Mesoporous Crystals and Related Nano-Structured Materials, Volume 148: Proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials, ... (Studies in Surface Science and Catalysis)

Studies in Surface Science and Catalysis 148 MESOPOROUS CRYSTALS AND RELATED NANO-STRUCTURED MATERIALS This Page Int...

Author: Osamu Terasaki

27 downloads 979 Views 21MB 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

DOWNLOAD PDF

Studies in Surface Science and Catalysis 148

MESOPOROUS CRYSTALS AND RELATED NANO-STRUCTURED MATERIALS

This Page Intentionally Left Blank

Studies in Surface Science and Catalysis Advisory Editors: 6.Delmon and J.T. Yates Series Editor: G. Centi Vol. 148

MESOPOROUS CRYSTALS AND RELATED NANO-STRUCTURED MATERIALS Proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials, Stockholm, Sweden, 1-5 June 2004

Edited by

Osamu Terasaki Structural Chemistw Arrhenius Laboratow Stockholm Univetx* Sweden

2004 ELSEVIER Amsterdam - Boston - Heidelberg - London - New York San Francisco - Singapore - Sydney -Tokyo

Stockholm,

- Oxford - Paris - San Diego

ELSEVIER B.V. Sara Burgerhartstmat 25 P.O. Box 211,1000 AE Amsterdam The Netherlands

ELSEVIER Inc. ELSEVIER Ltd 525 B Street, Suite 1900 The Boulevard, Langford Lane San Diego, CA 92101-4495 Kidlington, Oxford OX5 1GB USA UK

ELSEVIER Ltd 84 Theobalds Road London WClX ERR UK

O 2004 Elsevier B.V. All rights reserved. This work is protected under copyright by Elsevier B.V.. and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying. including multiple or systematic copying. copying for advertising or promotional purposes. resale. and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier's Rights Department in Oxford. UK: phone ( 4 4 ) 1865 843830. fax ( 4 4 ) 1865 853333. e-mail: [email protected]. Requests may also be completed on-line via the Elsevier homepage (http:ll www.elsevier.comllocate/permissions). In the USA. users may clear permissions and make payments through the Copyright Clearance Center. Inc.. 222 Rosewood Drive. Danvers. MA 01923. USA; phone: (+I) (978) 7508400. fax: (+I) (978) 7504744. and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS). 90 Tottenham Coua Road. London W1P OLP. UK, phone: ( 4 4 ) 20 763 1 5555; fax: ( 4 4 ) 20 763 1 5500. Other countries may have a local reprographic rights agency for payments.

Derivative Works Tables of contents may be reproduced for internal circulation. but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works. including compilations and translations.

Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work. including any chapter or part of a chapter. Except as outlined above. no part of this work may be reproduced. stored in aretrieval system or transmitted in any form or by any means. electronic. mechanical. photocopying. recording or otherwise. without prior written permission of the Publisher. Address permissions requests to: Elsevier's Rights Department. at the fax and e-mail addresses noted above.

Notice No responsibility is assumed by the Publisher for any injury andlor damage to persons or property as a matter of products liability. negligence or otherwise. or from any use or operation of any methods. products. instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences. in particular. independent verification of diagnoses and drug dosages should be made.

First edition 2004 Library of Congress Cataloging in Publication Data A catalog record is available from the Library of Congress British Library Cataloguing in Publication Data Meeting on Mesoporous Crystals and Related Nano-Structured Materials (2004: Stockholm. Sweden) Mesoporous crystals and related nano-structured materials: proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials. Stockholm. Sweden. 1-5 June 2004. - (Studies in surface science and catalysis; 148) 1. Porous materials C o n g r e s s e s 2. Nanocrystals C o n g r e s s e s 3. Catalysis C o n g r e s s e s I. Title 11. Terasaki. Osamu 660.1'1699

ISBN: 0-444-51720-0 (this volume) ISSN: 0167-2991 @ The paper used in this publication meets the requirements of ANSVNISO 239.48-1992 (Permanence of Paper). Printed in The Netherlands.

Preface The discovery of mesoporous solids has stimulated a wealth of interdependent scientific research over the last decade. Most discoveries of new materials arise by developing an unexpected new appearance with careful observation and thorough enquiry - this is certainly true for mesoporous solids. Mesoporous silica crystals form via self-assembly of amphiphilic molecules in water, and fortunately excellent studies had already been reported on selforganised structures in arnphiphile-water systems that paved the way to understanding the new structural types. A large research effort has not only suggested important and complex catalytic potential for these materials, but has gone much further, yielding new developments that have guaranteed the field of mesoporous materials as an active and important part of material research for years to come. Advances in understanding the formation mechanism and in structural characterisation; recent developments in replication or metal oxide based synthesis; and far-reaching applications in medicinal chemistry and nano-electronics, are but a few of the research areas where mesoporous materials are having an important impact. Commercially, apart from obvious catalytic potential the pores and cages of such materials can be tailored in such a manner to have potential use in both the electronics and pharmaceutical industries that have a huge impact on our society. A recent web search leads to more than 29000 'hits' for "mesoporous materials". This Symposium on Mesoporous Crystals and Related nano-structured Materials aims to highlight the core research that has led to such a fruitful and exciting field. For example we stress the importance of studies conducted long before the discovery of mesoporous materials in the fields of polymorphism in lipids and other self-assembly of surfactant molecules, structural work that led to the beautiful and complex spatial arrangements that we now observe replicated in mesoporous materials, and emphasis will be given to the first preparations of mesoporous materials by researchers from different sides of the world. Through this, we hope that the inspiration that led to the synthesis of such novel materials will be passed on, first hand, to the young researchers that will attend the Symposium and workshop. Although still in its infancy, mesoporous materials research has advanced at a rapid pace. After the discovery in the early 1990s, several groups took up the task of studying at a

fundamental level the synthesis and mechanisms that are involved in the formation of mesoporous materials, finding novel reaction pathways that have led to more complex nanostructures with a wider compositional range. To investigate the structural properties and porous characteristics of such materials the development of characterisation methods was, and still is, vital. Both synthesis and characterisation go hand-in-hand with applications research and in the early days the potential catalytic applications of mesoporous materials played an important role in the growth of this field. To learn about the key advances in these areas, we have invited speakers that in our view played an important role in these processes. We hope that the proceedings of this symposium, published in this booklet, which is composed of the speakers' notes, will become useful not only as important reference material but also as a source of scientific inspiration to future researchers in this field. We are very pleased to hold this symposium, "Mesoporous Crystals and Related nanostructured Materials", at the prestigious Lecture Hall of the Swedish Royal Academy of Sciences and we hope this venue will stimulate open discussion and future collaborations between international researchers attending this meeting. The symposium will be followed by a Workshop at the Department of Structural Chemistry, Stockholm University organised by and for young researchers. The aim of this workshop is to maximise the contact between researchers at an early stage in their scientific careers with world leading experts in an informal but informative atmosphere. I would like to gratefully acknowledge the Swedish Royal Academy of Sciences, the Nobel Committee of Chemistry for the support. Additional funding from following organisations is also acknowledged: Japan Science and Technology Agency (JST), the SSF-funded research school for Colloid and Interface Technology, Carl Tryggers Stiftelse (CTS) and the Bio Nanotech Research Institute (BNRI) Japan.

Osamu TERASAKI Stockholm University, SWEDEN

Stockholm, June 2004

vii

CONTENTS Preface

v

1. The evolution of ordered mesoporous materials Ferdi Schiith

2. The cubic phases of lipids Vittorio Luzzati, Hewe' Delacroix, Annette Gulik, Tadeusz Gulik-Krzywicki, Paolo Mariani and Rodolfo Vargas

17

3. Bicontinuous cubic lipid-water particles and cubosomal dispersions K6re Larsson

4. The discovery of ExxonMobil's M41S family of mesoporous molecular sieves Charles Kresge, James Vartuli, Wieslaw Roth and Mike Leonowicz 5. Discovery of mesoporous silica from layered silicates Kazuyuki Kuroda

6. FSM- 16 and mesoporous organosilicas Shinji Znagaki 7. Integrating interfaces and function with molecular assembly G.D. Stucky and J. Herbert Waite 8. Designer synthesis of mesoporous solids via block copolymer ternplating pathway D.-Y. Zhao, B. -Z. Tian and X.-Y. Liu 9. Significance of mesoporous crystals for catalytic application John Meurig Thomas and Robert Raja 10. Evaporation-induced self-assembly to functional nanostructures Hongyou Fan and Jeffrey Brinker 11. Nanostructured carbon materials synthesized from mesoporous silica crystals by replication Ryong Ryoo and Sang Hoon Joo 12. Structural study of meso-porous materials by electron microscopy Osamu Terasaki, Tetsu Ohsuna, Zheng Liu, Yasuhiro Sakamoto and Alfonso E. Garcia-Bennett Other volumes in the series

53

This Page Intentionally Left Blank

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

The evolution of ordered mesoporous materials Ferdi Schiith Max-Planck-Institut G r Kohlenforschung, 45470 Miilheim, Kaiser-WilhelmPlatz 1,45470 Miilheim, Germany

Since the discovery of ordered mesoporous silica in the early nineties the field has seen a tremendous development. Citation numbers of the seminal papers are above 3000 at present and climbing with a breathtaking rate. One can safely say that the initial discoveries have opened a completely new field of research, i.e. the structuration of inorganic matter on the mesoscale by the use of surfactants and related molecules. In the following pages it will be attempted to trace the major development lines which have emerged in the science of ordered mesoporous materials, with a clear focus on the synthetic methods and the materials obtained. Although I will try to give an objective account, such an attempt necessarily has to be heavily biased on the one hand and incomplete on the other. It is virtually impossible to list in a short survey all the important contributions which have determined the state of the art as we now see it, and the personal involvement in part of the story makes one see some of the developments clearer than others which also makes objectivity impossible. I would therefore like to apologize at this point for not mentioning contributions which other researchers would identi@ as crucial, but the judgement always is influenced by many factors, including personal predilection. There are many other review papers available which give more complete coverage or assessments of the field as seen from a different point of view [I]. The reader is referred to these publications to obtain a more balanced picture. 2.

BACKGROUND

Before ordered mesoporous materials were discovered as such in the early 1990s, a patent had been filed in 1971 [2] in which also the synthesis of a material comparable to MCM-41 was disclosed, although these early researchers did not recognize the true nature of the material produced [3]. Thus, before 1990, the

materials available with pores in the mesopore size range and the highest degree of order were anodic aluminas [4] and controlled pore glasses [ S ] . While the former have relatively narrow pore size distributions, albeit not as narrow as later reached in ordered mesoporous materials, anodic alumina is rather difficult to obtain as a bulk material. In addition, due to the less favourable walllpore ratio, the specific surface areas are typically substantially lower than for ordered mesoporous silica. In contrast, controlled pore glasses can be synthesized also on the commercial scale in larger amounts, however, the pore size distributions, although cutting edge before the discovery of surfactant templated silica, were still rather broad. The pore sizes available in zeolites increased only slowly, with the highest values obtained with the discovery of VPI-5 in the late 1980s, which has pore sizes of around 1.2-1.3 nm [6]. The advent of ordered mesoporous silica in the early nineties was therefore a major breakthrough which has triggered research activities in hundreds of laboratories around the world.

3.

THE EARLY DISCOVERIES

Silica materials with a regularly ordered pore system in the range between 2 and 10 nm and pore walls essentially composed of amorphous silica were independently discovered by two groups, i.e. the group working at Mobil Oil Corporation headed by C.T. Kresge [7,8,9] and the group at Waseda University headed by K. Kuroda [lo]. While the submission date of the Japanese publication was somewhat earlier than the filing date of the first Mobil patent, the pathway described in the publication by Yanagisawa et al. [lo] is rather specific and difficult to generalize, while the early publications of the Mobil group essentially describe a pathway which is generalizable - and has been generalized - to a wide variety of other materials in the years afterwards [l 1,121. The synthesis procedure described by Yanagisawa et al. relies on intercalation of kanemite, a sheet silicate, with surfactants of the alkyltrimethylammonium type. The mechanism suggested for the formation of these types of materials called FSM-n (folded sheet materials, n indicates the number of C-atoms in the surfactant chain) consists of intercalation of the surfactant between the kanemite sheets and subsequent warping of the sheets, followed by inter-sheet condensation by which the hexagonal array of pores is formed (Fig. 1). One may speculate that at the high synthesis pH the kanemite is dissolved and reassembled to give the hexagonal mesostructure under the influence of the surfactants. However, in subsequent studies it was shown that the properties of the kanemite derived materials were substantially different from directly templated materials [13]. There is thus probably indeed an alternative formation mechanism in operation.

Fig. 1: Mechanism for formation of FSM-n type materials. Kanemite is intercalated with the surfactant, then the kanemite sheets are corrugated and fuse by condensation reactions at the contact points (after Ref. [lo]).

Fig. 2: Solution hzed formation of ordered mesoporous silica. The upper pathway corres~ondsto the TI.CT mechanism with direct replication ofthe liquid crystal, the lower to the cooperative mechanism (after Ref. [9]).

However, most of the ordered mesostructured materials are formed via a solution mechanism where oligomeric inorganic species are organized into mesostructured arrays under the influence of surfactant molecules. Possible mechanisms were already suggested in the seminal papers by the Mobil group (Fig. 2) [9]. Principally, one could distinguish two different situations represented in the figure: The assembly could start from the already formed liquid crystal on which silicate species condensate to from the rigid inorganic structure, or a cooperative mechanism may operate, in which surfactant and silica co-assemble to form the organic-inorganic liquid crystal. In fact, in later studies it was found that both modes can be realized, depending on conditions. Attard and coworkers [14] introduced the so-called true liquid crystal templating (TLCT) mechanism, where indeed a preformed liquid crystal was used as the template. Although this liquid crystalline structure is intermittently destroyed and reformed, one could consider this as the prototype of the first pathway. In most cases, however, a cooperative mechanism which was first in more detail described by Monnier et al. [11,15] seems to be in operation. According to this mechanism, multicharged inorganic solution species interact with the surfactant species in solution and cooperatively form a liquid crystalline arrangement. Due to increased concentration and charge screening of the inorganic species in the interface, condensation reactions are facilitated and an extended inorganic network forms. Depending on the solution conditions and the charge matching between organic and inorganic species, eventually the final structures are created. In this early phase, it was also discovered that assembly mechanisms with variously differently charged species can be envisaged, such as positively

charged surfactant/negatively charged inorganic species (s'I-), SI' and also mediated pathways S'X-I' and S-X'I- where the X are ionic species of appropriate charge [12]. Also the concept of the surfactant packing parameter, which had been introduced in surfactant chemistry much earlier by Israelachvili [16], was for the first time applied in the synthesis of ordered mesoporous silica during these early studies [17].

4.

EXTENSION BEYOND SILICA

Already in some of the early publications the extension of the concepts beyond silica frameworks had been predicted [ I l l or even realized [12], although removal of the surfactants was not possible at this time. Substitution of part of the silicon atoms in the framework by other species, however, had already been reported in one of the first papers by the Mobil group, i.e. the formation of aluminum-containing frameworks [9]. Also titanium has been incorporated already very early by the groups of Pinnavaia [IS], Schulz-Ekloff [19] and C o m a [20], and subsequently a whole range of different metal ions were introduced into the framework of ordered mesoporous materials [Id]. The expectation in these studies, where part of the silica was substituted by heteroatoms, was the creation of materials analogous to fiamework-substituted zeolites. However, the materials rather resemble amorphous silicas containing heteroelements, and no materials with clearly superior catalytic properties seem to have been obtained, yet. The breakthrough with respect to fully non-siliceous frameworks was achieved in the middle of the 1990s [21,22]. Key to these developments was the fact that rather redox stable early transition metal oxides, such as titania or zirconia, were chosen for these approaches. All attempts to the synthesis of non-silica based materials face two major obstacles, in addition to the requirements that the solution chemistry has to be compatible with the assembly and condensation of the mesophase [lm]: (i) redox instability makes the template removal via calcination very difficult, since intermediate reduction/oxidation processes will lead to collapse of the structures; (ii) most materials crystallize much easier than silica, but crystalline structures are often not compatible with the high curvatures which need to be accommodated. However, in spite of these severe constraints many different framework compositions could be synthesized over the following years in the form of ordered mesoporous materials, which are reviewed in Ref. [lml. One example which should be specifically mentioned in this connection is the synthesis of ordered mesoporous metals by the group of Attard [23], because this substantially went beyond the oxidic frameworks synthesized thus far. The

synthesis relies on the TLCT approach. A preformed liquid crystalline phase was impregnated with a noble metal precursor which was then - while being present in the mesostructured state - reduced by a soluble reducing agent, such as hydrazine, or lumps of less noble metals which were added to the solution. After removal of the surfactant by a washing process (non-ionic surfactants are used in this process which can relatively easily be removed) a mesostructured, highly porous noble metal is obtained (Fig. 3). The textural properties such as surface area and porosities did not exceed those of Raney metals, but conceptually, this developmcnt was a major advance.

Fig. 3: SEM image of a mesoporous Pi-sample obtained via the true liquid crystal ternplating pathway (left) and TEM of such a sample (right) (from Ref. [23]).

Another major development line in the synthesis on non-silica materials was the creation of ordered organosilicas. Mesoporous silica containing pending organic groups on the surface had already been described in the seminal publications of the Mobil group [9]. In this study the groups were grafted by post-treatment via the conventional techniques used for the silanization of silica. Later ordered mesoporous silica with pending organic groups was prepared by hydrolysis and co-condensation of tetraethoxysilane and alkyltrialkoxysilane [24,25]. A true organosilica, however, was only realized in 1999 when three groups independently used silsesquioxanes of the type (RO)3Si-R'-Si(OR)3 which were assembled to mesostructures in the presence of surfactants [26, 27,281. Even with 100 % of organically bridged silicon atoms in the materials the structural perfection was still very high. Going even beyond the structural definition of the purely inorganic silica, organosilica synthesized with phenyl-bridged alkoxysilane moieties had periodicity within the walls, which is probably brought about by the x-stacking of the aromatic rings as integral part of the wall structure [29].

5.

POLYMERIC SURFACTANTS AS TEMPLATES

Parallel to the development of the non-silica frameworks another series of important discoveries was made and exploited, the use of neutral, polymeric surfactants, which expanded again the range of structures and framework compositions which could be synthesized in mesostructured form. The first publications in which neutral surfactants were used, came from the group of Pinnavaia [30,31] and Attard [14]. They used poly (ethylene oxide), Tergitol and Pluronics type surfactants. These studies demonstrated that it is possible to produce regular mesoporous silicas also in the absence of Coulombic interactions. The predominant forces if neutral surfactants are used are hydrogen bonding interactions. A big advantage of these surfactant types is the possibility to recover them by simple extraction procedures, which is not possible for the charged systems. However, the development with the most far reaching consequences was probably the synthesis of SBA-15 and related materials using the triblockcopolymer Pluronics P123 [32] which was discovered by the Santa Barbara team. These polymers contain blocks of poly (ethylene oxide) and poly (propylene oxide) with the general formula (PEO),(PPO),(PEO), where the PEO blocks are the hydrophilic parts and the PPO block is hydrophobic. SBA-15 has appreciably thicker walls as compared to MCM-41, the thickness of which can be tuned to some extent, and can be synthesized with much wider pores up to 20 nrn pore width. In addition, the walls of these materials contain a substantial fraction of micropores. Since the mesopores and the micropores can be made accessible in two separate steps [33], independent fbnctionalization of the different pore systems is possible, which in principle allows to design catalysts with spatially defined and separated functionalities. Following the initial synthesis of SBA-15, a similar synthesis strategy was devised by the groups of Stucky and Chrnelka to create non-silica systems with triblockcopolymers [34,35]. Due to the thicker walls, such materials could also accommodate crystalline phases in their walls, which are, for instance, present as small particles of crystalline transition metal oxides. The variability of the triblockcopolymer chain also allowed the creating of novel structures, such as SBA- 16 [36]. For these materials an innovative TEM technique has been used to directly elucidate the real space structure of mesostructured solids [37]. Fig. 4 shows this, as an example for a novel structure synthesized with triblockcopolymers as well as to illustrate the success of the TEM structure determination. Due to the superior properties of SBA-15 type materials compared to MCM-41 in various aspects, it seems that a substantial fraction of

Fig. 4: Direct image of 3D pore strnclure of SBA-16. a, Electrostatic potential map of SBA-I6 parallel to (110) lhrough the centrc of the cell. b, 3D arrangement of a cavity and its interconnection. Black corresponds to the cavity. c, Mean cavity surface for SBA-16 (from Kef. [37]).

the research activities nowadays focuses more on SBA-15 type materials than on MCM-41 and related solids. 6.

NANOCASTNG

About one year after the discovery of SBA-15 and related materials the group of Ryoo reported the first successful "nanocasting" (sometimes also called "hard templating") of a negative carbon replica, which was called CMK-1, from siliceous MCM-48 [38]. Independently the group of Hyeon synthesized a negative replica of HMS-type materials [39]. This process is surprisingly simple, although it is rather astonishing that it does work at all: A siliceous mesoporous material with a three-dimensionally connected pore system is filled with a carbon precursor. Originally, the precursor was sucrose solution, but by now various different precursor species have been used, such as furfuryl alcohol, resorcinol-formaldehyde resin, poly (acrylonitrile), acetylene and various others. In a next step, the precursors are converted to carbon by high temperature treatment in inert atmosphere, and in a last step, the silica matrix is removed by leaching with HF or NaOH. In the ideal case, this results in the formation of the negative replica of the original material. For the MCM-48 structure, however, the symmetry is not that of the parent, but is reduced due to the fact that the carbon frameworks which have filled the two independent pore systems of the MCM-48 precursors shift with respect to each other. Later on, a material in which the original space group was retained has been reported. This material had been created by using acetylene as the carbon source in a high temperature gas phase process [40].

On first sight surprisingly, also SBA- 15 can be obtained as a carbon negative by a nanocasting process, although it consists of a hexagonally packed array of cyclindrical channels. However, the tubes are connected with each other through the micropores in the walls, which provides the three-dimensional crosslinking necessary to obtain a replica with sufficient rigidity. The resulting carbons are called CMK-3 [41]. A special kind of cast from the SBA-15 structure is obtained if the channels of the SBA-15 are not filled completely, but only the walls are coated. Carbonization of the precursor and removal of the silica then gives a material with two independent pore systems which could be considered as a hexagonal packing of carbon nanotubes called CMK-5 (Fig. 5) [42]. However, one should keep in mind that the carbon in the walls is less defined than the well known single-walled or multi-walled carbon nanotubes. The materials have very high surface areas for non-microporous carbons (up to about 2500 m2/g). In

Fig. 5: TEM of an CMK-5 type carbon obtained by wall coating of the SBA-15 channels, carbonization and subsequent leaching of the silica.

addition, by varying the thickness of the coating or the wall thickness of the parent SBA-15, the two pore systems of the materials can be tailored independently [43]. This opens interesting routes for the creation of materials with spatially separated and defined chemical functionality, for instance, a hydrophobic and a hydrophilic pore system. Related to the nanocasting of carbon replicas is the replication of the pore structure of ordered mesoporous materials by noble metals [44]. Here a volatile noble metal precursor is evaporated into the pore system of the mesostructure and then pyrolized to form the noble metal replica. However, this technique was

more used as an analytical tool to elucidate details of the pore structure than for the production of materials. The carbons, which can be created by a nanocasting process, are in turn excellent hard templates for the production of other ordered mesoporous materials by a second nanocasting step. So far, this was only successhl for silica and this possibility was discovered at about the same time independently by two groups [45,46]. The perfection of the material after twice repeated nanocasting is surprisingly high (Fig. 6). The textural parameters of the positive replica after the repeated nanocasting are somewhat altered compared to the original template, which is due to the shrinkage occurring during each high temperature treatment step.

Fig. 6 : TEM images of original SBA-15 (left) and silica NCS-1 (right), obtained by replicating the SBA-15 first as carbon CMK-3 and then replicating the carbon by infiltration with tetraethoxysilane, hydrolysis, condensation and oxidative removal of carbon.

Nanocasting and repeated nanocasting are two relatively recent methods for the creation of ordered mesoporous materials. However, it can be expected that these techniques will become as versatile as the original solution phase methods to create materials with novel and fascinating properties. 7.

APPLICATIONS

This attempt to sketch the major development lines in the science of ordered mesoporous solids is strongly focussed on the synthetic aspects of such materials. However, at least some of the application perspectives should be sketched as well in this section.

It would be impossible to list all the applications of ordered mesoporous materials which have been proposed over the last ten years. Especially investigations into the catalytic properties are appearing at an astonishing rate. Even a first commercial application by Exxon-Mobil has been announced in 2002 during the 2nd IMMS by J. Vartuli, but no specifics were given. The developments in the field of catalysis are well covered in several review papers[ la,l,m,p] and addressing catalytic applications would by far exceed this short account. Some other interesting developments, however, will be highlighted in the following paragraphs. Applications in optics were made possible by the creation of well defined fibers [47]. The mesopores in these fibers can be loaded with laser dyes and the materials can then be used as fiber lasers [48,49]. The potential in optical applications, which such materials may have, can be fully realized only if the fibers could be integrated into extended structures. The possibility to do this, however, is dependent on supporting the lasing fibers on low refractive index materials to maintain the waveguiding properties. Such low index materials were, in fact, found in the calcined ordered mesoporous materials themselves [50] and first prototype structures were fabricated. The low refractive index is correlated with a low dielectric constant. Low-k materials are a central development target in the electronics industry. Ordered mesoporous materials seem to be very promising in such applications, since they contain a large void fraction and can be made hydrophobic to exclude water both of which substantially reduce the dielectric constant below the one of silica ( b u i ~vaiue: 3.8) Vaiues which have been reported are as iow as about i .5 [ 5 i]. Together with the excellent processability [52] these data suggest that ordered mesoporous materials may indeed find commercial uses in the semiconductor industry. Several other application perspectives have been published, such as for instance remediation of heavy metal ions by thiol hnctionalized materials [53]. However, it will be easiest to introduce the materials in "high-tech" applications since the costs of ordered mesoporous solids compared to cheaper alternatives will make it difficult to commercialize ordered mesoporous materials in high volume/low added value applications. 8.

SUMMARY

In the previous sections I have tried to sketch the major development lines in the field of ordered mesoporous materials, starting with the early work on silica mesophases, the subsequent expansion of the inorganic fi-amework composition, the introduction of polymeric surfactants with their much higher degree of

variability, and finally the nanocasting, which adds another new dimension to the creation of ordered mesoporous materials. It is clear that a whole new research field has been opened by the work done in the early 1990s, and it will be interesting to see how many novel developments, hopehlly also in commercial applications, will be realized in the future.

REFERENCES [I] a) A. Corma, Chem. Rev., 97 (1997) 2373; b) U. Ciesla, F. Schuth, Micropor. Mesopor. Mater., 27 (1999) 131; c) M. LindCn, S. Schacht, F. Schuth, A. Steel, K. Unger, J. Porous. Mater., 5, (1998) 177; d) A. Tuel, Micropor. Mesopor. Mater., 27 (1999) 151; e) P. Selvam, S. K. Bhatia, C. G. Sonwane, Ind. Eng. Chem. Res., 40 (2001) 3237; f) Y. Liu, T. J. Pinnavaia, J. Mater. Chem., 12 (2002) 3179; g) R. Anwander, Chem Mater., 13 (2001) 4419; h) J. Y. Ying, C. P. Mehnert, M. S. Wong, Angew. Chem. Int. Ed., 38 (1999) 56; i) A. P. Wight, M. E. Davis, Chem. Rev., 102 (2002) 3589; j) D. E. de Vos, M. Dams, B. F. Sels, P. A. Jacobs, Chem. Rev., 102 (2002) 3615; k) G. J. de A. A. Soler-Illia, C. Sanchez, B. Leveau, J. Patarin, Chem. Rev., 102 (2002) 4093; 1) D. T. On, D. Desplantier-Giscard, C. Danumah, S. Kaliaguine, Appl. Catal. A: Gen., 253 (2003) 545; m) A. Wingen, F. Kleitz, F. Schuth, in: M. Baems (Ed.), Basic Principles in Applied Catalysis, Springer, Berlin, 2003, pp. 281-319; n) F. Schuth, Chem. Mater., 13 (2001) 3184; o) F. di Renzo, A. Galarneau, P. Trens, F. Fajula, in: F. Schuth, K.S.W. Sing, J. Weitkamp (eds.), Handbook of Porous Solids, Wiley-VCH, Weinheim, 2002, p.1311-1394; p) A. Taguchi, F. Schuth, Micorpor.Mesopor.Mater., in press. [2] V. Chiola, J. E. Ritsko, C. D. Vanderpool, US patent No.3556725 (1971). [3] F. DiRenzo, H. Cambon, R. Dutartre, Micropor. Mesopor. Mater., 10 (1997) 283. [4] For a survey, see: D. Honicke, E. Dietzsch, in: F. Schuth, K.S.W. Sing, J. Weitkamp (eds.), Handbook of Porous Solids, Wiley-VCH, Weinheim, 2002, pp. 1395-143 1. [5] for a survey, see: F. Janowski, D. Enke, in: F. Schuth, K.S.W. Sing, J. Weitkamp (eds.), Handbook of Porous Solids, Wiley-VCH, Weinheim, 2002, pp. 1432-1590. [6] M.E. Davis, C. Saldarriaga, C. Montes, J. Garces, C. Crowder, Zeolites, 8 (1988) 362. 171 J.S. Beck and N.Y. Princeton, US Patent No. 5 057 296, 1991. [8] C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli, J. S. Beck, Nature, 359 (1992) 710. [9] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.-W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, J. Am. Chem. Soc., 114 (1992) 10834. [lo] T.Yanagisawa, T. Shimizu, K. Kuroda, C. Kato, Bull. Chem. Soc. Jpn., 63 (1990) 988. [ l l ] A. Monnier, F. Schuth, Q. Huo, D. Kumar, D. Margolese, R.S. Maxwell, G. Stucky, M. Krishnamurty, P. Petroff, A. Firouzi, M. Janicke, B. Chmelka, Science, 261 (1993) 1299. [12] Q. Huo, D. Margolese, U. Ciesla, P. Feng, T. Gier, P. Sieger, R. Leon, P.M. Petroff, U. Ciesla, F. Schuth, G. Stucky, Nature, 368 (1994) 317. 1131 a) S. Inagaki, Y. Fukushima, K. Kuroda, J.Chem.Soc., Chem.Sommun. (1993) 680; b) J.C. Vartuli, C.T. Kresge, M.E. Leonowicz, A.S. Chu, S.B. McCullen, I.D. Johnson, E.W. Sheppard, Chem.Mater, 6 (1994) 2070. [14] G.S. Attard, J.C. Glyde, C.G. Goltner, Nature, 378 (1995) 366.

1151 Q. Huo, D. Margolese, U. Ciesla, D. Demuth, P. Feng, T. Gier, P. Sieger, A. Firouzi, B. Chmelka, F. Schuth, G.D. Stucky, Chem. Mater., 6 (1994) 1176. [16] J.N. Israelachvili, D.J. Mitchell, B.W. Ninham, J.Chem.Soc.Faraday Trans. 11, 72 (1976) 1525. [17] Q. Huo, D.I. Margolese, G.D. Stucky, Chem.Mater. 8 (1996) 1147. [IS] P.T. Tanev, M. Chibwe and T.J. Pinnavaia, Nature, 368 (1994) 321. [19] 0 . Franke, J. Rathousky, G. Schulz-Ekloff, J. Starek, A. Zukal, Stud.Surf.Sci.Cata1 84 (1997) 77. [20] A. Corma, M.T. Navarro and J. Perez Pariente, Chem. Commun., (1994) 147. [21] D.M. Antonelli, J.Y. Ying, Angew.Chem.Int.Ed., 34 (1995) 2014. [22] U. Ciesla, S. Schacht, G.D. Stucky, K.K. Unger, F. Schuth, Angew.Chem.Int.Ed., 35 (1996) 541. [23] G.S. Attard, C.G. Goltner, J.M. Corker, S. Henke, R.H. Templer, Angew.Chem.Int.Ed., 36 (1997) 1315 1241 S.L. Burkett, S.D. Sims, S. Mann, Chem.Commun., (1996) 1367. [25] D.J. Macquarrie, Chem.Commun., (1996) 1961. [26] S. Inagaki, S. Guan, Y. Fukushima, T. Ohsuna, T. Terasaki, J.Am.Chem.Soc., 121 (1999) 961 1. [27] B.J. Melde, B.T. Holland, C.G. Blanford, A. Stein, Chem.Mater., 11 (1999) 3308. [28] T. Asefa, M.J. MacLachlan, N. Coombs, G.A. Ozin, Nature, 402 (1999) 867. [29] S. Inagaki, S. Guan, T. Ohsuna, 0 . Terasaki, Nature, 416 (2002) 304. [30] S.A. Bagshaw, E. Prouzet, T.J. Pinnavaia, Science, 269 (1995) 1242. [31] E. Prouzet, T.J. Pinnavaia, Angew. Chem. Int. Ed., 36 (1997) 516. [32] D. Zhao. J. Feng, Q. Huo, N. Melosh, G. H. Fredrickson, B. F. Chmelka, G. D. Stucky, Science, 279 (1998) 548. [33] C.-M. Yang, W. Schmidt, F. Schuth, Chem.Mater., 15 (2003) 3739. [34] P.D. Yang, D.I. Zhao, D.1 Margolese, B.F. Chmelka, G.D. Stucky, Nature, 296 (1998) 152. [35] P. Yang, D. Zhao, D. I. Margolese, B.F. Chmelka, G. D. Stucky, Chem. Mater., 11 (1999) 2813. [36] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J.Am.Chem.Soc., 120 (1998) 6024. [37] Y. Sakamoto, M. Kaneda, 0. Terasaki, D.Y. Zhao, J.M. Kim, G. Stucky, H.J. Shin, R. Ryoo, Nature, 408 (2000) 449. [38] R. Ryoo, S. 1-1. Joo, S. Jun, J. Phys. Chem. B, 103 (1999) 7743. [39] J. Lee, S. Yoo, T. Hyeon, S.M. Oh, K.B. Kim, Chem.Commun., (1999) 2177. [40] M. Kaneda, T. Tsubakiyama, A. Carlsson, Y. Sakamoto, T. Ohsuna, 0. Terasaki, S.H. Joo, R. Ryoo, J.Phys.Chem. B, 106 (2002) 1256. [41] S. Jun, S.H. Joo, R. Ryoo, M. Kruk, M. Jaroniec, Z. Liu, T. Ohsuna, 0. Terasaki, J.Am.Chem.Soc., 122 (2000) 10712. [42] S.H. Joo, S.J. Choi, I. Oh, J. Kwak, Z. Liu., 0. Terasaki, R. Ryoo, Nature, 412 (2001) 169. [43] A. Lu, W. Schmidt, B. Spliethoff, F. Schuth, Adv.Mater., 15 (2003) 1602. [44] H. Kang, Y.-W. Jun, J.-I. Park, K.-B. Lee, J. Cheon, Chem.Mater., 12 (2000) 3530. [45] A. Lu, W. Schmidt, A. Taguchi, B. Spliethoff, B. Tesche, F. Schuth, Angew.Chem.Int. Ed., 41 (2002) 3489.

[46] M. Kang, S.H. Yi, H.I. Lee, J.E. Yie, J.M. Kim, Chem.Commun., (2002) 1944. 1471 Q. Huo, D. Zhao, J. Feng, K. Weston, S.K. Buratto, G.D. Stucky, S. Schacht, F. Schiith, Adv. Mater., 9 (1997) 974. 1481 F. Marlow, M. D. McGehee, D. Zhao, B. F. Chmelka, G. D. Stucky, Adv. Mater., 11 (1999) 632. [49] J. Loerke, F. Marlow, Adv.Mater., 14 (2002) 1745. [50] P. Yang, G. Wirnsberger, H.C. Huang, S.R. Cordero, M.D. McGehee, B. Scott, T. Deng, G.M. Whitesides, B.F. Chmelka, S.K. Buratto, G.D. Stucky, Science, 287 (2000) 465. [511 a) D. Zhao, P. Yang, N. Melosh, J. Feng, B.F. Chmelka, G.D. Stucky, Adv.Mater., 10 (1998) 1380; b) C.M. Yang, A.T. Choi, F.M. Pan, T.G. Tsai, K.J. Chao, Adv.Mater., 13 (2001) 1099. [52] D.A. Doshi, N.K. Huesing, M.C. Lu, H.Y. Fan, Y.F. Lu, K. Simmons-Potter, B.J. Potter, A.J. Hurd, C.J. Brinker, Science, 290 (2000) 107. [53] X. Feng, G.E. Fryxell, L.Q. Wang, A.Y. Kim, J. Liu, K.M. Kemner, Science, 276 (1997) 923.

This Page Intentionally Left Blank

2. "Lipid polymorphism" Professor Vittorio Luzzati CNRS, Centre de GknLtique Molkculaire, FRANCE

1. V. Luzzati and A. Tardieu: "Lipid phases: structure and structural transition", Ann. Rev. Phys. Chem. (1974), 25,79-84 2. Vittorio Luzzati, H. Delacroix, A. Gulik-Krzywicki, P. Mariani and R. Vargas, "The cubic phases of lipids", Current Topics in Membranes, (1997), 44,3-24 These two review papers have been suggested, and number 2 (above) is as follows.

This Page Intentionally Left Blank

Shldies in Surface Science and CaWysis 148 Terasaki (Editor) 0 21104 Rlrcvier R.V. All rights reserved.

The Cubic Phases of Lipids Vittoriu Lurrati,* Herv6 Delacruix,* Annette Gulik,* Tadeusz Gulik-Knywicki,* Paolo Mariani,' and Rodolfo Vargas* *Cenlre de GCnCtique Mol6culaire. Laboratoire Propre du CNRS Associt B I'Universitt Pierre et Marie Curie, 91198 Gif-sur-Yvette Cedex, France, 'Istitnto di Scienze Fisiche and Islituto Nuionale per la Fisica della Materia, Universilb di Ancona, 60131 Ancona, Italia, and 'Centm d r Biofisica y Bioquimica, IVIC, Caracas 19020-A, Venezuela

I. Introduction U. Structure Analysis A . Crystallographic Techniques B. Freeze~EracturcElcclron Microscopy C. Lateral Diffusion Studies 111. Chemical Propcrlics 1V. Structure Keprcscntations A. Bicontinuous Phases: Networks of Rods, Convoluted Surfaces B. Micellar Phases V. Orderly Disposal of Short-Range Conformational Disorder: The Chaotic Zones VI. Biologjcal Implications Abbreviations References

I. INTRODUCTION

Phases with cubic symmetry have been observed in lipid-water systems since the early days of lipid polymorphism, at a time when it seemed all too natural to presume without closer inspection that phases with such high symmetry should consist of spherical micelles of type I or 11,orderly packed in a cubic lattice (Luzzati et ul.. 1960; Luzzati and Reiss-Husson, 1966). Reprinted fmm Lipid Polymorphism and Mcmhane Propmics, Volume 44: Cunrnt Topics in Membranes, V. Lurrati, H. Delacroin. A. tiulik, r. Gulik-Kizywick~.P. Marioni andR Vargas; The Cubic Phases orllpids, pp. 3-24. 1997. wit11permission fmm Elsrvicr

Eventually, careful crystallographic analyses revealed that cubic symmetry is the attribute of not one lipid phase but of a large family of phases. Moreover, the structure of these phases, which took quite a few years to determine, turned out to be unexpectedly complex and multifarious. At the date of the present review, seven cubic phases have been identified and their structure (at least tentatively) described: 1. Q230(space group Ia3d). This was the first of the cubic phases properly identified, and the first whose structure was firmly established (Luzzati and Spegt, 1967;Luzzati etal., 1968b;Delacroix etal., 1990,1993b).The structure (Fig. 1) was originally described in terms of two 3D networks of rods, mutually intertwined and unconnected, with the rods joined co-planarly three by three; each network is chiral, and the two are mirror images of each other. The surfaces of the rods may be visualized to sit at the polar-apolar interface; in some of the systems the interiors of the rods are occupied by the hydrocarbon chains and the interstices are filled by the polar medium (structure of type I), whereas in others the relative distribution of the polar and the apolar media is the reverse (type 11). Luzzati et al. (1968~)pointed out that the structure is "bicontinuous," in the sense that it consists of two media of the same polarity, separated by a medium of the opposite polarity, and that the three media are continuous throughout the 3D space. In 1976, Scriven put forward an alternative representation of the structure, in terms of the G-type IPMS (Schwarz, 1880; Schoen, 1970). 2. Q224 (space group Pn3m). This phase was originally identified and its structure analyzed by Tardieu (1972) (see also Tardieu and Luzzati, 1970).

FIGURE 1 Representation of the bicontinuous phases in terms of a pair of 3D networks of rods (drawn respectively in black and white), mutually intertwined and unconnected, that represent the skeletal graphs of the IPMS. QZ3O: the rods are joined co-planarly three by three. Qzz4: the rods are joined tetrahedrally four by four.

In 1983 Longley and McIntosh independently observed this phase and put forward a structure in close agreement with Tardieu's early proposal. The structure (Fig. I), like that of phase Q230,was originally described in terms of two networks of rods, mutually intertwined and unconnected. The rods in this phase are tetrahedrally joined four by four. This structure, like that of Q230,is bicontinuous and, as pointed out by Longley and McIntosh (1983), can also be described in terms of the IPMS (D-type). 3. Q229(space group lm3m).This phase is often mentioned in lipid literature, frequently in three-component systems (lipid-protein-water [GulikKrzywicki et al., 1984; Mariani et al., 19881; surfactant-oil-water [Barois et al., 1990; Maddaford and Topragcioglu, 1993]), although Caffrey (1987), Kekicheff and Cabane (1987), and Mirkin (1992) mention phases with this symmetry in a variety of lipid-water systems. The XRS study of a phase of this symmetry belonging to a ganglioside-water system (Gulik et al., 1995) has shown that the structure consists of identical quasi-spherical micelles packed in the cubic body-centered mode (represented in Fig. 2 in terms of space-filling polyhedra). It is worthwhile to stress that phase Q229 is often presented as a paradigm of the IPMS (P-surface, the "plumber's nightmare" cartoon), a claim that so far lacks firm experimental support (Luzzati et al., 1993; Luzzati, 1995). 4. Q225(space group Fm3m). This phase has been observed in the system C12EO12-water(CI2EOl2is a polyethylene glycol surfactant) (Mirkin, 1992) and in several ganglioside-water systems (Gulik et al., 1995). The structure, studied by XRS and FFEM methods, has been shown to consist of identical quasi-spherical micelles close-packed in the cubic face-centered mode (represented in Fig. 2 in terms of space-filling polyhedra) (Gulik et al., 1995). 5. Q223(space group Pm3n). This phase was discovered by Balmbra et al. (1969) in DTAC and studied by Tardieu and Luzzati (1970) in a variety of lipid-water systems. After some hesitation, and in agreement with earlier proposals (Fontell et al., 1985; Eriksson et al., 1987; Charvolin and Sadoc, 1988), XRS (Vargas et al., 1992) and FFEM (Delacroix et al., 1993a) studies showed that the structure contains two types of micelles, one quasi-spherical the other disk shaped (represented in Fig. 2 in terms of space-filling polyhedra). 6. Q227(space group Fd3m). This phase was discovered by Tardieu (1972) and described by Mariani et al. (1988). The phase has been observed in a variety of lipid-water systems, in all of which the lipid component is highly heterogeneous (Luzzati et al., 1992). The structure was first, and incorrectly, interpreted in terms of a 3D network of rods and a set of micelles (Mariani et al., 1990); a more recent analysis revealed that, in keeping with an early proposal by Charvolin and Sadoc (1988), the structure contains two types of quasi-spherical micelles (represented in Fig. 2 in terms of space-filling

mGURE 2 Representation of the micellar phases in terms of space-filling polyhedra. QZz5:face-centered cubic packing of regular rhombic dodecahedra. 0ZZ9: body-cmterrd cubic packing of regular truncated octahedra. QU3: primitive cubic packing of distorted dodecahedra face-centered cuhic packing of distorted dndccahcdra and hcxadeand tetradecahedra. QZz7: cahedra. (Redrawn with permission from Williams, R. [197Y]. "The Geometrical Foundation of Nurural Srrucrure." Duver Puhlicalions. Ncw York.)

polyhedra) (Lu7zati e l al., 1992). This structure has been confirmed by a rccent ETEM study (Delacroix et aL, 1996). 7. Q2" (space group P43.Z2). This phase was reported in the system monoolein-water-cytochrome C (Mariani et al., 1988). Those authors proposed a structure formed by a 3D nctwork olrods enclosing a set of identical miccllcs, each of which contains one protein molecule. It is appropriate to mention another phase of a lipid-water system that consists of identical quasi-spherical micellcs, in spitc of the fact that its symmetry is not cubic but hexagonal (space group P63/mmc) (Clerc, 1996). We discuss in this paper the structural and physical propcrties of the cubic phases, and also of other lipid phases of lower symmetry, in a scarch

for general criteria applicable to all lipid phases. We also tackle the longlasting problem of the correlations between the chemical composition of the lipid component and the physical structure of the phases. Finally, we envisage the possible biological implications of lipid polymorphism, with special emphasis on the cubic phases.

11. STRUCTURE ANALYSIS

Determining the structure of lipid phases, especially those of type a (i.e., with a disordered short-range conformation of the hydrocarbon chains), is a peculiar exercise, whose rules are rather different from those that prevail in other areas of structural chemistry. A few techniques are available.

A. Crystallographic Techniques

Although single crystals can sometimes be obtained (Clerc, 1996), their use is rarely compatible with the requirements of operating at variable and controlled conditions (usually temperature and water content). Most often, therefore, the experiments are performed on polycrystalline aggregates and the X-ray scattering data take the form of Debye-Scherrer rings. As a consequence, the determination of the space group is sometimes ambiguous (Mariani et al., 1988). Moreover, the highly disordered short-range conformation limits the resolution of the scattering data, and the number of observed reflections is small. Determination of the phase angle of the reflections cannot rely upon conventional methods, since heavy atom derivatives are rarely available in lipid systems. Trial-and-error procedures can only be used when the shape of the structure elements is sufficiently simple to be described in terms of simple geometric models (Gulik et al., 1995). In some cases these drawbacks can be overcome by a systematic study as a function of water content; as a rule, however, the concentration range of cubic phases is too narrow to implement that technique. Another approach is based upon a pattern recognition procedure, based on the axiom that, if two phases are available and if their chemical composition (nature of the lipids and water concentration) is the same, then the histograms of the electron density maps, properly normalized in shape and scale, are identical (Luzzati et al., 1988; Mariani et al., 1988; Luzzati et al., 1992;Vargas et al., 1992). The unknown structure of phase Xis thus "solved" by seeking the map compatible with the observed amplitude of the reflections whose histogram best agrees with that of the reference phase. Physical and chemical information is then retrieved from that map. By way of

verification, it is of the utmost importance to ascertain that the map is compatible with the physical and chemical properties of the system, especially when these properties are not involved in the crystallographic analysis. B. Freeze-Fracture Electron Microscopy

One of the novelties regarding lipid-water phases is the impact of imagefiltering FFEM. FFEM has been applied to lipid-containing systems for more than 25 years; combined low-temperature XRS and FFEM techniques, moreover, have been introduced (Gulik-Krzywicki and Costello, 1978) to avoid the effects of freezing artifacts. Although the FFEM images of the cubic phases are too complex to provide much useful information upon plain visual inspection, the presence of extended quasi-periodically ordered areas justifies the use of image-filtering techniques based upon correlation averaging (Saxton and Frank, 1977; Frank et al., 1978). The results yield a remarkable wealth of information. Image-filtering procedures have been applied to a variety of cubic phases: the bicontinuous phase Q230of types I and I1 (Delacroix et al., 1990,1993b); the micellar phases Q223(Delacroix et al., 1993a) and Q22sof type I (Gulik et al., 1995), the micellar phase Q227of type I1 (Delacroix et al., 1996). For structures of type I, in which the fracture occurs at the lipid-water interface, the filtered images display remarkable correlations with the electron density maps (Figs. 3 and 4; see also Delacroix et al., 1990, 1993a). For structures of type 11, in which the fracture occurs over the CH3-decorated surfaces, the comparison is less straightforward. Moreover, the images provide valuable information regarding symmetry. For example, Fig. 4 reveals the presence of mirror planes that escape XRS detection and help narrow down the choice of the space group, and also of fourfold mirror planes. C. Lateral Difision Studies

The diffusion properties of different lipid phases, cubic as well as noncubic, are related to the topological and geometric properties of their structures. Field-gradient NMR experiments have demonstrated that the lateral diffusion of the lipid molecules is much easier in phase Q230than in phase Q223(Charvolin and Rigny, 1973; Eriksson et al., 1987), whereas in other phases the results are not as clearcut (Hendrykx et al., 1994; Lindblom and Oradd, 1994; see also Chapter 3, this volume). FRAP experiments have shown that the lateral diffusion is fast in phases Q230and Q224and quite slow in phases Q223and Q227,in keeping with the bicontinuous structure

FIGURE 3 FFFM of ~ h a s cQZ30, twe .. I. Fracture lane .12111.. (A) . . Electron microeraoh - . of arotatory-shadowed rcplica. Note the presence of subdomains (labeled 0 to 3) corresponding to seauential steos of frcczc-fracture. Inset. Correlation averare of subdomain 2. IBj , , Crosscorrelation map showing the relative shift of the different suhdomains. (C) Sequence of sections of the electron density map normal to [211] mutually spaced by a1d6; the dots mark the projection of the origin of the unit cell. Note the similarity helwcen the correlationaveraged motif (inset in panel A) and the corresponding section of the electron density map 2, and also the faithful representalion of the sliding of the carrclation maps (panel R). (Reprinted with permission from Culik-Krzywicki, T., and Dclacroix, H. 119941. Combincd use of freeze-fracture electron microscopy and X-ray diffraction for thc structure determination of three-dimensionally ordered specimens. Biol. Cell 80, 193-201.)

of phases QZ30 and Q2%and the micellar structure of phases Qa3 and Q"' (Cribier et al., 1993).

Ill. CHEMICAL PROPERTIES

Phases QUO, QU4, QU9, QnS,and QZ2? have been observed in lipid-water systems containing a chemically homogcncous (or moderately hctcrogeneous: e.g., egg lecithin) lipid component. In contrast, phase Q227 seems to require a mixture of polar (MO, FA salt, PL) and apolar (FA, DAG) lipids (1.wzati et al., 1992). Phase QZz9 has been reported also in more

FIGURE 4 FFEM of phasc QZ2', type 1. Fracture plane [lM)]. (A) Electron micrograph of a rotatory-shadowed replica, with the subdomains (labeled 1 and 2) corresponding to sequential steps of freeze-fracture. (B) Quasi-optical Fourier filtrring of the whole frame (panel A) revealing the content and boundaries of the suhdomains. Inset, Cross-corrclationaveraged motifs, determined over the area laheled respectively 1 and 2 in psncl B. Note that motifs 1 and 2 are related to each other by a 90" rotation. (C) Contour plots of the inscts of pancl B with, in panel D, the carrcspnding sections of the electron density maps. The mirror planes, shown in panels C and D, distinguish space group QZz3 from space group Q2IR,also compatible with the powder X-ray scattering data. (Reprinted with permissk,n from Delacroix, H., Culik-Krzywicki, T., Mariani, P., and Luzzati, V. [1993]. Freeze-fracture electron microscope study of lipid systems: The cubic phase of space group Pmln. J . Mol. Biol. 229,526-539.)

heterogeneous systems (lipid-protein-water [Gulik-Krzywicki et al., 1984; Mariani et al., 19881; surfactant-oil-water [Barois et al., 1990, Maddaford and Topragcioglu, 19931). Phase Q212has been observed in only one lipidprotein-water system (Mariani et al., 1988). Phases Q229and Q225have been observed in lipid-water systems in which the polar headgroups of the lipids are particularly bulky (CI2EOl2[Mirkin, 19921; gangliosides [Gulik et aL, 19951). Phase Q230,and most likely phase Q22" have been reported of both type I and 11, according to the chemical nature of the lipid (Luzzati et al., 1968c, 1988; Gulik-Krzywicki et al., 1984). All the examples of the other cubic phases reported so far are either of type I (Q225,Q223)or of type I1 (Q212, Q224,Q227). Phases Q227and Q223display some more subtle correlations between chemical composition and phase behavior. In the case of phase Q223,the area-volume ratio of the micelles (Luzzati et aL, 1993) is almost the same in the two micelles. Moreover, since the lipid component is chemically homogeneous in this phase, the conclusion can be drawn that the area per molecule is the same in the two types of micelles. In contrast, and in keeping with the chemical heterogeneity of the lipid moiety, the area-volume ratio takes different values in the two types of micelles of phase Q227(Luzzati et aL, 1993). This observation can be explained by a difference of lipid composition between the two types of micelles, assuming that the area per molecule of the two lipid components is different. These observations have an interesting thermodynamic implication: The chemical potential of the lipid components-which by definition is constant throughout the volume of the phase-is directly related to the area per molecule at the polarapolar interface. Also noteworthy is the fact that in phase Q227the parameter <(Ap)4> is close to the minimum of all values compatible with the data (Luzzati et al., 1992), whereas in the two examples of phase Q223it is far from the minimum (Vargas et al., 1992). In other words (Mariani et aL, 1988), the "entropy" of the map is close to maximal in the case of phase Q227and quite far from it in phase Q223.This property, illustrated in Color Plate I, seems to be related to the chemical type (I or 11) of the structure. The argument runs as follows. In the absence of heavy atoms, the most conspicuous electron density fluctuations are those associated with the low-density CH3end groups. In structures of type I the CH3 end groups are clustered in a small volume surrounding the center of the hydrocarbon region; in structures of type 11, those groups are instead spread out over a far more extended volume. As a consequence, and other things being equal, the local density of CH3groups-and thus the amplitude of the electron density

fluctuations and the value of <(Ap)4>-is likely to be larger in structures of type I than in structures of type I1 (see later). These considerations are also relevant to the position of the fracture and to the aspect of the FFEM images in the phases of type I and 11, as discussed earlier.

IV. STRUCTURE REPRESENTATIONS A. Bicontinuous Phases: Networks o f Rods, Convoluted Sudaces

Two alternative representations, corresponding respectively to Schoen's skeletal graphs and infinite labyrinths (Schoen, 1970), have been proposed to describe the structure of the bicontinuous cubic phases, one in terms of rodlike elements and the other of folded surfaces (reviewed by Mariani et al., 1988). The rod representation was adopted in the early crystallographic study of phase Q230 for reasons that have lost none of their relevance. This cubic phase was observed originally in the anhydrous fatty acid salts of divalent cations, among a variety of 2D and 3D phases in all of which the cations (and, presumably, the polar headgroups) are clustered in quasi-crystalline linear aggregates (Luzzati et al., 1968b): a rod model of the cubic phase fitted nicely into that picture. Moreover, the number of polar headgroups per length of rod (a parameter easy to determine when the dimensions and symmetry of the unit cell and the partial volumes of the polar and apolar moieties are known) was found to take almost the same value in a variety of rod-containing phases, cubic and also of lower symmetry. Finally, the observed and the calculated intensity of the XRS reflections were found to be in excellent agreement with each other (Luzzati et al., 1968~). On the other hand, many examples of phase Q230 are known whose water content is high (Luzzati et al., 1988). The structure in these cases can be represented in terms of convoluted polar-apolar interfaces-phase Q230 is indeed related to one of the paradigms of IPMS, the gyroid G (also called the G-surface) (see Color Plate I). Q224, the other bicontinuous cubic phase, has been observed with a high water content: its structure has been described sometimes in terms of rods, sometimes of convoluted surfaces. The surface, in this phase, is related to another type of IPMS, the D-surface (reviewed by Hyde, 1990). The use for one and the same phase of a structure representation that varies with the degree of hydration is at odds with the gradual effects of water (see, e.g., the system egg lecithin-water, Luzzati et al., 1968a). It would be more satisfactory to produce physical and chemical arguments,

less formal than those invoked by Schoen (1970) and by Charvolin and Sadoc (1988), in support of one or the other of the two representations, or possibly reconciling the two. The problem is even more confusing if the same representation must also take into account the structure of the micellar phases.

B. Micellar Phases

Several facets of the structure of the micellar phases can be considered.

1. Packing of Rigid Spheres: Partial Miscibility of Polar Headgroups and Hydrocarbon Chains Knowing the chemical composition, the symmetry, and the unit cell dimension of the lipid-water phase, it is possible to determine the volume of the apolar micelles and, assuming that they are spherical, their radius (R,,,). A puzzling result of this calculation is that R,,, is always larger than the maximum length of the hydrocarbon chains. This anomaly cannot be explained by the polyhedral shape of the micelles nor by some wrinkling of their surface. In fact the critical assumption underlying the calculation is that the polar and apolar regions of the structure are sharply separated from each other. An obvious way to elude the paradox is to allow some of the polar headgroups to dip to some depth inside the hydrocarbon volume or a few of the headgroups to be embedded in the paraffin core of the micelles (Luzzati et al., 1996). Phases Q223and Q227,which contain two types of micelles, do not lend themselves to this type of analysis. 2. Space-Filling Polyhedra: An Analogy with Foams As depicted in Fig. 2, phases QZ2', QZ2', Q223,and Q227can be visualized in terms of space-filling assemblies of polyhedra: rhombic dodecahedra in Q225,truncated octahedra in Q229,a mixture of distorted dodecahedra and tetradecahedra (in the ratio 3:l) in Q223,and a mixture of distorted dodecahedra and hexadecahedra (in the ratio 1:2) in Q227. Three of these space-filling assemblies (Q225,Q229,and Q223)are sometimes evoked in foams (Wearie, 1994). Foams are similar to the micellar cubic phases in the sense that both systems consist of disjointed cells, containing respectively air or hydrocarbons, separated by septa formed respectively by the hydrated surfactant films or by the hydrated headgroups. In foams, and at the limit of vanishing water content, the geometry of the septa and of their junctions are generally assumed to obey Plateau's conditions: three faces join at each edge with mutual angles equal to 120';

four edges meet at each vertex with mutual angles equal to 109O28'. As the water content increases, these constraints are expected to relax. According to Wearie (1994), the lowest surface energy at the limit of vanishing water content corresponds to the body-centered packing of Kelvin's tetrakaidecahedra. These space-filling polyhedra derive from the regular truncated octahedra by a subtle distribution of curvature on the hexagonal faces that has the effect of bringing the angles at the vertices (which are of 120" and 90" in the truncated octahedron) closer to tetrahedral. At higher water content the lowest surface energy seems to correspond to the clathrate-like space-filling assembly of distorted dodecahedra and tetradecahedra (phase Q223). In this structure, it must be stressed, the faces meet at angles close to 120" and each vertex is a quasi-regular tetrahedral junction of four edges (Davidson, 1973). At still higher water content the lowest surface energy seems to correspond to a system of regular rhombic dodecahedra packed in the face-centered cubic mode, much like the micellar phase Q225. This last structure falls short of fulfilling Plateau's conditions: four edges join tetrahedrally at the vertices of each rhombic dodecahedra1 cell but eight edges join octahedrally at the six other vertices. The structure of the three micellar cubic phases (Gulik et al., 1995) and their sequence Q223 ~ 2 2 9 ~ 2 2 5 as a function of increasing water content (Mirkin, 1992) are consistent with these theoretical expectations. +

+

3. Infinite Periodic Minimal Surfaces The relevance of the IPMS to the bicontinuous cubic phases stems from the fact that these phases can be thought of as lipid bilayers folded in space according to an IPMS: D-surface in phase Q224, G-surface in phase Q230. The two labyrinths are directly congruent in phase Q224 and chirally congruent in Q230. The skeletal graphs of these surfaces coincide with the three- and four-connected systems of rods that are commonly used in the structural description of phases Q230 and Q224 (Fig. 1). The center of the bilayer, whose locus is the minimal surface, and the skeletal graph reside respectively in the polar and in the apolar region if the structure is of type I, and the other way around if the structure is of type I1 (Color Plate I). In contrast, the micellar cubic phases, which subdivide 3D space into an infinite number of disjointed compartments of one polarity embedded in a matrix of the opposite polarity, are by no means bicontinuous. One might thus expect these phases to be utterly unrelated with the IPMS. This is certainly the case for those IPMS whose labyrinths are congruent: the issue is not as clear cut for those IPMS whose labyrinths are not congruent. In particular, two of the IPMS, I-WP and F-RD, display striking correlations with the micellar cubic phases Q22y and Q225. These two surfaces can be visualized by reference to the polyhedra of Fig. 2 and to the pair of noncongruent skeletal graphs.

phase H

phase Glz3O

FIGURE 5 Phases H and Q2": composite representauon of the maps Ap(r). of the chaorrczones (xteen, ~ o l a rblue. : awlar) and of the wlar-wlar interfaces (red). The equidensitv lines Ao(r). c&stan;are k l equaliy spaced, with an-intcrvai of 0.5: negative values are dotted. NW, in all the cases, the deep troughs near the cmtcr of the hydrocarbon regions (see the text) and the fairly flat polar regions; the minimum at the center of the polar region corresponds w the high local concentration of water. The polar-apolar interfaces we assumed to coincide with the equidensity surface Ap -0. (Phase QZ3O)Sections n m a l to the four-fold (plan? r = al8) and to the three-fold axes; these planes contain some of the rods. Note, in the stmcture of type I& thar the electron density troughs fall exactly on the G-surface: in Utis case, therefore, the IPMS and the apolar chaotic zone coincide. The polar chuotiotic zone coincides with the networks of mds. (Phase Hj Sections normal to the six-fold axcs. In HIrhe polar and the apolar chaotic zones coincide respectively with the hexagonal honeycomb and with the six-fold axcs: Ihe opposite is the case in H ,. Note that the polar-apoh intcrfaccs an:far away from thc minimal aurfaccs (phasc Q230) and h m thc hexagonal honeycomb (phase H). The d o n s of the G-surface with the huo planes of the figure were computed and drawn by h.C. Oguey. The bar is 1CQ A long. (Reprinted with permission from Luzzati, V., Vargas, R., Mariani, P., Gulik, A,. wd Delacmix, H. [1993]. Cubic phases of lipid-containing systems: Elements of a theory and biological connotations. 3. Mol. Biol. 229, 540-551.) These representations were developed based on the following data: Lipid

c

T

a

Phase

Type

c(Ap)b

Reference

DTAC DTAC

20

37.2

H

I

2.90

Vargas et al. (1992)

20 60

79.6 62.9

Quo

I

PFL

0.90 0.90 0.65

H

2.91 148

Mmiani (unpublished) Mariani etal. (1990)

MO

0.66

25

143.0

Qm

U U

1.74

Luzzati PI at. (1988)

FIGURE6 [email protected] 11001, [ILO], aad 11111 rhrauph th~origin. N o l h md6cllcr as in Pig. 5. l k trace6 rdtbc lPMS !red linsu) WK m p u a d amaxling lo von SEmcnng snd N e p a (1991). Thsgccnsnss~sardmlhesstionaofihcpdlPr~h.~Phsaeq":UlcIPWShkP-ED snrfac5 theFmvh isblue. Md t t e X B S W ia &en Wi*bt) Phaac CWt We,iPMSZl lheT-WP s d . We

Regarding the I-WP surface (Schoen, 1970), one of the graphs (I-graph) is constructed by joining with straight lines all nearest neighbor points of the body-centered cubic lattice; the lines are eight-connected at both ends. The other graph (WP-graph) is the assembly of all the edges of the spacefilling assembly of regular truncated octahedra (see Fig. 2): the edges are four-connected at both ends. As for the F-RD surface (Schoen, 1970), one of the graphs (F-graph) consists of the straight lines joining all nearest neighbor points of the facecentered cubic lattice; the lines are 12-connected at both ends. The other graph (RD-graph) is the assembly of all the edges of a space-fillingassembly of regular rhombic dodecahedra (Fig. 2); the edges are 4-connected at one end and &connected at the other. The sections [loo], [110], and [111] of the two IPMS, computed according to von Schnering and Nesper (1991), are plotted in Color Plate I1 along with the electron density maps. Most remarkably, in all the cases the relevant IPMS is almost coincident with one particular equi-electron-density surface. In other words, one experimentally defined surface [p(r)= constant] exists whose mean curvature vanishes everywhere. This empirical observation hints at some subtle correlation between the structure of the phases and the IPMS. The structures that correspond to these virtual minimal surfaces (I-WP in Q229,F-RD in Q225)consist of two disjointed 3D labyrinths, one filled by water and the other by the highly hydrated lipid micelles fused together via their hexagonal (in Q229)or quadrilateral (in Q225)faces. The interfacial forces, supposed to be exerted on the virtual IPMS and thus to be minimal, would then be applied at an interface separating a polar labyrinth from a labyrinth containing the hydrocarbon chains and a fraction of the polar moiety of the system (Luzzati, 1995). Presumably, in order that the IPMS representation makes physical sense, the size and shape of the labyrinths must be such that water, paraffins, and headgroups properly fit together. All these conditions are not easy to meet; this may be the reason why the micellar cubic phases Q229and Q225seem to be so rare in lipid-water systems. V. ORDERLY DISPOSAL OF SHORT-RANGE CONFORMATIONAL DISORDER: THE CHAOTIC ZONES

Many of the ideas explored so far hinge upon the very notion of lipid film and implicitly ascribe to the interfacial interactions a predominant role among all the forces at play in lipid-water phases. The question arises of the chemical nature of the structure elements and of the position of the interface (Luzzati, 1995).

The arguments used by the authors who have sought in the IPMS a unified theory underlying lipid polymorphism (reviewed by Hyde, 1990) seem to entertain some confusion between the mathematical notion of surface and the physical concept of film. In the bicontinuous cubic phases the polar-apolar interfaces lie not only at some distance from the IPMS (Charvolin and Sadoc, 1988; Anderson et al., 1988; Hyde, 1990) but in some cases at a distance that is almost as large as is compatible with the space group and the cell dimension. The most striking example is phase Q230 of the anhydrous divalent cation soaps (Luzzati et al., 1968b), in which the polar-apolar interface is very close to the skeletal graph. Similarly, in phases Q227and Q223the polar-apolar interface, the most likely candidate for a film to which a "surface tension" is to be applied, lies at quite a distance from the polyhedral faces (see Fig. 2). Let us get away from the interface and move deeper into the hydrocarbon and the water media. The short-range conformation of the hydrocarbon chains (above the "melting" temperature) has been the matter of some controversy. In the early days, Hartley (reviewed by Hartley, 1977; Luzzati et aL, 1993) produced lucid arguments in support of a highly disordered conformation. After some confusion, the general consensus now seems to be that the chains are highly flexible (reviewed by Charvolin and Hendrikx, 1985) and that the stability of any one structure involves a compromise between the curvature of the polar-apolar interface and the flexibility of the chains (Luzzati and Spegt, 1967; Luzzati, 1968; Charvolin and Sadoc, 1988; Turner and Gruner, 1992; Turner et aL, 1992, and references cited therein). More precisely, the lateral order of chain packing is visualized to be high near the polar-apolar interface-to which the chains are anchored-and to decrease gradually as the distance to the interface increases. The hydrocarbon core of the structure, namely the region most distant from the interfaces, is occupied by segments of chains anchored at different interfaces, with a high concentration of CH3 end groups. The packing is inevitably disordered in those regions. A similar picture applies to the water regions. In the highly hydrated phases the water dipoles are strongly oriented in the immediate vicinity of the polar-apolar interfaces; the orientation becomes increasingly disordered as the distance to the interface increases. In the very center of the water region, the orienting effects of the different interfaces cancel out and the disorder probably becomes as high there as in liquid water. We call the regions where the short-range disorder is maximal chaotic zones. These zones, as we now show, occupy special positions (i.e., symmetry elements, IPMS, skeletal graphs) and coincide with the points, lines, and surfaces that play a conspicuous role in the structure representation. By

definition, on the other hand, the short-range order is maximal at the polar-apolar interface. The apolar chaotic zones are decorated by the CH3 end groups of the hydrocarbon chains and, as discussed in Section 11, coincide with the minima of the electron density maps. By way of illustration, we present a few examples in Color Plate I. In the maps of phase H the minima (and thus the apolar chaotic zones) coincide either with the six-fold axes (structure of type I) or with the hexagonal honeycomb (structure of type 11). In phase Q230,type I, the minima fall exactly on the rods of the two networks. In the case of phase Q230,type 11, the minima of the map fall sharply o n the Gsurface. This remarkable observation, which applies also to the bicontinuous phase Q224(result not shown), indicates that the chaotic zones are the support of the IPMS. As discussed in Section 11, the situation is not quite symmetrical for phases of type I and type 11, since the zones within the water matrix where the short-range disorder is maximal are not decorated by electron density maxima or minima. In the micellar cubic phases, on the other hand, the minima in the maps coincide with the centers of the micelles in phase Q225,Q229,and Q223,with the surface of the polyhedra in phase Q227(results not shown). By analogy with phases QZ3Oand H (see Color Plate I) we thus conclude that the polar and the apolar chaotic zones coincide respectively with the surface and with the center of the polyhedra in phase QZz3,QZz5,and QZz9and with the center and the surface of the polyhedra in phase Q227(see Fig. 2). Over the surface of the polyhedra, moreover, the disorder is likely to be higher along the edges and even higher at the vertices, where two and three quasi-planar chaotic zones meet. Alternatively, should one adopt the IPMS representation of phases Q225and Q229,then the chaotic zones would coincide with the skeletal graphs RD for the F-RD surface (phase Q225)and I and WP for the I-WP surface (phase QZz9).It thus appears that one of the constraints that these structures are bound to fulfill is to orderly distribute the short-range disorder in space. Table I presents a list of the chaotic zones, polar and apolar, for some of the lipid phases whose structures are firmly established. It is worthwhile to end this section with a few comments of general interest: 1. At variance with the widely accepted idea that the IPMS are related to the polar-apolar interfaces, we stress the fact that in the crystallographic sense the positions of the polar-apolar interfaces are in no way remarkable (see Color Plate I; see also Anderson et al., 1988; Turner and Gruner, 1992; Turner et al., 1992).

e

TABLE I Topological Properties and Chaotic Zonesa

Phase Isotropic sol. Isotropic sol. L H H 0230 0230

~ 2 2 1 ~ " 5

Type

Class

1

h/liccllar Micellar Larncllar Rodlike Rodlike Bicontinuous Bicontinuous Bicontinuous Micellar Bicunrinuous Micellar Bicontinuo~rs Micell ar Micellar

I1

I I1 I 11 11 I

~ 2 2 5 ~ 2 2 9

I

~ 2 2 9 ~ 2 2 3 ~ 2 2 7

I 11

Number of polar compartments

-

Number of apolar compartments

--

Number of polar-apolar interfaces

Polar chaotic zone

Apolar chaotic zune

None Center of micellcs Plane Hexagonal honeycomb 6-fold axes G-surface G-graph D-graph Polyhedra surfaces RD-graph Polyhedra surfaces WP-graph Polyhedra surfaces Center of micelles

Center of nliccllcs None Plane 6-fold axes Hexagonal huneycumb G-graph G-surface D-surface Center of micelles F-graph Center of micelles I-graph Center of micelles Polyhedra surfaces

--

"The numbers of polar and apolar compartments and of polar-apolar interfaces refer to a crystallite of infinite dimensions. In the micellar phases the matrix forms a unique compartment, polar in Q'?' (type I) and apolar in QZZ7(type 11); the other medium is split into an infinitc number of compartments. For phases QZ5and Q"' the alternative representation in terms of IPMS is presented in italics. Note that, in the bicontinuous phases with congruent labyrinths (Q230,G-surface: QZ4, D-surface), the number of disjointed polar-apolar interfaces is always two and the number of polar (or apolar) disjointed compartments is two (or one) or one (or two) according to whether the phase is of type I1 (or I). whereas the biwntinuous phases with noncongruent labyrinths (QZ25,F-RD surface: QZZ9,I-WP surface) contain one polar and one apolar compartment and one polar-apolar interface.

2. We introduce the notion of chaotic zones to designate the geometric singularities (points, lines, surfaces) that coincide with the loci of maximal short-range disorder. 3. We note that orderly disposing short-range disorder seems to bring an important contribution to the energy of the phases, much like minimizing the area of the polar-apolar interfaces. 4. Finally, the notion of chaotic zones has the rewarding effect of ascribing formally equivalent roles to the two alternative representations of the bicontinuous cubic phases-rods or IPMS. VI. BIOLOGICAL IMPLICATIONS

Another important aspect of the cubic phases of lipid-containing systems is their possible biological relevance. In a more general way, the biological significance of lipid polymorphism is a problem often evoked in the past (reviewed by Mariani et al., 1988). The point has also been stressed that, on account of their remarkable topological properties, the bicontinuous cubic phases are far better candidates for biological speculations than the lamellar and the hexagonal phases. Besides, if a biological function is to be ascribed to any of the lipid phases, then that phase is expected to be stable in the presence of excess water, as most biological systems are. Of all lipid phases, four have so far been observed in equilibrium with excess water: lamellar for most phospholipids devoid of a net electrical charge, hexagonal (HII)for some phospholipids (reviewed by Seddon, 1990), cubic Q224 for some lipids of biological interest (MO [Mariani et al., 19881; lipid extract from a thermoacidophilic archaebacterium [Gulik et al., 1985]), and cubic Q227 for lipid mixtures of the proper chemical composition (Luzzati et al., 1992). Phase Q224, and its spongy structure, have played a prominent role in biological speculations-for example, the plasma membranes of thermoacidophilic archaebacteria (Luzzati et al., 1987) and the digestion of fats (Mariani et al., 1988). We may also envisage a possible biological significance for phase Q2". Two properties of this phase are relevant to this question. One is chemical composition: phase Q227 contains one (or more) of the most common lipid components of biological membranes (PC, PE), plus one of the usual end products of many a lipid degradation (FA, DAG). The other property is watertightness, associated with the apolar 3D continuum. In contrast, apolar and polar continua are present in the bicontinuous phases. This last property suggests a sort of "patch-the-puncture" process, whereby the local formation of phase Q227 might have the effect of repairing the damage of some

lipolytic agents. Initially, an enzymatic attack would liberate FA and/or DAG, (locally) destabilize the bilayer, and make the membrane leaky; subsequently, FA and/or DAG would mix with the intact lipids and form a patch of watertight phase QZz7that would eventually stop the leak. A similar process, averting the danger of leaks, could come into play in other processes, for example, membrane budding and fusion, which involve a local disruption of the diffusion barrier. This hypothesis is corroborated by the close chemical and physical similarities of phases Q227and QZz4.The same lipid (e.g., MO) is indeed present in these two phases, and the two phases are stable in the presence of excess water. The addition of FA (or DAG) transforms the bicontinuous structure of QZz4,highly permeable to both water and oil, into the water-impermeable Q227.The process could thus be visualized as a switch between two states, one of high and the other of low conductivity, under the control of FA (andlor DAG). In support of these speculations is a study of phospholipase C-induced liposome fusion (Nieva et al., 1995) as well as the frequent observation of lipidic particles in membrane systems undergoing fusion (Verkleij, 1984). Moreover, DAG is known to be an essential component in the signal transmission pathway that cells utilize to recognize and respond to a variety of extracellular signals (Bell and Burns, 1991). This function is indeed preceded by a liberation of DAG that might well induce local and transient alterations of the membrane structure. Finally, the discovery in a wide variety of cells of membrane systems folded in space according to the symmetry of the IPMS (Hyde et al., 1997) is a fascinating generalization of the cubic phases (see also Luzzati et al., 1987). It is also worth mentioning that Delacroix, Nicolas, and GulikKrzywicki have observed that, in the photosomes of some annelids (Harmothoe lunulata), the lattice and symmetry of these "cubic membranes" may well vary with the physiological state of the organelles (in preparation).

Acknowledgments We are grateful to Christophe Oguey for kindly computing the sections of the G-surface of Fig. 5. This work was supported in part by grants from the Association Frangaise pour la Recherche Mtdicale, the Ligue Nationale contre le Cancer, and the Association Frangaise contre les Myopathies.

List of Abbreviations ID, etc. DAG DTAC FA FFEM FRAP

Average value of a function A(r) over the unit cell One-dimensional, etc. Diacyl glycerol Dodecyltrimethylammonium chloride Fatty acid Freeze-fracture electron microscopy Fluorescence recovery after bleaching

Infinite periodic minimal surface without intersection 1-D lamellar, 2-D hexagonal, 3-D cubic phase of space group NOn (International Tables, 1952) MO Monoolein NMR Nuclear magnetic resonance PC Phosphatidylcholine PE Phosphatidylethanolamine PL Phospholipid AP(~) Normalized, dimensionless expression of the electron density p(r): Ap(r) = [p(r) -

]l[ -

2]"2 Types I and Structures with the hydrocarbon chains inside the structure elements (oil-inI1 water) and vice versa (water-in-oil) XRS X-ray scattering IPMS L, H, Q"

References

Anderson, D. M., Gruner, S. M., and Leibler, S. (1988). Geometrical aspects of the frustrations of the cubic phases of lyotropic liquid crystals. Proc. Natl. Acad. Sci. U.S.A.85,5364-5368. Balmbra, R. R., Clunie, J. S., and Goodman, J. F. (1969). Cubic mesomorphic phases. Nature (London) 222,1159-1160. Barois, P., Eidam, D., and Hyde S. T. (1990). X-ray study of cubic phases in ternary systems of surfactant DDAB, water and oil. J. Phys. (France) Colloque 51(C7), 25-34. Bell, R. M., and Burns, D. J. (1991). Lipid activation of protein kinase C*. J. Biol. Chem. 266,4661-4664. Caffrey, M. (1987). Kinetics and mechanism of transitions involving the lamellar, cubic, inverted hexagonal and fluid isotropic phases of hydrated monoacylglycerides, monitored by time-resolved X-ray diffraction. Biochemistry 26, 6349-6363. Charvolin, J., and Hendrikx, Y. (1985). Amphiphilic molecules in lyotropic liquid crystals and micellar phases. In "Nuclear Magnetic Resonance of Liquid Crystals" (J. W. Emsley, ed.), pp. 449-471. D Reidel Publishing, Norwell, MA. Charvolin, J., and Rigny, P. (1973). Proton relaxation study of paraffin chain motions in a lyotropic liquid crystal. J. Chem. Phys. 58, 3999-4008. Charvolin, J., and Sadoc, J. F. (1988). Periodic systems of frustrated films and "micellar" cubic structures in liquid crystals. J. Phys. (France) 49, 521-526. Cierc, M. (1996). A new symmetry for the packing of amphiphilic direct micelles. J. Phys. I1 (France) 6,961-968. Cribier, S., Gulik, A., Fellmann, P., Vargas, R., Devaux, P. F., and Luzzati, V. (1993). The cubic phases of lipid-containing systems. A translational diffusion study by fluorescence recovery after photobleaching. J. Mol. Biol. 229,517-525. Davidson, D. W. (1973). Clathrate hydrates. In "Water: A Comprehensive Treatise" (F. Franks, ed.), Vol. 2, pp. 115-234, Plenum Press, New York. Delacroix, H., Mariani, P., and Gulik-Krzywicki,T. (1990). Image analysis of freeze-fractured lipid-water cubic phases of space group Ia3d. J. Phys. (France) Colloque 51(C7), 119-129. Delacroix, H., Gulik-Krzywicki, T., Mariani, P., and Luzzati, V. (1993a). Freeze-fracture electron microscope study of lipid systems: The cubic phase of space group Pm3n. J. Mol. Biol. 229,526-539. Delacroix, H., Gulik-Krzywicki, T., Mariani, P., and Risler, J. L. (1993b). Freeze-fracture electron microscopy of lyotropic lipid systems: Quantitative analysis of cubic phases of space group Ia3d (QZM).Liquid Crystals 15, 605-625. Delacroix, H., Gulik-Krzywicki, T., and Seddon, J. (1996). Freeze-fracture electron microscopy of lyotropic lipid systems: Quantitative analysis of the inverse micellar cubic phase of J. Mol. Biol. 25S,88-103. space group Fd3m (QZZ7).

Eriksson, P. O., Lindblom, G., and Arvidson, G. (1987). NMR studies of micellar aggregates in 1-acyl-sn-glycero phosphocholine systems. The formation of a cubic liquid crystalline phase. J. Phys. Chem. 91, 846-853. Fontell, K., Fox, K. K., and Hansson, E. (1985). On the structure of the cubic phase I1 in some lipid-water systems. Mol. Cryst. Liq. Cryst. Lett. 1,9-17. Frank, J., Goldfarb, W., Eisenberg, D., and Baker, T. S. (1978). Reconstruction of glutamine synthetase using computer averaging. Ultramicroscopy 3,283-290. Gulik, A,, Luzzati, V., DeRosa, M., and Gambacorta, A. (1985). Structure and polymorphism of bipolar isopranyl ether lipids from archaebacteria. J. Mol. Biol. 182, 131-149. Gulik, A., Delacroix, H., Kirschner, G., and Luzzati, V. (1995). Polymorphism of gangliosidewater systems: A new class of micellar cubic phases. Freeze-fracture electron microscopy and X-ray scattering studies. J. Phys. II (France) 5, 445-464. Gulik-Krzywicki, T., and Costello, M. J. (1978). The use of low temperature X-ray diffraction to evaluate freezing methods used in freeze-fracture electron microscopy. J. Micros. 112, 103-113. Gulik-Krzywicki, T., and Delacroix, H. (1994). Combined use of freeze-fracture electron microscopy and X-ray diffraction for the structure determination of three-dimensionally ordered specimens. Biol. Cell 80, 193-201. Gulik-Krzywicki, T., Aggerbeck, L. A,, and Larsson, K. (1984). The use of freeze-fracture and freeze-etching electron microscopy for phase analysis and structure determination of lipid systems. In "Surfactants in Solution" (K. L. Mittal and B. Lindman, eds.), pp. 237-257. Plenum Press, New York. Hartley, G. S. (1977). Micelles-retrospect and prospect. In "Micellization, Solubilization and Microemulsions" (K. L. Mittal, ed.), pp. 23-43. Plenum Press, New York. Hendrykx, Y., Sotta, P., Seddon, J. M., Dutheillet, Y., and Bartles, E. A. (1994). NMR measurements in inverse micellar cubic phases. Liquid Cryst. 16, 893-903. Hyde, S. T. (1990). Curvature and the global structure of interfaces in surfactant-water systems. J. Phys. (France) Colloque 51(C7), 209-228. Hyde, S., Andersson, S., Larsson, K., Blum, Z., Landh, T., Lidin, S., and Ninham, B. D. (1997). "The language of shape. The role of curvature in condensed matter: physics, chemistry and biology." Elsevier, Amsterdam. "International Tables for X-ray Crystallography." (1952). Kynoch Press, Birmingham, United Kingdom. Kekicheff, P., and Cabane, B. (1987). Between cylinders and bilayers: Structures of intermediate mesophases of the SDSIwater system. J. Phys. (France) 48, 1571-1583. Lindblom, G., and Oradd, G. (1994). NMR studies of translational diffusion in lyotropic liquid crystals and lipid membranes. Prog. NMR Spectros. 26, 483-515. Longley, W., and McIntosh, M. J. (1983). A bicontinuous tetrahedral structure in a liquidcrystalline lipid. Nature (London) 303, 612-614. Luzzati, V. (1968). X-ray diffraction studies of lipid-water systems. In "Biological Membranes" (D. Chapman, ed.), Vol. 1, pp. 71-123. Academic Press, London. Luzzati, V. (1995). Polymorphism of lipid-water systems: Epitaxial relationships, area-pervolume ratios, polar-apolar partition. J. Phys. II (France) 5, 1649-1669. Luzzati, V., and Reiss-Husson, F. (1966). Structure of the cubic phases of lipid-water systems. Nature (London) 210,1351-1352. Luzzati, V., and Spegt, P. A. (1967). Polymorphism of lipids. Nature (London) 215,701-704. Luzzati, V., Mustacchi, H., Skoulios, A. E., and Husson, F. (1960). La structure des colloi'des d'association. I. Les phases liquide-cristallines des systtmes amphiphile-eau. Acta Crystallog. 13, 660-667. - -

Luzzati, V., Gulik-Krzywicki, T., and Tardieu, A. (1968a).Polymorphism of lecithins. Nature (London) 218,1031-1034. Luzzati, V., Tardieu, A,, and Gulik-Krzywicki, T. (1968b). Polymorphism of lipids. Nature (London) 217,1028-1030. Luzzati, V., Tardieu, A., Gulik-Krzywicki, T., Rivas, E., and Reiss-Husson,F. (196%). Structure of the cubic phases of lipid-water systems. Nature (London) 220, 485-488. Luzzati, V., Gulik, A,, DeRosa, M., and Gambacorta, A. (1987). Lipids from Sulfolobus solfataricus, life at high temperature and the structure of membranes. Chem. Scripta 27B, 211-219. Luzzati, V., Mariani, P., and Delacroix, H. (1988). X-ray crystallography at macromolecular resolution: a solution of the phase problem. Makromol. Chem. Macsomol. Symp. 15,l-17. Luzzati, V., Vargas, R., Gulik, A,, Mariani, P., Seddon, J. M., and Rivas, E. (1992). Lipid polymorphism: A correction. The structure of the cubic phase of extinction symbol Fdconsists of two types of disjointed reverse micelles embedded in a 3D hydrocarbon matrix. Biochemistry 31, 279-285. Luzzati, V., Vargas, R., Mariani, P., Gulik, A., and Delacroix, H. (1993). Cubic phases of lipid-containing systems: Elements of a theory and biological connotations. 3. Mol. Biol. 229,540-551. Luzzati, V., Delacroix, H., and Gulik, A. (1996).The micellar cubic phases of lipid-containing systems: Analogies with foams, relations with the infinite periodic minimal surfaces, sharpness of the polarlapolar partition. 3. Phys. II (France) 6, 405-418. Maddaford, P. J., and Topragcioglu, C. (1993).Structure of cubic phases in the ternary system didodecyldimethylammonium bromide/water/hydrocarbon. Langmuir 3,2868-2878 Mariani, P., Luzzati, V., and Delacroix, H. (1988). Cubic phases of lipid-containing systems: Structure analysis and biological implications. 3. Mol. Biol. 204, 165-189. Mariani, P., Rivas, E., Luzzati, V., and Delacroix, H. (1990). Polymorphism of a lipid extract from Pseudomonas fiuorescens: Structure analysis of a hexagonal phase and of a novel cubic phase of extinction symbol Fd-. Biochemistry 29, 6799-6810. Mirkin, R. J. (1992). Ph.D. thesis, University of Southampton, United Kingdom. Nieva, J. L., Alonso, A., Basiiiez, A., Gulik, A., Vargas, R., and Luzzati, V. (1995).Topological properties of two cubic phases of a phospholipid:cholesterol:diacylglycerol aqueous system and their possible implications in the phospholipase C-induced liposome fusion. FEBS Lett. 368, 143-147. Saxton, W . O., and Frank, J. (1977).Motif detection in quantum noise-limited electron micrographs by cross-correlation. Ultramicroscopy 2, 219-227. Schoen, A. H. (1970). "Infinite Periodic Minimal Surfaces without Self-intersections." NASA Technical Note D-5541. National Aeronautics and Space Administration, Washington, DC. Schwarz, H. A. (1880). "Gesammelte Mathematische Abhaldungen," Vol. 1. Springer, Berlin. Scriven, L. E. (1976). Equilibrium bicontinuous structure. Nature (London) 263, 123-125. Seddon, J . M. (1990).Structure of the inverted hexagonal (Hrr) phase, and non-lamellar phase transitions of lipids. Biochim. Biophys. Acta 1031, 1-69. Tardieu, A. (1972). ~ t u d ecristallographique de systkmes lipide-eau. Ph.D. thesis, Universit6 Paris-Sud. Tardieu, A., and Luzzati, V. (1970). Polymorphism of lipids. A novel cubic phase, a cagelike network of rods with enclosed spherical micelles. Biochim. Biophys. Acta 219,ll-17. Turner, D. C., and Gruner, S. M. (1992). X-ray diffraction reconstruction of the inverted hexagonal ( H I I )phase in lipid-water systems. Biochemistry 31, 1340-1355. Turner, D. C., Gruner, S. M., and Huang, J. S. (1992). Distribution of decane within the unit cell of the inverted hexagonal ( H I I )phase of lipid-water-decane systems determined by neutron diffraction. Biochemistry 31, 1356-1363.

Vargas, R., Mariani, P., Gulik, A., and Luzzati, V. (1992). The cubic phases of lipid-containing systems. The structure of phase QZz3(space group Pm3n): An X-ray scattering study. J. Mol. Biol. 225, 137-145. Verkleij, A. J. (1984). Lipidic intramembranous particles. Biochim. Biophys. Acta 779,43-63. von Schnering, H. G., and Nesper, R. (1991). Nodal surfaces of Fourier series: Fundamental invariants of structured matter. Z. Phys. B Condens. Matter 83, 407-412. Wearie, D. (1994). Structure transformations in foams? Philos. Mag. Lett. 69,99-105.

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Bicontinuous cubic lipid-water particles and cubosomal dispersions KAre Larsson Camurus Lipid Research Foundation, Ideon Science Park S-223 70 Lund, Sweden 1. INTRODUCTION

Our understanding of liquid-crystalline phases of lipids is based on the fundamental work by Luzzati and coworkers on aqueous systems of fatty acid soaps, revealing the combination of long-range order with short-range (liquidlike) disorder, cf. [I]. Another simple lipid, glycerol monooleate (GMO), has been of particular significance in studies of cubic lipid-water phases. Their bicontinuous character was settled by NMR diffusion measurements [2], and the cubic phase coexisting with excess of water was shown to have space group Pn3m, consistent with the diamond type of minimal surface structure; C, [3]. Then, studies over the whole composition range of the GMO-water system showed that there are two epitaxially related cubic phases forming a coherent cubic region of the phase diagram, and the phase at lower water content with space group Ia3d was proposed to have the gyroid type of minimal surface structure; CG [4]. At the coexistence composition, the lattice of the two phases in equilibrium were consistent with the requirement of a Bonnet relation, which was taken as a strong indication of the minimal surface relevance of the minimal surface description [5].An isometric Bonnet transformation leaves the intrinsic geometry intact and preserves the zero mean curvature. It is now generally accepted that the center of the lipid bilayer of bicontinuous cubic lipid-water phases is curved as a minimal surface, separating the two congruent water pore systems on each side. A cubic phase (space group Im3m) corresponding to the third simple minimal surface structure C, (Schwarz' primitive surface) was first observed by Landh [6] when an amphiphilic polymer was introduced, the triblock polymer Pluronic F-127 (or Poloxamer 407) PE09,PP0,,PE0,,. The ternary system of GMO/water/Pluronic has then been used frequently in order to disperse cubic phases, as shown below.

Fig.1. The three cubic minimal surface conformations of the bilayer calculated according to the nodal surface approximation 171 (using Mathematica 2.2 for Macintosh). The three cubic phases C,, C, and C , - no other bicontinuous structures - have been observed in aqueous lipid systems. The reason is probably that these minimal surfaces are the simplest and most homogeneous ones, with the smallest variation of Gaussian curvature over the surface. The nodal surface approximation of the minimal surfaces introduced by von Schnering and Nesper [71 has been most useful in order to describe the minimal surface structures. The three surfaces corresponding to lipid-water phases are shown in Fig. I. This description might be relevant when the dynamic properties of the bilayer is taken into consideration. Lipid bilayers in lamellar liquid-crystalline phases may exhibit undulation motion. Similar motion in a cubic lipid-water phase must form standing waves. The surface description in Fig. 1 could therefore be regarded as the nodal surface of standing wave oscillation of the bilayer. Such a model of a breathing mode of bilayer motion has been calculated [8], and is shown in Fig 2. The bilayer around the flat points should be expected to have the highest freedom for transverse oscillations, and a corresponding breathing mode is shown.

2. DISPERSING PROCESS In connection with structural studies, the possibility to disperse these cubic phases was examined. In order to terminate the three-dimensional bilaycr structure and close it into particles, it seemed likely that fragmentation by fusion with liposomal bilayers might work. The three-phase region of the ternary system soy bean phosphatidylcholine (PC)/GMO/water consisting of the CDphase, the lamellar liquid-crystalline phase (La), and water, was therefore homogenized by ultrasonification. The resulting dispersions were quite stable

even at very low levels of PC, and the particles were isotropic in the polarizing microscope. A closing mechanism as shown in Fig. 3 was therefore proposed, and the particles obtained in this way were termed cubosomes [9].

Fig. 2. The structure unit of the C,-surface with indication of a breathing vibration as transparent regions located at the flat points, after 181.

Fig. 3. Proposed closing mechanism of a fragmented cubic phase by fusion with an Ln-type of bilayer, resulting in a cubic particle, after 191.

As mentioned above, Pluronic is an efficient dispersing agent, and superior to PC in this respect. A simple method is to solve the Pluronic into melted GMO, and add this liquid as droplets into the water phase so that a coarse dispersion is

formed. The particle size is then reduced in a microfluidizer or by a valve homogenizer [lo]. Other preparation methods have also been described [ l 1,121. The size of the particles can be reduced by multiple passage through the homogenizing equipment and be increasing the pressure gradient. The dispersing agent can also influence the particle size distribution. Current work has for example shown that dispersions of GMO/ethylhydroxyethylcellulose (EHEC) form very small particles. Under conditions applied to prepare the dispersions shown below, the cubic GMOIEHEC particles are frequently as small as one unit cell. It can be mentioned that water in these cubic phases can be replaced by nonaqueous polar solvents. The GMOIglycerol system for example shows a similar cubic region. Also such phases can be fragmented into dispersions.

3. STRUCTURE AND PHYSICAL PROPERTIES The textures of cubic particles as observed by cryo-transmission electron microscopy (cryo-TEM) are shown in Figs. 4 and 5. The inner bilayer conformations were identified as the C,- and the CD-structure respectively based on space group determination from X-ray diffraction data. Diffraction curves of the dispersion in relation to the corresponding bulk phase of the CD-structure are shown in Fig. 6 (indexing according to Pn3m gives a unit cell of 96 A). The two curves show good agreement, besides line broadening effects due to the particle size. The periodicity observed by cryo-TEM is consistent with the diffraction data, and the textures of the particles indicate that each one is a single crystal. 13c NMR was also used to compare the dispersed particles with the bulk phase [14]. The mobility of the carbon atoms of the acyl chain and of the glycerol group were quite similar, showing that the dynamic properties of the molecules persist after dispersion of the phase. It is reasonable to assume that there is no opening of the bilayers that exposes the hydrocarbon chain core towards water. As a consequence of this, the cubic particle consists of one inner pore system without contact with the outside water phase and another pore system that is directly connected with the outside continuous phase. This structure is also confirmed by release behavior of solubilized molecules. as described below.

Fig. 4. Cryo-TEM texture of C,particles containing 7.4 % pluronic and 92.6 % GMO, after [lo]. The bar is 1M) nm.

-

Fig. 5. Cryo-TEM texture of one C,-particle containing 4 % pluronic and % % GMO, after 1131.Thebaris l00nm.

Fig. 6. X-ray diffraction curve of C,particles with 4 % polymer 96 % GMO (upper curve) compared to the fully swollen C,-phase of GMO in water, after 1131.

Fig. 7. Non-contact A m image of dispersed C,panicles (like those in Fig. 4) on mica, after 1141. The vertical lines are cleavage planes of mica. The cubic particles of the C,-phase have been visualized by atomic force microscopy (AFM) [15]. An image of these dispersed particles in siru are shown in Fig. 7 . Their size vary between 300 and 750 nm., and this size distribution is in agreement with results from dynamic light scattering. The periodic structure at the surface could not be resolved by AFM. This was also the case with the vesicle-shaped expansions of the bilayer at comers, sometimes seen by cryoTEM (cf. Figs. 4 and 5). The shape of the particles reflects the inner organization. Thc C,-particles are shaped as cubes as seen above, whereas the CD-particles tend to form dodecahedra and the C,-particles usually are more irregular, forming globular aggregates. The aqueous pore system allows solubilization of large molecules. Thus proteins can be incorporated in their native state, with denaturation temperature and enthalpy similar to those in water solution 1161. By the addition of charged lipids to GMO the pore systems can be expanded. About 1 % of distearoyl-PC in relation to the amount of GMO is enough for accommodation of about twice as

much of water in the cubic lattice, compared to the fully swollen GMOIwater phase [17].

4. CUBIC CELL MEMBRANE ASSEMBLIES Cell membranes sometimes associate into organized bodies, such as the endoplasmatic reticulum. Their bilayer conformation may under certain conditions transform into the same types of cubic minimal structures as in the lipid-water systems described above, although the length scale is larger by about one order of magnitude. In the early studies of the cubic bicontinuous structures it was striking that similar electron microscopy conformations were seen in organized membrane assemblies, for example the reported "orthogonal arrays of particles" seen along various plasma membranes, which were proposed to have the C,- structure [9]. Landh later identified over thousand examples of cubic cell membrane conformations [18,19]. The fact that membrane systems in vivo exhibit the same three minimal surface conformations as observed in lipid-water systems demonstrates the biological relevance of these bicontinuous structures. An interesting mechanism behind formation of cubic membrane structures was reported in the mitochondria membranes of amoebae, cf. [20]. Starvation of the amoebae results in formation of cubic membranes, and this transition is reversible. These observed cubic organizations (involving all three structures C,,CG and C,) may reflect a vegetative state. Cubic lipid bilayer particles have been observed under conditions corresponding to gastrointestinal fat digestion and absorption [21]. This may explain the bioavailability of drugs solubilized in such particles at oral administration, as described in the next section.

5. TECHNOLOGY ASPECTS The possibility to solubilize hydrophilic molecules into the aqueous compartments and lipophilic or amphiphilic molecules into the bilayer opens various applications. Drug delivery by cubosomal dispersions has been studied for about one decade and some results will be summarized here. Significant peroral absorption of insulin in cubic particles has been reported [22]. Within an extensive research program on oral administration of peptides run by research groups at Lund University and the drug delivery company Carnurus, mechanisms behind the enhanced gastrointestinal uptake of calcitonin and other

peptides in cubosomal dispersions are studied [23]. One function of calcitonin a 32 amino acid peptide - is inhibition of bone resorption.

-

There are also promising results on parenteral drug administration. Somatostatin is an endogeneous peptide hormone with a half-life of about 55 seconds in the circulation. When this tetradecapeptide is injected in the circulation solubilized in cubic GMOIpluronic particles, it exhibits two decay processes. First a halflife of 1.5 minutes is seen, followed by a half-life of about 2.6 hours [24]. These two decay periods may reflect the two pore systems of the particles. The release from the pore system which is continuous with the outside aqueous phase should be expected to be much faster then the release from the other pore system, which is closed in relation to the outside water. Particles with biocompatible properties in the circulation, such as lipid-water particles, have a particular advantage in cancer therapy. Due to a mechanism called enhanced permeability retention, an incorporated drug will accumulate in solid tumors and reduce systemic toxicity. Among possible applications in food technology, controlled release of flavors and encapsulation of enzymes seem promising. A number of enzymes have been incorporated in the cubic GMOIwater phase for biosensor applications [25]. Cubic particles could in a similar way form mini-sensors in order to monitor bio-reactions. Future development may even lead to artificial organelles via step-wise fusion of different types of cubic aggregates, each with a particular contentlfunction. The solubility of GMO (about 1 0 . ~ M) might be a limitation in some applications. Polymerization of the bilayer within the cubic particle, however, has been achieved [26], which opens new possibilities.

6. OTHER MESOPOROUS LIPID-WATER PARTICLES It is also possible to form kinetically stable dispersions of the well-known inverse hexagonal phase (H,,), consisting of an hexagonal arrangement of water cylinders covered by lipid monolayers, and the L3-phase, which can be regarded as a "melted" cubic phase. Whereas the cubic particles are crystals (in crystallographic sense), the HI,- and L3-particles are liquid-crystalline and liquid respectively. The structures of the pore systems of L3-particles should be expected to be quite similar to those of cubic particles described above. A difference, however, is the relation between the pores and the continuous phase. The pores are thus assumed to open and close dynamically in relation to the

outside water phase, due to the disorder of this liquid structure compared to the cubic structure. Lipid-water H,,-phases can sometimes be dispersed by polymers such as Pluronics. Such a phase formed by mixtures of GMO and retinyl palmitate in the weight proportions 8 4 1 6 have been dispersed [13]. The hexagonal unit cell axis is 60 A, and the cryo-TEM texture of a characteristic particle from a dispersion containing Pluronicilipids in the weight ratio 5:95 is shown in Fig. 8. A hexagon-shaped face is formed in one direction, where the rod axes are roughly perpcndicular to the surface. A remarkable feature is that thc rods are not extended hut bent into circular or elliptic shapes, so that the rods opens at both ends at hexagon-shaped faces, as seen in Fig. 8. Many particles are therefore shaped roughly like half-spheres (or prolate half-ellipsoids) with a planar face formed by the cross-section. There are two alternative pore structures; open or closed in relation to the outside. In the second alternative, bilayer units form caps over the rod openings. In both alternatives, all hydrophobic regions, not to be exposed on the surface of the panicles, are covered by monolayers with their polar groups oriented towards outside water.

Fig. 8. A particles of the inverse hexagonal phase, after 1131. The water rods are oriented perpendicular to this hexagon-shaped face. The bar is 100 nm.

7. MATHEMATICAL MODELS OF CUBIC PARTICLES

A powerful method for modeling of periodic structures with finite periodicity has been developed by Sten Andersson, cf. 1271. As an example, the calculation of a particle with C,-structure, which contains 216 structure units, is shown in Fig. 9.

Fig. 9. Calculation of a C, type of cubic particle according to Andersson's approach, cf. 1271.

ACKNOWLEDGEMENT This review is based on fruitful collaboration over many years, involving Mats Almgren, Sten Andersson, Piero Baglione, Sven Engstrom, Jonas Gustavsson, Stephen Hyde, Tomas Landh, Marcus Larsson, Sven Iddin, Helena LjusbergWahren, Maura Monduzzi, Bany Ninham and Fredrik Tiberg.

REFERENCES 1. 2. 3. 4. 5.

V. Luzzati, H. Mustacchi, A. Skoulios and F. Husson. Acta Cryst. 13 (1960) 660. G.Lindblom, K. Larsson, L. Johansson, K. Fontell and S. F o n h . J. Am. Chem. Soc. lOl(1979) 5465. W. Longley, T.J. McIntosh. Nature 303 (1983) 612. K Larsson. Nature 304 (1983) 664. S.T. Hyde, S. Andersson, B. Ericsson and K. Lanson. Z. Kristallogr. 168 (1984) 213.

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

T. Landh. J. Phys. Chem. 98 (1994) 8453. H.G. von Schnering and R. Nesper. Z. Phys. B - Condensed Matter 83 (1991) 407. M. Jacob, K. Larsson and S. Andersson. Z. Kristallogr. 212 (1997) 5. K. Larsson. J. Phys. Chem. 93 (1989) 7304. J. Gustavsson, H. Ljusberg-Wahren, K. Larsson, M. Almgren. Langmuir 12 (1996) 461 1. 11. P.T. Spicer, K.L. Hayden, M. L. Lynch, A. Ofori-Boateng and J.L. Burns. Langmuir 17 (2001) 5748. M. Nakano, A. Sugita, H. Matsuoka and T. Handa. Langmuir 17 (2001) 3917. J. Gustavsson, H. Ljusberg-Wahren, K. Larsson, M. Almgren. Langmuir 13 (1997) 6964. M. Monduzzi, H. Ljusberg-Wahren and K. Larsson. Langmuir 16 (2000) 7355. C. Neto, G. Aloisi, P. Baglione and K. Larsson. J. Phys. Chem. B103 (1999) 3896. B. Ericsson, K. Larsson, K. Fontell. Biochim. Biophys. Acta 729 (1983) 23. J. Engblom, Y. Miezis, T. Nylander, V. Razumas and K. Larsson. Progr. Colloid Polym. Sci. 116 (2000) 9. T. Landh. Cubic Cell Membrane Architectures. Thesis, University of Lund, Food Technology, PO Box 124, S- 221 00 Lund, Sweden (1996). S.T. Hyde, S. Andersson, K. Larsson, Z. Blum, T. Landh, S. Lidin and B. Ninham. The Language of Shape. Elsevier, Amsterdam (1997). Y. Deng, M. Marko, K.F. Buttle, A. Leith, M. Mieczkowski and C. Mannella. J. Struct. Bio1.127 (1999) 23 1. J. Borne'. Lipid Self-Assembly and Lipase Action. Thesis, Lund Univesity, Physical Chemistry, PO Box 124, S-221 00 Lund, Sweden (2002). H. Chung, J. Kim, J.Y. Um, I.C. Kwon and S.Y. Jeong. Diabetologia 45 (2002) 448. F. Tiberg. Personal communication S. Engstrom, B. Ericsson and T. Landh. Proceed. Intern. Symp. Control. Rel. Bioact. Mater. 23 (1996) 89. V. Razumas, J. Kanapieniene, T. Nylander, S. Engstrom and K. Larsson. Anal. Chem. Acta 289(1994)155 D. Yang, D.F. O'Brien and S.R. Marder. J. Am. Chem. Soc. 124 (2002) 13388. S. Andersson, M. Larsson, K. Larsson and M. Jacob. Biomathematics: Mathematics of Biostructures and Biodynamics, Elsevier, Amsterdam (1999).

This Page Intentionally Left Blank

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

The Discovery of ExxonMobil's M41S Family of Mesoporous Molecular Sieves Charles T. Kresge+, James C. Vartuli*, Wieslaw J. Roth*, and Michael E. Leono~icz'~

Keywords: Mesoporous, MCM-41, Molecular Sieves, M41S, Mobil, ExxonMobil The quest for new molecular sieves in the late 1980's led a team of Mobil researchers to the discovery of a family of nanostructured mesoporous materials known as M41S [ I , 21. MCM-41, the hexagonal phase, is undoubtedly the best known and most widely studied of this family of materials. Other discrete members of the M41S family are the cubic (MCM48) and the lamellar (MCM-50) forms. Each is synthesized via a counterion initiated, selfassembled liquid crystal mechanism involving oxide precursors, which form an inorganic equivalent to a liquid crystal-micelle structure [3,4]. This manuscript describes the events that led to the discovery of M41S materials. It also summarizes the supporting characterization and mechanistic studies that led to a picture of how these of materials are actually formed. The mechanistic and characterization studies involved many researchers from ExxonMobil's Paulsboro and Princeton laboratories. Relevant publications by these ExxonMobil scientists include references [3-191. Beyond the initial discoveries of MCM-41, -48, and -50, these scientists contributed significantly to the fundamental understanding and refinement of this new class of materials. ExxonMobil has very recently scaled-up the synthesis and commercialized MCM-41 for an undisclosed application. This decade-long journey from the discovery to commercial application is similar in duration to that of many novel materials. Yet, there were many unique challenges posed by the synthesis and development of such a novel class of materials. Like most discoveries of novel materials, the discovery of the M41S family of mesoporous molecular sieves was an unanticipated outcome of the application of observational skills, prior knowledge, and novel synthesis approaches. Paralleling the efforts of several major petroleum companies, ExxonMobil had a materials synthesis effort attempting to identify new porous materials that could selectively convert bulky, high molecular weight petroleum molecules into more valuable fuel and lubricant products. In the mid-19801s, an ExxonMobil Research and Engineering predecessor, Mobil Research and Development Corporation, had a significant effort at its Paulsboro, NJ laboratory aimed at discovering and developing layered-type materials and converting them into stable porous catalysts by pillaring. Theoretically, pillared materials offered the ability to tune the pore size, the active site density, and the composition more widely than possible in traditional aluminosilicate zeolites. The pore systems could be tuned for the desired application by varying the pillar size and density. The pillar composition also appeared to be adjustable so that various reactivities and chemistries could be obtained [see for example ref. 201. 'Current addresses: The Dow Chemical Company, 566 Building, Midland Michigan 48674 *Current address: Exxon Mobil Research & Engineering Co., 1545 Route 22 East, Annandale, NJ 08801 (to whom correspondence should be sent) "Current address: Consultant, Charlotte, North Carolina

In the mid- to late 19801s,the authors, working together in what was then Mobil's Paulsboro Laboratory, approached the effort of synthesizing large pore frameworks by attempting to combine both the concepts of the pillared layered materials and the formation of zeolites. We worked together experimenting with novel synthesis approaches and discovering unique materials. One of these novel approaches was to consider that some zeolites were formed via layered intermediates. If these layered intermediates, so-called "zeolite precursors," could be isolated and used to form pillared porous materials, we postulated that the resultant product should have crystalline walls that would be thermally stable, catalytically active, and contain both micro- and mesoporosity. This concept had credibility because of a new framework that was discovered during that time that was found to transpose from a layered precursor to the zeolitic structure [21]. This material was designated as Mobil Composition of Matter number 22 or MCM-22. In the case of MCM-22, it was noted earlier in exploring its physical properties that upon thermal treatment of the as-synthesized product, some of the peaks in the X-ray diffraction pattern of the material had shifted to higher 2 0 values. We initially interpreted this as indicative of unit cell contraction, similar to that of swollen layered materials when the intercalate is removed. In layered materials the low angle lines associated with the interlayer distance shift to lower d-spacings consistent with the removal of the organic template intercalate and the collapse of the layers. However, in the as-synthesized MCM22 sample, this shift of the d-spacings, upon thermal treatment, was subtle, - 2 - 3 ~ ,while the peaks associated with the intra-layer framework remained relatively unaffected. This suggested that the MCM-22 zeolite was composed of crystalline layers that were linked together by weak chemical bonds during the synthesis. Upon thermal treatment, these chemical linkages became much stronger, as the layers condensed onto each other. Using the as-synthesized layered zeolite material, with the template intact and prior to any thermal exposure, we attempted to delaminate or separate these crystalline layers of the MCM-22 "precursor". A pillared layered material resulting from this delamination and subsequent pillaring was obtained and identified later as MCM-36 [22,23]. The process involved intercalation of the layers using an alkyltrimethylammonium compound followed by the insertion of stable inorganic oxide pillars using a reactive silica source such as tetraethylorthosilicate. This effort was expanded to other suspected layered zeolite precursor candidates and other areas involving surfactant-oxides chemistry (or systems). The layered zeolite precursors, exemplified by MCM-22, differed from other layered predecessors such as clays and layered silicates in possessing layers with high zeolite activity and porosity. Obviously, these were very attractive from a catalytic standpoint. A complicating factor distinguishing the layered zeolite precursors from the other layered materials, was their resistance to swelling by ion exchange with neutral or mildly basic media, such as quaternary ammonium salts or amines. These swelling agents afforded only partial exfoliation. We found that the introduction of a quaternary ammonium surfactant in a hydroxide form, was more effective in swelling of the layered zeolite precursors 1241.

This general approach of interrupting zeolite syntheses, isolating the layered zeolite precursors, and using these potential crystalline layered materials as reagents to form large pore active catalysts was investigatedfor other zeolite families such as ZSM-35, or synthetic ferrierite. To optimize the formation of these layered precursors, several reaction conditions were identified and imposed on the traditional zeolite synthesis. The zeolite synthesis was interrupted prior to any X-ray diffraction evidence of crystallinity. The interruption could be initiated at any point within -25 to 75% of the total expected synthesis time. High concentrations of the intercalate, an alkyltrimethylammonium salt, usually in the hydroxide form, at high pH were added to this interrupted zeolite precursor media. In other syntheses, a reactive silica source, tetramethylammonium silicate, was also added as a potential pillaring agent. These new synthesis mixtures were then subjected to additional hydrothermal treatment, usually at low temperatures -lOO°C, in an attempt to form the zeolite-layered hybrid. Many of the products exhibited some very unusual properties that later on were recognized as those of the mesoporous materials. The X-ray diffraction pattern of these interrupted zeolite preparations were essentially featureless except for one broad low angle peak at about 2" 2 0 . This X-ray diffraction pattern was intriguing since the original zeolite templating agent still existed in the synthesis composition. The other unusual properties of this unknown material were the extremely high BET surface areas and hydrocarbon sorption capacities. These BET surface area values, typically greater than 1000 m2/g, exceeded those normally observed for zeolite samples. The hydrocarbon (n-hexane and cyclohexane) sorption capacities were in excess of 50 weight percent, also abnormally high compared to our typical microporous samples. In fact, these sorption values were so remarkable that our analytical laboratories initially believed that their test equipment was broken or out of calibration. In a parallel and concurrent synthesis effort within our team, the cetyltrimetylammonium hydroxide reagent, which was developed for high efficiency swelling, was also used directly as a structure-directing-agent in zeolite-like hydrothermal syntheses. The properties of the products were similar to those generated in the layered zeolite precursor systems, i.e., characterized by a low angle line in an X-ray diffraction pattern corresponding to large d-spacing and unusually high BET surface area and adsorption capacities. Thus, both interrupted zeolite precursor systems and direct introduction of cetyltrimethylammonium hydroxide as a structure-directing-agent resulted in the new mesoporous molecular sieve products. As described below, subsequent detailed characterization studies allowed elucidation of the nature of these remarkable materials. Obviously, the aforementioned unusual physical properties are characteristic of M41S mesoporous molecular sieves. However, with a featureless X-ray diffraction pattern, except for one low angle line at a d-spacing of -40A, it was impossible to discern the nature of these materials. One early theory was that we had synthesized some kind of layered silicate precursor with crystalline domain sizes below X-ray detectability.

A key to the identificationof this new class of porous materials was the observation, by transmission electron microscope (TEM) analyses, of a trace amount of MCM-41 in one of our samples. The observation of trace quantities of MCM-41 as discerned by the uniform hexagonal pore structure in one of the interrupted synthesis preparations, provided us with hard evidence of this new class of materials (see Figure 1). Figure 1: The Initial TEM observation of MCM-41 [ref. I ] After the initial observation of MCM-41, we focused our synthesis effort on identifying the synthesis conditions required to produce this unique molecular sieve. In a relatively short time, we were able to produce enough excellent quality samples of MCM-41 for detailed analyses. After confirming reproducibility and many of the analyses, we filed our initial patent memorandum describing our observations. We concluded that we had discovered a new class of mesoporous molecular sieves, a class that would be useful for many petroleum processes. One of our first hypotheses, based on both the hexagonal ordering of the pores and variation of pore sizes, as seen in TEM analyses, and the XRD pattern, was that we had discovered one of the crystalline phases predicted by Smith and Dytrych, known as the 81(n) family of frameworks [25]. The theoretical XRD pattern of this family almost matched that of some of our best samples of MCM-41. However, it was not until later when we obtained the 2 9 ~NMR i data that we determined that our material was not like a typical crystalline framework. That the XRD patterns could be generated by the order of the pores and not by crystalline walls, was a unique feature of this new class of mesoporous materials. We presented our story to the research staffs of both the Paulsboro and Princeton Laboratories. Very rapidly, many individuals from both laboratories were involved in investigating this new family of materials. Their efforts ranged from synthesis efforts of new regions such as varying the surfactant chain length and solubilization, characterization of the products using sorption and NMR techniques, and catalytic testing. In all cases, we were analyzing a new class of materials that presented unique data. For example, the pore size distribution was remarkable; the narrow pore size appeared to be like that of microporous materials but within the mesopore range. As mentioned previously, the hydrocarbon sorption capacity was unique. Benzene sorption isotherms clearly indicated pore condensation inflections at benzene partial pressures indicative of mesopore size channels. These inflections were typically not observed with microporous materials due to the low partial pressures needed. By June of 1990, Jeff Beck, a key collaborator and a member of the Princeton Laboratory, was able to synthesize various pore size materials using both different alkyl chain lengths of the cationic surfactant as well as taking advantage of micellular swelling [26,27]. In recognition of his contributions to further advancing MCM-41, Jeff became a co-author of the seminal Nature article [3]. Another member of the Princeton Laboratory, Kirk Schmitt i that the walls of these materials were amorphous. was able to demonstrate by 2 9 ~NMR Since surfactants were used in the syntheses, other researchers investigated the

connection between our molecular sieves and micelles and liquid crystal chemistry. This knowledge base and the subsequent discoveries of the other unique structures, MCM-48 and MCM-50, helped to establish the basis for the mechanism of formation of these materials. In retrospect, the synthesis conditions that we were using in our aluminosilicate systems to obtain layered zeolite hybrids and/or larger pore materials, namely the high pH, high surfactant concentration, and a reactive silica source, were the very synthesis conditions conducive to the formation of the mesoporous molecular sieves. The discovery and identification of other members of this new class of porous materials, MCM-48 and MCM50, came during the middle half of 1990 as a result of a detailed study relating the effect of surfactant concentration on the silica reagent (Figure 2) 1281. The discovery of these additional two members of the mesoporous molecular sieve family was another key factor in supporting the proposed mechanism of formation. Figure 2: The M41S Family of Materials Including MCM-41, MCM-48, and MCM-50, The discovery of this new class of materials, mesoporous molecular sieves, posed several challenges for our understanding of the formation of porous materials. Understanding the mechanism of formation of these materials was a challenging task, which led to many long debates within the Paulsboro and Princeton research community. We initially approached the concept of formation of these materials like traditional zeolite chemists. One of our first proposals was that the materials might have been formed by some sort of templating structure or pore filling agent. This meant, in the case of the mesoporous molecular sieves, that the templating agent was an aggregation of molecules and not the discrete molecules that normally template microporous structures (see Figure 3). Based on our initially limited knowledge of liquid crystal structures and micelles, we concluded that the liquid crystal structure existed prior to the formation of the molecular sieve. In the case of the MCM-41, this would be the hexagonal liquid crystal phase. Figure 3: The role of quaternary directing agents [ref. 41 However, this simple and appealing mechanistic pathway was not universally accepted within our research community. Alternatively, it was proposed that the silicate reagent also affected the formation of these materials. It was this second proposed route that appeared operative in most systems as more data were obtained. A significant set of data that helped establish the preferred mechanistic route was a group of experiments that investigated the effect of various levels of silica at the same surfactant (SUR) concentration. By changing the SUR/Si molar ratio we were able to synthesize MCM-41, -48 and -50 while keeping the other synthesis conditions the same [28]. These conditions would then exclude the possibility of any preformed liquid crystalline phase prior to the formation of the silicon phase, since the same surfactant concentration was used for all experiments. Only the amount of silica added to each solution was changed. These data supported the concept that the anion, in this case, the silicate species, significantly affected the formation of the resultant template of the mesoporous molecular sieves (see

Figure 4). These data were some of the evidence that led to the proposed mechanisms of formation published in the initial MCM-41 journal articles [3,4]. Figure 4: The proposed mechanism of formation pathways [ref. 41 In retrospect both proposed pathways proved to be valid. The predominant pathway operating in most situations appears to be the anionic species initiated one (using cationic surfactants). This concept was explained and expanded upon by many researchers, specifically by the group at the University of Santa Barbara headed by Galen Stucky [29,30]. A Michigan State University group headed by Thomas Pinnavaia, expanded this mechanistic pathway concept further to include neutral directing agents such as polymers [31,32]. Later, researchers at the University of South Hampton demonstrated that the other proposed pathway, originally labeled the liquid crystal phase initiated pathway, can also function [33,34]. George Attard and his co-researchers used a preformed liquid crystal phase to synthesize both a silica and a metal (platinum alloy) mesoporous molecular sieve. The M41S mesoporous molecular sieves exhibit characteristics that are different than those generally attributed to typical zeolites. They contain little or no Bronsted acidity. They also contain amorphous walls, which are generally around 10A when synthesized using the initial cationic surfactants. These thin, amorphous walls limit both the thermal and hydrothermal stability under severe conditions (relative to crystalline structures). However, it was shown that increasing synthesis temperature and/or duration led to improved hydrothermal stability and quite robust silica MCM-41 structures were demonstrated [35]. Although the wall composition contains random bond angles and atomic location, there exists a uniform density of silanol groups (other than the silanols that exist due to incomplete condensation) within the channels due to surfactant packing requirements. These silanol groups provide unique anchoring sites for the functionalization of species within a mesopore channel. These functionalized products provide opportunities for designing unusual catalytic/sorptive materials for various applications and present a class of materials significantly different than other uniform porous materials [36-381. Although we published two initial papers, Exxon Mobil 's primary goal was to obtain broad applied IP coverage for the novel class of materials we discovered. We were allowed to write additional manuscripts for journals only after most of these patents were issued, i.e., after 1994. Throughout the early go's, many aspects of the synthesis, composition of matter, modification, and applications were covered in United States Patents (see Tables 1-3). Broad claims were allowed covering any inorganic, nonlayered material having a single X-ray diffraction peak below 5" 20, and a benzene sorption uptake of at least 15 grams per 100 grams of material at 25°C. Included are claims for both the assynthesized (organic containing) and the calcined (organic removed) forms. The functionalization claims are very broad including the functionalization of the material at various stages of the synthesis and with a wide variety of organic/inorganic compositions such as metal salts and complexes with attached or accompanying groups including

alkyls, alkoxides, amines, phosphines, sulfides, sulfonates, nitrates, carbonyls and cyanos compounds [36-381. Patent coverage for method of making is also broadly includes various organics that exhibit amphiphilic character [39]. Application patents covering a wide range of typical catalytic processes were also obtained. For those applications that were outside of ExxonMobil's general interest, we collaborated with outside experts. Specifically, early in our evaluation of MCM-41, we obtained patents covering sensors and optical devices with Professor Stucky of UC of Santa Barbara and in the area of separation with Professor D. L. C. Wang of MIT [40-421. The publication of our early results and the recognition of other related materials created an extraordinary interest in the scientific community 143, 441. More than 6000 publications covering all aspects of these materials, including synthesis, characterization, and applications have appeared in the literature since 1992. Separate sessions at international symposia dedicated to mesoporous materials began to appear regularly. Later, entire meetings were dedicated to the subject. A separate society (International Mesoporous Material Society) was formed. The interest in these materials continues to grow as shown in the number of citations in literature surveys (see Figure 5). Figure 5 Number of publications per year In summary, by the combination of prior knowledge, observation skills, and novel synthetic approaches, we discovered a family of mesoporous molecular sieves including discrete structures - MCM-41 (hexagonal), MCM-48 (cubic), and MCM-50 (lamellar). These materials were formed unlike that of our classical microporous structures involving reagent induced-macromolecular templating mechanism. Working with other,Mobil scientists and engineers, we were able to establish a predictive mechanism of formation and identify a broad class of templating reagents. We obtained extensive intellectual property coverage (Table 1-3) including composition of matter, methods of synthesis, templating reagents, functionalization, and a wide range of applications. Within the past several years, ExxonMobil has commercialized MCM-41 for an undisclosed application. Acknowledgments -

We want to thank the technical staff of the former Mobil Paulsboro and Princeton Laboratories. In particular, we want to thank C.D.Chang, R.M. Dessau and H.M. Princen for early discussions on surfactants and liquid crystal chemistry. We want to thank J.S. Beck for his early synthesis work on the effect of pore size using both the variation of the surfactant chain-length and solubilization. We also want to recognize I. D. Johnson for her early synthesis efforts, K. D. Schmitt for his * ' ~ iNMR data and early functionization work, and J.B. Higgins and J.L. Schlenker for their assistance in X-ray diffraction indexing.

References

1. Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; and J. C. Vartuli, "Synthetic Mesoporous Crystalline Material" US Patent 5,098,684, March 24, 1992 (assigned to ExxonMobil) 2. Kresge, C. T., Leonowicz, M. E.; Roth, W. J.; and J. C. Vartuli, "of Synthetic Porous Crystalline Material, Its Synthesis" US Patent 5,102,643, April 7, 1992 (assigned to ExxonMobil) 3. Kresge, C. T., Leonowicz, M. E., Roth, W. J., Vartuli, J. C, and J. S. Beck, "Ordered Mesoporous Molecular Sieves Synthesized by a Liquid-Crystal Template Mechanism," Nature, 1992, 359, 710-712. 4. Beck, J. S. Vartuli, J. C., Roth, W. J., Leonowicz, M. E., Kresge, C. T., Schmitt, K. D., Chu, C. T-W., Olson, D. H., Sheppard, E. W., McCullen, S. B., Higgins, J. B., and J. L. Schlenker, "A New Family of Mesoporous Molecular Sieves Prepared with Liquid Crystal Templates." J. Am. Chem. Soc. 1992, 114, 10834-10843. 5. J.C. Vartuli, K.D. Schmitt, C.T. Kresge, W.J. Roth, M.E. Leonowicz, S.B. McCullen, S.D. Hellring, J.S. Beck, J.L. Schlenker, D.H. Olson, and E. W. Sheppard, "Development of a Formation Mechanism for M41S Materials", Zeolites and Related Microporous Materials: State of the Art1994 (Proceedings of the 10th International Zeolite Conference, GarmischPartenkirchen, Germany, 7117-22/94), J. Weitkamp, H.G. Karge, H. Pfeifer, and W. Holderich, (eds.), Elsevier Science, 53(1994). 6. J.S. Beck, J.C. Vartuli, G.D. Kennedy, C.T. Kresge, W.J. Roth, and S.E. Schramm, "Molecular or Supramolecular Templating: Defining the Role of Surfactant Chemistry in the Formation of Microporous and Mesoporous Molecular Sieves", Chemistry of Materials, 6(1O), 1816-21 (1994). 7. J.C. Vartuli, C.T. Kresge, M.E. Leonowicz, A.S. Chu, S.B. McCullen, I.D. Johnson, and E.W. Sheppard, "Synthesis of Mesoporous Materials: Liquid Crystal Templating versus Intercalation of Layered Silicates", Chemistry of Materials, 6(1 I ) , 2070-77 (1994). 8. J.C. Vartuli, K.D. Schmitt, C.T. Kresge, W.J. Roth, M.E. Leonowicz, S.B. McCullen, S.D. Hellring, J.S. Beck, J.L. Schlenker, D.H. Olson, and E. W. Sheppard, "Effect of Surfactant:Silica Ratios on the Formation of Mesoporous Molecular Sieves: Inorganic Mimicry of Surfactant Liquid Crystal Phases and Mechanistic Implications", Chemistry of Materials, 6(12), 2317-26 (1994). 9. C.T. Kresge, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, J.S. Beck, K.D. Schmitt, C. T-W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, "M41S: A New Family of Mesoporous Molecular Sieves Prepared with Liquid Crystal Templates", Science and Technology in Catalysis 1994, Y. Izumi, H. Arai, M. lwamoto (eds.), Elsevier Science, 11 (1994). 10.J.S. Beck, J.C. Vartuli, G.D. Kennedy, C.T. Kresge, W.J. Roth, and S.E. Schrammn, "Molecular or Supramolecular Templating: Defining the Role of Surfactant Chemistry in

the Formation of Microporous and Mesoporous Molecular Sieves", Catalysis by Microporous Materials, Studies in Surface Science and Catalysis, Vol. 98, H. K. Beyer, H. G. Karge, I. Kiricsi, and J. B. Nagy (eds.), Elsevier Science, 15-16 (1995). 1I.S.B. McCullen, J.C. Vartuli, C.T. Kresge, W.J. Roth, J.S. Beck, K.D. Schmitt, M.E. Leonowicz, J.L. Schlenker, S.S. Shih, and J.D. Lutner, "A New Family of Mesopore Molecular Sieves", Proceedings of the Nanomaterials Symposium at Mich. State Univ. in May, 1995, Access in Nanoporous Materials, T. J. Pinnavaia and M. F. Thorpe (eds.), Plenum Press, 1-10 (1995). 12. C.T. Kresge, S.B. McCullen, J.C. Vartuli, W.J. Roth, J.S. Beck, K.D. Schmitt, M.E. Leonowicz, J.L.Schlenker, J.D. Lutner, "Molecular Sieves as Hosts for Nanophases Design, Preparation, and Properties", Proc. Robert A. Welch Found. Conf. Chem. Res. 1995, 39th (Nanophase Chemistry), 139-48 (1995). 13. J.S. Beck and J.C. Vartuli, "Recent advances in the synthesis, characterization, and applications of mesoporous molecular sieves", Current Opinion in Solid State and Materials Science, I , 76-87 (1996). 14. J.C Vartuli, C.T. Kresge, W.J. Roth, S.B. McCullen, J.S. Beck, K.D. Schmitt, M.E. Leonowicz, J.D. Lutner, and E.W. Sheppard, "Designed Synthesis of Mesopore Molecular Sieve Systems using Surfactant Directing Agents" Div. Pet. Chem. 1995, 40(1), 21-5 and Advanced Catalysis and Nanostructured Materials, W. Moser (ed.), Academic Press, 1-19 (1996). 15.J.S. Beck, S.D. Hellring and J.C. Vartuli, "Micellar Solubilization in the Synthesis of M41S Materials: To Swell or not to Swell, What are the Considerations?", Book of Abstracts, 213th ACS National Meeting, San Francisco, April 13-17 (1997), COLL-311. 16.J.C. Vartuli, W .J. Roth, J.S. Beck, S.B. McCullen, C.T. Kresge, "The Synthesis and Properties of M41S and Related Mesoporous Materials", Molecular Sieves - Science and Technology Volume 1 Synthesis, eds. H. G. Karge and J. Weitkamp, 97-121 (1998). 17 J.C. Vartuli, S.S. Shih, C.T. Kresge, J.S. Beck, "Potential Applications for M41S Type Mesoporous Molecular Sieves" Mesoporous Molecular Sieves 1998, Studies in Surface Science and Catalysis Volume 117, eds. L. Bonneviot, F. Beland, C. Danumah, S. Giasson, S. Kaliaguine, 13-23 (1998). 18 Chitnis, G. K., Vartuli, J. C., Kresge, C. T., and Johnson, I. D., "Latest Trends in M41S type Mesoporous Molecular Sieves and Their Novel Applications", Proceedings of the International Symposium on Frontiers in Catalysis in the 21st Century, India, 1999. 19 Roth, W. J., Vartuli, J. C., and Kresge, C. T. "Characterization of mesoporous molecular sieves: Differences between M41S and pillared layered zeolites", Stud. Surf. Sci. Catal. 129 Nanoporous Materials II. Proceedings of the Conference on Access in Nanoporous Materials, 2000, 501-08.

20. Landis, M. E., Aufdembrink, B. A,, Chu, P., Johnson, I. D., Kirker, G. W., and M. K. Rubin, Preparation of Molecular Sieves from Dense Layered Metal Oxides," J. Am. Chem. Soc, 1991, 113, 31 89 - 3192. 21. Leonowicz, M. E.; Lawton, J. A.; Lawton, S. L.; Rubin, M. K., MCM-22: A Molecular Sieve with Two Independent Multidimensional Channel Systems, Science, 1994, 264, 1910-1915. 22. Kresge, C. T.; Roth, W. J.; Simmons, K. G.; Vartuli, J. C., Crystalline Oxide Material. US Patent 5,229,341, July 20, 1993, Mobil Oil Corporation. 23. Roth, W. J.; Kresge, C. T.; Vartuli, J. C.; Leonowicz, M. E.; Fung, A. S.; McCullen, S. B., MCM-36.The First Pillared Molecular Sieve with Zeolite Properties", Catalysis by Microporous Materials,Studies in Surface Science and Catalysis, Vol. 94, H. K. Beyer, H. G. Karge, I. Kiricsi, and J. B. Nagy (eds.), Elsevier Science. 1995, 94, 301. 24. Roth, W. J.; Vartuli, J. C., The Effect of Stoichiometry and Synthesis Conditions on the Properties of Mesoporous M41S Family Silicates. Stud. Surf. Sci. Catal. 2002, 141, 273. 25. Smith, J. V.; Dydrich, W. J., Nets With Channels of Unlimited Diameter, Nature. 1984, 309, 607-08. 26. Beck, J. S., Method for Synthesizing Mesoporous Crystalline Material. US Patent 5,057,296, October 15, 1991, Mobil Oil Corporation. 27. Beck, J. S.; Vartuli, J. C.; Kennedy, G. J.; Kresge, C. T.; Roth, W. J.; Schramm, S. E., Molecular or Supramolecular Templating: Defining the Role of Surfactant Chemistry in the Formation of Microporous and Mesoporous Molecular Sieves. Chem. Mater. 1994, 6 (lo), 1816-1821. 28. Vartuli, J. C.; Schmitt, K. D; Kresge, C. T.; Roth, W. J.; Leonowicz, M. E.; McCullen, S. B.; Hellring, S. D.; Beck, J. S.; Schlenker, J. L.; Olsen, D. H.; Sheppard, E. W., Effects of SurfactantISilica Molar Ratios on the Formation of Mesoporous Molecular Sieves: Inorganic Mimicry of Surfactant Liquid-Crystal Phases and Mechanistic Implications. Chem. Mater. 1994, 6, 231 7-2326. 29. Stucky, G. D.; Monnier, A,; Schueth, F.; Huo, Q.; Firouzi, D. I.; Janicke, M.; Chmelka, B. F., Molecular and Atomic Arrays in Nano- and Mesoporous Materials Synthesis. Mol. Crysta. Liq. Cryst. 1994, 240, 187-196. 30. Firouzi, A.; Kumar, D.; Bull, L. M.; Besier, T.; Siegar, P.; Huo, Q.; Walker, S. A.; Zasadzinski, J.; Glinka, A. G.; Nicol, J.; Marogloese, D.; Stucky, G. D.; Chmelka, B. F., Cooperative Organization of lnorganic Surfactant and Biomimetic Assemblies. Science 1995, 267, 1138-1 143.

31. Tanev, P. T.; Chibwe, M.; Pinnavaia, T. J., Titanium-Containing Mesoporous Molecular Sieves for Catalytic Oxidation of Aromatic Compounds. Nature, 1994, 368, 321-324. 32. Pinnavaia, T. J.; Zhang, W., Catalytic properties of mesoporous molecular sieves prepared by neutral surfactant assembly, Mesoporous Molecular Sieves 1998, Studies in Surface Science and Catalysis, Vol. 117, L. Bonneviot, F. Beland, C. Danumah, S. Giasson, S. Kaliaguine (eds.), Elsevier Science, 1998, 117, 23-36. 33. Attard, G. S.; Glyde, J. C.; Goltner, C. G., Liquid-Crystalline Phases as Templates for the Synthesis of Mesoporous Silica. Nature. 1995, 378, 366 (1995). 34. Attard, G. S.; Leclerc, S. A. A,; Maniquet, S.; Russell, A. E.; Nandakumer, I.; Gollas, B. R.; Bartlett, P. N., Ordered Mesoporous Silicas Prepared from Both Micellar Solutions and Liquid Crystal Phases. Micro. and Meso. Mater. 2001, 44-45, 159 163. 35. Roth, W. J.; Vartuli, J. C., The effect of stoichiometry and synthesis conditions on the properties of mesoporous M41S family silicates, Zeolites and Mesoporous Materials at The Dawn of 21st Century ,Studies in Surface Science and Catalysis, Vol. 135, A. Galarneau, F. di Renzo, F. Fajula, J. Vedrine (eds.), Elsevier Science, 2001, 135, 134. 36. Beck, J.S.; Calabro, D. C.; McCullen S. B.; Pelrine, B. P.; Schmitt, K. D.; Vartuli, J. C., Method for Functionalizing Mesoporous Crystalline Material, US Patent 5,145,816, September 8, 1992. 37. idem, Catalytic Conversion over Modified Synthetic Mesoporous Crystalline Material, US Patent 5,200,058, April 6, 1993. 38. idem, Sorptionlseparationover Modified Synthetic Mesoporous Crystalline Material, US Patent 5,220,101, June 15, 1993. 39. Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C., Use of Amphiphilic Compounds to Produce Novel Classes of Crystalline Materials, US. Patent 5,250,282, October 5, 1993. 40. Olson, D. H.; Stucky, G. D.; Vartuli, J. C., Sensor Device Containing Mesoporous Crystalline Material, US Patent 5,348,687, Nov. 15, 1994. 41. Beck, J. S.; Kuehl, G. H.; Olsen, D. H.; Schlenker, J. L.; G. D. Stucky, G. D.; Vartuli, J. C., US Patent 5,364,797, Sept. 20, 1994. 42. Herbst, J. A.; Kresge, C. T.; Olson, D. H.; Schmitt, K. D.; Vartuli, J. C.; Wang, D. L. C., US Patent 5,378,440, Jan. 3, 1995.

43. Yanagisawa, T.; Shimizu, T.; Kazuyuki, K, Kato, K., Bull. Chem. Soc. Japan, 63, 1990, 988-992. 44. Inagaki, S.; Fukushima, Y.: Okada, A,; Kurauchi, T.; Kuroda, K.; Kato, C., New silicaalumina with nano-scale pores prepared from Kanemite, Proceedings from the Ninth International Zeolite Conference, R. von Ballmos, J. B. Higgins, M.M.J. Treacy, (eds.), 1994,305.

Table 1 Selected ExxonMobil Patents on the Synthesis of M41S Molecular Sieves U.S. Patent Number

Description

5,057,296

Use of Organic Additives to Vary Pore Size of M41S

5,098,684

MCM-41 Composition of Matter (Hexagonal)

5,102,643

Basic M41S Composition of Matter

5,108,725

Basic Synthesis of MCM-41 (Hexagonal) with CTMA

5,110,572

M41S Synthesis Using Organometallics

5,145,816

Addition of Functional Groups to M41S Materials

5,179,054

ShellICore Catalyst with M41S Shell

5,198,203

MCM-48 Composition (Cubic)

5,246,689

MeAlSiO M41S (Me = Co, Cr ...)

5,250,282

Synthesis of M41S with Amphiphilic Compounds

5,264,203

Synthesis of 2 SAP0 M41S Materials

5,300,277

Improved Synthesis of MCM-48 (Cubic)

5,304,363

Lamellar M41S Materials (MCM-50)

5,308,602

Synthesis of M41s using an Amphiphilic Compound

5,366,945

Heteropoly Acid Catalyst Supported on M41S

Table 2 Selected ExxonMobil Patents on Catalytic Applications of M41S Molecular Sieves

U.S. Patent Number

5,258,114 5,260,501

Description

EB SYNTHESIS WlTH M41S CATALYSTS CONVERSION OF PROPYLENE TO C4-C5 TERTIARY OLEFINS OVER MCM-41 OLlGOMERlZATlON OF OLEFINS IN A MIXED C3-C5 FEED OVER MCM-41 OLEFIN OLlGOMERlZATlON OVER MCM-41 ORGANIC CONVERSION OVER M41S HYDROCRACKING USING M41SREOLITE COMBINED CATALYST DEMETALLIZING HYDROCARBONS OVER M41S MATERIALS ALKYLATION OVER TO PRODUCE HIGH VI POLYALKYLATED NAPHTHALENES OLEFIN DISPROPORTIONATION OVER MCM-41 LEWIS AClD PROMOTED MCM-41 IN PARAFFlNlOLEFlN ALKYLATION ORGANIC CONVERSION OVER MCM-41 ORGANIC CONVERSION BIFUNCTIONAL HYDROPROCESSING CATALYST CONTAINING M41S MATERIALS SELECTIVE LIGHT OLEFIN PRODUCTION FROM NAPHTHA CRACKING WlTH M41S POST-SYNTHESIS ADDITION OF ACTIVATING METALS TO M41S PARAFFIN ISOMERIZATION OVER M41S MATERIALS CRACKING CATALYST INCLUDING M41S MATERIAL OLEFIN OLlGOMERlZATlON OVER Ni-MODIFIED MCM-41 HYDROCRACKlNGlHYDROlSOMERlZATlON TO PRODUCE LUBES AROMATICS SATURATION OVER M41S MATERIALS OLEFIN OLlGOMERlZATlON CATALYST COMPRISING M41S MATERIALS LUBE HYDROCRACKING OVER M41S MATERIALS HYDROCRACKING OVER M41S MATERIALS HYDROCRACKlNGlHYDROlSOMERlZATlON TO PRODUCE LUBES HYDROCRACKING CATALYST COMPRISING M41S WlTH SMALLER PORE SIEVE RESlD UPGRADING OVER M41S MATERIALS PHASE TRANSFER CATALYSIS WlTH AS-SYNTHESIZED MCM-41 USE OF M41S MATERIALS TO PRODUCE LOW AROMATIC DISTILLATES HYDROPROCESSING OF HYDROCRACKER BOTTOMS TO PRODUCE LUBES ORGANIC CONVERSION OVER HETEROPOLY AClD CATALYSTS SUPPORTED ON M41S MATERIALS HYDROGENATION OF PAO'S OVER Pt M41S MATERIALS HYDROGENATION OF LUBES OVER M41S SORPTION OF POLYNUCLEAR AROMATICS WlTH M41S MATERIALS METAL-CONTAINING M41S COMPOSITIONS AS HYDROPROCESSING CATALYSTS

Table 3 Selected ExxonMobil Patents on Other Applications of M41S Molecular Sieves U.S. Patent Number

Description

5,143,707

USE OF M41S IN NOx REDUCTION

5,348,687

M41S MATERIALS HAVING NONLINEAR OPTICAL PROPERTIES

5,364,797

SENSOR DEVICE CONTAINING M41S

5,378,440

SEPARATION OVER M41S MATERIAL

Figure 1 The initial TEM observation of MCM-41

I

MCM-41 (Hexagonal)

I

MCM-48 (Cubic)

lil

I

Degrees 2 8

Degrees 2 Q

I

I - D pore system

I

MCMdO (Stabilized Lamellar)

I

3-D pore system

I

Degrees 2 Q

I

Silica

I

Dimensionality Unknown

Figure 2 The M41S family of materials including MCM-41, MCM-48, and MCM-50

0

Isolated, short alkyl chain length quaternary ions direct the formation of microporous molecular

Self-assembling, long alkyl chain length quaternary ions direct the formation of mesoporous molecular sieves

Figure 3 The role of quaternary directing agents

Surfactant

Micelle

liquid crystal phase initiated Micellar Rod

wwwww*w w

• w

alcination

+

. w

=ww.*

I

I

Hexagonal Array

silicate counterion initiated

t

I

MCMQ~

Silicate

Mechanistic pathways for the formation of MCM-41 @ liquid crystal phase initiated @ silicate anion initiated Figure 4 The proposed mechanism of formation pathways

4

N

A" /

-

, ' /

/

-

-

-

-

-

-

C itationsofthe-paper Kresge et al, Nature, 1992 , '

-

Figure 5 Number of publications per year key word 'Mesoporous Materials' - Source: scifinderB

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Discovery of mesoporous silica from layered silicates Kazuyuki KURODA Department of Applied Chemistry and Kagami Memorial Laboratory for Materials Science and Technology, Waseda University, Ohkubo-3, Shinjuku-ku, Tokyo 169-8555, Japan, and CREST, Japan Science and Technology Agency, Japan

1. INTRODUCTION Porous materials with mesopores (diameter of 2 - 50 nm) have attracted great interests in the community of science and technology. The homogeneous pore size, unique and ordered pore arrangements, the presence of inner surface reactive groups (ex. silanol groups for mesoporous silica), and high surface area are quite attractive for a wide variety of applications including catalysts, catalyst supports, adsorbents, reaction vessels, matrices for photochemical species, optical and electronic devices, and biomedical applications including drug delivery. Materials with low dielectric constants andlor low refractive indices are also extensively studied using porous materials. Mesoporous materials can provide novel nanospaces that can not be supplied by microporous crystals like zeolites. Among mesoporous materials reported so far mesoporous silica is the most widely investigated. Silica is environmentally benign and biologically inactive, and the resources are virtually not limited. Therefore, the discovery of mesoporous silica has stimulated many new studies in diverse fields. A large number of microporous aluminosilicates (zeolites) have been synthesized since 1950s because they can be used as adsorbents, catalysts, builders for detergents, host materials for nanoscale fabrication, and so on. Porous crystals with novel structures and various compositions besides aluminosilicates have also been synthesized. The enlargement of pore size was a

keen requirement not only from scientific interests but also industrial necessities; therefore, preparation of porous crystals with larger pores was the most important target among a large number of studies on porous crystals. The important thing note here is that the enlargement of pore size was not realized by a simple extension of conventional zeolite syntheses. Zeolites are normally synthesized by hydrothermal reactions, and the basic concept of the synthesis is crystallization of "soluble" species in the presence of inorganic andlor organic structure directing agents [I]. The major breakthrough to create ordered mesoporous materials comes from the use of "molecular assembly" of organic substances. In addition, "nanosheets" derived from single layered polysilicate were used in our group. Crystalline layered silicates are desirable as a starting materials for constructing molecularly ordered silica walls of mesoporous silica. This type of materials can not be synthesized from soluble silicate species. It has been known that pillaring robust inorganic oligomers in the interlayer spaces of layered materials including layered polysilicates is useful for the preparation of porous materials with various compositions and has been investigated for a very large number combinations of host layered materials and guest pillaring agents [2]. However, both the pore sizes and the arrangements are not well controlled. Thermal stabilities of pillared clays are not so high. The procedure used for the preparation of mesoporous silica, described below, is completely dissimilar to the pillaring technique. The main process is based on the interactions of flexible layered polysilicates with assembled organic substances. Layered polysilicates, such as kenyaite, magadiite, layered octosilicate, and kanemite, are unique because the frameworks are composed of only tetrahedral Si04 units, and both hydrated alkali metal ions and silanol (SiOH) groups are present in the interlayer region [3]. Therefore, these layered polysilicates are quite attractive because the interlayer surfaces can be organically modified by ion exchange, adsorption, and direct derivatization such as silylation [4]. Intercalation compounds of layered polysilicates with several organic compounds have been known and the intercalation reactivities of layered polysilicates have been summarized by Lagaly [ 5 ] . Kenyaite, magadiite, layered octosilicate (ilerite), kanemite, and related disilicates are layered sodium (or potassium) polysilicates and their acid-treated layered polysilicic acids are also known. Cation exchange of interlayer cations in the polysilicates with organic cations can produce organoammonium-exchanged polysilicates. Acid-treated layered polysilicic acids can also form organic intercalation compounds with amines and other polar molecules, depending on the reactivities of layered

polysilicic acids. Before our discovery only lamellar intercalation compounds had been known. The conspicuous point of these materials, being different from clay minerals, is the presence of silanol groups in the interlayers. Interlayer silanol groups can be modified by grafting (ex. silylation) to form covalent bonds, which leads the formation of silicate-organic hybrid materials. During the course of our study on intercalation chemistry of various inorganic layered materials we happened to find that a single layered polysilicate kanemite was allowed to react with alkyltrimethylammonium ions to form silicate-organic complexes having a three-dimensional silica network [6]. They can be calcined to give mesoporous silica with narrow pore size distributions because occluded assemblies of alkyltrimethylammonium ions acted as templates. The formation of three dimensional silica networks can easily be monitored by 2 9 ~MAS i NMR [6]. The serendipitous finding of the conversion to three dimensionality triggered the preparation of novel ordered porous silica materials. The characteristics of mesoporous materials derived fiom kanemite have been described in this decade since the first successful report. Those results have elucidated that mesoporous materials of M4 1s series [7], which will be discussed in separate chapters, and the kanemite-derived materials are different in various aspects.

2. KANEMITE: A SINGLE LAYERED POLYSILICATE Kanemite was discovered as a natural mineral in Lake Chad by Johan et al. in 1972 [8] and the chemical synthesis and reactivity were reported by Beneke and Lagaly [9]. Kanemite (NaHSi205.3H20)is routinely prepared by dispersing 6Na2Si205 into water. Because a single crystal of kanemite is not large enough for X-ray crystal structure analysis, the detailed crystal structure had not been determined until Gies et al. solved the structure [lo]. Before this crystal structure determination, the silicate structure had been discussed on the basis of other related disilicates. The structure is basically similar to those of KHSi20j-I1 and H2Si205-111.The silicate structures of KHSi205-11and H2Si205-I11 were reported and they are thought to be composed of six-membered rings of Si04 tetrahedra and they are linked two-dimensionally [ll]. The acid treatment of both kanemite and KHSi205-I1 produces the same material of H2Si205-111. Kanemite can also be recovered by treating H2Si205-I11(prepared by the acid treatment of KHSi20j-11) with NaOH. Accordingly, the structure of kanemite is reasonably related to this model and composed of single layered silicate sheets of linked Si04 tetrahedra with hydrated Na ions in the interlayers. Fig. 1 shows a schematic model of kanemite, as determined by Gies et al. The single silicate

Fig. I. Schematic model of kanernitc. (Single layered silicate network and interlayer hydrated sodium cations are presented.)

sheets are wrinkled regularly and ion-exchangeable Na ions are present in the interlayer space. The solid state Z 9 ~MAS i NMR spectrum of kanemite exhibits only one peak due to Q' species ((Si0)SiONa and (Si0)3SiOH)) at -97 ppm, indicating that the structure is composed of only Q~ units of Si04 tetrahedra. 3. FORMATION O F KANEMITE-ALKYLTRIMETHYLAMMONIUM MESOSTRUCTURED MATERlALS AND THE CONVERSION TO MESOPOROUS SILICA -DISCOVERY O F MESOPOROUS SILICA~

~

Interactions of layered polysilicates with organic substances have not been investigated so extensively. The interactions of kanemite with organoammonium ions were firstly investigated by Beneke and Lagaly who reported the formation of intercalation compounds of kanemite with several organoammonium ions [a]. From the linear tendency of the increase in the basal spacings of the products with the chain length of alkyltrimethyl- and dialkyldimethyl-ammonium ions, they concluded the formation of lamellar intercalation compounds.

Fig. 2. Powder X-ray diffraction patterns in lower 20 region of kanemite (a) and dodecyltrimethylammonium-kanmitccomplcxcs reacted for 2 weeks (h), I h (c), 3 h (d), and 1 d (e). (Copyright: The Chemical Society ofJapan) However, thermally treated samples of those products told us a different story. We investigated the reaction of kanemite with several alkyltrimethylammonium ions under the same conditions as reported previously 161. When kanemite was allowed to react with dodecyltrimethylammonium ions, the powder XRD analysis o f the product showed that the peaks due to kanemite almost disappeared and that the intense peaks at around d-spacings of 3.1 nm and 3.7 nm appeared. (Fig. 2) The peak of 3.1 nm agreed well with that due to the lamellar phase reported previously [9]. The peak decreased with the reaction time and also disappeared by washing with acetone, supporting that the 3.lnm phase is lamellar. However, the Z 9 ~NMR i result of the reaction product clearly indicated the presence of a Q~ ((SiO)&) signal and we concluded that interlayer condensation between adjacent silicate layers occurred (Fig. 3), though the resolution of the spectra was low from the viewpoint of current performance of NMR apparatus. The reactions with other alkyltrimethylammonium ions (carbon number: 14, 16, and 18) also produced similar findings with increased dspacings according to the chain length. After the thermal treatment of those silicate-organic complexes at 700 OC, the XRD patterns were similar to those ofthe as-synthesized materials. This fact

Fig. 3. 2 9 ~ iMAS - ~ NMR ~ spectra of dodecyltrimethylamm~niumkanemite complexes. (Reaction time: a) Ih, b) 3h, and c) Id. ) (Copyright: The Chemical Society of Japan)

is not generally accepted if lamellar structures were retained. Under such a high temperature organoammonium cations can not survive and the layer structures should collapse. Therefore, it is quite reasonable to consider that the three dimensional silica networks were formed during the interactions with the organic substances and the structures were retained after the calcination. This finding opened a new pathway to produce novel porous materials. The specific surface areas determined by BET method were determined to be about 900 mZlg and, more importantly, the pore size was narrow and varied with the chain length of the organoammonium ions used. (Fig. 4) The pore size does not correspond to the size of each organoammonium ion; therefore, it is quite reasonable to regard the assemblies of the ammonium ions as templates for pore formation. Later we call this new Dorous materials as KSW-I. Additionally, the silicate-organic complexes were trimethylsilylated with chlorotrimethylsilane in order to modify the internal pore structure [12]. This method is used for the modification of silanol groups of layered polysilicates 1131. The 2 9 ~MAS i NMR spectrum of kanemite-hexadecyltrimethylammonium complex shows the split peaks due to the presence of both Q3 and Q4 units (Fig. 5b). The spectrum of the silylated derivative (Fig. 5c) shows the remarkable decrease in the intensity of the peak due to Q~unit, indicating the substantial

dlnm

Fig. 4. Pore size distributions in the calcined products obtained from a) dodecyl-, b) tetradecyl-, c ) hexadccyl-, and d) octadec~l-himethylammonium-kanemite complexes. (Copyright: The Chcmical Society of Japan)

Fig. 5. "~i-MAS NMR spectra of a) kanemite, h) hexadecyltrimethylammonium kanemite complex and c) the trimethylsilylatcd derivative. (Copyright: The Chemical Society of Japan)

Fig. 6. Average pore diameters d of -: calcined products from alkyltrimethylammoniumkanemite complexes and -: calcined products from trimethylsilylated derivatives. (n = the number of carbon atoms in alkyl chains). (Copyright: The Chemical Society of Japan)

conversion to Q4 units by silylation of Si sites with Q3 environments. The XRD patterns of the silylated products did not change so seriously and the specific surface areas and the pore sizes of the calcined products were tuned, indicating the controllability of the pore size by silylation. (Fig. 6.) Inagaki et al. extended this conclusion to more efficient production of highly ordered silica-based mesoporous materials [14- 161. They considered that the formation reaction of silicate-organic complexes used for the preparation of ordered mesoporous silica depended on the degree of ion exchange, expansion of interlayer spacing, and condensation between the layers. By choosing the optimum conditions, they obtained highly ordered mesoporous materials (called FSM) with homogeneous mesopores. In those reactions, kanemite was allowed to react with hexadecyltrimethylammonium ions at 70 OC for 3 h at relatively high pH values, and then the pH values of the suspension were adjusted at 8.5 by acid-treatment to precede condensation. The XRD patterns of the as-synthesized and calcined products were characterized as 2d-hexagonal, and the specific surface area of the calcined product reached to ca. 1100 m21g and the pore size was ca. 2.8 nm, which is similar to the feature of MCM-41. After the reaction with hexadecyltrimethylammonium ions, the reaction mixture was separated by centrifugation to remove dissolved silica, thus well resolved XRD pattern of the product was obtained [16]. The detailed discussion on this topic will be

separately described in another chapter. Chen and Davis reported an extensive study on the difference between MCM-4 1 and mesoporous materials derived from kanemite [I 71. The mesoporous materials derived from kanemite depend on the synthetic conditions used. In some cases, the thermal and hydrothermal stabilities of the materials are higher than those of MCM-41. They have claimed that, in the case of kanemitederived mesoporous materials, the layered structure of kanemite is locally rearranged by the reaction with alkyltrimethylammonium ions and that a hexagonal phase formed by condensation of the fragmented layered silicates after their wrapping rodlike micelles, although the conditions employed by Inagaki et al. to form a hexagonal phase are too drastic to retain local environments of the layered nature. This point was corrected afterwards by our in-situ XRD, indicating that even in the high pH conditions employed by Inagaki et al. the formation process was different from that of MCM-41, which will be mentioned later. The reason why kanemite derived materials can show higher thermal and hydrothermal stabilities is explained by higher condensed framework of SiOz derived from the layered silicate structure [17]. This difference can be related to the synthetic procedures. In the preparation of MCM-41, the proposed reaction mechanism suggests that silica source should be depolymerized to monosilicic and/or oligo- and poly-silicic acids before organization with surfactants. Inaki et al. also compared both of the materials from the point of photomethathesis of propene in mesopores and reported that the activity of FSM-16 is higher than that of MCM-41 because of the unique arrangements of surface Si-OH groups originated from kanemite [18]. In the procedure of the preparation of mesoporous materials including MCM-41, other phases such as cubic and lamellar phases have been reported depending on the reaction conditions such as the surfactantlsi ratio [7]. However, in all the cases of kanemite used in the preparation of mesostructured materials, cubic phases have not been observed, which should be noted as a remarkable difference between MCM series and kanemite-derived mesoporous silicas. On the other hand, lamellar phases are frequently observed in the case of kanemite. Inagaki et al. reported the formation of lamellar phases by the reaction of kanemite with dialkyldimethylammonium ions or alkyltrimethylammonium ions with alkylalcohols [19]. In these cases, the surfactants form a lamellar structure, thus inducing the formation of lamellar phases of the silica-surfactant systems. It should be noted that even in these lamellar phases there are some silicon atoms assignable to Q~ environments, indicating these lamellar phases

can not be interpreted as simple intercalation compounds [19]. Chen and Davis have suggested that the reaction conditions are critical for the differentiation between mesoporous materials derived from kanemite and MCM-4 1 [17]. They changed the reaction conditions and reported interesting results. They raised the reaction temperature at 80 OC and also changed the amounts of aqueous surfactant solutions to kanemite. In the initial stages of the reactions, they noticed that intercalation compounds formed as reported by Yanagisawa et al. In this early stage, initial condensation of Q~ sites occurred and it was supposed that lamellar phase different from simple intercalation compounds would be formed in this stage. After the long reaction times, three dimensional silicate networks formed and they produced silica-based mesoporous materials with high surface area. As described above, depending on the reaction conditions, the characteristics of the products derived from kanemte are different. The variation in the reaction conditions such as pH, surfactantlsi ratio, Na content in kanemite, and so on causes the variation in the structures of the silica-surfactant products. One of the important factors controlling the preparation of the silicateorganic complexes is pH value in the reaction media. The original reports by Lagaly et al. and by Yanagisawa et al. adopted relatively lower pH values around 8.5-9 where the reaction of kanemite with surfactants proceeded for 2 weeks with changing aqueous surfactant solutions once a week. The XRD patterns of the formed silicate-organic complexes showed a broad peak at around 3.5-4 nm, indicating the formation of disordered mesostructured materials. The reactivities of other layered polysilicates with alkyltrimethylammonium ions should be noted. Depending on the compositions and structures, there are several kinds of layered polysilicates. Kenyaite, magadiite, and octosilicate have thicker silicate layers, which are easily monitored by XRD and 2 9 ~ i -Kenyaite ~ ~ ~and . magadiite are layered polysilicates which show typical intercalation chemistry [13,20]. Octosilicate (ilerite) also forms intercalation compounds with organic substances, retaining the parent silicate sheet structure [21]. The 2 9 ~MAS i NMR spectra of these polysilicates after the reactions do not change, which demonstrates this account clearly. 4 . FORMATION MESOPHASES

PROCESSES

OF

SILICA-SURFACTANT

The process from kanemite to silicate-organic mesostructured materials is one of the key issues to understand the nature of the porous materials derived from

kanemite. In order to clarify the formation mechanisms of mesostructured materials from kanemite, it is necessary to follow the reaction processes in situ. O'Brien et al. have applied this technique to the study on the formation processes of silica-surfactant mesophases [22]. The reaction mixtures of kanemite and hexadecyltrimethylammonium (CI6TMA) ions were allowed to react at the same conditions described by Inagaki et al. [14]. The synthesis of the silica-organic mesophase yielding MCM-41 was also followed by the same conditions described previously [7]. In the energy dispersive X-ray powder diffraction spectra of kanemiteC16TMAsystem, the most striking feature of the phases formed by the reaction of kanemite with C16TMAions is the initial formation of a lamellar phase in the dispersion state. At the very initial stage of the reaction, the lamellar phase with d-spacing of ca. 3.0 nm appeared. The d-spacing increased gradually and became constant at ca. 3.3 nm. Examination on this phase separated as a solid sample indicated that the phase was undoubtedly a lamellar phase because the interlayer spacing expanded after swelling in organic solvents although the structure of the lamellar phase has somewhat condensed silica network because of the presence of a large amount of Q~ units. The other phase having d-spacing of ca. 3.6 nm became apparent with time although the phase can not be assigned definitely. As the reaction time increases, the phase with d-spacing of ca. 4.1 nm appeared. Since the other spectra showed the presence of the (110) and (200) peaks in addition to the main (100) peak at 4.1 nm, this peak is certainly assignable to the hexagonal phase. The phase increased with the time and became constant. It is questionable to describe that all of the silicates dissolves before forming a hexagonal phase in this very short reaction time under the reaction conditions used by Inagaki et al. [14,16]. These findings indicate that the formation of the hexagonal mesophase from kanemite with HDTMA ions is different from that of the mesophase of MCM-41. In the formation process of the mesophase related to MCM-4 1, the pattern clearly indicated that, in the very initial stage, there were no crystalline phases and the hexagonal phase appeared in the early stage and the intensity increased and became constant. This in-situ data did not show any sign of a lamellar phase as an intermediate. It has been described several times that fragmented silicate sheets wrap rodlike micelles to form a 2-d hexagonal structure in the formation of FSM-16. Recently we have reported the presence of fragmented silicate sheets in the formation process of mesostructured materials for FSM-16, and the finding is consistent with the report by O'Brien et a1 [22]. On the other hand, structurally

different mesoporous silica called KSW-2 [23] is also formed from kanemite. An intercalation compound composed of kanemite-derived silicate sheets and C16TMA ions can be transformed into silica-based mesostructured materials containing squared mesopores by acid treatment that induces the structural deformation. This reaction uses surfactant assemblies but the porous structures are not governed by the packing, which is only observed for the unique kanemite system. The powder XRD patterns of mesostructured precursors for KSW-2 exhibit several peaks at higher diffraction angles, suggesting the partial retention of the silicate framework of kanemite. Based on these findings, it is reasonable to say that mesoporous silicas derived from layered disilicate are structurally superior to those using soluble silicate species in thermal and hydrothennal stabilities as well as catalytic activity, and that the formation mechanisms between them are different. 5. LAMELLAR MESOPHASES DERIVED FROM KANEMITE Lamellar mesophases synthesized by the reaction of kanemite with surfactants have not been well investigated, though the presence of lamellar mesophases and the transformation have been inferred by in situ XRD study, as described above. On the basis of the in-situ results lamellar mesophases were further surveyed [24]. Detailed studies on the formation of species in lamellar mesophases should progress the understanding on previously reported interlayer condensation of adjacent silicate sheets of acid-treated H2Si205-I11[25]. Lamellar organoammonium silicates with variable silicon environments can be synthesized by the reaction of kanemite with an aqueous solution of hexadecyltrimethylammonium ( C 1 6 T W )chloride. Si04 units with both and environments were present in the silicate frameworks of the lamellar mesophases. The Q~ species mainly formed by "intralayer" condensation of the species in the individual silicate sheets in kanemite rather than by "interlayer" condensation between the adjacent sheets. The intralayer condensation can be suppressed by lowering the reaction temperature and/or shortening the reaction time, which results in the relative retention of the silicate framework of kanemite in the lamellar mesophase. This finding leads to the formation of new mesoporous silica KS W-2 described in the next section. The C16TMA/Simolar ratio is an important factor to direct the formation of mesophases. When the C l 6 T M / S imolar ratio is 0.2, all the XRD peaks in low scattering angles are assignable to a hexagonal phase with a lattice constant of 4.7 nm (dloo= 4.1 nm), which is consistent with the synthesis of a precursor of FSM-16. When the C16TMA/Si ratio increased, the XRD patterns of the

e4

e4 e3

e3

products changed. The peak at 3.2 nm and higher order diffractions were observed. The layered nature of the products was further proved by the increase in the d-spacings to 4.6 nm by intercalation of n-decylalcohol. In addition, the structures collapsed upon calcination at 700 OC for 6 h. All these findings indicate that the obtained materials are lamellar mesophases. In the present case, silicate species were detected in the 2 9 ~MAS i NMR spectrum of the lamellar mesophase, meaning that the lamellar mesophase is not simply composed of alternating C16TMAions and silicate sheets of kanemite. The XRD patterns of the lamellar CI6TMA-silicate mesophases (CI6TMA/Si= 2.0) prepared at room temperature (L-RT), 50 OC (L-50), 70 OC (L-70), and 90 "C (L-90) are surveyed. The peak at the d-spacing larger than 3 nm and the peaks of higher order diffractions were observed for each product. The 2 9 ~MAS i NMR spectrum of L-RT (C16TMA/Si= 2.0) showed that several peaks were mainly observed and that a small peak was detected. This peak did not increase remarkably in the spectrum of the product even after i NMR profile is related to the stirring for 50 days. The change of the 2 9 ~MAS variation in both the interaction of the silicate sheets with CI6TMAions and the bonding angle among tetrahedral Si04 units in the silicate sheets owing to the + the head group of C16TMAions. The difference in ionic radii between ~ a and TEM image of L-RT showed clear striped patterns with a repeated distance of ca. 3.0 nm. These results indicate that single silicate sheets in kanemite were almost retained by the synthesis at room temperature in the present system. The peak intensity due to silicate species increased and the peaks were broadened with the increase in the reaction temperature. Depending on the synthesis temperatures, silicate frameworks in the lamellar CI6TMA-silicate mesophases derived from kanemite involve Si04 species with environments in addition to silicate species. The structural change of the silicate frameworks has never been found for organoammonium intercalation compounds of other layered silicates. Accordingly, the formation of silicate species is related to the unique structure of kanemite. The the intralayer condensation is plausible because the individual silicate sheets of kanemite are composed of Si04 tetrahedra and wrinkled regularly, and the adjacent SiOH groups are alternatively confronted each other. The flexibility of the silicate framework in kanemite becomes lower by intralayer condensation and the lamellar mesophases occur more easily. In the present system, lamellar mesophase silicates were prepared by using quite a high C,,TMA/Si ratio (2.0) even at relatively high pH (ca. 11). On the basis of the insight on "intralayer condensation" described here, the structural variety of silicate-organic mesostructures derived from kanemite is restricted to the

e4

e3

e4

e3

e4

e4

e4

e4

e3

variation in the silicon environments. This finding widens the reaction system of kanemite with surfactants, which contribute to novel materials design toward both the preparation of layered materials with new silicate frameworks and the charge density control of layered silicates.

6 . KSW-2: MESOPOROUS SILICA DERIVED FROM KANEMITE BY BENDING SILICATE SHEETS The formation of KSW-2 truly shows the evidence and uniqueness of the use of kanemite for the formation of mesoporous silica [24]. All the structures of mesostructured and mesoporous silica reported so far have been governed by the geometrical packing of surfactants because the formation of mesostructured precursors relies on the cooperative organization of silicate species and surfactants. On the other hand, the formation of novel mesoporous silica (denoted as KS W-2) with square or lozenge one-dimensional (I-D) channels can be prepared by mild acid-treatment of a layered alkyltrimethylammonium (C,TMA)-kanemite complex. Mesostructured precursors of KSW-2 formed through the bending of individual silicate sheets of kanemite. The square- or lozenge-shape of the relatively ordered pores are quite unique and have never been found in all the reported mesoporous and mesostructured materials. A mesostructured precursor of KSW-2 is obtained from a layered C16TMA-kanemite complex by adjusting the pH value below 6. The preparation of the layered complex without a structural change of the silicate is the most important step to obtain the mesostructured precursor. Therefore, the synthesis procedure of KS W-2 is quite different from those of reported KS W- 1 and FSM16 derived fiom kanemite. The TEM image of the calcined KS W-2 (Fig. 7) exhibits relatively ordered square arrangements which show the periodic distance of adjacent pores of ca. 3.3 nm. Striped patterns with the same periodic distance are also observed, strongly supporting the presence of 1-D arrangement of mesopores. The electron diffraction pattern shows the angle of diagonal lines connecting the [I101 spots is close to 90'. On the basis of the TEM results, all the powder X-ray diffraction (XRD) peaks of the calcined KS W-2 are assigned as an orthorhombic structure (C2mm). Typical powder XRD patterns are shown in Fig. 8. Using the N2 adsorption isotherm of the calcined KSW-2, the BET surface area, the pore volume and the average pore diameter were 1100 m2 g-l, 0.46 cm3 g-l and 2.1 nrn, respectively. The mesopores are surrounded by relatively flat silicate walls and typical scanning electron micrographs showed that all the products do not

morphologically change, exhibiting all the images similar to that ofkanemite.

Fig. 7. a) Typical TEM image of KSW-2 (pH 4.0) and the corresponding ED pattern indexed as hkO projection. b) Typical TEM image and the col'responding pattern of as-synthesized KSW-2 (pH 6.0). c) another TEM image of as-synthesized KSW-2 (pH 6.0) showing the bending of silicate sheets. (Copyright: WII.EY-VCH Verlag GmbH)

This indicates that kanemite does not dissolve during the syntheses of both layered C16TMA-kanemite complex and as-synthesized KSW-2. Actually, the reactions were conducted relatively lower pH values where silica does not dissolve. In order to investigate the formation mechanism of the as-synthesized KSW-2 several samples were prepared at various pH values above 4.0. In the 2 9 ~MAS i NMR spectrum of the layered C16TMA-kanemite complex several peaks due to different Q' environments were mainly observed in the range from -95 to -105 ppm, whereas a broad peak centered at -110 ppm due to a Q4 environment was detected as a minor component. This result reveals that the

single silicate sheet structure in kanemite is retained essentially during the

Fig. 8. XRD patterns of a layered C16TMA-kanemite complex, and KSW-2 prepared at pH 6.0 heforc h) and after c) calcination. (Copyright: WILEY-VCH Verlag GmbH)

synthesis of the layered C16TMA-kanemite complex. At the pH value of 8, the spectrum of the product hardly changed and the peak intensity at -1 10 ppm increases slightly. In contrast, the profiles of the acid-treated products obtained at the pH values below 6.0 vary dramatically; both the Q3 and Q' peaks centered at -101 and -110 ppm are observed, respectively, being in accordance with the structural change from the layered C,sTMA-kanemite complex to assynthesized KSW-2. Soon after the structural change (pH = 6.0), the as-synthesized product was thoroughly observed by HRTEM. As well as those observed for the calcined KSW-2 (pH = 4), similar TEM images were observed and the periodic distance of adjacent pores was ca. 4.0 nm (Fig. 7b); the angle of diagonal lines connecting the [I101 spots was variable, ranging from nearly 90" to ca. 70" in the electron diffraction patterns. The images are reproducible and the possibility of superposition of striped patterns was denied because the angle of the diagonal lines falls into a limited value (70 - 90") and the images were observed at thin parts of the sample. The bending of individual silicate sheets, which is not fully linked between

adjacent layers, was partly observed (Fig. 7c). This observation is reproducible;

Fig. 9. XRD patterns (left) and 29Si MAS NMR spectra of samples during the acid-treatment of layered C16TMA-kanemite complex at various adjusted pH values with keeping the samples at those pH values for a very short period.

the bending was directly observed by HRTEM for KSW-2 synthesized in another batch. In addition, the range of pH values during the acid-treatment cannot lead to dissolution of silicate species, but lead to their condensation. The observed wall thicknesses of the products during the acid-treatment were almost constant (0.6 - 0.7 nm),being consistent with the thickness of the silicate sheet in kanemite. These strongly suggest that the as-synthesized KSW-2 can be obtained from the layered CI~TMA-kanemite by bending of the individual silicate sheets. Even in the TEM image of the calcined KSW-2 (pH = 4), bent silicate sheets were slightly observed. Fig. 9 shows that the structural variation of the samples treated at various pH. The transformation from a layered structure to three dimensional network is clearly observed. On the basis of the variations in the C16TMA/Siratio, derived i NMR (Q4/(Q3+Q4)ratio) of the products during from CHN data, and 2 9 ~MAS the acid-treatment (Fig. lo), the transformation steps of the layered complex into as-synthesized KSW-2 can be categorized as follows. (I) The Q ~ / ( Q ~ + Qratio ~) increased in the range of pH = 9.6 - 7.0 with the slightly decreasing C16TMA/Si

ratio. This observation suggests that the beginning ofthe structural change is 04

-

-0 s -am

1

-&I5

rn

-c.

Q.4QW)

aio

- > : *D 03

M (a0

1

- 0.a OM 9.D

8

7.0

SO

5511

4.0

5.0

-pn

Fig. 10. Variation in the amounts of C16TMA ions and the Q41(Q3+Q4) ratios during acid treatment. (Copyright: WILEY-VCH Verlag GmbH)

e4

e4

silicate species. The formation of species manifested by the formation of occurred at the intralayers because the structural change at pH = 8.0 was hardly observed by XRD. (11) In the range of pH = 7.0 - 6.0, the transformation of the layered complex into as-synthesized KSW-2 is caused by partial removal of C t 6 W A ions. The Q 4 / ( @ + ~ 4 )ratio increased further, indicating the condensation between adjacent layers as well as progressive intralayer condensation. (111) In the structural change at the pH values lower than 6.0, additional condensation among @silicate species occurs between the adjacent layers. KSW-2 with a square I-D arrangement is thought to be formed through these processes. The layered silicate network originating from the structure of kanemite connects two-dimensionally and is not destroyed under the conditions used. Thus, the bending of individual silicate sheets may be caused by the limited structural changes due to partial intralayer condensation and accompanying structural change of CL6TMAassemblies during the gradual leaching. On the basis of the crystal structure of kanemite, the intralayer condensation of S i 4 H groups is possible only in the limited direction. Further interesting result is found that the XRD profiles in the range of 15" to 30° are somewhat different from those observed for the mesoporous materials reported up to date. Though halo XRD peaks were observed for the KSW-2

materials obtained at pH = 4, the XRD patterns of the KS W-2 materials prepared at pH = 6 showed somewhat unusual peaks in the range of 15" to 30" (Fig. 8). The XRD pattern of the layered C16TMA-kanemite complex at higher angles showed a broad peak centered at 20 = 21. l o and a sharp one at 24.3", suggesting that the structural units of kanemite are partly retained in the layered C16TMA-kanemite complex at least. Even in the as-synthesized KSW-2, these peaks remained with some broadening. Although further broadening of such peaks was observed after calcination at 550 "C for 6 h, the profile in the range of 15" to 30" is somewhat different from those observed for the mesoporous materials reported so far. We utilized a layered CI6TMA-kanemite complex as the starting material, in which the basic structural framework was retained at least partly after the formation of the layered complex. The formation mechanism of KSW-2 proposed here is based on the bending of intralayer-condensed silicate sheets of kanemite. Consequently, the square pore system has never been found among the reported mesoporous materials and is not defined by the geometrical packing of surfactants molecules. Although the frameworks are not fully retained after calcination, this approach is a way to incorporate inorganic structural units to mesostructured materials. 7. REACTIONS OF KANEMITE WITH SURFACTANTS WITH VARIOUS PACKING PARAMETERS

In order to investigate the reactivity of kanemite, silica-based mesostructured materials were prepared by the reactions with various cationic surfactants such as alkyltrimethylammonium (C,TMA, n = 12-1 8, 22), alkyltriethylammonium (C,TEA, n = 14-22), and gemini-type diammonium (C16-3-1, C16-3-16, C 165-16) surfactants. The relation between the geometrical packings of surfactant assemblies and the mesostructures derived from kanemite was investigated. The surfactant molecules were assembled in the two-dimensionally limited space of kanemite, leading to the formation of lamellar, 2-d hexagonal, and disordered phases. The lamellar phases were synthesized by using C,TMA (n = 16, 18) and C,TEA (n = 20, 22) surfactants with longer alkyl chains where the N/Si molar ratios were 2.0. 2-d Hexagonal phases were formed where the NISi ratios were 0.2 (C,TMA, n = 12-18). Even in the case of N/Si = 0.2, lamellar phases were obtained by using C22TMA and C22TEA surfactants because the surface curvatures become lesser. The lamellar phases were transformed into 2-d hexagonal phases by acid treatment. The results by using C22TMA and &TEA prove that the 2-d hexagonal phases are formed through layered intermediates

composed of the surfactants and fragmented silicate sheets. The disordered phases were obtained by using C,TEA (n = 14-20) and C16-3-] surfactants with larger surface curvatures of the surfactant assemblies. The two-dimensionally limited space prevents C,TEA and C16-3-1surfactants from assembling spherically. Lamellar phases were also formed by using C16-3-16 and C16-5-16 surfactants and the acid treatment of the lamellar C16-5-16-silicate complex induced the transformation into a 2-d hexagonal phase. The formation of ordered and disordered mesostructured materials derived from kanemite can be simply summarized on the basis of surfactant assemblies in the two-dimensionally limited space. The details are presented below. By the presence of two-dimensionally connected silicate networks, ordered mesoporous silicas with three-dimensionally connected mesopores that are observed for MCM-48, SBA-1, and SBA-2 have never been synthesized in the surfactant-kanemite systems [24]. The formation of several mesostructures is possible in the CnTMA-kanemite systems, and the complicacy is due to the twodimensional nature of kanemite. The formation of an orthorhombic mesostructure (KSW-2) is understood on the basis of both the geometrical packing of CI6TMAand the interactions of the cationic headgroups with the silicate sheets. The assemblies of C16TMA molecules within semi-squared spaces cannot simply be explained by using the geometrical packing. The orthorhombic mesostructure is formed through the bending of the individual silicate sheets that are not so flexible as monomeric andfor oligomeric silica species. Thus, the surface curvature of the silicate sheets does not match that of CI6TMA assemblies completely and the C I ~ T M A molecules are encapsulated within the semi-squared spaces because of the interactions of the cationic headgroups of the C16TMA molecules with the silicate sheets. Therefore, the present study on the reactions of kanemite with various cationic surfactants that are assembled with larger surface curvatures is quite important for further understanding the surfactant-kanemite systems to produce mesostructured materials. All the reactions were performed at 70 OC for 3 h. The reactions of kanemite with monovalent ammonium surfactants (C,TMABr , C,,TEABr) were conducted according to the previous papers [14, 16, 241. The reaction conditions of kanemite with divalent ammonium surfactants (C16-3-17C1&3-16,C16-5-16), were selected on the basis of the solubility of the divalent ammonium surfactants. The reaction of kanemite with C16-3-1was performed by using the same method of the CnTMABr- and CnTEABr-kanemite systems. In addition, acid-treatment of lamellar phases obtained by the reactions with C16-3-16 and

C16-5-16was conducted by the addition of 2N HCI; the pH values of the suspensions were adjusted to 8.5 and the stirring was kept at 70 OC for 3 h. 7.1. Reactions with Alkyltrimethylammonium Surfactants. The XRD patterns of calcined materials prepared by using C,TMABr (n = 12-18), where the preparation was followed by the typical synthesis procedure of FSM-16 derived from kanemite (NISI = 0.2) [14], are shown in Figure I l(a)-(d). Four diffraction peaks assignable to 2-d hexagonal phases (space group; p6mm) are distinctly observed in low scattering angles. The dloo-spacings of the calcined materials are linearly increased with the increase in the alkyl chain lengths of the C,TMA surfactants used (C12TMA; 2.9 nm, CI4TMA;3.2 nm, Ci6TMA; 3.5 nrn, C18TMA;3.9 nm). All the N2 adsorption isotherms of the calcined materials showed type IV behaviors characteristic of ordered mesoporous silicas. The BET surface areas, the pore volumes, and the average pore sizes (r) are shown in Table 1. The unit cell parameters (ao) and the wall thicknesses (w) of the mesoporous silicas were calculated by the equations of a 0 = 2/43 X dloo and w = a. - r, respectively; the wall thickness is almost constant (Table 1). The dloo-spacingof the calcined material prepared by using C22TMABrwas 4.4 nm, being in agreement with the aforementioned relation between the dIoospacing of the FSM-type mesoporous silicas and the alkyl chain length of the C,TMA surfactants used [26]. However, the use of C22TMA cations for the synthesis of FSM-16 led to the broadening of the XRD peaks (Fig. 1l(e)). Even in the synthesis of 2-d hexagonal MCM-41, lamellar MCM-50-type silicas are formed by using C,TMA surfactants with longer alkyl chains such as C~OTMA and C22TMAcations. Although the use of C22TMAcations is not advantageous for the synthesis of MCM-41, MCM-41-type mesoporous materials can be obtained under optimal synthetic conditions [27]. The XRD patterns of the as-synthesized materials prepared by using C,TMABr (n = 12-18, 22) under the typical synthetic conditions for the synthesis of layered C16TMA-silicate complexes derived from kanemite (N/Si = 2.0) are shown in Fig. 1l(f)-('j). As shown in the figure, the main peaks assignable to (001) and the higher order diffractions are observed for the assynthesized materials prepared by using C16TMABr, C18TMABr, and C22TMABr (CI6TMA; 3.2 nm, C18TMA;3.4 nm, C22TMA;3.5 nm). However, in the cases of C12TMABrand CI4TMABr,broad peaks were collected in low scattering angles and the d-spacings (C14TMA; 3.4 nm, C18TMA; 3.7 nm) were larger than those observed for the layered C,TMA-silicates (n = 16, 18, 22) in

Fig. I I. (left) XRD patterns of calcined materials prepared by using (a) C12TMABr,(b) C14TMABr,(c) C16TMARr,(d) C18TMABr, and (e) Cz2TMABrwhere the N/Si molar ratios were 0.2. (right) XRD panerns of as-synthesized materials prepared by using (0 C,flMABr, (g) CMTMAB~, (h) C d M A B r , (i) ClnTMABr, and C2?TMABr where the N/Si molar ratios were 2.0.

u)

Table 1. Characteristics of FSM-type mesoporous silicas prepared by using C,,TMABr (n =

12-18) and C22TEABr. Surfnetant BET surface area /m2 g' CUTMA 650

Pore volume Average pore ImI. g' size (r) I nm 0.34 2.1

Unit cell parameter (00)1 ~n 3.3

Wall thickness (w)1 nm 1.3

spite of their shorter alkyl chains. T h e solubilities of C12TMABrand C,4TMABr are higher than those of C,TMABr ( n = 16, 18,22). In addition, the reactivity of layered materials with surfactant molecules with shorter alkyl chains is lower. Then, the amounts of introduced C12TMA and C14TMAcations between the silicate sheets of kanemite are not enough to be assembled as lamellar phases, meaning that the C12TMA- and CI4TMA-silicates are not layered materials.

7.2. Reactions with Alkyltriethylarnrnoniurn Surfactants. The XRD patterns of calcined materials prepared by using C,TEABr (n = 14-20) where the N/Si molar ratios were 0.2 are shown in Figure 12(a&(d). Being different from the C,TMA-kanemite system, disordered phases were obtained. However, the XRD pattern of the calcined material prepared by using C22TEABr showed a successful formation of a 2-d hexagonal phase (Figure 12(e)), as reported by us recently [26]. The TEM image of the calcined material revealed the presence of ordered hexagonal arrangements of mesopores and the N2 adsorption data showed that the material has high surface area, pore volume, and large pore size (Table 1). The formation processes of the 2-d hexagonal mesoporous silica obtained from the C22TEA-kanemite system were investigated by XRD and 2 9 ~MAS i NMR. During the synthesis of the mesoporous silica, samples were recovered before and after the pH adjustment at 8.5. The XRD pattern of the sample before the pH adjustment showed that a peak at the d-spacing of 4.3 nrn and only the higher order diffractions were observed, suggesting a layered CZ2TEA-silicate

Fig. 12. (left) XRD patterns of calcined materials prepared by using (a) CldTEABr, (b) CI~TEABT, (c) ClsTEABr, (d) CzoTEABr, and (e) CzzTEABr where the N/Si molar ratios were 0.2. (right) XRD patterns of as-synthesized materials prepared by using (0 ClrTEABr, (g) C I ~ T E A B(h) ~ , CI~TEABT, (i) CzoTEARr, and (i)CuTEABr where the N/Si molar ratios were 2.0.

can be synthesized by the reaction of kanemite with C22TEABr (before pH adjustment) [26]. Surfactant molecules with longer alkyl chains are likely to be assembled as lamellar phases because of their geometrical packings [28]. The XRD pattern of the sample after the pH adjustment showed the appearance of four diffraction peaks assignable to a 2-d hexagonal phase accompanied with a slight amount of the remaining layered phase. During the acid treatment (the pH adjustment), the Cz2TEA cations were partly removed out of the interlayer spaces and the silicate framework of kanemite was condensed further (as described below), meaning that lamellar assemblies of the C22TEA cations are changed into rod-like micelles to induce the 2-d hexagonal phase. The Z 9 ~MAS i NMR spectra of the samples recovered during the synthesis of the mesoporous silica are shown in Fig. 13. Before the pH adjustment (layered Cz2TEA-silicate), Q2 ((Si0)2Si02), Q3 ((Si0)3SiO), and Q4 ((Si0)aSi) signals were observed at around -90 ppm, -100 ppm, and -110 ppm, respectively. The 2 9 ~MAS i NMR measurement of the layered C22TEA-silicate was performed without drying the layered Cz2TEA-silicate. In addition to those

-70

-80

-90

-100

-110

-120

-150

140

Chemical shift /ppm

Fig. 13. 2 9 ~ MAS i NMR spectra of the samples obtained during the synthesis of a 2-d hexagonal mesoporous silica by using Cz2TEABr;before pH adjustment (a) without and (b) with drying, (c) aAer pH adjustment and (d) the calcined material.

peaks, Q0 (Si04) and Q' ((SiO)Si03) peaks were detected, indicating the and Q3 silicate presence of soluble silicate species. The presence of both species is the direct evidence on the fragmentation of the individual silicate sheets of kanemite in the layered Cz2TEA-silicate though the fragmentation size has not been clear. The condensed Q4 silicate species are formed by intralayer condensation and the reaction of the individual silicate sheets with the soluble ratios ~ ) were increased by pH adjustment (2-d silicate species. The Q ~ / ( Q ~ + Q hexagonal C22TEA-silicate) and the following calcination. Similar results were obtained in the Cz2TMA-kanemite system. Although the formation mechanism of FSM-16 has been proposed by TEM and in-situ XRD [17, 22, 291, the formation mechanism of FSM-type mesoporous silicas is proved by the results in this study. The XRD patterns of as-synthesized materials prepared by using C,TEABr (n = 14-22), where the NISi molar ratios were 2.0, are shown in Figure 12(f)-0). The main peaks assignable to (001) and the higher order reflections are observed for the as-synthesized materials prepared by using C20TEABr and C22TEABr (C20TEA; 3.7 nm, C22TEA; 3.9 nm). As in the case of the C,TMA-kanemite system, the XRD patterns showed that layered C,TEA-silicates cannot be obtained by using C,TEABr with shorter alkyl chains (n = 14-18) (C14TEA;3.7 i NMR spectra of the nm, CI6TEA; 3.9 nm, CI8TEA; 3.8 nm). The 2 9 ~MAS and Q4 silicate species layered C,TEA-silicates are shown in Figure 14. Both are present in the layered C,TEA-silicates though kanemite is composed of only silicate species [lo, 301. The Q4 silicate species are formed by intralayer condensation depending on the reaction temperatures.

e2

e3

e3

7.3. Reactions with Gemini Surfactants.

The XRD patterns of the as-synthesized materials prepared by using a gemini type C16-3-1 surfactant (N/Si = 0.2, 2.0) and the calcined materials are shown in Figure 15. In both of the cases, the broad peaks at the d-spacings of ca. 4.5 nm were observed for the as-synthesized materials. In addition to the main peaks, the higher order diffractions with very weak intensities are collected as shown in the figure. In spite of the same alkyl chain lengths, the dloo-spacingsof the C1c3-1-silicates were ca. 4.5 nm, being larger than that observed for 2-d hexagonal C16TMA-silicate (ca. 4 nm). Although it is possible to index the peaks to 2-d hexagonal phases, the peaks are broadened further after calcination. The C16-3-1 surfactants are assembled spherically, being useful for the synthesis of C1c3-1-silicate with 3-d hexagonal phase (SBA-2, space group; P631mmm) [3 11. The formation of such spherical assemblies are not conceivable within the two-dimensionally limited space of kanemite. Because the surface

(b,

A ; -80

-90

-100

-110

-120

-130

-140

Chemical shift /ppm

Fig. 14. 2 9 ~MAS i NMR spectra of the as-synthesized materials prepared by using (a) Cr4TEABr,(b) C!eTEABr, (c) ClsTEIZBr, (d) CzoTEAUr, and (e) CzzTEABrwhere the NISi molar ratios were 2.0.

Fig. 15. XRD patterns of the products ohtained duringthe synthesis of C16,.1-silicates where the NISi molar ratios wcre (a), (c) 0.2 and (b), (d) 2.0; (a), (b) after pH adjustment and (c), (d) the calcined material.

curvature of the spherical assemblies is higher than that of rod-like micelles are formed in the composed of C,TMA molecules, disordered C16_3_~-silicates C16_3_1-kanemite system. The XRD patterns of the as-synthesized materials prepared by using other gemini type C16-3-16 and C165 16 surfactants are shown in Figure 16. The peaks at the d-spacings of 3.5 nm and the higher order diffractions are observed in the XRD patterns of C 1 ~ _ 1 ~ s i l i c a recovered tes before and after pH adjustment. The C1&3-16surfactant has a tendency to be assembled as lamellar phase or rodlike micelles according to the synthetic conditions, being useful for the synthesis of C16_3-l-silicate with lamellar (MCM-50) and 2-d hexagonal phases (MCM41, SBA-3). Therefore, the formation of the layered C16-3 16-~ili~ate is advantageous within the limited interlayer space of kanemite. In contrast, in the C165.16-kanemite system, both lamellar (3.5 nm) and 2-d hexagonal phases (3.9 nm) were obtained as mixed products and the 2-d hexagonal phase (dloa= 4.0 nm, dll0 = 2.3 nm, d 2 =~2.0~ nm) was mainly observed after pH adjustment. The result indicates that the layered C165 16-sili~ate is also transformed into the 2-d hexagonal phase by acid treatment as well as layered Cz2TMA- and Cz2TEA-silicates. The use of C16d-16and CI66-lh surfactants is useful for the formation of C164-16-and C16.61~silicates with 2-d hexagonal phases (MCM41, SBA-3) mainly. By increasing the alkyl chain length between diammonium groups (spacer) in gemini type surfactants, the C16-5 l6 molecules are assembled as rod-like micelles in the present case.

Fig. 16. XRD patterns of the products obtained during the synthesis of (a), (c) GIG-16- (NISi = 0.32) and (b), (d) C1a.j.l6-silicates ( N i s i 0.2); (a), ( b ) before and (c), (d) aftcr pH

adjuslmcnt.

7.4. Formation of Mesostructured Materials Derived from Kanemite. In the C,TMA-kanemite system, the discussion on the formation of FSM16 has started as one of the research topics why the 2-d hexagonal phase with 3d silicate networks is allowed to form from kanemite with 2-d silicate networks. Originally, the folded sheets mechanism has proposed; rod-like micelles of the C,TMA molecules are formed with the cooperative bending of the individual silicate sheets of kanemite around the micelles. However, the presence of a layered intermediate during the formation of the 2-d hexagonal phase was claimed on the basis of in-situ XRD data [22]. Although both single and double layers of the silicate sheets must be observed by TEM if the 2-d hexagonal phase is formed through the folded sheets mechanism, such parts have never been observed and the wall thickness of FSM-16 is almost constant [29]. Thus, the presence of fragmented silicate sheets during the formation of FSM- 16 has been speculated, as was firstly pointed out that the disordered KSW-1 is formed through the fragmentation of the silicate sheets of kanemite. In this study, a layered intermediate was recovered and then the presence of fragmented silicate sheets was proved on the basis of the NMR data. The schematic formation routes of ordered mesostructured materials derived from kanemite are shown in Scheme 1 on the basis of the present and previous results [23, 24, 261. The formation of the 2-d hexagonal phase (FSM16) is explained above. Layered surfactant-silicates derived from kanemite are unique because the silicate framework contains condensed ordered silicate species. The formation of the condensed silicate species has already been proved; intralayer condensation occurs within the individual silicate sheets of kanemite and the degree of the condensation is controllable with the reaction temperature. A reaction of kanemite with C16TMAcations at room temperature leads to the formation of a layered C16TMA-silicate composed of mainly @ silicate species. Mild acid treatment of the layered CI6TMA-silicate with retaining the kanemite structure induces the mesostructural transformation into an orthorhombic phase (KSW-2, space group; C2mm) that is truly formed by the bending of the individual silicate sheets of kanemite. The orthorhombic structure can be formed because of the 2-d connecting silicate framework originated from kanemite. In the MCM-type mesostructured materials, mesostructural transformation is also observed during thermal and hydrothermal posttreatment~.However, various chemical reactions occur; silicate species are solubilized and bonded again, surfactant molecules are partly degraded, the derivative organic molecules are solubilized in the remaining surfactant assemblies, and so on. In contrast, in the C,TMA-kanemite systems, layered

,d //*Y.n

9

,, ?.

,*

-&@a

Kanemite

(FSM-16)

'intralayer condensation'

RT NISI = 2.0

i

'bendha or silicate 'Intralayer condensation'

Scheme I. Schematic formation routes of ordered mesostructured and mesoporous materials derived from kanemite. (Copyright: The Chcmical Society of Japan)

C,TMA-silicate phases are present as key materials during the formation of each mesostructured material. The silicate frameworks of the layered C.TMA-silicate phases have already been investigated in detail Consequently, in the present study, the mesostructural transformations can be simply summarized by using the structural change of the silicate frameworks of kanemite such as fragmentation, intralayer condensation and bending. 7.5. Surfactnnt Assemblies in the Two-DimensionallyLimited Space.

In the synthesis of MCM-type mesoporous silicas, the mesostructures formed through the cooperative organization have often been explained by the geometrical packing of the surfactant molecules. The presence of inorganic species attached to the hydrophilic headgroups induces mesostructural variation depending on the synthetic conditions that reflect the degree o f silica condensation. In the surfactant-kanemite systems, 2-d silicate networks originated from kanemite always affect the formation of mesostructured materials because of the interactions of the silicate frameworks with surfactants.

The formation of an orthorhombic mesostructure (KSW-2) is a good example proving that surfactant molecules are not freely assembled in accordance with the geometrical packing within the limited space. The ClaTMA molecules are allowed to accommodate in the semi-squared spaces because of the interactions of the cationic headgroups with the silicate sheets. Here, the formation of disordered phases is also discussed on the basis of both the geometrical packing of surfactants used and the interactions of the silicate frameworks with the surfactants. The formation routes of ordered and disordered materials derived from kanemite are summarized in Scheme 2. Disordered phases are formed by the reactions of kanemite with C,TEA (n =14-18) and C16-3-1surfactants. Although layered materials are very important for the formation of ordered phases in the surfactant-kanemite systems, these surfactant molecules cannot be assembled as lamellar phases. The C,,TEA and

Within2-0' Limited apace

2-6Hexagonal (FSM-16) 'Transformation'

(KSW-2)

-

Formation mute M OMered phases Formation route of dlsorderedphaoes

Scheme 2. Schematic formation mutes of ordered and disordered materials derived from kanemite. (Copyright: The Chemical Society of Japan)

C16-3-1surfactants tend to be assembled spherically because of the geometrical packings. One-directional bending of the silicate sheets is possible for the formation of ordered mesostructured precursors (KSW-2). However, twodirectional bending of the silicate sheets in order to match the surface curvature between the spherical surfactant assemblies and the bent silicate sheets is not rational because the spherical surfactant assemblies cannot be surrounded by the silicate sheets with long range networks. Thus, the disordered phases that afford mesoporous silicas are obtained through the bending of the silicate sheets according to the interactions of the silicate frameworks with those surfactants. This section can be concluded as follows. Silica-based mesostructures are derived from a layered polysilicate kanemite, and the formation is induced by the reactions with various cationic surfactants with different geometrical packings. In the surfactant-kanemite systems, several reactions such as the interactions of cationic surfactants with the silicate sheets, the fragmentation or bending of the silicate sheets, and the geometrical packing of surfactant molecules are complicated. However, the formation of ordered (lamellar, 2-d hexagonal, and 2-d orthorhombic) and disordered mesostructured materials can be simply summarized on the basis of the presence of layered surfactant-silicate intermediates. The obtained insights are potentially applicable for controlling mesostructured materials obtained from layered polysilicates in both short (framework) and long range scales (mesostructure) to realize the preparation of mesostructured and mesoporous materials with truly crystallized silicate frameworks.

8. MESOPOROUS SILICA DERIVED FROM a-DISODIUM DISILICATE It is quite interesting to survey whether the reactivity of kanemite (derived from 6-disodium disilicate) and a similar phase derived from a disodium disilicate is same or not to investigate further the usefulness of layered silicates for the formation of mesoporous silica. 6-Na2Si205(precursor for kanemite) and aNa2Si205 have different framework structures; the 6 phase has silicate sheets with boat-type 6-membered rings whereas the a phase possesses silicate sheets with chair-type 6-membered rings [32,33]. Therefore, we can expect that the degree of condensation during the formation of mesostructured materials should be different and that the assembling structure of surfactants should also be different, which may affect the structure of mesoporous silica after calcination. There have been no reports on the formation of mesostructured materials derived from a-Na2Si205and related materials, though Lagaly et al. reported

that acid-treated H2Si205-I(acid-treated phase of a-Na2Si205)reacted with some alkylamines to form intercalation compounds [5, 341. In the present study, based on the synthetic procedure established for FSM-16, hydrated a-sodium disilicate was used as a starting layered polysilicate for the formation of mesostructured materials and the products were characterized to find the influences of the different structural features of layered disodium disilicates on the formation of mesostructured materials. The composition of hydrated a-sodium disilicate prepared was Na, 2H08Si205.2.5H20.The 29Si MAS NMR spectrum of hydrated a-sodium and units, indicating the disilicate shows several signals in the region of diversity of Si environments, which is basically consistent with the previous reports [35, 361. The hydrated behavior is quite different fi-om that of kanemite. However, the Raman spectra of a-Na2Si205 and hydrated a-sodium disilicate show similar patterns, suggesting the retention of the original silicate framework in the hydrated form. The layered nature of the hydrate was also supported by the intercalation behavior with C16H33m2.The original silicate framework of aNa2Si205is probably preserved in its hydrated a-sodium disilicate. The influence of the different structural features of layered disodium disilicates on the formation of mesostructured materials is verified. In the CI6TMA/Simolar ratios of 0.2-1.0 (CI6TMA/Si = 0.2, 0.5, 0.7, and 1.0), 2-d hexagonal structures were formed from hydrated a-sodium disilicate though a less ordered structure was formed when the ratio was 2.0. The uniformly arranged straight channels are observed in the TEM image of mesoporous silica obtained by calcination of the precursor prepared at the ratio of 2.0. This is characteristic of the system using hydrated a-sodium disilicate which should bend along the a axis of the silicate structure. On the other hand, in the case of kanemite, lamellar mesostructures were formed at the ratios higher than 0.5, while a 2-d hexagonal structure was formed only at the ratio of 0.2. Longer distance between Si-0- sites along the bending direction in hydrated a-sodium disilicate than that in kanemite, making the headgroups of C,TMA ions apart, may be the reason for these differences. Mesoporous silica derived from hydrated a-sodium disilicate is prepared at lower pH than from kanemite at the C16TMA/Si ratio of 2.0. More importantly, the acid treatment, which is normally required for the preparation of mesoporous silica from kanemite, was not needed for the formation of 3-d silica network from hydrated a-sodium disilicate at the ratio of 2.0. These findings are quite useful for future design of the framework of mesostructured silica because mesostructured materials can be prepared under lower pH conditions that are advantageous for retaining the original silicate structures to some extent [37].

e2 e3

9. FUTURE OUTLOOK The use of layered silicates has advantages for controlling the wall nature of mesoporous silica. There are some reports on the different catalytic activities between MCM-4 1 and mesoporous silica derived from kanemite. Well-designed silicates with controlled structure and connectivity will be available by chemistry of silica and silicates. From this viewpoint the use of various types of layered silicates would provide other opportunities in future materials design. One of the other promising ways to construct novel silicate structure is the use of sol-gel process of well defined precursor molecules. We have recently reported that organotrialkoxysilanes with long alkylchains are utilized to construct lamellar mesophase composed of siloxane layers and alkyl groups [ 3 8 ] . When alkoxytrichlorosilanes are used lamellar silica-alcohol composites are obtained, which strongly indicates silica-organic nanomaterials are obtained by hydrolysis and condensation of only single molecules without any additives [ 3 9 ] . Very recently we have reported that lamellar and 2d-hexagonbal-like mesostructures are directly formed from tris(trialkoxysiloxy)alkylsilane by simple hydrolysis and condensation without any additives [40]. Such a molecular design would be utilized in a more effective way to construct novel nanomaterials with controlled pore morphology, pore arrangement, and wall composition and structures. Mesostructured materials before removal of organic parts can not be regarded as just a precursor for mesoporous materials and are important as organic-inorganic nanomaterials. The host-guest interactions between inorganic framework and guest species are quite important in the sophisticated design of nanomaterials.

10. CONCLUSIONS The discovery of the formation of mesoporous silica derived from a layered polysilicate kanemite and the subsequent development of mesoporous silica derived from layered polysilicates are reviewed. All the results accumulated up to the present stage show that layered silicates are one of the important starting materials for the preparation of mesoporous silica from not only the academic viewpoints including silicate chemistry but also practical applications including catalysts. The results on the formation of mesostructured silica-organic systems indicate the importance of the layered silicates. Layered silicates are unique starting materials for the formation of mesoporous silica. The silicate frameworks play crucial roles in forming the walls of mesoporous silica.

Mesoporous materials can not be categorized as a sub-member of zeolites. Mesoporous materials are related to much more diverse fields. The variations in the compositions, structure, and morphology are very promising for future technological advancement including the fields of catalysts, adsorbents, host matrices for host-guest interactions, and nanoscience and nanotechnology. ACKNOWLEDGEMENTS I express my deepest thanks to Professor Osamu Terasaki (Stockholm University) for his continuing fruitful discussion through his high resolution TEM images of many mesoporous silica and giving me an opportunity to write this review. Professor D. O'Hare and Dr. S. O'Brien (both at University of Oxford) did excellent in-situ work on the mesophase formation that I deeply appreciate. I thank Professor C. Kato (Emeritus professor of Waseda University) for his enchanting me to silicate chemistry. I also thank Dr. T. Yanagisawa (former PhD student) and Mr. T. Shimizu (former master course student) for their main experimental contribution to the discovery. The contribution by Dr. T. Kimura (AIST Nagoya, Japan) is quite large for the discovery of KS W-2 by his unparallel hard work. Some parts of this manuscript are also supported by his efforts. Dr. T. Shigeno (former PhD student), Mr. Daigo Itoh (former master course student), Ms. Nanae Okazaki (former master course student), Mr. T. Kamata (former master course student) Ms. Y. Takano (former master course student), Mr. K. Inoue (former master course student), and Mr. M. Kato (former master course student) are also acknowledged for their enthusiastic contributions to the formation of mesoporous silica. In particular, I also thank Professors Y. Sugahara and M. Ogawa (Waseda University) for their continuing supports and discussions. The work on mesoporous silica has been supported by many funds. Grants-in-Aid for Scientific Research by Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan are gratefully acknowledged. A Grantin-Aid for COE Research and the 21st century COE program by the same organization are also deeply acknowledged. Waseda University always supports our research through various supporting systems to which I have owed so much. REFERENCES [I] R.M. Barrer, Hydrothermal Chemistry of Zeolites, Academic Press (1982). [2] I. V. Mitchell ed., Pillared Layered Structures, Current Trends and Applications, Elsevier Applied Science (1990). [3] G. Lagaly, in "Chemical Reactions in Organic and Inorganic Constrained Systems", Ed by

R. Setton, D. Reidel Pub., Dordrecht (1986) p.361. [4] E. Ruiz-Hitzky and J. M. Rojo, nature, 287 (1980) 28. [5] G. Lagaly, Adv. Colloid Interface Sci. 11 (1979) 105. [6] a) T.Shimizu, T.Yanagisawa, K.Kuroda, and C.Kato, Abstract No. 1x11 D42(I-761), Annual Meeting of the Chemical Society of Japan (1988); b) K.Kuroda, T.Yanagisawa, T. Shimizu, and C.Kato, Abstracts of the 9th Int. Clay Conf. Strasbourg(l989) p.222; c)T.Yanagisawa, T.Shimizu, K.Kuroda, and C.Kato, Bull. Chem. Soc. Jpn. 63 (1990) 988. [7] a) C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, and J.S. Beck, Nature, 359 (1992) 710; b)J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T-W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, and J.L. Schlenker, J. Am. Chem. Soc. 114 (1992) 10834. [8] Z. Johan and G.F. Maglione, Bull. Soc. Fr. Miner. Cristallogr., 95 (1972) 371. [9] K. Beneke and G. Lagaly, Am. Mineral., 62, (1977) 763. [lo] S. Vortmann, J. Rius, B. Marler, H. Gies, Eur. J. Mineral., 11 (1999) 125. [ l l ] M.T. Le Bihan, A. Kalt, R. Wey, Bull. Soc. Fr. Miner. Cristallogr., 94 (1971) 15. [12] T. Yanagisawa, T. Shimizu, K. Kuroda, and C. Kato, Bull. Chem. Soc. Jpn. 63 (1990) 1535. [13] T. Yanagisawa, K. Kuroda, and C. Kato, React. Solids 5 (1988) 167. [14] S.Inagaki, Y.Fukushima, and K.Kuroda, J. Chem. Soc. Chem. Commun., (1993) 680. [15] Y. Fukushima, S. Inagaki, and K.Kuroda, Nippon Kagaku Kaishi, (1995) 327. [16] S.Inagaki, A. Koiwai, N. Suzuki, Y.Fukushima, K.Kuroda, Bull. Chem. Soc. Jpn., 69 (1996) 1449. [I71 C-Y. Chen, S.Q. Xiao, and M.E. Davis, Microporous Mater., 4 (1995) 1. [18] Y. Inaki, H. Yoshida, K. Kimura, S. Inagaki, Y Fukushima, T. Hattori, Phys. Chem. Chem. Phys., 2 (2000) 5293. [19] S. Inagaki, Y. Fukushima, and K. Kuroda, Stud. Surf. Sci. Catal., 84 (1994) 125. [20] T. Yanagisawa, K. Kuroda, and C. Kato, Bull. Chem. Soc. Jpn. 61 (1988) 3743. [21] K. Endo, Y. Sugahara, and K. Kuroda, Bull. Chem. Soc. Jpn. 67 (1994) 3352. [22] a) S. O'Brien, R.J. Francis, S.J. Price, D. O'Hare, S.M. Clark, N. Okazaki, and K. Kuroda, J. Chem. Soc., Chem. Commun. (1995) 2423 (b) S. O'Brien, R.J. Francis, A. Fogg, D. O'Hare, N. Okazaki, and K. Kuroda, Chem. Matel: 11 (I 999) 1822. [23] T. Kimura, T. Kamata, M. Fuziwara, Y. Takano, M. Kaneda, Y. Sakamoto, 0. Terasaki, Y. Sugahara, and K. Kuroda, Angew. Chem. Int. Ed., 39 (2000) 3855. [24] T. Kimura, D. Itoh, N. Okazaki, M. Kaneda, Y. Sakamoto, 0. Terasaki, Y. Sugahara, and K. Kuroda, Langmuir, 16 (2000) 7624. [25] Z. Q. Deng, J. F. Lambert, and J. J. Fripiat, Chem. Mater., 1 (1989) 375. [26] T. Kimura, D. Itoh, T. Shigeno, and K. Kuroda, Langmuir, 18 (2002) 9574. [27] S. Namba, A. Mochizuki, and M. Kito, Stud. Surf. Sci. Catal., 117 (1998) 257. [28] J. N. Israelachvili, D. J. Mitchell, and B. W. Ninham, J. Chem. Soc., Faraday Trans. 2, 72

(1976) 1525. [29] Y. Sakamoto, S. Inagaki, T. Ohsuna, Y. Fukushima, Y. Nozue, and 0. Terasaki, Microporous Mesoporous Mater., 2 1 (1998) 589. [30] L. A. J. Garvie, B. Devouard, T. L. Groy, F. Camara, and P. R. Buseck, Am. Mineral., 84 (1999) 1170. [31] a) Q. Huo, R. Leon, P. M. Petyroff, and G. D. Stucky, Science, 268 (1995) 1324. b) Q. Huo, D. I. Margolese, and G. D. Stucky, Chem. Mater., 8 (1996) 1147. [32] V. Kahlenberg, G. Dorsam, M. Wendschuh-Josties, and R. X. Fischer, J. Solid State Chem., 146, (1999) 380. [33] M. E. Fleet, J. Solid State Chem., 199, (1995) 400. [34] G. Lagaly, K. Beneke, P. Dietz, and A. Weiss, Angew. Chem. Int. Ed., 13 (1974) 819. [35] X. Ai, F. Deng, J. Dong, L. Chen, and C. Ye, J. Phys. Chem. B, 106 (2002) 9237. [36] W. Wieker, D. Heidemann, R. Ebert, and A. Tapper, Z. Anorg. Allg. Chem., 621 ( 1995) 1779. [37] M. Kato, T. Shigeno, T. Kimura and K. Kuroda, Chem. Mater., submitted. [38] a) A. Shimojima, Y. Sugahara, and K. Kuroda, Bull. Chem. Soc. Jpn, 70 (1997) 2847. b) A. Shimojima, Y. Sugahara, and K. Kuroda, J. Am. Chem. Soc., 120 (1998) 4528. c) A. Shimojima and K. Kuroda, Langmuir, 18 (2002) 1144. d) A. Shimojima, N. Umeda, and K. Kuroda, Chem. Matex, 13 (2001) 3610. 1391 Y. Fujimoto, A. Shimojima, and K. Kuroda, Chem. Mater., 15 (2003) 4768. [40] A. Shimojima and K. Kuroda, Angew. Chem. Int. Ed., 42 (2003) 4057.

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

FSM-16 and mesoporous organosilicas Shinji Inagaki Toyota Central R&D Labs., Inc., Nagakute, Aichi, 480-1 192 Japan

Contents 1. FSM-16 1.1. Introduction 1.2. Background 1.3. Highly ordered mesoporous silica from a layered silicate 1.4. Mesoporous aluminosilicates from a layered silicate 1.5. Applications 1.6. Conclusion 2. MESOPOROUS ORGANOSILICAS 2.1. Introduction 2.2. Mesoporous ethane-silicas 2.3. Extension of mesoporous organosilica system 2.4. Mesoporous aromatic-silicas with crystalline pore walls 2.5. Highly functionalized mesoporous organosilicas 2.6. Conclusion 3. ACKNOWLEDGEMENTS

1.1. Introduction Porous solids are important materials in science and technology due to their wide applications such as catalysis, adsorption and opto-electrical devices. There are many types of porous materials and a considerable effort has recently been made in order to fabricate novel porous materials. Crystalline porous materials have several advantages compared to amorphous porous materials. Zeolites, crystalline aluminosilicates, have a uniform pore at atomic scale and the

uniformity makes it possible to catalyze shape selective reactions obtaining high yields of up to 98% for p-xylene from toluene and methanol, for example. Such a high selectivity is not expected for an amorphous aluminosilicate. However, microporous catalysts show limitations when applied to larger molecules. The cracking reaction of heavy oil has become a necessity for an efficient utilization of petroleum sources. The synthesis of large molecular sized functional organic compounds has also been desired. The needs have arisen to synthesize materials with large pore size. Even in a case of usual small molecules, the molecules are sometimes difficult to difkse in micropores because the molecular dimension is almost equal to the zeolite pore. In 1988-1993, the first syntheses of ordered mesoporous materials with larger pore diameters of 2- 10 nrn were reported using surfactants by some independent research groups including our group [l-61. Since their discovery, several researchers have started to study such materials and the number of publication related has increased rapidly. The surfactant-mediated synthesis strategy has opened a new research field of mesoporous materials. Herein, I describe one of the first mesoporous materials, FSM-16 with highly ordered structure synthesized from a layered silicate material, and highly functionalized periodic mesoporous solids whose framework is composed of an organic-silica hybrid. The hybridization of organic and inorganic materials in the framework of mesoporous materials has brought about not only high functionalization of pore walls but also the formation of very unique structural feature, that is, the crystallization of pore walls. The new type of ordered mesoporous materials with crystalline pore walls are expected to exhibit unique chemical and physical properties due to their hierarchically ordered structure. 1.2. Background

Kuroda et al. reported the synthesis of mesoporous silicas with uniform pore-size distribution from a layered silicate, kanemite [the ideal composition is NaHSi205-3H20]in 1988 [1,2]. Kanemite was treated in a aqueous solution of cationic surfactant, alkyltrimethylammonium chloride [C,H~,+~N+(CH~)~C~-, n=12, 14, 16 or 181 at 60 "C for 2 weeks. Filtered material was calcined at 700 "C to yield mesoporous silica. The pore-size of the mesoporous modified kanemite was distributed in the narrow range and controlled between 1.8 and 3.2 nm by the variation of alkyl-chain length of alkyltrimethylammonium used. The X-ray diffraction pattern showed at least one strong peak at low angle region under 10 degrees (28) and transmission electron micrographs showed a structure

closslinked between the expanded interlayer of kanemite [2]. However, a highly ordered pore arrangement structure with hexagonal symmetry was not observed in the mesoporous material by transmission electron micrograph (TEM). On the basis of the results, mesoporous material was considered to be formed by expanding the interlayer of kanemite with surfactants followed by a partial closslinking between the silicate layers [2]. We have developed Kuroda's strategy and succeeded in the synthesis of highly ordered mesoporous silica, FSM-16 with hexagonal symmetry (published on June, 1993 [6]), and aluminosilicate mesoporous materials with strong acid properties (filing a patent on January, 1991 [7] and presenting at international conference on July, 1992 [3]). We designated the highly ordered mesoporous silica derived from kanemite as "Folded Sheet Mesoporous material'' (FSM- 16). Parallel to our study, researchers in Mobil Corporation synthesized mesoporous materials with hexagonal p6mm (MCM-4 I), cubic Ia-3d (MCM-48) and lamellar (MCM-50) symmetry and published in October, 1992 [4,5]. The family of mesoporous molecular sieves was designated as M41S. Mesoporous MCM-41 shows a two-dimensional hexagonal array of uniform channels whose diameter can be controlled between 1.5 and 10 nm. Although the structure of FSM- 16 resembled that of MCM-41, our work was independent of Mobil's work. We have proposed a "folded sheet" mechanism for the formation of mesoporous material from a layered silicate, while the Mobil group have proposed a "liquid crystal templating mechanism" for the formation of the M41S materials. Although FSM- 16 and MCM-4 1 have the same hexagonal pore arrangements, the structure of the pore walls is different between both materials. The different pore wall structures of siliceous FSM- 16 and MCM-4 1results in different catalytic activity as will be shown later.

1.3. Highly ordered mesoporous silica from a layered silicate We have synthesized highly ordered mesoporous silica FSM-16 with a hexagonal array of uniform channels from a layered silicate kanemite by modifying Kuroda's method [6]. Fig. 1 shows the comparison of synthesis conditions between original and modified methods [8]. We found that higher pH over 11 for the first stage of the reaction and the subsequent pH adjustment at 8.5 at the following stage were the best suited for the formation of highly ordered and stable mesoporous material. The precise pH management has brought about not only the formation of a highly ordered hexagonal mesophase but also a large reduction of synthesis time from 2 weeks to 3 hours. We

Original method Kanemite/C,,TMA

=0.47

pH=8.5 65'C x 2 weeks

b KSW-I

Modified method

Fig. 1 Comparison of synthesis conditions between original and modified methods.

employed higher kanemite/surfactant ratio of the mixture solution, which resulted in an increase in the initial pH of the solution. TEM images of FSM-16 showed clearly a hexagonal array of uniform channels (Fig.2). The XRD pattern also showed well-defined peaks indexed as two-dimensional hexagonal (p6mm) symmetry (Fig.3). The TEM image and the XRD pattern of FSM-16 are quite similar to those of MCM-41. However, the scanning electron micrograph indicated that the morphology of FSM-16 is quite different from that of MCM-41 [6]. The plate-like morphology of the original kanemite was preserved for FSM-16. While MCM-41 showed a hexagonal prism-like morphology [5]. It suggested a completely different formation mechanism of FSM-16 and MCM-41. The silicate sheets of kanemite are folded and cross-linked each other to form a three-dimensional framework (Figd) [8,9]. The interlayer cross-linking occurred by condensation of silanols on the silicate sheets. The increase in amount of occluded cationic surfactant with increasing the pH expanded the

Fig. 2 Transmission electron micrographs of FSM-16. Cross-section views of channels in (a) perpendicular and (b) parallel to the channel direction.

c

4 2

hkl d(m, im 177 n*43Snm 'lo UK] 1.88 ?lo

"'

interlayers of kanemite fully and consequently formed the regular hexagonal structure. The subsequent pH adjustment at 8.5 accelerated the cross-linking of the interlayers and stabilized the

framework three-dimensional structure. The dissolution of silicate from kanemite was 2 4 6 8 l o suggested at the initial high pH 20(CuKa) condition. However, the degree of Fig.3 X-ray powder diffraction pattern dissolution was limited because of FSM-I 6. the morphology of kanemite was preserved for FSM-16. Later Kuroda et al. clearly demonstrated the "folded sheet" formation mechanism by direct TEM observation of waved silicate sheets with no cross-linking in an intermediate material [lo]. The porous and the pore wall structures of FSM-16 were characterized by XRD [11,12], physisorption [12, 131, infrared spectroscopy [14], TEM [15], Z 9 ~MAS i NMR, modeling and simulations [12]. Kanemite (NaHSi20,.3H,0)

CnH,n+,N(CH,),CL Fig.4 Folded sheet formation mechanism of PSM-16

1.4. Mesoporous aluminosilicates from a layered silicate The introduction of acid properties to mesoporous materials was the most important subject for the application of mesoporous material to catalysis. The new solid-acid catalyst having larger pore size than that of zeolite has been of great interest, because of their availability for many useful reactions such as cracking of heavy oil and synthesis of pharmaceutical chemicals. We first reported the synthesis of mesoporous aluminosilicates and their acid properties at the 9'h International Zeolite Conference held in Montreal on July, 1992 [3].

Fig.5 Acid amount of mesoporous aluminosilicatcs prepared by (a) impregnation with AICI, solution and (b) conversion from layered aluminosilicate. Acid amounts were determined by MI3TPD method.

Fig.6 Infrared spectra of mesoporow aluminosilicate (15.8 wt?A AI,03) prepared by impregnation method (a) afier and (b) bcfore treatment with pyridine vapor.

We impregnated the as-synthesized mesoporous silica derived from kanemite with aluminum chloride solution and calcined the dry sample at 700 "C to remove the surfactant and incorporate in the silica framework [3]. The calcined material showed a large amount of acidity of 0.5-1.1 mmollg depends on the contents of aluminum (Fig.5). 60% of aluminum in the mesoporous material had tetrahedral 81 Si coordination which was confirmed by "AI MAS NMR, indicating the isomorphous substitution of A I ~ 'for si4'in the framework. Infrared spectra of adsorbed pyridine molecules on the mesoporous aluminosilicate showed that it had both Lewis and Br4nsted '" -lZu acid sites and the result was PP=' similar to the usual Fig.7 27Al and 29Si MAS NMR spectra of mesoporous aluminosilicate prepared from a m o ~ h O u s material (Fig.6). The kanemite containing Al.

aluminosilicate materials had a uniform pore size distribution centered on 3 nm and high surface areas of 600 m21g. The XRD pattern showed an intense peak with d-spacing of 3.8 nm, indicating an ordered mesoporous structure. Later we also reported the preparation of layered silicate materials containing ~ l 'in the silicate sheets and the conversion to mesoporous alurninosilicates by folding sheets method [16]. The mesoporous materials derived from layered aluminosilicates had a high level incorporation of ~ 1 in~ ' + the mesoporous the framework with SiIAI ratio of 7.2-1 88. Almost all of A I ~ in material is located in tetrahedral sites (Fig.7). Moreover, the mesoporous aluminosilicates showed highly ordered mesostructures with 2D-hexagonal symmetry. The successive formation of mesoporous aluminosilicates with high regularity and high aluminum incorporation is attributed to using a layered silicate containing A1 as a starting material. The topochemical synthesis method from layered materials is expected to apply for synthesizing various inorganic materials. 1.5. Applications The application studies of FSM-16 have been carried out in a various fields such as catalysis, adsorption and inclusion chemistry. In some studies, FSM-16 has shown better properties than MCM-41 although they had a similar framework structure. Yoshida et al. reported photocatalysis of siliceous FSM-16 for propene metathesis reaction [I 7, 181. Siliceous FSM- 16 exhibited higher catalytic activity than siliceous MCM-41 and amorphous silica. Active sites on the siliceous materials are the strained siloxane bridges generated by dehydroxylation of isolated hydroxy groups. The silica pore walls of FSM-16 are thermally stable and rigid because they retain the local ordered structure of the original crystalline kanemite. The pore walls of FSM- 16 with higher stability than MCM-41 would suppress the conversion of strained siloxane bridges into inactive unstrained bridges. Itoh et al. also reported the photocatalysis of siliceous FSM-16 such as oxidative photodecarboxylation of a-hydroxycarboxylic acids and photo-oxidation of arylmethyl bromide [19, 201. FSM-16 showed the highest activity in various solids including MCM-41, HMS, H-Y, Na-Y, H-ZSM-5, Si02 and AI2O3. Yamamoto et al. reported the acid property and catalysis of siliceous FSM-16 and discussed the differences between FSM- 16 and amorphous silica [2 1,221. FSM-16 is also an excellent support for enzyme stabilization [23,24]. Horseradish peroxidase (HRP) was easily adsorbed in the channels of FSM-16

by simple immersion method. The immobilized enzyme in FSM-I6 exhibited continuous enzymatic activity in an organic toluene, while naked enzyme lost the activity immediately. The Vessel enzyme in FSM-16 is also themally stable in Fig. 8 Two stage pulp bleaching system using aqueous solution at high Manganese peroxidase (MnP) immobilized in temperature of 70 "C. The FSM-16. pore size of FSM-16 affected the stabilized effects critically. FSM-16 with 5 nm in diameter showed the highest stability, because the pore size was the best fitting of HRP with molecular size of 4.3 x 6.4 nm. Similar stabilized effects were observed for MCM-41 but not for SBA-15 prepared using non-ionic surfactant. It suggested that not only size fitting but also interaction between pore surface and enzyme is important for stabilization of enzyme. We have applied the enzyme-immobilized FSM-16 to chlorine-free pulp bleaching system (Fig.8). The Manganese peroxidase (MnP) immobilized FSM-16 converts ~ n to~~ ' n by~ using ' H202and the produced ~ n bleaches ~ + pulp. The FSM-MnP showed excellent stability against Hz02 and continuous activity, while naked MnP deactivated by H~02. Stabilization of chlorophylls in the channels of FSM-16 was also observed [25, 261. Natural chlorophylls extracted form living leaves are unstable and discolored immediately. FSM-16 with larger pore size over 2.3 nm adsorbed large amount of chlorophyll (15.8-29.2 wt%) in the channels. The photostability of chlorophyll was enhanced largely by encapsulated in the channels of FSM-16. The large 2.7 nm shift of absorption band of chlorophyll molecules in FSM-16 suggests a strong Fig.9 Special arrangement of chlorophyll interactions between moleculcs in the channels of FSM-16 with chlorophyll-chlorophyll and pore diameter of 2.7nm. chlorophyll-FSM-16 (Fig.9).

Similar interactions are observed in a living plant leaf. Moreover, the chlorophyll-FSM-16 conjugate generated hydrogen from water in the presence of methyl viologen, 1-lysine, poly(vinylpyrrolidone), sodium carbonate, 2-mercaptoethanol and platinum under irradiation with Xe lamp [26]. The chlorophyll-FSM-16 composite has high potential for the application to artificial photosynthesis. We have studied the adsorption properties of water vapor for FSM-16 [27, 281, and the template synthesis of metal nanowires and dots in the channels of FSM-16 [29, 301.

1.6. Conclusion The mesoporous material FSM-16 discovered in this study was the first materials possessing a well-defined pore whose size was larger than that in usual zeolites and it has been used in various fields. FSM-16 is a suitable material for fbndamental studies in catalysis, adsorption, and host-gust chemistry. Especially, the material has brought about a progress in the understanding of the unique properties of nanospaces. FSM-16 is also expected to be applied to actual uses such as catalysts and nano-containers for bio-molecules because of the high surface area, adsorption capacity and thermal stability, besides the larger pore dimension with regular arrangement. The success in the synthesis of FSM-16 implies novel concepts to produce inorganic materials using surfactant self-assembly as a template and folding sheets of layered silicates. The synthesis method using templating with surfactant aggregations enables us to make materials with tailored mesoporous structure. A variation of the family of mesoporous materials has already been extended by applying this template concept. The folding sheet mechanism would offer possibilities of a rational structure design and an efficient route of making a desired material under moderate conditions. 2. MESOPOROUS ORGANOSILICAS 2.1. Introduction Since the discovery of ordered mesoporous silicates M41S and FSM-16, a variety of ordered mesoporous materials have been synthesized by a template method, using supramolecular assembly of surfactant molecules. These materials have a range of framework compositions, morphologies, and pore structures. The framework composition has been studied extensively, since that

Organic group (R'OhSidSi(OR3)

+

6; ?

Extraction of Surfactants

Sur&ctant 1 H,O

walls

Fig. 10 Synthesis of mesoporous organosilica from bridged organosilaneprecursor.

governs catalysis and adsorption properties. Mesoporous materials now include a variety of inorganic materials, e.g., non-Si transition-metal oxides [3 11, metals 1321, and carbon [33]. Recently, functionalization with organic groups of ordered inorganic mesoporous [34] and microporous [35] materials has attracted much attention because new catalytic and adsorption functions can be introduced onto the internal pore surfaces through the direct design of organic functional groups. These organic-tknctionalized mesoporous materials have a heterogeneous structure composed of an inorganic main kamework with an organic layer grafted onto the framework. Generally, they exhibit poorer structural ordering than nonfunctionalized inorganic mesoporous materials, evidenced by less-distinct X-ray diffraction patterns. On the hand, many kinds of amorphous inorganic oxides, containing organic groups in their framework have been derived by the sol-gel polymerization method [36, 371. Although these amorphous materials have a homogeneous distribution of organic groups and inorganic oxide in the framework, they have disordered structures and scattered pore-size distribution. Here we report the synthesis of novel mesoporous organosilica materials with a homogeneous distribution of organic fragments and inorganic oxide within the kamework rather than end-grafted, exhibiting a highly ordered structure of uniform pores, which are quite different from the conventional organic-functonalized ordered mesoporous materials and sol-gel derived porous hybrid organic-inorganic materials. The new type of mesoporous material was designated as organic-silica Hybrid Mesoporous Material (HMM). Some of HMMs have novel crystal-like pore walls that exhibit structural periodicity with 7.6-1 1.7 A along the channel direction.

Fig.11 Variation of produced rnesophaes of ethane-silicas,

2.2. Mesoporous ethane-silicas HMMs are synthesized from organosilane precursors in which two silicon alkoxides are attached at both sides of the organic group [(R'O),Si-R-Si(OR'),I as shown in Fig.10. In 1999 we were the first to report the synthesis of mesoporous ethane-silicas (Et-HMMs) in which ethane groups (-CH2CH2-)were uniformly distributed within the pore walls from 100% organosilane (CH~O),Si-CHzCH2-Si(OCH& [38]. Two types of mesophases with 2D- and 3D-hexagonal symmetry were obtained by controlling the synthesis conditions. Later we reported the synthesis of Et-HMM with cubic Pm-3n mesophase [39]. The Et-HMMs showed highly ordered reflections in XRD patterns, Fig.12 Transmission electron micrograph indicating higher than usual degrees of of mesoporous ethane-silica with 2D- structural order (Fig.11). Figure 12 hexagonal symmetry.

(a) 2D-hexagonal

(b) 30-hexagonal

(3) Cublc Pm-3n

Fig.13 Scanning electron micrographs of mesoporous ethane-silicas

shows an enlarged TEM image and an electron diffraction pattern of Et-HMM with 2D-hexagonal symmetry. A structure with pores in an ordered arrangement is shown. High-order spots with hexagonal symmetry can be seen in the electron diffraction pattern, indicating that the pore arrangement is highly ordered. The high degree of structural order is also confirmed by the distinct particle morphology of each of the mesoporous materials (Fig.13). The 2D-hexagonal, 3D-hexagonal, and cubic structures formed external morphologies of hexagonal rod, spherical, and decaoctahedral particles, respectively. Each of these have ideal morphologies that reflect the symmetry of the particle interior structure, demonstrating that highly ordered crystals with very few structural defects were being created. Especially, Et-HMM with cubic Pm-3n is composed of uniform particles, both in size (5pm in diameter) and shape (Fig.14) [39, 401. Recently, the synthesis of single crystalline-like mesoporous silica particles has been reported but this requires precise control of the synthesis conditions, and we believe that Et-HMMs can be synthesized more easily. This is because it is easier to eliminate distortions in the lattice and to produce crystals with few defects if highly flexible organic substances are introduced into the structure instead of just an inorganic substance. Fig. 14 Scanning electron micrograph of To date, mesoporous materials with mesoporous ethane-silica with cubic Pm-3n a variety of structures have been symmetry.

successfully synthesized using various surfactants in inorganic systems. There is a clear relationship between the molecular shape of the surfactant and the mesophase that is produced, and this relationship is expressed by the packing parameter of the surfactant [41]. The packing parameter (g=V/aol) regulates the packing geometry of the surfactant molecule in the micelle according to the and the hydrophilic group area (ao), hydrophobic group volume hydrophobic group length (I). A surfactant molecule with a small g parameter readily forms 3D-hexagonal and cubic Pm-3n mesophases comprised of spherical micelles with a relative high degree of curvature. Conversely, a large g parameter results in the formation of lamellar, cubic Ia-3d, and 2D-hexagonal mesophases comprised of lamellar or rod-shaped micelles with relatively little ATEA] [42, 431, in which curvature. Alkyltriethylammonium [CnH2n+IN(C2H5)3: the hydrophilic group (ao) is large, produces cubic Pm-3n (SBA-1) and 3D-hexagonal (SBA-2) mesophases. On the other hand, using ATMA), which has a relatively alkyltrimethylammonium (CnH2n+lN(CH3)3: small hydrophilic component, produces lamellar (MCM-50), cubic Ia-3d (MCM-48), and 2D-hexagonal (MCM-4 1) mesophases. However, in this study, when an HMM containing an ethylene group was used, cubic Pm-3n and 3D-hexagonal mesophases were produced in addition to the 2D-hexagonal mesophase, despite the fact that ATMA was used as the surfactant. This was the first time that cubic Pm-3n and 3D-hexagonal mesophases were produced in a system in which ATMA was used. The creation of mesophases not normally produced is attributed in this case to

(v,

Table 1 The relationship between geometry of precursor molecules and produced mesophase. Geometry of precursor molecules

?Me MeO-Si-OMe I OMe

?Me Me? MeO-S,i-CH,CH,-S,i-OMe Me0 OMe

Produced mesophases

Effective headgroup area (a,) of ATMA

Oo=small g * =large

Lamellar Cubic Ia-3d 2D-hexagonal

g =smal

Cubic Pm-3n 3D-hexagonal

the specific molecular structure of the precursor, (CH30)3Si-CH2CH2-Si(OCH3)3 (BTME). The anion charge density on organosilane species formed by hydrolysis and oligomerization of BTME is lower than that on the silicate species formed from monosilane precursor such as tetramethoxysilane, Si(OCH3)4, because the silyl groups are separated by ethane groups in BTME derivatives species. The low charge density enlarges the effective headgroup area of ATMA counteractions in the BTME-ATMA assembly in accordance with the charge-density matching at the interface, which induces the 3D-hexagonal and cubic Pm-3n mesophases. This result suggests that not only the geometry of the surfactant molecule but also the geometry of precursor molecule can control the structure of the inorganic-surfactant mesophase. 2.3. Extension of mesoporous organosilica system After our publication, mesoporous organosilicas with a variety of bridged organic groups inside the pore walls have been reported (Scheme 1). Stein et al. reported the synthesis of mesoporous materials containing ethane and ethylene groups (-CH=CH-) [44] They confirmed that organic groups were exposed on the pore surface by showing that the ethylene groups could be readily brominized. These mesoporous materials had disordered phases with wormlike pores. Next, Ozin et al. reported the synthesis of mesoporous materials from a mixed system of ethylene-bridged Si-CH2-Si Si-CH2CH2-Si Si-CH=CH-Si dialkoxysilane and TEOS [45]. Structural order increased with s i increasing TEOS mixing ratio. Later, Si Ozin et al. also reported the synthesis of mesoporous materials containing phenylene (-C6H4-) [46], CH3 OCH, thiophene (-C4H2S-) r461, and . - s i methylene (-CHI-) [47] groups. si \ / \ / They also attempted to synthesize Si the mesoporous materials (-C=C-), containing acetylene ferrocene (-C5H4-C5H4-), and H2 dithiophene (-C4H2S-C4H2S-) groups [46], but they were unable to Scheme 1. Bridged organosilanes used for the synthesis of ordered mesoporous materials. obtain ordered mesoporous Alkoxy groups (OCH,, 0C2H,) attached at Si materials. Brinker et al. were the were abbreviated.

Gsi

psi

first to report the synthesis of organic bridged mesoporous thin films [48]. They fabricated a homogeneous transparent film from a sol-gel solution of alkoxysilane containing ethylene groups using spin- and dip-coating techniques. The dielectric constant of the thin film was extremely low (1.98 to 2.15), and these materials demonstrated superior properties as low-k materials for semiconductors. Several groups reported the synthesis of ethane- and benzene-bridged mesoporous materials with larger pore sizes and thicker pore walls using nonionic triblock coplymers [49-541. We have also reported the synthesis of mesoporous phenylene- [55, 561 and biphenylene-silicas [57]. The mesoporous aromatic-silicas hybrid materials showed novel molecular-scale periodicity within the pore walls as shown later. Recently, Ozin et al. reported the mesoporous organosilica containing interconnected [Si(CH2)I3rings [58]. Bifunctionalized mesoporous materials were synthesized by co-condensation of bridging and terminal (or TEOS) precursors. The co-condensation approach resulted in the synthesis of various mesoporous materials containing both of bridging organic moieties inside the walls and terminal groups protruding into the channel space [59-621. The bifunctional mesoporous materials have unique structure in which bridging organics play a structural and mechanical role while the terminal groups are readily accessible for chemical transformation. Alvaro et al. reported a mesoporous material containing viologen units in the framework by co-condensation with TEOS [63]. The pore walls of the mesoporous material should show unique optoelectrical properties because viologenes are the most widely used electron acceptor units in a variety of charge transfer complexes and electron transfer processes. The bridged mesoporous materials containing cyclic m i n e complexes in the framework have been also synthesized by co-condensation with TEOS [64-661. The reports also exist on the synthesis and catalysis of organic-bridged mesoporous materials incorporating A1 or Ti in the framework [67-701. By combining these previous synthesis approaches it is possible to design unique mesoporous catalyst containing hydrophobic and hydrophilic sites, acid sites and organic functional sites. 2.4. Mesoporous aromatic-silicas with crystalline pore walls Almost all of mesoporous materials synthesized previously have amorphous pore walls, which restrict their application to limited uses because of amorphous-like surface properties and low stability. Although many efforts have been made to crystallize the pore wall of mesoporous material, the crystallization was limited in only small part of pore walls [30, 7 1, 721. Recently,

Sclf-assembly of sulfactant

Crystal-like pore walls

(~to)~.i~~i(o~t),

+

69

NaOH

Extraction o f Surfactants

Surfactant I H,O

-

Sell-assembly ofprecursor

(E~o),s~-s~(oE~)~),

""

Fig. 15 Syntheu~sof ordered mesoporous organosilica with crystal-like periodicity inside the walls by double self-assembly o f precursor and surfactant molecules.

we found the formation of atomic-scale periodicity in the pore walls in a whole region of mesoporous material containing phenylene groups inside the pore walls (Ph-HMM) [ 5 5 ] . It is the first synthesis of mesoporous materials possessing a crystal-like pore-wall structure. The novel mesoporous material 1,4-bis(triethoxysilyl)benzene [BTEB, was synthesized by adding (C2H50)3Si-C&-Si(OC2H5)3] to the mixture of octadecyltrimethylamrnonium (ODTMA), sodium hydroxide (NaOH), and water (Fig.15). White precipitate was recovered by filtration. Surfactant was removed by solvent extraction using HCI/EtOH solution. Self-assembly of the BTEB precursor molecules formed the periodic structure in the walls of mesoporous phenylene-silica because hydrophobic and hydrophilic interaction directed the self-assembly of BTEB

2H /degree (CuKa)

Fig. 16 X-ray powder diffraction pattern ofmesoporous phenylene-silica

Fig. 17 Transmission electron micrograph and electron diffraction of mesoporous phneylene-silica.

molecules. Ph-HMM had a highly ordered mesoporous structure with two-dimensional hexagonal symmetry (a=52.5 A), which was confirmed by well-defined XRD pattem in low angles (28<10) (Fig.16) and a clear hexagonal arrangement of i I3cNMR spectra clearly uniform pores observed in TEM images. 2 9 ~and showed that Ph-HMM had a hybrid organic-inorganic framework structure unit. The XRD pattem also consisting of covalently bonded 01.5Si-c6H4-Si01.5 showed several sharp peaks in a wide angle region (28=10-50 degree) at d=7.60, 3.80, 2.53 and 1.95 A, in addition of the low angle reflections (Fig.16). These difiaction peaks are due to periodic structure with a d-spacing of 7.6 A and it's higher index reflections. In the TEM image taken in the perpendicular to the channels axis, many lattice fringes corresponding to the 7.6 A periodicity were observed in the whole region of the pore walls of Ph-HMM (Fig.17). The lattice planes were perpendicular to the one-dimensional channel axis and stacked along the channels, which was also confirmed by electron difiaction (Fig.17). The periodicity is attributed to regular arrangement of 01.5Si-C6H4-SiOL.S unit in the pore walls due to the interaction among phenylene groups. The interaction among benzene rings was observed by the fluorescence spectrum of the Ph-HMM material. The structural model of Ph-HMM was constructed on the basis of the above experimental results and the well-defined structure of crystalline 1,4-bis(trihydroxysily1)benzene. Fig.18 shows the structural models showing hexagonal arrangement of channels and pore surface structure of Ph-HMM. Benzene rings are aligned in a circle around the pore, fixed at the

both sides by silicate chains. The silicate is terminated by silanol (Si-OH) at the surface. Hydrophobic benzene layers and hydrophilic silicate layers array alternatively at an interval of 7.6 A along the channel direction.

Plate like narticle

Aggregation of fibrous crystals

Fibrous crystal

Fibrous crystal

Fig. 19 HierarohicaUy ordered shuclures of mesopomus phenylene-silica

The periodically arranged hydrophobic-hydrophilic surface has a great advantage for using this material as catalyst and host material for inclusion chemistry because it could enable structural orientation of guest molecules or clusters enclosed in the pores. Ph-HMM has unique hierarchically ordered structures (Fig.19). Plate-like particles of Ph-HMM are composed of aggregation of fibrous crystals of mesoporous material with a diameter of 100 nm and a length of l p n . The fibrous crystals are composed many channels aligned in the direction of fiber. The pore walls of channels are composed of regularly arranged phenylene-silica units. The unique crystalline mesoporous material system has been extended to the other mesoporous organosilicas containing aromatic groups such as biphenylene [56] and 1,3 phenylene [57]. They also showed similar lamellar structure with periodicity of 11.9 and 7.6 A, respectively, within the pore walls, in addition of periodic mesostmctures. 2.5. Highly functionalized mesoporous organosilicas We have also attempted to design a highly functionalized acid-catalyst by introducing sulfonic groups (-S0,H) onto the periodic pore wall surface of mesoporous aromatic-silicas [55, 73, 741. Sulfuric acid (H2S04) is one of the most frequently used acid-catalyst for chemical processes. However sulfuric acid has a drawback because it is a toxic liquid and as a liquid catalyst it is very difficult to be separated from products and reused. Therefore, the fixation of the sulfuric acid to a solid has been one of the most important subjects to establish eco-friendly chemical processes. We have successfully developed two kinds of sulfuric-acid functionalized mesoporous benzene-silicas. The first is the mesoporous benzene-silica with sulfonic groups directly attached on ne in the walls [551. The Fig. 20 CG image of pore surrace of ~ h e n ~ l e groups sulfuric acid-functionalized mesopomus sulfonation was carried out on the benzene-silica . About 10% of phenylene mesoporous benzene+ilica using groups in the walls are sulfonated.

fuming sulfuric acid at 105-110 "C. In spite of the severe condition for the treatment, the mesoscopically ordered structure and the molecular-scale periodicity of the mesoporous material was preserved after the treatment. The content of sulfonic groups in the mesoporous material was 0.4 meq./g, which was estimated by the titration with sodium hydroxide solution. This indicates that approx. 10% of phenylene groups in the walls of mesoporous benzene-silica were attached with sulfonic groups (Fig. 20). The Fig.21 Structural image of mesoporous material should show strong bifunctional mesoporous benzeneacid catalysis due to the molecular structure silica attached with propylsulfonic of sulfonic groups directly bonded to acid groups on the silicate layers of phenylene and the periodic structure of the ordered pore surface. walls. The sulfonic groups were kept with pore walls at the high temperature under 480 OC. The excellent thermal stability allows us to apply this material as catalyst for both liquid and gas phase reactions. The other material we have developed is the mesoporous benzene-silica with propylsulfonic groups (-C3H6-S03H)attached on the hydrophilic silicate layers (Fig. 21) [73]. The material was synthesized in two steps; first the synthesis of mesoporous benzene-silica with mercaptopropyl groups (-C3Hs-SH) at the silicate layers, and followed by the oxidative transformation of thiol groups (-SH) to sulfonic groups. The mercaptopropyl-functionalized mesoporous benzene-silica was synthesized by co-condensation of BTEB and 3-mercaptopropyltrimethoxysilane [(CH30)3Si-C3H6-SH, MPTMS] showed similar molecular-scale periodicity (7.6 A) as observed for the mesoporous benzene-silica. Incorporation of mercaptopropyl groups in the framework could be confirmed by 2 9 ~and i NMR and IR measurements. The results clearly indicated that the mercaptopropyl groups are attached on the silicate layers o f mesoporous benzene-silica. The maximum contents of mercaptopropyl groups in the mesoporous materials was 1.68 mmol/g. Subsequent oxidation treatment with HN03 resulted in the formation of sulfonic groups. The conversion of oxidation of -SH to -S03H was 41.7%, indicating the maximum acid amount of 0.70 mmollg. Thus, in the material the catalytic acid sites and hydrophobic

benzene sites are separately designed and apart from each other on the mesoporous surface. Indeed, such dimensionally designed surface structure is a better catalytic environment. The mesoporous biphenylene-silica with propylsulfonic aicd was also prepared [74].

2.6. Conclusion The organic-bridged mesoporous material is a very interesting system because (i) chemical, electrical and optical functionalities could be integrated within the framework of mesoporous materials preserving highly ordered mesostructures, (ii) hierarchically ordered mesoporous material with crystal-like pore-walls can be synthesized by incorporating interactive bridging organic groups inside the pore walls, and (iii) organic groups within the pore walls can be modified by attaching with functional groups and further chemical transformation. 3. Acknowledgements The author acknowledges Prof. K. Kuroda, Prof. 0 . Terasaki, Dr. Y. Fukushima, Dr. S. Guan, Prof. T. Ohsuna, Dr. M. P. Kapoor, Prof. Q. Yang, Prof. A. Fukuoka, Prof. H. Yoshida, Prof. A. Itoh,, Dr. H. Takahashi, Dr. Y. Goto, Dr. A. Bhaumik, Dr. Y. Sakamoto, Dr. S. Yamamoto, Ms. Y. Yamada, Mr. N. Suzuki and Mr. H. Kadoura for their contributions and helpful discussions. The author would like to thank Toyota Central R&D Labs., Inc. for the permission to publish these works. This is partly supported by CREST, Japan Science and Technology Agency.

REFERENCES 1. T. Yanagisawa, T. Shimizu. K. Kuroda, C. Kato, Abstract of 56th National Meeting ofthe Chemical Society of Japan, Tokyo, No. 1XIID42 (1988) 761. 2. T. Yanagisawa, T. Shimizu, K. Kuroda, C. Kato, Bull. Chem. Soc. Jpn., 63 (1990) 988. 3. S. Inagaki, Y. Fukushima, A. Okada, T. Kurauchi, K. Kuroda, C. Kato, Proc. 9thInternational Zeolite Conf., Montreal, I (1992) 305. 4. C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli, J. S. Beck, Nature, 359,7 lO(1992) 5. J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge, K. D. Schmitt, C. T-W. Chu, D. H. Olson, E. W. Sheppard, S. B. McCullen, J. B. Higgins, J. L. Schlenker, J. Am. Chem. Soc., 114 (1992) 10834.

6. S. Inagaki, Y. Fukushima, K. Kuroda, J. Chem. Soc., Chem. Commun., (1993) 680. 7. S. Inagaki, Y. Fukushima, A. Okada, C. Kato, K. Kuroda, Japanese Patent No. H4-238810(1991), US Patent No. 5,508,08 l(1996). 8. S. Inagaki, Y. Fukushima, K. Kuroda, Stud. Surf. Sci. Catal., 92 (1994) 143. 9. S. Inagaki, A. Koiwai, N. Suzuki, Y. Fukushima, K. Kuorda, Bull. Chem. Soc. Jpn., 69 (1996) 1449. 10.T. Kimura, T. Kamata, M. Fuziwara, Y. Takano, M. Kaneda, Y. Sakamoto, 0 . Terasaki, Y. Sugahara, K. Kuroda, Angew. Chem. Int. Ed., 39 (2000) 3855. 11.Y. Fukushima, S. Inagaki, J. Mater. Sic. Eng., A2 171218 (1996) I 16. 12.S. Inagaki, Y. Sakamoto, Y. Fukushima, 0. Terasaki, Chem. Mater., 8 (1996) 2089. 13.P. J. Branton, K. Kaneko, N. Setoyama, K. S. W. Sing, S. Inagaki, Y. Fukushima, Langmuir, 12 (1996) 599. 14.T. Ishikawa, M. Matsuda, A. Yasukawa, K. Kanda, S. Inagaki, Y. Fukushima, S. Kondo, J. Chem. Soc., Faraday Trans., 92 (1996) 1985. 15.Y. Sakamoto, S. Inagaki, T. Ohsuna, N. Ohnishi, Y. Fukushima, Y. Nozue, 0. Terasaki, Microporous Mesoporous Mater., 21 (1998) 589. 16.S. Inagaki, Y. Yamada, Y. Fukushima, Stud. Surf. Sci. Catal., 105 (1996) 109. 17.Y. Inaki, H. Yoshida, K. Kimura, S. Inagaki, Y. Fukushima, T. Hattori, Phys. Chem. Chem. Phys., 2 (2000) 5293. 18.Y. Inaki, H. Yoshida, T. Hattori, J. Phys. Chem. B, 104 (2000) 10304. 19.A. Itoh, T. Kodama, S. Inagaki, Y, Masaki, Org. Lett., 2(2000) 331. 20.A. Itoh, T. Kodarna, S. Inagaki, Y. Masaki, Org. Lett., 2 (2000) 2455. 21.T. Yamamoto, T. Tanaka, T. Funabiki, and S. Yoshida J. Phys. Chem. B, 102 (1998) 6450. 22.T. Yamamoto, T. Tanaka, S. Inagaki, T. Funabiki, and S. Yoshida J. Phys. Chem. B, 103 (1999) 6450. 23.H. Takahashi, B. Li, T. Sasaki, C. Miyazaki, T. Kajino, S. Inagaki, Chem. Mater., 12 (2000) 330 1. 24.H. Takahashi, B. Li, T. Sasaki, C. Miyazaki, T. Kajino, S. Inagaki, Microporous Mesoporous Mater., 44-45 (200 1) 755. 25.T. Itoh, K. Yano, Y. Inada, Y. Fukushima, J. Mater. Chem., 12 (2002) 3275. 26.T. Itoh, K. Yano, Y. Inada, Y. Fukushima, J. Am. Chem. Soc., 124 (2002) 13437. 27.S. Inagaki, Y. Fukushima, K. Kuroda, J. Colloid Interface Sci., 180 (1996) 623.

28. S. Inagaki, Y. Fukushima, Microporous Mesoporous Mater., 2 1 (1998) 667. 29.T. Yamamoto, T. Shido, S. Inagaki, Y. Fukushima, M. Ichikawa, J. Am. Chem. SOC.,118 (1996) 5810. 30.M. Sasaki, M. Osada, N. Sugimoto, S. Inagaki, Y. Fukushima, A. Fukuoka, M. Ichikawa, Microporous Mesoporous Mater., 21 (1998) 597. 31.P. Yang, D. Zhao, D. I. Margolese, B. F. Chmelka, G. D. Stucky, Nature, 396 (1998) 152. 32.G. S. Attard et al., Science, 278 (1997) 838. 33.S. H. Joo, S J. Choi, I. Oh, J. Kwak, Z. Liu, 0. Terasaki, R. Ryoo, Nature, 412 (2001) 169. 34.S. L. Burkett, S. D. Sim, S. Mann, Chem. Commun., (1996) 1367. 35.C. W. Jones, K. Tsuji, M. E. Davis, Nature, 393 (1998) 52. 36.D. A. Loy, K. J. Shea, Chem. Rev., 95 (1995) 1431. 37.R. J. P. Corriu, Angew. Chem. Int. Ed. 39 (2000) 1376. 38.S. Inagaki, S. Gum, Y. Fukushima, T. Ohsuna, 0 . Terasaki, J. Am. Chem. SOC.,121 (1999) 96 11. 39.S. Guan, S. Inagaki, T. Ohsuna, 0. Terasaki, J. Am. Chem. Soc., 122 (2000) 5660. 40.S. Guan, S. Inagaki, T. Ohsuna, 0. Terasaki, Microporous Mesoporous Mater. 44-45 (2001) 165. 41.4. Huo, D. I. Margolese, G. D. Stucky, Chem. Mater., 8 (1996) 1147. 42.4. Huo, D. I. Margolese, U. Ciesla, P. Feng, T. E. Gier, P. Sieger, P. M. Petroff, F. Schuth, G. D. Stucky, Nature, 368 (1994) 317. 43.4. Huo, D. I. Margolese, U. Ciesla, D. G Demuth, P. Feng, T. E. Gier, P. Sieger, A. Firouzi, B. F. Chmelka, F. Schuth, G D. Stucky, Chem. Mater., 6 (1994) 1176. 44.B. J. Melde, B. T. Holland, C. F. Blanford, A. Stein, Chem. Mater., 11 (1999) 3302. 45.T. Aefa, M. J. MacLachlan, N. Coombs, G. A. Ozin, Nature, 402 (1999) 867. 46.C. Y. Ishii, T. Asefa, N. Coombs, M. J. MacLachlan, G. A. Ozin, Chem. Commun., (1999) 2539. 47.T. Asefa, M. J. MacLachlan, H. Grondey, N. Coombs, G. A. Ozin, Angew. Chem. Int. Ed., 39 (2000) 1808. 48. LaVan, C. J. Brinker, J. Am. Chem. Soc., 122 (2000) 5258. 49. E.-B. Cho, K.-W. Kwon, K. Char, Chem. Mater., 13 (2001) 3837. 50.H. Zhu, D. J. Jones, J. Zajac, J. Roziere, R. Dutartre, Chem. Commun., (2001) 2568.

5 1.0. Muth, C. Schellbach, M. Froba, Chem. Commun., (2001) 2032. 52.M. C. Burleigh, M. A. Markowita, E. M. Wong, J-S. Lin, B. P. Gaber, Chem. Mater., 13 (200 1) 44 11. 53.J. R. Matos, M. Kruk, L. 0. Mercuri, M. Jaroniec, T. Asefa, N. Coombs, G. A, Ozin, T. Kamiyama, 0. Terasaki, Chem. Mater., 14 (2002) 1903. 54.Y. Goto, S. Inagaki, Chem. Commun., (2002) 2410. 55.S. Inagaki, S. Guan, T. Ohsuan, 0. Terasaki, Nature, 416 (2002) 304. 56.M. P. Kapoor, Q. Yang, S. Inagaki, J. Am. Chem. Soc., 124 (2002) 15176. 57.M. P. Kapoor, Q. Yang, S. Inagaki, Chem. Mater. in press. 58.K. Landskron, B. D. Hatton, D. D. Perovic, G. A. Ozin, Science, 302 (2003) 266. 59.T. Asefa, M. Kruk, M. J. MacLachlan, N. Coombs, H. Grondey, M. Jaroniec, G. A. Ozin, J. Am. Chem. Soc., 123 (2001) 8520. 60.M. C. Burleigh, M. A. Markowitz, M. S. Spector, B. P. Gaber, J. Phys. Chem. B, 105 (2001) 9935. 61.M. C. Burleigh, M. A. Markowitz, M. S. Spector, B. P. Gaber, Chem. Mater., 13 (2001) 4760. 62.M. C. Burleigh, M. A. Markowitz, M. S. Spector, B. P. Gaber, Langmuir, 17 (2001) 7923. 63.M. Alvaro, B. Ferrer, V. Fornes, H. Garcia, Chem. Commun., (2001) 2546. 64.R. J. P. Corriu, A. Mehdi, C. Reye, C. Thieuleux, Chem. Commun., (2002) 1382. 65.R. J. P. Corriu, A. Mehdi, C. Reye, C. Thieuleux, Chem. Commun., (2003) 1564. 66.K. Z. Hossain, L. Mercier, Adv. Mater., 14 (2002) 1053. 67.K. Yamamoto, Y. Nohara, T. Tatsumi, Chem. Lett., (2001) 648. 68.M.P.Kapooq A. Bhaumik, S. Inagaki, K. Kuraoka, T. Yazawa, J. Mater. Chem.,l2 (2002) 3078. 69.M. P. Kapoor, A. K. Sinha, S. Seelan, S. Inagaki, S. Tsubota, H. Yoshida, M. Haruta, Chem. Commun., (2002) 2902. 70.A. Bhaumik, M. P. Kapoor, S. Inagaki, Chem. Commun., (2003) 470. 7 1.B. Lee, D. Lu, J. N. Kondo, K. Domen, Chem. Commun., (200 1) 2 118. 72.Y. Liu, W. Zhang, T. J. Pinnavaia, J. Am. Chem. Soc., 122 (2000) 8791. 73.Q. Yang, M. Kapoor, S. Inagaki, J. Am. Chem. Soc., 124 (2002) 9694. 74.M. P. Kapoor, Q. Yang, Y. Goto, S. Inagaki, Chem. Lett., 32 (2003) 914.

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Integrating Interfaces and Function with Molecular Assembly G. D. Stucky" and J. Herbert waiteb "Department of Chemistry & Biochemistry and Program in Biomolecular Science and Engineering, University of California, Santa Barbara, California 93 106 USA b~epartmentof Molecular Cell and Developmental Biology, University of California, Santa Barbara, California 93 106 USA 1. INTRODUCTION

The 3-D organized molecular assembly of components, which are both hnctionally and compositionally distinct, into an integrated system is a particularly intriguing challenge. In addition to the requirement of the multicompositional, hierarchical assembly of spatially distinct domains, there must be information exchange between the domains that can be selectively coupled with the external environment. Living systems provide a sophisticated model for this that is currently well beyond the best bench-top efforts. In biogenesis, the components of the organized, integrated system are created via non-linear parallel, multivalent syntheses and processing. The bioassembly is typically directed by entropy change and interface interactions in confined spaces. The spaceltime assembly definition of structure and function makes possible the closely coupled organization and processing of integrated organic and inorganic domains. It is also utilized to create, among other things, gradient interface structures that enable coupling of the components of the organism internally or with the external environment. This presentation describes selected biogenic model systems and the related use of kinetic control, competing processes, non-equilibria phenomena, multiphase media, entropy and organictinorganic interfaces to synthesize composite materials that have patterned 3-D structural and coupled hnctional properties.

2. INTERFACES

The creation and functional effectiveness of an integrated composite system is defined by the interfaces between the spatially defined components, and their efficacy with respect to information transfer, whether it be electro-optic, mechanical, or chemical. The system will only be useful if there is also coupling of the system to the external environment. The discussion that follows will consider first the organization of internal interfaces by molecular assembly between spatially distinct domains (atoms or nanoparticles) with different compositions that enable energy transfer and the coupling of local electronic transitions (nanoscale dimensionality) with photonic confinement (micron length scale). This will be followed by examples of how a soft biological system can be effectively coupled to its environment, even though there is an exceptional mismatch between the mechanical properties and chemical potentials. In the final part of the presentation, a confining environment is used to control the molecular assembly entropically by limiting the number of allowed configurations during the assembly (- In Q), and energetically by increasing the contribution of the system exterior surface interactions relative to the molecular interior interactions.

2.1. Interfaces and the Multicompositional, Hierarchical Assembly of Functional Units An obvious way to introduce coupled multiple subsystem functionalities is by the integrated, but spatially distinct, organization of domains with different composition. Block copolymers provide a particularly useful and powerful generic synthesis platform for the cooperative organization of inorganic and organic moieties into domains on nano- to macroscopic length scales with organized micro- through macro-phase separation [ 1-51. The distinct chemical potentials of the phase-separated domains can be used as reaction media for the molecular assembly or organization of functional units on sub-length scales within a given domain. Thus one can carry out higher levels of domain definition by creating sub-domains with different functionality and composition organized within initially created phase-separated domains. In this way inorganic nanoparticles with different composition and properties can be coassembled; and at the same time, organics with very different chemical or optical functionality can be integrated in an organized way with the multicompositional inorganics using a "single pot" configuration that combines both synthesis and processing of the composite system into a desired macroscale topological configuration. This is a good first approximation to the assembly that nature uses for living cells and biogenesis, since it permits the chemist to utilize on the bench-top the spatialltemporal kinetics and thermodynamics that are associated with a complex reactant mixture to selectively control competing

interactions, entropic frustration, and topology on the preferred multiple length scales. It should be noted and emphasized that this approach makes use of organic-inorganic, inorganic-inorganic and organic-organic interfaces. For example, PbS nanoparticles, which have an absorption spectrum profile that closely matches that of the solar spectrum and a conduction band at higher energy than that of anatase (TiOz), can be assembled and integrated in situ with anatase nanocrystals, which are assembled in parallel with the PbS nanocrystals, into a highly ordered 3-D mesostructured, flat thin film of macroscale dimensionality. Both nanocrystals are formed and homogeneously co-organized into a highly ordered 3-D mesostructured array during a continuous process [671. Another example is the use of block polypeptides to create multicomponent, nanoparticle hollow spheres in -90 seconds without any oillwater emulsion or preformed micro-template (e.g., polystyrene spheres). The nanoparticles can be metallic or magnetic, or can be wide-band-gap or narrow-band-gap semiconductors, or combinations of these. For example, capped quantum dots, with electron confinement, can be organized into hollow micro-spheres that are capped by silica nanoparticles. The photons emitted by the quantum dots can be amplified to give monomodal microcavity laser emission, thus coupling two different length scale functionalities -- the electronic quantum confinement and the photonic confinement of the microcavity. The spheres can be used as "Trojan horses", with separate tagging of, for example, 1) the individual gold nanoparticles, which collectively make up an interior spherical shell array, with up to even 100 kD proteins, and 2) the silica nanoparticles on the outside of the spheres so that the spheres can be targeted to specific cellular sites or as receptors for specific antigens. Disassembly and reassembly of the spheres to release the tagged gold particles can be achieved in a variety of ways, for example by pH or electrolyte changes.

2.2. Interfaces and Gradient Structures When two spatially distinct components, each with its own mechanical and chemical properties, are brought into intimate contact, their critical properties are challenged. What has been termed contact deformation occurs. On the atomic scale, one consequence can be described as electronegativity equalization. On the macroscopic scale, the major mechanical difficulty that arises from the apposition of a stiff, hard structure with a softer hydrogel is the disruption of the soft hydrogel [8]. A traditional way of dealing with mismatches such as this is to prime the hard surface with a polymer or some other interface agent that is semi-compatible with both of the components. This does not decrease the interfacial radial stress, but may increase the limit for the stress-to-failure ratio. We have found that nature uses a more sophisticated approach, namely the use of functional gradients. If the interface effectively

either incrementally or gradually transforms the hard structure into a soft material, the interfacial stress can be dissipated over a much larger area. Two biomolecular structures, mussel byssal threads [9-101, which interface between a rock and the living soft tissue of the mussel, and polychaete worm jaws [ll-121, resolve this problem by the judicious use of gradients and metals. This energy transfer interface hnctionally couples the organism with its environment. Probably the best bench-top analog of achieving gradient structures and functionality is "layer by layer" assembly [13-141. Less tedious and more efficient approaches are being developed.

2.3. Confining Mesoscale Periodic Order An important feature of biological molecular assembly processes is their use of spatially confined environments, in which entropy and well-engineered interfacial molecular interactions are used to create the desired sophisticated structural and hnctional system. This has been a strong driving force for our studies. We have developed a strategy based on confined molecular assembly that enables control of the structural phase generated using a given block copolymer, silica source, and set of reaction conditions. This is achieved by nanoscale tuning of the confinement parameters that determine the entropy loss and interface interactions. Structural phases obtained from a single reaction type by this method have included the selective isolation of a single cage wire array, double cage wire array, single helix, double helix, single helix with interior straight channel, triple helix, and concentric donut array. The mesostructures are easily backfilled with Group VIII metals, giving, for example, nanowire springs. Studies of the mechanical and electro-optical properties of these phases as well as permeation properties of membranes made from parallel arrays of these structures are in progress. 3. FUTURE DIRECTIONS The ability to use molecular assembly to synthesize 3-D material systems with built-in multi-component functionalities and substructures has the potential of offering an exceptionally rich generic platform for device fabrication. A waterbased, green chemistry approach that is coupled with combining synthesis and processing, ideally using a "single pot" approach, also suggests that it might be environmentally fiiendly and cost-effective. An important spin-off of this approach is that it encourages the "make every atom count" philosophy in material design and fabrication. The synthesis parameter space is rich and challenging to explore, as it is based on non-equilibrium, non-linear parallel molecular assemblies that utilize time and space to couple the multiple domain kinetics and interaction energies. Nevertheless, the feasibility and utility of complex, multivalent, competing chemistries that do just this are demonstrated

in the biology and biochemistry that are all around us. It is clear that we will succeed only by being in tune with what nature has to offer, and following her instruction.

REFERENCES [I] [2] [3] [4] 153 [6] [7] [8] [9] [lo] [ll] [12] [13] [14]

D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Fredrickson, B. F. Chmelka, and G. D. Stucky, Science, 279 (1998) 548. D. Zhao, Q. Huo, J. Feng, B. F. Chmelka, and G. D. Stucky, J. Am. Chem. Soc., 120 (1998) 6024. P. Yang, G. Wirnsberger, S. R. Cordero, H. C. Huang, B. Scott, M. D. McGehee, T. Deng, G. M. Whitesides, B. F. Chmelka, S. K. Buratto, and G. D. Stucky, Science, 287 (2000) 465. P. Yang, T. Deng, D. Zhao, P. Feng, D. Pine, B. F. Chmelka, G. M. Whitesides, and G. D. Stucky, Science, 282 (1998) 2244. P. Yang, D. Zhao, D. I. Margolese, B. F. Chmelka, and G. D. Stucky, Nature, 396 (1998) 152. M. H. Bartl, J. Tang, and G. D. Stucky, Angew. Chemie (2004), in press. J. N. Cha, M. H. Bartl, M. S. Wong, A. Popitsch, T. J. Deming, and G. D. Stucky, Nano Letters, 3 (2003) 907. S. Suresh, Science, 292 (2001) 2447. J. H. Waite, Results Problems Cell Differentiation, 19 (1992) 27. J.. H. Waite, X. Qin, a nd K. J. Coyne, Matrix Biology, 17 (1998) 93. H. Lichtenegger, T. Schoberl, M. H. Bartl, J. H. Waite, and G. D. Stucky, Science, 298 (2002) 389. H. C. Lichtenegger, T. Schoberl, J. T. Rukolainen, J. 0. Cross, S. M. Heald, H. Birkedal, J. H. Waite, and G. D. Stucky, Proc. Natl Acad. Sci. USA, 100 (2003) 9144. G. Decher and J. D. Hong, Berichte der Bunsen Gesellschaft f. Physik. Chem., 95 (1991) 1430. T. Mallouk, Acc. Chem. Res., 25 (1992) 420.

This Page Intentionally Left Blank

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Designer synthesis of mesoporous solids via block copolymer templating pathway D.-Y. Zhao*, B.-Z. Tian and X.-Y. Liu Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Department of Chemistry, Fudan University, Shanghai 200433, P. R. China. Tel: 86-2 1-65642036; Fax: 86-2 1-6564-1740; Email: [email protected] 1. INTRODUCTION The delicate construction of materials with tailored textures and functionalities is regarded as a key step toward new technologies. During the past decade, research on mesostructured materials has sparked great contributions, obviously due to their potentially valuable applications in areas, such as catalysis, sensors, semiconductors, and photonic and electronic microdevices. A challenging and profitable task for material scientists is: how to approach this 'lofty goal of smart materials by design'? Amphiphilic block copolymers, with at least two distinct moieties and regarded as macromolecular analogs of molecular weight surfactants, are the focus of a great wealth of research in contemporary materials and macromolecular science. Self-assembled macro- and meso-structures of block copolymers provide a tremendous tool for controlling the spatial formation of various nanomaterials [l-41. The study of amphiphilic block polymer templated mesostructured materials is an emerging research area offering enormous scientific and technological promise. Compared with ionic surfactants, block copolymers have become more and more popular and powerful in the synthesis of mesoporous solids because of their diverse structural characteristics, rich phase behaviors, low cost, nontoxic degradation and many other merits. Different synthetic methodologies have been developed to fabricate periodic mesoporous silica materials in the past few years. By carefully manipulating the processing variables such as temperature, pH, ionic strength, reaction time and solution composition, ordered mesoporous solids have been obtained from different nonionic amphiphilic block copolymers with variable structures and adjustable physical properties. Herein, a brief overview is provided of recent

developments in the use of block copolymer self-assembly to fabricate mesoporous solids with various compositions and topologies, with some important and interesting work highlighted. 2. SYNTHETIC PATHWAYS AND RELATED MECHANISMS 2.1. Aqueous media Up to now, most efficient syntheses of ordered mesoporous silica structures were carried out in acidic media. This is related to the assembly characteristics of PEO [poly(ethylene oxide)] type nonionic amphiphilic block copolymers. The assembly of the inorganic and organic periodic composite materials appears to take place by a hydrogen bonding (SOH') XI' pathway. Besides, it should be noted that some researchers have proposed several strategies to prepare ordered periodic mesoporous silica structures under neutral or alkaline conditions. An example is the mesoporous silica prepared by Voegtlin et al. under near-neutral conditions employing nonionic oligomeric surfactants in the presence of fluoride ions [5]. In a wide range of pH conditions (pH = 0 - 9), Stucky and coworkers have developed a one-step synthesis of ordered hexagonal silica-surfactant mesostructured composites by using nonionic amphiphilic block copolymers with fluorides by controlling the rates of hydrolysis and condensation of tetramethoxysilane (TMOS) as a silica source [6].Pinnavaia and coworkers also prepared ordered hexagonal MSU-H by using sodium silicate as a silica source and triblock copolymer Pluronic PI23 [poly(ethylene oxide)-block-poly (propylene oxide)-block-poly(ethy1ene oxide) as a structure-directing agent (PEO-PPO-PEO) copolymers] (E020P070E02~) (SDA) in the near-neutral media [7, 81. The basic understanding of the acidic assembly mechanism of block copolymer templated mesoporous silica structures may help us achieve more rational syntheses. Goldfarb and coworkers investigated the formation mechanism of mesoporous silica SBA-15 by using in situ X-band electron paramagnetic resonance (EPR) spectroscopy in combination with electron spinecho envelope modulation (ESEEM) experiments [9]. They observed a continuous depletion of water within the corona-core interface during the mesophase assembly. Before high temperature hydrothermal treatment, the majority of the PEO chains are located in the micropores. Moreover, they found that the extent of the PEO chains located within the silica micropores depended on the hydrothermal ageing temperature and on the Sip123 molar ratios. Employing time-resolved in situ 'H-NMR (nuclear magnetic resonance) and transmission electron microscopy (TEM), the dynamics of the formation of a mesoporous silica SBA-15 is recently investigated by Flodstrom et al. [lo] They observed four stages during the cooperative assembly: (i) silicates

adsorption on globulur micelles; (ii) association of the globular micelles into floes, (iii)precipitation offloes, and (iv) micelle-micelle coalescence. Based on detailed investigation of the formation of mesoporous crystals templated by nonionic surfactants, Yu et al. very recently proposed a Colloidal Phase Separation Mechanism (CPSM) [I I] (Figure 1). According to this concept, the meso-lmacro- topological evolution includcs the following three stages. (i) cooperative assembly, which takes place at the molecular level to form surfactantisilica composite aggregates. Charge density matching between the surfactant head-groups and hydrolyzed inorganic oligomers is important at this stage. The nanosized composites serve as the building blocks for later hierarchical assembly; (ii) colloidal-like interaction, further coalescence and condensation of the nano-building blocks give a new liquid crystal-like phase made up of the block copolymerlsilica hybrid species. As the silica species further condense, the new liquid phase grows denser with time and finally separates fiom the water phase. Moreover, such "liquid phase" with small diameters should grow larger in order to minimize the surface area and therefore the surface free energy (F). This newly formed liquid phase can he regarded as the precursor of the final highly ordered liquid crystal phase (socalled liquid crystal-like phase). (iii) rnultiphase energy competition. At this stage the mesostructure assembly is still under way, and the separated liquid crystal-like phase is further growing into the final solid mesostructure. Although the free energy of the mesophase formation (AG) is responsible for the final mesostructure, the competition between AG and the surface ftee energy (F) of this liquid crystal-like phase determines the morphology of final mesoporous materials. The general evolution process described in this mechanism that are much based on the observations of "in situ" TEM and X-ray diffractions (XRD) is somehow in agreement with that of Flodstrom [lo], but the critical analysis of F and AG of CPSM may shed light on several important issues of the mesostructured solids (e. g. morphology, additive effects).

Collnidal-likc interaction P11aw ~ ~ p i ~ t ~ : k l i o # ~

~llilcro-\IrucI~~rc*

Fig. 1. Schematic diagram of Colloidal Phase Separation Mechanism. Reprinted with permission from Ref. 36.

From the comparison of the critical micelle concentration (CMC) values of several block copolymers and the directed silica mesostructures [12, 131, it has been proposed that the CMC of block copolymers correlates to the final silica mesostructures and can be used as an important criterion for the rational design of ordered mesoporous materials especially when they are synthesized under the aqueous media. 2.2. Nonaqueous (or mixed with a little water) media Non-aqueous synthesis is a very convenient method for the preparation of ordered mesoporous materials, and this method has become more and more powerful and profitable. Most of the syntheses conducted in the nonaqueous media adopt the well-known evaporation induced self-assembly (EISA) process [141. The versatile EISA strategy was first used by Brinker and coworkers [14], in the preparation of mesoporous silica thin films. Stucky and co-workers further developed this method to synthesize large pore mesoporous silica and non-siliceous mesoporous materials [15-201. For preparation of mesostructured silica films, tetraethoxysilane (TEOS) dissolved in the organic solvent (normally ethanol, tetrahydrofuran, acetonitrile) was pre-hydrolysized with stoichiometic quantity of water (catalyzed by acids, such as HCl) at temperature of 25 70°C. Then low-polymerized silicate species could randomly couple together with the block copolymers. Upon solvent evaporation, the silicate species hrther polymerized and condensed, gradually forming the mesostructures via the disorder to order transition. Large pore (up to 14 nm) mesoporous nonsiliceous metal oxides can be similarly prepared in alcoholic media, by using the amphiphilic poly (alkylene oxide) block copolymers as the SDAs. They use unhydrous inorganic salts rather than alkoxides or organic metal complexes as the soluble and hydrolysable inorganic precursors to construct the metal oxide frameworks. This simple and versatile method has generated several large pore mesoporous composites, including Ti02, Zr02, Nb205, Ta205, A1203, SO2, Sn02, W03, HfOz, and mixed oxides SiAlO,, A12TiOy,ZrTiO,, SiTiO, and ZrWO,. Those materials are relatively thermally stable probably due to their much thick inorganic walls. This method is more efficacious for the preparation of high valence metal oxides such as Nb2O5, Ta205,and Zr02 than that for low valence metal oxides such as A1203and metal oxides whose precursors behave poorly from the sol-gel point of view. In the later case, only disordered mesoporous oxides or hybrid mesostructures can be obtained. The most important feature, is the semicrystalline framework where nano-crystallites nucleated within the amorphous inorganic mattress. This virtue is particularly much desired in high performance electronic and optical devices that fabricated with mesostructured metal oxide films (coatings).

-

Detailed and complete studies on EISA-based syntheses of inorganic species-block copolymer hybrid composite films (using PEO-PPO-PEO block copolymers, or PEO-alkyl surfactants as templates) have been carefully investigated by researchers in Sanchez's group [21-331. Mesoporous transition metal (Y-Zr, Ti, Zr, V, Al, Fe, etc.) oxide-based hybrid thin films have been prepared reproducibly with variable mesostructures. Many characterizations, such as Synchrotron-SAXS (Small Angle X-ray Scattering), WAXS (Wide Angle X-ray Scattering), interferometry, ultraviolet visible spectroscopy (Uvvis), NMR, EXAFS (Extended X-ray Absorption Fine Structure)/XANES (Xray Absorption Near Edge Structure) etc., were employed to monitor the earlier stages of film formation during the dip-coating process. Generally, the selfassembly leading to the organized mesopbase took place at the final stage of the drying process, which was after the formation of a disordered intermediate phase. The role of various synthesis parameters, such as surfactant, water and acid concentrations, processing humidity and evaporation temperatures, were thoroughly discussed to understand their influence on the biphasic assembly of ordered hybrid mesostructures. Moreover, they also employed some stepwise treatments [26, 271 to stabilize the films formed, thus making them more thermally stable upon calcination. Reproducible preparation and rational design of mesoporous materials, as well as better understanding of the EISA mechanism become possible, thanks to their critical and informative discussions.

Fig. 2. Schematic diagram of "acid-base pairs" concept and grouping principle. Reprinted with permission from Ref 36.

Recently, we proposed a new "acid-base pairs" concept [35, 361 to address the interplay of inorganic-inorganic species for the preparation of homogeneous non-siliceous mesoporous materials, especially with multi-components, via EISA. An "acid" mineral precursor was designed to couple with a "base" counterpart, forming the "acid-base pairs" (Figure 2). These pairs not only generate a proper acidic media for the self-assembly and the gelation of inorganic species, but are also crucial for the homogeneous mineral compositions within the whole frameworks. The proper acidic media was produced by tuning the ratio of "acid" to "base" precursors (e.g. for the preparation of TiPO, the starting precursors can be TiC14,Ti(OC3H7)4and PC13, and the acidity can well be tuned by changing the TiCWTi(OC3H7)4ratio). For assembling ordered mesostructures, generally the "acid-base pairs" formed from strong "acid" and strong "base", or strong "acid (base)" and medium "base (acid)" are required, which pave the formation of homogeneous multicomponent inorganic precursors. Various characterizations, such as liquid state and solid state NMR and UV-vis, conformed the pair formation within the mother solutions and as-prepared samples. Guided by this concept, we reported the successful syntheses of a wide variety of well ordered, large pore, homogeneous, and multi-component mesostructured solids, including metal phosphates (TiPO, AIPO, NbPO, ZrPO, CePO, LiTi2(P04)3, etc.) silicoaluminophosphates (SAPO, MeSAPO, Me = V, W, Sn, In, Mn, Fe, Co, Ni, etc.), mesostructured metal borates (AIBO, TiBO, etc.), as well as various metal oxides (Ti02, Zr02, etc.) and mixed metal oxides (ZrW20,, CeTi30x) mesostructured phases. Some composites such as TiPO and AlPO are stable up to 800°C. Moreover, such materials can be easily processed as thin films, monoliths and multi-scale ordered coatings. Besides EISA, the direct precipitation of mesostructured materials from nonaqueous media have been demonstrated by Tilley and coworkers [37, 381. The obtained inorganic frameworks are molecular homogeneous, but the lack of long range mesostructural ordering was reflected by the relatively poorly resolved XRD patterns and disordered patterns in TEM images. Unlike most of the synthesis routes reported, this route relies on the non-hydrolytic condensation of mixed alkoxides in non-polar media. 2.3. Supercritical fluid media Very recently, Pai et al. [39] have prepared well-ordered mesoporous solid films by infusion and selective condensation of silicon alkoxides within microphase-separated block copolymer templates dilated with supercritical carbon dioxide. By using a supercritical fluid (SCF) as the process medium, the principal challenge that to achieve efficient transport and reaction within the polymer films while preserve the organic template order has been well

overcome. Three-dimensional (3-D) and high-fidelity replication of the block copolymer morphology was obtained via the confinement of metal oxide deposition to continuous hydrophilic moieties of the pre-organized micelles. It should be highlighted that this strategy may offer the possibility to prepare 2-D mesoporous films with cylindrical channels oriented normal to the substrate surface. The mesoporous silica films are reported to have excellent mechanical properties and can survive the chemical-mechanical polishing step required for device manufacturing. Not only these structural features are of interest to microchip designers, but the synthesis rates are fast enough for use in a microelectronics fabrication facility. 2.4. Comparisons of aqueous and nonaqueous synthesis Compared with "hydrothermal" synthesis, non-aqueous EISA offers more convenient and faster preparation. As to the preparation of non-silica materials, it is better to conduct the syntheses under non-aqueous conditions, concerning the fact that the interactions between the nonionic block copolymers and the inorganic species are relatively weaker than that of ionic surfactants and that most metal alkoxides readily hydrolyze when attacked by water. EISA can generally and efficiently fabricate a large family of non-siliceous mesostructured materials. EISA do not require stringent selection of structure directing agents. For or F98 (E0123P047E0123) are very difficult to example, F 108 (E0132P050E0132) direct the formation of mesostructures under aqueous condition, while are much more arbitrary when via EISA. The yield of EISA prepared ordered mesoporous materials is generally very high, while much lower in some cases for hydrothermal synthesis. A typical example is the preparation of SBA-16 (templated by F108), normally only 20% yield can be reached under aqueous media, which is clearly caused by the striking difference of SiISDA ratios in starting reactants and as-precipitated (without hydrothermal treatment) products [l 11 and the narrow starting synthetic range. On the contrary, 3-D cubic F108lsilica hybrids (space group, Im-3m) can be easily and reproducibly prepared with at least 90% mesophase purity using EISA method. Moreover, EISA is a powerful strategy for preparing mesoporous films, monoliths and spheres, etc. Finally, there are several structural features of mesoporous materials prepared via EISA versus hydrothermal precipitation. a) Framework distortion The samples made by EISA normally require certain substrate for controlled deposition, which imposes a strain field, generating a uniaxial lattice distortion [22-27, 29, 31, 32, 401. This distortion can be qualitatively and quantitatively

-

analyzed and shown to have resulted in lowering the mesostructure symmetry. b) Microporosity Mesoporous materials prepared by EISA generally have much lower microporosity (without hydrothermal treatment) [40], which may due to the less inclusion of EO segments into the inorganic framework. The nonaqueous solvents may screen the charge coupling or other interplay between the inorganic species and the hydrophilic corona. This fact may also lead to dense inorganic framework. c) Flexibility Solid state 29 Si-NMR measurements [40] show that the as-made EISA silica membranes (templated by P123) has a (Q2+ Q3/Q4)ratio of 2.43 (the ratio of (H0)2-Si= and HO-Si= species to SO4), almost double that of fi-eshly precipitated SBA-15 powder prepared under aqueous media, which indicates a more flexible (lower cross-linkage) backbone of hybrid materials prepared via EISA. This good flexibility, combined with high content and great mobility of block copolymer templates makes possible the phase transformation from 2-D hexagonal mesostructure (space group p6mm) to 3-D bicontinuous cubic mesostructure (Ia-3d) upon solvothermal treatment [40]. Therein [40], the employment of non-polar solvents (such as hexane) instead of water as the heating media played a decisive role in the formation of gyroid cubic mesophase. Different from water that can greatly facilitate the hydrolysis and condensation of siliceous species and improve the regularity of 2-D hexagonal mesostrutures, non-polar solvents is more inert as to the reaction with inorganic siliceous species, therefore it may basically transfer energy to trigger the mesophase transition (from high to low mesophase curvature) at the early stage when the framework of siliceous hybrids are quite flexible. d) Others EISA hybrid materials usually contain more organic templates; Mixed mesophases co-exist especially for thick membranes or monoliths; It's relatively difficult to tune the pore wall thickness and pore size of mesoporous silica materials via EISA method. 2.5. Special crystallization techniques It has long been realized that fully crystallized porous materials are much more useful than amorphous or semi-crystallized ones. For many applications, such as photocatalysis, the percentage of nanocrystallites makes a strike difference. One has to address this problem probably by choosing more proper inorganic precursors or employing a different thermal treatment process.

Employing a so-called two-step 'delayed rapid crystallization' thermal treatment, mesoporous Ti02 optical thin films with fully nanocrystalline anatase framework that are ideal for energy transfer applications can be successfully prepared [27]. In-situ investigations performed during such treatments indicated that these Ti02 thin films retained most of their meso-skeletons up to 700°C. The anatase framework can be obtained via the rapid but homogeneous crystallization of the low temperature stabilized amorphous matrix. This emphasizes the role of the treatment method to stabilize transition metal oxide mesoporous materials over extended crystallization at high temperatures. Supported by embedded carbon rods, Katou et al. have prepared highly crystalline niobium-tantalum mixed oxides mesoprous solids [41]. It is interesting to note that some domains of the mesoporous solids exhibit single crystal features, which have never been observed for nonsiliceous mesoporous materials previously. This method may be applied to crystallize other mesoporous oxides, but their crystallization temperature may be within a proper range (550 750°C). Zhou and coworkers very recently demonstrated a synthetic methodology to create ordered mesoporous composites with crystalline oxides/amorphous phosphates framework [42], which are reminiscent of the 'brick and mortar' structure. They built a group of nano-architectures by means of surfactanttemplated self-assembly followed by the controlled in-situ crystallization. Functional nanocrystals are used as the building blocks of ordered mesopores, and the amorphous glass phase can act both as the 'glue' between nanocrystals and as a functionalized component in the composites. Ordered mesoporous ceramics with fully crystallized Ti02 and ZrOz 'bricks' have been prepared.

-

3. MESOSTRUCTURES 3.1. Siliceous mesoporous materials templated by block copolymers Mesoporous materials can be divided into ordered and disordered structures from the viewpoint of the porelchannel packing regularity. Typical ordered mesostructured silicas synthesized by using amphiphilic block copolymers are classified by their meso-scale topologies in Table 1. a) Ordered mesoporous siliceous solids C I 2EOs Employing nonionic oligomeric surfactants [C12H25(CH2CH20)80H] and Cl6EO8,Attard and coworkers have successfully prepared ordered mesoporous silica structures under acidic condition [43]. The resulted ordered materials exhibit pore sizes up to 3.0 nrn. In late 1997, in an acidic and non-aqueous medium, Wiesner and co-workers [44] have done excellent illuminating work of the preparation of large pore mesoporous

ceramics employing poly(isoprene-block-ethylene oxide) (PI-b-PEO) block copolymers as the SDAs. These works suggested that the use of higher molecular weight block copolymer mesophases instead of conventional low molecular weight surfactants provides a simple, easily controlled pathway for the preparation of various silica-type mesostmctures that extends the accessible length scale of these structures by about an order of magnitude. [45-481. The successful syntheses of a new family of mesoporous silica materials (such as SBA-15) were carried out under the acidic aqueous media by using nonionic alkyl poly(ethy1ene oxide) oligomeric surfactants, star diblock copolymers or triblock copolymers as the templates. A large number of highly ordered mesoporous silica materials with various mesopore packing symmetries and well defined pore connectivities have been prepared [49, 501. Under a mild acidic condition, Su and coworkers have obtained a wellordered mesoporous silica CMI-1,-2 and -4 fiom C16E010and ClsEOloin a relatively wide range of surfactant concentrations [5 1-54]. El-Safty et al. have done detailed and informative work on the fabrication of a family of nanostructured silica monoliths (HOM-n) with remarkable 2-D and 3-D structures and controllable pore size [55, 56, 571. They employed microemulsion lyotropic liquid crystal mesophase of Brij 56 ( C I ~ E O Ias~ )templates, and with the addition of alkanes with different molecular size (C6-CI9alkyl chains) in some cases. The mesostmctures of the well-defined HOM-n include body-centered cubic Im-3m (HOM-I), 2-D hexagonal p6mm (HOM-2), 3-D hexagonal P6Jmmc (HOM-3), primitive-centered cubic Pm-3m (HOM-4), Pn3m (HOM-7), Pm-3n (HOM-9) symmetries, bicontinuous cubic Ia-3d (HOM-5), lamellar L a (HOM-6) and face-centered cubic Fm-3m (HOM-10). This work is reminiscent of the previously reports on large pore mesoporous silica monoliths templated by triblock copolymers [58, 591, therein, Peng et al. also employed the microemulsion (ternary surfactant (Pluronic F127 or P123)-cosurfactant (butanol, pentanol, or hexano1)-water systems) templating strategy [58, 591. b) Disordered mesostructures Except ordered mesoporous silica materials, the syntheses of disordered sponge-like or wormhole-like mesoporous silica materials templated by nonionic amphiphilic block copolymers have also contributed a lot to the exploiting of organization principles in inorganic-surfactant reaction systems. Pinnavaia and coworkers have brought forward a neutral nonionic surfactant templating pathway to synthesize mesoporous materials. Disordered structures, designated as MSU-X [60], with uniform pore diameters in the range of 2.0-5.8 nm have been obtained by using PEO surfactants. Su and coworkers have also prepared disordered wormhole mesostmctures (DWM) analogous [53, 611 to

MSU type materials with the surfactant weight percentage of 50% by using ClsEOlo,CI3EOn(n = 6, 12, 18) or C16E010as templates. Table 1 List of typical ordered siliceous mesoporous silicate materials templated by nonionic block copolymers Space group

Fm-3m

Researchers or materials FDU- 1 FDU- 12 HOM-10 SBA- 16 Yu et.al.

Im-3m

ST-SBA-16 Wiesner and coworkers

Pm-3m

SBA- 11 HOM-4

Pn-3m

HOM-7

Pm-3n

HOM-9 SBA- 12

Pbdmrnc HOM-3

p6mm

Attard and coworkers SBA- 15 Wiesner and coworkers. Peng et a/.

Block copolymers

Remarks

E039P047E039 large caged E0106P07~E0106 ultra-large caged ClhEOln with additive alkanes TMOS is preferable E0106P07oE0106 than TEOS highly ordered E0132P05oE0132 crystals ClsEOloo small pore, think wall "the plumber's PI-b-PEO nightmare" with lower surfactant Ci6E01o concentration Ci6E01o ClnE01n with additive alkanes with higher surfactant C16Eoio concentration or with additive alkanes Ci6E0io with additive alkanes mixed hcp and ccp CixE01o vhases with medium Ci6E0io surfactant concentration CIZEOS,C16EOs E020P070E020

PI-b-PEO

Reference [a611 [641 I571 [491 ~651 [661 [45-481 [57, 55, 561 [@I 164, 551 157, 55, 561 [571 [@I [57, 55, 561

small pore highly ordered large pore materials organically modified aluminosilicate with organic additives

~49,501 [441 [58,591

HOM-2 Cho et al. MSU-H Chan et al. FDU-5

la-3d

Flodstrom et al. KIT-6 Schuth and coworkers Che et a1 HOM-5 FDU-5

L,

Wiesner and coworkers

with medium surfactant concentration PEO-PLGA-PEO organosilica E ~ z o P ~ ~ o E ~ z neutral o pH silicon-containing P(PMDSS)-DG triblock copolymer first acidi EOzoPO7oEOzo with MF c16E010

E020P070E020 with inorganic salts E020P070E020 with additive butanol with TEVS and E020P070E020 inorganic salts E0zoP07oE0zo with MPTS with medium CI~EOIO surfactant concentration post-solvothermal E0zoP07oE020 synthesis with higher surfactant CI~EOIO concentration organically modified PI-b-PEO aluminosilicate

[57,55, 561 ~671 [7,81 [681

1711 ~721 [731 [741 [57,55, 561 [401 [551 [441

Combined with previous literatures and experimental data in our group, the table shown above can help to inform us several general synthetic trends and designer ideas: (1) Block copolymers with high hydrophilic/hydrophobic ratios (such as F 108 (EOi32P050E0132), F98 (E0123P047E0123), F 127 (E0106P070E0106)and Brij700 (ClsEOloo))can be used to prepare caged mesoporous materials, whose topological curvatures are rather high. While block copolymers with medium hydrophilic/hydrophobic ratios (e.g. P 123 (E020P070E020), B50- 1500 (BO10EO16,BO = buthylene oxide) favor the formation of mesostructures with medium curvature (e.g. 2-D hexagonal or 3-D bicontinuous cubic). Namely, there is a rough structural correlation between the block copolymers and the obtained mesophase materials [49]. (2) For a specific block copolymer and within a certain synthetic range, normally the higher concentration leads to mesostructures with lower mesophase curvature. This is especially suited for the rational synthesis of

mesoporous solids prepared via the EISA and lyotropic liquid-crystal templating pathway [15, 751. (3) Additives have a strong influence on the assembly and packing symmetries of the organiclinorganic mesophase hybrids. For example, large pore mesoporous silica with double gyroidal structure (space group, Ia-34 can only be prepared in the presence of some additives (organic or inorganic) when triblock copolymer P123 were used as the template in acidic aqueous media [69, 721. The effects of various additives will be detailed in the following section. (4) Since the EO moieties of block copolymers are readily protonized in acidic media, the higher the [H'] is, generally the more hydrophilic the block copolymers are. This basic idea provides us a way to tune slightly the inorganiclorganic biphasic interplay in aqueous media [72]. (5) Although the aforementioned materials are siliceous and are mainly synthesized under aqueous conditions, the block copolymer packing rules and mesophase behaviours can also be applied to the preparation of non-siliceous mesoporous materials via EISA. For example, 2D-hexagonal and 3D-caged mesoporous Ti02 (or TiPO, AlPO, Zr02 etc.) can be well prepared templated by triblock copolymer P 123 and F 108, respectively [36, 761. 3.2. Triblock copolymer templated SBA-n and FDU-n series Mesoporous SBA-11 was synthesized over a wide range of reactant compositions at room temperature with a mean pore size of 2.5 nm [49]. SBA12 has a BET surface area of 1150 m21g, a pore volume of 0.83 cm31g and a mean pore size of 3.1 nm. SBA-12 was originally assumed to have a space group of P63/mrnc, later it was revealed by using TEM techniques that the SBA12 specimens usually contained mixed hcp and ccp phases [77]. Highly ordered mesoporous materials SBA- 15 [49, 501 have attracted more and more attention not only because they have high quality structure regularity, thick inorganic walls, excellent thermal and hydrothermal stability (the structure of SBA-15 can be thermally stable up to 1000°C. Calcined SBA-15 retains most of its structures after being heated in boiling water for more than one week), but also because the templates are economically cheap and nontoxic, and the synthesis is quite simple and reproducible. The pore (both primary and complementary pores) size of the materials can be tuned with ease through hydrothermal treatment, namely higher heating temperature achieves larger mesochannel sizes and larger openings within the mesopore walls. More importantly, people have revealed that the wall structure of SBA-15 is quite different from that of MCM-41, although the two materials have a same space group (p6mm). A large number of disordered micropores are distributed within the walls of SBA-15, even mesopores with the diameters between 2 3 nm can be observed. It has been revealed that, as the synthesislaging temperature

-

increased, the total volume of the complementary pores decreased only slightly relative to the primary mesopore volume, but the relative amount of the micropore volume within the complementary pores decreased significantly [78]. At even higher temperatures such as 130°C, the micropore region in the complementary pores is almost eliminated, leaving only relatively larger 8 nm) within the walls of SBA-15 materials [79]. mesopores (3 Complementary porosity of SBA-15 was retained to a significant extent even after calcination at 900°C, but most likely completely disappeared at 1000°C, resulting a wall structure and therefore nitrogen adsorption properties similar to those of MCM-4 1 materials [78]. These special topological feature make SBA15 a very good host to prepare self-supported metal oxides, metal sulfide, carbon even metal nanowire arrays. SBA-16 [49] was reported to have a cubic mesophase with Im-3m space group and a large cell parameter (a) of 16.6 nm. The most efficacious structuredirecting agent is triblock copolymer F127, E0106P070E0106, with long hydrophilic segments. With the aid of inorganic salts (KC1, &So4), highly ordered SBA-16 can be easily prepared. Moreover, TMOS is much preferable over TEOS for the precipitation of SBA- 16 powders. an amphiphilic triblock copolymer with more By using E039B047E039, hydrophobic moieties (BO), mesoporous materials with significantly larger pore size and unit cell dimension has been prepared (donated as FDU-1 [62]). This large caged materials was originally assigned to the space group Im-3m with unit cell parameter a = 22 nrn and the pore size is about 12 nm. This large caged mesoporous material has sparked great deal of research thereafter. Recently, with resort to high-resolution TEM and small-angle X-ray scattering, Jaroniec and coworkers suggested that FDU-1 silica was a face-centered cubic Fm-3m structure with 3-D hexagonal intergrowth. Previous efforts have demonstrated it extremely difficult to synthesize large pore 3-D bicontinuous cubic (Ia-3d) mesoporous materials under acidic conditions. FDU-5 [69] with highly branched channels and double gyroidal framework was prepared in a non-aqueous solution by using P123 as a template and by adding a small amount of organosilicates, for example, mercaptopropyltrimethoxysilane (MPTS), or a fraction of non-polar organic molecules, for example, ethylbenzene, toluene, 1,3,5-trimethylbenzene (TMB), etc., at room temperature under an acidic condition. The resultant materials have a bicontinuous cubic Ia-3d symmetry that is analogous to the structure of MCM48 prepared by using cationic surfactant under a basic condition (Figure 3). However, the pore size (up to 10 nm) of the materials is much larger than that for MCM-48. These large pore mesoporous silica materials FDU-5 have been used as hard templates to prepare mesoporous carbons, metal oxides and metal sulfides materials with bicontinuous cubic structure (Ia-34.

-

Fig. 3. TEM images of FDU-5, along a) [loo] and b) [I 111 incidence. c) TEM image of carbon replica, alone [311] incidence and d) model of the double gyroidal channels. Reprinted with permission from Ref. 40 and 69.

Mesoporous silica structures FDU-12 [64] were synthesized in acidic media by using nonionic block copolymer F127, E O I O ~ P O ~ ~asEaOtemplate, I ~ ~ TEOS as a silica source, and TMB together with inorganic salts such as KCl, as additives. The resulted silica materials present a cubic (Fm-3m) mesostructure with a large cavity size of 10 12.3 nm. Mostly importantly, the size of the entrances can be continuously adjusted in the range of 4 9 nm. It was also revealed that the entrance size of FDU-12 is a key factor for the immobilization of enzymes, and fabrication of cubic mesoporous carbon replicas with ordered large pores by using FDU-12 as the hard templates. Thus it was proposed that the entrance size of mesoporous materials with cage-like structures is a key factor for the applications in which mass transportation and diffusion are necessav.

-

-

4. ADDITIVES

Herein, we will emphasize the influence of two types of additives, organic and inorganic additives, on the mesoscopic and macroscopic topologies of mesoporous materials under aqueous and non-aqueous conditions.

4.1. Organic additives a) Aqueous media Various organic additives, such as TMB, alcohol (e.g. butanol, heptanol, hexanol), alkane (e.g. hexane, heptane, octane, nonane, decane, dodecane, pentadecane, heptadecane, and nonadecane) and N, N-dimethylformamide (DMF), have been added to the acidic cooperative assembly systems. Feng et al. [58, 591 reported a general method for the synthesis of periodic monolithic silica mesostructure by direct liquid crystal templating in multicomponents (TMOS, surfactants and alcohol) systems. Zhao et al. [89] prepared doughnut-like SBA-15 with ordered curve mesochannels, using DMF as the cosolvent. Assisted by the addition of alkanes with variable chain lengths, the lyotropic mesophase of Brij56 (C16E010)has been greatly enriched, leading to the successhl preparation of 3-D mesoporous silica materials (HOM-n) [55-

TMB basically plays two important roles in the acidic synthesis of mesoporous silica structures. (i) enlarge mesopore (primary pore)/micropore (secondary pore) size, window size (of caged mesostructures, e.g. FDU-12 [64]); (ii) increase mesostructure regularity. The latter is conditional when the SDAs are highly hydrophilic, e.g. Brij 700 [66] or F127 [64]. b) Nonaqueous media The most typical example is the synthesis of FDU-5 [69] mesoporous silicas. Bicontinuous cubic FDU-5 was synthesized via EISA by using P123 as a structure directing agent under an acidic condition. The double gyroidal mesophase formation is assisted by the addition of a small amount of mercaptopropyltrimethoxysilane (MPTS), or a fraction of non-polar organic molecules, for example, ethylbenzene, toluene, TMB, etc.. All these additives well altered (decreased) the curvature of inorganic/organic interface, leading to mesophase transition fromp6mm to Ia-3d. c) Interface Normally, the formation of crack-free mesoporous monoliths is timeconsuming, which takes several days or even one month. Zhao and co-workers created a liquid paraffin medium protected solvent evaporation method to fast yield transparent, crack-free, large sized, highly ordered mesoporous monoliths by using amphiphilic block copolymers as the templates [90]. The thin layer of paraffin that covered onto the freshly dried monoliths can significantly reduce the original surface tension of bare monoliths. This technique is expected to provide practical methods to the fabrication of monolithic materials aimed at optic and separation applications.

Fig. 4. SEM image of large pore mesoporous single crystals templated by FlO8. Reprinted with permission from Ref. 11.

4.2. Inorganic additives a) Aqueous media Highly hydrophilic block copolymers (e.g. F127, F108, F98 and Brij 700), though somewhat diff~cultto prepare ordered mesostructured solids practically, are ideal SDAs for the formation of caged mesoporous materials fiom the viewpoint of their intrinsic packing symmetries and mesophase behaviours. This conflict can be well overcome when 'salting out' inorganic salts (such as, KCI, K2S04,Na2S04) are added into the synthetic batches. In some cases, even highly ordered large pore single crystals can be obtained [ l l , 651 (Figure 4). This 'salting out' effect can also be expressed as the ability to lower the CMC and CMT values of block copolymers used, therefore highly ordered mesoporous silica can be prepared with lower block copolymer concentrations or at lower temperatures, when resorting to these inorganic salts [12, 13,651. Viewing the surfactant/silica composite aggregates initially formed in dilute solution fiom a colloidal point of view, it is energetically unfavorable for these colloidal particles to approach each other because of electrostatic repulsion. It may especially be the problem for the highly hydrophilic block copolymer templated hybrid colloidals (the EO segments of these SDAs are highly protonizedhydrated, hence rather difficult to couple with respect to each other). When the ionic strength of the solution is increased, the energy barrier is decreased. Finally, at a certain high ionic strength, the aggregation of colloidal particles becomes energetically favorable. According to Colloidal Phase Separation Mechanism, the evolution of morphologies of mesoporous materials can also be well clarified [ll]. For the synthesis systems of mesoporous silica with the same block copolymers under similar acidic condition, it is proposed that the time interval from addition of TEOS to phase separation (the induction time) may be used approximately to estimate the free energy change of the composite mesophase formation. AG

(free energy of mesophase formation) becomes less negative with increasing time because the condensation of inorganic silica species is progressing with time. When the phase separation occurs early (e.g. in the presence of inorganic salts), AG is dominant in the synthesis system, so that the morphology of the particles is developed together with the formation of ordered mesostructures (crystal-like). The final particle morphology reflects the intrinsic liquid crystal structure. On the other hand, if the phase separation occurs more slowly (e.g. in the presence of some organic additives, such as D m ) , F (surface free energy) will have considerable influence upon the macrostructure because the morphology is developed by surface energy consideration during transformation fiom liquid crystal-like phase to solid phase. With the increasing influence of F, the morphology with lower curvature will be generated in order to minimize the surface energy (e.g., spherical SBA- 15 synthesized at 20°C). On the contrary, some 'salting in' inorganic salts can be employed to increase the mesophase curvature, which has been demonstrated by Flodstrom et al. [71]. They used inorganic additive NaI to tune the mesophase from basically multilamellar vesicles to bicontinuous cubic phase, which situates between the lamellar and the 2-D hexagonal phases. The use of inorganic salts can also significantly facilitate the formation of highly ordered mesoporous organosilicas [91-931, which will not be detailed herein. b) Nonaqueous media Stucky and coworkers have produced a novel mesoporous silica membrane with 3-D sponge-like macrostructures [19]. The synthesis was especially interesting in that the macropore dimensions were established by the sizes of droplets of aqueous electrolytes, such as NaC1, LiC1, KC1, NH4C1, or MgS04. The hierarchical pore structures may be important for mass transportations in catalysis and separation.

5. APPLICATIONS Although much less exploited compared with mesoporous silica materials, non-siliceous mesoporous solids are potentially important for various applications. The preliminary works briefly reviewed below were conducted mainly by our group and might give some implications. Large pore mesoporous niobium oxide (MNO) films as matrix, successhl biocapsulation of cytochrome c (Cyt-c) has been demonstrated [94]. The ordered MNO has good electrochemical properties and can effectively promote the direct electron transfer of redox proteins though the protein redox sites that

are far fiom the surface of electrode. Moreover, the fabricated electrode was employed as a biosensor for the electrocatalysis of hydrogen peroxide, and we found that the adsorbed protein molecules still retain their electrocatalytic activity. Very recently, 3-D cage-like mesostructured W03-Ti02materials have been synthesized via "acid-base pair" strategy by using triblock copolymer Fl08 as a template. The 3-D cage-like mesostructure as well as the large surface area provides more surface active sites and leads to shorter diffusion length of lithium ions, which greatly improves the total electrochromic properties of mesoporous W03-Ti02 thin films. Worm-like mesoporous W03-Ti02films were also synthesized and applied, for the first time, to immobilize photosynthetic reaction center (RC) to construct nanosemiconductor-based Bio- photoelectrode (PE). These tailored mesoporous W03-Ti02 films exhibit unique structural characters, i.e. opened pore channels, proper pore size with narrow distribution and hydrophilic surfaces for specific entrapment of RC with high activity retained. As a consequence, the incident photon to photocurrent conversion efficiency detected at 860 nm for RC Bio-PE based on tailor-made mesoporous W03-Ti02 is about 49.2%, nearly 3 times larger than that based on electrodeposited Ti02 or A1203sol-gel.

6. OUTLOOK Despite the great advances that have been achieved in amphiphilic block copolymer templating systems, some challenges remain to be hrther explored. First of all, new synthesis strategies should be explored in order to obtain ordered, large surface area, and thermally stable mesoporous materials especially with Si (Ge), metal sulfide, metal nitride, metal carbide, organic polymer and other compositions that cannot be synthesized by the current methods. Secondly, PEO type block copolymers are usually utilized in acidic conditions. In order to synthesize mesoporous materials with crystalline frameworks, one strategy may include the use of ionic block copolymers with amine head-groups. It is expected that with such templating systems the synthesis may be carried out under basic conditions. The block copolymers backbone and the amine head groups may be used to induce the formation of mesostructures and crystalline (microporous) frameworks, respectively. Thirdly, mesoporous materials with highly ordered ultra-large pore sizes (50-100 nm) that bridge mesopore and macropore are much desired. This feature is important not only for fundamental research, but for practical applications as well (bioseparation, catalysis, etc.). Finally, the fabrication of high performance microdevices with mesostructured materials still require extensive research

efforts especially the interdisciplinary collaborations, and much more exciting success can be anticipated!

ACKNOWLEDGMENT This work was supported by National Natural Science Foundation of China, Shanghai Nanotech Promotion Center (0212nrn043), State Key Basic Research Program of PRC (001CB510202), 863 Project of China Sci. & Tech Ministry (2002AA321010), Shanghai Sci. & Tech. Committee (03DJ14004).

REFERENCE [I] I. W. Hamley, Nanotechnology, 14 (2003) R39. [2] M. Lazzari and M. A. Lopez-Quintela, Adv. Mater., 15 (2003) 1583. [3] C. Goltner-Spickermann, Colloid Chemistry 1,226 (2003), 29. [4] G. Soler-Illia, E. L. Crepaldi, D. Grosso and C. Sanchez, Curr. Opin. Colloid Interface Sci., 8 (2003) 109. [5] A. C. Voegtlin, F. Ruch, J. L. Guth, J. Patarin and L. Huve, Microporous Mater., 9 (1997) 95. [6] J. M. Kim, Y. J. Han, B. F. Chmelka and G. D. Stucky, Chem. Commun., (2000) 2437. [7] S. S. Kim, A. Karkamkar, T. J. Pinnavaia, M. Kruk and M. Jaroniec, J. Phys. Chem. B, 105 (2001) 7663. [8] S. S. Kim, T. R. Pauly and T. J. Pinnavaia, Chem. Commun., (2000) 1661. [9] S. Ruthstein, V. Frydman, S. Kababya, M. Landau and D. Goldfarb, J. Phys. Chem. B, 107 (2003) 1739. [lo] K. Flodstrom, H. Wennerstrom and V. Alfredsson, Langmuir, 20 (2004) 680. [11] C. Z. Yu, J. Fan, B. Z. Tian and D. Y. Zhao, Chem. Mater., 16 (2004) 889. [12] C. Z. Yu, B. Z. Tian, B. Fan, G. D. Stucky and D. Y. Zhao, Chem. Commun., (2001) 2726. [13] J. Y. Chang and D. Y. Jung, Bull. Korean Chem. Soc., 24 (2003) 613. [14] Y. F. Lu, R. Ganguli, C. A. Drewien, M. T. Anderson, C. J. Brinker, W. L. Gong, Y. X. Guo, H. Soyez, B. Dunn, M. H. Huang and J. I. Zink, Nature, 389 (1997) 364. [15] D. Y. Zhao, P. Yang, N. Melosh, J. Feng, B. F. Chmelka and G. D. Stucky, Adv. Mater., 10 (1998) 1380. [16] P. D. Yang, D. Y. Zhao, D. I. Margolese, B. F. Chmelka and G. D. Stucky, Nature, 396 (1998) 152. [17] P. D. Yang, T. Deng, D. Y. Zhao, P. Y. Feng, D. Pine, B. F. Chmelka, G. M. Whitesides and G. D. Stucky, Science, 282 (1998) 2244. [18] D. Y. Zhao, P. D. Yang, D. I. Margolese, B. F. Chmelka and G. D. Stucky, Chem. Commun., (1998) 2499. [19] D. Y. Zhao, P. D. Yang, B. F. Chmelka and G. D. Stucky, Chem. Mater., 11 (1999) 1174.

[20] P. D. Yang, D. Y. Zhao, D. I. Margolese, B. F. Chmelka and G. D. Stucky, Abstr. Pap. Am. Chem. Soc., 216 (1998) U314. [21] E. L. Crepaldi, G. Soler-Illia, D. Grosso, P. A. Albouy and C. Sanchez, Chem. Commun., (2001) 1582. [22] E. L. Crepaldi, G. Soler-Illia, D. Grosso, P. A. Albouy, H. Amenitsch and C. Sanchez, Nanoporous Materials Iii, 141 (2002), 235. [23] E. L. Crepaldi, D. Grosso, G. Soler-Illia, P. A. Albouy, H. Amenitseh and C. Sanchez, Chem. Mater., 14 (2002) 3316. [24] E. L. Crepaldi, G. Soler-Illia, D. Grosso, F. Cagnol, F. Ribot and C. Sanchez, J. Am. Chem. Soc., 125 (2003) 9770. [25] E. L. Crepaldi, G. Soler-Illia, A. Bouchara, D. Grosso, D. Durand and C. Sanchez, Angew. Chem. Int. Ed., 42 (2003) 347. [26] D. Grosso, G. Soler-Illia, F. Babonneau, C. Sanchez, P. A. Albouy, A. Brunet-Bruneau and A. R. Balkenende, Adv. Mater., 13 (2001) 1085. [27] D. Grosso, G. Soler-Illia, E. L. Crepaldi, F. Cagnol, C. Sinturel, A. Bourgeois, A. Brunet-Bruneau, H. Amenitsch, P. A. Albouy and C. Sanchez, Chem. Mater., 15 (2003) 4562. [28] D. Grosso, G. Illia, E. L. Crepaldi, B. Charleux and C. Sanchez, Adv. Funct. Mater., 13 (2003) 37. [29] L. Pidol, D. Grosso, G. Soler-Illia, E. L. Crepaldi, C. Sanchez, P. A. Albouy, H. Amenitsch and P. Euzen, J. Mater. Chem., 12 (2002) 557. [30] G. Soler-Illia, E. Scolan, A. Louis, P. A. Albouy and C. Sanchez, New J. Chem., 25 (2001) 156. [31] G. J. D. Soler-illia, C. Sanchez, B. Lebeau and J. Patarin, Chem. Rev., 102 (2002) 4093. [32] G. Soler-Illia, A. Louis and C. Sanchez, Chem. Mater., 14 (2002) 750. [33] N. Baccile, D. Grosso and C. Sanchez, J. Mater. Chem., 13 (2003) 301 1. [34] C. Sanchez, B. Lebeau, F. Chaput and J. P. Boilot, Adv. Mater., 15 (2003) 1969. [35] L. M. Wang, B. Z. Tian, H. Fan, X. Y. Liu, H. F. Yang, C. Z. Yu, B. Tu and D. Y. Zhao, Microporous Mesoporous Mater., 67 (2004) 123. [36] B. Z. Tian, X. Y. Liu, B. Tu, C. Z. Yu, J. Fan, L. M. Wang, S. H. Xie, G. D. Stucky and D. Y. Zhao, Nature Mater., 2 (2003) 159. [37] J. W. Kriesel, M. S. Sander and T. D. Tilley, Adv. Mater., 13 (2001) 331. [38] J. W. Kriesel, M. S. Sander and T. D. Tilley, Chem. Mater., 13 (2001) 3554. [39] R. A. Pai, R. Humayun, M. T. Schulberg, A. Sengupta, J. N. Sun and J. J. Watkins, Science, 303 (2004) 507. [40] B. Z. Tian, X. Y. Liu, L. A. Solovyov, Z. Liu, H. F. Yang, Z. D. Zhang, S. H. Xie, F. Q. Zhang, B. Tu, C. Z. Yu, 0. Terasaki and D. Y. Zhao, J. Am. Chem. Soc., 126 (2004) 865. [41] T. Katou, B. Lee, D. L. Lu, J. N. Kondo, M. Hara and K. Domen, Angew. Chem. Int. Ed., 42 (2003) 2382. [42] D. L. Li, H. S. Zhou and I. Honrna, Nature Mater., 3 (2004) 65. [43] G. S. Attard, J. C. Glyde and C. G. Goltner, Nature, 378 (1995) 366. [44] M. Templin, A. Franck, A. DuChesne, H. Leist, Y. M. Zhang, R. Ulrich, V. Schadler and U. Wiesner, Science, 278 (1997) 1795.

[45] P. F. W. Simon, R. Ulrich, H. W. Spiess and U. Wiesner, Chem. Mater., 13 (2001) 3464. [46] A. C. Finnefrock, R. Ulrich, A. Du Chesne, C. C. Honeker, K. Schumacher, K. K. Unger: S. M. Gruner and U. Wiesner, Angew. Chem. Int. Ed., 40 (2001) 1207. [47] A. C. Finnefrock, R. Ulrich, G. E. S. Toombes, S. M. Gruner and U. Wiesner, J. Am. Chem. Soc., 125 (2003) 13084. [48] U. Wiesner, Nanotechnology in Mesostructured Materials, 146 (2003), 407. [49] D. Y. Zhao, Q. S. Huo, J. L. Feng, B. F. Chmelka and G. D. Stucky, J. Am. Chem. Soc., 120 (1998) 6024. [SO] D. Y. Zhao, J. L. Feng, Q. S. Huo, N. Melosh, G. H. Fredrickson, B. F. Chmelka and G. D. Stucky, Science, 279 (1998) 548. [51] A. Leonard, J. L. Blin, G. Herrier and B. L. Su, Nanotechnology in Mesostructured Materials, 146 (2003), 243. [52] A. Leonard, J. L. Blin and B. L. Su, Nanoporous Materials Iii, 141 (2002), 109. [53] A. Leonard, J. L. Blin and B. L. Su, Scientific Bases for the Preparation of Heterogeneous Catalysts, 143 (2002), 1027. [54] J. L. Blin, A. Leonard and B. L. Su, Chem. Mater., 13 (2001) 3542. [55] S. A. El-Safty and T. Hanaoka, Nanotechnology in Mesostructured Materials, 146 (2003), 173. [56] S. A. El-Safty and T. Hanaoka, Chem. Mater., 15 (2003) 2892. [57] S. A. El-Safty and T. Hanaoka, Chem. Mater., 16 (2004) 384. [58] P. Y. Feng, X. H. Bu and D. J. Pine, Langmuir, 16 (2000) 5304. [59] P. Y. Feng, X. H. Bu, G. D. Stucky and D. J. Pine, J. Am. Chem. Soc., 122 (2000) 994. [60] E. Prouzet, F. Cot, G. Nabias, A. Larbot, P. Kooyman and T. J. Pinnavaia, Chem. Mater., 11 (1999) 1498. [61] J. L. Blin, A. Leonard, G. Herrier, G. Philippin and B. L. Su, Nanoporous Materials Iii, 141 (2002), 117. [62] C. Z. Yu, Y. H. Yu and D. Y. Zhao, Chem. Commun., (2000) 575. [63] Y. H. Yu, C. Z. Yu, Z. Yu and D. Y. Zhao, Chem. Lett., (2000) 504. [64] J. Fan, C. Z. Yu, T. Gao, J. Lei, B. Z. Tian, L. M. Wang, Q. Luo, B. Tu, W. Z. Zhou and D. Y. Zhao, Angew. Chem. Int. Ed., 42 (2003) 3146. [65] C. Z. Yu, B. Z. Tian, J. Fan, G. D. Stucky and D. Y. Zhao, J. Am. Chem. Soc., 124 (2002) 4556. [66] L. M. Wang, H. Fan, B. Z. Tian, H. F. Yang, C. Z. Yu, B. Tu and D. Y. Zhao, Microporous Mesoporous Mater., 67 (2004) 135. [67] E. B. Cho and K. Char, Chem. Mater., 16 (2004) 270. [68] V. Z. H. Chan, J. Hoffman, V. Y. Lee, H. Iatrou, A. Avgeropoulos, N. Hadjichristidis, R. D. Miller and E. L. Thomas, Science, 286 (1999) 1716. [69] X. Y. Liu, B. Z. Tian, C. Z. Yu, F. Gao, S. H. Xie, B. Tu, R. C. Che, L. M. Peng and D. Y. Zhao, Angew. Chem. Int. Ed., 41 (2002) 3876. [70] Y. T. Chan, H. P. Lin, C. Y. Mou and S. T. Liu, Chem. Commun., (2002) 2878. [71] K. Flodstrom, V. Alfredsson and N. Kallrot, J. Am. Chem. Soc., 125 (2003) 4402. [72] F. Kleitz, S. H. Choi and R. Ryoo, Chem. Commun., (2003) 2136. [73] Y. Q. Wang, C. M. Yang, B. Zibrowius, B. Spliethoff, M. Linden and F. Schuth, Chem. Mater., 15 (2003) 5029.

[74] V. Antochshuk, M. Kruk and M. Jaroniec, J. Phys. Chem. B, 107 (2003) 11900. [75] B. Z. Tian, X. Y. Liu, Z. D. Zhang, B. Tu and D. Y. Zhao, J. Solid State Chem., 167 (2002) 324. [76] B. Z. Tian, H. F. Yang, X. Y. Liu, S. H. Xie, C. Z. Yu, J. Fan, B. Tu and D. Y. Zhao, Chem. Commun., (2002) 1824. [77] Y. Sakamoto, I. Diaz, 0 . Terasaki, D. Y. Zhao, J. Perez-Pariente, J. M. Kim and G. D. Stucky, J. Phys. Chem. B, 106 (2002) 31 18. [78] M. Kruk, M. Jaroniec, C. H. KO and R. Ryoo, Chem. Mater., 12 (2000) 1961. [79] J. Fan, C. Z. Yu, L. M. Wang, B. Tu, D. Y. Zhao, Y. Sakamoto and 0. Terasaki, J. Am. Chem. Soc., 123 (2001) 12113. [80] P. I. Ravikovitch and A. V. Neimark, Langmuir, 18 (2002) 9830. [81] B. L. Newalkar, S. Komarneni, U. T. Turaga and H. Katsuki, J. Mater. Chem., 13 (2003) 1710. [82] J. R. Matos, M. Kruk, L. P. Mercuri, M. Jaroniec, L. Zhao, T. Karniyama, 0 . Terasaki, T. J. Pinnavaia and Y. Liu, J. Am. Chem. Soc., 125 (2003) 821. [83] J. R. Matos, M. Kruk, L. P. Mercuri, M. Jaroniec, T. Asefa, N. Coombs, G. A. Ozin, T. Kamiyama and 0. Terasaki, Chem. Mater., 14 (2002) 1903. [84] J. R. Matos, M. Kruk, L. P. Mercuri and M. Jaroniec, Nanoporous Materials Iii, 141 (2002), 61. [85] M. Kruk, V. Antochshuk, J. R. Matos, L. P. Mercuri and M. Jaroniec, J. Mater. Chem., 124 (2002) 768. [86] M. Jaroniec and M. Kruk, Nanotechnology in Mesostructured Materials, 146 (2003), 343. [87] E. B. Cho, K. W. Kwon and K. Char, Chem. Mater., 13 (2001) 3837. [88] F. Kleitz, D. N. Liu, G. M. Anilkumar, I. S. Park, L. A. Solovyov, A. N. Shmakov and R. Ryoo, J. Phys. Chem. B, 107 (2003) 14296. [89] D. Y. Zhao, J. Y. Sun, Q. Z. Li and G. D. Stucky, Chem. Mater., 12 (2000) 275. [90] H. F. Yang, Q. H. Shi, B. Z. Tian, S. H. Xie, F. Q. Zhang, Y. Yan, B. Tu and D. Y. Zhao, Chem. Mater., 15 (2003) 536. [91] K. Landskron, B. D. Hatton, D. D. Perovic and G. A. Ozin, Science, 302 (2003) 266. [92] W. P. Guo, J. Y. Park, M. 0 . Oh, H. W. Jeong, W. J. Cho, I. Kim and C. S. Ha, Chem. Mater., 15 (2003) 2295. [93] W. P. Guo, I. Kim and C. S. Ha, Chem. Commun., (2003) 2692. [94] X. Xu, B. Z. Tian, J. L. Kong, S. Zhang, B. H. Liu, D. Y. Zhao, Adv. Mater., 15 (2003), 1932.

This Page Intentionally Left Blank

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Significance of mesoporous crystals for catalytic application John Meurig

horna as"'^ and Robert RajaC

"Department of Materials Science, University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, U.K. E-mail: [email protected] b ~ a v Faraday y Research Laboratory, The Royal Institution of Great Britain, 2 1 Albemarle Street, London W 1S 4BS, U.K. "Department of Chemistry University of Cambridge, Lensfield Road, Cambridge, CB2 lEW, U.K. E-mail: [email protected] ABSTRACT

Ordered mesoporous silicas offer unprecedented opportunities not only in the preparation and clarification of the mode of action of well-defined, single-site catalytically active centres on high-area solid surfaces, they also afford means of designing and producing new, high-performance heterogeneous catalysts, some of which are already of considerable commercial significance. The large, adjustable pores of these ordered silicas, with their high surface density of pendant silanol groups, readily permit bulky organometallic precursors to be uniformly anchored on their inner walls and, by subsequent gentle therrnolysis, to be converted to well-defined active sites, the process of conversion as well as the ensuing catalysis being followed by in situ X-ray absorption and FTIR spectroscopy thereby elucidating the precise mechanism of the catalysis. A particular example is the ~i'"-centred epoxidation of olefins, a process that is already of significance in sustainable development, since the unsaturated components (e.g. fatty acid esters) of sunflower and soya bean oils (and other plant sources) may be converted industrially to useful products (by HO-Ti-(OSi=)3 active centres). Bulky mixed-metal carbonylate precursors (typified by [ R U ~ ~ P ~ ~ C ~ ( Cmay O ) also ~ ~ ]be ~ ' readily ) converted to nanoparticle (1 to 1.5

nm dia.) catalysts at the interior surfaces of mesoporous silicas. These exhibit exceptionally high activities and selectivities in low-temperature hydrogenations, many of which may be carried out under solvent-free conditions. A potentially important conversion of muconic acid (derived from plant sources) into adipic acid can be effected on RuIOPt2nanoparticle hydrogenation catalysts. The adjustable curvatures and diameters of mesoporous silica permit asymmetric organometallic catalysts to be anchored in a spatially constrained manner, thereby leading to enhanced performance enantioselective catalysts, again of potential commercial significance. In view of some of the practical disadvantages (e.g. stability, cost, decomposition products and other factors {see Kleitz et al, Micropor. Mesopor. Mat., 44 (2001) 95)) of organic-template-derived mesoporous silicas, there is merit in using (for potential industrial applications), less wellordered mesoporous silicas, also of sharply defined pore diameters, for enantioselective hydrogenations. Specialized techniques of high-resolution electron microscopy (e.g. scanning transmission electron tomography and high-angle annular dark field imaging) as well as in situ XAFS studies conducted in parallel with XRD measurements are required fully to characterize most of our catalysts. 1.

INTRODUCTION

In the context of catalysis, the full impact of the enfergence of mesoporous silicas that occurred in the early 1990s is best appreciated if one first briefly recalls the quite exciting situation that prevailed at that time concerning zeolitic and other microporous solids belonging to the aluminosilicate and aluminophosphate families. Unlike bulk metal, alloy and other binary solid catalysts, openstructure aluminosilicates (embracing natural and especially synthetic zeolites) as well as open-structure aluminophosphates (AlPOs), particularly framework-substituted variants (MAlPOs where M Mg, Co, Mn, Zn, ...) are prime examples of uniform heterogeneous catalysts. Not only are the active sites in these catalysts distributed in a spatially uniform manner [I], they also conform to Langmuir's classic assumption in that the energy released upon the uptake of adsorbate species is constant up to monolayer coverage. This energetic uniformity is seen vividly in the calorimetric work cited in Fig. 1. where there is a constancy in the heat of adsorption of pyridine on the Bronsted acid sites (of pentasil zeolitic catalysts) up until all

-

the sites are neutralized (i.e. up to monolayer coverage of the catalytically active acid centres). Such constancy is never seen with metal catalysts because even a single-crystal face has a number of closely-spaced, distinct energetic sites; moreover the various individual sites are so close together in a metal surface that mutual repulsion causes a diminution in heat of adsorption with increasing coverage. Calorimetric measurements (see Fig. I.) again reveal the nature of the surface, which is clearly not uniform energetically nor spatially. In zeolitic (and MAlPO) catalysts, the active sites are so far apart - and the higher the SiIA1 ratio in a pentasil, acidic zeolite, the greater the separation distance between equivalent =Si-O(H)-Al- that they effectively behave as "single-site" catalysts. Likewise, in the shapeselective, Bronsted acid catalysts of the MAlPO-18 family [2], for example, where the active sites are the protons loosely attached to the oxygens adjacent , ~ " etc) , that occupy a to the doubly-charged M" ions (typically CO", ~ n " M small percent of the sites normally occupied by framework ~ 1 " 'ions, there is again spatial and energetic uniformity in the active sites. And even for other types of zeolitic catalysts, such as "ship-in-bottle" (or "tea bag") ones (where for example Cu-perchlorophthalocyanine [3] entities are incarcerated in zeolite Y or cobalt salophen [4] are encapsulated in zeolite Y, used respectively for the aerial oxidation of methane to methanol and primary or secondary alcohols to aldehydes or ketones), the notion of well-separated "single-site" active centres remains valid. Although - up until the early 1990s - microporous zeolites and MAlPO-type catalysts, because of their open structures, offered ready access of many reactants to, and egress of products away from, the active sites, their pore diameters (always less than 10 A) were so small that they prevented any progress to be made in catalytically converting molecules other than those of modest size and shape. Strenuous efforts were made up to the 1990s, using bulky organic templates - many designed with inventive ingenuity - to generate larger-pore zeolitic and MAlPO-type catalysts. But this endeavor, now seems, in retrospect, to have been chimerical. No one succeeded to generate stable (de-templated) microporous catalysts possessing pore dimensions much greater than 10 A. On reading the work of U.S.-based workers at the Mobil Co., on surfactant-mediated synthesis of the M41S family [5] and also of the Japanese (Tokyo-Waseda) groups on folded sheet silicates (FSM) derived from the synthetic mineral kanemite [6], one sensed immediately that the stage was set for a dramatic surge in the teeming availability of new types of heterogeneous catalysts that could not have hitherto been contemplated. It was not simply a case of the door being opened to enable one to process,

catalytically, in a petrochemical sense, bulkier hydrocarbons for cracking and reforming than had been possible hitherto with the so-called "large-pore" zeolites typified by faujasite, one could now also entertain many other challenging and important possibilities. For example, one had previously yearned to conduct subtle, elegant, possibly enantioselcctive conversions, or to oxidize as well as to hydrogenate selectively molecules of importance in fine-chemical, agrochemical and pharmaceutical contexts.

-.. ..-...-.-..I. 2 \

jl001

yrn

,

.

200

400

0

0

-5 Z150O O

,

Typical locations of the Bronsted acid sites (=AIO(H)Si=)are indicated by filled circles

600

......... . ,

,

0.2

0.4

c rn

F 100

m

8o

50

0.0

0.6

Apparent coverage IML)

Fig. 1. (Top) In the acidic zeolitic catalyst H-ZSM-5 (H,Si..,Al,Oz,, with SYAI ranging from 10 to 500), the active centres are widely separated and are never less than 0.6 nm from one another. In this uniform catalyst, each active site has the same enthalpy of adsorption of pyridine up to their total coverage (after Parrillo et al, Appl. Catal. A,, 110 (1994) 67). (Bottom right) The (111) face of a fcc metal has approximately ten times as per I unit )} area many adsorption sites {of three distinct kinds, top 0,bridged (B), hole @ than any surface of H-ZSM-5. This energetic non-uniformity is reflected in the decline of the enthalpy of adsorption with coverage. (Bottom left data refer to CO adsorbed on Pt(l1 I}, after Yeo et al, J. Chem. Phys., 106 (1997) 393).

There was also the intriguing thought that, with such well-ordered, well-defined, mesoporous silicas, the large internal areas of which were replete with silanol groups, it should at least be possible to compare the catalytic performance of a single-site homogeneous catalyst with that very same catalytically active site when anchored (and hence rendered heterogeneous) on the surface of a mesoporous silica. (This hope was indeed later fulfilled as described in Section 5 below). But numerous other thoughts occurred to one. Indeed, when, in 1994, Galen Stucky, Ferdi Schiith and co-workers published [7] in Nature their work on mesoporous silica the Editor of that journal asked me (JMT) to write [8] an accompanying News and Views, an excerpt from which I reproduce below: "The prospect of producing open-structure networks of a wide variety of inorganic materials, with apertures in the range 20 to 200 A diameter, is brought a stage nearer by the work of Galen Stucky, Ferdi Schiith and their co-workers, reported on page 31 7 of this issue. Such mesoporous ordered solids are likely to be of great practical value in a host of applications in the physical and biological sciences, in engineering and conceivably in medicine. Their mode of formation also sheds new light on the important phenomenon of biomineralization." 2.

THE PRACTICAL ADVANTAGES OF MESOPOROUS SILICA

Although the theme of this Symposium encompasses mesoporous crystals in general, it is prudent to emphasize that, in a catalytic context, mesoporous silica - rather than any other mesoporous solid - has, so far, had the profoundest consequences. This is partly because silica itself has most of the attributes of an ideal catalyst support. The merits of having available a higharea silica with well-defined pore dimensions are obviously attractive features to the catalyst merchant. In addition, owing to its excellent thermal and chemical stability, ease of handling and profusion of exposed silanol groups, silica is ideal for the heterogenization of molecular catalysts. Moreover, silica as a support has a rather rigid structure and does not swell in solvents, so that it may conveniently be used at both high and low temperature and at high pressure, in sharp contrast to most organic supports (e.g. styrenes). Its very inflexibility and non-compressibility makes silicatethered (i.e. anchored) catalysts suitable for use in continuous-flow reactors, which is a potentially key factor in chemical engineering situations. As has been recognized by others, site-isolation of anchored molecular catalysts (or

of any active catalyst derived from a bulky anchored precursor) can be more carefully defined on silica than on a flexible polymer backbone, but the catalyst (or ligand) loading must not be too high so as to maintain the condition of active site isolation. Another key advantage possessed by silica is that a number of the elements of the Periodic Table - not just obvious candidates like germanium, carbon, aluminium or gallium but also a range of transition-metal ions may be substitutionally incorporated in place of silicon into its structure, thereby transforming it into an active catalyst of a kind governed by the nature of the substituting element. This was obvious [ 8 ] from the outset: "Thefeasibility of synthesizing mesoporous silicas with some redox element such as titanium incorporated into their inner surfaces is an idea that has occurred to several groups [9-111. After all, the remarkable properties of titanium silicalite, in which ~i'"ions are inserted into microporous silica thereby converting it into a highly selective oxidation catalyst for the generation of catechol and hydroquinone from phenol, is already [I21 harnessed industrially by the Enichem Company in Italy, where such workJirst began. Corma and his colleagues [IO], using their Ti-incorporated mesoporous silica (pore diameter approximately 20 A) have selectively oxidized rather large organic molecules such as norbornene." But what was perhaps most attractive of all (as judged in early 1990s) about mesoporous silica - provided it possessed the necessary mechanical strength and thermal stability in air and in contact with solvents - was the high surface concentration of pendant silanol groups (in the range 1 to 2 OH per 1 nm2). This meant that a clear route, easily trodden by preparative chemists, existed for introducing an almost limitless range of large organic, and in particular organometallic moieties into the inner walls of high-area "crystalline" silica supports. By appropriate modification of their conditions of preparation laid down by the American and Japanese instigators of mesoporous crystals, it soon became possible to prepare mesoporous silica-alumina as was done early on by Bellussi et a1 1131. And indeed it proved relatively easy, by judicious choice of the alkoxide of a particular element, to incorporate traces of transition-metal ions (notably titanium [9-111, but also chromium, vanadium [14] and other ions) into the walls of MCM-41 silicas. Such preparative ease stimulated a number of interesting catalytic ventures.

3.

SOME INITIAL CATALYTIC MESOPOROUS SILICA

APPLICATIONS

OF

Two preparative approaches were pursued: (i) the incorporation [14] during growth of 'redox' or 'acid' catalytically active sites into the (thin) walls of the mesoporous silica; and (ii) post growth grafting (or anchoring or tethering, the terms are used as synonyms here) of the catalytic centre on to the inner surfaces of the mesoporous silica. The catalytic activity of mesoporous silica-alumina (i.e. acidic solids) of the MCM-41 type reported by Bellussi et a1 [13] generated some early optimism because of their encouraging performance in propene oligomerization (to produce gasoline and middle distillates with unique selectivity towards C9 and CI2 hydrocarbons, as this is a significant component of industrial catalysis). Soon, however, the limitations in stability and general mechanical and thermal ruggedness of these filigree versions of (MCM-41) mesoporous catalysts became apparent. It is now known that comparatively little real headway has been made in the last decade on mesoporous silica-aluminas in high-temperature catalytic applications, where hydrothermal stability of the catalyst is a key requirement [15]. For milder applications, however, such as low-temperature alkylations [16], mesoporous silica-aluminas are viable replacements for mineral acids as industrial Bronsted acid catalysts. The catalytic activity, particularly in selective oxidation, of ~ i "ions embedded within the walls of MCM-41 mesoporous silicas was studied in several early investigations [9-111. And modest catalytic activity, but very high selectivities were reported, as illustrated in Fig. 2. A significant advance was made when ~ i "active centres were grafted on to the inner surfaces of mesoporous silicas using an organometallic l ~ was ) done by precursor, in particular titanocene dichloride ( T i ( c ~ ) ~ Cas Maschmeyer et a1 [17]. The key steps in the introduction of the isolated, single-site, active centers on to the inner walls of MCM-41 are shown in Fig. 3. The detailed course of this "heterogenization" of a TiIVactive centre was followed by in situ X-ray absorption spectroscopy combined with in situ Xray diffractometry [18]. (The former to track the precise environment of the Ti, its valence state and its extent of coordination to neighbouring atoms; the latter to record whether the crystallinity of the mesoporous host is retained). Two important points emerged from this work. First, the "halfsandwich" Ti" compound, where one of the cyclopentadienes is retained, serves as a device to secure 'single-site' ~ i "active centres.

Fig. 2a. Computer-graphic representation of a titanium-containing mesoporous silica catalyst which selectively oxidizes 2.5-di-tert.butyl phenol into the corresponding quinone in the presence of Hz02 (after Thomas 181). Fig. 2b. Illustration (after Thomas and Greaves, Science, 265 (1994) 1675) showing ease of titanium-catalyzed epoxidation of cyclohexene.

Fig.3. The grafting of ~i"-centred active sites to the inner walls of mesoporous silica occurs via the interaction of titanocene dichloride (Ti(Cp)~C12)and pendant silanol groups. A half-sandwich surface compound (bottom right) forms as an intermediate. (Colour code: Ti, 0.CI, H = white).

- R'OH

.R'OH

-4

112 intermediate

Fig. 4. Mechanism of epoxidation of cyclohexene by an alkyl hydroperoxide (HOOK) catalyzed by the ~i"-active centre tripodally bound to silica. In situ XAFS and Dm calculations indicate two plausible surface intermediates (ql and q2)(Thomas et al. [33]).

Table 1 Cyclohexene epoxidation at 298 K in acetonitrile [32] Catalyst

Oxidant

Cyclohexene otherLb1 oxide (%)["I products (%)

Selectivity w.r.t. epoxide

TBHP MPPH air TBHP MPPH air TBHP MPPH air [a] Percentages given are based on the amount of cyclohexene converted to cyclohexene oxide. [b] Other products derived from cyclohexene and in the case of MPPH reactions from benzyl radicals. [c] Number of moles of epoxide produced per mole Ti" per hour.

Since this half-sandwich is an essential feature of the process of heterogenization there is no possibility for the formation of dimeric (or higher) Ti" compounds as is invariably the case when so-called isolated Ti centres are prepared via aqueous solutions. Second, the EXAFS results (together with in situ FTIR) proves beyond doubt that the catalytic site is Ti" in tetrahedral coordination and that the Ti" ion is tripodally connected, via oxygens, to the surface of the silica as represented in Fig. 4. (It is noteworthy that our EXAFS results ruled out a previously proposed [20] "titanyl" (>Ti=O) structure for the active site). If we symbolize the catalyst in which TirV centres are embedded (during growth) as Ti+MCM-41 and the catalyst in which the TirVcentres are grafted to the walls as T~?MCM-41,we may compare their catalytic activity with one another, and with the Ti/Si02 catalysts used by the Shell Co. to epoxidize propene to propylene oxide [21,22]. If, furthermore, we mesoporous compare the catalytic performance of these two ~i'~-centred, catalysts in a typical epoxidation reaction, where cyclohexene is epoxidized either by tert. butyl hydroperoxide (TBHP) or by 2-methyl- 1-phenyl-2-propyl hydroperoxide (MPPH), the results (Table 1) unmistakeably reveal that the 'grafted' Ti" centre is superior, by a factor of ten or so, in its activity. (In fact, by comparing the T~?MCM-41with any other Tirv-centred catalyst,

including the industrially used preparation, we find that the grafting method yields the best ever recorded catalytic performance). 3.1.

Other initial ventures Three other early endeavours to capitalize upon the merits of the availability of mesoporous silicas are outlined in this section:

the design of a cobalt oxo-centred catalyst grafted on to silica to (i) achieve the selective, low-temperature oxidation of cyclohexane to cyclohexanone [23]; (ii) sulfonic acid functionalized mesoporous silicas as catalysts for condensation and esterification reactions [24]; and (iii) the production of organic bases (as mild catalysts) attached to the inner walls of mesoporous silica via alkylsiloxy chains. The catalytic activation, especially partial oxidation, of alkanes constitutes one of the major challenges of present-day chemistry; and the conversion of cyclohexane to cyclohexanone is among the principal target reactions since the latter is used as a feedstock in several industrial processes, including the production of nylon from E-caprolactam and adipic acid [25291. Maschmeyer et a1 [23] took as a point of departure the fact that several oxo-centred trimeric cobalt (111) acetates (coordinated with pyridine [30]) exhibit considerably more activity in selectively oxidizing the tertiary C-H bond in adamantane than their dimeric analogues. They therefore grafted the H ) to ( ~the ~ ) ~inner walls of following species: C O ~ ( ~ ~ - O ) ( O A C ) ~ ( ~ & on MCM-41 and monitored changes in its structure (by in situ EXAFS and XRD [18]) during its use in the oxidation of cyclohexane with tert. butyl hydroperoxide (to yield tert. butanol and a mixture of cyclohexanone and cyclohexanol). 29 Si MASNMR spectroscopy was also used to identify precisely the nature of the immobilization of the catalyst. Interesting results were obtained: there was appreciable catalysis, during the course of which EXAFS studies revealed a significant change in the structure of the oxocentred CO"' trimer. By functionalizing mesoporous silica with sulfonic acid groups, Van Rhijn et a1 [24] produced catalytic materials that were very effective for the formation of bisfurylalkanes and polyol esters. An outline of the nature of these catalysts is shown in Fig. 5.

1. neutral H202 2.0.2 M H ~ S O ~ 3. rinsing

(CH2)3 I

Fig. 5. Van Rhijn et al's method [24] of producing sulfonic acid functionalized mesoporous silica catalysts for condensation and esterifications.

One of the reactions that they catalyse is

which is the formation of 2,2-bis(5-methylfuryl) propane. (A bisfurylalkane of this kind is a key intermediate for macromolecular chemistry). Neithcr the acidic forms of zeolite Y or zeolite Beta are of any use for this reaction: they each yield tarry oligomeric products, which promptly deactivate the zeolitic (functionalized) surface of the catalysts. It seems that the hydrophobic . . mesoporous silica prevents too strong an adsorption and oligomerization of 2-methvlfuran, while its larger dimension facilitates vroduct desomtion. The MCM-SO~H(coated) catalyst of Van Rhijn achieves greater thanA80percent conversion to the bisfurylalkane with 95 percent selectivity towards the desired product. By using the procedure outlined in Fig. 6., amine or diamine functions may be directly grafted to the mesoporous silica. The Knoevenagel condensation (of active methylene compounds of the type Z-CH2-Z') with aldehydes or ketones yields olefinic products (such as R ' RC=CZZ' ) using these amine-functionalized silicas [31].

-

Fig. 6 . Bmnel et al [3 11 converted micelle-ternplatedsilicas (MTS) into catalysts rich in amine or diamine functions.

4.

ILLUSTRATIVE CASE HISTORIES: A SUMMARY

Here we deal first (Section 5) with ~ i ' ~ - c a t a l ~ z(mesoporous ed solids) for selective oxidation, focussing mainly, but not exclusively, on epoxidation. Apart from shedding much light on the principles of catalytic action, it transpires that these ~i'"-centred catalysts play an increasing role in sustainable development in that they can convert abundantly available feedstocks such as fatty acid methyl esters (obtained from plant sources exemplified by sunflower oil and soya bean oil) as well as the vast family of naturally-occurring terpenes into desirable products for the polymer, fabrics and foodstuffs industries. We then describe some other transition-metal-ion (mesoporous silica) catalysts, which again exhibit good performance, and which are also examples of "single-site" heterogeneous catalysts. In Section 6 we focus on the scope offered by mesoporous silicas to design and produce novel enantioselective catalysts, which are of great commercial potential. Here we show how chemical advantage of the concavity of the mesopores may be exploited to enhance the enantioselectivity of a chiral catalyst grafted on to the walls of the mesoporous silica. Finally, in Section 7 we summarize the great advantages offered by mesoporous silicas as support for bimetallic nanocatalysts that are extremely active in a variety of selective hydrogenation reactions. Again it will emerge that sustainable development looms significantly with these catalytic variants since some of the intermediate products of biocatalytic conversion of corn and other plant material may, by selective hydrogenation, be converted to bulk chemicals such as adipic acid, which has wide use in nylon and other textile manufacture and in the foodstuffs industry. In

addition, many of the selective hydrogenations, leading to commercially important products, may be effected in a solvent-free fashion, a procedure that is environmentally benign.

As explained elsewhere [17,19,32-361 our in situ (XAFS aided by FTIR and UV-Vis) studies of the TiIv-centred active site at the internal surfaces of l~) an unambiguous picture mesoporous silica (grafted via T i ( c ~ ) ~ Cproduced both of the tetrahedrally coordinated metal ion and paved the way to a deeper understanding of the mechanism of the epoxidation of alkenes by peroxidic reagents. Moreover, by knowing precisely the atomic environment of the active centre, it became possible to boost further its catalytic activity. For example, one of the three silicons in mesoporous silica to which the Ti" is linked, tripodally via oxygen atoms, may be replaced by germanium, thereby boosting the catalytic performance [37]. Furthermore, owing to our active site in the knowledge of the atomic environment of the ~i'~-centred the so-called heterogeneous catalyst, soluble molecular analogues silsesquioxanes also possessing well-defined (single) Ti" active sites (see Fig. 7.) - could be prepared, and their catalytic performance directly compared with their heterogeneous analogues. Very seldom, if ever, is it possible to make a direct comparison of the catalytic performance of a particular active site which has essentially the same atomic architecture in the heterogeneous and homogeneous case. This comparison, (facilitated by the use of pre-edge and near-edge X-ray absorption spectroscopy and by molecular dynamics calculations) provides [38] quantitative information pertaining to the tetrahedrally coordinated active site (see Table 2 and Fig. 7). Table 2 Comparison of the performance of insoluble heterogeneous, single-site ~ i ' " l ~ i 0 2 epoxidation catalysts with their homogeneous soluble molecular analogue^'^' Homogeneous Catalysts [c-C5H9)7Si7012Ti(OSiPh3)] [c-C5H9)7Si7012Ti(OGePh3)]

Heterogeneous Catalysts ~i'bi02 T~?MCM~I T ~ ? G ~ ? M C 1M ~

[a] See Ref [38] for reaction conditions.

TOF (h-') 18 52 26 34 40

.. Time I min

Fig. 7. (Top left) The performance of a ~i'~-cenh.ed catalyst grafted on silica (T~TMCM41) is lcss than that of a grafted catalyst in which one of the three silicons (in HOTi(OSi)3) is replaced by Ge ( T ~ ? G ~ ? M C M - ~Both ~ ) . are superior to an ordinary T~?s~o* catalyst. The activity (Table 2) of the heterogeneous catalysts may be directly compared with analogous homogeneous catalysts (prepared from an appropriate silsesquioxane; R @CsH9))(bottom left) {see [38] and S. Krijnen et al, Angew. Chem. Int. Ed. Engl., 37 (1998) 356; Phys. Chem. Chem. Phys., 1 (1999) 361).

-

Attfield et a1 [39] showed that grafting Ti-(OSiPh& onto the internal surface of MCM-41 (without further calcination) produces an epoxidation catalyst with high activity and high selectivity. This arises becausc the presence of the phenyl groups stabilizes the catalytic ~ i "centres towards attack from atmospheric moisture. Interestingly, the elegant work of Tilley and his colleagues [40-431, who have pioneered the so-called molecular precursor strategy for control of catalyst structure (to arrive, as with the T i ( C ~ ) ~precursor cl~ at well-spaced, single sites) also found that when they grafted -OS~(O'BU)~ groups on to their SBA-15 specimens of mesoporous silica (without calcination) they too observed enhanced stability in their ~ i catalytically active sites. (We shall return to Tilley's method of preparing highly effective, atomically dispersed active sites on mesoporous silica below - not least because it is applicable to other transition-metal-ions besides titanium - but it is instructive to emphasize here the advantages of using tris(tert-butoxy)siloxy titanium complexes to generate single-site catalysts). Note, for example, that in the molecular entity T~[os~(o'Bu)~]~ there is

'

~

already built into this precursor the stoichiometry and environment (i.e. tetrahedrally coordinated Ti surrounded by four O S i groups) that is desired in the ultimate active catalyst. (These complexes react with the pendant silanol groups of MCM-41 or SBA-15).

Grafted ~i'"-centred catalysts for the epoxidation of fatty acid methyl esters 1441 Epoxidized fatty acids and their derivatives have been used for many commercial applications such as plasticizers and stabilizers in chlorinecontaining resins, as additives in lubricants, as components in thermosetting plastics, in urethane foams and as wood impregnants. Vegetable oils and fats are renewable sources of two popular unsaturated fatty methyl esters: methyl(Z)-9-octadecanoate (methyl oleate, structure 1 in Fig. 8. and methyl-(E)-9octadecanoate, methyl elaidate, structure 2). In the past, an environmentally unfriendly "peracid" method was used to epoxidize the naturally occurring unsaturated compounds. 5.1.

Fi 8 Both methyl oleate (I)and methyl elaidate (2) are completely cpoxidized using A;. Ti -grafted catalysts and tert. butyl hydroperoxide (3) as oxidant (after Guidotti et a1 [441).

Now, however, as Ravasio and her coworkers have shown [44], the ~ i ' ~ - ~ r a f tactive e d site on mesoporous silica (via T i ( c ~ ) ~ ( C [17]) l ) ~ is an excellent and environmentally friendly method for converting the fatty acid methyl esters (FAME) into their epoxides. These workers have recently ted also effectively converts the (doubly) shown that the ~ i ' ~ - ~ r a f catalyst unsaturated components of soya bean oil into useful epoxides - another important step towards sustainable development.

Grafted ~i'"-centredcatalysts for the epoxidation of terpenes [45] Major sources of terpenes (which are natural products, the structure of which is built up from isoprene units) are balsams, natural resins and essential oils, but they are also by-products of lemon- and orange- juice production as well as of the pulp and paper industries. Some terpenes, notably (-)-a-pinene and (+)-limonene are among the more readily (naturally) available optically active substances and are therefore used for the syntheses of other optically active products. Here again it is obvious that catalysts capable of efficiently functionalizing terpenes are of value in the context of sustainable development. The work of Ravasio and her collaborators [45] has shown that the T~?MCM-41 (grafted) catalyst (derived from T i ( C ~ ) ~ ( c l [17]) )~ is particularly good in epoxidizing such important terpenes as a-terpinol, carveol and limonene (see Table 3) under mild conditions (i.e. at ca 85 "C using tert. butyl hydroperoxide, TBHP, in CH3CN). Indeed, in harmony with earlier work on the epoxidation of cyclohexene [32] also using TBHP, the T~?MCM-41surpasses the activity of the sol-gel grown Ti+MCM-41 by a factor of ten in the case of the a-terpinol, the main constituent of pine oil. 5.2.

Table 3 Turnover frequency (TOF) of terpene epoxidation on Ti-MCM-41 TOF (h- 1)

Substrate a-terpineol carve01 limonene

and T ~ ~ M C M - ~ I

Ti+ 2 15 4

~ i ? 20 33 20

T = 85°C; CH3CN solvent; 30 % wt catalyst; TBHP:terpene mole ratio = 1

Judging by the results of other workers who have compared the ~ ~ the grafted variety) with that of catalytic performance of T ~ ? M c M - (i.e.

the Ti+MCM-41 (sol-gel preparation) there is no doubt of the superiority of the former. Thus, in their study of the hydroxylation of benzene in the liquid phase (using aqueous HZ02)He et a1 [46] found both higher activity and enhanced selectivity to phenol (as well as greater chemical stability) with the grafted catalyst.

Other transition-metal ion, single-site catalysts supported on mesoporous silica Shortly after the titanocene method of introducing isolated ~i'~-centred active sites at the surfaces of mcsoporous silica was introduccd [17], the method was applied with success to the production of molybdenum and vanadyl centres (also on to MCM-41) - see Fig. 9. MO" active centres on silica are good catalysts for the oxidative dehydrogenation of methanol to formaldehyde [47]. Likewise vanadium (vV) centres on mesoporous silica are good catalysts for the epoxidation of alkenes and for oxidation of alkanes to alcohols and ketones [48]. Maschmeyer et a1 [49] subsequently used other Ti-containing precursors to produce novcl siliceous high-area supports, such as TUD-1, in which both mesopores and micropores were present. These materials were prepared (without any involvement of micclles or alkylammoniurn ions as templates) using metal-complexes of a benign kind. 5.3.

-

CpzMC1z

mesoporous silica

sio

_*

sioq \osi Sio M

lOl_

P"

sio@@i/"\si sio

M=V

sio

S~O

SiO

Fig. 9. Single-site selective oxidation catalysts on mesoporous silica may be formed from their parent cyclopentadiene analogues (see Refs. [17], [46] and [47]).

A different approach, alluded to earlier, was pioneered by Tilley et a1 [40-43, 50-571 in which a molecular precursor route is taken to arrive at a series of active catalysts on mesoporous (and certain other) supports. The metal ions in question cover those of Ti, Cr, Fe and vanadyl. And the essence of their preparation is that the desired atomic environment aimed at in the final catalyst (e.g. Ti-(OSi)4 or Ti-(OSQ3) is already present in the socalled thermolytic molecular precursor. Thus, by taking as the precursor ( ' P ~ o ) T ~ [ o s ~ ( o ~ B uthe ) ~ environment ]~ ultimately achieved in the single-site it is Ti-(OS& catalyst is Ti-(OSi)3, and from the precursor T~[os~(O'BU)~]~ [40]. Typical supports used by Tilley were the high-area mesoporous silicas MCM-41 and SBA-15, the latter being distinctly more thermally stable (owing to its thicker walls) even though their activity was roughly equal [40].

<

u II

o/"\~o

1

'BUO' 'BUO

S,i

/si\ 'BUO

I

O'BU O'BU

I O'BU O'BU

i

i:

'Q

As--------------.

, ,,-

Silica

silica surface

Fig. 10. The Tilley method [40,43] of preparing single-site catalysts on mesoporous silica via thermolytic molecular precursors such as M[os~(o'Bu)~],.

A general picture is given in Fig. 10. The precursor is bonded to the hydroxyl groups of the surface of the silica via protonolysis reactions. For the case of an alkoxy(si1oxy) species of the type M[OSi(Ot~u)3],,where M Ti, Fe, Cr, . .... this surface-attachment chemistry occurs with loss of HO~BU or H ~ S ~ ( O ' B Uto) ~result , in bonding to the surface through M-0-(surface) or Si-0-(surface) linkages, respectively. Calcination then leads to the highly dispersed supported metal of nominal composition MO,(n-1)SiO2. A typical situation, relevant to the case of isolated Fe atoms at some silica surfaces (namely xerogels, but applicable in principle to mesoporous silica) is shown in Fig. 11. The activities and selectivities of this catalyst for selective oxidation of three reactants with H202are also shown [58].

-

molecular precursor and spe~troscopicmodel

SBA-15 silica surface I.O OH nm-'

OH OH OH

0H

I

4 / H , O

- CH2,CMe2

I

I

OH

YHI

b I

I

OH

OH

OH] I

well-defined, isolated sites 0.23 Fe nm-'

0 isolated, pseudo-tetrahedral 0-Fe(OSi03) sites selective oxidation catalysts for various organic compounds with H202: selectivity 100%

TOF, mol (mol ~ e ) . 's.' 2.5

Fig. 11. Single-site 'Fe' catalysts on silica exhibit good activity and selectivity [56,58]

Nowotny et a1 [59], extending the work of others [60] on rhodiumcatalyzed hydrofonnylations (in which an alkene and a mixture of CO and H2 are catalytically converted to an aldehyde), compared the behaviour of Rh(I1) dinuclear complexes when they were separately grafted on to ordinary silica

-

and on to MCM-41 mesoporous silica.

The dinuclear complex was

[Rh2(p- PC )2(p-02CR)2] where p- PC is a bridging ortho-metalated arylphosphane ligand (see Fig. 12.). The performance of the immobilized catalysts 3 and 4 (Fig. 12.) was studied using styrene and 1-decene. The chemoselectivity towards the formation of aldehyde products was nearly quantitative in all the experiments employing styrene. Some catalyst leaching took place from each support, and the drop in activity was appreciably less in the case of the complex grafted inside the MCM-41.

,, ,104~~~

Q

'Rh1°

3-

'h2

1

rO

R(CHz)2PPhz

toluene I HOAC

‘rU

Pig. 12. This dinuclear Rh compound grafted on silica (see text) smoothly hydroformylates styrenc to its linear aldehyde.

To summarize, we show in Fig. 13. the many oxidative reactions (of considerable industrial significance) of unsaturated and saturated hydrocarbons that may be effected by transition-metal ion, single-site catalysts supported on mesoporous silica.

Fig. 13. A selection of the important selective oxidations that may be effected by a range of metal-centred, single-site catalysts grafted on mesoporous silica.

6.

DESIGNING CHIRAL CATALYSTS CONFINED WITHIN MESOPOROUS SILICA: THEIR SUPERIOR PERFORMANCE RELATIVE TO HOMOGENEOUS ANALOGUES

In the pharmaceutical and agrochemical industries, as well as in the expanding fields of fine chemicals generally [61], which encompasses fragrances and flavors, there is a growing demand for enantiomerically pure products, driven in part by ever-more exacting legislation and in part by stringent scientific criteria. To date, the asymmetric catalysts employed both on the laboratory and industrial scale have been homogeneous, largely because these possess well-defined, single-site active centres. No one doubts that, from the standpoints of ease of separation of products and regeneration of the catalyst, heterogeneous asymmetric catalysts would be far superior to their homogeneous counterparts. The cost alone of the sophisticated chiral ligands often exceeds that of the noble metal employed, so that catalyst

recovery is of cardinal importance for the application of enantioselective metal-centred catalysis to large-scale processes (particularly in continuousflow reactors) [62]. The problem, however, has been that, hitherto, almost all attempts to heterogenize (by immobilization on an appropriate support) homogeneous chiral catalysts has led to poor performance, principally because a spectrum [63] of different kinds of active site was generated by the very act of heterogenization. We recognized [64-671 quite early on that mesoporous silica, because of its large pores and profusion of functionable, pendant silanol groups presented unprecedented opportunities for designing powerful new types of chiral catalysts in which advantage could be taken of the spatial restrictions (for prochiral reactants) that exist after grafting and confining asymmetric (homogeneous) catalysts in the pores and channels of mesoporous silica. In other words, with such supports, no longer would a spectrum of active sites result: single sites would prevail. 6.1.

Strategic principle The large-diameter channels of MCM-41 family (see Fig. 13.) prompted us to graft quite sizeable chiral metal complexes and organometallic moieties on to the inner walls of these high surface area solids (see section 3.1 and ref. [23] above) by a variety of ways that included functionalizing pendant silanols with organic groups such as alkyl halides, amines, carboxylates and phosphanes. This opened the way to the preparation of novel catalysts consisting of quite large (surface) concentrations of accessible, well-spaced, and structurally well-defined active sites (As outlined in Section 9 below, the whole panoply of in situ and ex situ techniques of characterization, embracing spectroscopy, resonance and diffraction of diverse kinds could and were deployed for such purposes [67,68]). One expected, and we did indeed find, as shown in Sections 6.2 and 6.3, that such heterogeneous solid catalysts behave at least as efficiently as their homogeneous counterparts and sometimes with far superior enantioor regioselectivity. Various kinds of organometallic, chiral catalysts may be tethered to the inner walls of a mesoporous silica employing the strategy illustrated in Fig. 14. The key features here are the reactant's (i.e. the substrate's) interaction with both the pore walls and the chiral directing group. The confinement of the reactant (substrate) within the mesopore should lead to a larger influence of the chiral directing group on the orientation of the substrate (reactant) relative to the reactive catalytic centre when compared to the situation in solution.

.

"Chird Space"

Through-SpaInteractions

Fig. 14. Schematic representation of the confinement concept in which the substrate is incarcerated in the cavity of a chiral modified mesoporous host and leads to chiral heterogeneous catalysis [65].

Proof of principle in allylic amination To test the idea encapsulatcd in Fig. 14., we first decided to investigate the allylic amination reaction between cinnamyl acetate and benzylamine. This reaction has two possible products: a straight-chain one (which is favored as a result of the retention of the delocalized ?I system) and a chiralbranched one (Scheme 1):

6.2.

THF, 313 K

PhCH,NH,

[cat]'

Scheme 1. The allylic amination of cinnamyl acetate and benzylamine

The aim of the reaction is to produce (with an effective chiral catalyst) the greatest possible yield of the branched product with the highest possible

cnantiomeric exccss (ee). Three related chiral catalysts were chosen: one homogeneous; another the same homogeneous catalyst grafted on to a convex, non-porous silica surface (such as the commercial product known as Cabosil); and yet another the same chiral homogeneous catalyst grafted on to the inner walls of mesoporous (MCM-41) silica.

/ \

R = CH=CHICHIH~

2

3

5

,

1

4

7

iii

Fig. 15. Sequence of steps showing the immobilization of N-[1',2-bis(dipheny1phosphanyl)femcenyl]-ethyl-N,N'-dimelhylcthylediamine (2) in its chirally constrained and unconstrained (5) states. R = (CH&Br

We demonstrated [69] that a chiral ligand derived from 1,l'bis(diphenylphosphino)ferrocene (dppf) bonded to an active metal centre ( ~ d " )and tethered, via a molecular link of appropriate length, to the inner walls of a mesoporous silica (MCM-41 of g 30 A diameter) yields a degree of catalytic regioselectivity as well as an g that is far superior to either the homogeneous counterpart or the Cabosil-bound catalyst (We chose a chiral chelate based on dppf for several reasons: first, its planar chirality never undergoes racemization; second, it is synthetically very accessible; and third, dppf possesses functionalities suitable for reaction with pore-bound tethers). Care was taken to ensure that all activity is confined to the internal surface of the mesoporous silica. This was achieved by selectively deactivating the external surface of the support. Our overall approach to the comparisons between the three systems is summarized in Fig. 15. from which it is seen that the mesoporous framework was fust treated with Ph2SiC12to deactivate the exterior walls of the MCM4lsample. The interior walls of this material were then derivatized with 3-bromopropyltrichlorosilane to give the "prepared" MCM-41 designated 1 in Fig. 15.

dppf-diaminePd-catalyst

Conv

Straight Chain

(%)

r/.)

Homogeneous

76

99+

Tethered-Silica

98

98

Branched

ee

(%)

(%)

2

43

Fig. 16. Whereas the homogeneous dppf chiral catalyst (top left) yields no branched product (and no enantioselectivity) and the non-porous-silica tethered dppf catalyst yields but a small amount of the branched product, the spatially constrained f o m produces a substantial branched f o m and a high value of ee.

The ferrocenyl-based ligand (S)- 1-[(R)-lf,2-bis(diphenylphosphino)ferrocenyl]ethyl-N,Nf-dimeth-ylethylenediamine 2, was prepared by literature methods. On treatment of the activated MCM-41 with 1 with an excess of 2, the chiral catalytic precursor 4 is produced, and this, on reaction with PdC12CH3CN gives the required catalyst 6 . A separate related procedure yielded the Cabosil-supported catalyst 7. The grafted chiral catalyst 6 was fully characterized by MASNMR and EXAFS spectroscopy [67,68]. The mesopore-confined catalyst showed an enantioselectivity for superior to that of both the homogeneous (only linear product) and Cabosil-tethered analogue as shown in Fig. 16.

Exploiting confined chiral catalysts for enantioselective hydrogenations Having established the principle for the case of allylic amination, we then proceeded to take advantage of asymmetric catalysts grafted inside mesoporous silica to a number of industrially important hydrogenations. 6.3.

6.3.1. Conversion of ethyl nicotinate to ethyl nipecotinate [70]

\

ethyl nicotinate

N

/

chtral modifier

1,4,5,6-tetrahydronicotinate

N ''

ethyl nipecotinate

Scheme 2. The two-step hydrogenation of ethyl nicotinate to ethyl nipecotinate Ethyl nipecotinate is an important intermediate in biological and medicinal transformations. Previous efforts to hydrogenate enantioselectively an aromatic ring such as that in ethyl nicotinate had resulted [71] in values of ee that were less than 6 percent, but a two-step (Scheme 2) process (using a cinchonidine modified Pd catalyst supported on carbon) raised the to 19 percent at a conversion of 12 percent [72]. Using our Pd(dppf)-chiral catalyst confined within the 30 A pores of mesoporous MCM-41 we achieved (in a single-step) conversions in excess of 50 percent with an g of 17. The Pd(dppf) catalyst was also grafted to the vertex of an incompletely condensed silsesquioxane cube [73] (compare Fig. 7 above), the idea here being to

create a soluble (homogeneous) analogue of our confined chiral catalyst attached to a (0-Si-0)" framework. This catalyst resulted in a racemic product, thereby proving the chiral advantage achieved by confinement of the Pd(dppf) active centre inside siliceous mesopores (Fig. 17.).

dwf~finyl dl.mincWCa,a,p,

Humogmous

(I)

Subswale (rnEU",)

i

Canv

rr

(h)

(*o

(%)

-

ethyl

72

IS9

nlcotim

12L

21.2

Fig. 17. The dppf catalyst confined within mesoporous silica yields substantial ee of the desired nipecotinate product. Neither the homogeneous chiral catalyst not its nonconstrained, tethered form (on Cabosil) yield any significant ee of the nipecotinate.

6.3.2 Other enantioselective hydrogenations using confined, chiral diamino-type ligands [74-761 Comparatively few reports have hitherto been published in which ~ hor' ~ d " asymmetric complexes without phosphane ligands have been used to activate hydrogen, but a growing number employing nitrogen-containing ligands has appeared of late for the purpose of enantioselective conversions [77]. The chiral ligands shown in Fig. 18. have been used by us with ~ has'the metallic

core: a molecular diagram of the ~ hcomplex ' with ligand c (in Fig. 18.) is shown in Fig. 19. It is seen that the molecular cation (that functions as a chiral catalyst) is hydrogen-bonded (visible in the crystal structure [76]) to the BF; anion via the pendant nitrogen of the pyrrolidine.

Fig. 18. Some of the chiral ligands used in the constrained organometallic catalysts for hydrogenations described in the ensuing pages

Fig. 19. A molecular diagram of the heterogenized single-site ~h'-(s)-(-)-2-aminomethylI-ethyl pyrrolidine catalyst bonded to 1,5-cyclooctadiene (COD).

The catalyst itself is pseudo-square-planar where the ~ his' bonded to 1,5cyclooctadiene (COD). The hydrogenations (Scheme 3) investigated by us were:

Schcmc 3. Schematic representation of the hydrogenations of E-a-phenylcinnamic acid and methyl benzoylformate.

By grafting the ~ hchiral ' complex on both a concave silica (using MCM-41, 30 A diameter) and a convex silica (a non-porous silica) we established beyond doubt that the spatial restrictions imposed by the concave surface at which the active centre was located enhances the enantioselectivity of the catalyst - see Fig. 20.

silica

<&<

concave

silica

convex

Fig. 20 a. Graphical model (to scale) showing the constraint at the catalyst (see Fig. 19) when anchored on a concave silica in contrast to the situation on a convex (Cahosil) surface.

Heterogeneous (concave)

Heterogeneous (convex)

Fig. 20 b. The chiral diamine organometallic catalyst constrained at a concave silica surface surpasses the performance (selectivity and ee) of the same catalyst (shown in Fig. 20 a.) attachcd to a convex, non-porous silica.

(C). A deeper analysis of the beneficial use of asymmetric organometallic catalysts constrained within mesoporous silica [78] Convinced of the merit (in enantioselective syntheses and other organic processes) of constraining asymmetric organometallic catalysts within siliceous nanopores (so as to increase the interaction between the pore wall and the active centre and hence to restrict access of reactant to thc catalyst) we embarked on a systematic study 1781 in which a range of porous silicas were investigatcd. In each of these there is a very narrow spread of pore diameter. Rather than employing as porous siliceous supports the organic-template derived (MCM-41 or SBA-15) varieties, a set of commercially available desiccant silicas having narrow pore size distributions (Fig. 21.) (designated Davison 923, 634 and 654). (These are made by reacting sodium silicate with a strong mineral acid (usually sulfuric acid); the pore-size being controlled by gel time, final pH, temperature, concentration of reactants, etc). Compared to MCM-41 type silicas they are much lower in cost, more thermally and mechanically stable, less susceptible to structural collapse and available in a range of granularities. They also have some intersecting pores that facilitate the diffusion of the reactant spccies to the immobilized catalyst. The average diameters [79] of the pores of these silicas is, respectively, 38,60 and 250 A, and their respective surface arcas are 700,500 and 300 mZg-I.

Fig. 21. Pore-size distribution curve for the mesoporous sample (Davison 923) which has a value of 38 bi mean pore diameter.

Instead of grafting the cationic Rh(1) complex, containing the chiral diamino ligand (and cyclooctadiene, COD), using our customary covalent procedure (involving 3-bromopropyl-trichlorosilane to link up with a surface silanol group), we have insteud employed the non-covalent immobilization approach recently described by Rege et a1 [80,81]. In this method, a surfacebound triflate (CF3S0<) counter-ion securely anchors the cationic Rh(I)(COD) (Fig. 22.) or the Pd(ally1) diamino complcx to the inner wall. (This straightforward method circumvents the need for ligand modification to secure covalent tethering and its advantages are described fully elsewhere [751).

Fig. 22. (Left) A triflatc counter-ion, which is strongly hydrogen-bonded to the silanols of the silica, holds securely (by ionic interaction) the chiral organometallic ~h'-centred cation. (Right). A topographic view within a concave surface showing the constraints of the catalyst once anchored onto a mesoporons surface and the spatial restrictions that dominate the approach of an incoming prochiral reactant (E-a-phenylcinnamic acid).

The four constrained chiral catalysts were:

~Rh(COD~(S)-~+)-l-[2-~vrrolidinyl~ethy1)-~yrrolidine CF3SOY. . . .. . . . . . . . . I Pd(allyl) (S)-(-)-I -(2-l'yrrolidinylMcthyl)-Pyrrolidinc]. CI:,SO,', [Rh(COI)) (S)-(-)-2-Aminomethyl-1-Ethyl-Pyrrolidinc]' CF3S0,', and [Rh (COD) (IR,2R)- (7)-1.2-Iliphcnyl-Ethylenel)iamine]. CFxSO; which we abbreviatc to Rh(COL))PMP, Pd(allyl)PMP, Rh(C0D)AEP and Rh(C0D)DED respectively.' ~ n the d test-reaction was the asymmetric hydrogenation of methyl benzoylfonnate to its corresponding methyl mandelate (see Scheme 3 above)

Table 4 Asymmetric hydrogenation of methyl benzoylformate Catalyst

Cat

Silica (pore dia.)

Homo Het

Rh(C0D)AEP

Metal Rh(1)

Dav. 923 (38 4 Dav. 634 (60 4 Dav. 654 (250 A)

Homo

Rh(1)

Rh(1)

t (h) 2.0

Conv

ee

62

TOF (h-') 46

0.5 2.0

82.6 93.3

542 153

82 77

0.5 2.0

67.1 93.9

440 154

65 61

0.5 2.0 2.0

44.6 86.1 69.9

292 141 60

0 0 0

0.5 2.0

77.7 98.1

596 188

50 79

0.5 1.0

59.7 75.5

458 290

68 73

0.5 2.0 0.5

38.8 83.1 46.2

298 159 145

0 4 53

0.5 2.0

92.8 95.8

643 166

85 94

0.5 2.0

63.0 91.5

436 159

72 78

0.5 2.0 0.5

60.7 86.9 96.0

420 151 264

65 59 55

0.5 2.0

89.8 98.9

542 149

62 67

2.0

100

151

66

0

,Rh(N ph

Hit

Rh(C0D)DED Homo Het

Rh(C0D)PMP

Dav. 923 (38 A) Dav. 634 (60 A) Dav. 654 (250 A)

Rh(1)

Rh(1) Dav. 923 (38 A) Dav. 634 (60 A) Dav. 654 (250 A)

Rh(1)

Pd(I1)

Homo

Pd(I1)

Pd(ally1)PMP

Catalyst Recycled

-

substrate = 0.5 g; solvent = 30 ml; catalyst (homogeneous) 10 mg; (heterogeneous) = 1 -I 50 mg; H2 20 bar; T = 3 13 K; TOF = [(molsubstrate-converted)(m~lcom~lex(~h or ~d)/silica)h 1

In their homogeneous form, only the Rh(C0D)PMP and Pd(ally1)PMP exhibit any significant enantioselectivity (ee) under the reaction conditions (see Table 4) employed by us, whereas the other two homogeneous catalysts (namely Rh(C0D)AEP and Rh(C0D)DED) did not display any significant ee. This is probably because the bulkiness of PMP in comparison to AEP and DED exerts further spatial congestion in the vicinity of the active centre. Table 4 summarizes the results with all four chiral catalysts and shows that, as expected from arguments given above, chiral restriction does indeed boost the ee values in a manner that logically reflects the declining influence of spatial constraint in proceeding from the 38 A to the 60 A to the 250 A pore-diameter silica. For the heterogeneous catalysts the trend with Rh(C0D)PMP is mirrored by both AEP and DED ligands and it is clear that even when some of the asymmetric catalysts exhibit significant ee's under homogeneous conditions, their performance is much enhanced when immobilized in a constrained environment It is also noteworthy that the noncovalent method of anchoring the organometallic catalyst does not lead to facile leaching when the catalyst is recycled. Further experimental details are given in patents [82] recently filed by German industry. Although, in general, enzymes are at present more widely used industrially than asymmetric transition-metal complexes for enantioselective catalytic conversions involving pharmaceuticals and agrochemicals, the latter are of increasing importance and are more generally applicable especially for reactions that cannot be catalyzed enzymatically. In commenting on our earlier preliminary work on constrained catalysts, the authors [61] of a recent comprehensive text on the role of heterogeneous catalysts in the production of fine chemicals remarked: "This approach seems to hold considerable promise for meeting the future challenge of developing robust, recyclable catalysts for asymmetric syntheses." We believe that this view is vindicated by our subsequent endeavours. In particular, the convenience of using a noncovalently grafted chiral catalyst has obvious practical merit.

7.

NANOPARTICLE BIMETALLIC HYDROGENATION CATALYSTS SUPPORTED ON MESOPOROUS SILICA

Physical and chemical characteristics Finely divided metallic and bimetallic particles ranging in size from a few to several thousand atoms have long played an important part in laboratory and industrial catalysis. Two main problems concerned with their preparation and properties were recognized early on: first, the difficulty in preparing 'mono-disperse7 nanoparticles (i.e. where they are all of the same 7.1.

size and contain the same number of atoms), and second, their propensity to sinter on the underlying support when subjected to modest temperatures. Traditional methods for preparing nanoparticle catalysts have generally involved the procedure of so-called 'incipient wetness': a solution of the appropriate salt, containing the metal that is ultimately desired as a nanoparticle, is allowed to be absorbed andlor adsorbed (depending on the porosity) by the high-area support, typically alumina, silica or silica-alumina. After thorough drying, reduction and judicious heat-treatment the supported nanoparticles appear. But this method almost invariably produces a distribution of particle sizes, and very often the nanoparticles contain several hundred or so atoms.

Fig. 23. (Top left) Scanning transmission electron tomograph of mesoporous silica containing Rul#t2 nanoparticle catalysts. (Top right) HAADF images of Ru~oPtz nanoparticles with the electron-induced X-ray emission peaks (shown in bottom right) of three individual particles each consisting of 12 atoms (and each weighing 2 zeptograms {2 x 10.~'g}). Each nanoparticle is separately imaged and its precise composition may be determined from the X-ray spectra.

A bimetallic system, involving say, ideally nanoparticles of Cu and Ru in intimate contact with one another may prove very difficult to prepare in

this way. In recent years, other workers have used mixed-metal cluster compounds as precursors. Thus Nashner et al [83] used the cluster [PtRuS(C0)6] on carbon supports, but they found that the PtRuS 'cores' aggregated to produce relatively uniform nanoparticles ranging in diameter from g 8 to 23 A. And with precursors such as [R~~C(CO),,{~~-R~(CO)~ { p 3 - ~ r ( ~ 0 ) ~ the } ] 2resulting ~ material formed separated nanoparticles, the original metal ratio of Re and Ir having been lost. As soon as mesoporous silicas became available, with their pores large enough readily to allow access to quite bulky mixed-metal carbonylate P ~ ,C~ IR~U]~~(~C O and ) ~ ~with ] ~ - )their internal (typically [ R U ~ ~ C ~ ( C O ) ~ ~orC[U surfaces rich in silanol groups, it immediately became apparent [84] that a reliable method of introducing well-defmed, mono-disperse, uniformly distributed (spatially) nanoparticle bimetallic catalysts consisting of 4 or 6 or 12 or 16 metal atoms (in specific ratios such as RusPt pr RuloPt2)was open to us. Full details are contained in Refs. [84-881. The nature of the bonding of these nanoparticles, in precise atomic detail, is determined from XAFS spectroscopy, and scanning transmission electron microscopy yields their spatial distribution within the pores. Electron-induced X-ray emission (on individual nanoparticles) reveals 1891 the atomic ratio of the constituent elements and scanning transmission electron tomography shows the morphology of the nanoparticles within the nanoporcs see Fig. 23. -

Fig. 24. (Left) A single mesopore replete with its pendant silanol group, with which the carbonyl groups of the mixed-metal carbonylate carbonylate (such as that shown on the right with its molecular cation, PPN) may form hydrogen bonds of the kind schematized here.

If one of the components of the bimetallic nanoparticles is chosen carefully, its oxophilicity (e.g. of Cu or of Pd) secures the bimetallic entity firmly to the support. This endows the nanoparticles with a far greater resistance to sintering when the system is heated - see Fig. 24.

7.2.

High-performance nanocatalysts for single-step hydrogenations After demonstrating [90,911 that we could routinely prepare bimetallic catalyst particles that are (i) so small (1 to 1.5 nm diameter) that essentially all the atoms in them are exposed to reactant species; (ii) so firmly anchored (via Si-0-M bonds with M = Cu, Ag, Pd, etc) to the walls of the mesopores that their tendency to sinter and coalesce is minimal; and (iii) distributed in a spatially uniform manner so as to provide easy diffusional access of reactants to, and egress of products from, the nanoparticles, we proceeded to carry out a number of commercially relevant hydrogenations. Many of these could be effected in a solvent-free, single-step fashion, features that are invaluable in the context of clean technology and green chemistry. These nanoparticle bimetallic catalysts exhibit exceptional activity and are highly selective (as is shown below). Because of their minute size (which confers novel electronic properties upon them) each atom - and almost all are at the surfaces of the nanoparticles - is coordinatively unsaturated. Secondly, as was realized earlier by Sinfelt [91], bimetallic clusters are vastly superior catalytically to their monocomponent counterparts. This is well illustrated in our work on the hydrogenation of 1-hexene, 1-dodecene or naphthalene, where we examined the activity of bimetallic PdRu and monometallic Ru and Pd nanoparticles, derived from Pd6Ru6, Ru6 carbonylate clusters and a Pd complex, respectively. The turnover frequencies displayed by the Pd6Ru6 nanocluster (in hexene hydrogenation) was a factor of ten or more in excess of those for monometallic Ru or Pd clusters. It is not yet clear - more theoretical study is required - why the synergy between Pd and Ru is so pronounced. But it is relevant to note the well established ability of Ru to activate molecular hydrogen and of Pd to activate the olefinic bond. Bimetallic nanoparticles such as Ru6Pd6, Ru6s1-1, RuIOPt2,RuloPt, RuI2Cu4and RuI2Ag4,anchored within mesoporous silica all exhibit high activities and frequently high selectivities, depending upon the composition of the nanoparticles, in a number of single-step (and often solvent-free) hydrogenations at low temperatures (333 to 373 K). The selective hydrogenation of polyenes (such as 1,5,9-cyclododecatriene and 2,5norbornadiene) are especially efficient. Good performance is found with these nanoparticle catalysts in the hydrogenation of dimethyl terephthalate (DMT) to 1,4-cyclohexane dimethanol (CHDM) and of benzoic acid to

cyclohexane- 1-carboxylic acid, and also in the conversion of benzene to cyclohexene, the latter being an increasingly important reaction in the context of the production of Nylon. Table 5 Single-Step, Highly Active and Selective Nanoparticle Catalysts for the Hydrogenation of Some Key Organic Compounds Catalyst

Reaction

Solvent

t

TOF

Commercial Significance Polymer intermediates, ketones and polyesters Laurolactam, copolyamides, nylon intermediates Coatings, lactones, polymers Starting material in production of K-A oil Polyester fibres, polycarbonates polyurethanes Precursor for caprolactam and nylon Paints, waxes, lacquers, polishes and as substitutes for turpentine, Precursor for nylon and Ecaporolactam

hexadecane hexadecane

hexane hexane Ru~Pt1iSi02 Ru10Pt2ISi02 Pd6Ru6/Si02

H

O

O

COOH

c

C2HsOH C2H50H C2H50H

5 5

5

912 965 1012

Nylon 6.6, gelatins, jams, polyurethanes, lubricants

Catalyst = 20 mg; H2 pressure = 20 bar; TOF = [ ( m ~ l , , ~ ~ t ~ ) ( m ~ l ~ ~ ~ ~ t ~ ~ ) ~ l h ~ ~ ] Note: Mesoporous SiO2 used here is of the MCM-41 type.

Table 5, above, highlights some of these conversions, as well as the remarkably high turnover frequencies (TOF) and an indication of the commercial potential of the reactions for which we have found viable catalysts. Adipic acid from sustainable sources via a mesopore-supported RuloPt2nanoparticle catalyst [921 Muconic acid (see Fig. 25.) may be readily produced from corn using a biocatalyst (devised by J.W. Frost [93]). We have discovered a catalyst (nanoparticlc of RUIOP~Z generated by gentle thermolysis after introducing thc carbonylate precursor [ R U ~ ~ P ~ ~ C ~ into (CO mesoporous ) ~ ~ ] ~ - silica [92]) that is superior to all others (Rh/A1203,PtlSi02, P&RudSi02) in converting this acid, in a single-step, to the desired adipic acid. Adipic acid is a major stepping-stone in the production of Nylon and other fabrics and foodstuffs. Hitherto, it has been produced (by a sequence of environmentally aggressive steps [90]) from fossil-fuel sources, in particular benzene (which is converted to cyclohexane that is oxidized to cyclohexanone and cyclohexanol and these are then transformed to adipic acid). It is clear that bimetallic nanocatalysts, allied to the appropriate biocatalyst, have a major role to play in future environmentally benign industrial processes. Very recently [94] a highly selective nanoparticle colloidal catalyst, supported on mesoporous silica, has been developed to hydrogenate phenol prefcrentially to cyclohcxanone (with cyc1ohexanone:cyclohexanol ratios in excess of 100). This is deemed a major step forward in the context of industrial catalysis.

7.3

.. m860pomus

ei1,m

00 0 Si

i, A

OC

Fig. 25. Schematic diagram illustrating the process of converting trans, trans-muconic acid, derived from glucose, to adipic acid, which is used in the manufacture of nylon. Thc catalyst is composed of bimetallic RuloPtz nanoparticles anchored via two Pt-0 and one Rn-0 bonds (established by X-ray absorption) to mesoporous silica @ore diameter 3nm).

8.

MESOPOROUS AND NON-POROUS SILICA AS CATALYST SUPPORTS: A COMPARISON

Industrial chemists and laboratory researchers have recognized the great advantages of silica (outlined in Section 2 above) for over a hundred years, but it was not until the early 1970s that organometallic chemists began to explore the silica surface as a possible rigid ligand for their deeper study of catalysis and practical exploitation, especially in polymer science [95]. Twenty or so years ago, Basset and co-workers began a sustained, systematic and highly successful series of studies of surface organometallic chemistry [96-981 in which, inter alia, the transfer of concepts and practices of molecular organometallic chemistry were made to well-defined surfaces, socalled Aerosil silica. Organometallic compounds, especially metal carbonyls, anchored to silica surfaces have also been much investigated by Gates and co-workers [96]. Aerosil silica (produced by the Degussa Co.) has a surface area of some 200 mZg-',and, depending upon the temperature of its formation or subsequent treatment, it may have a variable amount of pendant silanol groups. The surfaces of Aerosil silicas heated to 700 OC or so are richer in siloxane linkages (=Si-0-Si=) than in silanols (=SOH). Basset quotes [98] a silanol surface concentration of 0.7 0.2 per nm2 (which is equivalent to 0.23 OH g-l when treated at 700 OC). The first key difference between mesoporous silica and Aerosil, therefore, is the far greater availability of pendant silanols in the former. With a surface area of 800 to 1000 m2g-1, typical mesoporous silicas have approximately ten times as many pendant silanol groups available for creating single-site, immobilized organometallic catalysts than a typical Aerosil. On the other hand, the thermal and general mechanical stability and lesser aptitude to collapse of Aerosil surpasses that of the MCM-41 family, and, to a lesser degree, of the SBA-15 family. But the mesoporous silica (prepared by the Davison division of the W.R. Grace Co.) matches and often exceeds the thermal and mechanical stability of Aerosil - see Section 6.3 C above). The opportunities for exploring single-site heterogeneous catalysis (of silica-functionalized surfaces) for both Aerosil and mesoporous silica are about equal - and both are very considerable. Basset et al. have already demonstrated [97] the especial merits of using non-porous silica for a range of pure and applied catalysis, embracing inter alia olefin polymerisation, olefin metathesis and even low-temperature hydrogenolysis of alkanes (catalyzed by a tripodally grafted Zr centre to the silica surface [98]). The quintessential difference between our own work on mesoporous silica and the previous work on the non-porous silica is that we have the

+

supreme advantage of being able to exploit the (adjustable) pore diameter (and hence the concavity of the surface) so as to achieve enhanced enantioselectivity using immobilized, organometallic chiral catalysts. In addition, we are also able to capitalize upon the greater surface concentration of silanols (with all that that permits for creating single-site heterogeneous catalysts). Mesoporous silicas have made it possible to winnow the grain of understanding from the chaff of overwhelming (often confusing) evidential fact, thereby deepening our knowledge of the fundamentals of heterogeneous catalysis [99] whilst at the same time opening up important new practical applications of the phenomenon. The preparative breakthroughs that led to the ready production of various kinds of mesoporous silica proved crucial. But equally important have been the techniques of catalyst characterization - some well-established but many of them new - that have placed this area of single-site, heterogeneous catalysis on such a firm platform. Our final section outlines these techniques. 9.

A SUMMARY OF THE TECHNIQUES USED CHARACTERIZE MESOPOROUS CRYSTALS CATALYTIC SIGNIFICANCE

TO OF

Adequate descriptions are already available [loo] concerning the standard methods of characterizing mesoporous (and microporous) solids. These embrace the use of gas adsorption isotherms, low-angle and ordinary X-ray diffraction, scattering methods (of neutrons) and the most popular forms of optical, scanning and transmission electron microscopy. Here we are concerned only with those techniques, not in widespread use as yet, that we ourselves have developed principally to elucidate the nature of nanoporous and nanoparticle catalysts. Ex situ methods include all the conventional, as well as some less commonly used spectroscopic procedures: FTIR; Raman; UV-Vis (Difhse-reflectance); mass spectrometry; ESR; multi-nuclear (solid-state) MASNMR (especially of 'H, 2 ~ , 13 C, 2 7 ~ 12, 9 ~and i 3 1 ~ )X-ray ; diffractometry and conventional transmission electron microscopy. We have relied heavily both on conventional high-resolution (HR) transmission electron microscopy [ l o1- 1041 (TEM) (see Fig. 26) and on high-resolution (HR) scanning transmission electron microscopy (STEM) with their allied technique of electron-induced X-ray emission (for analytical

powers down to the zeptogram (10-" g) level) 1105-1071. Thc STEM approach readily enables images to be recorded at high-angles and with an annular dark-field (HAADF), where so-callcd Z-contrast prevails [105-1071. (At high scattering angles, Rutherford scattering dominates - intensity is thus proportional to the square of the atomic number, Z). The HAADF method readily identifies (scc Fig. 23.) a few isolated atoms (of relative high Z) on a light background, such as silica - one atom of platinum produces an electron intensity equal to that of four hundred oxygen atoms.

7% thiolsiloxane la 3d

2-5% thiol p 6mm SEA-15

FDU-5 STA-I I

D. Zhao ef al. Angew. Chem. Int. Ed. 41 (2002) 3876 A.E. Garcia-Bennett (PhD Thesis, St. Andrews, 2002)

Fig. 26. HRTEM is indispensable in identifying (and solving [108]) the structure of new large-pore mesoporous crystals. The precise phase that forms depends of the conditions of crystallization. With 7 percent thiol-siloxanc thc STA-I I phase (identical to FDU-5) forms space group Ia 3d; with 2 to 5 percent thiol, the hexagonal phase (p 6mm). i.e. SBA-15, is formed. (Courtesy of Dr. P.A. Wright, St. Andrews University)

HRSTEM and HRTEM are uniquely well-suited to explore both the average structure and local infractions or other structural irregularities in the bulk and at the surface of mesoporous silica [102,103] (see Figs. 23. and 26). In the hands of Terasaki et a1 [log-1101, the conventional, transmission electron microscope, through the method of electron crystallography, is

capable of solving, de novo and in atomic detail, the structure of new microporous crystals. Tomography is also feasible using STEM HAADF imaging [107,111,112]. In general, this entails reconstructing the three-dimensional structure of an object from an angular series of two-dimensional images (projections). It has enabled us to determine the three-dimensional distribution of bimetallic nanocatalysts within mesopores of silica, and for the elemental composition of each nanoparticle to be evaluated see Fig. 23. In situ methods of characterizing catalysts have been evolved by us over the years - see Refs. [67,68] - and the most important tool deployed by us for this purpose is the combined use of X-ray absorption spectroscopy (XAFS) (embracing near-edge, i.e. XANES as well as extended-edge, i.e. EXAFS structures) and X-ray diffraction (XRD) (see Fig. 27). When the XAFS is recorded in an energy-dispersive mode (as indicated in Fig. 27.) rapid measurements are possible (giving rise to the acronym QUEXAFS - quick EXAFS [I 13,1141). The great merit of XAFS is that it can conveniently identify the immediate chemical environment of all elements {with Z above 10). Bond distances and coordination numbers - as well as valence states and the degree of flexibility or rigidity of the local structure are retrievable this way. The bonus offered by combining XAFS with XRD is that the entire structural integrity of the mesoporous crystal may be directly assessed [I 13,1141. (This combination, XAFS-XRD, proved invaluable in tracking the local environment of c o n ions prior to and during crystallization of a microporous solid from a nutrient gel [I 151.) -

Fig. 27. A set-up of this kind [18] enables the determination of both the immediate atomic environment of the catalytically active site (from XAFS) and the long-range structural integrity of the mesoporous crystal (from XRD) to be recorded in parallel under in situ conditions of catalysis.

In the actual testing of catalytic performance use is made of minirobots for the rapid transfer of minute (0.1 p1) aliquots of the reactantsproducts from high-pressure, PEEK-lined stainless-steel cells to either GC or LC-MS analysers. This enables kinetic data on conversion and growth and decay of intermediates to be rapidly determined. We acknowledge the support of the EPSRC (for a rolling grant to JMT), to the Royal Commission for the Exhibition of 1851 (for a Research Fellowship to RR), to Bayer and BP (for financial support) and to all our colleagues, especially Prof. B.F.G. Johnson and Dr. P.A. Midgley, for their cooperation. REFERENCES J.M. Thomas, Angew. Chem. Int. Ed. Engl., 27 (1988) 353. J. Chen and J.M. Thomas, J. Chem. Soc. Chem. Commun., (1994) 603. R. Raja and P. Ratnasamy, Appl. Catal. A, 158 (1997) L7. A. Zsigmond, F. Notheisz, G. Csjernyik and J.E. Backvall, Top. Catal., 19 (2002) 119. C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli and J.S. Beck, Nature, 359 (1992) 710. S. Inagaki, Y. Fukushima and K. Kuroda, J. Chem. Soc. Chem. Commun., (1993) 680; see also S. Inagaki, A. Koiwai, N. Suzuki, Y. Fukushima and K. Kuroda, Bull. Chem. Soc. Jpn., 69 (1996) 1449. Q. Huo, D. Margolese, U. Ciesla, P.Y. Feng, T. Geir, P. Sieger, R. Lyon, P.M. Petroff, F. Schiith and G.D. Stucky, Nature. 368 (1994) 3 17. J.M. Thomas, Nature, 368 (1994) 289. P.T. Tanev, M. Chibwe and T.J. Pinnavaia, Nature, 368 (1994) 321. A. Coma, M.T. Navarro and J.P. Pariente, J. Chem. Soc. Chem. Commun., (1994) 147. G. Sankar, F. Rey, J.M. Thomas, G.N. Greaves, A. Coma, B.R. Dobson and A.J. Dent, J. Chem. Soc. Chem. Commun., (1994) 2279. B. Notari, in P.J. Grobet, J.A. Martens and P.A. Jacobs (eds.), Innovation in Zeolite Materials Science, Elsevier, Amsterdam, 1987, pp. 143. G. Bellussi , C. Perego, A. Carati, S. Peratello, E.P. Massara and G. Perego, Stud. Surf. Sci. Catal., 84 (1994) 85. F. Rey, G. Sankar, T. Maschrneyer, J.M. Thomas, R.G. Bell and G.N. Greaves, Top. Catal., 3 (1996) 121. G. Bellussi, Lecture given at EuropaCat-VI, Innsbruck, Austria, September (2003). J.M. Thomas, Sci. Am., 266 (1992) 112. T. Maschmeyer, F. Rey, G. Sankar and J.M. Thomas, Nature, 378 (1995) 159. J.W. Couves, J.M. Thomas, D. Waller, R.H. Jones, A.J. Dent, G.E. Derbyshire and G.N. Greaves, Nature, 354 (199 1) 465. J.M. Thomas and G. Sankar, Acc. Chem. Res., 34 (2001) 571.

R.A. Sheldon, J. Mol. Catal., 20 (1983) 1. F. Wattimena and H.P Wulff, GB Patent No. 1 249 079 (1970). H.P. Wulff, US Patent No. 3 923 843 (1975). T. Maschmeyer, R.D. Oldroyd, G. Sankar, J.M. Thomas, I.J. Shannon, J.A. Kleptko, A.F. Masters, J.K. Beattie and C.R.A. Catlow, Angew. Chem. Int. Ed. Engl., 36 (1997) 1639. W.M. Van Rhijn, D.E. De Vos, B.F. Sels, W.D. Bossaert and P.A. Jacobs, J. Chem. Soc. Chem. Commun., (1998) 3 17. R. Raja and J.M. Thomas, J. Chem. Soc. Chem. Commun., (1998) 1841. J.M. Thomas, R. Raja, G. Sankar and R.G. Bell, Acc. Chem. Res., 34 (2001) 191. J.M. Thomas, R. Raja, G. Sankar and R.G. Bell, Nature, 398 (1999) 227. R. Raja, G. ~ a n k a and r J.M. Thomas, Angew. Chem. Int. Ed. Engl., 39 (2000) 2313. C.L. Hill (ed.), Activation and Functionalization of Alkanes, Wiley, Chichester, 1989. S. Kemura, A. Spencer and G. Wilkinson, J. Chem. Soc. Dalton T., (1973) 2565. D. Brunel, A. Carvel, F. Fajula and F. Di Renzo, Stud. Surf. Sci. Catal., 97 (1995) 173. R.D. Oldroyd, J.M. Thomas, T. Maschmeyer, P.A. MacFaul, D.W. Snelgrove, K.U. Ingold and D.D.M. Wayner, Angew. Chem. Int. Ed. Engl., 35 (1996) 2787. J.M. Thomas, C.R.A. Catlow and G. Sankar, Chem. Commun., (2002) 292 1. G. Sankar, J.M. Thomas, C.R.A. Catlow, C.M. Barker, D. Gleeson and N. Kaltsoyannis, J. Phys. Chem. B., 105 (2001) 9028. J.M. Thomas, Faraday Discuss., 105 (1996) 1. L. Marchese, E. Gianotti, V. Dellarocca, T. Maschmeyer, F. Rey, S. Coluccia and J.M. Thomas, Phys. Chem. Chem. Phys., 1 (1999) 585. R.D. Oldroyd, G. Sankar, J.M. Thomas and D. Ozkaya, J. Phys. Chem. B., 102 (1998) 1849. J.M. Thomas, G. Sankar, M. Klunduk, M.P. Attfield, T. Maschmeyer, B.F.G. Johnson and R.G. Bell, J. Phys. Chem. B., 103 (1999) 8809. M.P. Attfield, G. Sankar and J.M. Thomas, Catal. Lett., 70 (2000) 155. J. Jarupatrakorn and T.D. Tilley, J. Am. Chem. Soc., 124 (2002) 8380. M.P. Coles, C.G. Lugmair, K.W. Teny and T.D. Tilley, Chem. Mater., 12 (2000) 122. T.D. Tilley, J. Mol. Catal. A., 182-183 (2002) 17. K.L. Fujdala and T.D. Tilley, J. Catal., 216 (2003) 265. M. Guidotti, N. Ravasio, P. Psaro, S. Coluccia, L. Marchese and E. Gianotti, Green Chem., 5 (2003) 42 1. N. Ravasio, F. Zaccheria, M. Guidotti and P. Psaro, Top. Catal., 27 (2004) 157. J. He, H. Ma, Z. Guo, D.G. Evans and X. Duan, Top. Catal., 22 (2003) 41. I.J. Shannon, T. Maschmeyer, R.D. Oldroyd, G. Sankar, J.M. Thomas, H. Pernot, J.P. Balikdjian and M. Che, J. Chem. Soc. Faraday T., 94 (1998) 1495. R.D. Oldroyd, G. Sankar, J.M. Thomas, M. Hunnius and W.F. Maier, J. Chem. Soc. Faraday T., 94 (1998) 3177. Z. Shan, J.C. Jansen, L. Marchese and T. Maschmeyer, Micropor. Mesopor. Mat., 48 (2001) 181. R. Rulkens and T.D. Tilley, J. Am. Chem. Soc., 120 (1998) 9959.

R. Rulkens, J.L Male, K.W. Keny, B. Olthof, A. Khodakov, A.T. Bell, E. Iglesia and T.D. Tilley, Chem. Mater., 11 (1999) 2966. J.L. Male, H. Niessen, A.T. Bell and T.D. Tilley, J. Catal., 194 (2000) 43 1. M.P. Coles, R. Rulkens and T.D. Tilley, J. Chem. Soc. Dalton, (2000) 627. K.L. Fujdala and T.D. Tilley in N. Auner and J. Weis (eds.), Organosilicon Chemistry V - From Molecules to Materials, Wiley-VCH, Weinheim, 2003, pp. 379. C. Pak, A.T. Bell and T.D. Tilley, J. Catal., 206 (2002) 49. C. Nozaki, C.G. Lugmair, A.T. Bell and T.D. Tilley, J. Am. Chem. Soc., 124 (2002) 13194. K.L Fujdala and T.D. Tilley, J. Catal., 218 (2003) 123. A.T. Bell, Science, 299 (2003) 1688. M. Nowotny, T. Maschmeyer, B.F.G. Johnson, P. Lahuerta, J.M. Thomas and J.E. Davies, Angew. Chem. Int. Ed. Engl., 40 (2001) 955. V.A. Likholobov, in J.M. Thomas and K.I. Zamaraev (eds.), Perspectives in Catalysis, Blackwells and IUPAC, 2000, pp. 71. R. A. Sheldon and H, van Bekkum (eds.), Fine Chemicals Through Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001. C. Bianchini and P. Barbaro, Top. Catal., 19 (2002) 17. H. Brunner, in B. Cornils and W.A. Herrmann (eds.), Applied Homogeneous Catalysis with Organometallic Compounds, Vol. 1, Wiley-VCH, Weinheim, 1996, pp. 206. J.M. Thomas, T. Maschmeyer, B.F.G. Johnson and D.S. Shephard, J. Mol. Catal. A., 141 (1999) 131. T. Maschmeyer, Lecture at Davy Faraday Research Laboratory, London, March 1995. J.M. Thomas, Faraday Discuss., 100 (1995) C9. (a) J.M. Thomas, Angew. Chem. Int. Ed. Engl., 38 (1999) 3588; (b) J.M. Thomas and R. Raja, J. Chem. Soc. Chem. Commun., (2001) 675. J.M. Thomas, Chem.-Eur. J., 3 (1997) 1557. B.F.G. Johnson, S.A. Raynor, D.S. Shephard, T. Maschmeyer, J.M. Thomas, G. Sankar, S.T. Bromley, R.D. Oldroyd, L. Gladden and M.D. Mantle, J. Chem. Soc. Chem. Cornmun., (1999) 1167. S.A. Raynor, J.M. Thomas, R. Raja, B.F.G. Johnson, R.G. Bell and M.D. Mantle, J. Chem. Soc. Chem. Commun., (2000) 1925. R.M. Laine, G. Hum, B.J. Wood and M. Dawson, Stud. Surf. Sci. Catal., 7 (1981) 1478. H.U. Blaser, H. Honig, M. Studer and C. Wedemeyer-Exl, J. Mol. Catal. A., 139 (1999) 253. (a) F.J. Fehler, D.A. Newman and J.F. Walzer, J. Am. Chem. Soc., 111 (1989) 1741; (b) T. Maschmeyer, J.M. Thomas and A.F. Masters, NATO Adv. Series, 498 (1997) 46 1. M.D. Jones, R. Raja, J.M. Thomas, B.F.G. Johnson, D.W. Lewis, J. Rouzard and K.D.M. Harris, Angew. Chem. Int. Ed. Engl., 42 (2003) 4326. J. Rouzard, M.D. Jones, R. Raja, B.F.G. Johnson, J.M. Thomas and M.J. Duer, Helv. Chim. Acta., 86 (2003) 1753. M.D. Jones, R. Raja, J.M. Thomas and B.F.G. Johnson, Top. Catal., 25 (2003) 71.

(a) A. Trogni and L. Venanzi, Angew. Chem. Int. Ed. Engl., 106 (1994) 517; (b) F. Fache, E. Schultz, M.L. Tommasino and M. Lemaire, Tetrahedron-Asymmetr., 10 (1999) 1813. R. Raja, J.M. Thomas, M.D. Jones, B.F.G. Johnson and D.E.W. Vaughan, J. Am. Chem. Soc., 125 (2003) 14982. These values were obtained from the dV/d log(D) plots pertaining to the desorption isotherm of Nz on the silica. F.M. de Rege, D.K. Morita, K.C. Ott, W. Tumas and R.D. Broene, J. Chem. Soc. Chem. Commun., (2000) 1797. C. Bianchini, D.G. Burnaby, J. Evans, P. Frediani, A. Meli, W. Oberhauser, R. Psaro, L. Sordelli and F. Vizza, J. Am. Chem. Soc., 121 (1999) 5961. (a) R. Raja, J.M. Thomas, M.D. Jones and B.F.G. Johnson, Process for Reducing ketocarboxylic esters, German Patent Filing No. DE 1 030 5946 (2003); (b) R. Raja, J.M. Thomas, M.D. Jones and B.F.G. Johnson, Process for Preparing Enantiomerically Enriched a-P- hydroxycarboxylic esters, German Patent Filing No. DE 1 030 5943. (a) M.S. Nashner, A.I. Frenkel, D.M. Somerville, C.W. Hills, J.R. Sharpley and R.G. Nuzzo, J. Am. Chem. Soc., 120 (1998) 8093; (b) C.W. Hills, M.S. Nashner, A.I. Frenkel, J.R. Sharpley and R.G. Nuzzo, Langmuir, 15 (1999) 690. D.S. Shephard, T. Maschmeyer, J.M. Thomas, B.F.G. Johnson, G. Sankar, D. Ozkaya, W. Zhou, R.D. Oldroyd and R.G. Bell, Angew. Chem. Int. Ed. Engl., 36 (1997) 2242. S. Hermans, R. Raja, J.M. Thomas, B.F.G. Johnson, G. Sankar and D. Gleeson, Angew. Chem. Int. Ed. Engl., 40 (2001) 1211. R. Raja, T. Khimyak, J.M. Thomas, S. Hermans and B.F.G. Johnson, Angew. Chem. Int. Ed. Engl., 40 (2001) 4638. J.M. Thomas, B.F.G. Johnson, R. Raja, G. Sankar and P.A. Midgley, Acc. Chem. Res., 36 (2003) 20. J.M. Thomas, R. Raja, B.F.G. Johnson, S. Hermans, M.D. Jones and T. Khimyak, Ind. Eng. Chem. Res., 42 (2003) 1563. D. Ozkaya, W. Zhou, J.M. Thomas, V.J. Keast, P.A. Midgley and S. Hermans, Catal. Lett., 60 (1999) 101. J.M. Thomas, R. Raja, G. Sankar, B.F.G. Johnson and D.W. Lewis, Chem.-Eur. J., 7 (2001) 2973. J.H. Sinfelt, Intl. Rev. Phys. Chem., 7 (1988) 281. J.M. Thomas, R. Raja, B.F.G. Johnson, T.J. O'Connell, G. Sankar and T. Khimyak, J. Chem. Soc. Chem. Commun., (2003) 1126; see also Science, 300 (2003) 867. K.M. Draths and J.W. Frost, J. Am. Chem. Soc., 116 (1994) 399. R. Raja, V. Golovko, J.M. Thomas and B.F.G. Johnson, manuscript in preparation. (a) D.G.H. Ballard, Adv. Catal., 23 (1973) 263; (b) J.P. Candlin and H. Thomas, Adv. Chem. Ser., 132 (1974) 212. (a) J.M. Basset, A. Choplin, J. Mol. Catal., 21 (1983) 95; (b) J.M. Basset, J.P. Candy, A. Choplin, B. Didillon, F. Quignard and A. Theolier, in J.M. Thomas and K.I. Zamaraev (eds.), Perspectives in Catalysis, Blackwells and IUPAC, Oxford, 1992, pp. 125; (c) see also B.C. Gates, in L.L. Hegedus, A.T. Bell, N.Y. Chen,

W.O. Haag, J. Wu, R. Anes, M. Boudart and G.A. Somorjai (Eds.), Catalyst Design: Progress and Perspectives, J. Wiley, 1987, pp. 7 1. C. CopCret, M. Chabanas, R.P. Saint-Arroman and J.M. Basset, Angew. Chem. Int. Ed. Engl., 42 (2003) 156 and references therein. C. LCcuyer, F. Quignard, A. Choplin, D. Olivier and J.M. Basset, Angew. Chem. Int. Ed. Engl., 30 (1991) 1660. (a) J.M. Thomas and W.J. Thomas, Principles and Practice of Heterogeneous Catalysis, Wiley-VCH, Weinheim, 1997; (b) R.A. van Santen, P.W.N.M. van Leeuwen, J.M. Moulijn and B.A. Averill (eds.), Catalysis: An Integrated Approach, NIOK, The Netherlands, 2ndEdition, 1999. F. Schiith, K.S.W. Sing and J. Weitkamp (eds.), Handbook of Porous Solids, Vol. 1-5, Wiley-VCH, 2002. See, in particular, the sections written by B. Tesche (microscopy), J.D.F. Ramsey (scattering) and F. and J. Rouqueral (adsorption methods). J.M. Thomas, in M.H. Chisholm (ed.), Inorganic Chemsitry: Towards the 2lSt Century, ACS Symposium Series, 21 1 (1983) pp. 445. J.M. Thomas, Ultramicroscopy, 8 (1982) 13. J.M. Thomas, 0 . Terasaki, P.L. Gai, W. Zhou and J.M. Gonzalez-Calbert, Acc. Chem. Res., 34 (2001) 583. J.M. Thomas and P.A. Midgley, J. Chem. Soc. Chem. Commun., (2004) in press. W. Zhou and J.M. Thomas, Curr. Opin. Solid State Mat. Sci., 5 (2001) 75. J.M. Thomas and W. Zhou, Chem. Phys. Chem., 4 (2003) 927. P.A. Midgley, J.M. Thomas, R. Raja, L. Laffont, B.F.G. Johnson and T. Khimyak, J. Phys. Chem. B., (2004) in press. A. Carlsson, M. Kaneda, Y. Sakomoto, 0 . Terasaki, R. Ryoo and H. Joo, J. Electron Microscopy, 48 (1999) 795. Y. Sakomoto, M. Kaneda, 0 . Terasaki, D.Y. Zhao, J.M. Kim, G.D. Stucky, H.J. Shim and R. Ryoo, Nature, 408 (2000) 499. 0 . Terasaki and J.M. Thomas, Top. Catal., 21 (2002)155. P.A. Midgley, J.M. Thomas, M.A. Weyland and B.F.G. Johnson, J. Chem. Soc. Chem. Commun., (2001) 907. M.A. Weyland, P.A. Midgley and J.M. Thomas, J. Phys. Chem. B., 105 (2001) 7882. G. Sankar, P.A. Wright, S. Natarajan, J.M. Thomas, G.N. Greaves, A.J. Dent, B.R. Dobson and R.H. Jones, J. Phys. Chem., 97 (1993) 9550. B.S. Clausen, H. Topsoe and R. Frahm, Adv. Catal., 42 (1998) 315. G. Sankar, J.M. Thomas, F. Rey and G.N. Greaves, J. Chem. Soc. Chem. Commun., (1995) 2549.

This Page Intentionally Left Blank

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Evaporation-induced nanostructures

self-assembly

to

functional

Hongyou Fan and Jeffrey Brinker

Sandia National Laboratories, Chemical Synthesis and Nanomaterials Department, Advanced Materials Laboratory, 1001 University Blvd. SE, Albuquerque, NM 87 106. The University of New Mexico/NSF Center for Micro-Engineered Materials, Department of Chemical and Nuclear Engineering, Albuquerque, NM 87 131. Self-assembly employs pre-existing components "pre-programmed" to spontaneously organize under the influence of normally non-covalent bonding interactions such as hydrogen bonding, n-n: interactions, hydrophobic interactions, electrostatic forces, capillarity, etc. For example, amphiphilic molecules composed of hydrophilic and hydrophobic parts when added to water plus oil organize into micelles and, at higher concentrations, spatially extensive periodic networks with hexagonal, cubic, bicontinuous, or lamellar arrangements. Since the pioneering efforts of Kresge and co-workers[l], it has been understood that amphiphile selfassembly conducted in aqueous solutions of hydrophilic precursors such as silicic acid results in composite inorganictsurfactant mesophases[2-51. Calcination or extraction of the amphiphilic surfactant assemblies then yields so-called mesoporous materials characterized by periodic networks of monosized pores with 2- or 3-D connectivities. Although the original MCM (Mobil Composition of Matter) materials were mesoporous silica powders, a significant effort has been made during the past 12 years to prepare mesoporous films (Ogawa, etc.[6]), Spherical and other well-defined shapes[7-lo], and hierarchical assemblies (Liu, et al.[l I]). It is also now recognized that amphilphile self-assembly conducted in the presence of both hydrophilic inorganic and hydrophobic organic precursors provides a viable route to the formation of hybrid inorganictorganic nanocomposites[l2]. Evaporation-Induced Self-Assem bly (EISA)

Although films can be formed under the same batch conditions used to form mesoporous precipitates, in such cases, heterogeneous nucleation and growth leads to non-uniform nodular structures[l3- 151. It is significant

therefore that starting with dilute solutions of silicic acid plus surfactants (often prepared in alcohol/water solvents) solvent evaporation can drive the continuous self-assembly of homogeneous highly ordered thin film mesophases[l6]. This evaporation induced self-assembly (EISA) process has been employed to prepare porous and composite thin film nanostructures, as well as nanostructured spherical particles (using an aerosol-assisted EISA process)[l7]. Here we summarize several recent advances made since our earlier overview[l7]: 1) in-situ characterization of EISA. 2) Elaboration of EISA to patterned, multifunctional nanostructures, and 3) EISA of environmentally responsive composite nanostructures. 1. In situ Characterization of EISA

EISA starts with a homogeneous solution of soluble silica (oligo silicic acids) plus surfactant prepared normally in an ethanollwater solvent at an initial surfactant concentration less than the critical micelle concentration, CMC, and with an acid concentration (-0.01M) designed to minimize the siloxane condensation rate[l8]. During film[l6] or particle[l9] formation, preferential evaporation of ethanol and then water concentrates the system in surfactant and silica resulting in the self-assembly of micelles and their further organization into mesophases. Rather than random heterogeneous nucleation, the steep concentration gradient established by solvent evaporation results in an interfacially-mediated self-assembly process that proceeds from the liquid-vapor interface along concentration gradient contours. EISA is technologically important because it enables rapid efficient integration of well-defined nanostructures into microelectronic devices and microsystems using readily available processing methods. However understanding and ultimately control of self-assembly (directed assembly) is a critical need for its emergence from largely a laboratory practice to a reliable nanofabrication "tool". Recently, several in situ investigations[20251 have been conducted to monitor EISA with the aim of elucidating the self-assembly pathway and correlating mesostructure formation with compositional variations. Spatially resolved spectroscopic studies[l6, 26-28] and time resolved[20, 22, 231 and spatially resolved[25] grazing incidence small angle X-ray scattering (GISAXS) experiments performed in situ during dip-coating have established the critical micelle concentration and have observed the evolution of thin-film mesophases during EISA. These experiments have employed (i) non-steady-state coating conditions[20-231 in which the film is

draining and evaporating or (ii) steady state dip-coating conditions[25], but both have relied on optical interferometry[20-23, 251, which is limited in resolution to -A/2 (where A is the wavelength of the monochromatic light source) to monitor film thickness. The non-steady-state conditions[21] do not allow determination of composition from the observed film thickness. Under steady-state conditions, calculation of compositional changes occurring near and beyond the "drying line" (solid-liquid-vapor interface) where mesophase formation and transformations have been observed, is precluded by the limited resolution of optical interferometry. Thus these experiments could not correlate mesophase formation with compositional variation and thereby make valid comparisons with the phase behavior of bulk surfactantJwater/alcohol systems necessary to establish how silica and the presence of interfaces influence the self-assembly pathway. To derive structural and compositional information simultaneously, we recently combined GISAXS (X22B, National Synchrotron Light Source, Brookhaven National Laboratory) with gravimetric analysis to study selfassembly of a slowly evaporating film maintained in a horizontal geometry under controlled environmental conditions. Figure 1 shows a schematic of the setup and reciprocal space probed during the experiments. Samples prepared in ethanollwater solvents with different surfactant (CTAB)/Si ratios were investigated and compared to the corresponding samples prepared without silica (WS). Figure 2 shows how the system weight, and correspondingly, surfactant concentration change with time, where Regions I-IV correspond to the isotropic, lamellar, correlated micellar, and hexagonal mesostructural domains, respectively, based on the GISAX results (see Figure 4). Also plotted is the d-spacing of the lamellar, correlated micellar, and hexagonal mesophases. We attribute the steep initial weight loss to evaporation of ethanol and, beyond 600 s, the slower weight loss to evaporation of water. This allows us to plot the evaporation trajectories on the ethanol/water/surfactant phase diagram (Figure 3), where we consider water and the hydrophilic silicic acid precursor to be equivalent. Figure 4 shows GISAXS patterns corresponding to compositions a-f in Figure 3. This system evolves from an isotropic state (point a) through an interfaciallyoriented lamellar phase (b) and correlated micellar phase (c) to the final 2D hexagonal mesotructure (e and 0.Unanticipated from the phase diagram is the lamellar phase, which based on its orientation parallel to the substrate, we attribute to a transient mesophase that forms due to ethanollwater evaporation at the filmlvapor interface. The ensuing correlated micellar phase represents spherical or cylindrical micelles that organize into a proto-

hexagonal from which the hexagonal mesophase forms. The first appearance of the 2D hexagonal mesophase corresponds nearly to the hexagonal phase boundary for the ethanol/water/surfactant system suggesting that silicic acid does serve as a hydrophilic component similar to water. Notably the WS system prepared without silica does not form a hexagonal mesophase, presumably due to kinetic constraints associated with the removal of water. Performing EISA at -pH2 where siloxane condensation is suppressed maintains the system in a fluid state avoiding kinetic barriers to selfassembly. Figure 2 indicates that for the lamellar, correlated micellar and hexagonal mesophases the corresponding d-spaces increase with time. This dilation results from ethanol evaporation. Monte Carlo simulations of lamellar mesophases in bulk ethanollwaterl surfactant systems show ethanol molecules to be located adjacent to the surfactant headgroups with their hydrocarbon tails oriented toward the hydrophobic micellar core[25]. This arrangement promotes surfactant tail interdigitation in the lamellar mesophase due to favorable hydrophobic interactions[25]. Evaporation of ethanol results in an unfavorable interaction between the hydrophobic surfactant tails and the hydrophilic waterlsilica rich solvent, causing the bilayer spacing to increase (extent of interdigitation decreases) as we observed for both the silica containing and WS samples. Similarly, within cylindrical micelles, ethanol evaporation causes the surfactant tails to rearrange to become more radially oriented with respect to the cylinder axis. This causes an expansion of the micelle diameter. Our observation of a continuous reorganization of both mesostructure and lattice dimension within regions 11-IV emphasizes the need to suppress the siloxane condensation rate to allow self-assembly to proceed unimpeded. Table 1 summarizes the evolution of structure with time during evaporation.

2. EISA of patterned functional nanostructures Since the discovery of surfactant-templated silica mesophases[l], considerable effort has been devoted to the development of molecular-scale, organic modification schemes to impart useful functionality to the pore surfaces[29-341. Concurrently a variety of patterning strategies have been developed to define macroscopically the shapes of deposited thin-film mesophases and their locations on the substrate surface[35-371. Our recent work combined molecular-scale EISA of organically modified mesophases with macroscopic, evaporative printing procedures (micrio-pen lithography (MPL), ink-jet printing (IJP), and p-molding)[38]. We demonstrated the rapid fabrication of patterned structures exhibiting form and function on

multiple length scales and at multiple locations. At the molecular scale, functional organic moieties (Table 2) are positioned on pore surfaces; on the mesoscale, mono-sized pores are organized into one-, two- or threedimensional networks, providing size-selective accessibility from the gas or liquid phase; and on the macroscale, two-dimensional (2D) arrays and fluidic or photonic systems are developed. A second patterning scheme involved the co-self-assembly of amphiphilic photoacid generator molecules with surfactant and silica to form photosensitive thin film mesophases.[391. Patterning by ultraviolet exposure promoted localized acid-catalyzed siloxane condensation, enabling selective etching of unexposed regions. Standard calcinations resulted in thin film mesophases in which the refractive index, dielectric constant, pore size, and surface area depended on the UV exposure. Such materials merge top-down and bottom-up processing and provide a means to optically tune structure and function. Figure 5a shows a macroscopic pattern formed in several seconds by MPL of a rhodamine-B-containing solution on a hydrophilic surface. The inset in Fig. 5a shows the corresponding fluorescence image of several adjacent stripes acquired through a 610-nm bandpass filter, demonstrating retention of rhodamine-B functionality, and the transmission electron microscopy (TEM) image (Fig. 5b) reveals the ordered pore structure characteristic of a cubic thin-film mesophase. The MPL line width can vary from micrometers to millimeters. It depends on such factors as pen dimension, wetting characteristics, evaporation rate, capillary number (Ca = ink viscosity x substrate speedlsurface tension) and ratio of the rates of ink supply and withdrawal (inlet velocity/substrate velocity). The advantages of MPL are that we can use any desired combination of surfactant and functional silane as ink to print selectively different functionalities at different locations. Furthermore, we can use computeraided design (CAD) to define arbitrary 2D patterns that can be written on arbitrary surfaces. For example, we have demonstrated writing rhodaminecontaining mesophases (refractive index n = 1.2-1.3) on aerogel[40] and emulsion-templated thin films (n = 1.03-1. lo), thereby directly defining optical waveguide structures potentially useful for lasing[41]. Figure 6b shows an optical micrograph of an array of hydrophobic, mesoporous spots formed on a silicon substrate by IJP of a TFTS-modified ink. The IJP process dispenses the ink (prepared as for MPL) as monosized, spherical aerosol droplets. On striking the surface, the droplets adopt a new shape that balances surface and interfacial energies. Accompanying evaporation creates within each droplet a gradient in surfactant concentration that drives radially directed silica-surfactant self-assembly

inward from the liquid-vapour interface. The TEM micrograph (Fig. 5c) shows the ordered mesoporosity of a calcined, fluoroalkylated silica mesophase formed by IJP. The link to computer-aided design, greater printing resolution achieved compared to standard ink (see Fig. 5a), and our ability to selectively functionalize the ink, suggest applications in sensor arrays and display technologies. Figure 7 illustrates dip-coating on patterned SAMs. This rapid, parallel procedure uses micro-contact printing[42] or electrochemical patterning[43] of hydroxyl- and methyl-terminated SAMs to define hydrophilic and hydrophobic patterns on the substrate surface. Then, using inks identical to those employed for MPL and IJP, preferential ethanol evaporation during dip-coating enriches the depositing film in water, causing selective de-wetting of the hydrophobic regions and ensuring self-assembly of silica-surfactant mesophases exclusively on the hydrophilic patterns. In this fashion, multiple lines, arrays of dots, or other arbitrary shapes, can be printed in seconds[38]. Figure 7 also illustrates the formation of a patterned propylaminederivatized cubic mesophase by selective de-wetting followed by calcination to remove the surfactant templates (the organosilane used was aminopropyltrimethoxysilane - compound 3 in Table 2.) The TEM micrograph (inset A) and N,-sorption isotherm (based on surface acoustic wave, SAW, measurements) confirm the 3D accessibility of the mesopores. In order to make a pH-sensitive fluidic system, the covalently bound propylamine ligands were conjugated with a pH-sensitive dye, 5,6- carboxy fluorescein, succinimidyl ester (5,6-FAM, SE) introduced in the porechannel network of the cubic mesophase. After removal of any noncovalently-bonded dye, the uniform, the patterned, dye-conjugated array was used to monitor the pH of fluids introduced at terminal pads and transported by capillary flow into an imaging cell. Photosensitive thin film mesophases were prepared by adding molecular photoacid generators (PAC) (Figure 8) into initial sols[39]. By exploiting the pH sensitivity of both the siloxane condensation rate and the silica-surfactant self-assembly process, we were able to define film location, mesostructure, and properties. The procedure begins with a homogeneous solution of silica, surfactant, PAG (a diaryliodonium salt), and HCI, with initial acid concentration designed to minimize the siloxane condensation rate. Preferential ethanol evaporation during dip- or spin-coating concentrates the depositing solution in water and nonvolatile constituents, thereby promoting self-assembly into a photosensitive, 2D-hexagonal silicasurfactant mesophase. Because it bears a long-chain hydrocarbon, the PAG

serves as a co-surfactant during the assembly process, which promotes its uniform incorporation within the mesostructured channels of the 2D-H film. Irradiation of the PAG at a maximum wavelength of 256 nm (reaction 1) results in homolytic or heterolytic photodecomposition to yield the Bronsted superacid, H'SbF6-, plus an iodoaromatic compound and organic byproduct. Thus ultraviolet (UV) exposure of the photosensitive mesophase through a mask creates patterned regions of differing acid concentrations compartmentalized within the silica mesophase (Fig. 9). Suppression of the siloxane condensation rate during film deposition enables several modes of optically mediated patterning. Because acid generation promotes siloxane condensation, selective UV exposure results in patterned regions of more and less highly condensed silica. Differential extents of siloxane condensation result in turn in differential solubility, allowing selective etching of more weakly condensed regions in aqueous base (0.2 M NaOH) (scheme 1 of Fig. 9). An optical micrograph of a UVexposed and etched thin film mesophase after calcination to remove the surfactant templates (Fig. IOB) reveals that the film is present only in the exposed regions. The plan-view transmission electron microscopy (TEM) image (Fig. 10B, inset) reveals a striped mesoscopic structure consistent with a 1-dH mesophase. If the films are heated rather than etched, we find the densities of the UV-exposed regions to be greater than the unexposed regions (Fig. 9 scheme 2) - the extent of this difference depending on the UV exposure time. This 'gray-scale' patterning effect is explained in part by the greater volatility of the photolyzed PAG, which serves as a swelling agent. PAG volatilization from UV-exposed regions causes the pore volume, pore size, and surface area to decrease relative to the adjoining unexposed regions. If the system is prepared near a mesophase boundary UV exposure can alternatively results in a hexagonal to cubic transition (Fig. 9 Scheme 3). 3. Self-assembly of environmentally-responsive composite nanostructures The surfactant templating technique provides a versatile pathway to the synthesis of mesoporous metal oxides with high surface areas and controlled pore size. The resulting materials demonstrate great potential in catalysis, separation, adsorption, and sensing. To expand the range of applications as well as create new properties, various organic functional groups have been covalently incorporated onto the pore surfaces of mesoporous materials, including phenyl, octyl, aminoalkyl, cyanoalkyl, thioloalkyl, epoxyl, and vinyl groups (discussed above). However, to date, these modifications have provided mainly "passive" functionality, such as controlled wetting

properties, reduced dielectric constants, or enhanced adsorption of metal ions. By comparison, materials with "active" functionality would enable properties to be dynamically controlled by external stimuli, such as pH, temperature, or light. Here we summarize our recent work on fabrication of nanocomposite materials with "active" functionality.

3.1 Thermoresponsive nanocomposite multilayered thin films Polymers such as poly(methacry1ates) and poly(N-isopropylacrylamide) (PNIPAM) show a pronounced response toward changes in pH and temperature, respectively[44]. In water, PNIPAM exhibits a phase transition at the lower critical solution temperature (LCST) of approximately 32 "C. This temperature can be controlled through random copolymerization of NIPAM with methacrylic acid[451. Below the LCST, the hydrogel incorporates water and swells, whereas water release at higher temperatures causes shrinkage. The (de)swelling of these films upon change of pH or temperature is of interest for controlled release of molecules or for membranes with switchable permeabilities. Previously, we demonstrated that evaporation-induced self-assembly (EISA) can organize hydrophilic, inorganic and hydrophobic, organic precursors into ordered layered nanostructures[46]. We extended this approach to form films with pH- and temperature-responsive nanostructures[47]. Nanocomposite films were prepared by evaporationinduced self-assembly followed by polymerization and washing as depicted in Figure 11. EISA begins with a homogeneous solution of silicic acid, monomers (NIPAM and/or dodecyl methacrylate), surfactant (cetyltrimethylammonium bromide, CTAB), coupling agent (trimethoxy( 7octen-1-yl)silane, 7-OTS, featuring an alkoxysilane headgroup and a polymerizable double bond), and thermal initiator, 1 1 azobis(1cyclohexanecarbonitrile) (ACHN). Initiators such as persulfates were incompatible with the procedure. A heat treatment (120 "C for 3 h in N,) is then employed to initiate the free-radical polymerization of the confined monomers and promote the condensation of the silica framework. Finally, the films are washed successively in ethanol and acetone to remove the surfactant and any unreacted monomers. During EISA, the organic monomers partition within the hydrophobic domains of a lamellar mesophase. In-situ polymerization results in a 1-2 nm thick hydrogel phase sandwiched between layers of silica oriented parallel to the substrate surface. The thermoresponsiveness of PNIPAM is preserved in this confined environment, and the polymeric layers reversibly swell and deswell by a factor of two in water upon temperature changes around the transition

temperature of PNIPAM (32 "C). Figure 12 illustrates the swelling and deswelling behavior of resulting films and their mesostructures.

3.2 Photoresponsive nanocomposite thin film Photoresponsive thin-film nanocomposites[48] were fabricated by cocondensing a novel photoresponsive azobenzene-containing organosilane, 4(3-triethoxysilylpropylureido) azobenzene (TSUA) with tetraethyl orthosilicate (TEOS) to form an ordered, periodic silica framework. Azobenzene derivatives were selected because of their well-studied response to light[49-521. UV irradiation of the trans isomer causes transformation to the cis isomer. Removal of the UV radiation, heating, or irradiation with a longer wavelength switches the system back to the trans form. TSUA was mixed with tetraethyl orthosilicate (TEOS) in a homogeneous ethanollwater solution with an initial surfactant concentration less than the critical micelle concentration (cmc) and an acid concentration designed to minimize the siloxane condensation rate, thereby avoiding premature gelation, which would frustrate the self-assembly process. During EISA, ethanol evaporation accompanying spin or dip coating enriches the depositing film in water, surfactant, and silica constituents, which results in the self-assembly of silica-surfactant micelles and their further organization into liquid-crystalline mesophases. For TSUA-containing sols, we suspect that the amphiphilic nature of the hydrolyzed TSUA molecule will position the hydrophobic propylureidoazobenzene groups in the hydrophobic micellar cores and that the silicic acid groups will be co-organized with hydrolyzed TEOS moieties at the hydrophilic micellar exteriors. In this fashion, TSUA is ultimately incorporated onto the pore surfaces with the azobenzene ligands disposed towards the pore interiors. A subsequent solvent extraction results in a mesoporous silica framework derivatized with azobenzene ligands, which are isomerizable by light and thermal stimuli (Figure 13). The dimensional change of the propylureidoazobenzene ligand associated with this isomerization mechanism is estimated to be 3.4 A based on molecular modeling with Chem3D Pro 5.5 software (Figure 13). UVIVis spectroscopy was used to characterize the photo and thermal responsiveness of the incorporated azobenzene ligands. Figure 14 shows the UVIVis spectra of thin films after exposure to varying conditions of UV irradiation (8W, h= 302 nm), room light, or heat.

3.3 Self-assembly of ordered thermal-, mechanical- and chemicalchromatic polydiacetylenelsilica nanocomposites Above, we discussed nanocomposites synthesized by EISA that exhibited

a single specific functional response to thermal or light stimuli. In this third approach, we demonstrate the synthesis of multifunctional, conjugated polymer/silica nanocomposite films using polymerizable amphiphilic diacetylene molecules as both structure-directing agents and monomers(53, 541. Through synthesis of a family of polymerizable diacetylene (DA) surfactants with a systematic variation in the critical packing parameters g, we were able to control hexagonal, lamellar, and cubic mesostructures (Figure 15). The self-assembly procedure is rapid and incorporates the organic monomers uniformly within a highly ordered, inorganic environment. Polymerization results in polydiacetylene (PDA)/silica nanocomposites that are optically transparent and mechanically robust. Compared to ordered diacetylene-containing films prepared as Langmuir monolayers[55] or by Langmuir-Blodgett deposition[55], the nanostructured inorganic host alters the diacetylene polymerization behavior, and the resulting nanocomposite exhibits unusual chromatic changes in response to thermal, mechanical and chemical stimuli. The inorganic framework serves to protect, stabilize, and orient the polymer, and to mediate its function. The nanocomposite architecture also provides sufficient mechanical integrity to enable integration of conjugated polymers into devices and microsystems. Beginning with a homogeneous solution of silicic acid and surfactant prepared in a THFIwater solvent with initial surfactant concentration c, much less than the critical surfactant micelle concentration cmc, we use evaporative dip-coating, spin-coating, or casting procedures to prepare thin films on silicon or fused silica substrates. During deposition, preferential evaporation of THF concentrates the depositing film in water and nonvolatile silica and surfactant species. The progressively increasing surfactant concentration drives self-assembly of diacetylenelsilica micelles and their further organization into ordered liquid crystalline mesophases. Shape and concentration of the DA surfactants influence the mesophase obtained (lamellar, hexagonal, or cubic) (Figure 15). Ultraviolet (UV) lightinitiated polymerization of the DA units, accompanied by catalyst-promoted siloxane condensation, topochemically convert the colorless mesophase into the blue PDAIsilica nanocomposite, preserving the highly ordered, selfassembled architecture (Figure 16 and 17). It should be noted that these DA surfactants assembled as a Langmuir monolayer do not polymerize, implying that nano-confinement, presumably the curvature of the nanostructured host, influences strongly the polymerization pathway. The nanostructured polydiacetylene composites exhibit thermo, chemoand mechano-chromism similar to that of bulk polydiacetlyene, however unlike bulk PDA we observed in many cases reversibility. Overall

nanostructuring of conjugated polymers appears promising to control other properties like charge and energy transfer needed to improve conjugatedpolymer based devices. Acknowledgement: This work was supported by the U.S. Department of Energy Basic Energy Sciences Program, NASA, the DOE Nanoscience Engineering and Technology Program DE-FG03-02ER15368, the Air Force Office of Scientific Research F49620-01-1-0 168 and MURI F49602-01- 1-0352, the National Aeronautics and Space Administration NAG5-882 1, the U.S. Army Research Laboratory and the U.S. Army Research Office, the Sandia National Laboratories Laboratory-Directed Research and Development Program, the UNMINSF Center for Micro-engineered Materials. And the Defense Advanced Research Projects Agency Bio-Weapons Defense Program. TEM investigations were performed in the Department of Earth and Planetary Sciences at the University of New Mexico. Sandia is a multiprogram laboratory operated by Sandia Corp., a Lockheed-Martin Co., for the U.S. DOE under Contract DE-AC04-94AL85000.

Reference:

C. Kresge, M. Leonowicz, W. Roth, C. Vartuli, J. Beck, Nature 359 (1992) 710. S. Mann, S. Burkett, S. Davis, C. Fowler, N. Mendelson, S. Sims, D. Walsh, N. Whilton, CHEMISTRY OF MATERIALS 9 (1997) 2300. D. Zhao, P. Yang, Q. Huo, B. Chmelka, G. Stucky, Curr. Opin. Solid State Mat. Sci. 3 (1998) 111. I. Soten, G. Ozin, CURRENT OPINION IN COLLOID & INTERFACE SCIENCE 4 (1999) 325. S. Pevzner, 0 . Regev, R. Yerushalmi-Rozen, CURRENT OPINION IN COLLOID & INTERFACE SCIENCE 4 (1 999) 420. M. Ogawa, Chem. Commun. (1996) 1149. M. Kapoor, S. Inagaki, CHEMISTRY OF MATERIALS 14 (2002) 3509. S. Inagaki, S. Guan, T. Ohsuna, 0 . Terasaki, NATURE 416 (2002) 304. M. Kruk, M. Jaroniec, S. Guan, S. Inagaki, JOURNAL OF PHYSICAL CHEMISTRY B 105 (2001) 681. M. Kapoor, S. Inagaki, Chem. Lett. 33 (2004) 88. Z. Tian, J. Liu, J. Voigt, B. McKenzie, H. Xu, Angew. Chem.-Int. Edit. v.42 (2003) 4 14. A. Stein, B. J. Melde, R. C. Schroden, Advanced Materials 12 (2000) 1403. H. Yang, N. Coombs, I. Sokolov, G. Ozin, NATURE 381 (19%) 589. N. Yao, A. Ku, N. Nakagawa, T. Lee, D. Saville, I. Aksay, CHEMISTRY OF MATERIALS 12 (2000) 1536. H. Yang, N. Coombs, G. Ozin, J. Mater. Chem. 8 (1998) 1205. Y. Lu, R. Ganguli, C. Drewien, M. Anderson, C. Brinker, W. Gong, Y. Guo, H. Soyez, B. Dunn, M. Huang, J. Zink, NATURE 389 (1997) 364. C. Brinker, Y. Lu, A. Sellinger, H. Fan, ADVANCED MATERIALS 11 (1999) 579. C. J. Brinker, G. W. Scherer, Sol-gel science: the physics and chemistry of sol-gel processing, Academic Press INC,, San Diego, CA, 1990. Y. Lu, H. Fan, A. Stump, T. Ward, T. Rieker, C. Brinker, NATURE 398 (1999) 223. D. Grosso, G. Soler-Illia, F. Babonneau, C. Sanchez, P. Albouy, A. BrunetBruneau, A. Balkenende, ADVANCED MATERIALS 13 (2001) 1085. D. Grosso, A. Balkenende, P. Albouy, A. Ayral, H. Amenitsch, F. Babonneau, CHEMISTRY OF MATERIALS 13 (2001) 1848. D. Grosso, F. Babonneau, G. Soler-Illia, P. Albouy, H. Amenitsch, CHEMICAL COMMUNICATIONS (2002) 748. D. Grosso, F. Babonneau, P. Albouy, H. Amenitsch, A. Balkenende, A. BrunetBruneau, J. Rivory, CHEMISTRY OF MATERIALS 14 (2002) 931. D. Doshi, A. Gibaud, V. Goletto, M. Lu, H. Gerung, B. Ocko, S. Han, C. Brinker, JOURNAL OF THE AMERICAN CHEMICAL SOCIETY 125 (2003) 1 1646. D. Doshi, A. Gibaud, N. Liu, D. Sturmayr, A. Malanoski, D. Dunphy, H. Chen, S. Narayanan, A. MacPhee, J. Wang, S. Reed, A. Hurd, F. van Swol, C. Brinker, JOURNAL OF PHYSICAL CHEMISTRY B 107 (2003) 7683. M. Huang, B. Dunn, H. Soyez, J. Zink, LANGMUIR 14 (1998) 7331. M. Huang, H. Soyez, B. Dunn, J. Zink, CHEMISTRY OF MATERIALS 12 (2000) 23 1.

M. Huang, B. Dunn, J. Zink, JOURNAL OF THE AMERICAN CHEMICAL SOCIETY 122 (2000) 3739. T. Asefa, M. MacLachan, N. Coombs, G. Ozin, NATURE 402 (1999) 867. S. Inagaki, NANOTECHNOLOGY IN MESOSTRUCTURED MATERIALS 146 (2003) 1. S. Burkett, S. Sims, S. Mann, Chem. Commun. 11 (19%) 1367. C. Fowler, S. Burkett, S. Mann, Chem Commun (1997) 1769. M. H. Lim, C. F. Blanford, A. Stein, J Am Chem Soc 117 (1997) 4090. J. Liu, Feng, X., Fryxell, G.E., Wang, L.Q., Kim, A.Y., Gong, M., Adv. Mater. 10 (1998) 161 M. Trau, N. Yao, E. Kim, Y. Xia, G. Whitesides, I. Aksay, NATURE 390 (1997) 674. P. D. Yang, T. Deng, D. Y. Zhao, P. Y. Feng, D. Pine, B. F. Chmelka, G. M. Whitesides, G. D. Stucky, Science 282 (1998) 2244. H. Yang, N. Coombs, G. Ozin, Adv. Mater. 9 (1997) 81 1. H. Fan, Y. Lu, A. Stump, S. Reed, T. Baer, R. Schunk, V. Perez-Luna, G. Lopez, C. Brinker, NATURE 405 (2000) 56. D. Doshi, N. Huesing, M. Lu, H. Fan, Y. Lu, K. Simmons-Potter, B. Potter, A. Hurd, C. Brinker, SCIENCE 290 (2000) 107. S. S. Prakash, C. J. Brinker, A. J. Hurd, S. M. Rao, Nature 374 (1995) 439. P. Yang, G. Wirnsberger, H. Huang, S. Cordero, M. McGehee, B. Scott, T. Deng, G. Whitesides, B. Chmelka, S. Buratto, G. Stucky, SCIENCE 287 (2000) 465. Y. N. Xia, J. A. Rogers, K. E. Paul, G. M. Whitesides, Chemical Reviews 99 (1999) 1823. L. Tender, R. Worley, H. Fan, G. Lopez, Langmuir 12 (19%) 5515. H. Schild, Prog. Polym. Sci. 17 (1992) 163. E. Diez-Pena, I. Quijada-Garrido, P. Frutos, J. Barrales-Rienda, Macromolecules 35 (2002) 2667. A. Sellinger, P. Weiss, A. Nguyen, Y. Lu, R. Assink, W. Gong, C. Brinker, NATURE 394 (1998) 256. G. Garnweitner, B. Smarsly, R. Assink, W. Ruland, E. Bond, C. Brinker, JOURNAL OF THE AMERICAN CHEMICAL SOCIETY 125 (2003) 5626. N. Liu, Z. Chen, D. Dunphy, Y. Jiang, R. Assink, C. Brinker, ANGEWANDTE CHEMIE-INTERNATIONAL EDITION 42 (2003) 1731. M. Ogawa, K. Kurodacd, J. Moric, Chem Commun (2000) 2441. L. Siewierski, W. Brittain, S. Petrash, M. Foster, Langmuir (1996) 5838. G. Kumar, D. Neckers, Chem. Mater. 89 (1989) 1915. M. Ueda, H. Kim, T. Ikeda, K. Ichimura, Chem. Mater. (1992) 1229. Y. Yang, Y. Lu, M. Lu, J. Huang, R. Haddad, G. Xomeritakis, N. Liu, A. Malanoski, D. Sturmayr, H. Fan, D. Sasaki, R. Assink, J. Shelnutt, F. van Swol, G. Lopez, A. Burns, C. Brinker, JOURNAL OF THE AMERICAN CHEMICAL SOCIETY 125 (2003) 1269. Y. Lu, Y. Yang, A. Sellinger, M. Lu, J. Huang, H. Fan, R. Haddad, G. Lopez, A. Burns, D. Sasaki, J. Shelnutt, C. Brinker, NATURE 410 (2001) 913. D. Sasaki, R. Carpick, A. Burns, J. Colloid. Interf. Sci. (2000) 490.

Table 1. Temporal evolution of mesophases. Time at appearance of different mesophases T=3% 3 9 6 ~ 4 4 4s1 T=441s Lamellar Lamellar Appearance correlated CTAB Isotropic stage of a single micellar Brag spot T=570s T=585s T=630s Appearance Appearance Appearance SDS of lamellar of (OOl), of (002), (0031, (0'33) (0'321, (003) T=300s T=330s T=670s Incipient Correlated Hexagonal Brij56 lamellar micellar cubic correlated hexagonal micellar Td96s

With silica

+

Without silica CTAB

+

T=53 1s hexagonal

+ +

Isotropic +correlated micellar+lamellar and correlated micellar~crystallineCTAB

Table 2. Functional organosilanes and properties of resultant thin film mesophases

"Pore size and surface area were determined from N, sorption isotherms obtained at -196"C, using a surface acoustic wave (SAW) technique. § Functional groups are retained through selective surfactant removal during heat treatment in nitrogen. TGA and DTA were used to establish the appropriate temperature window enabling complete surfactant removal without silane decomposition. Additives investigated include rhodamine-B, cytochrome c (from Fluka), oil blue N, disperse yellow 3 (from Aldrich), silver ions and silver nanoparticles. # Dye molecule used here is 5,6-carboxyfluorecein, succinimidyl ester (5,6FAM, SE, from Molecular Probes).

Figure 1. Experimental setup and reciprocal space schematics. (A) The liquid spectrometer setup used for GISAXS experiments. (B) Schematic of the reciprocal space probed by the experiment, where R is the angle of incidence, B is the scattering angle in the specular direction, and re is the scattering angle in plane. The reciprocal lattice is defined by qz, normal to the substrate surface. qx and qy are the in-plane directions.

Figure 2. Temporal evolution of structure and composition for the sample CTABISi=0.12: (I) the isotropic phase; @I) the lamellar phase; 011) correlated micellar phase; (IV) hexagonal mesophase.

Ethanol

-

Figure 3. Bulk and thin-film t e m q phase diagram. The evaporationinduced compositional trajectories of the three CTABlsilica syskms ( CTAB/Si=O.lO, 0.12, and 0.16) andthe WS system are mapped onto the bulk water/ethanol/CTAB phase diagram, considering the hydrophilic silicic acid precursors to be equivalent to water.

:k 7:

a*

w

* m

.t

a.

08 .S&

..

%3*

-

*I

.

w

<*

4,

0

*<

,0

u

A-..

- 4,

m

2

0,

at

Figure 4. GISAXS patterns obtained after the specified times, i (seconds), are presented for sample 0.12, corresponding to the following: (a) the isotropic phase, (b) the lamellar mesophase, (c and d) the correlated micellar, and (e and f) the hexagonal mesophase.

Figure 5. Meandering patterned mesophase created by MF'L. a, Optical micrograph of patterned rhodamine-B-containing silica mesophase deposited on an oxidized [loo]- oriented silicon substrate at a speed of 2.54 cmls-I. Inset is a fluorescence image of rhodamine-B emission acquired through a 610-nm bandpass filter, demonstrating retention of rh0damine-B functionality. b, Representative E M micrograph of a fragment of the patterned rhodamine-B-containing Blm corresponding to a [I 1 01-oriented cubic mesophase with lattice constant a = 10.3 nm.

Figure 6. Patterned dot arrays created by ink-jet printing, IJP. a, Optical micrograph of a dot array created by IJP of standard ink (Hewlett-Packard Co.) on a non-adsorbent surface. b, Optical micrograph of an array of hydrophobic, mesoporous silica dots created by evaporation-induced silica/surfactant self-assembly during IJP on an oxidized [loo]-oriented silicon substrate followed by calcination. c, Representative TEM micrograph of a dot fragment prepared as in b.

Mircrfluidic system

Figure 7. Patterned functional mesostructure formed by selective de-wetting. Using micro-contact printing or electrochemical desorption techniques, substrates are prepared with patterns of hydrophilic, hydroxyl-terminated SAMs and hydrophobic methyl-terminated SAMs.

+ C,H,OH/H,O (Solvent)

R-Ar-I + Ar.

+ H + S ~ F +~ -Solvent-

R = alkyl chain

Ar = aryl

Figure 8. Photoacid generator and reaction 1.

Ah=?-yewfilm

Schema 2 : g a y scale patterning (lower surfactant conc.) "I,>

nu,

Yellow (IigMer) regions are UV exposed

1

Y

~ e m 1e: ssl&tiva etching

!

^..-...

Cnad phass

I..-.-. (higher surfaclmtcon%) "w > "un1,r

Blue (darker) regions are not exposad

Figure 9. Processing pathways for optically defined multifinctional patterning of thin-film silica mesophases. Conc., concentration; irr, irradiated; unirr, unirradiated.

Figure 10. Optical patterning of function and properties in thin-.lm silica mesophases. (A) Optical image of localized acid generation via wincorporation of a pH-sensitive dye (ethyl violet). (B) Optical micrograph of a UV-exposed and selectively etched mesostructured thin film (after calcination). Inset: TEM image of the -1m seen in (B), (C) Optical interference image showing thickness and refractive index contrast in a patterned calcined film. The green areas correspond to UV-exposed and calcined regions and the black areas to unexposed and calcined regions. (D) Optical image of an array of water droplets contained within patterned hydrophilic-hydrophobic corrals.

-

agankkL*

Figure 1 1. Preparation of environmentally responsive nanocomposite thin films

. .

Figure 12. SAXS data of a film prepared with NIPAM and DM before (A,i) and aRer (A,$ heat treatment, and in water at 30 "C (B,iii) and 40 "C (B,iv). Solid lines: fit based on ref 9. (C) pH sensitivity of PDM films. Curve c is scaled down by a factor of 2 for better visualization. (D) TEM micrograph of the nanocomposite film corresponding to Figure 2A,ii.

Figure 13. Photoresponsive nanocomposites prepared by EISA. The trans or cis conformation of azobenzene unit was calculated using Chem3D Pro 5.5 molecular modeling analysis s o h a r e . Atom labels: C: grey, 0: red, N: dark blue, Si: blue, H atoms are omitted.

Figure 14. Photo and thermal isomerization o f azobenzene ligands in the nanocomposite films prepared with Brij56 template. a) Prepared; b) after UV irradiation for 30 minutes; c, d, e, f) after room-light exposure of the sample (b) for 3, 15, 35, and 60 minutes, respectively; g) after heating the sample (b) to 100°C for 5 minutes.

C

u

r

C

n

~

~

-"=8 ..A!-

--

Figure 15. DA surfactants serve both as amphiphiles to direct the selfassembly of the silicic acid mesostructure and as monomeric precursors of the conjugated polymer, PDA. Increasing the number, n, of EO subunits comprising the hydrophilic surfactant headgroup resulted in the formation of higher-curvature mesophases: lamellar (n = 3) hexagonal (n = 5) +cubic (n = 10) due to a decreasing value of the surfactant packing parameter g. However, larger headgroups (n = 10) also served as spacers, preventing polymerization of the pure DA-EOIO surfactant. Addition of surfactants with smaller headgoups (e.g., 1 with n = 3 or 5, or 2) was necessary to form PDA in the cubic system.

+

a

Coiourlesa transparent nanocompositefilm8

L

-

Ultraviolet i (2 min)

d

---

, @C(Imin) , Ultraviolet irradiation (2 min) 100

Figure 16. Patterned polymerization induced by W irradiation and thermochromic and solvatochromic transition of hexagonal PDAIsilica nanocomposite fdms. (a) Schematic representation of the UV patterning procedure. (b) Optical image of patterned blue/colorless film formed by UV irradiation for 2 min. (c) Patterned red/colorless film formed by heating the film in (b) to 100 OC for 1 min. (d) Patterned redhlue film formed by UV exposure of film in (c) for 2 min.

Figure 17. Representative TEM images of PDNsilica nanocomposite thin films and particles (formed by a related aerosol assisted EISA approach[l9]).

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Nanostructured carbon materials synthesized from mesoporous silica crystals by replication Ryong Ryoo* and Sang Hoon Joo National Creative Research Initiative Center for Functional Nanomaterials and Department of Chemistry (School of Molecular Science-BK2 I), Korea Advanced Institute of Science and Technology, Daejeon, 305-701, Republic of Korea. E-mail: rryoo(tr)ltaist.ac.h Since the first synthesis of ordered mesoporous carbon in 1999, the field of ordered mesoporous carbon has made a significant progress during the last five years. This review presents the recent advances in the synthesis, properties and applications of ordered mesoporous carbons with an emphasis on our recent results. Ordered mesoporous carbons with various mesopore structures, designated as CMK-n (Carbon Mesostructured by KAIST), have been synthesized using mesoporous silicas or aluminosilicate templates constructed with 3-D pore connectivity such as MCM-48 (cubic Ia3d), SBA-1 (cubic Pm3n), SBA-15 (2-D hexagonal p6mm), SBA-16 (cubic Im3m) and KIT-6 (large pore cubic Ia3d). Mesoporous carbons composed of graphitic frameworks, namely CMK-nG, are synthesized via in situ conversion of aromatic organic compounds into mesophase pitch. The pore diameter control of mesoporous carbons has been achieved by developing synthetic route to the systematic control of wall thickness of hexagonal mesoporous silicas. New class of porous materials, ordered nanoporous carbon-polymer composites, has been synthesized by coating mesoporous carbon frameworks with polymeric entities. Applications of mesoporous carbons have been explored in areas such as catalyst supports, energy storage media, dye and biomolecule adsorbents, and templates for new inorganic porous materials.

1. POROUS CARBONS WITH IRREGULAR STRUCTURE There has been a growing interest on microporous (pore diameters less than 2 nm) carbonaceous materials in the fundamental scientific understanding as well as in the growing technology, owing to their remarkable properties such as high specific surface areas, large pore volumes, chemical inertness and good

mechanical stability [l-31. The potential applications of these materials include adsorption of large hydrophobic molecules, chromatographic separations, electrochemical double layer capacitors and lithium batteries. Generally, the microporous carbon materials are synthesized by direct pyrolysis of polymers such as polyvinylidene chloride, polyvinyl chloride and polyfurfuryl alcohol [4]. Physicochemical properties, especially porosity and pore size, of the so obtained carbon materials can be systemically modified by varying the temperature, time of pyrolysis and post-synthesis treatment. As an alternative approach, activated carbons can be used as precursors in the production of porous carbons. In order to control the pore size, activated carbons are modified by a coating of polymer or deposition of propylene gas [ 5 ] . However, for advanced applications such as adsorption of large hydrophobic molecules, chromatographic separations, electrochemical double layer and lithium batteries, the materials with wider and uniform pore sizes, preferably in the mesoporous range (pore diameters 2 - 50 nm), would be more advantageous. Several methods exist to achieve this goal [5]. Catalytic activation is the most often used method for the production of mesoporous carbons. With this view, various types of catalysts such as iron, nickel, cobalt, rare earth metals, titanium oxide, boron and phosphate-containing sodium compounds were widely used. Mesoporous carbons can also be synthesized by carbonizing a mesoporous organic aerogel or a polymer blend. Polymer aerogel is used for the synthesis of carbon aerogel in which the mesoporosity of the aerogel can be maintained during its carbonization. In the polymer blend method, a nanoscale mixture of thermosetting and pyrolyzing polymers is heat-treated; the former polymer forms a carbon matrix, and the latter one leaves pore in the matrix upon the heat treatment. In continuation, template-synthesis approach using inorganic templates has also been suggested to synthesize mesoporous carbon with uniform pore size distribution [6]. Multi-wall carbon nanotubes are also considered as a class of mesoporous carbon materials [7]. However, despite the ongoing research efforts on different synthetic approaches, there still remains an opening for the synthesis of mesoporous carbons with monodisperse pores of well-defined size and shape.

2. TEMPLATE-DIRECTED APPROACHES T O POROUS CARBONS Although various synthetic methodologies not using solid template exist for porous carbons with uniform carbon pore structure, the resultant carbon materials often do not exhibit the pore structural order. Many efforts on the development of template-directed carbonization techniques have been devoted in order to synthesize materials with pore structural order. The template-directed methods consist of the carbonization of organic compounds in the nanospace of an inorganic template and the removal of the resultant carbon structure from the

template. Knox et al., reported that silica gel could be impregnated with polymer precursors that were polymerized to form a continuous network surrounding silica particles [6]. The carbonization of the polymer coating and the subsequent dissolution of the silica gel template afforded a mesoporous carbon with a rigid structure featuring also some micropores. The material exhibited the specific surface area around 500 m2 g-' and large total pore volume up to 2 The work on templated carbon using amorphous silica gels was followed by the successful synthesis of microporous carbons using layered materials [8, 91. Kyotani and Tomita prepared ultra thin films fiom carbonization of organic polymers in the two-dimensional opening between the lamellae of layered clays such as montmorillonite and taeniolite [8]. Bandosz et al., synthesized microporous molecular sieve carbon by the carbonization of polyfurhryl alcohol between the lamellae of several types of pillared clay having an interlayer spacing as large as 2 nm [9]. On the other hand, anodic alumina with disconnected nanochannels was used as a template to generate carbon nanotubes [lo]. Despite the narrow distribution of pore diameters, these carbon materials failed to exhibit the pore structural order. Synthetic siliceous opals with ordered structures were successful in templating ordered macroporous carbons [ll]. For this purpose, Si02 sphere array was sintered which resulted in "necks" between the silica spheres. The void spaces between connected SiOz arrays were subsequently infiltrated with various carbon sources. The resulting carbon structure after HF washing was an inverse carbon opal. The pore size regime in the carbon structure is suitable for application as photonic band-gap materials. It is reasonable that the removal of the template would afford a highly ordered porous carbon material if the template materials are endowed with a three-dimensional (3-D) channel structure and if enough carbon can be filled into the 3-D regular channels. Based on this ground, previously, several research groups attempted to synthesize ordered microporous carbon replicas using zeolite templates with 3-D pore connectivity [12-141. The resultant carbons exhibit very high specific surface areas with a narrow pore size distribution. However, carbons obtained via zeolite templating did not retain their internal periodic structure but could retain the shapes of zeolite particles. This was because the amount of the carbon precursor infiltrated into the zeolite pores was not sufficient for the formation of rigid carbon frameworks. A large number of mesoporous silica and aluminosilicate materials with various structures was discovered in recent years using the cooperative assemblage between the silicates and surfactant micelles [15-171. The mesoporous materials have attracted considerable interests as host materials for various nanoarchitectures such as metals, metal oxides, semiconducors and

organic polymers due to their uniform and controllable pore diameters, typically, in the range of 2 to 15 nm and rigid frameworks with thickness, typically, 1 to 3 nrn [18]. In 1999, we have pointed out that such mesoporous structures are suitable as a template for the synthesis of ordered mesoporous carbons [19]. Since the first discovery of the ordered mesoporous carbon referred to as CMK1, a large number of mesoporous carbons were obtained following the same or similar synthesis routes. It is also noteworthy that Kyotani et al., later synthesized the ordered carbon replica of zeolite Y by improving carbon loading procedure through two-step infiltration [20]. The synthesis procedure consists of impregnation of Y zeolite with h r h r y l alcohol and subsequent carbonization followed by chemical vapor deposition of additional amount of carbon from propylene pyrolysis. This led to the formation of carbon that retained the (111) plane ordering of zeolite Y template. 3. ORDERED MESOPOROUS CARBONS 3.1. Synthesis strategy The synthesis scheme using mesoporous silica is shown in Fig. 1. In the first step of synthesis, a carbon source such as sucrose, h r h r y l alcohol, phenolresin monomers or acetylene is infiltrated into the template mesopores. The carbon sources are then converted to carbon by pyrolysis, similar to the preparation of ordinary porous carbons materials. However, in the case of template synthesis, the pyrolysis should be restricted within template pores. An effective means for such restriction of carbonization is to convert the carbon sources into cross-linked solid polymers before the pyrolysis reactions take place. An acid polymerization catalyst is dissolved in the carbon source, sulfuric acid being most suitable in the case of carbohydrates. Alternatively, a solid acid (typically, Al) can be incorporated within the mesoporous silica frameworks prior to the infiltration of the carbon source. The A1 catalyst is preferred in most of the cases because it can catalyze the polymerization process exclusively inside the templating region, so that subsequent carbonization of the crosslinked polymers leads to the formation of nanostructured carbons inside the pores. The template can easily be removed at room temperature or with boiling in ethanol-water solution of HF or NaOH. It should be noted that 3-D pore connectivity is prerequisite for the formation of carbon networks that can retain the structure after the removal of template. Two types of carbon products are obtained, depending on the synthesis conditions where carbon precursors are completely in template pores or coated as a thin film on the pore walls. Rod-type carbons are generated after silica template walls are removed in the former case while tube-type carbon can be

Carbon1 Silica

Mesoporous Car-"

9

@

Fig. 1. Schematic diagram of lhc synthesis strategy for ordered mesoporous carbons

obtained in the latter case. Generally, the synthesis of the tube-type carbons becomes progressively more difficult as the diameter of template pores decreases, which is due to capillary condensation of carbon sources. Even in the case where carbon precursors are initially coated as a film on pore walls, the carbon source can undergo the pore filling druing polymerization and carbonization processes.

3.2. Variety of mesostructures After the successflu1 synthesis of ordered mesoporous carbon CMK-I, our research interests emerged to explore the possibilities to synthesize various types of mcsoporous carbons with different mesostructures and pore diameters. Their synthetic and characteristic features are briefly outlined in the following discussion. CMK-1 was synthesized using MCM-48 mesoporous silica as the template and sucrose as the carbon source (Fig. 2a) [19]. The MCM-48 silica was impregnated with sucrose in the presence of sulfuric acid acting as a catalyst, the resulting mixture was dried and then the impregnation-drying step was repeated once as the amount of the carbon source impregnated in the first step was not sufficient for the formation of rigid carbon frameworks. The obtained sample was carbonized at 1173 K under vacuum or nitrogen and finally the silica

template was removed either by aqueous-ethanol solution of NaOH or HF. Other carbon precursors such as glucose, xylose, h r h r y l alcohol and in situ polymerized phenol resin can also be used for CMK-1 synthesis. However, in the case of h r h r y l alcohol and in situ polymerized phenol resin, the aluminosilicate form of the template is used to ensure facile carbonization, instead of sulhric acid as catalyst. The resulting carbon, CMK- 1, shows several distinct Bragg X-ray diffraction (XRD) lines below 28 = 5" indicating the highly ordered mesostructure. The periodic nature of CMK-1 was confirmed by transmission electron microscopy (TEM). CMK-1 exhibits high BET surface area (1200 - 1800 m2g-') and a large pore volume (0.9 - 1.2 cm3g-') [21]. Typical CMK-1 has uniform mesopores about 3 nm in diameter. It is noteworthy that the periodic structure of CMK-1 is different from that of the MCM-48 template. Considering the fact that the MCM-48 silica (cubic Ia3d) has a bicontinuous structure consisting of an enantiomeric pair of interpenetrating 3-D networks of mesoporous channels [22], the carbon networks formed in the two different kinds of mesoporous channel systems are not interconnected. Upon removal of separating silica frameworks, the two carbon networks can change their positions with respect to one another, probably, along the [OOl] crystal axis or [I101 [23, 241. Structural study by electron crystallography suggests that this structural transformation results in the symmetry change from cubic Ia3d of MCM-48 template to cubic I4,/a of CMK-1 [23]. CMK-1 mesoporous carbons were synthesized by the MCM-48 templates with various pore sizes [21]. Although the unit-cell parameters of the resulting CMK-1 carbons varied according to the MCM-48 templates, the pore sizes were found to be relatively constant because of the constant pore wall thickness of the MCM-48 templates irrespective of their varied pore sizes [2 11. Similar synthesis methods using MCM-48 with different carbon sources were also reported fiom other laboratories, but the carbon materials exhibited the same structural symmetry [25,26]. CMK-2 is an ordered mesoporous carbon obtained from sucrose as a carbon source and the SBA-1 silica as a template (Fig. 2b) [27]. The single crystal analysis using electron diffractions reveals that the SBA-1 carbon is composed of cages (typically 3.3 x 4.1 nm) interconnected with two different kinds of uniform pores (mesopores and micropores) [28]. Structurally, CMK-2 carbon can be identified as the inverse negative replica of SBA-1 template indicated by the XRD pattern and TEM images. CMK-3 is an ordered mesoporous carbon, which was synthesized using SBA-15 mesoporous silica as a template and sucrose as a carbon source (Fig. 2c) [29]. Initially, the choice of SBA-15 silica template was far from being obvious, since SBA-15 silica was reported and widely believed to exhibit a structure with disconnected 1-D channel-like pores, similar to MCM-41 [30]. Carbon templating

Fig. 2. Representative TEM images for ordered mesoporous carbons: (a) CMK-1 (cubic 14,/a structure), (b) CMK-2 (cubic Pm3n), (c) CMK-3 (2-D hexagonal, rod-type) and (d) CMK-4 (cubic Ia3d). with such a 1-D structure yielded a disordered, fiber-like carbon structures [21]. However, further efforts toward the elucidation of SBA-15 pore structure provided the evidence for the existence of connecting micropores and small mesopores in the walls of large-pore channels of SBA-15 silica [31, 321. These structural identification facilitated the synthesis of CMK-3 whose structure is the faithhl replica ofthe mesoporous silica template revealed by XRD and E M . The CMK-3 carbon exhibits high BET surface area around 1500 m2g1 and pore volume about 1.3 cmjg-l. The pore size of CMK-3 was around 4.5 nrn revealed by nitrogen adsorption and TEM. Recently, the CMK-3 mesoporous carbons with variable pore diameters were synthesized by the systematic control of pore wall thickness of hexagonal mesoporous silica templates [33]. Through the successful fabrication of CMK-3 mesoporous carbon, the carbon replication was found to be useful for the identification of exact pore structure of the block-copolymer-templated SBA-15-type mesoporous silicas. For example, the change of pore structure of SBA-15 silica treated under high temperature could be probed by the carbon replication [34]. The hexagonally ordered CMK-3 carbons were synthesized using SBA-15 templates calcined al a temperature of 1153 K, whereas a disordered carbon was obtained using SBA-15 calcined at 1243 K indicating that the connecting pores in SBA-15 are eliminated

close to 1243 K. As an another example of carbon replication, the carbon synthesis was performed with the SBA-15-type mesoporous silicas synthesized using different silica sources under acidic and neutral conditions [35]. The results indicated that, in all cases, hexagonally ordered CMK-3 carbons were obtained providing an evidence that the presence of connecting pores between ordered mesoporous channels is a general feature of the ordered silicas synthesized using polymer or oligomer templates with poly(ethy1ene oxide) blocks. Recently, hexagonally ordered mesoporous carbons containing nitrogen groups were synthesized using mesoporous silica SBA-15 as a template and polyacrylonitrile as a carbon source [36]. The CMK-4 carbon is prepared with partially disordered MCM-48 silica obtained by hydrothermal treatments of high-quality MCM-48 (Fig. 2d) [23]. The XRD pattern and TEM image reveals that the space group of templated carbon is preserved after the complete removal of the silica template indicating that CMK-4 carbon is a negative replica retaining the cubic Ia3d structure of the MCM-48 template, on the contrary to the case of CMK-1. This fact suggests that the two carbon network systems formed in MCM-48 channels are partly connected by pores that have been generated in the separating silica frameworks. The combination of the two carbon networks can also be prevented by a thick carbon coating at the external surface of the MCM-48 crystals [37]. The CMK-5 mesoporous carbon was synthesized using aluminosilicate SBA-15 and furfury1 alcohol (Fig. 3) [38]. The carbon source is converted to carbon through vacuum pyrolysis after being polymerized inside the mesopores of SBA-15. The carbonization under vacuum pyrolysis conditions results in the carbon films coated on the mesopore walls instead of complete filling. The structure of CMK-5 is composed of hexagonal arrays of carbon nanopipes originally formed inside the cylindrical nanotubes of SBA-15 template. Even after the template was completely removed, the carbon nanopipes are rigidly interconnected into a highly ordered hexagonal array by carbon spacers that are formed inside the complementary pores between the adjacent cylinders. The mesoscopic structural order between the carbon nanopipes gives rise to more than five Bragg X-ray diffi-action lines at small scattering angles below 20 = 5". The carbon atoms in the frameworks do not have sufficiently long-range atomic order to exhibit XRD peaks in larger scattering angles. Due to the short-range atomic order as in partially ordered graphite, ring patterns and fragmented graphitic fringes appear in the electron diffraction pattern and high-resolution transmission electron micrographs. It is noteworthy that in comparison to our previous work on the CMK-3 (rod-type) synthesis, the polymerization of hrfuryl alcohol by the designed catalytic function of the aluminosilicate template leads to the formation of the pipe-type carbons. The pore size distribution curve obtained by the N2

Nth)[pjq

--!

2 (a) E -E

.-

1

2

3

2eIdeamesl

4

5

2

1

5

8 1 0 1 2

Pore size lnml

Fig. 3. Tube-type ordered mcsoporous carbon CMK-5 (2-D hcxagonal, rod-type): (a) XRD pattern, (b) porc size distribution, (c) TEM image and (d) HRTEM image

adsorption exhibited two sharp peaks with the maxima corresponding to the inside diameter of the carbon nanopipes (5.9 nm) and the pore formed between the adjacent pipes (4.2 nm). The outside diameter of the nanopipes can be tailored by the choice of SBA-15 template with suitable diameter [39].The inner diameter and the pore wall thickness can be controlled in several ways. One such method is to change the amount of polymerized furfuryl alcohol using different polymerization temperature and time. Another method relies on the addition of more carbon source after the initial polymerization [39]. The tube-type mesoporous carbon was also synthesized via catalytic chemical vapor deposition (CCVD) using SBA-15 silica template [40].In this case, the carbon structure generated within silica pores were controlled by duration of the CCVD. Mesoporous carbons with cubic Im3m symmetry were synthesized from the mesoporous silica SBA-16. The SBA-16 silica templates were synthesized using nonionic surfactants such as oligomers and poly(ethy1enc oxide)-type triblock copolymers under acidic conditions [41]. Structurally, SBA-16 can be perceived as cage-like pores interconnected multidirectionally through the rather narrow pore entrances. Recent synthesis efforts toward SBA-16 silica indicated that the diameter of cages and pore entrances can be controlled via hydrothermal treatments 1421. Using these SBA-16 silicas as templates and furfulyl alcohol as the carbon source, two types of mesoporous carbons (CMK-6 and CMK-7) were

synthesized [43].CMK-6 mesoporous carbon was generated where the cage-like pores were completely filled with the carbons, whereas the CMK-7 carbon was generated with carbon coating on the templating pore walls after the carbonizations. The latter carbon exhibited the bimodal porosity as in the case of CMK-5 mesoporous carbon. IIighly ordered mesoporous carbons were also prepared from large pore mesoporous silicas of cubic Ia3d symmetry (Fig. 4). Recently, we developed a new synthesis route to high-quality large mesoporous cubic la3d silica, denoted as KIT-6, utilizing a triblock copolymer-butanol mixture as the structuredirecting agent [44]. The method has the advantage of high reproducibility in significantly large quantities, easy tunability of pore sizes from 4 to 12 nm via simple hydrothermal treatment. Template syntheses using these silicas generated rod-type and tube-type mesoporous cubic Ia3d carbons [44].The carbonization of sucrose led to the rod-type CMK-8 mesoporous carbon that was faithful negative replica of the mesoporous Ia3d silica with 3-D cubic arrangement of branched rods organized in two enantiomeric interwoven systems of KIT-6 silica. On the other hand, the use of furfulyl alcohol under controlled vacuum pyrolysis afforded the tube-type CMK-9 carbon. In contrast to MCM-48, KIT-6 silica possesses pores that are large enough to allow the walls of silica to be

Fig. 4. Ordcred mcsoporous carbon CMK-8 (cubic Za3d,rod-type) and CMK-9 (cubic Ia3d, tuhe-type) templated from KIT-6 mesoporous silica: (a) XRD pattern and TEM image of KIT6 silica template, (b) XRD pattern for CMK-8 and (d) XRD pattern for CMK-9 carbon.

preferentially coated with a film of carbon, generating a tube-type carbon. Interestingly, the structural symmetry of KIT-6 template was preserved after the carbon replications, contrary to the carbon synthesis using MCM-48. This feature was attributed to the presence of the porous bridges between the two independent channel systems [44]. Very recently, we succeeded in the synthesis of the KIT-6 silica with controlled degree of interconnections between the two channel systems by changing the hydrothermal temperature treatment [45]. The "structurally-transformed" mesoporous carbon structures can be obtained using KIT-6 silica template synthesized at 333 K and sucrose as a carbon source similar to CMK- 1.

3.3. Mesoporous carbons with graphitic frameworks As described in the previous section, the choice of suitable mesoporous silica templates can generate a variety of mesostructured carbons. In mesoporous carbon materials, the alterations of carbon framework structures should result in different physicochemical properties. However, the synthesis of mesoporous carbons with pore regularity was limited to carbon frameworks with an amorphous-carbon-like nature. Recently, we developed a synthetic strategy for mesoporous carbon materials composed of graphitic framework structures (CMK-nG) [46]. In this synthetic methodology, various aromatic compounds such as acenaphthene, acenaphthylene, indene, indan, and substituted naphthalenes were converted to in situ generate mesophase pitch within the pores of Al-incorporating silica templates. After carbonization of the mesophase pitch, the composite was further heated under vacuum for complete carbonization. The carbon product was then released by the treatment of HF or NaOH. The resulting CMK-nG mesoporous carbons constitute the first example of porous carbon materials with the structural regularity both on meso- and atomic scales. The synthesis strategy is generally applicable to the mesoporous silica templates with various mesostructures such as MCM-48, SBA-1, SBA-15, SBA16 and KIT-6 mesoporous silicas. In addition, tube-type carbon nanostructures can be generated by the judicious control of synthesis conditions. Fig. 5 shows the XRD pattern, TEM and high resolution (HR) TEM images of CMK-nG carbons. The CMK-nG carbons exhibit several XRD peaks below 20 = 5" characteristic of the highly ordered mesostructures and similar to CMK-n carbons synthesized using sucrose or furfuryl alcohol. Furthermore, the CMKnG carbons are characterized by intense peaks around 20 = 26, 45, 53, and 78' corresponding to the (002), (101), (004), and (1 10) diffractions of the graphitic frameworks respectively (not shown). The diffraction intensities and peak widths are comparable to those of multi-walled carbon nanotubes. The graphitic character of CMK-nG carbons revealed by XRD patterns was confirmed by the

(a1

-B: U)

E

3 E

-

2

4

6

8

10

20 (degrees)

Fig. 5. Ordered mesoporous carbon with graphitic fmmeworks CMK-nG: (a) XRD patterns for CMK-nG carbons and (band c) TEM images of CMK-3G.

Raman spectra of different carbon species. Interestingly, the carbon framework is constructed with a stacking of the discoid graphene sheets oriented perpendicular to the direction of the rods indicated by the HRTEM image. These mesoporous carbons exhibit enhanced mechanical and thermal stability compared to the previously reported mcsoporous carbons composed of amorphous frameworks [46]. Furthermore, we have successfully synthesized the CMK-nG-type mesoporous carbons using commercially available mesophase pitch that can provide a facile route to large scale, low-cost synthesis of graphitic mesoporous carbons [47]. Synthesis works using mesophase pitch in SBA-15 template werc reported in recent studies from other research groups, yet the graphitic framework structure was not confirmed [48,49]. 3.4. Pore size control of mesoporous carbons The mesoporous carbons can be prepared with a variety of pore shapes, connectivity and pore wall thickness, depending on pore structures and diameters of the silica templates. Apart from such structural variations, the control of pore diameters represents another important aspect for the applications of the mesoporous carbons. Obviously, in order to control the pore size of mesoporous carbons, the wall thickness of the mesoporous silica templates should be varied. However, this is not a simple problem. In the case of mesoporous silicas, the pore diameter can be readily controlled by the choice of surfactant containing different length of hydrophobic part or post-synthesis hydrothermal pore expansion. However, pore diameters of mesoporous carbons

are relatively more difficult to control, due to the lack of the effective method to control the pore wall thickness of silica templates. To this end, we developed a synthetic strategy for the systematic control of the pore wall thickness of hexagonal mesoporous silicas. These mesoporous silicas were successfully used as templates for the synthesis of mesoporous carbons with various pore diameters [33]. Our synthesis route to the pore wall thickness control relied on the use of the surfactant mixtures composed of cationic [hexadecyltrimethylammonium bromide (HTAB)] and nonionic [polyoxyethylene hexadecyl ether-type (Cl&O*)] surfactants. The synthesis of hexagonal mesoporous silicas was performed under acidic conditions with different molar ratio of the surfactant mixtures. The wall thickness of mesoporous silicas was systematically increased with the increase of C16E08 surfactant contents. The pore wall thickness was reasonablc in that the silica species interacting with the head-group corona through hydrogen bonding would increase in number with increasing EO segments per surfactant. The carbons synthesized using these mesoporous silica templates resulted in the CMK-3-type hexagonal structures as shown in Fig. 6(a). The pore diameters of mesoporous carbons are very narrow in distribution and the pore size at the maximum of the distribution is systematically shilled from 2.2 to 3.3 nrn against the HTAB: ClhEOn ratio used for the synthesis of templates [Fig. 6(b)]. The synthesis route to control the wall thickness as exemplified by HTAB-C16E08 system can be generalized to the mesoporous silicas synthesis employing the cationic-nonionic surfactant mixture systems. The silicas thus synthesized can successfully be used as templates for mesoporous carbons of different pore diameters. (a)

I

2

4

6

28 (degrees)

8

(b) 7 -

2

3

4

5

6

Pore Size (nm)

Fig. 6. CMK-)-type mesoporous carbons with differentpore diameters: (a) XRD patlem and (b) pore size distributions. (x:y) in CMK-3(x:y) refen to the HTAR: c16F.O~ratio uscd fur the synthesis uf hexagonal silica templates.

3.5. Various approaches to mesoporous and mesostructured carbons In the continuing efforts, various approaches concerning the synthesis of new mesoporous carbon structures using mesoporous silicas are also available. Hyeon and coworkers synthesized disordered mesoporous carbon with narrow pore size distribution using HMS silica template [50]. They reported the fabrication of mesocellular carbon foams with ultra large pore by templating mesoporous siliceous foam [5 11. They further synthesized a mesoporous carbon with disordered pore arrangement using a colloidal silica sol (Ludox) as template and resorcinol-formaldehyde (RF) resin as carbon source [52-541. Jaroniec group developed a colloid-imprinting method for nanoprous carbon structures [48, 55, 561. The colloid-imprinted carbons are synthesized by imprinting dry solid particles of the mesophase pitch as carbon precursor with colloidal silica particles. The colloidal imprinting method allows the creation of spherical pores in the volume of precursor particles as well on their surface depending on the synthesis conditions. The advantage to use the colloidal particles of different sizes and/or chemical nature allows one to design the pores of different sizes in the colloid-imprinted carbon materials as well as to tailor their sorption and catalytic properties. Fuertes reported on the fabrication of mesoporous carbons suing sol-gel derived silica xerogels as a template [57]. The resulting carbons showed that mesopore diameters could be controlled in the range between 3 to 4.5 nm by the proper selection of silica xerogel template. Interestingly, unimodal or bimodal mesoporous carbons could be obtained from the same silica xerogel template by varying the amount of carbon precursor. Other research groups reported that the direct carbonization of silica-organic composite materials as prepared could generate mesoporous carbons, with or without additional carbon sources [58-601. Kyotani et al., reported the synthesis of mesoporous carbon from silica-carbon composite obtained by sol-gel polymerization of tetraethylorthosilicate in the presence of furfuryl alcohol [58]. The resulting carbon structure after the removal of silica framework exhibits mesoporous structure with uniform pore size around 4 nm. Sayari and coworkers reported on the synthesis of nanoporous carbon materials via the preparation of cyclodextrin-silica nanocomposite and subsequent direct carbonization of occluded cyclodextrin [59].

4. NANOPOROUS POLYMER-CARBON COMPOSITES Organic Nanostructured materials with uniform pores have been pursued for a long time. There are reports on the synthesis of such nanostructured organic materials by liquid-crystal templating route and by polymerization using colloidal or mesoporous silica templates. The mesoporous organic materials

have the advantage of facile functionalization and high affinity with organic molecules, compared to inorganic materials. However, the variation of pore diameters and shapes is yet of limited success. The present situation stimulated us to develop a simple synthesis scheme for ordered nanoporous organic polymers using mesoporous carbon as the retaining framework [61]. The overall synthetic scheme for the nanoporous polymer-carbon composite is shown in Fig. 7. As shown in the scheme, ordered mesoporous carbons containing micropores in the mesopore walls are chosen for such synthetic purpose. The porous carbon is impregnated with organic monomers under controlled conditions so that the micropores are filled with organic monomers and that the mesopore walls are coated with a thin layer of monomers. The monomers are subsequently converted to cross-linked polymers through a thermal polymerization process. The resultant materials have the structure with ordered mesopores that are constructed by polymeric materials coated on the carbon fi-ameworks. The synthesis stratcgy can be generalized to various compositions of hydrophilic and hydrophobic organic polymers including crosslinked polystyrene, poly(2-hydroxyethyl methacrylate) and poly(methy1 methacrylate). The polymers can be synthesized on mesoporous carbons of various pore diameters, connectivity and pore shapes such as 2-D hexagonal, cubic Ia3d, cubic Im3m and mesocellular foam. The mesoporous composite structures thus formed can be characterized by changes in the XRD intensity, pore-size analysis using N2 adsorption isotherms and the adsorption of large organic molecules. For example, as shown in Fig. 8, the polystyrene-CMK-l composite exhibits a significant increase in the XRD intensity, compared to the pristine CMK-1. This phenomenon results from the preferential polymerization of styrene inside the micropores of carbon frameworks, which causes apparent density of the mesopore walls to increase. The mesopore vacancy of the

Fig. 7. Synthesis schcmc for ordcrcd nanoporous polymer-carbon composites

2

4

6

8

213 (degrees)

Fig. 8. XRD patterns of CMK-I mesoporous carbon and PS-CMK-I composite,

composite material can bc confirmed by a slight decrease in the mesopore diameter aficr the loading of polystyrene. The resultant polymer-carbon composite nanoporous materials exhibit the same chemical properties of the organic polymers but the stability of the pores against mechanical compression, thermal and chemical treatment is enhanced considerably. The resultant materials exhibiting surface properties of the polymers and the electrical conductivity of the carbon framework may provide new possibilities for advanced applications. Thus, the nanoporous composite system may lead to broad applications in many areas of material science including separation of biomaterials, removal of pollutants, selective ion exchange, manufacturing of high-performance catalysts and sensors due to their unique properties such as high surface area, regular pore structure, high capacity for metal dispersion and facile chemical functionalizability.

5. APPLICATIONS AND FUTURE POSSIBI1,ITIES The ordered structural characteristics with uniform pores, controllable pore diameters, high specific surface areas and large pore volumes impart the mesoporous carbons with a unique opportunity for the application as a standard or reference material in the studies of adsorption, catalysis and other numerous fundamental properties of mesoporous and microporous carbons. Owing to the recent development of the electron single crystallography for mesoshuctured materials [28], it is now possible to determine the pore structures absolutely at least for 3-D ordered CMK-n carbons. The pore shapes and diameters, solved by the electron crystallography, can be compared with the pore structural data obtained from other analysis techniques such as the BJH method, density

functional theory and Kelvin equation. It may be expected that new CMK carbons with various structures would be synthesized in the near future, and that the range of controllable pore diameters could be extended to the microporous range. The exact structural information would provide invaluable information for the investigation of various physicochemical phenomena as a function of nanoporous environments with different pore diameters. Even without the rigorous structural analysis, the presence of well-defined Bragg XRD lines of the CMK-n carbons offers many advantages for characterization. The XRD lines can be used for identification of the structures. The XRD lines increase in intensity with the addition of guest species inside mesopores, which can be used to monitor the adsorption. As seen in the case of the synthesis of polymer composite materials, the XRD intensity can also increase depending on the location of the guest species. Thus, the presence of the distinct XRD patterns in mesoporous carbons provides new opportunities for precisely monitoring various physico-chemical phenomena that take place inside the well-defined carbon pores or at the pore walls such as adsorption, impregnation, framework changes, formation of metal clusters and grafted functional groups. Accompanying with the opportunities for fundamental studies, the mesoporous carbon materials promise to be a suitable alternate for adsorbents, catalyst supports, energy storage media, and materials for advanced electronic applications. For catalytic applications, Pt, Ru and Pd metals have been prepared on CMK mesoporous carbons, and extremely high metal dispersion was already achieved. In particular, the pipe-type CMK-5 carbon showed intriguing capability of supporting well dispersed Pt nanoparticles with high loadings and the Pt loaded CMK-5 exhibited very high electrocatalytic activity toward the oxygen reduction reactions, which is relevant to the fuel-cell technology [38]. In the case of Ru clusters supported on the mesoporous carbon, this catalyst system exhibited remarkably higher catalytic activity in the aerobic oxidation of benzyl alcohols, compared to Ru particles supported on other carbons or inorganic supports [62]. The mesoporous carbons have been effective as energy storage materials. The mesoporous carbon structures templated from MCM-48 and HMS were applied to the electrical double layer capacitor (EDLC) [25, 43, 631. These carbons exhibited a more ideal behavior as capacitor compared to activated carbon. On the other hand, promising application of CMK-3 mesoporous carbon as Li-ion battery was demonstrated [64]. CMK-3 carbon exhibited high specific energy capacity about 1100 m ~ h ~for - ' lithium storage. Furthermore, after the first cycle, the discharge and charge remained at a reversible capacity level with a good cycle performance. The mesoporous carbons were also exploited as adsorbents for bulky

molecules. Hartmann and coworkers tested CMK-3 mesoporous carbon as the adsorbent for cytochrome c [62]. The adsorption capacity of the CMK-3 carbon was higher than that reported for mesoporous silicas such as MCM-41 and SBA15. On the other hand, silica sol-templated mesoporous carbons exhibited high adsorption capacity for bulky dyes [54] and humic acids [66]. Another interesting application of the mesoporous carbons is their utilization as templates for inorganic mesoporous structures. The Kim group [67] and Schuth group [68] demonstrated that the mesoporous silica structures could be templated from hexagonal mesoporous carbon CMK-3. To date, the syntheses of mesoporous materials composed of transition metal oxides are yet of limited success despite their wide applicabilities due to the facile crystallization during the mesostrcuture formation or template removal. The rigid pore walls and well-defined 3-D pore structure of mesoporous carbons is expected to help the formation and stabilization of the inorganic frameworks, facilitating the synthesis of mesoporous inorganic materials such as alumina, titania and zirconia. 6. CONCLUSIONS

This review addressed the recent developments in the synthesis and applications of ordered mesoporous carbons. Following the synthesis of new mesoporous silica structures, a wide variety of interesting mesostructures of mesoporous carbons has emerged. Ordered mesoporous carbons composed of graphite frameworks became available beyond amorphous activated carbon-like structures. Furthermore, mesoporous carbon-based novel nanoporous polymercarbon nanocomposites were synthesized which may pave a way for the exciting future applications. Accompanying with these synthetic approaches, wide applications of mesoporous carbons for catalyst supports, energy storage media, adsorbents for macromolecules and templates for new inorganic porous materials have been investigated. The field of ordered mesoporous carbons is rapidly growing and currently establishing itself as a new subdiscipline both of mesoporous and porous materials. There still remains much to be explored concerning various aspects of mesoporous carbons from findamental structural studies and the nanosize effects to new applications.

ACKNOWLEDGEMENTS This work was supported by the Creative Research Initiative Program of Korea Ministry of Science and Technology.

REFERENCES C. R. Bansal, J.-B. Donnet and F. Stoeckli, Active carbon, Marcel Dekker, New York, 1988. H. Marsh, E. A. Heintz and F. Rodriguez-Reinoso, Introduction to Carbon Technology, University of Alicante, 1997. H. C. Foley, Microporous Mater., 4 (1995) 407. E. Fitzer, K. Mueller and W. Schaefer, in: P. L. Walker Jr. (Ed.), Chemistry and Physics of Carbon, Vol. 7, Marcel Dekker, New York, 1971, pp 238-403. T. Kyotani, Carbon, 38 (2000) 269 and references therein. J. H. Knox, B. Kaur and G. R. Millward, J. Chromatogr., 352 (1986) 3. S. Iijima, Nature, 354 (1991) 56. T. Kyotani, T. Mori and A. Tomita, Chem. Mater., 6 (1994) 2138. T. Bandosz, J. Jagiello, K. Putyera and J. A. Schwarz, Chem. Mater., 8 (1996) 2023. T. Kyotani, T.; L.-f. Tsai and A. Tomita, Chem. Mater., 8 (1996) 21 09. A. A. Zakhidov, R. H. Baughman, Z. Iqbal, C. Cui, I. Khayrullin, S. 0 . Dantas, J. Marti and V. G. Ralchenko, Science, 282 (1998) 897. T. Kyotani, T. Nagai, S. Inoue and A. Tomita, Chem. Mater., 9 (1997) 609. S. A. Johnson, E. S. Brigham, P. J. Ollivier and T. E. Mallouk, Chem. Mater., 9 (1997) 2448. J. Rodriguez-Mirasol, T. Cordero, L. R. Radovic and J. J. Rodriguez, Chem. Mater., 10 (1998) 550. C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature, 359 (1992) 710. A. Corma, Chem. Rev., 97 (1997) 2373. J. Y. Ying, C. P. Mehnert and M. S. Wong, Angew. Chem., Int. Ed., 38 (1999) 57. K. Moller and T. Bein, Chem. Mater., 10 (1998) 2950. R. Ryoo, S, H. Joo and S. Jun, J. Phys. Chem. B, 103 (1999) 7743. Z. Ma, T. Kyotani, Z. Liu, 0 . Terasaki and A. Tomita, Chem. Mater., 13 (2001) 4413. M. Kruk, M. Jaroniec, R. Ryoo and S. H. Joo, J. Phys. Chem. B, 104 (2000) 7960. A. Carlsson, M. Kaneda, Y. Sakarnoto, 0 . Terasaki, R. Ryoo, and S. H. Joo, J. Electron Microscopy, 48 (1999) 795. M. Kaneda, T. Tsubakiyama, A. Carlsson, Y. Sakamoto, T. Ohsuna, 0 . Terasaki, S. H. Joo and R. Ryoo, J. Phys. Chem. B, 106 (2002) 1256. L. A. Solovyov, V. I. Zaikovskii, A. N. Shmakov, 0 . V. Belousov and R. Ryoo, J. Phys. Chem. B, 106 (2002) 12198. J. Lee, S. Yoon, T. Hyeon, S. M. Oh and K. B. Kim, Chem. Commun., (1999) 2177. S. B. Yoon, J. Y. Kim and J. S. Yu, Chem. Commun., (2001) 559. R. Ryoo, S. H. Joo, M. Kruk and M. Jaroniec, Adv. Mater., 13 (2001) 677. Y. Sakamoto, M. Kaneda, 0. Terasaki, D. Y. Zhao, J. M. Kim, G. Stucky, H. J. Shin, and R. Ryoo, Nature, 408 (2000) 449. S. Jun, S. H. Joo, R. Ryoo, M. Kmk, M. Jaroniec, Z. Liu, T. Ohsuna and 0 . Terasaki, J. Am. Chem. Soc., 122 (2000) 10712. D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Fredrickson, B. F. Chmelka and G. D. Stucky, Science, 279 (1998) 548. M. Kruk, M. Jaroniec, C. H. KO and R. Ryoo, Chem. Mater., 12 (2000) 1961. R. Ryoo, C. H. KO, M. Kruk, V. Antochshuk and M. Jaroniec, J. Phys. Chem. B, 104 (2000) 11465.

J.-S. Lee, S. H. Joo and R. Ryoo, J. Am. Chem. Soc., 124 (2002) 1156. H. J. Shin, R. Ryoo, M. Kruk, M. Jaroniec, Chem. Commun., (2001) 349. S. H. Joo, R. Ryoo, M. Kruk and M. Jaroniec, J. Phys. Chem. B, 106 (2002) 4640. A. Lu, A. Kiefer, W. Schmidt and F. Schiith, Chem. Mater., 16 (2004) 100. C. Vix-Guterl, S. Boulard, J. Parmentier, J. Patarin, and J. Werckmann, Chem. Lett., 10 (2002) 1062. S. H. Joo, S. J. Choi, I. Oh, J. Kwak, Z. Liu, 0. Terasaki and R. Ryoo, Nature, 412 (2001) 169. M. Kruk, M. Jaroniec, T.-W. Kim and R. Ryoo, Chem. Mater., 15 (2003) 28 15. W.-H. Zhang, C. Liang, H. Sun, Z. Shen, Y. Guan, P. Ying and C. Li, Adv. Mater., 14 (2002) 1776. D. Zhao, Q. Huo, J. Feng, B. F. Chmelka and G. D. Stucky, J. Am. Chem. Soc., 120 (1998) 6024. T.-W. Kim et al., manuscript in preparation. T.-W. Kim et al., manuscript in preparation. F. Kleitz, S. H. Choi and R. Ryoo, Chem. Commun., (2003) 2136. R. Ryoo et al., manuscript in preparation. T.-W. Kim, I.-S. Park and R. Ryoo, Angew. Chem. Int Ed., 42 (2003) 4375. R. Ryoo et al., manuscript in preparation. Z. Li and M. Jaroniec, J. Phys. Chem. B, 108 (2004) 824. C. Vix-Guterl, S. Saadallah, L. Vidal, M. Reda, J. Parmentier and J. Patarin, J. Mater. Chem., 13 (2003) 2535. J. Lee, S. Yoon, S. M. Oh, C.-H. Shin and T. Hyeon, Adv. Mater., 12 (2000) 359. J. Lee, K. Sohn and T. Hyeon, J. Am. Chem. Soc., 123 (2001) 5646. S. Han and T. Hyeon, Carbon, 37 (1999) 1645. S. Han and T. Hyeon, Chem. Commun., (1999) 1955. S. Han, K. Sohn and T. Hyeon, Chem. Mater., 12 (2000) 3337. Z. Li and M. Jaroniec, J. Am. Chem. Soc., 123 (2001) 9208. Z. Li and M. Jaroniec, Chem. Mater., 15 (2003) 1327. A. B. Fuertes, Chem. Mater., 16 (2004) 449. D. Kawashima, T. Aihara, Y. Kobayashi, T. Kyotani and A. Tomita, Chem. Mater., 12 (2000) 3397. B.-H. Han, W. Zhou and A. Sayari, J. Am. Chem. Soc., 125 (2003) 3444. S. B. Yoon, J. Y. Kim and J. S. Yu, Chem. Commun., (2002) 1536. M. Choi and R. Ryoo, Nature Mater., 2 (2003) 473. R. Ryoo et al., manuscript in preparation. S. Yoon, J. Lee, T. Hyeon and S. M. Oh, J. Electrochem. Soc., 147 (2000) 2507. H. Zhou, S. Zhu, M. Hibino, I. Honma and M. Ichihara, Adv. Mater., 15 (2003) 2107. A. Vinu, C. Streb, V. Murugesan and M. Hartmann, J. Phys. Chem. B, 107 (2003) 8297. S. Han, S. Kim, W. Choi, H. Park, J. Yoon and T. Hyeon, Microporous Mesoporous Mater., 58 (2003) 131. M. Kang, S. H. Yi, H. I. Lee, J. E. Yie and J. M. Kim, Chem. Commun., (2002) 1944. A,-H. Lu, W. Schmidt, A. Taguchi, B. spilethoff, B. Tesche and F. Schuth, Angew. Chem. Int. Ed., 41 (2002) 3489.

Studies in Surface Science and Catalysis 148 Terasaki (Editor) O 2004 Elsevier B.V. All rights reserved.

Structural study of meso-porous materials by electron microscopy Osamu era saki', Tetsu Ohsunal, Zheng ~ i u " Yasuhiro ~, sakamotol and Alfonso E ~arcia- en nett' 1

Structural Chem., Arrhenius Lab., Stockholm Univ., S- 10691 Stockholm, Sweden 2 Bio Nanotec Research Institute Inc., Hamacyo, Tokyo, Japan

1. INTRODUCTION Ever since mesoporous silica materials were synthesised[l,2,3], many mesoporous materials have been synthesised by using self-organisation of amphiphilic molecules, surfactants and polymers under either acidic or basic condition. These mesoporous crystals are structurally unique in that they exhibit order on the mesoscopic-scale and disorder on the atomic-scale. The channels, cages or pores formed within the materials and supportedlseparated by an amorphous silica wall, are arranged periodically on a lattice like artificial atoms or molecules in ordinary crystals. The materials might thus be thought of as "cavity-crystal". The meso-scale order is very sensitive to both synthesis conditions and synthesis time showing local structural variations caused by various local fluctuations in the former. X-ray diffkaction(XRD) powder profile of the mesoporous crystals in most cases show a few broad peaks and therefore structure determination by XRD alone is difficult. As electrons interact with materials much more strongly than X-ray and electron scattering amplitudes are approximately lo4 times larger those of X-rays, we can obtain the same structural information using electrons from a crystal volume lom8smaller than with X-rays. Crystals on the order of tens of nm are sufficiently large for single

crystal transmission electron microscopy(TEM) observation. The successful synthesis of both two-dimensional(2D-) and threedimensional(3D-) mesoporous silica structures, M41S, was reported by the first paper of Kresge et a1 [2]. It is difficult to determine crystal structures solely from powder X-ray difiaction data as mentioned above, even if a crystal has 2Dstructure such as p6mm. Two TEM images with incidences parallel and perpendicular to the channel give conclusive evidence ofp6mm symmetry and a one-dimensional(1D-) channel system. The situation is shown schematically in Fig. 1. TEM images in conjunction with simulation enable observation and discussion of (i) the 1D nature of the channels and (ii) their 2D hexagonal arrangement (p6mm) together with channel shape and wall thickness. Although SBA-15 has p6mm symmetry like MCM-41, it has channel connectivities through randomly arranged tunnels, hercafier, referred to as complementary pores, inside the silica wall, unlike MCM-41. Such a situation may only be clarified at present by observing replicas such as a Pt nano-network structure using TEM[4] and this will be shown later.

Electrons

Electrons

(a) (b) Figure 1. Schematic diagram for two HRTEM imaging modes from two principal directions for 2D-hexagonal p6mm structure, and the corresponding HRTEM images.

Kresge et a1 cleverly combined electron microscopy (EM) observations with powder XRD experiments in order to solve the structure of MCM-41 in their Nature paper[2]. Although this publication clearly indicates the importance of TEM in the structural analysis of mesoporous crystals many papers have since been published comprising an unfortunate mix of possible(speculative) structures and structure solutions. TEM observation is powerful, however it is noteworthy that a TEM image is essentially projected structural information of the specimen along the direction of incident electrons. Therefore, in order to obtain 3D structure it is necessary to combine images from a number of different incidences. A general approach for this has been given in terms of Radon transformation or Fourier transformation[5]. In general, we can determine structure of 3D objects by combining thousands of projected images through "filtered back projection" - so called tomography. This is widely used in medical imaging and will become very useful for certain types of nano-structured materials as shown by J.M. Thomas et a1[6]. If the material is a crystalline, periodic system, we can apply crystallography instead of tomography. Using crystallography, we can dramatically (i) reduce the number of images required to only a few depending on crystal symmetry (the higher the crystal symmetry the fewer images are required) and (ii) enhance S/N ratio, because all structural information concentrates only on reciprocal points. Then, information on the periodically averaged structure is collected over the region of high-resolution transmissionelectron microscopy(HRTEM) image where the Fourier diffractogram(FD) was obtained. In order to determine the 3D structure of mesoporous silica crystals, resolution of ca. 10 A is necessary. In order to determine space-group(SG) uniquely, from the possible SGs obtained by the extinction conditions, hrther information is needed such as point-group(PG) symmetry, which can be obtained from a synthesised mesoporous crystal with a nice morphology. The 3D crystal structure analysis based on Fourier analysis of a set of HRTEM images, electron crystallography (EC), which we have developed may be the most powerful approach currently available for mesoporous silica crystals. The result shows electrostatic potential maps giving channellcage size and

connectivity, and therefore giving fundamental information of pore volume/shape and surface area without the need for presumed structural models[7- lo]. Once the 3D structures of mesoporous silica crystals are solved, they are very useful templates for the tailored synthesis of nano-structured materials as shown by R. Ryoo[ll] and D. Zhao[l2]. Understanding of the formation mechanism of mesoporous crystals should lead to fine tuning of the structure by suitable adjustment of the synthesis conditions, such as amphiphilic molecular species, silica sources, compositions, temperature and basic or acidic media. The formation process is another important aspect to be studied because the mesoporous silica growth is dominated by kinetics rather than occurring in equilibrium. Structures are thus dependent not only on thermodynamic parameters but also on the historylroute of the synthesis, and structural study on transformation and time evolution of mesoporous crystals is an essential subject[13,14,15]. In this report, we try to follow up our approaches for structure analysis of the mesoporous crystals through the use of electron microscopy, both TEM and scanning electron microscopy(SEM).

2. SCANNING ELECTRON MICROSCOPY FOR SILICA MESOPOROUS CRYSTALS SEM has advantages over TEM for the determination of crystal morphology and fine surface structures, as SEM has a large depth of focus and surface topology can be observed as different contrast in the image. Furthermore, it is a relatively simple experiment compared to TEM requiring minimum sample preparation. There are two typical crystal morphologies: equilibrium and growth forms. In the case of equilibrium form, the crystal and its surroundings are thermodynamically in equilibrium, and morphology is governed by minimum surface-energy under constant volume and is well described by Wulff construction. The growth form is governed by anisotropy of growth rates in different crystallographic directions and is strongly dependent on growth conditions. In both cases, crystal morphologies should be commensurate with

PG symmetry of the crystal. In order to observe morphology of mesoporous silica crystals by SEM images, metal coating can be applied to overcome the charging problems common for insulators. Fig. 2(a),(b),(c) and (d) show typical morphologies of 2D-hexagonal p6mm, of 3D-hexagonal P63/mmc and 3D-cubic with MCM-48(Ia%) and SBA- 1( ~ r n 3 n )types, respectively. Both SBA- 1 and MCM-48 are commensurate with PG of m3m, and fifty-four surfaces are indexed for SBA-1 as shown in the Fig. 2(e). If SEM images are taken without metal coating from an as-synthesised crystal at high resolution, it is possible to observe not only surface fine structure, including growth steps, but also channel and cage opening and their arrangement at the external surface[l6]. Recent progress in SEM, especially using high brilliant electron source with small energy spread by field emission gun (FEG) and an objective lens with small chromatic aberration, makes it possible to observe a high-resolution SEM(HRSEM) image at low accelerating voltage. Conditions for HRSEM observation will be reported separately[l7]. From HRSEM images of SBA-15 shown in Fig. 3(a), the presence of interconnections between the hexagonally packed mesoporous channels and fine details of surface termination are observed[l8]. The former is consistent with previous work based on TEM imaging of the Pt replica[4], and the latter shows that the surfaces perpendicular to the direction of the channels are intricately composed of tubes with closed ends, open ends, and channels that curl back to the inside of the particles, which are shown schematically in Fig. 3(b) and (c), respectively. The channel openings, growth steps and heights are observed in HRSEM images of SBA-16 taken almost along the three-fold [Ill], and Fig. 4(b) is an enlarged image of Fig. 4(a). It is reasonable to conclude from the images that the crystal grows layer-by-layer, that is, by accretion of lumps of silica onto faces of specific index and lateral crystal growth follows gradually [16].

Figure 2.

(el SEM images showing the typical molphologies of 2D-hexagonalp6mmb), 3D-

hexagonal Phdmmc(h) and 3D-cubic with MCM-4X(c) and SRA-l(d) types, respectively. Schematic drawing for SBA-l(e).

the channel

(a) (b) Figure 4. HRSEM image of SRA-16(a), (b) is an enlarged image of (a)

3. TWO-DIMENSIONAL SILICA MESOPOROUS CRYSTALS

The essential structural features of 2D mesoporous crystals can be observed by TEM images taken along the channel direction, if there is no structural order inside silica wall. Drs. Fukushima and Inagaki interested us in the structural study of FSM-16 by TEM when we were working on zeolite structures. An electron difTraction(ED) patter of FSM-16, synthesised from layered silicate kanemite, showed 6-fold symmetry along the channel this is difficult to reconcile with their "folded sheet" formation mechanism. From a series of sequential TEM observations of the material (Fig. S), it was reasonable to conclude that the kanemite sheets were dissolved in some sense and that derivatives of the sheet formed 2D-hexagonal p6mm structure. Prof. Kuroda interested us in studying the synthesis procedure of KSW-2, which was also synthesised from same the kanemite as for FSM-16 but under more moderate condition as separately discussed by him in this symposium. A set of HRTEM -

images of KSW-2 shows the "folded sheet mechanism" originally proposed for FSM-16, and a schematic drawing is also shown in Fig. 6.

(c)

Figure 5. HRTEM images of PSM-I6 at dilferent synthesis steps Inagaki succeeded in preparing new hybrid inorganic-organic mesoporous crystals. He further succeeded in synthesising a hybrid benzene-silica mesoporous crystal with a crystal-like pore wall and 3D-structure. Reflections both from mesoscopic and atomic orders were observed in an ED pattcm. In order to confirm the atomic-scale order, it was necessary to take an HRTEM image showing both were from the same originlcrystal and not from two independent regions. Fig. 7 clearly shows orders on different length scales, atomic and mesoscopic. It is noteworthy that in taking HRTEM images,

1 Clyrral Shape

1. Lamellar ~ h n s e Surfaerant lnlercailation

3. 2d-phnsetKSW-2) Id-Lozenge Channel

(a)

(b) Figure 6 . HRTEM images of KSW-2(a) and schematic drawing corresponding to the

images@).

Figure 7. H K E M image of hybrid organic-inorganic mesoporous crystal. ED patterns showing mcsoscopic scale and atomic scale orders are inserted in a right orientation with the image.

conditions of very under(or over)-focus and Scherzer-focus give strong contrast for mesoscopic and atomic order and the conditions are very different. This is because of the dependence of contrast transfer function(CTF) on defocus value Af as shown in Fig. 8, which will be discussed in the next section.

(b) Figure 8. CTF functions for different objective focus conditions(b). First zero-cross and position of 211 reflection(the most important for MCM-48), are shown by arrows.

4. STRUCTURE DETERMINATION (ELECTRON CRYSTALLOGRAPHY) FOR 3-D STRUCTURE

Mesoporous silica crystal is a 3D periodic array of cages or channels supported by continuous amorphous silica wall. The structure, electrostastic crystal potential V(r), (r = xa + yb + zc) in a unit cell, is obtained from scattering experiment through crystal structure factor (CSF), F(h), for h (h=ha* + kb* + lc* ) reflection, which is the scattering amplitude and Fourier coefficient of V(r) as; J V(r) exp 2ni h r dr = F(h) = I F(h) / exp{ i 8(h) ), (eq. 1) where 6(h) is the phase of CSF for h and is a function of the coordinates of the origin. F(h) is complex in general. Only absolute values, moduli, / F(h) / can be obtained from diffraction intensity, I(h), for reflection h as given in eq.2 and phase information disappears. I(h)= F(h)* F(h)= [ I F(h) I 12. (es. 2) After obtaining phases of CSF, q h ) , by some methods, structure V(r) can be determined by an inverse Fourier transform straightforwardly as V(r) = J F(h) exp(- 2xir h) dh . (eq. 3) For a centrosymmetric crystal, by taking an origin at an inversion centre, 6(h) will be either O(+) or n(-) and F(h) becomes real. In the phase object approximation, electron wave function at the specimen exit surface for a crystal of thickness t, that is, the exit wave function is given by (x>Y)= exp [io t Vp (x,y)l (es. 4) where the projected potential for a crystal thickness t, Vp (x,y) t =Jot V(x,y,z)dz, and the interaction parameter o is equal to 0.00653 [~olt''nm-'1at 300 kV. 4r

For a thin specimen, weak phase object(WP0) approximation is applied and eq.4 is approximated by

4t (x,y) = 1 + i o t V, (x,y) .

(es.

5)

In the back focal plane of the object lens, the wave function becomes: F h (u,v) = F T { G r (x,Y) ) = 6 (u,v) + i o t V~(U,V) (eq. 6) where F T means Fourier transformation, and u and v are coordinates in reciprocal space.

Fh (u, V) = F(h) = (2xme/h2 ) Vh(u, V) . (eq. 7) In the image plane, the wave function @ (x,y) is modified through the objective lens and is given by @ (-,Y) = FT

{ F h (u, V) exp[i x(u, v)l ) (es. 8) x(u, v) = JC (C,h3(u2 + v2j2/4-Af h (u2 + v2)/2 ), (eq. 9) where C, and Af are a spherical aberration coefficient of objective lens and defocus value, respectively. The function sinx(u,v) is know as the contrast

transfer function(CTF) and shows the transfer ability of the objective lens in structural details. If the CTF = -1 for wide range of (u,v) would be ideal however it is a complex finction of C, , Af, u and v. A problem induced by this CTF in mesoporous crystals will be shown later. Then observed image, Image I(x,y,), will be given as:

The Fourier diffractogram of the HREM image, Iimag,(h),taken from such a thin region, is: Iimage(h) = FT {I(x,y)) = 2 0 t {F(h) / (2zme/h2)) sin X(U,V). (eq- 11) Therefore, Iimage(h)is proportional to the crystal structure factor F(h) and thickness t multiplied by the CTF. So if WPO is applicable, crystal structure factors can be obtained through Fourier transformation of HREM image after CTF correction, which is calculated with C, and A$ It is very important to note that HREM images should be taken from thin areas to fulfill the condition of WPO and at the same time to obtain the genuine extinction rule for the space-group determination through the Fourier diffractogram(FD), which can be obtained with enough resolution and intensity from about 10 x 10 times the unit cell size.

-'

5. SOME 3D-STRUCTURE SOLUTIONS The first 3D structure was reported by Kresge et a1 and they claimed the structure had cubic symmetry(la%') from a similarity in powder XRD pattern with that of a cubic liquid-crystal phase obtained by Luzzati[20]. Here we report our study of cubic mesoporous crystal with Ia3d symmetry in detail followed by

some of other typical structure solutions of 3D mesoporous silica crystals[21,22,23]. The ratio of observed d-spacings among the reflections in the powder XRD pattern is approximately 6Ii2 : 8" : 14In : 4 : 20In : 22" : 24'12 : 26". These reflections can be indexed as 211, 220, 321, 400, 420, 332, 422, 431 and so on, if we assume cubic crystal (at small scattering vector, experimental error in peak positions is relatively large). This is commensurate with IaTd, and it is very rare to observe so many reflections - in most cases observable peaks are limited to between 2 to 4 reflections. Therefore it is difficult to dcduce cvcn the crystal system solely from powder XRD pattern. Using TEM, we can obtain single crystal structural information either through ED patterns or HRTEM images. However, it is difficult to obtain an ED pattern free from multiple diffraction effects by the selected area ED method. This is because the minimum size of selected area aperture is approximately 200 nm and this is too large to obtain thin specimen information selectively. If we use FD of IIRTEM image instead of ED pattern, we can obtain diffraction information only from a thin area, which is much smaller than the aperture size but enough to give FD as mentioned in section 3. The difference between ED and FD for finding the extinction condition is shown as an example in Fig. 9. (200) reflections, which are forbidden for IaTd, are observed in ED pattern.

(a) Figure 9. ED(a) and FD(b) patterns of MCM-48.

(b)

(a) (b) Figure 10. HRTEM images of mesoporous silica crystal with IuTd(a) and its carbon

replica(b) taken along [ I l l ] .Corresponding FDs are inserted.

(a) (b) Figure 11. H R E M images of the same crystal as Fig. 10 taken along [1001(a)and [110](b).

Fig. 10 shows HREM images of the large pore mesoporous silica crystal (Ia7d) and its carbon replica made using the mesoporous silica as a template, taken with [ I l l ] incidence together with corresponding FDs. It is clear that both FDs are identical patterns, although structures are obviously complementary

being the template and the product (this is known in optics as Babinet's principle). The reason this is shown here and not in section 7 is to mention that the phase-relation of crystal structure factor(CSF) makes the structure unique from many possible structures. Figure 11 (a) & (b) show HREM images togethcr with corresponding FDs from the same mcsoporous silica crystal shown in Fig. 10. It is clear from Fig. 10 and 11 that HREM images of [100],[I101 and [ I l l ] incidences show plane groups ofp4mm, c2mm and p6mm, respectively, and that reflections with h+k+l = odd and in addition other reflections, such as 110, 200, 310, 222, 226, 2210, 334 are extinct. Therefore, possible reflection conditions are hkl: h+k+l =2n, Okl: k,l =2n, hhl: 2h+l=4n, h0O: h=4n. From these observations, the space-group symmetry was uniquely determined to be la3d. A 3D-data set of CSF of the mesoporous silica crystal with large pores was obtained by merging 2D-CSF data sets obtained from FDs of [100],[I101 and [ I l l ] incidences. The 3D-electrostatic potential-distribution was obtained by inverse FT. Fig. 12 shows the maps for the MCM-48 viewed along [loo] at the section of (a) z = 0 and (b) z = 118, where the origin was taken at the centre of inversion 3 point symmetry. Sections of 3D-periodic minimal Gyroid surface are overwritten by solid curves. It is clear that the silica wall exactly followed the Gyroid surface and that it separates two independent and interwoven channels with right and left handed chirality. We estimate the silica wall thickness to be ca 11 by taking N2 adsorption volume data and assuming an amorphous silicawall density of 2.2 gem". From similar analysis ofthe

(a) (b) Figure 12. Electrostatic potential density map obtained for MCM-48

mesoporous silica crystal with large pore, we can observe new complementary pores, which will interconnect two originally independent channel systems in MCM-48, at the special flat-point on the G-surface. This will be published separately [24]. In most cases, the extinction rules obtained from a series of FDs give possible space groups. Examples will be shown later for PmTn (SBA-1 and -6)

(c)

Figure 13. A sct of HRTEM images obtained from SBA-6, [1001(a),[110](b)and [ I l l ] ( c ) . Corresponding FDs are inserted.

and ImTm (SBA-16), and in these cases combining PG symmetry m7m from crystal morphology we could determine their SGs unambiguously. A set of HREM images of SBA-6 is shown in Fig. 13 together with corresponding FDs. From the observed extinction rule, both PmTn and PT3n were possible SGs. However, P m h was uniquely determined, as the crystal morphology suggested PG to be m7m. The 3D-electrostatic potential distribution map obtained is also unique solution for the structure. From N2 adsorption data and the silica density, the 3D silica structure of SBA-6 is determined as shown in Fig. 14(a). There are two cages, A and B, with different diameters and the cages are arranged in AIR

(b) Figure 14. Electrostatic potential density map obtained for SBA-6(a) and 3D-structure of SBA-6 (b).

configuration as shown in Fig. 14(b), where the A-cage is the larger with a diameter of 85 A at (1/2,0,1/4), (1/2,0,314), (0,114,112), (0,314,112), (114,112,O) and (314,112,0), and the B-cage is the smaller with a diameter of 73 A at (0,0,0) and (1/2,112,1/2). A B-cage is surrounded by 12 A-cages that are connected through openings of 20 A, while the openings between A-cages are about 33 x 41 A. Using the same approach, we have solved the 3D-structure of SBA-16 with Im5m symmetry. Using anionic surfactants and co-structure directing agents, we have recently reported a novel synthesis of mesoporous silica crystals, AMS-n, and have succeeded in characterising their structures[25,26]. Through choice of synthesis conditions, many different 3D-structures can be synthesised systematically. Here two HREM images as examples are shown in Fig. 15 to show high crystalline order in AMS-8 and to highlight the possibility of producing new structure types in AMS-2. Furthermore, we have recently succeeded in the synthesis and structural characterisation or chiral mesoporous silica crystal, and this will be reported separately[27].

(a) Figure 15. HRTEM images of AMS-2(a) and AMS-8(b).

6. STRUCTURAL TRANSFORMATIONS Structural transformations from one structure to another have been reported by powder XRD experiment before and after transformation, though the structural relationships occurring during the change were not be observed experimentally. The transformation among M41S, that is, lamellar, 2D-hexagonal p6mm and 3Dcubic IaTd, is a typical case. An in situ XRD experiment gives better understanding of the transformation, however, the basic issues underlying these transformations, such as the epitaxial relationships, remain unclear. This was studied by TEM, and the results on M41 S were reported[l2,15]. Here, HRTEM study of the transformation from 2D-hexagonal p6mm structure to cubic PmTn structure observed in SBA-1 is shown[l4]. We showed in a series of powder XRD patterns that the p6mm structure was first formed and the structure gradually changed to PmSn with increase of synthesis time. Both 2D- p6mm and 3D-Pm7n structures were observed right top and left-bottom in an HRTEM image taken from an intermediate sample(Fig. 16). Lattice fringes for I0 plane of p6mm and 211 plane of PmSn were drawn by lines. This showed that the structural change occurred not via a dissolution-recrystallisation process but via a solid-solid transformation, that is, the (211) plane of the cubic phase was formed via the topological changes involving silica restructuring along the cylinder axis of the 2D-hexagonal p6mm structure by keeping an epitaxial relation. The { I 0 ) and (211) reflections are the most important waves to produce p6mm and PmSn structures in Fourier sum(reconstruction), respectively. The CSF F(h,k,l) has the following phase relations for Pm7n and IaTd, F(h, k, l) = F(-h, -k,-1) = F(-h, k, 1) = F(h, -k, I ) = F(h, k,-1) for PmTn, and F(h,k,l) = F(-h,-k,-l) = -F(-h,k,l) =- F(h,-k,l) = F(h,k,-l) for (211) in 1a5d. When a mesoporous crystal transforms from one structure to another, the corresponding structural modulation, which can be described by waves (i.e., modulation waves), will become stable with time. We believe the phase transformation fiomp6mm to either PmSn or IaTd is induced by the same (211) waves but phase relations among them are different.

Figure 16. HRTEM image showing epitaxial crystal transformation from 2D-hexagonal to 3D-cubic. 10 plane ofp6mm and 211 plane ofPmTn arc drawn by lines.

7. NANO-STRUCTURED MATERIALS SYNTHESIZED WITHIN PORES We have been interested in electronic states of materials confined in periodically arrayed cagest281 or curved geometries Tor a long time and recently particularly in the latter case[29]. We now have real systems where the electrons are confined in well-defined geometries. Using mesoporous silica crystals with p6mm and IaTd structures as templates, carbon-, Pt- and metaloxides-nanowires were synthesised in their spaces, and their structures were studied by TEM. Here, the Pt-nanowires case is considered individually. The details of synthesis of Pt-nanowires can be found in the original paper by R. Ryoo et al[l 11. Fig. 17(a) and (b) are HRTEM images of Pt-nanowires extracted from channels of MCM-41 and SBA-15, respectively. The length of Pt-nanowires extracted from both MCM-41 and SBA-15 ranged from several tens to several hundreds of nanometers. The Pt-nanowires extracted from MCM-41 are single crystal-nanowires with fairly smooth surfaces. In

Figure 17. HRI'EM images of Pt-nanowires extracted from MCMdl(a) and SEA-IS@).

contrast, although the Pt-nanowires manufactured in the channels of calcined SBA-15 are close to single crystals, two different aspects Crom those synthesiscd in MCM-41 are clearly observed: (1) the outer-surfaces projections of the Pt rods are not straight but smoothly curved; (2) there are bridges between adjacent rods and small protrusions on the surfaces. As Pt-nanowires are replicas, the above two points are inherited from the channel structure with thc complementary pores of SBA-15 (Fig. 3(b)). Fig. 18(a) and (b) show TEM images taken by high-voltage, HREM(JEM-1250) from the same Pt-nanowires extracted from MCM-48 taken with the (a) [loo]and the (b) [ I l l ]incidences, respectively. It is clear &om these images that Pt-nanowires occupy one channel system, either right- or left-handed chirality except the area where the two channel systems are occupied simultaneously, pointed by arrows. An IIR'IEM image of the Pt-nanowire in one of the channels, which correspond to a space group 14132 is shown in Fig. 19 taken with [loo]incidence, which is along the four-fold screw axis, on a meso-scale crystal structure. In atomic-scale, the Pt-

(a) (b) Figure 18. TEM images of Pt-nanowire extracted from MCM-48 of [I001 (a) and [lll](b).

Figure 19. HRTEM imagc of Pt-nanowire extracted from MCM-48 of[100] together with a FD.

Figure 20. A pair of stereographic HRSEM imagc of Pt-nanowire to show 3D-stri~cturc.

Schematic Powder X R D patterns

A', i

Figure 21.

Cu Ku

(bf A schematic diagram to show allowed wave vectors in the material with atomic

and mesoscopic orders.

nanowire itself shows single crystalline feature with <110> atomic arrangement of face centred cubic, fcc, (a FD pattern is inserted). The handedness can be simply determined through a set of HRSEM images by tilting the Pt-nanowire, "topography ", and the images are shown in Fig. 20. The reason for the diffuse nature of the spots observed in the inserted FD (Fig. 19) is that the reciprocal points of the Pt-nanowire are a convolution of that for bulk Pt-single crystal with that for mesoscopic order. This is shown schematically in Fig. 21, and the powder XRD profile of this material gives rise to a few very difhse peaks, which clarifies one new problem to be solved.

8. FUTURE New nano-structured materials will provide new interesting physical properties, which are characteristic of electron confinement in curved spacelrod with 3D periodicity. First we must synthesise and solve the structure of well-crystalline materials with atomic and mesoscopic scale orders by developing a new approach to solving the problem as mentioned in the above section. We have recently proposed a new approach of diffraction-based 3Dmicroscopy[30]. This is by taking a tilting series of ED patterns with coherent beams (for example, from -70 deg to 70 deg in 5 deg increments along a single rotation axis), to obtain 3D atomic scale structures of nano-structured materials and to overcome resolution barriers inherent in HREM and tomography. By combining coherent ED patterns with the oversampling phasing method, we hope to show its power by solving the actual 3D structure of a nano-structured material. In-situ XRD experiments provide very important information of crystal growth or structural transformation in mesoporous crystals. The advantage of EM lies in the ability to show local spatial/structural information and it will be a new approach to study structures of non-periodic system or of "softer" material and crystal structures of time evolution by a "snap shot" or "freezing" TEM observation complementary to the "in situ" XRD experiment.

9. CONCLUSIONS It has been shown that EM is a very powerful approach for characterising mesoporous crystal structures and nano-structured materials by a collection of examples together with some basic background. Recent progress in EC for the structural solutions has been given. As we have novel materials with orders both at atomic and mesoscopic scales, we should continue to develop (crystallographic) methodologies for such materials.

ACKNOWLEDGMENT The authors acknowledge many collaborators who have contributed to the original papers, especially Profs. S. Inagaki, K. Kuroda, R. Ryoo, D. Zhao, G. Stucky, who are separately presenting in this symposium, and Prof. S. Che and T. Tatsumi. Financial supports from the Swedish Research Council VR and Japan Science and Technology Agency(JST) and Bio Nanotec Research Institute(BNRI), Japan are acknowledged.

REFERENCES [l] T. Yanagisawa, T. Shimizu, K. Kuroda & C. Kato, Bull. Chem. Soc. Jpn. 988-992 (1990). [2] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli & J.S. Beck, Nature 359, 710-712. [3] S. Inagaki, Y. Fukushima & K. Kuroda, Chem. Soc. Chem. Commun.680-682 (1993). [4] Z. Liu, 0 . terasaki, T. Ohsuna, K. hiraga, H.J. Shin & R. Ryoo, ChemPhysChem 2001, 229-23 1.

[ 5 ] DeRosier & A. Klug; Nature 217, 130(1968), R.A. Crowther, D.J. DeRosier & A. Klug Proc. Roy. Soc. Lond. A3 17, 3 19-340(1970) [6] P.A. Moidgley, J.M. Thomas, R. Raja, L. Laffont, B.F.G. Johnson & T. khimyak, J. Phys, Chem. B, (2004), in press. [7] 0 . Terasaki, Molecular Sieves, vol. 2, Heidelberg; Springer-Verlag,l999, pp 7 1- 1 12. [S] 0. Terasaki & T. Ohsuna, Handbook of Zeolite Science and Technology, eds by S.M. Auerbach, K.A. Carrado & P.K. Dutta, Marcel Dekker, Inc., 2003, pp291-3 15. [9] 0 . Terasaki & T. Ohsuna, , Studies in Surface Science and Ctalysis 135(2001), 61-71. [lo] 0. Terasaki, T. Ohsuna, Z. Liu, M. Kaneda, S. Kamiya, A. Carlsson, T. Tsubakiyama, Y.

Sakamoto, S. Inagaki, S. Che, T. Tatsumi, M.A. Camblor, R. Ryoo, D. Zhao, G. Stucky, D. Shindo & K.Hiraga, Studies in Surface Science and Catalysis 141(2002), 27-34. [ l l ] R. Ryoo, S.H. Joo, S. Jun, J. Phys. Chem. B103(1999),7743; Z. Liu, Y. Sakamoto, T. Ohsuna, K. Hiraga, 0 . Terasaki, C.H. KO, H.J. Shin & R. Ryoo, Angew. Chem., Int. Ed.,39(2000), 3107-3110.; H.J. Shin, R. Ryoo, Z. Liu & 0 . Terasaki, J. Am. Chem. SOC.123(2001), 1246 [12] B. Tian, X. Liu, L.A. Solovyov, Z. Liu, H. Yang, Z. Zhang, S. Xie, F. Zhang, B. Tu, C. Yu, 0 . Terasaki & D. Zhao, J. Am. Chem. Soc. 126(2004), 865-875. [13] Landry, S.H. Tolbert, K.W. Gallis, A. Monnier, G.D. Stucky, P. Norby & J.C. Hanson, Chem. Mater., 13(2001), 1600-1608. [14] S. Che, S. Kamiya, 0 . Terasaki, T. Tatsumi, J. Am. Chem. Soc. 123(2001), 12089-1290; S. Kamiya, H. Tanaka, S. Che, T. Tatsumi & 0. Terasaki, J. Solid. State Sciences5(2003), 197-204. [15] I.Diaz, J.Perez-Pariente & 0 . Terasaki, J. Mater. Chem. 14 (2004) 48-53. [16] 0.Terasaki & R. Ryoo, to be submitted. [17] S. Namba, 0 . Terasaki, to be submitted. [IS] S. Che, K. Lund, T. Tatsumi, S. Iijima, S.H. Joo, R. Ryoo & 0. Terasaki, Angew. Chem. Int. Ed., 42(2003), 2182-2185. [19] T. Kimura, T. Kamata, M. Fuziwara, Y. Takano, M. Kaneda, Y. Sakamoto, 0 . Terasaki, Y. Sugahara & K. Kuroda. Angew. Chem. Int. Ed. 39(2000), 3855-3859. [20] V. Luzatti & P. P. A. Speght, Nature 21 5(1967),701-704. [21] A Carlsson, M Kaneda, Y Sakarnoto, 0 Terasaki, R Ryoo, SH Joo. J Electron Microscopy 48: 795-798, 1999. [22] Y Sakamoto, M Kaneda, 0 Terasaki, DY Zhao, JM Kim, G Stucky, HJ Shin, R Ryoo, Nature 408: 449-453,2000. [23] M. Kaneda, T. Tsubakiyama, A. Carlsson, Y. Sakamoto, T. Ohsuna, 0. Terasaki, S.H. Joo, R. Ryoo. J. Phys. Chem. B106: 1256-1266,2002. [24] Y. Sakamoto, T. W. Kim, R. Ryoo & 0 . Terasaki, to be submitted. [25] S. Che, A.E. Garcia-Bennett, T. Yokoi, K. Sakamoto, H. Kunieda, 0 . Terasaki & T. Tatsumi, NatureMaterials 2(2003), 801-805. [26] A.E. Garcia-Bennett, 0 . Terasaki, S. Che & T. Tatsumi, Chem. Mater. 16(2004), 8 13-821. [27] S. Che, Z. Liu, T. Ohsuna, K. Sakamoto, 0 . Terasaki & T. Tatsumi, To be published. [28] Y Nozue, T. Kodaira, S. Ohwashi, T. Goto & 0 . Terasaki, Phys. Rev 48B(1993), 1225312261, Y. Nozue, T. Kodaira, 0 . Terasaki & H. Takeo, Springer Proceedings in Physics,

vol. 8 1, "Materials and Measurements in Molecular Electronics", (1996), 15 1- 162. [29] H. Aoki, M. Koshino, D. Takeda, H. Morise & K. Kuroki, Phys. Rev B65(2001), 035102; N. Fujita: Spin & Charge Transport in Nanostmctures, Braga 2003; N. Fujita & 0. Terasaki in preparation. [30] J. Miao, T. Ohsuna, 0.Terasaki & M. O'Keefe, Phys. Rev Letters 89(2002), 155502.

STUDIES IN SURFACE SCIENCE A N D CATALYSIS Advisory Editors: B. Delmon, Universite Catholique de Louvain, Louvain-la-Neuve, Belgium J.T. Yates, University of Pittsburgh, Pittsburgh, PA, U.S.A.

Volume 1

Volume 2

Volume 3

Volume 4

Volume 5

Volume 6

Volume 7

Volume 8 Volume 9

Volume 1 0

Volume 11

Volume 1 2

Volume 1 3 Volume 1 4

Volume 1 5

Preparation of Catalysts I. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the First lnternational Symposium, Brussels, October 14-1 7,1975 edited by B. Delmon, P.A. Jacobs and G. Poncelet The Control of the Reactivity of Solids. A Critical Survey of the Factors that Influence the Reactivity of Solids, w i t h Special Emphasis on the Control of the Chemical Processes in Relation t o Practical Applications by V.V. Boldyrev, M. Bulens and B. Delmon Preparation of Catalysts II.Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Second lnternational Symposium, Louvain-laNeuve, September 4-7, 1 9 7 8 edited b y B. Delmon, P. Grange, P. Jacobs and G. Poncelet Growth and Propertiesof Metal Clusters. Applications t o Catalysis and the Photographic Process. Proceedings of the 32"d lnternational Meeting of the Societe de Chimie Physique, Villeurbanne, September 24-28, 1 9 7 9 edited by J. Bourdon Catalysis by Zeolites.Proceedings of an lnternational Symposium, Ecully (Lyon), September 9-1 1, 1 9 8 0 edited by B. Imelik, C. Naccache,Y. BenTaarit, J.C.Vedrine, G. Coudurier and H. Praliaud Catalyst Deactivation. Proceedings of an lnternational Symposium, Antwerp, October 13- 15.1 9 8 0 edited by B. Delmon and G.F. Froment New Horizons in Catalysis. Proceedings of the 7'h lnternational Congress on Catalysis, Tokyo, June 30-July 4, 1980. Parts A and B edited by T. Seiyama and K.Tanabe Catalysis by Supported Complexes by Yu.l.Yermakov, B.N. Kuznetsov and V.A. Zakharov Physics of Solid Surfaces. Proceedings of a Symposium, BechyAe, September 29-October 3.1 9 8 0 edited by M. Lknitka Adsorption at the Gas-Solid and Liquid-Solid Interface. Proceedings of an lnternational Symposium, Aix-en-Provence, September 21 -23, 1 9 8 1 edited by J. Rouquerol and K.S.W. Sing Metal-Support and Metal-Additive Effects in Catalysis. Proceedings of an lnternational Symposium, Ecully (Lyon), September 14-1 6, 1 9 8 2 edited by B. Imelik, C. Naccache, G. Coudurier, H. Praliaud, P. Meriaudeau, P. Gallezot, G.A.Martin and J.C.Vedrine Metal Microstructures in Zeolites. Preparation - Properties - Applications. Proceedings of a Workshop, Bremen, September 22-24, 1 9 8 2 edited by P.A. Jacobs, N.I. Jaeger, P. JirD and G. Schulz-Ekloff Adsorption on Metal Surfaces. A n Integrated Approach edited b y J. Benard Vibrations at Surfaces. Proceedings of the Third lnternational Conference, Asilomar, CA, September 1-4, 1 9 8 2 edited by C.R. Brundle and H.Morawitz Heterogeneous Catalytic Reactions Involving Molecular Oxygen by G.I. Golodets

Volume 16

Volume 17

Volume 18

Volume 1 9

Volume 2 0

Volume 21

Volume 2 2 Volume 23 Volume 2 4

Volume 25

Volume 26

Volume 27 Volume 28

Volume 29 Volume 3 0

Volume 31

Volume 3 2 Volume 3 3 Volume 3 4

Volume 35

Preparation of Catalysts Ill.Scientific Bases for the Preparation of Heterogeneous Catalvsts. Proceedings of the Third lnternational Symposium, Louvain-la-Neuve. . . september 6-9, 1982 edited by G. Poncelet, P. Grange and P.A. Jacobs Spillover of Adsorbed Species. Proceedings of an lnternational Symposium, Lyon-Villeurbanne, September 12-1 6, 1983 edited by G.M. Pajonk,S.J.Teichner and J.E. Germain Structure and Reactivity of Modified Zeolites. Proceedings of an lnternational Conference, Prague, July 9-1 3, 1984 edited by P.A. Jacobs, N.I. Jaeger, P. JirG, V.B. Kazansky and G. Schulz-Ekloff Catalysis on the Energy Scene. Proceedings of the grh Canadian Symposium on Catalysis, Quebec, P.Q., September 30-October 3, 1984 edited by S. Kaliaguine and A.Mahay Catalysis by Acids and Bases. Proceedings of an lnternational Symposium, Villeurbanne (Lyon), September 25-27, 1984 edited by B. Imelik, C. Naccache, G. Coudurier,Y. Ben Taarit and J.C.Vedrine Adsorption and Catalysis on Oxide Surfaces. Proceedings of a Symposium, Uxbridge, June 28-29, 1984 edited by M. Che and G.C.Bond Unsteady Processes in Catalytic Reactors by Yu.Sh. Matros Physics of Solid Surfaces 1984 edited by j.Koukal ~eolites:~~nthesis, Structure,Technology and Application. Proceedings of an Portoroi -Portorose, September 3-8, 1984 lnternational Symposium, . . edited by B.Driaj,S. Hotevar and S. Pejovnik Catalytic Polymerization of Olefins. Proceedings of the lnternational Symposium on Future Aspects of Olefin Polymerization, Tokyo, July 4-6, 1985 edited by T.Keii and K.Soga Vibrations at Surfaces 1985. Proceedings of the Fourth lnternational Conference, Bowness-on-Windermere, September 15-1 9, 1985 edited by D.A. King, N.V. Richardson and S. Holloway Catalytic Hydrogenation edited by L. Cerveny New Developments in Zeolite Science and Technology. Proceedings of the 7th lnternational Zeolite Conference, Tokyo, August 17-22, 1986 edited by Y.Murakami,A. lijima and J.W.Ward Metal Clusters in Catalysis edited by B.C. Gates, L. Guczi and H. Knozinger Catalysis and Automotive Pollution Control. Proceedings of the First lnternational Symposium, Brussels, September 8-1 1, 1986 edited by A. Crucq and A. Frennet Preparation of Catalysts IV. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Fourth lnternational Symposium, Louvain-laNeuve, September 1-4, 1986 edited by B. Delmon, P. Grange, P.A. Jacobs and G. Poncelet Thin Metal Films and Gas Chemisorption edited by P.Wissmann Synthesis of High-silica Aluminosilicate Zeolites edited by P.A. Jacobs and J.A.Martens Catalyst Deactivation 1987. Proceedings of the 4'h lnternational Symposium, Antwerp, September 29-October 1, 1987 edited by B. Delmon and G.F. Froment Keynotes in Energy-Related Catalysis edited by 5. Kaliaguine

Volume 3 6

Volume 3 7

Volume 3 8

Volume 3 9

Volume 4 0

Volume 4 1

Volume 4 2 Volume 4 3 Volume 4 4

Volume 4 5 Volume 4 6

Volume 4 7 Volume 4 8

Volume 4 9

Volume 5 0

Volume 5 1 Volume 5 2

Volume 5 3

Volume 5 4

Methane Conversion. Proceedings of a Symposium on the Production of Fuels and Chemicals from Natural Gas, Auckland, April 27-30, 1 9 8 7 edited by D.M. Bibby, C.D.Chang, R.F.Howe and S.Yurchak Innovation in Zeolite Materials Science. Proceedings of an lnternational Symposium, Nieuwpoort, September 13-1 7, 1 9 8 7 edited by P.J. Grobet, W.J.Mortier, E.F.Vansant and G. Schulz-Ekloff Catalysis 1987.Proceedings of the lothNorth American Meeting of the Catalysis Society, San Diego, CA, May 17-22, 1 9 8 7 edited by J.W.Ward Characterization of Porous Solids. Proceedings of the IUPAC Symposium (COPS I), Bad Soden a. Ts., April 26-29.1 9 8 7 edited by K.K.Unger, J. Rouquerol, K.S.W.Sing and H. Kral Physics of Solid Surfaces 1987. Proceedings of the Fourth Symposium on Surface Physics, Bechyne Castle, September 7-1 1, 1 9 8 7 edited by J.Koukal Heterogeneous Catalysis and Fine Chemicals. Proceedings of an lnternational Symposium, Poitiers, March 15-1 7, 1 9 8 8 edited by M. Guisnet, J. Barrault, C. Bouchoule,D. Duprez, C. Montassier and G. Perot Laboratory Studies of HeterogeneousCatalytic Processes by E.G. Christoffel, revised and edited by Z. Pail Catalytic Processes under Unsteady-State Conditions by Yu. Sh. Matros Successful Design of Catalysts. Future Requirements and Development. Proceedings of the Worldwide Catalysis Seminars, July, 1988, on the Occasion of the 3Oth Anniversary of the Catalysis Society of Japan edited by T. lnui Transition Metal Oxides. Surface Chemistry and Catalysis by H.H.Kung Zeolites as Catalysts, Sorbents and Detergent Builders. Applications and Innovations. Proceedings of an lnternational Symposium, Wurzburg, September 4-8,1988 edited by H.G. Karge and J.Weitkamp Photochemistry on Solid Surfaces edited by M.Anpo and T. Matsuura Structure and Reactivity of Surfaces. Proceedings of a European Conference, Trieste, September 13-1 6, 1 9 8 8 edited by C.Morterra,A. Zecchina and G. Costa Zeolites: Facts, Figures, Future. Proceedings of the 8'h lnternational Zeolite Conference, Amsterdam, July 10-1 4, 1 9 8 9 . Parts A and B edited by P.A. Jacobs and R.A. van Santen Hydrotreating Catalysts. Preparation, Characterization and Performance. Proceedings of the Annual lnternational AlChE Meeting, Washington, DC, November 27-December 2, 1 9 8 8 edited by M.L. Occelli and R.G.Anthony New Solid Acids and Bases.Their Catalytic Properties by K.Tanabe, M. Misono,Y. Ono and H. Hattori Recent Advances in Zeolite Science. Proceedings of the 1 9 8 9 Meeting of the British Zeolite Association, Cambridge, April 17-1 9, 1 9 8 9 edited by J. Klinowsky and P.J. Barrie Catalyst in Petroleum Refining 1989. Proceedings of the First lnternational Conference on Catalysts in Petroleum Refining, Kuwait, March 5-8, 1 9 8 9 edited b y D.L.Trimm,S.Akashah, M.Absi-Halabi and A. Bishara Future Opportunities in Catalytic and Separation Technology edited by M. Misono,Y.Moro-oka and S. Kimura

292 Volume 55

Volume 56

Volume 57A

Volume 578

Volume 5 8 Volume 59

Volume 6 0

Volume 61

Volume 6 2

Volume 6 3

Volume 6 4 Volume 65

Volume 6 6

Volume 6 7

Volume 6 8

Volume 6 9

Volume 7 0

New Developments in Selective Oxidation. Proceedings of an lnternational Symposium, Rimini, Italy, September 18-22, 1989 edited by G. Centi and F.Trifiro Olefin Polymerization Catalysts. Proceedings of the lnternational Symposium on Recent Developments in Olefin Polymerization Catalysts, Tokyo, October 23-25, 1989 edited by T.Keii and K.Soga Spectroscopic Analysis of HeterogeneousCatalysts. Part A: Methods of Surface Analysis edited by J.L.G. Fierro Spectroscopic Analysis of HeterogeneousCatalysts. Part B: Chemisorption of Probe Molecules edited by J.L.G. Fierro Introduction to Zeolite Science and Practice edited by H. van Bekkum, E.M. Flanigen and J.C. Jansen Heterogeneous Catalysis and Fine Chemicals 11. Proceedings of the Znd lnternational Symposium, Poitiers, October 2-6, 1990 edited by M. Guisnet, J. Barrault, C. Bouchoule,D. Duprez, G. Perot, R.Maurel and C. Montassier Chemistry of Microporous Crystals. Proceedings of the lnternational Symposium on Chemistry of Microporous Crystals, Tokyo, June 26-29, 1990 edited by T. Inui,S. Namba and T.Tatsumi Natural Gas Conversion. Proceedings of the Symposium on Natural Gas Conversion, Oslo, August 12-1 7, 1990 edited by A. Holmen, K.-J. Jens and S.Kolboe Characterization of Porous Solids II.Proceedings of the IUPAC Symposium (COPS II), Alicante, May 6-9, 1990 edited by F. Rodriguez-Reinoso, J. Rouquerol, K.S.W. Sing and K.K.Unger Preparation of Catalysts V. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Fifth lnternational Symposium, Louvain-la-Neuve, September 3-6, 1990 edited by G. Poncelet, P.A. Jacobs, P. Grange and B. Delmon New Trends in CO Activation edited by L. Guczi Catalysis and Adsorption by Zeolites. Proceedings of ZEOCAT 90, Leipzig, August 20-23, 1990 edited by G. ~hlmann,H. Pfeifer and R. Fricke Dioxygen Activation and Homogeneous Catalytic Oxidation. Proceedings of the Fourth lnternational Symposium on Dioxygen Activation and Homogeneous Catalytic Oxidation, Balatonfijred, September 10-1 4, 1990 edited by L.I. Simandi Structure-Activity and Selectivity Relationships in HeterogeneousCatalysis. Proceedings of the ACS Symposium on Structure-Activity Relationships in Heterogeneous Catalysis, Boston, MA, April 22-27, 1990 edited by R.K. Grasselli and A.W.Sleight Catalyst Deactivation 1991. Proceedings of the Fifth lnternational Symposium, Evanston, IL, June 24-26, 1991 edited by C.H. Bartholomew and J.B. Butt Zeolite Chemistry and Catalysis. Proceedings of an lnternational Symposium, Prague, Czechoslovakia, September 8-1 3, 1991 edited by P.A. Jacobs, N.I. jaeger, L.Kubelkova and B.Wichterlova Poisoning and Promotion in Catalysis based on Surface Science Concepts and Experiments by M. Kiskinova

Volume 71

Catalysis and Automotive Pollution Control 11. Proceedings of the Znd lnternational Symposium (CAPoC 2), Brussels, Belgium, September 10-1 3,

1990 Volume 72

Volume 7 3

Volume 74 Volume 75

Volume 76 Volume 77

Volume 78

Volume 79

Volume 8 0

Volume 81

Volume 82

Volume 8 3

Volume 8 4

Volume 85 Volume 86 Volume 87

edited b y A. Crucq New Developments in Selective Oxidation by HeterogeneousCatalysis. Proceedings of the 3fdEuropean Workshop Meeting on N e w Developments in Selective Oxidation by Heterogeneous Catalysis, Louvain-la-Neuve, Belgium, April 8-10, 1991 edited by P. Ruiz and B. Delmon Progress in Catalysis. Proceedings of the 12Ih Canadian Symposium on Catalysis, Banff, Alberta, Canada, May 25-28, 1992 edited by K.J. Smith and E.C. Sanford Angle-Resolved Photoemission.Theory and Current Applications edited by S.D. Kevan New Frontiers in Catalysis, Parts A-C. Proceedings of the 1Oth lnternational Congress on Catalysis, Budapest, Hungary, 19-24 July, 1992 edited by L. Guczi, F. Solymosi and P.Tetenyi Fluid Catalytic Cracking: Science and Technology edited b y J.S.Magee and M.M. Mitchell, Jr. New Aspects of Spillover Effect in Catalysis. For Development of Highly Active Catalysts. Proceedings of the Third lnternational Conference on Spillover, Kyoto, Japan, August 17-20, 1993 edited b y T. Inui, K. Fujimoto,T.Uchijima and M. Masai Heterogeneous Catalysis and Fine Chemicals Ill. Proceedings of the 3'd lnternational Symposium, Poitiers, April 5 - 8, 1993 edited by M. Guisnet, J. Barbier, J. Barrault, C. Bouchoule,D. Duprez, G. Perot and C. Montassier Catalysis: An Integrated Approach to Homogeneous, Heterogeneousand Industrial Catalysis edited b y J.A. Moulijn, P.W.N.M. van Leeuwen and R.A. van Santen Fundamentals of Adsorption. Proceedings of the Fourth lnternational Conference o n Fundamentals of Adsorption, Kyoto, Japan, May 17-22, 1992 edited by M. Suzuki Natural Gas Conversion 11. Proceedings of the Third Natural Gas Conversion Symposium, Sydney, July 4-9, 1993 edited by H.E. Curry-Hyde and R.F.Howe New Developments in Selective Oxidation 11. Proceedings of the Second World Congress and Fourth European Workshop Meeting, Benalmadena, Spain, September 20-24, 1993 edited by V. Cortes Corberan and S.Vic Bellon Zeolites and Microporous Crystals. Proceedings of the lnternational Symposium on Zeolites and Microporous Crystals, Nagoya, Japan, August 22-25, 1993 edited byT. Hattori and T.Yashirna Zeolites and Related Microporous Materials: State of the Art 1994. Proceedings of the 1Oth lnternational Zeolite Conference, Garmisch-Partenkirchen, Germany, July 17-22, 1994 edited by J.Weitkamp, H.G. Karge,H. Pfeifer and W. Holderich Advanced Zeolite Science and Applications edited b y J.C. Jansen, M. Stocker, H.G. Karge and J.Weitkarnp Oscillating HeterogeneousCatalytic Systems by M.M. Slinko and N.I. Jaeger Characterization of Porous Solids Ill.Proceedings of the IUPAC Symposium (COPS Ill), Marseille, France, May 9-1 2, 1993 edited by J.Rouquerol, F. Rodriguez-Reinoso, K.S.W. Sing and K.K.Unger

294 Volume 8 8

Volume 8 9

Volume 9 0

Volume 91

Volume 9 2

Volume 93 Volume 9 4

Volume 9 5 Volume 9 6

Volume 9 7

Volume 9 8

Volume 9 9 Volume 1 0 0

Volume 1 0 1

Volume 1 0 2 Volume 103 Volume 1 0 4 Volume 1 0 5

Catalyst Deactivation 1994. Proceedings of the 6'h lnternational Symposium, Ostend, Belgium, October 3-5, 1 9 9 4 edited by B. Delmon and G.F. Froment Catalyst Design for Tailor-made Polyolefins. Proceedings of the lnternational Symposium on Catalyst Design for Tailor-made Polyolefins, Kanazawa, Japan, March 10-1 2, 1 9 9 4 edited by K.Soga and M.Terano Acid-Base Catalysis II. Proceedings of the lnternational Symposium on Acid-Base Catalysis II, Sapporo, Japan, December 2-4, 1 9 9 3 edited by H. Hattori,M. Misono and Y.Ono Preparation of Catalysts VI. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Sixth lnternational Symposium, Louvain-La-Neuve, September 5-8, 1 9 9 4 edited by G. Poncelet, J.Martens,B. Delmon, P.A. Jacobs and P. Grange Science and Technology in Catalysis 1994. Proceedings of the Second Tokyo Conference on Advanced Catalytic Science and Technology, Tokyo, August 2 1-26, 1 9 9 4 edited by Y. Izumi, H.Arai and M. lwamoto Characterization and Chemical Modificationof the Silica Surface by E.F.Vansant, P.Van Der Voort and K.C.Vrancken Catalysis by Microporous Materials. Proceedings of ZEOCAT'95, Szombathely, Hungary, July 9-1 3, 1 9 9 5 edited by H.K. Beyer, H.G.Karge, I.Kiricsi and J.B.Nagy Catalysis by Metals and Alloys by V. Ponec and G.C.Bond Catalysis and Automotive Pollution Control Ill.Proceedings of the Third lnternational Symposium (CAPoC3), Brussels, Belgium, April 20-22, 1 9 9 4 edited by A. Frennet and J.-M. Bastin Zeolites:A Refined Tool for Designing Catalytic Sites. Proceedings of the lnternational Symposium, QuBbec, Canada, October 15-20, 1995 edited by L. Bonneviot and S. Kaliaguine Zeolite Science 1994: Recent Progress and Discussions. Supplementary Materials t o the 1Oth lnternational Zeolite Conference, Garmisch-Partenkirchen, Germany, July 17-22, 1 9 9 4 edited by H.G. Karge and J.Weitkamp Adsorption on New and Modified Inorganic Sorbents edited by A.Dabrowski and V.A.Tertykh Catalysts in Petroleum Refining and PetrochemicalIndustries 1995. Proceedings of the 2"d lnternational Conference on Catalysts in Petroleum Refining and Petrochemical Industries, Kuwait, April 22-26, 1 9 9 5 edited by M.Absi-Halabi, J. Beshara, H. Qabazard and A. Stanislaus Ilthlnternational Congress on Catalysis 40mAnniversary. ICC, Baltimore, MD, USA, June 30-July 5, 1 9 9 6 Proceedings of the 1 Ith edited by J.W. Hightower,W.N. Delgass, E. lglesia and A.T. Bell Recent Advances and New Horizons in Zeolite Science and Technology edited by H. Chon,S.I.Woo and S. -E. Park Semiconductor Nanoclusters - Physical, Chemical, and Catalytic Aspects edited by P.V. Kamat and D. Meisel Equilibria and Dynamics of Gas Adsorption on HeterogeneousSolid Surfaces edited by W. Rudzi~iski,W.A.Steele and G. Zgrablich Progress in Zeolite and Microporous Materials Proceedings of the Ilth lnternational Zeolite Conference, Seoul, Korea, August 12-1 7, 1 9 9 6 edited by H. Chon$-K. Ihm and Y.S.Uh

-

Volume 106

Volume 107

Volume 108

Volume 109

Volume 1 1 0

Volume 11 1

Volume 1 12

Volume 1 1 3

Volume 1 1 4

Volume 1 15

Volume 1 1 6

Volume 1 17

Volume 1 18

Volume 1 19

Volume 120A

Hydrotreatment and Hydrocracking of Oil Fractions Proceedings of the 1" lnternational Symposium I 61h European Workshop, Oostende, Belgium, February 17-19, 1997 edited by G.F. Froment,B. Delmon and P. Grange Natural Gas Conversion IV Proceedings of the 41h lnternational Natural Gas Conversion Symposium, Kruger Park, South Africa, November 19-23, 1995 edited by M. de Pontes, R.L. Espinoza, C.P. Nicolaides, J.H. Scholtz and M.S. Scurrell Heterogeneous Catalysis and Fine Chemicals IV Proceedings of the 41h lnternational Symposium on Heterogeneous Catalysis and Fine Chemicals, Basel, Switzerland, September 8-1 2, 1996 edited by H.U. Blaser, A. Baiker and R. Prins Dynamics of Surfaces and Reaction Kinetics in HeterogeneousCatalysis. Proceedings of the lnternational Symposium, Antwerp, Belgium, September 15-17, 1997 edited by G.F. Froment and K.C.Waugh Third World Congress on Oxidation Catalysis. Proceedings of the Third World Congress on Oxidation Catalysis, San Diego. CA, U.S.A., 21 -26 September 1997 edited by R.K. Grasselli,S.T.Oyama, A.M. Gaffney and J.E. Lyons Catalyst Deactivation 1997. Proceedings of the 7Ih lnternational Symposium, Cancun, Mexico, October 5-8, 1997 edited by C.H. Bartholomew and G.A. Fuentes Spillover and Migration of Surface Species on Catalysts. Proceedings of the 41h lnternational Conference on Spillover, Dalian, China, September 15-18, 1997 edited by Can Li and Qin Xin Recent Advances in Basic and Applied Aspects of Industrial Catalysis. Proceedings of the 131h National Symposium and Silver Jubilee Symposium of Catalysis of India, Dehradun, India, April 2-4, 1997 edited by T.S.R. Prasada Rao and G.Murali Dhar Advances in Chemical Conversionsfor Mitigating Carbon Dioxide. Proceedings of the 4Ih lnternational Conference on Carbon Dioxide Utilization, Kyoto, Japan, September 7-1 1, 1997 edited by T. Inui, M.Anpo,K. Izui,S.Yanagida and T.Yamaguchi Methods for Monitoring and Diagnosing the Efficiency of Catalytic Converters. A patent-oriented survey by M. Sideris Catalysis and Automotive Pollution Control IV. Proceedings of the 4'h lnternational Symposium (CAPoC4), Brussels, Belgium, April 9-1 1, 1997 edited by N. Kruse, A. Frennet and J.-M. Bastin Mesoporous Molecular Sieves 1998 Proceedings of the IS' lnternational Symposium, Baltimore, MD, U.S.A., July 10-12, 1998 edited by L.Bonneviot, F. Beland, C.Danumah, S. Giasson and S. Kaliaguine Preparationof Catalysts VII Proceedings of the 71h lnternational Symposium on Scientific Bases for the Preparation of Heterogeneous Catalysts, Louvain-la-Neuve, Belgium, September 1-4, 1998 edited by B. Delmon, P.A. Jacobs, R. Maggi, J.A.Martens, P. Grange and G. Poncelet Natural Gas Conversion V Proceedings of the 51h lnternational Gas Conversion Symposium, Giardini-Naxos, Taormina, Italy, September 20-25, 1998 edited by A. Parmaliana, D. Sanfilippo, F. Frusteri, A.Vaccari and F.Arena Adsorption and its Applications in Industry and Environmental Protection. Vol I: Applications in Industry edited by A. Dqbrowski

296 Volume 1 2 0 B

Volume 1 2 1

Volume 1 2 2

Volume 1 2 3

Volume 1 2 4 Volume 125

Volume 1 2 6

Volume 1 2 7

Volume 1 2 8

Volume 1 2 9

Volume 1 3 0

Volume 1 3 1

Volume 1 3 2

Volume 1 3 3

Volume 1 3 4

Adsorption and its Applications in Industry and Environmental Protection. Vol I I : Applications in Environmental Protection edited by A. Dqbrowski Science and Technology in Catalysis 1998 Proceedings of the Third Tokyo Conference in Advanced Catalytic Science and Technology, Tokyo, July 19-24, 1 9 9 8 edited by H. Hattori and K. Otsuka Reaction Kinetics and the Development of Catalytic Processes Proceedings of the lnternational Symposium, Brugge, Belgium, April 19-21, 1999 edited by G.F. Froment and K.C.Waugh Catalysis:An Integrated Approach Second, Revised and Enlarged Edition edited by R.A. van Santen, P.W.N.M. van Leeuwen, J.A. Moulijn and B.A.Averill Experiments in Catalytic Reaction Engineering by j.M. Berty Porous Materials in Environmentally Friendly Processes Proceedings of the 1"'lnternational FEZA Conference, Eger, Hungary, September 1-4, 1 9 9 9 edited by I. Kiricsi, G. Pal-Borbely, J.B.Nagy and H.G. Karge Catalyst Deactivation 1999 Proceedings of the 81h lnternational Symposium, Brugge, Belgium, October 10-1 3, 1 9 9 9 edited by B. Delmon and G.F. Froment Hydrotreatment and Hydrocracking of Oil Fractions Proceedings of the 2ndlnternational S y m p 0 s i u m I 7 ~ ~ European Workshop, Antwerpen, Belgium, November 14-1 7, 1 9 9 9 edited by B. Delmon, G.F. Froment and P. Grange Characterisation of Porous Solids V Proceedings of the 51h lnternational Symposium on the Characterisation of Porous Solids (COPS-V), Heidelberg, Germany, May 30- June 2, 1 9 9 9 edited by K.K.Unger,G.Kreysa and J.P. Baselt Nanoporous Materials II Proceedings of the 2"d Conference on Access in Nanoporous Materials, Banff, Alberta, Canada, May 25-30, 2 0 0 0 edited byA. Sayari,M. jaroniec and T.J. Pinnavaia I2 Ih lnternational Congress on Catalysis Proceedings of the 1 2 th ICC, Granada, Spain, July 9-1 4, 2 0 0 0 edited byA. Corrna, F.V. Melo,S. Mendioroz and J.L.G. Fierro Catalytic Polymerization of Cycloolefins Ionic, Ziegler-Natta and Ring-Opening Metathesis Polymerization By V. Dragutan and R. Streck Proceedings of the lnternational Conference on Colloid and Surface Science, Tokyo, Japan, November 5-8,2000 2 5 Ih Anniversary of the Division of Colloid and Surface Chemistry, The Chemical Society of Japan edited by Y. Iwasawa, N.Oyama and H.Kunieda Reaction Kinetics and the Development and Operation of Catalytic Processes Proceedings of the 3rd lnternational Symposium, Oostende, Belgium, April 2225, 2001 edited by G.F. Froment and K.C.Waugh Fluid Catalytic Cracking V Materials and Technological Innovations edited by M.L. Occelli and P. O'Connor

Volume 135

Volume 136

Volume 137

Volume 138 Volume 139 Volume 140

Volume 141

Volume 142

Volume 143

Volume 144

Volume 145

Volume 146

Volume 147

Volume 148

Zeolites and Mesoporous Materials at the Dawn of the 21'' Century. Proceedings of the 1 3 ' ~lnternational Zeolite Conference, Montpellier, France, 8-13 July 2001 edited by A. Galameau, F. di Renso, F. Fajula ans J. Vedrine Natural Gas Conversion VI Proceedings of the 6'h Natural Gas Conversion Symposium, June 17-22, 2001, Alaska, USA. edited by J.J. Spivey, E. lglesia and T.H. Fleisch Introduction t o Zeolite Science and Practice. 2ndcompletely revised and expanded edition edited by H. van Bekkum, E.M. Flanigen, P.A. Jacobs and J.C. Jansen Spillover and Mobility of Species o n Solid Surfaces edited by A. Guerrero-Ruiz and I. Rodriquez-Ramos Catalyst Deactivation 2001 Proceedings of the gthlnternational Symposium, Lexington, KY, USA, October 2001 edited by J.J. Spivey, G.W. Roberts and B.H. Davis Oxide-based Systems at the Crossroads of Chemistry. Second lnternational Workshop, October 8-1 1, 2000, Como, Italy. Edited by A. Gamba, C. Colella and S. Coluccia Nanoporous Materials Ill Proceedings of the 3rdlnternational Symposium on Nanoporous Materials, Ottawa, Ontario, Canada, June 12-15,2002 edited by A. Sayari and M. Jaroniec Impact of Zeolites and Other Porous Materials o n the New Technologies at the Beginning of the New Millennium Proceedings of the 2" lnternational FEZA (Federation of the European Zeolite Associations) Conference, Taormina, Italy, September 1-5, 2002 edited by R. Aiello, G. Giordano and F.Testa Scientific Bases for the Preparation o f Heterogeneous Catalysts Proceedings of the 8thlnternational Symposium, Louvain-la-Neuve, Leuven, Belgium, September 9-12, 2002 edited by E. Gaigneaux, D.E. De Vos, P. Grange, P.A. Jacobs, J.A. Martens, P. Ruiz and G. Poncelet Characterization of Porous Solids VI Proceedings of the 6'h lnternational Symposium on the Characterization of Porous Solids (COPS-VI), Alicante, Spain, May 8-1 1, 2002 edited by F. Rodriguez-Reinoso, B. McEnaney, J. Rouquerol and K. Unger Science and Technology i n Catalysis 2002 Proceedings of the Fourth Tokyo Conference on Advanced Catalytic Science and Technology, Tokyo, July 14-19, 2002 edited by M. Anpo, M. Onaka and H. Yamashita Nanotechnology i n Mesostructured Materials Proceedings of the 3rdlnternational Mesostructured Materials Symposium, Jeju, Korea, July 8-1 1, 2002 edited by Sang-Eon Park, Ryong Ryoo, Wha-Seung Ahn, Chul Wee Lee and Jong-San Chang Natural Gas Conversion VII Proceedings of the 7'h Natural Gas Conversion Symposium, Dalian, China, June 5-1 1, 2004 edited by X. Bao and Y. Xu Mesoporous Crystals and Related Nano-Structured Materials Proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials, Stockholm, Sweden, 1-5 June, 2004 edited by 0. Terasaki

This Page Intentionally Left Blank

Nanostructured materials and nanotechnology

Zeolites and Microporous Crystals: Proceedings of the International Symposium on Zeolites and Microporous Crystals, Nagoya, August 22-25, 1993 (Studies in Surface Science and Catalysis)

Nanostructured materials and nanotechnology

Nanostructured Materials and Nanotechnology

Nanostructured Materials

Ordered Mesoporous Materials

Nanostructured Materials

Nanostructured Materials

Nano-Architectured and Nanostructured Materials

Advanced Catalysts and Nanostructured Materials

Nanostructured Materials in Electrochemistry

Nanostructured Materials and Nanotechnology III (Ceramic Engineering and Science Proceedings)

Chemistry of Nanostructured Materials

Sensors Based on Nanostructured Materials

The chemistry of nanostructured materials

Bulk Nanostructured Materials

Bulk Nanostructured Materials

Bulk Nanostructured Materials

Physics and Chemistry of Nanostructured Materials

Phonon Confinement in Nanostructured Materials

Catalysis by Ceria and Related Materials (Catalytic Science Series)

Catalysis by Ceria and Related Materials (Catalytic Science Series)

NANOSTRUCTURED MATERIALS Processing Properties and Potential Applications

Bioinspired Intelligent Nanostructured Interfacial Materials

Nanostructured Materials for Engineering Applications

Nanostructured Materials for Engineering Applications

Liquid Crystals: Materials Design and Self-assembly

Domain Structure in Ferroelectrics and Related Materials

Domain Structure in Ferroelectrics and Related Materials

Nanophotonic Materials: Photonic Crystals, Plasmonics, and Metamaterials

Conservation of Leather and Related Materials

Zeolites and Microporous Crystals: Proceedings of the International Symposium on Zeolites and Microporous Crystals, Nagoya, August 22-25, 1993 (Studies in Surface Science and Catalysis)

Mesoporous Crystals and Related Nano-Structured Materials, Volume 148: Proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials, ... (Studies in Surface Science and Catalysis)

Mesoporous Crystals and Related Nano-Structured Materials, Volume 148: Proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials, ... (Studies in Surface Science and Catalysis)

Nanostructured materials and nanotechnology

Nanostructured Materials and Nanotechnology

Nanostructured Materials

Ordered Mesoporous Materials

Nanostructured Materials

Nanostructured Materials

Nano-Architectured and Nanostructured Materials

Advanced Catalysts and Nanostructured Materials

Nanostructured Materials in Electrochemistry

Nanostructured Materials and Nanotechnology III (Ceramic Engineering and Science Proceedings)

Chemistry of Nanostructured Materials

Sensors Based on Nanostructured Materials

The chemistry of nanostructured materials

Bulk Nanostructured Materials

Bulk Nanostructured Materials

Bulk Nanostructured Materials

Physics and Chemistry of Nanostructured Materials

Phonon Confinement in Nanostructured Materials

Catalysis by Ceria and Related Materials (Catalytic Science Series)

Catalysis by Ceria and Related Materials (Catalytic Science Series)

NANOSTRUCTURED MATERIALS Processing Properties and Potential Applications

Bioinspired Intelligent Nanostructured Interfacial Materials

Nanostructured Materials for Engineering Applications

Nanostructured Materials for Engineering Applications

Liquid Crystals: Materials Design and Self-assembly

Domain Structure in Ferroelectrics and Related Materials

Domain Structure in Ferroelectrics and Related Materials

Nanophotonic Materials: Photonic Crystals, Plasmonics, and Metamaterials

Conservation of Leather and Related Materials

Zeolites and Microporous Crystals: Proceedings of the International Symposium on Zeolites and Microporous Crystals, Nagoya, August 22-25, 1993 (Studies in Surface Science and Catalysis)

Recommend Documents