Titania (TiO2) has similarly been synthesized with improved properties and a high surface area in the presence of chitosan. Retuert et al. used a pre-hydrolized solution of titanium (IV) isopropoxide which was added dropwise to a solution of chitosan in formic acid35. Once formed, the gel particles were recovered by centrifugation and dried for 16 hours at 80 °C. Calcination to temperatures between 500 °C and 800 °C for two hours produced anatase and mixed anatase/rutile phases respectively. After calcination, the TiO2 materials were porous and monolithic. The group found that the surface area of the monolithic material increased with the amount of chitosan used in the synthesis; the greater amount of interconnected chitosan in the composite resulting in the larger amounts of porosity produced on calcination. Wang et al. used chitosan as a route to the production of ZnS nanoparticles for applications in non-linear optics36. In a simple synthesis a solution of zinc acetate was added to chitosan under stirring. Once a homogeneous solution was achieved, the whole was cast onto glass slides and dried under vacuum. Once dried, the cast film was immersed in Na2S solution to induce the formation of ZnS nanoparticles. TEM studies revealed that the average nanoparticle size was 3.4 nm with a regular array of lattice fringes which suggested that the CS capped ZnS were highly crystalline and lacked surface defects. The group found that the small particle size produced ZnS/chitosan films which demonstrated nonlinear optical absorption with a two photon absorption optical absorption coefficient of 2.29 x 102 cm/Gw. They attribute this to the quantum confinement effect and surface modification of nanoparticulate ZnS. Iron oxide materials are able to be made on the nanoscale in a controlled fashion by synthesis in chitosan. By keeping the Fe2+ centres spatially separate, Janardhanan et al. found that there is a greater control over the calcination regime and less chance of ferric iron being formed37. The synthesis was achieved by the addition of FeSO4 to a solution of chitosan and calcined to 800 °C. The nanoparticles of hematite produced were found to have an average particle size of 102 nm, approximately half the size of analogous particles synthesized in the absence of chitosan. The group measured the zeta potential of the chitosansynthesized hematite and found that it was -10.8 mv. This large negative value is an indication that the nanoparticles are stable with regards to agglomeration and flocculation.
110
Biotemplating
Ganesan et al. studied the effect chitosan has on the growth of WO3 crystals38. This material is increasingly becoming important, as an electrocatalyst, owing to the ready formation of tungsten bronzes (HxWO3) which act as an effective material for the acceleration of electrocatalytic reactions39, 40. Additionally, in theory, tungsten bronze should provide an effective hydrogen storage medium, enabling the sequestration and transport of hydrogen for future use in devices of the ‘hydrogen economy’41. These materials are limited however, in that they are relatively poor at giving up their hydrogen when required. This can be improved by making the crystals nano-sized, leading to higher catalytic activity as well as a more effective hydrogen release. In their work, Ganesan et al. prepared tungsten trioxide nanoparticles by the simple addition of chitosan to an acidic solution of ammonium metatungstate (AMT). After gelation, the composite was calcined to 600 °C to form WO3. TEM images showed that the particles were of the order of 40 nm in diameter and highly crystalline, as evidenced by the clear appearance of lattice fringes in the crystallites. The group were able to index these fringes to the (200) planes of WO3. Hydrogen intercalation of the tungsten bronzes was monitored by cyclic voltammetry. It was found that for chitosan-templated WO3 nanoparticles, the onset voltage potential was -0.62 V and the current value reached 45 mA cm-2 at -0.8 V. This was found to be 4.5 times greater than that seen in WO3 grown in the absence of chitosan, which had a current value of 10 mA cm-2 at -0.8 V. The greater propensity for hydrogen intercalation in nanoparticulate WO3 leads therefore, to a four-fold increase in electrochemical hydrogen evolution by having controlled crystal growth in chitosan. One particularly interesting class of materials to have been templated with chitosan are the so-called high-temperature superconductors. Morphological control of superconductor growth is of prime importance as it has been shown previously that a significant reduction in superconducting properties occurs with an increase in the mis-alignment of grain boundaries of only a 2–7° 42. Hall showed that it was possible to control the crystal morphology of the hightemperature superconductor YBa2Cu4O8 (Y124) using chitosan to sequester metal ions, provide sites of preferred nucleation and growth and to prevent the sintering of particles during calcination43. In doing so, the group created the first hightemperature superconductor nanowires which were readily accessible, the growth template (chitosan) having been removed during synthesis. Experimentally, the synthesis was a simple addition of a solution of mixed metal nitrates in the correct stoichiometries for the formation of Y124, to a solution of chitosan in acetic acid. The chitosan was kept in excess with regards to the mixed-metal salt
Chitin/Chitosan
111
solution (10:1 by volume) to ensure homogeneity of the metal salt solution throughout the biopolymer. This was done in order to prevent the possibility of forming regions where one metal would be deficient, thereby resulting in an incorrect stoichiometry on calcination. Once dried, the composite material was a thin, flexible film which was then calcined to 920 °C in order to form the Y124 phase. The heating rate was found to produce different morphologies in the resulting superconductor. Crystalline superconductor nanoparticles were produced at a heating rate of 10 °C min-1 and nanowires at a heating rate of 1 °C min-1 (Figure 5.10). The calcination protocol in either case involved heating to 500 °C for two hours, followed by cooling to room temperature, then heating to 920 °C for a further two hours, followed again by cooling to room temperature. Energy-dispersive X-ray analysis (EDXA) showed the presence of Y, Ba, and Cu and powder X-ray diffraction (PXRD) confirmed the correct crystallinity, giving sharp peaks that could be indexed to the Y124 phase (selected reflections: d = 3.685 (102), 2.719 (111), 2.223 (117), 1.922 (200), 1.576 (217), 1.499 (219), 1.365 Å (220)).
Figure 5.10 – Transmission electron microscopy (TEM) images showing (a) the control Y124 sample, (b) a low-magnification image of the nanowires, (c) an image of a single nanowire, and (d) an image of a nanowire with corresponding electron diffraction pattern showing a view along the [010] zone axis of Y124 (inset). The crystallographic c-axis of the nanowire is shown. The serrated surface structure was observed after exposure of the nanowires to the electron beam for several minutes at 120 keV. No change to the electron diffraction pattern was observed during this process. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
112
Biotemplating
The nanowires that formed had mean widths of 50 ± 5 nm and were 600 nm to over 1000 nm in length. Electron diffraction analysis indicated that the nanowires were all single crystals of Y124, oriented along the crystallographic c-axis (Figure 5.10d, inset). SQUID magnetometry of the nanowires showed the typical temperature-dependent magnetic response for a superconductor, with a Tc of 85 K, which was within 4 K of the control sample, indicating that no significant loss of superconducting performance was associated with the nanowire morphology (Figure 5.11).
Figure 5.11 – SQUID magnetometry of (a) control Y124 and (b) nanowire Y124 samples. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
Nanowires formed specifically at a heating rate of 1 °C min–1 due to the nucleation and growth of discrete nanoparticles in the chitosan matrix from around 300 °C, followed by outgrowth of the nanowires at 350 °C onwards. This temperature range corresponds to the formation of metal oxide species, and the results suggest that sintering of the nucleated nanoparticles is prevented by the extended presence of the chitosan matrix producing small, nanoparticulate centers for subsequent Y124 outgrowth. The stability and retention of fibre structure in the chitosan matrix on calcination are key to this process. PXRD of pure chitosan films show broad peaks at around 2θ = 11°, 18°, and 22°, which are indicative of the crystalline alignment of the chitosan molecules. Diffraction
Chitin/Chitosan
113
patterns of chitosan–Y124 composite films calcined at 300 and 400 °C show that, although attenuated, reflections at around 11° and more broadly at around 22° indicate that the structural integrity of the matrix is not fully compromised and a degree of crystallinity remains. Nanowire growth is therefore able to occur preferentially along the Y124 crystallographic c-axis, directed and controlled by the fibre network of the chitosan matrix, which is still present at this elevated temperature. It is possible that a c-axis oriented nanowire could facilitate flux ordering by providing crystallographic constraint of flux domains; this remains an exciting possibility for further study. In samples with a heating rate of 10 °C min–1, there is no opportunity for c-axis growth of nanowires, owing to the relative rapidity of degradation of the chitosan template, and discrete nanoparticles of Y124 were produced (Figure 5.12). Control over the size of the nanoparticles was achieved by adjusting the amount of Y124 precursor sol added to the chitosan solution.
Figure 5.12 – TEM images showing (a) 10 ± 7 nm nanoparticles of Y124 formed from the addition of 0.5 mL of the Y124 sol to 10 mL of chitosan solution and (b) 28.5 ± 3 nm nanoparticles of Y124 formed by increasing the amount of Y124 sol to 1 mL. The corresponding electron diffraction pattern is indexed to the Y124 phase (inset). Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
114
Biotemplating
5.3 References 1. 2.
3. 4. 5.
6. 7. 8. 9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21.
Babel, S. & Kurniawan, T.A. Low-cost adsorbents for heavy metals uptake from contaminated water: a review. J. Hazard. Mater. 97, 219-243 (2003). Dambies, L., Vincent, T. & Guibal, E. Treatment of arsenic-containing solutions using chitosan derivatives: uptake mechanism and sorption performances. Water Res. 36, 36993710 (2002). Manoli, F., Koutsopoulos, S. & Dalas, E. Crystallization of calcite on chitin. J. Cryst. Growth 182, 116-124 (1997). Falini, G., Albeck, S., Weiner, S. & Addadi, L. Control of aragonite or calcite polymorphism by mollusk shell macromolecules. Science 271, 67-69 (1996). Kato, T. & Amamiya, T. A new approach to organic/inorganic composites. Thin film coating of CaCO3 on a chitin fibre in the presence of acid-rich macromolecules. Chem. Lett., 199-200 (1999). Falini, G., Fermani, S. & Ripamonti, A. Crystallization of calcium carbonate salts into betachitin scaffold. J. Inorg. Biochem. 91, 475-480 (2002). Wan Andrew, C.A., Khor, E. & Hastings, G.W. The influence of anionic chitin derivatives on calcium phosphate crystallization. Biomaterials 19, 1309-1316 (1998). Yamaguchi, I. et al. The chitosan prepared from crab tendon I: the characterization and the mechanical properties. Biomaterials 24, 2031-2036 (2003). Yi, H.M. et al. Biofabrication with chitosan. Biomacromolecules 6, 2881-2894 (2005). Mukherjee, D.P. et al. An animal evaluation of a paste of chitosan glutamate and hydroxyapatite as a synthetic bone graft material. J. Biomed. Mater. Res. Part B 67B, 603-609 (2003). Spasowka, E., Rudnik, E. & Kijenski, J. Biodegradable polymer nanocomposites. Part I. Methods of preparation. Polimery 51, 617-626 (2006). Habraken, W., Wolke, J.G.C. & Jansen, J.A. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv. Drug Deliv. Rev. 59, 234-248 (2007). Pinheiro, A.G. et al. Chitosan-hydroxyapatite-BIT composite films: Preparation and characterization. Polym. Compos. 28, 582-587 (2007). Ehrlich, H. et al. Chitosan membrane as a template for hydroxyapatite crystal growth in a model dual membrane diffusion system. J. Membr. Sci. 273, 124-128 (2006). Chesnutt, B.M. et al. Characterization of biomimetic calcium phosphate on phosphorylated chitosan films. J. Biomed. Mater. Res. Part A 82A, 343-353 (2007). Wang, L. & Li, C.Z. Preparation and physicochemical properties of a novel hydroxyapatite/chitosan-silk fibroin composite. Carbohydr. Polym. 68, 740-745 (2007). Tsai, M.S. & Li, M.J. A novel process to prepare a hollow silica sphere via chitosanpolyacrylic acid (CS-PAA) template. J. Non-Cryst. Solids 352, 2829-2833 (2006). Beppu, M.M. & Santana, C.C. PAA influence on chitosan membrane calcification. Mater. Sci. Eng. C-Biomimetic Supramol. Syst. 23, 651-658 (2003). Li, J.J., Chen, Y.P., Yin, Y.J., Yao, F.L. & Yao, K.D. Modulation of nano-hydroxyapatite size via formation on chitosan-gelatin network film in situ. Biomaterials 28, 781-790 (2007). Cheng, D.M., Xia, H.B. & Chan, H.S.O. Fabrication of polymeric hollow nanospheres, hollow nanocubes and hollow plates. Nanotechnology 17, 1661-1667 (2006). Marchetti, A.P. et al. Nanoregions of rocksalt AgI in AgBr microcrystals. Phys. Rev. B 69, 9, id. 094107 (2004).
Chitin/Chitosan 22.
23. 24.
25.
26.
27. 28. 29.
30.
31. 32. 33.
34.
35. 36. 37. 38. 39. 40. 41.
115
Green, D.W. et al. Biomineralized polysaccharide capsules for encapsulation, organization, and delivery of human cell types and growth factors. Adv. Funct. Mater. 15, 917-923 (2005). Ding, Y., Hu, Y., Zhang, L.Y., Chen, Y. & Jiang, X.Q. Synthesis and magnetic properties of biocompatible hybrid hollow spheres. Biomacromolecules 7, 1766-1772 (2006). Wong, D.W.S., Gastineau, F.A., Gregorski, K.S., Tillin, S.J. & Pavlath, A.E. Chitosan Lipid Films - Microstructure and Surface-Energy. J. Agric. Food Chem. 40, 540-544 (1992). Khiew, P.S., Radiman, S., Huang, N.M., Ahmad, M.S. & Nadarajah, K. Preparation and characterization of ZnS nanoparticles synthesized from chitosan laurate micellar solution. Mater. Lett. 59, 989-993 (2005). Inoue, K., Baba, Y. & Yoshizuka, K. Adsorption of Metal-Ions on Chitosan and Cross-Linked Copper(II)-Complexed Chitosan. Bull. Chem. Soc. Jpn. 66, 2915-2921 (1993). Liang, P. et al. Petunia-shaped superstructures of CaCO3 aggregates modulated by modified chitosan. Langmuir 20, 10444-10448 (2004). Kitano, Y. The Behaviour of Various Inorganic Ions in the Separation of Calcium Carbonate from a Bicarbonate Solution. Bull. Chem. Soc. Jpn. 35, 1973-1980 (1962). Rudloff, J. & Colfen, H. Superstructures of temporarily stabilized nanocrystalline CaCO3 particles: Morphological control via water surface tension variation. Langmuir 20, 991-996 (2004). Huang, Y.P. et al. Peanut-shaped aggregation of CaCO3 crystallites in the presence of an amphiphilic derivative of carboxymethylchitosan. Colloid Polym. Sci. 285, 641-647 (2007). Huang, H.Z. & Yang, X.R. Synthesis of polysaccharide-stabilized gold and silver nanoparticles: a green method. Carbohydr. Res. 339, 2627-2631 (2004). Trimm, D.L. & Stanislaus, A. The Control of Pore-Size in Alumina Catalyst Supports - A Review. Applied Catalysis 21, 215-238 (1986). Marquez-Alvarez, C., Zilkova, N., Perez-Pariente, J. & Cejka, J. Synthesis, characterization and catalytic applications of organized mesoporous aluminas. Catal. Rev.-Sci. Eng. 50, 222-286 (2008). Fajardo, H.V. et al. Synthesis of mesoporous Al2O3 macrospheres using the biopolymer chitosan as a template: A novel active catalyst system for CO2 reforming of methane. Mater. Lett. 59, 3963-3967 (2005). Retuert, J., Quijada, R. & Arias, V. Porous titania obtained through polymer incorporated composites. Chem. Mat. 10, 3923-3927 (1998). Wang, X.H. et al. Large two-photon absorbance of chitosan-ZnS quantum dots nanocomposite film. Physica E 30, 96-100 (2005). Janardhanan, S.K., Ramasamy, I. & Nair, B.U. Synthesis of iron oxide nanoparticles using chitosan and starch templates. Transit. Met. Chem. 33, 127-131 (2008). Ganesan, R. & Gedanken, A. Synthesis of WO3 nanoparticles using a biopolymer as a template for electrocatalytic hydrogen evolution. Nanotechnology 19, 5, id. 025702 (2008). Wang, Q. et al. Li-driven electrochemical properties of WO3 nanorods. Nanotechnology 17, 3116-3120 (2006). Chen, H.J. et al. Gasochromic effect and relative mechanism of WO3 nanowire films. Nanotechnology 18, 6, id. 205701 (2007). Tseung, A.C.C. & Chen, K.Y. Hydrogen spill-over effect on Pt/WO3 anode catalysts. Catal. Today 38, 439-443 (1997).
116 42. 43.
Biotemplating Klie, R.F. et al. Enhanced current transport at grain boundaries in high-Tc superconductors. Nature 435, 475-478 (2005). Hall, S.R. Biomimetic synthesis of high-Tc, type-II superconductor nanowires. Adv. Mater. 18, 487-490 (2006).
Chapter 6
Proteins and Lipids
In nature, structure is usually conferred by the use of one of two types of macromolecular assembly. We have already seen how one of these types, the polysaccharide is able through the glycosidic bond to provide complexity through this very simple bond formation. Variations on the building blocks and the way these are conjoined through the glycosidic bond provide a rich variation on a structural theme, which organisms (including the scientist) can utilize as templates in the creation of complex inorganic phases. The other main structural macromolecular assembly found in nature are the proteins. As the glycosidic bond is to polysaccharides, so the peptide bond is to proteins.
6.1 Proteins – structures and properties The sub-units of proteins are amino acids; molecules which contain both carboxyl and amine functional groups. There are a number of ways in which a molecule can be constructed to contain both functional groups, but perhaps the simplest are the α-amino acids, where both functionalities are attached to the same carbon (Figure 6.1).
Figure 6.1 – Molecular structure of an alpha-amino acid. 117
118
Biotemplating
These amino acids have the general formula H2NCHRCOOH, where R is any organic moiety. This is where the richness of form and function is generated in proteins, as the R group in the amino acids can vary in size from the simple proton in the case of the amino acid glycine, to large, bulky organic groups such as tyrosine and tryptophan. The R group can also be a long-chained aliphatic moiety, further extending the range of physical behaviour of these molecules, by making them (depending on the length of the aliphatic chain) exhibit simple surfactant-like behaviour. The side-chain group is important in that it largely determines the eventual structure of the protein. Differences in relative hydrophilicity and hydrophobicity of different side-groups leads to differences in polarity, which create attractive and repulsive interactions between amino acid residues inter- and intra-molecularly. The resulting tertiary structure of the protein is thus complex and well-defined for any given protein. These interactions are particularly difficult to simulate and research into predicting the way a protein will fold is an extremely active area of research. In addition, a further layer of complexity is afforded by the stereochemical environment of the α-carbon. From the structure of the amino acid in Figure 6.1, one can readily see that the α-carbon is a chiral centre, which means that the potential number of protein structures possible on peptide bond formation is effectively doubled. Of the two enantiomeric forms D and L, most of the naturally occurring amino acids (but not all) adopt the L form. Rather confusingly, the L and D labels for amino acids do not refer to the optical activity of the amino acid itself, but to the optical activity of the molecule glyceraldehyde having the same stereochemistry as the amino acid under consideration. One of the simplest amino acids, glycine does not exhibit isomerism, as with R = H, the α-carbon is no longer a chiral centre. The combination of amino acids to form proteins occurs with the condensation of the amine group on one amino acid, with the carboxyl group of another in the formation of a peptide bond (Figure 6.2). The resulting molecule can undergo further condensation reactions, thereby building up the number of amino acid residues in the molecule. From simple dipeptides, through polypeptides to proteins, an almost infinite number of complex molecules can be formed. In nature, proteins find a wide range of uses, most notably as enzymes or in mechanical applications such as the myosin and actin proteins which constitute muscle fibre1, 2. Wide use is also made of more complex protein assemblages to provide structure in cells and as vessels for the encapsulation of diverse inorganic
Proteins and Lipids
119
materials. It is the latter function of proteins which has been used by biomimetic chemists to produce biotemplated materials with complex form. This chapter deals with the use of the simpler forms of protein as templates. Complex proteinaceous structures generated by bacteria and viruses have been particularly fertile ground for biotemplating and so will be dealt with in another chapter.
Figure 6.2 – Molecular representation of a peptide bond.
Structural proteins were among the first to be investigated as biotemplates. In particular, fibrous proteins such as collagen and keratin and globular proteins which adopt a hollow tertiary structure in order to sequester and facilitate the transport of specific ions. But there is complexity in the crystalline form of the protein itself, which means that the possibility exists to use this as a morphological directing agent for subsequent inorganic crystallization. Surprisingly, this has not yet (as of 2009) been investigated. Freeman et al. used a crystalline form of the protein lysozyme as a template to form polyacrylamide replicas of the pore structure of the enzyme3. Void spaces in a protein crystal can occupy 30–60% of the total volume of the crystal and can be highly interconnected and complex in overall morphology, depending on the crystal under investigation. They found that although previous work had been done on the infiltration of these void spaces4, 5, no work had been done on the solidification on the infilling species; the void spaces of the enzyme crystal being used primarily as a vector for the catalysis of organic syntheses. In order to
120
Biotemplating
perform a chemical transformation within the pore space of an enzyme crystal, the protein first had to be stabilized in order to improve the structural stability of the crystal. Freeman et al. achieved this by cross-linking the lysozyme crystal with a solution of glutaraldehyde and sodium acetate at pH 4.8, followed by extensive washing in a sodium acetate/sodium chloride solution. These crosslinked crystals were then immersed in a solution of 15% acrylamide, 1% (v/v) ethyleneglycoldimethacrylate, and 0.25% (v/v) tetraethylenepentamine, dissolved in 25% aqueous ethylene glycol, for 20 hours at ambient temperature. Polymerization of the acrylamide solution within the pore space of the lysozyme crystal was then simply achieved by immersing the crystals in a solution of ammonium persulfate. The group were able to prove that the complex 3D structure of the lysozyme crystal was retained throughout the process of crosslinking and acrylamide infiltration/polymerization by monitoring the X-ray crystallography of the crystals. They found that the protein crystals had the same space group and unit cell dimensions in each case, with variations in the displacement of atoms in the crystal lattice (B-factor) as each step was performed (Table 6.1). Table 6.1 – Step-by-step characterization by X-ray diffraction of chemically crosslinked gel “filling” of lysozyme crystal. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. Crystal Native lysozyme Cross-linked lysozyme Cross-linked lysozyme following equilibration with 15% acrylamide Cross-linked lysozyme following polymerization with 15% acrylamide
Resolution (Å)
Space group
Unit cell dimensions (Å)
B-factora (Å2)
1.7 2.0 2.6
P43212 P43212 P43212
a = b = 78.883, c = 36.773 a = b = 78.559, c = 36.773 a = b = 79.45, c = 37.864
15.52 (15.52) 28.674 (13.78) 110.747 (25.51)
2.5
P43212
a = b = 78.958, c = 36.428
87.224 (23.698)
a
B-factor represents mean displacement of the equivalent atoms in the crystalline lattice (B = 8π 2 * U 2, U = mean atomic displacement). The value in parentheses is B-factor of the native structure at present resolution.
The value of the B-factor is a reflection of the thermal motion and disorder in the protein crystal. The data suggested to the group therefore that as cross-linking and acrylamide infiltration progressed, the lysozyme crystal was able to undergo increased conformational movements, whilst retaining the overall crystal structure. Once polymerization was affected however, the B-factor was reduced, indicating a restriction of the free fluctuations of the protein surface. This provides an elegant way of monitoring the interaction between a protein crystal and its local chemical/electrostatic environment and thereby act as a sensitive
Proteins and Lipids
121
sensor to chemical changes. This work immediately suggests that protein crystals can be used as a vector for the growth of inorganic phases within the protein crystal. In nature, it is after all, lysozyme which plays a key role in the control of calcite mineralization in hens egg-shells. Simple adaptation of the synthetic protocol above should therefore lead to inorganic phases being created within the cross-linked crystal voids. The more usual approach of control of crystallization by proteins and polypeptides is that of the simple adsorption of the molecule onto a crystal surface. This produces control over crystal growth through passivation of crystal surfaces. One of the earliest studies on the specificity of molecular configuration to spatially complementary crystal faces was by Heywood and Mann. In a review of their early work6, they demonstrate that certain macromolecular configurations would produce defined crystal morphologies, highly suggestive of a direct epitaxial matching of the organic to the inorganic phase (Figure 6.3).
Figure 6.3 – Oriented overgrowth by epitaxy involving geometric matching between the host substrate (Y) and a specific face (X) in the developing crystal. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
They realized however, that in reality, the assumption that a one-to-one lattice correspondence was clearly untenable in many cases. This was prompted by the
122
Biotemplating
observation that oriented overgrowth could be achieved with lattice mismatches approaching 30%. As observed in Chapter 1, it was this and analogous observations which led to the general consensus of biotemplating being a cooperative process involving selective charge accumulation, spatial and stereochemical disposition of active functional groups and the underlying symmetry of the crystallizing phase. In this regard, polypeptides have been used effectively in the control of mineralization, in particular those polypeptides which possess a well defined complex secondary structure. A good example of this was in the work of DeOliveira et al. who looked at the specific interaction of an R-helical peptide (CBP1), which was synthesized to have an array of aspartyl residues of the correct spacing to bind to the (110) faces of calcite7. When the group added the polypeptide to a saturated solution of calcium carbonate, they found that the specificity was highly dependent on the temperature of the growth medium. They observed that when the temperature was 3 °C, where CBP1 is in a helical conformation, the calcite crystals grew with a prismatic habit, with growth continuing along the c-axis, but inhibited parallel to the c-axis and prism faces. When the temperature was 25 °C, the polypeptide was no longer in a helical conformation and they observed that calcite growth produced unstructured, studded crystals, which they reasoned occurred from epitaxial growth off each of the six rhombohedral surfaces. The helical form of the polypeptide was therefore acting as a crystallochemical specific passivating agent, whilst the unstructured CBP1 was acting as a simple polyanionic, non-face specific passivating agent. The group found that when other non-helical peptides were employed as passivating agents, the same non-specific crystal growth was observed. A different way to ensure shape specificity is to use an imprinting technique on a pliant inorganic substrate. By impressing a siloxane surface with the protein lysozyme, Puleo et al. were able to create a precise inverse of the shape of the protein in order to produce a surface which would specifically recognize a lysozyme molecule on subsequent exposure to the protein8. This technique is prevalent in chromatography in order to produce a static phase (usually a methacrylate polymer) with the potential for separating specific small molecules by acting essentially as an ‘artificial antibody’, but is relatively unknown in biotemplating. The group first created a polysiloxane scaffold by synthesizing a tetraethoxysilane (TEOS) aminopropyltriethoxysilane (APS) gel via an acid catalyzed route. The addition of a surfactant, sodium docecylsulfate (SDS) enabled the creation of a porous structure in order to allow perfusion throughout
Proteins and Lipids
123
the structure. Addition of lysozyme to the APS resulted in the successful incorporation/imprinting of the protein once gelation was complete. The digestion of the lysozyme from the siloxane scaffolds was simply achieved by immersion of the material in a protease solution. SEM and TEM studies of the digested scaffold showed that the scaffolds were highly porous, with a wide distribution of pore sizes, ranging from nanosized up to 200 µm. The group examined the ability of imprinted scaffolds to preferentially bind lysozyme by immersing them in solutions of mixed protein. The proteins in the solutions were labelled with fluorophores of distinct wavelengths in order to distinguish which protein was adsorbed. The group labelled lysozyme with Alexa Fluor 488 (molecular probes; λex = 495 nm and λem = 519 nm), and the other protein (RNase A) with was labelled with Alexa Fluor 594 (molecular probes; λex = 590 nm and λem = 617 nm). Control (non-imprinted) scaffolds were found not to discriminate between lysozyme and RNase, and bound similar amounts of the different proteins. They found that in solutions containing only lysozyme (1:0), approximately 45% of the available protein was re-bound to the surface of lysozyme imprinted scaffolds. In a solution containing just the competitor protein RNase A (0:1), the amount of protein re-bound was found to be approximately 27%. This result means that almost twice as much template protein than competitor was re-bound to the surface. When equal amounts of the proteins (1:1) were present in the rebinding solution, about 2.5 times more lysozyme than RNase A was present on the surface. Lysozyme has also been used in the dual role of reaction catalyst and in situ templating agent in the formation of zirconia microparticles. In nature, the role of an enzyme is primarily catalytic, with over 3,500 known to perform critical roles in cell regulation, ion pumping and in digestion9. There are instances however, where the enzyme performs in a structure-directing role. A commonly encountered group of biomolecules which perform this dual role are the silaffins, proteins which are present in some marine invertebrates such as diatoms and sponges10. The polycationic nature of the sillafins has been postulated to act as a catalyst for the condensation of silicic acid, with the resulting silica forming around the structure of the silaffin itself. Silaffins have even been used to promote and control the formation of non-siliceous phases. For example, Wright et al. looked at the use of a naturally occurring peptide from a silaffin protein expressed in the diatom Cylindrotheca fusiformis in the templating of titanium dioxide (TiO2)11. They found that the peptide was able to form TiO2 even from a non-naturally occurring source, titanium bis(ammonium lactato)dihydroxide.
124
Biotemplating
Jiang et al. noted that lysozyme is also polycationic and should therefore be able to act in a similar way to the silaffins12. They demonstrated that lysozyme was able to catalyze the hydrolysis/condensation of potassium hexafluorozirconate (K2ZrF6) and template the formation of zirconia particles at room temperature. Control experiments (no lysozyme) produced no precipitate of zirconia even after two hours of incubation at room temperature. On addition of 1 ml of a 20 mg/ml lysozyme solution to 9 ml of a 0.01 M K2ZrF6 solution, a precipitate was formed within several minutes, which the group determined via XRD to be titania (Figure 6.4).
Figure 6.4 – Scanning electron microscopy (SEM) micrograph of zirconia particles prepared in the presence of native lysozyme. Reprinted with permission from12. Copyright 2008 American Chemical Society.
SEM studies revealed that the lysozyme precipitated particles were approximately 1 µm in diameter with a relatively narrow distribution of crystallite sizes. This contrasted with zirconia precipitated from the K2ZrF6
Proteins and Lipids
125
solution by ammonium hydroxide, which produced zirconia particles of the order of 5 µm in size and reasonably polydisperse in terms of particle size and morphology. Another group used a polymer-protein composite in order to mimic the silaffin-mediated condensation of silica, although using a pre-hydrolyzed silica source meant that the system was not behaving as a catalyst-template, but rather as a simple template in the directed formation of silica. Borner et al. used a poly(ethyleneoxide)-peptide (PEO-peptide) composite as a morphological directing agent for the condensation of silica from tetramethoxysilane (TMOS)13. The group used a PEO-peptide composite for the silaffin analogue, as it has been previously observed to self-assemble in water, producing nanotapes14. In addition, this previous work revealed that the peptide part of the composite had a tendency to form β-sheet secondary structures, which presented the cationic threonine groups prominently. This, coupled with the presence of N,N-dimethyl glycine residues on the peptide ensured that the molecule had regions of readily accessible positive charge density, enabling strong interactions with the anionic silica species in pre-hydrolyzed TMOS solution. The PEO fragment of the composite had the effect of shielding the planar β-sheet faces, thereby mimicking the anisotropy of the silica-fibre forming silaffins. TEM studies of the uncoated PEO-peptide self-assembled structures revealed the presence of tapes 13–17 nm wide and up to 2 µm long. The group then took an ethanolic solution of PEOpeptide at a concentration of 0.12 mg/ml and equilibrated it with pre-hydrolyzed TMOS. This resulted in a slicification of the nanotapes, producing tapes in the region of 15 nm thick, but with increased length of up to 5 µm long (Figure 6.5). The increased length of the nanotapes suggested to the group that silica incorporation was inside the PEO-peptide network, the stabilization due to silica incorporation leading to a stiffening of the nanotape structures and an increased resistivity to degradation of the PEO-peptide by thermal effects. The group note that as the silicic acid concentration is relatively low (approximately three times the equilibrium concentration of silicic acid in the oceans) and the incubation times are short at less than 30 seconds, that there is a strong binding capability of the PEO-peptide assemblage for silica. Stupp et al. used a similar composite assemblage in order to synthesize nanofibres of hydroxyapatite. In investigating the relationship between HAp and an organic phase, they hoped to mimic the interaction between collagen and HAp in bone. In order to do this, the group synthesized a peptide amphiphile (PA) which contained several design features to enable the generation of fibrous
126
Biotemplating
structures with particular affinity for HAp mineralization in aqueous environments. The structural and chemical motifs were demarked into discrete regions of the molecule. Region 1 was a long alkyl tail conveying a hydrophobic character to the molecule which, when combined with the peptide region, results in the amphiphilic nature of the molecule.
Figure 6.5 – Isolated silica composite tapes obtained after mixing a diluted PEO-peptide nanotape solution with a diluted solution of prehydrolyzed silica precursor (2.5 equiv. per threonine) for 10 s [AFM micrograph of uniform composite tapes (a); TEM image illustrating the low- and highcontrast areas, resulting from deposited silica components (b) [inset: SAED]; AFM micrograph of two closely packed composite tapes (c), and the height analysis, corresponding to the length profile of a composite tape (d)]. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
Region 2 was composed of four consecutive cysteine residues that when oxidized formed disulfide bonds. This has the effect of ‘cross-linking’ the structure, forming a strong polymerized assemblage of molecules. Disulfide bond formation is reversible, so the selection of these allowed for self-correction of improperly made bonds, or a ‘deassembling’ of the polymerized structure by
Proteins and Lipids
127
treatment with a mild reducing agent. Region 3 was a flexible linker region of three glycine residues which provided the hydrophilic head group. Region 4 was a single phosphorylated serine residue designed to interact strongly with calcium ions and help direct mineralization of hydroxyapatite. Phosphorylated proteins are associated with the collagen extracellular matrix in vivo and have been found to play a significant part in HAp mineralization15. Finally, Region 5 contained the cell adhesion ligand RGD as the group decided that with a view to potential biomedical applications, it would be advantageous for their molecule to promote the adhesion and growth of cells on the surface of the assemblages. It has been demonstrated previously that fibronectin, a protein closely associated with collagen contains the amino acid sequence Arg-Gly-Asp (RGD), so the engineering of this sequence into their PA structure could be an important factor in establishing cell adhesion. After assembling the PA structures, the group treated the molecules with dithiothreitol, at a pH of 8. This had the effect of reducing the cysteine residues in the PA to free thiol groups and thereby making the PA soluble in water (at concentrations greater than 50 mg/ml). TEM investigations of a frozen gel of the PA material revealed a network of fibres of approximately 7.6 nm in diameter, with lengths up to several microns. With increasing concentrations of PA in solution, the group found that the fibres were packing into flat ribbon assemblages, a structural motif common in many fibrous biopolymers. The group examined the mineralization potential of the PA molecules by exposing TEM-grid mounted samples to treatment with 10 mM CaCl2 and 5 mM Na2HPO4 and dropping the two solutions onto the grid from opposite sides. This ensured that mixing of the two solutions was limited to the holes in the carbon coating of the grids, thereby restricting uncontrolled calcium phosphate mineralization. Thirty minutes after the introduction of the mineralizing solution, the group found that a platy, polycrystalline material was visible on the surfaces of the PA fibres. Analysis of this inorganic phase by energy dispersion X-ray fluorescence spectroscopy (EDS) indicated a Ca:P ratio of 1.67 ± 0.08, indicative of HAp with a stoichiometry of Ca10(PO4)6(OH)2. Electron diffraction of the fibres showed that the HAp grew with preferred crystallization, the crystallographic c-axis being aligned with the fibre long axis. Another polypeptide, poly-L-lysine (PLL) has been used by Naik et al. to promote and control silica condensation16. Mineralization was achieved simply by incubation of PLL at concentrations of between 1 and 10 mg/ml in buffered 50–100 mM silicic acid. Precipitation of silica occurred within minutes of the introduction of silicic acid and varied in morphology depending on the PLL chain
128
Biotemplating
length. When the chain length of the PLL was between 100 and 840 amino acid residues, the silica formed as discrete platelets, whereas PLL chain lengths less than 100 gave a gel-like network of silica spheres (Figure 6.6).
Figure 6.6 – SEM micrographs of silica structures using PLL20 and PLL222. (A) Network of fused silica spheres using PLL20. (B) Hexagonal silica platelets obtained using PLL222. Scale bar 2 µm. (Inset) Low-magnification micrograph showing the aggregation of PLL222-derived silica platelets. Reprinted with permission from16. Copyright 2005 American Chemical Society.
Proteins and Lipids
129
As would be expected from an ambient temperature synthesis, the silica platelets produced were amorphous. The group surmise that formation of hexagonal PLL crystals was due to the supramolecular assembly of R-helical chains of PLL, assembling in the presence of the phosphate counterions in the buffer solution. Platelets were not observed when PLL was in a non-helical (random coil) formation, which suggested that the helicity of the PLL was an important factor in the assemblage of silica with defined morphology. This was confirmed with the formation of silica spheres when PLL was too short (<100 amino acid residues) to adopt a helical conformation. The group suggest that long-chain PLL undergoes a rapid secondary-structure transition from a random coil to a helical structure in the presence of silicic acid and phosphate ions. The negatively charged silicate precursor, in association with the phosphate ions, allows for the secondary-structure phase transition of PLL to occur at neutral pH. The formation of the helical PLL chains during the silicification reaction results in the formation of hexagonal silica platelets. Yu et al. used another polypeptide, poly(aspartic acid) (PAA) to induce the formation of calcitic spheres17. PAA has been employed numerous times in the control of calcium carbonate mineralization, as aspartic acid has been found in many biological calcitic mineralization systems and is therefore considered to play a significant role in the control of crystallization. In their work, Yu et al. studied the effect of PAA on the growth of CaCO3 phases from the initial nucleation of calcitic nanoparticles. They found that as CaCO3 phases grew, the PAA was able to effectively cap each nanoparticle, thereby ensuring that morphology was controlled. They consider that the ‘soft’ PAA capping layer binding on the growing CaCO3 nanoparticle surfaces resulted in an inhibition of the expression of preferred crystal faces, producing an isotropic, essentially ‘amorphous’ growth of spherical nanoparticles. Proteins which adopt a more complex tertiary structure are more useful for controlling the macromorphology of an assemblage of nanoparticles. Mukherjee et al. used antibodies to vascular endothelial growth factor-165 isoform (VEGF), 2C3 (AVF), epidermal growth factor receptor, C225, and immunoglobulin G (IgG) as morphological templates during the synthesis of gold nanoparticles18. When cooled to a temperature of -20 °C, these proteins adopt a rod-like shape, to which the gold nanoparticles were adsorbed. This was possible due to the overall cationic nature of the surface of the protein assembly, which was able to act as a site for the electrostatic binding of AuCl4- ions in solution. A typical synthesis involved the incubation of the protein with a solution of HAuCl4 for 30 minutes,
130
Biotemplating
followed by the introduction of the reducing agent sodium borohydride. After further incubation, the samples were frozen to -20 °C followed by thawing to room temperature for subsequent analysis. The group found that they were able to maintain a slight degree of control of morphology of the gold nanorods by variations in the ratio of HAuCl4 to protein. With increasing peptide length, the opportunity arises for proteinaceous assemblages to have ever more complex form. In nature, proteins with complex form which are involved in the creation/maintenance of structural features such as cell walls and ion transport are classified broadly into two types according to their quaternary structure. These proteins tend to form either fibrous or globular morphologies and thereby take quite distinct roles in vivo. Fibrous proteins are only found in the animal kingdom and can be characterized by the relatively restricted palette of amino acids which form the protein. They tend to feature large, hydrophobic R side-groups which induce selfaggregation and impart a high degree of insolubility to the protein, making it particularly useful as an invariant structural component in the aqueous environment of animal cells. Forming rod or wire-like morphologies, the fibrous proteins comprise the connective tissues in animals, such as collagen, keratin and elastin. With a well defined anisotropic morphology built in to the fibrous proteins and a physico-chemical preference for hydrophobic environments, it is not surprising that fibrous proteins have been used as biotemplates for the mineralization of complex inorganic phases. One of the most widely investigated, owing to its critical role in bone formation in the animal kingdom is collagen. The mineralization of bone is directed largely by the layout of the fibrils of collagen. Hydroxyapatite crystallizes on collagen, producing uniaxially aligned crystals which are oriented along the fibrils. This apparent epitaxial matching has been studied extensively in an attempt to uncover the fine control exerted by the collagen matrix on HAp formation19-21. The mode of formation of HAp in vivo is further complicated by the presence of other macromolecules and as such present a very complex system for investigation. This has led to the use of simplified in vitro systems as models for the formation of HAp. One example study investigated HAp mineralization of a collagen film. In this work, calcium chloride, disodium glycerophosphate and alkaline phosphatase were added sequentially to a collagen solution22. XRD of the calcium phosphate formed revealed that the mineral was indeed HAp. Although there was no attempt to indentify control over mineralization, it is apparent when comparing
Proteins and Lipids
131
the control and collagen-mediated HAp that the relative intensity of the peaks are different in the collagen mediated sample, which can be indicative of a preferred crystallization. Coradin et al. used collagen as morphological template for the growth of silica23. The group observed by birefringence studies that a solution of collagen in acetic acid naturally formed chiral nematic liquid crystal structures comprised of the collagen helices. Increasing pH by the addition of ammonia induced the helices to self-assemble into a gel consisting of fibrillar structures. The group then exposed these gels to the vapour of a volatile alkoxysilane in order that hydrolysis and condensation of silica would occur upon and within the fibrillar gel matrix. The group found that this could be simply achieved by the exposure of open vials of gels to the vapours of tetramethoxysilane (TMOS) in a closed system for six days. Confirmation of the retention of the collagen liquid crystalline structure by polarized light microscopy showed that the birefringence and long-range order was maintained in the silica/collagen hybrid material once condensation was complete. Analysis of the composite gels by TEM revealed a high degree of replication of the collagen structure, with silica forming along the collagen fibrils (Figure 6.7). As a final step, the group were able to calcine the hybrid materials resulting in fibrillar structures in the silica phase. The composite materials were first calcined at 120 °C for two days in order to increase the densification of the silica, followed by heating at 600 °C for two hours to remove the collagen template. Once calcined, BET of the silica material revealed that the average pore size was 2.2 nm ± 0.2 nm, which was slightly larger than the typical collagen helix diameter. The group also investigated the removal of the collagen template by more benign methods; digestion of the collagen by either 1M HCl or by the enzyme collagenase, both of which produced similar results to calcination. The group conclude that the fine pore size in the final material is suggestive of the silica condensing/templating single helices of collagen. Shinkai et al. used a similar approach to the formation of silica fibres, although without the formation of a liquid crystal gel by raising pH24. In their method, a buffered solution of collagen was left to spontaneously form fibres over a period of one month. In doing so, the group formed large fibrous structures, although no description was given as to any spontaneous ordering in solution. As with the work by Coradin et al., the collagen fibres were then exposed to a volatile siloxane, in this case by mixing the collagen solution with a solution of tetraethoxysilane (TEOS).
132
Biotemplating
Figure 6.7 – TEM micrographs of silica-concentrated collagen solution hybrid (a) before (bar = 1 µm) and (b) after calcination (bar = 50 nm) and (c) calcined silica–collagen gel hybrid (bar = 100 nm). Reproduced by permission of The Royal Society of Chemistry.
Proteins and Lipids
133
Hydrolysis and condensation of the silica was then able to occur during quiescent ageing of the mixed solution, in order that mechanical disruption of the fibre structures was minimized. Once silica condensation was complete, the group analysed the results by TEM. This revealed that individual fibres had been coated evenly with silica, producing 50 nm to 100 nm thick structures which contained a central core of collagen. Interestingly, the fibres had a serrated edge, with a period of 60 nm to 80 nm. This corresponds well to the groove in the collagen fibre coils, which was calculated to be 67 nm. Calcination of the composite material resulted in stable, hollow silica tubes without much structural collapse. Aside from collagen, the other fibrous protein most intensively investigated as a biotemplate is silk. Silk is a ubiquitous biopolymer and has been used by Man since time immemorial. It is a particularly useful material, as it is incredibly strong for its size, with some spider dragline silks reportedly having a higher tensile strength than a commensurate steel wire25. It is composed of a small number of amino acids, among which glycine comprises almost 50% of the protein; the small size of the glycine residue leading, in part to the efficient packing and thereby enhanced structural stability of the silk protein fibre through the formation of extensive hydrogen bonding. Silk is particularly useful for the creation of fibrous inorganic phases. Mao et al. investigated the use of spider dragline silk in the growth of HAp26. They found that on incubating strands of silk in a supersaturated HAp solution, elongated crystals grew along the silk fibres. Electron diffraction of these crystals showed that all of the crystals were HAp. In addition, the (002) reflection which the group normally observed (in control experiments) to form a ring, indicative of a random polycrystalline arrangement, was instead in the form of a very short arc, indicating that the HAp crystals were nucleated with preferred crystallographic orientation. Analysis of the disposition of these crystals on the spider silk showed that the crystals were oriented with their c axes preferentially oriented at an angle of 72.9° with respect to the long axis of the silks. The group rationalize the mechanism of preferred orientation as being due to crystal plane matching, stereochemistry matching, and the electrostatic interaction between the silk and the HAp phase. Experiments on the crystallization of other inorganic phases on the silk fibres, namely Au and CdS, showed an affinity for nucleation, but with no preferred orientation on growth. Invoking an epitaxial matching mechanism, the group rationalize the preferred orientation of HAp crystals as
134
Biotemplating
being due to the close match between the a-b plane of HAp matching the b-c plane of silk protein nanocrystals (Figure 6.8). The group found that the a-b plane of HAp and the b-c plane of the silk are almost structurally identical, with only a small degree of mismatch of around 7.5%. FTIR showed that the hydrogen bonding in the silk-HAp composite was linear in nature. This linearity means that the O-H group in HAp nuclei tends to be parallel to the C=O group of the protein at the organic-inorganic interface. It follows then, that the O-H group in HAp will have to be perpendicular to the c-axis of the protein fibres to satisfy the linearity of the hydrogen bonding. The group contend that because of the formation of hydrogen bonds between HAp and silk proteins at the HAp/silk interface, the c-axis of the HAp nuclei will have a tendency to be perpendicular to that of the spider silk, suggesting that the preferred orientation of the c-axis of the fibrous protein will determine that of HAp nucleated on the silk. Electrostatic attraction of Ca2+ ions to the oriented silk β-sheets results in strong binding and the creation of sites of preferred nucleation and subsequent growth of HAp. Once Ca2+ is bound to the silk surface, PO43- ions will be electrostatically attracted to the silk surface, leading to the eventual formation of oriented HAp. The group observe finally that as spider silk is totally biocompatible, the opportunity exists for their composite material to act as a bone implant and tissue engineering scaffold. Kaplan et al. went one stage further in controlled mineralization of HAp using silk. In this study, the group grew HAp on fibres spun from a mixture of silkworm (Bombyx mori) silk and polyethylene oxide (PEO)26. The PEO solution contained poly(L-aspartate (poly-Asp) which the group introduced as an analogue to the associate proteins found in natural bone. The solution of silk/PEO/poly-Asp was subjected to electrospinning at an applied voltage of 12 kV to generate an electric-field strength of 0.6 kV/cm. The flow rate of the solution was 0.02 ml/min. Once spun, the fibres were immersed in 100% methanol in order to induce transformation to the β-sheet structure. As before, a simple soaking of the silk in a solution of buffered CaCl2 (200 mM), followed by transfer to a Na2HPO4 solution (120 mM) produced pronounced mineralization of the fibres. At low poly-Asp concentrations, the group found that HAp mineralization occurred, albeit with a random orientation of the crystallites. Once the poly-Asp concentration rose above 200 mg/g of silk however, oriented growth along the fibres of the silk template was observed.
Proteins and Lipids
135
Figure 6.8 – Lattice matching between the a-b plane of HAp and the b-c plane of the nanocrystalline domain of the spider silks as well as the hydrogen bonds between them. (a) Unit cell of HAp crystal structure, which belongs to a hexagonal crystal system. The highlighted plane (in red) in part a is the a-b plane of HAp shown in part b. (c) Unit cell of nanocrystalline regions that are made of oriented β-sheets in the spider silk. The highlighted (in black) plane in part c is the b-c plane of silk proteins shown in part d. (e) Superimposition of the a-b plane of HAp (in red) and the b-c plane of the silk protein (in black), indicating that the a-b plane of HAp and the b-c plane of protein nanocrystals match each other approximately and that OH- perpendicular to the a-b plane of HAp can form hydrogen bonds with the O in the CdO group (i.e., O···H-O) of the β-sheet of silk protein nanocrystals. Reprinted with permission from26. Copyright 2007 American Chemical Society.
136
Biotemplating
In addition to its role as a director of crystallization, the fibrous nature of silk has been exploited as a sacrificial morphological template in a similar manner to collagen. Kunitake et al. used volatile butoxides of titanium and zirconium in order to selectively coat fibres of silk27. After soaking silk fibres in a solution of the butoxide, the fibres were removed and the volatile phase allowed to hydrolyze and condense in the moisture of the air. This cycle of soaking and exposure to air was repeated ten times to ensure a complete coverage. Calcination of the composite material at 450 °C for 4 hours resulted in the removal of the silk template and a consolidation of the inorganic phase. SEM investigations in the case of titania revealed a porous structure of filaments approximately 5 µm wide. This compares to the original width of the silk fibre of 9 µm, indicating a appreciable degree of shrinkage on mineralization. The titania coating was not even however, giving rise to an hierarchical pore structure comprising large pores of around 100 nm in diameter and smaller pores of approximately 10 nm in diameter. The group note that the porous structure of the silk may be aiding the production of porosity in the titania in addition to the intra-crystallite pores produced by discrete crystals of titania growing apart during the condensation step. Electron diffraction of these titania crystallites revealed that they were well crystallized anatase (Figure 6.9). Zirconia templated silk fibres resembled the titania fibres, albeit with a poorly-crystalline coating as zirconia is amorphous at the calcination temperatures used in the study. For both titania and zirconia, the group demonstrated their use as nano-reactors for the controlled growth of gold nanoparticles, producing nanoparticles of gold of sub-10 nm diameter with small standard deviations. These mineralizations of silk rely on the transformation of the amorphous silk protein to a β-sheet conformation. In doing so, the silk adopts the well-defined fibrous structure seen when this transformation occurs in nature. Mann et al. however, induced this transformation after the silk protein was incorporated within the voids of an opaline colloidal crystal, to create a silk inverse opal replica28. The replicas were synthesized by capillary infiltration of preformed colloidal crystals of 500 nm-diameter polystyrene microspheres with an aqueous 8%wt solution of regenerated Bombyx mori silk fibroin. The group used polystyrene for the creation of semi-regular opal structures in order to make use of the hydrophobic-hydrophobic interactions between the polymer and the extensive hydrophobic domains on the protein.
Proteins and Lipids
137
Figure 6.9 – SEM images of (a) titania filaments, (b)–(d) single filament at higher magnifications, and (e) titania helix. Insert in (a): photo of silk-templated titania. Reprinted with permission from27. Copyright 2004 American Chemical Society.
Once the opal monoliths had been completely infiltrated with the silk solution, the samples were air dried and immersed in ethanol to induce β-sheet transformation. Owing to the physically constrained volume within the opal, the silk transformation followed the surfaces of the polystyrene spheres with a high degree of accuracy. Once dry, the polystyrene spheres were able to be removed by washing the composite material in toluene. SEM images showed that the replication was of high quality, with the production of periodic arrays of spherical voids in a silk matrix (Figure 6.10). The voids were of the order of 400 nm in diameter, surrounded by a continuous silk framework approximately 50 nm thick. The production of 400 nm diameter spherical voids from 500 nm polystyrene spheres indicate that a small degree of shrinkage occurred either during removal of the polystyrene template, or by beam induced heating whilst under SEM observation.
138
Biotemplating
Figure 6.10 – SEM images of a silk inverse opal prepared using 8 wt% regenerated fibroin, (a) low magnification, and (b) high magnification. Scale bars = 1 µm. Reproduced by permission of The Royal Society of Chemistry.
The group found that the silk inverse opals were elastic and able to withstand and recover from compressive loads of up to 112 MPa without loss of pore structure. Although no photonic studies were undertaken, it is likely that these
Proteins and Lipids
139
structures could be used in optical devices which require tuneable band gaps. In this case, compression of the monoliths will result in the exclusion/admission of electromagnetic radiation of different wavelengths, depending on the direction and degree of compression. The group also determined that the silk monoliths were superhydrophobic, with water droplet contact angles in the region of 140°. This they attribute to the ‘lotus effect’ of the periodic rough surface texture of the inverse opals, aiding the repulsion of water from the already hydrophobic silk material. They note that this behaviour would make the silk inverse opals suitable for applications as biocompatible elastic scaffolds, storage–release, barrier and self-cleaning materials, and in the design of load-responsive microfluidic devices. After fibrous proteins, the other major group of structural proteins are the globular proteins. Adopting spherical or near-spherical morphology, these proteins are particularly effective in the uptake and sequestration of metal ions from solution. The structure of a globular protein is determined by the orientation of hydrophillic, polar R-groups on the peptide residues. When the protein folds, it is these groups which are disposed on the outer surface of the assemblage, where they are able to interact with the polar aqueous environment in vivo. Similarly, hydrophobic residues will be oriented toward the inside of the protein assembly, minimizing contact with the aqueous environment in which the protein finds itself. The creation of a protein which is water soluble and has a well-defined hydrophobic inner cavity enables the uptake and sequestration of ions from solution and the possibility of chemical transformation to take place which would not normally be possible in the aqueous environment surrounding the protein. In this regard, globular proteins find use as enzymatic catalysts, ion transport vectors and even as antibodies in the immune system of organisms. In many animals, it is the globular proteins myoglobin and haemoglobin which perform the most critical of physiological functions, namely the transportation of oxygen from the lungs to the cells of the body. Oxygen, being non-polar and largely insoluble in blood plasma is able to be sequestered by these globular proteins by chelation to an iron (II) centre in the protein. Intimately linked to the smooth functioning of the vascular system is another globular protein, ferritin. Of all the globular proteins, it is ferritin which has found the most use as a biotemplate owing to its stability over a wide range of pH and temperatures. Ferritin is used as a sequestration and transport device for iron, the accumulation of which in the blood plasma would be toxic. The protein resembles a hollow sphere around 12 nm in diameter, with the internal cavity of
140
Biotemplating
the order of 9 nm in diameter. When in vivo, the cavity is filled with a core of hydrated iron (III) oxide. The sequestration of iron by ferritin is possible due to the formation of channels in the globular protein, enabling ion transport to take place. It is the presence of these channels which has made ferritin a particular favourite for biotemplating studies, as it is possible to reductively dissolve the iron core and extract the solubilized iron through these channels. This nonmetallized, hollow protein capsid is called apoferritin and as it still retains metallophilicity, so can be subsequently remineralized. Most notably the apoferritin core has been used as a biotemplate for the sequestration and growth of manganese oxide29, uranyl oxide30 and the semiconductor cadmium sulfide31, all of which grow as uniform nanoparticles based on the geometry and dimensions of the hollow core of the ferritin protein assembly. One of the first remineralization studies on ferritin was by Mann et al. who investigated the replacement of the iron (III) oxide core by a magnetic form of iron, magnetite (Fe3O4) 32. This was a logical procedure to undertake, as the ferritin protein assembly is particularly siderophillic, so re-growth of an iron phase was always going to have a high probability of success. Experimentally, the procedure was a two-step process involving the demineralization then remineralization of ferritin. First, native ferritin from horse spleen was dialysed against a buffered solution of thioglycolic acid in order to dissolve the iron cores, whilst leaving the protein assemblage intact. Once complete, the apoferritin molecules were found to be quite stable in quiescent storage at low temperatures (5 °C) prior to remineralization. The second step was achieved by incubating the apoferritin with buffered solutions of trimethylamine oxide and ferrous ammonium sulfate which were added dropwise at a temperature of 65 °C over a period of approximately three hours at pH 8.6. The resulting magnetoferritin can be considered a ferrofluid and as such has found applications as a magnetic resonance contrast agent33. In another study by Mann et al., the group looked at the effect of controlling the size of the iron core on the magnetic properties of the ferrofluid34. They found that by carefully controlling the addition of the reagents and more importantly, introducing them in a step-wise manner, they were able to limit mineralization to the apoferritin core and minimize magnetite growth in bulk solution and on the surface of the capsid. The iterative addition allowed the group to terminate the supply of reagents at any time, thereby allowing them to finely control the size of the eventual nanoparticle obtained from 4 nm to 14 nm. SQUID magnetometry of the nanoparticles revealed that although the group intended to synthesize
Proteins and Lipids
141
magnetite, the magnetic behaviour indicated that the nanoparticles were in fact maghemite (γ-Fe2O3). Maghemite is formed by the aerial oxidation of magnetite, which the group deduced was enhanced by the nanoparticulate nature of the cores, the higher surface area leading to more rapid oxidation. Recently, ferritin has been used as a porogen in the formation of porous silica films. Kunitake et al. produced thin silica films with 12 nm pores by alternately spin coating solutions of ferritin and pre-hydrolyzed TEOS onto a polymercoated silicon wafer substrate35. Once the silica had condensed, the composite film was lifted off the substrate by immersion in ethanol and laid down on a TEM grid. Treatment of the composite film with UV-ozone for 15 minutes resulted in the decomposition of the ferritin protein and immersion in 1M HCl for a few minutes removed the iron cores. Whilst not exhibiting any periodicity, templating the pore structure with ferritin ensured a very tight range of pore diameters and as such, represents a relatively simple way to synthesize thin films with well-defined pore dimensions.
6.2 Lipids Another closely related group of molecules which exhibit complex behaviour in nature are the lipids. Lipids are prevalent as membrane-forming molecules in all biological membranes. What makes the lipids particularly suitable in this role is that their molecular structure affords regions of hydrophilicity and hydrophobicity, thereby allowing for the creation of volumes of controlled size.
6.3 Structure and properties Lipids are characterized by their ability to be solubilized in fats (triglycerides) and include a wide range of molecules such as sterols, vitamins A, D and E and phospholipids. Owing to their wide compositional variations, the lipids have found many diverse functions within organisms, most notably as energy storage media and in the maintenance of structural stability in cell membranes. It is in this latter role that the lipids have inspired use as complex biotemplates for mineralization. It is the phospholipids that are of particular interest, as their molecular structure contains both hydrophillic polar and hydrophobic non-polar regions, thereby allowing the molecule to exhibit surfactant-like behaviour (Figure 6.11).
142
Biotemplating
Figure 6.11 – Molecular image of a phospholipid.
When in vivo, it is this amphiphillic behaviour which is responsible for the organization of the lipids into bilayers; the hydrophobic ‘tails’ of the lipid molecules adopting a close-packed arrangement in order to minimize contact with the aqueous environment and present the hydrophillic ‘heads’ outwards. Initially forming bilayers, further addition of lipid molecules will result in the formation of vesicles, decreasing the interaction of the hydrophobic groups with the aqueous environment. In this way, a well-defined hydrophobic pocket can be created, in which (depending on the location of the lipid body in vivo) various molecules and/or inorganic materials can be sequestered and transported. Controlling inorganic growth by use of a lipid layer was demonstrated by Wong et al. where they used a composite organic template consisting of anionic DNA and cationic lipid membranes which were found to spontaneously self assemble into a multi-lamellar structure36. The lipid component comprised binary mixtures of the neutral lipid DOPC (dioleoyl-phosphatidylcholine) and the cationic lipid DOTAP (dioleoyltrimethylammonium propane), which were mixed with an aqueous DNA solution to form the self-assembled complexes. Chains of DNA were found to arrange themselves parallel across the lipid membrane structure, forming a 1D lattice. By using lipids to induce periodicity in the DNA molecules, the group were able to organize Cd2+ ions within the interhelical pores of the DNA chains, thereby providing sites of preferential reaction. On treatment with H2S, the Cd2+ formed CdS nanorods which were all aligned with the (002) planes parallel to the negatively charged sugar-phosphate DNA backbone, indicating the fine control over the crystallographic properties afforded by the preferential growth along the DNA chains (Figure 6.12). Confirmation of the nanoparticulate nature of the CdS nanorods was afforded both by TEM and wide-angle X-ray scattering, which showed broadened peaks, indicative of nanoscale crystallite size. The group note that the pronounced tilt of the nanorods of 60° matched the orientation of the DNA sugar-phosphate backbone, which is tilted by approximately 60° with respect to the helix axis in
Proteins and Lipids
143
B-form DNA when projected onto a 2D plane. The negatively charged ridge on the DNA molecule in this orientation is responsible for the nucleation of cations of Cd2+, leading to CdS growth with a pronounced tilt.
Figure 6.12 – (a) HRTEM of a typical templated CdS nanorod (scale bar is 5 nm): Note tilt of (002) planes relative to rod axis. (b) Schematic representation of crystal structure within nanorod (Cd, red; S, yellow) showing (002) planes. (c) Schematic representation of B-form DNA, showing the negatively charged phosphate groups (green) on the backbone, which organize the Cd2+ ions and guide the nucleation of CdS. Reprinted with permission from36. Copyright 2003 American Chemical Society.
Organized lipid films created using a Langmuir-Blodgett (LB) method have been utilized many times as a simple method to constrain crystallization and provide crystallochemical control over the inorganic phase37-40. Talham et al. used this technique to control the growth of cadmium dihalide phases, using LB films of cadmium arachidate over which hydrogen halide gas was passed41. TEM with electron diffraction showed that the crystals formed in each case had preferred crystallographic orientation. In the case of CdI2, all crystals were formed with their [001] crystal axes perpendicular to the plane of the LB film. Even with the LB film up to 10 layers thick, the crystals were all oriented the same way. When the halide was bromine or chlorine, the crystals of CdBr2 or CdCl2 formed had two orientations, either with the [001] axes parallel or perpendicular to the LB plane. The group reason that there could be two mechanisms responsible for the high degree of control over crystal growth. Firstly, a simple constraint of crystal growth as the LB cadmium arachidate film organizes the cadmium ions into a planar arrangement, diffusion of the halide
144
Biotemplating
into the film will occur naturally parallel to the plane. Diffusion of the halide is limited perpendicular to the plane of the lipid film, further reinforcing the formation parallel to the plane. Secondly, as there are domains in which discrete cadmium halide particles have the same in-plane orientation, some degree of lattice matching exists between the LB template and the cadmium halide and that this lattice matching is responsible for the directed crystal growth. As the organic template is not a static entity, domain size is restricted as defects form in the film, resulting in the formation of domain boundaries. The formation of CdBr2 and CdCl2 crystals in different orientations could therefore be thought to arise from these defect sites. The formation of bilayer assemblages was also used by Sastry et al. in the production of 2D bilayer stacks of the lipid steric acid, to induce a morphological change in the crystallization of BaSO4. The group deposited thin films of steric acid on silicon wafers, followed by immersion in a aqueous solution of BaCl2 in order to effect the electrostatic binding of Ba2+ ions with the carboxyl groups of the steric acid, effectively forming barium sterate42. Once barium uptake was complete, crystallization of BaSO4 in situ was performed by immersion of the films in dilute sulphuric acid. FTIR studies on the film at various stages of growth showed a prominent absorption seen at 1700 cm-1 in films immersed in BaCl2, indicative of the carbonyl stretch of carboxylic acid groups in the steric acid. After exposure to sulphuric acid, the carbonyl stretch disappeared as barium sulphate was formed. Before exposure to sulphuric acid, the group determined that the barium sterate was following the lamellar profile of the lipid, by the appearance of a set of c-axis plane reflections from (004) to (0010), characteristic of a lamellar morphology. Once formed however, the crystals of barium sulphate did not appear lamellar, but rather were of an elongated, hexagonal morphology. The group observe that although the precise nature of crystallization is not certain, it is likely that there are barium sterate sites where nucleation and growth of barium sulphate occurs, and whilst not replicating the underlying structure of the lipid bilayer, leads to outgrowth along preferred crystallographic directions. The flower-like barite crystals produced were argued to be likely due to hydrophobic interactions between individual crystallites covered by a monolayer of stearic acid. Another common morphology adopted by lipids is the lipid tubule, as amphiphilic behaviour of lipid bilayers lead naturally to the formation of curved surfaces in order to minimize unfavourable lipid-solvent interactions. By curling up in this manner, the lipid tubule is a ready-made hollow cylindrical template of
Proteins and Lipids
145
nanoscale dimensions. Many common inorganic phases have been templated by lipid tubules, among the first of which was silica, reported by Hofman et al. in the early 1950s43. Subsequent researchers have varied the dimensions of the tubes by assembling multi-layered lipid structures. A good example of the most usual synthetic approach is given by Shimizu et al. In their work, a simple lipid was synthesized to consist of tri-proline and glutamic acid dialkyl amide groups44. When added to water at a concentration of 5 mg per ml, and subjected to ultrasonication, the lipid formed tubules. The addition of TEOS to the lipid solution at a concentration of 50 mg per ml resulted in the formation of a gel after a week of quiescent storage. TEM of the tubules (stained with phosphotungstate) prior to TEOS addition revealed that they were multi-micron in length and approximately 200 nm wide. After gelation, TEM images of the composite material revealed that the gel consisted of high-yields of hollow nanotubes with a very tight size distribution. The nanotubes were all between 5 and 10 microns in length, with a 200 nm diameter and 20 nm wall thickness, indicating that the silica produced a particularly fine replication of the lipid. In many studies, calcination of the composite material to remove the lipid at this point in the synthesis results in a marked structural collapse of the silica replica. In this work however, Shimizu et al. lyophilized the aqueous gel under vacuum (1 Pa) for 48 hours prior to calcination. By forming a xerogel in this manner, the structure of the lipid could be preserved, as the destruction of morphology by outgassing during calcination is minimized. The wall thickness of the calcined nanotubes was now determined to be approximately 8 nm, indicating a degree of shrinkage, although overall morphology was retained. A more complex form of siliceous material was investigated by Mann et al. in the formation of an organoclay around a chiral lipid template. Organoclays are of particular interest, as their enhanced functionality by the incorporation of reactive functional groups into clay matrices, allow for a greater range of sequestration and drug release applications than for simpler, non-functionalized clays alone. Mann et al. used a chiral phospholipid, 1,2-bis(10,12-tricosadiyonyl)-sn-glycero3-phosphatidylcholine (DC8,9PC) as the templating phase for the creation of complex organoclay structures, owing to the ability of DC8,9PC to form helical ribbons and fibres by self-assembly45. A dispersion of the self-assembled lipid structures in an ethanol-water solution was able simply to be mineralized by the addition of 3-(2-aminoethyl-3-aminopropyl)trimethoxysilane (EDTMS) and magnesium chloride, producing high-fidelity replicas of the lipid consisting of Mg-phyllo(organo)silicate clay materials. The high quality of the replication was
146
Biotemplating
confirmed by TEM, which showed helical ribbons or tubules of an electron-dense material 0.55 µm to 0.8 µm in diameter, with lengths greater than several microns (Figure 6.13).
Figure 6.13 – TEM data on organoclay tubules. Samples prepared by treatment of organoclaycoated microtubules with ethanol for 24 hours showing (a) intact, open-ended tubule, (b) tubule with helical ribbon morphology and (c) tubule-within-tubule morphology. (d) Sample prepared in ethanol for 10 days showing increased wall thickness. All scale bars are 500 nm. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
XRD confirmed that the inorganic phase was indeed the desired phyllo(organo)silicate clay, but that also peaks due to the lipid lamellar mesophase were seen, confirming the retention of complex ordered structure in the replica material. From close examination of the TEM images, the inner and outer wall of the tubules was able to be identified, now decorated with the organoclay. This revealed that the wall thickness of the material was 35 nm,
Proteins and Lipids
147
which corresponded to a six lipid bilayer wide assembly. When the group increased the immersion time of the lipid in EDTMS, thicker structures of up to 60 nm were observed. This suggested to them group that clay clusters were interacting with the zwitterionic phosphocholine headgroups of the lipid on both inner and outer walls of the tubules. Changing the order of reactant addition produced similar assemblages of mineralized lipid tubules, which indicated a general electrostatic mechanism of mineralization, rather than a more active role of EDTMS as a facilitator of nucleation and growth of tubule assemblies. In a similar manner, Chappell et al. used DC8,9PC as a lipid template on which to grow aluminium phases, citing a marked affinity of the lipid for Al3+ cations46. Simple addition of aluminium chloride to a ethanolic solution of DC8,9PC produced an even coating of an aluminium-containing phase. The group were unable to obtain an electron diffraction pattern for this material, suggesting that it was a poorly or non-crystalline phase of an aluminium hydroxide-like material. The group found however, that these aluminium-coated tubules were eminently suitable on which to subsequently grow crystalline aluminium phases. Redispersion of the aluminium coated tubules in ethanol, followed by the addition of sodium bicarbonate produced in some instances a crystalline solid, which replicated the tubule structure completely. Again, the group found that electron diffraction and this time XRD produced inconclusive results, although a tentative assignment of the crystalline phase as a sodium hydroxyaluminium carbonate was made. The same DC8,9PC lipid was used by Price et al. in the creation of copper nanospirals47. In this work, use was made of the seams created by ribbon formation on the lipid, rather than the body of the lipid itself. It had been noted in previous studies on the interaction between DC8,9PC lipids and nanoparticulate metals that colloidal particles would adsorb preferentially on the seams of a tubule48. They note however, that as the interaction between the lipid and the metal occurs through van der Waals and hydrogen bonding, there is invariably enough of the metal adsorbing to the lipid walls to result in complete coverage of the lipid. It is likely that the enhanced binding of colloid at lipid seams is due to the preferential reduction of the metallic species by diacetylenic groups exposed at the edges of the lipid bilayer. Price et al. used this and other observations in order to restrict metallization to the lipid seams by the introduction of a polyelectrolyte which enhanced the specificity of the metal templating phase for the seam areas. By sequentially adsorbing oppositely charged polyelectrolyte layers of sodium
148
Biotemplating
poly(styrenesulfonate) (PSS) and poly(ethyleneimine) (PEI), onto lipid tubule seams, the binding of Pd(II) nanoparticles was able to be achieved preferentially at these sites owing to the strong electrostatic affinity of the negatively charged Pd(II) nanoparticles for the terminal positively charged PEI layer of the multilayer polyelectrolyte/lipid assembly. Areas of Pd were then utilized as catalytic sites for the preferential deposition of copper onto the composite assembly. SEM and TEM images of the copper-coated composite showed that after prolonged (2 hours) exposure to the copper deposition, some spiral structure are formed which mimic the underlying spiral structure of the lipid seams. The group note however, that the spirals are not produced in high yield, with a significant proportion of irregular-shaped metallic structures formed by nonspecific interactions between the lipid assembly and the copper. The group were able to optimize the preferential binding by increasing the charge density at the lipid seams. They achieved this by using a mixed lipid starting material, comprising DC8,9PC lipid doped with 2% by weight of a negatively charged lipid, 1,2-bis(tricosa-10,12-diynoyl)-sn-glycero-3-phosphohydroxyethanol (DC8,9PEOH). The increased negative charge density by having DC8,9PEOH lipid preferentially at the lipid tubule seams further increased the specificity of polyelectrolyte adsorption and subsequent Pd(II) and copper metallization. The success of replication of the lipid structure was determined by measurement of the pitch angle of the copper nanospirals. The group found that a pitch angle of 48° ± 9° in the copper nanostructures correlated well with that of the starting lipid (42° ± 6°). Anisotropic assemblages of different inorganic materials have also been realized using lipids. An early realization of this was by Mann et al. who used lipid tubules as a template for the formation of magnetic and non-magnetic iron oxides49. In other work, Shimizu et al. reported on the templated growth of vanadium oxide, titania and tantalum oxide using a self-assembled lipid tubule system which contained secondary ammonium sites as positively charged headgroups50. Their lipid assembly produced tubular structures with outer diameters of 80 nm and very thin walls, corresponding to only one bilayer in thickness. Addition of an ethanolic solution of titanium isopropoxide Ti(OPr)4 to an aqueous dispersion of the lipid did not produce templated mineralization, merely bulk condensation of titania which rapidly threw a sediment. Ti(OPr)4 will rapidly undergo hydrolysis and self-condensation when in contact with water and so the presence of large amounts in which the lipids are dispersed understandably resulted in uncontrolled titania formation. The group solved this
Proteins and Lipids
149
problem by carrying out the mineralization with vitrified lipid nanotubes in ice. They suggest that the presence of ethanol in the system at 0 °C results in a slow permeation of the alcohol throughout the ice, carrying the titania throughout the structure. Encountering an unfrozen film of water around each tubule would then induce condensation of the titania only in those regions in intimate contact with the tubule, thereby effectively constraining mineralization to the tubule surfaces. After lyophilization of the vitrified composite material, SEM revealed that there were a large number of intact tubular structures. Calcination of these structures at 500 °C resulted in pure titania hollow tubes. The group were also able to apply their methodology to the synthesis of tantalum and vanadium oxides, producing hollow tubes of the oxide in each case. True complexity arises from the organization of the polysaccharides, proteins and other molecules we have encountered thusfar, by living organisms. At ambient temperature and pressure, organisms are able to assemble macromolecular structures with a level of complexity which far exceeds that attainable (thusfar) in the laboratory. Much can be made therefore of these preformed structures as templates, rather than relying on the molecules themselves as templates. The remaining sections of this book deal with these ‘higher biological constructs’.
6.4 References 1. 2.
3. 4.
5. 6. 7.
Rayment, I. et al. Structure of the Actin-Myosin Complex and its Implications for MuscleContraction. Science 261, 58-65 (1993). Spudich, J.A. & Watt, S. Regulation of Rabbit Skeletal Muscle Contraction. 1. Biochemical Studies of Interaction of Tropomyosin-Troponin Complex with Actin and Proteolytic Fragments of Myosin. J. Biol. Chem. 246, 4866-4871 (1971). Cohen-Hadar, N. et al. Monitoring the stability of crosslinked protein crystals biotemplates: A feasibility study. Biotechnol. Bioeng. 94, 1005-1011 (2006). Cvetkovic, A. et al. Quantifying anisotropic solute transport in protein crystals using 3-D laser scanning a confocal microscopy visualization. Biotechnol. Bioeng. 86, 389-398 (2004). Margolin, A.L. & Navia, M.A. Protein crystals as novel catalytic materials. Angew. Chem.Int. Edit. 40, 2205-2222 (2001). Heywood, B.R. & Mann, S. Template-Directed Nucleation and Growth of Inorganic Materials. Adv. Mater. 6, 9-20 (1994). DeOliveira, D.B. & Laursen, R.A. Control of calcite crystal morphology by a peptide designed to bind to a specific surface. J. Am. Chem. Soc. 119, 10627-10631 (1997).
150 8. 9. 10. 11. 12. 13. 14.
15. 16. 17. 18. 19. 20.
21. 22.
23.
24. 25. 26. 27. 28.
Biotemplating Lee, K., Itharaju, R.R. & Puleo, D.A. Protein-imprinted polysiloxane scaffolds. Acta Biomater. 3, 515-522 (2007). Bairoch, A. The ENZYME database in 2000. Nucleic Acids Res. 28, 304-305 (2000). Kroger, N., Lorenz, S., Brunner, E. & Sumper, M. Self-assembly of highly phosphorylated silaffins and their function in biosilica morphogenesis. Science 298, 584-586 (2002). Sewell, S.L. & Wright, D.W. Biomimetic synthesis of titanium dioxide utilizing the R5 peptide derived from Cylindrotheca fusiformis. Chem. Mat. 18, 3108-3113 (2006). Jiang, Y. et al. Facile synthesis and novel application of zirconia catalyzed and templated by lysozyme. Ind. Eng. Chem. Res. 47, 1876-1882 (2008). Kessel, S. & Borner, H.G. High rate silicification of peptide-polymer assemblies toward composite nanotapes. Macromol. Rapid Commun. 29, 419-424 (2008). Eckhardt, D., Groenewolt, M., Krause, E. & Borner, H.G. Rational design of oligopeptide organizers for the formation of poly(ethylene oxide) nanofibres. Chem. Commun., 2814-2816 (2005). Veis, A., Sabsay, B. & Wu, C.B. Phosphoproteins as Mediators of Biomineralization. Acs Symposium Series 444, 1-12 (1991). Tomczak, M.M. et al. Polypeptide-templated synthesis of hexagonal silica platelets. J. Am. Chem. Soc. 127, 12577-12582 (2005). Yu, J.G., Lei, M., Cheng, B. & Zhao, X.J. Effects of PAA additive and temperature on morphology of calcium carbonate particles. J. Solid State Chem. 177, 681-689 (2004). Bhattacharya, R. et al. Assembly of gold nanoparticles in a rod-like fashion using proteins as templates. Adv. Funct. Mater. 16, 395-400 (2006). Anderson, H.C. Electron Microscopic Studies of Induced Cartilage Development and Calcification. J. Cell Biol. 35, 81-101 (1967). Krebsbach, P.H. et al. Bone formation in vivo: Comparison of osteogenesis by transplanted mouse and human marrow stromal fibroblasts. Transplantation 63, 1059-1069 (1997). LeGeros, R.Z. Properties of osteoconductive biomaterials: Calcium phosphates. Clin. Orthop. Rel. Res., 81-98 (2002). Tomomatsu, O., Tachibana, A., Yamauchi, K. & Tanabe, T. A film of collagen/calcium phosphate composite prepared by enzymatic mineralization in an aqueous phase. J. Ceram. Soc. Jpn. 116, 10-13 (2008). Eglin, D., Mosser, G., Giraud-Guille, M.M., Livage, J. & Coradin, T. Type I collagen, a versatile liquid crystal biological template for silica structuration from nano- to microscopic scales. Soft Matter 1, 129-131 (2005). Ono, Y. et al. Preparation of novel hollow fibre silica using collagen fibres as a template. Chem. Lett., 475-476 (1999). Sponner, A. Spider silk as a resource for future biotechnologies. Entomological Research 37, 238-250 (2007). Cao, B. & Mao, C. Oriented nucleation of hydroxylapatite crystals on spider dragline silks. Langmuir 23, 10701-10705 (2007). He, J.H. & Kunitake, T. Preparation and thermal stability of gold nanoparticles in silktemplated porous filaments of titania and zirconia. Chem. Mat. 16, 2656-2661 (2004). Swinerd, V.M., Collins, A.M., Skaer, N.J.V., Gheysens, T. & Mann, S. Silk inverse opals from template-directed beta-sheet transformation of regenerated silk fibroin. Soft Matter 3, 1377-1380 (2007).
Proteins and Lipids 29.
30.
31. 32. 33. 34.
35. 36.
37. 38. 39. 40. 41. 42.
43. 44.
45. 46. 47.
151
Mackle, P., Charnock, J.M., Garner, C.D., Meldrum, F.C. & Mann, S. Characterization of the Manganese Core of Reconstituted Ferritin by X-ray Absorption Spectroscopy. J. Am. Chem. Soc. 115, 8471-8472 (1993). Meldrum, F.C., Wade, V.J., Nimmo, D.L., Heywood, B.R. & Mann, S. Synthesis of Inorganic Nanophase Materials in Supramoleculr Protein Cages. Nature 349, 684-687 (1991). Wong, K.K.W. & Mann, S. Biomimetic synthesis of cadmium sulfide-ferritin nanocomposites. Adv. Mater. 8, 928-932 (1996). Meldrum, F.C., Heywood, B.R. & Mann, S. Magnetoferritin- In vitro Syntheses of a Novel Magnetic Protein. Science 257, 522-523 (1992). Bulte, J.W.M. et al. Magnetoferritin - Characterization of a Novel Superparamagnetic MR Contrast Agent. JMRI-J. Magn. Reson. Imaging 4, 497-505 (1994). Wong, K.K.W., Douglas, T., Gider, S., Awschalom, D.D. & Mann, S. Biomimetic synthesis and characterization of magnetic proteins (magnetoferritin). Chem. Mat. 10, 279-285 (1998). Fujikawa, S., Muto, E. & Kunitake, T. Embedding of individual ferritin molecules in large, self-supporting silica nanofilms. Langmuir 23, 4629-4633 (2007). Liang, H.J., Angelini, T.E., Ho, J., Braun, P.V. & Wong, G.C.L. Molecular imprinting of biomineralized CdS nanostructures: Crystallographic control using self-assembled DNAmembrane templates. J. Am. Chem. Soc. 125, 11786-11787 (2003). Smotkin, E.S. et al. Size Quantization Effects in Cadmium-Sulfide Layers Formed by a Langmuir-Blodgett Technique. Chem. Phys. Lett. 152, 265-268 (1988). Mann, S. Molecular Tectonics in Biomineralization and Biomimetic Materials Chemistry. Nature 365, 499-505 (1993). Aizenberg, J., Black, A.J. & Whitesides, G.M. Control of crystal nucleation by patterned self-assembled monolayers. Nature 398, 495-498 (1999). Schreiber, F. Structure and growth of self-assembling monolayers. Prog. Surf. Sci. 65, 151-256 (2000). Pike, J.K., Byrd, H., Morrone, A.A. & Talham, D.R. Oriented Cadmium Dihalide Particles Prepared in Langmuir-Blodgett Films. Chem. Mat. 6, 1757-1765 (1994). Rautaray, D., Kumar, A., Reddy, S., Sainkar, S.R. & Sastry, M. Morphology of BaSO4 crystals grown on templates of varying dimensionality: The case of cysteine-capped gold nanoparticles (0-D), DNA (1-D), and lipid bilayer stacks (2-D). Cryst. Growth Des. 2, 197-203 (2002). Nemetschek, T. & Hofmann, U. Feine Fasern aus Kieselsaure. Z.Naturforsch.(B) 8, 410-412 (1953). Ji, Q. et al. Direct sol-gel replication without catalyst in an aqueous gel system: From a lipid nanotube with a single bilayer wall to a uniform silica hollow cylinder with an ultrathin wall. Chem. Mat. 16, 250-254 (2004). Patil, A.J., Muthusamy, E., Seddon, A.M. & Mann, S. Higher-order synthesis of organoclay pipes using self-assembled lipid templates. Adv. Mater. 15, 1816-1819 (2003). Chappell, J.S. & Yager, P. Formation of Mineral Microstructures with a High Aspect Ratio from Phospholipid-Bilayer Tubules. J. Mater. Sci. Lett. 11, 633-636 (1992). Price, R.R., Dressick, W.J. & Singh, A. Fabrication of nanoscale metallic spirals using phospholipid microtubule organizational templates. J. Am. Chem. Soc. 125, 11259-11263 (2003).
152 48. 49. 50.
Biotemplating Schnur, J.M. et al. Lipid-Based Tubule Microstructures. Thin Solid Films 152, 181-206 (1987). Archibald, D.D. & Mann, S. Template Mineralization of Self-Assembled Anisotropic Lipid Microstructures. Nature 364, 430-433 (1993). Ji, Q.M. & Shimizu, T. Chemical synthesis of transition metal oxide nanotubes in water using an iced lipid nanotube as a template. Chem. Commun., 4411-4413 (2005).
Chapter 7
Viruses and Bacteria
“As scientists attempt to understand a living system, they move down from dimension to dimension, from one level of complexity to the next. I followed this course in my own studies. I moved from anatomy to the study of tissues then to electron microscopy and chemistry and finally to quantum mechanics. This downward journey through the scale of dimensions has its irony, for in my search for the secret of life I ended up with atoms and electrons, which have no life at all. Somewhere along the line, life had run out through my fingers.” Albert Szent-Györgyi It is not a huge conceptual step to move from polypeptides and proteins synthesized either in the lab or by the animal kingdom to the larger constructs which are of themselves ‘living’. Although no one would consider the protein capsid ferritin as a living entity, when capsids combine with the long chained molecules RNA or DNA, the assemblage that arises is considered by many to be ‘living’. Wherever the vital ‘spark’ may be, in its simplest form, a capsid of protein containing genetic material is able to make replicas of itself by associating with a cell of the host in which it finds itself, thereby allowing the virus to evolve and adapt to its surroundings. The proteinaceous capsid which acts as a delivery vector for the genetic material can adopt many structures and the morphological complexity which results can be quite stunning. Importantly from the point of view of the scientist, these complex forms are able to be reproduced in vast numbers, with each new individual virtually identical to the other. Bacteria on the other hand can be more comfortably considered to be alive as they are entirely self-contained entities which do not require the assistance of another cell to survive and replicate. As fast growing unicellular organisms, bacteria are found in every habitat on Earth, where their high turnover enables 153
154
Biotemplating
them to very quickly adapt to changes in their environment or to environments of extreme temperature, pH, salinity or even radioactivity. Morphologically, they exhibit a wide range of shapes and sizes, from small spheres to elongated rods. The surface of a bacterium is a complex mixture of polysaccharides which are cross-linked by amino acids. As such, the surface layers of bacteria are particularly amenable to mineralization owing to the large number of reactive functional groups which are present. For both viruses and bacteria, the complex morphologies adopted by their proteinaceous or polysaccharide coats have made them attractive to scientists interested in biotemplating. All of the chemistry we have seen thusfar in the earlier chapters on these respective biopolymers can therefore be done on viruses and bacteria, this time with the added level of complexity that the macromorphology brings.
7.1 Viruses – structure and properties At its very simplest, a virus consists of genetic material, either RNA or DNA, surrounded by a protein coat. The interest for the biotemplator comes in the structures adopted by the protein capsid. There are many variations, but the main structure types can be summarized as helical or filamentous and spherical/polyhedral. Helical capsids consist of a single type of protein molecule which selfinteracts and forms a cylindrical assembly around the genetic material via electrostatic interactions. In this manner, the helical viruses are able to form highly anisotropic assemblies, sometimes many microns in length, depending on the length of the genetic material. This cooperative assembly between the protein and genetic material creates a tubular structure which, if the genetic material is subsequently removed, can be used as a constrained reaction environment, in addition to the opportunities for mineralization afforded by the outer surface of the protein coat. Spherical or more correctly, polyhedral viruses are comprised of a number of protein sub-units which self-assemble into an efficient space filling arrangement in order to encapsulate the genetic material. The interactions between the protein sub-unites is non-covalent in nature and as such, represents a relatively fluxional system which is a able to spontaneously rearrange, allowing for the delivery of the genetic material once in an appropriate cellular environment (Figure 7.1).
Viruses and Bacteria
155
a
b
Figure 7.1 – Schematic comparing the morphologies of (a) helical and (b) polyhedral viruses. The structures marked in (a) are the (1) nucleic acid, (2) capsomer (protomer) and (3) the protein capsid. The structures in (b) represent (1) penton capsomeres, (2) hexon capsomeres and (3) the viral genome. Both images were created by ‘Y_tambe’ and are freely available on Wikipedia under the Creative Commons Attribution ShareAlike 3.0 Licence.
Viruses are available in large number and monodisperse in terms of morphology and chemical reactivity. In particular, their marked resistivity to many extremes of environment make them amenable to a range of chemical reactions without appreciable loss of morphology. Moreover, the relationship between shape and chemical reactivity (the precise spatial ordering of the same reactive functional group on the protein capsid) means that viruses are able to direct mineralization in a more morphologically defined manner than the equivalent protein when in solution.
156
Biotemplating
7.2 Spherical/polyhedral viruses Morphologically, the simplest form of virus which can be used in biotemplating is the spherical or polyhedral type. The production of capsids with well defined interior dimensions and reactivity enable a range of different mineralizations to be carried out, often with particular viruses being particularly suited chemically to particular reactions. An additional advantage for the biotemplating chemist is the fact that the viral capsid presents an inner and outer surface to the environment, meaning that there exist well defined regions of hydrophilicity and hydrophobicity. One of the first studies on the mineralization of a spherical viral capsid involved the cowpea chlorotic mottle virus (CCMV). In this work, CCMV was used as it presented a heterogeneous reaction environment which directed mineralization almost exclusively to the interior of the capsid1. The CCMV capsids are 26 nm in diameter, with an inner cavity size of approximately 20 nm in diameter. The protein shell comprises 180 identical protein subunits, which spontaneously self-assemble into capsid form. On assembly, arginine and lysine amino acid residues adopt a conformation towards the interior of the capsid, thereby imbuing the interior with an appreciable positive charge. The outer surface of the capsid is not as highly charged, so by creating a charge gradient, reactions, in particular those involving anionic metallic species, will proceed almost exclusively within the cavity. Douglas et al. exploited the high interior positive charge density by incubating CCMV capsids with WO42-, VO3- and MoO42- species in pH neutral solutions. At around pH 7, the structure of the CCMV capsids is swollen, leading to an open pore structure and concomitant facile infiltration of anionic species. The group then lowered the pH of the solution to 5, resulting in the formation of polyoxometallate species which readily crystallized within the interior of the capsid. The lowering of the pH had the additional effect of de-swelling the capsid and thus closing the pore structure, effectively trapping the metal species within the interior. TEM studies revealed that discrete metal cores had formed, which were 18 nm in diameter, a close match to the inner diameter of the CCMV capsid. Staining of the TEM grid revealed that the protein capsid remained intact throughout the process, so the resultant composite has the potential to be used for the creation of isolated metal nanoparticles without the possibility of sintering increasing the overall metallic grain size if further processing/assembly is required. The group also found that it was possible to genetically engineer the
Viruses and Bacteria
157
virus so that the inner surface of the capsid was highly negatively charged, allowing for the sequestration and growth of transition metal oxides. This adaptation is covered in a later chapter. The group also investigated the assembly of a protein capsid around a pre-formed inorganic nanoparticle. They found that they were able to induce the self-assembly of subunits of the brome mosaic virus (BMV) around colloid gold nanoparticles. The inducement for protein assembly was the favourable electrostatic charge interaction between the citrate capped (negatively charged) gold and the positively charged residues on the protein subunits which normally comprise the interior surface of the capsid. Whilst encapsulation of gold in this way was possible, the yield of protein encapsulated nanoparticles was quite low. The group was able to improve the encapsulation by producing gold nanoparticles with a more negative surface by the passivation of the colloidal particles with polyethylene glycol. The surface of icosahedral viruses have also been used for the nucleation and growth of inorganic phases. The majority of the work in this area has focused on the genetic modification of the capsid surface and so will be discussed in another chapter.
7.3 Helical/filamentous viruses Adopting an anisotropic configuration is an efficient way for protein subunits to pack around the long chains of genetic material of the virus. One of the most commonly investigated, owing to its size, stability over a range of pH (3–9), temperatures up to 90 °C and organic solvents is the tobacco mosaic virus (TMV). Consisting of approximately 2,000 protein subunits which assemble around an RNA core in a ‘lock washer’ configuration, the protein forms a structure 18 nm wide and up to 300 nm in length. The core in which the RNA sits is 4 nm in diameter and can therefore be used as a constrained reaction volume on removal of the genetic material. As the chemistry of the amino acid subunits determines the tertiary structure of the protein, it is important that any mineralizing system into which the virus is introduced, is investigated for its effects on the viability of the protein and whether or not the surface charge of the capsid is altered appreciably, which may lead to spontaneous aggregation in solution. A study by Harris et al. looked into the effect of incubating TMV with CuCl2 a commonly used precursor material in the mineralization of copper compounds2. In their work, they looked at the
158
Biotemplating
hydrodynamic radius of TMV as an indicator of the aggregation of TMV capsids. They found that at concentrations of CuCl2 higher than 0.5 mM, the hydrodynamic radius of TMV increased substantially, indicating that spontaneous aggregation had occurred. Furthermore, the group observe that complete mineralization of the TMV capsids was not possible in their system at CuCl2 concentrations lower than 0.5 mM. They conclude that in a system with no other additive, it is not possible to achieve dense mineralization of TMV capsids without aggregation. Their studies were complicated by the fact that they found is necessary to replace the buffered solution containing the TMV capsids, with water in order to achieve reproducible coatings and to make the analysis of the colloidal interaction between TMV capsids simpler. No such complications were observed by Mann et al. as they were able to keep the TMV capsids in buffered solutions and yet achieve appreciable mineralization of the surface of the virion surfaces3. In their work, they were able to mineralize the surface of TMV capsids with CdS, PbS, iron oxides and silica. In the case of sulfide compounds, incubation of either chlorides or nitrates of cadmium or lead with TMV capsids, followed by diffusion of H2S gas through the system, produced nanocrystals of CdS or PbS. TEM studies of the CdS mineralized capsids revealed that the capsids were coated with a 16 nm thick crust, which lattice images revealed composed of CdS nanoparticles in the zinc blende structure, approximately 5 nm in diameter. The group note that mineralization appeared to be restricted to the outer surface of the capsid, although given the high electron density of the CdS, a clear picture of the inner surfaces of the virion could not be obtained. For PbS, the particles coating the capsid surface were much larger, approximately 40 nm in diameter and appeared either prismatic or irregularly shaped. Electron diffraction data of these nanoparticles showed that they were single domain crystals and possessed the rock salt structure (Figure 7.2). In the case of silica mineralization, a suspension of TMV was incubated in an ethanol/water mixture acidified to pH 2.5. On addition of TEOS, hydrolysis and condensation of silica species resulted in deposition of silica preferentially on the surfaces of the viral capsids. As expected for an amorphous mineral, the coating of the TMV surface was more complete and smooth, resulting in an even coating of a silica shell of 3 nm in width. The group found that in addition to an even mineralization, a number of the virions self-assembled under these conditions in an end-to-end configuration. This sort of self-assembly has been observed
Viruses and Bacteria
159
previously as a result of hydrophobic interactions between the dipolar ends of the TMV tubules.
Figure 7.2 – TEM micrographs of CdS-mineralized TMV particles. (a) Low magnification image showing dense mineral coating of the external surface of the virion, scale bar = 50 nm. Inset: Corresponding EDX spectrum (Ni and Cu peaks arise from TEM grid and specimen holder, respectively). (b) High resolution lattice image of composite nanotube showing single-domain CdS nanocrystallites (white arrows) on TMV surface. Scale bar = 10 nm. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
160
Biotemplating
In this work, the group rationalize this spontaneous assembly as being encouraged due to the acidic media causing mutual repulsion between amino acid residues. For iron oxide mineralization, the group mixed a suspension of TMV with an acidic solution of 1 mM ammonium iron sulfate and iron chloride. Addition of NaOH raised the pH of the solution to 9 and threw a brown precipitate. TEM revealed to the group that the TMV virions were coated with an irregular, poorly crystalline iron oxide, which SAED confirmed was ferrihydrite. Ageing the system for six months produced a more regular coating, with a gradual thinning of the iron oxide to around 2 nm thick. Whilst the previous work avoided (or at least did not observe) mineralization of the inner surface of the TMV capsid, with judicious choice of synthetic protocol, it has been shown that the hollow, inner tube of TMV can be used as a nano-reaction vessel. Again, Mann et al. used TMV as template, this time to encourage the deposition of Ag, Au or Pt nanoparticles4. Simple reduction of [PtCl6]2- or [AuCl4]- ions in acidic solution with TMV produced analogous results to those seen previously with other metal systems, namely the decoration of the surfaces of the TMV protein with discrete nanoparticles approximately 10 nm in diameter. However, the group found that by mixing TMV with a solution containing Ag(I) ions and allowing the whole to undergo photochemical reduction at pH 7, growth of silver nanoparticles within the hollow interior of the virion was achieved. In many instances, well separated silver nanoparticles were observed within the interior of the virion (Figure 7.3). In some cases, a ‘plug’ of uranium (from the uranyl acetate stain) was observed, unable to progress further along the interior of the TMV on encountering a silver nanoparticle. In this regard, the precipitated silver appears to be very effective at sealing off the interior of the virion. The group extended their work on TMV/silica materials, inducing TMV to form into liquid crystalline arrangements which were subsequently mineralized5. Making use of the tendency to align in strong magnetic fields, the group placed a buffered aqueous suspension of TMV inside the bore of an NMR machine for one hour. This produced a viscous gel which exhibited birefringence, indicative of nematic liquid crystalline ordering. Addition of a mixture of tetraethoxysilane (TEOS) and aminopropyltriethoxysilane (APTES) to the gel, followed by gentle manipulation of the NMR tube allowed siliceous species to infiltrate the entire liquid crystal. After quiescent storage for four days, the group ensured full condensation of the silica by placing the tube in a bath of water at 45 °C for two days.
Viruses and Bacteria
161
Figure 7.3 – TEM micrograph of ~ 5 nm silver nanoparticles grown inside the hollow channel of wild-type TMV. Arrows indicate nanoparticles that prevented the stain (uranyl acetate) from penetrating further in the cavity. One of these is magnified in the inset. Scale bar for the main image is 50 nm. (b) EDXA spectrum showing the presence of silver and uranium (stain). Supporting grid: † Cu. (c) UV-visible absorption spectrum displaying a plasmon band at 422 nm, characteristic of metallic silver colloid. Reprinted with permission from4. Copyright 2003 American Chemical Society.
Once complete, the gel was removed from the tube and dried under ambient conditions overnight. Finally, the TMV was removed from the system to leave hollow inorganic replicas by calcining the gel to 540 °C. TEM of the composite material revealed to the group that the sample comprised fragments of silica/TMV composite with pronounced lattice fringes in the silica of approximately 18 nm in width. This they interpret to be due to a periodic array of co-aligned TMV particles joined end-to-end and intercalated within a continuous framework of amorphous silica. On removal of the TMV template, the structure
162
Biotemplating
remained largely intact, leaving a mesoporous siliceous body. Porosity was determined to be regular rather than disordered, as viewed end-on, the silica replicas showed a hexagonal array of 11 nm channels, separated by silica walls of approximately 10 nm thickness. One surprising result obtained by the group came when they reduced the concentration of siloxane precursor in the system. When the concentration was lowered by over 40%, a large number of spherical TMV/silica nanoparticles were formed. The group explain this behaviour by considering that as the liquid crystalline state is partially disrupted on siloxane addition, there is the possibility for the formation of radial arrays of TMV rods anchored to the surface of primary silica nanoparticles. Russell et al. used a similar forced aggregation in their work on the ordering of CPMV virions by dispersing them at an oil/water interface and crosslinking the proteins with gluteraldehyde in order to fix the assembly in place6. This higher order assembly is a synthetic analogue of the structures found in nature, in particular of the surface layers of many bacteria, which are complex threedimensional regular assemblages of proteins. These constructs represent the next level of complexity which have been exploited as biotemplates and bring us to those templates which are of, or can only be produced by, living things.
7.4 Bacteria – structure and properties Bacteria are considered to be living, although they still lack many of the important features of multicellular organisms. Most obviously, they do not possess individual organelles, instead they are formed of a cytoplasm into which the various intercellular structures are incorporated. The main feature of bacteria is that they possess a bacterial cell wall, which can adopt a range of morphologies, depending on species, from the simple coccus (spherical) through rod-shaped bacilli to complex, tightly coiled spirochaetal species. In addition, bacteria can spontaneously self-assemble to form even more complex structure in the form of bacterial filaments and mats7, 8. The macromorphology of the bacterium is a fruitful area of biotemplate investigation, but the cell wall itself is a complex structure, made up from polysaccharides cross-linked with D-amino acids. This assemblage is termed peptidoglycan and provides a useful way of characterizing bacteria into two classes, depending largely on its thickness. Bacteria with a thick cell wall are able to retain a methyl violet stain (gram positive) whereas those with thin peptidoglycan cannot (gram negative). Over the surface of the peptidoglycan in
Viruses and Bacteria
163
many bacteria is a highly ordered array of protein (S-layer) of which serves as a further chemical barrier to the cell interior. It is this S-layer which has been intensively investigated as a template, owing to the fact that the protein sub-units will spontaneously assembly in vitro to reform the structure they adopt on the bacteria9. This structure can either exhibit oblique (p1, p2), square (p4) or hexagonal (p3, p6) symmetry depending on the species of bacteria. In forming an array, the protein subunits collectively will enclose pore structures of between 2 nm and 8 nm in diameter. One of the first studies into this intricate and regular structure as a substrate for directed mineralization was by Mann et al., who used the S-layer of Bacillus stearothermophilus and Bacillus sphaericus to form oblique and square lattices respectively, of nanoparticles of the semiconductor CdS10 (Figure 7.4).
(a)
(b)
Figure 7.4 – Substrate-assisted self-assembly of bacterial S-layer monolayers used for biocrystal templating of CdS superlattices. (a) Recrystallization of S-layers on electron microscope grids produces a nanoporous monolayer with the corrugated negatively charged inner face exposed to the external medium. (b) Corresponding orientation of a recrystallized S-layer formed under a Langmuir monolayer of the lipid, dipalmitoylphosphatidylethanol (DPPE) and transferred by horizontal dipping to an electron microscope grid. By this method, the charge neutral outer face of the self-assembled S-layer is exposed. Reprinted by permission from Macmillan Publishers Ltd.10, copyright 1997.
The S-layer protein sub-units were allowed to self-assemble on TEM grids and then subsequently incubated with a solution of CdCl2. On exposure to H2S gas, reduction of the cadmium salt occurred, forming CdS nanoparticles at well defined sites on the protein substrate. The group note that the highly specific interactions between cadmium and functional groups on the protein layer limit
164
Biotemplating
the nucleation and growth to discrete places on the lattice and mimics the underlying geometry almost perfectly. This structure has also been utilized by Sletyr et al. to induce the nucleation and growth of ordered arrays of nanoparticulate gold11. In this work, the group incubated a square lattice of S-layer protein from the bacteria Bacillus sphaericus which had previously been exposed to thiol groups in order to further facilitate the binding of gold to the protein layer. The composite was left for four days at 4 °C in order for reduction of the tetrachloroaurate gold salt to gold (0) on the surface of the protein to take place. TEM studies of the resulting material revealed that nanoparticles of gold were formed in the region of the pores of the S-layer and that the entire surface of the protein layer was covered. Engelhardt et al. used a similar method to produce a regular array of platinum nanoparticles12. In their work, they used the S-layer of the bacteria Sporosarcina ureae, which possessed a p4 symmetry and a lattice spacing of 13.2 nm. Reduction of a platinum salt in situ produced preferential growth of platinum nanoparticles in a perfect register with the underlying protein template. The platinum nanoparticles were well separated, as they had an average diameter of approximately 2 nm and were situated in the pores and gaps of the S-layer structure (Figure 7.5).
Figure 7.5 – Spatial distribution of metal clusters (black) on the protein crystal as revealed by image processing. The unit cell of the cluster lattice is marked and possesses a size of 13.2 nm x 13.2 nm. Reproduced by permission of Springer Publishing Co. Ltd.
Viruses and Bacteria
165
A titania material has also been synthesized in a regular array using an S-layer protein as substrate13. Moore et al. deposited S-layer protein from the bacteria Sulfolobus acidocaldarius onto a silicon wafer, allowing the protein to selfassemble to form a p6 hexagonal array. Very simply, the group then shadowed the composite at an angle of 40° by electrodeposition of titanium in vacuo. On exposure to the air, the titanium spontaneously oxidized to form a 3.6 nm thick film. When the group used ion milling to etch away at the surface of the titania, they arrived eventually at a thin sheet which contained periodic arrays of pores with the same periodicity as the underlying S-layer. When thinner layers of titanium were deposited on the protein, ion milling resulted in the formation of nanoparticles of titania also with the same periodicity. It was shown subsequently, that in addition to performing reduction/oxidation of a metal species on the surface of the S-layer, pre-formed metal nanoparticles could be preferentially adsorbed to the protein, again making use of preferred sites of binding on the protein surface. Hall et al. demonstrated this first, with the creation of regular arrays of gold nanoparticles on the S-layer of the extremophilic bacterium Deinococcus radiodurans14. The S-layer of this particular bacillus consists of a hexagonal p6 arrangement of proteins with a lattice constant of 18 nm. The S-layer protein was allowed to self-assemble on a TEM grid, so that the hydrophilic outer surface of the protein layer was uppermost. Incubation of this layer was then achieved by the simple exposure of the TEM grid to a drop of citrate capped colloidal gold nanoparticles of a range of diameters between 5 nm and 20 nm. It was found that in the case of 5 nm and 10 nm colloidal gold, regular arrays of nanoparticles were produced which replicated the underlying hexagonal periodicity of the S-layer (Figure 7.6). Owing to the difference on size of the nanoparticles, the interparticle spacing could be controlled to a certain degree. Where 5 nm and 10 nm gold colloids were bound to the S-layer, the edge-to-edge separation was 13 nm and 10 nm respectively. What was of particular note was that the interparticle separation of the 10 nm gold colloids were larger than expected because the mean size of the particles assembled on the HPI layer (8.0 nm) was significantly smaller (17.5% less) than the mean diameter (9.7 nm) of particles in the corresponding sol. This was not the case for superlattices containing the 4.8 nm-sized particles, which suggests that the HPI layer exhibits a degree of preferential binding for the smaller nanoparticles for size distributions centred around 10 nm. When nanoparticles greater than the periodic lattice parameter of 18 nm were incubated with the S-layer, binding was observed, but without the periodicity observed for smaller colloids.
166
Biotemplating
Figure 7.6 – (a) Low magnification TEM image showing micrometre-sized domains of ordered 5 nm gold nanoparticles. Scale bar 100 nm. (b, c) High resolution TEM images showing twodimensional hexagonal arrays of Au nanoparticles with mean sizes of approximately 8 and 5 nm, respectively, formed by site-specific organization on the outer surface of self-assembled HPI layers. Vectors a and b show an average centre-to-centre spacing of 18 nm (±1 nm) and a^b ≈ 60°, in register with the underlying p6 lattice of the HPI layer (Scale bars are 7 and 10 nm, respectively). Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
Viruses and Bacteria
167
This suggests that there is still preferential electrostatic interaction between the colloid and S-layer albeit without the site specificity of the smaller nanoparticles. It was further observed that aggregation of the larger gold colloids occurred on the S-layer, often with the formation of chains of nanoparticles up to 20 particles in length. A similar effect was observed when the surface charge of the colloidal gold was cationic; binding to the protein layer, but without the one to one register with the underlying periodicity. It was subsequently demonstrated in an extension to this work by Bergkvist et al. that the gold nanoparticles were bound to the vertex points where each hexameric protein cone met15. The group found that due to repulsive forces between the citrate-capped gold nanoparticles that adsorption of the gold tended to be at every second vertex point. This was confirmed as it was observed that when increasing the ionic strength of the system, thereby lowering the mutual repulsion between nanoparticles, adsorption of the gold nanoparticles occurred at each vertex point. The minimization of interparticle repulsion was also achieved by the same group in the case of deposition of platinum onto S-layers by encapsulation of the metal nanoparticle within a dendrimer. Deposition of metal nanoparticles has also been achieved by using another bacterial protein complex structure, namely the bacterial enzyme lumazine synthase. Weinkauf et al. successfully grew gold and silver nanoparticles on the surface of this globular protein via a vacuum condensation approach16. In doing so, they discovered a specificity for different sites on the protein’s surface, a specificity which was replicated even for capsids from different bacteria. The group found that gold would preferentially nucleate and grow on areas of the protein which consisted of polar but uncharged amino acid residues, whilst silver would form around the pore structures of the capsid. These pore structures were found to be rich in charged and hydrophilic residues and were therefore attractive sites for metal nucleation. These observations lead the group to the conclusion that for both gold and silver, no binding was observed at hydrophobic sites on the protein’s surface which lacked both polar and charged residues. The interior of the lumazine synthase capsids are also amenable to mineralization, in much the same way that the globular protein ferritin was successfully used to form nanoparticles by sequestration and growth of metal species. This was demonstrated by Shenton et al. in the growth of iron oxide17. In a very simple experiment, the group took buffered solutions of lumazine synthase and incubated them with aliquots of deaerated 1 mM ammonium iron sulfate. Slow addition of the iron solution over a prolonged period ensured that
168
Biotemplating
uncontrolled mineralization was kept to a minimum. Over a period of 48 hours, the solution was stirred in air, to induce the formation of iron oxide. The group found that although the control (no capsids) solution threw a red-brown precipitate, solutions containing protein remained clear and free from sediment, indicating that iron binding to the protein was successful. TEM images confirmed that the iron mineralization took place inside the capsid cavities, producing iron nanoparticles of approximately 13 nm in diameter. Electron diffraction of the cores showed that the iron was present as lepidocrocite (γ-FeOOH). The group confirmed that the protein capsids were still intact by staining the samples on TEM grids with uranyl acetate, which showed the presence of capsids of up to 30 nm in diameter. As well as proteins produced by bacteria, the bacteria themselves can be employed in the creation of complex mineral morphologies. Certain common soil bacteria are known for their ability to sequester heavy metals, even to the extent that they are able to precipitate metal particles from their salts. A good example is the bacteria Ralstonia metallidurans, a bacteria able to produce micron sized gold particles when in the presence of dissolved gold salts such as gold (III) tetrachloride18. The affinity of bacteria for inorganic salts has been exploited by Mann et al. who used a mutant form of the bacteria Bacillus subtilis to create threads of cells many tens of centimetres long19. When in culture, this mutant strain is not able to undergo cell separation after reproduction, instead producing elongated, tangled filaments. These filaments are able to be drawn out to form a multi-bacterial thread around 0.2 mm wide and to virtually any desired length. Once formed, the group found that these threads were able to be dried, then rehydrated with colloidal solutions of either magnetite, titania or cadmium sulfide. The adherence of the inorganic phase and interpenetration of the solutions within the thread was determined to be most successful when the nanoparticles possessed an overall negative charge (magnetite) whereas neutral colloids (CdS) nucleated at the surface of the thread. In contrast, the positively charged TiO2 colloidal nanoparticles showed no penetration into the body of the bacterial thread, but managed to produce an even coating on the surface. Unfortunately, when the group attempted to remove the bacterial thread template, the inorganic superstructure underwent structural collapse. This they attributed to the fact that the inorganic phase was not continuous throughout the bacterial thread, either surface bound only in the case of titania and CdS, or as non-sintered grains in the case of magnetite. The same group had found however, that silica posed no such problems, owing to the total infiltration by silicate
Viruses and Bacteria
169
species and the subsequent formation of a fully-condensed silica network throughout the bacterial thread assembly. On calcination at 600 °C, the bacterial template was removed, leaving a silica replica which was highly porous. The pore structure consisted of 500 nm wide channels separated by silica walls of between 50 nm and 200 nm in thickness. Addition of a surfactant to the synthesis produced a greater fidelity of replication of individual bacterial filaments within the thread and resulted in the silica particles having mesoporosity. As well as amorphous silica and mesoporous silica, the group found in another study that the bacterial threads could be successfully used to immobilize zeolite nanoparticles. The affinity for negatively charged oxide precursors was used by Zhang et al. to form ZnO structures with spherical form using a coccus bacterium as template20. In their work, they incubated the bacterium Streptococcus thermophilus in a solution of zinc acetate and triethanolamine. Adjusting the pH of the system to 9 and stirring the whole for three hours at 90 °C induced the precipitation of zinc oxide on the surfaces of the bacteria. Once retrieved and dried in vacuo, the coated bacteria were calcined at 600 °C to remove the organic part of the composite. The group found that XRD confirmed the structure of the inorganic replicas as ZnO (wurtzite structure) with broadened peaks indicative of nanocrystallinity. SEM studies showed that the original spherical morphology of the bacteria was retained in the calcined products, with diameters between 500 nm and 900 nm. TEM studies on the replicas suggested to the group that the replicas were hollow and that the shell thickness was of the order of 200 nm to 400 nm in width. The group determined that the shells were indeed composed of nanocrystallites of ZnO approximately 30 nm in diameter. The group rationalize the mechanism of mineralization as a coordination and/or electrostatic interaction between the functional groups of the bacterial peptidoglycan and anionic zinc species in solution. The group use triethanolamine as a promoter to further encourage zinc binding to the surface of the bacteria and suggest a two-step process for the mineralization. Firstly, triethanolamine is thought to adsorb onto the carboxyl, hydroxyl and phosphoryl groups of the cell walls, thereby providing a more amenable surface for the zinc acetate to bind to. Once zinc is bound, the triethanolamine can then be considered to behave as a base, reacting with bound zinc acetate to form the first layer of zinc oxide. The formation of nanosized zinc oxide crystallites will then provide sites of preferred nucleation for zinc species, thereby enabling the growth of crystallites through ripening processes.
170
Biotemplating
The group determined that after calcination, the replicas possessed a type-IV isotherm indicative of mesoporosity, with a bimodal pore structure consisting of small pores 2.5 nm in diameter and larger pores 11 nm in diameter. The group rationalize that the small pores are likely to have been formed by spaces between ZnO crystallites and the larger pores from gaps produced by the incomplete surface passivation of the bacteria by triethanolamine.
7.5 DNA In addition to work done with DNA and lipids in co-assembly, the DNA molecule itself has been used directly as a template to form anisotropic structures. The amount of work that has been done on the templating of the DNA molecule is vast and it is beyond the scope of this review to cover them all in detail, so the reader is directed to some recent reviews on this subject for an indepth treatment21-24. Some of the most recent work on the templating of DNA is presented here however, to give a flavour of the current state of the field. Silica has been successfully grown around DNA to produce nanowires. Numata et al. were able to achieve this, but only by solving a couple of critical organic-inorganic compatibility issues25. Firstly, in neutral or alkaline media, both DNA and the silica precursor TEOS are anionic. This is a significant problem, as charge complimentarity is the major driving force in mineralization. Secondly, TEOS is insoluble in water, whilst DNA is virtually insoluble in any solvent other than water. The group solved both of these issues by replacing the sodium counterions associated with the DNA molecule with a guanidinium/ammonium complex group, thereby rendering the surface of the DNA cationic and imbuing the molecule with a certain degree of amphiphilicity. This allowed the DNA to be solubilized in a chloroform:methanol mixture, into which TEOS was incorporated. Sol-gel hydrolysis and condensation of the TEOS produced rod-like silica structures of up to 1 mm in length. The group confirmed that DNA was acting as a template via UV-vis spectrometry of the product, which showed a broadened spectrum, but which contained the basic spectral features of the cationic DNA solution, indicating that the DNA was still present encased within the silica matrix. TEM images of the calcined silica nanorods proved that they were hollow, with an inner core diameter of 10 nm, very close to the diameter of the cationic DNA (7 nm). Calcination of the composite structures removed the DNA successfully, with little or no structural collapse of the tubular structure.
Viruses and Bacteria
171
Another alteration of the DNA molecule prior to mineralization was demonstrated by using palladium nanoparticles as catalytic sites of nucleation and growth of cobalt26. By complexing Pd(II) to the DNA molecule, the group were able to reduce the metal in situ to Pd(0) with dimethylamine borane to produce Pd nanoparticles which decorated the surface of the DNA. On incubation in a CoCl2/boric acid/sodium citrate bath, cobalt nanoparticles were able to be grown preferentially on the palladium nanoparticles. After incubation, the group found that the DNA molecules had been completely metalized, resulting in cobalt nanowires between 40 nm and 60 nm in width and many microns in length. In control experiments (without decoration of the DNA by palladium) the group were unable to produce cobalt metallization of the DNA molecules; the solution remaining pink throughout, indicative of un-complexed hydrated cobalt species. The sol-gel approach was employed in a rather simpler method to achieve the coating of DNA by titania. By fixing DNA molecules on a flat substrate, Fujikawa et al. were able to deposit drops of ethanolic titanium n-butoxide (Ti(OnBu)4) which effectively formed a titania covering over the entire surface27. Once dry, the composite material was calcined to leave titania wires with diameters between 10 nm and 180 nm. DNA has also been used to infiltrate the gallery spaces of certain clays. In doing so, the clay provides an ‘armour-plating’ to the DNA molecule, rendering it much more resistive to thermal and chemical denaturation. This was achieved by the addition of DNA to an aminopropyl-functionalized magnesium phyllosilicate (AMP) clay suspended in water28. Prior to addition, the clay is exfoliated by sonication, resulting in nanosheets and clusters of clay layers, thereby facilitating the adsorption of DNA directly into the gallery spaces. Once incorporated, the composite material formed lamellar structures and highly elongated nanowires and nanofilaments that were flexible and continuous over several microns in length. The group determined that these filaments were comprised of individual nanowires between 15 nm and 30 nm in width. On drying, the clay materials were found to still contain DNA by FTIR spectroscopy (which showed characteristic absorbances for both DNA and AMP clay) and by XRD, which showed that the AMP gallery spacings had increased to 3.6 nm from 1.6 nm. The fact that DNA was so effectively trapped within the clay gallery spaces was demonstrated by recording the melting temperature profiles of free and intercalated DNA molecules at 260 nm via a UV-vis spectrophotometer. In each instance, the group obtained sigmoidal melting curves, indicative of denaturation of the DNA double helix, but the temperature at which melting
172
Biotemplating
occurred differed significantly. In free DNA, the melting temperature was determined to be 63 °C, whereas in intercalated DNA it was 83 °C. The group postulate that the significant thermal resistivity afforded to the DNA by the AMP clay was a result of multipoint electrostatic interactions between pendent aminopropyl groups covalently attached to the clay and counter-charged entrapped DNA molecules. Now that we have reached the paradigm of using whole cellular structures as templates, the use of multi-cellular organisms as templates is the next logical step to take. Still firmly rooted in the organic/inorganic interactions between mineral phases and proteins and polysaccharides, in theory any multi-cellular construct is a potential template; the interest lies in the relationship between complex form and function.
7.6 References 1. 2.
3.
4. 5.
6. 7. 8.
9. 10. 11. 12.
Douglas, T. & Young, M. Host-guest encapsulation of materials by assembled virus protein cages. Nature 393, 152-155 (1998). Lee, S.Y., Culver, J.N. & Harris, M.T. Effect of CuCl2 concentration on the aggregation and mineralization of Tobacco mosaic virus biotemplate. J. Colloid Interface Sci. 297, 554-560 (2006). Shenton, W., Douglas, T., Young, M., Stubbs, G. & Mann, S. Inorganic-organic nanotube composites from template mineralization of tobacco mosaic virus. Adv. Mater. 11, 253-256 (1999). Dujardin, E., Peet, C., Stubbs, G., Culver, J.N. & Mann, S. Organization of metallic nanoparticles using tobacco mosaic virus templates. Nano Lett. 3, 413-417 (2003). Fowler, C.E., Shenton, W., Stubbs, G. & Mann, S. Tobacco mosaic virus liquid crystals as templates for the interior design of silica mesophases and nanoparticles. Adv. Mater. 13, 1266-1269 (2001). Russell, J.T. et al. Self-assembly and cross-linking of bionanoparticles at liquid-liquid interfaces. Angew. Chem.-Int. Edit. 44, 2420-2426 (2005). Canfield, D.E. & Desmarais, D.J. Aerobic Sulfate Reduction in Microbial Mats. Science 251, 1471-1473 (1991). Ward, D.M., Ferris, M.J., Nold, S.C. & Bateson, M.M. A natural view of microbial biodiversity within hot spring cyanobacterial mat communities. Microbiol. Mol. Biol. Rev. 62, 1353-1370 (1998). Sara, M. & Sleytr, U.B. S-layer proteins. J. Bacteriol. 182, 859-868 (2000). Shenton, W., Pum, D., Sleytr, U.B. & Mann, S. Synthesis of cadmium sulphide superlattices using self-assembled bacterial S-layers. Nature 389, 585-587 (1997). Dieluweit, S., Pum, D. & Sleytr, U.B. Formation of a gold superlattice on an S-layer with square lattice symmetry. Supramol. Sci. 5, 15-19 (1998). Mertig, M., Kirsch, R., Pompe, W. & Engelhardt, H. Fabrication of highly oriented nanocluster arrays by biomolecular templating. Eur. Phys. J. D 9, 45-48 (1999).
Viruses and Bacteria 13. 14. 15.
16. 17. 18. 19. 20. 21. 22.
23. 24. 25.
26. 27.
28.
173
Moore, J.T., Beale, P.D., Winningham, T.A. & Douglas, K. Controlled morphology of biologically derived metal nanopatterns. Appl. Phys. Lett. 71, 1264-1266 (1997). Hall, S.R., Shenton, W., Engelhardt, H. & Mann, S. Site-specific organization of gold nanoparticles by biomolecular templating. ChemPhysChem 2, 184-186 (2001). Bergkvist, M., Mark, S.S., Yang, X., Angert, E.R. & Batt, C.A. Bionanofabrication of ordered nanoparticle arrays: Effect of particle properties and adsorption conditions. J. Phys. Chem. B 108, 8241-8248 (2004). Braun, N. et al. Formation of metal nanoclusters on specific surface sites of protein molecules. J. Mol. Biol. 321, 341-353 (2002). Shenton, W., Mann, S., Colfen, H., Bacher, A. & Fischer, M. Synthesis of nanophase iron oxide in lumazine synthase capsids. Angew. Chem.-Int. Edit. 40, 442-445 (2001). Reith, F., Rogers, S.L., McPhail, D.C. & Webb, D. Biomineralization of gold: Biofilms on bacterioform gold. Science 313, 233-236 (2006). Davis, S.A., Burkett, S.L., Mendelson, N.H. & Mann, S. Bacterial templating of ordered macrostructures in silica and silica-surfactant mesophases. Nature 385, 420-423 (1997). Zhou, H., Fan, T.X. & Zhang, D. Hydrothermal synthesis of ZnO hollow spheres using spherobacterium as biotemplates. Microporous Mesoporous Mat. 100, 322-327 (2007). Niemeyer, C.M. Self-assembled nanostructures based on DNA: towards the development of nanobiotechnology. Curr. Opin. Chem. Biol. 4, 609-618 (2000). Sharma, J., Chhabra, R., Liu, Y., Ke, Y.G. & Yan, H. DNA-templated self-assembly of twodimensional and periodical gold nanoparticle arrays. Angew. Chem.-Int. Edit. 45, 730-735 (2006). Tanaka, K. & Shionoya, M. Bio-inspired programmable self-assembly on DNA templates. Chem. Lett. 35, 694-699 (2006). Sotiropoulou, S., Sierra-Sastre, Y., Mark, S.S. & Batt, C.A. Biotemplated nanostructured materials. Chem. Mat. 20, 821-834 (2008). Numata, M., Sugiyasu, K., Hasegawa, T. & Shinkai, S. Sol-gel reaction using DNA as a template: An attempt toward transcription of DNA into inorganic materials. Angew. Chem.Int. Edit. 43, 3279-3283 (2004). Gu, Q. & Haynie, D.T. Palladium nanoparticle-controlled growth of magnetic cobalt nanowires on DNA templates. Mater. Lett. 62, 3047-3050 (2008). Fujikawa, S., Takaki, R. & Kunitake, T. Nanocopying of individual DNA strands and formation of the corresponding surface pattern of titania nanotube. Langmuir 21, 8899-8904 (2005). Patil, A.J., Li, M., Dujardin, E. & Mann, S. Novel bioinorganic nanostructures based on mesolamellar intercalation or single-molecule wrapping of DNA using organoclay building blocks. Nano Lett. 7, 2660-2665 (2007).
Chapter 8
Complex Biostructures as Templates
8.1 Complex biostructures Through millions of years of evolution, biological structures represent the optimum fit of form to function. Enhanced functionality through architectural elegance, always using the minimal amount of material is a feature of the natural world and has long fascinated scientists, who strive to recreate these structures in the laboratory. When environmental constraints are the blueprints of construction, the structures that result are ones which are produced with the utmost efficiency in terms of amounts of materials and energetics of assembly. This was first stated explicitly by D’Arcy Thompson, in his seminal work in the early years of the 20th Century ‘On Growth and Form’1. Since then, ‘biomimetics’ has developed as a paradigm to utilize complex biostructures to produce complex inorganic phases, usually (but not always) imbuing the inorganic phase with a functionality that it would not normally have. One of the first structures investigated as a biotemplate, owing to its familiarity and ease of use was wood. Wood is primarily cellulose in composition, and as we have seen from the previous chapter on polysaccharides, is therefore able to interact chemically with many inorganic materials, forming composite materials with ease. This favourable interaction is therefore able to be exploited just as easily when the cellulose is pre-organized in a complex form. At a cellular level, wood adopts a great many different morphologies, ranging from the woody grasses such as bamboo, which consists of highly ordered fibres of cellulose, to cork, which consists of spherical cells of cellulose in a spongelike morphology. Each type of wood therefore will possess different mechanical properties according to the disposition of the cells within the body of the plant. The simplest use of wood as a biotemplate is in the formation of a carbon replica of the wood through thermal treatment. In their work, Nagle and Byrne summarize this protocol2, in particular drawing attention to the fact that the 174
Complex Biostructures as Templates
175
manufacture of charcoal is essentially a biotemplating process; one which has been known since time immemorial. They investigate the production of charcoal from a materials chemistry point of view and under controlled laboratory conditions. In this way, they are able to produce crack-free porous carbon monoliths from any given piece of wood. By heating for several days (analogous to making charcoal) they find that crack formation is minimized and that the resulting carbon structure is almost indistinguishable in terms of morphology from the original wood structure. In addition, they find that there is a linear relationship between the density of wood and the density of the carbon replica. The group found that they were able to use these carbon replicas as starting materials for further syntheses. In particular, they were able to demonstrate three different protocols for the infiltration of the porous replicas, potentially enabling the formation of a wide range of carbide materials. They looked at infiltration by molten and colloidal precursors and also by exposure to reagents in the vapour phase. They demonstrated the efficacy of their technique by synthesizing silicon carbide (SiC). The group placed carbonized wood replicas in a vacuum tube assembly with pieces of silicon. Heating to 1450 °C under argon caused the silicon to melt and be absorbed into the porous carbon replica. The carbon and silicon were then able to undergo spontaneous reaction to form SiC. SEM images indicated that the crystallites of SiC followed the underlying cellular structure with high fidelity. The use of wood as a template allows for a very open structure to be formed and the resulting inorganic phase will possess a higher surface area than would be expected in the bulk phase. A high surface area (> 300 m2g-1) opens up new vistas of applicability for the material under investigation, such as catalysis, chromatographic separation techniques and tailored drug delivery. The high surface area material is also more chemically reactive, owing to micro- and nanoscale features providing greater numbers of reactive sites and points of nucleation, enabling a high surface area material to outperform a chemically identical, low surface area material. Carrying this motif through all length scales, carbon replicas of wood have been used as a template on which a zeolite was grown. Zeolites are high surface area materials, with a microporous (< 5 nm) structure. In addition to the 48 naturally occurring examples, more than 150 structurally distinct zeolite types have been synthesized in laboratories3-5. They are highly prized for their ion-exchange capabilities and have found uses in water purification, molecular sieving and as high surface area catalyst supports. Traditionally, in a laboratory synthesis, a structure directing agent such as a
176
Biotemplating
bulky organic cation for example, tetrapropylammonium hydroxide (TPAOH) will be used, with an aluminosilicate gel condensing around it to form the zeolite. These materials have a monodisperse pore structure, the size of which is dependent on the size of the bulky organic cation. Extension of the pore size would allow for a greater range of reactions/applications for zeolites and much research has been devoted to the exploration of different templating agents. Greil et al. achieved the formation of zeolitic phases around carbonized wood replicas6. Using rattan palm as a precursor material, pyrolysis to carbon replicas followed by an infiltration with liquid silicon at 1550 °C led to a biomorphous interconnected SiC skeleton, with no change in the overall morphology of the carbonized wood template. Crystallization of the zeolitic phase was then performed by incubating carbonized replicas in an autoclave at 175 °C for 48 hours, with a basic solution of TPAOH to which aluminium nitrate had been added. The successful formation of a crystalline zeolitic phase was confirmed by X-ray diffraction which showed peaks due to crystalline Si and SiC in the replicas, and the presence of an MFI-type (10 membered oxygen ring system) zeolite material (Figure 8.1).
Figure 8.1 – Biotemplating LSI process of rattan-derived SiSiC samples. Pictures of (a) rattan wood cylindrical samples. (b) Rattan-derived biocarbon preform. (c) Rattan-derived SiSiC and polished SiSiC.Silicalite-1 composite. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
The use of carbonized wood as a complex, porous biotemplate has been explored in analogous work on the formation of TiC7, TiN8, Al2O39 and ZrO2 materials10. The process is not just restricted to wood though, as carbonized replicas of wood products possess a structural complexity of their own. In particular paper and cardboard have been used as templates in the formation of
Complex Biostructures as Templates
177
complex ceramic phases11. In his excellent review, Sieber gathers together similar work done on the mineralization of carbonized cellular plant material and in particular, the practical use to which these materials can be turned12. As with the majority of biotemplates identified in this book, it is the low cost of wood and wood products which make them attractive as templates for complex materials. Naturally occurring cellular materials are not restricted to the world of wood. Greil et al. have also demonstrated that a zeolite can be grown effectively in a high surface area morphology using luffas13. Luffas are a family of vine plants, which produce fruits (luffa sponges) which are highly porous, containing channels of between 10 µm and 20 µm in diameter. This makes them excellent absorbers of moisture and have been used in a wide range of applications from bath sponges to scaffolds for the regeneration of cell damage in the body14. The luffa sponge, by nature of its thick walled vascular structure is resistant to mechanical stress, able to deform and regain its shape with the minimum amount of loss of integrity. Rather than pre-carbonize the luffa sponges prior to zeolite growth, the group found that the spongy luffa material was sufficiently structurally stable to undergo the synthesis of zeolite seeds under hydrothermal conditions, without disruption to the vascular structure of the luffa. The group simply incubated luffa sponges in a hydrothermal bomb device, with a zeolite synthesis solution comprising TPAOH and a silica source. Once the reaction was complete, the luffa sponges were found to have been seeded with zeolite crystals, which could then be subsequently used as sites of preferential nucleation and growth of a thicker zeolitic coating. A cycle of the seeded replicas under the first synthesis conditions produced an even coating of zeolite throughout the vascular structure of the luffa sponges. The group were able to remove the luffa sponge template after this second growth step, by calcining the composite material in air at 600 °C for 10 hours. The presence of the first seeded layer of zeolite crystals ensured a high degree of crystal intergrowth, thereby leading to a structurally stable inorganic replica post-calcination, with a cellular zeolitic material of between 5 and 30 µm thick. There are also indications that the internal chemistry of plants can be used to control the morphology of inorganic phases. Sastry et al. used extracts of the lemongrass plant (Cymbopogon flexuosus) in order to alter the growth of gold nanoparticles from aqueous solutions of tetracholoraurate (AuCl4)15. By boiling lemongrass leaves, the group managed to extract a substance which they determined to be rich in aldehydes and ketones. The reducing power of aldehydes was then used in the formation of gold nanoparticles on addition of the extract to
178
Biotemplating
a solution of tetracholoraurate ions. The group found that after 6 hours of reaction between the two solutions, there were appreciable numbers of spherical and triangular gold nanoparticles present. The triangular particles ranged in size from 50 nm to 1.8 µm and were determined to be single crystals via electron diffraction. The group surmise that the hexagonal disposition of the diffraction spots are a good indication that the prisms are highly [111] oriented, with the flat surface normal to the electron beam. The group also report that the reduction of gold by different plant extracts, such as that derived from geranium leaves can produce different morphologies (sheets and rods). As well as cellular plant material, which is polysaccharide writ large, protein aggregates with complex form can be used as templates. Of particular note is the silk generated either by spiders or from the cocoons of silk moths. Again, invoking all of the co-operative assembly stratagems of the protein molecule seen in an earlier chapter, the application to silk threads with organized structures produce another level of complexity in the eventual inorganic replica. Mann et al. used spider dragline silk as a template on which to grow primarily (but not exclusively) magnetic materials with the eventual form of hollow fibres of the same dimensions as the original spider silk (0.2 mm to 10 mm in diameter)16. Spider silk is of interest as it is often quoted as having a higher tensile strength than steel or Kevlar® fibres, although this is on a weight-forweight basis, which means that in practice the exceedingly light spider silk fibres are easy to break. Nevertheless, research interest is driven by the possibility of using such a high resistivity material as a template, as the range of syntheses which can be conducted on it without compromising structural integrity are broad. In their study, Mann et al. took bundles of spider dragline silk and suspended them in colloidal solutions of either magnetite (Fe3O4), cadmium sulfide (CdS) or gold in order to absorb the nanoparticles by natural contraction of the silk on immersion in polar solvents. After a few minutes in the solutions, the silk fibres were simply withdrawn slowly and allowed to dry in air. SEM images of the composite materials showed that the surface of the fibres was quite rough, suggesting that the uptake of nanoparticles had been substantial. Despite the thickness of the coating, the silk fibres retained their flexibility as was evidenced by the presence of twisted fibres with crack-free coatings (Figure 8.2). In the case of magnetite coated silk fibres, the integrity and amount of coating was able to be demonstrated when the group found that they were able to orient the fibres in a magnetic field.
Complex Biostructures as Templates
179
Figure 8.2 – SEM image of magnetite-coated silk fibre displaying retained flexibility without significant loss of the mineral phase, scale bar = 10 µm. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
Whilst replication of fibrous and cellular morphology are useful, the prime reason for using a complex biostructure as template is to make use of a morphology that would otherwise be impossible (or at least extremely improbable) to achieve in vitro. The replication of complex form will afford the inorganic replica a function such as high surface area or tailored porosity that is simply beyond the reach of current non-biotemplated methods. A prime example of the creation of tailored porosity and morphology was demonstrated by the work of Sandhage et al. in the replication of diatoms to produce silicon replicas which retained the morphology and features of the original17. In their highly original work, the group took diatomaceous earth which comprised large amounts of the cylindrically-shaped Aulacosira species and heated it in air to 600 °C in order to remove any organic material. The material left after this process was entirely silica in composition. The group then heated these treated diatoms to 900 °C under argon in a sealed steel container to which magnesium powder had been added. In the presence of magnesium vapour, the silica was converted to silicon by an exchange process: 2 Mg(g) + SiO2(s) → 2MgO(s) + Si(s)
180
Biotemplating
The conversion of silicious diatoms by this process preserved the morphology of the frustules (shells) perfectly. The group found that subsequently, the MgO in the product could be dissolved, leaving a highly porous silicon replica which retained the overall morphology of the original diatom (Figure 8.3).
Figure 8.3 – (a) Scanning electron microscopy image of an Aulacoseira frustule after reaction for 4 h with Mg(g) at 900 °C. (b) Bright-field transmission electron microscopy image of a cross section of an Aulacoseira frustule after reaction for 4 h with Mg(g) at 900 °C. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.
Complex Biostructures as Templates
181
The production of metallic replicas of diatoms was not just restricted to silicon however. Using the same methodology, the group also created diatom replicas in a wide range of compositions, thereby demonstrating that this approach is a general one. The power of this approach comes in the observation that as there are over 10,000 unique species of diatom, each with their own particular morphology, porosity and size (ranging from less than 1 µm to over 100 µm) it follows that a huge choice is afforded to the researcher who requires a given morphology for a given product. The group observe further that as asexual reproduction in a given species occurs several times a day, thirty reproductions will yield a maximum of 230 individuals. This results in over a billion diatom frustules, all with the same morphology and surface features. Such specificity and control over such a complex morphology is rare and as such is a most attractive feature for groups interested in biotemplating. The group highlight potential uses for diatom replica materials in low-cost, genetically engineered meso/nanodevices such as optical gratings, filters and catalytic reactors. The biotemplated replication of diatoms is driven largely by the need to produce complex, high surface area materials. In this case, silica is the starting material which is subsequently transformed into silicon. Silica however, is a useful material in its own right, due to its refractory nature and its ability to be amorphous in the solid state at ambient temperatures and pressures. With no preferred crystalline form at room temperature and pressure, silica materials are essentially isotropic in their physical behaviour. This means that the propensity for deformation by mechanical stresses along a given axis is eliminated. In addition, by introducing porosity, either by forming silica phases around sacrificial templates or by incorporating porogenic materials into existing silica networks, a high surface area can be attained. Even amorphous silica, which has had no treatment can have a surface area over 300 m2/g, owing to the presence of micropores in the condensed structure. With the use of surfactants, amorphous silica has been made to condense around organic molecules to form materials well above 1000 m2/g. The very high porosity in the product is therefore a result of the fine structural feature of the template. When looking for a natural material which can be used to create high surface area inorganic materials, one seeks materials which exhibit structural stability and intricate morphological features. One material which suits this criteria particularly well is pollen. Pollen grains, typically with dimensions in the size range 25–35 µm, are plant germ cells surrounded by a layer of cellulose known as the intine, and a waxy, resistant coat,
182
Biotemplating
the exine, which consists of a polymeric substance called sporopollenin. The function of the exine is to protect the pollen grain from desiccation and microbial attack, which accounts for its durability and resistivity/tolerance to chemical processing. It is this resistivity which make pollen incredibly persistent in the biosphere without progressive loss of morphology with time. For example, identifiable pollen grains have even been found in the tombs of Egyptian pyramids, their unchanged morphology a testament to their resistance to biodegradation even after almost 5,000 years of quiescent storage. Many pollen species have a high surface area and an intricate design, so can be selectively chosen for tailored applications in the same way that diatom replicas were used above. Hall et al. achieved the first considered replication of pollen grains using a simple immersion of pollen grains in metastable solutions of silica, calcium phosphate or calcium carbonate18. Figure 8.4 shows representative scanning electron microscope (SEM) images of native and mineralized pollen grains from mustard plants (Brassica). The native pollen grains were approx. 25 µm in length and exhibited a characteristic ellipse-like morphology consisting of four longitudinal segments with foam-like surface structure. These features were also displayed in the hollow inorganic replicas, along with an additional macroporosity associated with removal of the biological template and reduction in the size of the particles to 10–15 µm due to the thermal processing. Energy dispersive X-ray mapping of individual grains confirmed that the calcined replicas consisted of silica, calcium carbonate or calcium phosphate with negligible organic content. The group found that thermogravimetric studies of the silica replicas indicated a typical organic weight loss during heating of approx. 70% and that 29Si NMR spectra of the silica replicas showed Q4, Q3 and Q2 (Qn = Si(OSi)n(OH)4–n, n = 2–4) compositions of approximately 75%, 9% and 16% respectively. This shows that even after calcination at 600 °C, approximately 25% of the siloxane centres were not fully condensed and therefore could potentially be used in further functionalization. This is an interesting area for further exploration, as the attachment of reactive functional groups to the surfaces of these replicas could extend their range of applications considerably. The silica pollen replicas were found to have high surface areas, ranging from 300 m2g-1 to as high as 817 m2g–1. It is likely that the high surface area is a function of the complex foam-like surface morphology of the native pollen grains and outgassing of organic components during thermal degradation.
Complex Biostructures as Templates
183
Figure 8.4 – SEM images of (a) Uncoated pollen grain; (b)–(d) inorganic replicas consisting of (b) calcium phosphate, (c) calcium carbonate and (d) silica. Scale bars; (a) and (c) 5 µm, (b) and (d) 1 µm. Reproduced by permission of The Royal Society of Chemistry.
The group utilized this high surface area as a ‘nano-reactor’ in the synthesis of silver or magnetite (Fe3O4) nanoparticles to produce metallic or magnetic derivatives, respectively. In the former, in situ photoreduction of adsorbed Ag(I) ions resulted in a homogeneous dispersion throughout the silica matrix of silver nanoparticles with mean size of 11 nm. When the silica pollen replicas were immersed in a sol of magnetite nanoparticles, sequestration of the colloid resulted in the uptake of nanoparticles with mean size of 7.6 nm through capillary forces into the macro- and mesopores of the silica matrix. In a final step, the group proved the efficacy of these high surface area materials in the controlled release of the drug molecules ibuprofen and chlorpheniramine (an anti-histamine, used in the treatment of hay-fever!). Samples of the silica pollen replicas were immersed in solutions of the drugs, which were then recovered, washed and dried and immersed in solutions of simulated body fluid (SBF). The release of sequestered drug molecules was monitored by following the increase in absorbance of characteristic peaks of each drug in the UV-vis profile. For each drug, the silica
184
Biotemplating
replicas released approx. 50% of the entrapped drug within the first 10 minutes of immersion, after which the rate of release decreased significantly, such that 70 to 80% of the drug was transferred into the SBF after 3 hours (Figure 8.5).
Figure 8.5 – UV-vis spectra showing release of ibuprofen from (▲) silica, (♦) calcium carbonate and (■) calcium phosphate replicas. Reproduced by permission of The Royal Society of Chemistry.
The group also produced replicas of pollen grains which comprised titania (TiO2)19. Titania is of particular interest, as it can act a photocatalytic agent for the degradation of organic molecules. The theme of complexity as function enhancer is nowhere more starkly demonstrated than in the complex biostructures as templates. With complex form at a macroscopic scale, many other emergent properties are seen. With periodic and/or roughened surfaces, many biological materials are able to alter the play of light over them. A matte surface will reflect much less light than a surface with a gloss finish. This is just the sort of developmental feature that is driven by natural selection; the most efficient and high performance surface providing an evolutionary advantage for the organism. Wang et al. used the compound eyes of the common housefly Musca domestica as a template for the formation of
Complex Biostructures as Templates
185
alumina films, which demonstrated an anti-reflective effect20. The group attributed this to the roughened surface of the fly’s eye. The compound eye of the fly is composed of many small (20 µm diameter) optical units (ommatidia) arranged hexagonally across the eye. In addition, these ommatidia are themselves covered with close-packed nanosized hemispheres of around 350 nm in diameter and 120 nm in height. It is this nano-sized roughened surface which acts as an optical grating for the eye, most likely enhancing the absorption of certain wavelengths of light and improving the fly’s perception of its environment. Wang et al. took samples of compound eyes from flies and fixed them on an alumina substrate. The group were then able to coat the surface of the eye using atomic layer deposition (ALD). In this process, they used a solution of Al(CH3)3 (trimethylalumina – TMA) as the source and under vacuum at 80 °C, subjected the fixed compound eye to pulsed deposition of TMA. The thickness of the depositing layer could be finely tuned as under the experimental conditions, a 1 Å thick layer of alumina was deposited with each pulse. Once the required thickness had been reached, the composite samples were annealed at 500 °C for 15–20 hours in order to densify the alumina layer and simultaneously remove the eye template. The fine control over film thickness resulted in an almost perfect replication of the underlying structural features of the compound eye. From Figure 8.6 it can be seen that the even the nano-sized hemispheres on the surface of the ommatidia were reproduced. The group performed angular diffraction measurements on the alumina replicas, covering wavelengths from the near-IR to the near-UV. By measuring reflectance at a 30° angle to the normal, an indication of the anti-reflective nature of the replica could be obtained. When compared with a uncoated eye, an eye coated with 100 nm of alumina exhibited similar optical properties, with the exception of a red shift in the peak in the UV at 330 nm. In the replica, this peak was shifted to 375 nm, which was due to the change in refractive index afforded by the alumina coating. In addition, the hemispherical ommatidia acted as a antireflective material, analogous to a simple diffraction grating composed of a periodic array. The group gauged the effectiveness of the fly eye replica by considering the property of antireflection as determined by the grating equation: sin ϕ m = sin θ + mλ/d where θ, ϕ m, m and λ correspond to incident angle, diffraction angle, diffraction order and wavelength, respectively. In the case of the replicas, d was determined to be 350 ± 42 nm. In the case of perfect illumination, m = −1, and θ and ϕ −1
186
Biotemplating
kept at 30 and −30 respectively, the constructive diffraction was determined to be 350 ± 42 nm, which correlated well with the observed reflectance in the UV spectrum. Angular reflectance measurements showed that an untemplated alumina film gave a reflectance of 6%, whereas the roughened surface of the alumina fly eye replica showed a reflectance of only 0.7%. This decrease of an order of magnitude in the templated alumina is an indication of the success in physical property enhancement that biotemplating can afford.
Figure 8.6 – Structure of the alumina replica of the fly eyes. (a) SEM image of the alumina replica of a fly compound eye. (b) SEM image of the alumina replica showing preserved ommatidia. (c) EDS of the alumina replica. (d) High magnification SEM image of the alumina replicated ommatidium surface. (e) Enlarged SEM image showing the replicated protuberance nanostructures. (f) TEM image of a piece of replicated ommatidium structure showing the linear arrangement of spherical the protuberances. Reproduced by permission of IoP Publishing Ltd.
Complex Biostructures as Templates
187
A similar result was found by another group who used cicada eyes as a template for epoxy resin replicas21. Although they demonstrated that as roughened surfaces are not restricted to compound eyes, the wing of the cicada could also be used22. In this work the group took cicada wings and sonicated them in acetone/water in order to remove physisorbed contaminants. The group were then able to coat the wings with gold using a thin-film coating system. Once formed, the gold replica was able to be used as a resist to form a PMMA replica which was flexible and retained all of the fine structural features of the original wing. The group chose PMMA for its optical transparency in the IR- to visible region, but mainly for its high mechanical strength and flexibility. The latter property was critical for the future formation of an anti-reflective layer on a curved surface. The group found that the reflectivity of the PMMA wing replica was around 30% less than a corresponding unpatterned PMMA surface. The fine structure of butterfly wings was also investigated as an antireflective pattern. The rationale behind this work by Wang et al. was to replicate the structure of the butterfly wing scales, which in certain species such as the Morpho butterflies, exhibit photonic behaviour23. The complex interplay of light with the fine, periodic structures of the chitinous wing produce, in the case of the Morpho butterfly, a dazzling electric blue colour when viewed perpendicular to the wing surface. As the angle of view changes, so the colour changes, indicative that the colour observed is due to a physical phenomenon, rather than the result of a pigment. On closer examination, the reason for the colour play on the wing can be seen in the structure of the scales which cover the entire dorsal surface of the wing (Figure 8.7). Through SEM studies, the group determined that these scales have a structural periodicity through arrays of vertical spars of the order of 100 nm. These alter the permitted wavelengths of light which can interact with the wing scale, producing photonic effects such as band gaps (non-allowed wavelength propagation) and iridescence. The group made use of this natural photonic material as a template and given the delicacy of the wing scale structure, opted for an ALD approach as described above. Using this method, the group were able to produce alumina replicas of the wing scale structure, with control over the thickness of the alumina coating by additional iterations of the coating procedure. On coating the wing scale, it was found that the optical properties of the structure could be changed radically. By progressively increasing the thickness of the coating, the photonic properties of the wing scale composite with regards to the permitted wavelengths of light could be controlled. As the thickness of the coating increased from 10 nm to 40 nm, the group found
188
Biotemplating
that the reflected colour of the wing scale changed from blue to green, then yellow, orange and finally pink.
Figure 8.7 --- Images of the alumina replicas of the butterfly wing scales. (a) An optical microscope image of the alumina coated butterfly wing scales. (b) A low-magnification SEM image of the alumina replicas of the butterfly wing scales on silicon substrate after the butterfly template was completely removed. (c) The energy dispersive X-ray (EDX) spectrum of the alumina replica shown in part b. (d) A higher magnification SEM image of an alumina replicated scale, where the replica exhibits exactly the same fine structures. (e) A SEM image of two broken rib tips on an alumina replica. Reprinted with permission from23. Copyright 2006 American Chemical Society.
Complex Biostructures as Templates
189
The replication of the underlying wing scale structure was particularly fine using this approach, which was enhanced by the removal of the organic component by calcination to 800 °C for three hours. In addition to the removal of the original wing scale, the heat treatment had the effect of crystallizing the amorphous alumina, thereby ensuring that there was little or no structural collapse of the structure on template removal. TEM studies on the calcined alumina replica showed that the perpendicular lamella structures of the wing scale were retained, indicating that the material was an inverse 2D photonic crystal. Further investigations into the photonic properties of the replicas were undertaken by angular reflection measurements of single wing scales. As it was determined that the main ribs which give rise to the photonic properties were oriented 30° to the surface of the scale, the group mounted individual wing scale replicas on substrates and measured the interaction of them with light at an angle of 60° to the normal to ensure full interaction with the photonic structure. The group found that as expected for a purple/blue coloured Morpho butterfly, the original wing scales showed the highest reflection at 390 nm, with a large component in the blue region of the visible spectrum. Once coated with 40 nm alumina however, the reflection peak was red-shifted to 600 nm, which accounts for the red/pink colour. After calcination, the structure of the wing scale replica underwent densification and its dimensions were essentially the same in terms of size as the original wing scale, albeit an inverse. This resulted in a blue-shift of the peak reflectance to 420 nm. The group note that the performance of these replicas as photonic crystals could lead to applications in photonic integrated circuits, light emitters, waveguides, splitters, and detectors that produce, control and process light signals and/or perform logical computations. As with the diatoms discussed previously, careful selection of the wing scale from different species has the potential to provide the scientist or engineer with a material which could control light of any desired wavelength; it is merely a question of selecting the species of butterfly with the right wing scale morphology. Butterfly wing scales are therefore ideal as complex biostructures; the reactive/receptive interaction between the chitin of the wing scale and the inorganic material, coupled with the functional property of a photonic material. This template has the potential to be used to create a wide range of photonic inorganic materials. For example, it has been demonstrated that silica replicas can be formed through the application of silane gas to wing scales which have been previously soaked in a solution of hydrogen peroxide24. Similarly ZnO replicas have been created through immersion of wing scales in a zinc nitrate solution25. In both cases, the
190
Biotemplating
composite material was calcined in order to both remove the template and achieve the correct crystallinity of the inorganic phase. An exceedingly simple method of replication of butterfly wing scales was demonstrated by Li et al. in the creation of lead lanthanum zirconate titanate (Pb0.91La0.09(Zr0.65Ti0.35)0.9775O3 - PLZT) structures, a material with a tunable band gap26. The group first produced a gel mat of the precursor material using zirconium propoxide (Zr(OC3H7)4, 70 wt% solution in 1-propanol; titanium butoxide (Ti(OC4H9)4; lead acetate (Pb(C2H3O2)2 • 3H2O; and lanthanum nitrate (La(NO3)3 • 6H2O. Once dissolved together, the precursor materials were aged for two days and cast to form a thin film. Butterfly wing scales were then able to be simply placed onto the wet gel, leading to suffusion of the precursor material throughout the organic structure, where it formed a dry, compact coating. In order to obtain the ferroelectric phase of PLZT, the group calcined the composite material to 750 °C, producing an inverse replica of the required stoichiometry. SEM images of the product showed that the PLZT had adopted the form of the wing scale perfectly, albeit with a 20% to 30% shrinkage on calcination. The group demonstrated the efficacy of their technique by replicating wing scales from two other butterfly species, each with marked differences in morphology. In each case, replication was successful, the degree of shrinkage not altering the macromorphology of the materials post-calcination. Elemental analysis by EDXA revealed that the wing scale replicas consisted as expected of lead, zirconium, titanium and lanthanum, although analysis by X-ray diffraction was not reported. The utilization of a complex, porous periodic biostructure as a template can result in a very efficient synthesis, as in addition to a high surface area, reactants are able to penetrate the open framework of the structure and ensure homogeneity. If calcination is required, the open structure again provides a benefit, in that oxygen is able to access all areas of the composite and ensure total conversion of the inorganic species and removal of the template. Another chitinous biostructure with an open, porous morphology is the cuttlebone. This is a periodic material composed of calcium carbonate in the form of aragonite, which is formed on an organic framework of β-chitin. The aragonite is arranged in a lamellar structure, with mineralized sheets separated from each other by approximately 300 µm by S-shaped pillars of chitin. The cuttlebone is highly porous, with approximately 93% of the structure void space. This results in an extremely low density, with the cuttlebone possessing a specific gravity of 0.1927. This allows the cuttlefish to use the structure as an
Complex Biostructures as Templates
191
internal rigid buoyancy tank and to maintain a fixed position in a column of seawater by adjusting the fluid pressure inside the structure. As the aragonite is formed around the β-chitin matrix, it is a simple matter to demineralize the cuttlebone by immersion in acidic solutions. Once demineralization is complete, the β-chitin structure which remains still retains the complex lamellar structure of the original. This is an ideal template for bioreplication as previous chapters have shown that chitin/metal cation chelation is strong and driven by favourable hydrogen bonding and electrostatic interactions. Mann et al. undertook the earliest demineralization-remineralization study on cuttlebone, when they created silica replicas of the structure through a simple acid etch followed by silica deposition in alkaline media28. The group observed some collapse of the β-chitin structure on dissolution of the aragonite, although the organic fragment still retained the 3D structure of the original cuttlebone. The group then remineralized the organic matrix by incubating it in a solution of sodium silicate at pH 11.5, followed by immersion in ethanol/water mixtures. The ethanol reduced the solubility of the silica species, thereby inducing precipitation of silica onto the β-chitin matrix. The group consider that specific interaction between silica species in solution and glucopyranose rings exposed at the surface of the β-chitin matrix through polar and hydrogen bonding interactions produced high-fidelity replicas. Unfortunately, calcination of the silica-chitin composite material resulted in total structural collapse of the material, albeit with the retention of interconnected sheets of amorphous silica. Calcination of the cuttlebone without the first demineralization step is one way to ensure that no structural collapse will occur. This would be particularly useful if the inorganic phase to be templated could undergo favourable (or at least not be affected by) incorporation of calcium species into the material. This concept was exploited by Hall et al. in their creation of superconducting phases by templating around a native cuttlebone29. Research into superconductivity is largely driven by the need to increase the critical current density of the material. This key physical property of a superconductor is a measure of the magnitude of current that can be passed through the material before it reverts to a non-superconductive state. Increasing the critical current density (Jc) is therefore a very active area of research. In their work, Hall et al. synthesized a superconductor precursor solution using mixed metal nitrates to achieve the correct stoichiometry for Y123. A sol-based approach was chosen, owing to the ability of the cuttlebone to admit fluids efficiently throughout its porous structure. The sol was made by dissolving barium nitrate, yttrium nitrate and copper nitrate in a distilled water–
192
Biotemplating
ethylene glycol (50 : 50) solution, forming a clear, blue sol. The addition of ethylene glycol was deemed necessary to act as a gelation agent on solvent evaporation. Into this solution, a piece of cuttlebone was soaked for approximately 2 hours under vacuum to ensure complete infiltration of the superconductor precursor throughout the cuttlebone structure. Once complete, the cuttlebone was dried at 40 °C until the sol-gel transition had caused the surfaces of the cuttlebone to become completely covered in precursor gel. This procedure was undertaken three times in order to ensure total coverage of all the inner surfaces of the cuttlebone. The composite materials were then calcined to 920 °C at a rate of 1°/min, with a four hour hold once the final temperature had been reached. The group found that SEM images of the calcined replica showed an exact replication of the underlying cuttlebone (Figure 8.8). As the cuttlebone was not demineralized prior to infiltration of the superconductor precursor, the structure was able to withstand calcination without shrinkage or structural collapse owing to the in situ transformation of aragonite to calcium oxide. Both TEM and SEM analyses showed the presence of sub-micron sized crystallites which were confirmed as Y123 by electron diffraction. SQUID magnetometry showed that the superconducting critical temperature (Tc) of the replica was 93 K, a value which confirms optimum oxygenation of the Y123. The efficient oxidation of the superconductor species was possible due to the open architecture of the cuttlebone and as such represents an improvement over many Y123 syntheses which require calcination under flowing oxygen in order to produce a high-quality Y123 material. The group discovered that the effect on Jc of synthesizing the superconductor in this open morphology was considerable. The critical current density of the cuttlebone templated Y123 was measured at 1.6 MAcm-2 at 10 K and 1 T field. This is almost two orders of magnitude higher than that observed in a commercially available Y123 powder (Aldrich 99.9% — average particle size 5 µm), for which SQUID magnetometry revealed a Tc of 92 K and a critical current density of 0.02 MAcm-2 at 10 K and 1 T field. As synthesized, these superconducting cuttlebone replicas were found to be unsuitable for potential applications however, owing to their extreme fragility. The group therefore undertook to add a small amount (10 wt%) of silver nitrate to the superconductor precursor sol, as it was previously determined that the addition of silver would imbue cuprate superconductors with a higher degree of mechanical strength. The replicas which resulted were shown to be able to withstand a compressive strength of 28 kPa. This figure is still relatively low for many applications, but is comparable to the compressive resistance of many light roofing materials.
Complex Biostructures as Templates
193
Figure 8.8 – SEM images showing Ag-doped Y123 cuttlebone replicas. Scale bar in (a) is 200 µm, in (b) 50 µm. Reproduced by permission of The Royal Society of Chemistry.
As a final thought, the group note that as the weight of a 1 cm3 portion of their material is 0.06 g, compared to 6.38 g for the equivalent sized monolith of pure Y123, their materials may well find application in areas where weight is of critical importance, such as space-based and mobile device technologies. The replication of complex biostructures to give inorganic materials with enhanced form and function is still a relatively young field. As a result, there is a wealth of complex, naturally synthesized materials out there just waiting to be pressed into use as templates. All one has to do is find them.
194
Biotemplating
8.2 References 1. 2. 3.
4. 5. 6.
7. 8. 9. 10. 11. 12. 13.
14. 15. 16. 17.
18. 19.
20.
Thompson, D.A.W. On Growth and Form, Edn. 1st. (Cambridge University Press, 1917). Byrne, C.E. & Nagle, D.C. Cellulose derived composites - A new method for materials processing. Mater. Res. Innov. 1, 137-144 (1997). Breck, D.W., Eversole, W.G., Milton, R.M., Reed, T.B. & Thomas, T.L. Crystalline Zeolites 1. The Properties of a New Synthetic Zeolite, Type-A. J. Am. Chem. Soc. 78, 5963-5971 (1956). Davis, M.E. & Lobo, R.F. Zeolite and Molecular-Sieve Synthesis. Chem. Mat. 4, 756-768 (1992). Kokotailo, G.T., Lawton, S.L., Olson, D.H. & Meier, W.M. Structure of Synthetic Zeolite ZSM-5. Nature 272, 437-438 (1978). Zampieri, A. et al. Biomorphic cellular SiSiC/zeolite ceramic composites: From Rattan palm to bioinspired structured monoliths for catalysis and sorption. Adv. Mater. 17, 344-349 (2005). Niyomwas, S. in Epd Congress 2008. (ed. S.M. Howard) 345-354 (Minerals, Metals & Materials Soc, Warrendale; 2008). Luo, M. et al. Biomorphic TAX ceramics prepared by reduction-nitridation of charcoal/titania composite. J. Mater. Sci. 42, 3761-3766 (2007). Cao, J., Rambo, C.R. & Sieber, H. Preparation of porous Al2O3-Ceramics by biotemplating of wood. J. Porous Mat. 11, 163-172 (2004). Rambo, C.R., Cao, J. & Sieber, H. Preparation and properties of highly porous, biomorphic YSZ ceramics. Mater. Chem. Phys. 87, 345-352 (2004). Luo, M., Gao, J.Q., Qiao, G.J. & Jin, Z.H. Synthesis of wood-derived ceramics from biological templates. Prog. Chem. 20, 989-1000 (2008). Sieber, H. Biomimetic synthesis of ceramics and ceramic composites. Mater. Sci. Eng. A-Struct. Mater. Prop. Microstruct. Process. 412, 43-47 (2005). Zampieri, A. et al. Biotemplating of Luffa cylindrica sponges to self-supporting hierarchical zeolite macrostructures for bio-inspired structured catalytic reactors. Mater. Sci. Eng. C-Biomimetic Supramol. Syst. 26, 130-135 (2006). Chen, J.P., Yu, S.C., Hsu, B.R.S., Fu, S.H. & Liu, H.S. Loofa sponge as a scaffold for the culture of human hepatocyte cell line. Biotechnol. Prog. 19, 522-527 (2003). Shankar, S.S. et al. Biological synthesis of triangular gold nanoprisms. Nat. Mater. 3, 482-488 (2004). Mayes, E.L., Vollrath, F. & Mann, S. Fabrication of magnetic spider silk and other silk-fibre composites using inorganic nanoparticles. Adv. Mater. 10, 801-805 (1998). Sandhage, K.H. et al. Novel, bioclastic route to self-assembled, 3D, chemically tailored meso/nanostructures: Shape-preserving reactive conversion of biosilica (diatom) microshells. Adv. Mater. 14, 429-433 (2002). Hall, S.R., Bolger, H. & Mann, S. Morphosynthesis of complex inorganic forms using pollen grain templates. Chem. Commun., 2784-2785 (2003). Hall, S.R., Swinerd, V.M., Newby, F.N., Collins, A.M. & Mann, S. Fabrication of porous titania (brookite) microparticles with complex morphology by sol-gel replication of pollen grains. Chem. Mat. 18, 598-600 (2006). Huang, J.Y., Wang, X.D. & Wang, Z.L. Bio-inspired fabrication of antireflection nanostructures by replicating fly eyes. Nanotechnology 19, 6 (2008).
Complex Biostructures as Templates 21. 22. 23. 24. 25. 26. 27. 28.
29.
195
Gao, H.J., Liu, Z., Zhang, J., Zhang, G.M. & Xie, G.Y. Precise replication of antireflective nanostructures from biotemplates. Appl. Phys. Lett. 90, 3 (2007). Xie, G. et al. The fabrication of subwavelength anti-reflective nanostructures using a biotemplate. Nanotechnology 19 (2008). Huang, J.Y., Wang, X.D. & Wang, Z.L. Controlled replication of butterfly wings for achieving tunable photonic properties. Nano Lett. 6, 2325-2331 (2006). Cook, G., Timms, P.L. & Spickermann, C.G. Exact replication of biological structures by chemical vapor deposition of silica. Angew. Chem.-Int. Edit. 42, 557-559 (2003). Zhang, W. et al. Fabrication of ZnO microtubes with adjustable nanopores on the walls by the templating of butterfly wing scales. Nanotechnology 17, 840-844 (2006). Li, B. et al. Ordered ceramic microstructures from butterfly bio-template. J. Am. Ceram. Soc. 89, 2298-2300 (2006). Denton, E.J. & Gilpinbrown, J.B. Buoyancy of the Cuttlefish. Nature 184, 1330-1331 (1959). Ogasawara, W., Shenton, W., Davis, S.A. & Mann, S. Template mineralization of ordered macroporous chitin-silica composites using a cuttlebone-derived organic matrix. Chem. Mat. 12, 2835-2837 (2000). Culverwell, E., Wimbush, S.C. & Hall, S.R. Biotemplated synthesis of an ordered macroporous superconductor with high critical current density using a cuttlebone template. Chem. Commun., 1055-1057 (2008).
Chapter 9
Into the Future – Genetic Engineering and Beyond
9.1 Genetic engineering Technological stagnation has never been an option for mankind. Despite having an extremely large pool of biotemplates available, the chances are that the interaction between template and inorganic phase is not quite strong enough, or specific enough, or produces a morphology which is not exactly the desired one. It is only a small leap of imagination therefore, to imagine actively tailoring the biotemplate in order to totally satisfy the needs of a particular application. As highlighted in the previous chapter, in the case of diatoms, there may be over ten thousand different species, all with slightly different morphologies, suited to slightly different applications, but to be able to request a diatom with morphology ‘made to measure’ for an application would be a technological advantage many would gladly pay for. It is entirely possible, given that many species of plant and animal remain undiscovered, that the particular template you require already exists somewhere in the biosphere, but has yet to be discovered. In ‘Life, the Universe and Everything’, the author Douglas Adams wrote: “Very few things actually get manufactured these days, because in an infinitely large Universe such as, for instance, the one in which we live, most things one could possibly imagine, and a lot of things one would rather not, grow somewhere1.” Rather than rely on the right template to be discovered somewhere in the biosphere (or elsewhere) however, scientists are increasingly taking it upon themselves to engineer the exact biotemplate they require. It is only very recently though, with advances in genetic engineering, that the active construction of a biotemplate has become a very real possibility. On a fundamental level, researchers can create genetically engineered proteins which can be used to confirm our understanding of the interaction between organic and inorganic matter. Hunter et al. investigated how certain 196
Into the Future – Genetic Engineering and Beyond
197
sequences of phosphorylated amino acid residues control the growth of crystals of calcium oxalate2. The group engineered sequences of aspartic- and glutamic acid rich amino-acid sequences found in rat bone proteins. On addition of these sequences to calcium oxalate solutions, the crystals all had the growth of their (100) faces inhibited by adsorption of the engineered proteins. They found that by increasing the degree of phosphorylation the peptides would bind more readily to the (100) face, providing confirmation that the formation of stable, face-specific interactions through complimentary charge-matching is indeed of critical importance in biomineralization. Perhaps the simplest place to start on the engineering of living entities to produce specific templating effects is to consider the virus. With both interior and exterior surfaces of viral capsids available for mineralization to greater or lesser degrees, it is possible to improve the degree and specificity of binding for particular inorganic species by altering certain amino acid residues in specific locations on the capsid. It has been determined that there are certain amino acid sequences which show a particular affinity for certain inorganic ions. In a thorough review of specificity of binding, Sarikaya et al. set the scene for the future of genetically engineered biotemplating. In their work, they discuss the creation of short protein sequences using the combinatorial biological methods of bacterial cell surface and phage-display technologies3. Using these protocols, it is possible to screen large numbers of short amino acid sequences and thereby design proteins which exhibit preferential binding of certain metals and semiconductor quantum dots. These sequences can then be used in the assembly of functional nanostructures in a direct analogy of biomineralization in nature. The group used a disulphide-constrained M13 peptide library, consisting of a seven amino-acid chain length to determine specificity of binding for platinum or palladium metal. They also demonstrated that there were particular sequences which showed affinity for a range of other metals and compounds4. In general terms, in addition to allowing researchers to engineer these exact sequences into their proteins, this table allows the selection of existing amino acids for specific binding by noting common features which appear time and again for any given inorganic species. For example, the group note that in the case of protein sequences which show high specificity for the noble metals, there is a predominance of serine and threonine amino acid residues, both structurally similar in that they contain aliphatic hydroxyl groups in their side chains. It appears that when taking an overview of the noble metal binding, the sequences consist of hydrophobic and hydroxl-containing polar amino acids. If it is not
198
Biotemplating
possible to directly engineer the sequences that the group determined, a researcher will be able to look for these general features in an existing protein and select that for noble metal binding affinity. The group also analyze the table for common features for specificity of binding for other inorganic groups. They deduce that for metal oxide and zeolite materials, basic amino acids such as arginine and lysine and hydoxyl residues were the most efficacious for selectivity. This contrasts with amino acid sequences in nature which show preferential binding for e.g. calcium carbonate, which tend to be rich in asparagines and glutamates. By incorporating these sequences into E. coli. bacteria, proteins can be expressed in large quantities, giving rise to a library of short-chained proteins with high specificities for a wide range of inorganic phases. In addition to the academic interest generated from these materials, sequences can be engineered to produce remediation of metal waste streams by producing biopolymers which contain amino-acid sequences specific for heavy metals such as mercury and cadmium. One of the earliest studies on genetic alteration of a virus with a view to improving the specificity of mineralization was by Douglas et al. In this work, the group took the protein cage of the cowpea chlorotic mottle virus (CCMV) and altered the charge characteristics of the interior of the protein in order to make the cage behave like an analogue of ferritin and sequester iron5. They achieved this by genetically engineering the N-terminus of the protein, which they had previously determined was not essential for the self-assembly of the capsid, thereby allowing the group to retain the morphological features of the cage. By exchanging the basic residues at the N-terminus of the protein for glutamic acid residues, the group produced mutant CCMV capsids in which the electrostatic charge of the capsid inner surface was significantly different. By replacing 1620 cationic sites with 1620 anionic ones, the capsids were able to easily sequester iron on incubation with ammonium iron sulphate solutions. The group observe that once iron nanoparticle nucleation is instigated within the capsid, the iron undergoes oxidative hydrolysis and grows autocatalytically as iron oxide. Thus, whilst being a successful mimic for ferritin, the mutant CCMV does not require an enzyme to catalyse the oxidation of iron in situ, the high interior surface charge of the capsid will induce nucleation and growth. The group found that on examination of the mineralized capsids via TEM, high electron contrast spherical cores of 8.2 ± 1.6 nm had been formed. By adding further aliquots of iron to the mineralized cages, the size of the cores were increased to a maximum of 24.0 ± 3.5 nm, the diameter of the inner surface of
Into the Future – Genetic Engineering and Beyond
199
the capsids. Lattice imaging of the core nanoparticles revealed that they were single crystals of γ-FeOOH (lepidocrocite). Other groups have shown that it is possible to genetically engineer spherical capsids for the preferential sequestration of silver6, titanium7 and gold8. Once the genetic engineering of the inner cavity of protein capsids was shown to be possible, the genetic engineering of virus protein surface structures could then be applied and also extended to those with other capsid morphologies, mirroring the existing research on wild-type viruses. One such study involved the adaptation of the tobacco mosaic virus (TMV) to improve the stability of metal species when used as a surface coating on the protein. Harris et al. used TMV which had two cysteine residues introduced into the N-terminus of the capsid protein9. The insertion of the cysteine residues was possible without compromising the structural integrity of the capsid, as the N-terminus of the protein is situated on the outer surface of the virion, thereby causing a minimum amount of disruption to the overall structure. Gold, silver and palladium species were then introduced and reduced in situ. In the case of gold, the group found that nanoparticles of approximately 3 nm in diameter were formed and disposed evenly over the surface of the mutant TMV. They note that although gold decoration of the wild type TMV had been demonstrated previously, it was at relatively low pH values (around 2.5). In their method, the use of a mutant species allows for a favourable electrostatic interaction with the metal species to occur at much higher pH values, ranging from 3.9 to 5.6. This indicated to the group that nanoparticle formation at this pH was a consequence of the presence of the thiol groups provided by the cysteine residues on the virion surface (Figure 9.1). The group were also able to demonstrate the decoration of the mutant TMV with palladium through an analogous in situ reduction of a tetrachloropalladate salt. Many other studies have altered or augmented the amino acid sequence on the surfaces of a range of virus capsids10 in order to improve and/or extend the binding of metal species to the capsid. This improved interaction can also be a method for generating regular arrays of nanoparticles by ensuring highly specific binding to the protein which is itself formed as a regular array on a surface. McMillan et al. demonstrated this concept by the formation of regular arrays of preformed nanoparticles on templates constructed from hollow double ring protein structures called chaperonins11. Chaperonins are found in nature as subcellular ‘heat shock’ protein structures in extremophiles. These protein structures allow the bacterium which possesses them to survive in areas of high thermal activity such as hot springs.
200
Biotemplating
Figure 9.1 – (a) Ribbon diagram of TMV2cys coat protein. Added cysteine residues are shown as green ball-and-stick structures. (b) Cross-sectional diagram for a single turn of the TMV2cys virion. Added cysteine residues are shown as yellow spheres. (c) TEM image of TMV2cys, stained with 2% phosphotungstic acid solution. The scale bar equals 100 nm. Reproduced by permission of IoP Publishing Ltd.
The group took chaperonins from the bacteria Sulfolobus shibatae and expressed it as a recombinant protein in E. coli. By heating the cell extracts which result to 85 °C, the group were able to denature and precipitate the E. coli proteins, but as the chaperonin is stable at this temperature, it remains soluble and was therefore relatively easy to extract as a pure protein. They were then able to genetically modify the chaperonin by inserting cysteine residues at exposed sites on the surface of the protein. The group found that under the reducing conditions of their synthesis, the chaperonins assembled into disc-shaped, hexagonally packed 2D regular arrays of up to 20 µm in diameter. Once formed, these arrays could then act as sites for the binding of nanoparticles, with specificity of binding resulting from the regularity of cysteine thiol groups on the
Into the Future – Genetic Engineering and Beyond
201
surface of the protein. Gold nanoparticles exhibited a strong preference for binding to the cysteine surface sites, with both 5 nm and 10 nm diameter nanoparticles forming hexagonal arrays in register with the underlying chaperonin template. The group also demonstrated that other nanoparticles could be bound to the mutated chaperonin array by incubating the protein with core-shell CsSe-ZnS nanoparticles. They also explored the removal of a loop of amino acids which normally occlude the core of the chaperonin protein. By modifying the protein in this way, the metallophillic core of the chaperonin was exposed, thereby increasing the affinity of the array for metal nanoparticles. The group discovered that this newly exposed core was particularly efficacious as a site for the reduction of transition metal ions to form bimetallic nanoparticles. The group demonstrated this by synthesizing Ni-Pd and Co-Pd nanoparticles on the protein surface. The nanoparticles produced retained the underlying structure of the chaperonin array12. The formation of metallic nanoparticle arrays on genetically engineered protein surfaces has also been demonstrated for various other phage templates13. The way is now clear for genetic engineering to produce designed materials of more complex morphology, eventually moving towards engineered higher biological constructs. One of the first steps along the way has been taken by Matsui et al. with their genetically engineered polypeptide, a triple-helix collagen assembly14. The group used a naturally occurring collagen molecule and inserted cysteine residues in order to effectively cross-link the triple helices and thereby provide enhanced thermal stability and metallophilicity. Once expressed from E. coli, the genetically engineered collagen was found to form triple helices in solution of approximately 40 nm in length and 4 nm in diameter. The group were then able to use these engineered collagen molecules as templates for the deposition of gold by incubating the helices with trimethylphosphinegold chloride for four days, followed by reduction by hydrazine hydrate for one day at 48 °C. TEM images confirmed to the group that crystals of gold did indeed grow around the collagen helices. Coating the collagen with the aurophilic peptide sequence HRE (Ala-His-His-Ala-His-His-Ala-Ala-Asp) produced a more even coating of the gold over the surface of the collagen helix. Owing to the crosslinking by the cysteine residues, the collagen molecules were rigid, rather than adopting the usual twisted conformation often adopted by long-chained biomolecules in solution. A rigid conformation is desired if these are to be used as templates for nanowires in electronics applications.
202
Biotemplating
It is here that the state-of-the-art for the controlled engineering of biomaterials to produce a more precisely tailored template resides. It will surely not be long though, before the genetic alteration of the sequences in biomineralizing organisms themselves will enable scientists to direct organisms to grow inorganic materials to order. This will signal a paradigm shift in the use of biological systems from ‘passive’ to ‘active’ templates, thereby obviating the need for the scientist to ‘do chemistry’ on the template. The biological entity acting both as tailor made template and replicator would appear to be the next logical step. It is clear that the hand of Man is being successively removed as a controlling force in the engineering of these materials (Table 9.1). Table 9.1 – The controlling force in templated mineralization. Protocol
Assembly of Template
Mineralization of Template
Bioinspired Biotemplate Engineered passive template Engineered biomineralizing template
Man Nature
Man Man
Nature/Man
Man
Nature/Man
Nature
We are now much closer to the fully independent, self-replicating genetically engineered entities which are capable of building structures to our blueprints. Lincoln and Joyce have recently (2009) reported on the creation of the first synthetic RNA which can undergo self-sustaining replication15. It is still very much a preliminary step, but the creation of the first proto- ‘artificial life’ means that the introduction of genetic sequences which can encode for the considered deposition/sequestration of inorganic matter cannot be far behind. In the meantime, as long as nature can provide the complex precursors, Man will continue to improve the procedures involved in creating nano- and microengineering marvels through biotemplating.
9.2 References 1. 2.
Adams, D. Life, the Universe and Everything. (Pan Books, 1982). Grohe, B. et al. Control of calcium oxalate crystal growth by face-specific adsorption of an osteopontin phosphopeptide. J. Am. Chem. Soc. 129, 14946-14951 (2007).
Into the Future – Genetic Engineering and Beyond 3.
4. 5. 6. 7. 8. 9.
10. 11. 12. 13.
14.
15
203
Tamerler, C., Kacar, T., Sahin, D., Fong, H. & Sarikaya, M. Genetically engineered polypeptides for inorganics: A utility in biological materials science and engineering. Mater. Sci. Eng. C-Biomimetic Supramol. Syst. 27, 558-564 (2007). Sarikaya, M., Tamerler, C., Jen, A.K.Y., Schulten, K. & Baneyx, F. Molecular biomimetics: nanotechnology through biology. Nat. Mater. 2, 577-585 (2003). Douglas, T. et al. Protein engineering of a viral cage for constrained nanomaterials synthesis. Adv. Mater. 14, 415-418 (2002). Kramer, R.M., Li, C., Carter, D.C., Stone, M.O. & Naik, R.R. Engineered protein cages for nanomaterial synthesis. J. Am. Chem. Soc. 126, 13282-13286 (2004). Sano, K. et al. Endowing a ferritin-like cage protein with high affinity and selectivity for certain inorganic materials. Small 1, 826-832 (2005). Slocik, J.M., Naik, R.R., Stone, M.O. & Wright, D.W. Viral templates for gold nanoparticle synthesis. J. Mater. Chem. 15, 749-753 (2005). Lee, S.Y., Royston, E., Culver, J.N. & Harris, M.T. Improved metal cluster deposition on a genetically engineered tobacco mosaic virus template. Nanotechnology 16, S435-S441 (2005). Portney, N.G. et al. Organic and inorganic nanoparticle hybrids. Langmuir 21, 2098-2103 (2005). McMillan, R.A. et al. Ordered nanoparticle arrays formed on engineered chaperonin protein templates. Nat. Mater. 1, 247-252 (2002). McMillan, R.A. et al. A self-assembling protein template for constrained synthesis and patterning of nanoparticle arrays. J. Am. Chem. Soc. 127, 2800-2801 (2005). Lee, S.K., Yun, D.S. & Belcher, A.M. Cobalt ion mediated self-assembly of genetically engineered bacteriophage for biomimetic Co-Pt hybrid material. Biomacromolecules 7, 14-17 (2006). Bai, H.Y., Xu, K., Xu, Y.J. & Matsui, H. Fabrication of an nanowires of uniform length and diameter using a monodisperse and rigid biomolecular template: Collagen-like triple helix. Angew. Chem.-Int. Edit. 46, 3319-3322 (2007). Lincoln, T.A. & Joyce, G.F. Self-Sustained Replication of an RNA Enzyme. Published Online January 8, 2009. Science DOI: 10.1126/science.1167856.
This page intentionally left blank
Index
Chitosan, 70, 79, 80, 81, 92, 93, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 107, 108, 109, 110, 112, 113 Chromium, 70 Cicada, 187 Clay, 19, 20, 145, 146, 147, 171, 172 Cobalt, 67, 68, 69, 70, 105, 171 Collagen, 21, 64, 81, 82, 95, 119, 125, 127, 130, 131, 133, 136, 201 Copper, 23, 32, 39, 41, 70, 105, 147, 148, 157, 191 Crystallization, 5, 6, 10, 11 Cu2O, 23, 86 Curdlan, 64, 88 Cuttlebone, 92, 190, 191, 192
Acacia gum, 64, 89 Ag2S, 35 Agar, 64, 88 AgBr, 99 AgO, 35 Alginate, 1, 64, 74, 75, 76, 77, 78, 79, 80, 81, 86, 101 Alumina, 22, 23, 78, 79, 108, 185, 186, 187, 189 Aluminum, 22, 70, 79, 85, 89, 108, 147, 176 Bacteria, 30, 38, 63, 86, 119, 153, 154, 162, 163, 164, 165, 167, 168, 169, 170, 197, 198, 200 BaSO4, 8, 73, 144 Biomimetics, 4, 11, 174 Biomineralization, 3, 4, 5, 7, 11, 69, 92, 197 Butterfly, 187, 189, 190
D’Arcy Thompson, 2, 10, 174 Dextran, 30, 38, 39, 41, 42, 43, 44, 46, 48, 49, 50, 51, 52 Diatoms, 4, 10, 11, 36, 123, 179, 180, 181, 196 DNA, 142, 153, 154, 170, 171, 172 Douglas Adams, 196
Cadmium, 70, 86, 140, 143, 144, 158, 163, 168, 178, 198 Calcium carbonate, 1, 7, 8, 9, 11, 30, 70, 71, 76, 77, 94, 95, 105, 107, 108, 121, 122, 129, 182, 190, 198 Calcium oxalate, 197 Calcium phosphate, 11, 77, 79, 81, 95, 97, 98, 99, 101, 108, 127, 130, 182 Carbon, 16 Carrageenan, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 86, 87 CCMV, 156, 198 CdS, 35, 37, 86, 133, 142, 158, 163, 168, 178 Cellulose, 1, 30, 31, 32, 33, 35, 36, 37, 38, 52, 70, 92, 174, 181 Chitin, 1, 11, 92, 93, 94, 95, 189, 190, 191
Fe2O3, 35, 141 Fe3O4, 42, 101, 140, 178, 183 Ferritin, 9, 139, 140, 141, 153, 167, 198 Gelatin, 64, 81, 82, 83, 84, 85, 86, 87, 99 Gellan, 89 George Eastman, 82 Glucose, 14, 16, 22, 23, 24, 25, 29, 38, 52, 88 Gold, 38, 41, 42, 56, 108, 129, 130, 136, 157, 164, 165, 167, 168, 177, 178, 187, 199, 201 205
206
Biotemplating
Housefly, 184 Hydrocolloids, 63, 64, 65, 81, 86, 87, 89 Hydrogen storage, 17, 110 Hydroxyapatite, 10, 20, 21, 22, 35, 43, 77, 81, 83, 85, 97, 98, 99, 125, 126, 127, 130, 133, 134 Iron, 3, 10, 11, 18, 42, 67, 68, 69, 70, 109, 139, 140, 141, 148, 158, 160, 167, 168, 198 Isambard Kingdom Brunel, 3 Johannes Kepler, 2 Joseph Paxton, 2 K2ZrF6, 124 Kodak Process, 82 Lead titanate, 22 Lepidocrocite, 168, 199 LiFePO4, 18 Lipid, 103, 141, 142, 143, 144, 145, 146, 147, 148, 149, 170 Luffa, 177 Lysozyme, 119, 120, 121, 122, 123, 124 Magnetite, 42, 67, 69, 102, 103, 140, 141, 168, 178, 183 Manganese, 70, 140 MCM-48, 16, 17 Mercury, 70, 198 Microcapsules, 81, 101 Monosaccharides, 4, 14 Nanoparticle, 12, 20, 24, 33, 35, 36, 41, 42, 43, 46, 50, 51, 52, 53, 54, 67, 68, 69, 72, 75, 85, 86, 98, 100, 101, 102, 103, 105, 108, 109, 110, 111, 112, 113, 129, 136, 140, 141, 148, 156, 157, 158, 160, 162, 163, 164, 165, 167, 168, 169, 171, 177, 178, 183, 198, 199, 201 Nanorod, 24, 37, 130, 142, 170 Nanotube, 18, 19, 23, 24, 36, 37, 84, 86, 89, 145, 149 Nanowire, 33, 37, 43, 56, 57, 58, 59, 75, 110, 111, 112, 113, 125, 170, 171
Nickel, 23, 53, 67, 68, 69, 70, 105, 108 NiO, 53, 54 Oligosaccharides, 14, 15, 16, 17, 23 Pectin, 64, 89 Platinum, 164, 167, 197 PLZT, 190 PMMA, 187 Pollen, 1, 4, 181, 182, 183, 184 Polypyrrole, 32, 56, 99 Polysaccharides, 28, 29, 30, 59, 63, 70, 80, 82, 89, 92, 117, 149, 154, 162, 172, 174 Porosity, 16, 17, 20, 22, 24, 25, 37, 39, 43, 46, 49, 52, 59, 79, 83, 85, 98, 108, 109, 122, 123, 136, 141, 169, 175, 176, 177, 180, 190, 191 Protein, 1, 7, 10, 21, 97, 117, 118, 119, 120, 121, 122, 123, 125, 127, 129, 130, 133, 134, 136, 139, 140, 141, 149, 153, 154, 155, 156, 157, 160, 162, 163, 164, 165, 167, 168, 172, 178, 196, 197, 198, 199, 200, 201 Pyrrole, 32, 56, 57 R.J.P. Williams, 3 Selenium, 23, 24 SiC, 18, 19, 175, 176 Silaffin, 123, 125 Silica, 1, 16, 18, 20, 24, 36, 51, 71, 84, 85, 87, 88, 98, 123, 125, 127, 129, 131, 141, 145, 158, 160, 168, 170, 177, 179, 181, 182, 189, 191 Silk, 94, 97, 98, 133, 134, 136, 137, 138, 139, 178 Silver, 35, 38, 39, 41, 43, 46, 48, 52, 53, 82, 100, 160, 167, 183, 192, 199 S-layer, 163, 164, 165 Starch, 52, 53, 54, 55, 56, 57, 58, 59, 62 Stephen Mann, 3, 11 Sucrose, 16, 17, 18, 19, 20, 22, 23, 25, 38 Superconductor, 44, 46, 49, 72, 75, 110, 111, 191, 192 Tellurium, 57, 58
Index TEOS, 24, 36, 59, 72, 87, 88, 122, 131, 141, 145, 158, 160, 170 THEOS, 36, 71, 72 Titania, 33, 38, 41, 53, 54, 55, 56, 109, 123, 124, 136, 148, 149, 165, 168, 171, 184 Titanium, 56, 165, 199 TMOS, 72, 125, 131 TMV, 157, 158, 160, 199 Virus, 153, 154, 156, 157, 197, 198, 199 WO3, 110 Wood, 11, 174, 175, 176, 177 World Health Organization, 70
207 Xanthan gum, 30, 64, 65, 89 Y123, 44, 46, 48, 49, 191, 192 Y124, 76, 110, 111, 112 YAG, 85 Zeolite, 17, 49, 50, 51, 169, 175, 177, 198 Zinc, 23, 70, 83, 105, 109, 158, 169, 189 Zirconia, 23, 123, 124, 125, 136 ZnO, 23, 82, 83, 85, 169, 170, 189 ZnS, 103, 104, 109, 201