Springer Series in Wood Science

Series Editors

T. E. TimellState University of New YorkCollege of Environmental Science and ForestrySyracuse, NY 13210, USA

Professor Dr. Rupert WimmerBio-based Fibre MaterialsDepartment of Material Sciences and Process EngineeringUniversity of Natural Resources and Applied Life SciencesBOKU-ViennaPeter-Jordan-Strasse 821190 Vienna, Austria

Springer Series in Wood Science

Editors: T.E. Timell, R. Wimmer

L.W. Roberts/P.B. Gahan/R. AloniVascular Differentiation and Plant Growth Regulators (1988)

C. SkaarWood-Water Relations (1988)

J.M. HarrisSpital Grain and Wave Phenomena in Wood Formation (1989)

B.J. Zobel/J.P. van BuijtenenWood Variation (1989)

P. HakkilaUtilization of Residual Forest Biomass (1989)

J.W. Rowe (Ed.)Natural Products of Wood Plants (1989)

K.-E.L. Eriksson/R.A. Blanchette/P. AnderMicrobial 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. TorgovnikovDielectric Properties of Wood and Wood-Based Materials (1993)

F.H. SchweingruberTrees and Wood in Dendrochronology (1993)

P.R. LarsonThe Vascular Camblum: Development and Structure (1994)

M.-S. Ilvessalo-PfaffliFiber Atlac Identification of Papermaking Fibers (1995)

B.J. Zobel/J.B. JettGenetics of Wood Production (1995)

C. Matteck/H. KabierWood - The Internal Optimization of Wood (1997)

T. HiguchiBiochemistry and Molecular Biology of Trees (1997)

B.J. Zobel/J.R. SpragueJuvenile 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. WalkerKiln-Drying of Lumber (2000)

S. CarlquistComparative Wood Anatomy, 2nd ed. (2001)

M.T. Tyree/M.H. ZimmermannXylem Structure and the Ascent of Sap. 2nd ed. (2002)

T. Koshijima/T. WatanabeAssociation Between Lignin and Carbohydrates in Wood and Other Plant Tissues (2003)

V. BucurNondestructive Characterisation and Imaging of Wood (2003)

V. BucurAcoustics of Wood (2006)

F.H. SchweingruberWood Structure and Environment (2007)

P. ZugenmaierCrystalline Cellulose and Derivatives

Peter Zugenmaier

Crystalline Cellulose and Derivatives

Characterization and Structures

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 5 4 3 2 1 0

springer.com

Professor Peter ZugenmaierInstitute of Physical ChemistryTU ClausthalArnold-Sommerfeld-Str. 4D-38678 Clausthal-ZellerfeldGermany

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

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 sup-port 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

vi Preface

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 7.2J 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, A2Biomacromolecules: Figures 5.20, 5.33-36

ElsevierElsevier Publishing Company, Inc. New York-Amsterdam-London-Brussels: Physics and chemistry of cellulose fibres (1949): Figures 2.14, 5.19Academic Press, London-New York: Polymer and fibre diffraction (1972): Figure 3.6 Advances in Carbohydrate Chemistry and Biochemistry: Figure 3.4J Struct Biol: Figures 2.20, 5.7, 5.17J Mol Struct: Figure 3.11, A1 Prog Polym Sci: Figures 3.14, 4.4, 4.4Polymer: Table 5.11, Figure 5.24Polymer Comm: Figure 6.5Solid State Nucl Magn Reson: Figure 5.21Carbohydr Res: Table 7.2

John Wiley & Sons Ltd, New YorkEllis 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, 3.4John Wiley, New York: Cellulose and wood – chemistry and technology (1989): Figures A7, A8Ber Deutsch Chem Ges: Figure: 2.8

vii

Bioploymers: Figures 5.38, 5.43Helv Chim Acta: Figures 2.10, 2.17J Appl Polym Sci: Figures 7.7, 7.8J Polym Sci: Figures 2.19, A3-A6J Polym Sci B: Figures 5.9, 5.37J Polym Sci A: Figures 6.1, 6.2, 6.21J Polym Sci Phys Ed: Figure 6.19cMakromol Chem: Figure 7.1Verlag Chemie: Figure 2.7

Springer Verlag Berlin-Heidelberg:Cellulose: Table 5.10, Figures 5.18, 5.20, 5.28, 6.10, 6.24Polym Bull: Table 3.2, Figure 3.13Colloid & Polym Sci (Kolloid Z, Kolloid Z u Z Polymere):Table 2.2, Figures 6.3, 6.4, 6.21J Materials Sci: Figure 7.4Plenum Press, New York: Cellulose and other natural polymer systems (1982): Figure 7.5, Structural electron crystallography.(1995): Figure 6.20

Das Papier, Darmstadt: Figure 7.3Francis and Taylor, London: J Carbohydr Chem. Figure 3.12Hanser Publishers, Munich: Cellulosic polymers (1994): Table 5.2, Figures 5.1, 5.3 International Union of Pure and Applied Chemistry: Pure Appl Chem: Figure 3.4Oldenbourg Wissenschaftsverlag, München:: Z Phys Chem: Figures 2.5b, 2.9, 2.18The Society of Biotechnology, Osaka: J Biosci Bioeng: Table 5.20, Figure 5.44The Society of Polymer Science, Tokyo: Polymer J: Figure 5.5

The drawings of the models were produced with software fromKeller E (1992) Schakal 92: A computer program for graphic representation of molecules and crystallographic models. Freiburg

viii Acknowledgement for Copyrights

ix

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

General Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 History of Cellulose Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.1 The Concept of Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72.2 Concepts in Structural Research of Cellulose. . . . . . . . . . . . . . . . . . . 82.3 Arrangements of the Cellulose Molecules in the Solid State . . . . . . . 162.4 Chemical Constitution of Cellulose as a Macromolecule. . . . . . . . . . 21

2.4.1 Linkage of Cellulose – the Chain Structure of Cellulose (Freudenberg, Haworth) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.4.2 Macromolecule Formation – Size of the Chains (Staudinger) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.5 Historical Development of X-ray Models for Native Cellulose . . . . . 27References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.1 Diffraction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533.2 Model Building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603.3 Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

3.3.1 Optical Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653.3.2 NMR Spectroscopy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

3.4 Convention for the Description of Cellulosic (Chiral) Structures . . . 72References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

4 Model Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

4.1 Conformation and Packing Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 774.2 Monomers and Dimers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784.3 Trimers and Tetramers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

4.3.1 Conformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834.3.2 Packing Arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

4.4 Acetyl Derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

5 Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

5.1 Cellulose Polymorphy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1015.2 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

5.2.1 X-ray Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035.2.2 Spectroscopic Characterization . . . . . . . . . . . . . . . . . . . . . . . 105

5.3 Molecular and Crystal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125.3.1 Cellulose Iβ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135.3.2 Cellulose Iα . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1225.3.3 Cellulose II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1295.3.4 Cellulose III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385.3.5 Cellulose IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

5.4 Cellulose Solvent Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1515.4.1 Cellulose II–Hydrazine Complex . . . . . . . . . . . . . . . . . . . . . . 1515.4.2 Cellulose II Hydrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1525.4.3 Cellulose I–Ammonia I Complex . . . . . . . . . . . . . . . . . . . . . . 1565.4.4 Cellulose I–Ethylenediamine Complex . . . . . . . . . . . . . . . . . 160

5.5 Sodium Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1645.5.1 Sodium Cellulose I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1655.5.2 Sodium Cellulose IV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

6 Cellulose Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

6.1 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1756.1.1 Cellulose Triacetate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1756.1.2 Experimental Data for Cellulose Tripropionate and Cellulose

Acetate Dipropionate and Further Cellulose Esters . . . . . . . . 1796.2 Conformation and Packing Arrangement of CTA I, CTA II and CTA-N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1856.3 Conformation and Packing Arrangement of CDAP . . . . . . . . . . . . . . 1976.4 Conformation and Packing Arrangement of Cellulose Tribenzoate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1996.5 Trimethyl Cellulose and 6-O-Acetyl-2,3-di-O-methyl Cellulose . . . . 200References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204

7 Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

7.1 Crystalline Domain Sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2077.2 Microfibrils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2077.3 Microfibrils and Fibrils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2127.4 Parallel and Antiparallel Packing Arrangements of Microfibils. . . . . 216References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

x Contents

Chapter 1Introduction

Cellulose represents a naturally occurring linear macromolecular chain of 1–4-linked b-d-glucopyranose and exhibits great chemical variability and potential in applica-tions. 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 addi-tives 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 consid-erable 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. 1Springer Series in Wood Science.© Springer-Verlag Berlin Heidleberg 2008

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 struc-tures 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 deriva-tives. Cellulose fibers and films show excellent tensile properties. Thin sheets of cellulose acetate serve as optical compensators and shields for compounds evapo-rating 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 microcrystal-line 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

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

Composition (%)

Source Cellulose Hemicellulose Lignin Extract

Hardwood 43–47 25–35 16–24 2–8Softwood 40–44 25–29 25–31 1–5Bagasse 40 30 20 10Coir 32–43 10–20 43–49 4Corn cobs 45 35 15 5Corn stalks 35 25 35 5Cotton 95 2 1 0.4Flax (retted) 71 21 2 6Flax (unretted) 63 12 3 13Hemp 70 22 6 2Henequen 78 4–8 13 4Istle 73 4–8 17 2Jute 71 14 13 2Kenaf 36 21 18 2Ramie 76 17 1 6Sisal 73 14 11 2Sunn 80 10 6 3Wheat straw 30 50 15 5

2 1 Introduction

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 macromole-cules 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 solu-tions 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 dispersedsolution. 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

In this book we will present a short overview of basic principles in the develop-ment of macromolecular science, in particular the development of the macromo-lecular 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.

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 6,000–10,000Pulp 500–2,000Sulfate pulp 950–1,300Chemical pulp bleached 700Cotton 10,000–15,000Cotton linters bleached 1,000–5,000Valonia 25,000Bacterial cellulose 4,000–6,000Ramie 10,000Textile flax 9,000Rayon 300–500Cellophane 300Cellulose acetate 200–350

A comprehensive evaluation of molecular masses of native celluloses is provided by Schulz and Marx (1954) and Marx-Figini (1982).

IR, Fourier transform IR, Raman and NMR spectroscopic techniques have always served as complementary tools, especially if the long-range three- dimensional 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, BerlinFyfe CA (1983) Solid state NMR for chemists. CFC, GuelphHaworth WN (1929) The constitution of sugars. Edward &Arnold, LondonHaworth WN (1932) Die Konstitution der Kohlenhydrate. Steinkopff, DresdenHermans PH (1949) Physics and chemistry of cellulose fibres. Elsevier, New YorkHess K (1928) Die Chemie der Zellulose und ihrer Begleiter, XX. Akademische Verlagsgesellschaft,

LeipzigHon 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

General Literature 5

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, DresdenMarchessault RH, Sundararajan PR (1983) Cellulose. In: Aspinall GO (ed) The polysaccharides,

vol 2. Academic, New York, pp 11–95Mark H (1932) Physik und Chemie der Cellulose, XV. In: Herzog RO (ed) Technologie der

Textilfasern, vol I, part 1. Springer, BerlinMarsh JT, Wood FC (1939) An introduction to the chemistry of cellulose. Van Nostrand,

New YorkMarx-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 YorkMeyer KH, Mark H (1930) Der Aufbau der hochpolymeren organischen Naturstoffe. Akademische

Verlagsgesellschaft, LeipzigMeyer KH, Mark H (1950) Makromolekulare Chemie, 2nd edn. Geest & Portig, LeipzigMorawetz H (1985) Polymers – the origin and growth of a science. Wiley, New YorkOtt E (ed) (1943) Cellulose and cellulose derivatives, vol V. Interscience, New YorkOtt E, Spurlin HM, Grafflin MW (eds) (1955) Cellulose and cellulose derivatives, part III. High

polymers, vol 5, 2nd edn. Interscience, New YorkPriesner C (1980) H. Staudinger, H. Mark und K. H. Meyer – Thesen zur Größe und Struktur der

Makromoleküle. Verlag Chemie, WeinheimPummerer R (ed) (1953) Chemische Textilfasern – Filme und Folien. Enke, StuttgartPurves 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, 88–112Saechtling H (1935) Hochpolymere organische Naturstoffe. Vieweg, BraunschweigSchulz GV, Marx M (1954) Über Molekulargewichte und Molekulargewichtsverteilungen nativer

Cellulosen. Makromol Chem 14:52–95Sisson WA (1946) X-ray examination. In: Ott E (ed) Cellulose and cellulose derivatives. High

polymers, vol 5. Interscience, New York, pp 203–292Staudinger H (1932) Die hochpolymeren Verbindungen. Kautschuk und Cellulose, XV. Springer,

BerlinStuart HA (ed) (1955) Die Physik der Hochpolymeren, vol 3. Springer, BerlinWalton AG, Blackwell J (1973) Biopolymers. Academic, New York

Chapter 2History 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 origi-nal 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 treat-ment might modify its state of aggregation. In contrast, Frémy (1859), who investi-gated 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 mon-osaccharides at that time. Acid hydrolyzates of many materials consisting of cellu-loses 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 man-nose (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” fol-lowing Payen and further Schulze (1891), who used drastic extraction reagents, was

P. Zugenmaier, Crystalline Cellulose and Derivatives: Characterization and Structures. 7Springer Series in Wood Science.© Springer-Verlag Berlin Heidleberg 2008

8 2 History of Cellulose Research

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 experi-mental 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 sci-ence 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 paral-lel 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.

Fig. 2.1 Chemical constitution of cellulose as 1-4-linked b-d-anhydroglucopyranose and num-bering 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

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

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 sur-passed 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 mole-cules was quite common to visualize the placement of atoms according to the evalu-ation 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 prop-erties. 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 com-monly accepted today (Fig. 2.5).

2.2 Concepts in Structural Research of Cellulose 9

10 2 History of Cellulose Research

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

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

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

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)

2.2 Concepts in Structural Research of Cellulose 11

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 differentlywith 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, additionalinformation 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 omittingalso 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 build-ing 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)

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 determina-tions 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 cor-rected 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 struc-ture 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 cellu-lose 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 morpho-logical 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

2.2 Concepts in Structural Research of Cellulose 13

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 thesefactors must be recognized in any comparison of the molecular weight of cellulosedetermined by physical and chemical methods. All recent experiments in my labo-ratory 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 macromo-lecular structures.”

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 discerni-ble 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 poly-mers 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 valen-cies 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 mac-romolecules and the conceptual differences in the organic chemistry of macromol-ecules in relation to low molecular weight compounds regarding reactions and behavior. Staudinger had no doubts concerning primary bonds of the huge mole-cules. 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)

2.2 Concepts in Structural Research of Cellulose 15

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 (H

10C

6O

5) 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 pentag-onal 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 per-pendicular to the chain direction of the microfibrils are devoid of space for branch-ing. 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 specu-lated 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) pro-posed 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 crystallizationout 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

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)

2.3 Arrangements of the Cellulose Molecules in the Solid State 17

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

direction and upon crystallization antiparallel packing occurs. The plausible argu-ment 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 argu-mentation 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 intro-duction 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 polariza-tion 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.

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)

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 2

1 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-4C

1 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 essen-tial for such a description. The accepted 2

1 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 pro-posed (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

2.3 Arrangements of the Cellulose Molecules in the Solid State 19

20 2 History of Cellulose Research

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 monomericunit 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 cel-lulosic 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.

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