Crystalline Cellulose and Derivatives: Characterization and Structures (Springer Series in Wood Science) (Springer Series in Wood Science)

Citation preview

Springer Series in Wood Science

Series Editors T. E. Timell State University of New York College of Environmental Science and Forestry Syracuse, NY 13210, USA Professor Dr. Rupert Wimmer Bio-based Fibre Materials Department of Material Sciences and Process Engineering University of Natural Resources and Applied Life Sciences BOKU-Vienna Peter-Jordan-Strasse 82 1190 Vienna, Austria

Springer Series in Wood Science Editors: T.E. Timell, R. Wimmer L.W. Roberts/P.B. Gahan/R. Aloni Vascular Differentiation and Plant Growth Regulators (1988) C. Skaar Wood-Water Relations (1988) J.M. Harris Spital Grain and Wave Phenomena in Wood Formation (1989) B.J. Zobel/J.P. van Buijtenen Wood Variation (1989) P. Hakkila Utilization of Residual Forest Biomass (1989) J.W. Rowe (Ed.) Natural Products of Wood Plants (1989) K.-E.L. Eriksson/R.A. Blanchette/P. Ander Microbial and Enzymatic Degradation of Wood and Wood Components (1990) R.A. Blanchette/A.R. Biggs (Eds.) Defense Mechanisms of Woody Plants Againts Fungi (1992) S.Y. Lin/C.W. Dence (Eds.) Methods in Lignin Chemistry (1992) G. Torgovnikov Dielectric Properties of Wood and Wood-Based Materials (1993) F.H. Schweingruber Trees and Wood in Dendrochronology (1993) P.R. Larson The Vascular Camblum: Development and Structure (1994) M.-S. Ilvessalo-Pfaffli Fiber Atlac Identification of Papermaking Fibers (1995) B.J. Zobel/J.B. Jett Genetics of Wood Production (1995) C. Matteck/H. Kabier Wood - The Internal Optimization of Wood (1997) T. Higuchi Biochemistry and Molecular Biology of Trees (1997) B.J. Zobel/J.R. Sprague Juvenile Wood in Forest Trees (1998) E. Sjostrom/R. Alén (Eds.) Analytical Methods in Wood Chemistry, Pulping, and Papermaking (1999) R.B. Kery/T.A.G. Langrish/J.C.F. Walker Kiln-Drying of Lumber (2000) S. Carlquist Comparative Wood Anatomy, 2nd ed. (2001) M.T. Tyree/M.H. Zimmermann Xylem Structure and the Ascent of Sap. 2nd ed. (2002) T. Koshijima/T. Watanabe Association Between Lignin and Carbohydrates in Wood and Other Plant Tissues (2003) V. Bucur Nondestructive Characterisation and Imaging of Wood (2003) V. Bucur Acoustics of Wood (2006) F.H. Schweingruber Wood Structure and Environment (2007) P. Zugenmaier Crystalline Cellulose and Derivatives

Peter Zugenmaier

Crystalline Cellulose and Derivatives Characterization and Structures

Professor Peter Zugenmaier Institute of Physical Chemistry TU Clausthal Arnold-Sommerfeld-Str. 4 D-38678 Clausthal-Zellerfeld Germany

Cover: Transverse section of Pinus lambertiana wood. Courtesy of Dr. Carl de Zeeuw, SUNY College of Environmental Science and Forestry, Syracuse, New York.

ISBN 978-3-540-73933-3

e-ISBN 978-3-540-73934-0

Springer Series in Wood Science ISSN 1431-8563 Library of Congress Control Number: 2007933158 © 2008 Springer-Verlag Berlin Heidelberg This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: WMXDesign GmbH, Heidelberg, Germany Printed on acid-free paper springer.com

5 4 3 2 1 0

Preface

“On making many books there is no end” but we trust that no excuse is needed for the present work. The subject of cellulose chemistry is among the simplest of studies, but the important advances of recent years have clarified it to such an extent that we feel the time is ripe for publishing a relatively simple book which may act as a guide to younger chemists who are entering those branches of our great industries which are concerned with cellulose. J.T. Marsh and F.C. Wood (1939) An Introduction to the Chemistry of Cellulose

Recent progress in crystalline polysaccharide structure determination and the publication of numerous models of cellulose and cellulose derivatives by improved methods made a critical and up-to-date survey of structures and characterization of cellulose possible and necessary. Structural evaluations by refined experimental and computer-aided modeling represent the prerequisite for many research and testing areas of cellulosic materials, e.g., for establishing structure–property relationships (tensile strength, sorption, solubility, etc.), chemical reactivity and derivatization as well as the composition of cell wall materials and the orientation of microfibrils in cellulose fibers. Modern materials science needs tailored materials, linked to the structure for improvements and for new developments. The active species for enantiomeric separation in gel permeation chromatography columns are, e.g., microcrystalline cellulose derivative beads of a particular structure, which produce optimal results. Composites with soft materials and cellulose or cellulose derivatives exhibit enhanced properties strongly dependent on the stiff cellulosic backbone and can be improved by optimizing the interaction parameters. This book is concerned with the crystalline structure and characterization of cellulose, cellulose complexes and cellulose derivatives. The principles of structure determination of polymers rely on fiber diffraction combined with computer-aided modeling and also on spectroscopy. The various now-available results are evaluated and compared with oligomeric structures from which invariants are derived and used as standards. Suitable models were chosen and the geometric data are compared and best models according to standards shown in graphs and the coordinates are collected in an v

vi

Preface

Appendix. Representative X-ray, solid-state 13C NMR and IR patterns are provided for characterizing cellulosic structures. Research on cellulose started as soon as appropriate methods were available at the beginning of the twentieth century. The history of polymer science is closely linked to the development of the structure of native cellulose and in the beginning this was controversially discussed as aggregates of small molecular units or as macromolecules on one side. The macromolecular concept grew out of these controversies. On the other side the crystal packing arrangement was proposed to occur by parallel chains with all the nonreducing cellulose ends on one tip of the native microcrystals or by antiparallel arrangements with adjacent nonreducing ends on opposite tips. It was not until improved crystalline fibers and finer detection methods were available that conclusive computer-aided conformation and packing analysis led to a decisive proposal for the parallel packing of the native structure of cellulose. This development is critically overviewed in a chapter devoted to the history of cellulose research. However, antiparallel arrangements are observed for soft treated mercerized native cellulose and most derivatives. A conversion mechanism is needed and is presented to describe the conversion from native to mercerized cellulose by preserving the orientation of chains in fibers during this conversion. This book is a valuable, concise and up-to-date guide for the materials and life science community involved with cellulose and related materials. A rigorous description of the refinement procedures for structure determination is not presented here but may be found in the original publications. This book represents a collection of critically selected structures and is directed towards students, scientists and researchers in materials quality control who are interested in or depend on knowledge of crystalline cellulosic structures and who need reference data for characterizing materials. This book was initiated by Tore E. Timell, Syracuse, NY, USA, late editor of Springer Series in Wood Science. Without his enthusiastic encouragement and support this book would never have been finished. Cellulose as an abundant renewable material has stimulated basic and applied research throughout the years, as addressed in the historical review, and has inspired significant progress in polymer science. In recent years cellulose has gained renewed significance as a raw material and still possesses high potential for future applications. Academia and industry may equally profit from this comprehensive survey. Peter Zugenmaier

Acknowledgement for Copyrights

I wish to acknowledge permission from the following publishers to reproduce the copyrighted material indicated. Acknowledgments to the original sources are given in the figure captions: American Chemical Society Washington DC: ACS Symposium Series: Cellulose derivatives (1998): Figure 7.2 J Amer Chem Soc: Figures 5.9, 5.14 Macromolecules: Scheme 5.1, Figures 5.4, 5.6, 5.8, 5.13, 5.16, 5.25, 5.26, 5.40, 5.47, 5.51, 6.8, 6.19 a,b, A2 Biomacromolecules: Figures 5.20, 5.33-36 Elsevier Elsevier Publishing Company, Inc. New York-Amsterdam-London-Brussels: Physics and chemistry of cellulose fibres (1949): Figures 2.14, 5.19 Academic Press, London-New York: Polymer and fibre diffraction (1972): Figure 3.6 Advances in Carbohydrate Chemistry and Biochemistry: Figure 3.4 J Struct Biol: Figures 2.20, 5.7, 5.17 J Mol Struct: Figure 3.11, A1 Prog Polym Sci: Figures 3.14, 4.4, 4.4 Polymer: Table 5.11, Figure 5.24 Polymer Comm: Figure 6.5 Solid State Nucl Magn Reson: Figure 5.21 Carbohydr Res: Table 7.2 John Wiley & Sons Ltd, New York Ellis Horwood Ltd, Chichester, Cellulose: Structural and functional aspects (1989): Figures 6.6, 6.12, 6.13 Wiley Interscience, New York: The use of X-ray diffraction study of protein and nucleic structures (1966): Figures 3.3, 3.4 John Wiley, New York: Cellulose and wood – chemistry and technology (1989): Figures A7, A8 Ber Deutsch Chem Ges: Figure: 2.8 vii

viii

Acknowledgement for Copyrights

Bioploymers: Figures 5.38, 5.43 Helv Chim Acta: Figures 2.10, 2.17 J Appl Polym Sci: Figures 7.7, 7.8 J Polym Sci: Figures 2.19, A3-A6 J Polym Sci B: Figures 5.9, 5.37 J Polym Sci A: Figures 6.1, 6.2, 6.21 J Polym Sci Phys Ed: Figure 6.19c Makromol Chem: Figure 7.1 Verlag Chemie: Figure 2.7 Springer Verlag Berlin-Heidelberg: Cellulose: Table 5.10, Figures 5.18, 5.20, 5.28, 6.10, 6.24 Polym Bull: Table 3.2, Figure 3.13 Colloid & Polym Sci (Kolloid Z, Kolloid Z u Z Polymere):Table 2.2, Figures 6.3, 6.4, 6.21 J Materials Sci: Figure 7.4 Plenum Press, New York: Cellulose and other natural polymer systems (1982): Figure 7.5, Structural electron crystallography.(1995): Figure 6.20 Das Papier, Darmstadt: Figure 7.3 Francis and Taylor, London: J Carbohydr Chem. Figure 3.12 Hanser Publishers, Munich: Cellulosic polymers (1994): Table 5.2, Figures 5.1, 5.3 International Union of Pure and Applied Chemistry: Pure Appl Chem: Figure 3.4 Oldenbourg Wissenschaftsverlag, München:: Z Phys Chem: Figures 2.5b, 2.9, 2.18 The Society of Biotechnology, Osaka: J Biosci Bioeng: Table 5.20, Figure 5.44 The Society of Polymer Science, Tokyo: Polymer J: Figure 5.5 The drawings of the models were produced with software from Keller E (1992) Schakal 92: A computer program for graphic representation of molecules and crystallographic models. Freiburg

Contents

1

2

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

General Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5

History of Cellulose Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

2.1 2.2 2.3 2.4

3

4

The Concept of Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concepts in Structural Research of Cellulose. . . . . . . . . . . . . . . . . . . Arrangements of the Cellulose Molecules in the Solid State . . . . . . . Chemical Constitution of Cellulose as a Macromolecule. . . . . . . . . . 2.4.1 Linkage of Cellulose – the Chain Structure of Cellulose (Freudenberg, Haworth) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2 Macromolecule Formation – Size of the Chains (Staudinger) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Historical Development of X-ray Models for Native Cellulose . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7 8 16 21

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

53

3.1 3.2 3.3

Diffraction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Model Building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Optical Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 NMR Spectroscopy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Convention for the Description of Cellulosic (Chiral) Structures . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

53 60 64 65 68 72 74

Model Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

77

4.1 4.2 4.3

77 78 83 83 87 94 98

Conformation and Packing Analysis . . . . . . . . . . . . . . . . . . . . . . . . . Monomers and Dimers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trimers and Tetramers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Conformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Packing Arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Acetyl Derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

22 24 27 46

x

Contents

5

Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 5.1 5.2

Cellulose Polymorphy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 X-ray Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Spectroscopic Characterization . . . . . . . . . . . . . . . . . . . . . . . 5.3 Molecular and Crystal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Cellulose Iβ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Cellulose Iα . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Cellulose II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 Cellulose III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.5 Cellulose IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Cellulose Solvent Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Cellulose II–Hydrazine Complex . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Cellulose II Hydrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 Cellulose I–Ammonia I Complex . . . . . . . . . . . . . . . . . . . . . . 5.4.4 Cellulose I–Ethylenediamine Complex . . . . . . . . . . . . . . . . . 5.5 Sodium Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Sodium Cellulose I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Sodium Cellulose IV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Cellulose Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Cellulose Triacetate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Experimental Data for Cellulose Tripropionate and Cellulose Acetate Dipropionate and Further Cellulose Esters . . . . . . . . 6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Conformation and Packing Arrangement of CDAP . . . . . . . . . . . . . . 6.4 Conformation and Packing Arrangement of Cellulose Tribenzoate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Trimethyl Cellulose and 6-O-Acetyl-2,3-di-O-methyl Cellulose . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1

7

101 103 103 105 112 113 122 129 138 143 151 151 152 156 160 164 165 169 171

175 175 179 185 197 199 200 204

Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 7.1 Crystalline Domain Sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Microfibrils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Microfibrils and Fibrils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Parallel and Antiparallel Packing Arrangements of Microfibils. . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

207 207 212 216 220

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

Chapter 1

Introduction

Cellulose represents a naturally occurring linear macromolecular chain of 1–4-linked b-d-glucopyranose and exhibits great chemical variability and potential in applications. The cell walls of all plants contain fibers of cellulose. Cellulose has long been harvested as commercial fibers from the seed hairs of cotton (over 94% cellulose), as bast fibers (60–80% cellulose) from flax, hemp, sisal, jute and ramie or as wood (40–55% cellulose), which is a common building material or is used as a source for purified cellulose. The chemical compositions of some known species are collected in Table 1.1, which, when purified, serve as cellulose sources. Wood represents a composite material with cellulose as a major part combined in excellent form with lignin and hemicelluloses, creating a unique high-strength and durable material, and recently came again into focus as a renewable energy resource. Land plants such as forest trees and cotton synthesize cellulose from glucose, produced in the plant cells by photosynthesis. Unicellular plankton or algae in the ocean also generate cellulose by fixation of carbon dioxide as do land plants. Therefore, vast resources of cellulose are available and serve as food for animal life in the ocean or can be harvested. However, cellulose may be also assembled by several animals, fungi and bacteria, which are devoid of photosynthetic ability and require glucose or other organic substrates and are dependent on other organisms. In the nineteenth century, methods were developed to separate wood cellulose from lignin chemically and to regenerate the cellulose for use as fibers (rayon) and plastics (cellophane). Later, ester and ether derivatives of cellulose were developed and the esters were predominantly used as fibers and plastics and the ethers as binders and additives for special mortars or special chemicals for building and construction as well as viscosity stabilizers in paint, oil exploration, food and pharmaceutical products, etc. Cellulose nitrate (nitrocellulose, made into celluloid) and cellulose acetate (fibers, films and plastics) are important derivatives for solid-state applications. The properties of both these chemical derivatives are based on the cellulose chain structure. Cellulose also represents the basic materials in papermaking. Its fibers have high strength and durability. They are readily wetted by water, exhibiting considerable swelling when saturated, and are hygroscopic, i.e., they absorb appreciable amounts of water when exposed to the atmosphere. Even in the wet state, natural cellulose fibers show almost no loss in strength. It is the combination of these P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

1

2

1 Introduction

Table 1.1 Chemical composition of some typical cellulose-containing materials. (Adapted from Hon 1996) Composition (%) Source

Cellulose

Hemicellulose

Lignin

Extract

Hardwood Softwood Bagasse Coir Corn cobs Corn stalks Cotton Flax (retted) Flax (unretted) Hemp Henequen Istle Jute Kenaf Ramie Sisal Sunn Wheat straw

43–47 40–44 40 32–43 45 35 95 71 63 70 78 73 71 36 76 73 80 30

25–35 25–29 30 10–20 35 25 2 21 12 22 4–8 4–8 14 21 17 14 10 50

16–24 25–31 20 43–49 15 35 1 2 3 6 13 17 13 18 1 11 6 15

2–8 1–5 10 4 5 5 0.4 6 13 2 4 2 2 2 6 2 3 5

qualities with strength and flexibility that makes cellulose of unique value for paper manufacturing. Dry cellulose has thermosetting behavior, i.e., it forms permanent, bonded structures that cannot be loosened by heat or common solvents without causing chemical decomposition. Its thermosetting behavior arises from strong dipolar attractions that exist between cellulose molecules, imparting properties similar to those of interlinked polymer networks. In regenerated form, cellulose is used for textile fibers and for producing derivatives. Cellulose fibers and films show excellent tensile properties. Thin sheets of cellulose acetate serve as optical compensators and shields for compounds evaporating from the polarizer in modern liquid-crystalline displays, and cellulose acetate sheets are the base for photographic films. Composite materials with cellulose are widely used as wood; also worthy of mention are cactus thorns used by the Indians of South America as nails. Synthetic composites have been developed by extruding polypropylene with microcrystalline cellulose obtained by hydrolytic degradation of native cellulose or with regenerated cellulose. These easily accessible composites point towards promising applications. The superiority of cellulose derivatives as the stationary phase in chromatographic procedures separating enantiomeric molecular species should be mentioned as well. Cellulose as an abundant natural material serving mankind for centuries became the subject of science as soon as appropriate tools for scientific investigations

1 Introduction

3

became available to satisfy human curiosity and to improve the existing properties of materials as well. Therefore, cellulose has to be considered as a major subject in the history of polymer science in the development of the concept of macromolecules and the determination of polymeric crystal structures. These developments represent excellent examples of how science proceeds and develops with the ideas and contributions of many researchers. An involvement with the history of science may rediscover ideas which have been forgotten and lost. At certain times ideas were impossible to follow up because the necessary tools were not available or science took another route. All native celluloses are organized in fibrils, which represent the association of cellulose molecules and contain ordered and less ordered regions. From a structural point of view cellulose represents a semiflexible molecule and can be described as an extended wormlike chain for short molecular length but as a Gaussian coil with loops and intermolecular contacts for long chains as represented for cellulosic solutions in Fig. 1.1. Stiff wormlike chains are also the prerequisite for the formation of lyotropic liquid crystals with amazing physical properties. The diversity of appearance of cellulose and cellulose derivatives in various polymorphs in the solid state on a molecular and a supermolecular level as well as in solution made it difficult to obtain a clear picture of these structures for a long period of time. Native cellulose in fibers of higher plants possesses a very high degree of polymerization in contrast to treated cellulose used as sources in applications (Table 1.2), and various kinds of morphological structures may occur depending on the chain length. In addition, structural investigations have suffered extensively from insufficient data owing to imperfect structures and methods. In recent years improvements in structural research on the molecular level have led to the proposal of valuable models for the conformation and packing arrangements of chain molecules.

Fig. 1.1 Cellulosic molecules of short and long molecular chain lengths in molecularly dispersed solution. The broken lines indicate the solvent shell. Extended molecules are present for short chains and Gaussian coils with loops and intermolecular contacts for long chains

4

1 Introduction Table 1.2 Average degree of polymerization (average number of monomeric units in one chain, evaluated from the molecular mass distribution) of some selected celluloses Cellulose Degree of polymerization Wood of various species 6,000–10,000 Pulp 500–2,000 Sulfate pulp 950–1,300 Chemical pulp bleached 700 Cotton 10,000–15,000 Cotton linters bleached 1,000–5,000 Valonia 25,000 Bacterial cellulose 4,000–6,000 Ramie 10,000 Textile flax 9,000 Rayon 300–500 Cellophane 300 Cellulose acetate 200–350 A comprehensive evaluation of molecular masses of native celluloses is provided by Schulz and Marx (1954) and Marx-Figini (1982).

In this book we will present a short overview of basic principles in the development of macromolecular science, in particular the development of the macromolecular concept and of the crystal structure of native cellulose. The description of the crystal structures of various polymorphs of cellulose requires some insight into the methods applied for a basic understanding, for a judgment of the goodness of the available data and for further possible improvements of structures as well as developments to proceed to further fields. A survey of cellulosic structures may also lead to extraction of general structural features of cellulosic materials. Polymeric chain structures are composed of many monomeric units containing numerous atoms but the experimental data sets are very limited. Invariants have to be introduced, such as configuration, bond lengths and angles, derived from oligomeric compounds, and regarded as a necessity to implement the determination procedure of polymeric crystal structures and to supplement missing experimental data of polymeric compounds, e.g., cellulose. The determination of crystalline structures predominantly rests on diffraction methods such as X-ray, synchrotron, electron and neutron scattering of highly crystalline, highly orientated samples with little disorder. It took a long time until it was realized that native cellulose consists of two crystalline polymorphs, now termed cellulose Iα and Iβ. The cellulose microfibrils from the cell walls of the algae Cladophora, Halicystis, Valonia, etc., contain predominantly well-oriented cellulose Iα. In contrast, the microfibrils of cotton, ramie (China-grass), further bast fibers and the tunicin animal cellulose from the mantle of tunicates serve as sources for oriented and highly crystalline cellulose Iβ.

General Literature

5

IR, Fourier transform IR, Raman and NMR spectroscopic techniques have always served as complementary tools, especially if the long-range threedimensional order is disturbed or totally missing. Some background knowledge will be provided for a fast characterization and judgment of the materials as well as the discussion of the structures. An overview of cellulosic structures and the representation of experimental data may fulfill many purposes. It serves as a data base for characterizing naturally occurring substances, i.e., for differentiating between various cellulose polymorphs and other polysaccharides. For such a discrimination only fingerprint patterns of diffraction experiments or spectroscopic traces are necessary. A more detailed evaluation of the experimental data is needed to gain insights into the structure at the molecular or the morphological level and to extract information for studying the formation of the structure by nature or to invoke possible changes in the structure to improve properties and find new applications. Such comprehensive structural knowledge is also required for the prediction of pathways of chemical reactions and for the specific interaction sites of small molecules in inclusion complexes or at the surfaces of the crystallites. Science will rapidly develop further and with the advent of improved and new tools, techniques and with combinations of them, including experiments with further data, such a survey can only serve as a snapshot of our knowledge in this field at the present time.

General Literature Atalla RH (ed) (1987) The structures of cellulose – characterization of the solid states. ACS symposium series no 340. American Chemical Society, Washington Atalla RH (1999) Celluloses. In: Pinto BM (ed) Comprehensive natural products chemistry, vol 3: carbohydrates and their derivatives including tannins, cellulose, and related lignins. Elsevier, Amsterdam, pp 529–598 Bikales NM, Segal L (eds) (1971) Cellulose and cellulose derivatives. Wiley-Interscience, New York French AD (2000) Structure and biosynthesis of cellulose. Part I: structure. In: Kung S-D, Yang S-F (eds) Discoveries in plant biology, vol 3. World Scientific, Singapore, pp 163–197 French AD, Gardner KH (eds) (1980) Fiber diffraction methods. ACS symposium series no 141. American Chemical Society, Washington Freudenberg K (1933) Tannin, Cellulose, Lignin. Springer, Berlin Fyfe CA (1983) Solid state NMR for chemists. CFC, Guelph Haworth WN (1929) The constitution of sugars. Edward &Arnold, London Haworth WN (1932) Die Konstitution der Kohlenhydrate. Steinkopff, Dresden Hermans PH (1949) Physics and chemistry of cellulose fibres. Elsevier, New York Hess K (1928) Die Chemie der Zellulose und ihrer Begleiter, XX. Akademische Verlagsgesellschaft, Leipzig Hon DN-S (1996) Functional polymers: a new dimensional creativity in lignocellulosic chemistry. In: Hon DN-S (ed) Chemical modification of lignocellulosic materials. Dekker, New York, pp 1–10 Klemm D, Philipp B, Heinze T, Heinze U, Wagenknecht W (1998) Comprehensive cellulose chemistry, vols 1 and 2. Wiley-VCH, Weinheim

6

1 Introduction

Klemm D, Schmauder H-P, Heinze T (2004) Cellulose. In: de Baets S, Vandamme E, Steinbüchel A (eds) Biopolymers, vol 6. Polysaccharides II: polysaccharides from eukaryotes. Wiley-VCH, Weinheim, pp 275–319 Krässig HA (1993) Cellulose structure, accessibility, and reactivity. Gordon and Breach, New York Krüger D (1933) Zelluloseazetate. Steinkopff, Dresden Marchessault RH, Sundararajan PR (1983) Cellulose. In: Aspinall GO (ed) The polysaccharides, vol 2. Academic, New York, pp 11–95 Mark H (1932) Physik und Chemie der Cellulose, XV. In: Herzog RO (ed) Technologie der Textilfasern, vol I, part 1. Springer, Berlin Marsh JT, Wood FC (1939) An introduction to the chemistry of cellulose. Van Nostrand, New York Marx-Figini M (1982) The control of molecular weight and molecular-weight distribution in the biogenesis of cellulose. In: Brown RM Jr (ed) Cellulose and other natural polymer systems. Plenum, New York, pp 243–271 Meyer KH (1950) Natural and synthetic high polymers, vol IV. Interscience, New York Meyer KH, Mark H (1930) Der Aufbau der hochpolymeren organischen Naturstoffe. Akademische Verlagsgesellschaft, Leipzig Meyer KH, Mark H (1950) Makromolekulare Chemie, 2nd edn. Geest & Portig, Leipzig Morawetz H (1985) Polymers – the origin and growth of a science. Wiley, New York Ott E (ed) (1943) Cellulose and cellulose derivatives, vol V. Interscience, New York Ott E, Spurlin HM, Grafflin MW (eds) (1955) Cellulose and cellulose derivatives, part III. High polymers, vol 5, 2nd edn. Interscience, New York Priesner C (1980) H. Staudinger, H. Mark und K. H. Meyer – Thesen zur Größe und Struktur der Makromoleküle. Verlag Chemie, Weinheim Pummerer R (ed) (1953) Chemische Textilfasern – Filme und Folien. Enke, Stuttgart Purves CB (1946) Chemical nature of cellulose and its derivatives. In: Ott E (ed) Cellulose and cellulose derivatives. High polymers, vol 5. Interscience, New York, pp 29–76, 88–112 Saechtling H (1935) Hochpolymere organische Naturstoffe. Vieweg, Braunschweig Schulz GV, Marx M (1954) Über Molekulargewichte und Molekulargewichtsverteilungen nativer Cellulosen. Makromol Chem 14:52–95 Sisson WA (1946) X-ray examination. In: Ott E (ed) Cellulose and cellulose derivatives. High polymers, vol 5. Interscience, New York, pp 203–292 Staudinger H (1932) Die hochpolymeren Verbindungen. Kautschuk und Cellulose, XV. Springer, Berlin Stuart HA (ed) (1955) Die Physik der Hochpolymeren, vol 3. Springer, Berlin Walton AG, Blackwell J (1973) Biopolymers. Academic, New York

Chapter 2

History of Cellulose Research

2.1

The Concept of Cellulose

Cellulose is defined as a macromolecule, a nonbranched chain of variable length of 1-4-linked b-d-anhydroglucopyranose units (Fig. 2.1). In contrast, cellulose pulp represents purified cellulosic materials, and still contains other carbohydrates. These definitions are not trivial, since Payen (1838), who coined the term “cellulose” for sufficiently purified plant tissues, used the term “cellulose” for what is nowadays called pulp. Other researchers continued to use the term “cellulose” in Payen’s original definition (Purves 1946). Payen found 43.6–45% carbon, 6.0–6.5% hydrogen and the remainder was oxygen (theoretical C 44.4%, H 6.2%) for the extraction of the fibrous wrap and wood of all young plant cells but for also seeds, cotton linters as well as a few mosses and lichens. The purified residue represented dextrorotatory, gummy materials. These observations convinced Payen that the purified materials contained one uniform chemical species, which was a carbohydrate, based on glucose residues similar to starch: “In fact, wood contains a substance isomeric with starch, which we call cellulose and a material filling the cells, the real ligneous substance.” His idea was that cellulose was a more highly aggregated isomer than starch and when opponents disputed the uniformity of cellulose, he replied that chemical treatment might modify its state of aggregation. In contrast, Frémy (1859), who investigated enzymatic conversions, insisted that the differences in the properties of cellulose and starch are due to isomeric states of these substances. Payen’s opinion that cellulose was constituted invariably of glucose residues was based on inadequate data and experimental methods for discriminating monosaccharides at that time. Acid hydrolyzates of many materials consisting of celluloses were later found to contain substantial amounts of galactose, arabinose, mannose or xylose as well as glucose. The endosperm of ivory nut, which Payen had thought consisted of especially pure cellulose, yielded almost exclusively mannose (Purves 1946). Later, the less resistant carbohydrates, other than glucose, were given the generic name of hemicelluloses. Cotton was then considered as a standard because it consisted almost entirely of glucose residues. The term “cellulose” following Payen and further Schulze (1891), who used drastic extraction reagents, was

P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

7

8

2 History of Cellulose Research

Fig. 2.1 Chemical constitution of cellulose as 1-4-linked b-d-anhydroglucopyranose and numbering of carbon atoms in the representation of Haworth (1929, 1932). The equatorial position (b-position) of C1–O1 is given by O1 above the ring with O5 at the back

reserved for the portion of the cell wall resembling cotton cellulose in its physical and chemical properties. Controversial ideas about cellulose occurred throughout all stages in the history of science. Progress and clarification mostly came through the introduction of novel ideas influenced by general progress in science or by the invention of new experimental methods invoking better data sets for interpretation. Former ideas in science are often judged from today’s knowledge and it is not recognized that some of the background information was not available at that time. Science does not always proceed on a straight route and turn-off tracks may lead to dead ends. Published novel and exceptional ideas are sometimes not accepted because they run against conventional and established concepts or against the ideas of prominent scientists. It may also occur that an experiment does not qualify for a single interpretation and a second explanation cannot be excluded. Progress in science is the result of controversies and the contributions of numerous researchers.

2.2

Concepts in Structural Research of Cellulose

Cellulose research offers a wide field of aspects for how science proceeds with all human vanity and influence including priority claims and disputes over ideas. At times the excellent contribution of Sponslor and Dore (1926) has been totally neglected in the development of a structural model of cellulose, the first proposed chain model based on glucopyranose, later accepted by Meyer, who was also involved in priority claims (Kiessig 1939). They overturned the widely accepted idea that the size or length of a molecule is limited by the unit cell of crystalline domains and introduced primary valencies along the cellulose chain running parallel to the fiber axis in the crystalline part of ramie, leading to nondetermined chain lengths but chain lengths that were at least longer than the identity period (unit-cell dimension) of 10.25 Å along the fiber direction (Fig. 2.2). And also essential, they established a pyranose ring in a chair conformation for the glucose monomer (Figs. 2.3, 2.4). It is important to note that they came to their conclusion by model building with ball-and-stick models as well as space-filling models with the known atomic distances and valencies at the time and by comparing the estimated X-ray intensities with experimental data from various layer planes.

2.2 Concepts in Structural Research of Cellulose

9

They introduced secondary valencies with longer spacing between chains. It remained a mystery that the three-dimensional structure of diamond outgrew the unit cell in all three directions and the two-dimensional layers of graphite also surpassed the dimensions of the unit cell in two directions but that the one-dimensional chain with primary valence bonds of cellulose should be limited or contained in a small unit cell. The construction of models of inorganic and small organic molecules was quite common to visualize the placement of atoms according to the evaluation of X-ray data at the beginning of the 1920s but it was not extended to polymers nor was it used as a primary tool to construct a model with realistic properties. Sponsler and Dore proposed an orthorhombic unit cell which contained four chains but which can be reduced to a two-chain monoclinic one, which is commonly accepted today (Fig. 2.5).

Fig. 2.2 Representation of chain arrangements for native cellulose (ramie) in the crystalline microfibril. (From Sponsler and Dore 1926)

10

2 History of Cellulose Research

Fig. 2.3 Ball-and-stick model of glucopyranose as a monomeric unit of cellulose. (From Sponsler and Dore 1926)

Fig. 2.4 Chair conformation of a glucopyranose unit. (From Sponsler and Dore 1926)

The model building, today successfully performed by computer simulations and regarded as a most valuable tool, was still rejected by Meyer (1950) in 1950 (see also Meyer and Mark 1950) with the following argument: “Attempts to determine the position of the atoms based solely on models, for example Stuart’s hemisphere model (e.g. P. H. Hermans, Kolloid-Z., 102, 169 (1943)) are of no value and must be regarded as a step in the wrong direction. Not only must the spacing of the atoms obey the usual rules governing interatomic

2.2 Concepts in Structural Research of Cellulose

11

Fig. 2.5 a Unit cell of Sponsler and Dore (1926) in projection down the fiber axis. Subcell 5.40 Å, 6.10 Å. b Unit cell of Sponsler and Dore (1926) in comparison with that of Meyer and Mark (1928a). These two unit cells are related by a simple transformation. (a From Sponsler and Dore 1926; b from Kiessig 1939)

12

2 History of Cellulose Research

distances, but also it must be consistent with the physical properties of the material. Of these, the most important is its behavior under x-rays; whether or not a suggested structure agrees with theory can only be verified by measurements of the intensity. It is not possible to deduce the polymorphic character of cellulose solely from a consideration of wooden models nor can one invent models for each of the different modifications, which must however differ in structure, because they behave differently with x-rays. Arbitrary modifications of the crystal structure proposed by Meyer and Misch are valueless if they cannot be justified by intensity measurements.” There is no doubt that the X-ray method is the method of choice for crystal structure determinations, if enough data are available, but with limited data, additional information can be provided by model building (Fig. 2.6) and other means. It should be added that at that time the intensity calculations for cellulose consisted of a constant atomic scattering factor of f(C) = 6 and f(O) = 8, omitting H, and omitting also the temperature and disorder (lattice distortion) factor (Andress 1929a). In 1953, only 3 years after Meyer’s negative opinion on model building, the structure of DNA was proposed by model building with ball-and-stick models as a double helix by Watson and Crick (1953) and they were rewarded with the Nobel Prize in Chemistry. The much-criticized Hermans introduced through model building the bent conformation of the cellulose chain and the valuable O5–O3'(the prime denotes an adjacent residue) hydrogen bond along the cellulose chain, which was confirmed not only in structural studies of oligomers of b-d-glucose but also in

Fig. 2.6 Ball model representing the chain arrangements of cellulose. (From Sponsler and Dore 1926)

2.2 Concepts in Structural Research of Cellulose

13

cellulose. Concerning native cellulose, the data supplied by X-ray studies till the 1970s were not sufficient to discriminate between parallel and antiparallel chain arrangements nor could one rely on supplementary data from other sources, and this is true even today with a much broader data base. Model building as a serious tool cannot be neglected today in all kinds of molecular and crystal structure determinations of small and large molecules and in describing their interaction. The cellulose chain model of Sponslor and Dore (1926) had two drawbacks. Cellobiose was not a basic structural unit in their model; rather the glucose units have been linked through an alternating 1-1 and 4-4 linkage (Fig. 2.2), which did not affect the length of their dimeric unit with regard to cellobiose and was corrected by the work of Haworth and Freudenberg. The second point concerns the shape of the chain. They saw no need to introduce a “bent” structure as later proposed by Hermans (1943) on the basis of improved stereochemical data of interatomic distances and angles, which was required by a longer virtual bond length of 5.45 Å between the glycosidic or bridging oxygens. This virtual bond length by Sponslor and Dore amounts to 5.13 Å, half the distance of the fiber repeat of 10.25 Å. Their four-chain orthorhombic unit cell can be judged indifferently since it can be converted to the nowadays accepted monoclinic two-chain unit cell (Bragg 1930; Kiessig 1939) and as long as no space group assignment is provided this four-chain unit cell can be used for a description of the crystalline structure. Another point should be mentioned, namely, the correlation of molecular structure with morphology and further with properties of the materials. Sponslor and Dore (1926) first addressed this difficult task as they explained the anisotropy of strength, swelling and thermal expansion by parallel-running, main valence cellulose chains along the fibers (Meyer 1950). Further needed information on the size of macromolecules in general but also in the solid state or in solution was another point of long-lasting controversy and was addressed for some time with inadequate data. In the late 1920s neither a morphological model for crystallites, also termed “micelles” or “microfibrils” in the field of cellulosics, existed, such as the later-introduced concept of fringed micelles (Herrmann et al. 1930) with amorphous and crystalline parts along the chains, nor was chain folding developed with tie molecules between the folded lamellae. The chain size or length was considered as the length of a crystallite for some time, which one group of scientists regarded as an aggregation of chain molecules and another group as an aggregation of small molecules. The noncrystalline portion of scattering observed was thought to originate from the interactions or glue, which fixes the various crystallites. Staudinger represented a third opinion. He worked with organic chemical methods (building macromolecules with defined monomeric units or derivatizing by polymer analogous reactions) and introduced the intrinsic viscosity (Staudinger and Heuer 1930) as a tool for the determination of the degree of polymerization (DP). He proposed very long rod-like molecules in solution, but his methods were disputed by many of his colleagues. A final breakthrough came with the comparison of the DP of chain molecules in molecularly dispersed solutions of polymer analogous derivatized compounds on one hand and on the

14

2 History of Cellulose Research

other hand with the comparison of molecular mass or DP by various methods such as osmometry, intrinsic viscosity and later ultracentrifugation and light scattering. One further point should be mentioned as Haworth (1937, 1966) made clear in his Nobel Lecture in 1937 that the problem of linkage in cellulose was not solved in his opinion in 1937: “..it was not until 1925 that a precise structural model of any sugar was clearly and finally determined.” And “In 1934 I pointed out that this picture of xylan was probably typical of other polysaccharides in that these chains of limited length aggregated to form a larger entity and the nature of the bonds effecting the union of adjacent chains was discussed. It was suggested that these might be either united by principal valency links or by some other type of bond such as that which is responsible for coordination. Whatever this kind of agency or link may be, I prefer to describe it as the polymeric bond and as such it may differ from ordinary valency bonds and may find currency in the whole field of polymeric substances.” Concerning specifically cellulose he continued: “In this connection I suggested in 1935 that the molecular aggregate of cellulose may comprise an aggregation which not only increases the length of the chain, but also the width, by the lateral combination of adjacent chains. I pointed out that these factors must be recognized in any comparison of the molecular weight of cellulose determined by physical and chemical methods. All recent experiments in my laboratory have fully confirmed these conclusions. There can be no doubt that those forces which I describe as polymeric bonds are active in linking together adjacent chains of cellulose as in the case of xylan, glycogen, and starch. I do not share the view recently expressed that cellulose is constituted on the plan of a continuous loop of glucose units, this single loop being of a size to correspond with the high molecular weight found for cellulose by physical methods, although in my book on the constitution of sugars published in 1929 I suggested that this conception must be fully explored.” In 1930 Meyer and Mark expressed their view that with chemical methods no conclusion about the chain length can be drawn but the results are consistent with linked glucose rings as proposed by various researchers, e.g., Tollens, Irvine, Bertrand, Hibbert, Pringsheim and Karrer. These results cannot be contradicted by chemical methods but are not consistent with X-ray analysis. The chemical results leave room for discussion as Freudenberg (1933) emphasized that in the moment of scission a change may occur in structure, e.g., a linkage between C1 and C4 or C5 is possible. For Flory (1974, 1993) in his Nobel Lecture in 1974 the chain structure by primary valencies was trivial for macromolecules: “The concept of a chain molecule consisting of atoms covalently linked is as old as modern chemistry. It dates from the origins of the graphic formula introduced by Couper in 1858 and advanced by Kekulé, Loschmidt and others shortly thereafter. Nothing in chemical theory, either then apparent or later revealed, sets a limit on the number of atoms that may be thus joined together. The rules of chemical valency, even in their most primitive form, anticipate the occurrence of macromolecular structures.”

2.2 Concepts in Structural Research of Cellulose

15

And “The prevailing structural motif is the linear chain of serially connected atoms, groups or structural units. Departures from strict linearity may sometimes occur through the agency of occasional branched units that impart a ramified pattern to the over-all structure. Linearity is predominant in most macromolecular substances, however.” “It is noteworthy that the chemical bonds in macromolecules differ in no discernible respect from those in “monomeric” compounds of low molecular weight. The same rules of valency apply; the lengths of the bonds, e.g., C–C, C–H, C–O, etc, are the same as the corresponding bonds in monomeric molecules within limits of experimental measurement. This seemingly trivial observation has two important implications: first, the chemistry of macromolecules is coextensive with that of low molecular substances; second, the chemical basis for the special properties of polymers that equip them for so many applications and functions, both in nature and in the artefacts of man, is not therefore to be sought in peculiarities of chemical bonding but rather in their macromolecular constitution, specifically, in the attributes of long molecular chains.” In 1926 Staudinger (1926) gave a presentation at a scientific meeting in Düsseldorf and got strong opposition when presenting his ideas on primary valencies of macromolecules since polymerization by primary valency bonds was not generally accepted and 10 years later in 1936 at a meeting in Munich (Staudinger 1937) he still had to convince some of his fellow scientists of the size of the macromolecules and the conceptual differences in the organic chemistry of macromolecules in relation to low molecular weight compounds regarding reactions and behavior. Staudinger had no doubts concerning primary bonds of the huge molecules. At the same meeting Kuhn (1937) represented his statistical chain model, which was not accepted by Staudinger for cellulosics. Staudinger preferred rods as the idealized models for sol and gel solutions (Fig. 2.7). Hess (1937) stressed in his

Fig. 2.7 Representation of a cellulosic sol (monodisperse solution) (left) and a cellulosic gel (right). (From Staudinger 1937)

16

2 History of Cellulose Research

paper at that meeting that no end groups by chain degradation of cellulose are found and proposed the chains of cellulose as loops. The constituent and linkage of cellulose chains was a wide subject. Tollens (1883, 1895, 1914) proposed chain molecules of anhydroglucose (H10C6O5) units but had the wrong vision of the linkage and the number of atoms in the cyclic ring. Almost the full amount of cellulose was found by hydrolysis to glucose. Missing aldehyde reactions by the glucose compound led to the proposal of a cyclic pentagonal semi-acetal form with a longer side group. Irvine and Hirst (1923) found by hydrolysis of trimethylcellulose exclusively 2,3,6-trimethylglucose but a clear argument for a linear chain on one side and the 1-4-linked b-d-glucopyranose was still missing. A nonbranched cellulose chain in the crystal portion of the cellulosic materials can be confirmed by the unit-cell size in X-ray investigations. The dimensions perpendicular to the chain direction of the microfibrils are devoid of space for branching. The linear linkage between glucose units was deduced by consideration of oligomeric hydrolysis products, which represent a continuous transition to cellulose (cellobiose, Freudenberg 1921; tri- and tetraasaccharids, Willstätter and Zechmeister 1929; tri-, tetra and pentasaccharids, Zechmeister and Tóth 1931; Staudinger and Leopold 1934; cellotriose, Zechmeister et al. 1933), but the length of the chain remained obscure. Tollens considered four or 20 glucose units to constitute cellulose. The exclusive linkage and the chain length have been solved with joint chemical and physical approaches as mentioned above.

2.3

Arrangements of the Cellulose Molecules in the Solid State

Packing arrangements of cellulose chains is another topic which has been speculated upon till the present time. Since the cellulose chain has a directional sense with a nonreducing and a reducing end, the chains in a crystallite can be arranged in a parallel (all reducing ends on one side) or an antiparallel (on alternating sides) manner. This kind of problem was avoided by Sponslor and Dore but appeared when Meyer and Mark (Meyer and Mark 1928a; Mark and Meyer 1929) took up the idea of Haworth (1928) and Freudenberg and Braun (1928) with a 1-4-linked b-d-glucopyranose chain. In their first publication, Meyer and Mark (1928a) proposed parallel packing with an exact shift of one residue between the corner the and center chain (Fig. 2.8). The parallel arrangement was thought to be the result of the synthesizing mechanism in nature for native ramie cellulose. Later Mark and Meyer (1929) corrected their model (Fig. 2.9) with only a quarter shift of the center chain. In 1937 Meyer and Misch (1937) came to the conclusion that native cellulose is packed in antiparallel fashion (Fig. 2.10) because native cellulose can be converted with intact crystallites to mercerized cellulose II (called cellulose hydrate at that time). But this polymorph with the same unit cell can also be achieved by crystallization out of solution. Meyer and Misch argued that chains are randomly distributed in solution: half of the molecules point in one direction and the other half in the opposite

2.3 Arrangements of the Cellulose Molecules in the Solid State

17

Fig. 2.8 Model of cellulose arrangements by Meyer and Mark (1928a). The glucose residues are represented as regular hexagons. (From Meyer and Mark 1928a)

Fig. 2.9 Model of cellulose by Mark and Meyer (1929). Carbons of the rings, which are placed in a plane, are claimed to be strain-free. The glucose residues are represented as regular hexagons. (From Mark and Meyer 1929)

18

2 History of Cellulose Research

Fig. 2.10 Antiparallel chain arrangements as proposed by Meyer and Misch (1937). The pyranose rings are idealized as regular hexagons but actually are in the chair conformation. (From Meyer and Misch 1937)

direction and upon crystallization antiparallel packing occurs. The plausible argument of growing chains in parallel fashion by nature for native cellulose I was dropped and not discussed anymore. Neither of the two ideas used in their argumentation withstands a rigorous consideration. A change of parallel to antiparallel packing in the crystalline domains is quite common in a number of polysaccharides such as chitin, amylose and cellulose with no change of orientation in the fibrils. This transformation can be explained by interdiffusion of chains from neighboring crystallites whose parallel-packed chains are running in the opposite direction and intermingle driven by enthalpic effects. And parallel packing for solution-grown crystals occurs for low to medium molecular weight (DP 20–40 up to 250) double helical amylose with parallel strands. Parallel packing of cellulose chains was of some concern right from the introduction of this idea because it should lead to a polarization of chiral materials and to an increase of the energy density. As is known from recent studies on liquid crystals, a compensation of polarization can occur on scale larger than the molecular scale, e.g., as a supermolecular helicoidal structure. It is feasible that the polarization of the parallel arrangement of cellulose chains perpendicular to the fiber axis of the crystallites may be compensated by a twist of the microfibrils along the fiber axis. The polarization along the microfibrils finds its counterpart in the antiparallel, microfibrillar arrangements of neighboring crystallites, a necessity for the conversion of parallel to antiparallel packing by interdiffusion.

2.3 Arrangements of the Cellulose Molecules in the Solid State

19

With the work of Scherrer an estimation of the size of the crystalline domains (micelles, microfibrils) was possible and led to a correlation length of more than 100 glucopyranose units or 500 Å and a width perpendicular to the microfibrils of about 50 Å for ramie. The noncrystalline part was thought to originate from the interfaces of the microfibrils gluing the microfibrils to strong units (fibrils). With these dimensions in mind about 40–60 glucopyranose chains form a microfibril, of which about half of the hydroxyl groups lie at the interface as deduced from possible water adsorption (Meyer and Mark 1930). These hydroxyls at the interfaces may not be involved in the same intermolecular hydrogen-bonding system occurring inside the crystallite. Deviations of placement of hydroxyls at the interface may violate existing symmetries concerning the crystallites as the proposed 21 screw axis along the molecular chains and some required extinguished reflections due to symmetry might be observable. The pyranose rings of the cellulose chain consist of 4C1 chair configurations (Sachse–Mohr trans configuration) with carbon atom C4 high and carbon atom C1 low. The properties of slightly derivatized cellulose chains in solution are best described with very few non-4C1 forms (Brant and Christ 1990). A discussion about possible shapes of the pyranose residue is provided by Reeves (1950; 4C1 called C1) and the 4C1 chair configuration was experimentally determined by crystallographic means (Beevers and Cochran 1947; McDonald and Beevers 1950, 1952; b-d-glucose and cellobiose, Chu and Jeffrey 1968; cellobiose, Jacobson et al. 1961). A change in the residue configuration along the chain represents a defect. Already Schulz and Husemann (Schulz and Husemann 1942; Schulz 1946; Husemann 1947) had stated that about every 500th residue in a linear cellulose chain may deviate from the normal shape and is easily susceptible to scission. The conformation of the cellulose chain is needed for a complete description of the cellulose shape in solution or in the solid state. Besides the configuration of a single residue, knowledge of the twist between two neighboring residues is essential for such a description. The accepted 21 screw axis in the solid state has to be expanded to threefold, fivefold or eightfold helices for certain polymorphs and derivatives. For the accommodation of nonconventional, e.g., fivefold, helices by stick model building, a twist and bent conformation of the pyranose ring was proposed (Watanabe et al. 1968) but with today’s computer modeling techniques and data from model compounds, the available pyranose rings in the chair conformation can easily accommodate this kind of helical structure. In the 1920s a controversy arose regarding whether the chains of cellulose in the crystalline domains are straight or bent. As illustrated in Fig. 2.11 even a claimed straight chain has a bent virtual bond O4..O4'(adjacent bridge oxygen) and it is only a question about the size of bending by comparing the two drawings in Fig. 2.11. In the structural models in Figs. 2.8–2.10 the bridge oxygens are erroneously placed on the twofold axis (cf. Figs. 2.13, 2,15, 2.16). This problem is obsolete today, since the virtual bond of b-d-glucopyranose in cellulose requires bending. An almost constant virtual bond length of 5.45 ± 0.04 Å has been experimentally confirmed for oligomeric compounds of b-d-glucopyranose as well as for similar structures of chitobioside and mannose. This length of the virtual bond (5.45 ± 0.04 Å) has to fit a

20

2 History of Cellulose Research

Fig. 2.11 Straight (left) and bent (right) conformations according to Meyer and Mark (1930). (From Meyer and Mark 1930)

5–5.25-Å distance of the unit-cell axis for a monomeric unit of cellulosics and always requires a bent structure (Fig. 2.11). A flexible pyranose ring will be adjusted in the modeling procedure according to the bond length, bond angles and torsion angles known for a low-energy pyranose chair and overall minimal energy requirement. From today’s point of view it is obscure that only identity periods of 2, 3 and 4 times the projected length of a glucose unit on the fiber axis were accepted lying on twofold, threefold and fourfold axes of the unit cells, but that a fivefold helix axis placed between symmetry elements of the space group was beyond imagining. Hengstenberg (1927) found for polyoxymethylene 9 times the length of a monomeric unit as the identity period, which was attributed to the interaction of the chains (Meyer and Mark 1930). Today we know that a 9/5 helix causes this identity period. Nevertheless, the fiber repeat of crystalline celluloses served for a classification of cellulose structures without a structural vision or interpretation concerning the cellulosic chains (Table 2.1). In 1955 Kratky and Porod (1955) still excluded threefold and fivefold cellulose helices without strong deformation of valencies, which led Watanabe et al. (1968) to propose a bent and twisted pyranose ring as already pointed out. The third or fifth meridional reflections observed and explained by the helix theory of Cochran et al. (1952) did not lead Kratky and Porod to reconsider their negative opinion of threefold and fivefold helices. With the introduction of computer modeling these problems have been solved and a tension-free pyranose ring was established in these helical structures.

2.4 Chemical Constitution of Cellulose as a Macromolecule

21

Table 2.1 Classification of cellulose crystal structures according to fiber repeat. (Reproduced from Meyer and Mark 1940)

Mankind has used cellulose as an abandoned natural high polymer for centuries in its native or derivatized form. With progress in science it was a key material to investigate its structure and improve its properties with application of this knowledge. For a molecular structural evaluation, the chemical constitution and configuration have to be known and serve as prerequisites for today’s application of X-ray techniques in the solid state. On the other hand, a complete description of the chain arrangements and morphology in a fibril requires knowledge of the shape and length of the chain molecules. Cellulose played an exceptional role besides starch and rubber in developing the macromolecular concept and only through a combined effort of organic and physical chemistry as well as crystallography these structural problems could be solved. It is interesting to follow the routes science took to achieve the present-day picture for cellulose and high polymers. In the development of macromolecular science many important lines can be followed, of which three important ones will be considered next in detail. First the development of the concept of the constitution and configuration of cellulose (Freudenberg, Haworth), second that of giant molecules (Staudinger, Svedberg) and third the structural evaluation of conformation and packing of native cellulose I in the solid state (Sponsler and Dore, Mark and Meyer, etc).

2.4

Chemical Constitution of Cellulose as a Macromolecule

The discovery and development of new experimental methods started a more detailed investigation into biological and synthetic materials at the beginning of the twentieth century. The discovery of X-rays and their useful application in structural research in the field of solid-state matter as well as the development of IR spectroscopy for dynamical studies, of the ultracentrifuge, electrophoresis and improvement of

22

2 History of Cellulose Research

osmotic cells for investigations in solutions started the discussion of the macromolecular concept and the introduction of chain molecules.

2.4.1 Linkage of Cellulose – the Chain Structure of Cellulose (Freudenberg, Haworth) In the representation of the historical development given by scientists who were involved in the process, personal feelings were occasionally expressed and Freudenberg (1933) stated that the acting persons are visible in their accounts: for some it was trivial what others proposed with verve. Subordinate or erroneous facts were emphasized and important progress was overlooked. In 1933 as the goal of a concept had essentially been reached only few knew the path on which it was achieved. Freudenberg (1933, 1967) describes the proofs of organic chemistry in great detail which led to the chemical constitution of cellulose. And he evaluates their origin and the overcoming of the arguments for the hypothesis of small aggregates. One of the main problems of cellulose research in the 1920s consisted of the linkage of the cellobioses as units of cellulose and the possible unique kind of linkage as a chain. According to Freudenberg (1933), small molecules linked by lattice forces should not have been considered as the alternative from the beginning as a challenge of the chain concept and should not have been discussed at all, if the experimental findings had been correctly evaluated. Before 1920 it was known that cellulose consisted of glucose. Three hydroxyl groups on each glucose unit were detected by derivatization and degradation. The coupling of the di-, tri- and tetrasaccharides was discussed and in comparison with peptides it was thought that cellulose consisted of at least six glucose units. Tollens (1914) discussed a “relatively high number” of glucose units, which were for him four or 20 units. Nastukoff (1900) proposed 40 units by acetal not semi-acetal connections but as a large ring. Freudenberg (1921) stated that 60% cellobiose was created by acetolysis from cellulose by statistical scission, of which 20% was lost during the reaction and the remaining amount of 40% cellobiose was verified. The discussion about the constitution of cellulose focused on the lower amount of cellobiose obtained by hydrolysis at that time (Freudenberg 1921), which was carefully evaluated and explained. According to Freudenberg, the considered alternative of small aggregates in comparison with large chains was never the problem of polysaccharide chemistry. This problem had been addressed at a time when it had already been solved in its essential parts. The experimental findings were in full agreement with continuously running chains in which every glucose residue was linked to the next structurally and configuratively in the same manner as in cellobiose. This point of view was in agreement with X-ray results (Polanyi 1921). The problem to be discussed consisted only of whether the existing chains are connected in a uniform or a nonuniform manner. The existence of triaccharides or other oligosaccharides was not questionable but rather their constitution and configurative relation to cellulose were. The

2.4 Chemical Constitution of Cellulose as a Macromolecule

23

experimental fact that no more than 60% of cellobiose materials were formed by statistical scission in hydrolysis pointed towards a uniform connected chain, which proves that every glucose residue is structurally and configuratively similarly connected to the next. Polanyi (1921) presented data of Herzog and Jancke (Herzog and Jancke 1920 a, b; Herzog et al. 1920) and concluded from their X-ray experiments that the symmetry of the unit cell belongs to the orthorhombic class, but he did not exclude the monoclinic space group, and that cellulose consists either of chains in the form shown in Structure 2.1 or of rings in the form shown in Structure 2.2.

C6H10O4 O

O

O4H10C6

Polanyi could not know that a few weeks before his work was published, the idea of uniform chains had been developed by preparative chemistry, and the idea of a ring-shaped cellobiose anhydride, which should lead to 100% biose, was dropped. The formation of such tetraosanes to higher aggregates requires a radical and improbable enlargement of the concept of the chemical bond. Therefore, there was no need for this hypothesis to be discussed further besides the well-based chain formula. Irvine and Hirst (1923) further clarified the linkage of glucose, establishing that from trimethylcellulose only 2,3,6-trimethylglucose was obtained, suggesting a linkage in the 4- or 5-position. And in 1926–1927 Haworth reported the structure of cellobiose as 1-4-linked glucopyranose rings, which is in full agreement with the identity period established by X-ray data. Up to 1925 glucose was regarded as pentagonal ring (Böeseken 1913; p. 349 in Marsh and Wood 1939). Sponslor and Dore explicitly rejected such a pentagonal ring and came to the same conclusion of a hexagonal ring for glucose (Fig. 2.3) as Haworth did. The futile search for tetramethylglucose as a degradation product from the chain ends led to the proposal of large chains probably of several hundred glucose units (Freudenberg and Braun 1928). Haworth established a lower limit of 100–200 units taking possible errors into consideration by experimenting with large amounts of trimethylcellulose. The biose anhydride established by Bergmann and Knehe (1925) and Hess and Friese (1926) as small aggregates forming cellulose was later identified as units of four residues (Meyer and Mark 1928b) or ten to 16 residues (Freudenberg 1929) of the cellulose chain. With the discovery of the pyranose form of sugars and the configuration of cellobiose by Haworth (Haworth 1925; Charlton et al. 1926), it became clear that cellulose consisted of 1-4 linearly linked b-d-glucopyranose (Freudenberg 1928). But the actual size of the molecule as well as the shape were not clear despite a

24

2 History of Cellulose Research

lower limit for the size being established by organic chemical means. We will turn to these problems in the next sections, which were the primary research field of Staudinger in Freiburg, who faced strong opposition from researchers favoring a buildup of aggregates by strong forces of small molecules (anhydrides) or short chains hooked up by “polymeric bonds” (Haworth). The imprecise term “polymerization” at that time and its relationship to the term “molecule as units” especially in connection to starch with crystallizable Schardinger dextrins of small units pointed towards the presence of association as polymerization products and not to primary valencies as in molecules. Still in 1919, the term “polysaccharide” was in use for bioses, trioses up to cellulose and starch. The oligosaccharides found in nature and those synthesized were thought to be coupled hexoses.

2.4.2 Macromolecule Formation – Size of the Chains (Staudinger) Staudinger (1938) developed another concept to solve the pending problem of macromolecules. He regarded the year 1921 as a turning point in the field of cellulose chemistry, since in this year Karrer (Karrer and Nägeli 1921; Karrer 1925) provided a new concept about the constitution of starch and cellulose, which led to a decade of controversial discussions. Karrer proposed that starch and cellulose consist of anhydrides of maltose or cellobiose, respectively, as result of his degradation experiments. The colloidal solutions formed consist of aggregates of secondary bonded small molecules, so-called micelles in his view. This idea was supported by the small unit cell of cellulose determined by X-ray diffraction and the molecular weight determination by Hess et al. (Hess and Schulze 1926; Hess and Pichlmayr 1926; Hess and Friese 1926) and Bergmann (1926) on cellulose acetates and cellulose ethers by the cryoscoptic method. A large depression of the freezing point led Hess to the conclusion that this observation is caused by glucose anhydride. Years later it was found that their method cannot be applied to cellulose acetate and that the osmotic pressure method actually led to the expected high molecular weights. Biologists supported Karrer’s opinion, since they confirmed the ideas of Nägeli, who found in the 1860s that cotton and other high molecular biopolymers crystallize. He concluded that the crystalline domains, which he called micelles, caused the colloidal character in solutions. In 1925 Herzog (Herzog 1925; Herzog and Krüger 1925) observed that the crystallite size of solid cellulose determined by the X-ray methods agreed with the size of the micelles of nitrocellulose in solution deduced by diffusion. These findings were corrected several years later (Herzog and Kudar 1934). The idea that cellulose is a low molecular weight compound was generally accepted in the years 1921–1927, since the arguments by Karrer, Hess and Bergmann could not easily be contradicted in the field of cellulosics. In 1920 Staudinger had already proposed that a large number of small basic molecules are linked together to form macromolecules (Staudinger 1920; Staudinger and Fritschi 1922), and he was able to confirm the high molecular weight by end-

2.4 Chemical Constitution of Cellulose as a Macromolecule

25

group analysis in the case of polyoxymethylene and considered this compound as a model for cellulose (Staudinger and Lüthy 1925; Staudinger 1929). Some of the supporting arguments for aggregation of small molecules lost their validity when it was found that a small unit cell for crystalline cellulose as established by X-ray diffraction does not contradict the presence of long-chain molecules (Staudinger et al. 1927a, b; Staudinger and Signer 1929). Sponsler and Dore (1926) published a model of cellulose as a chain molecule derived from X-ray experiments based on atomic distances of organic compounds with linked pyranose rings and explained the 10.25-Å fiber repeat by primary valencies between the linked b-anhydroglucopyranose units (AGU). The mistake of alternating 1-1, 4-4 linkage of the AGU was corrected by Haworth (1927, 1928) and others. This corrected formula represents the basis for structural considerations by Meyer and Mark (1928a) and by others since 1928. The chain structure of cellulose was further supported by hydrolytic scission leading to oligosaccharides (Willstätter and Zechmeister 1929; Zechmeister and Tóth 1931) and the b-linkage was further supported by Freudenberg (1933). The length of the chain was not determined in these investigations. Meyer and Mark proposed a model for this problem, especially for colloidal solutions. Taking the proposal of Herzog that the solid-state micelles resemble those in solutions as well as the result of Sponsler and Dore, they proposed that the primary valence chains of definite length form blocks or micelles and are bound together by strong cohesion forces, so-called micellar forces. They will not be separated in solution as occurs for small molecules (Meyer 1928; Mark 1928). The micelle was regarded as the basic unit describing high molecular weight compounds (Fig. 2.12). It was assumed from X-ray investigations (line width) that 30–50 AGU (Meyer and Mark 1928a), later 60–100 AGU (Mark 1928; Meyer and Mark 1931), form the primary valence chain (the projection of one AGU on the chain axis is about 5 Å) and about 40–60 chains built a single irreversible micelle. The molecular weights of the “macromolecules” determined by osmotic pressure experiments were termed “micellar weights.” The high viscosity of cellulose and its derivatives in solution and also the swelling seemed to be consistent with the micellar model of long chains. This idea was well accepted by colloidal chemists and biologists but was proved to be wrong by measurements of

Fig. 2.12 Model of the morphological buildup of a ramie fiber according to Seifriz (1929), Meyer (1930) and Mark. (From Staudinger 1938)

26

2 History of Cellulose Research

the molecular weights of cellulose triacetate, which were obtained by polymer analogous reaction. Meyer and Mark later abandoned the micellar model and turned to the macromolecular concept. They supported the fiber model of FreyWyssling (1936), which originated from the fringed micellar concept of Herrmann et al. (1930) as well as Staudinger’s result that the cellulose chains have to be longer than the micelles. Therefore, chains might pass through several micelles. The idea of Staudinger proved to be correct that the entities or basic structural units in solution are dissolved macromolecules and not micelles. He was awarded the Nobel Prize in Chemistry in 1953 “for his discoveries in the field of macromolecular chemistry” (Staudinger 1953, 1964). He was able to prove his early assumptions with methods of organic and physical chemistry and set the foundation of concepts in this field that are still recognized today. He introduced what he called the polymer analogous derivatization of cellulose, i.e., he derivatized cellulose without changing the DP and measured the same DP (actually the corresponding molecular weight) starting from cellulose and going to cellulose triacetate and back to cellulose or to cellites (not fully substituted cellulose acetate), methyl cellulose and methylacetyl cellulose. The DP was determined by osmotic pressure experiments and later quite often by viscosity by applying the intrinsic viscosity [h] relationship: hsp [h ] = lim c→0

c = Km M a ,

(2.1)

where c is concentration, hsp is specific viscosity, Km is the experimentally determined constant for a specific solution adjusted sometimes from low molecular weights to higher molecular weights, M is molecular weight and the value of a depends on shape and solution properties of the molecules. Staudinger assumed a = 1 for cellulosics as rod-like macromolecules in solution (Staudinger and Heuer 1930) (a ≠ 1 for polyester) even in his later years (Staudinger 1953, 1964). He argued that the cellulose molecules are present as stiff rods in the solid fiber and in the same extended form in solution (Fig. 2.7). This value of a= 1 is correct for cellulose trinitrate for not too high DP (DP < 1,000). For cellulose acetates a value of a = 0.9 was proposed by careful evaluation of experiments in 1951 (Philipp and Bjork 1951a, b) after an extensive discussion of various experimental results, but it was not accepted by Staudinger. Staudinger’s main idea dwelled on a molecularly dispersed distribution of macromolecules in dilute solution. Nevertheless, he considered aggregation for higher concentrations. From today’s point of view, it turns out that polymer aggregation in dilute solution is neither as general as assumed by Meyer and Mark and others in the early days of polymer science nor generally absent as claimed by Staudinger. The breakthrough of the macromolecular concept was achieved by the systematic study of the chemical and physical properties of a homologous series from oligomers to high polymers of polyoxymethylene. All these materials have the same structural principle as is also verified for polymer analogous substitution of cellulose to obtain cellulose triacetate or nitrate and reactions back to cellulose obtaining the same DP. Important steps in cellulose research to establish the macromolecular concept were the determination of the configuration of the monomers and the link-

2.5 Historical Development of X-ray Models for Native Cellulose

27

age between them as well as the size (molecular weight) and shape of the cellulosic chain by osmometry, light scattering, small-angle X-ray scattering, ultracentrifugation and viscometry. Further information on the crystalline structure of fibers and films can be extracted by evaluation of X-ray data and spectroscopy.

2.5 Historical Development of X-ray Models for Native Cellulose The structure determination of cellulose by X-rays played an important role at the beginning of polymer science and the concept of long-chain molecules. The progress of structural research and the impact thereof can be evaluated and studied at certain stages by this development. But it should be emphasized that X-ray investigations cannot provide clues for the molecular mass of cellulose. The first diffraction patterns of bamboo and hemp were published shortly after the discovery of the interference phenomena of X-rays but did not contain detailed structural data (Nishikawa and Ono 1913; Nishikawa 1914) except the orientation of crystallites along the fibers. Scherrer (1918) applying the technique of what we call today a Debye–Scherrer diagram carried out a further investigation on randomly distributed cellulose. He claimed an amorphous cellulose structure. Independently Herzog and Jancke (1920a) and also Scherrer (1920) observed an interference pattern from powdered native cellulose and from a film cast from viscose and concluded that a crystalline structure was present in agreement with birefringence experiments. Fiber patterns were obtained by irradiation of bundles of fibers of various native sources with monochromatic X-rays by Herzog and Jancke (1920b) and Herzog et al. (1920). Native cellulose of different sources, including wood, showed the same pattern, i.e., exhibited the same structure, and it was also concluded that wood contains crystallites or microfibils of cellulose. The diagrams of, e.g., ramie resemble those of rotation exposures of single crystals but with nonideal orientation, which has to be considered when correcting the intensities of X-ray reflections (Lorentz factor; De Wolff 1962). The correction factors will be different from the ones for rotation exposures of single crystals. Herzog et al. (1920) and Weissenberg (1921) discussed the arrangements of the crystallites in these fibers and the degree of orientation or the deviation from the ideal orientation from the arcing of the reflections. They found agreement with the results from the optical rotation data of Ambronn (1916a, b, 1917). A quantitative evaluation of the reflections was carried out by Polanyi (1921) in part with Weissenberg (Polanyi and Weissenberg 1922) and an orthorhombic unit cell was proposed but a monoclinic one was not excluded (Table 2.2). Polanyi also discussed the consequences of his work and found that a decisive discrimination of long main valence chains versus small main valence complexes cannot be achieved with the X-ray method (see the discussion in Sect. 2.2). A long-lasting debate followed between supporters of these two concepts in the 1920s. Improved exposures and use of higher-oriented cellulose samples widened the data base and placed the

Table 2.2 Unit-cell dimensions of native cellulose. Historic survey of Schiebold (1944), completed to 2004. (Reproduced in part from Schiebold 1944)

28 2 History of Cellulose Research

19. 20. 21. 22. 23. 24. 25.

Nature (London) 181: 326 (1958). J. Polym. Sci. 42: 189 (1960). J. Polym. Sci. 32: 371 (1958). J. Polym. Sci 56: 339 (1962). J. Appl. Polym. Sci 18: 3373 (1974). Biochim Biophys.Acta. 222: 109 (1970). Macromolecules 7: 486 (1974).

b/ Å

materials

monoclinic (α=90, β=57.5) triclinic monoclinic monoclinic P21 monoclinic (reduced cell) monoclinic (α=90, β=57.0) triclinic monoclinic P21 (EM) (α=117, β=113) triclinic and monoclinic P21 monoclinic P21 (T=293 K) triclinic (α=118.08(5), β=114.80(5))

symmetry

Tunicate Glaucocystis

Microdictyon

Valonia Valonia Valonia Ramie Valonia

Chaetomorpha

materials

monoclinic (electron diffraction, −100o C) Valonia triclinic, P1 monoclinic Ramie monoclinic (Sponslor, Dore type) monoclinic, not P21(EM) Ramie, cotton

symmetry

26. Biopolymers 13: 1975 (1974). 27. Biopolymers 15: 1903 (1976). 28. Macromolecules 13: 1183 (1980). 29. Macromolecules 23: 3196 (1990). 30. Macromolecules 24: 4168 (1991). 31. JACS 124: 9074 (2002) 32. JACS 125:14300 (2003)

8 2 8 8 2 2 2 8 2 1 2 2 1

97.0 96.2 96.8 97.0 97.04 96.5 96.6 97.3 97.3 81 97.3 96.55(5) 80.37(5)

a/ Å

16.43 15.70 10.33 Nieduszynski, Atkins24 1970 1974 9.41 8.15 10.34 Sarko, Muggli25 1974 15.76 16.42 10.34 or Gardner, Blackwell26 1974 16.34 15.72 10.38 1976 8.18 7.84 10.38 Claffey, Blackwell27 1980 Woodcock, Sarko28 8.20 7.78 10.34 9.54 8.25 10.36 Sugiyama et al.29 Iα 1990 or 15.84 16.44 10.36 Iβ 7.92 8.22 10.36 6.74 5.93 10.36 Sugiyama et al.30 Iα 1991 1991 Iβ 8.01 8.17 10.36 Nishiyama et al.31 Iβ 2002 7.784(8) 8.201(8) 10.38(1) Nishiyama et al.32 Iα 2003 6.717(7) 5.962(6) 10.400(6)

Z

Z

10.58 10.34 10.34 0.3 1 10.3

β/o

γ/o

15.88 15.68 7.85 12.08 7.9

b/ Å

c/ Å

year

Author

16.78 16.40 8.17 10.85 8.35

c/ Å 8 8 2 4 2

1958 1960 1958 1962 1974

Honjo, Watanabe19 Fisher, Mann20 Jones21 Ellis, Warwicker22 Hebert, Muller23

a/ Å 82 82 83.6 93.2 84

year

Author

b:

Table 2.2 (continued)

2.5 Historical Development of X-ray Models for Native Cellulose 29

30

2 History of Cellulose Research

indexing and determination of the unit cell on a secure basis (Table 2.2). Besides the fiber period of b=10.3 Å (by Sponsler and Dore and today denoted as the c-axis) along the chain axis, the other two dimensions were determined to a = 8.3 Å and c = 7.9 Å and the angle between these two axis was determined to be b = 84°, space group P21 (Table 2.2). Andress (1928) first proposed a monoclinic unit cell with a monoclinic angle of 78°, which was miscalculated as pointed out by Schiebold and Andress himself and he was able to index all observed reflections on an X-ray diagram. Mark and Meyer (1929) reported an angle of 84° and referred to the work of Andress (1928). It was assumed from the cell volume of approximately 670 Å3 and the experimental density of the fibers that the unit cell contains four glucose residues. Herzog and Krüger (1925, 1929) estimated from diffusion experiments and Hengstenberg and Mark (1928) determined from the width of the X-ray reflections the size of the crystallites along the fiber axes to about 600 Å and the diameter of a crystallite to be 55 Å. A second polymorph was found for cellulose called cellulose hydrate, but it is now called cellulose II, as proposed by Hess et al. (Hess and Trogus 1935; Hess and Gundermann 1937) or Hermans and Weidinger (1946), since this structure contains no water. This material is obtained by regeneration of cellulose derivatives or by mercerization and both procedures lead to the same lattice. Cellulose II was first described by Herzog and Jancke (1920a, b) and the unit cell of mercerized cellulose was determined by Andress (1929b) and that of regenerated cellulose by Burgeni and Kratky (1929). The conformation of the cellulose II chain reported by Andress resembles that of native cellulose I with all the drawbacks of this model. He placed the chains in the corners and in the center of the unit cell of cellulose II and assumed packing arrangements in parallel fashion. Figure 2.13 shows parallelrunning chains and was drawn with the coordinates of Andress (1929b).

Fig. 2.13 Model for cellulose II according to Andress (1929b). Conformation and packing are similar to those of cellulose I, parallel chains and a quarter shift of the center chain. Only two corner chains and the center chain have been depicted for better visualization

2.5 Historical Development of X-ray Models for Native Cellulose

31

Burgeni and Kratky discussed two possibilities for the size of the unit cell both belonging to the monoclinic lattice type. The smaller unit cell contains four glucose units and the second one is exactly twice the base plane size. This evaluation resembles the proposals for cellulose I with a two-chain or a four-chain unit cell. The smaller unit cell was preferred because all X-ray reflections can be indexed with this unit cell and space group P21 was assigned. The structure determination of polymers by X-rays suffers from the unavailability of a sufficient numbers of reflections in three-dimensional space in contrast to the structure determination for many small molecules by the single-crystal method. Generally, uniaxial fibers and by special treatment sometimes poor biaxial fibers are available. The uniaxial orientation of extended polymer chains leads to few reflections in two dimensions (fiber pattern), which can be collected on a screen (photographic film, imaging plate, etc.; Fig. 2.14). Special polymerization processes may produce single crystals, which are the exception not the rule. The polymer fibers exhibit imperfect crystals and are additionally disturbed by some amorphous parts present in the materials. Therefore, further information is needed on the materials in addition to X-ray data, such as the chemical constitution, configuration and intramolecular and intermolecular interactions provided by spectroscopy and other means. The conformation of cellulose as chains of 1-4-linked b-d-glucose in the 4C1 chair configuration with primary bonds between residues was not known at the beginning of the structural research on native fibers of cellulose and caused many controversies. This kind of long polymeric chain arrangement was supported by the

Fig. 2.14 Fiber X-ray pattern of native cellulose I on a flat film with indicated layer lines (ramie, fiber axis b). (From Hermans 1949)

32

2 History of Cellulose Research

results of the basic research of Staudinger on synthetic and biological polymers, especially on polyoxymethylene as a model for cellulose after 1920. A chainlike constitution of cellulose was discussed almost at the same time by Freudenberg (1921) and Haworth and Hirst (1921) as well as by Herzog and Jancke (1921), but not in the precise form as by Staudinger. In a fundamental paper, Sponsler and Dore (1926) proposed for native ramie cellulose continuous chains joined by glucoside-like primary valence linkages. They not only could explain the fiber period of 10.25 Å by primary valence bond chains but also proposed a chair conformation of b-d-glucose as a pyranose ring. They started to build ball-and-stick models to a scale with the atomic radii known at that time in a proposed orthorhombic unit cell, which contained four chains running through this cell. A monoclinic unit cell of almost the size represents a better solution. And they came up with a three-dimensional structural model for cellulose. An estimation of X-ray intensities of various scattering planes led them to propose a repeating 1-1 and 4-4 linkage of the glucose units to explain the intensities of the visible odd-numbered meridional reflections. A 21 screw axis along the chain is avoided by this idea. They placed the bridge oxygen on the molecular axis (c-axis in their nomenclature). This placement of the bridge oxygen was erroneously kept on the molecular axis in subsequent schematic drawings of structures, also when a 21 screw axis of the cellulose chain coincided with the symmetry axis of the unit cell, but actually was not in agreement with the published coordinate sets (compare Figs. 2.9, 2.10 with Figs. 2.15, 2.16). It was emphasized by Hermans (1943), driven by improved knowledge of stereochemistry, that realistic stereochemical models with the given fiber repeat cannot be obtained with the oxygen positioned on the axes. It then still took a further 10 years for the proposal to be made that the helical axis of a chain molecule does not have to coincide with symmetry elements of the space group and, e.g., a fivefold helix may be present in a unit cell of a crystal. As

Fig. 2.15 Mark and Meyer (1929) model of native cellulose redrawn with the published coordinates. Only two corner chains and one center chain are shown for better visualization

2.5 Historical Development of X-ray Models for Native Cellulose

33

Fig. 2.16 Model of antiparallel-arranged cellulose (Meyer and Misch 1937) redrawn with the published coordinates. Only two corner chains and one center chain are shown for better visualization. Some of the bonds are outside the established limits of the time as already recognizable from the drawing

Kiessig (1939) reported, Sponslor and Dore missed some innermost strong reflections because of their equipment, which led them to propose the 1-1, 4-4 linkage along the cellulose chain. Their orthorhombic unit cell can be transformed to the Andress (1928) or Mark and Meyer (1929) two-chain monoclinic unit cell (Table 2.2) as shown later (Bragg 1930; Kiessig 1939) with the nowadays accepted monoclinic cell parameters. Schiebold (1944) provided a further argument for supporting the monoclinic space group: the position and correlation of the cellulose molecules are not in agreement with the symmetry elements of an orthorhombic space group. The wrong primary linkages given by Sponslor and Dore avoided the later longlasting discussion of chain arrangement, parallel or antiparallel. With a continuous 1-4 linkage the cellulose chain exhibits two different terminal residues at the two ends, reducing and nonreducing, and a chain direction has to be introduced. The chains can be arranged in parallel or antiparallel fashion. It has to be stressed and was expressed by many researchers (Kiessig 1939; Schiebold 1944; Zugenmaier 1981) that the Sponsler–Dore model was the basis for all crystalline models that were later proposed. Sponsler and Dore (1926) explained with their model many of the physical properties of cellulose fibers with the chains running along the fiber axis, such as the anisotropy of the swelling predominantly perpendicular to the fiber axis, the tensile strength and the conversion into ethers and esters. Meyer and Mark (1928a) took the Sponsler and Dore model and introduced the 1-4 linkage as established by Haworth and Freudenberg but did not give enough credit to the original work from which they took the lattice arrangement of the chains and the proposed distances as outlined by Kiessig (1939). Priesner (1980) in his survey on the development of polymer science wrongly assumed that the Sponsler and Dore paper was not readiliy available for German scientists, and missed that it actually had been cited in the paper of Meyer and Mark (1928a). Marsh and Wood (1939) explicitly state that

34

2 History of Cellulose Research …the proposed structure of Sponsler and Dore was criticized by Haworth at the Annual Meeting of the Society of Chemical Industry at Edinburgh in 1927, and also before the University of Basle and again to the Swiss Chemical Society at Neuchatel (Helv. Chim. Acta, 1928, 11, 547). At Edinburgh, Haworth acquainted Meyer with the work of Sponsler and Dore suggesting at the same time that the 1:1; 4:4 linkage must be replaced by the 1:4;1:4 linkings. A note to this effect is also given by Freudenberg (J.S.C.I., 1931, 50, 287. c.f. Annalen, 1928, 461, 130).

The advantage of the model of Meyer and Mark lies in the correct description of the linkage of the cellulose chain and that all inferences are produced with the given monoclinic unit cell as well as the estimation of their X-ray intensities being acceptable. Two papers by the two authors with slightly different models were published. In the paper of 1928 (Meyer and Mark 1928a) the parallel-running corner chain and the center chain are shifted by one residue in the chain direction. This shift was reduced to a stagger of half a residue in a paper of 1929 (Mark and Meyer 1929). However, Meyer and many authors up to the present day refer to the first paper and provide the model of the second paper (1929) as a figure. A drawback of their work, in contrast to the Sponsler- Dore model, is certainly that all six carbon atoms and the hydroxyl group of C6 are considered to lie in one plane. Only the ring oxygen is placed somewhat off this plane, which causes a highly strained ring. However, the ring was claimed to be strain-free by the authors. Further the bond lengths and bond angles were outside of the known range at that time as listed, e.g., by Meyer (1928) and provide a somewhat strange projection of the pyranose ring (Figs. 2.13, 2.15, 2.16). With this geometry the glycosidic bridge angle of 100.2° is far from an acceptable value. Andress (1929b) also provided a coordinate set for native cellulose I, which agrees with the set of Mark and Meyer except for the coordinates of the glycosidic oxygen O4. It seems that Mark and Meyer knew of and corrected the coordinate set of Andress, since it was mentioned in their paper. The coordinates of Andress led to an acceptable bridge angle but resulted in a short distance between O4 and C2'of the adjacent residue of less than 2.0 Å. This model of Andress concerning the bond lengths and angles was strongly criticized by Meyer and Misch (1937) in 1937 but they were not concerned with the same drawbacks of Meyer’s own work (Schiebold 1944). A point of criticism in the 1940s was also the parallel chain arrangement, which turned out to be correct from today’s point of view. Since the chains lie on 21 screw axes, odd meridional reflections should not appear. One anhydroglucose unit on each of the twofold axes of the corner chain as well as the center chain in the unit cell suffices for a complete description of the crystal structure. Here, we omit further proposed models similar to the Meyer–Mark model, which have influenced less the development of structural research of cellulose over the years. It interesting that Haworth (1928) himself proposed a space model for cellulose with a 1,4-glycosidic linkage of the main chains after rejecting the model of Sponslor and Dore. Figure 2.17 represents this model, which has a 21 screw axis but it exhibits van der Waals contacts that are too close, such as C2..C5' of approximately 1.5 Å, and no space is left for the hydroxyl groups. The fiber repeat could not be explained nor could the X-ray intensities. Haworth dropped his proposal later.

2.5 Historical Development of X-ray Models for Native Cellulose

35

Fig. 2.17 Space model of cellulose according to Haworth (1928)

The next major step in structural research on cellulose occurred with the proposal of a different model by Meyer and Misch (1937). The unit cell, space group and the general placement of cellulose as corner and center chains within the unit cell remained as for previous models. However, antiparallel chain arrangements were introduced as well as a strain-free conformation of the glucose ring. This model led to fair agreement between the calculated X-ray reflection intensities and the observed ones. But as Schiebold (1944) remarked, these calculations are relatively unreliable since the atomic scattering factors for C and O applied were not correct and the temperature factors and the hydrogen atoms were completely omitted. The antiparallel chain arrangement was deduced from the fact that by spinning fibers of cellulose II out of solution chains of both senses of direction are equally probable and upon crystallization lead to antiparallel packing. Since native cellulose fibers can be converted topochemically into cellulose II by preservation of the external form, native cellulose has to be arranged in antiparallel fashion. As we know today, this argumentation is not correct and chains may well interdiffuse originating from up and down strands (microfibrils) of parallel-packed chains. This antiparallel model allowed a hydrogen-bonding network between chains (corner–corner, center–center and corner–center) to form strong interactions perpendicular to the fiber axis. Nevertheless, this new model was also not up to date in terms of chain conformation at that time, i.e., in bond lengths and angles. The C6–O6 bond lies cis to C5–O5 (torsion angle approximately −10°), which causes a short van der Waals contact between O6..O5 of 2.67 Å. The torsion angle s (C3–C4–C5-C6) amounts to 180° and a hydrogen bond is absent between O3..O5 with a distance of 3.47 Å. In addition the pyranose ring has the l configuration and problems arise in contacts between atoms concerning packing. In a later paper, Meyer and van der Wyk (1941) suspected that a statistical distribution may be present in ordered cellulose modifications, since little differences exist in interaction for the up and down chains. The idea of a statistical distribution of chains was recently taken up by Blackwell (2000) for the explanation of the combined existence of native cellulose in two polymorphic forms, Iα and Iβ. However, the statistical consideration is now applied to two different parallel chain arrangements. Selecting one or the other model, parallel or antiparallel, was not

36

2 History of Cellulose Research

possible at the time the models were proposed. The calculated X-ray intensities differ little for the various models. A decision solely relying on diffraction was rather ambiguous at that time. Schiebold (1944), reviewing all available native cellulose models, suggested that the Meyer–Misch model represents a workable approximation but several points need to be improved. The given symmetry with a 21 screw axis as a space group element does not correspond to the observations of odd meridional reflections, and it may be that a lower-symmetry space group has to be assumed. Further, discrepancies exist between the ratios of the relative intensities of the meridional reflections as well as between reflections for certain layer lines. The hydrogen-bonding scheme is not fully developed, as it should be when compared with crystal structures of low molecular mass sugars. A remark is needed at this point. Kiessig (1939) carefully investigated the meridional reflections and clearly showed that odd-numbered reflections are present. He came to the conviction that a 21 screw axis along the cellulose fibers cannot be present and concluded that a completely different cellulose model has to be considered. He suggested that chains with two different conformations in a statistical manner might be present in the cellulose fiber, one of which exhibits a conformation with a 21 screw axis, and the other one has no helical symmetry. Such a model seemed to be compatible without an additional hypothesis based on X-ray data, and as we may conclude that it resembles the later proposal based on NMR data (Atalla and VanderHart 1984). Another remark of Kiessig is worth mentioning. Kiessig did not accept that other than twofold and threefold molecular axes are present in a crystal lattice. He envisioned that these symmetry elements have to coincide with space group symmetry elements. He believed that it was impossible for a fivefold helical axis to exist in any crystal lattice. Later it was recognized that the axes of helical structures might lie in the interstitial spaces between the symmetry elements of the space group. On the other hand, the helical descriptions of polymers (nomenclature u/t; motifs u and turn t) might be only approximations as proven for polyoxymethylene whose established 9/5 helix of a second modification is really an approximation. Actually, the glucose residues in most of the celluloses lie on a possible screw axis of proposed specific space groups, e.g., on the 21 screw axis of P21. This position of the chains was recently confirmed by a thorough investigation and discussion for native cellulose Iβ (Kono and Numata 2006) and for cellulose II (Kono and Numata 2004). In contrast, native cellulose Iα crystallizing in a space group without symmetry elements does not have any helical symmetry and shows two different residue conformations as a chain repeat. It seems difficult to come to a firm conclusion regarding the screw axis present in cellulose fibers from the evaluation of fiber X-ray data only and/or electron polymer single crystal diagrams. Extinct reflections might be present owing to defects or reflections might be accidentally extinct. Amazingly, Kiessig as well as Sponsler and Dore considered two different crystal forms or statistical arrangements of chains as a possibility to overcome the discrepancies concerning the presence of extinct reflections by the proposed space group, which had little impact on the discussion of the cellulose crystal structure over the years.

2.5 Historical Development of X-ray Models for Native Cellulose

37

Fig. 2.18 The crystallite orientation of stretched tunicin cellulose and corresponding X-ray patterns. The arrows indicate the X-ray beam direction, which led to the corresponding diagrams shown. d direction of stretching, q transverse, s perpendicular to the plane. (Reproduced from Mark and von Susich 1929)

Two different orientations of crystallites (micelles) were proposed for highly oriented animal cellulose tunicin by Mark and von Susich (1929). Their investigations led them to the orientation model shown in Fig. 2.18. The dominating orientation of the chain in the stretching direction is denoted by 1. But some crystallites were also oriented with the chain axis perpendicular to the stretching direction denoted by 2. It was not possible to discriminate between different structural features of the orientations 1 and 2. This idea of different structures of native cellulose was beyond imagination in their time. In 1943 stereochemistry had advance to a point where Hermans (1943) could carry out an investigation of an improved stereo model for the conformation and packing of cellulose I (termed cellulose Iβ today) with the available knowledge of bond lengths and bond angles. He again stressed the existence of the meridional reflections (010) and (030) (fiber axis b) and came up with a chair conformation of the anhydroglucopyranose ring for cellulose forming a straight-chain molecule with possible hydrogen bonds along the chain from O3 to O5' of the next residue and O6 to the bridge oxygen. He also clearly demonstrated that the bridge oxygen cannot lie on a 21 screw axis, which would not comply with the fiber repeat nor with a resulting bridge angle of 140° rather then the required tetrahedral one. He concluded that the two residues that form the cellobiose unit have to be bent (virtual bond O4–O1 not along the molecular axis). By a rotation of the anhydroglucose residues around the virtual bond, the bridge angle can be fixed at a suitable value. His model was not flexible enough to accommodate steric hindrance and, therefore,

38

2 History of Cellulose Research

he strongly favored antiparallel packing of the chains and found a difference in shift of about 3–4 Å of the two antiparallel-running chains. Nevertheless, a shift of the almost-parallel sheet-like corner and center chains in the a-direction by units of a resulted in an intermolecular hydrogen bonds between corner and corner chains as well as center and center chains from O6 to O2. In the 1950s Jones (e.g., Jones 1958, 1960) produced a number of papers and conducted extensive experimental studies and model building on cellulose. The device for establishing the coordinates is depicted in Fig. 2.19. He explored the stereochemistry of anhydrocellobiose taking into account IR data and X-ray intensities and discussed the controversial meridional reflections, stating that the odd meridional reflections are weak or missing according to some researchers and can be set to zero for an idealized structure. Space group discussion led him to either P21 with two anhydroglucose asymmetric units placed in the corner and in the center of the unit cell or P1 with two anhydrocellobiose units in the corner and center with no screw axes. He followed the line of Hermans (1943) in stereochemical model building without citing him and pointed towards, as Hermans did, close contacts between H1 and H4 in the vicinity of the bridge oxygen along the cellulose chain. These contacts cause restrictions in the free rotation around the bonds forming the glycosidic bridge angle (F, Y; Fig. 3.14) as well as restrictions on the placement of the residue on the 21 screw axis to maintain other contacts (e.g., hydrogen-bonding O3..O5'). The detailed structures proposed up to his time with nearly no exception had relied on

Fig. 2.19 Device for establishing the coordinates in projections for a cellulose chain. (Reproduced from Jones 1958)

2.5 Historical Development of X-ray Models for Native Cellulose

39

qualitative X-ray intensity data only, including the Meyer–Misch model. By careful evaluation of X-ray and IR data, Jones came up with a hydrogen-bonding scheme in cellulose I which withstood the test of time (intramolecular O3..O5', O6tg..O2', and intermolecular O6tg..O3) in the (020) layer (fiber axis c; tg means bond C6–O6 trans to C5–O5 and gauche to C5–C4). Jones did not check on the packing of the ribbonlike molecules in the direction perpendicular to the ribbon planes and left the question of parallel versus antiparallel packing open for further studies. This investigation was published 2 years later (Jones 1960) with a striking remark that the calculated reflections for the first layer line are exceptionally high and that no conventional P21 structure built of screw-related glucose residues can be expected to yield such weak first layer line reflections as observed in the experiment. He concluded that a statistical structure would be less in conflict with randomness of chain polarity in which adjacent chains have one characteristic shift along the c-axis. He also speculated on the larger unit cell reported for Valonia ventricosa (Table 2.2), which showed strong differences in IR spectra compared with ramie (Marrinan and Mann 1956). This larger unit cell could be formed from small cells by introducing a distinct shift in the c-direction and/or polarity between otherwise equivalent chains. A semidisordered structure based on the original cell may provide approximately the same intensities as an ordered one based on a larger repeating unit in the base plane. This idea may have influenced the structure determination later carried out by Sarko and Muggli (1974) or Gardner and Blackwell (1974), who reduced their investigation for Valonia to a two-chain unit cell instead of a larger one. Marrinan and Mann (1956) carried out IR spectroscopic investigations on cellulose polymorphs. They came to the conclusion that different interactions occur in native cellulose originating from Valonia and from plants in the OH stretching region of the spectrum (Fig. A.3). They termed the structure for Valonia A, nowadays predominantly Iα, and that for plant cellulose B, today termed Iβ. Conflicting statements have been found concerning the symmetry of the chains even for the same measurement technique and source of cellulose. An electron diffraction study by Fisher and Mann (1960), and also by Macchi (1990), clearly found no (003) reflection, which was observed by Honjo and Watanabe (1958) as well as by Hebert and Muller (1974), but all agree that the large space group of Valonia belongs to the triclinic space group P1. This is an important statement, since later space group P21 was wrongly assigned as the space group for the large eight-chain unit cell of Valonia and space group P21 was also wrongly assigned to a two-chain subcell of the large eight-chain triclinic P1 space group. The findings concerning the (003) reflection were extended to other cellulose species such as ramie and cotton (Hebert and Muller 1974) for which the generally accepted twochain unit cell was proposed but space group P21 was rejected. Nevertheless, it seems that these published measurements were not sensitive enough for a decision to be made on the chain symmetry (Fisher and Mann 1960). There were no doubts at that time as a result of IR and electron diffraction studies that two types of cellulose I structures are actually present in nature. In 1962 Ellis and Warwicker (1962) stated that no really satisfactory solution had yet been reached for the structure of native cellulose from the many investigations

40

2 History of Cellulose Research

with two-chain unit cells. They conducted an investigation themselves assuming a four-chain monoclinic unit cell comparable to the one proposed by Sponslor and Dore (1926) but with a monoclinic base plane angle to avoid the difficulties mentioned by Jones. With this four-chain unit cell they increased the degree of freedom for the cellulose chains but this did not alter too much the positions of the calculated reflections, which appeared now as overlaps of various layer planes. They adopted in their chain conformation an intramolecular hydrogen bond between O6 and O2, proposed by Mann and Marrinan (1958) from IR studies, and concluded that the structure of cellulose I can at least qualitatively best be described with parallel chain arrangements. A 21 screw axis is not a necessity either for the chains or for the unit cell. But the differences between the spectra of tunicin, ramie, etc, on one hand and of bacterial and algal cellulose (Valonia ventricosa) on the other hand remained obscure. A big step forward in the structural investigation of polysaccharides was achieved with the advent and availability of computers. Not only could the tedious model building with sticks and space-filling models and the determination of the coordinates thereof be simplified but applied refinement techniques allowed for models adjusted according to certain criteria. And the refinement procedure could also be extended to intensity analysis of X-ray data. In a preliminary step, fixed dimeric models with standard residues were used to search for possible conformations by rotating each residue around the bonds forming the glycosidic linkage (C1–O1 and O1–C4'), i.e., the torsion angles F and Y. All contacts between the two residues were assigned atom pair potentials and the sum of the contact potentials was plotted as contour lines in a so-called F, Y map (Sarko and Muggli 1974). Lines can be drawn for possible helices in such a map and for possible rise per residue along the helix axis. If low potential energy regions within such a map can be brought in agreement with experimental data as rise per residue or symmetry along the cellulose chain, these data may serve as a starting model conformation for further evaluation and refinement. Full search programs for conformation and packing with flexible rings and a choice of parameters to be varied against atom pair potentials and/or structure factors (X-ray intensities) as well as refinement towards given parameters were developed at the same time at various places. A simplex refinement technique was introduced and a program called PS79 (Zugenmaier and Sarko 1980) was established in a joint effort by the laboratories at Syracuse and Freiburg. At Purdue University a linked atom least squares refinement program (LALS; Smith and Arnott 1978) was developed for model evaluations applying the Lagrange multiplier technique for a selection of parameters and for closing the anhydroglucose ring for cellulose. The techniques as well as the models derived with these procedures for Valonia, ramie, etc. withstood the test of time with some smaller changes for the structural models as pointed out by French et al. (1987) and Finkenstadt and Millane (1998). A new era of structural investigations began with the availability of these computer-aided modeling programs, which allowed the combination of fiber diffraction and molecular modeling techniques. At the beginning computer capabilities were a limiting factor, but this disappeared over the years. Therefore, the structure

2.5 Historical Development of X-ray Models for Native Cellulose

41

evaluation for Valonia, although regarded as an eight-chain unit cell at that time and space group P1 as the sole possibility, was carried out for the smaller twochain monoclinic unit cell with space group P21 by Gardner and Blackwell (1974) and Sarko and Muggli (1974). The differences in chain arrangements of up and down chains for chiral polymers in a given monoclinic unit cell (monoclinic angle more than 90°) were recognized (Gardner and Blackwell 1974) and discussed but contradictory results in this respect were obtained as pointed out by French et al. (1987). These differences are now being resolved in a reinvestigation and agreement has been achieved as concluded by a comparison of the data and structures (Finkenstadt and Millane 1998). French et al. also stated that ramie possesses the same “parallel-up” structure, now called Iβ, as proposed for Valoni´. However, the major portion of a Valonia fiber consists of the polymorph Iα that differs in conformation and packing from Iβ. The basic unit of the Iα chain is the anhydrocellobiose with two different glucose conformations and a missing 21 screw axis along the chain. The basic units for the Iβ structure consist of one basic anhydroglucose residue along each of the two chains with a 21 screw axis. It is conceivable that the wrongly assigned 21 screw axis for the cellulose chain of Valonia and the assumed packing in a unit cell similar to the structure of Iβ led to an excellent proposal for the structure of native cellulose Iβ. The X-ray intensity data of Valonia might have been underestimated by a too high weighing of the low-energy conformation and packing considerations in the two-chain unit cell of ramie and might have resulted in an average crystalline structure of cellulose Iβ. The second minor coexisting portion of crystalline Valonia, which is actually cellulose Iβ, cannot influence the structure determination to a large extent. The two crystalline coexisting polymorphs of Valonia were considered as one crystalline form at that time and a large eight-chain unit cell was wrongly proposed, which was split up into four two-chain unit cells similar in size to the unit cell of ramie The agreement in goodness-of-fit parameters seems to clearly favor an up-chain packing arrangement (cf. Fig. 3.14) although the differences from other possibilities are small. In an electron diffraction study Claffey and Blackwell (1976) confirmed their original parallel-down arrangement from X-ray data (Gardner and Blackwell 1974), now recognized as not correct. They explicitly pointed out that no odd meridional reflections are observed. Direct observation in the electron microscope of stained fibrils of native cellulose at one end or enzymatic degraded microfibrils clearly favors the parallel packing within single fibrils. These findings are supported by an evaluation of up to the 12th-order meridional reflections of Valonia, which also points towards a parallel packing arrangement (Macchi 1990). High-resolution solid-state 13C NMR investigations (Atalla and VanderHart 1984) of a variety of native celluloses resulted in the proposal that native cellulose consists of two distinct crystalline forms derived from the fact that different ratios of the two components in various celluloses are expressed by distinct NMR peaks in the spectra. Since no changes in unit cells except for the established difference for Valonia and Chaetomorpha in contrast to ramie were observed, this proposal

42

2 History of Cellulose Research

resulted in doubts among crystallographers. A variation in crystalline structure should be mirrored in a change of the unit cell and X-ray intensities (e.g., the unit cell for cellulose I differs from that for another polymorph called cellulose III; Wada et al. 2006), which was not reported. Another explanation is also possible without the proposal of two crystalline forms being present. A crystalline structure may be capable of accommodating some distortions by building similar conformers into the unit cell. In this case the intensities also have to change and a small variation in unit-cell parameters should be observed. These predictions rely on theoretical considerations that reflection intensities are related by the Fourier transform to the electron density in a unit cell (Chap. 3). This electron density is different for mixed crystals and pure ones and changes in intensities should be observed. On the other hand, the minimum amount of steric interaction relates to optimal unit-cell parameters, which again are different for different contents of unit cells (see the discussion below). The solution of this problem of two crystal structures in native cellulose fibers was presented by Sugiyama et al. (1990) at the ACS Meeting in Boston (spring 1990) when they showed by electron diffraction that native cellulose I consists of two unit cells, a monoclinic two-chain arrangement Iβ and a triclinic form Iα. The final proof was then provided by Sugiyama et al. (1991) when they presented evidence from electron microdiffraction on single-crystal patterns that along the microfiber of Microdictyon tenuius small domains of two crystalline components of Iα and Iβ exist side by side. They also proposed that the Iβ form consists of a monoclinic unit cell belonging to space group P21 and the Iα form consists of a one-chain triclinic structure, space group P1. However, no changes in structure have been observed over the surface of Valonia crystallites despite there being a 65:35 ratio of Iα to Iβ (Baker et al. 1997). Nishiyama et al. (2002, 2003) determined the crystal and molecular structures of pure cellulose Iβ and cellulose Iα from synchrotron X-ray and neutron diffraction. For cellulose Iβ a two-chain monoclinic unit cell of space group P21 was confirmed and for cellulose Iα a one-chain triclinic unit cell of space group P1 was proposed on the basis of electron diffraction data. Both crystalline structures have parallel chain arrangements in common (Table 2.2). According to these investigations the chain distances in the two unit cells are almost the same and only the shifts of the chains in the fiber direction differ for the two polymorphs (Fig. 2.20). Therefore, the projections down the chain axes resemble each other for the two unit-cell proposals, Iα and Iβ, and lead to almost identical equatorial X-ray intensities. Owing to a shift of approximately c/2 between respective pairs of cellulose chains for Iα (Fig. 2.20), the additional reflections for the triclinic unit cell compared with the monoclinic one for Iβ on the second layer and the fourth layer are not observed (Sugiyama et al. 1991). Since these and especially the equatorial reflections dominate the X-ray fiber pattern, a discrimination between the two structures is rather difficult. Macchi (1990) on the other hand studied the meridional reflections of Valonia and found good agreement between a model calculation with parallel chains and experimental intensities with a two-chain unit cell, space group P21, which actually represents Iβ and not the predominantly present

2.5 Historical Development of X-ray Models for Native Cellulose

43

portion of Iα in Valonia. No odd-numbered meridional reflections were detected up to 12th order, which points towards a 21 screw axis along the cellulose chain also for Iα, but is not a necessary requirement for the triclinic space group P1. In contrast, an interpretation of the NMR data by Kono and Numata (2006) favors anhydrocellobiose with two different anhydroglucoses as basic units for the Iα chain, which excludes a 21 screw axis. However, the two chains of cellulose Iβ in the unit cell require for each chain one basic residue, which is compatible with 21 screw axes along the two chains. Another point should be mentioned. According to Macchi (1990) the shift of neighboring chain axes has to deviate from a quarter shift of the repeat distance c, if a meridional (002) reflection is observed. In the case of an exact 0.25c shift of the cellulose chains, the (002) reflection will vanish. Actually, the (002) reflection is missing in the triclinic structure according to Sugiyama et al. (1991) but is weak in the monoclinic Iβ structure. The coexistence of structures Iβ and Iα in native cellulose made a determination of the structures of the pure forms rather difficult for a long time. It may not have been appropriate to solve the crystal structure of the monoclinic form Iβ of cellulose (space group P21) with the data sets of Valonia, which actually provided an excellent data base on one hand but on the other hand consisted predominantly of the triclinic form Iα (space group P1) with a different molecular shift in chain direction (Fig. 2.20). Yet an excellent model for cellulose Iβ was achieved using the data set of Valonia and neglecting a few additional very weak reflections of cellulose Iα and assuming the correct size and symmetry of the unit cell of cellulose Iβ. Data from molecular mechanics calculations support the parallel-up model

Fig. 2.20 The chain arrangements of cellulose Iβ and Iα. The two distances and angles depicted represent the values for 100/010 faces. Note the different shift of the cellobiose unit across the structure for cellulose Iα compared with that for cellulose Iβ. (From Baker et al. 1997)

44

2 History of Cellulose Research

for cellulose Iβ, but since no coordinates have been published, fine details cannot be evaluated. Nevertheless, it should be emphasized that the cellulose structure determinations by Sarko and Muggli (1974) and Gardner and Blackwell (1974) assume properties which are not supported by a rigorous evaluation: A two-chain subcell of an assumed eight-chain Valonia unit cell and a 21 screw axis along the cellulose chain is not justified by space group P1 of Valonia. These assumptions seem reasonable, since the reflections on the equator of the fiber diffractogram can be indexed with a two-chain cell very similar to the two-chain unit cell of ramie. However, Nieduszynski and Atkins (1970) refer to this fact for another kind of algal cellulose (Chaetomorpha melagonium) and pointed out differences in the intensity distribution compared with ramie. Woodcock and Sarko (1980) determined the structure of ramie and explicitly found agreement with the structure of the two-chain subcell of Valonia cellulose. The 21 screw axes for corner and center chains have explicitly been confirmed for native cellulose Iβ, the major portion of crystalline ramie, by Kono and Numata (2006). Finally, cellulose still has some unsolved problems especially considering the morphology and superstructure. An explanation and experimental evidence for the long-disputed (001) and (003) reflections are missing. Some authors find these reflections by careful evaluation of the X-ray diagrams, others, and especially in electron diffraction diagrams, find that these reflections are not observed. A small distortion as a twist of the fibrils or partial occupancy etc. may cause deviations from an exact 21 screw axis of a cellulosic chain but the proposed space group may be able to accommodated some disorder. Lack of precise knowledge also exists beyond the packing arrangements in a three-dimensional lattice of native cellulose concerning the chain ends, the connections from one crystallite to the next one along the fibrils. These connections are responsible for the tensile strength and influence the lateral placement of microfibrils (crystallites) adjacent to each other to form a bundle of microfibrils or fibrils and chain folding of extremely long chains of regenerated cellulose. The two-phase model with the distribution of crystalline and noncrystalline parts in fibers is still being discussed and an explanation is needed for the observation that almost the same density occurs in both phases that is comparable to the density of extended chains. Lateral crystallite dimensions of native cellulose vary between 40 Å (cotton, dissolving pulp, etc.) and 100–200 Å (Valonia, other algal celluloses); those of regenerated cellulose are between 30 and 50 Å. The driving force for such small microfibrils is not known. The kind of disorder, the ratio of cellulose Iα and Iβ and the large number of chains at the surface influence the structure, properties and chemical behavior of cellulose fibers. Electron diffraction on a bundle of microfibrils of Valonia results in a fiber pattern with the crystallites twisted around the fiber axis and a microdiffraction of a microfibril shows a pattern of single polymeric crystals along the fiber. In contrast, single crystals of b-chitin are observed over an enormously long microfibril distance.

2.5 Historical Development of X-ray Models for Native Cellulose

45

The driving force and extensive experimental data concerning the morphological changes for the conversion of cellulose I to cellulose II need to be investigated. Certainly, ideas of interdiffusion of chains have been proposed but the mechanism, if this process is entropy-driven or enthalpy-driven, is unknown. The resulting size of crystalline domains is difficult to understand as well as the reversion of the process for which some claims are found in the literature and which contrasts with the statement of others of an irreversible transition. The changing domain size during the parallel to antiparallel transition has been addressed but still needs further confirmation and explanations. Looking back at the development of the structural research on native cellulose, it seems from today’s point of view that several proposals for certain features have been made on wrong or imprecise assumptions. The bridge oxygen should lie on the helix axes, thus fitting the repeat distance of the pitch. It was wrongly assumed that the helix axis has to be a symmetry element of the space group, thus excluding, e.g., a fivefold helix in crystal structures. The missing odd meridional reflection is only a necessary condition for space group P21 but is not a sufficient one. Kiessig (1939) proposed a two-phase structure for native cellulose, now forgotten, on the ground that a P21 unit cell cannot account for odd meridional reflections and Atalla and VanderHart (1984), who came up with the same idea, did not discriminate against the possibility that cellulose may be a mixed crystal representing a single phase. Nevertheless, during the long debates about the structure of native cellulose great progress has been achieved. We now know that native cellulose consists of two families, one with Iα-rich structure content (algal, bacterial cellulose) and one with Iβ-rich structure content (animal, higher plant cellulose, e.g., ramie), Iβ being the thermodynamically more stable one. The long-debated parallel chain arrangement has been confirmed by present-day X-ray analysis, by modeling studies and by direct evidence, and it can be expected that only the fine structure has some room for improvement. Model building proved to be an important tool to supplement X-ray fiber evaluation from the beginning of structural research. Better structural models of polymers appeared with improved data from model compounds by single-crystal analysis, first by analysis of monomers, then of oligomers. Especially the availability of data for longer oligomers is quite useful, since they already represent an important part of the polymer regarding conformation and packing, and by careful screening these data can be carried over to polymeric structures. Recent investigations on degraded and oriented cellulosics resulted in improved experimental scattering data sets from synchrotron radiation or neutron experiments with deuterated materials. High-resolution 13C solid-state NMR experiments led to valuable information concerning the characterization of polymorphs, the symmetry of chains and the number of basic residues within the unit cell. Evaluations of crystal structures from NMR data are in progress in combination with dynamic modeling procedures but these molecular structure determinations actually should include all available theoretical and experimental data. It is then expected that excellent structural models will be obtained.

46

2 History of Cellulose Research

References Ambronn H (1916a) Über das Zusammenwirken von Stäbchendoppelbrechung und Eigendoppelbrechung, I. Kolloid-Z 18:90–97 Ambronn H (1916b) Über das Zusammenwirken von Stäbchendoppelbrechung und Eigendoppelbrechung, II. Kolloid-Z 18:273–281 Ambronn H (1917) Über das Zusammenwirken von Stäbchendoppelbrechung und Eigendoppelbrechung, III. Kolloid-Z 20:73–185 Andress KR (1928) Über die Einwirkung von mässig konzentrierter Salpetersäure auf Cellulose. Z Phys Chem 136:279–288 Andress KR (1929a) Das Röntgendiagramm der nativen Cellulose. Z Phys Chem B 2:380–394 Andress KR (1929b) Das Röntgendiagramm der mercerisierten Cellulose. Z Phys Chem B 4:190–206 Atalla RH, VanderHart DL (1984) Native cellulose: a composite of two distinct crystalline forms. Science 223:283–284 Baker AA, Helbert W, Sugiyama J, Miles MJ (1997) High resolution atomic force microscopy of native Valonia cellulose I microcrystals. J Struct Biol 119:129–158 Beevers CA, Cochran W (1947) The crystal structure of sucrose sodium bromide dihydrate. Proc R Soc Lond Ser A 190:257–272 Bergmann M (1926) Allgemeine Strukturchemie der komplexen Kohlenhydrate und der Proteine. Ber Dtsch Chem Ges 59:2973–2981 Bergmann M, Knehe E (1925) I. Über ein Anhydrid der Cellobiose. Liebigs Ann Chem 445:1–17 Blackwell J (2000) Modeling ordered arrays of cellulose chains. In: Abstracts of papers of the American Chemical Society 219, Cell, part 1, San Francisco, 26 March 2000, p 217 Böeseken J (1913) Über die Lagerung der Hydroxyl-Gruppen von Polyoxy-Verblndungen im Raum. Die Konfiguration der gesättigten Glykole und der a-und b-Glykose. Ber Dtsch Chem Ges 46:2612–2628 Bragg WH (1930) New data on cellulose space lattice. Nature 125:633–634 Brant DA, Christ MD (1990) Realistic conformational modeling of carbohydrates – application and limitations in the context of carbohydrate-high polymers. In: French AD (ed) Computer modeling of carbohydrate molecules. ACS symposium series no 430. American Chemical Society, Washington, pp 42–68 Burgeni A, Kratky O (1929) Röntgenspektrographische Beobachtungen an Cellulose. V. Über das Gitter der Hydratcellulose. Z Phys Chem B 4:401–430 Charlton W, Haworth WN, Peat S (1926) A revision of the structural formula of glucose. J Chem Soc Lond 128:89–101 Chu SC, Jeffrey GA (1968) The refinement of the crystal structures of b-D-glucose and cellobiose. Acta Crystallogr Sect B 24:830–838 Claffey W, Blackwell J (1976) Electron diffraction of Valonia cellulose. A quantitative interpretation. Biopolymers 15:1903–1915 Cochran W, Crick FHC, Vand V (1952) The structure of synthetic polypeptides.1. The transform of atoms on a helix. Acta Crystallogr 5:581–586 De Wolff PM (1962) Lorentz factor for integrated intensities from azimuthal and radial diffractometer records of fiber patterns. J Polym Sci 60:S34 Ellis KC, Warwicker JO (1962) A study of the crystal structure of cellulose I. J Polym Sci 56:339–357 Finkenstadt VL, Millane RP (1998) Crystal structure of Valonia cellulose Iβ. Macromolecules 31:3776–7783 Fisher DG, Mann J (1960) Crystalline modification of cellulose. Part VI. Unit cell and molecular symmetry of cellulose I. J Polym Sci 42:189–194 Flory PJ (1974) Spatial configuration of macromolecular chains. Nobel Lecture. The Royal Swedish Academy of Sciences, Stockholm, 11 December 1974 Flory PJ (1993) Spatial configuration of macromolecular chains. In: Forsén S (ed) Nobel lectures, chemistry 1971–1980. World Scientific, Singapore, pp 156–180

References

47

Frémy E (1859) Remarques à l’occasion d’une communication de M. Payen sur les tissus des végétaux. C R Acad Sci 48:325–360 French AD, Roughead WA, Miller DP (1987) X-ray diffraction studies of Ramie cellulose I. In: Atalla RH (ed) The structures of cellulose – characterization of the solid states. ACS symposium series no 340. American Chemical Society, Washington, pp 15–37 Freudenberg K (1921) Zur Kenntnis der Cellulose. Ber Dtsch Chem Ges 54:767–773 Freudenberg K (1928) Nachtrag zu der Mitteilung über Methylcellulose. Liebigs Ann Chem 461:130–131 Freudenberg K (1929) Cellulose (9. Mitteilung über Lignin und Cellulose). Ber Dtsch Chem Ges 62:383–386 Freudenberg K (1933) Tannin, Cellulose, Lignin. Springer, Berlin Freudenberg K (1967) Von Emil Fischer zur molekularen Konstitution der Cellulose und Stärke. Ber Dtsch Chem Ges 100:422–438 Freudenberg K, Braun E (1928) Methylcellulose. Liebigs Ann Chem 460:288–304 Frey-Wyssling A (1936) Der Aufbau der pflanzlichen Zellwände. Protoplasma 25:261–300 Gardner KH, Blackwell J (1974) The structure of native cellulose. Biopolymers 13:1975–2001 Haworth WN (1925) A revision of the structural formula of glucose. Nature 116:30 Haworth WN (1927) Structural relationship in the carbohydrate group. J Soc Chem Ind 46:295–300T Haworth WN (1928) The structure of carbohydrates. Helv Chim Acta 11:534–548 Haworth WN (1929) The constitution of sugars. Edward & Arnold, London Haworth WN (1932) Die Konstitution der Kohlenhydrate. Steinkopff, Dresden Haworth WN (1937) The structure of carbohydrates and of vitamin C. Nobel Lecture. The Royal Swedish Academy of Sciences, Stockholm, 11 December 1937 Haworth WN (1966) The structure of carbohydrates and of vitamin C. In: Nobel lectures, chemistry 1922–1941. Elsevier, Amsterdam, pp 414–430 Haworth WN, Hirst EL (1921) The constitution of the disaccharides. Part V. Cellobiose (cellase). J Chem Soc 119:193–201 Hebert JJ, Muller LL (1974) An electron diffraction study on crystal structure of native cellulose. J Appl Polym Sci 18:3373–3377 Hengstenberg J (1927) Röntgenuntersuchungen über die Struktur der Polymerisationsprodukte des Formaldehyds. Ann Phys 84:245–278 Hengstenberg J, Mark H (1928) Über die Form und die Größe der Mizelle von Zellulose und Kautschuk. Z Kristallogr 69:271–284 Hermans PH (1943) Über die Gestalt und die Beweglichkeit des Moleküls der Zellulose. Kolloid-Z 102:169–180 Hermans PH (1949) Physics and chemistry of cellulose fibres. Elsevier, New York Hermans PH, Weidinger A (1946) The hydrates of cellulose. J Colloid Sci 1:185–193 Herrmann K, Gerngross O, Abitz W (1930) Zur röntgenographischen Strukturforschung des Gelatinemicells. Z Phys Chem B 10:371–394 Herzog RO (1925) Zur Erkenntnis der Cellulose Faser. Ber Dtsch Chem Ges 58:1254–1262 Herzog RO, Jancke W (1920a) Röntgenspektrographische Beobachtungen an Zellulose. Z Phys 3:196–198 Herzog RO, Jancke W (1920b) Über den physikalischen Aufbau einiger hochmolekularer organischer Verbindungen. Ber Dtsch Chem Ges 53:2162–2164 Herzog RO, Jancke W (1921) Röntgenographische Untersuchungen hochmolekularer Verbindungen. Z Angew Chem 34:385–392 Herzog RO, Krüger D (1925) Depolymerisation oder Dispergierung der Cellulose? Naturwissenschaften 13:1040–1042 Herzog RO, Krüger D (1929) Nitrocellulose diffusion experiments. J Phys Chem 33:179–189 Herzog RO, Kudar H (1934) Diffusion stäbchenförmiger Kolloide (Bestimmung der Teilchengröße durch Diffusion). Z Phys Chem A 167:343–353 Herzog RO, Jancke W, Polanyi M (1920) Röntgenspektrographische Beobachtungen an Zellulose. II. Z Phys 3:343–348

48

2 History of Cellulose Research

Hess K (1937) Problematisches bei den hochpolymeren organischen Naturstoffen. In: Zur Entwicklung der Chemie der Hochpolymeren. Verlag Chemie, Berlin, pp 158–165 Hess K, Friese K (1926) Über die Acetolyse der Cellulose (II). Liebigs Ann Chem 450:40–58 Hess K, Gundermann J (1937) Über die Einwirkung von flüssigem Ammoniak auf Cellulosefasern. Ber Dtsch Chem Ges 68:1986–1988 Hess K, Pichlmayr K (1926) Über kristallisierte Trimethylcellulose. Liebigs Ann Chem 450:29–40 Hess K, Schulze G (1926) Über das kryoskopische Verhalten kristallisierter Acetylcellulosen. Liebigs Ann Chem 448:99–120 Hess K, Trogus C (1935) Über Ammoniak-Cellulose. Ber Dtsch Chem Ges 70:1788–1799 Honjo G, Watanabe M (1958) Examination of cellulose fibre by the low-temperature specimen method of electron diffraction and electron microscopy. Nature 181:326–328 Husemann E (1947) Über Lockerstellen und ihre Spaltungsgeschwindigkeit in hydrolytisch abgebauten Ramiecellulosen. Makromol Chem 1:140–157 Irvine JC, Hirst EL (1923) The constitution of polysaccharides. Part VI. The molecular structure of cotton cellulose. J Chem Soc 123:518–532 Jacobson RA, Wunderlich JA, Lipscomb WN (1961) The crystal and molecular structure of cellobiose. Acta Crystallogr 14:598–607 Jones DW (1958) Crystalline modifications of cellulose. Part III. The derivation and preliminary study of possible crystal structures. J Polym Sci 32:371–394 Jones DW (1960) Crystalline modifications of cellulose. Part V. A crystallographic study of ordered molecular arrangements. J Polym Sci 42:173–188 Karrer P (1925) Einführung in die Chemie der polymeren Kohlenhydrate. Akademische Verlagsgesellschaft, Leipzig Karrer P, Nägeli C (1921) Polysaccharide II. Zur Konstitution der Diamylose. Helv Chim Acta 4:169–173 Kiessig H (1939) Untersuchungen über die Gitterstruktur der natürlichen Cellulose. Z Phys Chem B 43:79–102 Kono H, Numata Y (2004) Two-dimensional spin-exchange solid-state NMR study of the crystal structure of cellulose II. Polymer 45:4541–4547 Kono H, Numata Y (2006) Structural investigation of cellulose Iα and Iβ by 2D RFDR NMR spectroscopy: determination of sequence of magnetically inequivalent D-glucose units along cellulose chain. Cellulose 13:317–326 Kratky O, Porod G (1955) Gitterbestimmungen an Polysacchariden, Chitin und Kautschuk. In: Stuart HA (ed) Die Physik der Hochpolymeren, vol 3. Springer, Berlin, pp 127–134 Kuhn W (1937) Gestalt und Eigenschaften fadenförmiger Moleküle in Lösungen (und im elastisch festen Zustand). In: Zur Entwicklung der Chemie der Hochpolymeren. Verlag Chemie, Berlin, pp 180–196 Macchi EM (1990) The polarity of chain packing in native cellulose. A meridional electron diffraction analysis on Valonia fibres. Macromol Chem 191:2217–2226 Mann J, Marrinan HJ (1958) Crystalline modifications of cellulose. Part II. A study with planepolarized infrared radiation. J Polym Sci 32:357–370 Mark H (1928) Die physikalischen Grundlagen der Naegelischen Mizellartheorie. Naturwissenschaften 16:892–900 Mark H, Meyer KH (1929) Über den Bau des kristallisierten Anteils der Cellulose II. Z. Phys Chem B 2:115–145 Mark H, von Susich G (1929) Über den Bau des kristallisierten Anteils der Cellulose. III. Z Phys Chem B4:431–439 Marrinan HJ, Mann J (1956) Infrared spectra of crystalline modifications of cellulose. J Polym Sci 21:301–311 Marsh JT, Wood FC (1939) An introduction to the chemistry of cellulose. Van Nostrand, New York McDonald TRR, Beevers CA (1950) The crystal structure of a-D-glucose. Acta Crystallogr 3:394–395

References

49

McDonald TRR, Beevers CA (1952) The crystal and molecular structure of a-D-glucose. Acta Crystallogr 5:654–659 Meyer KH (1928) Neue Wege in der organischen Strukturlehre und in der Erforschung hochpolymerer Verbindungen. Z Angew Chem 41:935–946 Meyer KH (1930) Räumliche Vorstellungen über den Bau der Kohlenstoffverbindungen und ihre Verwendung in der Chemie der Hochpolymeren. Kolloid-Z 53:8–19 Meyer KH (1950) Natural and synthetic high polymers. Interscience, New York Meyer KH, Mark H (1928a) Über den Bau des krystallisierten Anteils der Cellulose. Ber Dtsch Chem Ges 61:593–614 Meyer KH, Mark H (1928b) Zur Cellulosefrage. Bemerkungen zu einer gleichbenannten Arbeit von Kurt Hess und Carl Trogus. Ber Dtsch Chem Ges 61:2432–2436 Meyer KH, Mark H (1930) Der Aufbau der hochpolymeren organischen Naturstoffe. Akademische Verlagsgesellschaft, Leipzig Meyer KH, Mark H (1931) Bemerkungen zu den Arbeiten von H Staudinger “Über die Struktur hochpolymerer Verbindungen”. Ber Dtsch Chem Ges 64:1999–2002 Meyer KH, Mark H (1940) Hochpolymere Chemie, vol II. Akademische Verlagsgesellschaft, Leipzig Meyer KH, Mark H (1950) Makromolekulare Chemie, 2nd edn. Geest & Portig, Leipzig Meyer KH, Misch L (1937) Positions des atomes dans le nouveau modèle spacial de la cellulose. Helv Chim Acta 20:232–244 Meyer KH, van der Wyk AJA (1941) Über den Feinbau der Cellulosefaser. Z Elektrochem 47:353–360 Nastukoff A (1900) Über einige Oxycellulosen und über das Molekulargewicht der Cellulose. Ber Dtsch Chem Ges 33:2237–2243 Nieduszynski IA, Atkins EDT (1970) Preliminary investigation of algal cellulose I. X-ray intensity data. Biochem Biophys Acta 222:109–118 Nishikawa S (1914) On the spectrum of X-rays obtained by means of lamellar or fibrous substances. Proc Math Phys Soc Tokyo 7:296–298 Nishikawa S, Ono S (1913) Transmission of X-rays through fibrous, lamellar and granular substances. Proc Math Phys Soc Tokyo 7:131–138 Nishiyama Y, Sugiyama J, Chanzy H, Langan P (2002) Crystal structure and hydrogen-bonding system in cellulose Iβ from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 124:9074–9082 Nishiyama Y, Chanzy H, Langan P (2003) Crystal structure and hydrogen-bonding system in cellulose Iα from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 125:14300–14306 Payen A (1838) Mémoire sur la composition du tissu propre des plantes et du ligneux. C R Acad Sci 7:1052–1056 Philipp HJ, Bjork CF (1951a) Role of the semipermeable membrane in the osmotic molecular weight determination of cellulose acetate. J Polym Sci 6:383–396 Philipp HJ, Bjork CF (1951b) Viscosity-molecular weight relationship for cellulose acetate in acetone. J Polym Sci 6:549–562 Polanyi M (1921) Die chemische Konstitution der Cellulose. Naturwissenchaften 9:288–288 Polanyi M, Weissenberg K (1922) Das Röntgen-Faserdiagramm. Z Phys 9:123–130 Priesner C (1980) H. Staudinger, H. Mark und K. H. Meyer – Thesen zur Größe und Struktur der Makromoleküle. Verlag Chemie, Weinheim Purves CB (1946) Chemical nature of cellulose and its derivatives. In: Ott E (ed) Cellulose and cellulose derivatives. High polymers, vol 5. Interscience, New York, pp 29–76, 88–112 Reeves RE (1950) The shape of pyranoside rings. J Am Chem Soc 72:1499–1506 Sarko A, Muggli R (1974) Packing analysis of carbohydrates and polysaccharides. III Valonia cellulose and cellulose II. Macromolecules 7:486–494 Scherrer P (1918) Bestimmung der Grösse und der inneren Struktur von Kolloidteilchen mittels Röntgenstrahlen. Göttinger Nachr, 98–100

50

2 History of Cellulose Research

Scherrer P (1920) Bestimmung der inneren Struktur und der Größe von Kolloidteilchen mittels Röntgenstrahlen. In: Zsigmondy R (ed) Kolloidchemie: ein Lehrbuch, 3. edn. Spamer, Leipzig Seifriz W (1929) The contractility of protoplasm. Am Nat 63:423–434 Schiebold E (1944) Beitrag zur Struktur der Zellulose, I. Kolloid-Z 108:248–265 Schulz GV (1946) Die Kinetik des Celluloseabbaus und die langperiodische Struktur des Cellulosemoleküls. J Polym Sci 3:365–370 Schulz GV Husemann E (1942) Über die Verteilung der Molekulargewichte in abgebauten Cellulosen und ein periodisches Aufbauprinzip im Cellulosemolekül. Z Phys Chem B 52:23–49 Schulze E (1891) Zur Kenntnis der chemischen Zusammensetzung der pflanzlichen Zellmembranen. Ber Dtsch Chem Ges 24:2277–2287 Smith PJC, Arnott S (1978) LALS: a linked-atom least squares reciprocal-space refinement system incorporating stereochemical restraints to supplement sparse diffraction data. Acta Crystallogr Sect A 34:3–11 Sponsler OL, Dore WH (1926) The structure of Ramie cellulose as derived from X-ray data. Colloid Symp Monogr 4:174–265 Staudinger H (1920) Über Polymersiation. Ber Dtsch Chem Ges 53:1073–1085 Staudinger H (1926) Die Chemie der organischen hochmolekularen Stoffe im Sinne der Kekuléschen Strukturlehre (Vortrag auf der Versammlung Deutscher Naturforscher und Ärzte in Düsseldorf). Ber Dtsch Chem Ges 59:3019–3043 Staudinger H (1929) Die Chemie der hochmolekularen organischen Stoffe im Sinne der Kekuléschen Strukturlehre (I und II). XII Mitteilung. Z Angew Chem 42:37–40 Staudinger H (1937) Über die makromolekulare Chemie. In: Zur Entwicklung der Chemie der Hochpolymeren. Verlag Chemie, Berlin, pp 119–157 Staudinger H (1938) Über die Zusammenhänge zwischen der Konstitution der Cellulose und ihren physikalischen Eigenschaften. Papierfabr Cellulosechem 36:1–83 Staudinger H (1953) Macromolecular chemistry. Nobel Lecture. The Royal Swedish Academy of Sciences, Stockholm, 11 December 1953 Staudinger H (1964) Macromolecular chemistry. In: Nobel lectures, chemistry 1942–1962, Elsevier, Amsterdam, pp 379–419 Staudinger H, Fritschi G (1922) Über Isopren und Kautschuk. 5. Mitteilung, Über die Hydrierung des Kautschuks und über seine Konstitution. Helv Chim Acta 5:785–806 Staudinger H, Heuer W (1930) Über hochmolekulare Verbindungen. 33. Mitteilung. Beziehungen zwischen Viskosität und Molekulargewicht bei Poly-styrolen. Ber Dtsch Chem Ges 63:222–234 Staudinger H, Leopold EO (1934) Über hochmolekulare Verbindungen, 90. Mitteilung: Über das Cellopenta-acetat und die Konstitution der Cellulose. Ber Dtsch Chem Ges 67:479–486 Staudinger H, Lüthy M (1925) Hochpolymere Verbindungen. 3. Mitteilung. Über die Konstitution der Poly-oxymethylene. Helv Chim Acta 8:41–64 Staudinger H, Signer R (1929) Über den Kristallbau hochmolekularer Verbindungen. Z Kristallogr 70:193–210 Staudinger H, Johner H, Signer R, Mie G, Hengstenberg J (1927a) Der polymere Formaldehyd, ein Modell der Zellulose. Naturwissenschaften 15:379–380 Staudinger H, Johner H, Signer R, Mie G, Hengstenberg J (1927b) Der polymere Formaldehyd, ein Modell der Zellulose. Z Phys Chem 126:425–448 Sugiyama J, Okano T, Yamamoto H, Horii F, Odani H (1990) Experimental evidence for a triclinic system in native cellulose. In: Abstracts of papers of the American Chemical Society 199, Cell, part 1, Boston, 22 April, 1990, p 31 Sugiyama J, Vuong R, Chanzy H (1991) Electron diffraction study on the two crystalline phases occurring in native cellulose from an algal cell wall. Macromolecules 24:4168–4175 Tollens B (1883) Über das Verhalten der Dextrose zu ammonalkalischer Silberlösung. 16:921–924 Tollens B (1895) Kurzes Handbuch der Kohlenhydrate, vol II. Trewendt, Breslau

References

51

Tollens B (1914) Kurzes Handbuch der Kohlenhydrate, 3rd edn. Barth, Leipzig Wada M, Nishiyama Y, Langan P (2006) X-ray structure of ammonia-cellulose I: new insights into the conversion of cellulose I to cellulose III1. Macromolecules 39:2947–2952 Watanabe S, Hayashi J, Imai K (1968) A study of cellulose trinitrate structure. J Polym Sci C 23:808–823 Watson JD, Crick FHC (1953) Molecular structure of nucleic acids – a structure for deoxyribose nucleic acid. Nature 171:737–738 Weissenberg K (1921) 1. Statistische Anisotropie in kristallinen Medien und ihre röntgenographische Bestimmung. Ann Phys 69:409–435 Willstätter R, Zechmeister L (1929) Zur Kenntnis der Hydrolyse von Cellulose. Ber Dtsch Chem Ges 62:722–725 Woodcock C, Sarko A (1980) Packing analysis of carbohydrates and polysaccharides. 11. Molecular and crystal structure of native ramie cellulose. Macromolecules 13:1183–1187 Zechmeister L, Tóth G (1931) Zur Kenntnis der Hydrolyse von Cellulose und die dabei auftretenden Zwischenprodukte. Ber Dtsch Chem Ges 64:854–870 Zechmeister L, Mark H, Tóth G (1933) Cellotriose und ihre Bedeutung für das Strukturmodell der Cellulose. Ber Dtsch Chem Ges 66:269–275 Zugenmaier P (1981) Present views of the conformation and packing of cellulose molecules. In: Robinson DG, Quader H (eds) Cell walls’ 81. Proceedings of the second cell wall meeting. Wissenschaftliche Verlagsgesellschaft, Stuttgart, pp 57–65 Zugenmaier P, Sarko A (1980) The variable virtual bond. Modeling technique for solving polymer crystal structures. In: French AD, Gardner KH (eds) Fiber diffraction methods. ACS symposium series no 141. American Chemical Society, Washington, pp 225–237

This page intentionally blank

Chapter 3

Background

3.1

Diffraction

Our knowledge of the structure of materials is predominantly influenced by imaging. The pictures taken with light show the shape and roughness of the surface. Going to smaller scales, the light and the electron microscopes using electromagnetic or particle waves provide informative pictures representing the structure. Modern techniques might also utilize other tools, such as NMR, IR and positron decay spectroscopy, to gain insights into the interior properties of materials and living tissue. The method of imaging does not work anymore and other techniques have to be employed if knowledge of the constitution of materials at the atomic level is the goal of investigations, such as the crystal structure, the placement of atoms and building defects. These techniques still rely on the scattering of electromagnetic or particle waves both of small wavelengths comparable to the size of the object. Powerful methods have been developed to study the placements of atoms in single crystals and have considerably increased our knowledge of biomacromolecules and their functioning, especially of enzymes, RNA, etc. However, these methods only work with appropriate experimental data, which are not always available, e.g., not for fibrous materials such as polysaccharides or DNA, and a different path had to be developed to solve the crystal structure of these materials. In this chapter we will briefly describe and illustrate the basic features practiced today for solving polymeric crystal structures. It should be noted that further experimental methods have also been employed for molecular structure determination, such as IR, NMR or additional spectroscopic investigations. These methods depend on interactions of atoms or orientations of groups and may serve as a useful supplement to the scattering methods, which are independent of atomic interactions. They allow the determination of positions of atoms and molecules in space rather than of correlations between atoms as do the spectroscopic methods. The diffraction method can be demonstrated and conclusions drawn by application of the so-called two-dimensional optical transform (optical Fraunhofer diffraction). Figure 3.1 represents a simple device to experimentally display the expected diffraction of models. A low-powered He–Ne laser serves as a light source. A mask M with motifs reproduced on a film is placed in the path of a P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

53

54

3 Background

Fig. 3.1 A device for producing optical transforms. (From Harburn et al. 1975)

parallel-aligned light beam. The image of the motifs appears at position I after the beam has passed the lens L4. The diffraction pattern can be collected at the focal plane F or with more intensity on a smaller scale at the focal plane F′. All the electromagnetic waves necessary for creating the image at I pass through the focal plane F. Considering this information, it is clear that a unique diffraction pattern representing the motifs can be obtained for any model by this experimental technique. A Fourier transform of the density distribution of the two-dimensional mask is the equivalent mathematical procedure, which can also be applied to a three-dimensional mask (structure). The procedure described resembles the X-ray diffraction of materials. The light passing through the focal plane F contains all the information to create the unique image at I. However, as soon as the pattern at F is experimentally processed, i.e., detection of the intensity of the light (square of the electric field of the waves) and not the unavailable electromagnetic components of the waves, some information is lost concerning the phases of the waves. This information is needed to construct the image of the motifs from a diffraction pattern. Therefore, the structure determination of materials from the diffraction pattern is rather difficult, since the phases of the electromagnetic waves from various points are needed but are lost. Only the intensities of the waves, still representing the unique structure, are accessible, which means the intensity distribution of a diffraction pattern uniquely characterizes the original structure. Therefore, the problem of solving structures by diffraction consists of the determination of the lost phases to obtain the image of the motifs and will be addressed later. The effect of diffraction from a single motif and from regularly repeating motifs in one or two dimensions as well as increasing the number of motifs is represented in Fig. 3.2. The diffraction pattern of a single circle is represented by broad concentric rings. If two or more circles as motifs are lined up in one dimension, the overall shape of the diffraction pattern of a single circle remains unchanged but will exhibit extinctions of intensities. The sharpness of the extinctions depends on the number of motifs lined up. The repeat distance of the extinct lines correlates with the distance of the motifs in the mask, which is called direct space. Optical transform also shows that the wider the distances of the motifs along a line in direct space, the smaller the corresponding distances between the diffracted spots on the

3.1 Diffraction

55

Fig. 3.2 Masks (left) and corresponding diffraction patterns by optical transforms (right). (From Harburn et al. 1975)

photographic picture, called reciprocal space. A lattice net as represented in the lower-right corner in direct space (Fig. 3.2) shows also a net of diffraction spots in reciprocal space. Such a net of points is always expected for a well-aligned threedimensional finite crystal, but then in three-dimensional space. We now turn to helical structures as proposed for many linear polymers or DNA, which can be simulated by a two-dimensional model by drawing a sinusoidal line. An X-shaped diffraction pattern is obtained as reproduced in Fig. 3.3. Placing motifs along the sinusoidal line, representing atomic units, the diffraction pattern changes somewhat (Fig. 3.4, left). The X-shaped pattern through the center is repeated at higher layer lines (Fig. 3.4, left). If a discontinuous helix contains ten motifs in the repeat, the center of the repeated X is placed on the meridian of the tenth-layer line (note the fifth layer is missing), which means a tenth meridional reflection is present. Two motifs in the repeat of the mask lead to the appearance of every second meridional reflection (Fig. 3.4, right). This observation can be generalized: the appearance of successive nth meridional reflections suggests an n-fold helix. The results of the two-dimensional diffractions can be verified theoretically for three-dimensional helices (Cochran et al. 1952). Cellulose in its native form as ramie, cotton, etc. consists of anisotropic fibrous materials and has to be investigated by diffraction, most commonly by X-rays, to gain insight into the ordered structure of the fibers. Cellulose derivatives, crystallized from solution as isotropic films, can be stretched to obtain fibers or fibrous materials or native cellulose fibers can be derivatized by a heterogeneous procedure in which the orientation is retained. The experimental set-up for obtaining

56

3 Background

Fig. 3.3 Diffraction pattern of a continuous sinusoidal helix produced by optical transform. (From Holmes and Blow 1966)

Fig. 3.4 Diffraction patterns by optical transforms of discontinuous helices with ten motifs on the repeat distance (left, a) and two motifs (right, a). Note the presence of a meridional reflection on the tenth layer line (left, b) and on every second layer line (right, b) (Left: From Holmes and Blow 1966; right: from Taylor 1969, 1972)

diffraction patterns of fibers is shown in Fig. 3.5. It is common practice to orient the fiber axis vertically. The random but oriented arrangements of elongated crystallites around the fiber axis are depicted in Fig. 3.6a with the corresponding schematic diffractogram, which is termed fiber pattern. Isotropically arranged crystallites in space lead to a so-called Debye–Scherrer pattern shown in Fig. 3.6b. In general, the crystal structure of materials is not dependent on the kind of orientation of the crystallites present.

3.1 Diffraction

57

Fig. 3.5 Experimental set-up for a diffraction experiment on fibrous materials such as native cellulose. The diffraction pattern can be split up into hyperbolic layer lines, if a flat film is used as the detection system or into straight lines for a cylindrical-shaped detection device

As demonstrated by optical transforms, a structure with a certain density distribution (blackening) leads to a unique diffraction pattern. This statement also holds for atomic dimensions, since the diffractogram can be calculated by a Fourier transformation of the electron density function representing the positions and the kinds of atoms in the structure. Actually the density distribution of an individual atom can be represented by a single atomic scattering factor (weight), which depends on the scattering angle 2Q (defined in Fig. 3.5) and is listed in tables. The intensity of a reflection is determined by taking into account the positions of all the individual atoms and their scattering power as well as the interference of the scattered waves. Moreover, only the intensity of the reflections (integral over the shape) not the shape is needed for comparison of the experimental and theoretical diffraction data for structure determination. The shape of the reflections contains further information on crystallite size, disorder and motion of atoms, which will not be discussed here (cf. Fig. 3.7). The arcing of the actual, not point-like reflections represents the orientation distribution of crystallites in a fiber. For single-crystal diffraction experiments normally a large amount of experimental data are available and computer-aided methods have been developed to deduce the actual crystal structure from the intensities of the reflections that represent the position of every single atom in the unit cell of the crystal. Single crystals used in atomic or molecular structure determination are composed of well-aligned mosaic crystallites of finite sizes. The sizes of the crystallites predominantly determine the widths of the reflections and are influenced by the perfection of the crystallites, which on the other hand influence the quality of the diffraction patterns. The description of the structure of the crystallites can be reduced to a geometrical unit, the so-called unit cell, since a crystallite consists of unit cells shifted by a simple translation in all three directions of space. This implies that the structure determination of a single crystal as well as of a crystalline fiber, which also consists of crystallites, requires the determination of the atomic positions in the unit cell only.

58

3 Background

Fig. 3.6 The distribution of crystallites and the resulting X-ray diagrams. a Uniaxially distributed crystallites around A leading to a fiber pattern. b Isotropically arranged crystallites producing a Debye–Scherrer pattern of concentric rings. (From Chandrasekaran 1997)

Fig. 3.7 Realistic representation of two-dimensional ordered domains of various sizes

3.1 Diffraction

59

Cellulosics cannot be obtained as single crystals of appropriate size, but can be obtained as fibers representing the best available order and orientation of the materials and only a few reflections are observed. They are not sufficient for developing starting models to overcome the phase problem as in the case of single crystals. An alternative procedure had to be introduced to obtain the crystal structure of polymeric fibers. The position of every atom in a molecular or crystal structure is determined by the three Cartesian coordinates x, y, z with the origin of the three-dimensional coordinate system fixed in space or alternatively in another coordinate system, e.g., the three polar coordinates by moving from one atom to the next. If we consider a macroscopic crystal, there are too many atoms to find a solution, i.e., to establish the positions of all atoms. In general, the number of atoms can be greatly reduced to the number of atoms contained in the repeating unit cell. This statement indicates the importance of knowing the shape, symmetry and size of the unit cell. Owing to symmetry restrictions of packing chiral biological compounds, e.g., an inversion symmetry element is not allowed, a selection of possible crystal lattices (space groups) is very helpful in the discussion of packing arrangements (Table 3.1). Table 3.1 List of 65 “biological” space groups Space group

Minimum symmetry

Triclinic Monoclinic

None Twofold axis parallel to b

Orthorhombic

Tetragonal

Trigonal

Hexagonal

Cubic

P1 P2, P21 C2 P222, P212121, P2221, P21212 C222, C2221 I222, I212121 F222 P4, P41, P43, P42 I4, I41 P422, P4122, P4322, P4222 P4212, P41212, P43212, P42212, I422, I4122 R3 P3, P31, P32 R32 P321, P312 P3121, P3221, P3112, P3212 P6, P61, P65 P63, P62, P64 P622, P6122, P6522 P6322, P6222, P6422 P23 P213 I23, I213 F23 P432, P4132, P4332 P4232 I432, I4132 F432, F4132

3 mutually perpendicular twofold axes

Fourfold axis parallel to c

Threefold axis parallel to c

Sixfold axis parallel to c

Threefold axes along diagonals

60

3 Background

As discussed above, the sites of the reflections on the film (reciprocal lattice) reflect the unit-cell parameters in direct space and can be determined by evaluating the positions of the reflections of the reciprocal lattice. Since six parameters (three axes and three angles between the axes) determine the shape and the size of the unit cell, the available reflections from a fiber pattern normally suffice for their determination, although difficulties may occur in obtaining a unique solution. It should be noted that, e.g., a doubling of the unit cell still leads to a correct solution of the crystal structure and for the atomic coordinates fixed within this larger unit cell as long as the actual present and imposed symmetry elements of the arrangement are not violated. However, more atomic coordinates are needed and their accuracy may be reduced owing to a poorer ratio of experimental data to variables.

3.2

Model Building

The few reflection intensities observed for a fiber X-ray pattern are not sufficient for a crystal structure determination and the data set has to be substituted by additional data of various kinds. Generally, for macromolecules the basic monomeric unit is known, which means bond lengths, bond angles and maybe torsion angles (dihedral angles) are available at least for parts of the molecule and help in reducing the number of parameters to be determined and in finding the molecular and crystal structure of this macromolecule. The bond lengths and angles are accessible by known crystal structures of model compounds in a narrow range and only the torsion angles may vary on a wider scale. Such considerations reduce the set of positional parameters for the atoms of the polymer in the unit cell considerably and suggest model building as a tool for providing insight into the polymer conformation and the subsequent necessary packing calculations. It is sometimes possible to find a few solutions for realistic structures, constrained by the experimentally determined unit cell, and to use these probable structures as starting models for testing the structure by X-ray diffraction. Model building with hard spheres or balls and sticks as the starting point for cellulose structure determination was first introduced by Sponsler and Dore (1926) as addressed in Chap. 2. This method also led to the proposal in 1953 for the double helical structure of DNA by Watson and Crick (1953). Today, refined methods are used with computer-aided procedures and energetic considerations. A flow diagram for such a procedure is given in Fig. 3.8 (Zugenmaier and Sarko 1980). The procedures and the energy functions may vary depending on developments that have taken place and authors, but the general idea is more or less the same. Here we will describe the steps which were followed in our own procedure in computer simulation and that have been applied in many laboratories. Arnott and Scott (1972) reported average bond lengths, bond angles and torsion angles for a b-glucopyranose ring, established limits in which these quantities vary and created an average b-d-glucose unit, which may serve as a starting monomeric

3.2 Model Building

61

Fig. 3.8 Flow diagram of a computer-aided procedure for the determination of polymeric crystal structures

unit of cellulose. In those early days it was overlooked that more than one stable pyranose ring may exist in the crystal structure and, therefore, all the early modeling procedures tried to fit the established average glucose ring. On the other hand some authors took a residue from a specific oligomeric crystal structure and used this fixed residue as the basic unit or the results of the structural evaluation of F, Y maps may represent a valuable starting model (Sect. 2.5; Fig. 3.14). The conformation of a single chain in a crystal is constrained by a basic unit repeat (equivalent to the fiber repeat or helix repeat) and the symmetry of the chain (helix) determined by the meridional reflections. These constraints lower the possible conformations considerably, which are obtained by model building, formally by hardsphere models. Nowadays, computer-aided calculations realistically describe the

62

3 Background

interactions between atoms with potential energy considerations and the parameters chosen can be optimized according to certain criteria. The density of the sample helps to determine the number of monomeric units in the unit cell or chain repeat in critical cases when solvent-complexed structures are considered. The possible conformations are then packed within the unit cell determined and the best model is evaluated by energy minimization of atom pair interactions by refining the appropriate parameters. The X-ray intensities are then calculated for these stereochemically reasonable models and are checked and refined towards the experimentally determined set of reflection data. The best fit of calculated and observed intensity data represents the targeted molecular and crystal structure. The goodness of fit is judged by the wellestablished crystallographic R factor. Large deviations from the geometric average values for the structural parameters may occur, and are caused by insufficient X-ray data. They can be avoided by a simultaneous analysis of conformation and packing as well as refinement against intensity data in one refinement procedure. Too few known oligomeric crystal structures represented a serious drawback in the application of the procedure described in the early days. Uncertainties were caused in the knowledge of the glycosidic bridge angle and the virtual bond length between adjacent glycosidic oxygen and further quantities essential for creating the helical structure in the procedure proposed. The following modeling technique was established. The virtual bond vector (see Fig. 3.14, e.g., for cellulose) creates the skeleton of the desired helix. The pyranose ring is then rotated around this vector until the desired bridge angle range is obtained. The rotational freedom of the primary hydroxyl groups is then fixed by packing analysis. Today the crystal structures of several longer cello-oligomeric compounds are known and the problem is not so serious anymore concerning the virtual bond length and the bridge angle as well. Care has to be taken because in some of these compounds two different conformational structures may exist for the chains in the unit cell, e.g., as found by the single-crystal structures of cellotetraose. Additional experimental data can be obtained from further diffraction techniques such as synchrotron scattering with a variation of wavelength and the advantage of extremely high primary beam intensities. In neutron scattering experiments, deuterons can be easily detected and by an exchange of hydroxyl hydrogen by deuterons these hydrogen atoms can be located. Neutron scattering provides further simplifications in the evaluation of the experimental data but is expensive. In contrast, the reflection intensities of the electron diffraction are difficult to evaluate. If so-called polymer single crystals can be grown from solution (Fig. 3.9), which nevertheless are of small size (height approximately 100–200 Å, lateral dimension in the micron range), the positions of electron diffraction spots of these single crystals represent most valuable data for a unique determination of the unit cell. In conclusion, it should be stressed that every crystal structure with a unique distribution of atoms in space leads to a unique diffraction pattern in one-, two- or threedimensional reciprocal space (detection device). This means that materials, including polymers, can be characterized by a fingerprint method as a Debye–Scherrer diagram

3.2 Model Building

63

Fig. 3.9 Electron micrograph and electron diffraction pattern (insert) of single crystals of trimethylcellulose II

or a fiber diagram without knowing the exact crystal structure. Furthermore, X-ray diagrams without detailed knowledge of the crystalline structure can be used for determination of important specific structural properties such as the crystallite sizes, different types of distortions or disorder by an investigation of the reflection line width. A crystallinity index can be established by a comparison of the intensity of the X-rays scattered into the reflections representing the crystalline part and into the background representing the noncrystalline part of cellulosic materials (reviewed in Fink 1990; Thygesen et al. 2005). Some other physical methods such as NMR and IR spectroscopy can also yield a crystallinity index, which differently weights the crystalline and noncrystalline portions of the specimens (Horii et al. 1982; Nelson and O’Connor 1964). The diffraction diagrams may serve for determination of crystal structures, if sufficient intensity data are available. Deviations from the ideal crystal structures are observed in the small-molecule regime and can also be anticipated for polymer structures. Cocrystallization in single crystals can occur with two compounds in various ways. The crystal lattice of a compound can incorporate a similar compound to a certain extent with small changes in the unit-cell size and reflection intensities but without any changes in the crystal class, which allows the localization of the compound incorporated. This means that the extinction rules caused by the symmetry elements of the space group are no longer precisely fulfilled; rather faint reflections appear at these extinction sites, e.g., faint odd meridional

64

3 Background

reflections are observed in the case of a 21 screw axis along the fiber axis. Disorder or partial occupancy of certain molecular groups may cause the same effect. The creation of pairs of molecules represents another possibility of cocrystallization, which may be especially caused by hydrogen bonding. These pairs have to be regarded as new structural units in the packing arrangement. In this case the size of the unit cell and the symmetry of the space are altered and a new crystal structure has to be discussed. Care has to be taken to ensure that the experimental reflection intensities are corrected by the appropriate procedures for polarization, absorption and Lorentz effects. Especially caution has to be observed for the Lorentz correction, which reflects the moving of the sample through the reflecting position of a crystallite in an X-ray experiment, if the single crystal is rotated or if statistically distributed crystallites in a fiber are present, which may lead to a different correction factor (Schiebold 1944). Today advanced experimental techniques in sample preparation and reflection data collection from X-ray, synchrotron and neutron scattering devices allow the application of single-crystal refinement procedures with application of constraints also for fiber diffractograms. It is sometimes difficult to judge the goodness of the resulting structures in comparison with older proposals as well as from various authors. Invariants may serve as criteria for preferring a certain crystal and molecular cellulosic structure deduced from comparable oligomeric compounds, such as a range of bond lengths, bond angles and torsion angles in similar hydrogen-bonding networks. Further invariants can be useful, such as virtual bond length, atom–atom contacts of adjacent residues and the long-known fact of an optimal hydrogenbonding network in polysaccharide structures. It should be added that model evaluations by spectroscopic means are a great advantage in securing derived models by diffraction techniques and should be incorporated in the procedure discussed. Spectroscopy will be introduced in the next section.

3.3

Spectroscopy

The molecular and crystal structure of highly crystalline materials can be obtained by diffraction experiments and the atomic position provided within a unit cell, i.e., the conformation or shape of the molecules described. In contrast to diffraction studies, spectroscopic investigations on crystalline materials rely on energy transfer to certain atoms and groups of a molecule, which are bound and influenced by force fields of the materials. The transfer of energy by electromagnetic waves of a certain frequency may then serve for an evaluation of the interactions between atoms or groups in a molecule and for structural insights into conformation and packing arrangements of a molecule. Frequency-related quantities instead of the frequency itself are generally used for a description of the measurements such as wavelength or wavenumber for IR and Raman investigations or the relative quantity of chemical shift is generally used for NMR investigations. For solid-state cellulosics, vibrational

3.3 Spectroscopy

65

IR and Raman or high-resolution solid-state cross-polarization/magic angle spinning (CP/MAS) 13C NMR experiments are appropriate methods for gathering spectroscopic data, which can be utilized for a characterization or to supplement X-ray data in a determination of polymeric structures. The chemical shift of NMR experiments describes the shielding of a nucleus by its environment and represents a structure-sensitive quantity. The vibrational frequencies of IR and Raman investigations lie in the IR wavelength region and powerful spectrometers with Fourier transform techniques are available. Raman spectroscopy can be performed in the visible wavelength region, since the difference in energy transfer from the applied wavelength as a standard is the quantity needed and can be detected with application of modern lasers and high-resolution apparatus. A change of the polarization of the material, i.e., electric dipoles, during energy transfer is required for use of IR spectroscopy and a change of polarizability, i.e., induced dipoles by an electric field, is required for Raman studies. Therefore, these two spectroscopic techniques are complementary, which means that IR absorbance bands are particularly sensitive to the vibrations of polar groups such as OH and Raman bands to the nonpolar skeletal bonds such as C–C and C–O, and both sets of information can be processed to gain structural information. Bonds that are highly polar and possess relatively high dipole moments result in intense bands in the IR spectrum and are relatively weak in the Raman spectrum and bonds that are primarily covalent with a high polarizability generally result in intense bands in the Raman spectrum and in weak bands in the IR spectrum. Structural information can be evaluated, if the modes in the vibrational spectra or the chemical shift in the NMR spectra have been assigned to certain groups, which is a prerequisite for the quantitative use of these data.

3.3.1

Optical Spectroscopy

Both IR and Raman spectroscopy of cellulosics provide information about chemical functionality, molecular conformation and hydrogen bonding and in particular the ratio can be determined of the two coexisting cellulose structures Iα and Iβ in native cellulose of various sources. For a discussion of the spectral features it is important to introduce some classes of motions and their range of appearance in the spectra. The IR bands at about 2,900 cm−1 in the spectra of cellulose represented in Fig. 3.10 can be assigned to C–H stretching vibrations and those beyond 3,000– 3,600 cm−1 to O–H stretching vibrations. Hydroxyls O–H involved in hydrogen bonding exhibit more intense bands than those which are not involved in these bonds and, furthermore, the stronger the hydrogen bond the greater the band intensity and the greater the shift to lower wavenumbers (Maréchal and Chanzy 2000). The band occurring approximately below 1,450 cm−1 can be assigned to the methylene bending vibration of H–C6–H and the deformation, wagging and twisting modes of the anhydroglucopyranose residue occur from 600 to 1,800 cm−1. Different polymorphic forms can easily be differentiated by the bands of the OH

Fig. 3.10 IR spectra of fibers and films of cellulose, not fully substituted cellulose acetate (CTA) and cellulose nitrate (CTN). Native ramie was used as cellulose fiber. Specific modes are marked. (From Temming and Grunert 1972)

66 3 Background

3.3 Spectroscopy

67

stretching region (; Marrinan and Mann 1956; Figs. A.3–A.6). These modes are also used for checking the degree of derivatization of cellulose derivatives. A trisubstituted compound should not show any OH stretching modes. The pendant groups (nitro, acetyl, etc.) introduced by substitution appear in the range from 600 to 1,800 cm−1 and may serve for identification of these groups or compounds. This kind of characterization of specific cellulosic materials by the fingerprint method is well known for IR spectroscopy as represented in Fig. 3.10. The hydrogen-bond direction in a fiber can be evaluated by application of polarized IR radiation (Mann and Marrinan 1958; Liang and Marchessault 1959a, b; Marchessault and Liang 1960; Kondo 1998). The most detailed spectra have been obtained for highly crystalline Valonia ventricosa cellulose (Blackwell et al. 1970; Walton and Blackwell 1973) and the hydrothermally derived cellulose Iβ thereof (Maréchal and Chanzy 2000; Fig. A.1). A detailed assignment for most of the IR bands above 800 cm−1 was achieved by a critical analysis of the spectra of hydrothermally treated Valonia cellulose Iβ microcrystals and is represented in Fig. 3.11 (Maréchal and Chanzy 2000). The corresponding IR spectra of nonpolarized, polarized in fiber direction c and polarized perpendicular to c are shown in Fig. A.1. Three conformations of primary C6H2–O6–H alcohols are displayed and labeled I, II and III. Conformations II and III are deduced from conformation I by rotating the C6H2–O6–H group around the C5–C6 bond. Conformation I represents at least two thirds of the total conformations, conformation II less than 10%. Hydrogen bonds are established by these primary alcohols to oxygens of other chains. The C2–O2–H alcohols establishing weak hydrogen bonds are drawn with free OH groups. This proposed conformation of a cellulose chain disagrees with the structure determined by Nishiyama et al. (2002) deduced from an

Fig. 3.11 Tentative assignment of IR bands of cellulose Iβ from heat-treated Valonia. Stretching bands are shown as two-headed arrows and bending bands as single-headed arrows. All three possible rotational positions of O6 (I, II, III) are observed with various probabilities. (From Maréchal and Chanzy 2000)

68

3 Background

excellent X-ray data base, which exhibits only one position of the primary hydroxyl group represented by conformation III and an intramolecular hydrogen bond to O2′ (the prime denotes an adjacent residue) of approximately 2.8-Å length (see also Sect. 5.3.1). Since the internal structural features of this Valonia cellulose Iβ specimen are not precisely known, the contribution of cellulose chains at the surface of the crystallites cannot be estimated, nor can that of the noncrystalline materials. It can be speculated that the two methods, IR and X-ray evaluations, for the determination of this Valonia cellulose Iβ conformation detect different and not comparable regions in the materials and lead to noncomparable results. The discrimination of the two structures of native cellulose Iα and cellulose Iβ can be achieved by a comparison of the bands in the O–H stretching region, for cellulose Iα at 3,240 cm−1 and for cellulose Iβ at 3,270 cm−1, and simultaneously in the O–H out of plane bending region, for cellulose Iα at 750 cm−1 and for cellulose Iβ at 710 cm−1, as described for Cladophora cellulose by Wada et al. (2003). This species contains predominantly the Iα form but can be converted to the Iβ form as the major part after hydrothermal treatment and the bands characteristic for cellulose Iα are considerably reduced (Fig. 5.5). The essential differences between the Iα and Iβ forms can be reduced to the described differences in IR bands of the hydroxyl groups, which are certainly involved in hydrogen bonding (Atalla 1999). The small differences in wavenumbers between cellulose Iα and cellulose Iβ suggest differences in hydrogen-bonding strengths for the two forms as found by the X-ray structure determination (Sect. 5.3.2).

3.3.2

NMR Spectroscopy

In contrast to IR, Raman and other optical spectroscopic methods, the NMR technique relies on transitions between various energy states of the nuclei induced with an additional magnetic field or pulse of matching energy, i.e., an electromagnetic wave of the corresponding frequency. The chemical shift describes the chemical environment of a nucleus and provides information on structural details of a molecule. Much chemical and structural information has been gained from isotropic chemical shift measurements, averaged over the anisotropic chemical shift by the isotropic motion of the molecules, by high-resolution NMR in solution. CP/MAS experiments extend these studies to the solid state. Generally, the NMR lines are broad and featureless in the solid state owing to various interactions and restricted mobility of the molecules, including the anisotropy of the chemical shift, but can be considerably reduced through application of special techniques. Dilution of nuclei, e.g., consideration of the 13 C nucleus of natural abundance of about 1% of all carbon nuclei in organic materials, reduces the homonuclear dipolar interaction of spins almost to zero, since interactions of the nuclei are more or less improbable. This dipolar interaction makes it difficult to obtain high-resolution NMR spectra of abundant nuclei in solid materials. The heteronuclear dipolar interaction C–H may be removed by powerful decoupling fields applied at the proton resonance frequency. The remaining line-broadening interaction is the chemical shift anisotropy, which can be reduced to an isotropic one

3.3 Spectroscopy

69

by magic-angle spinning. With the spinning of a probe at high speed in the spectrometer at a well-defined angle to the applied magnetic field, the dipolar interaction is reduced to its average and the chemical shifts are averaged to their isotropic values. Thus, the isotropic chemical shift of the solid materials and the corresponding solution data may be used for structure determination. However, the concept of magnetic dilution has an important disadvantage in that it reduces the sensitivity of the NMR signals, which is difficult to compensate even with the special pulse Fourier transform technique. Nevertheless, an enhancement of the signal of the rare spins and a very substantial increase in the signal-to-noise ratio is provided by the so-called crosspolarization technique. It is accomplished by a change in orientation and switching of a special pulse sequence. High-resolution CP/MAS 13C NMR measurements as described on polycrystalline materials yield isotropic chemical shifts, which may be used for structural and other studies. Generally, the solid-state chemical shifts are very similar to those in solution. In most cases the relation between solution and solid-state structures is relatively straightforward. One important use of the CP/MAS technique is that it provides a bridge between solid-state structures determined by diffraction techniques and those which exist in solution. In addition, it can act also as a bridge between solid materials whose structures cannot be determined and simpler model compounds, which are crystalline and whose structure can be solved by diffraction techniques. In these areas the technique is far superior in its diagnostic capabilities to other spectroscopic methods. High-resolution 13C NMR spectra can be collected from solid materials using a combination of high-power decoupling, magic-angle spinning and cross-polarization. The influence of the environment on the 13C nucleus leads to distinguishing between variously bonded carbons of the cellulosic residue and thus to a characterization of cellulose polymorphs. High-resolution CP/MAS 13C NMR spectra, quantitatively reliable, can be collected from solid cellulose materials and their derivatives and yield considerable information. Here we will discuss in some more detail the spectra of the crystal structure of cellulose II and cellotetraose, both showing similar spectra. Further cellulose spectra are represented and conclusions are drawn when discussing the crystal structures of various polymorphs. CP/MAS 13C NMR spectra of solid glucose and cellobiose do not show the same chemical shift values as the solution spectra of these compounds but nevertheless may serve as sensitive probes for structure evaluations in the solid state. As can be seen from Fig. 3.12, sufficient resolution is obtained for cellulose II, cellotetraose and cellotriose from the solid-state spectra to assign most of the resonances by comparing the chemical shifts of the same compounds in solution. The C1 and C4 resonances of cellulose II are both split into two peaks of equal intensities with similar relaxation behavior. Similar chemical shifts and multiplicity are observed in the spectrum of cellotetraose, which can then be regarded as a good model for the crystal structure of cellulose II as was predicted by Fyfe et al. (1984). The doublet of C1 and C4 of cellotetraose, and extending the previous conclusion to cellulose II, can be interpreted as two different glucose residues being present in the unit cell and their chemical shifts are a particular diagnostic of this polymorph. The 1:1 ratio

70

3 Background

Fig. 3.12 13C cross-polarization magic angle spinning NMR spectra of cello-oligomers and of cellulose II of low degree of polymerization. Signals from the reducing residues of cellotriose are labeled C-1′ and those from the nonreducing ones are labeled C-4˝. (From Fyfe et al. 1984)

of the two peaks of C1 for cellotetraose, checked for quantitative reliability, favors two inequivalent chains in the unit cell of A–A–A–A1 and B–B–B–B1 conformations with the signal of C1 from the reducing ends A1, B1 clearly separated in the spectrum. The close correspondence between the CP/MAS 13C NMR spectra of cellotetraose and cellulose II indicates that cellotetraose is a sound model for the cellulose II structure and Fyfe et al. concluded that the extrapolation of the singlecrystal structure of cellotetraose to that of cellulose II is justifiable. The molecular

3.3 Spectroscopy

71

and crystal structure of cellotetraose was later determined (Chap. 4) and served as an excellent model for the structure of cellulose II (Langan et al. (2001). It should also be noted that a quantitative 13C NMR investigation in combination with modeling procedures led to a proposal for the crystal and molecular structures for several polymorphs of cellulose and will be discussed in Chap. 5 (Sternberg et al. 2003). Horii et al. (1983) investigated the relationship between the 13C chemical shift obtained by CP/MAS 13C NMR spectra and the torsion angle c about the exo-cyclic C–C bond for different monosaccharides, oligosaccharides and cellulose. They found a simple linear relationship between the chemical shift of the CH2OH carbon and the torsion angle c about the exo-cyclic C–C bond determined by single-crystal X-ray experiments. The chemical shifts fell into three groups of 60–62.6, 62.5–64.5 and 65.5–66.5 ppm, which are related to gauche–gauche, gauche–trans, and trans–gauche conformations, respectively (Fig. 3.13, Table 3.2). On the basis of these results they predicted the rotational position of the torsion angle c of cellulose II along the two

Fig. 3.13 13C chemical shifts of the CH2OH carbons versus torsion angles c around the exo-cyclic C–C bonds determined by single-crystal X-ray experiments for various monoaccharides and oligosaccharides, most of them with pyranose rings (Horii et al. 1983). tg means C6–O6 is placed trans to C5–O5 and gauche to C5–C4. The other two possible placements of C6–O6 are gt and gg

72

3 Background

Table 3.2 Conformations of the CH2OH groups of cellulose in the solid state (Horii et al. 1983) 13 C chemical Conformation shift (ppm) CP/MAS NMR X-ray Native cellulose I Crystalline Noncrystalline Regenerated cellulose II Crystalline

66.0–66.4 63.4–63.9

trans–gauche gauche–trans

trans–gauche

63.9–64.3

gauche–trans

gauche–trans trans–gauche

Noncrystalline 62.9–63.0 CP/MAS cross-polarization/magic-angle spinning

gauche–trans

chains of the unit cell as gt, which turned out to be correct as later determined by Langan et al. (2001) with X-ray evaluation. At the time of the prediction, one chain was thought to expose O6tg, the other chain O6gt. For native cellulose I the NMR and X-ray data concerning the O6 position were in agreement. A similar linear correlation between the chemical shift of 13C carbon atoms and torsion angles was also observed for the ring carbons C1 and C4 as a function of the bridge torsion angles F and Y, respectively (Horii et al. 1987; Fig. 3.14). Highresolution CP/MAS 13C NMR spectra provide valuable data needed in many ways to complement diffraction experiments of solid-state materials. Some of them, such as symmetry elements and torsion angles, will be discussed in Chaps 5 and 6 and applied in the crystal and molecular structure determination.

3.4 Convention for the Description of Cellulosic (Chiral) Structures As pointed out by French et al. (1987) the description of native cellulose structures in the monoclinic crystalline state is confusing and can be avoided if the following conventions are used (Fig. 3.14). A monoclinic unit cell of chain molecules is uniquely defined by the unit cell vectors a and b (ab) in their original paper on native cellulose, which led to c pointing in the opposite direction to the above definition and their “up chains” are in the convention given “down chains.” The original residue labeling of the oligomeric model compounds, cellotrioside and cellotetraose, to be discussed in Sect. 4.3 starts from the nonreducing end (cf. Tables A.5–A.7). If the residue labeling is compared with the introduced definition

3.4 Convention for the Description of Cellulosic (Chiral) Structures

73

Fig. 3.14 Two cellulose up chains (or cellobiose molecules) with atom labeling. (From Zugenmaier 2001)

of up and down directions, the labeling has to be transformed. For cellotrioside this means that residue r = 3 with the reducing end has to be labeled as residue r = 1 and the original residue r = 1 with the nonreducing end is transformed to residue r = 3. Such relabeling is also necessary for cellotetraose. Here, the original residue r = 4 with the reducing end has to be transformed to residue r = 1, and the succeeding residues have to be labeled with 2, 3 and 4. The original residue numbering is preserved in Sect. 4.3.2 to facilitate the comparison of geometric data with the original literature. According to the definition introduced Figs. 4.5a and 4.6a represent down chains. The placement of adjacent anhydroglucopyranose residues is described by the following torsion angles. The torsion angle Y denotes the rotation around O1–C4′ (sometimes O1 is called O4′), and F denotes rotation around C1–O1 (or O4′) with various neighboring ring atoms or hydrogen used for an exact definition (Fig. 3.14). In the listings in the tables sometimes all these different notations are provided for the torsion angles. These two torsion angles F and Y essentially describe the relative position of succeeding residues, including the glycosidic bridge angle τ(C1–O4′–C4′) and the type of helix (Fig. 3.14, the residue is characterized by a second number attached to the label, e.g., C2 of residue 1 is characterized by C21). The placement of the primary hydroxyl oxygen O6 is described by the torsion angle c around C5–C6. Two conventions are used: c(O5–C5–C6–O6) or c ′(C4–C5–C6– O6). A simpler notation of the O6 position is generally applied by stating the placement of the O6–C6 bond in comparison with that of the adjacent bonds. If O6–C6 is placed trans to O5–C5 and gauche to C4–C5, then the position of this oxygen is called O6tg as in the structure in Fig. 3.14. The other two possible placements of O6–C6 are gt and gg.

74

3 Background

References Arnott S, Scott WE (1972) Accurate X-ray diffraction analysis of fibrous polysaccharides containing pyranose rings. Part I. The linked-atom approach. J Chem Soc Perkin Trans II 324–335 Atalla RH (1999) Celluloses. In: Pinto BM (ed) Comprehensive natural products chemistry. Carbohydrates and their derivatives including tannins, cellulose, and related lignins, vol 3. Elsevier, Amsterdam, pp 529–598 Blackwell J, Vasko PD, Koenig JL (1970) Infrared and Raman spectra of cellulose from cell wall of Valonia Ventricosa. J Appl Phys 41:4375–4379 Chandrasekaran R (1997) Molecular architecture of polysaccharide helices in oriented fibers. In: Horton D (ed) Advances in carbohydrate chemistry and biochemistry. Academic, San Diego, pp 311–439 Cochran W, Crick FHC, Vand V (1952) The structure of synthetic polypeptides. 1. The transform of atoms on a helix. Acta Crystallogr 5:581–586 Fink H-P (1990) Röntgenbeugungsuntersuchungen zur übermolekularen Struktur von Cellulose und deren Veränderung infolge einer Alkalisierung. Dissertation (B). Akademie der Wissenschaften, Teltow-Seehof French AD, Roughead WA, Miller DP (1987) X-ray diffraction studies of Ramie cellulose I. In: Atalla RH (ed) The structures of cellulose – characterization of the solid states. ACS symposium series no 340. American Chemical Society, Washington, pp 15–37 Fyfe CA, Stephenson PJ, Veregin RP, Hamer GK, Marchessault RH (1984) Insights into the lattice structure of cellulose II from the high resolution CP/MAS solid state 13C NMR spectrum of cellotetraose. J Carbohydr Chem 3:663–673 Gardner KH, Blackwell J (1974) The structure of native cellulose. Biopolymers 13:1975–2001 Harburn G, Taylor CA, Welberry TR (1975) Atlas of optical transforms. Bell, London Holmes KC, Blow DM (1966) The use of X-ray diffraction study of protein and nucleic structures. Wiley-Interscience, New York Horii F, Hirai A, Kitamaru R (1982) Solid-state high-resolution 13C-NMR study of regenerated cellulose samples with different crystallinities. Polym Bull 8:163–170 Horii F, Hirai A, Kitamaru R (1983) Solid-state 13C-NMR study of conformations of oligosaccharides and cellulose. Conformation of CH2OH group about the exo-cyclic C-C bond. Polym Bull 10:357–361 Horii F, Hirai A, Kitamaru R (1987) Cross-polarization-magic angle spinning carbon-13 NMR approach to the structural analysis of cellulose. In: Atalla RH (ed) The structures of cellulose – characterization of the solid states. ACS symposium series no 340. American Chemical Society, Washington, pp 119–134 Kondo T (1998) Hydrogen bonds in cellulose and cellulose derivatives. In: Dumitriu S (ed) Polysaccharides. Dekker, New York, pp 131–172 Langan P, Nishiyama Y, Chanzy H (2001) X-ray structure of mercerized cellulose II at 1 Å resolution. Biomacromolecules 2:410–416 Liang CY, Marchessault RH (1959a) Infrared spectra of crystalline polysaccharides. I. Hydrogen bonds in native cellulose. J Polym Sci 37:385–395 Liang CY, Marchessault RH (1959b) Infrared spectra of crystalline polysaccharides. II. Native celluloses in the region from 640 to 1700 cm−1. J Polym Sci 39:269–278 Mann J, Marrinan HJ (1958) Crystalline modifications of cellulose. Part II. A study with planepolarized infrared radiation. J Polym Sci 32:357–370 Marchessault RH, Liang CY (1960) Infrared spectra of crystalline polysaccharides. III. Mercerized cellulose. J Polym Sci 43:71–84 Maréchal Y, Chanzy H (2000) The hydrogen bond network in Iβ cellulose as observed by infrared spectrometry. J Mol Struct 523:183–196 Marrinan HJ, Mann J (1956) Infrared spectra of crystalline modifications of cellulose. J Polym Sci 21:301–311

References

75

Nelson ML, O’Connor RT (1964) Relation of certain infrared bands to cellulose crystallinity and crystal lattice type. Part II. A new infrared ratio for estimation of crystallinity in cellulose I and II. J Appl Polym Sci 8:1325–1341 Nishiyama Y, Sugiyama J, Chanzy H, Langan P (2002) Crystal structure and hydrogen-bonding system in cellulose Iβ from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 124:9074–9082 Schiebold E (1944) Beitrag zur Struktur der Zellulose, I. Kolloid-Z 108:248–265 Sponsler OL, Dore WH (1926) The structure of ramie cellulose as derived from X-ray data. Colloid Symp Monogr 4:174–265 Sternberg U, Koch F-T, Prieß W, Witter R (2003) Crystal structure refinements of cellulose polymorphs using solid state 13C chemical shifts. Cellulose 10:189–199 Taylor CA (1969) Optical methods as an aid in structure determination. Pure Appl Chem 18:533–550 Taylor CA (1972) Polymer and fibre diffraction. In: Lipson H (ed) Optical transforms. Academic, London, pp 115–151 Temming H, Grunert H (1972) Temming-Linters. Temming, Glückstadt Thygesen A, Oddershede J, Lilholt H, Thomsen AB, Stahl K (2005) On the determination of crystallinity and cellulose content in plant fibres. Cellulose 12:563–576 Wada M, Kondo T, Okano T (2003) Thermally induced crystal transformation from cellulose Iα to Iβ. Polymer J 35:155–159 Walton AG, Blackwell J (1973) Biopolymers. Academic, New York Watson JD, Crick FHC (1953) Molecular structure of nucleic acids – a structure for deoxyribose nucleic acid. Nature 171:737–738 Zugenmaier P (2001) Conformation and packing of various crystalline cellulose fibers. Prog Polym Sci 26:1341–1417 Zugenmaier P, Sarko A (1980) The variable virtual bond. Modeling technique for solving polymer crystal structures. In: French AD, Gardner KH (eds) Fiber diffraction methods. ACS symposium series no 141. American Chemical Society, Washington, pp 225–237

This page intentionally blank

Chapter 4

Model Compounds

4.1

Conformation and Packing Analysis

Experimental X-ray and spectroscopic data of cellulosics are not sufficient for a precise structure determination as accessible in single-crystal analysis of low molecular compounds. Therefore, additional data are needed and might be provided by a comparison with similar, known structures in the low molecular, oligomeric range. Such data, e.g., bond lengths, bond angles and torsion angles, can be carried over to polymeric structures or at least can be used as standards in refinement procedures. A structural data base was established in the 1970s for a standard 1-4-linked a-dglucose or b-d-glucose residue and was used in refinement of cellulosic structures at that time (Arnott and Scott 1972; cf. current data on oligomers – Raymond et al. 1995a). Very little was known about the linkage of two anhydroglucose residues, since only a few dimers had been solved by single-crystal analysis. In recent years, trimers and tetramers have been studied and a larger structural data base is now available. These oligomers can be regarded as models for cellulose, and some extracted data as invariants. Such a data base is of great value in establishing starting models for cellulosics. The trimers and tetramers also represent a longer section of the polymeric chain and their conformation or shape resembles more the cellulose chain than a dimer or monomer. The influence by surrounding molecules on the conformation in solid materials is less severe but nevertheless will affect the bond and torsion angles in oligomeric compounds to some extent, especially by hydrogen bonding. However, it should be stressed that even longer oligomers do not represent a continuous chain and possess no helical axis. Therefore, packing of oligomers may considerably differ from that of a continuous chain because of the influence of end groups, since the oligomers are connected to the next unit by van der Waals or strong dipolar forces (e.g., hydrogen bonds) instead of by chemical bonds. Looking for conformational invariants of oligomers may be safer than looking for packing invariants as can already be demonstrated by the two published, slightly different cellotetraose hemihydrate structures, which differ more in packing and intermolecular hydrogen bonding than in conformation (Gessler et al. 1994; Raymond et al. 1995b).

P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

77

78

4 Model Compounds

It should also be stressed that different procedures to complement X-ray data of crystalline cellulose have been used: (1) a refinement procedure or simulation to fit standard values (bond lengths, angles, torsion angles and further invariants) gathered as average values from many monomeric and oligomeric pyranose residues within a given range from the mean values by a weighting procedure (energy, etc.) or (2) a certain fixed geometry of a pyranose residue, which seems to fit a cellulosic structure (conformation) with certain criteria, may be a better choice. Such a proposal can be supported by spectroscopic data (NMR, etc.) or similar structural appearance. The use of cellotetraose as model for cellulose II may serve as an excellent example of selecting a model from contradictory ideas and results from various studies. The distances between the axes of the cellotetraose molecules as well as the atom–atom contacts fit quite well those of cellulose II and, therefore, the oligomeric structure was considered as model for cellulose II. Both molecules in the unit cell exhibit the same C6–O6 rotational position at all glucose residues. In contrast, an earlier refinement of cellulose II with the use of an average glucose structure led to the consideration of different rotational positions of the C6–O6 groups along two adjacent chains and the same C6–O6 rotational position as deduced from the tetramer structure was explicitly excluded. Nevertheless, a recent structural model of cellulose II primarily relying on NMR data favors the earlier proposed two rotational positions of the primary alcohol groups and not the one found in the analysis of single crystals of cellotetraose with only one rotational position (Chap. 5).

4.2

Monomers and Dimers

A broader representation of model compounds should lay a foundation for the use of structural data in the evaluation of models for cellulosics. Table 4.1 lists the average bond lengths and angles of a standard b-d-glucose unit established by Arnott and Scott (1972). In Table 4.2 further structural data of this standard b-dglucose (Arnott and Scott 1972) are compared with the refined data of a model simulation of methyl b-d-cellobioside by MM3 (92) (Rivet et al. 2001) exhibiting two low-energy conformations (1 and 2). Table 4.3 lists data for the experimentally determined single-crystal structures (packing and conformation) of methyl b-cellobioside complexed with methanol (Ham and Williams 1970; converted to d configuration) and of methyl 4-O-methyl-b-d-glucopyranosyl-(1-4)-b-dglucopyranoside (Mackie et al. 2002; in short form, dimethyl cellobioside). Methyl b-cellobioside is of special interest, since it exhibits a bifurcated hydrogen bond also found in some higher oligomeric compounds and in some cellulose polymorphs. The average C–C bond length in the pyranose ring in these structures is 1.523 Å, the average C–O bond length is 1.426 Å, except for C1–O1 of 1.389 Å, and the ring bond angles are 110.3°, except for C1–O5–C5 of 112.0°. The angles involving pendent oxygen are about 1° smaller and some small deviations are present for angles involving the primary hydroxyl group. These values can be taken as starting values for calculations of average pyranose residues in cellulosic

4.2 Monomers and Dimers

79

Table 4.1 Average bond lengths (in angstroms) and bond angles (in degrees) established by Arnott and Scott (1972) for a standard b-d-glucose unit Bond length Bond angle C1–C2 C2–C3 C3–C4 C4–C5 C5–O5 O5–C1 C6–C5 O1–C1 O2–C2 O3–C3 O4–C4 O6–C6 O1..O4

1.523 1.521 1.523 1.525 1.436 1.429 1.514 1.389 1.423 1.429 1.426 1.427 5.468

C1–C2–C3 C2–C3–C4 C3–C4–C5 C4–C5–O5 C5–O5–C1 O5–C1–C2 C6–C5–O5 O1–C1–C2 O2–C2–C1 O3–C3–C2 O4–C4–C3 O6–C6–C5 C6–C5–C4 O1–C1–O5 O2–C2–C3 O3–C3–C4 O4–C4–C5

110.5 110.3 110.2 110.2 112.3 109.3 106.9 108.4 109.3 109.6 110.4 111.8 112.7 107.3 110.8 109.7 108.6

Table 4.2 Comparison of selected geometrical data of the standard pyranose residue (b-dglucose) of Arnott and Scott (1972), the calculated methyl b-d-cellobioside of Rivet et al. (2001) (two lowest-energy conformations 1 and 2) and the experimentally determined ones of methyl 4-O-methyl-b-d-glucopyranosyl-(1-4)-b-d-glucopyranoside (Mackie et al. 2002). Bond and torsion angles in degrees, distances in angstroms; residue r. The glycosidic bridge oxygen can be labeled either O42 or O11; see Fig. 3.14 Arnott Rivet Rivet Mackie and Scott et al. (1) et al. (2) et al. Residue r

Blank

2

Ring torsion angles C1r–C2r–C3r–C4r −53.2 −52.2 C2r–C3r–C4r–C5r 52.2 50.5 C3r–C4r–C5r–O5r −56.0 −56.1 C4r–C5r–O5r–C1r 62.4 66.3 C5r–O5r–C1r–C2r −62.8 −67.2 O5r–C1r–C2r–C3r 57.5 58.8 O1r–C1r–C2r–C3r 174.2 171.2 O1r–C1r–O5r–C5r 179.9 177.6 O4r–C4r–C5r–O5r −177.0 −175.0 O4r–C4r–C3r–C2r 172.2 165.8 C1r–O5r–C5r–C6r −174.8 −173.4 C3r–C4r–C5r–C6r −175.3 −172.8 Torsion angles of the pendant C6–O6 group c ′(C4r–C5r–C6r–O6r) 61.2 −175.5 c(O5r–C5r–C6r–O6r) −60.0 66.2

1

2

1

2

1

−53.9 54.6 −58.7 65.3 −64.9 57.6 170.2 −179.2 −176.3 172.0 −174.3 −175.6

−52.6 51.2 −56.1 65.6 −66.9 58.8 171.3 177.8 −172.3 166.5 −174.1 −172.8

−54.5 54.4 −58.0 64.8 −65.4 58.7 171.3 −179.7 −175.7 171.9 −175.1 −175.1

−49.6 47.4 −52.9 64.2 −67.5 58.5 175.3 175.8 −174.2 164.7 −172.3 −171.6

−52.8 46.9 −49.4 61.6 −68.3 62.2 −179.5 −155.6 −169.7 165.1 −174.9 −169.5

−175.8 66.6

−175.0 67.0

−173.9 68.5

65.7 −54.3

−178.9 59.0

80

4 Model Compounds

Table 4.3 Selected hydrogen-bonding distances (in angstroms) between molecules of the crystal structures of methyl b-d-cellobioside–methanol (Ham and Williams 1970) and of methyl 4-Omethyl-b-d-glucopyranosyl-(1-4)-b-d-glucopyranoside (Mackie et al. 2002). (The symmetry operation for the second atom is given in parentheses) Mackie et al. (P21: a = 6.606 Å, Ham and Williams (P21: a = 7.652 Å, b = 25.532 Å, c = 4.496 Å, b = 101.84°, b = 14.074 Å, c = 9.318 Å, b = 108.95°, −193°C) ambient temperature) O41..O1Me (−x+1, y+0.5, −z+1) O62..O22 (x−1, y, z−1) O22..O21 (x+1, y, z+1) O21..O32 (x−1, y, z) O61..O31 (x+1, y, z) O62..O1Me (x, y, z)

2.696 2.715 2.748 2.741 2.648 2.700

O22..O62 (−x+1, y+0.5, −z+1) O22..O21 (−x+1, y+0.5, −z+1) O61..O31 (−x+1, y+0.5, −z+2) O31..O62 (x, y, z+1) O41..O21 (x−1, y, z)

2.776 2.807 2.769 2.673 2.805

Fig. 4.1 Constitution, conformation and the atom labeling of the lowest-energy conformation, termed 1, of methyl b-d-cellobioside (Rivet et al. 2001). Labeling of the skeleton is also correct for methyl b-d-cellobioside–methanol and methyl 4-O-methyl-b-d-glucopyranosyl-(1-4)-b-d-glucopyranoside. The glycosidic bridge oxygen can be labeled either O42 or O11; see also Fig. 3.14

chains. The fractional coordinates of selected dimers for which all bond lengths, bond angles and torsion angles can be derived are listed in Tables A.1–A.4. The constitution of methyl b-d-cellobioside and the atom labeling are shown in Fig. 4.1. Dimethyl cellobioside possesses an attached methyl group at O41 in addition. The ring torsion angles of the average b-d-glucose compare well with the calculated model of methyl b-d-cellobioside (residues 1 and 2) of the two dimeric molecules, with the exception of the torsion angles of the centered bonds C5–O5 and O5–C1 (Table 4.2). However, larger deviations are observed by a comparison

4.2 Monomers and Dimers

81

of these conformations with the experimentally determined dimethyl cellobioside of which the interaction with the surroundings by van der Waals and hydrogen bonding strongly influences the molecular structure. The differences are even more striking when comparing the torsion angles involving pendant atoms. The rotational position of the C6–O6 group differs for the two residues of dimethyl cellobioside. The relative position of the two residues in a dimeric molecules can be described by the torsion angles F and Y (Fig. 3.14). Various definitions of these two torsion angles are in use and are provided in Table 4.4. The rotational position of the C6r– O6r group (residue r, r=1 or 2) can be described by the torsion angle c'(C4r–C5r–C6r– O6r) or c(O5r–C5r–C6r–O6r). Except for the conformation of the higher-energy structure of Rivet et al. (2), the bridge torsion angles F(O51–C11–O42–C42) and Y(C11–O42–C42–C32) are similar in Table 4.4, apart from a larger deviation for Y for dimethyl b-d-cellobioside. The higher-energy conformation of Rivet et al. (2) does not represent a suitable conformation for the packing arrangement of this dimeric molecule compared with cellulose and as concluded from the deviations of the F Table 4.4 Comparison of selected geometrical data between the calculated methyl b-d-cellobioside of Rivet et al. (2001) (two lowest-energy conformations 1 and 2) and the experimentally determined ones of methyl b-d-cellobioside–methanol (Ham and Williams 1970) and of methyl 4-O-methylb-d-glucopyranosyl-(1-4)-b-d-glucopyranoside (Mackie et al. 2002). Torsion angles F and Y in degrees describing the relative positions of the two residues 1 and 2 in dimers including methyl groups (residue adjacent to residue r is denoted r′). Selected distances in angstroms and angles in degrees within the molecule Rivet Rivet Ham and Mackie et al. (1) et al. (2) Williams et al. Residue r→r′

1→2

1→2

1→2

F(H1r–C1r–O4r′–C4r′) F(O5r–C1r–O4r′–C4r′) F(C2r–C1r–O4r′–C4r′) Y(C1r–O4r′–C4r′–H4r′) Y(C1r–O4r′–C4r′–C3r′) Y(C1r–O4r′–C4r′–C5r′) O52–C12–O12–C12M C12–O12–C12M–H12M C31–C41–O41–C41M C41–O41–C41M–H41M O12..O42 O42..O41 O12..O41 O32..O51 O32..O61 H42..H11 C11–H11 t(C42–O42–C11) C12–O12–C12M C41–O41–C41M O12..O42..O41

32.3 −89.6 155.5 −42.6 80.9 −160.5 −79.7 178.7

42.1 −79.9 165.2 −10.4 111.4 −130.4 −79.4 178.7

24.7 −91.1 152.0 −47.7 80.3 −160.7 −76.1 172.7

5.459 5.448 10.343 2.705 3.119 2.282 1.109 116.6 112.4

5.448 5.445 10.386 2.769 3.461 2.240 1.121 115.1 112.4

5.470 5.463 10.144 2.763 2.914 2.065 1.045 115.8 113.2

143.0

144.9

136.2

1→2 30.1 −89.1 153.2 −33.1 86.7 −152.0 −67.8 −173.9 98.0 −178.2 5.436 5.443 10.261 2.813 3.156 2.175 0.984 117.0 113.1 116.0 141.2

82

4 Model Compounds

Fig. 4.2 Potential energy map of methyl b-D-cellobioside with isoenergy lines as a function of F(O51–C11–O42–C42) (abscissa) and Y(C11–O42–C42–C52) (ordinate). The four minima are marked with open circles and a closed circle (lowest energy), respectively. (From Rivet et al. 2001)

and Y values from the experimental ones determined for the polymeric structures. Potential energy maps, an example of which is depicted for methyl b-d-cellobioside in Fig. 4.2 shows four energy minima with four possible F and Y pairs. The structures described by the two pairs of lowest-energy minima have been evaluated and the data representing the geometry of the rings are collected in Tables 4.2 and 4.4. The virtual bond length O4..O1 (O1 also labeled O4') may be introduced as an invariant for describing structures as may the hydrogen bond O5..O3' (the prime denotes an adjacent residue) and size of the glycosidic bridge angle t (C4–O4–C1'), which all vary in narrow ranges in known crystal structures. As expected all hydroxyl groups are involved in intramolecular or intermolecular hydrogen bonding in the crystalline state. The two calculated model conformations exhibit the same valence angle C12–O12–C12M of 112.4° connecting the methyl group to the pyranose ring. This angle agrees with the two experimentally determined ones of 113.1° (Table 4.4). For dimethyl cellobioside this valence angle for the methyl group C41–O41–C41M increases to 116.0°, influenced by a rare hydrogen bond between a hydroxyl group and the bridge oxygen connecting the methyl group to the pyranose ring (here O41..O21(x-l,y,z); cf. Table 4.3). As expected the bond lengths and bond angles of the residues of the dimeric molecules lie within the range established by Arnott and Scott (1972) although they are sometimes at the extreme limit.

4.3 Trimers and Tetramers

4.3 4.3.1

83

Trimers and Tetramers Conformations

Valuable molecular data for two cellotetraoses, which differ slightly, are collected in Table 4.5 and may serve as guidelines for cellulose models. The fractional and Cartesian coordinates, which are the starting sets for all structural data evaluations, including calculations of spectroscopic features, are collected in the Appendix. The virtual bond length (O1..O4) within a monomer unit, which is an important quantity in some of the refinement procedures, exhibits values from 5.43 to 5.49 Å for b-glucopyranose units. The value for a b-(1-4)-linked disaccharide, a-N,N'diacetylchitobiose monohydrate, also lies in this range for a single residue (Mo and Jensen 1978). Narrow limits have been established by monomer and dimer analysis and a mean value of 5.45 Å is taken as standard for refinement procedures (Zugenmaier 1974). This limited standard value is confirmed by the data set for methyl bcellotrioside monohydrate 0.25 ethanolate (Raymond et al. 1995a; abbreviated in the text as cellotrioside) and collected in Table 4.6. The lengths of the dimer units O41.. O43 and also the corresponding lengths of the trimer and tetramer units of these model compounds are unexpectedly almost constant, ranging from 10.29 to 10.39 Å for the dimer unit, from 15.65 to 15.66 Å for the trimer unit and from 20.62 to 20.71 Å for the tetramer molecules. Limits can also be provided for the glycosidic bridge angle t (C4–O4–C1'), with a lower limit of 115° and an upper one of 119°. Hydrogen bonding within each oligomeric molecule between the succeeding residues O31..O52, O32.. O53, etc. is present in the tetramer as well as in the trimer, with bond lengths ranging from 2.80 to 2.90 Å. Hermans (1943) proposed such a hydrogen bond for cellulose in 1943 and in 1950 was criticized with the argument that model building cannot be used for establishing structural features (Meyer 1950). A critical distance for oligomeric and polymeric cellulosic chains is the distance between the two neighboring hydrogen atoms next to the bridge oxygen, i.e., H4..H1' of adjacent residues. The corresponding distances range from 2.03 to 2.24 Å in the model compounds; the upper limit may be reduced by 0.05 Å when the hydrogen are positioned with a more suitable standard bond length. At the end of the column for selected distances in Tables 4.5 and 4.6, one C–H distance is listed, which serves as an estimated standard length for the placement of all hydrogen atoms in C–H bonds. The relative position of two adjacent anhydropyranose units can be described by the respective torsion angles y(C1r'–O4r–C4r–C3r) and F(O5r'–C1r'–O4r–C4r) (residue r' = r+1), which are mainly responsible for intramolecular nonbonded contacts of a given conformation. They determine the conformation and to a certain extent the packing of the molecules. In some studies these angles are defined as y(C1r'–O4r–C4r–H4r) and F(H1r'–C1r'–O4r–C4r) and used as variables for plotting the potential energies in respective maps. These y and F torsion angles are provided in only one representation in Tables 4.5 and 4.6. The other two representations are collected in Zugenmaier (2001). The torsion angles listed may vary considerably along the chain but even more from molecule to molecule of the

84

4 Model Compounds

Table 4.5 Selected distances (in angstroms), bond and torsion angles (in degrees) for the model compound cellotetraose (for the labeling and coordinates of atoms see Tables A.6, A7) Raymond et al. (1995b) Gessler et al. (1995) Atoms

Molecule u

Molecule d

Intramolecular distances including hydrogen bonds O31..O52 2.864 2.805 O32..O53 2.884 2.799 O33..O54 2.905 2.796 O62..O31 3.053 3.319 O63..O32 3.104 3.314 O64..O33 3.303 3.607 O11..O41 5.488 5.450 O41..O42 5.487 5.475 O42..O43 5.485 5.487 O43..O44 5.491 5.449 O11..O42 10.339 10.325 O41..O43 10.398 10.378 O42..O44 10.353 10.382 O11..O44 20.687 20.707 H41..H12 2.181 2.098 H42..H13 2.165 2.081 H43..H14 2.095 2.133 C11-H11 0.980 Glycosidic bond angles t(C41–O41–C12) 118.6 118.9 t(C42–O42–C13) 115.0 118.0 t(C43–O43–C14) 115.9 118.4 Torsion angles C31–C41–C51–C61 −173.3 −179.3 C32–C42–C52–C62 −173.9 −177.0 C33–C43–C53–C63 −173.0 −177.9 C34–C44–C54–C64 −175.8 −170.7 c′(C41–C51–C61–O61) 169.6 −172.9 c′(C42–C52–C62–O62) 170.5 −176.9 c′(C43–C53–C63–O63) 172.2 −175.6 c′(C44–C54–C64–O64) 177.5 −162.6 Y(C12–O41–C41–C31) 82.5 96.2 Y(C13–O42–C42–C32) 86.5 94.7 Y(C14–O43–C43–C33) 89.9 96.8 F(O52–C12–O41–C41) −89.1 −93.1 F(O53–C13–O42–C42) −92.9 −91.4 F(O54–C14–O43–C43) −97.6 −93.4 Ring torsion angles C11–C21–C31–C41 −43.5 −56.6 C12–C22–C32–C42 −45.2 −54.6 C13–C23–C33–C43 −47.8 −57.1 C14–C24–C34–C44 −60.3 −53.6 C21–C31–C41–C51 45.7 57.8 C22–C32–C42–C52 47.1 55.2

Molecule a 2.799 2.819 2.826 3.306 3.271 3.660 5.467 (4.645a) 5.479 5.477 5.465 10.329 10.348 10.352 20.681 2.028 2.085 2.081

Molecule b 2.854 2.876 2.924 3.030 3.094 3.359 5.480 5.436 5.485 5.471 10.292 10.382 10.340 20.624 2.236 2.203 2.065 1.045

116.2 115.9 116.7

116.3 117.5 116.3

−178.8 −177.0 −177.5 −172.4 −171.3 −177.8 −177.3 −158.7 99.2 97.0 95.3 −93.2 −94.7 −95.3

−174.6 −173.9 −175.1 −172.1 173.0 168.7 169.9 −175.9 86.1 89.4 88.2 −91.9 −93.2 −98.3

−56.4 −57.9 −56.7 −52.9 57.3 59.0

−42.2 −46.4 −46.2 −58.9 47.3 48.1

(continued)

4.3 Trimers and Tetramers

85

Table 4.5 (continued) Raymond et al. (1995b) Atoms Molecule u C23–C33–C43–C53 48.1 C24–C34–C44–C54 54.8 C31–C41–C51–O51 −58.1 C32–C42–C52–O52 −57.3 C33–C43–C53–O53 −55.6 C34–C44–C54–O54 −56.7 C41–C51–O51–C11 69.5 C42–C52–O52–C12 68.6 C43–C53–O53–C13 66.6 C44–C54–O54–C14 66.6 C51–O51–C11–C21 −73.7 C52–O52–C12–C22 −68.3 C53–O53–C13–C23 −67.2 C54–O54–C14–C24 −68.3 O51–C11–C21–C31 56.4 O52–C12–C22–C32 54.7 O53–C13–C23–C33 55.8 O54–C14–C24–C34 64.6 a O11 partially in the α-position

Molecule d 58.1 52.5 −59.6 −58.2 −58.1 −56.1 59.8 62.4 60.3 63.1 −59.9 −63.8 −59.6 −61.8 57.4 57.0 57.5 55.8

Gessler et al. (1995) Molecule a 57.5 51.3 −60.5 −57.4 −58.2 −52.3 64.5 60.6 62.2 60.6 −64.0 −61.9 −64.4 −63.8 58.1 58.0 59.8 57.1

Molecule b 49.5 56.2 −60.0 −57.5 −59.6 −55.6 69.7 68.3 67.8 64.9 −66.4 −68.7 −65.8 −67.5 51.6 55.8 53.6 61.4

asymmetric unit. It should be pointed out that Y and F are strongly correlated with the glycosidic bridge angle t. Two distinct molecular conformations are found for cellotetraose and assigned to each of the two molecules in the unit cell as clearly expressed by the ring torsion angles (Table 4.5). For cellotrioside two pairs of molecules serve as an asymmetric (basic) unit and each molecule of a pair shows one distinct shape (Table 4.6). The origin of the two different conformations of cellotetraose or cellotrioside can be traced back to a probable bifurcated hydrogen bond along one oligomeric chain but not along the other molecule of the pair. This special hydrogen-bonding scheme does not include the reducing glucopyranose residue of cellotetraose, i.e., the fourth residue (r = 4). As deduced from Table 4.5 molecule u and molecule b of the two cellotetraose structures show the hydrogen bonds O3r..O5r' (r = 1, 2, 3, r' =r+1) along the molecule and in addition between O31..O62 (3.05 Å for molecule u and 3.03 Å for molecule b) and also between O32..O63 (3.10 Å for u and 3.09 Å for b) but the comparable distances between O33..O64 are too long for a hydrogen bond. These two shorter intramolecular distances (O31..O62; O32..O63) are absent in molecules d and a, respectively. The constraints imposed on molecule u or molecule b by the additional two hydrogen bonds lead to smaller ring torsion angles for the first three residues (r = 1, 2, 3; starting from the nonreducing end) compared with molecules d and a. But not all ring torsion angles show this effect, rather only those which include the bonds C1r–C2r, C2r–C3r and C3r–C4r as central parts. Comparing the ring torsion angles of the fourth residue with those smaller three torsion angles of the residues r = 1, 2, 3, differences occur up to 12° for similar angles. Such drastic changes are not found for the reminder of the ring torsion angles.

Table 4.6 Selected distances (in angstroms), bond and torsion angles (in degrees) for the model compound methyl b-cellotrioside (for the numbering and coordinates of atoms see Table A.5; Raymond et al. 1995a) Atoms Molecule u Molecule v Molecule d Molecule e Intramolecular distances including hydrogen bonds O31..O52 2.854 2.851 O32..O53 2.835 2.890 O31..O62 3.356 3.076 O32..O63 3.494 3.315 O11..O41 5.470 5.497 O41..O42 5.454 5.478 O42..O43 5.482 5.473 O11..O42 10.383 10.391 O41..O43 10.353 10.345 O11..O43 15.658 15.649 H41..H12 2.084 2.195 H42..H13 2.094 2.169 C11–H11 1.050 Glycosidic bond angles t(C41–O41–C12) 117.3 117.3 t(C42–O42–C13) 117.2 117.6 Torsion angles C31–C41–C51–C61 −177.3 −173.2 C32–C42–C52–C62 −176.3 −174.2 C33–C43–C53–C63 −173.6 −168.0 c′(C41–C51–C61–O61) −174.0 174.9 c′(C42–C52–C62–O62) −177.5 171.8 c′(C43–C53–C63–O63) −168.2 −179.2 Y(C12–O41–C41–C31) 96.0 86.0 Y(C13–O42–C42–C32) 97.1 87.0 F(O52–C12–O41–C41) −93.7 −91.7 F(O53–C13–O42–C42) −96.7 −95.5 Ring torsion angles C11–C21–C31–C41 −54.1 −43.3 C12–C22–C32–C42 −56.2 −47.0 C13–C23–C33–C43 −54.3 −50.3 C21–C31–C41–C51 56.0 45.3 C22–C32–C42–C52 57.9 47.5 C23–C33–C43–C53 53.0 46.6 C31–C41–C51–O51 −58.7 −57.3 C32–C42–C52–O52 −57.9 −56.5 C33–C43–C53–O53 −54.6 −51.0 C41–C51–O51–C11 62.1 69.1 C42–C52–O52–C12 60.7 67.2 C43–C53–O53–C13 60.9 63.7 C51–O51–C11–C21 −66.0 −67.3 C52–O52–C12–C22 −62.3 −65.8 C53–O53–C13–C23 −63.7 −67.8 O51–C11–C21–C31 59.7 52.7 O52–C12–C22–C32 58.3 54.2 O53–C13–C23–C33 59.0 58.7

2.833 2.900 3.076 3.313 5.487 5.485 5.471 10.391 10.345 15.643 2.149 2.162

2.811 2.841 3.314 3.487 5.465 5.460 5.473 10.383 10.353 15.650 2.046 2.086

117.3 117.7

116.8 116.6

−172.0 −173.9 −168.4 173.5 169.9 −179.5 86.1 85.7 −91.5 −96.3

−176.8 −177.6 −172.9 −172.8 −178.2 −168.4 97.5 97.8 −93.4 −96.8

−43.7 −44.8 −50.9 46.5 47.0 47.1 −57.0 −57.1 −51.9 68.8 66.3 64.4 −66.5 −64.5 −68.3 53.4 53.3 59.6

−55.4 −56.5 −53.3 55.2 58.1 52.2 −57.4 −58.8 −54.8 62.6 60.6 61.8 −65.9 −60.9 −65.1 60.5 58.0 58.3

4.3 Trimers and Tetramers

87

A similar observation holds for the ring torsion angles of molecules v and d of cellotrioside compared with molecules u and e (Table 4.6). Here only one short O31..O62 distance is detected between two of the three residues of each molecule of one pair. The differences for the above mentioned three ring torsion angles of the tetramer are not as large for molecules v and d as those for molecules u and e, but nevertheless they amount to 7–10°. The influence of the constraints discussed on all Y torsion angles of the two pairs of molecules of cellotetraose (d and a in contrast to u and b) as well as cellotrioside (v and d in contrast to u and e) appears especially severe, with differences between 7 and 14°. The differences in F can be considered as small, i.e., 1–4° (Tables 4.5, 4.6). These changes do not correlate with the glycosidic bridge angle t. In contrast, model compounds of acetyl derivatives do not exhibit such two distinct conformations in the crystalline state, since a single molecule represents the asymmetric motif in the unit cells. The conformation of the nonderivatized oligomer molecules closely resembles a 21 screw axis, i.e., the Y and F values of the 21 screw axis of a cellulose chain. These Y and F torsion angles are in reasonable agreement with the minimum pairs extracted from the calculated conformational Y, F energy maps with two adjacent fixed glucopyranose rings representing cellobiose (Sarko and Muggli 1974; Fig. 4.2). All the primary hydroxyl oxygen O6 are placed in a single site for cellotrioside and cellotetraose with similar torsion angles c'(C4–C5–C6–O6) and are termed in short O6gt, i.e., the bond O6–C6 gauche to O5–C5 and trans to C4–C5. The torsion angles deviate from their “ideal” position of 180° by approximately ±10°. The actual data are listed in Tables 4.5 and 4.6. O6gg and O6tg are the other two lowenergy positions, of which O6gg is found in one of the two residues of dimethyl cellobioside listed in Tables 4.2 and 4.3.

4.3.2

Packing Arrangements

Packing invariants are more difficult to establish. It is evident from the trimer and tetramer structures that they pack in an antiparallel fashion compensating dipole moments along the directional chains. By defining the vector from C4 to C1, these vectors point in opposite directions of appropriate adjacently placed molecules. The unit cell parameters in the base plane, i.e., the a–b plane for cellotetraose and the a–c plane for cellotrioside, resemble the base plane of cellulose II (Table 4.7). This fact suggests a similar antiparallel packing arrangement for cellulose II and supports the established packing arrangement determined by structural investigations of cellulose II proposed some time ago. Analyzing the relative position of two adjacent antiparallel molecules, the smallest stagger in the direction of the long molecular axis of two neighboring O4 amounts to 2.40 ± 0.05 Å for all of the three structures considered. The two ribbonlike planes each containing the adjacent up or down molecules lie very much parallel for all structures (Fig. 4.3). However, cellotetraose exhibits a two-molecule unit cell of space group P1 with two independent molecules. The unit

88

4 Model Compounds

Table 4.7 Unit cells of model compounds and comparison with cellulose II Unit cell Compounds Cellotetraose (Gessler et al. 1995) Cellotetraose, 2nd setting (Gessler et al. 1995) Cellotetraose (Raymond et al 1995b) Cellotrioside (Raymond et al 1995a) Cellulose II Regenerated, rayon (Kolpak and Blackwell 1976) Mercerized, cotton (Kolpak et al. 1978) Regenerated, Fortisan (Stipanovic and Sarko 1976) Mercerized, ramie (Langan et al. 2001) Regenerated, Fortisan (Langan et al. 2005)

b (Å)

c (Å)

a (°)

b (°)

8.023

8.951

22.445

89.26

85.07

63.93

8.023

8.951

22.445

89.26

94.93

116.07

8.045

9.003

22.51

89.66

94.83

115.80

7.998

76.380

90.0

116.40

90.0

8.01

9.04

10.36

90.0

90.0

117.1

8.02

8.99

10.36

90.0

90.0

116.6

7.94

9.09

10.31

90.0

90.0

117.3

8.10

9.05

10.31

90.0

90.0

117.1

8.03

9.04

10.35

90.0

90.0

117.11

a (Å)

8.9908

g (°)

cell of cellotriosode contains eight molecules belonging to space group P21 with four independent molecules. These four independent molecules are grouped in two neighboring blocks containing two pairs of molecules and each block resembles very much the placement of the cellotetraose molecules as shown in Fig. 4.3. This observation is supported by the same O4 shift of the two adjacent antiparallel molecules. The molecular axes are about 4.5 Å apart for the adjacent up and down chains in one block. This distance closely corresponds to the distance from the corner to the center of the base planes of the model compounds in the unit cell and to the equivalent distance in cellulose II. But the cellotrioside molecules of one pair in one block are twisted compared with the molecules of the adjacent block around the b-axis in the unit cell considered as shown in Fig. 4.3c and d. The projections of the structures on the a–b plane, or the a–c plane, are provided in Fig. 4.3. The base plane (a–b plane) in the original settings of the two cellotetraose structures does not provide similarity with cellulose II with corner and center chains. However, a transformation of the base plane, as indicated by dashed lines with a new b-axis, leads to a unit cell which exactly resembles the current model of cellulose II with corner and center chains (regenerated cellulose II; Kolpak and Blackwell 1976). The transformed b parameter (short diagonal of the given unit cell) and the angle or complementary angle g closely agree with the former ones (Zugenmaier 2001; also compare with the second setting, Gessler et al. 1995, in Table 4.7). The unit cell of cellotrioside contains two pairs of independent molecules with almost parallel running planes (ribbons) of each of the pairs. By separating the two adjacent blocks along b containing the two pairs, i.e., taking molecules u and d and

4.3 Trimers and Tetramers

89

Fig. 4.3 Projection of the structure of model compounds onto the base plane. a Cellotetraose (asymmetric unit of two molecules; Gessler et al. 1994), dashed lines indicate a unit cell arrangement similar to that of cellulose II. b Cellotetraose (asymmetric unit of two molecules; Raymond et al. 1995b). c Pair of molecules u and d of methyl b-cellotrioside (Raymond et al. 1995a) in one block along the b-axis. d Pair of molecules v and e of methyl b-cellotrioside (Raymond et al. 1995a) in the adjacent block (asymmetric unit of two pairs of molecules in P21). Note the different unit-cell notation as well as the placement of molecules d and u and molecules e and v within the unit cell. (From Zugenmaier 2001)

molecules v and e and drawing them in Fig. 4.3c and d, it is apparent that in Fig. 4.3d center and corner chains can be defined by a small shift of the unit cell to place one molecule exactly at the corner. In Fig. 4.3c the molecules are placed at the corner and on c/2 (note that b and c are interchanged for cellotrioside and cellotetraose). An equivalent transformation as applied in Fig. 4.3a of the base plane leads to corner and center chains in Fig. 4.3c with almost the published unit-cell parameter. Figure 4.4 reveals that in the projection down the a-axis the cellotetraose molecular axis lies parallel to c, which is not true in the projection down the b-axis, where there is a small tilt angle between these two axes. Figure 4.4b clearly expresses the sheet-like structure of identical conformations along the a-axis. These sheets containing the ribbonlike cellotetraose chains with a negligible tilt of the cellotetraose ribbons out of the plane are running parallel, one through the corner and the other through the center of the unit cell (Fig. 4.3) and may be called intrasheets in contrast to intersheets, which connect parallel intrasheets.

90

4 Model Compounds

Fig. 4.4 Projections of cellotetraose (Raymond et al. 1995b). a Asymmetric unit on the b–c plane; b the cellotetraose molecule on the a–c plane (chain shifted in the a-direction to demonstrate the sheet-like arrangement of the structure). Note that the chain axes lie parallel to c in a but are somewhat tilted towards c in b. (From Zugenmaier 2001)

The intermolecular hydrogen-bonding scheme of the model compounds is indicative of the packing arrangement of the molecules. As shown in Fig. 4.5 for cellotrioside, the interchain hydrogen bonding occurs in intrasheets and intersheets for different pairs of molecules, which are representative for the description of this strong dipolar interaction. Families of parallel intrasheets and intersheets can be defined, which rely on basic single sheets (Fig. 4.3c, d). Four unique single intrasheets are formed by the four unique pairs of molecules (dd or ee or uu or vv) as the center of the sheets. The parallel-arranged ribbonlike molecules at (x ± n, y, z), n = 0, ± 1, ± 2, etc., with (x, y, z) denoting the fractional coordinates of the molecules d or e or u or v, form the four basic single intrasheets. Adjacent molecules strongly interact edge on within each of the four sheets with hydrogen bonds (Table 4.8). A second type of sheet, called intersheets, contains pairs of specific molecules d and u [molecule u at (x+1, y, z)] or molecules v and e [molecule e at (x−1, y, z)],

4.3 Trimers and Tetramers

91

Fig. 4.5 Sheets of cellotrioside (Raymond et al. 1995a) in two planes demonstrating the intramolecular and intermolecular hydrogen-bonding scheme. a Sheet (001) of molecules ddd; b sheet (102) of molecules dud [shift of the original d-chain by (1,0,1) and (−1,0,0), respectively]

92

4 Model Compounds

Table 4.8 Intermolecular hydrogen bonding distances of methyl b-cellotrioside and cellotetraose in angstroms (for labeling and coordinates of atoms see Tables A.5–A.7). Second atom shifted in units of (a,b,c) Methyl b-cellotrioside (Raymond et al. 1995a) Parallel arrangement Atoms dd uu ee vv O61..O21 2.582(1,0,0) 2.728(−1,0,0) 2.732(1,0,0) 2.558(−1,0,0) O21..O61 2.582(−1,0,0) 2.728(1,0,0) 2.732(−1,0,0) 2.558(1,0,0) O22..O62 2.634(1,0,0) 2.669(−1,0,0) 2.675(1,0,0) 2.613(−1,0,0) O62..O22 2.634(−1,0,0) 2.669(1,0,0) 2.675(−1,0,0) 2.613(1,0,0) O63..O23 2.744(1,0,0) 2.753(−1,0,0) 2.747(1,0,0) 2.734(−1,0,0) O23..O63 2.744(−1,0,0) 2.753(1,0,0) 2.747(−1,0,0) 2.734(1,0,0) Antiparallel arrangement Atoms

du

ud

ev

ve

O22..O22 O62..O62 O21..O23 O23..O21 O61..O63 O63..O61 O61..O32 O32..O61 O31..O62 O62..O31 Cellotetraose

2.727(1,0,0) 2.683(−1,0,−1) 2.693(−1,0,−1) 2.814(−1,0,−1) 2.763(1,0,0) 2.707(1,0,0) 3.306(1,0,0) 2.971(1,0,0) 3.185(−1,0,−1) 3.217(−1,0,−1)

2.727(−1,0,0) 2.683(1,0,1) 2.814(1,0,1) 2.693(1,0,1) 2.707(−1,0,0) 2.763(−1,0,0) 2.971(−1,0,0) 3.306(−1,0,0) 3.217(1,0,1) 3.185(1,0,1)

2.725(1,0,0) 2.674(0,0,1) 2.818(0,0,1) 2.694(0,0,1) 2.705(1,0,0) 2.767(1,0,0) 2.971(1,0,0) 3.284(1,0,0) 3.238(0,0,1) 3.194(0,0,1)

2.725(−1,0,0) 2.674(0,0,−1) 2.694(0,0,−1) 2.818(0,0,−1) 2.767(−1,0,0) 2.705(−1,0,0) 3.284(−1,0,0) 2.971(−1,0,0) 3.194(0,0,−1) 3.238(0,0,−1)

Raymond et al. (1995b)

Gessler et al. (1995)

Parallel arrangement; second atom shifted by (1,0,0) Atoms

uu

dd

O21..O61 2.612 2.727 O62..O22 2.591 2.704 O23..O63 2.653 2.714 O64..O24 2.768 2.714 Antiparallel arrangement; second atom shifted by

aa

bb

2.743 2.661 2.677 2.749

2.604 2.611 2.620 2.712

(−1,0,0)

(−1,−1,0)

(−1,0,0)

(−1,1,0)

Atoms

du

ud

ab

ba

O61..O64 O61..O33 O22..O23 O63..O62 O63..O31 O24..O21

2.703 3.027 2.753 2.715 3.239 2.801

2.847 3.274 2.755 2.693 3.274 2.777

2.722 2.990 2.722 2.665 3.220 2.756

2.836 3.285 2.726 2.699 3.277 2.790

one molecule pointing down, and one molecule pointing up. These two pairs can be regarded as the center of each of the two intersheets. The pairs of molecules du positioned at [x+(−1)m2n, y, z+(−1)mn], n = 0, 1, 2, etc., m = 0 or 1, define the complete

4.3 Trimers and Tetramers

93

intersheet. The pairs ve positioned at [x-(−1)mn, y, z + (−1)mn] define the second basic intersheet. The hydrogen bonds of adjacent molecules that occur in the two intersheets are listed in Table 4.8. The dense hydrogen-bonding scheme between the terminal residues of various molecules of adjacent blocks along the chain axis b is of minor interest for our discussion and can be found in Raymond et al. (1995a). However, it is noteworthy that a hydrogen bond is directed towards the bridge oxygen connecting a methyl group of cellotrioside to the pyranose ring (Fig. 4.5a) as also occurs in the crystal structure of dimethyl cellobioside (Mackie et al. 2002). The general hydrogen-bonding scheme in cellotrioside is similar for all contacts within the sheets (dd, uu, ee, vv), with small variations in the distances between the corresponding oxygen (Table 4.8) and is represented in Fig. 4.5a as an example for the combination dd (down chains). Some differences occur in oxygen–oxygen contacts for the du chains as compared with the ud (up, down chains) chains of this intrasheet owing to different conformations of the d and u molecules (Fig. 4.5b, Table 4.8). This geometry remains unchanged for cellotetraose (interchanges in the unit-cell axis should be taken into consideration; see Fig. 4.6, Table 4.8). Hydrogen bonding occurs also between the terminal groups of the molecules placed in the adjacent unit cell in the c-direction and with water, which is located at the terminals of the chains in the interstitial spaces only. Differences between the two solved cellotetraose structures are detected, reflected in different intermolecular hydrogen-bond lengths. Since the hydroxyl hydrogens are poorly resolved, a precise determination of hydrogen bonding cannot be provided for longer oxygen–oxygen distances for which an exact hydrogen position is required.

Fig. 4.6 Sheets in two planes of cellotetraose (Gessler et al. 1995) to demonstrating the intramolecular and intermolecular hydrogen-bonding scheme. a Sheet of parallel-running molecules; b sheet of antiparallel-running molecules. Dashed lines represent possible hydrogen bonds (for atom labeling, see Table A.6)

94

4 Model Compounds

Interchain hydrogen-bonding patterns are similar for cellotrioside and cellotetraose, forming a dense network, despite the fact that the diagonal intersheets of the molecular pairs du and ve in cellotrioside are pointing in two different directions within the unit cell. Two distinct conformations between the up and down chain molecules as well as a similar length of the a-dimension of the unit cell lead to comparable oxygen– oxygen distances of adjacent molecules in cellotrioside (Fig. 4.5) and cellotetraose (Fig. 4.6) within the intersheets formed solely by the up and down chains, respectively. Specifically, hydrogen bonds occur within the intrasheets between O2..O6 and O6..O2 from the same central residues (Table 4.8). The primary hydroxyl O6 and O2 are further involved in hydrogen bonding within the intersheet between adjacent residues (Figs. 4.5, 4.6, Table 4.8). The oxygen O3 forms either an intramolecular bond to O5 of the adjacent residue along one chain or a bifurcated one to the adjacent O5 and O6, excluding the reducing residue. Actually, the two distinct conformations are defined by this intramolecular hydrogen-bonding scheme. Owing to uncertainties in positioning the hydroxyl hydrogen, only strong hydrogen bonds below 3.0 Å can be proposed with certainty. The relative shift of up chains versus down chains defined by adjacent bridge oxygens in the chain direction amounts to approximately 2.4 Å for all model compounds considered and points towards a c/4 stagger in cellulose II of the antiparallel chains.

4.4

Acetyl Derivatives

The influence of side groups on monomeric and oligomeric glucopyranose structures will be discussed for acetyl derivatives such as methyl tetraacetylb-d- glucoside (Zugenmaier and Rappenecker 1978), (1-4)-b-d-xylobiose hexaacetate (Leung and Marchessault 1973), b-d-acetyl cellobiose (Leung et al. 1976) and b-cellotriose undecaacetate (Pérez and Brisse 1977), for which selected fractional coordinates are listed in Tables A.8–A.10. Geometric data on ring conformation, the rotational position of the side groups and the relative position of the residue in oligomers as well as for cellulose triacetate I (CTA I; Sikorski et al. 2004) are listed in Tables 4.9–4.12; the crystal structure of CTA I will be discussed later. In the search for invariants for cellulose derivatives, it is observed that the influence of hydrogen bonding of nonderivatized compounds is less severe than expected on bond lengths, bond angles and especially on ring torsion angles. These geometric values for the derivatives lie in a comparable range compared with those for nonderivatized molecules but differ to some extent for various derivative compounds owing to packing arrangements. For model calculations, the acetyl side groups Ci–Oi–CiC–OiC..CiM (i is 2, 3, 6, Ci is carbon, C is carbonyl, M is methyl) connected to the pyranose ring can be placed in a plane with appropriate bond lengths and bond angles (Ci–Oi 1.445 Å, Oi–CiC 1.36 Å, CiC–OiC 1.195 Å, CiC–CiM 1.495 Å, angles Ci–Oi–CiC 112.4°, Oi–CiC–OiC 123.0°, Oi–CiC–CiM

4.4 Acetyl Derivatives

95

110.5°). The methyl hydrogens are in staggered positions. The location of the side group can be approximated by the bonds CiC = OiC cis to Ci–Hi (i is 2,3) and C6C = O6C cis to C5–H5 for O6gg but staggered for O6gt, which can be more precisely expressed by the appropriate torsion angles listed in Tables 4.11 and 4.12. Generally, the rotational site for O6 is gg, except in one residue of b-d-acetyl cellobiose and b-cellotriose undecaacetate for which O6gt was determined. The glycosidic bridge angles t are comparable with the nonderivatized compounds but sometimes the F and Y torsion angles describing the relative position of adjacent residues deviate considerably and favor a different kind of helix, such as left-handed 5/3 or 8/5. Only the F and Y torsion angles for xylobiose hexaacetate and the first two residues of b-cellotriose undecaacetate are quite similar to the nonderivatized model compounds and exhibit values that are compatible with a 2/1 helix as expressed by the F and Y values listed for CTA I (2/1 helix; Table 4.12). In both residues 1 and 2 of acetyl cellotriose, the rotational positions of O6 are gg and for the nonreducing residue 3 is gt. Therefore, the dimer formed by residues 1 and 2 of cellotriose undecaacetate or xylobiose hexaacetate is suitable to create a preliminary conformational model for the cellulose triacetate II (CTA II) polymorph for which the asymmetric unit consists of a dimer. Nevertheless, the conformation of this dimeric structure has to be constrained to the experimental fiber repeat and to appropriate geometrical data for representing a model for CTA II. For the CTA I polymorph only one residue is needed as an asymmetric unit and the residue has to be constrained to form a 21 screw axis along the chain within the given fiber repeat and to a geometry comparable to that of model molecules. The best model compound for CTA I seems to be b-d-xylobiose hexaacetate according to a comparison and overall evaluation of structural data, despite the fact that the side group at C5 is missing in this molecule (Tables 4.9–4.12).

Table 4.9 Selected torsion angles of the basic residues r of acetylated derivatives Tetraacetyl-b-db-d-Xylobiose b-d-Acetyl glucoside hexaacetate cellobiose Residue r Blank 1 2 1 2 C1r–C2r–C3r–C4r −52.2 −55.1 −48.5 C2r–C3r–C4r–C5r 49.3 54.1 48.0 C3r–C4r–C5r–O5r −53.1 −57.5 −54.4 C4r–C5r–O5r–C1r 64.2 62.7 62.9 C5r–O5r–C1r–C2r −67.8 −63.8 −65.6 O5r–C1r–C2r–C3r 59.4 60.3 57.7 C3r–C4r–C5r–C6r −172.8 C4r–C5r–C6r–O6r 56.1 O5r–C5r–C6r–O6r −64.1 O4r–C4r–C3r–C2r 168.2 174.6 166.2 O4r–C4r–C5r–O5r −170.7 −175.6 −176.7 O1r–C1r–C2r–C3r 176.0 177.5a 173.8 O1r–C1r–O5r–C5r 173.9 178.4a 177.0 a The glycosidic bridge oxygen can be labeled either O42 or O11

−49.0 48.9 −57.1 67.8 −70.1 59.0 −177.5 48.6 −70.9 164.4 −175.2 176.3a 173.5a

−46.3 46.9 −56.5 67.8 −69.9 56.8 −170.8 155.6 40.9 166.5 −174.2 169.3 175.7

96

4 Model Compounds

Table 4.10 Selected torsion angles of the basic residues r of acetylated derivatives

Residue r

1

Cellulose triacetate I 1

b-Cellotriose undecaacetatea 2 3

C1r–C2r–C3r–C4r −53.8 −49.4 −50.5 C2r–C3r–C4r–C5r 58.7 52.0 53.2 C3r–C4r–C5r–O5r −65.5 −61.1 −60.7 C4r–C5r–O5r–C1r 67.5 73.4 65.2 C5r–O5r–C1r–C2r −62.4 −70.3 −61.0 O5r-C1r-C2r-C3r 52.1 56.8 53.4 C3r–C4r–C5r–C6r 177.8 −178.7 −176.2 C4r–C5r–C6r–O6r 63.9 51.5 175.8 O5r–C5r–C6r–O6r −52.2 −64.4 59.7 O4r–C4r–C3r–C2r 171.3 164.2 173.0 O4r–C4r–C5r–O5r −178.7 −177.4 −178.2 Residue r→r′ 1→2 2→3 3→4 O4r′–C1r–C2r–C3r 166.0 174.8 168.6 O4r′–C1r–O5r–C5r −177.9 173.0 179.4 a The adjacent glycosidic bridge oxygen can be labeled either O44 or O13

−49.6 48.0 −55.6 65.5 −65.7 57.6 −174.0 57.9 −62.5 163.4 −174.1 1→3 170.6 178.7

Table 4.11 Selected torsion angles (in degrees) of oligomeric glucose acetates to describe the rotational placements of C6–O6 and carbonyl side-groups as well as of adjacent residues by F and Y. For comparison between the compounds certain distances (in angstroms) and glycosidic bridge angles (in degrees) are included. Atoms of adjacent residues are denoted by r′ = r + 1 Tetraacetyl-b-Db-d-Xylobiose b-d-Acetyl glucoside hexaacetate cellobiose Residue r Blank 1 2 1 2 Torsion angle C1r–C2r–O2r–C2rC 108.7 C2r–C3r–O3r–C3rC −122.6 C3r–C4r–O4r–C4rC 127.5 C5r–C6r–O6r–C6rC −153.9 O5r–C1r–O1r–C1rC c′(C4r–C5r–C6r–O6r) 56.1 O6r..O4r 3.58 O3r..O4r 2.90 C1r–H1r 1.02 Residue r→r′ Y(C1r–O4r′–C4r′–C3r′) F(O5r–C1r–O4r′-C4r′) t(C1r–O4r′–C4r′) O5r..O3r′ O2r..O6r′ O4r..O4r′ 5.45 (O1) O4r′..O1r′ O4r..O1r′ H1r..H4r′

118.3 −121.1 135.0

136.5 −105.1

−89.8

2.80 0.96

2.97 1.07 1→2 89.5 −104.7 117.9 3.03 5.45 5.42 10.47 2.15

114.3 −133.7 107.1 −141.5

111.5 −143.0

76.9 −85.0 48.6 155.6 3.39 3.00 3.01 2.91 1.00 1.35 1→2 134.3 −77.7 116.8 3.32 5.10 5.41 5.46 10.30 2.57

4.4 Acetyl Derivatives

97

Table 4.12 Selected torsion angles (in degrees) of b-cellotriose undecaacetate and cellulose triacetate I to describe the rotational placements of C6–O6 and carbonyl side-groups as well as of adjacent residues by F and Y. For comparison between the compounds certain distances (in angstroms) and glycosidic bridge angles (in degrees) are included. Atoms of adjacent residues are denoted by r′ = r + 1 for b-cellotriose undecaacetate and r′ = r + 2 for cellulose triacetate I β-Cellotriose undecaacetate Cellulose triacetate I Residue r 1 2 3 1 Torsion angle C1r–C2r–O2r–C2rC C2r–C3r–O3r–C3rC C3r–C4r–O4r–C4rC C5r–C6r–O6r–C6rC O5r–C1r–O1r–C1rC c′(C4r–C5r-C6r-O6r) O6r..O4r O3r..O4r C1r–H1r Residue r→r′ Y(C1r–O4r′–C4r′–C3r′) F(O5r–C1r–O4r′–C4r′) t(C1r–O4r′–C4r′) O5r..O3r′ O2r..O6r′ O4r..O4r′ O4r′..O1r′ O4r..O4(r+2) O4r..O4(r′+2) O4r..O1(r+2) H1r..H4r′

141.0 −119.0 111.8 160.1

129.9 −104.8

116.0 −131.9

168.0

63.9 3.59 2.89 1.05

88.2 −96.1 175.8 2.90 2.90 1.09

51.5 3.39 3.03 1.05 1→2 101.3 −97.8 116.9 3.04 3.56 5.44 10.57

2→3 133.9 −75.1 115.6 3.27 4.77 5.44 5.44 10.15

145.3 −101.2 165.6 57.9 3.52 3.01 1.10 1→3 95.6 −99.1 118.8 3.01 3.74 5.42

10.46 15.56 2.14

2.34

2.18

The unit cell of cellotriose undecaacetate, space group P21, contains one molecule as an asymmetric unit with two quite different pairs of F and Y torsion angles, two positions of O6 in gg and the third one in gt but with similar glycosidic bridge angles (Fig. 4.7). In contrast to this conformation, both basic molecules of cellotrioside, the underivatized trimer, show a more uniform shape with two pairs of similar F and Y torsion angles and a uniform placement of O6 in gt along the molecular axis. An unusual rotational position of O6 is detected for residue 2 of b-d-acetyl cellobiose, with c '= 155.6°, far from the ideal position of 180° (Table 4.11). Since also the bond lengths and angles deviate in this region of the molecule from standard values, it is likely that disorder causes this deviation that is not taken into consideration during structure refinement. An antiparallel packing arrangement can only be observed for the acetylated monomers and dimers, which are packed in appropriate space groups, e.g., P212121, and then one molecule may be needed as the basic unit. The structure of the acetylated trimer, cellotriose undecaacetate, also requires one molecule as an asymmetric unit in the polar space group P21 but with parallel-arranged molecules

98

4 Model Compounds

Fig. 4.7 b-Cellotriose undecaacetate projected on the b-c plane. Note the parallel chain packing in space group P21

(Fig. 4.7). In contrast, the nonderivatized trimeric and tetrameric compounds are packed in antiparallel fashion with pairs of antiparallel-arranged molecules in P21 (trimer) or P1 (tetramer). It is also remarkable that some O5..O3′ atoms are only little more than 3 Å apart (Tables 4.11, 4.12). Such distances still fall within the range of hydrogen bonding of the nonderivatized compounds. This short distance seems to be a prerequisite for a 2/1 helix of the cellulosic chain as does a short H1.. H4′ distance of about 2.15 Å between acetylated pyranose rings.

References Arnott S, Scott WE (1972) Accurate X-ray diffraction analysis of fibrous polysaccharides containing pyranose rings. Part I. The linked-atom approach. J Chem Soc Perkin Trans II 324–335 Gessler K, Krauß N, Steiner T, Betzel C, Sandmann C, Sänger W (1994) Crystal structure of b-d-cellotetraose hemihydrate with implications for the structure of cellulose II. Science 266:1027–1029

References

99

Gessler K, Krauß N, Steiner T, Betzel C, Sarko A, Sänger W (1995) b-d-Cellotetraose hemihydrate as a structural model for cellulose II. An X-ray diffraction study. J Am Chem Soc 117:11397–11406 Ham JT, Williams DG (1970) The crystal and molecular structure of methyl b-cellobiosidemethanol. Acta Crystallogr Sect B 26:1373–1383 Hermans PH (1943) Über die Gestalt und die Beweglichkeit des Moleküls der Zellulose. Kolloid-Z 102:169–180 Kolpak FJ, Blackwell J (1976) Determination of the structure of cellulose II. Macromolecules 9:273–278 Kolpak FJ, Weih M, Blackwell J (1978) Mercerization of cellulose: 1. Determination of the structure of mercerized cotton. Polymer 19:123–131 Langan P, Nishiyama Y, Chanzy H (2001) X-ray structure of mercerized cellulose II at 1 Å resolution. Biomacromolecules 2:410–416 Langan P, Sukumar N, Nishiyama Y, Chanzy H (2005) Synchrotron X-ray structures of cellulose Iβ and regenerated cellulose II at ambient temperature and 100 K. Cellulose 12:551–562 Leung F, Marchessault RH (1973) Crystal structure of b-d, 1→4 xylobiose hexaacetate. Can J Chem 51:1215–1222 Leung F, Chancy HD, Pérez S, Marchessault RH (1976) Crystal structure of b-d-acetyl cellobiose, C28H38O19. Can J Chem 54:1365–1371 Mackie ID, Röhrling J, Gould RO, Pauli J, Jäger C, Walkinshaw M, Potthast A, Rosenau T, Kosma P (2002) Crystal and molecular structure of methyl 4-O-methyl-b-d-glucopyranosyl-(1-4)b-d-glucopyranoside. Carbohydr Res 337:161–166 Meyer KH (1950) Natural and synthetic high polymers, vol IV. Interscience, New York Mo F, Jensen LH (1978) The crystal structure of a b-(1→4) linked disaccharide, α-N,N'diacetylchitobiose monohydrate. Acta Crystallogr Sect B 34:1562–1569 Pérez S, Brisse F (1977) The crystal and molecular structure of a trisaccharide, b-cellotriose undecaacetate: 1,2,3,6-tetra-O-acetyl-4-O-[2,3,6-tri-O-actyl-4-O-(2,3,4,6-tetra-O-acetylb-d-glucopyranosyl)-b-d-glucopyranosyl]-b-d-glucopyranose. Acta Crystallogr Sect B 33:2578–2584 Raymond S, Henrissat B, Tran Qui D, Kvick Å, Chanzy H (1995a) The crystal structure of methyl b-cellotrioside monohydrate 0.25 ethanolate and its relationship to cellulose II. Carbohydr Res 277:209–229 Raymond S, Heyraud A, Tran Qui D, Kvick Å, Chanzy H (1995b) Crystal and molecular structure of b-d-cellotetraose hemihydrate as a model of cellulose II. Macromolecules 28:2096–2100 Rivet A, Sabin C, Mazeau K, Imberty A, Perez S (2001) Disaccharides database. http://www. cermav.cnrs.fr. Cited 2 Sept 2004 Sarko A, Muggli R (1974) Packing analysis of carbohydrates and polysaccharides. III. Valonia cellulose and cellulose II. Macromolecules 7:486–494 Sikorski P, Wada M, Heux L, Shintani H, Stokke B (2004) Crystal structure of cellulose triacetate I. Macromolecules 37:4547–4553 Stipanovic AJ, Sarko A (1976) Packing analysis of carbohydrates and polysaccharides. 6. Molecular and crystal structure of regenerated cellulose II. Macromolecules 9:851–857 Zugenmaier P (1974) Conformation and packing analysis of polysaccharides. I: mannan. Biopolymers 13:1127–1139 Zugenmaier P (2001) Conformation and packing of various crystalline cellulose fibers. Prog Polym Sci 26:1341–1417 Zugenmaier P, Rappenecker G (1978) The crystal and molecular structure of methyl tetraacetylb-d-glucoside. Acta Crystallogr Sect B 34:164–167

This page intentionally blank

Chapter 5

Cellulose

5.1

Cellulose Polymorphy

Pure cellulose exists in several crystalline polymorphs (Hayashi et al. 1975; Sarko 1986) with different packing arrangements. The unit cells or the diffraction patterns of these crystalline structures represent a quantity which mirrors these differences. The unit cells are listed in Table 5.1. The conformation and packing of the cellulose chains are necessary quantities for a complete description of the polymorphs to evaluate their differences in behavior and properties. However, the determination of the cellulose structures is difficult to achieve, with only a few X-ray reflections often observed at low diffraction angles. Nevertheless, these few observed data normally suffice for the determination of the unit cell. Small differences in the size of the unit cells of the same polymorph are found in studies of various cellulose species and by various authors for the same species and may be caused by differences in supermolecular, i.e., morphological structures. Native cellulose I is now recognized to simultaneously crystallize in a one-chain triclinic structure Iα and a two-chain modification Iβ, both polymorphs packed in a parallel chain arrangement, but of various ratios in a fiber, depending on the origin. An eight-chain unit cell of Valonia cellulose proposed by Honjo and Watanabe (1958) actually represents the coexistence of two modifications, Iα and Iβ, and the eight-chain unit cell splits up in an overlap of a one-chain and a twochain unit cell. In contrast, the chains of regenerated or mercerized cellulose II are arranged in antiparallel fashion in a two-chain unit cell. Treatment of cellulose with dry liquid ammonia leads to cellulose III. Starting from native cellulose I, the conversion is denoted as cellulose III1, since this polymorph may be converted back to cellulose I. On the other hand, if cellulose II is the starting material for the liquid ammonia treatment, cellulose II may be regained, and this polymorph is termed III2. Heat treatment of cellulose III1 or cellulose III2 leads to cellulose IV1 or cellulose IV2 and these polymorphs can be converted back to the original celluloses. Recently it was proposed that cellulose IV1 actually consists of cellulose Iβ, if the conversion is performed with highly crystalline samples. Native cellulose can undergo reversible and/or irreversible conversions to other polymorphs (Scheme 5.1). Cellulose Iα converts by heat treatment irreversibly to P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

101

Cellulose Iα Cellulose Iβ Cellulose II mercerized (from flax) Cellulose II mercerized (from ramie) Cellulose II regenerated (from Fortisan) Cellulose III1 Cellulose IV1 Cellulose IV2 Cellulose II –hydrazine Cellulose hydrate I Cellulose I–ammonia Cellulose I–EDA (with a/2) Sodium cellulose I Sodium cellulose IV EDA ethylenediamine

6.717 7.784 8.01

8.10

8.03

4.450 8.03 7.99 9.37 9.02 4.47 4.76 8.83 9.57

1 2 2

2

2

1 2 2 4

2 1 1

4 2

P1 P21 P21

P21

P21

P21 P1 P1 P21

P21 P21 P21

P21 P21

25.28 8.72

9.63 8.81 12.88

7.850 8.13 8.10 19.88

9.04

9.03

5.962 8.201 9.04

10.29 10.35

10.34 10.34 10.35

10.31 10.34 10.34 10.39

10.35

10.31

10.400 10.38 10.36

90 90

90 90 90

90 90 90 90

90

90

118.08 90 90

Table 5.1 Unit cells for polymorphs of cellulose (recent data used for structure determinations) Unit cell Space Number of Type group chains a (Å) b (Å) c (Å) a (°) b (°)

90 90

90 90 90

90 90 90 90

90

90

114.80 90 90

90.0 122.0

116.4 92.7 118.8

105.10 90 90 120.0

117.11

117.1

80.37 96.55 117.1

g (°) Reference

Nishimura et al. (1991) Nishimura and Sarko (1991)

Lee and Blackwell (1981b) Wada et al. (2006) Lee et al. (1984)

Wada et al. (2004b) Gardiner and Sarko (1985) Gardiner and Sarko (1985) Lee et al. (1983)

Langan et al. (2005)

Langan et al. (2001)

Nishiyama et al. (2003) Nishiyama et al. (2002) Langan et al. (1999)

102 5 Cellulose

5.2 Characterization

103

Scheme 5.1 Polymorphy and conversions of pure cellulose. (From Kroon-Batenburg et al. 1996)

cellulose Iβ, which, therefore, is considered the more stable form (Horii et al. 1987b). Cellulose I can be transformed to cellulose II by regeneration, which involves dissolution, or by mercerization (treatment in a 20% sodium hydroxide solution), retaining the original morphological form. Mercerization is regarded as an irreversible process, although some reports claim that cellulose II can be converted back to cellulose I (Macchi et al. 1968; Atalla and Nagel 1974; Hess and Gundermann 1937; Watanabe et al. 1968). Additional reports are collected in Hermans (1949, p 154), some of which have been contradicted by Hess et al. (1941). Further conversions are depicted in Scheme 5.1 and discussed later.

5.2 5.2.1

Characterization X-ray Characterization

The structure of materials in small dimensions is uniquely imaged by the diffraction patterns of single crystals, of fibers or of powders (Debeye–Scherrer method) as discussed in Chap. 3. Therefore, the diffraction patterns can be used for identification of crystalline structures of all kinds. They are predominantly used and are very helpful in identifying biological materials such as cell walls and in establishing their source as cellulose, chitin, etc. For this purpose, standards of cellulose X-ray

104

5 Cellulose

patterns of various structures are necessary to identify unknown materials, which might be available in the form of fibers or of randomly oriented materials. Herzog and Jancke (1920a, b) recognized by X-ray diffraction in 1920 that cellulose exists in two crystalline modifications, now termed cellulose I and cellulose II, previously named native cellulose, in contrast to cellulose hydrate for mercerized as well as regenerated cellulose – both structures, mercerized and regenerated cellulose, represent the same crystalline modification. Actually, mercerized or regenerated cellulose II is not a hydrate as was later recognized, and the term hydrate was dropped in favor of cellulose II. However, Sakurada and Hutino (1930) found that cellulose II might incorporate water to form a genuine cellulose hydrate, which can be divided according to Hermans and Weidinger (1946) into two modifications with different amounts of water. Two more types of cellulose can be characterized by the X-ray method as shown by Hess and Trogus (1935) and at the same time by Barry et al. (1936), namely, cellulose III, after treatment of native cellulose I with dry ammonia, and by Hess et al.(1941), who established the X-ray pattern of cellulose IV after high-temperature treatment of cellulose III. Since cellulose III and cellulose IV may be obtained from cellulose I and cellulose II as starting materials, respectively, they will subsequently be marked with subscripts 1 or 2. All six modifications can be discriminated by X-ray analysis (Fig. 5.1). Debye–Scherrer patterns of isotropically oriented samples as shown in Fig. 5.1 represent a simple method of characterizing crystalline cellulose polymorphs

Fig. 5.1 Intensity traces of Debye–Scherrer patterns versus diffraction angles of various polymorphs of cellulose: native cellulose I, regenerated or mercerized cellulose II, ammonia-treated cellulose of original cellulose I or cellulose II resulting in cellulose III1 and cellulose III2 and heat-treated cellulose IV1 and cellulose IV2. (From Isogai 1994)

5.2 Characterization

105

Fig. 5.2 Fiber X-ray photographs for a native cellulose I, b cellulose II, c cellulose III1 and d cellulose IV1 (left) and cellulose IV2 (right). Fiber axis vertical. (Patterns courtesy of A. Sarko)

(Isogai et al. 1989; Isogai 1994). A fingerprint procedure suffices for identifying a certain polymorph. An X-ray pattern of an unknown substance is visually compared with patterns of known structures as shown in Fig. 5.1 and the unknown material can be identified upon agreement with the known pattern. Two-dimensional fiber patterns (Fig. 5.2) allow a more precise identification because of fewer overlapping reflections and a better observation of reflections of higher-layer lines. Instead of the traces in a graph, it is easier to compare the blackening on a film (negative or positive prints are produced) or signals of other detection devises. All cellulose modifications I–IV have a fiber period of approximately 10.3 Å in common, which generally leads to a second-order meridional reflection of 5.15 Å. The first-order meridional reflection is normally missing owing to a twofold screw axis of the cellulose chain along the fiber axis in many patterns or owing to symmetry elements of the space group.

5.2.2

Spectroscopic Characterization

In addition, the characterization of cellulosic materials by X-ray investigations can be supported by spectroscopic fingerprint patterns or these patterns may solely serve for a discrimination of various polymorphs. X-ray studies need well-ordered

106

5 Cellulose

structures to obtain a sufficient number of reflections necessary for an analysis, as demonstrated in Chap. 3. In contrast, the spectroscopic investigations can be carried out also for less crystalline or noncrystalline materials and are superior in this field. This means spectroscopic studies are complementary to X-ray analysis. Actually, the limited experimental data base for cellulose forces us to combine both methods for a better analysis. The cross-polarization/magic angle spinning (CP/MAS) 13C NMR spectra of various cellulose polymorphs are collected in Fig. 5.3 and the frequency range for the signals of various carbon atoms are listed in Table 5.2. In contrast to X-ray techniques, high-resolution solid-state 13C NMR spectra are a better choice in detecting cellulose Iα and cellulose Iβ, which coexist in native cellulose fibers, and their ratios can be evaluated.

Fig. 5.3 NMR spectra of various cellulose structures: celluloses I, II, III1, III2, IV1 and IV2, noncrystalline cellulose and cellulose of short chain length (low degree of polymerization, DP) in dimethyl sulfoxide (DMSO). (From Isogai 1994)

5.2 Characterization

107

Table 5.2 Chemical shifts of carbons of the anhydroglucopyranose residue of cellulose polymorphs in solid-state 13C NMR spectra. (From Isogai 1994) Chemical shift (ppm) Polymorph

C1

C4

C6

Cellulose I Cellulose II Cellulose III1 Cellulose III2 Cellulose IV1 Cellulose IV2 Amorphous Cellulose in dimethyl sulfoxide

105.5–106.0 105.8–106.3 105.3–105.6 106.7–106.8 105.6 105.5 ≈105 102.7

89.1–89.8 88.7–88.8 88.1–88.3 88.0 83.6–84.4 83.5–84.6 ≈84 80.1

65.5–66.2 63.5–64.1 62.5–62.7 62.1–62.8 63.3–63.8 63.7 ≈63 60.6

Figure 5.4 shows several traces of solid-state CP/MAS 13C NMR spectra of higher resolution for some cellulose polymorphs and significant conclusions can be derived concerning the number of asymmetric or basic residues in the unit cell. The NMR spectrum of cellulose III1 (Fig. 5.4, spectrum a) shows only singlet peaks for all carbon atoms, which can be interpreted as meaning that one unique residue is needed for a description of the asymmetric unit of the unit cell. The doublets in spectra b–d for microcrystalline cellulose II, for cellulose Iβ from tunicate and for cellulose Iα from Glaucosytis clearly point towards two residues as asymmetric units or to overlaps of structures. Nevertheless, the spectra can serve for identification

Fig. 5.4 Cross-polarization/magic angle spinning (CP/MAS) 13C NMR spectra of a cellulose III1, b microcrystalline cellulose II, c cellulose Iβ from tunicate and d cellulose Iα from Glaucocystis. (From Wada et al. 2001)

108

5 Cellulose

of the source of the celluloses and for characterization of the polymorphs Iα and Iβ. The spectrum of cellulose Iα shows one peak for C1 (for the assignment, see Fig. 5.3), which represents an overlap of two peaks, and a doublet is seen for C1 in the spectrum of cellulose Iβ. Tunicate has a high content of cellulose Iβ and Glaucosytis has a high content of cellulose Iα. If the spectra do not represent pure crystalline structural forms, a determination of the ratio of cellulose Iα to cellulose Iβ is possible by a deconvolution of the peaks. IR spectroscopy as well as NMR spectroscopy can serve for an evaluation of the ratio of cellulose Iα and cellulose Iβ in various native fibers. In Fig. 5.5 two bands of the Fourier transform IR (FT-IR) spectra are marked that are characteristic for the crystalline Iα and Iβ structures in the OH stretching region (3,000–3,700 cm−1) and the OH out-of-plane bending region (650–800 cm−1). The traces were obtained from the Cladophora species. Both crystalline modifications are present in the original specimen, but cellulose Iα has almost disappeared after hydrothermal treatment, which converts cellulose Iα to the more stable cellulose Iβ form. The two characteristic peaks for cellulose Iα and cellulose Iβ in close vicinity point towards small differences in the hydrogen-bonding scheme and a facile rearrangement without big changes in the structures.

Fig. 5.5 Fourier transform IR (FT-IR) spectra of Cladophora cellulose (ratio of cellulose Iα to cellulose Iβ 0.63:0.37) before (a) and after (b) heat treatment at 300°C (ratio of cellulose Iα to cellulose Iβ 0.25:0.75). (From Wada et al. 2003)

5.2 Characterization

109

Various ratios of cellulose Iα and cellulose Iβ structures are observed in different algal celluloses and in the animal cellulose from tunicates, which consists of almost pure cellulose Iβ. The IR absorbance spectra are compared in the range 800– 500 cm−1 in Fig. 5.6 and the effect of heat treatment is represented for Valonia. A quantitative evaluation of the peak areas for the various species led to an estimation of the fraction of cellulose Iα, which is listed in Table 5.3. Included in this Table is also the evaluation of the NMR measurements for which the procedure was applied to Oocystis and Glaucosistis in Fig. 5.7. The conversion of native cellulose to cellulose III1 and back to cellulose Iβ was achieved by Wada et al. (2004a) by means of the FT-IR bands at 3,240 and 3,270 cm−1 in the OH stretching region (Fig. 5.8), which were assigned to the Iα and Iβ phases (see also Fig. 5.5). Highly crystalline Cladophora cellulose containing cellulose Iα (peak at 3,240 cm−1) and cellulose Iβ (peak at 3,270 cm−1) (Fig. 5.8a)

Fig. 5.6 IR absorbance spectra in the range 800–500 cm−1 from various specimens for the identification of cellulose Iα and cellulose Iβ. (From Imai and Sugiyama 1998)

110

5 Cellulose Table 5.3 Estimated fraction of Iα structure from crosspolarization/magic angle spinning 13C NMR and Fourier transform IR (FT-IR) measurements of various cellulose sources consisting of Iα and Iβ polymorphs (Imai et al. 1999; Fink et al. 1999) Iα fraction Sample

13

FT-IR

Halocynthia Annealed Valonia Valonia Cladophora Oocystis Glaucocystis Acetobacter Flax Hemp Ramie Linters

0–0.10 0.12 0.64 0.65 0.83 0.88 0.80 0.55 0.50 0.40 0.36

0 0.20 0.64 0.76 0.80 0.81

C NMR

Fig. 5.7 CP/MAS 13C NMR spectra of Glaucocystis and Oocystis (a), enlarged C4 region with the deconvoluted peaks after line shape analysis with Lorentzian functions (b). TMS tetramethylsilane. (From Imai et al. 1999)

5.2 Characterization

111

Fig. 5.8 FT-IR spectra of highly crystalline Cladophora cellulose in the OH stretching region: a initial cellulose, b hydrothermally treated cellulose, c supercritical ammonia treated cellulose III1 and d supercritical ammonia treated and further glycerol treated cellulose. Bands at 3,240 and 3,270 cm−1 are assigned to the cellulose Iα and cellulose Iβ phases, respectively. (From Wada et al. 2004a)

was hydrothermally treated and cellulose Iβ was obtained (Fig. 5.8b); supercritical ammonia treatment and evaporation of the ammonia led to the metastable cellulose III1 (Fig. 5.8c); and further high-temperature treatment in glycerol led back to cellulose Iβ (Fig. 5.8d).

112

5 Cellulose

IR and Raman spectra were for a long time the only spectroscopic means for characterizing cellulosic materials. The FT-IR spectra of cellulose Iβ, an unpolarized spectrum and two polarized spectra, are shown in Fig. A.1 (Maréchal and Chanzy 2000). Cellulose Iβ was obtained by high-temperature treatment from Valonia cellulose Iα. The same representation was chosen for the FT-IR spectra of cellulose III1 (Wada et al. 2001; Fig. A.2). The OH stretching region of the spectrum may be considered, i.e., the range from 2,800 to 3,600 cm−1, for a discussion of the kind and the strength of the hydrogen bonds, which can be used for the determination of a certain polymorph. Marrinan and Mann (1956) recorded the spectra in this region and observed pronounced differences for the various polymorphs (Figs. A.3–A.6). They distinguished between native bacteria or algal and native plant cellulose I, nowadays termed cellulose Iα and cellulose Iβ. And they also stated that cellulose Iα can be irreversibly converted to cellulose Iβ and discussed the transitions shown in Scheme 5.1. Raman spectra for the cellulose polymorphs I, II and III are presented in Fig. A.7 in the conformation-sensitive region. Differences between cellulose Iα and cellulose Iβ are found in the OH stretching region in Fig. A.8 (Atalla and VanderHart 1989).

5.3

Molecular and Crystal Structure

A detailed investigation is needed for the determination of the conformation and packing arrangements of the various cellulose polymorphs, as discussed in Chap. 3. The goodness of a structure strongly depends on the experimental data available as well as on the model evaluation by computer-aided calculations. The limited lateral order of various cellulose species reduces the number of reflections that can be collected. Lateral order rarely exceeds 50 Å for higher-plant cellulose and 200 Å for algal or tunicin cellulose and, therefore, various highly oriented algal or tunicin celluloses are preferred in diffraction experiments. A second point has to be considered. Cellulose should be obtained in pure form by a mild extraction that does not perturb or alter the original state of aggregation. However, the complex architecture of cell walls of higher plants sometimes needs harsher extraction conditions and may change the state of aggregation. It should also be emphasized that pure polymorphs and not mixtures are a prerequisite for such studies, or at least the ratio of the mixed polymorphs should be known to evaluate the common interferences. The separation of native cellulose into the two polymorphs Iα and Iβ was a problem for a long time. In this overview of crystal and molecular structures we will not describe experimental details such as the elaborate procedure of preparing specimens or refining structures, rather we will present the results of the structure determination as models and list the coordinates in the Appendix. Both the Cartesian and the fractional coordinates are provided, since the coordinates necessary for X-ray refinement are the fractional ones, and for modeling as well as for an evaluation of the structure (shape of the chain, etc.) the Cartesian coordinates are needed. Judgment concerning suitable structures is sometimes rather difficult especially

5.3 Molecular and Crystal Structure

113

for chain molecules. In many instances only averaged conformations are available concerning the targeted goal with the use of the experimental data collected. Besides excellent experimental data, which can be visualized by exceptional X-ray fiber patterns and are nowadays often available from degraded cellulose samples, some requirements derived from small molecular compounds have to be fulfilled for bond lengths, bond angles and torsion angles as well as certain invariants (virtual bond lengths, not too short van der Waals contacts, etc.). These structural data have to lie within established limits. Therefore, some tables listing these data are a necessity for a comparison of structures. The structures are drawn in various projections and when necessary the hydrogen bonds are visualized in the figures. Only strong hydrogen bonds confirmed by the geometry of the chains are listed and those weak bonds within a residue which possess no equivalents in low molecular weight crystalline structures should be looked up in the original literature. Thermal treatment of cellulose Iα leads to cellulose Iβ and it was concluded that cellulose Iβ can be regarded as the more stable form of the two polymorphs (Sugiyama et al. 1990; Wada et al. 2003; Scheme 5.1). Conversions within one family of structures of either parallel or antiparallel arrangement are quite common and the transformation sometimes occurs via solvent complexes. But a crossover from one family of structures to the other has also been observed, i.e., from cellulose I to cellulose II, with the overall morphology being left intact as occasionally observed. Only a few papers report a transformation of cellulose II to cellulose I (Macchi et al. 1968; Atalla and Nagel 1974). The crystalline structures of celluloses Iα, Iβ, II, III1, IV1 and IV2 have been determined by various authors and will be discussed in detail. All these polymorphs have a fiber repeat of about 10.3 Å in common, indicating a two-residue repeat, which is supported by the virtual bond length of a cellobiose unit in single crystal structure investigation (Tables 4.5, 4.6). But the polymorphs differ in conformation and packing arrangement and can be identified by the size of the unit cells (Table 5.1) and hydrogen-bonding arrangements detected in the OH stretching region of IR spectra (Figures A.3–A.6) or by different chemical shifts from high-resolution solid-state CP/MAS 13C NMR spectra.

5.3.1

Cellulose Ib

Highly oriented microcrystals of mantels of tunicate treated with sulfuric acid and checked to exhibit pure cellulose Iβ were used as starting materials for an X-ray (synchrotron source) and neutron diffraction study to determine the conformation and packing arrangement of this polymorph (Nishiyama et al. 2002). The experimental and calculated fiber X-ray patterns (synchrotron radiation, wavelength l = 0.72080 Å) are presented in Fig. 5.9, and for of comparison, a diagram obtained with Ni-filtered CuKα (l = 1.5418Å) radiation is also included (Wada 2002; Fig. 5.9, right). From the synchrotron X-ray pattern 312 reflections were collected and

114

5 Cellulose

Fig. 5.9 X-ray fiber pattern of cellulose Iβ (Halocynthia) obtained from synchrotron diffraction (left, upper half; l = 0.72080 Å) and a calculated pattern with the coordinates provided in Table A.11 (left, lower half) in comparison with an X-ray fiber pattern of cellulose Iβ of a thermally treated specimen (right; l = 1.5418 Å). (Left: From Nishiyama 2002; right: from Wada 2002)

from the neutron diffraction 216 intensity data were available for the determination of the structure with single crystal refinement procedures using SHELX-97 (Sheldrick 1997) ). The unit-cell dimensions are listed in Table 5.1 and space group P21 was proposed, which means a corner and a center chain are running through the unit cell both having a 21 screw axis along the chain. The 21 screw axis of each chain, which exhibits different conformations, is supported by 2D 13C–13C rotorsynchronized radiofrequency-driven recoupling (RFDR) NMR experiments (Kono and Numata 2006). The fractional and Cartesian coordinates of the final model are listed in Table A.11. They were refined in a two step procedure. In a first step the data set used was from the synchrotron experiment. The atomic positions of C and O were optimized against the mean values of the bond lengths and angles established for the structure of cellotetraose (Chap. 4) and H atoms were fixed at the carbons with respect to known bond angles and lengths but not refined. In a second step deuterium atoms were experimentally substituted for the hydrogen of the hydroxyl groups and the neutron data set was used for the refinement of the deuterons with the geometry of the basic residues fixed. Except for D03 (or H03, atom attached to O3; Fig. 5.11) the other hydroxyl deuterium atoms are in partial occupancy positions. Table 5.4 lists selected bond and torsion angles for the conformation of this model and a comparison is provided with data from other authors to be discussed later. The basic residue of the corner chain is numbered with 1, that of the center chain with 2. The 21 screw symmetry related residue of the corner chain is residue 3 (further residues 5, 7, etc.) and those of the center chain are residues 4, 6, 8, etc. to facilitate a comparison with the one-chain unit cell of cellulose Iα and cellulose III1. Various projections of the crystal structure are represented in Figs. 5.10–5.12 to visualize the hydrogen bonding occurring in this structure. The projection onto the

5.3 Molecular and Crystal Structure

115

Table 5.4 Comparison of selected torsion and bond angles of cellulose Iβ (in degrees) for some proposed structural models (Nishiyama et al. 2002, 21 screw axes along the chains; Sternberg et al. 2003, 21 screw axis missing in space group P1; Finkenstadt and Millane 1998, 21 screw axes along the chains). The adjacent residue of residue r in the chain is denoted by r′ (r′ = r + 2) Nishiyama et al. (tunicin) 1 2

Sternberg et al. Residue r 1 3 2 Torsion angle C1r–C2r–C3r–C4r −46.4 −57.3 −61.4 −51.9 −59.6 C2r–C3r–C4r–C5r 46.0 57.1 58.5 55.7 48.1 C3r–C4r–C5r–O5r −52.0 −55.1 −62.1 −60.2 −47.6 C4r–C5r–O5r–C1r 62.8 54.6 64.0 63.3 62.2 C5r–O5r–C1r–C2r −65.0 −54.9 −61.7 −60.3 −65.2 O5r–C1r–C2r–C3r 55.4 55.9 63.3 54.2 63.2 C3r–C4r–C5r–C6r −170.6 −170.5 179.5 −179.6 −167.7 Residue r→r′ 1→3 2→4 1→3 3→5 2→4 Y(C1r–O4r′–C4r′–C3r′) 90.5 94.7 94.5 93.6 91.0 F(O5r–C1r–O4r′–C4r′) −98.5 −88.7 −95.9 −94.3 −91.6 c′(C4r–C5r–C6r–O6r) −70.5 −82.5 −69.6 −69.0 −42.4 Glycosic bridge angle t(C1r–O4r′–C4r′) 115.1 116.2 114.0 116.6 115.7 Further bond angles involving the bridge oxygen O4r′–C1r–C2r 106.7 105.4 109.6 107.7 112.1 O4r′–C1r–O5r 106.3 105.9 113.0 113.1 113.9 Residue r 1 2 1 3 2 O4r–C4r–C3r 111.7 110.7 112.7 112.7 118.3 O4r–C4r–C5r 109.6 105.3 109.8 109.8 115.6

4

Finkenstadt and Millane (Valonia) 1 2

−62.4 −55.6 −52.3 53.3 52.5 52.6 −52.4 −53.5 −56.2 58.8 59.5 61.2 −63.6 −62.4 −61.0 66.4 59.7 55.7 −170.6 −172.8 −175.5 4→6 1→3 2→4 91.3 94.7 94.2 −94.7 −95.4 −95.2 −50.4 −74.3 −75.4 108.6

116.0

116.1

105.9 111.2 4 112.7 109.7

108.5 107.3 1 110.3 107.7

108.2 107.1 2 110.4 107.7

base plane (Fig. 5.10) suggests that the crystal structure is composed of sheets placed parallel to the unit cell axis b. These sheets are also present in cellulose Iα but parallel to the diagonal (Fig. 5.10, right). From a comparison two neighboring sheets, one through the corner and the other through the center of the unit cell as shown in Figs. 5.11 and 5.12, it is apparent that one sheet is shifted relative to the adjacent one by approximately one quarter in the c-direction. Hydrogen bonds of cellulose Iβ are only detected within these sheets through the corner and the center of the unit cell (Figs. 5.11, 5.12, Table 5.5). If one concedes that the conformation of the glucopyranose units in small molecules or of chains in cellulose is predominantly determined by hydrogen bonds, it is expected that the two conformations of the cellulose chain in the two sheets through the corner and the center should be similar and not different as determined by X-ray analysis (Table 5.4). The importance of the van der Waals interactions seems to be underestimated. However, intersheet hydrogen bonding may enforce different conformations of neighboring chains as observed for cellotetraose (Chap. 4) or the structure of cellulose II derived therefrom. Nevertheless, for cellulose Iβ the starting model of cellotetraose was chosen and the model was refined with bond lengths and angles restrained to the mean values of the single-crystal structure of cellotetraose.

116

5 Cellulose

Fig. 5.10 Projection of the cellulose Iβ chains (coordinates from Nishiyama et al. 2002) in the [001] direction on the a–b plane (left) in comparison with cellulose Iα chains (coordinates from Nishiyama et al. 2003) in the [001] direction (right). Sheet-like arrangements in both crystal structures can be recognized with a one-chain unit cell for cellulose Iα and a two-chain unit cell for cellulose Iβ. Hydrogen bonds only occur within the sheets. Note the similar arrangements of projected chains in cellulose Iα and cellulose Iβ despite the different unit cells

Fig. 5.11 A sheet through the corner of the unit cell along b for cellulose Iβ (coordinates from Nishiyama et al. 2002) in the [100] direction. The hydroxyl hydrogens (deuterium) are omitted

5.3 Molecular and Crystal Structure

117

Fig. 5.12 Representation in the [100] direction of a sheet through the center of the unit cell extended in the b-direction for cellulose Iβ (coordinates from Nishiyama et al. 2002). The hydroxyl hydrogens (deuterium) are omitted. Note an approximately quarter shift of the molecules in c of the unit cell compared with Fig. 5.11

Table 5.5 Selected contact distances of various models of cellulose Iβ (in angstroms). (Unit cell of Sternberg et al. (2003): a = 8.12 Å, b = 10.39 Å, c = 8.02 Å, a = 90.0°, b = 82.5°, g = 90.0°, space group P1; unit cell of Finkenstadt and Millane (1998): a = 7.85 Å, b = 8.27 Å, c = 10.38 Å, a = b = 90.0°, g = 96.8, space group P21; and unit cell of Nishiyama et al. (2002): a = 7.784 Å, b = 8.201 Å, c = 10.38 Å, a = b = 90°, g = 96.55°, space group P21; see Table 5.1

O61..O31 O33..O63 O61..O21 O23..O63 O62..O32 O34..O64 O62..O22 O24..O64

Shift of second atom in units of (a,b,c) (0,1,0) (0,0,1), fiber axis b (0,1,0) (Finkenstadt (Nishiyama et al.) (Sternberg et al.) and Millane) 2.89 2.84 2.90 2.89 2.83 2.90 3.50 3.44 3.44 3.50 3.68 3.44 2.71 2.67 2.87 2.71 2.69 2.87 3.21 3.32 3.44 3.21 3.45 3.44

118

5 Cellulose

The intrasheet hydrogen-bonding scheme should compare with that of cellulose Iα with a similar chain arrangement in the intrasheet. Only one unique conformation of the chains with cellobiose as a basic unit is possible owing to the one-chain unit cell. In contrast to cellulose Iβ, differences for cellulose Iα are expected owing to the missing 21 screw axis along the chain and different van der Waals interactions, which influence the dissimilar strength of the hydrogen bonding detected in the IR stretching mode for cellulose Iα in contrast to cellulose Iβ (Figs. A.3, A.8). However, it should be stressed that various cellulose specimens of the same polymorph differ to some extent in unit-cell dimensions because of various influences. The morphology and the surface constitution of the crystallites can constrain the conformation and packing arrangement of the cellulose I structure, including the chain length of the cellulose molecules. A shorter chain length leads to more perfect crystal structures, which are less disordered. The hydrogen-bonding scheme shown in Figs. 5.11 and 5.12 is complete, which means all the hydroxyl groups are involved in hydrogen bonds, as deduced by a comparison of low molecular compounds. However, the position of the partial occupancy of the deuterium atoms may lead to further possible hydrogen bonds, which have not been included in the figures. Hydrogen bonds involving the glycosidic oxygen or within a molecule have not yet been detected in the small-molecule regime (Chap. 4). The α-cellobiose 2NaI 2H2O complex (Peralta-Inga et al. 2002) represents an exception because strong polar interactions with sodium leave some hydroxyl hydrogens free for the proposed weak hydrogen bonds within a molecule. The quite general O5..O3′ hydrogen bond is also missing in this structure, which in general is present for all neighboring glycopyranose rings in cello-oligomers and cellulose polymorphs. Sternberg et al. (2003) performed the determination of the cellulose Iβ structure solely from high-resolution solid-state CP/MAS 13C NMR experimental data combined with a molecular dynamics simulation or geometric optimization to minimize the energy. As seen in several studies with solid-state 13C NMR experiments on cellulose Iβ, more than two components are necessary for a description of the doublets of C1 and C4 in the spectra, suggesting that the number of magnetically nonequivalent sites in the unit cell exceeds 2 (Fig. 5.13). Atalla (1987) questioned a 21 screw axis of the cellulose backbone in native cellulose and mercerized cellulose II as a result of his Raman and NMR studies and proposed anhydrocellobiose as the basic unit of the cellulose chain. From a crystallographic point of view, the cellulose Iβ structure with an arbitrary shift between the corner and the center chain in c has to possess a 21 screw axis if a monoclinic two-chain space group is proposed (Table 3.1). A symmetry axis between the two chains is not possible. Therefore, Sternberg et al. proposed space group P1 for cellulose Iβ with four independent anhydropyranose residues in the unit cell, which has a monoclinic character (Table 5.5) and they also chose b as the fiber axis. A 21 screw axis along the cellulose chain is then no longer a necessity. In this case a larger monoclinic unit cell with four chains can also be considered, e.g., as proposed by Ellis and Warwicker (1962). Selected distances along the cellulose chain, packing contacts of oxygen and bond angles as well as torsion angles are listed in Tables 5.4–5.6.

5.3 Molecular and Crystal Structure

119

Fig. 5.13 CP/MAS 13C NMR spectra of native cellulose polymorphs: cellulose Iβ and cellulose Iα in comparison with low degree of polymerization cellulose II. (From VanderHart and Atalla 1984)

An overview by Finkenstadt and Millane (1998) on Valonia cellulose summarizes the structure determinations of cellulose Iβ (algal, ramie, etc,) by several authors and provides a new refined model for this structure (unit cell, Table 5.5; for coordinates, see also Zugenmaier 2001). It should be emphasized that Valonia exhibits predominantly the cellulose Iα polymorph. Nevertheless, the Iβ structure of ramie and that determined by investigations of Valonia are almost identical, if a 21 screw axis is assumed along the cellulose chain. As expressed in Chap. 2, this structure has been disputed over a long period of time and seems now to have been solved to an extent to fit the current experimental data, including biosynthesis (Koyama et al. 1997) and morphological studies. The accepted structure of the Iβ polymorph was proposed for Valonia in the monoclinic space group P21 by Gardner and Blackwell (1974) and Sarko and Muggli (1974) using for the first time modern fiber diffraction methods. Fewer experimental data and less computer capacity was available at that time. The two proposals converge with currently available refinement techniques on modern computers and improved data processing to an almost identical model although they were originally published with slight differences (parallel-up versus parallel-down chain arrangements). The a- and b-dimensions have to be interchanged and small differences occur in the size of the unit cell. Cellulose Iβ was described by Finkenstadt and Millane (1998) as follows (Tables 5.4–5.6). Owing to the given space group P21, one anhydroglucopyranose unit is placed on each of the twofold screw axis in the corner and the center of the unit

120

5 Cellulose

cell. Two residues, i.e., the cellobiose unit, represent the fiber repeat of approximately 10.3 Å and intramolecular hydrogen bonding occurs between O5 and the adjacent O3′ of the next residue in the cellulose chain to stabilize the conformation. The chains running through the corner and the center of the unit cell have almost the same conformation as indicated by the ring torsion angles and by F (O5–C1– O4–C4) and Y (C1–O4–C4–C5) as well as by the glycosidic bridge angle t. The primary hydroxyl group is placed in the unusual tg position with a uniform torsion angle c′(C4–C5–C6–O6) of about −75° and allows an intramolecular hydrogen bond O2..O6tg′. The similarity of the two chains is caused by the optimization procedure, which introduced the same structural target values as a prerequisite. The large differences between the ring torsion angles and further structural features of the recently determined conformation of cellulose Iβ by Nishiyama et al. (2002) can be traced back to the two quite different molecules of cellotetraose and cellotriose, which served as the optimization goal (Table 5.4). These two different chain conformations are supported by NMR investigations (Fig. 5.13). The packing arrangements evaluated by the O..O contacts and hydrogen-bond distances of the two proposals discussed (Finkenstadt and Millane as compared with Nishiyama et al.) are much closer than expected. The model proposed by Sternberg et al. (2003) deviates from the established size of invariants by model compounds for cellulose, such as the virtual bond length of 5.45 ± 0.04 Å, and also confirmed by diacetylchitobiose (Mo and Jensen 1978) as well as xylobiose hexaacetate (Leung and Marchessault 1973) and the glycosidic bridge angle of approximately 116° (Table 5.5). Further, two H1..H4′ distances lie below the cutoff of 2.00 Å and one bond angle t involving the bridge oxygen has a value of 118.3° and further values lie beyond the acceptable range. The packing arrangement is dominated by a sheet-like structure of all proposals as shown in Figs. 5.10–5.12 in which the ribbonlike molecules are placed with a parallel-up arrangement in intrasheets. The shift between the positions of O4 along c in the corner versus the center chain molecules amounts to 2.58 Å (Sarko and Muggli 1974), 2.66 Å (Gardner and Blackwell 1974), 2.57 Å (Finkenstadt and Millane 1998), 2.57 Å (Nishiyama et al. 2002) and 2.14 Å (Sternberg et al. 2003). Hydrogen bonds are only present within the defined sheets, i.e., edge on, between O3..O6 and O6..O3. The model of Finkenstadt and Millane exhibits only small differences of intrasheet oxygen–oxygen distance for the sheets running through the corner and the center of the unit cell owing to the same conformation of the individual chains and parallel arrangements of the ribbonlike molecules. These differences are larger in the model of Nishiyama et al. with two different chain conformations of corner and center chains. The models of cellulose Iβ of Nishiyama et al. and Finkenstadt and Millane lie within the limiting values of bond distances, bond angles and torsion angles established by single crystal structure determination of oligomeric compounds (compare the data in Tables 4.2–4.5 with those in Tables 5.4–5.6). However, both anhydroglucopyranose units of Nishiyama et al. (2002) serving as a basic entity of the corner and center chains of cellulose Iβ considerably deviate as expected in terms of ring conformation from the molecules of cellotetraose (Raymond et al. 1995b; Gessler et al. 1995) and of cellotrio-

5.3 Molecular and Crystal Structure

121

Table 5.6 Distances along the cellulose chain (in angstroms). Residue r, adjacent residue r′ = r + 2

Residue r→r′ O5r..O3r′ O2r..O6r′ H1r..H4r′ O4r..O4r′ Residue r O6r..O4r O3r..O4r C1r–H1r Residue r→r′ O5r..O3r′ O2r..O6r′ H1r..H4r′ O4r..O4r′ Residue r O6r..O4r O3r..O4r C1r–H1r

Corner chain Finkenstadt Nishiyama et al. and Millane 1→3 1→3 2.77 2.72 2.76 2.71 2.05 2.07 5.48 5.46 1 1 2.79 2.84 2.95 2.92 0.98 1.10 Sternberg et al. 1→3 3→5 2.77 2.75 2.79 2.73 2.02 1.98 5.52 5.51 1 3 2.74 2.74 2.84 2.85 1.10 1.10

Center chain Finkenstadt Nishiyama et al. and Millane 2→4 2→4 2.70 2.71 2.86 2.80 2.18 2.08 5.44 5.45 2 2 2.89 2.82 2.99 2.91 0.98 1.10 2→4 2.71 2.65 2.02 5.73 2 2.79 3.00 1.10

4→6 2.68 2.69 1.97 5.42 4 2.79 2.89 1.11

side (Raymond et al. 1995a; for the labeling, see Tables A.5–A.7). Extreme deviations occur in torsion angles with the central bonds C4–C5, C5–O5 and O5–C1 as well as in the torsion angles F. Minor differences are observed for the further three ring torsion angles. From a comparison of the two model compounds cellotriose and cellotetraose, all ring torsion angles as well as Y are in satisfactory agreement, probably owing to the intersheet hydrogen bonding. The deviations concerning the cellulose Iβ conformations might be due to the different hydrogenbonding scheme. A synchrotron reinvestigation of the structure of cellulose Iβ from tunicin (Langan et al. 2005) at ambient temperature and at −173°C led to the conclusion that the molecular conformation of cellulose Iβ essentially remains unchanged at these two temperatures. The deviations from the structure of tunicin (Nishiyama et al. 2002) established some years earlier are satisfactory but lie outside the expected differences for the structures at the two temperatures. The packing geometry, especially concerning the intermolecular hydrogen bonds, differs to the extent that shifts can be expected in the peaks of the IR spectrum of the OH stretching region.

122

5.3.2

5 Cellulose

Cellulose Ia

The experimental data evaluation of X-ray fiber patterns (Fig. 5.14) and of solidstate CP/MAS 13C NMR spectra (Fig. 5.13) has recently led to proposals of the conformation and packing arrangement of the cellulose Iα polymorph (freshwater alga Glaucocystis nostochinearum, Nishiyama et al. 2003, coordinates in Table A.12; Acetobacter, Sternberg et al. 2003). Kono and Numata (2006) have shown for Iα-dominant Acetobacter cellulose that the two residues in the basic cellobiose unit of the one-chain unit cell possess different conformations, which represents a strong argument for the proposed space group P1 (Table 5.1). The structure-solving procedures for both models (Nishiyama et al. and Sternberg et al.) follow very closely those applied for cellulose Iβ by these research groups. An average conformation of the two chains of cellulose Iβ serves as an optimization target in the procedure of Nishiyama et al. The refinement of the X-ray structure was completed by the use of neutron scattering experiments and replacing the hydrogen of the hydroxyl groups by the deuterium atoms for localizing these atoms. Partial occupancies for D06 and D02 have been proposed but those deuterium atoms are not considered here in the presentation of the hydrogen-bonding scheme for reasons pointed out earlier. The one-chain unit cell is triclinic, space group P1, and the conformation of a chain can be approximated by a 21 screw axis. A projection of the molecules down the c-axis ([001] direction) is shown in Fig. 5.10 (right) and resembles the projection of cellulose Iβ, except for the unit-cell dimensions. The ribbonlike chains are connected by hydrogen bonds all lying within intrasheets as in cellulose Iβ (Iα model of Nishiyama et al. 2003, cf. Fig. 5.15). The molecular axes of both cellulose polymorphs possess approximately the same lateral distance within the intrasheets.

Fig. 5.14 Synchrotron X-ray fiber pattern of cellulose Iα (Glaucosystis) (upper half) and a calculated pattern (lower half). (From Nishiyama et al. 2003)

5.3 Molecular and Crystal Structure

123

Fig. 5.15 Projection of a sheet of cellulose Iα in the [110] direction. The hydroxyl hydrogens (deuterium) are omitted. The relative positions of chains within a sheet are similar to those of cellulose Iβ (Fig. 5.12)

The adjacent molecules through the corners of the unit cell are shifted by 2.8 Å, calculated with the triclinic angles α and β of the unit cell in a Cartesian coordinate system, and the interactions are of van der Waals type. In contrast, the adjacent sheet of cellulose Iβ is shifted +2.5 Å, but the following one is shifted −2.5 Å, thus leading to a monoclinic two-chain unit cell with an a–b base plane perpendicular to the chains (Fig. 2.20). The packing arrangement in the projection in Figs. 5.10 and 5.15 does not show these shifts. The intensities and positions of the X-ray reflections caused by the base plane projection are almost identical and, therefore, discrimination between the two polymorphs is rather difficult. The model of Nishiyama et al. agrees with the invariants derived from model compounds but large deviations are found for the model of Sternberg et al. (Tables 5.7–5.9) concerning the virtual bond length O4..O4′, the distance H1..H4′, the glycosidic bridge angle τ(C1–O4′–C4′) and some angles involving the glycosidic oxygen. Hydrogen bonds between the chains in the packing arrangement are absent for the model of Sternberg et al. Discrepancies have been pointed out by Sternberg et al. between the crystallographic models of cellulose I and the calculated chemical shifts of C4. Sternberg et al. suggest a shorter bond distance of O4–C4 from 1.43 to 1.36 Å. However, such a short bond has not been observed for model compounds for which precise bond distance data are available. Some concluding remarks are necessary concerning the structures of native cellulose Iα and cellulose Iβ and their determination. The X-ray data sets for cellulose Iβ of Gardener and Blackwell (1974) and Sarko and Muggli (1974) as well as Finkenstadt and Millane are based on the evaluation of X-ray reflections of Valonia,

124

5 Cellulose

Table 5.7 Ring torsion angles and further selected angles (in degrees) for the one-chain unit cells of cellulose Iα, space group P1, as well as of cellulose III1, space group P21. Coordinates for cellulose Iα from Glaucocystis from Nishiyama et al. (2003) (Table A.12) and for cellulose Iα from Acetobacter from Sternberg et al. (2003), and for cellulose III1 from Cladophora from Wada et al. (2004b) (Table A.14). Atoms of the residue adjacent to r are termed r′ (r′ = r + 2) Cellulose Iα Nishiyama et al. Sternberg et al. Torsion angle Residue r C1r–C2r–C3r–C4r C2r–C3r–C4r–C5r C3r–C4r–C5r–O5r C4r–C5r–O5r–C1r C5r–O5r–C1r–C2r O5r–C1r–C2r–C3r C3r–C4r–C5r–C6r c′(C4r–C5r–C6r–O6r) Residue r→r′ Y(C1r–O4r′–C4r′–C3r′) F(O5r–C1r–O4r′–C4r′) Glycosidic bridge angle t(C1r–O4r′–C4r′) Angle including the glycosidic oxygen O4r′–C1r–C2r O4r′–C1r–O5r Residue r O4r–C4r–C3r O4r–C4r–C5r

Cellulose III1 Wada et al.

1 −54.4 49.9 −49.1 59.2 −66.4 61.1 −165.6 −74.3 1→3 99.2 −97.7

3 −57.8 54.3 −51.0 57.2 −63.4 61.2 −169.3 −73.9 3→5 94.6 −99.1

1 −50.3 46.6 −54.3 66.3 −62.2 55.2 −171.5 −69.1 1→3 96.2 −99.6

3 −50.4 47.6 −55.2 64.0 −57.9 52.4 −175.2 −58.3 3→5 97.1 −102.2

1 −45.1 49.7 −57.9 64.1 −61.7 50.5 −176.4 162.7 1→3 93.0 −92.0

116.3

115.8

112.0

107.1

117.4

105.9 106.5 1 113.1 106.3

107.3 106.5 3 111.7 106.7

108.0 111.3 1 114.7 111.8

109.7 113.2 3 113.2 113.3

114.7 108.3 1 111.2 105.3

Table 5.8 Selected intramolecular conformation distances (in angstroms) of cellulose Iα (unit cell of Sternberg et al. 2003: a= 6.57 Å, b= 5.87 Å, c= 10.45 Å; a = 115°, b = 113°, g = 91°, space group P1; two different residue conformations along the chain) and cellulose III1 (one symmetry-related residue defines the chain). Atoms of the residue adjacent to r are termed r′ (r′ = r+2) Cellulose Iα Cellulose Iα

Residue r→r′ O5r..O3r′ O6r..O3r′ O2r..O6r′ H1r..H4r′ O4r..O4r′ Residue r O6r..O4r O3r..O4r C1r–H1r O41..O45

Nishiyama et al.

Sternberg et al.

Nishiyama et al.

Sternberg et al.

Cellulose III1 Wada et al.

1→3

1→3

1→3

3→5

3→5

2.87

2.82

2.92

2.82

2.47 2.10 5.45 1 2.79 2.99 0.98 10.40

2.65 2.02 5.56 1 2.91 2.97 1.10 10.45

2.87 3.02 4.91 2.07 5.46 1 4.27 3.04 0.98 10.31

2.48 2.07 5.46 3 2.81 2.94 0.98

2.78 1.77 5.60 3 2.91 2.91 1.10

5.3 Molecular and Crystal Structure

125

Table 5.9 Selected intermolecular packing distances (in angstroms) for cellulose Iα (unit cell of Sternberg et al. 2003: a = 6.57 Å, b = 5.87 Å, c = 10.45 Å; a = 115°, b = 113°, g = 91°), space group P1; two different residue conformations along the chain and cellulose III1 (one symmetry-related residue defines the chain). Shift of second atom in units of (a,b,c). Note the considerable differences in the size of the unit cells in the two studies of cellulose Iα Cellulose Iα Cellulose III1 O61..O31 (1,−1,0) O61..O21 (1,−1,0) O61..O21 O33..O63 (1,−1,0) O23..O63 (1,−1,0)

Nishiyama et al.

Sternberg et al.

Wada et al.

2.82 3.61

3.56 3.91

3.63 (−1,−1,0) 2.64 (0,−1,0) 2.62 (−1,−1,0)

2.77 3.64

3.45 4.10

which consists in major parts of cellulose Iα according to IR and NMR data (approximately 65%). The unit-cell dimension and symmetry of the monoclinic space group P21 (two-chain unit cell) were assumed for conformation and packing analysis as well as for structure factor refinement. This means the structure of cellulose Iβ was predominantly determined with the data set of cellulose Iα but was constrained according to the unit cell and symmetry of cellulose Iβ. The determination of the structure of ramie, predominantly consisting of cellulose Iβ, with the same methods essentially led to the same structural model (Woodcock and Sarko 1980). This observation is remarkable, since it contrasts with the idea that every structure exhibits a unique X-ray pattern (Chap. 3). This discrepancy is caused by too few observed reflections, striking similarities between the two structures and the fact that the unit cell and symmetry are taken as a prerequisite in the structure determination of cellulose Iβ from Valonia cellulose. The excellent correspondence of the net of reflection spots of cellulose Iβ and cellulose Iα in the two diffractograms, including the electron fiber diffraction patterns of Fig. 5.16, points towards the same distances between all lattice planes of cellulose Iβ and cellulose Iα, if some additional weak reflections of cellulose Iα are omitted. In this approximation both crystalline structures can be described by parallel intrasheets of the same spacing in the a-direction in case of cellulose Iβ and in a diagonal direction in case of cellulose Iα (Fig. 5.10). Since all layer spacings are the same, the position of the chain axes in space can be regarded as being equivalent, as shown in the projections in Fig. 5.10. Differences in intensities of the reflection spots for the two structures are mainly due to changes in chain conformation and the relative shifts of the parallel intrasheets in the c-direction. The relative shift of succeeding sheets of cellulose Iβ amounts to approximately c/4, 0, c/4, 0, etc. and that of the sheets of cellulose Iα to c/4, c/2, 3c/4, c, etc. Nevertheless, no big changes in the reflection intensities are found between the two structures. One pair of chains forming the set of parallel intra-sheets of the two structures differs in the shift in the c-direction, namely, that with the relative shifts of c/2 and 3c/4 for cellulose Iα as compared with 0 and c/4 for cellulose Iβ. This means that a relative shift of this cellulose Iα pair of cellulose chains of c/2 occurs and the chains have to be rotated by 180° to come to an identical position with the chains of cellulose Iβ. The special

126

5 Cellulose

Fig. 5.16 Electron fiber diffraction diagram of a bundle of Valonia microfibrils (Fig. 5.17). a Initial sample approximately 65% cellulose Iα. b Annealed sample at 240°C, approximately 75% cellulose Iβ Note a few additional reflections on the first, third and fifth layer lines on the diagram for cellulose Iα. (From Sugiyama et al. 1990)

placement of this pair of chains results in small differences in reflection intensities but does not affect the intensities on the equator that represent the projection of the chains on the base plane where the relative shift of the chains does not show up, as presented in Fig. 5.10. These stated dissimilarities lead to small differences in the diffraction patterns when comparing cellulose Iα and cellulose Iβ as illustrated in Fig. 5.16 (Sugiyama et al. 1990) but were not detected by limited X-ray investigations in 1974. Therefore, the X-ray data set from the reflections of Valonia cellulose Iα with the assumptions introduced above led to a reasonable model of cellulose Iβ and to good agreement with the model of ramie cellulose Iβ. Figure 5.17 shows electron micrographs of a disintegrated cell wall at two magnifications for which a fiber pattern or patterns of a single microfibril were taken (Imai et al. 1999). Imai and Sugiyama (1998) investigated microfibrils of the algae Valonia, Cladophora, etc. by microdiffraction with an electron beam of 20-nm diameter and were not able to observe segregated domains consisting of pure cellulose Iβ and cellulose Iα as observed for Microdictyon tenius cellulose (Sugiyama et al. 1991). They assumed multiple intermingled domains separated by crystallographic boundaries alternating either longitudinally or laterally with interfaces. Since the beam size is comparable with the microfibril diameter determined by the reflection width by coherent scattering across the entire microfibril, the presented observations include multiple interfaces of cellulose Iα and cellulose Iβ structures in the microfibril. Therefore, a determination of the fine structure of Valonia cellulose Iα, especially the positions of the hydroxyl hydrogens, and some other algal celluloses

5.3 Molecular and Crystal Structure

127

Fig. 5.17 Electron micrographs of cellulose Iα from Glaucocystis cell walls disintegrated by acid hydrolysis. a General appearances. b Higher magnification. Left insert: Electron fiber diagram of a bundle of microcrystals. Right insert: Electron diffraction diagram of a microcrystal, width approximately 10 nm. (From Imai et al. 1999)

should include, besides a consideration of the two structures, also the interfaces between them. A random selection of chains of two quite different conformations of cellulose Iα and cellulose Iβ but with a fixed correlation as observed for frustrated crystalline polymers (Cartier et al. 1996; Blackwell 2000) or for single crystals of small molecules and described by partial occupancy should not be excluded at the present time. Nevertheless, the unit cells of cellulose Iα and cellulose Iβ may be selected as the basic lattice depending on the necessity for a description of the structures and agreement with the accessible data. The big differences in NMR (Figs. 5.13, 5.18, Table 5.10) and IR (Figs. 5.5, 5.6, A.3) spectra of the two polymorphs Iα and Iβ have not yet been successfully correlated to the refined geometrical data of the corresponding conformation and packing arrangement for both polymorphs in detail. For cellulose Iβ two different asymmetric units in the unit cell and consequently as basic units of the two cellulose chains have been proposed by Nishiyama et al. (2002) on the basis of diffraction experiments as well as by Kono and Numata (2006) interpreting NMR spectra. For cellulose Iα two different asymmetric units are also found in the one-chain unit cell but this time along a single chain increasing the basic unit to anhydrocellobiose (Nishiyama et al. 2003; Kono and Numata 2006). Although the hydrogen-bonding scheme is the same for both polymorphs Iα and Iβ, the strengths of the bonds are different, as concluded from the O..O distances. These differences can qualitatively explain the appearance of the two different IR spectra in the OH stretching region (Fig. A.3). The torsion angle Φ is weakly correlated to the chemical shift of C1 and Ψ is strongly correlated to the chemical shift of C4 as proposed by Horii et al. (1987a), but the differences in the NMR spectra in Fig. 5.18 and Table 5.10 cannot be described quantitatively. It might be that these torsion angles cannot be determined with the necessary accuracy or that the influence of further ring torsion angles is underestimated. The differences between the torsion angles listed for the two asymmetric units for cellulose Iβ in Table 5.4 are greater than those for cellulose Iα collected in Table 5.7 and might describe the differences in the NMR

128

5 Cellulose

Fig. 5.18 CP/MAS 13C NMR spectra of the two asymmetric units (residue conformation) of a Iα-dominant cellulose and b Iβ-dominant cellulose. (From Kono and Numata 2006)

Table 5.10 13C chemical shifts for the two nonequivalent residues (asymmetric units) of cellulose Iα (A and A′) and cellulose Iβ (B and B′) from rotor-synchronized radiofrequency-driven recoupling NMR spectra. (From Kono and Numata 2006) 13 C chemical shifts (ppm) Polymorph

Residue

C1

C2

C3

C4

C5

C6

Cellulose Iα Cellulose Iβ

A A′ B B′

105.0 105.0 106.1 104.0

70.1 71.6 71.3 71.0

73.9 74.7 74.9 74.2

89.1 90.0 88.0 88.9

72.6 70.1 70.6 72.2

65.2 65.2 65.6 65.0

spectra of Fig. 5.13. The single narrow line for the ring carbons C1 in Fig. 5.13, spectrum B and Fig. 5.18, spectrum a represents an overlap of two lines originating from two different ring conformations as confirmed by a two-dimensional NMR analysis (Kono and Numata 2006) and is not as might be expected caused by a single asymmetric anhydroglucose unit. A doublet for all the other ring carbons supports this interpretation. A complete assignment of the chemical shifts is realized for the CP/MAS 13C NMR spectra in Fig. 5.18 for all carbon atoms of the two asymmetric units along the chain for a Iα-dominant sample as well as of the two

5.3 Molecular and Crystal Structure

129

asymmetric or basic units belonging to two different chains for a Iβ-dominant sample. The exact values of the chemical shifts are collected in Table 5.10.

5.3.3

Cellulose II

Mercerization by intracrystalline swelling of native cellulose in NaOH, washing and drying of the samples results in cellulose II. The same structure is also obtained by spinning cellulose out of solution, i.e., regeneration. Both procedures lead to almost the same unit cells (Andress 1929; Burgeni and Kratky 1929; Tables 4.7, 5.1). The differences in positions and intensities of reflections from X-ray diffraction on the equator and the packing arrangement in the unit cell with regard to cellulose Iβ are depicted schematically in Fig. 5.19 and actual fiber X-ray patterns are shown in Fig. 5.20. Figure 5.21 represents a CP/MAS 13C NMR spectrum of high-crystallinity cellulose II. Figure 5.19 only represents the projections of the chains on the base plane, now defined as the a–b plane (see also Fig. 5.22), and the corresponding reflections on the equator. Owing to the different lattice spacings of native cellulose Iβ and cellulose II, the reflections appear at different sites and exhibit different intensities on the diffractogram for the two polymorphs. One of the main features of chain arrangements is missing in these figures: the chains of cellulose Iβ are arranged in parallel fashion, and those of cellulose II are arranged in antiparallel fashion as shown in Fig. 5.23. In both structures Iβ and II the corner and center chains are

Fig. 5.19 The equator of the X-ray patterns of cellulose Iβ (a) and cellulose II (b) and the corresponding arrangements of chains in the respective projection into the base plane of the unit cells (c, d). Note the old nomenclature: b fiber axis. (From Hermans 1949)

130

5 Cellulose

Fig. 5.20 Left: Synchrotron fiber diffraction pattern (radiation wavelength 1 Å) of repeatedly mercerized cellulose II at room temperature. Right: Synchrotron fiber diffraction pattern (radiation wavelength 0.7208 Å) of mercerized ramie cellulose II. (Left: From Horii and Wada 2006; right: from Langan et al. 2001)

Fig. 5.21 CP/MAS VanderHart 1999)

13

C NMR spectrum of high crystallinity cellulose II. (From Atalla and

placed on 21 screw axes of the monoclinic space group P21 and in solid-state NMR spectra both structures exhibit a splitting of the ring carbon C1 signal (Fig. 5.21). The residue in the center of the unit cell of cellulose II is shifted by 2.73 Å along c with regard to the corner residue, if the placement of the glycosidic O41 and O46 is considered (Fig. 5.23, bottom). One anhydroglucopyranose unit on each of the two screw axes suffices for establishing the crystal structure. The second unit to represent the anhydrocellobiose residue in the unit cell is obtained by applying symmetry relations. Unit-cell sizes from various authors and specimens are collected in Tables 4.7 and 5.1.

Fig. 5.22 Projection of the cellulose II chains in the [001] direction on the a–b plane. A sheet-like structure can be defined along the a-direction through the corner and through the center of the unit cell and also in diagonal direction [1–10]. Hydrogen bonds occur within all sheets along the a-direction through the origin of the unit cell as well as through the center and between corner and center chains (diagonal) and are represented in Fig. 5.23

Fig. 5.23 Sheets through the corner along the a-direction of cellulose II projected in the [010] direction and through the center along a projected in the [110] direction of the unit cell (upper two pictures). The lower picture represents a projection in the [110] direction of a sheet running through the corner and the center of the unit cell. Note the antiparallel chain arrangement and shift of chains within the unit cell

132

5 Cellulose

Numerous structure determinations have been performed on cellulose II from different sources with different methods. A comparison of these structures shows the progress in the applied experimental and theoretical methods and represents a rigorous verification of the agreement in crystalline structures of mercerized and regenerated cellulose fibers. In recent years excellent crystalline fibers (Langan et al. 1999, 2001, 2005) have been obtained to conduct an X-ray refinement of cellulose II on a secure basis and allowing a comparison with structural data of oligosaccharides (cellotrioside and cellotetraose). The knowledge of oligomeric structures brought into focus that two quite different conformations of cellulose chains might exist in the solid state and should be discussed in evaluation of the model. Sternberg et al. (2003) reported NMR spectroscopic data that, with the use of computer simulations, led to coordinates of cellulose II from which the conformation and packing of the chains can be derived. Therefore, the progress in refinement of structural models can be followed for this polymorph as compared with the first models originating from energy minimization and X-ray evaluation for regenerated rayon by Kolpak and Blackwell (1976) and for regenerated Fortisan by Stipanovic and Sarko (1976). These early models faced the limitations of using two identical glucopyranose backbones and also the lack of computer capabilities. In contrast, the refined models of today are developed by improved modern experimental and simulation techniques. The ring torsion angles as sensitive quantities and some selected intramolecular and intermolecular distances are collected in Tables 5.11–5.13 and will be discussed. Cellulose II crystallizes in a monoclinic two-chain unit cell, which consequently requires the assignment of space group P21 (Table 3.1), if two chains with different chain polarities are present. The chains have to be placed in the corner and the center of the unit cell. The space group assignment is supported by the observation of Kono and Numata (2004) by RFDR NMR experiments that two different anhydroglucose conformations A and B of the corner chain –A–A– and the center chain –B–B– were revealed for cellulose II (Fig. 5.24, Table 5.14). The older structural models of cellulose II (Kolpak and Blackwell 1976; Stipanovic and Sarko 1976; Table 5.11) have been evaluated with two similar backbone conformations of the corner and center chains of the unit cell but with different rotational positions of O6 of the corner and center chains in agreement with space group P21. The conformation and packing analysis places the primary hydroxyl group of the corner chain in O6tg position and that of the center chain in O6gt position (or vice versa, resulting in the same structure). Corner and center chains with the same rotational O6gt positions were explicitly excluded by Stipanovic and Sarko (1976) and also by Pertsin et al. (1984). Single crystal structure evaluation of cellotrioside and cellotetraose with an almost identical unitcell base plane as cellulose II (Chap. 4) actually led to the result that two different extended molecular conformations exist but with all O6 in gt positions (Figs. 4.5, 4.6). One molecule exhibits a weak bifurcated hydrogen bond (O6′..O3) and a strong one (O5′..O3) and the second adjacent molecule lacks this weak hydrogen bond (Tables 4.6, 4.7) probably owing to lower overall (conformation and packing) energy. A bifurcated hydrogen bond is preferred by energetic considerations over a linear

Center chain

Corner chain

Center chain

Corner chain

Center chain

Corner chain

Center chain

Corner chain

Center chain

Corner chain

Center chain

Residue r 1 2 1 2 1 2 1 2 1 2 1 2 Torsion angle C1r–C2r–C3r–C4r −56.1 −51.3 −54.0 −52.0 −52.8 −47.6 −53.1 −53.6 −61.4 −51.9 −55.5 −55.6 C2r–C3r–C4r–C5r 54.2 47.2 45.9 51.0 51.9 52.6 52.2 52.1 59.8 52.6 54.4 54.5 C3r–C4r–C5r–O5r −55.6 −54.4 −49.2 −55.6 −55.7 −64.8 −55.9 −55.8 −58.1 −58.5 −55.1 −55.2 C4r–C5r–O5r–C1r 63.4 65.1 61.9 63.5 62.5 69.0 62.2 62.2 60.6 68.7 61.0 61.0 C5r–O5r–C1r–C2r −67.9 −68.0 −64.5 −64.4 −63.0 −66.0 −62.8 −63.1 −66.3 −74.8 −64.4 −64.3 O5r–C1r–C2r–C3r 62.2 61.8 59.7 57.5 57.4 56.2 57.4 58.2 65.4 63.7 59.6 59.6 C3r–C4r–C5r–C6r 179.6 179.0 −170.6 −174.5 −174.9 176.2 −175.3 −175.3 179.0 −171.1 −173.5 −173.6 c′(C4r–C5r–C6r–O6r) −164.5 −175.0 −176.4 −67.9 −172.5 148.5 173.8 −69.7 −174.3 173.0 −74.8 −176.2 Residue r→r′ 1→3 2→4 1→3 2→4 1→3 2→4 1→3 2→4 1→3 2→4 1→3 2→4 Y (C1r–O4r′–C4r′–C3r′) 95.0 86.5 89.5 88.2 92.4 84.9 93.6 94.0 98.1 92.6 95.9 95.8 F (O5r–C1r–O4r′–C4r′) −96.8 −93.5 −92.2 −91.5 −95.3 −92.0 −95.9 −96.5 −92.0 −93.2 −97.6 −97.6 Residue r→r′ (cellobiose unit !) 3→5 4→6 F (O5r–C1r–O4r′–C4r′) 102.3 87.2 Residue r→r′ (cellobiose unit !) −96.3 −90.6 a The coordinates of this model were provided by Sarko, since the coordinates of the publication did not allow a realistic cellulose model to be created. Also note that only one residue of the collobiose unit of the corner and center chain are listed and some deviations of residues 3 and 4 are detected.

Corner chain

Table 5.11 Ring and further selected torsion angles (in degrees) of cellulose II for the structure proposed by Langan et al. (2001) (the coordinates are listed in Table A.13, from which more structural data can be extracted) in comparison with various proposals (Sternberg et al. 2003; Langan et al. 1999; Kolpak and Blackwell 1976; Stipanovic and Sarko 1976; Gessler et al. 1995). The residue adjacent to r along the chain is termed r′ (r′ = r + 2) Regenerated Regenerated Mercerized Regenerated Mercerized (rayon) Regenerated (Fortisan) ramie (Langan (Sternberg flax (Langan (Kolpak and (Gessler (Stipanovic and Sarko 1976) et al. 2001) et al. 2003) et al. 1999) Blackwell 1976) et al. 1995a)

5.3 Molecular and Crystal Structure 133

Residue r→r′ 1→3 O5r..O3r′ 2.80a O6r..O3r′ 3.38 O2r..O6r′ 4.61 O4r..O4r′ 5.44 H1r..H4r′ 2.07 Glycosidicbridge angle τ(C1r–O4r′–C4r′) 115.6 Residue r 1 O6r..O4r 4.06 O3r..O4r 2.95 C1r–H1r 0.98 a Hydrogen bonds

2.66a 3.31 4.76 5.49 2.08

113.9 1 4.19 2.91 1.05

2.66a 3.03a 4.89 5.54 2.08

110.7 1 4.18 2.99 1.09

114.8 1 4.28 2.91 1.05

2.69a 3.20 4.82 5.47 2.02 116.8 1 2.93 2.87 1.05

2.70a 5.07 2.76a 5.40 2.08 116.0 1 4.21 2.93 1.05

2.89a 3.27 4.62 5.47 2.15 112.4 3 4.08 3.07 1.05

3→5 2.72a 3.25 4.64 5.45 1.99 115.1 2 4.18 3.01 0.98

2→4 2.75a 3.06a 4.97 5.45 2.12 115.3 2 2.68 2.94 1.10

2.72a 5.15 2.70a 5.48 2.08

122.3 2 4.23 2.95 1.05

2.74a 3.20 4.83 5.39 2.33

115.1 2 2.77 2.91 1.05

2.70a 5.13 2.73a 5.46 2.02

116.7 2 4.17 2.87 1.05

2.70a 3.43 4.59 5.40 2.08

111.1 2 4.27 3.05 1.05

2.87a 3.05a 4.76 5.46 2.08

118.3 4 4.05 3.12 1.05

4→6 2.85a 3.30 4.96 5.45 2.23

Kolpak Kolpak Langan Sternberg Langan and Stipanovic Gessler et al. Langan Sternberg Langan and Stupanovic Gessler et al. et al. et al. et al. Blackwell and Sarko (1995) (basic et al. et al. et al. Blackwell and Sarko (1995) (basic (2001) (2003) (1999) (1976) (1976) unit cellobiose!) (2001) (2003) (1999) (1976) (1976) unit cellobiose!)

Table 5.12 Selected distances (in angstroms) and bond angles (in degrees) for various models of cellulose II (atoms of the residue adjacent to r are termed r′; r′ = r + 2) Corner chain Center chain

O61..O31 3.82 4.27 4.01 O61..O21 2.78b 2.94b 2.74* O64..O24 2.53b 3.45 2.71b O64..O34 3.61 2.69b 3.50 O32..O63 2.85b 3.12 3.03b O22..O21 2.92b 2.78b 2.76b O64..O31 3.17 3.33 2.95b O63..O33 O63..O23 O62..O22 O62..O32 O34..O65 O24..O23 O66..O33 a These data are due to the introduction of cellobiose as the screw axis. b Hydrogen bonds

2.65b 3.44 2.97b 4.01 2.84b 2.62b 2.80b

3.52 2.37b 2.75b 4.30 2.64b 2.76b 2.59b

(1,0,0)

3.44 2.61b 2.55b 4.33

Gessler et al. (1995)a (0,1,0)

(0,1,−1)

2.67b 2.65b 2.81b basic unit. They differ somewhat from those obtained with a basic glucopyranose unit and a 21

4.07 2.76b 3.45 2.67b 3.08 2.78b 3.25

Table 5.13 Packing contacts, including hydrogen bonds. Atoms of the second chain shifted in units of (a,b,c) Langan Sternberg Langan Kolpak and Stipanovic et al. et al. et al. Blackwell and Sarko (2001) (2003) (1999) (1976) (1976) (1,0,0) (1,0,0) (1,0,0) (1,0,0) (0,−1,0) (1,0,0) (0,−1,0)

5.3 Molecular and Crystal Structure 135

136

5 Cellulose

Fig. 5.24 CP/MAS NMR spectrum of uniformly 13C-enriched cellulose II (mercerized cellulose from Acetobacter xylinum).(From Kono and Numata 2004)

Table 5.14 13C chemical shift of the asymmetric units A and B, which represent the basic units for the 21 screw axes of the corner and center chains of cellulose II. (From Kono and Numata 2004) 13 C chemical shift (ppm) Residue A Residue B

C1

C2

C3

C4

C5

C6

107.1 105.0

72.8 74.8

75.1 76.5

87.5 88.7

74.3 71.7

62.6 63.0

one (Pertsin et al. 1984). These structural features of oligosaccharides led to the consideration of two distinct conformations of the cellulose molecules, especially since the chains, cellulose II and cellotrioside or cellotetraose, are packed in an antiparallel arrangement. Therefore, this idea was followed up by Gessler et al. (1995) and Langan et al. (1999, 2001, 2005) and new models of cellulose II structures with all O6 in gt position were proposed. The two quite dissimilar cellulose chain conformations allow a different type of packing and avoid clashes of atoms in the simulated models. In contrast to the model of Langan et al. (2001) with a 21 screw axis along the cellulose chain, the model of Gessler et al. (1995) relies on cellobiose as the basic structural unit in the fiber repeat. Here, the cellobiose unit has to be regarded as a disordered structure close to the proposed 21 symmetry in space group P21. This model has the advantage that the appearance of the odd meridional reflections observed in all samples investigated can be explained (Stipanovic and Sarko 1976). The center chain of the unit cell with the cellobiose basic unit shows a weak hydrogen bond of 3.06 Å succeeded by a long O..O distance of 3.30 Å along the

5.3 Molecular and Crystal Structure

137

cellulose chain (Table 5.12). The center chain (Langan et al. 2001) with the 21 symmetry element along the chain exhibits two successive weak hydrogen bonds of 3.03 Å in the fiber repeat (Table 5.12). A bifurcated hydrogen bond along the chain was also modeled for the mannan structure (Zugenmaier 1974) and was observed in methyl cellobioside (Ham and Williams 1970). The ring torsion angles of the mannan structure and of one residue of the methyl cellobioside structure agree well – except for the torsion angle s(O5–C1–C2–C3) – with those of the center chain of the cellulose II model of Langan et al. (2001). The conformation of the ring seems to be influenced and additionally stabilized by this week hydrogen bond along the center chain besides the much stronger O5..O3′ hydrogen bonds existing along all cellulose chains, corner and center chains. The two unlike conformations of corner and center chains, one without bifurcated hydrogen bonds and the other with these bonds, differ considerably in the models of Langan et al. (2001, 2005) but not as much as anticipated from the oligomeric model compounds with almost exactly the same base planes of the unit cells (a–b plane, see Table 4.7). The chain conformations of Langan et al. (1999), Kolpak and Blackwell (1976) and Stipanovic and Sarko (1976) do not exhibit this weak hydrogen bond, but O6′ in tg position forms a strong intramolecular hydrogen bond with O2 instead. The structural model of Langan et al. (1999) represents an exception owing to a disordered O6 site of the center chain, which is not involved in any intramolecular hydrogen bond but is involved in an additional intermolecular polar interaction. The ring torsion angles as well as Y and F of Kolpak and Blackwell or Stipanovic and Sarko are similar. The conformations of the two basic residues for both cellulose II structures can be regarded as being averaged between the two different basic conformations of the model of Langan et al. (2001) or cellotretraose with a few not serious deviations as expected for the assumed equivalent cellulose backbones of corner and center chains. The older models (Langan et al. 1999; Gessler et al. 1995) with intrinsic differences in the conformations of the corner and center chains do not compare well with the recent ones (Langan et al. 2001, 2005), which might be due to changes in model refinement. Evaluation of NMR data with an energy-minimization procedure within the given unit cell of Kolpak and Blackwell (1976) led Sternberg et al. (2003) to propose two different cellulose corner and center chain conformations, one chain with O6 in tg position and the other with O6 in gt position along each chain following the older ideas. The corner chain with O6gt then forms a weak intramolecular hydrogen between O6..O3′ of 3.03 Å and the center chain with O6tg exhibits a strong intramolecular hydrogen bond of O2..O6′ of 2.70 Å. Both chains then show two intramolecular hydrogen bonds with O5..O3′ included (Table 5.12). The packing arrangement of the various cellulose II models, represented by the intermolecular hydrogen-bonding scheme (Table 5.13), does not provide any preferences for one or the other model owing to the fact that X-ray evaluation cannot locate the hydroxyl hydrogen, i.e., establishing a “true” hydrogen-bonding scheme. Therefore, well-known invariants from the oligomeric model compounds

138

5 Cellulose

as well as some selected atomic distances (Table 5.12) have to be considered for a discussion and selection of the best model. The requirements proposed by comparison with model compounds (virtual bond length, glycosidic bond angle, H1..H4′ contact, etc.) are all fulfilled by the model of Langan et al. 2001 for mercerized ramie. The short intermolecular hydrogen bond of O64..O24 with 2.53 Å finds its counterpart in the cellotriose structure with 2.56 Å. Langan et al. (2005) recently determined the structure of regenerated cellulose II (Fortisan) at low (−173°C) and ambient temperature and found only small differences on comparing the structures at these two temperatures. These models match well with some geometrical deviations to the model of mercerized ramie (Langan et al. 2001) in terms of conformation and packing arrangements. On one hand, the already short intermolecular hydrogen bond in mercerized ramie of 2.53 Å reduces to 2.41 Å at −173°C and to 2.34 Å at ambient temperature. On the other hand, the torsion angle σ(O52–C12–C22–C3) of 61.8° for mercerized ramie decreases to 52.7° for Fortisan to the expected range as observed for cellotetraose or cellotrioside with bifurcated hydrogen-bonded residues (3.10 Å for Fortisan at ambient temperature). The model of Gessler et al. 1995 deviates from the introduced standards especially in terms of too low glycosidic bridge angles (112.4°, 111.1°) as does the model of Sternberg et al. (2003) with 110.7° (Table 5.12), which also exhibits a too long virtual bond length of 5.54 Å. An earlier published structure by Langan et al. (1999) with two different chain conformations forms a high-energy O6 position close to gt and the glycosidic bridge angle of the center chain of 122.3° deviates beyond an acceptable range. The two structures of Kolpak and Blackwell (1976) or Stipanovic and Sarko (1976) fulfill the proposed requirements from the geometrical point of view but the doublets in the NMR spectrum for C1 and C4 cannot be explained with their assumption of similar residue conformations of the two basic chains in the unit cell (Fig. 5.24, Table 5.14). If metastable structures depending on the specimen and treatment exist in cellulose II, which remains to be investigated, this might reconcile the two different views on the O6 positions (gt+gt or gt+tg) along the two chains. In conclusion, structural refinement of X-ray data provides strong evidence that mercerized and regenerated cellulose II exhibit the same crystal structure. The Cartesian and fractional coordinates for the cellulose II model of mercerized ramie (Langan et al. 2001) are collected in Table A.13 and these coordinates were used for the drawings of Figs. 5.22 and 5.23.

5.3.4

Cellulose III

The two crystalline cellulose structures III1 and III2 can be prepared by the same treatment in dry liquid ammonia starting from native or regenerated (mercerized) cellulose, leading to cellulose–ammonia complexes. Evaporation of ammonia may result

5.3 Molecular and Crystal Structure

139

in a new polymorph called cellulose III1, if the starting materials are native ramie, cotton or hemp (cellulose Iβ) or algal cellulose Iα, and cellulose III2 starting from mercerized ramie or Fortisan and rayon (Marrinan and Mann 1956; Hayashi et al. 1975). Sarko et al. (1976) found the unit cells for both crystalline polymorphs to be very similar, with some differences in intensities of the meridional reflections only. These findings were recently supported by a comparison of the two similar X-ray fiber patterns of cellulose III1 and cellulose III2 (Fig. 5.25). However, it is assumed that the two structures pack in quite different fashion (Sarko et al. 1976), parallel arrangements in cellulose III1 and antiparallel ones in cellulose III2, concluded from the fact that cellulose III1 can easily be converted by mild heat treatment or through a solvent complex to the parallel-packed cellulose I and that cellulose III2 can be converted by the same mild heat treatment to antiparallel-packed cellulose II. The molecular and crystal structure of cellulose III1 was determined with improved synchrotron X-ray and neutron fiber diffraction data (Fig. 5.26) and the coordinates are listed in Table A.14 (Wada et al. 2004b). The structure of cellulose III2 has yet to be solved. Cellulose III1 consists of a one-chain unit cell (Fig. 5.26) in monoclinic space group P21, as suggested by the NMR spectrum, which shows only singlets for all carbons (Fig. 5.27). The cellulose chain conformation resembles very much the center chain of cellulose II in the model of Langan et al. (2001) with the primary hydroxyl group in gt position and a proposed bifurcated hydrogen bond between O33..O51 and O33..O61 (Fig. 5.28). The same conclusions are reached by a comparison of NMR chemical shifts for the carbons C1, C4 and C6 of the basic residues of cellulose III1 and cellulose II (Wada et al. 2001). This study establishes differences between the spectra of cellulose I, from which cellulose III1 is derived, especially concerning the chemical shift of C6. In both cellulose I polymorphs O6 is positioned in tg in contrast to gt for cellulose III1. A resemblance of Raman bands between cellulose II and cellulose III1 has also been observed (Fig. A.7).

Fig. 5.25 Comparison of the fiber synchrotron patterns (wavelength of radiation 1 Å) of a cellulose III2 and b cellulose III1 at room temperature. Differences are encircled. (From Horii and Wada 2006)

140

5 Cellulose

Fig. 5.26 X-ray fiber pattern of cellulose III1 (left). Projection of the cellulose III1 chains in the [001] direction on the a–b plane (right; coordinates from Wada et al. 2004b;). Two intrasheet-like structures can be defined: (1) along b through the corner and (2) in diagonal direction [110]. Hydrogen bonds occur within both intrasheets and are represented in Fig. 5.28. (Left: From Wada et al. 2001)

Fig. 5.27 CP/MAS 13C NMR spectrum of cellulose III1 (origin Cladophora) and assignment of all peaks. Chemical shifts (ppm): C1 106.6; C2 75.3; C3 77.7; C4 89.6; C5 74.6; C6 64.0. (From Kono et al. 2003)

5.3 Molecular and Crystal Structure

141

Fig. 5.28 A sheet through the corner along b of cellulose III1 projected in the [100] direction (top) and along the diagonal projected in the [−210] direction (bottom; coordinates from Wada et al. 2004b)

Two intrasheets can be defined, one along the b-axis of the unit cell, the other in a diagonal direction. Because of a small twist of the ribbonlike chains towards the a-axis (Fig. 5.26), similar distances are observed between the parallel cellulose chain axes in the two intrasheets and simultaneously the same hydrogen-bonding scheme occurs between O6 and O2 of adjacent parallel cellulose chains (Fig. 5.28). This geometry actually resembles the scheme of the intrasheet through the center of cellulose II in the a-direction (Fig. 5.22), which contains parallel chains with a similar bifurcated hydrogen bond along the chain. Nevertheless, the conformation of the pyranose residues of the two cellulose molecules III1 and II (center chain) are different as expressed by the torsion angles (Tables 5.7, 5.11, Langan et al. 2001).

142

5 Cellulose

It should also be emphasized that the parallel-packed chains are not staggered in the proposed monoclinic one-chain unit cell of cellulose III1. All O4 of the cellulose chains lie in planes parallel to the a–b base plane. Sarko et al. (1976) found very similar X-ray diagrams for cellulose III1 and cellulose III2 and proposed a two-chain unit cell for cellulose III1. This idea was motivated by the necessity for two antiparallel-running chains in cellulose III2 and the almost identical X-ray patterns of the two structures. Although the two-chain unit cell of cellulose III1 can be converted to a comparable sized one-chain unit cell as proposed by Wada, the two structures established by Wada et al. (2004b) and Sarko et al. (1976) disagree in conformation and packing arrangements of the chains. Nevertheless, the structure determination procedure applied by Sarko et al. (no space group symmetry assumed) should lead to a solution similar to that proposed by Wada et al., which actually is not the case The structure and the coordinates of the original model of Sarko et al. are summarized in an overview on crystalline cellulose polymorphs (Zugenmaier 2001). It seems that the structural model of Sarko et al. represents an intermediate form between cellulose Iβ from which it is derived and the recent proposal by Wada et al. (2004b). The role of the cellulose source and the conversion process is not clear in the two proposals and it can be speculated that implied constraints during the transformation from cellulose Iβ cause the observed differences. The model of Sarko et al. shows a conformation similar to that of cellulose Iβ with O6tg for center and corner chains and both chains exhibit similar backbone conformations. This conformation promotes intramolecular hydrogen bonding between O2 and O6tg′ and intermolecular hydrogen bonding between O3 and O6tg for both residues along the fiber repeat and in the two intrasheets as for cellulose Iβ. Only a small stagger of −0.9 Å occurs between corner and center chains in the c-direction. A model calculation of antiparallel chains tested by Sarko et al. (1976) required, as proposed for cellulose II at that time, a corner chain with all O6 in tg and a center chain with O6 in gt. Rotational disorder of O6 was also tested for cellulose III1 and it was found that a small percentage of such a mix can explain the presence of the odd-order meridional reflections detected in the X-ray pattern. Minor deviations in the overall conformation occur between center and corner chains and along one chain from a 21 screw axis of a possible P21 space group. The deviation of the 21 screw axis, especially for the rotation angle χ′, results in minor differences in the hydrogenbonding scheme along the chain when the residues above and below the cellobiose unit are considered. The conversion of cellulose Iβ to cellulose III1, as introduced by Wada et al., might proceed through possible intermediate steps or structures. It can be speculated that shifts of van der Waals connected intrasheets in b- and c-directions (Fig. 5.10) may lead to the proposal of Sarko et al. in a first step, in particular since several ammonia-complexed structures exists (Sect. 5.4.3). In a second step, a rotation of the primary hydroxyl groups from the tg to the gt positions requires a different hydrogen-bonding network and causes a small twist of the chains. This final cellulose III1 structure seems to be a reasonable stable packing arrangement at a first glance. However, this structure can be easily converted back to cellulose Iβ.

5.3 Molecular and Crystal Structure

143

The reason may lie in the higher packing density of cellulose Iβ of 1.64 g cm−3 compared with 1.55 g cm−3 for cellulose III1, which allows the storage of more energy per unit volume in cellulose Iβ. Unfortunately, both the structure of Sarko et al. and that of Wada et al. are termed cellulose III1.

5.3.5

Cellulose IV

Cellulose IV is produced by heating cellulose III in glycerol for 20 min at 260°C (Hayashi et al. 1975) and, as expected, two polymorphs are formed and these are denoted with reference to the starting materials, cellulose I or cellulose II, as cellulose IV1 and IV2. These two subgroups of cellulose IV exhibit two distinct IR spectra in the OH stretching region (Marrinan and Mann 1956; Fig. A.6). Their poor diffraction diagrams (Fig. 5.2d) and their unit cells are very similar but their derived structures can be distinguished upon heterogeneous acetylation. Cellulose IV1 reversibly transforms to parallel-packed cellulose triacetate I and cellulose IV2 to antiparallel-packed cellulose triacetate II (Gardiner and Sarko 1985). Space group P1 has to be assumed for both structures as a consequence of packing considerations. Figure 5.29 shows the structure of cellulose IV1 in projection on the a–b base plane. A similar representation is obtained for cellulose IV2, now with the corner and center chains pointing in opposite directions (Fig. 5.30). The unit-cell parameter a equals almost b within experimental error and a = b = g = 90°. A 21 screw axis along the cellulose chains can be excluded because the O6 torsion angles c do not follow such a screw axis. Nevertheless, the chain backbone was placed on such a screw axis for simplicity. The fractional coordinates and unit cells (almost identical cells for cellulose IV1 and cellulose IV2) are collected in Tables 5.1, A.15, and A.16. A projection in the [001] direction of the structure for cellulose IV1 is presented in Fig. 5.29 and a quite similar one for cellulose IV2 is shown in Fig. 5.30. A sheet-like structure appears in the projections as for most celluloses. The cellobiose unit serves as a true fiber repeat in a two-chain unit cell. Therefore, four torsion angles c describe the placement of O6 and four F and Y and four bridge angles t describe the geometry of the cellobiose residues in the unit cell for both cellulose IV structures (Fig. 5.31, Table 5.15). Owing to the assumed 21 screw axes of the backbone of both chains, the description of the chain skeleton can be reduced to a set of two torsion angles F and Y and two bridge angles t only, which are listed in Table 5.15. Up to 50° differences occur in the O6 rotations c from the ideal sites along the chain. These placements are far from the preferred conformational energyminimum positions. Only intrasheet hydrogen bonds are present, with the exception of cellulose IV2, for which an intersheet hydrogen bond was found (Fig. 5.32, Table 5.16). Amazingly, O21 and O22 for cellulose IV1 and O23 and O22 for cellulose IV2 are not involved in hydrogen bonding. The size of the unit cell for cellulose IV1 resembles very much that of cellulose Iβ with a monoclinic angle g of 90° instead of 96.3°. The hydrogen-bonding

144

5 Cellulose

Fig. 5.29 Projection in the [001] direction of the parallel chains of cellulose IV1 on the a–b base plane (coordinates from Gardiner and Sarko 1985; Table A.15)

Fig. 5.30 Projection in the [001] direction of the antiparallel-running corner and center chains of cellulose IV2 on the a–b base plane (coordinates from Gardiner and Sarko 1985)

5.3 Molecular and Crystal Structure

145

Fig. 5.31 Projection in the [100] direction of the parallel-running cellulose IV1 up chains in a sheet along b through the origin (left) and those in a sheet through the center along b (right)

scheme follows the line of cellulose Iβ with one additional hydrogen bond compared with cellulose Iβ in each of the two intrasheets in the unit cell: O23..O63 and O24..O64, making the hydrogen bonding very dense (Table 5.16). The stagger of 2.3 Å between corner and center chains also compares well with cellulose Iβ. Recently an investigation was undertaken on the polymorphism of the cellulose I family (Wada et al. 2004a). Cellulose samples of the green alga Cladophora sp. exhibiting predominantly cellulose Iα was hydrothermally converted to cellulose Iβ and further by super-critical ammonia treatment as well as conventional liquid ammonia treatment to cellulose III1. In a further step cellulose III1 was heated in glycerol to 260°C and as clearly demonstrated by X-ray diffraction, FT-IR and NMR spectroscopy, cellulose Iβ was obtained and not cellulose IV1, which should show a single innermost equatorial reflection (Figs. 5.33, 5.34). The two innermost reflections partly overlap and they may appear as a single reflection with increasing reflection width caused by decreasing crystallite size. The solid state 13C NMR as well as the FT-IR spectra of the glycerol heat treated sample confirm the assignment to cellulose Iβ, since they resemble the spectra of the original cellulose Iβ (Fig. 5.35). The X-ray diffraction patterns for initial higher-plant ramie cellulose and the corresponding ammonia and succeeding heat-treated samples in glycerol are represented in Fig. 5.34 and the solid-state 13C NMR spectra are shown in Fig. 5.36. Here, the two innermost X-ray reflections overlap and Wada et al. argue that the smaller crystallite size causes this effect but that nevertheless the polymorph obtained is cellulose Iβ as also indicated by the NMR spectrum. The X-ray patterns of Gardiner and Sarko are of much poorer quality than the ones now presented and exhibit a very small crystallite size. However, the question of whether a metastable cellulose IV1 structure exists still remains open. The broad reflections in Fig. 5.2d for cellulose IV indicate small crystallites. The number of chains at the surface of the crystallites may be the same as or may exceed the number of chains in bulk and a very imperfect array of chains may result, to which the polymorph IV is assigned.

Distances along one chain Residue r→r′ O5r..O3r′ O2r..O6r′ O4r..O4r′ H1r..H4r′ C11–H11 Glycosidic bond angles t(C1r–O4r′–C4r′) Torsion angles Residue r c′(C4r–C5r–C6r–O6r) c(O5r–C5r–C6r–O6r) Residue r→r′ Y(C1r–O4r′–C4r′–C3r′) F(O5r–C1r–O4r–C4r′) Alternative Y(C1r–O4r′–C4r′–C5r′) Y(C1r–O4r′–C4r′–H4r′) F(C2r–C1r–O4r′-C4r′) F(H1r–C1r–O4r′-C4r′) 2→4 2.72 3.67 5.45 2.01

115.8 3 −99.6 139.8 2→4 95.4 −97.6 −143.5 −24.5 144.6 22.0

1→3 2.72 3.31 5.45 2.01

115.7

1 −80.4 158.9 1→3 95.5 −97.7

−143.4 −24.4 142.9 22.0

2 −75.0 164.3

115.8

1.05

3→5 2.72 2.87

4 −115.2 124.2

115.8

4→6 2.72 2.75

−146.2 −28.1 145.7 26.2

1 −101.5 137.3 1→3 92.4 −93.6

116.9

1→3 2.69 2.81 5.45 2.04

−143.8 −24.6 143.4 22.6

3 −76.3 162.5 2→4 95.6 −96.7

115.9

2→4 2.69 3.71 5.45 2.01

2 −68.1 170.7

116.9

3→5 2.69 3.38

4 −115.9 122.8

115.9

1.05

4→6 2.69 2.64

Table 5.15 Selected distances (in angstroms) and bond and torsion angles (in degrees) of the two subgroups of cellulose IV (Gardiner and Sarko 1985). For atom labeling see Figs. 5.31 and 5.32. The basic unit for the backbone is a glucopyranose residue with a close 21 screw axis, and for the pendant O6 it is a cellobiose unit. The cellobiose units above or below the unit cell have to be considered for contacts with O6 Cellulose IV2 Cellulose IV1

146 5 Cellulose

5.3 Molecular and Crystal Structure

147

Fig. 5.32 Projection in the [100] direction of the parallel-up chains on a sheet through the corner (left) and the parallel down chains on a sheet through the center (right) for cellulose IV2 Table 5.16. Intermolecular hydrogen bonds for the two subgroups of cellulose IV (Gardiner and Sarko 1985) in angstroms. Shift of second atom in units of (a,b,c) Cellulose IV2 Cellulose IV1 Shift of 2nd atom (0,1,0) Intrasheet (100) through origin O23..O63 2.84 O33..O63 2.63 O61..O21 3.26a O61..O31 2.70 Intrasheet (100) through center O24..O64 2.70 O34..O64 2.93 O62..O32 2.93 Intersheet hydrogen bond O64..O51 a Too long a distance for a hydrogen bond

Shift of 2nd atom (0,1,0)

Shift of 2nd atom (1,0,0)

3.31a 2.70 2.77 2.59 2.49 2.69 2.79 2.91

Chanzy et al. (1978, 1979) came to similar conclusions by investigating disencrusted cellulose subelementary fibrils of the primary cell wall of cotton and rose cells. The two observed diffraction spots on the equator agree with those of cellulose IV1 but not with those of either cellulose Iβ or cellulose II. The subelementary fibrils contain about 12–25 cellulose chains, possess a width of about 20–30 Å and a lateral disorganization is present. But nevertheless, an almost perfect scattering correlation length of about 200 Å was found along the fibril direction. Taking these data into account, it is clear that approximately 80% or more of the chains partially lie at the surface of the subelementary fibrils and cause the disorder. The cellulose molecules will not be locked with neighboring chains by hydrogen bonds as in thicker microfibrils

148

5 Cellulose

Fig. 5.33 X-ray fiber diffraction patterns of well-oriented Cladophora cellulose. a Initial sample with 46% cellulose Iβ (reflections of cellulose Iα are indicated by arrowheads); b hydrothermally treated sample with 92% cellulose Iβ; c cellulose III1 after supercritical ammonia treatment; d cellulose III1 after conventional liquid ammonia treatment; e, f samples represented in c and d treated in glycerol at 260°C both showing the cellulose Iβ form. The splitting of the innermost equatorial reflections is indicated by arrowheads and they point to the polymorph Iβ. (From Wada et al. 2004a)

Fig. 5.34 X-ray fiber diffraction patterns of ramie cellulose. a Initial fiber, cellulose Iβ (90%), b after supercritical ammonia and successive heat treatment, cellulose Iβ, and c after liquid ammonia and successive heat treatment in glycerol, cellulose Iβ (more than 99%). The overlapping of the innermost equatorial reflections is indicated by arrowheads in b and c for cellulose Iβ. (From Wada et al. 2004a)

comparable to cellulose Iβ and cellulose Iα. An uncompleted hydrogen-bonding arrangement may result as determined for cellulose IV1. The polymorph of cellulose Iβ can be suitably described by the monoclinic space group P21 and all hydroxyl groups are involved in strong hydrogen bonding, leading

5.3 Molecular and Crystal Structure

149

Fig. 5.35 Spectroscopic data for the samples shown in Fig. 5.33a–d. Left: Solid-state CP/MAS 13 C NMR spectrum. Right: FT-IR data of randomly oriented thin films in the OH stretching region. Bands at 3,240 and 3,270 cm−1 are assigned to cellulose Iα and cellulose Iβ, respectively. N.C. noncrystalline portions. (From Wada et al. 2004a)

to a low-energy structure. For cellulose IV1 with a two-chain orthorhombic unit cell, only space group P1 was proposed, since the two glucopyranose units in the fiber repeat are not symmetry-related. In addition some deviations occur from the usual rotational position of the primary hydroxyl groups. On the other hand, the morphology of the heavily treated initial cellulose may facilitate a rearrangement of the chains during the conversions as observed by Wada et al. (2004a), which may be hindered through morphological restraints, on the other hand. In the case of the subelementary fibrils studied by Chanzy et al. the lateral disorder seems to be an intrinsic property of the fibrils and cannot be erased by treatments normally leading to cellulose Iβ. Single polymeric crystals of cellulose IV2 have been grown with cellulose of low molecular weight by deacetylation of cellulose triacetate in a mixture of methylamine, dimethyl sulfoxide and water at 150°C (Buléon and Chanzy 1980). The poor electron diffraction pattern confirms two perpendicular axis corresponding to a= 7.99 Å and b=8.11 Å of the base plane. When the temperature was set below 90°C only cellulose

150

5 Cellulose

Fig. 5.36 Solid-state CP/MAS 13C NMR spectra of ramie cellulose shown in Fig. 5.34a and b. (From Wada et al. 2004a)

II was obtained and between 90 and 150°C hybrid crystals containing cellulose II and cellulose IV were obtained. It was demonstrated by Gardiner and Sarko (1985) that the structure of cellulose IV2 resides in an energy minimum, which could not be found for cellulose IV1. Nevertheless, the claim was rejected that cellulose IV1 is not a unique structure on the basis of crystallographic results and other observations. At that time cellulose IV1 was regarded as a poorly crystalline mixture of two or more crystal lattices. This claim was again taken up again in an overview of polymorphs of cellulose (Isogai 1994) in which cellulose IV1 is considered as a mixture of cellulose I and cellulose IV2. The crystalline cellulose polymorphs can be divided and classified into two families that differ in chain polarity: the parallel-chain family (celluloses Iα, Iβ, III1 and IV1) and the antiparallel-chain family (celluloses II, IV2). The probable antiparallel structure of cellulose III2 has not been determined. The chain conformation is formed by a 21 screw axis or is very close to it in all cases. Deviations are caused by rotational disorder of O6 hydroxyls. Differences can also occur because of the rotational positions of O6 of adjacent residues or chains. The unit cells of all polymorphs appear to be true two-chain or one-chain cells and not subcells of larger lattices as proposed in earlier studies for Valonia and some other algal celluloses. The fibrillar structure (morphology), that is the arrangement of a bundle of microfibrils (crystallites), has not been discussed but knowledge of it, including the amorphous part as well as the surface of the crystallites, may provide the starting point for insight into the mechanism of cellulose interconversion from one family to the other, e.g., mercerization. It

5.4 Cellulose Solvent Complexes

151

should also be pointed out that diffraction and spectroscopic methods may detect different domains in the fibrils. Diffraction responds to crystalline regions only; spectroscopy may include noncrystalline regions, e.g., surface chains, as well.

5.4

Cellulose Solvent Complexes

A few crystal structures with solvent molecules incorporated into the crystal lattice have been determined as the cellulose II-hydrazine complex (Lee et al. 1983), cellulose IIhydrate (Lee, Blackwell 1981b), the cellulose I-ammonia I complex (Wada et al. 2006) and the cellulose I-ethylenediamine complex (Lee et al. 1984). In addition a large number of cellulose complexes have been characterized by X-ray patterns. The structure of the cellulose II-hydrazine complex has been determined using a sub-cell consisting of half of the actual unit cell content i.e. the parameter a is cut in half. This approximation was justified by reasonable agreement between observed and calculated X-ray intensities. The relationship between the unit cell proposed (Table 5.1) and the subcell is not known. An alteration of the unit cell as a possibility cannot be excluded as e.g. in the case of Valonia cellulose Iα, which was first regarded as an eight-chain unit cell but turned out as a triclinic one-chain unit cell. The refinement in a sub-cell may influence especially the position of the complexing molecules and the hydrogen bonding network.

5.4.1

Cellulose II–Hydrazine Complex

The cellulose II-hydrazine complex is formed by swelling Fortisan fibers or mercerized Ramie in hydrazine (H2N-NH2) and then dried in vacuum. Two different complexes are formed and may be related to different morphologies of the two starting materials, regenerated and mercerized cellulose. The most detailed X-ray diagram was obtained from Fortisan rayon (Fig. 5.37) and used for further investigation (Lee et al. 1983). The unit cell listed in Table 5.1 contains segments of four chains and one hydrazine molecule per glucose unit. A structure determination was carried out in a subcell, half of the original one, with a/2 due to limited computational power and the coordinates are listed in Table A.19a. The space group of this two-chain subcell is approximated by P21 with some disorder. Antiparallel cellulose chain arrangement with two very similar ribbonlike cellobiose units led to intramolecular hydrogen bonds and hydrogen bonded hydrazine molecules (Figs. 5.45a, 5.46a). Table 5.17 lists possible hydrogen bonds and some noncritical short distances between nonbonded atoms. The hydrazine molecules are located in the interstitial spaces in the sheets. The glycosic bond angle of 114.4° lies on the low side with regard to the model compounds (Chap. 4) and the torsion angles Φ, Ψ compare with the molecules d and a (Table 4.5) and u and e (Table 4.6) of the model compounds. The final proposed structure shows O6 in gt and tg position along adjacent chains, respectively, which led to an intramolecular hydrogen bond of the center chain from O22..O64 of 3.01 Å as proposed in the early cellulose II conformation

152

5 Cellulose

Table 5.17 Distances in Å of possible hydrogen bonds for cellulose II-hydrazine and short distances. Second atom is shifted by (a,b,c) units Atom pairs distance atoms pairs, shift distance O51..O33 O52..O34 O22..O64 N21..N24 O21..N11 O21..N21 O21..N14 O21..N24 O31..N14 O31..N24 O63..O62 O23..N13 O23..N23 O22..N12 O22..N22 O24..N24 O64..N12 Short distances N11..H21 N11..C21

2.71 2.71 3.01 2.67 2.65 2.49 2.95 2.60 2.73 2.74 3.02 2.65 2.49 2.85 2.73 2.73 2.80

O21..N11(−1,0,0) O64..O61(1,1,0) N12..N13(2,1,0) N12..O33(2,1,0) O22..N13(2,1,0) O32..N13(2,1,0) N22..N13(2,1,0) O22..N23(2,1,0) O32..N23(2,1,0)

2.96 3.02 2.62 2.96 2.83 2.87 2.83 2.78 2.69

2.23 2.88

H51..H61A(−1,0,0)

1.88

(Kolpak and Blackwell 1976). The conformation of the two cellulose chains resembles the one of the corner chain of the cellulose I-ethylenediamine complex (Table 5.18). The interaction of cellulose II with hydrazine involves scission of intermolecular hydrogen bonds between the cellulose chains and disrupts the stacks of quarter staggered chains along the chain direction. This effect may be the reason for hydrazine to act as a cellulose solvent.

5.4.2

Cellulose II Hydrate

Crystalline cellulose hydrates are not formed with cellulose I, rather they are only formed with cellulose II (Sakurada and Hutino 1936) and two forms have been identified. It should be remembered that cellulose II without containing water has been termed cellulose hydrate for a long time. Mercer wrongly believed that mercerized cellulose (cellulose II) contains chemically bound water. One type of cellulose II hydrate has been obtained by washing alkaline cellulose with cold water until it does not contain any alkaline anymore (Sect. 5.5.2). This structure decomposes to what is called cellulose hydrate I. The hydrate of cellulose II investigated by Lee and Blackwell (1981b) was formed by swelling Fortisan fibers in hydrazine and then washing in water. The hydrate obtained is stable at 93% relative humidity and exhibits a monoclinic, antiparallel two-chain unit cell, space group P21 with disorder present and with

5.4 Cellulose Solvent Complexes

153

Fig. 5.37 X-ray fiber diffraction patterns of native Ramie-hydrazine complex (a), mercerized Ramie-hydrazine complex (b) and Fortisan-hydrazine complex (c). (From Lee, Blackwell 1981a)

the parameters a = 9.02 Å, b = 9.63 Å, c(fiber axis) = 10.34 Å, γ = 116.0° and V = 807.3 Å3 (X-ray pattern Fig. 5.38). The fractional and Cartesian coordinates determined by Lee and Blackwell are listed in Table A.17. The coordinates were transformed to the accepted convention with the chain at the origin in the up direction. The four water molecules in the unit cell were not located by the refinement technique used but have been incorporated in the X-ray evaluation by assuming average distributed water throughout the unit cell. The two cellobiose backbones in the unit cell have the same conformation and were taken from the cellulose II model of Kolpak and Blackwell (1976) as well as for the cellulose II–hydrazine complex and the cellulose I–ethylenediamine complex but with O6gt in both chains. Therefore, the two basic anhydrogluopyranoses show similar ring torsion angles as collected in Table 5.11. The corresponding glycosidic bridge angles and torsion angles F and Y for the two chains are listed in Table 5.18 with

154

5 Cellulose

Table 5.18 Selected distances (in angstroms), bond and torsion angles (in degrees) of cellulose solvent complexes (cellulose II hydrate, cellulose I–EDA, cellulose I–ammonia I) and Na IV-cellulosea Cellulose C II-hydrate C I–EDA C II-Na IV C I-NH3 Intramolecular distances O33..O51 2.68 2.69 2.60 2.94 O34..O52 2.68 2.60 O61..O33 3.12 3.31 3.27 3.25 O62..O34 3.13 3.11 O41..O43 5.47 5.46 5.45 5.46 O42..O44 5.46 5.45 H11..H43 2.01 2.01 2.13 2.10 H12..H44 2.02 2.13 C11–H11 1.05 1.05 1.05 0.98 Glycosidic bond angles τ(C11–O43–C43) 113.9 114.1 117.1 117.4 τ(C12–O44–C44) 114.3 117.1 Torsion angles c′(C41–C51–C61–O61) 162.3 −175.4 −170.2 −171.2 c′(C43–C53–C63–O63) 177.7 c′(C42–C52–C62–O62) 166.9 174.4 c′(C44–C54–C64–O64) 170.4 Y(C11–O43–C43–C33) 93.3 94.1 89.6 89.5 Y(C12–O44–C44–C34) 92.9 89.7 F(O51–C11–O43–C43) −95.4 −96.2 −92.4 −96.5 F(O52–C12–O44–C44) −95.0 −92.4 a Note that the glucopyranose unit of cellulose II-hydrate, now in d configuration, was originally published in l configuration.

some additional geometric data. The torsion angles c′ describing the position of O6 are different for the two basis residues and also deviate from the value for cellulose II and are not involved in a bifurcated hydrogen bond. The H1..H4 distance of adjacent residues in both chains as well as the bridge angles lie just beyond the lower limit of those for the oligomeric compounds and correspond to the geometric data of the cellulose I–ethylenediamine complex. Unfortunately the basic glucose residues of cellulose II hydrate were published for the l configuration as deduced from the original ring torsion angles. The chains can be converted to the d configuration by taking the mirror images, which means the application of a simple symmetry operation. This does not present a major problem concerning the conformation and arrangement of the chains as long as the determination of the structure predominantly relies on X-ray data for which a discrimination between the l and d configurations cannot be achieved at this level of refinement. The fractional and Cartesian coordinates are collected in Table A.17 for the d configuration. The primary O6 of both chains is located in gt position, which seems to be preferred for the equilibrium cellulose II structure (Langan et al. 2001) or related complexes, if sufficient space between the chains is available.

5.4 Cellulose Solvent Complexes

155

Fig. 5.38 X-ray diffraction pattern from Fortisan cellulose II hydrate (left) and a projection in the [001] direction of the structure on the base plane (right). (Left: From Lee and Blackwell 1981b)

Fig. 5.39 Projection in the [100] direction on the b–c plane of the antiparallel-running chains of cellulose II hydrate placed in a sheet through the origin and (1/2,1,0)

The packing of this hydrate structure (d configuration) is depicted in Figs. 5.38 and 5.39 showing the hydrogen bonds of O21..O64 (0,1,0) of 2.63 Å and O63..O22 (0,1,0) of 2.67 Å besides the intramolecular hydrogen bond O51..O33. The hydrogen-bonding scheme remains the same for the chains in the parallel intrasheet running through the two points (1/2,0,0) and (1,1,0) of the unit cell. All these O..O interactions do not complete the hydrogen-bonding scheme and leave possibilities open for hydrogen bonds with water. Lee and Blackwell (1981b) note that the model presented is not the only possible solution to explain the X-ray data. It was selected because it shows similarities with cellulose II, into which the hydrate can be converted.

156

5.4.3

5 Cellulose

Cellulose I–Ammonia I Complex

Ammonia can be incorporated into the crystal lattice of cellulose, enlarging the unit cell, as discovered by Hess and Trogus (1935), Hess and Gundermann (1937) and at the same time by Barry et al. (1936). X-ray investigations by Hess and Gundermann on native ramie cellulose established two completely different types of cellulose I–ammonia complexes termed cellulose I–ammonia II, a lowtemperature form, and cellulose I–ammonia I, a form above −20 to −30°C. The low-temperature form cellulose I–ammonia II exhibits a fiber repeat of 15.2 Å and contains a large number of ammonia molecules in the unit cell. In contrast the higher-temperature form cellulose I–ammonia I can be characterized by a fiber repeat of 10.3 Å and a ratio of glucose to ammonia of 1:1, at least for the complex whose crystalline structure was recently determined by Wada et al. (2006). The published data of the early studies, including the ones of Clark and Parker (1937), suggest that the intake of ammonia may be increased and several crystalline structures of this form exist with various contents of ammonia per glucose unit leading to different sizes of unit cells. The two types of ammonia cellulose, cellulose I–ammonia I and cellulose I–ammonia II, can be converted reversibly and used to obtain other cellulose structures. A cellulose–ammonia complex can also be obtained with mercerized cellulose II as starting material as well. The importance of X-ray investigations of cellulose–ammonia complexes lies in the discovery of conversions from and into various other crystalline forms and a description of the changes occurring. Clark and Parker (1937) found the same X-ray pattern for ammonia-treated cellulose derived from native cellulose or from mercerized ramie. Slow evaporation of ammonia from the crystal lattice of ammonia cellulose does not immediately lead to the starting material, cellulose I or cellulose II, rather it leads to a metastable modification, cellulose III, which was described in Sect. 5.3.4. However, the older studies relied on two chains in the unit cell instead of the nowadays accepted one-chain unit cell but with twice the volume. A complete onestep conversion of freshly prepared cellulose I–ammonia I complex to the native crystal can be obtained by treatment of the sample in concentrated aqueous ammonia (Clark and Parker 1937). However, the same treatment of ammonia cellulose from mercerized ramie (cellulose II–ammonia) results in 70% cellulose II and 30% cellulose III. Annealing both ammonia complexes, originating from cellulose I or cellulose II, at 105 C leads to cellulose III (Clark and Parker 1937). Treatment of cellulose III, originating from native or mercerized samples, in 20% sodium hydroxide solution (mercerizing strength) always results in cellulose II. Hess and Gundermann (1937) claimed almost complete correspondence in conversion behavior of cellulose III with that of mercerized ramie or native cellulose as starting material to the native cellulose form at higher temperature in the presence of water. Small differences between ammonia I originating from cellulose I and cellulose II cannot be excluded and may hinder a full conversion in the case of cellulose II–ammonia. In contrast Legrand (1951) proposed that the two lattices of cellulose III originating from native or mercerized ramie through ammonia cellulose conversion differ and termed them cellulose III1 and cellulose III2.

5.4 Cellulose Solvent Complexese

157

Davis et al. (1943) expressed the idea that ammonia bridges the ribbonlike cellulose chains placed edge on by hydrogen bonds. This idea can be verified if geometric details of the cellulose chains involved are known, e.g., if the crystal structure is known. Wada et al. (2006) determined the crystal structure of the cellulose I–ammonia I complex with a 1:1 ratio of glucose and ammonia units and discussed the transformations back to cellulose I via the cellulose III path. They used oriented cellulose samples of the green alga Cladophora sp. as starting material, immersed the fibers into liquid ammonia in a sealed vessel and in a final step annealed the material at 140°C for 1 h (supercritical treatment). The fibers were taken out of the vessel and placed in an X-ray vacuum camera for diffraction measurement. The ammonia complex was stable for several hours. A fiber X-ray diagram is shown in Fig. 5.40 and the unit cell parameters are listed in Table 5.1. The crystal and molecular structure was determined on a low data to parameter ratio (1:1) with almost the same number of constraints and is mirrored in numerous large deviations from the average geometric data (Chap. 4). Only one chain with a 21 screw axis passes through the monoclinic unit cell, space group P21. The fractional and Cartesian coordinates of this structure are listed in the Table A.18. Three projections of the parallel packing arrangements of the cellulose chains and the ammonia molecules without hydrogen are provided in Figs. 5.41 and 5.42 and selected geometric data are listed in Tables 5.18 and 5.19 to facilitate the discussion of the structure and of the conversions involving the cellulose I–ammonia I complex. The single cellulose chain (O6gt) with one anhydroglucose as the basic unit and one ammonia molecule are placed on a 21 screw axis in the corner of the unit cell. The intramolecular hydrogen bonding resembles that of molecule d or molecule a of cellotetraose on comparing the O51..O33 (2.94 Å) and O61..O33 (3.25 Å) distances, one below 3 Å and one above 3.2 Å (Tables 4.5, 5.18). In contrast, the chain conformation of cellulose III1 forms a bifurcated hydrogen bond (Table 5.8) similar to molecules u or b of cellotetraose. The glycosidic bridge angles τ of both structures lie in the expected range but larger deviations occur for several ring torsion angles compared with the respective cellotetraose molecules. The two

Fig. 5.40 X-ray fiber pattern of cellulose I–ammonia I. (From Wada et al. 2006)

158

5 Cellulose

Fig. 5.41 Projection of the chains and NH3 of cellulose I–ammonia I (Wada et al. 2006) in the [001] direction on the a–b plane

Fig. 5.42 Projection of the up chains of the cellulose I–ammonia I complex arranged in a sheet along the b-axis into the b–c plane ([100] direction a, and of the up chains of the diagonal sheet in the same [100] direction b. Note the different hydrogen bonding schemes. A dashed line marks the hydrogen bonds below 3.0 Å

different cellotetraose conformations in a unit cell show a large deviation in the bridge torsion angle Y(C1–O4–C4′–C3′) of about 10°, which is less between cellulose III1 and ammonia cellulose with comparable intramolecular hydrogen bonds. The packing arrangement of the cellulose I–ammonia I complex is shown in Figs. 5.41 and 5.42 in three projections and the essential O..O and O..N distances are listed in Table 5.19. The cellulose chains are arranged in parallel fashion owing to the one-chain unit cell. Only one ammonia molecule N11 is present in the crystallographic asymmetric unit and the second ammonia molecule N13 is related by the symmetry element of the space group. The ammonia molecules are placed between the ribbonlike cellulose chains and are involved in hydrogen bonding as

5.4 Cellulose Solvent Complexes

159

Table 5.19 Selected distances (in angstroms) concerning the structure of the cellulose I–ammonia I complex Distance Shift (2nd atom) O51..O33 O61..O33 O23..O63 (0,−1,0) O61..O21 (0,1,0) O61..N13 (0,−1,0) N13..O61 (0,1,0) O31..N11 (1,1,0) N13..O33 (1,1,0) N11..O23 (−1,0,−1) N13..O21 (1,0,0) O31..N11 (0,1,0) N13..O33 (0,1,0) O21..O61 (−1,1,0) H51..H61B (−1,0,0) H61B..H51 (1,0,0) O41..O43 O41..O45 H11..H43 C11–H11

2.942 3.246 2.914 2.914 2.761 2.761 2.861 2.861 3.142 3.142 3.195 3.195 4.384 2.12 2.12 5.46 10.34 2.10 0.98

indicated by the dashed lines in Fig. 5.41. The distance of O31 of the corner chain to N11 (lower central ammonia molecule in the unit cell) shifted from the original position by (1,1,0) amounts to 2.86 Å and can be regarded as a hydrogen bond. The distance from the same O31 to N11 (1,0,0) is 3.19 Å and it is difficult to decide if this is still a weak hydrogen bond. Translation of these two bond in the a-direction leads to a zigzag array of these bonds. However, some interactions with N11 (1,1,0) are missing in Fig. 5.41 owing to the limitation of two residues of the cellulose chain. These additional interactions can be taken into account in this drawing by consideration of the ammonia molecule N13 placed adjacent to N11 (1,1,0) in this projection. The distances N13..O61 (0,1,0) of 2.76 Å and N13.. O33 (0,1,0) of 3.19 Å are marked by dashed lines besides the symmetry-related bonding already discussed and described by the zigzag dashed lines. One hydrogen bond is retained in the asymmetric unit between the two cellulose chains, namely, O61..O21 (0,1,0) of 2.91 Å in the intrasheet. From a comparison with similar structures, e.g., cellulose III1, intrasheet and intersheet hydrogen bonding are the only possibilities for strong interactions between the chains. As listed in Table 5.9 hydrogen bonds for cellulose III1 occur between O61..O21 (0,−1,0) of 2.64 Å and O61..O21 (−1,−1,0) of 2.62 Å. In contrast hydrogen bonds between chains are completely missing for the cellulose I–ethylenediamine complex to be discussed in the following section. The conversion of ammonia cellulose to cellulose III1 occurs upon evaporation of ammonia by a decrease of the b-dimension of ammonia cellulose, an increase of the monoclinic angle and a twist of the cellulose chains leading to an interlocking system.

160

5 Cellulose

The twist and shift causes a decrease of the O21..O61 (−1,1,0) distance from 4.38 Å in cellulose I–ammonia I to a second hydrogen bond to 2.62 Å in cellulose III1 and reduces the existing hydrogen-bond distance of 2.91 Å in the intrasheet to 2.64 Å in cellulose III1.(compare Figs. 5.26 and 5.41 and Tables 5.9 and 5.19). The cellulose I–ammonia I complex can also be converted reversibly to cellulose Iβ, which is the most stable structure for cellulose. The intermediate structure cellulose III1 is a metasTable form with a density of 1.549 g cm−3, compared with 1.636 g cm−3 for cellulose Iβ and 1.463 g cm−3 cellulose I–ammonia I. The structure of cellulose Iβ was discussed in Sect. 5.3.1 and the unit cell is listed in Table 5.1. The plane of the ribbonlike cellulose Iβ chains lies almost parallel to the intrasheets passing through the corner and the center of the unit cell and two chains are 8.2 Å apart in the intrasheet. Placing ammonia molecules between the two chains in an intrasheet causes a breakdown of intramolecular hydrogen bonds O63tg..O21 along the chain as well as the intermolecular one O61tg..O31 (0,1,0) and replaces them by the intermolecular hydrogen bonds O61gt..O21 (0,−1,0), O61gt..N13 (0,−1,0) and N11..O31 (0,−1,0) and some weak polar interactions (Table 5.19). The intramolecular hydrogen bond O51..O33 remains untouched. The intake of NH3 enlarges the unit cell in the b-dimension to 8.8 Å, rotates the cellulose chain out of the intrasheet planes and increases the distance between the two intrasheets to about 4.47 Å. To achieve a threedimensional lattice of ammonia cellulose, the chains in the plane through the center parallel to b of cellulose Iβ have to be shifted along the intrasheet by b/2 and along the chain axis by c/4 to result in a one-chain unit cell. The difference in density, i.e., the favorable storage of energy per unit volume, seems to be the driving force for the conversion of cellulose III1 to cellulose Iβ. However, experimental evidence is still missing for the description of the actual pathway for this conversion.

5.4.4

Cellulose I–Ethylenediamine Complex

The crystalline cellulose I-ethylenediamine (C I-EDA) complex was obtained by swelling native ramie fibers in ethylenediamine (H2N-CH2-CH2-NH2). The complex was then dried in vacuum and was investigated utilizing the X-ray diffraction data (Figure 5.43). In addition, recently a study by CP/MAS 13C NMR was also undertaken of 13C enriched Valonia cellulose I, converted to cellulose I-EDA (Figure 5.44, Table 5.18). This sample was then further converted to cellulose III1 by alternating treatments of cellulose I in anhydrous EDA and washing in anhydrous methanol (Numata et al. 2003). The fiber diffraction data on highly crystalline fibers led to the determination of the crystal structure (Lee et al. 1984). The two-chain unit cell is monoclinic and the space group was approximated by P21 due to some disorder and is listed in Table 5.1. The EDA content is stoichiometric with one EDA molecule per anhydroglucopyranose unit. Cartesian and fractional coordinates are collected in Table A.19b. A parallel arrangement of the chains was proposed driven by the observation that the C I-EDA complex originates from parallel packed cellulose I and can be converted back to this structure by washing

5.4 Cellulose Solvent Complexes

161

Fig. 5.43 X-ray fiber diffraction pattern of cellulose I-ethylenediamine(EDA) complex (from Lee et al. 1984)

Fig. 5.44 CP/MAS 13C NMR spectra of (a) cellulose I, (b) cellulose I–EDA and (c) cellulose III1. (From Numata et al. 2003)

162

5 Cellulose Table 5.20 13C chemical shifts (δ/ppm) of cellulose Iα, cellulose Iβ, cellulose I–EDA and cellulose III1. (From Numata et al. 2003) Chemical shift (ppm) Cellulose Iα Cellulose Iβ Cellulose I–EDA Cellulose III1

C1

C2

C3

C4

C5

C6

106.9

73.6 72.6 73.0

76.5 76.2 76.8 76.0 79.4 77.7

91.6 90.8 90.6 90.0 86.4 89.6

74.4 72.6 74.2 73.0 78.3 74.6

67.1

107.6 105.9 107.0 106.6

73.4 75.3

67.5 66.9 63.8 64.0

Table 5.21 Hydrogen-bonding and short distances (in angstroms) for cellulose I–EDA Distance O51..O33 N11..O31 N23..O61 N13..O61 2nd atom shifted by (a,b,c) units O21..N23(1,1,0) N21..O23(1,1,−1) Short distances C81..O23(1,1,−1) C71..N21(−1,0,0)

2.69 2.71 2.82 2.78 2.69 2.69 2.70 2.99

with water. The packing of cellulose chains have been converted from a parallel down fashion to the now commonly accepted parallel up fashion. The cross-polarization/magnetic-angle spinning 13C NMR resonance study suggests that only one single glucopyranose conformation exists in the crystal structure, which is in agreement with a one-chain unit cell of space group P21 (Numata et al. 2003) and requires a cellulose chain with a 21 screw axis at the origin.. Such a unit cell can be easily constructed by cutting the proposed unit cell in one half (a = 4.762 Å, b = 12.876 Å, c = 10.353 Å, γ = 118.82°), which also satisfies the symmetry conditions. This procedure does not influence the published structure by Lee et al. (1984), since the originally two chains placed on 21 screw axis are identical in the z-coordinates and in ring conformation except small differences exist in the placement of the primary hydroxyl groups. The modeling and refinement techniques for the structure determination led only to rough models at the time when they had been conducted. An evaluation of packing arrangement of the cellulose chains with EDA in the small one-chain unit cell with the coordinates provided by Lee et al. (1984) came to acceptable results. Figures 5.45b and 5.46b,c represent the structure in three projections. The cellulose chains only form hydrogen bonds through the ethylenediamine molecules, which are involved in four donor and two acceptor hydrogen bonds and satisfactorily interact with all hydroxyl groups of the cellulose chains (Table 5.21). Hydrogen bonds are not formed between EDA molecules. Two close contacts are observed with EDA and the cellulose chain (see Table 5.21). A continuous interconnection exists between EDA molecules and cellulose chains for the chains lying in the (1–10) plane but not for the ones in the a-c plane (see Figure 5.46b,c).

Fig. 5.45 (a) Projection of the crystalline cellulose II–hydrazine complex in [001] direction on the a-b plane (Lee et al. 1983). (b) Projection of the crystalline cellulose I–EDA complex in [001] direction on the a–b plane (Lee et al. 1984)

Fig. 5.46 a Projection of the antiparallel chains of the cellulose II-hydrazine complex of the diagonal plane (cf. Figure 45a) in [100] direction on the b-c plane in comparison to cellulose I-EDA complex. b Projection of the up chains of the cellulose I-EDA complex in [100] direction on the b-c plane and c the up chains of the diagonal plane in the [100] direction. Note the additional hydrogen bond between O21..N23′ and N21..O23′ in c

164

5 Cellulose

A discrepancy exists concerning the ratio of EDA molecules and anhydroglucose units between the structural investigation by X-ray and by NMR. A 1:1 ratio was established by Lee et al., a 1:2 ratio by Numata et al. concluded from the integrated fractions of the respective NMR intensities. A monoclinic chiral one-chain unit cell requires space group P21 (see Table 3.1) with a 21 screw axis along the chain and is supported by a cellobiose unit in the 10.35 Å fiber repeat of the diffraction pattern and a single basic glucose residue also by NMR data. This means that a minimum of two EDA molecules are required in the unit cell of this space group and these two molecules are related by a 21 screw axis as shown in Figures 5.45b and 5.46b,c. One EDA molecule in the unit cell can only be accommodated in the triclinic space group P1 but then the constraint of a unique glucose residue as basic unit of the cellulose conformation has to be dropped and a conformation is possible comparable with the one of cellulose Iα and the NMR spectrum should show doublets. The conformation of the chains in the structure of cellulose I-EDA as proposed by Lee et al. resembles very much the conformation of cellulose Iβ (Gardener and Blackwell 1974). At the time the model was evaluated, the conformations of the glucopyranose rings were kept similar to a standard residue for all celluloses in the refinement procedure or sometimes a special residue conformation was selected. Therefore, all cellulose conformations tend to mimic one shape except the primary hydroxyl group, which was rotated to appropriate sites, for C I-EDA to the gt position in contrast to tg as for cellulose Iβ. The shape of the two independent EDA molecules in the original unit cell are similar in the proposal of Lee et al. and can be regarded as symmetry related (translation), which also points towards the onechain unit cell with one EDA molecule in the asymmetric unit. The NMR experiments are very much complimentary to X-ray results inasmuch as the local environment influences the spectrum to a large extent and the positions of the peaks in the spectrum are not strongly dependent on long range order often poorly developed in polymeric structures. X-ray determination relies on perfect long range order and short range order in the local environment does not provide a fully developed X-ray pattern. Therefore, the position and splitting of the signals for the ring carbons in the NMR spectra exhibit valuable information, which is shown in Figure 5.44 for the conversion of cellulose I to cellulose I-EDA to cellulose III1 and conclusions can be drawn regarding the various structures (Numata et al. 2003). A comparison of chemical shifts in Table 5.20 indicates differences in glucopyranose conformations of all four structures, which are due to different ring torsion angles caused by different interactions by packing arrangements, especially hydrogen bonding. Some essential details of the structure for cellulose I-EDA are presented in Tables 5.18, 5.21.

5.5

Sodium Cellulose

Controlled mercerization of cellulose to cellulose II with sodium hydroxide proceeds through several solid-state transformations as studied for ramie cellulose I. A mechanism was proposed for these transformations (Okano and Sarko 1984,

5.5 Sodium Cellulose

165

1985; Nishimura and Sarko 1987a, b). The first crystalline structure observed applying this procedure is sodium cellulose I (Nishimura et al. 1991) and the final one is sodium cellulose IV (Nishimura and Sarko 1991). These conversions have been studied by fiber diffraction techniques and antiparallel chain arrangements have been proposed for both sodium cellulose structures. During the transformation from cellulose I to cellulose II an intermediate crystalline state can be obtained known as sodium cellulose II. This crystalline structure contains hydroxyl and sodium ions. The hexagonal unit cell with a = b = 10.0 Å and a fiber repeat of c = 15.1 Å deviates from that of other chain repeats of cellulose polymorphs and is a clear indication that three residues per turn are present – approximately 5 Å per residue – forming a left-handed threefold helix. This structure was studied by Whitaker et al. (1974) and the torsion angles between two adjacent residues proposed were FH(H1–C1–O4′–C4′) = 30.4° and YH(C1–O4′–C4′–H4′) = 41.8° but the coordinates provided did not result in a reasonable glucopyranose residue for describing the helical structure. For reasons of comparison with sodium cellulose, it may be suitable to discuss a few geometrical data for the dimeric crystal structure of the α-cellobiose⋅2NaI⋅2H2O complex recently determined (Peralta-Inga et al. 2002). This structure can serve as a model compound concerning ring conformation and special features for sodium cellulose. The 12 O..Na+ distances have a narrow range of 2.27–2.50 Å. No direct hydrogen bonds among cellobiose molecules are formed and the usual intramolecular hydrogen bond bridging O3..O5′ is missing because a sixfold coordination of Na+ with five oxygen atoms and one iodine atoms can be achieved instead. The linkage torsion angles FH = 47.1° and YH = 14.6° point towards 2.9 residues per turn, if a continuous chain is considered, and the structure is close to that of a left-handed threefold helix of cellulosic chains as observed for some cellulose derivatives. Intramolecular hydrogen bonds of O3′–H′..O4′, O2′–H′..O4 and O3–H..O2 with modest stabilizing ability have been proposed because of long H..O distances of 2.5–2.6 Å and small O–H..O angles of 95–107°.

5.5.1

Sodium Cellulose I

Crystalline fibers of sodium cellulose I are formed by treatment of ramie fibers with aqueous NaOH, resulting in the most likely composition C6H10O5·NaOH·2H2O with a calculated density of 1.38 g cm−3 (X-ray fiber pattern in Fig. 5.47). Four chains run through the unit cell with dimensions a = 8.83 Å, b = 25.28 Å and c(fiber axis) = 10.29 Å and angles a = b = g = 90°. Nevertheless, as a probable space group the monoclinic space group P21 was proposed (Nishimura et al. 1991). Cartesian and fractional coordinates are listed in Table A.20. Significant structural data are collected in Tables 5.22 and 5.23 and the arrangements of the chains and sodium are presented in Figs. 5.48–5.50. The chain conformation differs in the rotational positions of the O6 hydroxyls, for which O6tg was determined in chains 1 and 2, i.e., the up chains, and O6gt in chains 3 and 4, i.e., the down chains. However, each

166

5 Cellulose

Fig. 5.47 X-ray fiber diffraction patterns of sodium cellulose I (left) and ramie cellulose I swollen with 1.5 aqueous NaOH (right). (From Nishimura et al. 1991) Table 5.22 Selected distances (in angstroms) and bond and torsion angles (in degrees) (labeling of residues by r1 and r2; note adjacent residues along the chains are 1, 5 or 2, 6 etc.; for coordinates see Table A.20; C–H distance 1.05 Å) Chain 1 Chain 2 Chain 3 Chain 4 Residues r1→r2 O5r1..O3r2 O2r1..O6r2 O4r1..O4r2 H1r1..H4r2 Glycosidic bridge angle τ(C1r1–O4r2–C4r2) Torsion angles c′(C4r1–C5r1–C6r1–O6r1) c′(C4r2–C5r2–C6r2–O6r2) c(O5r1–C5r1–C6r1–O6r1) c(O5r2–C5r2–C6r2–O6r2) Y(C1r1–O4r2–C4r2–C5r2) Y(C1r1–O4r2–C4r2–C3r2) Y(C1r1–O4r2–C4r2–H4r2) F(O5r1–C1r1–O4r2–C4r2) F(C2r1–C1r1–O4r2–C4r2) F(H1r1–C1r1–O4r2–C4r2)

1→5 2.620 2.884 5.425 2.09

2→6 2.621 4.509 5.425 2.09

3→7 2.620 4.909 5.425 2.09

4→8 2.621 2.766 5.425 2.09

116.6

116.6

116.6

116.6

−73.3 −74.8 166.7 165.2 −148.8 90.8 −29.6 −92.8 148.5 27.4

−163.9 −151.4 76.2 88.6 −148.8 90.8 −29.6 −92.7 148.5 27.5

172.6 176.8 56.6 56.9 −148.7 90.9 −29.5 −92.9 148.5 27.3

−74.9 −68.8 165.1 171.2 −148.8 90.8 −29.5 −92.8 148.5 27.6

chain can be described by an approximate twofold screw axis and the skeletons of all four chains are the same (Table 5.22). Such different rotational location of O6 was also discussed for cellulose II (Kolpak and Blackwell 1976; Stipanovic and Sarko 1976; Sternberg et al. 2003; Sect. 5.3.3). Intramolecular hydrogen bonds occur between O3 and O5′ of the adjacent residue and in addition for the tg positions between O6tg and O2′ of the neighboring residue. All eight Na+ ions form strong polar interactions with O2 and O6, but not with O3 hydroxyls. The ribbonlike cellulose chains are arranged in sheets about a/2 apart and interact through van der

5.5 Sodium Cellulose

167 Table 5.23 Distances (in angstroms) between sodium. Second atom shifted by (a,b,c) units Distance Na1..Na2(−1,0,0) Na1..Na2 Na1..Na4 Na1..Na4(0,0,−1) Na2..Na3(0,0,−1) Na3..Na4 Na5..Na6 Na5..Na8 Na6..Na7 Na7..Na8 Na1..O21 Na2..O63 Na3..O28 Na3..O64(0,0,1) Na4..O26(0,0,1) Na5..O22 Na6..O24 Na6..O68 Na7..O67(0,1,1) Na8..O25(0,1,0)

3.86 5.02 5.73 5.59 4.69 4.21 3.82 4.34 4.31 4.57 2.42 2.62 2.31 2.46 2.70 2.46 2.37 2.39 2.44 2.39

Fig. 5.48 Projection of the antiparallel chains of sodium cellulose I on the a–b base plane in the [001] direction

Waals forces. Hydrogen bonds in intrasheets between cellulose chains, edge on, as in cellulose I cannot occur, since the Na+ ions and probably water separate the cellulose chains. An all-parallel and an antiparallel model describe the X-ray data equally well. But the antiparallel chain arrangement is favored by the stereochemistry and the binding of the Na+ ions as well as by a better match of the symmetry elements. Some selected distances, bond angles and torsion angles are presented in Tables 5.22 and 5.23 as well as all strong polar interactions with Na+ (Nishimura and Sarko 1987b), the distances of which lie well within the range observed for the

168

5 Cellulose

Fig. 5.49 Projection in the [100] direction of the antiparallel chains of sodium cellulose I on the b–c plane of a sheet through the origin along b

Fig. 5.50 Projection in the [100] direction of the antiparallel chains of sodium cellulose I on the b–c plane of a sheet through a/2 along b

α-cellobiose 2NaI 2H2O complex. However, in contrast to the dimeric structure with the FH and YH angles close to those for a threefold helix, the conformation is represented by a 21 screw axis along the cellulose chain, which leads to torsion angles F and Y comparable with those of other celluloses and to the esTablished O5..O3′ intramolecular hydrogen bond.

5.5 Sodium Cellulose

5.5.2

169

Sodium Cellulose IV

Five crystalline sodium cellulose complexes have been observed during controlled mercerization for which sodium cellulose IV is the final crystalline structure before conversion to cellulose II occurs. This structure is obtained by washing sodium celluloses I, II and III, eliminating all NaOH and after drying conversion to cellulose II. Ramie fibers were the starting materials for the conversions. The most probable crystal structure represents a two-chain, monoclinic unit cell, space group P21, with antiparallel chain arrangements and unit cell parameters a= 9.57 Å, b= 8.72 Å, c(fiber axis) = 10.35 Å and g = 122.0° (Nishimura and Sarko 1991). Two water molecules are present in the unit cell. The X-ray pattern (Fig. 5.51) resembles that of cellulose II. The two antiparallel chains in the unit cell possess nearly identical conformations, a 21 screw axis of the cellulose chains except for some minor deviations of O6gt from this symmetry element, which causes certain disorder (Figs. 5.52, 5.53; Table 5.18). In contrast to recent models of cellulose II, none of the two-chain conformations exhibit an intramolecular bifurcated hydrogen bond nor is a change of the rotational position of O6 observed as proposed for the older cellulose II models for the corner chain in contrast to the center chain. The two chains are staggered by 1.3 Å in the c-direction on comparing O45 and O42 of the two chains, less than in cellulose II (Fig. 5.53). The assignment of space group P21 is also supported by the fact that the two water molecules represented by the oxygen atoms in Fig. 5.52 are closely related by a 21 screw axis. This means the structure of sodium cellulose IV can be described by two independent anhydroglucopyranose residues and one water molecule as an asymmetric unit in space group P21. Two sheets can be defined for the discussion of the packing arrangements (Fig. 5.52): one along the b-direction running through the origin of the unit cell and the

Fig. 5.51 X-ray fiber diffraction pattern of sodium cellulose IV. (From Nishimura and Sarko 1991)

170

5 Cellulose

Fig. 5.52 Projection in the [001] direction of the antiparallel-running chains of sodium cellulose IV with water molecules included on the a–b plane

Fig. 5.53 Projection in the [110] direction of the up chains of sodium cellulose IV a lying in the sheet through the origin along b and of the down chains b in the diagonal sheet through the center

other one parallel to b through the center of the unit cell (Figs. 5.52, 5.53). The corresponding O..O distances below 3.0 Å are listed in Table 5.24 and can be regarded as hydrogen bonds. The ribbonlike cellulose molecules are twisted to a dissimilar extent out of the two sheets, which shows up in a different hydrogen-bonding scheme. Cellulose chains in the sheet through the center are connected by hydrogen bonds only, in contrast to chains in the sheet passing through the origin, which are related by hydrogen bonds through the water molecules. Hydrogen bonding also occurs between corner and center chains (Table 5.24). This hydrogen-bonding network requires that each water molecule takes part in four hydrogen bonds.

5.5 Sodium Cellulose

171 Table 5.24 Intermolecular hydrogen-bonding distances (in angstroms) for sodium cellulose IV (Nishimura and Sarko 1991). Shift of second atom in units of (a,b,c). Coordinates are listed in Table A.21 Distance Intrasheet through origin O21..O1W O31..O1W O63..O2W O1W..O61(0,1,0) O2W..O23(0,1,0) O2W..O33(0,1,0) Intrasheet through center O62..O22(0,1,0) O24..O64(0,1,0) Intersheet through corner O32..O63(1,0,0) O64..O1W(1,0,0) O62..O2W O34..O61(0,1,0)

2.94 2.41 2.65 2.65 2.35 2.56 2.84 2.80 2.69 2.83 2.76 2.82

References Andress KR (1929) Das Röntgendiagramm der mercerisierten Cellulose. Z Phys Chem B 4:190–206 Atalla RH (1987) The structures of cellulose. In: Atalla RH (ed) The structures of cellulose – characterization of the solid states. ACS symposium series no 340. American Chemical Society, Washington, pp 1–14 Atalla RH, Nagel SC (1974) Cellulose – its regeneration in native lattice. Science 185:522–523 Atalla RH, VanderHart DL (1984) Native cellulose: a composite of two distinct crystalline forms. Science 223:283–284 Atalla RH, VanderHart DL (1989) Studies on the structure of cellulose using Raman spectroscopy and solid state 13C NMR. In: Schuerch C (ed) Cellulose and wood – chemistry and technology. Wiley, New York, pp 169–188 Atalla RH, VanderHart DL (1999) The role of solid state 13C NMR spectroscopy in studies of the nature of native celluloses. Solid State Nucl Magn Reson 15:1–20 Barry AJ, Peterson FC, King AJ (1936) X-ray studies of reactions of cellulose in non-aqueous systems. I. Interaction of cellulose and liquid ammonia. J Am Chem Soc 58:333–337 Blackwell J (2000) Modeling ordered arrays of cellulose chains. Abstracts of papers of the American Chemical Society 219, Cell, part 1, San Francisco, 26 March 2000, p 127 Buléon A, Chanzy H (1980) Single crystals of cellulose IV2: preparation and properties. J Polym Sci A-2 18:1209–1217 Burgeni A, Kratky O (1929) Röntgenspektrographische Beobachtungen an Cellulose. V. Über das Gitter der Hydratcellulose. Z Phys Chem B 4:401–430 Cartier L, Spassky N, Lotz B (1996) Structures frustrées de polymèrs chireaux. C R Acad Sci Paris Sér II b 322:429–435 Chanzy H, Imada K, Vuong R (1978) Electron diffraction from the primary wall of cotton fibers. Protoplasma 94:299–306

172

5 Cellulose

Chanzy H, Imada K, Mollard A, Vuong R, Barnoud F (1979) Crystallographic aspects of subelementary cellulose fibrils occurring in the wall of rose cells cultured in vitro. Protoplasma 100:303–316 Clark GL, Parker EA (1937) An X-ray diffraction study of the action of liquid ammonia on cellulose and its derivatives. J Phys Chem 41:777–786 Davis DE, Barry AJ, Peterson FC, King AJ (1943) X-ray studies of reactions of cellulose in nonaqueous systems. II. Interaction of cellulose and primary amines. J Am Chem Soc 65:1294–1299 Ellis KC, Warwicker JO (1962) A study of the crystal structure of cellulose I. J Polym Sci 56:339–357 Fink H-P, Walenta E, Kunze J (1999) Zur Struktur cellulosischer Naturfasern. Papier 9:534–542 Finkenstadt VL, Millane RP (1998) Crystal structure of Valonia cellulose Iβ. Macromolecules 31:3776–7783 Gardiner ES, Sarko A (1985) Packing analysis of carbohydrates and polysaccharides. 16. The crystal structures of cellulose IV1 and IV2. Can J Chem 63:173–180 Gardner KH, Blackwell J (1974) The structure of native cellulose. Biopolymers 13:1975–2001 Gessler K, Krauß N, Steiner T, Betzel C, Sarko A, Sänger W (1995) b-d-Cellotetraose hemihydrate as a structural model for cellulose II. An X-ray diffraction study. J Am Chem Soc 117:11397–11406 Ham JT, Williams DG (1970) The crystal and molecular structure of b-cellobioside-methanol. Acta Crystallogr Sect B 26:1373–1383 Hayashi J, Sufoka A, Ohkita J, Watanabe S (1975) The confirmation of existence of cellulose III1, III2, IV1 and IV2 by the X-ray method. Polym Lett 13:23–27 Hermans PH (1949) Physics and chemistry of cellulose fibres. Elsevier, New York Hermans PH, Weidinger A (1946) The hydrates of cellulose. J Colloid Sci 1:185–193 Herzog RO, Jancke W (1920a) Röntgenspektrographische Beobachtungen an Zellulose. Z Phys 3:196–198 Herzog RO, Jancke W (1920b) Über den physikalischen Aufbau einiger hochmolekularer organischer Verbindungen. Ber Dtsch Chem Ges 53:2162–2164 Hess K, Gundermann J (1937) Über die Einwirkung von flüssigem Ammoniak auf Cellulosefasern. Ber Dtsch Chem Ges 68:1986–1988 Hess K, Trogus C (1935) Über Ammoniak-Cellulose. Ber Dtsch Chem Ges 70:1788–1799 Hess K, Kiessig H, Gundermann J (1941) Röntgenographische und elektronenmikroskopische Untersuchungen der Vorgänge beim Vermahlen von Cellulose. Z Phys Chem B 49:64–82 Honjo G, Watanabe M (1958) Examination of cellulose fibre by the low-temperature specimen method of electron diffraction and electron microscopy. Nature 181:326–328 Hori R, Wada M (2006) The thermal expansion of cellulose II and III2 crystals. Cellulose 13:281–290 Horii F, Hirai A, Kitamaru R (1987a) Cross-polarization-magic angle spinning carbon-13 NMR approach to the structural analysis of cellulose. In: Atalla RH (ed) The structures of cellulose – characterization of the solid states. ACS symposium series no 340. American Chemical Society, Washington, pp 119–134 Horii F, Yamamoto H, Kitamaru R, Tanahashi M, Higuchi T (1987b) Transformation of native cellulose crystals induced by saturated steam at high temperatures. Macromolecules 20:2949–2951 Imai T, Sugiyama J (1998) Nanodomains of Iα and Iβ cellulose in algal microfibrils. Macromolecules 31:6275–6279 Imai T, Sugiyama J, Itoh T, Horii F (1999) Almost pure Iα cellulose in the cell wall of Glaucocystis. J Struct Biol 127:248–257 Isogai A (1994) Allomorphs of cellulose and other polysaccharides. In: Gilbert RD (ed) Cellulosic polymers. Hanser, Munich, pp 1–24 Isogai A, Usuda M, Kato T, Uryu T, Atalla RH (1989) Solid-state CP/MAS 13C NMR study of cellulose polymorphs. Macromolecules 22:3168–3172 Kolpak FJ, Blackwell J (1976) Determination of the structure of cellulose II. Macromolecules 9:273–278

References

173

Kono H, Numata Y (2004) Two-dimensional spin-exchange solid-state NMR study of the crystal structure of cellulose II. Polymer 45:4541–4547 Kono H, Numata Y (2006) Structural investigation of cellulose Iα and Iβ by 2D RFDR NMR spectroscopy: determination of sequence of magnetically inequivalent d-glucose units along cellulose chain. Cellulose 13:317–326 Kono H, Erata T, Takai M (2003) Complete assignment of the CP/MAS 13C NMR spectrum of cellulose III1. Macromolecules 36:3589–3592 Koyama M, Helbert W, Imai T, Sugiyama J, Henrissat B (1997) Parallel-up structure evidences the molecular directionality during biosynthesis of bacterial cellulose. Proc Natl Acad Sci USA 94:9091–9095 Kroon-Batenburg LMJ, Bouma B, Kroon J (1996) Stability of cellulose structures by MD simulations. Could mercerized cellulose II be parallel? Macromolecules 29:5695–5699 Langan P, Nishiyama Y, Chanzy H (1999) A revised structure and hydrogen-bonding system in cellulose II from a neutron fiber diffraction analysis. J Am Chem Soc 121:9940–9946 Langan P, Nishiyama Y, Chanzy H (2001) X-ray structure of mercerized cellulose II at 1 Å resolution. Biomacromolecules 2:410–416 Langan P, Sukumar N, Nishiyama Y, Chanzy H (2005) Synchrotron X-ray structures of cellulose Iβ and regenerated cellulose II at ambient temperature and 100 K. Cellulose 12:551–562 Lee DM, Blackwell J (1981a) Cellulose-hydrazine complexes. J Polym Sci B 19:459–465 Lee DM, Blackwell J (1981b) Structure of cellulose II hydrate. Biopolymers 20:2165–2179 Lee DM, Blackwell J, Litt MH (1983) Structure of a cellulose II-hydrazine complex. Biopolymers 22:1383–1399 Lee DM, Burnfield KE, Blackwell J (1984) Structure of a cellulose I-ethylenediamine complex. Biopolymers 23:111–126 Legrand C (1951) Recherches sur la cellulose III régénérée de l’ammoniac-cellulose. J Polym Sci 7:333–339 Leung F, Marchessault RH (1973) Crystal structure of b-d, 1→4 xylobiose hexaacetate. Chem 51:1215–1222 Mo F, Jensen LH (1978) The crystal structure of a b-(1→4) linked disaccharide, α-N,N′diacetylchitobiose monohydrate. Acta Crystallogr Sect B 34:1562–1569 Macchi E, Marx-Figini M, Fischer EW (1968) Elektronenbeugungsuntersuchungen an nativer und umgefällter Cellulose. Makromol Chem 120:235–237 Maréchal Y, Chanzy H (2000) The hydrogen bond network in Iβ cellulose as observed by infrared spectrometry. J Mol Struct 523:183–196 Marrinan HJ, Mann J (1956) Infrared spectra of crystalline modifications of cellulose. J Polym Sci 21:301–311 Nishimura H, Sarko A (1987a) Mercerization of cellulose. III. Changes in crystallite sizes. J Appl Polym Sci 33:855–866 Nishimura H, Sarko A (1987b) Mercerization of cellulose. IV. Mechanism of mercerization and crystallite sizes. J Appl Polym Sci 33:867–874 Nishimura H, Sarko A (1991) Mercerization of cellulose. 6. Crystal and molecular structure of Na-cellulose IV. Macromolecules 24:771–778 Nishimura H, Okano T, Sarko A (1991) Mercerization of cellulose. 5. Crystal and molecular structure of Na-cellulose I. Macromolecules 24:759–770 Nishiyama Y, Sugiyama J, Chanzy H, Langan P (2002) Crystal structure and hydrogen-bonding system in cellulose Iβ from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 124:9074–9082 Nishiyama Y, Chanzy H, Langan P (2003) Crystal structure and hydrogen-bonding system in cellulose Iα from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 125:14300–14306 Numata Y, Kono H, Kawano S, Erata T, Takai M (2003) Cross-polarization/magic-angle spinning 13C nuclear magnetic resonance study of cellulose I-ethylenediamine complex. J Biosci Bioeng 96:461–466 Okano T, Sarko A (1984) Mercerization of cellulose. I. X-ray diffraction evidence for intermediate structures. J Appl Polym Sci 29:4175–4182

174

5 Cellulose

Okano T, Sarko A (1985) Mercerization of cellulose. II. Alkali-cellulose intermediates and a possible mercerization mechanism. J Appl Polym Sci 30:325–332 Peralta-Inga Z, Johnson GP, Dowd MK, Rendleman JA, Stevens ED, French AD (2002) The crystal structure of the α-cellobiose·2·NaI·2·H2O complex in the context of related structures and conformational analysis. Carbohydr Res 337:851–861 Pertsin AJ, Nugmanov OK, Marchenko GN, Kitaigorodsky AI (1984) Crystal structure of cellulose polymorphs by potential energy calculations: 1. Most probable models for mercerized cellulose. Polymer 25:107–114 Raymond S, Henrissat B, Tran Qui D, Kvick Å, Chanzy H (1995a) The crystal structure of methyl b-cellotrioside monohydrate 0.25 ethanolate and its relationship to cellulose II. Carbohydr Res 277:209–229 Raymond S, Heyraud A, Tran Qui D, Kvick Å, Chanzy H (1995b) Crystal and molecular structure of b-d-cellotetraose hemihydrate as a model of cellulose II. Macromolecules 28:2096–2100 Sakurada J, Hutino K (1936) Über die Quellung der Zellulose durch Wasser. Kolloid-Z 77:346–351 Sarko A (1986) Recent X-ray crystallographic studies of celluloses. In: Young RA, Rowell RM (eds) Cellulose – structure, modification and hydrolysis. Wiley, New York, pp 29–49 Sarko A, Muggli R (1974) Packing analysis of carbohydrates and polysaccharides. III. Valonia cellulose and cellulose II. Macromolecules 7:486–494 Sarko A, Southwick J, Hayashi J (1976) Packing analysis of carbohydrates and polysaccharides. 7. Crystal structure of cellulose III1 and its relationship to other cellulose polymorphs. Macromolecules 9:857–867 Sheldrick GM (1997) SHELX-97. Program for the refinement of crystal structures. University of Göttingen Sternberg U, Koch F-T, Prieß W, Witter R (2003) Crystal structure refinements of cellulose polymorphs using solid state 13C chemical shifts. Cellulose 10:189–199 Stipanovic AJ, Sarko A (1976) Packing analysis of carbohydrates and polysaccharides. 6. Molecular and crystal structure of regenerated cellulose II. Macromolecules 9:851–857 Sugiyama J, Okano T, Yamamote H, Horii F (1990) Transformation of Valonia cellulose crystals by an alkaline hydrothermal treatment. Macromolecules 23:3196–3198 Sugiyama J, Vuong R, Chanzy H (1991) Electron diffraction study on the two crystalline phases occurring in native cellulose from an algal cell wall. Macromolecules 24:4168–4175 VanderHart DL, Atalla RH (1984) Studies of microstructure in native cellulose using solid-state 13 C NMR. Macromolecules 17:1465–1472 Wada M (2002) Lateral thermal expansion of cellulose Iβ and III1 polymorphs. J Polym Sci B 40:1095–1102 Wada M, Heux L, Isogai A, Nishiyama Y, Chanzy H, Sugiyama J (2001) Improved structural data of cellulose III1 prepared in supercritical ammonia. Macromolecules 34:1237–1243 Wada M, Kondo T, Okano T (2003) Thermally induced crystal transformation from cellulose Iα to Iβ. Polymer J 35:155–159 Wada M, Heux L, Sugiyama J (2004a) Polymorphism of cellulose I family: reinvestigation of cellulose IV. Biomacromolecules 5:1385–1391 Wada M, Chanzy H, Nishiyama Y, Langan P (2004b) Cellulose III1 crystal structure and hydrogen bonding by synchrotron X-ray and neutron fiber diffraction. Macromolecules 37:8548–8555 Wada M, Nishiyama Y, Langan P (2006) X-ray structure of ammonia-cellulose I: new insights into the conversion of cellulose I to cellulose III1. Macromolecules 39:2947–2952 Watanabe S, Takai M, Hayashi J (1968) An X-ray study on cellulose triacetate. J Polym Sci C 23:825–835 Whitaker PM, Nieduszynski IA, Atkins EDT (1974) Structural aspects of soda-cellulose II. Polymer 15:125–127 Woodcock C, Sarko A (1980) Packing analysis of carbohydrates and polysaccharides. 11. Molecular and crystal structure of native ramie cellulose. Macromolecules 13:1183–1187 Zugenmaier P (1974) Conformation and packing analysis of polysaccharides. I: mannan. Biopolymers 13:1127–1139 Zugenmaier P (2001) Conformation and packing of various crystalline cellulose fibers. Prog Polym Sci 26:1341–1417

Chapter 6

Cellulose Derivatives

6.1

Characterization

The three hydroxyls of the monomer units of cellulose can be substituted fully or in part. Generally, only regular derivatization along the chain leads to crystalline structures, and only these will be considered. Blockwise but regular derivatization of parts of the molecule can fulfill the proposed requirement of one-dimensional order along the chain molecule to a certain extent and an ordered three-dimensional structure appears in certain regions of the materials. Trisubstituted cellulose esters, ethers and carbamates are known to crystallize in various structures including regioselectively substituted molecules. Many of the derivatives can be drawn to form fibers and show excellent X-ray fiber patterns as demonstrated by Hess and Trogus in the 1920s and 1930s. These patterns might be classified according to Table 2.1 and categorized into various families depending on the fiber repeat. The kind of conformation is easily deduced from the fiber repeat of approximately 10, 15, 25 or 40 Å and, therefore, 2/1, 3/2, 5/3 and 8/5 helices can be assumed, since the monomeric repeat amounts to approximately 5 Å. These helices are all left-handed as determined by conformational analysis. A few reliable crystalline structures have been solved in detail and these will be discussed in the following sections but the fingerprint method allows the characterization of derivatives without knowing the molecular and crystal structure.

6.1.1

Cellulose Triacetate

Cellulose triacetate (CTA) is the product if all hydroxyls of the cellulose chain are substituted. A characteristic feature of CTA is its solubility in chloroform and its insolubility in the more common solvent acetone. Therefore, CTA is referred to as “chloroform-soluble acetate” or “primary acetate” in contrast to cellulose acetate mostly used in industrial processes. Cellulose acetate is obtained by deacetylation of CTA to a degree of substitution of approximately 2.5 and is soluble in acetone and, P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

175

176

6 Cellulose Derivatives

therefore, is termed “secondary acetate” or “acetone-soluble acetate.” The properties of cellulose acetate depend on the degree of substitution and the average distribution of the acetyl groups at the three possible sites as well as the distribution of the acetyl groups along the cellulose chain. Many of the industrially produced secondary acetates show X-ray diagrams similar to the CTA patterns but with less pronounced reflections. Two kinds of crystalline structures were established by Hess and Trogus (1929) for CTA: CTA I and CTA II. CTA I can be obtained by a so-called heterogeneous acetylation of native cellulose I without dissolving the materials at any time and CTA II can be obtained by a homogeneous acetylation, the cellulose acetate being completely dissolved in solution. Solid films are cast by evaporation of the solvent and fibers for X-ray diagrams are obtained by stretching these films. CTA II can also be produced by a heterogeneous reaction of cellulose II. The two polymorphs return to their original polymorphs, cellulose I and cellulose II, upon saponification (Sprague et al. 1958). These authors also found that heat treatment in the range from 210 to 280°C for a few minutes leads to high-quality X-ray diagrams of CTA I and CTA II. Figure 6.1 shows the X-ray intensity traces for randomly oriented CTA I and CTA II in comparison with oligomeric CTA and Fig. 6.2 shows the corresponding cross-polarization/magic angle spinning (CP/MAS) 13C NMR spectra (Kono et al. 1999; VanderHart et al. 1996). In Fig. 6.3 the Debye–Scherrer patterns of the two CTA polymorphs are represented recorded on a flat film detecting device and a pattern of a CTA– nitromethane (CTA-N) complex is also shown. This complex can be obtained by placing either CTA I or CTA II over nitromethane vapor. These conversions of fibers were studied by Kuppel et al. (1973) and the corresponding fiber patterns are shown in Fig. 6.4. Native ramie cellulose was heterogeneously acetylated to form the CTA I polymorph (Fig. 6.4a). Placing this fiber over nitromethane vapor gave a new structure, CTA-N (Fig. 6.4b). Polymeric single crystals of CTA-N were discovered by Manley (1963a, b) and single crystals with tetrachloroethane were discovered by Rånby and Noe (1961). The solvent is incorporated into the crystal lattice and both materials exhibit tetragonal-shaped single crystals in the electron microscope. However, in both investigations the authors did not recognize that these structures were complexed with solvent molecules and they obtained by exposure to X-rays the CTA II polymorph because the solvent included in the crystals evaporated during the exposure. Annealing the solvent-complexed fiber of Fig. 6.4b led to CTA II (Fig. 6.4c). When this CTA II fiber was placed over nitromethane again, the solvent-complexed structure was regained, but it was not possible to return to CTA I. This CTA I to CTA II transformation is interesting in many respects. First, a transformation occurs from CTA I to CTA II by leaving the orientation of the original ramie fiber intact. As will be discussed later, this transformation starts with parallel-packing arrangements of the chains and ends with an antiparallel chain arrangement. The packing arrangement of CTA-N has not been published but was determined to be antiparallel (Zugenmaier 2006) and will be discussed later. Second, the line width of the reflections changes. From a broad appearance for CTA I to a narrower one for CTA-N and to a broader one again for CTA II. Since the line width is directly inversely proportional to the crystallite size (Chap. 3, Fig. 3.2),

6.1 Characterization

177

Fig. 6.1 X-ray diffractograms of oligomeric cellulose triacetate (CTA) (degree of polymerization 2–6), CTA I and CTA II. (From Kono et al. 1999)

this means that CTA-N possesses the largest crystallite size, resulting from a more perfect packing arrangement compared with that of the starting and final fibers. It is interesting that the line width changes despite a constant orientation distribution. From a comparison of the Debye–Scherrer diagram of Fig. 6.3a for CTA I with the fiber pattern of Fig. 6.4a, it is apparent that between the first two reflections in Fig. 6.3a two more reflections appear on the comparable equator in Fig. 6.4a. These additional reflections have been proposed to belong to the CTA II polymorph (Roche et al. 1986), which might be created by the heat treatment necessary to obtain high-quality patterns of either CTA I or CTA II. But they might have a different origin as suggested by a series of annealing investigations by Sprague

178

6 Cellulose Derivatives

Fig. 6.2 Cross-polarization/magic angle spinning (CP/MAS) 13C NMR spectra of CTA I, CTA II and oligomeric CTA (degree of polymerization 2–6). Peak intensities of the methyl region are shown at 40% reduction. (From Kono et al. 1999)

et al. (1958). CTA I annealed at 250°C shows no additional reflections, at 275°C additional reflections appear but vanish on annealing at 290°C and a transformation to CTA II takes place when annealing occurs at 315°C. A metastable structure was suggested after heterogeneously derivatized ramie fibers were investigated by X-ray exposure without heat treatment as shown in Fig. 6.5b. This structure seems to be the active species used in chromatographic separation procedures of enantiomeric compounds as a stationary phase (Wolf et al. 1992). Reflections between the first two equatorial reflections of CTA I are clearly

6.1 Characterization

179

Fig. 6.3 Debye–Scherrer patterns of a CTA I, b CTA–nitromethane (CTA-N) and c CTA II. (a, b From Kuppel et al. 1972)

Fig. 6.4 Fiber patterns with the fiber axis diagonal of a CTA I, b CTA-N and c CTA II from a ramie fiber heterogeneously acetylated (a), placed over nitromethane vapor (b), and subsequent annealing (c). (From Kuppel et al. 1973)

visible compared with the diagram in Fig. 6.5a. This pattern was obtained by pulling fibers from a CTA liquid-crystalline dope in trifluoroacetic acid/H2O (100:8 w/w; Roche et al. 1986) and clearly represents a pattern comparable to the CTA I fiber diagram. Here, a parallel-packing arrangement is obtained from dissolved materials, which is a rare observation.

6.1.2 Experimental Data for Cellulose Tripropionate and Cellulose Acetate Dipropionate and Further Cellulose Esters Excellent fiber X-ray diffractograms (Shuto et al. 1986, 1989a, b; Shuto 1990; Zugenmaier 1983; Eggert 1985) and electron diffraction patterns (Shuto et al. 1987, 1989 a, b; Shuto 1990) have been published for cellulose tripropionate (CTP) as have IR (Fig. A.9) and NMR spectra (Shuto 1990) and conformation and packing models have been proposed. However, these models do not obey the rules established from oligomeric compounds and too many close-packing contacts are

180

6 Cellulose Derivatives

Fig. 6.5 Fiber X-ray diagrams: a for CTA (degree of polymerization 300) produced from a liquidcrystalline dope (trifluoroacetic acid/H2O, 100:8 w/w) by film-casting and hand-pulling (note the close similarity with a CTA I polymorph); b for CTA I (origin native ramie fiber) as obtained from a heterogeneous derivatization reaction without annealing (a From Roche et al. 1986)

Fig. 6.6 Cellulose tripropionate (CTP): a X-ray fiber diffraction pattern from a fiber annealed at 205°C and b electron diffraction pattern of a single polymeric crystal (see Fig. 6.8) grown at 210°C from a mixture of dibenzyl ether and n-tetradecane (15:85 v/v); CTP concentration 0.02%, crystallization time 3 h. (From Shuto et al. 1989b)

present in the structures. Therefore, the coordinates have been omitted but the X-ray fiber patterns of the two different types of structures as well as an electron and a Debye–Scherrer diffraction pattern are presented in Figs. 6.6 and 6.7. A comparison of the two fiber diffraction diagrams, which are obtained at different annealing temperatures, shows a different overall intensity distribution and differences on the equator. In Fig. 6.7 (left), the fiber axis of the sample is tilted at an angle corresponding to the third meridional reflection to verify this faint

6.1 Characterization

181

Fig. 6.7 CTP annealed at 223°C for approximately 25 min: X-ray fiber diffraction pattern with calibration ring of CaF2 (left) and Debye–Scherer pattern of CTP (right). (From Eggert 1985)

reflection. This procedure is necessary since all meridional reflections lie outside the reflecting range of the X-ray fiber technique and are missing. These reflections have to be brought into a reflecting position by a tilt of the fiber. Sometimes this occurs internally, which means some fibrils of the fiber are not perfectly oriented and lie in a reflecting position. These few fibrils contribute to the intensity of the whole fiber to a certain extent but the intensities cannot be compared with those of the materials contributing fully. In addition to the faint third meridional reflection, Shuto (1990) has shown that both fiber patterns, representing the two different structures, exhibit a strong sixth-order meridional reflection. According to the discussion in Sect. 3.1, a threefold helical structure can be proposed, but a sixfold helix with some structural disorder cannot be excluded. Deviations or defects in the crystal structures are often observed for cellulose and its derivatives. The determination of the exact symmetry of a helix might be difficult and superimposed by space group symmetry elements as observed for the CTA-N complex. An eightfold helix is the most probable structure and will be discussed in Sect. 6.2. Space group symmetry elements are actually fourfold and twofold symmetry axes for this structure and a square (tetragonal) shape of the single crystals is observed in electron micrographs. The shape of single crystals of CTP (Fig. 6.8) as well as the electron diffraction pattern (Fig. 6.6b) do not show any trigonal or hexagonal symmetry elements, but rather show elements of an orthorhombic or monoclinic crystal. Rectangular angles in the electron diffraction pattern as well as in the electron micrograph are observed and an unusual monoclinic two-chain space group for polymers was proposed with the monoclinic angle b between the chain axis and the base plane (Table 6.1). In contrast to a threefold helix, a sixfold one also contains twofold symmetry elements. The CP/MAS 13C NMR spectrum of highly crystalline CTP shows a splitting of the C1 signal (Fig. 6.9), which points towards at least two different residues in the asymmetric unit in contrast to the present model.

182

6 Cellulose Derivatives

Fig. 6.8 Electron micrograph of single crystals of CTP grown at 205°C from a mixture of dibenzyl ether and n-tetradecane (20:80 v/v); CTP concentration 0.01%, crystallization time 4 h. (From Shuto et al. 1987) Table 6.1 Unit cells of various cellulose esters and proposed space groups Compound a (Å) b (Å) c (Å) γ (°) Space group References Cellulose triacetate I Cellulose triacetate II

5.94 24.68

11.43 11.52

10.46 10.44

95.4 90

P21 P212121

Sikorski et al. (2004) Roche et al. (1978), Zugenmaier, (unpublished results 2005) Cellulose tripropionate 11.76 15.31 15.14 106.0a P21 Shuto 1990 Cellulose tributyrate 25.37 15.34 10.33 90 P212121 Shuto et al. (1989b) Cellulose trivalerate 29.40 15.40 10.36 90.0 P21 Shuto et al. (1989b) CDAP 24.98 12.39 10.44 90 P212121 Iwata et al. (1996b) CADP 10.88 15.93 15.09 94.1a P21 Iwata et al. (1996c) CDAB 21.47 15.44 10.62 90 P212121 Iwata et al. (1996a) Cellulose tribenzoate 12.20 12.20 15.21 120 P32 Steinmeier (1988) Cellulose trinitrate 9.00 14.60 25.40 90 Orthorhomic Meader et al. (1978) CDAP 2,3-di-O-acetyl-6-O-propanoyl cellulose, CADP 6-O-acetyl-2,3-di-O-propanoyl cellulose, CDAB 2,3-di-O-acetyl-6-O-butyryl cellulose a Monoclinic angle b

Comparable electron micrographs and electron diffraction patterns were obtained for the mixed esters 6-O-acetyl-2,3-di-O-propanoyl cellulose (CADP) and 6-O-acetyl-2,3-di-O-butyryl cellulose, which are shown in Fig. 6.10 (Iwata et al. 1994). The X-ray fiber pattern of CADP resembles in the intensity distribution very much the diffractogram of CTP (Fig. 6.6) but is of poorer quality (Shuto et al. 1996b). The known sizes of unit cells and space groups of crystalline cellulose esters are listed in Table 6.1.

6.1 Characterization

183

Fig. 6.9 CP/MAS 13C NMR spectrum of highly crystalline CTP. (From Shuto 1990)

Fig. 6.10 Single crystals and corresponding electron diffraction patterns of cellulose heteroesters. Left: 6-O-Acetyl-2,3-di-O-propanoyl cellulose; right: 6-O-acetyl-2,3-di-O-butyryl cellulose. (From Iwata et al. 1994)

In contrast, the shape of the single crystals of 2,3-di-O-acetyl-6-O-propanoyl cellulose (CDAP) as well as the intensity distribution of the X-ray diffractograms and the symmetry elements of the electron diffraction pattern resemble those obtained for CTA II (Sect. 6.2). The same space group P212121 was assigned to this crystal structure (Table 6.1), and will be discussed in Sect. 6.3. Further, the electron diffraction pattern of 2,3-di-O-acetyl-6-O-butyryl cellulose (CDAB; Iwata et al. 1996a) also resembles that of CTA II as well as those of CDAP in terms of symmetry and appearance. Evaluation of the X-ray fiber diffraction pattern resulted in the size and the symmetry of this structure in three dimensions still being found to be

184

6 Cellulose Derivatives

similar to those of CTA II and the comparable unit cell is listed in Table 6.1. However, the electron micrograph and diffraction pattern of the single crystal of CDAB in the paper of Iwata et al. (1994) do not match the shape expected for the given space group nor does the symmetry of the electron diffraction pattern (compare with CTA II and CDAP in Figs. 6.19, 6.24). The X-ray diffraction patterns of cellulose tributyrate (CTB; Shuto et al. 1989b, Eggert 1985) are presented in Fig. 6.11 and the electron diffraction diagram is shown in Fig. 6.12, which compares with that of CTA II regarding the symmetry elements. Similarly, the reflections can be indexed by an orthorhombic unit cell and space group P212121 was assigned (Table 6.1). From these similarities the arrangement of chains can be predicted to resemble the arrangements of CTA II, which will be discussed in Chap. 6.2. In contrast, the fiber X-ray pattern of cellulose trivalerate differs from those of CTA II and CTB and is shown in Fig. 6.13 (Shuto et al. 1989b; Iwata et al. 1996d). The proposed unit cell is listed in Table 6.1.

Fig. 6.11 Debye–Scherrer diagram (left picture) and fiber X-ray pattern (right picture) of cellulose tributyrate CTB. (From Eggert 1985)

Fig. 6.12 Electron diffraction pattern of cellulose tributyrate. (From Shuto et al. 1989b)

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N

185

Fig. 6.13 X-ray fiber diffraction pattern of cellulose trivalerate. (From Shuto et al. 1989b)

Immersing native cellulose in aqueous solution of mixtures of nitric, acetic and phosphoric acid provides a series of quite different X-ray diffraction patterns of which one with 13.9% nitrogen by elemental analysis was investigated by Meader et al. (1978). The nitration process with this method stops between dinitrate and trinitrate (14.14% nitrogen). A meridional reflection of the X-ray fiber pattern is observed on the fifth layer line with a 25.4-Å layer line spacing. These data suggest a fivefold helical chain, which was considered as a 5/2 righthanded helix packed in antiparallel fashion. Unique unit-cell dimensions are difficult to obtain from the poorly resolved X-ray diagram. A more sophisticated model-building procedure led Zugenmaier (1983) to propose a left-handed fivefold helix (5/3), which was later confirmed by French et al. (2002). This observation supports the general idea that all helical cellulosic conformations are left-handed with a rise per residue of about 5 Å. This fivefold helical structure of cellulose trinitrate led Watanabe et al. (1968) to propose a bent and twisted conformation of the glucose ring because in their modeling approach the ring was too stiff to accommodate this kind of helix. Modern computer modeling can accommodate a fivefold helix with a 25.4-Å repeat distance along the chain without constraining the glucopyranose residue.

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N CTA plays a significant role in applications and, therefore, its detailed structure is of great interest. In particular, a renewed curiosity for structural features of CTA arose from the observation that enantiomeric small molecular compounds can be separated chromatographically with heterogeneously produced microcrystalline CTA I as a stationary phase but not with the homogeneously derivatized crystalline form CTA II.

186

6 Cellulose Derivatives

The first crystal and molecular structure of CTA I was reported by Stipanovic and Sarko (1978), who proposed parallel-arranged cellulosic chains running through a two-chain orthorhombic unit cell. Their model showed that the two chains in theunit cell are almost identical in conformation and are related by almost translational symmetry. Nevertheless, they relied on the two-chain structure because indexing the X-ray pattern required the proposed unit-cell size. A one-chain unit cell and space group P21 is presently favored by improved X-ray analysis (Sikorski et al. 2004; starting material green alga Cladophora sp.) and an evaluation of NMR spectra (Fig. 6.2). The symmetry elements of the packing arrangement agree with the space group assignment of a one-chain unit cell. The proposed model is illustrated in two projections in Figs. 6.14 and 6.15. Cartesian and fractional coordinates of the structure are listed in Table A.22. One very short packing contact is detected between O33C..C63M (0,−1,0) of 2.88 Å (Fig. 6.15). A further attempt at structure determination of CTA I was undertaken by Wolf et al. (1992) by pure computer

Fig. 6.14 Projection of the sheet-like structures of CTA I chains (coordinates from Sikorski et al. 2004) in the [001] direction on the a–b plane. A plane (intrasheet) can be defined passing through the origin in diagonal direction [110]

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N

187

Fig. 6.15 Projection of a diagonal intrasheet of CTA I on the b–c plane

modeling. In this approach, the packing arrangement and the unit cell were deduced from low-energy data. The unit cells of the proposed crystalline CTA I structures of various authors are listed in Table 6.2 as are those of CTA II and CTA-N. Table 6.3 represents some selected torsion angles of the pendant acetyl groups and F and Y values as well as the glycosidic bridge angle t of the various CTA models. These values have to match comparable ones determined for the low molecular weight compounds within a reasonable range (Sect. 4.4). A first structural approach to CTA II was started by Dulmage (1957). An extensive experimental investigation followed by Sprague et al. (1958), resulting in the suggestion of a monoclinic unit cell being abandoned today. Dulmage based his considerations on the orthorhombic space group P212121 despite a weak (100) reflection and established the conformation and packing of a model by a balland-stick approach. Owing to the proposed space group, the chains run in antiparallel fashion. A qualitative discussion of the intensity distribution of the X-ray pattern (Fig. 6.16) led to the positioning of the chains in the unit cell and this model was confirmed by an electron density map in a–b projection. However, a 21 screw axis along the chain was excluded owing to too close contacts of the primary acetyls. A monoclinic unit cell with space group P21 was also discussed with random distribution of chain directions but was not further evaluated. Discrepancies are found for the calculated densities of CTA I (Table 6.2) but not for CTA II derived from the size and content of the unit cells and the experimentally determined values. The experimental density of CTA I (Stipanovic and Sarko 1978) and CTA II (Dulmage 1957) was determined to be 1.29 g cm−3 for both materials. This value is only slightly larger than the value of 1.28 g cm−3 measured for a nonoriented amorphous CTA sheet. An extensive experimental study on densities of cellulose esters by Malm et al. (1947) agrees with the experimental findings of

188

6 Cellulose Derivatives

Table 6.2 Unit cells of various polymorphs of cellulose triacetate (CTA). Parallel chain arrangements for all CTA I structures, antiparallel chain arrangements for CTA II and CTA–nitromethane (CTA-N) Space Materials a (Å) b (Å) c (Å) a (°) b (°) g (°) r (g cm−3) group References CTA I

23.63

6.27

10.43 90

90

CTA Ia CTA I

11.82 11.27

6.27 6.00

10.43 90 10.47 90

90 90

90

1.24

P21

~90 1.24 86.15 1.34

P21 P21

Stipanovic and Sarko (1978)

Roche and O’Brien (1985) CTA I 5.94 11.43 10.46 90 90 95.4 1.375 P21 Sikorski et al. (2004) CTA Ib 24.31 5.15 10.53 90.1 90.0 83.9 1.46 P21 Wolf et al. (1992) CTA II 24.68 11.52 10.54 90 90 90 1.29 P212121 Dulmage (1957), Roche et al. (1978) CTA-N 15.01 15.01 41.14 90 90 90 1.18c P41212 Zugenmaier, (unpublished results 2005), Kuppel et al. (1972) a Omitting the broad (300) reflection in the data of Stipanovic and Sarko (1978) b Computer modeling only c Assuming two nitromethane molecules per residue

Dulmage for CTA (1.29 g cm−3) but gives a value of 1.30 g cm−3 for 2.5 acetate, which still exhibits the X-ray pattern of CTA II. A further increase in density was observed for lower-substituted cellulose. The calculated density for CTA II of 1.28–1.30 g cm−3 (see also Table 6.2) using the size of the unit cell agrees with the experimental data within the estimated errors. However, the calculated densities of CTA I from 1.24 to 1.46 g cm−3 (Table 6.2), with 1.375 g cm−3 for the newly established one-chain structure (Sikorski et al. 2004), cannot be explained. Progress in single crystal structure determination as outlined in Chap. 4 and in model simulation with the use of computers triggered a refinement of the crystalline structure of CTA II by Roche et al. (1978). The central triacetyl glucose residue of cellotriose undecaacetate appeared to be a good but fixed starting residue to generate a model of CTA II in space group P212121. An approximately 21 screw axis was assumed for the chain skeleton conformation with a deviating C61C (carbonyl group at O61) torsion angle (Table 6.3) from this symmetry element. This conformation agrees with the proposal of Dulmage as does the general position of the antiparallel-running chains in the unit cell. The hydrogen atoms were omitted in this study but when they are introduced they lead to unreasonable short van der Waals contacts. A recently performed multidimensional refinement with all hydrogen included for CTA II and using the fixed minimum conformation of CTA I with

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N

189

Table 6.3 Selected torsion angles (in degrees) for the description of the rotational position of the carbonyl side-groups. The torsion angles F and Y express the relative position of two adjacent residues of CTA I and CTA II chains. Only the pendant groups at C6 of CTA II (Roche et al. 1978) do not follow a 21 screw axis. The two chains in the model of Stipanovie and Sarko (1978) are identical with a 21 screw axis each. The residue adjacent to r along the chain is termed r′ (r′ = r + 2) CTA II CTA I CTA I (Zugenmaier, (Stipanovic CTA II (Sikorski unpublished and Sarko (Roche Compound et al. 2004) results 2005) 1978) et al. 1978) Residue r 1 Torsion angle C1r–C2r– 145.3 O2r–C2rC C2r–C3r– −101.2 O3r–C3rC C5r–C6r– 165.6 O6r–C6rC H2r–C2r.. 28.2 C2rC–O2rC H3r–C3r.. 10.7 C3rC–O3rC H5r–C5r.. −7.2 C6rC–O6rC c′(C4r–C5r– 57.9 C6r–O6r) Residue r→r′ 1→3 F(O5r–C1r– −99.1 O4r′-C4r′) Y(C1r–O4r′– 95.6 C4r′-C3r′) Glycosidic bridge angle t(C1r–O4r′– 118.8 C4r′)

1

1

2

1

3

145.3

141.8

141.8

130.4

129.3

−101.3

−94.5

−94.6

−119.4

−110.4

165.6

137.8

137.8

134.9

64.6

27.6

12.3

12.3

15.6

15.9

10.9

2.2

2.2

−5.2

−4.7

−7.2

−13.5

−13.4

−34.9

−97.7

57.9

68.0

68.0

42.0

32.3

1→3 −98.3

1→3 −97.4

2→4 −97.6

1→3 −106.0

3→5 −107.2

94.7

94.8

95.1

99.6

100.2

119.0

118.4

118.3

118.6

119.9

an exact 21 screw axis of the chains (data of Sikorski et al. 2004 for CTA I) resulted in a model which is represented in Figs. 6.17 and 6.18. An experimental fiber repeat of 10.44 Å was introduced (Zugenmaier, unpublished results 2005), which compares well with the value of Dulmage (10.43 Å) but is somewhat shorter than the value of Roche et al. (10.54 Å). The coordinates of this model are listed in Table A.23. This new model was obtained by placing the fixed monomeric unit of CTA I on a somewhat smaller fiber repeat and moving one fixed chain with regard to best packing within the given four symmetry-related chains of the unit cell (Figs. 6.17, 6.18). Space group P212121 only requires a cellobiose unit as an asymmetric unit but the additional constraint of a 21 screw axis with a glucose residue as the basic unit seems to be a good first approximation and leaves small changes in the conformation open

190

6 Cellulose Derivatives

Fig. 6.16 X-ray fiber diffraction pattern of CTA II

Fig. 6.17 Representation of crystalline CTA II projected in the [001] direction on the a–b plane. The chain in the light drawing represents an additional molecule to the four ones required by the symmetry elements of space group P212121

The two adjacent parallel chains on the a-axis in Fig. 6.17 and a translated one by a shift in the b-direction represented in the light drawing resemble very much the placement of the chains of the CTA I structure in Fig. 6.14. The distances between the cellulosic chain axes in consideration for the two polymorphs match

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N

191

Fig. 6.18 The conformation and packing of two parallel chains placed edge on in Fig. 6.17. Note a pure translational shift between the molecules owing to an assumed 21 screw axis along the chains

up well. A short distance of 6.04 Å is found between the two symmetry-related CTA II chains on the a-axis compared with a distance of 5.94 Å along the a-axis for CTA I. Actually, the two chains for CTA II on the a-axis are related by a translation, i.e., a shift in the direction connecting their chain axes caused by the assumed 21 screw axis along the cellulose chain axes. Two neighboring chains in the intrasheet of CTA I (diagonal of unit cell) are 12.4 Å apart and are 13 Å apart in the two corresponding molecules placed edge on for CTA II in Fig. 6.17. The two symmetry-related (21 screw axis and translation) up chains by translation on the a-axis can be shifted along b, representing nearly the one-chain unit cell of CTA I and regarded as occupying approximately half of the unit cell of CTA II. The other half of the unit cell is occupied by the down chains created by the 21 symmetry element of the space group applied to the up chains and interlocked with the strip of up chains (Fig. 6.17). Therefore, a strong correlation exists in part between CTA I and CTA II for the packing arrangement and consequently for the unit cells. The unit-cell dimension b of CTA II corresponds to b in CTA I and the distance of the two symmetry-related molecules on a of CTA II corresponds to the a-dimension of CTA I. Small deviations are expected in the twist of the chains and the monoclinic angle resulting also from minor shifts of the chain axes from the line of the a-axis. Nevertheless, the model proposed for CTA II has to be regarded as preliminary since it was not refined against X-ray intensities. One short packing contact remains with this fixed conformation between O43..H21C (methyl hydrogen of the carbonyl group at O21; see Fig. 6.15, Table A.23 for nomenclature) of 2.19 Å, which might be released by introducing a more sophisticated model with more parameters

192

6 Cellulose Derivatives

Table 6.4 Solid-state 13C NMR spectral data (d, ppm) for CTA I, CTA II and the hexamer. (From Kono et al. 1999) Carbonyl Ring carbon Compound carbon C1 C2–C5 C6 Methyl carbon CTA I

170.9 172.2

103.2

CTA II

169.8 170.6 172.6 172.9

101.8

Hexamer

170.9 173.0

103.2

80.6 76.3 72.9 80.8 78.2 76.0 75.0 73.2 80.6 76.3 73.0

62.7

23.2 22.3

65.8 61.9

21.8

62.6

23.2 22.3

refined. It should also be remarked that a multidimensional refinement of the primary acetyl groups with the geometrical data of the chain proposed by Roche et al. tends to symmetry-related sites of the primary acetate groups along the cellulosic chain. The symmetry of the unit cell does not require a 21 axis along each cellulosic chain rather than only a 21 screw axis between neighboring chains, which means that a cellobiose unit represents the basic unit. Therefore, the relative positions (F, Y values) between adjacent residues in a chain may slightly change along the chain as may the conformation of the two glucose residues as required by NMR investigations (Fig. 6.2, Table 6.4). The coordinates listed should be considered as averaged. The critical contact distances between O5..O3′ and H1..H4′ compare well with the low values of the acetylated model compounds. In Table 6.3 the torsion angles of the pendant acetyl groups are compared for various models for CTA I and CTA II. Considerable deviations of the acetyl torsion angles are found for the two CTA II conformations listed as well as a difference in chain shift along the fiber axis of about 0.8 Å and some shifts of chain sites in the a–b plane. Nevertheless, all these angles for CTA I and CTA II are close to those established for oligomeric structures and demonstrate that essentially the proposed cellulosic chain follows a 21 screw axis in contrast to the primary acetyl group in the model of Dulmage and Roche et al. All the pendant acetyl groups are planar and the carbonyl bond C = O is placed almost cis to the corresponding C–H bond of the pyranose ring. The glycosidic bridge angle is somewhat on the high side for all CTA I and CTA II model chains. A one-chain unit cell as proposed for CTA I establishes parallel chain arrangements in contrast to antiparallel arrangements for CTA II with four chains running through the unit cell. Nevertheless, a conversion of parallel chains in CTA I to antiparallel chains in CTA II can be achieved by maintaining the orientation of a fiber as shown by the experimental data in Fig. 6.4. From this investigation it can be concluded that

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N

193

Fig. 6.19 Shape of a polymeric single crystal of CTA II viewed through an electron microscope (a) and placement of the CTA chains on the crystallographic a–b plane (b). (From Roche et al. 1978) The observed electron diffraction pattern of a tip of a single crystal of CTA II (insert) is shown in the a–b plane in c. (Chanzy et al. 1973)

the transformation of a parallel to an antiparallel chain arrangement occurs without chain folding. Chain folding was an idea followed up with great verve in the 1960s and 1970s. This negative finding does not generally exclude that chain folding is possible under special experimental conditions, e.g., long chains crystallizing from solution. Additional information on structure and morphology may be obtained by electron-microscopic studies. Figure 6.19 shows a polymeric single crystals of micron size of CTA II, the chain arrangement within the projected morphology and the electron diffraction pattern supporting the symmetry elements and the size of the CTA II unit cell projected onto the a–b plane. The placement of the chains within the unit cell is supported by an evaluation of the electron densities from diffraction patterns and is shown in Fig. 6.20. In contrast, the growth of single crystals of quadratic or tetragonal shape is observed in a mixture of a polar solvent and a nonsolvent such as nitromethane (Manley 1963a, b) or tetrachloroethane (Rånby and Noe 1961) and n-butanol as shown in Fig. 6.21. First, they were identified as CTA II structures until it was recognized that they contained solvent (e.g., nitromethane), which evaporated during the investigations. By careful preparation to retain the solvent within the crystal lattice, the electron diffraction pattern (Fig. 6.21, right) differs from that of CTA II

194

6 Cellulose Derivatives

Fig. 6.20 Electron density in the a–b plane of CTA II by evaluation of the diffraction pattern. Note the placements of the chains are clearly expressed. (From Dorset 1995)

Fig. 6.21 Electron micrograph (left) and electron diffraction pattern (right) of single crystals of CTA-N. (Left: from Manley 1963a; right: from Chanzy et al. 1971)

(Fig. 6.19c) as does the fiber X-ray pattern in Fig. 6.22 compared with that in Fig. 6.16. Evaluation of the X-ray fiber pattern of this crystalline cellulose acetate solvent complex with nitromethane (CTA-N) reveals a meridional reflection on the eighth layer line, suggesting an eightfold helix. Some innermost reflections on higher-layer lines, closer to the meridian than those on the equator, lead to the assignment of the reciprocal axis of the unit cell in diagonal directions in Fig. 6.21 (right). With this assignment it is observed that every second reflection is missing on the a-axis and the b-axis. The tetragonal space groups P43212 and P41212 seem to be the most probable ones and a unit cell with a = b = 15.01 Å, c(fiber axis) = 41.14 Å was determined. A decision between the two space group possibilities can only be reached by a full evaluation of the packing arrangements of the molecules, for which the shape (conformation) of the chain is a prerequisite.

6.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N

195

Fig. 6.22 X-ray fiber diagram of CTA-N. Note that the innermost reflections on higher-layer lines require a unit cell the axes of which lie in a diagonal direction for the electron diffraction pattern in Fig. 6.21

The conformation of CTA-N as a left-handed 8/5 helix was proposed some time ago (Zugenmaier 1986), but it was not until recently that space group P41212 and a packing model was introduced. The monomeric model compound methyl tetraacetyl-b-d-glucoside (Zugenmaier and Rappenecker 1978; coordinates in Table A.8) was found to be suitable for evaluating an 8/5 left-handed helix as a preliminary conformation and packed in the appropriate space group P41212. The dimer serves as a basic unit and only the torsion angles of O6 and C6C have to be altered by a few degrees to obtain a reasonable model (Fig. 6.23). No too short contacts have been encountered. A complete structural model with the nitromethane included and refinement against X-ray data has not been completed. The torsion angles Y = 127.9° and F = −80.2° describing the relative placement of the residues of the helix as well the glycosidic bridge angle t = 116.3° agree well with residues 2→3 of b-cellotriose undecaacetate and 1→2 of b-d-acetyl cellobiose (Sect. 4.4). Nevertheless, from the data available some remarkable conclusions can be drawn: 1. The symmetry requirements of the space group demand an antiparallel packing of adjacent chains for CTA-N. The change from parallel-arranged chains CTA I to antiparallel ones of CTA II during the transformation from CTA I to CTA-N to CTA II as depicted in Fig. 6.4 occurs at the conversion from CTA I

196

6 Cellulose Derivatives

Fig. 6.23 Representation of the crystalline CTA-N structure in two projections. Left: Projection in the [001] direction on the a–b plane. Right: Approximately in diagonal direction [110]. Note the antiparallel packing of the two chains. Nitromethane is not included

to CTA-N. The arrangement of chains in CTA-N supports the antiparallelpacked chains for CTA II and confirms the experimental evidence for the irreversible transformation of CTA I to CTA-N and for the transformation from CTA-N to CTA II being reversible. 2. Holes in the center and the corners of the unit cell in Fig. 6.23 provide the space for the placement of the nitromethane molecules. 3. According to space group symmetry only a fourfold helix is required. The suggested 8/5 (left-handed) helix with a basic monomeric residue can be modeled as a 4/1 (right-handed) helix with a dimeric residue (cellobiose) as a basic unit. The 8/5 helix is justified by an eighth meridional reflection, missing a fourth-order meridional one. The tetragonal space group P41212 accommodates the left-handed eightfold helix. The helix axes of the opposite-running chains are placed on (1/2, 0) and (0,1/2), respectively, i.e., on the middle of the a-axis and the b-axis in Fig. 6.23. The preliminary coordinates of CTA-N including all hydrogen atoms for a dimer, which describes the content of the complete unit cell, are collected in Table A.24 (nitromethane omitted).

6.3 Conformation and Packing Arrangement of CDAP

197

Fig. 6.24 X-ray fiber pattern (left), electron micrograph (right) and electron diffraction pattern (insert) of cellulose-2,3-di-O acetyl-6-O-propanoyl (CDAP). (From Iwata et al. 1996b)

6.3

Conformation and Packing Arrangement of CDAP

The regioselective derivatzation and the characterization of cellulose-2,3-diO-acetyl-6-O-propanoyl (CDAP) were described by Iwata et al. (1996b). The fiber X-ray pattern obtained by stretching a film cast from a chloroform solution is shown in Fig. 6.24 (left) as are an electron micrograph and an electron diffraction diagram from single polymeric crystals are shown. The appearance of the X-ray pattern as well as the shape of the single crystals and the corresponding electron diffractogram resemble very much those of CTA II (Figs. 6.16, 6.19). The same orthorhombic space group P212121 as for CTA II was proposed containing four chains. The unit cell with a=24.98 Å, b=12.39 Å and c(fiber axis)=10.44 Å is very much comparable with that of CTA II (Table 6.1), with a 0.9 Å longer a-dimension to accommodate the bigger propanoyl group at O6. The experimental density was determined to be 1.239 g cm−3; the calculated density was 1.241 g cm−3. The proposed model (Iwata 1996b) does not agree with the strict proposed requirements of cellulosic conformation and packing arrangements but the overall positioning of the chain molecules in the unit cell is correct. A simple revaluation of the conformation and packing by modeling procedures with the starting set of CTA II led to the coordinates listed in Table A.25. As for CTA II, a twofold screw axis of the cellulosic chain was assumed, which is not a necessity for the proposed space group. The evaluations of conformation and packing arrangements with these data are illustrated in Figs. 6.25 and 6.26 and compare well with the model developed for CTA II.

198

6 Cellulose Derivatives

Fig. 6.25 Crystalline CDAP. Projection in the [001] direction on the a–b plane. The chain in the light drawing represents an additional molecule to the four ones required by the symmetry elements of space group P212121

Fig. 6.26 The conformation and packing of two parallel chains placed edge on of CDAP in Fig. 6.25; projection in the [−110] direction. Note a pure translational shift between the molecules owing to an assumed 21 screw axis along the chains

6.4 Conformation and Packing Arrangement of Cellulose Tribenzoate

199

6.4 Conformation and Packing Arrangement of Cellulose Tribenzoate Cellulose tribenzoate (CTBe) was obtained in two different procedures by substitution in heterogeneous or homogeneous reaction. These two procedures led to a different crystal structure in CTA but resulted in a similar X-ray fiber pattern and unit-cell dimensions for CTBe (Steinmeier and Zugenmaier 1987; Steinmeier 1988; Zugenmaier 1983). An excellent X-ray fiber pattern of homogeneously derivatized CTBe termed CTBe II as shown in Fig. 6.27 led to the following conclusions. The pattern can be indexed with a trigonal, one-chain unit cell as for the heterogeneously derivatized CTBe I. The appearance of a third-order meridional reflection and a conformational analysis suggest a threefold left-handed

Fig. 6.27 X-ray fiber pattern of cellulose tribenzoate II

Fig. 6.28 Projection of left-handed threefold helices (3/2) in the trigonal space group P32 in the [001] direction on the a–b plane (left) and two molecules projected in the [010] direction to show the conformation (right)

200

6 Cellulose Derivatives

helical chain. Consequently, the chains have to be packed in parallel fashion for a one-chain unit cell, which is unusual for a structure crystallized from solution. An extensive investigation with consideration of several larger orthorhombic unit cells and comparison of the best conformation and packing as well as agreement with X-ray data resulted in parallel-packing arrangements in a one-chain unit cell of space group P32 (Riehl 1992; Zugenmaier 1995; coordinates in Table A.26). All reflections in the fiber X-ray pattern of CTBe I and CTBe II can be indexed with this one-chain unit cell. Figure 6.28 shows the conformation and parallel-packing arrangements in two projections. The parallel-packing arrangement of CTBe I is further supported by the fact that the X-ray fiber pattern of the original ramie cellulose was recovered after saponification of CTBe I (Steinmeier 1988). Since both structures, CTBe I and CTBe II, have comparable X-ray patterns, parallel packing is proposed for CTBe I and CTBe II.

6.5 Trimethyl Cellulose and 6-O-Acetyl-2,3-di-O-methyl Cellulose Trimethyl cellulose (TMC) was synthesized from ramie fibers by a heterogeneous process and an orthorhombic unit cell was determined by Trogus and Hess (1929). They found for this TMC I modification a= 21.3 Å, b=25.6 Å, c(fiber axis) =11.3 Å and reported an experimental density of 1.27 g cm−3. The fiber repeat was later corrected to c=10.3 Å by von Susich (Meyer and Mark 1930). A second crystalline modification, TMC II, was obtained by stretching a film produced from a TMC solution by evaporation of the solvent with an orthorhombic unit cell, a= 45.3 Å, b= 4.58 Å, c(fiber axis) = 10.4 Å (Zugenmaier and Kuppel 1986), and space group P21221 that

Fig. 6.29 X-ray Debye–Scherrer pattern of collected polymer single crystals (left) and of a powder (right) of trimethyl cellulose II (TMC II)

6.5 Trimethyl Cellulose and 6-O-Acetyl-2,3-di-O-methyl Cellulose

201

is now proposed. This elongated unit cell could only be established through electronmicroscopic investigations by growing single crystals and performing electron diffraction studies (Fig. 3.9). A first attempt to explain the unusual electron diffraction pattern with remarkable repeated extinction was undertaken by Zugenmaier and Kuppel (1986) and some general conclusions were drawn on the packing of the chains. Figure 6.29 shows the Debye–Scherrer patterns on a flat film for a powder of TMC II crystals in comparison with a sample of randomly oriented polymer single crystals. Although the reflection angles compare quite well, if the same sample-to-film distance is taken into account, the line width is much broader for the single-crystal materials, indicating smaller crystallite sizes in contrast to the randomly oriented TMC II sample. The fiber X-ray pattern of an oriented sample of TMC II shows relatively few reflections (Fig. 6.30). First- and third-order meridional reflections are observed despite the fact of a 10.4-Å fiber repeat suggesting an approximate twofold fiber axis as in most cellulosic structures. An exact two-fold axis of each chain in the unit cell is not required by space group P21221, since the 21 screw axis lies between the chains but, as for CTA II, it seems to be a good approximation. Traces of biaxiality are also observed, which means that the intensities on the film are not the same if the fiber in the diffraction experiment is rotated to various positions around its long axis. An idealized structure with a 21 fiber axis was assumed in the evaluation of the structure as in many cases for cellulosics. Further polymorphs have been found and some of them have been illustrated by X-ray fiber patterns (Zugenmaier 1983). It is likely that the formation of these structures depends on the kind of methylation procedure applied. Using partially methylated products from industry as starting materials and by trimethylation of

Fig. 6.30 Fiber X-ray diffraction pattern of a stretched TMC II film. Note that the fiber might be biaxially oriented to a certain extent

202

6 Cellulose Derivatives

Fig. 6.31 The crystal and molecular structure of TMC II in space group P21221. a Projection in the [010] direction on the a–c plane. Note the presence of a translational shift in the a-direction of the up as well as the down chains in this projection. b Projection in the [001] direction on the a–b plane

these specimen, one can obtain crystalline polymorphs, which are scarcely found by trimethylation of native cellulose. The conclusions drawn by Zugenmaier and Kuppel led to two different enantiomeric space groups (Table 3.1) being considered: C2 if the unit cell is monoclinic, leading to a parallel-packing arrangement, and P21221 for an orthorhombic one but with antiparallel-packed chains. A conformation and packing analysis led to the proposal of P21221. The coordinates obtained by a refinement procedure are listed in Table A.27 and models assuming a 21 screw axis as an approximation along the cellulosic chains are drawn in two projections in Fig. 6.31. Excellent X-ray and electron diffraction patterns have also been obtained for another cellulose triether, triethyl cellulose, with trigonal symmetry elements and a fiber repeat of approximately 15 Å. It is rather difficult to establish a 3/2 helical model for the chain conformation which also meets the packing requirements. The electron diffraction of some polymer single crystals suggests a one-chain trigonal unit cell, while other diffraction patterns suggest a larger unit cell as supported by additional reflections. The X-ray fiber pattern seems to be consistent with an overlap of two crystal structures, one of which resembles the one-chain structure observed by the electron diffraction and the second seems to point towards a frustrated structure as proposed by Cartier et al. (1996) with a random selection of up and down chains but with a fixed correlation between the chains. This proposal

6.5 Trimethyl Cellulose and 6-O-Acetyl-2,3-di-O-methyl Cellulose

203

Fig. 6.32 Fiber X-ray diagram of a stretched film of 6-O-acetyl-2,3-di-O-methyl cellulose (DMAC). Note the similarities with the pattern of TMC II in Fig. 6.30

Fig. 6.33 The crystal and molecular structure of DMAC in space group P21221. a Projection in the [010] direction on the a–c plane. Note the presence of a translational shift in the a-direction of the up as well as the down chains in this projection. b Projection in the [001] direction on the a–b plane. Note the similarities with Fig. 6.31

would explain the difficulty in finding a reasonable space group and unit cell for this crystal structure (Zugenmaier 1983). A regioselective mixed cellulose ester-ether (6-O-acetyl-2,3-di-O-methyl cellulose) was investigated (Möller 1982) and similarities were recognized with the

204

6 Cellulose Derivatives

X-ray fiber pattern of TMC II (compare Figs. 6.30, 6.32). It seems that the methyl groups at the O2 and O3 positions determine the similar packing arrangement with regard to TMC and that the acetyl group at C6 can accommodate to this packing arrangement. A conformation and packing analysis in space group P21221 with a 21 symmetry axis of the chain led to a reasonable structure with no too short van der Waals contacts and a realistic chain conformation. The coordinates are listed in Table A.28 and the two main projections of the model down the helix axes on the a–b plane and down the b-axis on the a–c plane are illustrated in Fig. 6.33 (Zugenmaier, unpublished results 2005).

References Cartier L, Spassky N, Lotz B (1996) Structures frustrées de polymèrs chireaux. C R Acad Sci Paris Sér II b 322:429–435 Chanzy H, Roche E, Vuong R (1971) Electron diffraction of cellulose triacetate single crystals. Kolloid Z Z Polym 248:1034–1035 Chanzy HD, Roche EJ, Vuong RK (1973) Crystallization of cellulose triacetate from solution at high temperature. J Polym Sci Polym Phys Ed 11:1859–1861 Dorset DL (1995) Structural electron crystallography. Plenum, New York Dulmage WJ (1957) The molecular and crystal structure of cellulose triacetate. J Polym Sci 26:277–288 Eggert U (1985) Untersuchungen zur Struktur von Cellulosetriestern. Doctoral dissertation, TU Clausthal, Clausthal-Zellerfeld French AD, Johnson GP, Peralta-Inga Z, Stevens ED, Dowd MK (2002) Implications of the crystal structure of NAI-cellulose and a hybrid potential energy surface for cellulose structures. Polym Prepr 43(1):195–196 Hess K, Trogus C (1929) Über reversible und irreversible Gitteränderungen von Triacetylcellulose. (Röntgenographische Untersuchungen an Cellulosederivaten. III). Z Phys Chem B 5:161–176 Iwata T, Okumura K, Azuma J, Chanzy H, Tanaka F (1994) Single crystals of regio-selectively substituted cellulose hetereo-esters. Cellulose 1:67–76 Iwata T, Fukushima A, Okumura K, Azuma J (1996a) X-ray and electron diffraction study on cellulose butyrate diacetate (CBDA, 2,3-di-O-acetyl-6-O-butyryl cellulose). Mokuzai Gakkaishi 42:289–292 Iwata T, Okumura K, Azuma J, Tanaka F (1996b) Molecular and crystal structure of cellulose propanoate diacetate (CPDA, 2,3-di-O-acetyl-6-O-propanoyl cellulose). Cellulose 3:91–106 Iwata T, Okumura K, Azuma J, Tanaka F (1996c) Molecular and crystal structure of cellulose acetate dipropanoate (CADP, 6-O-acetyl-2,3-di-O-propanoyl cellulose). Cellulose 3:107–124 Iwata T, Tanaka F, Okumura K, Azuma J (1996d) Conformation analysis of cellulose tributyrate and cellulose trivalerate. Sen’i Gakkaishi 52:423–429 Kono H, Numata Y, Nagai N, Erata T, Takai M (1999) CPMAS 13C NMR and X-ray studies of cellooligosaccharide acetates as a model for cellulose triacetate. J Polym Sci A 37:4100–4107 Kuppel A, Bittiger H, Husemann E (1972) Das Faserdiagramm des TriacetylcelluloseNitromethankomplexes. Kolloid Z Z Polym 250:623–624 Kuppel A, Husemann E, Seifert E, Zugenmaier P (1973) Über die Umwandlung von Triacetylcellulose I in Triacetylcellulose II und die Packung der Cellulose in der nativen Faser. Kolloid Z Z Polym 251:432–433 Malm CJ, Genung LB, Fleckenstein JV (1947) Densities of cellulose esters. Ind Eng Chem 39:1499–1504

References

205

Manley RStJ (1963a) Growth and morphology of single crystals of cellulose triacetate. J Polym Sci A 1:1875-1892 Manley RStJ (1963b) Hydrolysis of cellulose triacetate crystals. J Polym Sci A 1:1893–1899 Meader D, Atkins EDT, Happey F (1978) Cellulose trinitrate: molecular conformation and packing considerations. Polymer 19:1371–1374 Meyer KH, Mark H (1930) Der Aufbau der hochpolymeren organischen Naturstoffe. Akademische Verlagsgesellschaft, Leipzig Möller R (1982) Untersuchungen zur Struktur von Cellulosederivten. Diploma dissertation, TU Clausthal, Clausthal-Zellerfeld Rånby BG, Noe RW (1961) Crystallization of cellulose and cellulose derivatives from dilute solution. I. Growth of single crystals. J Polym Sci 51:337–347 Riehl K (1992) Strukturuntersuchungen an Cellulosederivaten mit stereoselektiver Trennwirkung. Doctoral dissertation, TU Clausthal, Clausthal-Zellerfeld Roche EJ, O’Brien JP (1985) Crystalline polymorphism of cellulose triacetate fibers spun from lyotropic solutions. In: Proceedings of the meeting of the Fiber Society of Japan, Hakone, Japan, p II-12 Roche E, Chanzy H, Boudeulle M, Marchessault RH, Sundarajan P (1978) Three-dimensional crystalline structure of cellulose triacetate II. Macromolecules 11:86–94 Roche EJ, O’Brien JP, Allen SR (1986) Preparation of cellulose triacetate I from solution. Polym Commun 27:138–140 Shuto Y (1990) Molecular and crystal structure analysis of cellulose tripropionate. Doctoral dissertation, Kyoto University, Kyoto Shuto Y, Okamura K, Tanaka F, Norimoto M (1986) Conformational analysis of cellulose tripropionate. Bull Kyoto Univ For 58:280–288 Shuto Y, Sugiyama J, Harada H, Okamura K (1987) Single crystals of cellulose tripropionate. Macromolecules 20:2317–2318 Shuto Y, Okamura K, Azuma J, Tanaka F, Chanzy H (1989a) A combined electron and X-ray diffraction study of cellulose tripropionate. In: Schuerch C (ed) Cellulose and wood – chemistry and technology. Wiley, New York, pp 207–220 Shuto Y, Okamura K, Azuma J, Tanaka F, Chanzy H (1989b) X-ray and electron diffraction study of some cellulose derivatives. In: Kennedy JF, Phillips GO, Williams PO (eds) Cellulose: structural and functional aspects. Harwood, Chichester, pp 283–288 Sikorski P, Wada M, Heux L, Shintani H, Stokke B (2004) Crystal structure of cellulose triacetate I. Macromolecules 37:4547–4553 Sprague BS, Riley JL, Noether HD (1958) Factors influencing the crystal structure of cellulose triacetate. Text Res J 28:275–287 Steinmeier H (1988) Synthese von Cellulosederivaten und Untersuchungen zur Struktur im kristallinen und flüssigkristallinen Zustand. Doctoral dissertation, TU Clausthal, Clausthal-Zellerfeld Steinmeier H, Zugenmaier P (1987) “Homogeneous” and “heterogeneous” cellulose triesters and a cellulose triurethane: Synthesis and structural investigations of the crystalline state. Carbohydr Res 164:97–105 Stipanovic AJ, Sarko A (1978) Molecular and crystal structure of cellulose triacetate I: a parallel chain structure. Polymer 19:3–8 Trogus C, Hess K (1929) Das Translationsgitter der Methylcellulose. Z Phys Chem B 4:331–345 VanderHart DL, Hyatt JA, Atalla RH, Tirumalai VC (1996) Solid-state 13C NMR and Raman studies of cellulose triacetate: Oligomers, polymorphism, and inferences about chain polarity. Macromolecules 29:730–739 Watanabe S, Hayashi J, Imai K (1968) A study of cellulose trinitrate structure. J Polym Sci C 23:809–823

206

6 Cellulose Derivatives

Wolf RM, Francotte E, Glasser L, Simon I, Scheraga HA (1992) Computation of low-energy crystalline arrangements of cellulose triacetate. Macromolecules 25:709–720 Zugenmaier P (1983) Structural investigations on cellulose derivatives. J Appl Polym Sci Appl Polym Symp 37:223–238 Zugenmaier P (1986) Structural investigations on some cellulose derivatives in the crystalline and liquid crystalline state. In: Young RA, Rowell RM (eds) Cellulose – structure, modification and hydrolysis. Wiley, New York, pp 221–248 Zugenmaier P (1995) Novel results of structural investigations on crystalline and liquid crystalline cellulose derivatives and their potential application. In: Kennedy JF, Phillips GO, Williams PO, Piculell L (eds) Cellulose and cellulose derivatives: physico-chemical aspects and industrial applications. Woodhead, Cambridge, pp 381–392 Zugenmaier P (2006) Order in cellulosics: historical development of crystal structure research on cellulose. In: Abstracts of papers of the American Chemical Society 231, Cell, 26 March 26 2006, p 101 Zugenmaier P, Kuppel A (1986) Diffraction studies on trimetylcellulose and trimetylmannan. Colloid Polym Sci 264:231–235 Zugenmaier P, Rappenecker G (1978) The crystal and molecular structure of methyl tetraacetylb-d-glucoside. Acta Crystallogr Sect B 34:164–167

Chapter 7

Morphology

7.1

Crystalline Domain Sizes

The crystalline structures of cellulosics have been discussed in the preceding chapters and models for chain conformation and packing arrangements established. The crystalline domains are of very limited size in cellulosics and these domains are arranged with noncrystalline materials in fibrils and fibers and predominantly determine the mechanical properties, such as tensile strength. The ratio of crystalline to total materials is termed the crystallinity index xcr and can be measured by a variety of methods relying on different structural features. This ratio differs for native and regenerated cellulose of various origins to a large degree as shown in Table 7.1 and is caused by different distributions of crystalline and noncrystalline materials in fibrils and fibers. The lateral sizes of the microfibils and bundle of microfibrils are also included in this table as determined by the X-ray line width of the (110) reflection and electron-microscopic observations. In this chapter we will not discuss in detail the size of domains and their aggregation that is the morphology and the structure of the fibers but rather will introduce some current models necessary for a discussion of properties.

7.2

Microfibrils

Mechanical solid-state properties of cellulosic materials are strongly influenced by the interactions of small domains, crystalline and/ or noncrystalline, which represent the basis for the morphology of the materials (fibrils, fibers and lamellae). However, the thermal behavior of materials is predominantly determined by the small domains, e.g., the melting of the crystallites and their glass transition. Materials with coexisting crystalline and noncrystalline domains are called partially crystalline and are commonly described by a two-phase model. Two extreme representations should be mentioned: The fringed-micelle model with molecular chains forming successive crystalline and noncrystalline (amorphous) domains of limited sizes (Fig. 7.1a) and the folded-chain model in which the chains of a crystalline domain are folded back (Fig. 7.1b, c) with only a few tie chains passing to P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. Springer Series in Wood Science. © Springer-Verlag Berlin Heidleberg 2008

207

208

7 Morphology

Table 7.1 Crystallinity index xcr and lateral dimensions of microfibrils of native cellulose I and cellulose II by the line-width method of reflection d(110) and by electron-microscopic measurements d; size of bundles of microfibrils D (Klemm et al. 2005; Ganster and Fink 1999) d(110) (nm) d (nm) D (nm) Source xcr (%) Algal cellulose Bacterial cellulose Cotton linters Ramie Hemp Flax Dissolving pulp Viscose (rayon) cellulose II Cotton linters mercerized cellulose II Lyocel filament cellulose II Carbamat filament Cellophane cellulose II

>80 65–79 56–65 44–47 60 56 43–56 27–31 45 35 34–43 45

10 5 5 5 3–5 4–5 4–5 5 5 4

10–35 4–7 7–9 3–12 3–18 3–18 10–30

35–95 50–80 40–100 25–80 30–80 ≤100

5

Fig. 7.1 Idealized models of microfibrils by various authors: a Hess et al. (1957); b Marx-Figini and Schulz (1966); c Manley 1964); d Frey-Wyssling and Mühlethaler (1963). Chain ends are incorporated in the microfibril. (From Muggli et al. 1969)

the adjacent crystalline domain, responsible for the toughness of the materials. The folded-chain model describes antiparallel chain packing, while the fringed-micelle model is able to represent parallel and antiparallel chain arrangements. The noncrystalline and disordered chains at the surface of the crystalline domains are omitted in the drawings in Fig. 7.1.

7.2 Microfibrils

209

The differences between amorphous regions of synthetic flexible chain molecules and of semiflexible chains of cellulosics will be briefly addressed and discussed. In contrast to the structure of amorphous regions of flexible chains, where the chains can be described by a Gaussian coil, the cellulosic chains are in extended form and represent rod-type structures as revealed by an investigation by Fink et al. (1987) for “amorphous” cellulose I and cellulose II of low degree of polymerization (DP). These ribbonlike chains are distributed isotropically in the sample. The crystallization procedure and the characteristics of the chain molecules influence the morphology. Bittiger and Husemann (1966) performed a wide-ranging investigation on soluble semiflexible cellulose tricarbanilate. They found that crystallization from dilute solution of high molecular mass compounds (DPw = 1,500–7,000) resulted in rod-type structures of a few or maybe only one folded single molecule. Crystallization from dilute solution of low molecular mass (DPw < 1,000) can create regular-shaped polymeric single crystals (lamellae) as shown in Fig. 3.9. However, concentrated solution led to fibrillar structures. In such solutions or in solutions with nondispersed macromolecules, the chains of low molecular mass already form elongated aggregates and by a crystallization process fibrils grow as shown in Fig. 7.2. Kuga et al. (1993) investigated a mutant strain of Acetobacter xylinum, which produces cellulose of band-like morphology. Bands of 10-nm width were observed with the chains oriented perpendicular to the strand extrusion axis. Since the chains are much longer than the bands, chain folding is assumed with the occurrence of antiparallel chains packed as in cellulose II. This anomalous band-like structures contrasts with the straight crystalline microfibrils of cellulose I produced by the wild type of this bacterial cellulose (Fig. 7.6).

Fig. 7.2 Idealized structures of microfibrils in the solid state formed out of solution (a) and aggregates in solution (b). (From Schulz et al. 1998)

210

7 Morphology

Details of morphological structures of cellulosic materials have been studied by investigations with polarized light, including swelling experiments and with X-rays (Sisson 1940). Nägeli (1858) showed that materials like starch and cellulose are birefringent as a result of internal arranged anisotropic, submicroscopic particles, which he called micelles. From investigations of the birefringence, especially by Ambronn (1916a, b, 1917) and his coworkers, it was concluded that the micelles consist of partially parallel and elongated arranged units. Details of the internal structure of the micelles remained obscure, but it was proposed that they are similar to those of crystals. These results were rejected for polysaccharides by many researchers in the 1920s, since the high molecular weight biopolymers showed different physical behavior from the low molecular materials and in addition no sharp shape of the proposed crystallites has been observed as found for low molecular materials. An excellent method for the estimation of the size of the micelles is represented by the line-width broadening of X-ray reflections from the crystalline part of the structure. The elongated crystalline domains (microfibrils) of native ramie cellulose have been proposed to be over 600-Å long and 50-Å wide by this X-ray method (Hengstenberg and Mark 1928). With the optical transforms introduced in Chap. 3 (Fig. 3.2), it can be illustrated that the size of the ordered domain, which is the number of motifs in a certain direction, correlates with the line width of the reflections. The noncrystalline or amorphous portion of a domain (surface, disordered region, etc.) is excluded in these estimations. However, if the domains with the crystalline and noncrystalline parts form regularly arranged particles, they can be regarded as ordered scattering units themselves and then may lead to so-called Bragg reflections at very low angles (small-angle X-ray scattering, SAXS). The particle size detected with this method includes the crystalline and noncrystalline parts of the particles. The size and the shape of the crystallites of native cellulose vary for the same cellulose polymorph from species to species (Table 7.1). From experiments with X-ray and neutron diffraction as well as from electron micrographs, the cross section of the elongated domains is about 49×66 Å2 for cotton (Müller et al. 2000), 41×44 Å2 for flax (Müller et al. 2000), 150×150 Å2 for tunicin (Favier et al. 1995), 25×25 Å2 for wood (Jakob et al. 1994) and 200×200 Å2 for Valonia (Sugiyama et al. 1985). For native ramie the average crystallite length in the fiber direction amounts to 485 Å and the average fiber diameter is 52 Å, both quantities recently obtained from X-ray line widths. In contrast, the length of the crystallites along the fiber of regenerated cellulose II (viscose) is 112 Å by the line-width method (Haase and Renwanz 1972). But 56 Å has to be added for the noncrystalline disordered region to result in regular ordered and repeated particle sizes of 168 Å as detected by SAXS measurements. The average fiber diameter amounts to 41 Å for this regenerated cellulose II sample (Haase et al. 1973). The lateral crystallite or microfibril sizes for cellulose II vary between 33 and 58 Å for samples of different origin (Hofmann et al. 1989). In addition, some larger fibrillar diameters have been found by a careful analysis with SAXS. These relate to the small ones with a ratio of 1:2:4:8 for ramie and rayon (viscose cellulose II) but with a ratio of 1:2:3 for

7.2 Microfibrils

211

bacterial cellulose and Fortisan (deacetylated cellulose II of different origin) (Haase et al. 1974). These larger particles for ramie and rayon do not coherently scatter X-rays and, therefore, in contrast to some synthetic polymers cannot be detected by wide-angle experiments. For algal and bacterial cellulose a recrystallization of the small particles takes place to form larger, coherently scattering particles. Smallangle neutron scattering (SANS) of selectively deuterated cellulose (Fischer et al. 1978), which provides a higher contrast by diffusion of deuterium into the noncrystalline regions of cellulose II, exhibits a long-period Bragg reflection of 165 Å for Fortisan and 193 Å for rayon fibers, values quite similar to that of 168 Å found as the length of the particles by Haase et al. (1973). The different cross sections of native cellulose I can be traced back to differences in imperfections of chain packing by disorder and defects as proposed by Hosemann et al. (1967). Differences in chain packing will also weakly influence the size of the unit cells and have actually been found in native cellulose I from different sources (Wellard 1954). For 11 samples of cellulose I prepared from cotton, flax or ramie, the unit-cell dimension a varied from 7.85 to 7.93 Å, the b dimension from 8.16 to 8.21 Å but the fiber repeat c stayed constant with a mean value of 10.34 ± 1 Å (Ioelovich and Larina 1999). These changes of unit-cell parameters can be explicitly correlated with the lateral dimension of the cellulose crystallites. A recent study of cellulose I from a wider source also led to conclusions regarding changes in the fiber repeat as well (Davidson et al. 2004; Table 7.2). However, the change of the fiber repeat, i.e., the c-dimension, may, besides pure packing effects, predominantly be influenced by the conformation of the cellulose chains as deduced from the conformational analysis of cellulosic chains and a comparison with oligomeric compounds. In Table 7.2 the line width b of the (200) X ray reflection is listed after correction for the instrumental broadening, the fraction of interior chains of the crystallites X and the fiber repeat determined from the meridional (004) reflection. An inverse correlation can be deduced between b and X, which also represents a measure of the lateral dimension of the crystallites. It should be remarked that the fiber repeat did not depend on the ratio of the cellulose polymorphs Iα and Iβ nor on moisture content. Table 7.2 X-ray diffraction data for ten samples of cellulose I. (From Davidson et al. 2004) Sample b(200) (°) X Fiber repeat c (Å) Pinus radiata wood 3.31 0.362 10.34 Sisal twine 3.30 0.363 10.33 Castanea sativa wood 3.26 0.370 10.33 Phormium tenax leaf fiber 2.86 0.430 10.35 Jute twine 2.86 0.430 10.37 Linen thread 1.82 0.610 10.36 Commercial cellulose powder 1.33 0.706 10.37 Cotton thread 1.22 0.729 10.40 Pyura pachydermatina (tunicate) 1.19 0.734 10.39 Chaetomorpha coliformis (alga) 0.64 0.852 10.39 b(200) line width of the (200) reflection, X fraction of crystallite-interior chains

212

7 Morphology

Fig. 7.3 The fringed-micelle model of cellulose fibers. (From Fink and Walenta 1994)

The elongated crystallites or microfibrils (Hearle 1958) are joined in some sections to form wider microfibrils or bundles of microfibrils as shown schematically in Fig. 7.3. At a larger scale by considering many of them, they form fibrils and fibers. The interconnections of the mircofibrils give rise to the high tenacity observed in cellulose fibers. In the next section the relation of the microfibrils to further larger units or supermolecular structures will be discussed.

7.3

Microfibrils and Fibrils

The physical properties of a single polymeric chain, such as chain length, stiffness, perfection in repeat units, the crystallization procedure, such as nucleation and growth, as well as the interaction within a crystallite and with noncrystalline regions, influence the shape and size of the morphological and supermolecular structure. Insights into the morphological structure may be obtained by a variety of methods: X-ray, optical and electron-microscopic investigations, and chemical and physical studies, such as degradation, molecular weight determination and mechanical measurements. All the data gathered have to be incorporated into a model for the fibrils. The results from line-width determination of X-ray wide-angle reflections (wide-angle X-ray scattering, WAXS) clearly establish an elongated crystallite and the low-angle scattering (SAXS) exhibiting Bragg reflections demonstrates that piles of fairly regular crystallites are present. The difference in the size of the crystallites determined by WAXS and the so-called long period determined by SAXS suggests some noncrystalline material is present between the crystallites. It is rather difficult to link all the results from various sources in one model and it may well be that several morphological structures exist. The biosynthesis of cellulose by nature to form long microfibrils, fibrils and fibers generally supports the idea of extended chains without chain folding (Fig. 7.4). Chain folding can also be

7.3 Microfibrils and Fibrils

213

Fig. 7.4 Electron micrographs of a dilute suspension of fibrils of native cellulose (sugar beet) (a) and whiskers of tunicin (b); length from 100 nm to several microns, width 10–20 nm. (From Eichhorn et al. 2001)

excluded by short chains forming cellulose II hindered by the stiffness of the chains (Fig. 7.2). Cellulosic chains in a monodisperse solution are extended as illustrated in Fig. 1.1. They will preserve their shape upon crystallization and the microfibrils obtained may resemble those of native cellulose. Chain folding cannot be avoided for regenerated cellulose of very long chain length, since its shape in solution resembles a Gaussian coil (Fig. 1.1) and the loops in solution will appear as folds in the solid state. A notable change in the shape of the backbone of the cellulosic chain in solution was proposed to occur at a DP of approximately 1,000 from rheological investigations (Marx-Figini and Gonzalez 1988), which may correlate with the change of the extended to the coiled form. The model of bricklike micelles of native cellulose first developed in the 1920s and illustrated in Fig. 2.12 had to be abandoned because of the fact that the native cellulosic chains are much longer than the size proposed for the micelles. The scattering method of electromagnetic or particle waves represents the method of choice for establishing structural models but suffers from drawbacks in the smallangle region for cellulose. The noncrystalline region separating the crystallites along a fiber has almost the same electron density as the crystalline region and, therefore, the necessary scattering contrast is low for a discrimination between the two regions. The nonconformity of the crystallites also makes their determination difficult. Additional experiments, such as increasing the scattering density of the noncrystalline region or attacking and degrading these regions, have been employed to create the needed contrast and to obtain the required information for establishing structural models. The noncrystalline spaces between the crystallites in all three dimensions are easily accessible to water or heavy water as well as to small molecules, causing swelling or, e.g., an exchange between the hydrogen of the hydroxyl of the cellulose and the deuterium of the heavy water, resulting in the necessary contrast for neutron scattering experiments. The penetration of small molecules into the crystal lattice is much more difficult and needs more time to open and widen the lattice of the cellulosic structure. A further source of information is represented by the degradation by aging or hydrolysis, but the kinetic mechanism of the degradation

214

7 Morphology

process is not clear. With application of this method a discrimination is essential between structural features of the chain, such as periodically occurring chemical weak bonds, activated inductive effects of carbonyl groups brought in by chemical or technological processes and features of the arrangements of chains, such as differences in noncrystalline and crystalline regions. Marx-Figini and CounMatus (1981) came to the conclusion that the fast hydrolysis at an early stage is a consequence of weak links, which can be regarded as an alteration of the 1-4 linked unit in the cellulose chain, probably a xylan unit. Extensive studies on cotton linters have shown that along a native cellulose chain about five weak linkages occur and the materials can suffer considerable scission during storage without any chemical and/or mechanical treatment (Schulz and Husemann 1942). Generally, a fast degradation was observed in the beginning of a hydrolysis reaction, which then slows down in solution or the solid state after a short period of time. This fast reaction was attributed to defects in the chemical structure of the cellulose molecule. Therefore, a precise model evaluation of fibrils is rather difficult by this method, since a rearrangement and the growth of crystals during the reaction also have to be taken into consideration. Schulz and Husemann (1942) proposed long periods in cotton linters of 3,100 glucose units (DP), which were divided into six regions of DP of 510. The degradation continues by hydrolysis and stops for these materials at a DP of about 200, called the leveling-off DP (LODP), which may decrease to a DP of about 70 for cellulose from a different origin (cellulose II). The periodic disorder was recently studied along microfibrils of ramie cellulose by SANS by Nishiyama et al. (2003). The labile hydrogens were replaced by deuterium in the noncrystalline regions to increase the contrast and a long-spacing Bragg reflection observed at 150 nm corresponding to a DP of 300. Additional studies were undertaken by electron microscopy and acid hydrolysis. The LODP of approximately 300 obtained by degradation of the hydrolyzed samples matches exactly the periodicity observed in diffraction studies (SAXS), which suggests the presence of regular arranged regions of this size along the microfibrils. It was concluded from the chain length of a DP of 1,600 that four to five residues are misaligned on this length and that faults are expected every 300 glucose units, which are responsible for the LODP behavior as well as for the fast start of the hydrolysis. The lateral dimension of the fibrils amounts to 129 Å for the deuterated sample, which is about twice the width observed by scanning electron microscopy for the individual microfibrils in the ramie fiber. In contrast, the SANS experiment on deuterated cellulose II, Fortisan and rayon, resulted in a long period of 165 and 193 Å, respectively, and was considered as support for the fringed-micelle model (Fig. 7.3). The lateral dimension of Fortisan was determined to be 34 Å, in excellent agreement with 35 Å obtained by electron-microscopic measurements (Fischer et al. 1978). But larger entities of this small unit have also been found, as was observed by X-ray line-width (WAXS) evaluation. Lateral crystallite dimensions of native cellulose strongly depend on its source and vary between 4 nm (dissolving pulps) and 10–15 nm (Valonia). For Valonia

7.3 Microfibrils and Fibrils

215

cellulose with a high content of cellulose Iα (58–64%) a small difference in the position of the reflections led to a correction of about 20% smaller lateral crystallite sizes. Generally, the lateral sizes of the microfibrils from X-ray reflection broadening are smaller than the values obtained from electron microscopy, except for bacterial cellulose (Fink et al. 1995). This result, besides a rather broad crystallite size distribution, supports the fringed fibrillar model of the morphological structure of cellulose. The lateral size distribution of regenerated cellulose II is narrower and amounts to an average between 3 and 5 nm. The lattice distortions are larger for smaller crystallite sizes as expected. The cellulose reactivity is enhanced by decreasing lateral order or lateral dimension, respectively. An extensive investigation of bacterial cellulose (Acetobacter strains) was carried out concerning the morphology with X-ray and synchrotron wide-angle scattering, cross-polarization/magic angle spinning 13C NMR and transmission electron microscopy. The ratio of cellulose Iα to cellulose Iβ (approximately 80% cellulose Iα) was explicitly considered. This cellulose type behaves exceptionally compared with other forms of cellulose. A cell of Acetobacter xylinum with an attached ribbon is represented in Fig. 7.5 (White 1982) and in Fig. 7.6 a supermolecular structural model of the early stages of the structure-forming process is presented taken from an overview on cellulose as a fascinating biopolymer and sustainable material by Klemm et al. (2005) (see also Bohn 2001). Huge amounts of water (water content more than 90%) can be incorporated into bacterial cellulose, and the water is placed between the water-free microfibrils of 7×13-nm2 cross section. About five to 12 microfibils aggregate to form flat fibrils in a water-swollen fibrillar state. A shell of noncrystalline cellulose chains passes around neighboring fibrils to produce a band of fibrils with a width of about 500 nm. Ellipsoidal, straight or slightly bent rods in the range 0.6–0.8 × 1.0–4.0 µm2 are formed and randomly laid down as a biofilm of various thicknesses. The polymerization and crystallization of the cellulose are closely linked and microfibrils of a high degree of crystallinity (60–90%) are formed by high molecular weight cellulose (DP 2,000–8,000). A uniplanar oriented foil is formed with the fibrils partially twisted around the longitudinal axis after drying the initial wet and highly swollen cellulose

Fig. 7.5 A high-resolution electron micrograph of an Acetobacter xylinum cell with attached cellulose ribbon (scale bar 1 µm). The ribbons are between 40- and 60-nm wide and are composed of 50–80 microfibrils, each 3.0–3.5-nm wide, and are twisted with a period of approximately 1 µm. (From White 1982)

216

7 Morphology

Fig. 7.6 A bacterial cellulose fibrillar band in the initial moisture state. (From Bohn 2001)

fleece. Bacterial cellulose behaves like plant cellulose after complete removal of water and then exhibits a water uptake of only about 6%. To some extent the twisting of the fibrils seems to be a general structural feature of cellulose of various species as observed in the micrograph in Fig. 7.5. Electron diffractograms of fibrils normally show a fiberlike pattern and only with the special technique of microdiffraction single crystal patterns can be obtained. Besides a compensation of the polarization along the fibrils of the parallel-packed chains, it can be speculated that the mechanical properties are enhanced and bending is allowed as for steel ropes manufactured from a number of thin twisted steel rods. The simplified models of microfibrils and fibrils are not able to theoretically describe all aspects of the experimental investigations of highly oriented cellulosic fibers and the morphology as revealed by modern experiments has to be taken into account. The noncrystalline parts (amorphous) of the microfibrils of native cellulose are placed as thin layers between the microfibrils as detected by Earl and VanderHart (1981) by NMR studies and by Müller et al. (2000) by inelastic neutron scattering. Large noncrystalline domains were not found and the molecules of the noncrystalline domains are oriented along the fiber axis, in agreement with model calculations (Ganster et al. 1994; Eichhorn and Young 2001). The small difference in density between the crystalline and noncrystalline domains supports this model as does the missing gap of electron density that produces an easily observable peak in SAXS of the microfibrils.

7.4 Parallel and Antiparallel Packing Arrangements of Microfibils Native cellulose I can be converted by treatment in the solid state with reaction of sodium hydroxide, so-called mercerization, to cellulose II; therefore, it was concluded for a long period of time that cellulose I and cellulose II should possess the

7.4 Parallel and Antiparallel Packing Arrangements of Microfibils

217

same parallel or antiparallel packing arrangement of chains in the microfibils. But with progress in structural research, it became clear that microfibrils of native cellulose I pack in a parallel chain arrangement and those of cellulose II in antiparallel chain arrangements. It was a puzzle how cellulose I can be converted to cellulose II without changing the appearance of the morphology or the orientation distribution of the fibers as shown in Fig. 6.4. The cellulose triacetate I (CTA I) fiber derived from ramie cellulose I (parallel packing) was converted to CTA II (antiparallel packing) and further to cellulose II through an intermediate step involving a CTA–nitromethane (CTA-N) complex with the preservation of the orientation and shape of the original ramie fiber. An interdiffusion model for the mercerization from cellulose I to cellulose II was proposed by Okano and Sarko (1985) as shown in Fig. 7.7. Starting with parallelpacked cellulose I chains of an antiparallel array of nearby mircofibils in a fiber (Fig. 7.7, left), uptake of sodium hydroxide beginning at the noncrystalline surface of the microfibrils, interdiffusion of chains, then washing and drying results in antiparallel cellulose II microfibrils (Fig. 7.7, right). A prerequisite for this model of native cellulose is arrays of parallel-packed chains in a single microfibril and the arrays have to be placed in up and down directions in a fiber as illustrated in Fig. 7.8. This requirement rests upon the mechanism of the biosynthesis of cellulose chains in native fibers. However, this simple interdiffusion model cannot provide clues about changes in crystallite sizes actually observed during the transformation from parallel CTA I to antiparallel CTA II in the intermediate crystalline solvent complex CTA-N. The uptake of solvent into the lattice of CTA I causes a considerable increase of the cross section of the microfibrils of CTA-N as concluded from a significant decrease of the line width of the reflections in Fig. 6.4. An increase in the number of coherently scattering chains in the crystallites is responsible for this effect and not the extension of the lattice by incorporating solvent, which amounts to less than 15%. The loss of solvent in the antiparallel-packed CTA-N leads to a considerable decrease of the cross section again for the antiparallel-arranged CTA II. This observed behavior may originate from differences in the interactions in the structure and also from differences in the perfection of the packing allowing more chains to be packed in the microfibrils.

Fig. 7.7 Model for the conversion of parallel-packed arrays of microfibrils of up and down chains of cellulose I to antiparallel-packed fibrils of cellulose II during mercerization. (From Okano and Sarko 1985)

218

7 Morphology

Fig. 7.8 Model of a cellulose fiber consisting of parallel chains placed in microfibrils of antiparallel arrays of microfibrils. (From Okano and Sarko 1985)

A simple lattice model has been developed to explain the phenomenon of different packing arrangement in a fiber, which also exhibits various cross sections of the microfibrils (Fig. 7.9). A trigonal or hexagonal grid across the fiber was introduced to simplify the theoretical considerations with chains of different polarity, i.e., up and down chains placed on the grid. Since the crystal structure of cellulose tribenzoate (CTBe) exhibits a trigonal base plane (Sect. 6.4), the dimensions of this base plane were used in the explicit estimation of interactions between chains. Further, some insight may be gained about the unusual parallel packing arrangement of this structure. The central chain of trigonal space group P32 is surrounded by a first shell of six chains with equal distances. The second shell of chains, also all at equal distances, can be neglected owing to the larger distances from the central chain and the resulting minor contribution to the interaction energy. A different interaction (crystallization) energy was assigned for a neighboring single antiparallel chain in comparison with a parallel one. A central chain was chosen on the grid and the growth of the crystal was simulated by placing neighboring chains on the grid. A probability parameter r was introduced, the value of r depending on the relationship of the interaction energy of chains of different polarities, to describe the

7.4 Parallel and Antiparallel Packing Arrangements of Microfibils

219

Fig. 7.9 Models of chain packing arrangements for various interaction parameters between chains of different polarity in projection on a trigonal grid. Black up chains, white down chains. Probability parameter r=1 for parallel-running chains and r=0 for antiparallel-running ones. a r = 0; b r = 0.25; c r = 0.5; d r = 0.75; e r = 0.85

statistical growth of the crystal with up and down chains on the two-dimensional lattice,. The probability parameter r is 1 for parallel-running chains with regard to a given chain, r = 0 for antiparallel-running ones and 0 1 for statistical packing with increasing preference for parallel packing with increasing r value (Riehl 1992; Zugenmaier 1995). An illustration of statistically distributed up and down chains is shown on a trigonal grid in Fig. 7.9. These patterns can be calculated by starting from various nuclei, e.g., the central chain placed at the center or at the brim as realized in Fig. 7.9. A criterion for the size of a domain is the extension of parallel-running chains in the up or the down direction, respectively. Figure 7.9a represents an antiparallel arrangement with r = 0. The domain size amounts to the average cross section of a single chain, which is about 12 Å for CTBe. For r = 0.85, as depicted in Fig. 7.9e, large areas of parallel chains are obtained of approximately 170 Å average domain size for CTBe chains in the up direction as well as in the down direction. For deceasing r the statistically arranged areas decrease, e.g., for r = 0.6 the average domain size is 34 Å for CTBe. It should be noted that an equal number of chains point in up and down directions for all probability parameters r, which means that the prerequisite for interdiffusion is fulfilled for a transformation from an all-parallel to an all-antiparallel chain arrangement in the microfibrils (Figs. 7.7, 7.8) and a change of crystallite size can be explained by a variation of the interactions between chains. The theoretical model of packing arrangements of chains proposed can be regarded as a first approximation, since it only describes the packing on a fixed trigonal grid as observed for CTBe. It does not include a three-dimensional interaction or packing of chains, and does not account for different sizes and symmetries of the unit cells as observed for the transformation from cellulose I to cellulose II or from CTA I to CTA-N.

220

7 Morphology

References Ambronn H (1916a) Über das Zusammenwirken von Stäbchendoppelbrechung und Eigendoppelbrechung, I. Kolloid-Z 18:90–97 Ambronn H (1916b) Über das Zusammenwirken von Stäbchendoppelbrechung und Eigendoppelbrechung, II. Kolloid-Z 18:273–281 Ambronn H (1917) Über das Zusammenwirken von Stäbchendoppelbrechung und Eigendoppelbrechung, III. Kolloid-Z 20:173–185 Bittiger H, Husemann E (1966) Elektronenmikroskopische Untersuchungen über die Kristallisation von Makromolekülen aus verdünnter Lösung. Makromol Chem 96:92–99 Bohn A (2001) Röntgenuntersuchungen zur Vorzugsorientierung und übermolekularen Struktur nativer und regenerierter Cellulose. Doctoral thesis TU Berlin Davidson TC, Newman RH, Ryan MJ (2004) Variations in the fibre repeat between samples of cellulose I from different sources. Carbohydr Res 339:2889–2893 Earl WL, VanderHart DL (1981) Observations by high-resolution carbon-13 nuclear magnetic resonance of cellulose I related to morphology and crystal structure. Macromolecules 14:570–574 Eichhorn SJ, Young RJ (2001): The Young’s modulus of a microcrystalline cellulose. Cellulose 8:197–207 Eichhorn SJ, Baillie CA, Zafeiropoulos N, Mwaikambo Y, Ansell MP, Dufresne A, Entwistle KM, Herrera- Franco PJ, Escamilla GC, Groom L, Hughes M, Hill C, Rials TG, Wild PM (2001) Review: current international research into cellulosic fibres and composites. J Mater Sci 36:2107–2131 Favier V, Chanzy H, Cavaillé JY (1995) Polymer nanocomoposites reinforced by cellulose whiskers. Macromolecules 28:6365–6367 Fink H-P, Walenta E (1994) Röntgenbeugungsuntersuchungen zur übermolekularen Struktur von Cellulose im Verarbeitungsprozeß. Papier 48:739–748 Fink H-P, Philipp B, Paul D, Serimaa R, Paakkari T (1987) The structure of amorphous cellulose revealed by wide-angle X-ray scattering. Polymer 28:1265–1270 Fink H-P, Hofmann D, Philipp B (1995) Some aspects of lateral chain order in cellulosics from X-ray scattering. Cellulose 2:51–70 Fischer EW, Herchenröder P, Manley RStJ, Stamm M (1978) Small-angle neutron scattering of selectively deuterated cellulose. Macromolecules 11:213–217 Frey-Wyssling A, Mühlethaler K (1963) Die Elementarfibrillen der Cellulose. Makromol Chem 62:25–30 Ganster J, Fink H-P (1999) Physical constants of cellulose. In: Immergut EH, Grulke EA (eds) Polymer handbook, sect V, 4th edn. Wiley, New York, pp 135–157 Ganster J, Fink H-P, Fraatz J, Nywlt M (1994) Relation between structure and elastic constants of man-made cellulosic fibers: I. A two phase anisotropic model with contiguity parameter. Acta Polym 45:12–318 Haase J, Renwanz B (1972) Profilanalyse der Röntgenweitwinkelreflexe in Kettenrichtung von Cellulose II. Kolloid Z Z Polym 250:503–505 Haase J, Hosemann R, Renwanz B (1973) Parakristalline Gitterstörungen in Kettenrichtung von Cellulose I und II. Kolloid Z Z Polym 251:871–875 Haase J, Hosemann R, Renwanz B (1974) Röntgenweit- und Röntgenkleinwinkeluntersuchungen an Zellulose. Colloid Polym Sci 252:712–717 Hearle JWS (1958) A fringed fibril theory of structure of crystalline polymers. J Polym Sci 28:432–435 Hengstenberg J, Mark H (1928) Über die Form und die Größe der Mizelle von Zellulose und Kautschuk. Z Kristallogr 69:271–284 Hess K, Mahl H, Gütter E (1957) Elektronenmikroskopische Darstellung großer Langperioden in Zellulosefasern und ihr Vergleich mit Perioden anderer Faserarten. Kolloid-Z 155:1–19

References

221

Hofmann D, Fink H-P, Philipp B (1989) Lateral crystallite size and lattice distortions in cellulose II samples of different origin. Polymer 30:237–241 Hosemann R, Lemm K, Schönfeld A, Wilke W (1967) Teilchengrößen und parakristalline Gitterstörungen. Kolloid Z Z Polym 216:103–110 Ioelovich M, Larina E (1999) Parameters of crystalline structure and their influence on the reactivity of cellulose. Cellulose Chem Technol 33:3–12 Jakob HF, Fratzl P, Tschegg SE (1994) Size and arrangement of elementary cellulose fibrils in wood cells: a small-angle X-ray scattering study of Picea abies. J Struct Biol 113:13–22 Klemm D, Heublein B, Fink H-P, Bohn A (2005) Cellulose: fascinating biopolymer and sustainable raw material. Angew Chem Int Ed Engl 44:3358–3593 Kuga S, Takagi S, Brown RM Jr (1993) Native folded-chain cellulose II. Polymer 34:3293–3297 Manley RStJ (1964) Fine structure of native microfibrils. Nature 204(496):1155–1157 Marx-Figini M, Coun-Matus M (1981) On the kinetics of hydrolytic degradation of native cellulose. Macromol Chem 182:3603–3616 Marx-Figini M, Gonzalez FR (1988) Rheologische Untersuchungen an verdünnten Cellulosenitratlösungen. Papier 42:660–664 Marx-Figini M, Schulz GV (1966) Zur Biosynthese der Cellulose. Naturwissenschaften 53:466–474 Muggli R, Elias H-G, Mühlethaler K (1969) Zum Feinbau der Elementarfibrillen der Cellulose. Makromol Chem 121:290–294 Müller M, Czihak C, Schober H, Nishiyama Y, Vogl G (2000) All disordered regions of native cellulose show common low frequency dynamics. Macromolecules 33:1834–1840 Nägeli C (1858) Die Stärkekörner. Schulthess, Zurich Nishiyama Y, Kim U-J, Kim D-Y, Katsumata KS, May RP, Langan P (2003) Periodic disorder along Ramie cellulose microfibrils. Biomacromolecules 4:1013–1017 Okano T, Sarko A (1985) Mercerization of cellulose. II. Alkali-cellulose intermediates and a possible mercerization mechanism. J Appl Polym Sci 30:325–332 Riehl K (1992) Strukturuntersuchungen an Cellulosederivaten mit stereoselektiver Trennwirkung. Doctoral thesis, TU Clausthal, Clausthal-Zellerfeld Schulz GV, Husemann E (1942) Über die Verteilung der Molekulargewichte in abgebauten Cellulosen und ein periodisches Aufbauprinzip im Cellulosemolekül. Z Phys Chem B 52:23–49 Schulz L, Burchard W, Dönges R (1998) Evidence of supramolecular structures of cellulose derivatives. In: Heinze TJ, Glasser WG (eds) Cellulose derivatives. ACS symposium series no 688, American Chemical Society, Washington, pp 218–238 Sisson WA (1940) X-ray studies regarding the structures and behavior of native cellulose membranes. Chem Rev 26:187–201 Sugiyama J, Harada H, Fujiyoshi Y, Uyeda N (1985) Lattice images from ultrathin sections of cellulose microfibrils in the cell wall of Valonia macrophysa Kütz. Planta 166:161–168 Wellard HJ (1954) Variation in the lattice spacing of cellulose. J Polym Sci 13:471–476 White AR (1982) Visualization of cellulases and cellulose degradation. In: Brown RM Jr (ed) Cellulose and other natural polymer systems. Plenum, New York, pp 489–509 Zugenmaier P (1995) Novel results of structural investigations on crystalline and liquid crystalline cellulose derivatives and their potential application. In: Kennedy JF, Phillips GO, Williams PO, Piculell L (eds) Cellulose and cellulose derivatives: physico-chemical aspects and industrial applications. Woodhead, Cambridge, pp 381–392

This page intentionally blank

Fig. A.1 Fourier transform IR spectra of cellulose Iβ derived from a heat-treated Valonia cellulose Iα specimen. U nonpolarized radiation, C polarization along the cellulose chain axes, B polarization perpendicular to the chain axes, A difference spectrum of U and B + C. (From Maréchal and Chanzy 2000)

Appendix

223

224

Appendix

Fig. A.2 Fourier transform IR spectra from oriented films of cellulose III1 obtained from cellulose treated in supercritical ammonia: a nonpolarized and b polarized radiation. (From Wada et al. 2001)

Appendix

225

Fig. A.3 Spectra of cellulose I: upper graph cellulose Iα (originally termed type A, Valonia); lower graph cellulose Iβ (originally type B, plant cellulose). (From Marrinan and Mann 1956)

Fig. A.4 Spectra of cellulose II: upper graph oriented film; lower graph nonoriented film. (From Marrinan and Mann 1956)

226

Appendix

Fig. A.5 Spectra of the two subgroups of cellulose III: upper graph cellulose III1 prepared from cellulose I; lower graph cellulose III2 prepared from cellulose II. (From Marrinan and Mann 1956)

Fig. A.6 Spectra of the two subgroups of cellulose IV: upper graph cellulose IV1 prepared from cellulose I (note the similarity with the spectra of cellulose Iβ); lower graph cellulose IV2 prepared from cellulose II. (From Marrinan and Mann 1956)

Appendix

227

Fig. A.7 Raman spectra of cellulose I (Cladophora), cellulose II and cellulose III1 in the conformation-sensitive region. (From Atalla and VanderHart 1989)

Fig. A.8 Raman spectra of native cellulose Iα and cellulose Iβ in the OH stretching region. (From Atalla and VanderHart 1989)

228

Fig. A.9 IR spectra of cellulose tripropionate and cellulose. (From Shuto 1990)

Appendix

Appendix

229 Table A.1 Cartesian coordinates of methyl (1–4)-b-d-cellobiose [MM3 (92)]. (From Rivet et al. 2001) X Y Z Atoms −0.634 −2.021 −2.853 −2.870 −1.443 −1.408 3.722 3.467 1.981 1.445 1.837 1.475 5.465 −0.001 −1.869 −4.206 −3.497 −0.789 −0.046 5.117 3.841 1.893 3.262 1.764 −0.046 −2.540 −2.439 −3.428 −0.900 −1.848 −2.002 3.232 4.073 1.407 1.851 1.339 2.030 0.393 6.567 5.139 4.999 −1.517 −4.208 −4.391 −0.034 3.769 1.014 2.698

1.259 1.076 2.337 2.715 2.842 3.114 −1.518 −0.128 0.216 −0.002 −1.395 −1.589 −3.091 −0.003 0.893 2.051 4.001 1.580 3.317 −1.712 −0.150 1.613 −1.528 −2.936 2.027 0.197 3.178 1.952 3.654 2.262 4.026 −2.314 0.627 −0.408 0.770 −2.183 −0.875 −1.405 −3.180 −3.712 −3.482 0.026 1.858 3.917 3.506 0.735 1.792 −3.071

0.001 −0.605 −0.411 1.065 1.591 3.091 0.926 1.486 1.413 −0.004 −0.507 −1.978 0.616 −0.006 −2.021 −0.795 1.187 1.388 3.491 0.824 2.872 1.740 −0.428 −2.371 −0.541 −0.144 −1.019 1.661 1.049 3.660 3.336 1.546 0.925 2.147 −0.705 0.108 −2.630 −2.140 0.520 1.480 −0.318 −2.168 −1.723 0.884 4.420 3.204 2.051 −2.277

C11 C21 C31 C41 C51 C61 C12 C22 C32 C42 C52 C62 C12M O42 O21 O31 O41 O51 O61 O12 O22 O32 O52 O62 H11 H21 H31 H41 H51 H61A H61B H12 H22 H32 H42 H52 H62A H62B H12M H22M H32M H021 H031 H041 H061 H022 H032 H062

230

Appendix Table A.2 Cartesian coordinates of a standard b-d-glucose residue. (From Arnott and Scott 1972) X Y Z Atoms 0.000 −0.504 0.041 1.559 1.973 3.475 −0.413 −1.927 −0.330 2.057 1.429 4.034 −0.367 −0.185 −0.394 2.001 1.605

0.000 1.437 2.194 2.074 0.608 0.429 −0.610 1.436 3.571 2.626 0.000 1.047 −0.518 1.938 1.781 2.625 0.076

0.000 0.000 −1.201 −1.257 −1.181 −1.117 1.178 −0.007 −1.109 −2.474 0.000 0.042 −0.898 0.926 −2.124 −0.414 −2.071

C1 C2 C3 C4 C5 C6 O1 O2 O3 O4 O5 O6 H1 H2 H3 H4 H5

Table A.3 Fractional coordinates of methyl (1–4)-b-d-cellobiose methanol (converted to d configuration). Monoclinic unit cell, space group P21: a=7.652 Å, b=25.532 Å, c=4.496 Å, a =90.00°, b =101.84°, g =90.00°, V=859.6 Å3, rcalc=1.500 g cm−3 (T=−193°C). (From Ham and Williams 1970) Atoms x y z C11 C21 C31 C41 C51 C61 C12 C22 C32 C42 C52 C62 C12M O41 O21 O31 O51 O61

0.6350 0.4450 0.4297 0.5756 0.7563 0.9062 1.0495 1.1134 0.9783 0.7890 0.7475 0.5635 1.1465 0.5793 0.3193 0.2586 0.7532 1.0701

0.8299 0.8501 0.9046 0.9428 0.9151 0.9477 0.6657 0.7198 0.7631 0.7478 0.6929 0.6730 0.5782 0.9866 0.8161 0.9270 0.8669 0.9186

0.6311 0.6084 0.4615 0.6103 0.6488 0.8312 0.9114 0.8460 0.8650 0.7029 0.8044 0.6668 0.9358 0.4209 0.4298 0.4645 0.8091 0.8947 (continued)

Appendix

231 Table A.3 (continued) Atoms x

y

z

O42 O12 O22 O32 O52 O62 C1Me O1Me H11 H21 H31 H41 H51 H61A H61B H12 H22 H32 H42 H52 H62A H62B H12Ma H12Mb H12Mc H021 H031 H041 H061 H022 H032 H062 H1Mea H1Meb H1Mec H0Me

0.7822 0.6302 0.7293 0.8080 0.6573 0.6656 0.5530 0.5608 0.8250 0.8510 0.9000 0.9530 0.9060 0.9570 0.9800 0.6620 0.7180 0.7710 0.7480 0.6930 0.6380 0.7070 0.5760 0.5480 0.5610 0.8100 0.9250 1.0120 0.9270 0.7590 0.8340 0.6920 0.5653 0.5749 0.5130 0.5920

0.7829 0.8254 1.0606 0.7317 0.7260 0.3442 0.1961 0.2865 0.4250 0.8260 0.2450 0.8050 0.4550 1.0500 0.7400 1.1400 0.6350 1.0950 0.4700 1.0100 0.7400 0.6850 1.1460 0.8500 0.8800 0.4940 0.3340 0.4700 0.7080 1.1600 0.8260 0.2400 −0.0302 0.3389 0.2141 0.2700

0.6574 1.1657 1.2776 1.0435 0.8755 0.5444 0.7369 0.5713 0.6760 0.4000 0.4580 0.5480 0.7880 0.8660 0.9300 1.0400 1.1200 0.9900 0.7600 0.7500 0.5400 0.4740 1.1580 1.2240 1.0220 0.2330 0.1900 0.5870 1.1480 1.2880 1.0040 0.5060 0.7244 0.8361 0.7695 0.5500

232

Appendix Table A.4 Fractional coordinates of dimethyl (1–4)-b-d-cellobioside. Monoclinic unit cell, space group P21: a=6.6060 Å, b=14.074 Å, c=9.318 Å, a =90.00°, b=108.95°, g =90.00°, V=819.4 Å3, rcalc=1.501 g cm−3. (From Mackie et al. 2002) Atoms x y z C11 C21 O21 C31 O31 C41 O41 C41M C51 O51 C61 O61 C12 O12 C12M C22 O22 C32 O32 C42 O42 C52 O52 C62 O62 H11 H21 H021 H31 H031 H41 H41Ma H41Mb H41Mc H51 H61A H61B H061 H12

0.5554 0.6731 0.7808 0.5093 0.6081 0.3694 0.1853 0.1987 0.2857 0.4569 0.1809 0.1015 0.7102 0.6714 0.8574 0.5044 0.4453 0.5314 0.3254 0.6237 0.7020 0.8186 0.7609 0.8977 0.7475 0.4430 0.7810 0.9100 0.4200 0.6400 0.4530 0.0810 0.2140 0.3270 0.1870 0.2810 0.0650 0.1910 0.8290

0.1445 0.0827 0.0068 0.0414 −0.0171 0.1185 0.0767 0.0628 0.1874 0.2200 0.2743 0.3358 0.3238 0.3494 0.3457 0.3395 0.4369 0.3050 0.3103 0.2052 0.1818 0.1946 0.2259 0.0938 0.0349 0.1084 0.1177 0.0190 0.0050 0.0140 0.1573 0.0350 0.1370 0.0410 0.1525 0.3070 0.2560 0.3840 0.3540

0.8089 0.9446 0.9001 1.0114 1.1389 1.0455 1.0700 1.2235 0.9125 0.8619 0.9522 0.8267 0.3645 0.2140 0.1693 0.4015 0.3899 0.5606 0.5771 0.5872 0.7462 0.5328 0.3787 0.5352 0.4306 0.7330 1.0220 0.9360 0.9360 1.2240 1.1310 1.2300 1.2700 1.2800 0.8310 1.0350 0.9840 0.8420 0.4300 (continued)

Appendix

233 Table A.4 (continued) Atoms x H12Ma 0.9100 H12Mb 0.8190 H12Mc 0.9720 H22 0.3940 H022 0.3840 H32 0.6290 H032 0.3460 H42 0.5060 H52 0.9350 H62A 1.0230 H62B 0.9220 H062 0.7000

y 0.2870 0.3720 0.3900 0.3020 0.4520 0.3423 0.3070 0.1590 0.2339 0.0920 0.0680 −0.0010

z 0.1780 0.0630 0.2270 0.3320 0.3030 0.6200 0.6720 0.5310 0.5990 0.5110 0.6360 0.4710

x

0.3110 0.4743 0.4175 0.2408 0.0868 −0.0903 0.2074 0.3506 0.6315 0.5714 0.1716 0.1544 −0.2321 0.2202 0.0635 0.1273 0.3082 0.4561 0.6347 −0.0951 −0.0152 0.3818 0.3846 0.7675 0.3402 0.5040

Atoms

C11 C21 C31 C41 C51 C61 C71 O11 O21 O31 O41 O51 O61 C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 C13 C23

0.4656 0.4538 0.4357 0.4299 0.4432 0.4385 0.4948 0.4822 0.4607 0.4248 0.4136 0.4591 0.4506 0.3982 0.3851 0.3678 0.3621 0.3758 0.3719 0.3912 0.3554 0.3463 0.3918 0.3849 0.3304 0.3179

y

0.4609 0.5403 0.4523 0.4596 0.3850 0.3959 0.4470 0.5386 0.5252 0.5286 0.3669 0.4757 0.3092 0.4557 0.3786 0.4735 0.4695 0.5430 0.5350 0.3956 0.4057 0.5642 0.4498 0.6119 0.4761 0.5543

z 0.1999 0.0673 0.1048 0.3144 0.4245 0.6341 0.2799 0.1757 −0.1181 0.0058 0.3545 0.3850 0.7306 0.3416 0.4758 0.4459 0.2420 0.1247 −0.0834 0.6617 0.5531 0.2071 0.1552 −0.1890 0.2251 0.0872

x 0.3256 0.3386 0.3576 0.3618 0.3476 0.3495 0.2952 0.3097 0.3342 0.3689 0.3772 0.3313 0.3347 0.3933 0.4060 0.4250 0.4296 0.4159 0.4191 0.4003 0.4359 0.4457 0.3994 0.4047 0.4615 0.4746

y −0.0567 −0.0424 −0.0841 −0.0030 −0.0275 0.0678 −0.0162 0.0087 −0.1471 −0.0275 −0.0735 0.0406 0.0537 −0.0069 −0.0226 0.0205 −0.0591 −0.0358 −0.1289 0.0792 −0.0293 0.0085 −0.1024 −0.1268 −0.0617 −0.0557

z 0.6499 0.5664 0.5336 0.7187 0.7986 0.9861 0.7384 0.6881 0.3942 0.4563 0.6951 0.8210 1.0403 0.7359 0.8164 0.8472 0.6610 0.5878 0.3983 0.9902 0.9213 0.6834 0.5651 0.3450 0.6370 0.5505

x 0.0733 0.0852 0.1032 0.1096 0.0959 0.1006 0.0443 0.0570 0.0785 0.1147 0.1255 0.0799 0.0884 0.1409 0.1542 0.1715 0.1772 0.1633 0.1675 0.1480 0.1838 0.1929 0.1472 0.1539 0.2089 0.2209

y 0.4594 0.5397 0.4517 0.4595 0.3862 0.3970 0.4462 0.5396 0.5250 0.5257 0.3670 0.4757 0.3090 0.4557 0.3801 0.4746 0.4685 0.5425 0.5354 0.3941 0.4042 0.5646 0.4503 0.6125 0.4758 0.5544

z 0.2454 0.3891 0.3106 0.1825 0.0487 −0.0641 0.2041 0.3336 0.4711 0.4666 0.0719 0.1555 −0.1764 0.1512 0.0007 0.0748 0.1991 0.3404 0.4533 −0.0834 −0.0821 0.3017 0.2425 0.5637 0.2140 0.3568

x 0.2132 0.2003 0.1814 0.1773 0.1921 0.1896 0.2437 0.2296 0.2049 0.1703 0.1620 0.2075 0.2044 0.1459 0.1331 0.1145 0.1094 0.1237 0.1200 0.1389 0.1032 0.0936 0.1395 0.1345 0.0776 0.0645

y

−0.0567 −0.0436 −0.0844 −0.0033 −0.0267 0.0678 −0.0159 0.0081 −0.1479 −0.0273 −0.0737 0.0411 0.0546 −0.0074 −0.0242 0.0225 −0.0587 −0.0363 −0.1294 0.0779 −0.0290 0.0089 −0.1021 −0.1251 −0.0612 −0.0572

z

Table A.5 Fractional coordinates of methyl b-cellotrioside. Monoclinic unit cell, space group P21: a=7.998 Å, b=76.380 Å, c=8.991 Å, b=116.40°, V=4,919 Å3, rcalc=1.48 g cm−3, antiparallel molecule arrangements. (From Raymond et al. 1995a) Molecule u Molecule d Molecule e Molecule v

234 Appendix

C33 C43 C53 C63 O23 O33 O43 O53 O63 O1 O1p C1 C1p C2 O1wp O1w O3w H11 H21 H31 H41 H51 H61A H61B H12 H22 H32 H42 H52 H62A

0.3003 0.2936 0.3071 0.3020 0.3250 0.2884 0.2781 0.3232 0.3133

0.4671 0.4523 0.4373 0.4278 0.4453 0.4382 0.4260 0.4006 0.3833 0.3700 0.3599 0.3779 0.3707

0.4498 0.2693 0.1163 −0.0569 0.6588 0.5969 0.2048 0.1816 −0.2057

0.2777 0.5077 0.3891 0.2695 0.0568 −0.0647 −0.1340 0.2473 0.0299 0.1549 0.2821 0.4838 0.6093

0.3347 0.6665 0.3273 0.5841 0.2602 0.5210 0.3440 0.5794 0.2527 0.5976 0.3458 0.6672 0.4104

0.4662 0.4675 0.3997 0.4156 0.5366 0.5327 0.3670 0.4900 0.3214

0.1737 0.0851 0.0515 0.3604 0.3861 0.6706 0.6753 0.3697 0.4547 0.4972 0.1977 0.1627 −0.1181

0.1159 0.3199 0.4492 0.6506 −0.1004 0.0002 0.3398 0.4080 0.7737

0.3241 0.3384 0.3589 0.3638 0.3460 0.3519 0.3604 0.3919 0.4058 0.4260 0.4313 0.4144 0.4225

0.4925 0.4983 0.4836 0.4874 0.4688 0.5049 0.5125 0.4681 0.4739

−0.1812 0.0806 −0.2134 0.1248 −0.1546 0.1936 0.0204 0.1182 −0.1465 0.1499 −0.1870 0.0913 −0.2523

−0.1221 −0.0377 −0.0320 0.0787 −0.1506 −0.1033 −0.1314 0.0392 0.0846

0.5180 0.6988 0.7834 0.9742 0.3786 0.4368 0.6628 0.8081 1.0275 0.3350 0.4638 0.2989 0.2999 0.1797 0.6208 0.3935 0.7953 0.5573 0.6596 0.4377 0.8138 0.7043 1.0858 0.9783 0.8322 0.7236 0.9436 0.5639 0.6848 0.2985

0.2389 0.2458 0.2319 0.2372 0.2145 0.2508 0.2610 0.2160 0.2259 0.0047 0.0346 0.0041 0.0340 0.0202 0.5059 0.5336 0.2923 0.0717 0.0867 0.1016 0.1118 0.0939 0.1008 0.1131 0.1385 0.1561 0.1694 0.1795 0.1612 0.1687

0.4676 0.4685 0.3984 0.4163 0.5368 0.5333 0.3678 0.4894 0.3222 0.3694 0.6295 0.5025 0.4962 0.4980 0.7294 0.2746 −0.1299 0.3333 0.6658 0.3266 0.5840 0.2612 0.5220 0.3456 0.5794 0.2545 0.5990 0.3447 0.6667 0.4110

0.0470 0.0409 0.0558 0.0518 0.0704 0.0339 0.0267 0.0711 0.0652

0.2459 0.2696 0.2145 0.2004 0.1800 0.1749 0.1939 0.1872 0.1787 0.1462 0.1331 0.1137 0.1073 0.1253 0.1169

0.2632 0.1428 0.0180 −0.0735 0.4501 0.3960 0.0310 0.1305 −0.1904

0.5768 0.8117 0.1467 0.4942 0.2348 0.2641 −0.0400 0.0258 −0.1513 0.2477 −0.1015 0.1524 0.1148 0.4300 0.3632

(continued)

−0.1293 0.1195 −0.1805 0.0794 −0.2137 0.1245 −0.1535 0.1934 0.0192 0.1185 −0.1483 0.1520 −0.1857 0.0905 −0.2533

−0.1217 −0.0391 −0.0325 0.0781 −0.1504 −0.1027 −0.1299 0.0382 0.0839

Appendix 235

Atoms H62B H13 H23 H33 H43 H53 H63A H63B

0.6874 0.3079 0.5394 0.4212 0.2962 0.0811 −0.0302 −0.0940

x

y 0.3600 0.3326 0.3163 0.3024 0.2906 0.3093 0.3026 0.2891

Table A.5 (continued) Molecule u

z 0.5955 0.3510 0.6808 0.3417 0.5900 0.2740 0.5409 0.3723

x −0.1173 0.2045 0.1116 0.0764 0.3543 0.4313 0.6680 0.6873

y 0.4297 0.4596 0.4762 0.4911 0.5025 0.4808 0.4901 0.4987

Molecule d z −0.0727 −0.1843 0.0683 −0.2496 0.0836 −0.1526 0.1993 0.0337

x 0.4068 0.5441 0.6416 0.4224 0.7950 0.6931 1.0720 0.9685

y 0.1793 0.2067 0.2224 0.2369 0.2489 0.2297 0.2365 0.2501

Molecule e z 0.5978 0.3507 0.6811 0.3428 0.5907 0.2726 0.5417 0.3737

0.5410 0.1121 0.4564 0.1757 0.2297 −0.0846 0.0295 −0.1549

x

y 0.1093 0.0789 0.0629 0.0484 0.0358 0.0586 0.0491 0.0405

Molecule v z −0.0742 −0.1837 0.0667 −0.2492 0.0823 −0.1529 0.1987 0.0324

236 Appendix

Appendix

237

Table A.6 Fractional coordinates of cellotetraose. Triclinic unit cell, space group P1: a=8.023 Å, b=8.951 Å, c=22.445 Å, a =89.26°, b =85.07°, g =63.93°, V=1,443.4 Å3, rcalc=1.56 g cm−3, antiparallel molecule arrangements. (From Gessler et al. 1995) Molecule a Molecule b Atoms

x

y

z

x

y

z

C11 C21 C31 C41 C51 C61 O11β O11α O21 O31 O41 O51 O61 C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 C14 C24 C34 C44 C54 C64 O24 O34

−0.1857 −0.0143 −0.0607 −0.2304 −0.3934 −0.5663 −0.1537 −0.2245 0.1328 0.0953 −0.2879 −0.3402 −0.7217 −0.2288 −0.3722 −0.2899 −0.1087 0.0272 0.2088 −0.5373 −0.4291 −0.0201 −0.0590 0.3282 −0.0530 0.1182 0.0721 −0.0982 −0.2608 −0.4340 0.2729 0.2327 −0.1552 −0.2056 −0.5802 −0.0886 −0.2379 −0.1602 0.0229 0.1614 0.3355 −0.3989 −0.2881

0.0179 −0.0542 0.0400 0.0300 0.1019 0.0936 −0.0727 0.1400 −0.0344 −0.0242 0.1246 0.0051 0.1838 0.0357 0.1101 0.0128 0.0179 −0.0540 −0.0475 0.0932 0.0840 −0.0798 0.0416 −0.1227 0.0089 −0.0676 0.0261 0.0181 0.0897 0.0810 −0.0537 −0.0438 0.1126 −0.0043 0.1587 0.0248 0.1017 0.0170 0.0168 −0.0498 −0.0321 0.0820 0.0810

0.8777 0.8333 0.7750 0.7519 0.8013 0.7827 0.9287 0.9024 0.8580 0.7340 0.6991 0.8529 0.8257 0.6453 0.5999 0.5422 0.5212 0.5688 0.5545 0.6218 0.4997 0.4668 0.6239 0.6003 0.4131 0.3693 0.3113 0.2892 0.3366 0.3202 0.3925 0.2697 0.2363 0.3900 0.3666 0.1814 0.1374 0.0771 0.0559 0.1039 0.0887 0.1588 0.0331

0.0250 0.1462 0.0954 −0.1188 −0.2177 −0.4268 0.0567

0.4992 0.5066 0.4641 0.5483 0.5212 0.6145 0.5701

0.1671 0.2115 0.2757 0.2899 0.2394 0.2489 0.1141

0.3366 0.1866 −0.1748 −0.1653 −0.5113 −0.1747 −0.3188 −0.3038 −0.1041 0.0269 0.2296 −0.4983 −0.4178 −0.0814 0.0067 0.3477 −0.1054 0.0177 −0.0291 −0.2370 −0.3417 −0.5539 0.2085 0.0699 −0.2862 −0.2981 −0.6434 −0.2825 −0.4371 −0.4225 −0.2307 −0.0793 0.1143 −0.6156 −0.5622

0.3976 0.5212 0.4779 0.5947 0.6009 0.5482 0.5299 0.5756 0.4958 0.5200 0.4279 0.6307 0.5216 0.5681 0.4493 0.4272 0.4982 0.5119 0.4674 0.5516 0.5256 0.6212 0.4102 0.5189 0.4827 0.5950 0.6149 0.5544 0.5567 0.6347 0.5381 0.5275 0.4152 0.6448 0.6412

0.1936 0.3140 0.3426 0.1869 0.1973 0.3985 0.4411 0.5051 0.5213 0.4733 0.4840 0.4231 0.5428 0.5742 0.4180 0.4337 0.6297 0.6724 0.7355 0.7528 0.7049 0.7181 0.6526 0.7746 0.8069 0.6512 0.6667 0.8600 0.9067 0.9634 0.9867 0.9368 0.9522 0.8858 1.0089 (continued)

238

Appendix

Table A.6 (continued) Molecule a Atoms O44 O54 O64 H11 H21 H31 H41 H51 H61A H61B H21O H31O H61O H12 H22 H32 H42 H52 H62A H62B H22O H32O H13 H23 H33 H43 H53 H63A H63B H23O H33O H63O H14 H24 H34 H44 H54 H64A H64B H24O H64O O1W

x

Molecule b

y

z

0.1111 0.0744 0.4832

−0.0824 0.0403 −0.1432

0.0039 0.1592 0.1205

0.0269 −0.0962 −0.1938 −0.4233 −0.5432 −0.5957

−0.1804 0.1659 −0.0943 0.2261 −0.0306 0.1461

0.8245 0.7822 0.7411 0.8095 0.7803 0.7411

−0.7237 −0.2109 −0.4065 −0.2573 −0.1403 0.0558 0.1854 0.2718 −0.6283 −0.3833 −0.0824 0.1558 0.0382 −0.0610 −0.2928 −0.4076 −0.4737 0.3371 0.2111 −0.6735 −0.0580 −0.2754 −0.1359 −0.0120 0.2030 0.3107 0.3741 −0.4985 0.5179 0.4347

0.1216 −0.0861 0.2356 −0.1129 0.1422 −0.1793 0.0765 −0.1112 0.1906 0.0156 0.1333 −0.1938 0.1513 −0.1070 0.2153 −0.0434 0.1425 −0.1776 −0.0389 0.0985 −0.1004 0.2286 −0.1064 0.1412 −0.1771 0.0898 −0.0569 0.2070 −0.0735 0.3667

0.8420 0.6514 0.5934 0.5473 0.5132 0.5719 0.5511 0.5141 0.6324 0.4583 0.4206 0.3619 0.3176 0.2795 0.3423 0.3147 0.2806 0.4055 0.2371 0.3721 0.1867 0.1335 0.0834 0.0469 0.1070 0.0990 0.0432 0.1825 0.1285 0.0598

y

z

−0.2141 −0.1083 0.2488 0.0552 0.1248 0.1388 −0.1563 −0.1802 −0.4639 −0.4745 0.3936 0.1530 −0.5088 −0.2037 −0.2958 −0.3484 −0.0731 −0.0081 0.2592 0.2548

x

0.6266 0.4537 0.4182 0.3742 0.6303 0.3367 0.6747 0.3940 0.7399 0.5653 0.4776 0.4778 0.6896 0.6737 0.4057 0.7037 0.3705 0.6470 0.3051 0.4853

1.0367 0.8844 0.9074 0.1607 0.2130 0.2807 0.2967 0.2357 0.2580 0.2851 0.2193 0.3508 0.1552 0.3963 0.4404 0.5100 0.5267 0.4726 0.4941 0.5199

−0.0715 −0.0049 0.0111 −0.2726 −0.3036 −0.5892 −0.5979

0.3719 0.6359 0.3398 0.6777 0.3982 0.7451 0.5673

0.6248 0.6741 0.7383 0.7592 0.7020 0.7293 0.7536

−0.3003 −0.4199 −0.4548 −0.2153 −0.0907 0.1288 0.1360

0.6752 0.4357 0.7621 0.4203 0.6474 0.2938 0.4554

0.8515 0.9143 0.9568 0.9992 0.9307 0.9557 0.9929

0.2355

0.3449

0.8748

Appendix

239

Table A.7 Fractional coordinates of cellotetraose. Triclinic unit cell, space group P1: a=8.045 Å, b =9.003 Å, c =22.510 Å, a =89.66°, b =94.83°, g =115.80°, V=1,461.8 Å3, rcalc=1.535 g cm−3, antiparallel molecule arrangements. (From Raymond et al. 1995b) Molecule u Molecule d Atoms

x

y

C11 C21 C31 C41 C51 C61 O11 O21 O31 O41 O51 O61 C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 C14 C24 C34 C44 C54 C64 O24 O34 O44 O54

0.7041 0.8190 0.7659 0.5622 0.4498 0.2534 0.7284 1.0028 0.8558 0.4991 0.5056 0.1608 0.5004 0.3576 0.3684 0.5654 0.6999 0.8978 0.1725 0.2508 0.5934 0.6779 1.0184 0.5645 0.6869 0.6423 0.4346 0.3247 0.1190 0.8739 0.7409 0.3807 0.3762 0.0244 0.3763 0.2329 0.2425 0.4381 0.5757 0.7739 0.0522 0.1050 0.4537 0.5544

0.2929 0.2765 0.3165 0.2422 0.2654 0.1786 0.2108 0.3789 0.2659 0.3029 0.1856 0.1804 0.2384 0.2537 0.2056 0.2815 0.2610 0.3507 0.1486 0.2542 0.2064 0.3342 0.3485 0.2765 0.2611 0.3111 0.2297 0.2484 0.1534 0.3625 0.2610 0.2908 0.1809 0.1564 0.2100 0.2197 0.1417 0.2358 0.2396 0.3417 0.1274 0.1406 0.1528 0.3141

z 0.8579 0.8149 0.7518 0.7366 0.7875 0.7758 0.9122 0.8329 0.7135 0.6842 0.8388 0.8280 0.6298 0.5857 0.5211 0.5066 0.5537 0.5427 0.6028 0.4833 0.4513 0.6084 0.5932 0.3977 0.3536 0.2911 0.2728 0.3222 0.3096 0.3738 0.2516 0.2205 0.3758 0.3595 0.1666 0.1209 0.0630 0.0411 0.0896 0.0732 0.1419 0.0181 −0.0105 0.1431

x 0.4848 0.6468 0.6040 0.4340 0.2755 0.1046 0.5167 0.8040 0.7634 0.3811 0.3282 −0.0503 0.4426 0.3013 0.3803 0.5599 0.6968 0.8783 0.1340 0.2431 0.6478 0.6100 0.9950 0.6182 0.7896 0.7431 0.5726 0.4154 0.2403 0.9403 0.9028 0.5169 0.4639 0.0953 0.5762 0.4373 0.5092 0.6960 0.8288 1.0035 0.2698 0.3841 0.7816 0.7474

y

z

0.7580 0.8252 0.7289 0.7409 0.6711 0.6807 0.8425 0.8104 0.7997 0.6556 0.7623 0.5948 0.7424 0.6713 0.7646 0.7604 0.8313 0.8202 0.6802 0.6966 0.8552 0.7337 0.8970 0.7752 0.8510 0.7573 0.7665 0.6950 0.7032 0.8331 0.8296 0.6731 0.7817 0.6242 0.7558 0.6803 0.7574 0.7577 0.8344 0.8103 0.6975 0.6869 0.8560 0.7450

0.1499 0.1926 0.2511 0.2728 0.2247 0.2425 0.0975 0.1685 0.2914 0.3259 0.1723 0.1988 0.3796 0.4234 0.4852 0.5042 0.4553 0.4720 0.4034 0.5252 0.5579 0.4014 0.4255 0.6107 0.6560 0.7137 0.7357 0.6881 0.7040 0.6329 0.7557 0.7887 0.6339 0.6602 0.8434 0.8879 0.9476 0.9688 0.9202 0.9352 0.8681 0.9913 1.0207 0.8646 (continued)

240

Appendix

Table A.7 (continued) Molecule u Atoms O64 O1W H11 H21 H31 H41 H51 H61A H61B H11O H21O H31O H61O H12 H22 H32 H42 H52 H62A H62B H22O H32O H62O H13 H23 H33 H43 H53 H63A H63B H23O H33O H63O H14 H24 H34 H44 H54 H64A H64B H24O H34O H44O H64O H1OW H2OW

x 0.9071 0.0875 0.7288 0.8028 0.8137 0.5198 0.4880 0.2122 0.2202 0.6499 1.0097 0.8608 0.1351 0.4776 0.3755 0.3204 0.6049 0.6668 0.9207 0.9256 0.1747 0.2940 1.0457 0.5904 0.6681 0.6883 0.3964 0.3571 0.0891 0.0772 0.9367 0.6692 0.0240 0.3591 0.2564 0.2206 0.4598 0.5574 0.7936 0.7899 0.0132 0.0759 0.5533 1.0070 0.1455 0.0304

y 0.3525 0.4145 0.4079 0.1622 0.4367 0.1233 0.3827 0.2293 0.0653 0.1145 0.4617 0.1818 0.0941 0.1226 0.3685 0.0852 0.3996 0.1435 0.3004 0.4642 0.0706 0.3552 0.2717 0.3923 0.1461 0.4313 0.1115 0.3654 0.0398 0.2002 0.3884 0.1743 0.2469 0.0964 0.3350 0.0274 0.3487 0.1266 0.2917 0.4519 0.0297 0.2146 0.2083 0.4295 0.3759 0.3563

Molecule d z 0.1201 0.9679 0.8657 0.8144 0.7497 0.7299 0.7966 0.7433 0.7636 0.9105 0.8502 0.7257 0.8456 0.6329 0.5869 0.5176 0.5009 0.5587 0.5082 0.5337 0.6213 0.4838 0.5890 0.4044 0.3514 0.2907 0.2658 0.3294 0.2988 0.2760 0.3452 0.2340 0.3628 0.1724 0.1153 0.0700 0.0315 0.0975 0.0381 0.0635 0.1320 0.0257 −0.0239 0.1146 0.9545 1.0067

x 1.1494

y 0.9197

z 0.9023

0.4559 0.6804 0.5744 0.4647 0.2469 0.0753 0.1273 0.5864 0.7647 0.7864 −0.0325 0.4680 0.2746 0.4045 0.5373 0.7215 0.9354 0.8567 0.1561 0.2151 0.9537 0.5899 0.8229 0.7150 0.6005 0.3878 0.2061 0.2632 1.0107 0.9454 0.1191 0.6015 0.4066 0.5310 0.6790 0.8590 1.0409 0.9782 0.2958 0.4278 0.7366 1.1717

0.6426 0.9418 0.6130 0.8574 0.5555 0.6319 0.7954 0.9402 0.7351 0.8960 0.6448 0.8580 0.5550 0.8802 0.6456 0.9465 0.8770 0.7057 0.7760 0.5979 0.8395 0.6593 0.9681 0.6414 0.8820 0.5792 0.6513 0.8178 0.9210 0.9305 0.6744 0.8718 0.5625 0.8731 0.6444 0.9517 0.8312 0.6971 0.7758 0.7390 0.9207 1.0154

0.1406 0.2013 0.2428 0.2803 0.2160 0.2810 0.2461 0.1050 0.1432 0.2990 0.1676 0.3736 0.4281 0.4801 0.5105 0.4488 0.5100 0.4749 0.3962 0.5281 0.3947 0.6029 0.6636 0.7057 0.7441 0.6805 0.7418 0.7085 0.6195 0.7550 0.6291 0.8388 0.8898 0.9450 0.9753 0.9164 0.9776 0.9256 0.8460 1.0230 1.0252 0.9108

Table A.8 Cartesian and fractional coordinates of tetraacetyl-b-d-glucoside. Orthorhombic unit cell, space group P212121: a =8.089 Å, b =13.251 Å, c =17.359 Å, a =b = g =90.0°, V=1,860.6 Å3, rcalc=1.293 g cm−3, antiparallel packing of molecules. (From Zugenmaier and Rappenecker 1978) X Y Z Atoms x y z 3.310 2.359 2.882 2.045 3.295 4.650 4.320 0.760 3.888 4.603 2.707 3.010 2.428 2.868 2.659 3.239 2.817 3.132 1.928 2.683 5.213 2.306 2.572 1.695 6.642 1.699 3.964 1.456 3.810 1.634 2.799 3.155 2.944 3.414 1.893 1.383 2.661 2.249 3.122 1.852 3.236 1.205 2.111 1.165 7.078 7.029 6.714

4.532 4.705 2.365 0.005 2.270 −0.191 5.812 2.312 −1.205 −2.369 3.480 3.565 2.355 1.059 1.102 −0.093 4.690 5.748 2.369 −1.081 −1.403 6.773 2.459 −2.050 −1.370 3.472 3.591 2.465 0.835 1.126 −0.928 −0.013 3.750 5.420 5.009 6.811 7.672 6.559 1.683 2.624 3.167 −1.617 −2.677 −2.359 −0.663 −2.160 −1.391

9.079 6.338 4.861 6.324 8.900 8.777 6.173 4.107 5.836 9.193 8.410 6.930 6.225 6.885 8.386 9.105 10.426 5.959 3.885 5.803 8.844 5.244 2.543 5.235 8.410 8.575 6.839 6.232 6.701 8.558 8.818 10.138 10.988 10.884 10.502 5.572 5.416 4.444 2.309 1.927 2.552 4.531 4.687 5.746 8.870 8.732 7.395

O1 O2 O3 O4 O5 O6 O2C O3C O4C O6C C1 C2 C3 C4 C5 C6 C1M C2C C3C C4C C6C C2M C3M C4M C6M H1 H2 H3 H4 H5 H6A H6B H1Ma H1Mb H1Mc H2Ma H2Mb H2Mc H3Ma H3Mb H3Mc H4Ma H4Mb H4Mc H6Ma H6Mb H6Mc

0.4092 0.2916 0.3563 0.2528 0.4074 0.5749 0.5340 0.0940 0.4807 0.5690 0.3347 0.3721 0.3001 0.3545 0.3287 0.4004 0.3482 0.3872 0.2383 0.3317 0.6444 0.2851 0.3180 0.2096 0.8211 0.2100 0.4900 0.1800 0.4710 0.2020 0.3460 0.3900 0.3640 0.4220 0.2340 0.1710 0.3290 0.2780 0.3860 0.2290 0.4000 0.1490 0.2610 0.1440 0.8750 0.8690 0.8300

0.3420 0.3551 0.1785 0.0004 0.1713 −0.0144 0.4386 0.1745 −0.0909 −0.1788 0.2626 0.2690 0.1777 0.0799 0.0832 −0.0070 0.3539 0.4338 0.1788 −0.0816 −0.1059 0.5111 0.1856 −0.1547 −0.1034 0.2620 0.2710 0.1860 0.0630 0.0850 −0.0700 −0.0010 0.2830 0.4090 0.3780 0.5140 0.5790 0.4950 0.1270 0.1980 0.2390 −0.1220 −0.2020 −0.1780 −0.0500 −0.1630 −0.1050

0.5230 0.3651 0.2800 0.3643 0.5127 0.5056 0.3556 0.2366 0.3362 0.5296 0.4845 0.3992 0.3586 0.3966 0.4831 0.5245 0.6006 0.3433 0.2238 0.3343 0.5095 0.3021 0.1465 0.3016 0.4845 0.4940 0.3940 0.3590 0.3860 0.4930 0.5080 0.5840 0.6330 0.6270 0.6050 0.3210 0.3120 0.2560 0.1330 0.1110 0.1470 0.2610 0.2700 0.3310 0.5110 0.5030 0.4260

242

Appendix

Table A.9 Cartesian and fractional coordinates of b-d, 1→4 xylose hexaacetate. Orthorhombic unit cell, space group P212121: a =10.936 Å, b =8.377 Å, c =29.976 Å, a =b =g =90.0°, V=2,746.1 Å3, antiparallel packing arrangements. (From Leung and Marchessault 1973) X Y Z Atoms x y z −0.394 −0.531 −1.628 −1.844 −2.127 −1.037 −1.874 −2.735 −3.183 −1.173 −4.030 −3.248 −1.171 −2.955 −2.197 −0.909 −2.958 −1.664 −3.945 2.451 2.701 1.456 0.919 0.822 3.591 4.064 1.388 4.964 4.289 1.847 3.668 3.022 1.786 2.094 2.554 4.686 0.731 0.252 −2.362 −1.050 −3.073 −0.241 −1.203 −1.422

2.088 3.470 3.903 5.406 5.740 5.204 2.714 6.704 7.710 2.347 7.169 9.213 3.585 5.810 7.171 3.786 2.309 7.119 7.060 −0.781 0.401 1.272 1.551 0.277 −2.823 0.518 2.771 −3.461 −0.146 4.098 −1.493 −0.076 2.525 −0.398 −3.415 1.403 2.000 3.920 3.376 5.973 5.311 5.772 5.252 1.608

3.180 3.255 2.296 2.404 3.846 4.736 0.261 0.650 4.625 −0.983 0.009 4.499 0.980 1.604 3.933 4.562 0.597 0.342 5.294 3.591 4.481 4.517 3.150 2.332 3.267 6.436 6.385 3.189 7.737 6.859 3.462 5.773 5.144 2.299 3.171 5.974 7.029 2.938 2.548 2.098 4.197 4.646 5.695 −1.259

O42(O11) C11 C21 C31 C41 C51 C21C C31C C41C C21M C31M C41M O21 O31 O41 O51 O21C O31C O41C C12 C22 C32 C42 C52 C12C C22C C32C C12M C22M C32M O12 O22 O32 O52 O12C O22C O32C H11 H21 H31 H41 H51A H51B H21Ma

−0.0360 −0.0486 −0.1489 −0.1686 −0.1945 −0.0948 −0.1714 −0.2501 −0.2911 −0.1073 −0.3685 −0.2970 −0.1071 −0.2702 −0.2009 −0.0831 −0.2705 −0.1522 −0.3607 0.2241 0.2470 0.1331 0.0840 0.0752 0.3284 0.3716 0.1269 0.4539 0.3922 0.1689 0.3354 0.2763 0.1633 0.1915 0.2335 0.4285 0.0668 0.0230 −0.2160 −0.0960 −0.2810 −0.0220 −0.1100 −0.1300

0.2492 0.4142 0.4659 0.6453 0.6852 0.6212 0.3240 0.8003 0.9204 0.2802 0.8558 1.0998 0.4279 0.6936 0.8560 0.4519 0.2756 0.8498 0.8428 −0.0932 0.0479 0.1518 0.1852 0.0331 −0.3370 0.0618 0.3308 −0.4132 −0.0174 0.4892 −0.1782 −0.0091 0.3014 −0.0475 −0.4077 0.1675 0.2388 0.4680 0.4030 0.7130 0.6340 0.6890 0.6270 0.1920

0.1061 0.1086 0.0766 0.0802 0.1283 0.1580 0.0087 0.0217 0.1543 −0.0328 0.0003 0.1501 0.0327 0.0535 0.1312 0.1522 0.0199 0.0114 0.1766 0.1198 0.1495 0.1507 0.1051 0.0778 0.1090 0.2147 0.2130 0.1064 0.2581 0.2288 0.1155 0.1926 0.1716 0.0767 0.1058 0.1993 0.2345 0.0980 0.0850 0.0700 0.1400 0.1550 0.1900 −0.0420 (continued)

Appendix

243

Table A.9 (continued) X Y −1.323 −0.372 −4.604 −3.926 −4.134 −3.128 −3.992 −2.452 1.804 3.390 0.886 1.498 0.612 0.087 4.965 5.534 5.151 3.489 4.582 4.790 1.454 1.487 2.570

3.150 2.077 6.635 7.523 7.925 9.491 9.583 9.630 −1.525 0.955 0.905 2.178 0.486 −0.356 −4.281 −2.815 −3.778 −0.612 0.285 0.913 4.859 4.130 4.205

Z −1.589 −0.869 0.210 −1.019 0.420 3.417 4.766 5.043 4.017 4.167 5.096 2.728 1.349 2.733 4.017 3.807 2.278 7.944 8.273 7.434 6.385 7.674 6.954

Atoms

x

y

z

H21Mb H21Mc H31Ma H31Mb H31Mc H41Ma H41Mb H41Mc H12 H22 H32 H42 H52A H52B H12Ma H12Mb H12Mc H22Ma H22Mb H22Mc H32Ma H32Mb H33Mc

−0.1210 −0.0340 −0.4210 −0.3590 −0.3780 −0.2860 −0.3650 −0.2242 0.1650 0.3100 0.0810 0.1370 0.0560 0.0079 0.4540 0.5060 0.4710 0.3190 0.4190 0.4380 0.1330 0.1360 0.2350

0.3760 0.2480 0.7920 0.8980 0.9460 1.1330 1.1440 1.1495 −0.1820 0.1140 0.1080 0.2600 0.0580 −0.0425 −0.5110 −0.3360 −0.4510 −0.0730 0.0340 0.1090 0.5800 0.4930 0.5020

−0.0530 −0.0290 0.0070 −0.0340 0.0140 0.1140 0.1590 0.1682 0.1340 0.1390 0.1700 0.0910 0.0450 0.0912 0.1340 0.1270 0.0760 0.2650 0.2760 0.2480 0.2130 0.2560 0.2320

Table A.10 Cartesian and fractional coordinates of β-cellotriose undecaacetate. Monoclinic unit cell, space group P21: a =5.675 Å, b =37.216 Å, c =11.709 Å, a =g =90.0°, b =94.78°, V=2,464.3 Å3, rcalc=1.303 g cm−3, parallel-down arrangement of molecules. (From Pérez and Brisse 1977) X Y Z Atoms x y z 1.353 2.036 2.641 0.390 0.352 0.193 4.278 1.687 2.090 −1.438 0.787 1.893 1.440 0.991 −0.199 −0.762 3.342 3.237

10.138 11.719 14.410 15.489 11.998 13.643 11.797 15.046 16.472 13.744 11.358 12.214 13.677 14.183 13.264 13.733 11.649 11.355

7.912 10.276 9.445 7.997 7.024 4.920 10.137 11.340 6.903 3.402 8.290 8.927 9.048 7.742 7.245 5.939 10.769 12.260

O42(O11) O21 O31 O41 O51 O61 O21C O31C O41C O61C C11 C21 C31 C41 C51 C61 C21C C21M

0.2393 0.3601 0.4670 0.0689 0.0622 0.0341 0.7565 0.2983 0.3695 −0.2543 0.1391 0.3347 0.2547 0.1752 −0.0352 −0.1348 0.5910 0.5724

0.2724 0.3149 0.3872 0.4162 0.3224 0.3666 0.3170 0.4043 0.4426 0.3693 0.3052 0.3282 0.3675 0.3811 0.3564 0.3690 0.3130 0.3051

0.6854 0.8922 0.8255 0.6858 0.6024 0.4216 0.8963 0.9805 0.6045 0.2803 0.7136 0.7759 0.7830 0.6683 0.6173 0.5018 0.9436 1.0702 (continued)

244 Table A.10 (continued) X Y 2.598 15.035 3.920 15.720 1.107 16.565 0.261 17.838 −0.288 13.632 0.825 13.461 −0.193 4.965 −0.744 6.777 −1.186 9.453 0.696 6.632 1.066 8.325 −2.939 6.963 −0.445 9.926 2.551 8.180 0.422 6.178 −0.575 7.190 −0.019 8.601 0.515 8.973 1.504 7.853 2.060 8.076 −2.037 6.658 −1.930 6.077 −1.281 9.981 −2.661 10.625 1.434 8.299 0.301 8.645 1.913 0.000 −0.291 0.919 −0.059 3.766 1.714 1.872 2.056 3.052 0.957 −1.474 1.117 0.622 −2.162 4.131 0.158 3.684 1.011 1.020 0.678 1.757 −0.075 3.033 0.651 3.859 0.884 3.000 1.669 3.781 1.870 −1.146 3.050 −2.062 0.045 0.458 −1.003 −0.246 −1.201 4.317 −1.043 4.998

Appendix

Z 10.593 10.857 7.476 7.782 3.675 2.694 7.284 5.218 6.405 8.529 10.763 5.422 4.296 12.336 7.228 6.593 6.605 8.010 8.480 9.820 4.801 3.370 5.153 4.969 12.088 12.916 8.015 9.691 9.915 6.797 4.218 6.750 11.313 9.879 3.278 7.741 8.930 8.666 7.604 6.392 5.403 7.385 7.681 10.876 11.583 10.439 11.653

Atoms C31C C31M C41C C41M C61C C61M O43(O12) O22 O32 O52 O62 O22C O32C O62C C12 C22 C32 C42 C52 C62 C22C C22M C32C C32M C62C C62M O13 O23 O33 O53 O63 O13C O23C O33C O63C C13 C23 C33 C43 C53 C63 C13C C13M C23C C23M C33C C33M

x

y

0.4594 0.6931 0.1958 0.0462 −0.0509 0.1459 −0.0342 −0.1316 −0.2098 0.1231 0.1885 −0.5197 −0.0786 0.4510 0.0747 −0.1016 −0.0034 0.0911 0.2660 0.3642 −0.3602 −0.3412 −0.2265 −0.4705 0.2536 0.0532 0.3383 −0.0514 −0.0104 0.3030 0.3635 0.1692 0.1976 −0.3823 0.0279 0.1788 0.1199 −0.0132 0.1151 0.1563 0.2951 0.3306 0.5394 0.0079 −0.1774 −0.2123 −0.1844

0.4040 0.4224 0.4451 0.4793 0.3663 0.3617 0.1334 0.1821 0.2540 0.1782 0.2237 0.1871 0.2667 0.2198 0.1660 0.1932 0.2311 0.2411 0.2110 0.2170 0.1789 0.1633 0.2682 0.2855 0.2230 0.2323 0.0000 0.0247 0.1012 0.0503 0.0820 −0.0396 0.0167 0.1110 0.0990 0.0274 0.0472 0.0815 0.1037 0.0806 0.1016 −0.0308 −0.0554 0.0123 −0.0066 0.1160 0.1343

z 0.9232 0.9552 0.6464 0.6665 0.3118 0.2360 0.6207 0.4403 0.5385 0.7334 0.9268 0.4421 0.3637 1.0718 0.6203 0.5590 0.5640 0.6878 0.7350 0.8534 0.3955 0.2740 0.4309 0.4054 1.0426 1.1052 0.6982 0.8256 0.8464 0.5927 0.3749 0.5833 0.9742 0.8283 0.2811 0.6683 0.7675 0.7396 0.6541 0.5522 0.4734 0.6441 0.6778 0.9292 0.9821 0.8830 0.9878 (continued)

Appendix Table A.10 (continued) X Y 1.177 3.055 1.542 2.252 −0.020 11.221 2.760 12.289 0.633 13.628 1.674 14.336 −0.950 13.145 −1.083 14.728 −1.586 13.132 2.256 11.284 3.665 12.065 3.665 10.502 4.575 15.564 4.343 15.378 3.806 16.717 −0.633 17.536 −0.079 18.318 0.690 18.504 1.714 13.405 0.854 14.261 0.696 12.661 1.305 6.121 −1.403 7.302 0.690 8.642 −0.266 9.114 2.268 7.934 2.754 8.753 2.596 7.227 −0.973 5.962 −2.358 6.706 −2.398 5.180 −3.207 10.540 −3.150 10.242 −2.551 11.619 −0.520 8.865 0.509 9.498 0.090 7.934 0.057 0.566 1.504 1.943 −0.933 2.910 1.568 4.205 0.153 2.650 2.358 4.213 1.080 4.659 3.687 −1.593 3.552 −2.263 2.731 −2.895

245

Z 3.193 1.981 8.948 8.387 9.829 7.073 7.854 6.033 5.688 12.527 12.784 12.492 10.120 11.721 10.981 8.284 6.962 8.384 3.124 2.036 2.108 6.661 7.307 5.949 8.698 7.901 9.886 10.063 3.079 2.691 3.292 5.795 4.209 4.756 12.385 13.470 13.575 7.466 9.335 8.403 7.982 6.017 5.880 5.272 8.356 6.869 8.132

Atoms C63C C63M H11 H21 H31 H41 H51 H61A H61B H21Ma H21Mb H21Mc H31Ma H31Mb H31Mc H41Ma H41Mb H41Mc H61Ma H61Mb H61Mc H12 H22 H32 H42 H52 H62A H62B H22Ma H22Mb H22Mc H32Ma H32Mb H32Mc H62Ma H62Mb H62Mc H13 H23 H33 H43 H53 H63A H63B H13Ma H13Mb H13Mc

x

y

0.2081 0.2726 −0.0035 0.4880 0.1120 0.2960 −0.1680 −0.1914 −0.2804 0.3990 0.6480 0.6480 0.8090 0.7680 0.6730 −0.1120 −0.0140 0.1220 0.3030 0.1510 0.1230 0.2307 −0.2480 0.1220 −0.0470 0.4010 0.4870 0.4590 −0.1720 −0.4170 −0.4240 −0.5670 −0.5570 −0.4510 −0.0920 0.0900 0.0160 0.0100 0.2660 −0.1650 0.2772 0.0270 0.4170 0.1910 0.6520 0.6280 0.4830

0.0821 0.0605 0.3015 0.3302 0.3662 0.3852 0.3532 0.3958 0.3529 0.3032 0.3242 0.2822 0.4182 0.4132 0.4492 0.4712 0.4922 0.4972 0.3602 0.3832 0.3402 0.1645 0.1962 0.2322 0.2449 0.2132 0.2352 0.1942 0.1602 0.1802 0.1392 0.2832 0.2752 0.3122 0.2382 0.2552 0.2132 0.0152 0.0522 0.0782 0.1130 0.0712 0.1132 0.1252 −0.0428 −0.0608 −0.0778

z 0.2811 0.1802 0.7640 0.7360 0.8440 0.6160 0.6640 0.5075 0.4745 1.0860 1.1180 1.0930 0.8970 1.0320 0.9650 0.7030 0.5940 0.7210 0.2790 0.1800 0.1850 0.5782 0.6140 0.5130 0.7409 0.6910 0.8640 0.8780 0.2560 0.2130 0.2640 0.4720 0.3370 0.3880 1.0540 1.1540 1.1600 0.6380 0.8080 0.7110 0.6929 0.5150 0.5190 0.4580 0.7400 0.6120 0.7140 (continued)

246 Table A.10 X −1.855 −1.244 −0.735 −0.096 −1.612 −1.340 2.426 1.595 0.848

Appendix (continued) Y −0.328 0.194 −1.183 5.032 4.622 5.999 1.831 2.873 1.571

Z 11.044 12.469 11.806 1.951 12.371 11.552 2.116 1.178 1.814

Atoms H23Ma H23Mb H23Mc H33Ma H33Mb H33Mc H63Ma H63Mb H63Mc

x −0.3280 −0.2200 −0.1300 −0.0170 −0.2850 −0.2370 0.4290 0.2820 0.1500

y −0.0088 0.0052 −0.0318 0.1352 0.1242 0.1612 0.0492 0.0772 0.0422

z 0.9300 1.0560 1.0030 1.0200 1.0450 0.9770 0.1980 0.1120 0.1610

Table A.11 Cartesian and fractional coordinates of native cellulose Iβ. Monoclinic unit cell, space group P21: a =7.784 Å, b =8.201 Å, c(fiber axis)=10.38 Å, a =b =90.00°, g =96.55°, V=658.3 Å3, rcalc=1.636 g cm−3, parallel-up chain arrangements (T=20°C). Note the partial occupancy of some hydroxyl deuteriums; residues 3 and 4 are symmetry-related to residues 1 and 2, respectively, by a 21 screw axis along the chain axes. (From Nishiyama et al. 2002) X Y Z Atoms x y z Corner chain 0.107 −0.201 0.312 −0.055 0.200 −0.361 0.471 −0.207 0.608 −0.430 0.371 1.069 −1.166 1.289 −1.018 1.165 −1.293 −0.315 0.283 −0.001 0.367 −0.111 0.226 −0.107 0.201 −0.312 0.055 −0.200

−0.353 −1.488 −1.159 0.256 1.284 2.655 −2.632 −2.114 0.640 0.862 3.262 −0.277 −1.651 −1.233 0.261 1.360 2.575 3.212 −2.652 −2.901 −2.010 4.051 2.755 0.353 1.488 1.159 −0.256 −1.284

0.449 −0.536 −1.918 −2.336 −1.279 −1.587 −0.003 −2.864 −3.554 −0.055 −2.637 0.621 −0.567 −1.910 −2.517 −1.131 −1.846 −0.794 0.955 0.800 −3.651 −2.958 −3.456 5.639 4.654 3.272 2.855 3.911

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61 H11 H21 H31 H41 H51 H61A H61B D211 D221 D031 D611 D621 C13 C23 C33 C43 C53

0.0138 −0.0260 0.0403 −0.0071 0.0258 −0.0467 0.0609 −0.0268 0.0786 −0.0556 0.0480 0.1382 −0.1508 0.1667 −0.1316 0.1507 −0.1672 −0.0407 0.0366 −0.0001 0.0475 −0.0144 0.0292 −0.0138 0.0260 −0.0403 0.0071 −0.0258

−0.0415 −0.1843 −0.1369 0.0304 0.1594 0.3187 −0.3144 −0.2607 0.0865 0.0991 0.4030 −0.0188 −0.2177 −0.1323 0.0176 0.1821 0.2959 0.3872 −0.3194 −0.3537 −0.2400 0.4924 0.3391 0.0415 0.1843 0.1369 −0.0304 −0.1594

0.0433 −0.0516 −0.1848 −0.2250 −0.1232 −0.1529 −0.0003 −0.2759 −0.3424 −0.0053 −0.2540 0.0598 −0.0546 −0.1840 −0.2425 −0.1090 −0.1778 −0.0765 0.0920 0.0771 −0.3517 −0.2850 −0.3329 0.5433 0.4484 0.3152 0.2750 0.3768 (continued)

Appendix

247

Table A.11 (continued) X Y 0.361 −2.655 −0.471 2.632 0.207 2.114 −0.608 −0.640 0.430 −0.862 −0.371 −3.262 −1.069 0.277 1.166 1.651 −1.289 1.233 1.018 −0.261 −1.165 −1.360 1.293 −2.575 0.315 −3.212 −0.283 2.652 0.001 2.901 −0.367 2.010 0.111 −4.051 −0.226 −2.755 Center chain 4.121 3.669 4.214 3.730 4.183 3.498 4.081 3.762 4.350 3.751 4.191 5.086 2.691 5.194 2.754 5.158 2.584 3.473 3.894 3.609 3.947 4.009 4.133 3.612 4.064 3.519

3.257 2.185 2.490 3.888 4.911 6.283 0.896 1.476 4.306 4.547 7.012 3.202 2.199 2.482 3.898 5.009 6.135 6.800 0.859 0.236 1.788 7.966 6.532 4.056 5.128 4.823

Z 3.603 5.187 2.326 1.636 5.135 2.553 5.811 4.623 3.280 2.673 4.059 3.344 4.396 6.145 5.990 1.539 2.232 1.734

Atoms C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B D213 D223 D033 D613 D623

x 0.0467 −0.0609 0.0268 −0.0786 0.0556 −0.0480 −0.1382 0.1508 −0.1667 0.1316 −0.1507 0.1672 0.0407 −0.0366 0.0001 −0.0475 0.0144 −0.0292

y −0.3187 0.3144 0.2607 −0.0865 −0.0991 −0.4030 0.0188 0.2177 0.1323 −0.0176 −0.1821 −0.2959 −0.3872 0.3194 0.3537 0.2400 −0.4924 −0.3391

3.159 2.174 0.794 0.389 1.441 1.168 2.616 −0.084 −0.837 2.739 0.175 3.322 2.130 0.823 0.300 1.422 0.879 1.989 3.579 2.070 −0.998 0.338 −0.671 8.349 7.364 5.984

C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 H12 H22 H32 H42 H52 H62A H62B D212 D222 D032 D612 D622 C14 C24 C34

0.5329 0.4745 0.5449 0.4823 0.5409 0.4523 0.5277 0.4865 0.5625 0.4850 0.5420 0.6577 0.3480 0.6716 0.3561 0.6670 0.3341 0.4491 0.5035 0.4667 0.5104 0.5184 0.5345 0.4671 0.5255 0.4551

0.4548 0.3178 0.3626 0.5263 0.6574 0.8151 0.1664 0.2326 0.5859 0.6070 0.9137 0.4616 0.3058 0.3753 0.5139 0.6830 0.7843 0.8778 0.1593 0.0793 0.2733 1.0275 0.8543 0.5452 0.6822 0.6374

z 0.3471 0.4997 0.2241 0.1576 0.4947 0.2460 0.5598 0.4454 0.3160 0.2575 0.3910 0.3222 0.4235 0.5920 0.5771 0.1483 0.2150 0.1671 0.3043 0.2094 0.0765 0.0375 0.1388 0.1125 0.2520 −0.0081 −0.0806 0.2639 0.0169 0.3200 0.2052 0.0793 0.0289 0.1370 0.0847 0.1916 0.3448 0.1994 −0.0961 0.0326 −0.0646 0.8043 0.7094 0.5765 (continued)

248

Appendix

Table A.11 (continued) X Y 4.003 3.425 3.550 2.402 4.235 1.030 3.652 6.417 3.971 5.837 3.383 3.008 3.983 2.766 3.542 0.301 2.647 4.111 5.042 5.114 2.540 4.832 4.979 3.415 2.575 2.304 5.150 1.178 4.260 0.513 3.840 6.454 4.124 7.077 3.786 5.525 3.724 −0.653 3.600 0.782

Z 5.579 6.631 6.358 7.806 5.106 4.353 7.929 5.365 8.512 7.320 6.013 5.490 6.612 6.069 7.179 8.769 7.260 4.192 5.528 4.519

Atoms C44 C54 C64 O24 O34 O44 O54 O64 H14 H24 H34 H44 H54 H64A H64B D214 D224 D034 D614 D624

x 0.5177 0.4591 0.5477 0.4723 0.5135 0.4375 0.5150 0.4580 0.3423 0.6520 0.3284 0.6439 0.3330 0.6659 0.5509 0.4965 0.5333 0.4896 0.4816 0.4655

y 0.4737 0.3426 0.1849 0.8336 0.7674 0.4141 0.3930 0.0863 0.5384 0.6942 0.6247 0.4861 0.3170 0.2157 0.1222 0.8407 0.9207 0.7267 −0.0275 0.1457

z 0.5375 0.6388 0.6125 0.7520 0.4919 0.4194 0.7639 0.5169 0.8200 0.7052 0.5793 0.5289 0.6370 0.5847 0.6916 0.8448 0.6994 0.4039 0.5326 0.4354

Table A.12 Cartesian and fractional coordinates of native cellulose Iα. Triclinic unit cell, space group P1: a =6.717 Å, b =5.962 Å, c =10.400 Å, a =118.08°, b =114.80°, g =80.37°, V=333.3 Å3, rcalc = 1.616 g cm−3, parallel-up chain arrangements. Note the partial occupancy of some hydroxyl deuteriums. (From Nishiyama et al. 2003) X Y Z Atoms x y z −0.332 −0.872 −1.062 0.211 0.839 2.276 −2.109 −1.481 0.019 0.924 2.263 −0.966 −0.233 −1.768 0.859 0.296 2.766 2.720 −1.804

0.176 1.262 0.638 −0.024 −0.954 −1.328 1.726 1.669 −0.802 −0.267 −2.258 −0.563 2.003 −0.039 0.682 −1.764 −0.532 −1.718 2.172

2.704 1.755 0.396 −0.092 0.930 0.566 2.293 −0.502 −1.280 2.184 −0.506 2.818 1.696 0.454 −0.296 1.028 0.303 1.335 3.102

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61 H11 H21 H31 H41 H51 H61A H61B H211

−0.0545 −0.1431 −0.1743 0.0346 0.1377 0.3736 −0.3461 −0.2431 0.0031 0.1516 0.3714 −0.1585 −0.0383 −0.2902 0.1410 0.0485 0.4540 0.4464 −0.2960

0.0310 0.2337 0.1136 −0.0030 −0.1754 −0.2361 0.3131 0.3067 −0.1524 −0.0442 −0.4130 −0.1140 0.3792 −0.0201 0.1359 −0.3333 −0.0814 −0.3072 0.4000

0.2536 0.1930 0.0215 −0.0003 0.0794 0.0919 0.2112 −0.0314 −0.1634 0.2391 −0.0595 0.1973 0.2550 −0.0404 0.0464 0.0220 0.1302 0.1664 0.3260 (continued)

Appendix Table A.12 X −2.620 −1.146 2.620 1.523 0.291 0.896 1.096 −0.228 −0.902 −2.329 2.146 1.643 −0.055 −0.965 −2.357 0.896 0.294 1.726 −0.824 −0.380 −2.803 −2.780 2.029 2.748 1.840 −3.095 −1.493

249 (continued) Y 1.887 1.279 −3.039 −1.898 −0.145 −1.179 −0.496 0.050 0.953 1.304 −1.594 −1.433 0.827 0.279 2.241 0.614 −1.947 0.247 −0.703 1.776 0.501 1.679 −1.286 −1.313 −0.848 2.799 2.108

Z 1.453 −1.347 −0.048 −1.064 7.907 6.956 5.623 5.109 6.122 5.755 7.508 4.704 3.915 7.385 4.691 8.041 6.858 5.735 4.912 6.220 5.491 6.528 8.433 6.759 3.953 5.002 4.276

Atoms H221 H031 H611 H621 C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B H213 H223 H033 H613 H623

x −0.4300 −0.1880 0.4300 0.2500 0.0478 0.1470 0.1799 −0.0375 −0.1480 −0.3823 0.3522 0.2697 −0.0090 −0.1584 −0.3869 0.1471 0.0483 0.2833 −0.1353 −0.0623 −0.4600 −0.4562 0.3330 0.4510 0.3020 −0.5080 −0.2450

y

z

0.3400 0.2350 −0.5590 −0.3500 −0.0255 −0.2178 −0.0864 0.0079 0.1747 0.2313 −0.2877 −0.2607 0.1569 0.0461 0.4092 0.1231 −0.3680 0.0593 −0.1395 0.3350 0.0751 0.2993 −0.2300 −0.2300 −0.1480 0.5100 0.3900

0.1150 −0.1170 −0.0390 −0.1290 0.7664 0.6499 0.5661 0.4832 0.5957 0.5122 0.7397 0.4550 0.4163 0.6796 0.4567 0.8462 0.5732 0.6442 0.3980 0.6716 0.4236 0.5849 0.8390 0.7100 0.4220 0.4810 0.4500

Table A.13 Cartesian and fractional coordinates of cellulose II. Monoclinic unit cell, space group P21: a=8.10 Å, b=9.03 Å, c(fiber axis)=10.31 Å, a =b =90.0°, g =117.10°, V=671.3 Å3, rcalc=1.604 g cm−3, antiparallel chain arrangements. Residues 3 and 4 are symmetry-related to residues 1 and 2, respectively, by a 21 screw axis. (From Langan et al. 2001) X Y Z Atoms x y z Corner chain −0.310 0.222 −0.901 1.238 −1.089 0.530 0.245 −0.053 0.851 −0.950 2.149 −1.578 −2.156 1.663 −1.615 1.450 0.079 −0.862 0.959 −0.184 2.430 −2.643

3.928 2.949 1.608 1.155 2.227 1.887 3.444 0.680 −0.010 3.433 2.784

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61

−0.0430 −0.1250 −0.1510 0.0340 0.1180 0.2980 −0.2990 −0.2240 0.0110 0.1330 0.3370

0.0070 0.0860 −0.0030 0.0080 −0.0570 −0.0530 0.0620 0.0690 −0.0910 0.0340 −0.1550

0.3810 0.2860 0.1560 0.1120 0.2160 0.1830 0.3340 0.0660 −0.0010 0.3330 0.2700 (continued)

250

Appendix

Table A.13 (continued) X Y −0.906 −0.296 −1.728 0.864 0.214 2.116 2.855 −2.741 −1.751 2.485 0.310 0.901 1.089 −0.245 −0.851 −2.149 2.156 1.615 −0.079 −0.959 −2.430 0.906 0.296 1.728 −0.864 −0.214 −2.116 −2.855 2.741 1.751 −2.485 Center chain 3.836 4.918 4.341 3.540 2.596 1.752 5.552 5.408 2.733 3.310 0.822 3.135 5.593 3.750

Z

Atoms

x

y

z

−0.543 2.002 −0.204 0.680 −1.673 −1.921 −0.917 1.489 1.067 −2.352 −0.222 −1.238 −0.530 0.053 0.950 1.578 −1.663 −1.450 0.862 0.184 2.643 0.543 −2.002 0.204 −0.680 1.673 1.921 0.917 −1.489 −1.067 2.352

4.046 2.850 1.729 0.958 2.392 0.976 1.932 2.893 −0.037 3.550 9.083 8.104 6.763 6.310 7.382 7.042 8.599 5.835 5.145 8.588 7.939 9.201 8.005 6.884 6.113 7.547 6.131 7.087 8.048 5.118 8.705

H11 H21 H31 H41 H51 H61A H61B H021 H031 H061 C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B H023 H033 H063

−0.1257 −0.0411 −0.2396 0.1198 0.0297 0.2935 0.3960 −0.3801 −0.2428 0.3446 0.0430 0.1250 0.1510 −0.0340 −0.1180 −0.2980 0.2990 0.2240 −0.0110 −0.1330 −0.3370 0.1257 0.0411 0.2396 −0.1198 −0.0297 −0.2935 −0.3960 0.3801 0.2428 −0.3446

−0.1115 0.2049 −0.1205 0.1243 −0.1731 −0.0928 0.0603 0.0096 0.0189 −0.1196 −0.0070 −0.0860 0.0030 −0.0080 0.0570 0.0530 −0.0620 −0.0690 0.0910 −0.0340 0.1550 0.1115 −0.2049 0.1205 −0.1243 0.1731 0.0928 −0.0603 −0.0096 −0.0189 0.1196

0.3924 0.2764 0.1677 0.0929 0.2320 0.0947 0.1874 0.2806 −0.0036 0.3443 0.8810 0.7860 0.6560 0.6120 0.7160 0.6830 0.8340 0.5660 0.4990 0.8330 0.7700 0.8924 0.7764 0.6677 0.5929 0.7320 0.5947 0.6874 0.7806 0.4964 0.8443

2.353 1.944 1.761 2.992 3.421 4.594 0.744 1.540 2.764 3.607 4.835 1.670 2.659 0.978

3.609 4.598 6.000 6.382 5.268 5.536 4.206 6.908 7.547 4.042 4.464 3.529 4.633 6.001

0.5320 0.6820 0.6020 0.4910 0.3600 0.2430 0.7700 0.7500 0.3790 0.4590 0.1140 0.4348 0.7757 0.5200

0.4780 0.4940 0.4410 0.5320 0.5260 0.6080 0.3970 0.4770 0.4610 0.5870 0.5820 0.3626 0.6114 0.3208

0.3500 0.4460 0.5820 0.6190 0.5110 0.5370 0.4080 0.6700 0.7320 0.3920 0.4330 0.3423 0.4494 0.5821

C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 H12 H22 H32

(continued)

Appendix

251

Table A.13 (continued) X Y 4.159 3.727 1.986 2.668 1.258 4.453 2.312 5.374 5.482 0.193 5.113 1.448 1.224 5.153 3.375 2.987 2.293 3.396 2.870 3.579 3.670 2.348 4.615 1.919 5.459 0.746 1.658 4.596 1.803 3.800 4.478 2.576 3.901 1.733 6.389 0.505 4.076 3.670 1.617 2.681 3.461 4.362 3.052 1.613 5.225 2.672 5.953 0.887 4.899 −0.034 1.728 5.147 2.098 3.892 5.987 0.187

Z 6.567 5.118 6.357 5.649 4.812 7.664 3.827 −1.546 −0.557 0.845 1.227 0.113 0.381 −0.949 1.753 2.392 −1.113 −0.691 −1.626 −0.522 0.846 1.412 −0.037 1.202 0.494 −0.343 2.509 −1.328

Atoms H42 H52 H62A H62B H022 H032 H062 C14 C24 C34 C44 C54 C64 O24 O34 O44 O54 O64 H14 H24 H34 H44 H54 H64A H64B H024 H034 H064

x 0.5768 0.2754 0.1744 0.3206 0.7603 0.7091 0.1697 0.4680 0.3180 0.3980 0.5090 0.6400 0.7570 0.2300 0.2500 0.6210 0.5410 0.8860 0.5652 0.2243 0.4800 0.4232 0.7246 0.8256 0.6794 0.2397 0.2909 0.8303

y 0.6484 0.4080 0.5644 0.7261 0.3321 0.4501 0.6400 0.5220 0.5060 0.5590 0.4680 0.4740 0.3920 0.6030 0.5230 0.5390 0.4130 0.4180 0.6374 0.3886 0.6792 0.3516 0.5920 0.4356 0.2739 0.6679 0.5499 0.3600

z 0.6370 0.4964 0.6166 0.5479 0.4667 0.7434 0.3712 −0.1500 −0.0540 0.0820 0.1190 0.0110 0.0370 −0.0920 0.1700 0.2320 −0.1080 −0.0670 −0.1577 −0.0506 0.0821 0.1370 −0.0036 0.1166 0.0479 −0.0333 0.2434 −0.1288

Table A.14 Cartesian and fractional coordinates of cellulose III1. Monoclinic unit cell, space group P21: a =4.450 Å, b =7.850 Å, c(fiber axis)=10.31 Å, a =b =90.0°, g =105.10°, V=347.7 Å3, rcalc=1.549 g cm−3, parallel chain arrangements (T=20°C). One residue is sufficient to describe the structure in P21 of a one-chain unit cell. Residue 3 is symmetry-related to residue 1 by a 21 screw axis. (From Wada et al. 2004) X Y Z Atoms x y z 0.219 0.822 0.203 0.040 −0.639 −0.757 0.780 1.043 −0.821

0.391 1.318 1.204 −0.256 −1.067 −2.529 2.670 1.931 −0.376

3.923 2.960 1.582 1.163 2.252 1.948 3.420 0.686 0.007

C11 C21 C31 C41 C51 C61 O21 O31 O41

0.0510 0.1914 0.0472 0.0093 −0.1487 −0.1761 0.1816 0.2427 −0.1910

0.0573 0.1961 0.1604 −0.0313 −0.1579 −0.3482 0.3670 0.2818 −0.0761

0.3805 0.2871 0.1534 0.1128 0.2184 0.1889 0.3317 0.0665 0.0007 (continued)

252

Appendix

Table A.14 (continued) X Y 0.166 −1.023 −0.707 1.767 −0.681 0.918 −1.530 −1.473 0.069 1.719 0.644 −0.516 −0.219 −0.822 −0.203 −0.040 0.639 0.757 −0.780 −1.043 0.821 −0.166 1.023 0.707 −1.767 0.681 −0.918 1.530 1.473 −0.069 −1.719 −0.644 0.516 −0.821

−0.946 −3.269 0.682 1.071 1.627 −0.643 −0.698 −2.670 −2.844 3.006 1.946 −4.100 −0.391 −1.318 −1.204 0.256 1.067 2.529 −2.670 −1.931 0.376 0.946 3.269 −0.682 −1.071 −1.627 0.643 0.698 2.670 2.844 −3.006 −1.946 4.100 −0.376

Z 3.436 3.134 4.059 2.874 1.598 0.962 2.424 1.309 1.548 3.402 −0.206 3.093 9.078 8.115 6.737 6.318 7.407 7.103 8.575 5.841 5.162 8.591 8.289 9.214 8.029 6.753 6.117 7.579 6.464 6.703 8.557 4.949 8.248 10.317

Atoms

x

y

z

O51 O61 H11 H21 H31 H41 H51 H61A H61B D021 D031 D061 C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B D023 D033 D063 O45

0.0386 −0.2382 −0.1645 0.4112 −0.1585 0.2137 −0.3562 −0.3429 0.0160 0.4000 0.1500 −0.1200 −0.0510 −0.1914 −0.0472 −0.0093 0.1487 0.1761 −0.1816 −0.2427 0.1910 −0.0386 0.2382 0.1645 −0.4112 0.1585 −0.2137 0.3562 0.3429 −0.0160 −0.4000 −0.1500 0.1200 −0.1910

−0.1148 −0.4516 0.0626 0.1972 0.1839 −0.0503 −0.1415 −0.3908 −0.3599 0.4420 0.2700 −0.5400 −0.0573 −0.1961 −0.1604 0.0313 0.1579 0.3482 −0.3670 −0.2818 0.0761 0.1148 0.4516 −0.0626 −0.1972 −0.1839 0.0503 0.1415 0.3908 0.3599 −0.4420 −0.2700 0.5400 −0.0761

0.3333 0.3040 0.3937 0.2788 0.1550 0.0933 0.2351 0.1270 0.1501 0.3300 −0.0200 0.3000 0.8805 0.7871 0.6534 0.6128 0.7184 0.6889 0.8317 0.5665 0.5007 0.8333 0.8040 0.8937 0.7788 0.6550 0.5933 0.7351 0.6270 0.6501 0.8300 0.4800 0.8000 1.0007

Table A.15 Cartesian and fractional coordinates of cellulose IV1. Space group P1: a =8.03 Å, b =8.13 Å, c(fiber axis)=10.34 Å, a =b =g =90.0°, V=675 Å3, rcalc=1.595 g cm−3, parallel chain arrangements. (From Gardiner and Sarko 1985) X Y Z Atoms x y z Corner chain −0.043 0.340 0.186 1.478 −0.389 1.109 0.129 −0.239

3.916 2.933 1.583 1.122

C11 C21 C31 C41

−0.0054 0.0232 −0.0484 0.0161

0.0418 0.1818 0.1364 −0.0294

0.3787 0.2837 0.1531 0.1085 (continued)

Appendix

253

Table A.15 (continued) X

Y

Z

Atoms

x

y

z

−0.073 0.538 −0.413 −0.058 −0.565 0.545 −0.319 −1.072 1.219 −1.435 1.153 −1.102 0.703 1.458 0.052 −0.181 0.400 −0.122 0.081 −0.531 0.413 0.074 0.574 −0.539 0.467 1.082 −1.215 1.445 −1.146 1.110 −0.989 −1.258

−1.294 −2.626 2.668 2.102 −0.651 −0.861 −3.357 0.198 1.635 1.056 −0.159 −1.423 −3.189 −2.458 −0.353 −1.492 −1.125 0.223 1.281 2.615 −2.686 −2.118 0.636 0.848 3.514 −0.212 −1.644 −1.069 0.142 1.410 3.035 2.465

2.200 1.831 3.426 0.610 −0.047 3.416 0.943 4.066 2.830 1.666 0.905 2.365 2.701 1.354 9.088 8.100 6.753 6.289 7.371 7.005 8.593 5.778 5.122 8.588 6.506 9.236 7.994 6.841 6.071 7.537 7.851 6.262

C51 C61 O21 O31 O41 O51 O61 H11 H21 H31 H41 H51 H61A H61B C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B

−0.0091 0.0670 −0.0514 −0.0072 −0.0704 0.0679 −0.0397 −0.1335 0.1518 −0.1787 0.1436 −0.1372 0.0875 0.1816 0.0065 −0.0225 0.0498 −0.0152 0.0101 −0.0661 0.0514 0.0092 0.0715 −0.0671 0.0582 0.1347 −0.1513 0.1800 −0.1427 0.1382 −0.1232 −0.1567

−0.1592 −0.3230 0.3282 0.2585 −0.0801 −0.1059 −0.4129 0.0244 0.2011 0.1299 −0.0196 −0.1750 −0.3923 −0.3023 −0.0434 −0.1835 −0.1384 0.0274 0.1576 0.3216 −0.3304 −0.2605 0.0782 0.1043 0.4322 −0.0261 −0.2022 −0.1315 0.0175 0.1734 0.3733 0.3032

0.2128 0.1771 0.3313 0.0590 −0.0045 0.3304 0.0912 0.3932 0.2737 0.1611 0.0875 0.2287 0.2612 0.1309 0.8789 0.7834 0.6531 0.6082 0.7129 0.6775 0.8310 0.5588 0.4954 0.8306 0.6292 0.8932 0.7731 0.6616 0.5871 0.7289 0.7593 0.6056

4.408 5.509 5.214 3.812 2.790 1.394 6.764 6.159 3.488 3.144 0.821 4.394

1.052 −0.069 −1.281 −1.742 −0.664 −1.033 0.562 −2.254 −2.911 0.552 −2.016 1.202

C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 H12

0.4998 0.5453 0.4686 0.5121 0.4710 0.5263 0.4895 0.5247 0.4200 0.5540 0.4177 0.3702

0.5422 0.6776 0.6413 0.4689 0.3432 0.1715 0.8320 0.7576 0.4290 0.3867 0.1010 0.5405

0.1017 −0.0067 −0.1239 −0.1685 −0.0642 −0.0999 0.0544 −0.2180 −0.2815 0.0534 −0.1950 0.1162

Center chain 4.013 4.379 3.763 4.112 3.782 4.226 3.931 4.213 3.373 4.449 3.354 2.973

(continued)

254

Appendix

Table A.15 (continued) X Y 5.424 2.722 5.139 2.745 4.231 5.199 4.022 3.651 4.273 3.920 4.251 3.806 4.094 3.828 4.661 3.582 4.935 5.061 2.607 5.316 2.894 5.287 3.173 3.279

5.538 5.296 3.766 2.789 0.789 1.438 3.709 2.607 2.900 4.302 5.327 6.726 1.349 1.954 4.627 4.973 7.595 3.723 2.583 2.827 4.347 5.329 7.100 6.691

Z

Atoms

x

y

z

−0.034 −1.172 −1.959 −0.499 −0.174 −1.425 6.224 5.236 3.889 3.425 4.507 4.141 5.729 2.914 2.258 5.724 3.991 6.373 5.131 3.977 3.208 4.673 4.891 3.234

H22 H32 H42 H52 H62A H62B C14 C24 C34 C44 C54 C64 O24 O34 O44 O54 O64 H14 H24 H34 H44 H54 H64A H64B

0.6755 0.3390 0.6400 0.3418 0.5269 0.6474 0.5009 0.4547 0.5321 0.4882 0.5294 0.4740 0.5098 0.4767 0.5804 0.4461 0.6146 0.6303 0.3247 0.6620 0.3604 0.6584 0.3951 0.4083

0.6812 0.6514 0.4632 0.3431 0.0970 0.1769 0.4562 0.3207 0.3567 0.5292 0.6552 0.8273 0.1659 0.2403 0.5691 0.6117 0.9342 0.4579 0.3177 0.3477 0.5347 0.6555 0.8733 0.8230

−0.0033 −0.1133 −0.1895 −0.0483 −0.0168 −0.1378 0.6019 0.5064 0.3761 0.3312 0.4359 0.4005 0.5541 0.2818 0.2184 0.5536 0.3860 0.6163 0.4962 0.3846 0.3103 0.4519 0.4730 0.3128

Table A.16 Cartesian and fractional coordinates of cellulose IV2. Space group P1: a =7.99 Å, b =8.10 Å, c(fiber axis)=10.34 Å, a =b =g =90.0°, V=669 Å3, rcalc=1.61 g cm−3, antiparallel chain arrangements. (From Gardiner and Sarko 1985) X Y Z Atoms x y z Corner chain 0.060 −0.148 0.433 −0.126 0.059 −0.573 0.475 0.155 0.557 −0.544 0.417 1.089 −1.178

−0.352 −1.490 −1.102 0.228 1.288 2.614 −2.670 −2.102 0.658 0.846 3.547 −0.193 −1.666

4.458 3.470 2.129 1.662 2.747 2.388 3.962 1.147 0.493 3.968 1.934 4.594 3.360

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61 H11 H21

0.0075 −0.0185 0.0542 −0.0158 0.0074 −0.0717 0.0594 0.0194 0.0697 −0.0681 0.0522 0.1363 −0.1474

−0.0435 −0.1840 −0.1360 0.0281 0.1590 0.3227 −0.3296 −0.2595 0.0812 0.1044 0.4379 −0.0238 −0.2057

0.4311 0.3356 0.2059 0.1607 0.2657 0.2309 0.3832 0.1109 0.0477 0.3838 0.1870 0.4443 0.3250 (continued)

Appendix

255

Table A.16 (continued) X Y 1.474 −1.148 1.086 −1.065 −1.275 −0.051 0.157 −0.424 0.135 −0.049 0.582 −0.464 −0.145 −0.511 0.554 −0.219 −1.079 1.187 −1.466 1.160 −1.077 0.688 1.531 Center chain 3.962 4.232 3.631 4.117 3.875 4.434 3.676 3.963 3.447 4.501 3.708 2.928 5.270 2.586 5.147 2.842 4.378 5.440 4.036 3.766 4.368

Z

Atoms

x

y

z

−1.011 0.122 1.434 3.006 2.461 0.345 1.482 1.094 −0.236 −1.297 −2.621 2.662 2.094 −0.696 −0.854 −3.327 0.189 1.656 1.004 −0.125 −1.433 −3.205 −2.446

2.229 1.449 2.909 3.228 1.622 9.628 8.640 7.296 6.832 7.917 7.558 9.132 6.317 5.663 9.138 6.601 9.779 8.531 7.395 6.634 8.079 8.424 7.146

H31 H41 H51 H61A H61B C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B

0.1845 −0.1437 0.1359 −0.1333 −0.1596 −0.0064 0.0196 −0.0531 0.0169 −0.0061 0.0728 −0.0581 −0.0181 −0.0640 0.0693 −0.0274 −0.1350 0.1486 −0.1835 0.1452 −0.1348 0.0861 0.1916

−0.1248 0.0151 0.1770 0.3711 0.3038 0.0426 0.1830 0.1351 −0.0291 −0.1601 −0.3236 0.3286 0.2585 −0.0859 −0.1054 −0.4107 0.0233 0.2044 0.1240 −0.0154 −0.1769 −0.3957 −0.3020

0.2156 0.1401 0.2813 0.3122 0.1569 0.9311 0.8356 0.7056 0.6607 0.7657 0.7309 0.8832 0.6109 0.5477 0.8838 0.6384 0.9457 0.8250 0.7152 0.6416 0.7813 0.8147 0.6911

3.705 2.581 2.937 4.296 5.345 6.701 1.369 1.954 4.720 4.935 7.281 3.806 2.463 2.971 4.240 5.435 7.336 6.593 4.407 5.533 5.176

7.476 8.464 9.807 10.272 9.187 9.546 7.972 10.787 11.441 7.966 10.637 7.325 8.573 9.709 10.470 9.025 8.711 9.825 2.306 3.294 4.635

C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 H12 H22 H32 H42 H52 H62A H62B C14 C24 C34

0.4959 0.5297 0.4544 0.5153 0.4850 0.5549 0.4601 0.4960 0.4314 0.5633 0.4641 0.3665 0.6596 0.3237 0.6442 0.3557 0.5479 0.6809 0.5051 0.4713 0.5467

0.4574 0.3186 0.3626 0.5304 0.6599 0.8273 0.1690 0.2412 0.5827 0.6093 0.8989 0.4699 0.3041 0.3668 0.5235 0.6710 0.9057 0.8140 0.5441 0.6831 0.6390

0.7230 0.8186 0.9485 0.9934 0.8885 0.9232 0.7710 1.0432 1.1065 0.7704 1.0287 0.7084 0.8291 0.9390 1.0126 0.8728 0.8425 0.9502 0.2230 0.3186 0.4483 (continued)

256

Appendix

Table A.16 (continued) X 3.882 4.124 3.564 4.324 4.036 4.587 3.497 4.615 5.072 2.729 5.412 2.855 5.517 2.891 3.053

Y 3.818 2.770 1.412 6.744 6.160 3.426 3.178 0.443 4.304 5.652 5.142 3.869 2.680 1.102 1.484

Z 5.102 4.017 4.376 2.802 5.617 6.271 2.796 4.505 2.170 3.403 4.535 5.315 3.855 3.632 5.290

Atoms C44 C54 C64 O24 O34 O44 O54 O64 H14 H24 H34 H44 H54 H64A H64B

x 0.4859 0.5161 0.4461 0.5412 0.5051 0.5741 0.4377 0.5776 0.6348 0.3416 0.6773 0.3573 0.6905 0.3618 0.3821

y 0.4714 0.3420 0.1743 0.8326 0.7605 0.4230 0.3923 0.0547 0.5314 0.6978 0.6348 0.4777 0.3309 0.1360 0.1832

z 0.4934 0.3885 0.4232 0.2710 0.5432 0.6065 0.2704 0.4357 0.2099 0.3291 0.4386 0.5140 0.3728 0.3513 0.5116

Table A.17 Cartesian and fractional coordinates of cellulose II hydrate. Monoclinic unit cell, space group P21: a =9.02 Å, b =9.63 Å, c(fiber axis)=10.34 Å, g =116.0°, V=807.3 Å3, rcalc=1.48 g cm−3, antiparallel chain arrangements. One water molecule per residue. Atoms assigned labels 3 and 4 are symmetry-related to those with labels 1 and 2 by a 21 screw axis, respectively. The anhydroglucose units have been changed from the original l configuration to d configuration. (From Lee and Blackwell 1981) X Y Z Atoms x y z Corner chain −0.884 −0.099 0.170 0.331 1.151 0.941 1.581 2.205 0.486 1.072 1.451 1.517 −0.097 −0.251 −0.997 −0.814 −1.524 −2.194 −2.019 −2.867 −0.276 −0.934 −0.640 0.975 1.985 0.308 −0.285 1.783 0.684 −0.928 −1.809 −0.163 −2.282 −2.125 −0.738 −2.762 0.884 0.099

−1.996 1.996 1.003 1.479 −0.352 −1.313 −0.796 0.279 −0.031 1.106 1.520 2.167 0.914 −0.287 −0.980 0.419 −0.755 −0.432 3.174

O41 C11 C21 O21 C31 O31 C41 C51 C61 O61 O51 H11 H21 H31 H41 H51 H61A H61B O43

−0.1090 0.0210 0.1420 0.1950 0.0600 0.1790 −0.0120 −0.1230 −0.1880 −0.2490 −0.0340 −0.0790 0.2448 −0.0351 0.0844 −0.2231 −0.2815 −0.0910 0.1090

−0.0550 0.0430 0.1560 0.3090 0.1360 0.2310 −0.0310 −0.1350 −0.3050 −0.4000 −0.1110 0.0688 0.1325 0.1707 −0.0617 −0.1085 −0.3362 −0.3242 0.0550

−0.1930 0.1930 0.0970 0.1430 −0.0340 −0.1270 −0.0770 0.0270 −0.0030 0.1070 0.1470 0.2096 0.0884 −0.0278 −0.0948 0.0405 −0.0730 −0.0418 0.3070 (continued)

Appendix

257

Table A.17 (continued) X Y

Z

Atoms

x

y

z

−0.170 −1.151 −1.581 −0.486 −1.451 0.097 0.997 1.524 2.019 0.276 0.640 −1.985 0.285 −0.680 1.809 2.282 0.738 −0.884

−0.331 −0.941 −2.205 −1.072 −1.517 0.251 0.814 2.194 2.867 0.934 −0.975 −0.308 −1.783 0.929 0.163 2.125 2.762 −0.099

7.166 6.173 6.649 4.818 3.857 4.374 5.449 5.139 6.276 6.690 7.337 6.084 4.883 4.177 5.589 4.415 4.738 8.344

C13 C23 O23 C33 O33 C43 C53 C63 O63 O53 H13 H23 H33 H43 H53 H63A H63B O45

−0.0210 −0.1420 −0.1950 −0.0600 −0.1790 0.0120 0.1230 0.1880 0.2490 0.0340 0.0790 −0.2448 0.0351 −0.0839 0.2231 0.2815 0.0910 −0.1090

−0.0430 −0.1560 −0.3090 −0.1360 −0.2310 0.0310 0.1350 0.3050 0.4000 0.1110 −0.0688 −0.1325 −0.1707 0.0620 0.1085 0.3362 0.3242 −0.0550

0.6930 0.5970 0.6430 0.4660 0.3730 0.4230 0.5270 0.4970 0.6070 0.6470 0.7096 0.5884 0.4722 0.4040 0.5405 0.4270 0.4582 0.8070

Chain on a/2 3.851 3.810 3.567 2.521 3.210 3.121 4.256 4.483 5.602 5.991 4.840 2.929 4.444 2.281 5.157 3.596 5.292 6.431 4.256 4.297 4.540 5.586 4.897 4.986 3.851 3.624 2.505

−1.117 −2.253 −3.377 −4.215 −2.798 −3.853 −1.806 −0.761 0.205 0.968 −1.407 −1.708 −3.946 −2.313 −2.309 −0.220 0.851 −0.343 −2.837 −1.701 −0.577 0.261 −1.156 −0.101 −2.148 −3.193 −4.159

10.340 6.349 7.341 6.876 8.717 9.668 9.151 8.065 8.396 7.259 6.835 6.182 7.430 8.655 9.347 7.921 9.162 8.733 5.170 1.179 2.171 1.706 3.547 4.498 3.981 2.895 3.226

O42 C12 C22 O22 C32 O32 C42 C52 C62 O62 O52 H12 H22 H32 H42 H52 H62A H62B O44 C14 C24 O24 C34 O34 C44 C54 C64

0.4750 0.4700 0.4400 0.3110 0.3960 0.3850 0.5250 0.5530 0.6910 0.7390 0.5970 0.3613 0.5482 0.2814 0.6361 0.4435 0.6527 0.7933 0.5250 0.5300 0.5600 0.6890 0.6040 0.6150 0.4750 0.4470 0.3090

0.0790 −0.0410 −0.1700 −0.3100 −0.1280 −0.2420 0.0280 0.1480 0.3050 0.4040 0.0990 −0.0290 −0.1847 −0.1246 0.0214 0.1593 0.3564 0.2901 −0.0790 0.0410 0.1700 0.3100 0.1280 0.2420 −0.0280 −0.1480 −0.3050

1.0000 0.6140 0.7100 0.6650 0.8430 0.9350 0.8850 0.7800 0.8120 0.7020 0.6610 0.5979 0.7186 0.8370 0.9040 0.7661 0.8861 0.8446 0.5000 0.1140 0.2100 0.1650 0.3430 0.4350 0.3850 0.2800 0.3120 (continued)

258

Appendix

Table A.17 (continued) X Y

Z

Atoms

x

y

z

2.116 3.267 5.178 3.663 5.826 2.950 4.512 2.816 1.676 3.851

2.089 1.665 1.012 2.260 3.485 4.166 2.751 3.992 3.563 0.000

O64 O54 H14 H24 H34 H44 H54 H64A H64B O46

0.2610 0.4030 0.6387 0.4518 0.7186 0.3639 0.5565 0.3473 0.2067 0.4750

−0.4040 −0.0990 0.0290 0.1847 0.1246 −0.0210 −0.1593 −0.3564 −0.2901 0.0790

0.2020 0.1610 0.0979 0.2186 0.3370 0.4029 0.2661 0.3861 0.3446 0.0000

−4.923 −2.547 −2.246 −0.008 −1.642 −1.641 −3.735 −4.805 −3.611 −1.117

Table A.18 Cartesian and fractional coordinates of ammonia cellulose I. Monoclinic unit cell, space group P21: a = 4.47 Å, b = 8.81 Å, c(fiber axis) = 10.34 Å, a = b= 90.0°, g =92.70°, V = 406.7 Å3, rcalc = 1.463 g cm−3, parallel chain arrangements (T = 20°C). One residue is sufficient to describe the structure in P21 of a one-chain unit cell. (From Wada et al. 2006) X Y Z Atoms x y z 0.000 0.112 −0.313 0.237 0.134 0.799 −0.670 0.134 −0.473 0.764 0.491 −0.943 1.052 −1.293 1.181 −0.814 0.517 1.761 −2.237 0.000 −0.112 0.313 −0.237 −0.134 −0.799 0.670 −0.134 0.473

0.414 1.554 1.142 −0.205 −1.240 −2.540 2.666 2.161 −0.744 −0.697 −3.512 0.155 1.833 1.101 −0.102 −1.403 −2.859 −2.415 −4.678 −0.414 −1.554 −1.142 0.205 1.240 2.540 −2.666 −2.161 0.744

5.164 4.198 2.812 2.393 3.496 3.205 4.653 1.903 1.253 4.684 4.188 5.237 4.161 2.786 2.151 3.687 2.330 3.177 1.065 10.334 9.368 7.982 7.563 8.666 8.375 9.823 7.073 6.423

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61 H11 H21 H31 H41 H51 H61A H61B N11 C13 C23 C33 C43 C53 C63 O23 O33 O43

0.0000 0.0250 −0.0700 0.0530 0.0300 0.1790 −0.1500 0.0300 −0.1060 0.1710 0.1100 −0.2112 0.2357 −0.2896 0.2644 −0.1822 0.1157 0.3944 −0.5010 0.0000 −0.0250 0.0700 −0.0530 −0.0300 −0.1790 0.1500 −0.0300 0.1060

0.0470 0.1770 0.1280 −0.0220 −0.1400 −0.2840 0.2990 0.2460 −0.0870 −0.0750 −0.3960 0.0125 0.2137 0.1180 −0.0053 −0.1636 −0.3217 −0.2647 −0.5430 −0.0470 −0.1770 −0.1280 0.0220 0.1400 0.2840 −0.2990 −0.2460 0.0870

0.4994 0.4060 0.2720 0.2314 0.3381 0.3100 0.4500 0.1840 0.1212 0.4530 0.4050 0.5065 0.4024 0.2694 0.2080 0.3566 0.2253 0.3073 0.1030 0.9994 0.9060 0.7720 0.7314 0.8381 0.8100 0.9500 0.6840 0.6212 (continued)

Appendix Table A.18 X −0.764 −0.491 0.943 −1.052 1.293 −1.181 0.814 −0.517 −1.761 2.237 −0.473

259 (continued) Y 0.697 3.512 −0.155 −1.833 −1.101 0.102 1.403 2.859 2.415 4.678 −0.744

Z 9.854 9.358 10.407 9.331 7.956 7.321 8.857 7.500 8.347 6.235 11.593

Atoms O53 O63 H13 H23 H33 H43 H53 H63A H63B N13 O45

x −0.1710 −0.1100 0.2112 −0.2357 0.2896 −0.2644 0.1822 −0.1157 −0.3944 0.5010 −0.1060

y 0.0750 0.3960 −0.0125 −0.2137 −0.1180 0.0053 0.1636 0.3217 0.2647 0.5430 −0.0870

z 0.9530 0.9050 1.0065 0.9024 0.7694 0.7080 0.8566 0.7253 0.8073 0.6030 1.1212

Table A.19a Cartesian and fractional coordinates of cellulose II-hydrazine complex (Lee et al. 1981). Monoclinic subcell, space group P21: a = 4.69 Åa, b = 19.88 Å, c = 10.39 Å, a = b = 90.0°, g = 120.0°; V= 838.9 Å3, rcalc = 1.54 g cm−3; robs = 1.49 g cm−3, antiparallel chain arrangements X

Y

−0.024 0.485 0.185 −0.151 0.438 0.053 −0.697 −0.414 −0.133 −0.760 0.261 2.818 2.123 −1.054 1.526 −1.180 1.073 −1.447 0.904 −0.501 0.024 −0.485 −0.185 0.151 −0.438 −0.053 0.697

0.359 1.398 2.697 1.170 2.055 −0.268 −0.540 −1.229 −2.679 −3.568 −0.952 2.932 4.197 0.477 1.301 1.369 −0.430 −1.106 −2.845 −2.880 −0.359 −1.398 −2.697 −1.170 −2.055 0.268 0.540

Z 4.001 3.008 3.487 1.640 0.688 1.192 0.000 2.278 1.945 2.865 3.512 3.721 3.948 4.167 2.919 1.702 1.003 2.420 1.949 0.982 9.196 8.203 8.682 6.835 5.883 6.387 5.195

Atoms

x

y

z

C11 C21 O21 C31 O31 C41 O41 C51 C61 O61 O51 N11 N21 H11 H21 H31 H41 H51 H61A H61B C13 C23 O23 C33 O33 C43 O43

−0.0060 0.1193 0.0456 −0.0372 0.1079 0.0130 −0.1717 −0.1020 −0.0327 −0.1870 0.0642 0.6938 0.5228 −0.2596 0.3758 −0.2906 0.2642 −0.3564 0.2226 −0.1233 0.0060 −0.1193 −0.0456 0.0372 −0.1079 −0.0130 0.1717

0.0173 0.0844 0.1410 0.0545 0.1161 −0.0119 −0.0474 −0.0738 −0.1386 −0.2016 −0.0403 0.2293 0.2728 −0.0066 0.1098 0.0346 0.0096 −0.0977 −0.1168 −0.1594 −0.0173 −0.0844 −0.1410 −0.0545 −0.1161 0.0119 0.0474

0.3851 0.2895 0.3356 0.1578 0.0663 0.1147 0.0000 0.2193 0.1872 0.2758 0.3380 0.3582 0.3800 0.4011 0.2809 0.1638 0.0966 0.2329 0.1876 0.0945 0.8851 0.7895 0.8356 0.6578 0.5663 0.6147 0.5000 (continued)

260 Table A.19a (continued) X Y Z 0.414 1.229 7.473 0.133 2.679 7.140 0.760 3.568 8.060 −0.261 0.952 8.707 −2.818 −2.932 8.916 −2.123 −4.197 9.143 1.054 −0.478 9.362 −1.526 −1.302 8.114 1.180 −1.369 6.897 −1.073 0.429 6.199 1.448 1.107 7.615 −0.904 2.844 7.144 0.501 2.880 6.177 −0.697 −0.540 10.390 2.347 8.939 6.597 2.954 9.924 7.589 4.213 10.367 7.111 3.102 9.266 8.958 3.535 10.237 9.909 1.776 8.669 9.405 1.947 7.890 10.597 1.213 7.761 8.319 −0.165 7.222 8.652 −0.099 6.080 9.502 1.086 8.478 7.086 4.787 13.089 6.491 3.402 12.968 6.942 2.998 8.132 6.430 2.315 10.753 7.679 3.819 8.502 8.896 1.094 9.445 9.595 1.869 6.953 8.177 −0.667 6.970 7.765 −0.709 7.974 9.143 1.714 8.596 1.402 1.108 7.611 2.395 −0.152 7.169 1.916 0.960 8.269 3.762 0.527 7.298 4.714 2.285 8.867 4.210 2.115 9.645 5.402 2.848 9.775 3.124 4.227 10.313 3.458 4.160 11.454 4.307 2.976 9.057 1.891

Appendix

Atoms C53 C63 O63 O53 N13 N23 H13 H23 H33 H43 H53 H63A H63B O13 C12 C22 O22 C32 O32 C42 O42 C52 C62 O62 O52 N12 N22 H12 H22 H32 H42 H52 H62A H62B C14 C24 O24 C34 O34 C44 O44 C54 C64 O64 O54

x 0.1020 0.0327 0.1870 −0.0642 −0.6938 −0.5228 0.2596 −0.3758 0.2906 −0.2642 0.3564 −0.2226 0.1233 −0.1717 0.5779 0.7273 1.0374 0.7637 0.8702 0.4373 0.4793 0.2987 −0.0406 −0.0243 0.2674 1.1786 0.8375 0.7381 0.5700 0.9404 0.2694 0.4602 −0.1641 −0.1746 0.4221 0.2727 −0.0374 0.2363 0.1298 0.5627 0.5207 0.7013 1.0406 1.0243 0.7326

y 0.0738 0.1386 0.2016 0.0403 −0.2293 −0.2728 0.0066 −0.1098 −0.0346 −0.0096 0.0977 0.1168 0.1594 −0.0474 0.5178 0.5850 0.6438 0.5562 0.6176 0.4876 0.4534 0.4256 0.3585 0.3030 0.4580 0.7974 0.7511 0.4961 0.6081 0.5386 0.5069 0.4040 0.3312 0.3805 0.4822 0.4150 0.3562 0.4438 0.3824 0.5124 0.5466 0.5744 0.6415 0.6970 0.5420

z 0.7193 0.6872 0.7758 0.8380 0.8582 0.8800 0.9011 0.7809 0.6638 0.5966 0.7329 0.6876 0.5945 1.0000 0.6349 0.7305 0.6844 0.8621 0.9537 0.9052 1.0199 0.8007 0.8328 0.9145 0.6820 0.6247 0.6681 0.6189 0.7391 0.8562 0.9234 0.7870 0.7474 0.8800 0.1349 0.2305 0.1844 0.3621 0.4537 0.4052 0.5199 0.3007 0.3328 0.4145 0.1820 (continued)

Appendix

261

Table A.19a (continued) X Y Z Atoms x y −0.725 4.447 1.296 N14 −0.1786 0.2026 0.660 4.567 1.747 N24 0.1625 0.2489 1.064 9.403 1.235 H14 0.2619 0.5039 1.747 6.783 2.484 H24 0.4300 0.3919 0.242 9.033 3.701 H34 0.0596 0.4614 2.967 8.090 4.399 H44 0.7306 0.4931 2.192 10.583 2.982 H54 0.5398 0.5960 4.728 10.566 2.570 H64A 1.1641 0.6688 4.771 9.561 3.948 H64B 1.1746 0.6195 1.947 7.890 0.207 O46 0.4793 0.4534 a The refinement was performed with a subcell with a/2 of the original unit cell.

z 0.1247 0.1681 0.1189 0.2391 0.3562 0.4234 0.2870 0.2474 0.3800 0.0199

Table A.19b Cartesian and fractional coordinates of cellulose I–ethylenediamine. Monoclinic unit cell, space group P21: a=4.762 Å (dimension half the original one of Lee et al 1984 of 9.524 Å), b =12.876 Å, c = 10.353 Å, a = b = 90.0°, g = 118.82°, V=556.2 Å3, rcalc=1.33 g cm−3, parallel chain arrangements. One residue is sufficient to describe the structure in P21. (From Lee et al. 1984) X Y Z Atoms x y z −0.016 0.516 0.247 −0.125 0.484 0.047 −0.708 −0.441 −0.195 −0.752 0.239 0.430 0.415 1.860 2.552 −1.042 1.554 −1.149 1.063 −1.471 0.838 −0.639 0.016 −0.516 −0.247 0.125 −0.484

0.356 1.377 2.684 1.163 2.038 −0.257 −0.512 −1.212 −2.674 −3.539 −0.956 4.708 6.029 6.522 5.588 0.499 1.257 1.383 −0.439 −1.065 −2.847 −2.885 −0.356 −1.377 −2.684 −1.163 −2.038

4.002 3.009 3.488 1.641 0.688 1.193 0.000 2.279 1.945 2.924 3.513 0.244 −0.375 −0.634 −1.535 4.171 2.920 1.703 1.003 2.422 1.871 1.018 9.179 8.185 8.664 6.817 5.865

C11 C21 O21 C31 O31 C41 O41 C51 C61 O61 O51 N11 C71 C81 N21 H11 H21 H31 H41 H51 H61A H61B C13 C23 O23 C33 O33

−0.0038 0.1236 0.0591 −0.0299 0.1161 0.0112 −0.1698 −0.1057 −0.0468 −0.1803 0.0573 0.1030 0.0995 0.4457 0.6116 −0.2498 0.3725 −0.2753 0.2547 −0.3526 0.2008 −0.1532 0.0038 −0.1236 −0.0591 0.0299 −0.1161

0.0270 0.1290 0.2190 0.0850 0.1790 −0.0180 −0.0700 −0.1130 −0.2160 −0.3070 −0.0640 0.3840 0.4860 0.5860 0.5430 −0.0058 0.1640 0.0583 0.0113 −0.1456 −0.1853 −0.2514 −0.0270 −0.1290 −0.2190 −0.0850 −0.1790

0.3866 0.2906 0.3369 0.1585 0.0665 0.1152 0.0000 0.2201 0.1879 0.2824 0.3393 0.0236 −0.0362 −0.0612 −0.1483 0.4029 0.2820 0.1645 0.0969 0.2339 0.1807 0.0983 0.8866 0.7906 0.8369 0.6585 0.5665 (continued)

262 Table A.19b (continued) X Y −0.047 0.257 0.708 0.512 0.441 1.212 0.195 2.674 0.752 3.539 −0.239 0.956 −0.430 −4.708 −0.415 −6.029 −1.860 −6.522 −2.552 −5.588 1.042 −0.499 −1.554 −1.257 1.149 −1.383 −1.063 0.439 1.471 1.065 −0.838 2.847 0.639 2.885

Appendix

Z 6.369 5.176 7.455 7.122 8.100 8.689 5.421 4.802 4.543 3.641 9.348 8.096 6.880 6.180 7.598 7.047 6.194

Atoms C43 O43 C53 C63 O63 O53 N13 C73 C83 N23 H13 H23 H33 H43 H53 H63A H63B

x −0.0112 0.1698 0.1057 0.0468 0.1803 −0.0573 −0.1030 −0.0995 −0.4457 −0.6116 0.2498 −0.3725 0.2753 −0.2547 0.3526 −0.2008 0.1532

y 0.0180 0.0700 0.1130 0.2160 0.3070 0.0640 −0.3840 −0.4860 −0.5860 −0.5430 0.0058 −0.1640 −0.0583 −0.0113 0.1456 0.1853 0.2514

z 0.6152 0.5000 0.7201 0.6879 0.7824 0.8393 0.5236 0.4638 0.4388 0.3517 0.9029 0.7820 0.6645 0.5969 0.7339 0.6807 0.5983

Table A.20 Cartesian and fractional coordinates of sodium cellulose I. Monoclinic unit cell; space group P21: a = 8.83 Å, b = 25.28 Å, c(fiber axis)=10.29 Å, a = b =g =90.0°, V=2,207 Å3, antiparallel chain arrangements, dimers serve as a basic building unit, chain backbones obey a 21 screw axis. (From Nishimura et al. 1991) X Y Z Atoms x y z Chain 1 (corner) −0.129 0.387 0.154 1.497 −0.363 1.090 0.197 −0.270 −0.061 −1.299 0.608 −2.627 −0.476 2.690 0.021 2.073 −0.458 −0.728 0.481 −0.832 −0.057 −3.342 −1.164 0.255 1.190 1.656 −1.411 1.031 1.229 −0.192 −1.095 −1.446 0.619 −3.206 1.599 −2.447 0.129 −0.387

3.940 2.940 1.573 1.181 2.274 1.992 3.372 0.617 0.000 3.510 0.954 4.060 2.884 1.605 1.005 2.379 2.876 1.695 9.085

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61 H11 H21 H31 H41 H51 H61A H61B C15

−0.0146 0.0174 −0.0411 0.0223 −0.0069 0.0689 −0.0539 0.0024 −0.0519 0.0545 −0.0065 −0.1318 0.1348 −0.1598 0.1392 −0.1240 0.0701 0.1811 0.0146

0.0153 0.0592 0.0431 −0.0107 −0.0514 −0.1039 0.1064 0.0820 −0.0288 −0.0329 −0.1322 0.0101 0.0655 0.0408 −0.0076 −0.0572 −0.1268 −0.0968 −0.0153

0.3829 0.2857 0.1529 0.1148 0.2210 0.1936 0.3277 0.0600 0.0000 0.3411 0.0927 0.3946 0.2803 0.1560 0.0977 0.2312 0.2795 0.1647 0.8829 (continued)

Appendix

263

Table A.20 (continued) X Y

Atoms

x

y

−1.497 −1.090 0.270 1.299 2.627 −2.690 −2.073 0.728 0.832 3.357 −0.255 −1.656 −1.031 0.192 1.446 3.195 2.445

8.085 6.718 6.326 7.419 7.137 8.517 5.762 5.145 8.655 6.123 9.205 8.029 6.750 6.150 7.524 8.018 6.816

C25 C35 C45 C55 C65 O25 O35 O45 O55 O65 H15 H25 H35 H45 H55 H65A H65B

−0.0174 0.0411 −0.0223 0.0069 −0.0689 0.0539 −0.0024 0.0519 −0.0545 0.0087 0.1318 −0.1348 0.1598 −0.1392 0.1240 −0.0726 −0.1802

−0.0592 −0.0431 0.0107 0.0514 0.1039 −0.1064 −0.0820 0.0288 0.0329 0.1328 −0.0101 −0.0655 −0.0408 0.0076 0.0572 0.1264 0.0967

0.7857 0.6529 0.6148 0.7210 0.6936 0.8277 0.5600 0.5000 0.8411 0.5950 0.8946 0.7803 0.6560 0.5977 0.7312 0.7792 0.6624

Chain 2 (at b/2) 0.130 13.027 −0.148 14.139 0.367 13.727 −0.198 12.370 0.056 11.338 −0.618 10.016 0.487 15.330 −0.013 14.713 0.455 11.909 −0.485 11.811 −0.087 8.969 1.165 12.893 −1.184 14.301 1.415 13.666 −1.230 12.453 1.089 11.189 −1.650 10.104 −0.468 9.771 −0.130 12.253 0.148 11.141 −0.367 11.553 0.198 12.910 −0.056 13.942 0.618 15.264 −0.487 9.950 0.013 10.567 −0.455 13.371

3.693 4.693 6.060 6.452 5.359 5.641 4.261 7.016 7.633 4.123 4.831 3.573 4.749 6.028 6.627 5.253 5.467 6.650 −1.452 −0.452 0.915 1.307 0.214 0.496 −0.884 1.871 2.488

C12 C22 C32 C42 C52 C62 O22 O32 O42 O52 O62 H12 H22 H32 H42 H52 H62A H62B C16 C26 C36 C46 C56 C66 O26 O36 O46

0.0147 −0.0168 0.0416 −0.0224 0.0063 −0.0700 0.0552 −0.0015 0.0515 −0.0549 −0.0099 0.1319 −0.1341 0.1602 −0.1393 0.1233 −0.1869 −0.0530 −0.0147 0.0168 −0.0416 0.0224 −0.0063 0.0700 −0.0552 0.0015 −0.0515

0.5153 0.5593 0.5430 0.4893 0.4485 0.3962 0.6064 0.5820 0.4711 0.4672 0.3548 0.5100 0.5657 0.5406 0.4926 0.4426 0.3997 0.3865 0.4847 0.4407 0.4570 0.5107 0.5515 0.6038 0.3936 0.4180 0.5289

0.3589 0.4561 0.5889 0.6270 0.5208 0.5482 0.4141 0.6818 0.7418 0.4007 0.4695 0.3472 0.4615 0.5858 0.6440 0.5105 0.5313 0.6463 −0.1411 −0.0439 0.0889 0.1270 0.0208 0.0482 −0.0859 0.1818 0.2418

−0.154 0.363 −0.197 0.061 −0.608 0.476 −0.021 0.458 −0.481 0.077 1.164 −1.190 1.411 −1.229 1.095 −0.641 −1.591

Z

z

(continued)

264

Appendix

Table A.20 (continued) X Y

Z

Atoms

x

0.485 −0.089 −1.165 1.184 −1.415 1.230 −1.089 1.601 0.654

13.469 16.354 12.387 10.979 11.614 12.827 14.091 15.244 15.408

−1.022 −0.093 −1.572 −0.396 0.882 1.482 0.108 0.125 1.535

O56 O66 H16 H26 H36 H46 H56 H66A H66B

0.0549 −0.0101 −0.1319 0.1341 −0.1602 0.1393 −0.1233 0.1813 0.0741

0.5328 0.6469 0.4900 0.4343 0.4594 0.5074 0.5574 0.6030 0.6095

−0.0993 −0.0090 −0.1528 −0.0385 0.0857 0.1440 0.0105 0.0121 0.1492

Chain 3 (at a/2) 4.476 −0.404 5.224 −1.272 4.579 −1.135 4.474 0.331 3.785 1.138 3.793 2.622 5.192 −2.622 5.362 −1.845 3.685 0.447 4.480 0.958 3.216 3.365 3.491 −0.748 6.222 −0.951 3.616 −1.552 5.434 0.720 2.794 0.806 4.780 2.943 3.237 2.801 4.362 0.404 3.614 1.272 4.259 1.135 4.364 −0.331 5.053 −1.138 5.045 −2.622 3.646 2.622 3.476 1.845 5.153 −0.447 4.358 −0.958 5.628 −3.365 5.347 0.748 2.616 0.951 5.222 1.552 3.404 −0.720 6.044 −0.806 4.057 −2.945 5.597 −2.801

3.290 4.290 5.657 5.869 4.956 5.238 3.858 6.613 7.230 3.720 4.167 3.170 4.346 5.625 6.224 4.850 5.396 6.110 −1.855 −0.855 0.512 0.904 −0.189 0.093 −1.287 1.468 2.085 −1.425 −0.975 −1.975 −0.799 0.480 1.079 −0.295 0.246 0.968

C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B C17 C27 C37 C47 C57 C67 O27 O37 O47 O57 O67 H17 H27 H37 H47 H57 H67A H67B

0.5069 0.5916 0.5186 0.5067 0.4287 0.4296 0.5880 0.6072 0.4173 0.5074 0.3642 0.3954 0.7046 0.4095 0.6154 0.3164 0.5413 0.3666 0.4940 0.4093 0.4823 0.4942 0.5723 0.5713 0.4129 0.3937 0.5836 0.4935 0.6374 0.6055 0.2963 0.5914 0.3855 0.6845 0.4595 0.6339

−0.0160 −0.0503 −0.0449 0.0131 0.0450 0.1037 −0.1037 −0.0730 0.0177 0.0379 0.1331 −0.0296 −0.0376 −0.0614 0.0285 0.0319 0.1164 0.1108 0.0160 0.0503 0.0449 −0.0131 −0.0450 −0.1037 0.1037 0.0730 −0.0177 −0.0379 −0.1331 0.0296 0.0376 0.0614 −0.0285 −0.0319 −0.1165 −0.1108

0.3197 0.4169 0.5498 0.5704 0.4816 0.5090 0.3749 0.6427 0.7026 0.3615 0.4050 0.3081 0.4224 0.5466 0.6049 0.4713 0.5244 0.5938 −0.1803 −0.0831 0.0498 0.0879 −0.0184 0.0090 −0.1251 0.1427 0.2026 −0.1385 −0.0948 −0.1919 −0.0776 0.0466 0.1049 −0.0287 0.0239 0.0941

y

z

(continued)

Appendix

265

Table A.20 (continued) X Y

Z

Atoms

Chain 4 (center) 4.340 13.039 4.760 14.106 4.196 13.765 4.580 12.347 4.195 11.358 4.691 9.958 4.286 15.370 4.701 14.695 3.873 11.975 4.792 11.755 3.917 9.318 3.297 13.042 5.808 14.132 3.149 13.841 5.614 12.296 3.150 11.343 4.656 9.389 5.688 10.013 4.498 12.241 4.078 11.174 4.642 11.515 4.258 12.933 4.643 13.922 4.147 15.322 4.552 9.910 4.137 10.585 4.965 13.305 4.046 13.525 4.827 15.911 5.541 12.238 3.030 11.148 5.689 11.439 3.224 12.984 5.688 13.937 4.281 15.919 3.121 15.277

4.765 3.765 2.398 2.006 3.099 2.817 4.197 1.442 0.825 4.335 1.805 4.885 3.707 2.430 1.830 3.204 3.699 2.492 9.910 8.910 7.543 7.151 8.244 7.962 9.342 6.587 5.970 9.480 6.855 10.030 8.852 7.575 6.975 8.349 8.814 7.741

C14 C24 C34 C44 C54 C64 O24 O34 O44 O54 O64 H14 H24 H34 H44 H54 H64A H64B C18 C28 C38 C48 C58 C68 O28 O38 O48 O58 O68 H18 H28 H38 H48 H58 H68A H68B

Sodium −0.135 4.866 3.905 0.025 1.026 4.847 4.147 −0.414

3.334 3.698 10.127 8.556 5.113 5.070 8.309 8.043

Na1 Na2 Na3 Na4 Na5 Na6 Na7 Na8

5.084 5.349 7.834 7.435 17.572 17.501 20.252 20.429

x

y

z

0.4915 0.5391 0.4752 0.5187 0.4751 0.5313 0.4854 0.5324 0.4386 0.5427 0.4436 0.3734 0.6578 0.3566 0.6358 0.3567 0.5273 0.6442 0.5094 0.4618 0.5257 0.4822 0.5258 0.4696 0.5155 0.4685 0.5623 0.4582 0.5467 0.6275 0.3431 0.6443 0.3651 0.6442 0.4848 0.3535

0.5158 0.5580 0.5445 0.4884 0.4493 0.3939 0.6080 0.5813 0.4737 0.4650 0.3686 0.5159 0.5590 0.5475 0.4864 0.4487 0.3714 0.3961 0.4842 0.4420 0.4555 0.5116 0.5507 0.6061 0.3920 0.4187 0.5263 0.5350 0.6294 0.4841 0.4410 0.4525 0.5136 0.5513 0.6297 0.6043

0.4631 0.3659 0.2330 0.1949 0.3012 0.2738 0.4079 0.1401 0.0802 0.4213 0.1754 0.4747 0.3603 0.2362 0.1778 0.3114 0.3595 0.2422 0.9631 0.8659 0.7330 0.6949 0.8012 0.7738 0.9079 0.6401 0.5802 0.9213 0.6662 0.9747 0.8603 0.7362 0.6778 0.8114 0.8566 0.7523

−0.0153 0.5511 0.4422 0.0028 0.1162 0.5489 0.4696 −0.0469

0.2011 0.2116 0.3099 0.2941 0.6951 0.6923 0.8011 0.8081

0.3240 0.3594 0.9842 0.8315 0.4969 0.4927 0.8075 0.7816 (continued)

266

Appendix

Table A.21 Cartesian and fractional coordinates of sodium cellulose IV. Monoclinic unit cell; space group P21: a = 9.57 Å, b = 8.72 Å, c(fiber axis) = 10.35 Å, a = b= 90°, g = 122.0°, V=733 Å3, rcalc=1.55 g cm−3, antiparallel chain arrangements. (From Nishimura and Sarko 1991) X Y Z Atoms x y z Chain 1 (corner) −0.218 0.348 −0.178 1.483 −0.602 0.967 0.252 −0.233 0.233 −1.289 1.164 −2.444 −1.046 2.519 −0.429 1.995 −0.275 −0.809 0.650 −0.703 0.962 −3.511 −1.200 −0.009 0.802 1.856 −1.612 0.682 1.240 0.078 −0.745 −1.656 2.159 −2.113 0.985 −2.801 0.210 −0.337 0.168 −1.464 0.581 −0.947 −0.267 0.255 −0.225 1.298 −1.155 2.451 1.035 −2.503 0.404 −1.988 0.265 0.833 −0.650 0.718 −1.172 3.411 1.192 0.017 −0.811 −1.838 1.593 −0.665 −1.259 −0.048 0.755 1.659 −2.128 2.087 −0.834 2.929

3.947 2.936 1.572 1.187 2.290 2.020 3.375 0.599 0.000 3.530 2.943 4.058 2.875 1.605 1.017 2.396 2.079 1.049 9.127 8.106 6.739 6.348 7.465 7.183 8.541 5.781 5.165 8.705 8.238 9.239 8.048 6.765 6.182 7.568 7.026 6.305

C11 C21 C31 C41 C51 C61 O21 O31 O41 O51 O61 H11 H21 H31 H41 H51 H61A H61B C13 C23 C33 C43 C53 C63 O23 O33 O43 O53 O63 H13 H23 H33 H43 H53 H63A H63B

Chain 2 (center) 4.060 1.413 4.704 0.478 4.070 0.685 4.145 2.156 3.562 3.037 3.728 4.512

5.085 6.095 7.459 7.844 6.741 7.012

C12 C22 C32 C42 C52 C62

−0.0269 −0.0219 −0.0742 0.0311 0.0287 0.1434 −0.1289 −0.0529 −0.0339 0.0801 0.1185 −0.1479 0.0988 −0.1986 0.1528 −0.0918 0.2660 0.1214 0.0259 0.0207 0.0716 −0.0329 −0.0277 −0.1423 0.1275 0.0498 0.0327 −0.0801 −0.1444 0.1469 −0.0999 0.1963 −0.1551 0.0930 −0.2622 −0.1028

0.0243 0.1573 0.0678 −0.0087 −0.1311 −0.1969 0.2139 0.1980 −0.1125 −0.0340 −0.3337 −0.0870 0.2703 −0.0373 0.0978 −0.2433 −0.0876 −0.2506 −0.0236 −0.1559 −0.0670 0.0101 0.1327 0.1983 −0.2129 −0.1990 0.1145 0.0358 0.3072 0.0874 −0.2689 0.0379 −0.0957 0.2444 0.0868 0.2761

0.3814 0.2837 0.1519 0.1147 0.2213 0.1952 0.3261 0.0579 0.0000 0.3411 0.2843 0.3921 0.2778 0.1551 0.0983 0.2315 0.2009 0.1014 0.8818 0.7832 0.6511 0.6133 0.7213 0.6940 0.8252 0.5586 0.4990 0.8411 0.7959 0.8927 0.7776 0.6536 0.5973 0.7312 0.6788 0.6092

0.5003 0.5796 0.5015 0.5107 0.4389 0.4593

0.4530 0.3919 0.3702 0.5443 0.6035 0.7845

0.4913 0.5889 0.7207 0.7579 0.6513 0.6775 (continued)

Appendix

267

Table A.21 (continued) X Y 4.527 4.767 3.391 4.228 3.293 3.041 5.731 3.064 5.146 2.540 4.737 3.157 4.054 3.414

−0.863 −0.089 2.359 2.766 5.314 1.188 0.689 0.383 2.423 2.821 4.718 4.761 2.221 3.150

Z 5.656 8.433 9.031 5.501 5.917 4.974 6.156 7.427 8.014 6.635 7.219 7.856 −0.096 0.925

Atoms

x

y

O22 O32 O42 O52 O62 H12 H22 H32 H42 H52 H62A H62B C14 C24

0.5578 0.5874 0.4178 0.5210 0.4058 0.3747 0.7062 0.3775 0.6341 0.3130 0.5837 0.3890 0.4995 0.4207

0.2254 0.3314 0.5135 0.6202 0.8454 0.3542 0.4897 0.2635 0.6466 0.5055 0.8805 0.7722 0.5452 0.6059

z 0.5465 0.8148 0.8726 0.5315 0.5717 0.4806 0.5948 0.7176 0.7743 0.6411 0.6975 0.7590 −0.0093 0.0894

Table A.22 Cartesian and fractional coordinates of cellulose triacetate I. Monoclinic unit cell, space group P21: a = 5.939 Å, b = 11.431 Å, c(fiber axis)=10.460 Å, a = b = 90.0°, g = 95.40°, V=707.0 Å3, T=20°C, rcalc=1.354 g cm−3, parallel chain arrangements (T=20°C). One residue is sufficient to describe the structure in P21 of a one-chain unit cell. Residue 3 is symmetry-related to residue 1 by a 21 screw axis.. (From Sikorski et al. 2004) X Y Z Atoms x y z −0.700 0.096 0.102 1.133 0.645 −0.863 −0.215 1.348 3.515 2.599 2.779 1.799 1.131 1.976 3.296 −1.316 −0.165 −1.517 −0.400 0.813 −0.794 2.060

−0.142 −0.298 0.368 0.984 1.052 −0.855 −0.925 2.331 2.275 2.864 4.226 1.350 3.482 2.629 2.922 −2.263 −3.126 −4.923 −4.468 −5.347 1.002 0.376

0.175 1.358 4.152 3.232 1.794 2.405 3.671 3.729 3.116 3.640 4.245 0.959 0.665 0.518 −0.147 2.072 1.957 2.048 1.963 1.859 4.260 3.286

O41 C41 C11 C21 C31 C51 O51 O21 O21C C21C C21M O31 O31C C31C C31M C61 O61 O61C C61C C61M H11 H21

−0.1184 0.0162 0.0173 0.1916 0.1091 −0.1460 −0.0364 0.2280 0.5945 0.4396 0.4700 0.3043 0.1913 0.3342 0.5574 −0.2226 −0.0279 −0.2566 −0.0677 0.1375 −0.1343 0.3484

−0.0182 −0.0253 0.0330 0.0955 0.0974 −0.0819 −0.0827 0.2151 0.2281 0.2720 0.3927 0.1330 0.3140 0.2463 0.2829 −0.2089 −0.2748 −0.4432 −0.3942 −0.4610 0.0811 0.0499

0.0167 0.1298 0.3969 0.3090 0.1715 0.2299 0.3510 0.3565 0.2979 0.3480 0.4058 0.0917 0.0636 0.0495 −0.0141 0.1981 0.1871 0.1958 0.1877 0.1777 0.4073 0.3141 (continued)

268

Appendix

Table A.22 (continued) X Y

Z

−0.151 0.927 −1.760 −1.885 −1.999 2.880 3.702 1.912 3.241 4.084 3.622 1.684 0.627 1.080 0.700 −0.096 −0.102 −1.133 −0.645 0.863 0.215 −1.348 −3.515 −2.599 −2.779 −1.799 −1.131 −1.976 −3.296 1.316 0.165 1.517 0.400 −0.813 0.794 −2.060 0.151 −0.927 1.760 1.885 1.999 −2.880 −3.702 −1.912 −3.241

1.702 1.201 2.479 1.122 2.860 5.345 3.869 4.026 −0.771 0.616 −0.789 2.376 2.306 0.792 5.405 6.588 9.382 8.462 7.024 7.635 8.901 8.959 8.346 8.870 9.475 6.189 5.895 5.748 5.083 7.302 7.187 7.278 7.193 7.089 9.490 8.516 6.932 6.431 7.709 6.352 8.090 10.575 9.099 9.256 4.459

1.817 −1.016 −0.205 −2.267 −2.632 4.142 4.713 4.885 3.839 3.090 2.076 −4.890 −6.346 −5.492 0.142 0.298 −0.368 −0.984 −1.052 0.855 0.925 −2.331 −2.275 −2.864 −4.226 −1.350 −3.482 −2.629 −2.922 2.263 3.126 4.923 4.468 5.347 −1.002 −0.376 −1.817 1.016 0.205 2.267 2.632 −4.142 −4.713 −4.885 −3.839

Atoms H31 H41 H51 H61A H61B H21Ma H21Mb H21Mc H31Ma H31Mb H31Mc H61Ma H61Mb H61Mc O43 C43 C13 C23 C33 C43 O53 O23 O23C C23C C23M O33 O33C C33C C33M C63 O63 O63C C63C C63M H13 H23 H33 H43 H53 H63A H63B H23Ma H23Mb H23Mc H33Ma

x

y

z

−0.0255 0.1568 −0.2977 −0.3188 −0.3381 0.4871 0.6261 0.3234 0.5481 0.6907 0.6126 0.2848 0.1060 0.1827 0.1184 −0.0162 −0.0173 −0.1916 −0.1091 0.1460 0.0364 −0.2280 −0.5945 −0.4396 −0.4700 −0.3043 −0.1913 −0.3342 −0.5574 0.2226 0.0279 0.2566 0.0677 −0.1375 0.1343 −0.3484 0.0255 −0.1568 0.2977 0.3188 0.3381 −0.4871 −0.6261 −0.3234 −0.5481

0.1577 −0.0812 −0.0325 −0.2139 −0.2468 0.3862 0.4429 0.4432 0.3626 0.3041 0.2116 −0.4139 −0.5500 −0.4715 0.0182 0.0253 −0.0330 −0.0955 −0.0974 0.0819 0.0827 −0.2151 −0.2281 −0.2720 −0.3927 −0.1330 −0.3140 −0.2463 −0.2829 0.2089 0.2748 0.4432 0.3942 0.4610 −0.0811 −0.0499 −0.1577 0.0812 0.0325 0.2139 0.2468 −0.3862 −0.4429 −0.4432 −0.3626

0.1627 0.1148 0.2370 0.1073 0.2734 0.5110 0.3699 0.3849 −0.0737 0.0589 −0.0754 0.2272 0.2205 0.0757 0.5167 0.6298 0.8969 0.8090 0.6715 0.7299 0.8510 0.8565 0.7979 0.8480 0.9058 0.5917 0.5636 0.5495 0.4859 0.6981 0.6871 0.6958 0.6877 0.6777 0.9073 0.8141 0.6627 0.6148 0.7370 0.6073 0.7734 1.0110 0.8699 0.8849 0.4263 (continued)

Appendix

269

Table A.22 (continued) X Y

Z

Atoms

x

y

z

−4.084 −3.622 −1.684 −0.627 −1.080 −0.700

5.846 4.441 7.606 7.536 6.022 10.635

H33Mb H33Mc H63Ma H63Mb H63Mc O45

−0.6907 −0.6126 −0.2848 −0.1060 −0.1827 −0.1184

−0.3041 −0.2116 0.4139 0.5500 0.4715 −0.0182

0.5589 0.4246 0.7272 0.7205 0.5757 1.0167

−3.090 −2.076 4.890 6.346 5.492 −0.142

Table A.23 Cartesian and fractional coordinates of cellulose triacetate II (preliminary model). Orthorhombic unit cell, space group P212121: a = 24.68 Å, b = 11.52 Å, c(fiber axis) = 10.44 Å, V = 2,968 Å3, rcalc=1.29 g cm−3, antiparallel chains (T=20°C). A dimer is required as the basic unit to describe the structure in P212121. (Zugenmaier, unpublished results 2005) X Y Z Atoms x y z 8.468 9.243 9.314 10.403 9.924 8.233 8.870 10.751 12.903 12.049 12.365 11.104 10.652 11.407 12.750 7.641 8.700 7.174 8.332 9.450 8.524 11.239 9.290 9.956 7.461 7.118 6.982 12.456 11.597 13.270 12.789

−0.399 −0.623 0.063 0.564 0.669 −1.071 −1.196 1.887 1.609 2.291 3.635 0.841 3.027 2.092 2.245 −2.432 −3.405 −5.057 −4.716 −5.715 0.751 −0.068 1.495 −1.375 −0.363 −2.380 −2.715 3.550 4.312 3.986 3.159

3.464 4.650 7.438 6.515 5.076 5.700 6.969 7.000 6.390 6.907 7.501 4.240 3.925 3.788 3.122 5.380 5.273 5.379 5.291 5.196 7.516 6.569 4.945 4.483 5.766 4.470 6.145 8.602 7.264 7.104 2.490

O41 C41 C11 C21 C31 C51 O51 O21 O21C C21C C21M O31 O31C C31C C31M C61 O61 O61C C61C C61M H11 H21 H31 H41 H51 H61A H61B H21Ma H21Mb H21Mc H31Ma

0.3431 0.3745 0.3774 0.4215 0.4021 0.3336 0.3594 0.4356 0.5228 0.4882 0.5010 0.4499 0.4316 0.4622 0.5166 0.3096 0.3525 0.2907 0.3376 0.3829 0.3454 0.4554 0.3764 0.4034 0.3023 0.2884 0.2829 0.5047 0.4699 0.5377 0.5182

−0.0346 −0.0541 0.0055 0.0490 0.0581 −0.0930 −0.1038 0.1638 0.1397 0.1989 0.3155 0.0730 0.2628 0.1816 0.1949 −0.2111 −0.2956 −0.4390 −0.4094 −0.4961 0.0652 −0.0059 0.1298 −0.1194 −0.0315 −0.2066 −0.2357 0.3082 0.3743 0.3460 0.2742

0.3318 0.4454 0.7125 0.6240 0.4862 0.5460 0.6675 0.6705 0.6121 0.6616 0.7185 0.4061 0.3760 0.3628 0.2990 0.5153 0.5051 0.5152 0.5068 0.4977 0.7199 0.6292 0.4737 0.4294 0.5523 0.4282 0.5886 0.8239 0.6958 0.6805 0.2385 (continued)

270 Table A.23 X 12.947 13.490 10.361 9.667 9.144 9.884 9.107 9.035 7.947 8.426 10.119 9.480 7.599 5.449 6.301 5.987 7.249 7.698 6.942 5.600 10.709 9.650 11.178 10.018 8.900 9.828 7.113 9.080 8.394 10.891 11.232 11.370 5.894 6.752 5.079 5.563 5.402 4.859 7.989 8.682 9.206 8.468

Appendix (continued) Y 1.401 2.318 −5.342 −5.902 −6.611 −0.211 0.013 −0.674 −1.174 −1.279 0.462 0.586 −2.498 −2.220 −2.902 −4.245 −1.452 −3.637 −2.703 −2.856 1.821 2.796 4.448 4.108 5.106 −1.361 −0.541 −2.093 0.765 −0.247 1.771 2.105 −4.160 −4.921 −4.595 −3.768 −2.010 −2.927 4.732 5.292 6.002 −0.399

Z 2.530 3.864 5.711 4.186 5.650 8.684 9.870 12.658 11.735 10.296 10.920 12.189 12.220 11.610 12.127 12.721 9.460 9.145 9.008 8.342 10.600 10.493 10.599 10.511 10.416 12.736 11.789 10.183 9.703 10.986 9.690 11.365 13.822 12.484 12.324 7.710 7.750 9.084 10.931 9.406 10.870 13.904

Atoms H31Mb H31Mc H61Ma H61Mb H61Mc O43 C43 C13 C23 C33 C53 O53 O23 O23C C23C C23M O33 O33C C33C C33M C63 O63 O63C C63C C63M H13 H23 H33 H43 H53 H63A H63B H23Ma H23Mb H23Mc H33Ma H33Mb H33Mc H63Ma H63Mb H63Mc O45

x 0.5246 0.5466 0.4198 0.3917 0.3705 0.4005 0.3690 0.3661 0.3220 0.3414 0.4100 0.3841 0.3079 0.2208 0.2553 0.2426 0.2937 0.3119 0.2813 0.2269 0.4339 0.3910 0.4529 0.4059 0.3606 0.3982 0.2882 0.3679 0.3401 0.4413 0.4551 0.4607 0.2388 0.2736 0.2058 0.2254 0.2189 0.1969 0.3237 0.3518 0.3730 0.3431

y 0.1216 0.2012 −0.4637 −0.5123 −0.5739 −0.0183 0.0011 −0.0585 −0.1019 −0.1110 0.0401 0.0509 −0.2168 −0.1927 −0.2519 −0.3685 −0.1260 −0.3157 −0.2346 −0.2479 0.1581 0.2427 0.3861 0.3566 0.4432 −0.1181 −0.0470 −0.1817 0.0664 −0.0214 0.1537 0.1827 −0.3611 −0.4272 −0.3989 −0.3271 −0.1745 −0.2541 0.4108 0.4594 0.5210 −0.0346

z 0.2423 0.3701 0.5470 0.4010 0.5412 0.8318 0.9454 1.2125 1.1240 0.9862 1.0460 1.1675 1.1705 1.1121 1.1616 1.2185 0.9061 0.8760 0.8628 0.7990 1.0153 1.0051 1.0152 1.0068 0.9977 1.2199 1.1292 0.9754 0.9294 1.0523 0.9282 1.0886 1.3239 1.1958 1.1805 0.7385 0.7423 0.8701 1.0470 0.9010 1.0412 1.3318

Appendix

271

Table A.24 Preliminary Cartesian and fractional coordinates of an asymmetric unit (two residues) of cellulose triacetate–nitromethane without placement of nitromethane. Tetragonal unit cell, space group P41212: a = b = 15.01 Å, c(fiber axis) = 41.14 Å, V = 9,269 Å3, antiparallel chains (T=20°C). (Zugenmaier, unpublished results 2005) X Y Z Atoms x y z 7.655 7.371 6.919 5.748 6.024 8.472 8.080 4.557 3.576 3.681 2.370 4.989 4.264 4.481 3.191 9.795 9.589 10.558 11.570 10.184 7.054 5.586 6.262 7.320 8.598 10.152 10.489 2.602 1.654 2.094 3.520 2.846 2.325 9.878 9.374 10.770 6.719 7.544 7.853 8.851 8.195 6.665 7.435 9.243

−0.961 −0.080 −0.094 0.147 −0.505 −0.221 0.476 −0.490 0.300 1.497 −0.523 −0.111 −1.107 −2.266 −0.520 0.349 1.730 2.309 1.739 3.696 −1.121 1.177 −1.477 0.914 −1.238 −0.204 0.307 −1.428 −0.576 0.114 −0.008 0.140 −0.999 4.162 3.510 4.398 0.574 0.152 0.481 1.139 1.405 −0.527 −0.743 2.431

−1.282 −0.166 2.662 1.734 0.394 0.866 2.046 2.234 2.725 2.849 3.067 −0.523 −1.111 −0.944 −1.964 0.412 0.010 −0.709 −0.988 −1.118 2.829 1.612 0.714 −0.500 1.096 −0.406 1.199 3.362 2.302 3.855 −2.729 −1.223 −2.316 −0.347 −1.759 −1.633 3.860 4.977 7.805 6.877 5.536 6.009 7.188 7.376

O41 C41 C11 C21 C31 C51 O51 O21 C21C O21C C21M O31 C31C O31C C31M C61 O61 C61C O61C C61M H11 H21 H31 H41 H51 H61A H61B H21Ma H21Mb H21Mc H31Ma H31Mb H31Mc H61Ma H61Mb H61Mc O43 C43 C13 C23 C33 C53 O53 O23

0.5010 0.4910 0.4609 0.3830 0.4013 0.5644 0.5383 0.3036 0.2382 0.2452 0.1579 0.3324 0.2841 0.2985 0.2126 0.6526 0.6388 0.7034 0.7708 0.6785 0.4700 0.3722 0.4172 0.4877 0.5728 0.6764 0.6988 0.1734 0.1102 0.1395 0.2345 0.1896 0.1549 0.6581 0.6245 0.7175 0.4477 0.5026 0.5232 0.5897 0.5460 0.4440 0.4953 0.6158

−0.0641 −0.0053 −0.0062 0.0098 −0.0337 −0.0148 0.0317 −0.0327 0.0200 0.0997 −0.0348 −0.0074 −0.0738 −0.1509 −0.0346 0.0232 0.1153 0.1538 0.1159 0.2462 −0.0747 0.0784 −0.0984 0.0609 −0.0825 −0.0136 0.0205 −0.0951 −0.0384 0.0076 −0.0006 0.0093 −0.0666 0.2773 0.2338 0.2930 0.0382 0.0101 0.0320 0.0759 0.0936 −0.0351 −0.0495 0.1620

−0.0312 −0.0040 0.0647 0.0422 0.0096 0.0211 0.0497 0.0543 0.0662 0.0693 0.0746 −0.0127 −0.0270 −0.0229 −0.0477 0.0100 0.0003 −0.0172 −0.0240 −0.0272 0.0688 0.0392 0.0174 −0.0122 0.0266 −0.0099 0.0291 0.0817 0.0560 0.0937 −0.0663 −0.0297 −0.0563 −0.0084 −0.0428 −0.0397 0.0938 0.1210 0.1897 0.1672 0.1346 0.1461 0.1747 0.1793 (continued)

272

Appendix

Table A.24 (continued) X Y

Z

Atoms

x

y

z

10.495 11.267 10.766 9.206 9.013 8.042 10.188 6.132 7.255 6.979 5.860 8.224 7.031 9.694 7.340 8.282 5.857 5.488 5.612 9.962 11.236 11.412 10.317 10.898 10.461 8.770 8.665 8.307 8.466

7.868 7.991 8.210 4.619 4.032 4.199 3.179 5.554 5.153 4.433 4.154 4.025 7.972 6.754 5.856 4.643 6.238 4.736 6.341 8.504 7.445 8.997 2.413 3.919 2.826 4.796 3.383 3.510 9.003

C23C O23C C23M O33 C33C O33C C33M C63 O63 C63C O63C C63M H13 H23 H33 H43 H53 H63A H63B H23Ma H23Mb H23Mc H33Ma H33Mb H33Mc H63Ma H63Mb H63Mc O45

0.6992 0.7506 0.7173 0.6133 0.6005 0.5357 0.6788 0.4085 0.4833 0.4649 0.3904 0.5479 0.4684 0.6458 0.4890 0.5518 0.3902 0.3657 0.3739 0.6637 0.7485 0.7603 0.6873 0.7261 0.6969 0.5843 0.5773 0.5534 0.5641

0.1710 0.1096 0.2665 0.1238 0.2048 0.2492 0.2277 −0.1243 −0.1797 −0.2526 −0.2734 −0.3003 0.0741 0.0350 0.1281 −0.0344 0.0069 −0.1151 −0.1550 0.2982 0.3028 0.2495 0.1881 0.2129 0.2911 −0.3078 −0.2534 −0.3610 0.0100

0.1912 0.1943 0.1996 0.1123 0.0980 0.1021 0.0773 0.1350 0.1253 0.1078 0.1010 0.0978 0.1939 0.1642 0.1424 0.1129 0.1516 0.1151 0.1541 0.2067 0.1810 0.2187 0.0587 0.0953 0.0687 0.1166 0.0822 0.0853 0.2188

2.567 1.646 4.000 1.858 3.075 3.741 3.418 −1.866 −2.697 −3.791 −4.104 −4.507 1.112 0.525 1.923 −0.516 0.103 −1.728 −2.327 4.477 4.545 3.746 2.823 3.195 4.369 −4.620 −3.803 −5.419 0.150

Table A.25 Cartesian and fractional coordinates of 2,3-di-O-acetyl-6-O-propanoyl cellulose II (preliminary model). Orthorhombic unit cell, space group P212121: a = 24.98 Å, b = 12.39 Å, c = 10.44 Å, V = 3,231 Å3, rexp=1.239 g cm−3, rcalc=1.241 g cm−3), antiparallel chains. A dimer is required as the basic unit to describe the structure in P212121. (Zugenmaier, unpublished results 2005) X Y Z Atoms x y z 8.651 9.430 9.490 10.567 10.084 8.431 9.070 10.884 13.042 12.175

−0.449 −0.657 0.032 0.555 0.649 −1.127 −1.238 1.885 1.653 2.317

3.544 4.729 7.518 6.594 5.155 5.780 7.048 7.079 6.470 6.987

O41 C41 C11 C21 C31 C51 O51 O21 O21C C21C

0.3463 0.3775 0.3799 0.4230 0.4037 0.3375 0.3631 0.4357 0.5221 0.4874

−0.0362 −0.0530 0.0026 0.0448 0.0524 −0.0910 −0.0999 0.1521 0.1334 0.1870

0.3395 0.4530 0.7201 0.6316 0.4938 0.5536 0.6751 0.6781 0.6197 0.6693 (continued)

Appendix

273

Table A.25 (continued) X

Y

Z

Atoms

x

y

z

12.460 11.261 10.761 11.536 12.877 7.869 8.945 7.457 8.608 9.747 9.355 8.686 11.416 9.432 10.159 7.644 7.344 7.214 12.555 11.681 13.359 12.895 13.092 13.614 10.589 9.977 10.109 8.456 9.225 10.062 9.283 9.225 8.146 8.628 10.282 9.642 7.829 5.670 6.537 6.252 7.454 7.951 7.177 5.835 10.844 9.767 11.256

3.666 0.846 3.022 2.104 2.286 −2.498 −3.451 −5.134 −4.768 −5.743 −7.070 0.703 −0.059 1.461 −1.394 −0.435 −2.458 −2.795 3.583 4.327 4.037 3.199 1.446 2.374 −5.348 −5.909 −7.783 −7.415 −6.924 −0.230 −0.024 −0.712 −1.235 −1.329 0.446 0.558 −2.565 −2.333 −2.997 −4.346 −1.526 −3.703 −2.784 −2.966 1.819 2.770 4.454

7.580 4.319 4.005 3.867 3.201 5.459 5.352 5.457 5.370 5.275 5.929 7.594 6.649 5.025 4.562 5.845 4.550 6.225 8.681 7.345 7.184 2.569 2.609 3.943 5.762 4.264 5.773 5.511 6.960 8.764 9.949 12.738 11.814 10.375 11.000 12.268 12.298 11.690 12.207 12.800 9.539 9.225 9.087 8.421 10.679 10.573 10.678

C21M O31 O31C C31C C31M C61 O61 O61C C61C C61D C61M H11 H21 H31 H41 H51 H61A H61B H21Ma H21Mb H21Mc H31Ma H31Mb H31Mc H61Da H61Db H61Ma H61Mb H61Mc O43 C43 C13 C23 C33 C53 O53 O23 O23C C23C C23M O33 O33C C33C C33 C63 O63 O63C

0.4988 0.4508 0.4308 0.4618 0.5155 0.3150 0.3581 0.2985 0.3446 0.3902 0.3745 0.3477 0.4570 0.3776 0.4067 0.3060 0.2940 0.2888 0.5026 0.4676 0.5348 0.5162 0.5241 0.5450 0.4239 0.3994 0.4047 0.3385 0.3693 0.4028 0.3716 0.3693 0.3261 0.3454 0.4116 0.3860 0.3134 0.2270 0.2617 0.2503 0.2984 0.3183 0.2873 0.2336 0.4341 0.3910 0.4506

0.2959 0.0683 0.2439 0.1698 0.1845 −0.2016 −0.2785 −0.4144 −0.3848 −0.4635 −0.5706 0.0567 −0.0048 0.1179 −0.1125 −0.0351 −0.1984 −0.2256 0.2892 0.3492 0.3258 0.2582 0.1167 0.1916 −0.4316 −0.4769 −0.6282 −0.5985 −0.5588 −0.0186 −0.0019 −0.0575 −0.0997 −0.1073 0.0360 0.0450 −0.2070 −0.1883 −0.2419 −0.3508 −0.1232 −0.2989 −0.2247 −0.2394 0.1468 0.2236 0.3595

0.7261 0.4137 0.3836 0.3704 0.3066 0.5229 0.5126 0.5227 0.5144 0.5053 0.5679 0.7274 0.6369 0.4813 0.4370 0.5599 0.4358 0.5963 0.8315 0.7035 0.6881 0.2461 0.2499 0.3777 0.5519 0.4084 0.5530 0.5279 0.6667 0.8395 0.9530 1.2201 1.1316 0.9938 1.0536 1.1751 1.1780 1.1197 1.1693 1.2261 0.9137 0.8836 0.8704 0.8066 1.0229 1.0127 1.0228 (continued)

274

Appendix

Table A.25 (continued) X

Y

Z

Atoms

x

y

z

10.107 8.968 9.355 10.029 7.297 9.298 8.553 11.071 11.368 11.498 6.160 7.034 5.353 5.818 5.620 5.098 8.126 8.736 8.598 10.254 9.487 8.651

4.090 5.063 6.390 −1.383 −0.619 −2.131 0.712 −0.247 1.779 2.116 −4.263 −5.007 −4.717 −3.879 −2.126 −3.054 4.666 5.229 7.103 6.736 6.243 −0.449

10.591 10.496 11.149 12.815 11.869 10.264 9.783 11.066 9.770 1.445 13.901 12.565 12.404 7.789 7.829 9.163 10.983 9.485 10.993 10.731 12.180 13.984

C63C C63D C63M H13 H23 H33 H43 H53 H63A H63B H23Ma H23Mb H23Mc H33Ma H33Mb H33Mc H63Da H63Db H63Ma H63Mb H63Mc O45

0.4046 0.3590 0.3745 0.4015 0.2921 0.3722 0.3424 0.4432 0.4551 0.4603 0.2466 0.2816 0.2143 0.2329 0.2250 0.2041 0.3253 0.3497 0.3442 0.4105 0.3798 0.3463

0.3301 0.4086 0.5157 −0.1116 −0.0500 −0.1720 0.0575 −0.0199 0.1436 0.1708 −0.3441 −0.4041 −0.3807 −0.3131 −0.1716 −0.2465 0.3766 0.4220 0.5733 0.5437 0.5039 −0.0362

1.0145 1.0054 1.0679 1.2275 1.1369 0.9831 0.9371 1.0600 0.9358 1.0963 1.3315 1.2035 1.1881 0.7461 0.7499 0.8777 1.0520 0.9085 1.0530 1.0279 1.1667 1.3395

Table A.26 Cartesian and fractional coordinates of a monomeric unit of cellulose tribenzoate. Trigonal space group P32: a = b= 12.20 Å, c(fiber axis) =15.21 Å, a = b=90.00°, g = 120.00°, V =1,941 Å3, rexp=1.24 g cm−3, rcalc=1.218 g cm−3, parallel chains (T=20°C). One residue is sufficient to describe the one-chain unit cell. (From Riehl 1992; Zugenmaier, unpublished results 2005) X Y Z Atoms x y z −0.333 −0.269 −0.885 −1.689 −1.694 0.474 0.444 −3.026 −3.494 −2.954 −2.329 −3.421 −3.794 1.936 2.553 3.835 4.591

−1.101 −0.209 0.061 0.839 0.117 −0.922 −0.137 1.005 2.267 3.221 0.935 0.438 −0.697 −1.154 0.012 −0.092 −0.940

0.000 1.116 3.927 2.891 1.553 2.242 3.444 3.371 3.483 2.996 0.570 −0.046 0.084 1.920 1.389 0.978 1.367

O41 C41 C11 C21 C31 C51 O51 O21 C21A O21A O31 C31A O31A C61 O61 C61A O61A

−0.0315 −0.0255 −0.0838 −0.1599 −0.1603 0.0449 0.0420 −0.2864 −0.3307 −0.2796 −0.2204 −0.3238 −0.3591 0.1832 0.2416 0.3630 0.4345

−0.1060 −0.0299 −0.0369 −0.0112 −0.0706 −0.0531 0.0098 −0.0608 0.0205 0.1242 −0.0336 −0.1260 −0.2367 −0.0030 0.1218 0.1740 0.1402

0.0000 0.0734 0.2582 0.1901 0.1021 0.1474 0.2264 0.2216 0.2290 0.1970 0.0375 −0.0030 0.0055 0.1262 0.0913 0.0643 0.0899 (continued)

Appendix

275

Table A.26 (continued) X

Y

Z

Atoms

x

y

z

−1.345 −1.254 −2.105 0.237 0.013 2.021 2.435 −4.702 −5.306 −6.488 −7.068 −6.463 −5.279 −4.097 −5.249 −5.911 −5.420 −4.268 −3.605 4.210 5.512 5.880 4.948 3.647 3.278 −4.874 −6.943 −7.954 −6.895 −4.826 −5.617 −6.776 −5.917 −3.899 −2.742 6.210 6.856 5.225 2.949 2.304 −0.787

−0.858 1.787 −0.822 0.671 −1.847 −1.942 −1.427 2.363 3.610 3.700 2.546 1.300 1.208 1.391 1.017 1.950 3.259 3.631 2.696 0.910 0.930 1.911 2.871 2.851 1.870 4.474 4.637 2.616 0.436 0.272 0.037 1.672 3.958 4.612 2.979 0.210 1.927 3.607 3.570 1.856 0.839

4.140 2.765 1.780 0.849 2.424 1.229 2.803 4.244 4.446 5.191 5.733 5.530 4.787 −0.870 −1.571 −2.377 −2.482 −1.781 −0.973 0.029 −0.487 −1.416 −1.827 −1.308 −0.382 4.040 5.342 6.291 5.936 4.634 −1.494 −2.905 −3.089 −1.859 −0.447 −0.179 −1.804 −2.523 −1.617 0.008 5.069

H11 H21 H31 H41 H51 H61A H61B C2b1 C2b2 C2b3 C2b4 C2b5 C2b6 C3b1 C3b2 C3b3 C3b4 C3b5 C3b6 C6b1 C6b2 C6b3 C6b4 C6b5 C6b6 H2b2 H2b3 H2b4 H2b5 H2b6 H3b2 H3b3 H3b4 H3b5 H3b6 H6b2 H6b3 H6b4 H6b5 H6b6 O43

−0.1273 −0.1187 −0.1992 0.0224 0.0012 0.1913 0.2305 −0.4450 −0.5022 −0.6141 −0.6690 −0.6117 −0.4996 −0.3878 −0.4968 −0.5595 −0.5130 −0.4040 −0.3412 0.3985 0.5217 0.5565 0.4683 0.3452 0.3103 −0.4613 −0.6571 −0.7528 −0.6526 −0.4568 −0.5316 −0.6413 −0.5600 −0.3690 −0.2595 0.5878 0.6489 0.4945 0.2791 0.2181 −0.0745

−0.1340 0.0871 −0.1670 0.0662 −0.1508 −0.0635 −0.0017 −0.0288 0.0448 −0.0038 −0.1258 −0.1993 −0.1508 −0.0799 −0.1650 −0.1199 0.0106 0.0956 0.0504 0.2738 0.3371 0.4349 0.4695 0.4063 0.3084 0.1361 0.0515 −0.1620 −0.2906 −0.2061 −0.2628 −0.1836 0.0444 0.1935 0.1144 0.3111 0.4824 0.5429 0.4322 0.2612 0.0315

0.2722 0.1818 0.1170 0.0558 0.1594 0.0808 0.1843 0.2790 0.2923 0.3413 0.3769 0.3636 0.3147 −0.0572 −0.1033 −0.1563 −0.1632 −0.1171 −0.0640 0.0019 −0.0320 −0.0931 −0.1201 −0.0860 −0.0251 0.2656 0.3512 0.4136 0.3903 0.3047 −0.0982 −0.1910 −0.2031 −0.1222 −0.0294 −0.0118 −0.1186 −0.1659 −0.1063 0.0005 0.3333

276

Appendix

Table A.27 Cartesian and fractional coordinates of trimethyl cellulose (preliminary model). Orthorhombic unit cell, space group P21221: a = 43.18 Å, b = 4.644 Å, c =10.45 Å, a = b = g = 90.0°, V = 2,096 Å3, rcalc=1.294 g cm−3, antiparallel packing of molecules (T=20°C). A dimer is required as the basic unit to describe the structure in P21221. (Zugenmaier, unpublished results 2005) X Y Z Atoms x y z 5.597 5.667 5.157 4.224 4.248 6.575 6.469 2.859 2.119 3.501 2.198 8.030 8.498 9.875 4.847 4.508 3.798 6.045 6.259 8.127 8.594 2.128 2.552 1.128 1.570 2.244 1.816 10.454 10.070 10.119 5.213 5.143 5.653 6.586 6.562 4.235 4.341 7.951 8.691 7.309 8.612 2.780 2.312 0.935

0.506 1.362 1.376 2.134 1.505 0.716 1.444 2.024 3.230 2.345 1.908 0.708 2.080 2.306 0.378 3.144 0.557 2.304 −0.273 0.257 0.175 3.738 3.837 3.008 1.952 0.920 2.531 1.650 2.141 3.297 1.997 1.142 1.127 0.369 0.999 1.788 1.059 0.479 −0.727 0.158 0.595 1.796 0.423 0.197

0.841 2.012 4.829 3.911 2.532 3.039 4.259 4.359 4.476 1.636 1.278 2.631 2.563 2.825 4.929 3.855 2.576 1.744 3.201 1.688 3.339 3.557 5.216 4.742 2.117 0.925 0.523 2.246 3.843 2.578 6.066 7.237 10.054 9.136 7.757 8.264 9.484 9.584 9.701 6.861 6.503 7.856 7.788 8.050

O41 C41 C11 C21 C31 C51 O51 O21 C21M O31 C31M C61 O61 C61M H11 H21 H31 H41 H51 H61A H61B H21Ma H21Mb H21Mc H31Ma H31Mb H31Mc H61Ma H61Mb H61Mc O43 C43 C13 C23 C33 C53 O53 O23 C23M O33 C33M C63 O63 C63M

0.1296 0.1312 0.1194 0.0978 0.0984 0.1523 0.1498 0.0662 0.0491 0.0811 0.0509 0.1860 0.1968 0.2287 0.1123 0.1044 0.0879 0.1400 0.1450 0.1882 0.1990 0.0493 0.0591 0.0261 0.0364 0.0520 0.0421 0.2421 0.2332 0.2343 0.1207 0.1191 0.1309 0.1525 0.1520 0.0981 0.1005 0.1841 0.2013 0.1693 0.1994 0.0644 0.0535 0.0216

0.1091 0.2932 0.2963 0.4596 0.3240 0.1541 0.3110 0.4358 0.6956 0.5051 0.4109 0.1524 0.4480 0.4966 0.0813 0.6769 0.1200 0.4961 −0.0587 0.0553 0.0377 0.8049 0.8262 0.6478 0.4203 0.1982 0.5449 0.3552 0.4611 0.7100 0.4300 0.2458 0.2428 0.0794 0.2151 0.3850 0.2281 0.1032 −0.1566 0.0340 0.1281 0.3867 0.0911 0.0424

0.0805 0.1925 0.4621 0.3742 0.2423 0.2908 0.4075 0.4171 0.4283 0.1566 0.1222 0.2518 0.2452 0.2703 0.4716 0.3689 0.2465 0.1669 0.3064 0.1615 0.3195 0.3404 0.4991 0.4538 0.2026 0.0885 0.0500 0.2149 0.3677 0.2467 0.5805 0.6925 0.9621 0.8742 0.7423 0.7908 0.9075 0.9171 0.9283 0.6566 0.6223 0.7518 0.7452 0.7703 (continued)

Appendix

277

Table A.27 (continued) X 5.962 6.301 7.012 4.764 4.550 2.683 2.216 8.681 8.257 9.682 9.240 8.565 8.994 0.355 0.740 0.690 5.597

Y 2.126 −0.640 1.946 0.200 2.776 2.247 2.328 −1.234 −1.334 −0.505 0.552 1.583 −0.027 0.854 0.362 −0.794 0.506

Z 10.154 9.080 7.801 6.969 8.426 6.913 8.564 8.782 10.441 9.967 7.342 6.150 5.748 7.471 9.068 7.803 11.291

Atoms H13 H23 H33 H43 H53 H63A H63B H23Ma H23Mb H23Mc H33Ma H33Mb H33Mc H63Ma H63Mb H63Mc O45

x 0.1381 0.1459 0.1624 0.1103 0.1054 0.0621 0.0513 0.2010 0.1912 0.2242 0.2140 0.1984 0.2083 0.0082 0.0171 0.0160 0.1296

y 0.4578 −0.1379 0.4191 0.0430 0.5977 0.4838 0.5013 −0.2658 −0.2872 −0.1087 0.1188 0.3409 −0.0059 0.1838 0.0780 −0.1710 0.1091

z 0.9716 0.8689 0.7465 0.6669 0.8064 0.6615 0.8195 0.8404 0.9991 0.9538 0.7026 0.5885 0.5500 0.7149 0.8677 0.7467 1.0805

Table A.28 Cartesian and fractional coordinates of 6-O-acetyl-2,3-di-O-methyl cellulose (preliminary model). Orthorhombic unit cell, space group P21221: a = 48.04 Å, b = 4.53 Å, c(fiber axis) =10.47 Å, a = b = g = 90.0°, V = 2,278 Å3, rcalc=1.354 g cm−3, antiparallel packing of molecules (T=20°C). A dimer is required as the basic unit to describe the structure in P21221. (Zugenmaier 2005, unpublished results 2005) X Y Z Atoms x y z 6.264 6.192 5.722 4.703 4.766 7.144 7.023 3.392 2.666 3.882 2.642 8.604 9.392 11.208 10.694 11.707 5.486 4.953 4.434 6.485 6.884

0.007 0.854 0.790 1.463 0.844 0.252 0.933 1.363 2.573 1.564 0.944 0.321 −0.623 −0.105 −0.730 −1.614 −0.228 2.479 −0.151 1.831 −0.755

0.496 1.649 4.491 3.571 2.190 2.682 3.940 4.117 4.275 1.335 1.019 2.288 3.005 1.781 2.669 3.316 4.605 3.491 2.253 1.405 2.825

O41 C41 C11 C21 C31 C51 O51 O21 C21M O3 C31M C61 O61 O61C C61C C61M H11 H21 H31 H41 H51

0.1304 0.1289 0.1191 0.0979 0.0992 0.1487 0.1462 0.0706 0.0555 0.0808 0.0550 0.1791 0.1955 0.2333 0.2226 0.2437 0.1142 0.1031 0.0923 0.1350 0.1433

0.0015 0.1885 0.1743 0.3229 0.1864 0.0557 0.2059 0.3009 0.5680 0.3452 0.2084 0.0708 −0.1376 −0.0232 −0.1612 −0.3564 −0.0503 0.5473 −0.0333 0.4043 −0.1667

0.0474 0.1575 0.4289 0.3411 0.2092 0.2562 0.3763 0.3932 0.4083 0.1275 0.0973 0.2185 0.2870 0.1701 0.2549 0.3167 0.4398 0.3334 0.2152 0.1342 0.2698 (continued)

278

Appendix

Table A.28 (continued) X

Y

Z

Atoms

x

y

z

8.969 8.686 1.720 2.546 3.190 2.085 2.104 2.825 12.332 12.279 11.217 5.722 5.794 6.264 7.283 7.220 4.842 4.963 8.594 9.320 8.104 9.344 3.382 2.594 0.778 1.292 0.279 6.500 7.033 7.547 5.501 5.102 3.017 3.300 10.266 9.435 8.796 9.901 9.882 9.161 −0.346 −0.293 0.769 6.264

1.289 0.114 2.364 3.037 3.214 1.571 0.769 0.030 −1.038 −2.090 −2.338 1.424 0.577 0.641 −0.032 0.586 1.178 0.498 0.067 −1.143 −0.134 0.487 1.110 2.054 1.535 2.160 3.045 1.658 −1.049 1.581 −0.401 2.185 0.141 1.317 −0.934 −1.607 −1.783 −0.141 0.661 1.400 2.468 3.520 3.769 0.007

2.465 1.262 4.682 3.340 4.920 0.387 1.902 0.534 3.929 2.576 3.898 5.731 6.884 9.726 8.806 7.425 7.917 9.175 9.352 9.510 6.570 6.254 7.523 8.240 7.016 7.904 8.551 9.840 8.726 7.488 6.640 8.060 7.700 6.497 9.917 8.575 10.155 5.622 7.137 5.769 9.164 7.811 9.133 10.966

H61A H61B H21Ma H21Mb H21Mc H31Ma H31Mb H31Mc H61Ma H61Mb H61Mc O43 C43 C13 C23 C33 C53 O53 O23 C23M O33 C33M C63 O63 O63C C63C C63M H13 H23 H33 H43 H53 H63A H63B H23Ma H23Mb H23Mc H33Ma H33Mb H33Mc H63Ma H63Mb H63Mc O45

0.1867 0.1808 0.0358 0.0530 0.0664 0.0434 0.0438 0.0588 0.2567 0.2556 0.2335 0.1191 0.1206 0.1304 0.1516 0.1503 0.1008 0.1033 0.1789 0.1940 0.1687 0.1945 0.0704 0.0540 0.0162 0.0269 0.0058 0.1353 0.1464 0.1571 0.1145 0.1062 0.0628 0.0687 0.2137 0.1964 0.1831 0.2061 0.2057 0.1907 −0.0072 −0.0061 0.0160 0.1304

0.2845 0.0251 0.5219 0.6704 0.7095 0.3469 0.1698 0.0067 −0.2291 −0.4613 −0.5162 0.3143 0.1273 0.1415 −0.0071 0.1293 0.2600 0.1099 0.0149 −0.2523 −0.0295 0.1074 0.2450 0.4534 0.3389 0.4769 0.6721 0.3660 −0.2316 0.3490 −0.0886 0.4824 0.0312 0.2907 −0.2062 −0.3547 −0.3937 −0.0312 0.1460 0.3090 0.5449 0.7771 0.8319 0.0015

0.2354 0.1205 0.4472 0.3190 0.4699 0.0370 0.1817 0.0510 0.3753 0.2460 0.3723 0.5474 0.6575 0.9289 0.8411 0.7092 0.7562 0.8763 0.8932 0.9083 0.6275 0.5973 0.7185 0.7870 0.6701 0.7549 0.8167 0.9398 0.8334 0.7152 0.6342 0.7698 0.7354 0.6205 0.9472 0.8190 0.9699 0.5370 0.6817 0.5510 0.8753 0.7460 0.8723 1.0474

Appendix

279

References Arnott S, Scott WE (1972) Accurate X-ray diffraction analysis of fibrous polysaccharides containing pyranose rings. Part I. The linked-atom approach. J Chem Soc Perkin Trans II 324–335 Atalla RH, VanderHart DL (1989) Studies on the structure of cellulose using Raman spectroscopy and solid state 13C NMR. In: Schuerch C (ed) Cellulose and wood – chemistry and technology. Wiley, New York, pp 169–188 Gardiner ES, Sarko A (1985) Packing analysis of carbohydrates and polysacchaides. 16. The crystal structures of cellulose IV1 and IV2. Can J Chem 63:173–180 Gessler K, Krauß N, Steiner T, Betzel C, Sarko A, Sänger W (1995) b-d-Cellotetraose hemihydrate as a structural model for cellulose II. An X-ray diffraction study. J Am Chem Soc 117:11397–11406 Ham JT, Williams DG (1970) The crystal and molecular structure of b-cellobioside-methanol. Acta Crystallogr Sect B 26:1373–1383 Langan P, Nishiyama Y, Chanzy H (2001) X-ray structure of mercerized cellulose II at 1 Å resolution. Biomacromolecules 2:410–416 Lee DM, Blackwell J (1981) Structure of cellulose II hydrate. Biopolymers 20:2165–2179 Lee DM, Blackwell J, Litt MH (1983) Structure of a cellulose II-hydrazine complex. Biopolymers 22:1383–1399 Lee DM, Burnfield KE, Blackwell J (1984) Structure of a cellulose I-ethylenediamine complex. Biopolymers 23:111–126 Leung F, Marchessault RH (1973) Crystal structure of b-d, 1→4 xylobiose hexaacetate. Can J Chem 51:1215–1222 Mackie ID, Röhrling J, Gould RO, Pauli J, Jäger C, Walkinshaw M, Potthast A, Rosenau T, Kosma P (2002) Crystal and molecular structure of methyl 4-O-methyl-b-d-glucopyranosyl(1–4)-b-d-glucopyranoside. Carbohydr Res 337:161–166 Maréchal Y, Chanzy H (2000) The hydrogen bond network in Iβ cellulose as observed by infrared spectrometry. J Mol Struct 523:183–196 Marrinan HJ, Mann J (1956) Infrared spectra of crystalline modifications of cellulose. J Polym Sci 21:301–311 Nishimura H, Sarko A (1991) Mercerization of cellulose. 6. Crystal and molecular structure of Na-cellulose IV. Macromolecules 24:771–778 Nishimura H, Okano T, Sarko A (1991) Mercerization of cellulose. 5. Crystal and molecular structure of Na-cellulose I. Macromolecules. 24:759–770 Nishiyama Y, Sugiyama J, Chanzy H, Langan P (2002) Crystal structure and hydrogen-bonding system in cellulose Iβ from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 124:9074–9082 Nishiyama Y, Chanzy H, Langan P (2003) Crystal structure and hydrogen-bonding system in cellulose Iα from synchrotron X-ray and neutron fiber diffraction. J Am Chem Soc 125:14300–14306 Pérez S, Brisse F (1977) The crystal and molecular structure of a trisaccharide, b-cellotriose undecaacetate: 1,2,3,6-tetra-O-acetyl-4-O-[2,3,6-tri-O-actyl-4-O-(2,3,4,6-tetra-O-acetyl-b-d-glucopyranosyl)-b-d-glucopyranosyl]-b-d-glucopyranose. Acta Crystallogr Sect B 33:2578–2584 Raymond S, Henrissat B, Tran Qui D, Kvick Å, Chanzy H (1995a) The crystal structure of methyl b-cellotrioside monohydrate 0.25 ethanolate and its relationship to cellulose II. Carbohydr Res 277:209–229 Raymond S, Heyraud A, Tran Qui D, Kvick Å, Chanzy H (1995b) Crystal and molecular structure of b-d-cellotetraose hemihydrate as a model of cellulose II. Macromolecules 28:2096–2100 Riehl K (1992) Strukturuntersuchungen an Cellulosederivaten mit stereoselektiver Trennwirkung. Doctoral dissertation, TU Clausthal, Clausthal-Zellerfeld Rivet A, Sabin C, Mazeau K, Imberty A, Perez S (2001) Disaccharides database. http://www. cermav.cnrs.fr. Cited 2 Sept 2004

280

Appendix

Shuto Y (1990) Molecular and crystal structure analysis of cellulose tripropionate. Doctoral dissertation, Kyoto University Sikorski P, Wada M, Heux L, Shintani H, Stokke B (2004) Crystal structure of cellulose triacetate I. Macromolecules 37:4547–4553 Wada M, Heux L, Isogai A, Nishiyama, Chanzy H, Sugiyama J (2001) Improved structural data of cellulose III1 prepared in supercritical ammonia. Macromolecules 34:1237–1243 Wada M, Chanzy H, Nishiyama Y, Langan P (2004) Cellulose III1 crystal structure and hydrogen bonding by synchrotron X-ray and neutron fiber diffraction. Macromolecules 37:8548–8555 Wada M, Nishiyama Y, Langan P (2006) X-ray structure of ammonia-cellulose I: new insights into the conversion of cellulose I to cellulose III1. Macromolecules 39:2947–2952 Zugenmaier P, Rappenecker G (1978) The crystal and molecular structure of methyl tetraacetylb-d-glucoside. Acta Crystallogr Sect B 34:164–167

Index

A b-d-acetyl cellobiose geometrical data 95, 96 intramolecular distances 96 aggregation, aggregate 7, 13, 14, 22-26, 209, 216 algal cellulose 40, 44, 109, 150, 208 amylose 18 analogous substitution 13, 26 animal (tunicin) cellulose 4, 37, 40, 45, 109, 112, 115, 120 anisotropic properties 13 application 1, 2 composites 2 enantiomeric separation 2, 178, 185 paper 1 textile fibers 2 wood 1, 2

B bacterial cellulose 4, 40, 45, 208, 209, 211, 215, 216 “biological” space groups 59

C cellobiose 13, 16, 19, 22-24, 69, 73, 87, 113, 118 cellotetraose geometrical data 84-85 hydrogen-bonding 84, 92 intramolecular distances 84 models 89, 90, 93 NMR 70 unit cell 88 b-cellotriose undecaacetate geometrical data 96, 97 intramolecular distances 97

models 98, 188 unit cell 243 cellulose bast fiber 1, 4 chain size 3, 13-15, 24, 27 chemical constitution 8, 22, 31 classification 20, 21 concept 7 cotton 1, 2, 4, 7, 24, 147, 208 domain size 207, 208, 210-215, 219 Gaussian coil 3, 213 hydrolysis 2, 16, 213, 214 length 3, 16, 19, 118, 212, 214 linkage 13, 14, 16, 22-25, 32, 214 microcrystalline 2 molecular weight (mass) 4, 14, 15, 25, 26, 209, 210 native fiber 3 purified 1 straight/ bent chain 13, 19, 20, 37 synthesis 1 unit cells 28, 29, 88, 102 virtual bond 13, 19, 83 worm-like chain 3 cellulose esters plastics 1 unit cells 182, 188 cellulose Iα electron fiber diffraction 126 geometrical data 124 hydrogen-bonding 124, 125 intrachain distances 124 IR 108, 149, 225 models 116, 123 NMR 107, 119, 128, 162 origin 4 Raman 227 unit cell 28, 29, 102 X-ray 122, 148

281

282 cellulose Iβ antiparallel packing 16, 18 electron fiber diffraction 126 geometrical data 115 hydrogen-bonding 117, 121 intrachain distances 121 IR 66, 108, 149, 223, 225 models 116, 117 NMR 107, 119, 149, 150, 162 origin 4 parallel packing 16, 17, 40, 42 Raman 227 unit cell 28, 29, 102 X-ray 114, 148 cellulose II antiparallel packing 87, 101, 132 geometrical data 133 hydrogen-bonding 134, 135 intrachain distances 134 IR 66, 225 models 131 NMR 107, 119, 130, 136 parallel packing 30 Raman 227 unit cell 88, 102 X-ray 130 cellulose III1 geometrical data 124 hydrogen-bonding 124, 125 intrachain distances 124 IR 149, 224, 226 models 140, 141 NMR 107, 140, 149, 161, 162 Raman 227 unit cell 102, 251 X-ray 139, 140, 148 cellulose III2 IR 226 X-ray 139 cellulose IV1 hydrogen-bonding 146, 147 intrachain distances 146 IR 226 models 144, 145 unit cell 102, 252 X-ray 105 cellulose IV2 hydrogen-bonding 146, 147 intrachain distances 146 IR 226 models 144, 147 unit cell 102, 254 X-ray 105 cellulose I-ammonia geometrical data 154

Index hydrogen-bonding 154, 159 intramolecular distances 154, 159 models 158 unit cell 102 X-ray 157 cellulose II-ethylenediamine geometrical data 154 hydrogen-bonding 154, 162 intramolecular distances 154 models 163 NMR 161, 162 unit cell 102 X-ray 161 cellulose II-hydrate geometrical data 154 hydrogen-bonding 154, 155 intramolecular distances 154 models 155 unit cell 102 X-ray 155 cellulose II-hydrazine hydrogen-bonding 152 models 163 unit cell 102, 259 X-ray 153 cellulose-6-O-acetyl-2,3-di-O-butyryl (CADB) 183 cellulose-6-O-acetyl-2,3-di-O-methyl (DMAC) models 203 unit cell 277 X-ray 203 cellulose-6-O-acetyl-2,3-di-O-propanoyl (CADP) EM observation 183 unit cell 182 cellulose-2,3-di-O-acetyl-6-O-butyryl (CDAB) 183 unit cell 182 cellulose-2,3-di-O-acetyl-6-O-propanoyl (CDAP) EM observation 197 models 198 unit cell 182 X-ray 197 cellulose nitrate 1, 36, 185 IR 66 unit cell 182 cellulose triacetate I (CTA I) 94, 95, 176, 217 density 188 geometrical data 189 models 186, 187 NMR 178, 192 unit cell 182, 188, 267 X-ray 177, 179, 180

Index cellulose triacetate II (CTA II) 184, 187, 191, 217 density 188 electron diffraction 193 geometrical data 189 IR 66 models 190, 191, 193, 194 NMR 178, 192 unit cell 182, 188, 269 X-ray 177, 179, 190 cellulose triacetate-nitromethane (CTA-N) 195, 196 electron diffraction 194 models 196 unit cell 188, 271 X-ray 179, 195 cellulose tribenzoate II (CTBe II) 197, 228 models 199 unit cell 182, 274 X-ray 199 cellulose tributyrate (CTB) 184 electron diffraction 184 unit cell 182 X-ray 184 cellulose tripropionate (CTP) 179 electron diffraction 180, 182 IR 228 NMR 183 unit cell 182 X-ray 180, 181 cellulose trivalerate (CTV) 185 unit cell 182 cell walls 1, 4, 103, 127 chemical composition 1, 2 chain defect 19, 36, 181, 211, 214 chain folding 13, 44, 193, 209, 213 chair configuration (conformation) 8, 10, 18, 19, 31, 32, 37 chain model continuous 14, 22, 32, 33, 165 rods 15, 26 statistical 15, 35, 36, 39, 219 characterization cellulose 104-112 cellulose acetate 175-179, 193 cellulose acetate-nitromethane 179, 194, 195 cellulose tributyrate 184 cellulose tripropionate 180-183 cellulose trivalerate 185 Debye-Scherrer 58, 62, 104 fiber diagram (pattern) 57, 58, 63, 105 fingerprint 66 IR 65, 67, 68

283 mixed cellulose ester 183 NMR 68-72 spectroscopy 64, 65 chemical shift 64, 65, 68, 69, 71, 72 chitin 18, 44, 103 cocrystallization 63, 64 conformation bent and twist 20, 185 conversion cellulose I-cellulose II 16, 30, 103, 217 cellulose Iα-cellulose Iβ 103, 108, 113, 145 cellulose Iβ-cellulose I-ammonia 109, 111, 145 cellulose I-ammonia -cellulose III1 109, 111, 145 cellulose III1-cellulose Iβ 103, 109, 111, 145 CTA I-CTA-N -CTA II 176, 179 parallel/ antiparallel 16-18, 30, 45, 176, 217 coordinates cellulose 246-256 cellulose complexes 256-267 cellulose derivatives 267-278 model compounds 229-246 crystallinity 63, 208, 215 crystallinity index 63, 207, 208 crystallite 13, 16, 27, 57, 212, 213 orientation 13, 27, 37, 58, 68 size 30, 44, 63, 145, 176, 210, 211 surface 5, 19, 118, 147, 208, 210

D degradation enzymatic 41 hydrolytic 2, 22-24, 212, 214 degree of polymerization (DP) 4, 13, 14, 26, 209, 213, 214 derivatization heterogeneous 55, 143, 176, 179, 180, 185, 197 homogeneous 176, 185, 197, 199 deuterium exchange 114, 122, 214, 246, 248 diffraction 53-60 diffusion 24, 30, 211 deuterium 211, 213 dimethyl-b-d-cellobioside geometrical data 79, 81 hydrogen-bonding 80, 81 intramolecular distances 81 unit cell 80 dipole 65, 87 disorder 57, 64, 97, 118, 136, 137, 142, 147, 211, 214

284 E electron diffraction 39, 41, 42, 44, 63, 127, 149, 193 energy map 82, 87

F fiber diffraction 31, 40, 57, 58, 119, 126 fibrils 3, 18, 19, 21, 41, 44, 151, 207, 212, 214-216 fringed micelle 13, 26, 207, 212, 214, 215

G glucopyranose 1, 7, 10, 16, 19, 23, 25, 37, 60 glucose anhydride 24

H helix 8/5, 5/3 symmetry 95, 175, 185, 194-196 theory 20, 55 hemicellulose 1, 2, 7

I identity (fiber) period 8, 20, 23, 30, 32, 105 indexing 30, 31, 186 interdiffusion 18, 45, 217, 219 intersheet 89, 90, 94, 115 intrasheet 89, 90, 118 IR spectroscopy 21, 63-68

L layer lines 31, 55-57, 105, 126 leveling-off DP (LODP) 214 linkage (1-1, 4-4) 13, 25, 32, 33 (1-4) 1, 7, 8, 16, 23, 31 long period 211, 212, 214 Lorentz correction 27, 64

M methyl b-d-cellobioside conformation 80 geometrical data 79 hydrogen-bonding 81 intramolecular distances 81 methyl b-d-cellobioside-methanol geometrical data 81 hydrogen-bonding 80, 81 intramolecular distances 81 unit cell 80

Index methyl b-cellotrioside geometrical data 86 hydrogen-bonding 86, 92 intramolecular distances 86 models 89, 91 unit cell 88 micelle 13, 19, 24-26, 210 microfibril models 13, 18, 19, 44, 208, 209 antiparallel 18, 218 diameter 30, 126, 207-210 electron diffraction 126, 213, 215 model building 12, 55, 60-64, 188 model compounds dimer 79-82, 94-96 monomer 95, 96 standard b-d-glucose 79 tetramer 84, 85, 88-90, 92, 93 trimer 86, 88-92, 96-98 models ball-and-stick 8, 10, 12 computer-aided 10, 19, 20, 40, 57, 60, 61, 112 historical development 27-45 space-filling 8, 12, 40 morphology bacterial cellulose 215, 216 bricklike 25, 213 fibrillar arrangement 126, 127, 150 mercerization 217

N neutron scattering 4, 42, 45, 62, 64, 114, 210, 211, 213, 216 NMR spectroscopy 68-72

O optical transform 54, 55 helices 56 orientation 4, 18, 27, 37, 55, 57, 59, 186 biaxial 31, 201 degree 27, 217 uniaxial 31 osmometry 14, 22, 24-26

P packing arrangement antiparallel 13, 16, 18, 33, 35, 39, 218 parallel 13, 16, 18, 30, 34, 39, 218 parallel-down 41, 72, 119, 217 parallel-up 72, 73, 119, 217 statistical 35, 36, 39, 219 partial crystalline 207

Index partial occupancy 44, 64, 114, 118, 122, 127, 246, 248 polarizability 65 polarization 18, 65, 214 polymorphy 101, 103 pulp 4, 7, 44, 208, 214

R Raman spectroscopy 5, 64, 65, 227 ratio Iα/Iβ 108-110 residue labeling 72, 114

S single crystal 27, 31, 53, 57, 63 polymeric 36, 42, 62, 63 sodium cellulose I dipolar-bonding 167 geometrical data 166 models 167, 168 unit cell 102 X-ray 166 sodium cellulose IV geometrical data 154 hydrogen-bonding 171 intramolecular distances 154 models 170 unit cell 102 X-ray 169 solution 1, 16, 209 aggregates 25, 26, 209 colloidal 24, 25 molecularly dispersed 3, 13, 15, 26 structure determination 4, 5 subelementary fibrils 147

T tetraacetyl-b-d-glucoside geometrical data 95, 96

285 intramolecular distances 96 unit cell 241 torsion angles (definition) position O6 (χ, χ’) 73 (Φ, Ψ) 73 trimethyl cellulose (TMC II) chain packing 219 EM observation 63 models 202 unit cell 276 X-ray 200, 201 2,3,6-trimethylglucose 16, 23

V valencies primary 8, 14, 15, 24, 32 secondary 9 viscosity 13, 26

X X-ray 25, 62-64 Debye-Scherrer 103-105 equator, cellulose I, II 42, 44, 126, 129 fiber pattern 31, 105, 129 meridional reflections 20, 32, 34, 36, 55, 56, 61, 63 reflection intensity 40, 41, 57, 62 reflection width 14, 19, 25, 57, 147, 176, 208, 210 SAXS 210, 212, 214 WAXS 212 Xylan 14, 214 b-d-xylobiose hexaacetate geometrical data 95, 96 intramolecular distances 96 unit cell 242