Cellulose Derivatives

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.fw001

August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.fw001
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 ACS SYMPOSIUM SERIES 688

Cellulose Derivatives
Modification, Characterization,
and Nanostructures
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.fw001

Thomas J. Heinze, EDITOR

Friedrich-Schiller-Universität Jena

Wolfgang G. Glasser, EDITOR

Virginia Polytechnic Institute and State University
August 31, 2012 | http://pubs.acs.org

Developed from a symposium sponsored by the Division

of Cellulose, Paper, and Textiles at the 212th National Meeting
of the American Chemical Society,
Orlando, Florida,
August 25-29, 1996

American Chemical Society, Washington, DC

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 QD 323 .C39 1998 Copy 1

Cellulose derivatives

Library of Congress Cataloging-in-Publication Data

Cellulose derivatives : modification, characterization, and nanostructures /

Thomas Heinze, editor, Wolfgang Glasser, editor.

p. cm.—(ACS symposium series, ISSN 0097-6156; 688)

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.fw001

"Developed from a symposium sponsored by the Division of Cellulose,

Paper, and Textiles at the 212th National Meeting of the American Chemical
Society, Orlando, Florida, August 25-29, 1996."

Includes bibliographical references and indexes.

ISBN 0-8412-3548-1

1. Cellulose—Congresses. 2. Cellulose—Derivatives—Congresses.

I. Heinze, Thomas, 1958- . Glasser, Wolfgang G., 1941- . III. American

Chemical Society. Cellulose, Paper, and Textiles Division. IV. American
th
Chemical Society. Meeting (212 : 1996 : Orlando, Fla.) V. Series.

QD323.C39 1998
661'.802—dc21 98-14355
August 31, 2012 | http://pubs.acs.org

CIP

This book is printed on acid-free, recycled paper.

Copyright © 1998 American Chemical Society

Distributed by Oxford University Press

All Rights Reserved. Reprographic copying beyond that permitted by Sections 107 or 108 of the U.S.
Copyright Act is allowed for internal use only, provided that a per-chapter fee of $20.00 plus $0.25 per
page is paid to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, M A 01923, USA.
Republication or reproduction for sale of pages in this book is permitted only under license from ACS.
Direct these and other permissions requests to ACS Copyright Office, Publications Division, 1155 16th
Street, N.W., Washington, DC 20036.

The citation of trade names and/or names of manufacturers in this publication is not to be construed as
an endorsement or as approval by ACS of the commercial products or services referenced herein; nor
should the mere reference herein to any drawing, specification, chemical process, or other data be
regarded as a license or as a conveyance of any right or permission to the holder, reader, or any other
person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted
work that may in any way be related thereto. Registered names, trademarks, etc., used in this
publication, even without specific indication thereof, are not to be considered unprotected by law.

PRINTED IN THE UNITED STATES OF AMERICA

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Advisory Board
ACS Symposium Series

Mary E . Castellion Omkaram Nalamasu

ChemEdit Company AT&T Bell Laboratories
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.fw001

Arthur B. Ellis Kinam Park

University of Wisconsin at Madison Purdue University

Jeffrey S. Gaffney Katherine R. Porter

Argonne National Laboratory Duke University

Gunda I. Georg
Douglas A. Smith
University of Kansas
The DAS Group, Inc.
Lawrence P. Klemann
Martin R. Tant
Nabisco Foods Group
Eastman Chemical Co.
Richard N. Loeppky
University of Missouri Michael D. Taylor
Parke-Davis Pharmaceutical
Research
August 31, 2012 | http://pubs.acs.org

Cynthia A. Maryanoff
R. W. Johnson Pharmaceutical
Research Institute Leroy B. Townsend
University of Michigan
Roger A. Minear
University of Illinois William C. Walker
at Urbana-Champaign DuPont Company

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Foreword

± H E ACS SYMPOSIUM SERIES was first published in 1974 to provide

a mechanism for publishing symposia quickly in book form. The pur-
pose of the series is to publish timely, comprehensive books devel-
oped from ACS sponsored symposia based on current scientific re-
search. Occasionally, books are developed from symposia sponsored
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.fw001

by other organizations when the topic is of keen interest to the chem-

istry audience.
Before agreeing to publish a book, the proposed table of contents
is reviewed for appropriate and comprehensive coverage and for in-
terest to the audience. Some papers may be excluded in order to better
focus the book; others may be added to provide comprehensiveness.
When appropriate, overview or introductory chapters are added.
Drafts of chapters are peer-reviewed prior to final acceptance or re-
jection, and manuscripts are prepared in camera-ready format.
As a rule, only original research papers and original review pa-
pers are included in the volumes. Verbatim reproductions of previ-
ously published papers are not accepted.
August 31, 2012 | http://pubs.acs.org

ACS BOOKS DEPARTMENT

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Preface

IMPRESSIVE ADVANCES in the ability to modify, regenerate, and reshape cellulose and
polysaccharide derivatives with unique chemical, physical, and physiological properties have
raised the interest in this most important biological macromolecule over the past decade. Cel-
lulose derivatives have received much attention from authors with diverse research, clinical,
and business interests; and this interest has created an opportunity for a broad display of topi-
cal discussions with varying degrees of technical depth. The launching of a new, nationwide
research program in Germany focusing on the design of molecular and supramolecular struc-
tures based on cellulose and cellulose derivatives, by the Deutsche Forschungsgemeinschaft
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.pr001

(DFG, the equivalent of the National Science Foundation in the United States), has attracted
scientists from a variety of disciplines.
This book was developed from a symposium titled "Recent Advances in Cellulose Modi-
fication", held at the 212th National Meeting of the American Chemical Society, in Orlando,
Florida, August 25-29, 1996. The symposium provided a forum for organizing an integrated
discussion of the current state of the art. It was organized with the intent of bringing together
scientists from academia and industry in the expectation that the new insights gained would be
useful for the development of novel, value-added materials from this polymer which is basic
to all plants.
Impulses for the new focus on cellulose derivatives originated from several sources:

• availability of new cellulose sources, especially bacterial cellulose

• new cellulose solvents and their corresponding regenerated fibers
• new regioselective modification methods
August 31, 2012 | http://pubs.acs.org

• new insights into the anisotropic solution states of (especially lyotropic liquid-crystalline)
cellulose derivatives
• a new understanding of the enzyme systems involved in cellulose degradation
• new, chirally active cellulose-based separation materials
• new, highly ordered thin film architectures of cellulose derivatives prepared by the Lang-
muir-Blodgett technique

This book highlights advances in (1) both homogeneous and heterogeneous phase modi-
fication of cellulose to create unusual derivatives, often with regioselective substitution pat-
terns, (2) analysis of selectively and specifically modified derivatives, (3) issues such as the
self-assembly of cellulosic macromolecules in dilute and concentrated solutions as well as in
solids, and (4) supramolecular architectures potentially useful in novel sensors, immunoas-
says, membranes, and biocompatibilized surfaces.
The first section of this book, "Modification Chemistry", shows that recent research on
chemical conversion of cellulose is mainly directed toward the synthesis of functionalized de-
rivatives with well-defined primary structures, both within the anhydroglucose repeat unit and
along the polymer chain. Moreover, nonconventional functional groups with special properties
are playing a dominant role in advanced cellulosics. In order to design new polymers based on
cellulose, special synthesis concepts are introduced to create reactive microstructures by, for

xi

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 example, induced phase separation. Nucleophilic displacement reactions and selective cellu-
lose oxidations are discussed as well. Examples are given for the advanced manufacture of
conventional esters and ethers.
A fast and reliable supply of comprehensive analytical data is an indispensable prerequi-
site for considering and pursuing new routes of synthesis and for controlling chemical proc-
esses in cellulose functionalization. In the section "Chemical and Molecular Structure", recent
results in adapted instrumental techniques are presented. In particular, NMR, Fourier-
transform infrared, and chromatographic techniques are providing new insights into molecular
structures and intermolecular interactions. The determination of molecular weights and mo-
lecular-weight distributions, and their changes during chemical, physical, and enzymatic
modifications, is discussed as well.
Based on the synthesis and structure characterization of fiinctionalized cellulose deriva-
tives, the processing of materials with defined supramolecular architectures is one of the most
important areas of cellulose research. The present state of the art in the design of supra-
molecular structures in the liquid and solid state, and the search for promising applications of
these structures, is summarized in the section "Supramolecular Structures". The successful
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.pr001

engineering of nanostructures and defined colloids yields insight into self-organization princi-
ples. The study of mesophase formation of cellulose derivatives with a defined primary struc-
ture provides a better understanding of liquid-crystalline systems. Tailored cellulosic com-
pounds are employed to design ordered supramolecular structures that can find application as
sensors, light-wave conductors, and selective membranes.
We express our appreciation to the American Chemical Society's Cellulose, Paper,
and Textile Division for sponsoring the symposium. The editors are indebted to their respec-
tive institutions, the Friedrich Schiller University of Jena, Germany and the Virginia Poly-
technic Institute and State University, Blacksburg, VA, for financial and logistic support of
this endeavor. We also thank Mark Fitzgerald and David Orloff of the American Chemical
Society Books Department, and Mary Holliman of Pocahontas Press, Blacksburg, VA for their
conscientious efforts to ensure timely review and completion of the book.

THOMAS HEINZE
August 31, 2012 | http://pubs.acs.org

Institut für Organische Chemie und

Makromolekulare Chemie der
Friedrich Schiller-Universität
Humboldtstrasse 10
D-07743 Jena, Germany

WOLFGANG G. GLASSER
Department of Wood Science and Forest Products
Virginia Polytechnic Institute and State University
Blacksburg, V A 24061-0324

xii

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 1

The Role of Novel Solvents and Solution Complexes

for the Preparation of Highly Engineered Cellulose
Derivatives

1 2
Thomas Heinze and Wolfgang G. Glasser

1
Institute of Organic Chemistry and Macromolecular Chemistry, University of Jena,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Humboldstrasse 10, D-07743 Jena, Germany

2
Biobased Materials/Recycling Center, and Department of Wood Science and Forest
Products, Virginia Polytechnic Institute and State University,
Blacksburg, VA 24061

Novel solvent systems for cellulose are reviewed in terms of their

potential for supporting chemical modification reactions that lead to
highly engineered derivatives. Systems capable of sustaining
homogeneous phase conditions include (1) dissolution with non-
-derivatizing solvents; (2) dissolution with partial derivatization with
reactive solvents; (3) dissolution by derivatization with protective
substituents; and (4) dissolution by derivatization with solubilizing
August 31, 2012 | http://pubs.acs.org

substituents that provide access to subsequent replacement reactions

(ie., leaving groups). Numerous examples and applications are cited
for each type of homogeneous phase reaction system.

Chemical modification continues to provide a dominant route towards cellulose

utilization in polymeric materials. The discovery of novel solvents and solution
complexes for cellulose in the past three decades has created opportunities for the
application of significantly more diverse synthesis pathways and derivative types.
Although fiinctionalization opportunities with cellulose are limited to OH-groups,
significant distinctions exist in different solvent systems between the reactivities of
primary and secondary OH groups, and between two different secondary OH-
groups located on the anhydroglucose repeat unit (AGU). The control of
substituent distribution within the AGU and along the polymer chain presents a
relatively recent challenge to the polysaccharide chemist. Homogeneous phase
chemistry has opened doors for the derivatization with more than one functional
group as well as for both partial and complete fiinctionalization of OH-groups with
such functionalities as chromophores and fluorophores, redox-active groups as well
as substituents with special magnetic, optical, and biological activity that are
indispensable prerequisites for the design of highly-engineered materials as well as

2 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 3

"smart materials." Access to novel reaction methodologies requires the support

from analytical tools qualified to describe quantitatively chemical and molecular
structures in a cost effective and timely manner. This support has become available
l 13
through product-adapted H-, C NMR- and two-dimensional NMR-
spectroscopy, either applied to the parent polymer (1-5) or to the partially or
completely depolymerized molecule (6), and through FAB and MALDI-TOF-
mass spectroscopy (7,8), as well as through HPLC following hydrolytic chain
degradation (9,10).
Highly-engineered cellulose derivatives, with carefully selected sites of
functionalization within the AGU and with controlled pattern of derivatization,
offer the potential for materials with specific solubility, enzyme degradability ( Π ­
Ι 5) (carboxymethyl and methyl cellulose), and blend compatibility
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

(cellulose/poly(ethylene oxide)) (16). Physiological activities have recently been

found to be significantly related to the distribution of sulfuric half ester functions
within the AGU, which seem to influence the interaction with human blood (17).
In addition, the potential to self-organized in liquid crystalline and monomolecular
ultra thin film structures also appears to be influenced by the specific design of
cellulose derivatives (18-21).
Highly engineered cellulose derivatives by homogeneous phase modification
can, in general, be accessed via the following four strategies:

• Dissolution with non-derivatizing solvents;

• Dissolution with partial functionalization, i.e., by the application of
derivatizing solvents;
• Dissolution by derivatization, with the introduction of specific solubilizing
August 31, 2012 | http://pubs.acs.org

substituents that protect specific OH functional groups during subsequent

reactions; and
• Dissolution by chemical modification with solubilizing substituents that
provide access to subsequent replacement reactions (i.e., leaving groups).

The intent of this review is to highlight examples of recent research which

illustrate the variety of the different synthesis chemistries, and to focus on the
opportunities and limitations they present. Comprehensive reviews of the field
have recently appeared elsewhere (22-25).

1. Dissolution With Non-derivatizing Solvents

Cellulose dissolves in a wide variety of non-aqueous solvents and solvent

complexes (Table I). These may involve single or multiple components. Although
the list of potential solvent systems for cellulose is impressive, indeed (Table I),
only a few systems have demonstrated the capability for supporting controlled and
homogenous chemical modification. Among the limitations are high toxicity, high

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Table I. Typical Non-Aqueous Cellulose Solvents.*

Number of components Solvent group Examples

Single (solvent) N-alkylpyridinium halogenide Ethylpyridinium chloride

Oxide of tertiary amine Triethylamine-N-oxide
N-methylpiperidine-N-oxide
N-methylmorpholine-N-oxide
N,N-dimethylcyclo-hexylamine-N-oxide
Multiple (solvent complex) Dimethyl sulfoxide (DMSO)-containing solvents DMSO/methyl amine
DMSO/KSCN
DMSO/CaCl 2

Liquid ammonia/sodium or ammonium salts NH3/NaI (NH I) 4

NH /NaSCN (NH SCN)

3 4

NH /NaN0
3 3

NH /N(C H ) Br
3 3 5 4

Dipolar aprotic sol vents/Li CI Ν,Ν-Dimethyl acetamide

(DMA)/LiCl
N-Methylpyrrolidone/LiCl
N-Methylcaprolactam/LiCl
N-Hexamethyl posphoric acid
triamide/LiCl
Dimethyl urea/LiCl

In Cellulose Derivatives; Heinze, T., et al.;

Pyridine or quinoline containing systems Pyridine/resorcinol
Quinoline/Ca(SCN) 2

Liquid S0 /secondary or tertiary amines

3 S0 /triethyl amine
3

NH or amine/salt/polar solvent
3 NH /NaCl/morpholine
3

NH /NaCl/DMSO
3

Ethylene diamine/NaI/N,N-dimethyl

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

formamide
NH, or amine/SO, or SOCl^/polar solvent diethylamine/SO^/DMSO
*A preactivation of the cellulose is required in most cases.
 5

reactivity of the solvent leading to undesired side reactions, and loss of solubility
resulting in non-homogeneous reaction conditions. Two solvent systems have
found considerable interest for cellulose modification; these are N,N-
dimethylacetamide (DMAc) in combination with lithium chloride (LiCl) and
dimethyl sulfoxide (DMSO) in combination with S0 and diethylamine.
2

The synthesis of cellulose esters, carbamates, and lactones in the

DMAc/LiCl solvent system yields derivatives with high purity and high
uniformity, in high yields, at moderate temperatures and with modest reagent
concentrations (26,28). Prominent known side reactions involve derivatives which
are susceptible to nucleophilic attack in the reaction system. Examples (Scheme 1)
include the formation of chlorodeoxy celluloses during the reaction with N-
chlorosuccinimide-triphenylphosphine (29), and the homogeneous bromination of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

cellulose with N-bromosuccinimide-triphenylphosphine in DMAc/LiBr (30).

However, relatively pure cellulose-p-toluenesulfonates (la) have been prepared
with a DS as high as 2.3 with low chlorine content (less than 0.5%) (31).
Tosylation proved more effective at 0-6 than at 0-2,3. Cellulosemethylsulfonates
have also been reported (32). The DMAc/LiCl solvent system also has supported
the derivatization of cellulose with the fluorophore 5-dimethylamino-l-
naphthalenesulfonate (lb) which allows studying cellulose solution properties by
means of absorption and fluorescence spectroscopy (33).
The synthesis of uniform unconventional cellulose esters, having fatty and
waxy ester substituents as large as C-20 (eicosanic acid, lc), has been supported by
the DMAc/LiCl solvent system in combination with acid anhydrides in the
presence of Ν,Ν-dicyclohexylcarbodiimide and 4-pyrrolidinopyridine and/or p-
toluenesulfonic acid (34-36). The homogeneous phase carbanilation of cellulose in
August 31, 2012 | http://pubs.acs.org

DMAc/LiCl with phenylisocyanate produces a completely substituted derivative

which has been widely adopted for the determination of molecular weights in
nonaqueous solvents by means of gel permeation chromatography (GPC) (37).
The synthesis of cellulose ethers in DMAc/LiCl has been accomplished
using triethylamine and pyridine as catalysts (38). Films of cellulose 4,4'-
bis(dimethylamino) diphenylmethyl ether (Id) cast from DMF-solution showed
photo conducting behavior (39). Although this example provides testimony to the
ability of DMAc/LiCl to support cellulose etherification reactions, the lack of
solubility of strong bases in this solvent system presents a significant limitation to
most typical ether-forming reactions. There has not appeared to exist a particular
advantage of this homogeneous phase reaction system over conventional
heterogeneous processes for cellulose ether production (40). However, comparing
conventionally prepared carboxymethyl cellulose derivatives (CMC) with
corresponding derivatives synthesized in a suspension of solid NaOH in
DMAc/LiCl, it was revealed (by HPLC analysis), that the substituent pattern may
vary dramatically in relation to reaction conditions (10,41). The CMC prepared in
the presence of DMAc/LiCl contained a significantly higher amount of both

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 6

CH:, NiCH?r
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001
August 31, 2012 | http://pubs.acs.org

Scheme I Representative monomelic units of: cellulose-/?-toluenesulfonate la,

cellulose-5-dimethylamino-l-naphthalenesulfonate lb, cellulose
,
eicosate lc, 4,4-bis(dimethylamino)diphenylmethyl cellulose Id,
tri-O-naphthylmethyl cellulose 13, triphenylcarbinol containing
cellulose If

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 7

tricarboxymethylated and unsubstituted units than those derivatives obtained by

the conventional slurry method of cellulose in isopropanol/water. The DMAc/LiCl
solvent system seems to give rise to a non-statistical distribution of monomeric
units, and this is responsible for several unconventional properties. It is suggested
that reactions using the induced phase separation method are not limited to CMC
synthesis, and that this may represent a new synthesis strategy for cellulose esters
with unconventional distribution of functional groups.
A more universally adaptable homogeneous phase cellulose etherification
system involves the DMSO/S0 /diethylamine solvent. The conversion of cellulose
2

dissolved in this system with suspended sodium hydroxide and benzyl chloride has
been studied extensively by Isogai et al (42). This system revealed superiority by
comparison to the solvents N2O4/DMF and DMAc/LiCl in terms of reaction rates
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

and yields. Among the derivatives reported were tri-O-arylmethylethers containing

double bonds (43) and tri-O-a-naphthylmethyl ether (le) which exhibits liquid
crystalline behavior (44). A cellulose ether derivative containing a triphenyl moiety
(If) assumed conformations which were subject to photoregulation (45).

2. Dissolution With Derivatizing Solvents (Reactive Intermediates)

Effective cellulose dissolution requires the complete disruption of all

cellulose to cellulose hydrogen bonds during solvation. Few solvent systems exist
which accomplish this task without at least partial polymer derivatization.
Examples of these derivatizing solvents, their reactive intermediates and typical
reactions for which these intermediates are used, are given in Table Π. Although
cellulose dissolution by the formation of reactive (soluble) intermediates by partial
August 31, 2012 | http://pubs.acs.org

derivatization is an acceptable method for cellulose regeneration, the use of these

solvent systems for chemical modification has remained limited. Unreproducibility
has been attributed to the undefined structure of the reactive intermediates, to the
limited stability of the intermediates during rapidly-changing reaction conditions,
and to the general dynamics of the complexation with the solubilizing derivatizing
group during the process of modification. Among the reactions listed in Table Π,
the Ν,Ν-dimethylformamide (DMF)/N 0 solvent system yielding cellulose nitrite
2 4

as intermediate has found considerable interest in the synthesis of inorganic

cellulose esters despite its highly toxic nature. Various cellulose sulfates and
phosphates have been prepared by this method (46-48).
The derivatization of cellulose via the formation of reactive intermediates
proceeds with greater predictability if the respective intermediate is isolated (and
characterized) following formation and prior to conversion into a final cellulose
derivative product. Reactive intermediates are typically soluble in common organic
solvents, and they can be the starting materials for a wide variety of highly
engineered cellulose derivatives. Among popular reactive intermediates are the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

QO

Table II. Examples of Cellulose Dissolution with Partial Derivatization; their Reactive Intermediates
Formed; and Typical Subsequent Conversions in situ.

Dissolving agent Reactive cellulose Subsequent reactions Ref.

intermediate in situ

Formic acid Formate

Trifluoroacetic acid Trifluoroacetate Acylation 50
Trichloroacetic acid/polar solvent Trichloroacetate
Ν,Ν-dimethyl formamide (DMF)/N 0> 2 4 Nitrite Sulfation, Phosphation 94
Paraformaldehyde/DMSO Methylol Etherification 95
Silylation 96
Trimethylchlorsilane/DMF Trimethylsilyl
Chloral/DMSO Trichloroacetal Chlorination 97

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 9

formate, the trifluoroacetate, and the trimethylsilyl ether (Table III). These reactive
cellulose intermediates vary widely in DS as well as in derivative stability.
Trifluoroacetylation takes place preferably at C-6 (49,50). The derivatives
are conveniently characterized regarding their structures following permethylation,
13
saponification and degradation by means of C-NMR and HPLC analysis (51).
The dissolution involves a mixture of trifluoroacetic acid and trifluoroacetic
anhydride at room temperature for 4 h, and this results in DS-values of circa 1.5
with complete substitution at C-6. Higher DS-values are obtained by the addition
of chlorinated hydrocarbons as co-solvents during trifluoroacetylation. Owing to
its high C-6-selectivity, the DS 1.5-trifluoroacetate of cellulose has been used as
starting material for a wide variety of esters involving subsequent esterification with
acid chlorides, mixed anhydrides of p-toluenesulfonic and carboxylic acids, or
carboxylic acids in combination with N-N-carbonyldiimidazole (52) (Table ΙΠ).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

(Trifluoroacetates do not transesterify under these conditions as was determined by

the reaction of pure 2,3-di-0-methyl-6-0-trifluoroacetyl cellulose (2a) in aprotic
medium with acetic acid/acetic anhydride in the presence of pyridine at 80°C with
subsequent analysis by ^^H-COSY-NMR spectroscopy (Figure 1)). A
remarkable 0-3 selectivity was established for the sulfation of cellulose
trifluoroacetate with pyridine/S0 (53). The primary trifluoroacetate substituent is
3

easily removed in an aqueous work-up procedure. In addition to the use of

cellulose formates for cellulose regeneration (54), these DMF-soluble derivatives
: 13
were isolated with DS values of up to 1.2 (55). C-NMR spectroscopy revealed
that the ease of formylation is in the order of 0-6 > than 0-2 > than 0-3 (56,57).
Sulfation of the reactive formate intermediate was found to occur with partial
substituent removal (transesterification) as well as at the free OH-groups (58).
August 31, 2012 | http://pubs.acs.org

Since transesterification can be avoided by choice of reaction conditions, (52) an

inverse pattern of substituention can be obtained after subsequent removal of the
formate substituents by treatment with water. This route, however, is limited by
the high DS-value of the starting formate. Both trifluoroacetylation and
formylation are highly degradative reactions which often produce cellulose
derivatives of low DP. Depolymerization can be avoided by silylation.
Trimethylsilyl (TMS) celluloses werefirstdescribed by Schuyten (59), and
they were extensively studied for their regeneration potential by simple treatment
with acids (60). A wide variety of silylated cellulose derivatives with a broad range
of DS-values has been reported (61-63). TMS cellulose has shown to react either
with its free OH-groups or with its trimethylsiloxy groups during treatment with
acid chlorides, depending on reaction conditions (64,65). Whereas tertiary amines
were found to catalyze the reaction of free OH-groups with acid chlorides,
nitrobenzene at 160°C caused reaction only at the TMS functionalities (63). This,
however, is not universally true since a partially silylated TMS-polyvinylalcohol
has recently been shown to produce esters at both the TMS and OH-fiinctional
groups by acylation in nitrobenzene at 160°C (66).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Table III. Examples of Subsequent Reactions on Isolated Cellulose Intermediates.

Reactive cellulose Reagents Cellulose derivative produced Ref.

intermediate*
0
Trifluoroacetate (1.5) Pyridine/S0 3 Cellulose sulfate 53
c
N,N-carbonyldiimidazole/4- Cellulose-4-nitrobenzoate 52
nitrobenzoic acid
d
4-Nitrobenzoic acid/tosyl Cellulose-4-nitrobenzoate 52
chloride/
d
Palmitoyl chloride/tosyl chloride Cellulose palmitate 52
d
Phenyl isocyanate Cellulose phenylcarbamate 52
0
Na-monochloroacetate/NaOH Carboxymethyl cellulose 67
1
Formate (2.2) Pyridine/S0 3 Cellulose sulfate* 52
0
Ν,Ν-dimethyl formamide/S0 3 Cellulose sulfate 58
d
Phenyl isocyanate Cellulose phenylcarbamate 52
0
Na-monochloroacetate/NaOH Carboxymethyl cellulose 67
d
Trimethylsilyl cellulose ( 1.6) 3,4-Dinitrobenzoyl chloride (in Cellulose-3,4-dinitrobenzoate 62
triethylamine)
d
Trimethylsilyl cellulose (2.0) 4-Bromobenzoyl chloride/4- Cellulose-4-bromobenzoate 62
dimethylamino pyridine (in

In Cellulose Derivatives; Heinze, T., et al.;

benzene/triethyl amine)
Trimethylsilyl cellulose (2.5) 4-Nitrobenzoyl chloride Cellulose-4-nitrobenzoate 63
0
Trimethylsilyl cellulose (1.1) Na-monochloroacetate/NaOH Carboxymethyl cellulose 68
"Degree of substitution given in ( ).
b
0-3 substitution
°Partial or total removal of functional groups of the reactive intermediate during the reaction.

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

d
Inverse pattern of functionalization.
 11

XCCF3

2a OCH3

Ν
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Η-2

3.0
(y

3.5
I 3 00
August 31, 2012 | http://pubs.acs.org

β / 4.0

04
ζ * _

ό/ppm
ό/ppm US 40 3.5 3,0
'H-'H-COSY-NMR spectrum of 2,3-di-0-methyl-6-O-
trifluoroacetyl cellulose 2a after treatment with acetic acid/acetic
anhydride in the presence of pyridine for 8 h at 80°C proving that
neither transesterification nor split off of trifluoroacetyl functions
occurred.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 12

The soluble reactive intermediates of Table II, and even cellulose acetate, can
be converted into cellulose ethers with high DS by removing the solubilizing ester
groups in a one step synthesis in DMSO with suspended solid NaOH powder as
base (67). This pseudo-homogeneous conversion proceeds to cellulose ethers that
have a substitution pattern that is differentfromthat of a heterogeneously prepared
products (68).

3. Dissolution by Derivatization with Protective Groups

Two major groups of protective agents have been of interest to cellulose

chemists. The triphenylmethyl (trityl) group is a widely used protective group for
the C-6 position which provides for subsequent selective functionalization of the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

remaining secondary OH-groups (69-71). The reaction of cellulose with methoxy-

substituted triphenyl chlorides in homogeneous DMAc/LiCl solution proceeds with
little degradation and high reproducibility (72). The monomethoxy-trityl (3a), in
particular, combines a fast and specific blocking step of well-soluble polymers with
an easy subsequent deblocking reaction. This pathway has recently been employed
for the preparation of 2,3-di-O-carboxymethyl cellulose (73), among many others
(74-77).
Another OH-protective agent with great regio-selectivity is 6-0-
thexyldimethylchlorosilane (TDMS). 6-0-TDMS-celluloses (3b) were synthesized
using a heterogeneous phase reaction in the presence of ammonia-saturated polar-
aprotic solvents at -15°C. The conversion of cellulose with TDMS chloride in N-
methylpyrrolidone/NH was found not to proceed past a total D S of 1.0; it
3 a

yielded a derivative with 96% silylation at 0-6 (78). This well-protected cellulose
August 31, 2012 | http://pubs.acs.org

derivative is a useful starting material for the subsequent synthesis of various regio-
specifically modified cellulose materials (79). This field has been subjected to
several recent reviews (12,25).

4. Dissolution by Derivatization with Leaving Groups

Homogeneous phase reaction conditions with cellulose provide convenient

access to derivatives which may be subject to subsequent substitution by
nucleophiles. The chemical activation of cellulose proceeds at low temperatures
and short reaction times, and it is virtually free of side-reactions and associated
impurities. The active groups of particular interest for subsequent nucleophilic
replacement are sulfonates (especially tosylate) and halodeoxy cellulose derivatives.
In contrast to heterogeneous sulfonation, which is subject to numerous
potential side reactions, homogeneous mesylation (32) and tosylation (80,81) in
DMAc/LiCl produces uniform and well-defined products with DS rangingfrom0.4
to 2.3 (31). The derivatives are typically soluble in a variety of organic solvents
and contain only traces of halogen and no nitrogen. The use of cellulose sulfonates

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 13

.0...

3b
3a
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Scheme 2. Representative monomelic units of monomethoxytrityl cellulose (3a)

and 6-0-texyl dimethylsilyl cellulose (36).

for a variety of nucleophylic substitutions has been summarized by Belyakova (82)

and Hon (83). In contrast to cellulose triflates, which are highly reactive and
sensitive to even mildly nucleophylic species, cellulose tosylates are rather stable.
Various novel cellulose derivatives were synthesized from cellulose tosylates of
varying DS-levels and molecular weights by both acylation (84) and sulfation of the
unsubstituted OH-groups. The amphiphilic cellulose sulfate tosylates are soluble
in both water and DMSO at an appropriate DS-balance, and they are promising
August 31, 2012 | http://pubs.acs.org

starting materials for self-organizing polymer systems (85).

Chlorodeoxy cellulose can be prepared heterogeneously (in an inert medium
such DMF or pyridine/chloroform with thinoyl chloride, phosphorus oxychloride,
sulfuryl chloride or methanesulfonyl chloride), or homogeneously in DMF/chloral
(29,86). More elegantly, cellulose chlorination may also proceed in DMAc/LiCl
with sulfuryl chloride (87) or tosyl chloride via the in-situ formed cellulose
sulfonates (88). A convenient and relatively selective chlorination of cellulose in
DMAc/LiCl involves N-chlorosuccinimide-triphenylphosphine. While the reaction
initially derivatizes only the C-6 position, DS-levels as high as 1.86 have been
reported (29). The corresponding N-bromosuccinimide-triphenylphosphine gave
rise to bromodeoxy cellulose in three different solvents (DMF, N-
methylpyrrolidone, and DMAc), in combination with LiBr. The derivatives had a
maximum DS of 0.9 and were selectively brominated at C-6 (30). An alternative
bromination reagent consists of tribromoimidazole, triphenylphosphine and
imidazole (90). Homogeneous reactions in DMAc/LiBr produced polymers with
DS-values of up to 1.6. The halodeoxy celluloses are also accessible via
nucleophilic substitution of tosylate functions by halides (91). A large number of

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 14

follow-up replacement reactions using halodeoxy celluloses have been described,

and these include conversions with thio-urea, with iminodiacetic acid (92) or with
thiols (93).
Replacement reactions based on cellulose silylethers have been discussed in
a previous section.

CONCLUSION

With the advancement of solvents and solution complexes for cellulose,

modification reactions have become feasible that provide the tools necessary for
tailoring highly engineered cellulose derivatives to specific end-uses. The pathways
to selectively and specifically modified cellulose derivatives involve dissolution in
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

single component solvents; dissolution via reactive intermediate by partial

derivatization (i.e., derivatizing solvents); dissolution as protected cellulose
derivatives; and dissolution by chemical activation for subsequent substituent
replacement by nucleophilic attack. These systems provide a wide range of
opportunities for the preparation of highly-engineered cellulose derivatives, and
they make it possible that cellulose derivatives be used in such molecular
recognition concepts as self-organizing supra-molecular systems, in nano-
structures, in environmentally responsive (smart) materials, and in materials useful
in sensors as well as in chiral templates. Much future work is needed to explore
and exploit the full benefits of homogeneous phase reactions with polysaccharides.

Literature Cited
August 31, 2012 | http://pubs.acs.org

1. Nehls, I.; Wagenknecht, W.; Philipp, B.; Stscherbina, D. Progr. Polym.

Sci. 1994, 19, 29.
2. Buchanan, C.M.; Hyatt, J.Α.; Lowman, D.W. J. Am. Chem. Soc. 1989,
111, 7312.
3. Stein, Α.; Klemm, D. Papier (Darmstadt) 1995, 49, 732.
4. Tezuka, Y.; Tsuchiya, Y. Carbohydr. Res. 1995, 273, 83.
5. Tezuka, Y.; Tsuchiya, Y.; Shiomi, T. Carbohydr. Res. 1996, 291, 83.
6. Baar, A.; Kulicke, W.M.; Szablikowski, K.; Kiesewetter, R. Macromol.
Chem. Phys., 1994, 195, 1483.
7. Arisz, P.W.F. Ph. D. Thesis, University of Amsterdam, The Netherlands,
1995.
8. Mischnick, P.; Kühn, G. Carbohydr. Res. 1996, 290, 199.
9.
Erier, U.; Mischnick, P.; Stein, A ; Klemm, D. Polym. Bull. 1992,29, 349.
10. Heinze, Th.; Erler, U.; Nehls, I.; Klemm, D. Angew. Makromol.
Chem.1994, 215,93.
11. Kamide, K.; Saito, M. Macromol. Symp. 1994, 83, 233.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 15

12. Klemm, D.; Stein, Α.; Heinze, Th.; Philipp, Β.; Wagenknecht, W.
Polymeric Materials Encyclopedia: Synthesis, Properties and
Applications, Salamone, J.C. Ed.; CRC Press, Inc., Boca Raton, USA, vol.
2, 1996, p. 1043-1053.
13. Gelman, R.A. J. Appl Polym. Sci. 1982, 27, 2957.
14. Takahaishi, S.-I.; Fujimoto, T.; Miyamoto, T.; Inagaki, H. J. Polym. Sci.,
Part A: Polym. Chem. 1987, 25, 987.
15. Nojiri, M.; Kondo, T. Macromolecules 1996, 29, 2392.
16. Kondo, T.; Sawatari, C. Polymer 1994, 35, 4423.
17. Klemm, D.; Heinze, Th.; Wagenknecht, W. Ber. Bunsenges. Phys. Chem.
1996, 100, 730.
18. Guo, J.X.; Gray, D.G. Lyotropic Cellulosic Liquid Crystals, In: Cellulosic
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Polymers, Blends and Composites, Gilbert, R.D. Ed.; Hanser Publ.,

Munich, Vienna, New York, 1994, p. 25.
19. Zugenmaier, P. Polymer Sovent Interactions in Lyotropic Liquid Crystalline
Cellulose Derivative Systems, In: Cellulosic Polymers, Blends and
Composites, Gilbert, R.D. Ed.; Hanser Publ., Munich, Vienna, New
York, 1994, p. 71.
20. Schaub, M.; Fakirov, C.; Schmidt, Α.; Lieser, G.; Wenz, G.; Wegner, G.;
Albony, P.A.; Wu, H.; Foster, M.D.; Majrkzak, M.; Satija, S.
Macromolecules 1995, 28, 1221.
21. Wegner, G.; Schaub, M.; Wenz, G.; Stein, A.; Klemm, D. Adv. Mater.
1993, 5, 919.
22. Philipp, B. J.M.S.-Pure Appl. Chem. 1993, A30, 703.
23. Johnson, D.C. Solvents for Cellulose, In. Cellulose Chemistry and its
August 31, 2012 | http://pubs.acs.org

Applications, Nevell, T.P.; Zeronian, S.H. Eds.; E. Horwood Ltd.,

Chichester, 1985, p. 181.
24. Herlinger, H.; Hengstberger, M. Lenzinger Ber. 1985, 59, 96.
25. Philipp, B.; Wagenknecht, W.; Nehls, I.; Klemm, D.; Stein, Α.; Heinze, Th.
Polymer News 1996, 21, 155.
26. McCormick, C.L.; Lichatowich, D.K. J. Polym. Sci., Polym. Lett. Ed.
1919, 17, 479.
27. Dawsey, T.R. Applications and Limitations of LiCl/N,N­
-Dimethylacetamide in the Homogeneous Derivatization of Cellulose, In:
Cellulosic Polymers, Blends and Composites, Gilbert, R.D. Ed.; Hanser
Publ., Munich, Vienna, New York, 1994, p. 157.
28. Morgenstern, B.; Kammer, H.-W. TRIP 1996, 4, 87.
29. Furuhata, K.-I.; Chang, H.-S.; Aoki, N.; Sakamoto, M. Carbohydr. Res.
1992, 230, 151.
30. Furuhata, K.-I; Koganai, K.; Chang, H.-U.; Aoki, N.; Sakamoto, M.
Carbohydr. Res., 1992, 230, 165.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 16

31. Rahn, Κ.; Diamantoglou, M.; Klemm, D.; Berghmans, H.; Heinze, Th.
Angew. Makromol. Chem. 1996, 238, 143.
32. Frazier, C.E.; Glasser, W.G. Polymer Preprints 1990, 31, 634.
33. Heinze, Th.; Camacho Gomez, J.A.; Haucke, G. Polym. Bull. 1996, 37,
743.
34. Samaranayake, G.; Glasser, W.G. Carbohydr. Polym. 1993, 22, 1.
35. Glasser, W.G.; Samaranayake, G.; Dumay, M.; Dave, V. J. Polym. Sci.,
Part Β: Polym. Phys. 1995, 33, 2045.
36. Sealey, J. E.; Samaranayake, G.; Todd, J.G.; Glasser, W.G. J. Polym.
Sci., Part B: Polym. Phys. 1996, 34, 1613.
37. Terbojevich, M.; Cosani, Α.; Camilat, M.; Focher, B. J. Appl. Polym. Sci.
1995, 55, 1663.
38. Erler, U.; Klemm, D.; Nehls, I. Makromol. Chem., Rapid Commun. 1992,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

13, 195.
39. Heinze, Th.; Erler, U.; Heinze, U.; Camacho,J.;Grummt, U.-W.; Klemm,
D. Macromol. Chem. Phys. 1995, 196, 1937.
40. Dawsey, T.R. Polym. Fiber Sci.: Recent Adv. 1992, Forners, R.E.; Gilbert
R.D. Eds.; VCH New York, p. 157.
41. Heinze, Th.; Heinze, U.; Klemm, D. Angew. Makromol. Chem. 1994,
220, 123.
42. Isogai, Α.; Ishizu, Α.; Nakano, J. J. Appl. Polym. Sci 1984, 29, 2097,
3873.
43. Isogai, A.; Ishizu, A.; Nakano, J. J. Appl. Polym. Sci 1986, 31, 341.
44. Dave, V.; Frazier, C.L.; Glasser, W.G. J. Appl. Polym. Sci. 1993, 49,
1671.
August 31, 2012 | http://pubs.acs.org

45. Arai, K.; Kwabata, Y.Macromol. Chem. Phys. 1995, 196, 2139.
46. Akelah, Α.; Sherrington, D.C. J. Appl. Polym. Sci. 1981, 26, 3377.
47. Wagenknecht, W.; Nehls, I.; Philipp, B. Carbohydr. Res. 1992, 237, 211.
48. Wagenknecht, W.; Nehls, I.; Philipp, B. Carbohydr. Res. 1993, 240, 245.
49. Hasegawa, M.; Isogai, Α.; Onabe, F.; Usuda M., J. Appl. Polym. Sci
1992, 45, 1857.
50. Salin, B.N.; Cemeris, M.; Mironov, D.P.; Zatsepin, A.G. Khim. Drev.
1991, 3, 65, CA 116(8): 61812b.
51. Liebert, T.; Schnabelrauch, M.; Klemm, D.; Erler, U. Cellulose 1994, 1,
249.
52. Liebert, T. Ph. D. Thesis, University of Jena, Germany, 1995.
53. Klemm, D.; Heinze, Th.; Stein, Α.; Liebert, T. Macromol. Symp. 1995,
99, 129.
54. Rudy, H. Cellulosechemie 1931, 13, 49.
55. Schnabelrauch, M.; Vogt, S.; Klemm, D.; Nehls, I.; Philipp, B. Angew.
Makromol. Chem. 1992, 198, 155.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 17

56. Takahashi, S.I.; Fujimoto, T.; Barua, B.M.; Miyamoto, T.; Inagaki, H. J.
Polym. Sci., Polym. Chem.Ed.1986, 24, 2981.
57. Fujimoto, T.; Takahashi, S.I.; Tsuji, M.; Miyamoto, T.; Inagaki, H. J.
Polym. Sci., Polym. Lett. 1986, 24, 495.
58. Philipp, B.; Wagenknecht, W.; Nehls, I.; Ludwig, J.; Schnabelrauch, M.;
Rim, K.H.; Klemm, D. Cellul. Chem. Technol. 1990, 24, 667.
59. Schuyten, H.A.; Weaver, J.W.; Reid, J.D.; Jürgens, J.F. J. Am. Chem.
Soc. 1948, 70, 1919.
60. Weigel, P.; Gensrich, J.; Wagenknecht, W.; Klemm, D.; Erler, U.; Philipp,
B. Papier (Darmstadt) 1996, 50, 483.
61. Schemp, W.; Krause, Th.; Seifried, U.; Koura, A. Papier (Darmstadt)
1984, 38, 607.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

62. Klemm, D.; Schnabelrauch, M.; Stein, A.; Philipp, B.; Wagenknecht, W.;
Nehls, I. Papier (Darmstadt) 1990, 44, 624.
63. Stein, A.; Klemm, D. Makromol. Chem., Rapid Commun. 1988, 9, 569.
64. Wagenknecht, W.; Nehls, I.; Stein, A.; Klemm, D.; Philipp, B. Acta
Polymerica 1992, 43, 266.
65. Klemm, D.; Stein, A.; Erler, U.; Wagenknecht, W.; Nehls, I.; Philipp, B.
New Precedures for regioselective synthesis and modification of
trialkylsilylcelluloses, In: Cellulosics: Materials for Selective
Separation and Other Technologies, Kennedy, J.F.; Phillips, G.O.;
Williams P.A. Eds.; E. Horwood Ltd., New York, London, Toronto,
Sydney, Tokyo, Singapore 1993, p. 221.
66. Mormann, W.; Wagner, Th. Macromol. Chem. Phys. 1996, 197, 3463.
67. Liebert, T.; Klemm, D.; Heinze, Th. J.M.S.-Pure Appl. Chem. 1996, A33,
August 31, 2012 | http://pubs.acs.org

613.
68. Heinze, Th.; Liebert, T. ACS Symp. Ser. 1997, in press.
69. Green, J.W. Triphenylmethyl Ethers, In: Methods in Carbohydr. Chem.,
Whistler, R.L.; Green, J.W.; BeMiller, J.N. Eds.; vol 3, 1963, p. 327.
70. Harkness, B.R.; Gray, D . G Macromolecules 1991, 24, 1800.
71. Kondo, T.; Gray, D.G. Carbohydr. Res. 1991, 220, 173.
72. Camacho Gomez, J.A.; Klemm, D.; Erler, U. Macromol. Chem. Phys.
1996, 197, 953.
73. Heinze, Th.; Röttig, K.; Nehls, I. Macromol. Rapid Commun. 1994, 15,
311.
74. Iwata, T.; Azuma, J.-I.; Okamura, K.; Muramoto, M. B. Chun, Carbohydr.
Res. 1992, 224, 277.
75. Kondo, T. Carbohydr. Res. 1993, 238, 231.
76. Itagaki, H.; Takahashi, I.; Natsume, M.; Kondo, T. Polym. Bull. 1994, 32,
77.
11. Kasuya, N.; Iiyama, K.; Meshituska, G.; Ishizu, A. Carbohydr. Res. 1994,
260, 251.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 18

78. Klemm, D.; Stein, A. J.M.S.-Pure Appl. Chem. 1995, A32, 899.
79. Koschella, Α.; Klemm, D. Macromol. Symp. 1997, in press.
80. Dawsey, T.R.; Newman, J.K.; McCormick, C.L. Polym. Prepr. (Am.
Chem. Soc., Div. Polym. Chem.) 1989, 30, 191.
81. McCormick, C.L.; Dawsey, T.R.; Newman, J.K. Carbohydr. Res. 1990,
208, 183.
82. Belyakova, M.K.; Gal'braikh, L.S.; Rogovin, Z.A. Cellul. Chem. Technol.
1971, 5, 405.
83. Hon, D. N.-S. Chemical Modification of Cellulose, In: Chemical
Modification of Lignocellulosic Materials, Hon D. N.-S. Ed.; Marcel
Dekker, New York, Basel, Hong Kong, 1996, p. 114.
84. Heinze, Th.; Rahn, K.; Jaspers, M.; Berghmans, H. Macromol. Chem.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch001

Phys. 1996, 197, 4207.

85. Heinze, Th.; Rahn, K. Macromol. Rapid Commun. 1996, 17, 675.
86. Schnabelrauch, M.; Heinze, T.; Klemm, D. Acta Polymerica 1990, 41, 113.
87. Furubeppu, S.; Kondo, T.; Ishizu, A. Sen'i Gakkaishi 1991, 47, 592.
88. McCormick, C. L.; Callais, P. Polymer 1987, 28, 2317.
89. Furuhata, K.-I.; Aoki, N.; Suzuki, S.; Arai, N.; Sakamoto, M.; Saegusa, Y.;
Nakamura, S. Carbohydr. Res. 1994, 258, 169.
90. Furuhata, K.-I.; Aoki, N.; Suzuki, S.; Sakamoto, M.; Saegusa, Y.;
Nakamura, S. Carbohydr. Polym. 1995, 26, 25.
91. Heinze, Th.; Rahn, K. Papier (Darmstadt) 1996, 50, 721.
92. Mentasti, E.; Sarzanini, C.; Gennora, M.C.; Porta, V. Polyhedron 1987, 6,
1197.
93. Aoki, N.; Koganei, K.; Chang, H-S.; Furuhata, K.; Sakamoto, M .
August 31, 2012 | http://pubs.acs.org

Carbohydr. Polym. 1995, 27, 13.

94. Philipp, B. Polymer News 1990, 15, 170.
95. Nicholson, M.D.; Johnson, D.C. Cellul. Chem. Technol. 1977, 11, 349.
96. Shiraishi, N.; Miyagi, Y. Sen-i Gakkaishi 1979, 35, 466.
97. Ishii, T.; Ishizu, Α.; Nakano, J. Carbohydr. Res. 1977, 59, 155.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 2

Regiocontrol in Cellulose Chemistry: Principles and

Examples of Etherification and Esterification

D. O. Klemm

Institute of Organic and Macromolecular Chemistry,

Friedrich Schiller University, D-07743 Jena, Germany
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

Regiocontrol means site-selective reactions of cellulose and related

polysaccharides that form functionalization patterns essential for
new properties. Experimental results are presented in synthesis,
analysis, and subsequent reactions of silylethers, cellulose p­
-toluenesulfonates, and deoxy-thiosulfates (Bunte salts). Regiocon­
trol succeeds by selective loosening supramolecular structures, as
well as by selective protection, activation, and migration of
functional groups.

The topics of this paper relate to the essential interaction of cellulose

August 31, 2012 | http://pubs.acs.org

functionalization with organic synthesis and structure analysis, as well as with

product properties and applications. After some remarks on principles of regio-
selectivity in cellulose modification, the selective synthesis and subsequent
reactions of cellulose ethers, the functionalization via ester intermediates of
controlled stability and reactivity, and examples of structure properties relation­
ships will be described.

General Principles of Regioselective Cellulose Chemistry

Up to now there have been known four principal synthesis pathways to obtain
functionalized celluloses (Figure 1): The first is the well-known polymeranalogous
reaction of cellulose after isolation from plants or bacteria culture media. The
second is the biosynthesis of functionalized celluloses using, for example, the
copolymerization of β-D-glucose with N-acetylglucoseamine by Acetobacter
xylinum (/). The third way is enzymatic in vitro synthesis starting, for example,
from 6-0-methyl-B-cellobiosyl fluoride and purified cellulases in a stereo- and
regioselective polymerization (2). The fourth is the chemical synthesis starting
from glucose. Results from last year demonstrated the first chemosyntheses of
functionalized celluloses by ring-opening polymerization of 3,6-di-O-benzyl-a-D-

©1998 American Chemical Society 19

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 20

3 OH
Cellulose

Polymeranalogous reactions
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

Biosynthesis Chemosynthesis
S. TOKURA F. N A K A T S U B O
M. TAKAI T. NISHIMURA

Acetobacter xylinum
Ph C©BF ©
3 4

R 2
- NHCOCH3 0 - .

Functionalized celluloses
OH R - OH, R \ R 2

Ο
HO BnO ^ O B n
.OH
HO .0
August 31, 2012 | http://pubs.acs.org

OH
+ Cellulase R 1
- OCH 3

OH

HO OH
HO
Enzymatic synthesis
Glucose S .KOBAYASHI Glucose
derivatives derivatives

,R
1

HO
HO
OH \ 0

Cellobiose derivatives

Figure 1. Principal synthesis pathways to obtain functionalized celluloses.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 21

glucose- 1,2,4-orthopivalate (3), and by stepwise reaction of selectively protected

allylethers of β-D-glucose (4).
In the case of polymeranalogous functionalization of cellulose, the chemical
reaction proceeds in a statistic or regioselective way. The term "regioselectivity''
means an exclusively or significantly preferential reaction at one or two of the
three sites 2,3, and 6 of the anhydroglucose unit (AGU) as well as along the
polymer chains. Typically simple examples are selectively C-6 or C-2,3 modified
celluloses (A) and copolymers with block-like structures (B) as shown schema­
tically in Figure 2. The symbols used ( · , * ) denote different types of functional
groups.
The energy profile of a typical regioselective reaction demonstrates that steric
hindrance (by bulky groups), entropie acceleration (caused, for example, by H O —
H bonds), or electronic promotion (e.g., electronically withdrawn substituents) are
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

essential to differentiate among the free enthalpies of activation as well as of rate

constants and concentrations of the isomeric products formed. From the point of
view of such differences, the varying reactivity of the O H groups in the A G U of
cellulose - primary 6-OH, more acidic 2-OH - may be used for selective
functionalization of cellulose. Moreover, a selective loosening of the supra-
molecular structure may be of importance in regiocontrol.
4
In regard to the question: 'Why regiocontrol in cellulose chemistry?" three
important fields should be pointed out:
- Basic investigations, e.g., on structure and interaction in solution, as well as on
formation of well-defined supramolecular structures using cellulose derivatives
with known patterns of functionalization;
- Design of advanced materials and nanoscale architectures in interdisciplinary
research at the interface of organic and macromolecular chemistry. Potential fields
August 31, 2012 | http://pubs.acs.org

of application are liquid crystalline polymers, selective membranes, multilayered

assemblies, sensor matrices, recognition devices, and bioactive materials;
- Better knowledge of reactions, mechanisms, and product structures and properties
of present as well as of new commercial types of cellulosics of industrial large-
scale processes. Further knowledge of the control of processing and end-use
properties by the pattern of functionalization.
For example, regioselectively modified celluloses are of importance in investi­
gations into solution properties as well as into enzymatic degradation of cellulose
derivatives. This work deals with fringed micelles at different aggregation stages in
solution as demonstrated by Burchard (5), with the characterization of the reaction
behaviour of cellulases using 6-0- and 2,3-di-O-methyl celluloses (6), and with the
investigation of HO-—H bond systems and gelation of cellulose derivatives (7, 8)
as published by Kondo.
Our investigations on regiocontrol in polysaccharide chemistry have been based
up to now on cellulose and starch including β-cyclodextrine as a model compound.
Regioselectivity in the A G U and along the polymer chains may be low, high, or
100 % and may lead to copolymers (degree of substitution, DS, < 1) or homo-
polymers (DS = 1) by reaction at one preferred site of the A G U . Regiocontrol of
the primary and of subsequent reactions succeeds by the reagent (e.g., steric

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 22
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002
August 31, 2012 | http://pubs.acs.org

Figure 2. Examples of regioselectively functionalized celluloses.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 23

hindrance) and by substituent effects (protection, activation, migration) and by the

supramolecular structure in the reaction media (stepwise loosening of HO—H
bonds, phase separation during reaction, structure in solution). Preferentially
used cellulose types are cellulose powder obtained by acid hydrolysis and
mechanical disintegration (degree of polymerization, DP, 150-300), spruce sulfite
dissolving pulp, pine sulfate dissolving pulp (DP 600-900), scoured and
bleached cotton linters (DP 800-2000), and cellulose prepared by Acetobacter
xylinum (DP 1000-2000).
Important synthesis routes to obtain regioselectively functionalized celluloses
have been developed: Firstly, the complete functionalization of cellulose followed
by a site selective reaction of the introduced groups and by the elimination of the
primary groups. Secondly, the regioselective partial functionalization of cellulose
and subsequent reactions of the free or functionalized OH groups followed by the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

elimination of the primary groups. Typical results of these investigations are

summarized in Klemm et al. (Ρ). Silylcelluloses (soluble in organic solvents, DS
0.4-3.0, deblocking by HC1/H 0 resp. fluoride ions), tritylethers (soluble,
2

detritylation by HC1), and cellulose p-toluenesulfonates (DS up to 2.3, soluble in

organic solvents, reactive in nucleophilic substitutions), have been developed as
suitable cellulose intermediates for regioselective functionalization.

Regioselective Synthesis and Subsequent Reactions of Cellulose Ethers

The silylation of cellulosic O H groups leads to a preferred 6-0- or 2,6-di-O-
functionalization. These silylether intermediates open up a wide field of subse­
quent reactions. The chemical modification of starch is included in the investiga­
tions.
The trimethylsilylation (R = Me, cf. Figure 3) of cellulose dissolved in N -
August 31, 2012 | http://pubs.acs.org

methylpyrrolidone (NMP)/LiCl leads to DS values up to 3.0. At DS values of

about 1.0, 60 % of the silylether groups are located at C-6. In a cellulose
suspension in N-methylpyrrolidone/NH at -33 to -15 °C, the trimethylsilyl-
3

cellulose dissolves at DS 1.3 after a heterogeneous reaction, and the silylation takes
place up to DS values of 3.0. At DS values of about 1.0, 80 % of the silylether
groups are located at C-6.
In the case of thexyldimethylsilylation (R = CMe CHMe = Thx), this higher 6-
2 2

O-regioselectivity leads to a very high selectivity at the primary O H groups and to

the formation of pure 6-O-thexyldimethylsilyl cellulose (10-12). This thexyldi­
methylsilylation in N M P / N H stops at the DS of 1.0. No further silylation takes
3

place, even with additional amounts of chlorosilane, either at higher temperatures

or after isolation and dissolution of the polymer in pyridine. We assume a low
accessibility of the O H groups at positions 2 and 3 after swelling and decry-
stallization with ammonia, in contrast to OH groups in the cellulose dissolved in
N M P / L i C l or Ν,Ν-dimethylacetamide (DMA)/LiCl (75).
The low selectivity in case of dissolved cellulose is suitable to prepare 2,6-di-
O-thexyldimethylsilyl cellulose even with the bulky reagent thexyldimethyl-
chlorosilane.
As summarized in Figure 3, starting from primary 6-O-protected celluloses we

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 24
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002
August 31, 2012 | http://pubs.acs.org

1
R - CH 3I H C-^
2

2
R = CH , - ^ - N 0 , CH = C H - @
3 2

3
R - CCI ,-@-N0 , (CH ) CH
3 2 2 14 3

Figure 3. Subsequent reactions of 6-0-silylated celluloses.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 25

developed effective ways to obtain different types of 2,3-di-O-esters and -ethers

by reactions of the free O H groups as well as of esters by reaction of the silylether
group (14).
As in the case of the cellulose products, the subsequent acylation of 6-0- and
2,6-di-O-thexyldimethylsilyl starch leads to the 2,3-di-O- resp. 3-0-starch esters.
Whereas the alkylation of 2,6-di-0-silyl celluloses is a suitable way to prepare the
3-0-ethers, the related starch derivatives form the 2-0-ethers with migration of
the silyl groups to position 3. The same results could be observed in the
benzylation of β-cyclodextrine thexyldimethylsilylated resp. t-butyldimethyl-
silylated in positions 2 and 6. The formation of the 2-0-alkyl-3,6-di-0-silyl
derivatives was mainly concluded from Ή N M R spectra by the typical high field
shift of the proton at C-2 after alkylation (75).
In cooperation with P. Mischnick (16), we investigated the silyl group migration
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

in alkylation of oligo- and polysaccharides with oc-glycosidic linkages as a suitable

synthesis way to obtain the corresponding 2-0-ethers. Figure 4 shows 2-0-
methylation of 2,6-di-O-thexyldimethylsilyl starch as a typical example. The cyclic
intermediates possessing two trans-diequatorial O H groups are formed under the
reaction conditions because of interaction with the neighboring Ια-oxygen. After
desilylation of the 2-0-ethers, a subsequent acylation resp. alkylation proceeds to
form regioselectively trifunctionalized starches (cf. Figure 12).
A further and well-known 6-0-blocking group in cellulose chemistry represents
the trityl group suitable for the preparation of many of 2,3-di-O-derivatives (77-
20). With the aim of a complete detritylation under mild conditions, we
investigated the formation and properties of methoxy substituted tritylethers with
this result: The higher the methoxy substitution, the higher is the rate of tritylation
and detritylation (27, 22). With respect to product stability, the p-mono-
August 31, 2012 | http://pubs.acs.org

methoxytrityl group is very suitable for 6-0-protection, e.g., for synthesis of 2,3-
di-O-carboxymethyl cellulose, including a complete deprotection with ethanolic
HC1 (23) as demonstrated in Figure 5.

Functionalization via Ester Intermediates of Controlled Stability and

Reactivity
The introduction of the p-toluenesulfonyl (tosyl) group into different types of
celluloses in the DP range of 150-2000 is a suitable way to prepare 6-
functionalized derivatives under "Umpolung" of the reactivity at position 6.
Donor agents such as sodium thiosulfate lead to cellulose-deoxy-thiosulfates
(Bunte salts) with high 6-regioselectivity (22).
The cellulose p-toluenesulfonates were synthesized in an N,N-dimethylacet-
amide/LiCl solution of cellulose in the presence of triethylamine (TEA). In the DS
range of about 1 - controlled by the mol equivalents of the reagent tosyl chloride -
80 % of the sulfonate groups are located at position 6 (24, 25). The subsequent
reaction (DS values of cellulose p-toluenesulfonates up to 2.3) with sodium thio­
sulfate in a mixture of dimethylsulfoxide and water resulted in nucleophilic
substitution of the primary tosylate groups forming water-soluble polymers (Figure
6). This Bunte salt formation takes place in the same way after subsequent functio­
nalization of free O H groups in the cellulose p-toluenesulfonates.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Ο —Si-R Ο —Si-R
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

R R
R - Thx, Bu' ii

O —Si-R
August 31, 2012 | http://pubs.acs.org

Figure 4. Silyl group migration in alkylation of 2,6-di-O-thexyldimethylsilyl

starch.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 27

H CO 3
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

H CO-H C ^
3 4 e C,H S

CI-C-C H 6 5

°--« 70°C

CICH COONa/
2

NaOH (DMSO)
70°C, 29h

H,CO
August 31, 2012 | http://pubs.acs.org

OH
HCI

C H OH/H 0
5 5 2

CH COONa 2
CH,COONa
CH COONa
2
CHoCOONa

Figure 5. Methoxytrityl ethers as very suitable 6-O-protective group in cellu-

lose chemistry.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 OH O-Tos
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002
August 31, 2012 | http://pubs.acs.org

Figure 6. Synthesis of cellulose-deoxy-thiosulfates (Bunte salts) via cellulose

p-toluensulfonates.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 29

Typical examples are the formation of the corresponding esters, cellulose-0-(4-

chloro-phenylcarbamate)-deoxy-thiosulfate and cellulose acetate-deoxy-thiosulfate,
1
soluble in organic solvents such as tetrahydrofurane and acetone (cf. R in Figure
8).
Further synthesis pathways to obtain Bunte salts of cellulose and
cellulose derivatives consist in the reaction of different types of chlorinated and
unsaturated cellulose esters described in Camacho Gomez (22). Examples are
substitution reactions of cellulose-0-(3-chloropropionate), cellulose-0-(4-chloro-
methylbenzoate), cellulose-0-(2-chloroethylcarbonate), and cellulose-0-(N-2-
chloroethylcarbamate) with sodium thiosulfate in a mixture of dimethylsulfoxide
2
and water at 80 °C for 24 h (cf. R in Figure 8), as well as addition reactions of
cellulose-O-monomaleate with this reagent.
Three important types of subsequent reactions of low-molecular Bunte salts are
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

known in relation to the conditions (26) as shown in Figure 7.

For the polymeric and watersoluble cellulose-deoxy-thiosulfates, the introduced
anionic S-S(0 )0-groups are useful as protected alkene and thiol functions,
2

suitable for subsequent crosslinking reactions in solution, as well as in films,

ultrathin layers, recognition matrices, and spherical supports. These investigations
are part of our ongoing research on regioselective functionalized celluloses and
starches, including the preparation of artifical supramolecular architectures (27).
First results demonstrate effective crosslinking of Bunte salts of cellulose and
cellulose derivatives in the presence of oxidizing reagents such as iodine or
hydrogen peroxide (cf. Figure 8). As a result of the crosslinking, the polymers
precipitated from solutions and layers formed from water or organic solvents were
insoluble after the described treatment.
August 31, 2012 | http://pubs.acs.org

Analysis of the Molecular Structure

An important part of an investigation on regiocontrol in polysaccharide chemistry
is the analysis of the functionalization patterns prepared. As suitable methods we
used C and Ή N M R spectroscopy of the polymers and HPLC after chain
l 3

degradation. In the case of HPLC a complete methylation of the free O H groups

takes place (methyl inflate, ditert. butyl pyridine, 4 h 60 °C, 16 h 25 °C) before
acidic degradation (0.2 Ν trifluoroacetic acid in water) occurs. Typical examples
are described in Erler et al. (28).
A suitable way to take Ή Ή COSY spectra of high molecular cellulose and
starch derivatives is based on the stepwise introduction of methyl and acetyl groups
under complete functionalization of free OH groups. Starting from the
regioselectively functionalized silylethers, the subsequent chemical modification
leads, e.g., to 6-0-acetyl-2,3-di-0-methyl cellulose (Figure 9).
The ' H N M R spectra of the prepared polymers show a high resolution. The very
intensive signals of the original silyl groups are eliminated, the methylation causes
a signal shift of the proton at the functionalized C atom to higher field, and the
acetylation of the former silylated OH groups a signal shift to lower field. Typical
examples are presented in Figure 10 as well as in (12, 13, 29).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 30
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

e
R-X-CH,-CH,-S-S-O Na®

NaOH H 0
2 2

Η , Ο , H®

R-X-CH = CH, R-X-CH -CH -S 2 2

R X~"CH2 CHg S
+ Na S 0
2 2 3

+ H0 + 2NaHS0
August 31, 2012 | http://pubs.acs.org

2 4

R-X-CH -CH -SH2 2

+ NaHS0 4

Figure 7. Subsequent reactions of Bunte salts.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 31
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002
August 31, 2012 | http://pubs.acs.org

— C - 0 - C H 2 C H 2 — , — C-NH-CH CH
2 2 — ,

R 3
= Η or R S 0 Na
2
2 3

Figure 8. Crosslinking of Bunte salts of cellulose in the presence of oxidizing

reagents.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 32

\ \
V
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

Si Si
/ I /I
Ο
NaH, CH I 3

Ο. (THF)
HO H,CO
OH 25°C, 24h + OCH,
50°C, 48h
Bu N® F©
4

(THF)

Y ο
OH
5 0 ° C . 24h

ο
August 31, 2012 | http://pubs.acs.org

(CH CO) 0,
3 2

Ν Ο-. pyridine
H CO
a
H C0
3

OCH, 25°C, 24h + OCH,

100°C, 4h

Figure 9. Synthesis of acetyl-methyl celluloses for Ή Ή COSY NMR spectros­

copy.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002
August 31, 2012 | http://pubs.acs.org

Figure 10. Ή Η COSY NMR spectrum of 6-O-acety 1-2,3-di-O-methylcellulose

preparedfrom6-O-thexyldimethylsilyl cellulose (DS = 0.78).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 34

Conclusion
The silylation of cellulose and starch with the bulky thexyldimethylchlorsilane
represents an important example of a regioselective polysaccharide functiona-
lization. This etherification is controlled by a partial (6-O-silylation) or complete
(2,3-di-O selectivity) loosening of the supramolecular structure caused by
decrystallization and solvation of the polymers.
In the case of cellulose these silylethers are suitable intermediates to prepare 2,3-
di-O and 3-0 ethers and esters (protection group technique) as well as to synthesize
6-0 and 2,6-di-O esters (activation by silylation of the O H groups) as summarized
in Figure 11.
The corresponding starch functionalization additionally leads to 2-0 ethers as a
result of migration of the 2-0-silyl groups to position 3 (Figure 12).
In all cases the subsequent reactions may be used to prepare different types of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

completely functionalized polymers suitable for *H N M R analysis and investi-

gation into structure properties relationships.
As an important example of ester intermediates in regioselective cellulose
functionalization, water-soluble and film-forming Bunte salts (deoxy-thiosulfates)
represent crosslinkable polymers with a wide range of additional functional groups.
Prepared from cellulose p-toluenesulfonates, Bunte salts react like anionic
protected alkene and thiol groups suitable for controlled polymer modification.

Acknowledgements
The author is indebted to Dr. Th. Heinze, Dr. Katrin Petzold, Dr. Armin Stein,
Dipl.-Chem. Andreas Koschella, Dipl.-Chem. Juan Chamacho Gomez, and Dipl.-
Chem. Kerstin Rahn for their creative work in the field of the described ethers and
esters of cellulose and starch.
August 31, 2012 | http://pubs.acs.org

The financial support of the work on regioselective functionalization by the

Deutsche Forschungsgemeinschaft, by the Bundesministerium fur Ernàhrung,
Landwirtschaft und Forsten, as well as by the Fonds der Chemischen Industrie of
the FRG, is gratefully acknowledged.

Literature Cited
(1) Shirai, Α.; Takahashi, M.; Kaneko, H.; Nishimura, S.; Ogawa, M.; Nishi, N.;
Tokura, S. Int. J. Biol.Macromol.1994, 16, 297.
(2) Kobayashi, S.; Kashiwa, K.; Kawasaki, T.; Shoda, S. J. Am. Chem. Soc. 1991,
113, 3079.
(3) Nakatsubo, F.; Kamitakahara, H.; Hori, M. J. Am. Chem. Soc.1996, 118, 1677.
(4) Nishimura, T.; Takano, T.; Nakatsubo, F.; Murakami, K. Mokuzai Gakkaishi
1993, 39, 40.
(5) Burchard, W. Adv. In Colloid and Interface Sci. 1996, 64, 45.
(6) Nojiri, M.; Kondo, T. Macromolecules 1996, 29, 2392.
(7) Kondo, T. J. Polym. Sci., Polym. Phys. 1994, B32, 1229.
(8) Hagaki, H.; Takahashi, I.; Natsume, M.; Kondo, T. Polymer Bull. 1994, 32, 77.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 35

1
R'-Hal ^OSiMe R 2

NaH
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

2
R0 OR 2

N® B u F ®
4
2
R - Me

.OH

ACTIVATION MeO OMe

2,3-di-O-
1
SO, I R - CH,
August 31, 2012 | http://pubs.acs.org

e
^OS0 3 Ν a® .OAc

e
HO OH(OS0 Na®) 3 MeO OMe

6-0-, 2,6-di-O-
1
R - Me, Thx, Bu'
2
R = Me, Bn, Ac
Figure 11. Regiocontrol by selective protection and activation.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 36

MIGRATION
1
0-SiMe R 2 1
OSiMe,R
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

ΗΟ^Τ-ή N a H
R'Me Si0' 2
N
OR 2

1 ,° Ο
R Me,S
2
N® B u F © 4
R - Bn
2,6-di-O-

1
^OSiMe R 2
HO OBn

AcO OSiMe,R 1
2-0-
August 31, 2012 | http://pubs.acs.org

3-0- (i) 1
R Me SiCI 2 . AcO or M e l
Me, Thx, Bu' (") A c
2 °

R = Me, Bn 1
^OSiMe R 2 ^OAc(Me)

AcO OBn (Me) AcO OBn

Figure 12. Regie-control by selective protection and silyl group migration.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 37

(9) Klemm, D.; Stein, Α.; Heinze, Th.; Philipp, Β.; Wagenknecht, W. In Polymeric
Materials Encyclopedia; Salamone, J.C., Ed.; CRC Press, Inc.: Boca Raton,
FL, USA, 1996, Vol. 2c; pp 1032-1054.
(10) Stein, A . PhD Thesis; University of Jena, Germany, 1991.
(11) Henze-Wetkamp, H.; Zugenmaier, P.; Stein, Α.; Klemm, D.Macromol.Symp.
1995, 99, 245.
(12) Klemm, D.; Stein, A. J. Macromol. Sci., Pure Appl. Chem. 1995, A32, 899.
(13) Klemm, D.; Heinze, Th.; Stein, Α.; Liebert, T. Macromol. Symp. 1995, 99,
129.
(14) Stein, Α.; Klemm, D. Macromol. Chem., Rapid Commun. 1988, 9, 569.
(15) Petzold, Κ. PhD Thesis; University of Jena, Germany, 1996.
(16) Mischnick, P.; Lange, M.; Gohdes, M . ; Stein, Α.; Petzold, Κ. Carbohydr. Res.
1995, 277, 179.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch002

(17) Harkness, B.R.; Gray, D.G. Macromolecules 1990, 23, 1452.

(18) Harkness, B.R.; Gray, D.G. Macromolecules 1991, 24, 1800.
(19) Kondo, T.; Gray, D.G. Carbohydr. Res. 1991, 220, 173.
(20) Kondo, T. Carbohydr. Res. 1993, 238, 231.
(21) Camacho, Gómez, J.A.; Erler, U.W.; Klemm, D. Macromol. Chem. Phys. 1996,
197, 953.
(22) Camacho Gómez, J. PhD Thesis; University of Jena, Germany, 1997
(23) Heinze, Th.; Röttig, Κ.; Nehls, I. Macromol. Rapid Commun. 1994, 15, 311.
(24) Rahn, K.; Diamantoglou, M.; Klemm, D.; Berghmans, H.; Heinze, Th. Angew.
Makromol. Chem 1996, 238, 143.
(25) Heinze, Th.; Rahn, K.; Jaspers, M.; Berghmans, H. J. Appl. Polym. Sci. 1996,
60, 1891.
(26) Milligan, B.; Swan, J.M. Rev. pure appl. Chem. 1962, 12, 72.
August 31, 2012 | http://pubs.acs.org

(27) Schaub, M.; Wenz, G.; Wegner, G.; Stein, Α.; Klemm, D. Adv. Mater. 1993, 5,
919.
(28) Erler, U . ; Mischnick, P.; Stein, Α.; Klemm, D. Polymer Bulletin 1992, 29, 349.
(29) Stein, Α.; Klemm, D. Das Papier 1995, 49, 732.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 3

Long-Chain Cellulose Esters: Preparation,

Properties, and Perspective

1 1 2
Kevin J. Edgar , Thomas J. Pecorini , and Wolfgang G. Glasser

1
Eastman Chemical Company, P.O. Box 511, Kingsport, TN
2
Biobased Materials/Recycling Center, and Department of Wood Science and Forest
Products, Virginia Polytechnic Institute and State University,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Blacksburg, VA 24061

Obstacles to the heterogeneous esterification of cellulose with long-

-chain alkanoic acids were overcome (a) by reacting cellulose in
homogeneous phase solution (in DMAc/LiCl) in the presence of
dicyclohexyldicarbodiimide (DCC) and 4-pyrrolidinopyridine (PP)
or tosyl chloride; or (b) by reaching cellulose heterogeneously, in
DMAc-suspension, with anhydrides in the presence of Ti (IV)
isopropoxide. Long-chain cellulose esters (LCCEs) were prepared
with constant and uniform degree of substitution (DS), and with
stoichiometric control. Cellulose ester structure was found to
August 31, 2012 | http://pubs.acs.org

influence solubility, flexural modulus, melt viscosity, thermal

transitions, and enzyme recognition. LCCEs showed melt-rheology
behaviors that are normally found only in plasticized esters. Melt
viscosity and flexural modulus of these internally plasticized
cellulose esters were found to be related to the solubility parameter
over a broad range of ester-substituents and DS. Practical synthesis
of LCCEs is possible leading to melt-processable (uncompounded)
thermoplastic polymers on cellulose basis.

Classically, cellulose esters are prepared using mineral acid catalysts and carboxylic
anhydrides in the corresponding carboxylic acid diluent with acetic, propionic,
butyric and valeric acids, or mixtures thereof. Due to slow reaction rates and
competitive cellulose chain cleavage, it is difficult to prepare esters with longer
chain acids by this method (1). Cellulose esters with short chain carboxylic acids
(C-2 and C-3) have high glass transition and melting temperatures and relatively low
decomposition temperatures. In some cases (low-DS-acetates, for example), the
decomposition temperature is actually lower than the T and T , which is similar to
g m

parent cellulose. Cellulose esters are therefore normally melt-processed in the

38 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 39

presence of plasticizers that enlarge the processing window. Attachment of longer

chain ester groups, if it could be accomplished efficiently, should provide an
effective internal plasticizer which potentially eliminates the need for external
plasticizers which may be prone to extraction or volatilization. This would also
substantially reduce the solubility parameter of the cellulose ester, enhancing (a) the
solubility in less polar solvents, and (b) the blending with less polar polymers than
are miscible with currently available materials.
Previous attempts to prepare long-chain cellulose esters have been severely
limited in scope and in practicality. The first route was the so-called "impeller"
method by Clarke and Malm in 1932 (2). They attempted to compensate for the
low reactivity of the sterically encumbered long-chain acyls by converting them to
mixed anhydrides with reactive acyls such as chloroacetyl or methoxyacetyl. Thus,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

reaction of cellulose with acetic acid, stearic acid, and chloroacetic anhydride in
chloroacetic acid, with magnesium perchlorate catalysis, gave a cellulose acetate
stéarate. The "impeller" acids efficiently catalyze chain cleavage and make the
degree of polymerization (DP) difficult to control. Residues of the impeller acid
may also be found on the product. A modern version of the impeller method,
which uses trifluoroacetic anhydride as the impeller (3), effectively conserves DP
but provides access only to fully substituted esters.
A second approach to the preparation of long-chain cellulose esters was
described by Malm and coworkers in 1951 (1). They found that acylation with the
appropriate acid chloride in 1,4-dioxane with pyridine as an acid acceptor and
catalyst afforded cellulose triesters ranging from acetate to hexanoate to palmitate.
Surprisingly, only a moderate amount of DP-reduction occurred in this reaction.
Although this is a convenient method to form triesters, this method is limited to
August 31, 2012 | http://pubs.acs.org

high-DS cellulose esters, and furthermore it requires amorphous, and thus highly
reactive, regenerated cellulose as starting material. This introduces the inefficiency
of having to acylate cellulose twice, with an intervening deacetylation step.
However, the availability of this method allowed Malm et al. (4) to prepare the
entire series of cellulose triesters and investigate their properties. A profound
impact on such properties as melting point, solubility, moisture regain, density, and
tensile strength was firmly established. This research indicated an area of promising
new materials which was limited only by the missing dimension of DS-
modification.
Recently Tao et al. have described a new and more direct method to
synthesize cellulose esters with long-chain acids (5). The extremely simple method
involves the reaction of mercerized cellulose with an acid chloride at elevated
temperature under vacuum to facilitate the removal of the by-product HC1.
Partially substituted cellulose esters with a long chain (palmitate) were prepared in
this manner for the first time, with DS ranging from zero to 2.5. Although simple,
the method is limited by the tendency of the co-produced HC1 to cause chain
cleavage and loss of DP. Furthermore, esterification does not proceed past DS 1.5

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 40

unless at least 10 equivalents of the acid chloride are used, even if no solvent is
used. Most seriously, the product esters are not homogeneously substituted and
are not soluble in organic solvents (6).

Long-chain Cellulose Esters (LCCEs) by Solution Acylation

In the 1980s, Turbak (7) and McCormick et al. (8) showed that cellulose
dissolves in a solution of lithium chloride (LiCl) in N,N-dimethylacetamide
(DMAc) under well-defined conditions. Of equal importance was the subsequent
work by several investigators, with major contributions from the McCormick
laboratory, which showed that reactions of cellulose could be performed in this
solvent system (9). With respect to ester synthesis, McCormick, Diamantoglou
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

(10), Samaranayake and Glasser (11,12), and others (13-15) showed that cellulose
could be induced to react with carboxylic anhydrides, acid chlorides, and other
electrophilic acyl derivatives using mineral acid or alkaline catalysts to afford
partially substituted cellulose esters directly, without the need for a hydrolysis
step. This was significant because conventional cellulose esterification is a
heterogeneous reaction until the cellulose is nearly fully reacted; therefore, the only
previous way to obtain processable (solvent-soluble or melt-flowable) cellulose
esters was to esterify the cellulose fully and then back-hydrolyze the derivative to
the desired DS. By reacting cellulose in D M A c / L i C l solution with 3.8 eq of
propionic anhydride and 0.1 eq of acetic anhydride at 100°C, and without any
catalyst, we now confirmed that cellulose may react with short-chain anhydrides to
form a mixed cellulose acetate propionate with D S 2.5 and D S 0.10.
pr ac

Application of the D M A c / L i C l solution process to long-chain cellulose

August 31, 2012 | http://pubs.acs.org

esters is complicated by two factors that are related (a) to solubility of reagents and
reaction products, and (b) to the susceptibility of cellulose to depolymerization.
As chain-length gets higher than hexanoic anhydride, the requisite anhydrides either
are poorly soluble or insoluble in the reaction mixture, or they contribute to a
substantial inefficiency of esterification due to the loss of the by-product carboxylic
acid as a consequence of esterification. Loss of solubility results in heterogeneous
acetylation and a heterogeneous and insoluble product (see Entry 1, Table I). We
were able to address this problem by adapting the impeller method of Clarke and
Malm. Mixed anhydrides of reactive acyl groups were formed with either 4-
pyrrolidinopyridine (PP) or with tosic acid; or by using the corresponding acid
chlorides with either pyridine or triethylamine as acid acceptor. The use of the
impeller method or the acyl chloride improved efficiency by preventing the loss of
the by-product carboxylic acid during esterification with anhydrides as well as
contributing to solubility in general.
The use of the mixed anhydride with PP was pioneered by Samaranayake
and Glasser (11). The reaction is mediated by the presence of dicyclohexyl-

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Table I. LCCEs by Solution Esterification in DMAc/LiCl.

Entry Acylating Equivalents Acid Reaction Reaction DS IV Solubility
Agent per A H G Scavenger Time(h) Temp. (C) (DMSO) (iœ)

a
1 Stearic 1.00 None 1 110 0.95 Insoluble
Anhydride
b
2 Hexanoyl 1.00 Pyridine 0.5 60 0.89 1.10 41 D M S O , NMP, Pyridine
Chloride
Acetone, M E K , CHC1 ,
3
b
3 Hexanoyl 2.00 Pyridine 0.5 60 1.70 0.76 90 HOAc, THF, DMSO, N M P ,
Chloride Pyridine

Pyridine
b c
4 Lauroyl 2.00 Pyridine 0.5 60 1.83 1.00 67
Chloride Acetone, M E K , CHC1 ,3

HOAc, THF, DMSO, N M P ,

e
5 Stearoyl 1.00 Pyridine 1 105 0.79" 0.11 27 Pyridine
Chloride

In Cellulose Derivatives; Heinze, T., et al.;

b
Acetic 3.00 1.93
Anhydride
a b c

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

DS by alcoholysis/GC. D S by Ή NMR. Phenol/Tetrachloroethane. ^ by GPC in T H F except Entry 2 (NMP).
Tartly insoluble. A H G = Anhydroglucose
 42

1.DCC V ρ 2.DCU

RCOOH * R—ΰ—0—H-

l PP
6
RCOOH + RJ-N^-NQ + R-oJ-fl — RJ^O^Q
RCO<?
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Scheme 1. Mechanism of esterification with DCC/PP reagent.

carbodiimide (DCC), a powerful condensation agent which is well-known to couple

amines and carboxylic acids in peptide and protein chemistry (16), but it is rarely
used in polymer modification reactions. When DCC is applied in combination with
anhydrides, the by-product carboxylic acid is recycled by formatting a mixed
anhydride with PP (Scheme 1). Mixed anhydrides are soluble in the reaction
August 31, 2012 | http://pubs.acs.org

mixture up to hexanoic acid; beyond C-6 carboxylic acids, the reaction mixtures
become heterogeneous (11).
The second application of the impeller method has involved the use of tosyl
chloride (TsCl) (15). When we added a solution of long-chain alkanoic acids (C-12
to C-20) in D M A c to a cellulose solution in DMAc/LiCl, the non-polar aliphatic
carboxylic acid precipitated (14). This process was subsequently reversed by
adding TsOH in D M A c . The homogenization of the reaction mixture indicates the
formation of a mixed anhydride of the alkanoic acid with TsOH. By heating the
reaction mixture to 50 to 70°C in the presence of sufficient acid acceptor (pyridine
or equivalent), a long-chain cellulose ester (LCCE) is formed that is (a) soluble in
the reaction mixture; (b) that has a well-preserved DP; and (c) that is completely
devoid of tosyl or Cl-substitution. C-12 to C-20 derivatives with DS between 2.8
and 2.9 were obtained using a stoichiometric ratio of two equivalents of acid per
cellulose hydroxyl (14).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 43

Another method of overcoming the solubility problems routinely

encountered during cellulose esterification in DMAc/LiCl involves the use of acid
chlorides combined with pyridine (or equivalent) as acid acceptor. Although acid
chlorides are soluble in the DMAc/LiCl solvent mixture, they quickly become
insoluble when the necessary acid acceptor is added. Triethylamine is less useful
for this purpose than is pyridine. Reaction mixture homogeneity could also be
triggered by product insolubility. We found that this difficulty could be
circumvented by adding the acid chloride in solution in a small amount of a co-
solvent which was a better solvent for the product; tetrahydrofuran (THF) proved
useful for this purpose. Soluble and melt-processable products were obtained that
are represented by Entries 2 to 5 of Table I. We were able to make long-chain
esters and mixed esters of cellulose with hexanoate, laurate, and stéarate groups,
with DS ranging from 0.9 to 2.7, using this procedure.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

These procedures provide access to a broad range of materials, with one or

more long-chain ester substituents, having any DS and any chain length. This
flexibility has not previously been available to cellulose chemists.

Heterogeneous Acylation Using Ti Catalyst in Amide Solvent

The methods we have so far described provide access to a class of cellulose

esters which was previously inaccessible, and so expand the armament of the
materials scientist. These methods have the advantage that they make available the
entire range of ester-DS and chain length by simply choosing the acylating agent and
stoichiometry; however, the DMAc/LiCl system is less than ideal from the
perspective of efficiency. Because of the viscosity of cellulose solutions in this
August 31, 2012 | http://pubs.acs.org

system, cellulose concentrations above about 8% are impractical. The procedure

for dissolving cellulose in this system is complex, all of the components are quite
hygroscopic, and the expense of lithium chloride is a practical issue. It would be
desirable to find a more efficient route to these long-chain esters.
The use of titanium (IV) alkoxides as catalysts for cellulose esterification in
carboxylic acid diluents has been known since the work of Tamblyn and Touey in
the 1960s (17). These Lewis acid catalysts require much higher temperature (120 -
180°C) than conventional mineral acid catalysts. We have found that titanium (IV)
isopropoxide is an excellent catalyst for the reaction of cellulose in the appropriate
carboxylic acid solvent with short-chain anhydrides for preparing triesters. We
have not found this heterogeneous system to be a practical direct route to partially
substituted esters. We wondered if a more powerful solvent for partially
substituted esters, such as a polar aprotic solvent like a dialkylcarboxamide, would
facilitate their preparation as homogeneous materials. Reaction of cellulose with a

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 44

mixture of hexanoic and acetic anhydrides (2 eq each), in D M A c diluent, with

titanium (IV) isopropoxide catalyst, afforded a homogeneous solution. Isolation of
the product by precipitation gave a cellulose acetate hexanoate, with D S 1.88, ac

DShex 0.91 (Entry 1, Table Π). Several features of this reaction are of interest; the
product is partially substituted and apparently homogeneous, as judged by its
solubility in a wide range of solvents. Despite the high reaction temperature, a high
molecular weight product was obtained, which is perhaps attributable to the
moderating effect of the mildly alkaline amide solvent on the acidic catalyst. While
the cellulose clearly reacted preferentially with the shorter-chain anhydride, a
substantial amount of reaction with the hexanoic anhydride did occur.
The generality of this method is illustrated in Table Π. In each case (except
Entry 7), D M A c was the diluent and titanium (IV) isopropoxide (3% by weight
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

based on cellulose) was the catalyst. These seven examples illustrate the fact that it
is possible by this method to esterify cellulose with acyl groups as large as
palmitate (C-16). High temperatures are required, but even so, the products are of
relatively high molecular weight. The total DS of these products ranges from 2.5-
2.7. It should be noted here that the esterification with this catalyst in the amide
solvent is rather slow, with reaction times of 6-12 h common. This slow
esterification, however, is what permits easily repeatable isolation of partially
substituted esters. The esterification rate, when the cellulose fully dissolves, is
sufficiently slow that dissolution may be used as an endpoint, with highly
repeatable results in terms of product DS. The product T varies in a predictable
g

way depending on the relative content of acetyl and long-chain acyl. All of these
products are soluble in a wide range of solvents, including acetone, acetic acid,
N M P , CHC1 , and THF. Even where only long-chain anhydride is used (Entry 3),
August 31, 2012 | http://pubs.acs.org

a partially substituted, homogeneous long-chain ester is smoothly obtained. It is

especially interesting that in this case the product contains a DS 0.12 of acetyl,
even though no acetic anhydride was used. We believe that the acetyl in the
product must have come from reaction with the Ν,Ν-dimethylacetamide solvent.
The results of Entry 4, Table III, support this contention. The hypothesis is
further confirmed by Entry 7 of Table II in which a solution is identified that allows
exclusive attachment of a long-chain acyl. In this reaction, we substituted a
tetraalkylurea, N,N-dimethyl-2-imidazolidinone (DMI), for D M A c as diluent.
Because there is no solvent acetyl to react with cellulose, a cellulose hexanoate
without acetyl groups is expected and that is what is obtained. There does appear
to be slightly more chain cleavage when the D M I diluent is used; the reaction rate
and efficiency with respect to the anhydride ( D S 2.73 obtained with only 3 eq
hex

hexanoic anhydride used) are also substantially higher than with D M A c . It is clear
from the results in Table II that the cellulose acylation rate with a given anhydride
declines predictably as the anhydride chain length increases. Comparing the results
of reacting cellulose with two equivalents each of acetic anhydride and a long-chain
anhydride, proceeding up the series from hexanoic to nonanoic to lauric to palmitic

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Table II. LCCEs by Titanium (IV) Isopropoxide-Catalyzed Esterification in DMAc.

Entry Anhydrides) Equivalents Reaction Reaction DS IV Mw

e
per A H G Time(h) Temp. (C) (Ή NMR) (NMP) /1000 /lOOO* (D&Q

1 Aœtic 2.00 9 155 1.91 1.39 40 164 149°C

Hexanoic 2.00 0.75
2 Aœtic 1.00 9 155 1.38 0.90 35 113 122°C
Hexanoic 3.00 1.36
3 Acetic 0.00 6 155 0.12 0.94 33 245 119°C
Hexanoic 4.50 2.39
4 Acetic 2.00 11 145 2.03 1.18 44 177 129°C
Nonanoic 2.00 0.70
5 Acetic 3.50 12 140 2.40 2.12 96 295 165°C
Laurie 1.00 0.20
b
6 Acetic 2.00 12 145 2.06 0.29 33 125 156°C
Palmitic 2.00 0.42
C

In Cellulose Derivatives; Heinze, T., et al.;

7 Acetic 0.00 7 140 0.00 0.44 23 61 104°C
Hexanoic 3.00 2.73
a b c
By GPC in NMP. Partly insoluble. N,N-Dimethylimidazolidinone (DMI) used as solvent instead of DMAc.
A H G = Anhydroglucose.

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Table III. Cellulose Acetate Nonanoates by Titanium (IV) Isopropoxide-Catalyzed Esterification in DMAc.

Entry Anhydride Equivalents Reaction Reaction DS IV Mn T

m

per A H G Time (h) Temp. (C) ( Ή NMR) (NMP) /1000 /1000 (Difc) (DSC)
a
1 Aœtic 2.00 11 145 2.03 1.18 44 177 129 183
Nonanoic 2.00 0.70
2 Aœtic 3.00 8 145 2.44 1.71 43 220 161 180*
Nonanoic 1.00 0.26
b
3 Aœtic 1.00 13 155 1.59 1.16 44 182 118 174
Nonanoic 3.00 1.11
b
4 Aœtic 0.00 13 160 1.11 0.89 31 200 110 167
Nonanoic 4.00 1.35

In Cellulose Derivatives; Heinze, T., et al.;

b
"First scan only. B y modulated DSC. A H G = Anhydroglucose. M , M by GPC in N M P .
n w

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

 47

(Entries 1, 4, 5, and 6, Table II) shows that the DS of long-chain acyls in the
product declines with increasing chain length. It remained to be seen what chain
length was necessary to obtain sufficient processability for a self-plasticized
polymer.
We performed a series of experiments (Table III) reacting cellulose with
acetic and nonanoic anhydrides with titanium (IV) isopropoxide in D M A c diluent
to learn more about the nature and properties of materials available via this
chemistry. To maintain practical reaction rates, the reaction temperatures ranged
from 145°C for the more acetic anhydride-rich runs, to 160°C for the run with only
nonanoic anhydride. The cellulose acetate nonanoates obtained were all of
reasonably high molecular weight, as indicated by IV and GPC. The T decreasesg

with increasing nonanoyl content, as expected; nonanoyl content can in turn be

rationally controlled by the ratio of anhydrides used. The maximum nonanoyl
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

content obtainable is DS 1.35, with a D S of 1.11 (Entry 4); clearly, nonanoic

ac

anhydride has sufficient bulk (hydrophobicity may also be a factor in early stages
of the reaction with the hydrophilic cellulose) that the reaction of cellulose with
D M A c becomes competitive in rate. The melting temperature also seemed to
decline with increasing nonanoyl substitution; modulated DSC was an invaluable aid
in identifying the T in the high-nonanoyl esters.
m

Mechanical and Rheological Properties

In several respects, esterification of cellulose with long-chain substituents

can provide superior properties over their conventional short side-chain
counterparts. To their disadvantage, currently commercial cellulose esters (CA,
August 31, 2012 | http://pubs.acs.org

CAP, CAB) possess melt viscosities that are too high to permit melt processing at
reasonable temperatures without adding external plasticizer. The compounding step
associated with that addition not only increases the cost of the final product, but
external plasticizers may also leach out of the compounded plastic, producing
undesired smell, taste or feel. Furthermore, selection of an effective plasticizer is
limited by solubility. Therefore, the only good plasticizer for a particular mixed
ester may possess undesirable properties.
In contrast, increasing the amount of long-chain ester essentially increases
the amount of covalently bonded internal plasticizer, and does not produce the
problems cited above. The chemical structure of a typical L C C E with covalently
bonded internal plasticizer is illustrated in Scheme 2. The waxy substituents that
surround the cellulose backbone provide significant plasticization at temperatures at
which most cellulose esters are glassy. We found that LCCEs with fatty acid
substituents larger than C-12 had distinctly dual morphology, with both the
cellulosic and the waxy phase undergoing separate glass-to-rubber and rubber-to-
melt transitions (14). The T -, T - and T -transitions of the waxy (plasticizing)
g c m

phase of LCCEs having substituents in the range of C-12 to C-20 reveal significant

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 48
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Scheme 2. Molecular model of cellulose eicosanoate (C-20) with ester substituents

aligned regularly as they might be in a crystal lattice (From ref. 14).
August 31, 2012 | http://pubs.acs.org

molecular mobility in the temperature range between -60 to +60°C (Fig. 1) (14). A s
expected, the mechanical and Theological properties of several of the LCCEs, listed
in Tables IV and V, vary with changes in ester side-chain length and degree of
substitution similar to the effect of adding greater amounts of external plasticizer.
Indeed, with sufficient side-chain length and degree of substitution, the properties
(particularly modulus and melt viscosity) equal those of plasticized C A and CAP,
affording LCCEs that can be easily processed without external plasticizer. These
properties all reflect the ability of a plasticizer, both external and internal, to
increase free volume. For example, T decreases with increasing external plasticizer
g

content as well as with increasing DS and chain length (see Figure 2). Resins with
lower T should exhibit lower melt viscosities when measured at 220°C. However,
g

as shown in Figure 3, at a given Tg, cellulose esters with longer side-chain lengths
and higher DS have lower melt viscosities than externally plasticized resins of
shorter chain length and lower DS. This suggests that long side-chains are more

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 49
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003
August 31, 2012 | http://pubs.acs.org

N o . of Carbons i n A c y l Substituent

Figure 1. Thermal transitions of the waxy phase of LCCEs in the temperature range
of -60 to +70°C. T by DSC: -A-; T by D M T A : -Δ-; T by DSC: -0-.
m g c

(In accordance with ref 14).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Table IV. Rheological and Mechanical Data for CA, CAP, CAB, CAH and CAN.
Material DS of Plasticizer Absolute Viscosity at Viscosity @ 100 rad/ s T.byDMTA Modulus
long chain Content Mw 1/s (Pa- s) 1/100 s normalized to Mw = (Q (mPa)
(%) (g/mole) (Pas) 100,000 (Pa-s)

CA 0 33 61000 2783 329 1766 103 1650

CA 0 28 61000 4009 317 1702 107 2000
CA 0 18 61000 10320 1243 6673 132 2910
CA 0 0 61000 47500* 203 4410·
CAP 0.73 25 79000 919 261 582 120 1520
CAP 0.73 20 79000 2126 419 934 125 1930
CAP 0.73 0 79000 8500· 188 3580·
CAP 2.60 14 103000 600 306 277 114 1430
CAP 2.60 10 103000 1350 450 407 122 1660
CAP 2.60 6 103000 3000 830 751 131 1960
CAP 2.60 0 103000 10600 1258 1138 156 2410·
b
CAB 1.67 0 90000 141 2070

CAH 0.75 10 164000 2971 702 131 129 1670

CAH 0.75 5 164000 7032 1107 206 137 2120
b

In Cellulose Derivatives; Heinze, T., et al.;

CAH 0.75 0 164000 340· 149 2410·
b
CAH 1.36 0 113000 855 353 233 128/122 1310
b
CAH 2.39 0 245000 1175 104 5 lll/119 620
b
CAN 0.70 0 177000 10300 952 137 129
b
CAN 1.11 0 182000 1929 371 48 118

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

b
CAN 1.35 0 200000 547 73 7 110 690
; D;
• Extrapolated; byDSC.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Table V. Physical Property Data for Compounded C A H and CAN.

Cellulose Ester CAH CAH CAH CAH CAN CAN

DS %DOA 2.39 1.36 0.75 0.75 1.35 1.35

Plasticizer 0.0 0.0 5 10 0.0 5

PROPERTY UNITS
3
Density g/cm 1.17 1.21 1.198 1.11 1.1
IV after molding 0.877 0.998 1.194 1.194 0.77 0.73
Flex Strength MPa 20.1 40.7 57.7 45.7 19.0 13.4
Rex Modulus MPa 668 1322 2115 1674 606 434
Creep Modulus* MPa 179 627 792 578 158 124
0.45 MPa HDT C 61 82 100 87 65 57
1.82 MPa HDT C 43 62 68 58 51 47
Hardness R,L scale R28.3 R87.3 L36.6 L23.5 R23.6 R8.0
23C Notched Izod J/m 194 184 231 230 48 123
OC Notched Izod J/m 10 136 134 127 8 58
-40C Notched Izod J/m 23 24 67 102 10 9

In Cellulose Derivatives; Heinze, T., et al.;

23C UnNotched Izod J/m NB NB 2016 NB NB NB
OC UnNotched Izod J/m NB NB 1663 NB NB NB
-40C UnNotched Izod J/m 267 706 1512 1260 121 207

a)
6.89 MPa, 23 C, 200 h; NB = No Break.

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

 52
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

"100000
cd
August 31, 2012 | http://pubs.acs.org

ο 10000
CV2
CV2

cd
1000

100^
T3
CD
Ν

10

100 120 140 ' 160

§
180 ' ' 200 220
Tg by DMTA or DSC (C)

Figure 3. Normalized viscosities of plasticized and non-plasticized cellulose

esters plotted as a function of T . g

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 53

effective as flow aids than are external plasticizers, which may be related to reduced
hydrogen bonding or the creation of free space between the cellulose backbones due
to the addition of long non-polar side chains. In addition to the plasticizing effect
of bulky substituents, it can not be ruled out that liquid crystallinity factors may
not also contribute to the facilitation of melt processing. Although we did not
detect any direct, overt evidence of liquid crystallinity, we did not specifically and
systematically look for it.
Because the side chains are comprised of CH -units having a low solubility
2

parameter, one would expect the overall solubility parameter of the LCCEs to
decrease with increasing DS and chain length. This expectation is borne out when
solubility parameters calculated by the method of Coleman (18) are related to the
DS (Table VI). Long-chain esterification, therefore, could allow the use of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

alternative external plasticizers that are insoluble in currently commercial cellulosic

materials and may also offer the opportunity to create miscible blends of cellulosics
with more hydrophobic polymers than are compatible with conventional cellulose
esters such as CA, C A P and C A B . Furthermore, the solubility parameter as it is
calculated for these materials, in essence measures how much (internal) plasticizer is
present. Material properties that are controlled by plasticization, such as T , g

normalized melt viscosity, and modulus, all strongly correlate with the solubility
parameter (Figure 4). It is possible, therefore, to design a cellulosic material with
any desired properties (as far as they are related to solubility parameter) simply by
identifying the matching solubility parameter from Figure 4, and esterifying with
the associated combination of DS and chain length. The principal advantage of
selecting a low-DS long-chain cellulose derivative over a high-DS short-chain ester
with similar effective plasticization for melt processing rests with (a) the long side
August 31, 2012 | http://pubs.acs.org

chains effectiveness as processing aids; (b) with their non-fugitive (permanent)

character, and (c) with their likely better compatibility with more hydrophobic
polymers in blending. While the latter is supported by solubility parameter data,
the experimental evidence for it still needs to be developed.

Other Properties

Mixed cellulose esters show crystallinity across the entire spectrum of

substituent mixing. T declines sharply as DS of the bulky substituent rises from 0
m

to 1, followed by a more moderate decline as DS increases from 1 to 3 (Figure 5)

(13). When the number of carbons in the long-chain ester substituent increases
beyond 12 (laurate), a series of thermal transitions is reported which represents
motion by both ester substituents and the cellulosic main chain (14). Broad
crystallization and melting transitions attributed to side-chain crystallinity range
between -19 to +55°C, and these side-chain T - and T - transition temperatures
m c

increased by 10°C per carbon atom of the ester substituent (13, 14).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 54

U220

9.0 9.5 10.0 10.5 11.0

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Solubility Parameter

i100000f

ο 10000^
Β
ο
CV2
C\2
^ 1000
ed

ci 100
>

u
ο
I I I I I I I I I I I I I I I I ' I I I ' 'I ι ι ι ι ι I I

9.4 9.8 10.2 10.6 11.0

August 31, 2012 | http://pubs.acs.org

Solubility Parameter
£5000

9.0 9.4 9.8 10.2 10.6 11.0

Solubility Parameter

Figure 4. Properties for various cellulose esters without plasticizer plotted as a

function of calculated solubility parameter: A) T -values, B) normalized g

viscosities, and C) flexural modulus.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 55
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003
August 31, 2012 | http://pubs.acs.org

Figure 5. Relationship between thermal transitions (T , T and T - T ) and DS

m g m g

with hexanoyl groups, of CAHs (In accordance with ref. 13).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 56

Table V L Solubility Parameters of Selected Cellulose Esters.

Material DS(H,N,P,B) Solubility Parameter

3 1/2
Total DS = 2.7 (cal/cm )
CA 0.00 10.53
CAP 0.73 10.36
CAP 2.60 10.04
CAB 1.70 9.90
CAH 0.75 9.98
CAH 1.36 9.67
CAH 2.39 9.33
CAN 0.70 9.74
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

CAN 1.11 9.47

CAN 1.35 9.35
Cellulose 14.02
Polyethylene 8.00

The biodegradability of cellulose esters is limited by the ability of

cellulolytic enzymes to recognize cellulose derivatives as degradable
polysaccharides. Cellulase enzyme-biodegradability (CEB) was probed using a
series of LCCEs with variable substituent content and substituent size (19). Our
results suggest that the ability of cellulase enzymes to recognize cellulose esters
declines rapidly with both substituent content and substituent size (Fig. 6). A s
August 31, 2012 | http://pubs.acs.org

substituent size increases from C-4 to C-20, the DS at which a significant (i.e.,
greater than 20% glucose formation) degradation occurs decreases from 1.8 to 0.5.
This rate of decline is linearly related to the length of cellulose ester side chain (Fig.
6) (19).

Conclusions:

1. The synthesis of unusual cellulose esters is possible by homogeneous-phase

reaction chemistry using D M A c / L i C l as a solvent system, or by
heterogeneous reaction of cellulose with titanium (IV) isopropoxide as
catalyst and D M A c as diluent.
2. Homogeneous-phase reactions may involve carboxylic anhydrides as well as
acyl chlorides and a series of mixed anhydrides. These may involve the
traditional (impellers) of Clarke and Malm (i.e., chloroacetyl or
methoxyacetyl) or 4-pyrrolidino pyridine (in combination with DCC) or
tosyl chloride.
3. The new reaction systems afford greater reagent efficiency, greater
uniformity, and greater product control via stoichiometric reagent mixing.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 57

% Degradability (CEB)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Slope: d(D)/d(DS)
August 31, 2012 | http://pubs.acs.org

Figure 6. Cellulolytic enzyme biodegradability (CEB) of cellulose esters in relation

to DS (A); and the relationship between the slope of CEB/DS vs.
substituent size in number of carbons per acyl substituent (Β) (From réf.
19).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 58

4. Cellulose ester structure dramatically influences solubility, flexural modulus,

melt viscosity, thermal transitions, and enzyme recognition.
5. Cellulose esters with variable substituent content and substituent type, and
cellulose mixed esters, represent a novel class of engineered biobased
materials with potential use in biodegradable and biocompatible
thermoplastic materials.
6. Relationships of certain physical properties (e.g. flexural modulus, melt
viscosity) with easily calculated solubility parameters have been established
for cellulose esters over a broad range of ester chain-length and DS. The
materials scientist can now design a cellulose ester whose properties (those
related to solubility parameter) may be predicted with confidence before
preparation in the lab.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

APPENDIX

Experimental Evaluation of Mechanical Properties

To identify the mechanical properties of the LCCEs, larger quantities of

several of the cellulose ester derivatives were pelletized on a two-roll mill. After
the compounded pellets were dried at 80°C for 16 h, test specimens were molded
on a Newbury molding machine using a melt temperature of 220°C and a mold
temperature of 35°C. The C A H (DS 0.75) was compounded with 5% or 10%
hex

di(2-ethylhexyl)adipate (DOA) plasticizer. The other two C A H materials ( D S hex

2.39 and D S 136) were compounded without adding plasticizer. The C A N

h e x

(DS 1.35) was split into two halves for compounding with either 0% or 5%
August 31, 2012 | http://pubs.acs.org

non

D O A plasticizer. In addition, data are also shown for several commercial cellulose
acetate (CA), cellulose acetate propionate (CAP) and cellulose acetate butyrate
(CAB) materials. The C A material was plasticized with DEP and the C A P
materials (DSp, 0.73 and D S 2.60) were plasticized with D B P and D O A ,
pr

respectively. The C A B was also plasticized with DOA. (The total DS of all these
materials was 2.7.) Physical properties were measured by standard A S T M methods
using standard conditions. Dynamic Mechanical Thermal Analysis (DMTA) was
run on a Polymer Laboratory D M T A in flexure at 1 Hz with a heating rate of
4°C/min. Creep modulus was measured after 200 h of flexural creep at 1000 ρ si
applied load and at 23°C.
Melt rheology was performed on a Rheometrics Dynamic Analyzer (RDA)
in a nitrogen blanket. Frequency scans were taken between 1/s and 1/100 s. The
viscosities were all measured at 220°C, which is a reasonable processing
temperature for cellulose esters, sufficiently below the temperature at which
significant degradation occurs (~260-280°C). Table IV lists melt viscosity at 1/s
and 1/100 s, as well as melt viscosities at 1/100 s normalized to a molecular weight
of 100,000. Because melt viscosity is influenced by molecular weight in addition to
ester type, an attempt to filter out the effect of molecular weight was performed by

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 59

normalizing all the data to a molecular weight of 100,000 using the relation η C *
3 4
Mw (20). Prior investigations using a variety of cellulose esters (CAP, C A B and
CA) had demonstrated a strong linear relation between plasticizer content and
log(viscosity); this relation was then applied to the plasticized CA, C A P and C A B
materials to obtain values of melt viscosity for these materials at 0% plasticizer
content. These extrapolations are shown in Figure 7, and also listed in Table IV.
The flexural modulus values for unplasticized CA, CAP and C A B shown in Table
IV were also generated by extrapolating known values back to the 0% plasticizer
condition, as shown in Figure 8.

Literature Cited
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

1. Malm, C.J.;Mench, J. W.; Kendall, D. L.; Hiatt, G. D. Ind. Eng. Chem.

1951, 43, 684-688.
2. Clarke, H. T.; Malm, C. J. U. S. 1,880-808, 1932.
3. Morooka, T.; Norimoto, M.; Yamada, T.; Shiraishi, N. J. Appl. Poly. Sci..
1984, 29, 3981-3990.
4. Malm,C.J.;Mench, J. W.; Kendall, D. L.; Hiatt, G. D. Ind. Eng. Chem.
1951, 43, 688-6921.
5. Kwatro, H. S.; Caruthers, J. M.; Tao, Β. Y. Ind. Eng. Chem., Res. 1992, 31,
2647-2651.
6. Tao, Β. Y., private communication.
7. Turbak, A. F.; El-Kafrawy, Α.; Snyder, F. W.; Auerbach, A. B. U. S. Patent
4,302,252, 1981.
8. McCormick, C. L. U. S. Patent 4,278,790, 1981.
August 31, 2012 | http://pubs.acs.org

9. Dawsey, T. R.; McCormick, C. L. Rev. Macromol. Chem. Phys. 1990, C30,

405-420.
10. Diamantoglou, M.; Kuhne, H. Das Papier 1988, 42, 690-696.
11. Samaranayake, G., W. G. Glasser. Carbohydrate Polymers, 1993, 22, 1-7.
12. Samaranayake, G., W. G. Glasser. Carbohydrate Polymers, 1993, 22, 79-
86.
13. Glasser, W. G., G. Samaranayake, M. Dumay, V. Dave, J. Polym. Sci.: Pt.
B, 1995, 33, 2045-2054.
14. Sealey, J. E., G. Samaranayake, J. G. Todd, W. G. Glasser. J. Polym. Sci.:
Pt. B, 1996, 34, 1613-1620.
15. Shimizu, Y., J. Hayashi, Cell. Chem. Technol., 1989, 23, 661.
16. Haslam, E., Tetrahedron, 1980, 30, 2409-33.
17. Touey, G. P.; Tamblyn, J. W., 1961, U. S. Patent 2,976,277.
18. M. M. Coleman, C. J., Serman, D. E. Bhagwager, and P. C. Painter Polymer,
1990, 31, 1187.
19. Glasser, W. G., B. McCartney, G. Samaranayake. Biotech. Progr. 1994, 10,
214-219.
20. U. W. Gedde, "Polymer Physics," 1995, Chapman & Hall, London.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 60
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch003

Plasticizer Content (%)

Figure 7. Normalized viscosities of plasticized cellulose esters extrapolated back to

the zero plasticizer condition.

5000 -τ
August 31, 2012 | http://pubs.acs.org

10 20 30
Plasticizer Content (%)

Figure 8. Flexural moduli of plasticized cellulose esters extrapolated back to the

zero plasticizer condition.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 4

Induced Phase Separation: A New Synthesis Concept

in Cellulose Chemistry

1
T. Liebert and Thomas Heinze

Institute of Organic Chemistry and Macromolecular Chemistry, University of Jena,

Humboldstrasse 10, D-07743 Jena, Germany
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

A new synthesis concept in cellulose chemistry is described. Cellulose

dissolved in the solvent system N,N-dimethylacetamide/LiCl as well as
organo-soluble cellulose derivatives of different hydrolytic stability
(dissolved in dimethyl sulfoxide) were converted into cellulose ethers of
unusual substituent patterns after an induced phase separation with
solid NaOH particles. As confirmed by means of HPLC analysis, these
cellulose ethers contain a significantly higher amount of both 2,3,6-tri­
-O-functionalized and unsubstituted units in the polymer chain than
those obtained by conventional etherification reactions in cellulose
slurries. Moreover, the cellulose ethers show a preferred
functionalization of O-6 within the modified anhydroglucose units as
1
revealed by means of H-NMR-spectroscopy.
August 31, 2012 | http://pubs.acs.org

Cellulose is a very uniform homopolymer composed of D-anhydroghicopyranose units

(AGU) which are linked together by β-1->4 grycosidic bonds. Based on this unique
naturally occurring structure it is challenging to develop synthesis paths for the design
of advanced cellulosic materials usable for self-organizing and controlled interactions
in solutions.
Recent research in cellulose chemistry is in particular directed to the
development of synthesis capabilities for a regiocontrolled introduction of functional
groups in the polymer (7). The majority of methods studied so fer are based on the
difference in reactivity of the primary and secondary O H groups within the A G U .
Consequently, protective group techniques have found considerable interest.
Especially the bulky triphenylmethyl moiety (2,3) was established to protect the
primary positions. Recently, trialkylsilyl derivatives of cellulose have been synthesized
and it was found that the activated polymer (ammonia-saturated dipolar aprotic
solvent) converted under heterogeneous conditions yields a 6-0-thexyldimethylsUyl
cellulose of remarkable high uniformity (see chapter 1). The free OH-groups of these
protected polymers could be modified to regioselectivery substituted ethers and esters

'Corresponding author.

©1998 American Chemical Society 61

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 62

and the obtained deblocked products show differences in properties in comparison

with conventionally prepared samples (4,5).
Besides the controlled distribution of functional groups on the level of the
anhydroglucose unit (AGU) a more complex problem is the regioselective conversion
of segments of the polymer chains as well as the analytical determination of such
structures. Recently, a model consisting of oUgoacetylcelhilose- and oligodihexanoyl-
chitin blocks was synthesized using non-glycosidic urethane linkages (6). Poly-
saccharide block copolymers containing glycosidic linkages only are expected to
possess unique properties.
In general, cellulose reactions are carried out with dissolved reagents and
highly swollen or dissolved (so far in lab-scale synthesis only) cellulose polymers to
avoid both concentration and accessibility gradients. Consequently, no differences in
reactivity of the chemically equivalent glucose repeating units exist. Thus, it is
necessary to establish new synthesis concepts employing segments of different
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

reactivity within the polymer chains of the polysaccharide.

We introduce a new and general approach for the preparation of cellulosics
with an unconventional distribution of substituents along the polymer chains. This
approach is based on an induced high concentration gradient (phase separation), i.e.
the formation of specific areas of reactivity (regioselectivity) within the polymer
chains. The results of structure determination studies are compared with those of
conventional products.

Results and Discussion

Today large-scale production of cellulose ethers like the most important ionic one,
carboxymethyl cellulose (CMC), is exclusively carried out by slurry processes, i.e. by
conversions of alkali cellulose swollen in an organic liquid and aqueous NaOH with an
August 31, 2012 | http://pubs.acs.org

appropriate etherifying agent. In the case of C M C it was revealed by means o f l 3 -

C

and ÎH-NMR spectroscopy of hydrolytically degraded samples that the carboxymethyl

functions are distributed in the order 0-2 > 0-6 > 0-3 within the A G U (7,8).
Although a time-consuming mathematical processing of the ^ C - N M R spectra reveals
the monomer composition, i.e. the molar ratio of all differently modified units
(anhydroglucose unit, A G U ; 2-, 3-, and 6-mono-OAGU; 2,3-, 2,6-, and 3,6-di-O-
A G U as well as 2,3,6-tri-O-AGU), more appropriate methods have to be developed to
gain these information. Using a convenient HPLC method we confirmed for a broad
variety of C M C that by the conventional slurry process no significant deviation from a
statistical pattern of fiinctionalization is achievable (9). This rapid and convenient
method represents an inalienable prerequisite for studying the influence of synthesis
conditions on the amount of the different monomer units. It was useful to determine
the un-, mono-, di- and tricarboxymethylated units only, i.e. to neglect the different
positions of both the mono- and dicarboxymethylated units. Figure 1 shows a typical
HPL chromatogram as well as the assignment of the peaks.
It seems appropriate to control the fiinctionalization patterns within the
cellulose chains by using differences in reactivity due to their accessibility gradients
caused by the well known supramoleculare structure. To activate the polymer and to
initiate the etherification reaction it is common to use aqueous sodium hydroxide
solutions. Dependent on the concentration of alkali, the cellulose swells to various
extents and yields a more or less decrystallized polymer. That means the degree of

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 63

activation (accessibility) may be controlled by the alkali concentration and thereby a

control of the content of the different monomelic units might be possible. In a first
series of experiments, cellulose 1 (spruce sulfite pulp) slurried in isopropanol was
converted with monochloroacetic acid after activating the polymer with aqueous
NaOH in the concentration range from 5 to 30 % (m/v). Regarding the results (Table
I), it has to be underlined that the total degree of substitution (DSÇMC) reached a
maximum value of 1.24 at a concentration of 15 % (m/v) aqueous NaOH. In order to
analyze the mole fractions of the monomelic units, the CMCs 6a-f were degraded
using HCIO4. Subsequent removal of most of the perchloric acid as KCIO4 after
precipitation with K O H gives solutes which were analyzed directly by HPLC. A
separation on a polystyrene-based strong cation-exchange resin resulted in the desired
mole fractions of unsubstituted-, mono-, di-, and tricarboxymethylated glucose units
(see Figure 1). The results obtained are graphically displayed as a function of the total
DSCMC in Figure 2. The curves (shown in Figure 2 as well) were calculated on the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

basis of a statistical model (bionominal distribution) for the arrangement of

substituents in cellulose derivatives (9, JO). Obviously, the determined mole fractions
are in good agreement with the statistical model independence of the aqueous NaOH
concentration used (see Experimental). That means that the totally heterogeneous
carboxymethylation is mainly determined by statistics even at a low activation.

0
Table L Degree of substitution (DSHPLC) * carboxymethyl cellulose dependent
on concentration of aqueous NaOH (reaction of cellulose 1 in isopropanol and
aqueous NaOH with monochloroacetic acid for 5 h at 55 °C)

NaOH concentration Carboxymethyl cellulose

August 31, 2012 | http://pubs.acs.org

(%, w/v) No. DSHPir

5 6a 0.59
8 6b 0.93
10 6c 1.00
15 6d 1.24
20 6e 1.03
30 of 0.95
a
see Experimental

Thus, new synthesis concepts had to be established employing segments of

different reactivity within the polymer chain.
In a recent research project on carboxymethylation of cellulose it was observed
that a treatment of cellulose dissolved in the solvent system A^A^-dimethylacetamide
(DMA)/LiCl with solid NaOH-particles (size about 1 mm) suspended in D M A induces
a phase separation (77). The gel particles formed can be isolated by solidification with
diethyl ether. From deconvolved FTIR-analysis of the gel particles it was concluded
that cellulose II is regenerated. The phase separated systems were converted with
monochloroacetic acid resulting in C M C with unconventional properties. While C M C
samples, prepared via the conventional totally heterogeneous path in

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 64

Inorganic salts

I Mono-O-CM-gIc
A
DI-O-CM- η
glc

ο
α
CO

1
12 14 16 18 20 22 24
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

Retention time (min)

Figure 1. HPLC analysis of the carboxymethyl cellulose sample 6d after hydrolysis

with dilute HCIO4.
August 31, 2012 | http://pubs.acs.org

i · 1 • 1 ' 1 ' Γ—Η

0,50 0,75 1,00 1,25 1,50

D S
HPLC

Figure 2. The mole fractions of repeating units ( • , glucose, O , mono-0-

carboxymethyl-, Δ, di-0-carboxymethyl-, and V , 2,3,6-tri-O-carboxymethylated
glucose) in hydroryzed C M C samples (CMCs were synthesized via the
conventional slurry process) plotted as function of DS^PLC- dependent on the
aqueous NaOH concentration. The curves are calculated as described in the text,
see ref (9) as well.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 65

A ) 0 U T w a t e r
isopropanol/aqueous NaOH, dissolve already at a DSCMC °^ ' ^ » the
samples synthesized in D M A / L i C l are water-soluble not until 1.5 (72). Representative
results are listed in Table Π. A n analysis of the C M C polymers by means of HPLC
after degradation with HCIO4 reveals a significant deviation of the amounts of the
monomelic units with values calculated (see Figure 3). These CMCs contain a
significantly higher amount of both tricarboxymethylated and unsubstituted units than
those obtained in a slurry of cellulose in isopropanol/aqueous NaOH at comparable
D ^ C M C values, Le. a non statistic distribution of the monomelic units occurs.
To confirm that these unconventional patterns of functionalization are due to
the induced phase separation before the chemical conversion, various cellulose
intermediates resp. derivatives of different hydrorytic stability were reacted under
comparable water-free conditions. Solutions of cellulose trifluoroacetate 2, prepared
15 D P 4 6 0
with trifluoroacetic acid/trifluoroacetic anhydride, DSÇJFA >
2 2 D P 2 6 0
cellulose formate 3, prepared with formic acid/POCli 5(OH)i 5; DSCF · >
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

8 2 2 a s w e a s
(77), commercial cellulose acetate 4 (DSÇA 1· > DP ° ) ^ trimethylsilyl
(TMS) cellulose 5, prepared with TMS chloride after activation with NH3, DS 1.1,
D P 220 (14), in DMSO (5.7 %, w/v, polymer) were treated with solid NaOH particles
suspended in dimethyl sulfoxide (DMSO).
In any case, phase separation and gel formation upon addition of the NaOH
particles was observed. An interesting way to study the phase behavior during this first
stage of reaction was found to be the application of polarized-light microscopy. While
the solutions of the cellulosics mentioned represent a homogeneous system, after
addition of the NaOH/DMSO suspension a growth of crystals (sodium salts of
trifluoroacetic-, formic- and acetic acid, respectively) was observed. The regenerated
cellulose Π is not detectable as a polymeric particle. Consequently, the polymer has to
be fixed mainly on the solid NaOH. The phase separated systems formed were allowed
to react with monochloroacetic acid. Typical experimental data and analytical results
August 31, 2012 | http://pubs.acs.org

o r % 2 2 w a s r e a c n e
are listed in Table Π. It can be seen that a maximum DSÇMC d even
in an one step synthesis. The samples obtained become water soluble starting from 1.5.
A number of representative data of the mole fractions determined by means of HPLC
are graphically displayed as a function of the total DSflpjx (Figure 3). As can be
concluded from the comparision with the values calculated according to the binominal
distribution, the polymers consist of higher amounts of glucose and
tricarboxymethylated anhydroglucoses and lower amounts of mono- and
dicarboxymethylated anhydroglucoses. The deviation from the statistics is exact in the
same range as found for CMCs prepared in DMA/LiCl. Consequently, the induced
phase separation and subsequent carboxymethylation yield samples with a non-
statistical distribution of substituents within the polymer chains.
An important question was, how the system behaves i f the alkali becomes
mobile in the reaction mixture by addition of an appropriate sovent like e.g. water. For
this purpose a cellulose acetate (DS 0.8) was converted in the above mentioned
manner with solid NaOH (20 mol/mol modified A G U ) and monochloroacetic acid (10
mol/mol modified A G U ) but in the presence of 1 % (v/v) of water. Results of the
HPLC analysis are shown in Figure 4 as absolute deviation from the statistic values.
The CMCs obtained exhibit a much higher DS (2.1) compared to a synthesis under
water-free conditions as well as a strictly statistical pattern of substitution. It is
believed that this observation is due to the fact that the water in the system is able to
partially dissolve the solid NaOH and does thereby level out the phase separation

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

,0
0,0 0,5 1,0 1,5 2,0 2,5 3,0

D S
H P L C

Figure 3. The mole fractions of repeating units ( • , glucose, O , mono-0-

carboxymethyl-, Δ, di-0-carboxymethyl-, and V, 2,3,6-tri-O-carboxymethylated
glucose) in hydrolyzed C M C samples (CMCs were synthesized via induced phase
e c u r v e s a r e
separation) plotted as function of DSHPLO T h calculated as
described in the text, see réf. (9) as well.
August 31, 2012 | http://pubs.acs.org

Glc MCMG DCMG TCMG

Figure 4. Influence of water on the absolute deviation of the mole fractions of
glucose, mono-0-carboxymethyl-, di-(3-carboxymethyl-, and 2,3,6-tri-O
carboxymethylated glucose from the binomial distribution. The C M C were
synthesized starting from cellulose acetate 4 via induced phase separation. ^
water-free (DSHP 0.45), ^ 1%, v/v water (DSHP 1.70).
LC LC

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 67

induced concentration gradient. It has to be mentioned that this effect is less drastic in
case of more hydrophobic cellulose intermediates like, e.g., TMS-cellulose. The TMS-
cellulose dissolved in DMSO yields products of significant deviation of the amount of
the mole fractions even in the presence of 1 % of water.
The addition of phase transfer catalysts does not influence the deviation of the
amount of the monomelic units from the statistics. Carboxymethylation via phase
separation starting from a TMS-cellulose (DS 1.1) in the presence of dibenzo-18-
crown-6 yields a polymer of DSÇMC 1-^6 and with a significant deviation from
statistically calculated mole fractions.
It was important to investigate the influence of the size of the NaOH particles.
The results obtained with CTFA 2 dissolved in DMSO (5.7 %, w/v) are summarized in
Figure 5. The phase separation and carboxymethylation reactions were carried out
under equal conditions. By decreasing the size from < 1 mm to < 0.63 mm to < 0.25
mm the total DSÇMÇ values increased from 0.62 to 0.97 to 1.12. The mole fractions
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

of the monomelic units determined by means of HPLC show that a deviation from the
statistic values occurs as expected. However, as can be seen from Figure 5, the
n e
decreasing particle size has mainly an influence on the total DSÇMC- ^ influence on
the amount of the monomelic units within the chains is rather small.
Carboxymethylation reactions via the cellulose intermediates mentioned of
appropriate DS in unpolar solvents does not yield CMCs. Thus, the conversion of
TMS-cellulose (DS 2.8, DP 460) dissolved in methylene chloride with dispersed solid
NaOH (20 mol/mol modified A G U ) and monochloroacetic acid (10 mol/mol modified
A G U ) under reflux for 16 h gives practically no C M C (DSÇMC determined by means
of HPLC are less than 0.01).
Besides the determination of the amounts of monomelic units that built up the
polymer, the distribution of carboxymethyl groups within the A G U was a goal of our
studies. For this purpose the polymers were degraded with deuterated sulfuric acid and
August 31, 2012 | http://pubs.acs.org

the corresponding hydrolysates were analyzed by means of *H-NMR spectroscopy.

The partial degree of substitution calculated from the spectra indicates a substitution in
the order 0-6 > 0-3 > 0-2 in case of samples synthesized starting from CTFA and
CF. The CMCs prepared in D M A / L i C l possess a distribution 0-6 > 0-2 > 0-3. In
comparison the carboxymethyl groups of conventionally prepared samples are
distributed in the order 0-2 > 0-6 > 0-3. Consequently, the phase separation process
does not just effect the regioselectrvity on the level of the polymer chain. It has also an
remarkable effect on the distribution of functional groups on the level of the
monomelic unit.
An important feature due to the new synthesis concept using induced phase
separation is the application to other etherification reactions. Preliminary experiments
of methylation and ethylation prove the usefulness. For example the conversion of a
cellulose acetate (DS 0.8) dissolved in DMSO (5.7 %, w/v, polymer) with methyl
iodide (10 mol/mol modified AGU) after induced phase separation with solid NaOH
1 1 2 e
(20 mol/mol modified AGU) yields a methyl cellulose of a DS^ethyl °f · ^
product was analyzed by means of HPLC after degradation with trifmoroacetic acid
(75). The absolute deviation of the amount of mole fractions of unsubstituted and
methylated glucoses from calculated values according to the binominal distribution is
shown in Figure 6. It is obvious that the deviation is comparable with results obtained
from C M C synthesized via phase separation. Additionally, analytical results of a
commercial methyl cellulose (DS 1.38) and of a methyl cellulose (DS 1.85) prepared

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 68
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

Glc MCMG DCMG TCMG

Figure 5. Influence of the NaOH particle size on the absolute deviation of the
mole fractions of glucose, mono-0-carboxymethyl-, di-O-carboxymethyl-, and
2,3,6-tri-(9-carboxymethylated glucose from the binomial distribution. The C M C
were synthesized starting from cellulose trifluoroacetate 2 via induced phase
separation. ^ 1.00 - 0.63 mm (DSHP 0.64), ^ 0.63 - 0.25 mm (DSHPLC
LC

12
0.97), Ο < 0.25 mm (DSHPLC 1- )
August 31, 2012 | http://pubs.acs.org

Glc MMGIc DMGIc TMGIc

Figure 6. Absolute deviation of the mole fractions of glucose, mono-0-methyl-,
di-O-methyl-, and 2,3,6-tri-O-methylated glucose from the binomial distribution
of methyl celluloses, |3 synthesized from TMS cellulose 5 via induced phase
separation ( D S 1.12), ^ from C A 4 via induced phase separation
M e t h v l
DS 4 5
( Methyl ° · λ HT commercial sample ( D S i 1.38), and |fll| synthesized via
M e t h y

induced phase separation in the presence of 1 % (v/v) of water ( D S h i 1.85).

Met y

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 69

via phase separation, however with 1 % water in the reaction mixture are included in
Figure 6 as well. Both do not exhibit significant deviations from the calculated values.
It is worth mentioning that the water leads to a sample of higher total D S ] y f y i as
em

already found for the carboxymethylation.

Conclusion

It was shown that induced phase separation processes are useable for the preparation
of cellulose products with a non statistic distribution of functional groups within the
polymer chains. These samples already show unexpected properties. Thus, CMCs
prepared via phase separation dissolve in water starting from DS 1.2 ... 1.6. In
comparison conventionally prepared samples are water soluble at about 0.4. On the
other hand, solutions of samples described here show a comparable lower decrease in
viscosity upon an addition of electrolytes (72).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

The validity of the synthesis concept was shown recently for the preparation of
methyl starches. Products synthesized starting from a solution of the polymer and
subsequent phase separation yields methyl starch with a block-like structure as
concluded from FAB-MS analysis and statistical calculations (16). Moreover,
Miyamoto et al. have shown that a methyl cellulose prepared via phase separation
possess a very different water solubility as well as thermotropic gel formation tendency
compared with a sample prepared by the so-called alkali cellulose process, which is a
totally heterogeneous reaction (77).
In contrast to common knowledge that the etherification of cellulose in the
solvent system D M A / L i C l has no particular advantage over conventional
heterogeneous processes for cellulose ether production (18) our results show that new
cellulosic polymers with both unconventional distribution of functional groups and
properties may be designed. A first indication for such alternative molecular structures
August 31, 2012 | http://pubs.acs.org

was already found by us in 1994 (9).

Besides the necessary more detailed elucidation of the structural features that
are under progress, a very helpful tool for the understanding of the reaction
mechanism would be a mathematical model suitable for the simulation of non
statistical distributions of substituents along the polymer chains. One approach is the
establishment of a series of rate laws and a mathematical handling that yield equations
useable for the calculation of the mole fractions of repeating units dependent on the
rate constants. By iteration and fitting of the analytical data with the calculated values,
essential drawbacks concerning the mechanism of the reaction path are possible. These
results will be published elsewhere.

Experimental

Materials. Spruce sulfite pulp, Λ^Λ^-dimethylacetamide (DMA), NaOH,

monochloroacetic acid and LiCl were purchased from F L U K A . D M A was dried over
CaH and distilled under reduced pressure. A l l other chemicals were used for the
carboxymethylation after drying at 105°C for 5 h in vacuum
The cellulose intermediates were synthesized and analyzed according to (73):
cellulose trifluoroacetate, (77): cellulose formate, (14): trimethylsilyl cellulose.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 70

Carboxymethylation of cellulose in isopropanol/aqueous NaOH. The

carboxymethylation was carried out by a standard solvent method. 5 g of air-dry
cellulose 1 (spruce sulfite pulp) in 150 ml isopropanol was stirred vigorously, while
13.3 ml of a 5 to 30 % (w/v) aqueous NaOH (see Table I) was added dropwise during
10 min at room temperature. Stirring was continued for 1 h and 6 g of
monochloroacetic acid was then added. The mixture was placed for 5 h on a water
bath at 55 °C with stirring. The mixture was filtrated, suspended in 300 ml of aqueous
methanol and neutralized with dilute acetic acid. The product (CMCs 6a-f) was
washed three times with 80 % aqueous ethanol and with ethanol and dried at 60 °C.
DScMCa r e Φ*® 1m T a W e 1
Π* (KBr): 1630, 1410 cm" (C=0, carboxylate group).

Carboxymethylation of cellulose in DMA/LiCl. For a typical preparation, 1 g of

cellulose 1 (spruce sulfite pulp) and 60 ml DMAc was kept at 130 °C for 2 h under
stirring. After the slurry was allowed to cool to 100 °C, 3 g of anhydrous LiCl were
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

added. By cooling down to room temperature under stirring the cellulose dissolved
completely. After standing overnight, a suspension of pulverized NaOH (4 - 10
mol/mol A G U ) in 20 ml D M A and a suspension of monochloroacetic acid (2-5
mol/mol A G U ) in 20 ml D M A were added under vigorous stirring. The temperature
was raised to 70°C. After 48 h reaction time the mixture was cooled to room
temperature and was precipitated into 300 ml ethanol. The precipitates were filtered
of£ suspended or dissolved in 75 ml distilled water, neutralized with acetic acid and
reprecipitated into 300 ml ethanol. After filtration the products were washed with
ethanol and dried in vacuum at 50°C.
DScMC g™ a r e 011 m T a D l e 1
Π, IR (KBr): 1630, 1410 cm" (C=0, carboxylate group).

Carboxymethylation of a cellulose intermediates 2-5, typical example. 1 g of the

intermediate (2-5) was dissolved in 17.5 ml DMSO under nitrogen. A suspension of
August 31, 2012 | http://pubs.acs.org

dried pulverized NaOH (10-40 mol/mol modified A G U ; dried under vacuum at 45 °C,
5 h; see Table II) in DMSO (2.75 ml per g NaOH) was added to the solution within 10
minutes, followed by sodium monochloroacetate (5-20 mol/mol modified A G U , dried
under vacuum at 45 °C, see Table II) under vigorous stirring. The temperature was
raised to 70 °C. After various reaction times (Table II) the reaction mixture was
cooled to room temperature and precipitated into 75 ml methanol. The precipitate was
filtered oflÇ dissolved or suspended (in dependence on the D S ç ^ ç ) in water,
neutralized with acetic acid, and reprecipitated into 100 ml of 80 % (v/v) aqueous
ethanol. The products obtained are summarized in Table II.
D S
CMC a r e
Φ®1
m T a b l e
Π, IR (KBr): 1620, 1410 cm-1 (C=0, carboxylate group).

Measurements. The HPLC analysis of the C M C was carried out as described in ref.
(9). However, the samples were hydrolysed with perchloric acid. 0.1 g of C M C were
dispersed in 2 ml HCIO4 (70 %) and after 10 min at room temperature dilute with
18 ml distilled water. This mixture was kept at 100 °C for 16 h. The solution obtained
was carefully neutralized with 2 M K O H and keept at 4°C for 1 h to guarantee a
complete precipitation of the KCIO4. The salt was filtered off and washed three times
with distilled water. The obtained solution was reduced to approximately 3 ml and
dilute with distilled water to give exactly 5 ml sample. A Jasco HPLC equipment with
+
two Bio-Rad Aminex HPX-87 columns ( H form) was used. The ^H-NMR analyses
were carried out according to ref. (//).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 71

Table Π. Conditions and results of carboxymethylation of cellulose (1) dissolved

in 7V,7V-dimethylacetamide (DMA)/LiCl as well as cellulose trifluoroacetate
(CTFA, 2), cellulose formate (CF, 3), cellulose acetate (CA 4), and trimethylsilyl
9

cellulose (TMSC, 5) via induced phase separation with NaOH particles (size <
0.25 mm).

Staring cellulosic Molar Reaktion Carboxymethyl cellulose

a c
material ratio time No DStn>LC Solubility
(h)b in water

Cellulose (1) in 1:2:4 48 7a 1.13 -

DMA/LiCl 1:4:8 48 7b 1.88 +
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

1:5:10 48 7c 2.07 +
CTFA (2) 1:5:10 2 8a 0.11 -
1:10:20 4 8b 1.86 +
1:10:20 16 8c 1.54 +
l:10.20 d
4 8d 1.36 -
1:20:20 e
2 8e 0.62 -
l:10:20 f
2 8f 0.97 -
CF (3) 1:10:20 2 9a 1.46 +
1:10:20 4 9b 1.91 +
August 31, 2012 | http://pubs.acs.org

1:15:30 4 9c 1.36 -
1:20:40 2 9d 2.21 +
CA(4) 1:10:20 2 10a 0.36 -
1:10:20 4 10b 0.45 -
TMSC (5) 1:10:20 0.5 11a 2.04 +
1:10:20 1 lib 1.91 +
1:10:20 2 11c 1.97 +

a
Molar ratio: Modified anhydroglucose unit (AGU) : C l C H C O O H (Na) : NaOH.
2

b Reaction temperature 70 °C.

c 1
DSjjPLC degree of substitution determined by means of HPLC (see ref. 9).
d First addition of ClCF^COONa and subsequent phase separation with solid
NaOH particles.
e
NaOH particle size: 0.63- 1.00 mm
f
NaOH particle size: 0.25 - 0.63 mm

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 72

Acknowledgment

The general financial support of the German 'Oeutsche Forschungsgemeinschaft

(DFG; Schwerpunktprogramm Cellulose)" is gratefully acknowledged. We would like
to thank the "Stifterverband fur die Deutsche Wissenschaft" for generous financial
support.

Literature Cited

1. Klemm, D.; Stein, Α.; Heinze, Th.; Philipp, B.; Wagenknecht, W. In Polymeric
Materials Encyclopedia: Synthesis, Properties and Applications; Salamone,
J.C. Ed.; CRC Press, Inc., Boca Raton, USA, vol. 2, 1996, 1043-1053.
2. Harkness, B.R.; Gray, D.G. Macromolecules 1991, 24, 1800.
3. Heinze, Th.; Röttig, Κ.; Nehls, I. Macromol. Rapid Commun. 1994, 15, 311.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch004

4. Kamide, K.; Saito, M . Macromol. Symp. 1993, 83, 233.

5. Klemm, D.; Heinze, Th.; Wagenknecht, W. Ber. Bunsenges. Phys. Chem.
1996, 100, 730.
6. Kadakowa, J.-I.; Karasu, M.; Tagaya, H.; Chiba, K. J.M.S.-Pure Appl. Chem.
1996, A33, 1735.
7. Baar, Α.; Kulicke, W.-M.; Szablikowski, K.; Kiesewetter, R. Macromol.
Chem. Phys. 1994, 195, 1483.
8. Reuben, J.; Conner, H.T. Carbohydr. Res. 1983, 115, 1.
9. Heinze, Th.; Erler, U . ; Nehls, I.; Klemm, D. Angew. Makromol. Chem. 1994,
215, 93.
10. Spurlin, H . M . J. Am. Chem. Soc. 1939, 61, 2222.
11. Liebert, T.; Klemm, D.; Heinze, Th. J.M.S.-Pure Appl. Chem. 1996, A33. 613.
12. Heinze, Th.; Heinze, U . ; Klemm, D. Angew. Makromol. Chem. 1994, 220,
August 31, 2012 | http://pubs.acs.org

123.
13. Liebert, T.; Schnabelrauch, M.; Klemm, D.; Erler, U . Cellulose, 1994, 249.
14. Klemm, D.; Stein, A. J.M.S.-Pure Appl. Chem. 1995, A32. 899.
15. Erler, U . ; Mischnick, P.; Stein, A . ; Klemm, D. Polymer Bull. (Berlin), 1992,
29, 349.
16. Mischnick, P.; Kühn, G. Carbohydr. Res. 1996, 290, 199.
17. Miyamoto, T.; Donkai, N.; Nishimura, H . Proceedings of "Japanese-German
seminar on future developments of polysaccharides. Fundamentals and
applications", 1996, Hokkaido University, Sapporo, Japan, p.35.
18. Dawsey, R.D. In Polymer Fiber Sci.: Recent Adv.; Forners, R.E., Gilbert,
E D . , Ed.; V C H , New York, USA, 1992; pp. 157-176.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 5

Methods for the Selective Oxidation of Cellulose:

Preparation of 2,3-Dicarboxycellulose and
6-Carboxycellulose

1 1 2
A. C. Besemer , A. E. J. de Nooy , and H. van Bekkum

1
TNO Nutrition and Food Research Institute, P.O. Box 360, 3700 AJ, Zeist,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

Netherlands
2
Delft University of Technology, Julianalaan 136, 2628 BL, Netherlands

Three methods for the selective oxidation of cellulose are described. The
classical method consists of consecutive oxidation with sodium periodate,
leading to 2,3-dialdehyde cellulose and sodium chlorite, giving 2,3-
dicarboxy cellulose. This material, which is obtained in high yield and has a
high carboxylate content (7.6 mmol COONa/g; 90% of the theoretical
value), has a very good calcium sequestering capacity.
The second method is by oxidation of the substrate, dissolved in concen-
trated phosphoric acid with nitrite/nitrate, leading to the selective oxidation
of the substrate at the 6-position of the glucose unit. Generally, the yields are
August 31, 2012 | http://pubs.acs.org

higher than 80%, and the degree of oxidation is 80-90%. However, the reac-
tion is not completely specific, since some oxidation at the secondary
hydroxylic groups occurs. Borohydride reduction of the product restores the
diol configuration and also ß-elimination is avoided and thereby
depolymerization. Oxidation with sodium hypochlorite and bromide as a
catalyst and TEMPO as a mediator appears also to be applicable to cellu-
lose. Selectivity of oxidation at the 6-CH OH group is somewhat lower than
2

that obtained earlier for glucans like starch and pullulan. Products with a
degree of oxidation of 80% are obtained in 90% yield or higher.

Oxidation of (poly)saccharides has been studied in detail by numerous investi-

gators, but, because of the presence of several reactive groups, it is not easy to
attain high selectivity, and only a limited number of reagents are available for this
purpose (1). Because it is insoluble in water and most common organic solvents,
there are especially difficulties with cellulose.
In view of the structure of an anhydroglucose unit in glucans such as cellulose and
starch, one has to account for the presence of three reactive groups: one primary
and two secondary OH-groups. Usually the oxidation of the secondary hydroxyl

©1998 American Chemical Society 73

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 74

groups in glucans results in ring cleavage. Well-known methods for this conver-
sions are reactions with sodium periodate or with lead(IV) tetraacetate, which lead
to the formation of the corresponding 2,3-dialdehyde derivatives (2-4). Upon
subsequent oxidation of this material with sodium chlorite the corresponding
dicarboxy derivatives are obtained. The conversion of starch has been studied in
detail because the oxidation products have excellent calcium binding properties and
therefore may be used as a substitute for builders in laundry detergents (3,4). Good
results can also be obtained with cellulose. Floor et al. (4) improved the procedure
by using hydrogen peroxide in the second step. For the quantitative conversion of
dialdehyde derivatives 6 mol instead of 2 mol of sodium chlorite are required:

Dialdehyde polysaccharide + 2 NaC10 - Dicarboxy polysaccharide + 2 NaOCl

2
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

because the reaction product (NaOCl) decomposes sodium chlorite according to

NaOCl + 2NaC10 + H 0 - 2 C10 + NaCl + 2 NaOH.

2 2 2

Use of hydrogen peroxide has two advantages: less reagent is needed, since sodium
hypochlorite reacts faster with hydrogen peroxide than with sodium chlorite
according to

NaOCl + H 0 2 2 - NaCl + H 0 + 0
2 2

and, better products are obtained.

About twenty years ago the most attractive way to selectively oxidize polysac-
charides such as starch and cellulose at the 6-position of the anhydroglucose unit
August 31, 2012 | http://pubs.acs.org

was by exposing them to (gaseous) N 0 or N 0 (5). Satisfactory results can be

2 2 4

obtained with cellulose and starch; i.e., the corresponding 6-carboxy polysaccharide
can be prepared with a high carboxylate content and with a satisfactory yield.
An alternative, which has been studied by a few authors, is oxidation with nitrous
acid (6,7). In this system, the substrate is dissolved in concentrated phosphoric acid
and allowed to react with sodium nitrite. In the highly viscous solution a foam
develops in which various oxidizing species like N 0 and N 0 are present. A
2 3 2 4

drawback of these methods is that considerable depolymerization occurs and that

the degree of oxidation is no higher than approximately 80% (6).
Recently, we improved the latter oxidation and we developed a new method for
oxidation of primary alcohol groups in polysaccharides using hypochlorite as the
primary oxidant and bromide and 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO)
as the catalysts. In (8-11) we presented the results of this reaction with respect to
selectivity, depolymerization, and scope. Mainly water soluble-polymers, such as
starch, inulin, and pullulan, were investigated.
In this study we describe results obtained by applying these oxidation methods to
poorly water-soluble polymers such as cellulose. A comparison is made with some
model compounds.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 75

Results and discussion

a. Periodate/clorite/hydrogen peroxide oxidation of cellulose

Table I shows the results of the glycol cleavage of starch and cellulose, using 2 mol
of sodium chlorite and 2 mol hydrogen peroxide. For comparison we have also
presented data on the method in which 6 mol of sodium chlorite are used (and no
hydrogen peroxide). Generally, in polysaccharide oxidation, method A gives the
best results. However, it is seen that oxidation of cellulose according to method Β
can give somewhat better results.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

Table I. Dicarboxy polysaccharides obtained by sodium periodate/sodium

chlorite oxidation
b
Substrate Method 3
Yield(%) CC(%) SC (mmol Ca/g)

Cellulose A 93 82 2.20

Cellulose Β 91 86 2.29

Starch A 98 79 2.51

Starch Β 93 86 2.39

Amylose A 87 74 2.39

Amylose Β 95 80 2.29
August 31, 2012 | http://pubs.acs.org

a
Method A: 2 mol of sodium chlorite and 2 mol of hydrogen peroxide per anhydro-
glucose unit
Method B: 6 mol of sodium chlorite per anhydroglucose unit
b
SC = sequestering capacity defined as the mmol of Ca bound by 1 gram of
5
material until the concentration is below 10" M Ca(H). C C = carboxylate content
(at 100 % conversion the C C = 8.4 mmol COONa/g)

b. 6-carboxy cellulose (oxidation in phosphoric acid with sodium nitrate/nitrite)

The results of the oxidation of cellulose in the oxidation with sodium nitrate are
shown in Table Π. Generally, satisfactory results are obtained. Good results were
also obtained in the oxidation of amylose (yield 80%; carboxylate content 95%). It
can be concluded that this method is very suitable for the oxidation of poorly
soluble biopolymers in water. Yield and carboxylate content are high. Although
some may occur (see Figure 1), depolymerization appears to be very modest,
especially in comparison to former procedures (5) . From the kinetic experiments

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 76

(see Figure 2) it can be seen that nitrite has a catalytic effect. The reduction of nitric
acid by NO, leading to the formation of N 0 , is thought to play an important role.
2

An important advantage of this method over the TEMPO method is that no glycolic
oxidation can occur. Table ΙΠ shows some results of the oxidations of β-
cyclodextrin using the phosphoric acid/nitrate/nitrite method.

From the results presented in Table IV, it appears that only the stoichiometric
amount of nitrate is needed, according to the theoretical reaction equation

3 -CH OH + 4 H N 0
2 3 - 3 -COOH + 4NO + 5H 0, 2

indicating 1.33 N a N 0 per CH OH-group.

3 2
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

Table II. Oxidation of cellulose with sodium nitrate/sodium nitrite in

phosphoric acid
a
Substrate T(°C) Time DO (%) Yield(%)
(hours)
cellulose 20 3 23 77
cellulose 20 6 72 85
cellulose 20 9 88 82
cellulose 20 20 99 74
cellulose 4 6 <10 nd
August 31, 2012 | http://pubs.acs.org

cellulose 4 24 90 85
cellulose 4 40 97 88
a
DO = Degree of oxidation determined by the Blumenkrantz method (12).

It is important to note that when using only sodium nitrite, as in the literature
method (7), a much larger amount of NO will be produced per primary alcohol
group:

-CH OH + 4 H N 0 -
2 2 -COOH + 4 N O + 3H 0.
2

c. TEMPO-oxidation of starch, pullulan and cellulose

The results of the TEMPO-oxidation of cellulose and amylose are shown in Table
IV.

It is observed that the results with cellulose are not quite reproducible; i.e. yield and
carboxylate content vary to some extent, and seem to depend on some unknown
variables. No clear explanation exists for the fact that some material remains undis­
solved and that the composition varies. The fact that a yield of higher than 100%

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 1.0x10 4
1.0Χ10 5
1.0x1 o c
1.0x10'
Molecular weight
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

Figure 1. Molecular weight distribution of pullulan (right peak)

and completely oxidized samples at 4 °C (middle peak) and at
20 °C (left peak) as obtained by SEC-MALLS (oxidation method
sodium nitrate/sodium nitrite in phosphoric acid).

0.8
August 31, 2012 | http://pubs.acs.org

0 30 60 90 120
Time (min)

Figure 2. Influence of the initial concentration of sodium nitrite

on the reaction rate (measured with the Blumenkrantz assay).
Conditions: 0.70 g β-cyclodextrin and 0.70 g sodium nitrate
dissolved in 5 ml 85% phosphoric acid with 0 mg ( r), 20 mg (+),
40 mg ( *), 60 mg ( *) sodium nitrite, respectively.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 78

Table III. Degree of oxidation of β-cyclodextrin in the nitrate/nitrite oxidation

in relation to the nitrate/substrate molar ratio
a
Entry NaN0 /primary
3 DO(%)
alcohol(molar ratio)
1 0.27 17
2 0.54 41
3 0.81 56
4 1.10 68
5 1.38 77
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

6 1.93 80
a
Based on the assumption that under optimum conditions the degree of oxidation
per anhydroglucose unit is 80%.

can be obtained with a low uronic acid content of the materials points to the occur­
rence of a competing oxidation (glycol cleavage). Although the oxidation has not
been studied in detail, evidence exists that the reagent sodium hypochlorite should
be added gradually to avoid undesired reactions, such as glycol cleavage, due to the
presence of excess sodium hypobromite (13).
To overcome this complication, three options should be considered for better
results:
August 31, 2012 | http://pubs.acs.org

avoiding excess of reagent by adding only small amounts and waiting until
the reaction has ceased
use of a higher TEMPO-concentration
use of "activated" cellulose.

Use of activated (water-swollen) cellulose will lead to a more accessible substrate,

which combined with a higher TEMPO-concentration is expected to favour the
desired reaction. Use of a lower hypochlorite concentration will suppress glycol
cleavage. It is therefore supposed that all these measures will be effective.

d. Molecular weight

The molecular weights of some representative samples have been measured and,
again, the high selectivity is shown; i.e. the molecular weight decreases only from
300.000 to 150.000.

In Figure 3 the molecular weight distribution is given for pullulan converted with
the TEMPO/hypochlorite system to varying degrees of oxidation. It can be seen that
loss of molecular weight is moderate. Whereas the maximum in pullulan is found
at 300.000, the maximum of the TEMPO-oxidized material (100%) is found at
170.000. In the N 0 oxidation (under optimum conditions, 4°C reaction
2 4

temperature), the maximum shifts from 300,000 to 100,000 (see Figure 3). A

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 79

Table IV. Oxidation of cellulose and amylose with the sodium hypo-
chlorite/bromide/TEMPO system
b
Substrate time(hours) T(°C) Yield 3
Yield" CC 3
CC
(%) (%) (%) (%)
cellulose 20 20 86 13 70 35

cellulose 24 20 82 28 48 20

cellulose 24 20 80 25 30 25

cellulose 24 5(20) c
85 10 73 35

amylose 24 5(20) c
95 - 52 -
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

3
Yield and degree of oxidation (DO) of water-soluble material
b
Yield and carboxylate content (CC) of non-dissolved material
c
Initial temperature. The reaction temperature was allowed to rise to 20 °C.

1.2
August 31, 2012 | http://pubs.acs.org

Molecular weight

Figure 3. Molecular weight distribution of pullulan with 0 (line

1), 25 (line 2), 50 (line 3), 75 (line 4) and 100% (line 5) degree of
oxidation. Oxidation method hypochlorite/NaBr/TEMPΟ.

higher reaction temperature leads to severe depolymerization (from 300,000 to

approximately 30,000). From this result it appears that for retention of a high
molecular weight the TEMPO-method is preferred. However, so far, no data are
available of the molecular weights of cellulose, oxidized before and after TEMPO-
oxidation. The competing glycol cleavage accompanied by depolymerization may
especially lead to severe depolymerization. In this respect it is expected that
optimization of the conversion of cellulose with the TEMPO-system will lead to
better results (see recommendations above).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 80

Conclusion

There is no doubt that the 2,3-oxidation reaction is highly selective, a fact that can
be attributed to the first step: it is postulated that a cyclic ester of periodate with the
diol group is formed (14).
The behaviour of cellulose in the two oxidation methods directed on oxidation of
the primary alcohol group is different, mainly because the substrate is weakly
soluble in water. In the TEMPO-oxidation it is seen that a part of the cellulose does
not react or does not dissolve.
The yield of the TEMPO-oxidized material is 70-75%, but the conversion is not
quantitative (75%). A part of the material remains undissolved and appears to be
only partly oxidized. It may be recalled that the TEMPO-oxidation gives better
results with water-soluble or swellable polymers. It should also be kept in mind that
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

the N0 -oxidation is in fact not selective. Some non-specific oxidation can be

2

repaired by sodium borohydride reduction. It is to be concluded that the phosphoric

acid/NaN02/NaN0 method gives better results, so far.
3

Up to now no data are available with regard to the molecular weight of TEMPO-
oxidized cellulose. During TEMPO-oxidation hypochlorite/hypobromite oxidation
may be an important side reaction, which may lead to the glycol cleavage reaction.
Some measures to improve the results of the TEMPO-catalysed cellulose oxidation
are indicated.

Experimental

Materials. The cellulose used was a highly purified cotton wool; microcrystalline
August 31, 2012 | http://pubs.acs.org

cellulose (Avicel) was purchased from Merck; Amylose-V (amorphous), potato

starch, and β-cyclodextrin were gifts of Avebe, Veendam, The Netherlands.
Sodium hypochlorite solution (150 gram of active chlorine/L was a gift of A K Z O -
Nobel (Hengelo, The Netherlands). Sodium nitrite, sodium nitrate, sodium
borohydride, and sodium hydroxide were obtained from Merck. Sodium chlorite
(80% purity) was obtained from Aldrich. TEMPO (tetramethylpiperidine-N-oxyl)
was a Sigma Chemicals product. Pullulan was purchased from Hayashibara
(Japan).

Oxidation methods

2,3-dicarboxy cellulose (sodium periodate/sodium chlorite). Method A . To a

suspension of 5.0 g of cellulose in 100 ml of water (pH 5), cooled to 5°C, sodium
periodate (6.64 g) was added. The mixture was stirred in the dark for 168 hours.
The product (dialdehyde cellulose) was filtered off, washed several times with
water, and dried in vacuo. The yield was about 98 %. This material was dispersed
in water, and a solution of hydrogen peroxide (6.3 ml 30% w/w) was added. In the
course of two hours 6.8 g sodium chlorite was added. The mixture was allowed to
react for 48 hours. During the reaction the pH was kept constant at 5 by addition of
0.5 M NaOH-solution. To precipitate the dicarboxy-cellulose, the solution was

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 81

poured out into ethanol. The product was filtrated, redissolved in water, and again
precipitated.

Method B . Procedure as described in method A , however, no hydrogen peroxide

was applied and 20 g of sodium chlorite was used (molar ratio starch/NaC10 = 2

1:6).

Carboxylate content of the products was measured by ion exchange (strong acid),
followed by freeze drying and titration with NaOH. The calcium sequestering
capacity was measured using a calcium ion selective electrode. Two other
substrates, starch and amylose, were subjected to the same oxidation procedures.

6-carboxy cellulose (phosphoric acid/sodium nitrate/nitrite). Cellulose (1.5 g) was

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

dissolved in 30 ml 85% phosphoric acid at 4 °C, which took about 4-6 hours of
stirring. To the viscous solution 1.5 g sodium nitrate was added and stirring was
continued for 30 minutes to dissolve the salt. Then sodium nitrite (40-60 mg) was
added. After stirring for another 60 minutes, the mixture was left at 4°C for 40
hours. A foam developed from which (the toxic) N O escaped. The product was iso­
lated by pouring out the solution while stirring in cold ethanol. A solid precipitated,
which was collected by filtration. The material was dissolved in water and the sol­
ution was brought to pH 7 with sodium carbonate. To this solution 100 mg sodium
borohydride was added to remove any carbonyl function present. After 20 hours the
solution was brought to pH 5-6 with acetic acid. Any salts present were removed by
nanofiltration (Toray, UTC 260 membrane filter). The final solution was freeze
dried. Generally, 6-carboxy-cellulose was obtained in the sodium form as a white
product and in a yield of approximately 80%. The uronic acid content was
August 31, 2012 | http://pubs.acs.org

measured with the Blumenkrantz method (12).

Some kinetic measurements were carried out with β-cyclodextrin in order to prove
the catalytic effect of NO (induced by the presence of nitrite). The experiments
were conducted in the same way as described above. β-Cyclodextrin (0.70 g) was
dissolved in 5 ml phosphoric acid. To this solution 0.70 g sodium nitrate was
added, and after dissolution of this compound, the catalyst sodium nitrite was
added. During the reaction the uronic content was measured as a function of time.
Another experiment was conducted with the objective to study the influence of the
amount of nitrate on the degree of oxidation.

6-carboxy cellulose (oxidation with sodium hyrx>chlorite/bromide/TEMPO). In 100

ml of water 15 mg of TEMPO and 100 mg of sodium bromide were dissolved. In
this solution 2 g of the substrate was suspended and 1 ml of an aqueous solution of
sodium hypochlorite solution( 1 M) was added. Within a few minutes the pH starts
to drop. During the reaction the pH was kept constant by the addition of 0.5 M
NaOH solution (when the rate of the consumption of NaOH-solution decreased,
again a small amount of hypochlorite was added). In this way the competing reac­
tion, glycolic oxidation by sodium hypochlorite, is largely prevented. For a
complete conversion 12 ml of hypochlorite solution (2M), was used. It was
observed that a homogeneous mixture could not be obtained; i.e. a small amount of

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 82

non-reacted or partially converted material did not dissolve. This material was
separated from the solution by centrifugation. The supernatant liquid was desalted
through nanofiltration. The solution was freeze dried. A white material was
obtained of which weight, yield and carboxylate content were determined. The
pellet was washed with ethanol (96%) and dried. From this material the weight and
carboxylate content were also determined.

Literature cited

1. Radley, J.A. Starch and its derivatives, Chapter 11, Chapman and Hall,
London (1968).
2. Nieuwenhuizen, M; Kieboom, A.P.G.; Van Bekkum, H.; Starch/Stärke
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch005

1985, 37, 192.

3. Floor, M; Thesis, Delft University of Technology Delft, The Netherlands
(1989).
4. Floor, M; Kieboom, A.P.G.: Van Bekkum, H.; Rec. des Travaux Chim.
Pays-Bas 1989, 108, 384.
5. Yackel, E.C.; Kenyon, W.O.; J. Am. Chem. Soc. 1942, 64, 121.
6. Painter, T.J.; Carbohydr. Res. 1977, 55, 95.
7. Painter, T.J.; Cesaro, Α.; Delben, F.; Paoletti, S.; Carbohydr. Res. 1985, 61,
140.
8. De Nooy, A.E.J.; Besemer, A.C.; Van Bekkum, H.; Rec. des Travaux Chim.
Pays-Bas 1994, 113, 165.
9. De Nooy, A.E.J.; Besemer, A.C.; Van Bekkum, H; Carbohydr. Res., 1995,
269, 89.
August 31, 2012 | http://pubs.acs.org

10 De Nooy, A.E.J.; Besemer, A.C.; Van Bekkum, H; Tetrahedron 1995, 51,

8023.
11. De Nooy, A.E.J.; Besemer, A.C.; Van Bekkum, H.; Van Dijk, J.A.P.P.;
Smit, J.A.M.; Macromolecules 1996, 29, 6541.
12. Blumenkrantz, N.; Asboe-Hansen, G.; Anal. Biochem. 1973, 54, 484
13. Besemer, A.C.; Van Bekkum, H; Starch/Starke 1994, 46, 95
14. De Wit, D.; Thesis, Delft University of Technology, Delft, The Netherlands
(1990).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 6

Reaction of Bromodeoxycellulose

1 2 3
N. Aoki , M . Sakamoto , and K. Furuhata

1
Molecular Engineering Division, Kanagawa Industrial Research Institute, 705-01,
Shimo-imaizumi, Ebina-shi, Kanagawa Prefecture 243-04, Japan
2
College of School Education, Joetsu University of Education, Yamayashiki-machi,
Joetsu-shi, Niigata Prefecture 943, Japan
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

3
Department of Organic and Polymeric Materials, Faculty of Engineering, Tokyo
Institute of Technology, O-okayama, Meguro-ki, Tokyo 152, Japan

The aim of this paper is to show that bromodeoxycellulose, whose C-6

hydroxyl groups are regioselectively and quantitatively substituted
with bromine atoms, is useful for the synthesis of cellulose derivatives.
The comparison of rate constants of nucleophilic halogen substitution
of halogenated methyl glycosides revealed that the rates for
bromodeoxysaccharides were about 1000 times higher than those of
corresponding chlorodeoxysaccharides. Bromodeoxycellulose was
converted effectively to S-substituted deoxymercaptocellulose
derivatives by the reaction with thiols under homogeneous conditions.
Deoxymercaptocellulose samples having high degrees of substitution
were obtained by the reaction of bromodeoxycellulose with thiourea
August 31, 2012 | http://pubs.acs.org

and consecutive alkali treatment.

Chlorodeoxycellulose (Cell-Cl) and cellulose tosylate have been used for the
syntheses of many kinds of cellulose derivatives (7). Cell-Cl is potentially more
useful than cellulose tosylate because the regioselective and quantitative substitution
of hydroxyl groups at C-6 alone or both at C-6 and C-3 is possible (2, 3). However,
the use of Cell-Cl is somewhat limited because of its relatively low reactivity.
Bromodeoxycellulose (Cell-Br) is considered to be a better cellulose derivative than
Cell-Cl for further reactions because bromine is a better leaving group than chlorine.
Until recently, however, it was difficult to obtain Cell-Br samples with both excellent
regioselectivity and high degree of bromine substitution.
Several organic solvent systems for cellulose have been developed in these 20
years and some of them are used as the reaction media in the chemical modification of
cellulose (4). Lithium halide-MN-dimethylaœtamide (DMAc) systems were found to
be very suitable for the substitution of hydroxyl groups with halogen atoms because
they include high concentrations of halide ion. The degree of substitution (DS) of
halodeoxycellulose has been improved remarkably under homogeneous conditions in
the lithium halide-DMAc systems. The substitution occurs first at C-6, and next at C-
3 but not at C-2. The maximum DS by chlorine achieved for Cell-Cl in LiCl-DMAc
was 1.8 (2) or 1.9 (3) while that of Cell-Br in LiBr-DMAc was 0.9 (regioselectively
substituted at C-6) with the ^V-bromosuccinimide-triphenyl-phosphine (TPP) reagent
system (5) or 1.6 with tribromoimidazole-TPP (6).
This paper describes three topics relating to the reactions of Cell-Br; (i)
evaluation of reactivity of halodeoxycellulose using model saccharides, (ii) reaction of

©1998 American Chemical Society 83

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 84

Cell-Br with thiols in a homogeneous system and (iii) synthesis of

deoxymercaptocellulose.

Experimental

Syntheses. Methyl 3,6-dichloro-3,6-dideoxy-p-D-alloside (Me 3,6-α -β-Α11)(7), 2

methyl 6-chloro-6-deoxy-a(P)-D-glucoside (8) and methyl 6-bromo-6-deoxy-a(P)-D-

glucoside (9) were synthesized by the methods described in the literature, as were
methyl 3,6-dibromo-3,6-dideoxy-P-D-alloside (Me 3,6-ΒΓ -β-Α11) and methyl 3,6-
2

dibromo-3,6-dideoxy-P-D-glucoside (Me 3,6-Br -f*-Glc)(70).

2

Bromodeoxycellulose (Cell-Br) used in this study was synthesized with N-

bromosuccinimide and triphenylphosphine in the UBr-MN-dimethylacetamide
(DMAc) solvent system as described previously (5). In the Cell-Br samples obtained
by this method, only hydroxyl groups at C-6 were substituted with bromine. The
degree of substitution by bromine was 0.85 - 0.97 which was determined based on the
elemental analysis (5).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

For the reaction with a thiol, Cell-Br was dissolved in LiBr-DMAc. After
stirring for 1 h at 60°C, the thiol and triethylamine were added to the solution at
reaction temperatures. The product was recovered by reprecipitation with excess
acetone. The precipitates were treated with a dilute Na C0 solution and dialyzed
2 3

against distilled water and freeze-dried.

Deoxymercaptocellulose (Cell-SH) was synthesized from Cell-Br by a method
similar to that from Cell-Cl (77). A typical reaction procedure is as follows; Cell-Br
(1 g) was treated with thiourea (1.77 g) in 100 mL of dimethyl sulfoxide (DMSO) at
80 °C for 24 h. The mixture became homogeneous as the reaction proceeded.
Deoxyisothiouroniumcellulose bromide (Cell-TU) was recovered as a precipitate by
pouring the reaction solution into 1 L of acetone. Cell-TU was dissolved into water
and the solution was treated with aqueous alkali. Cell-SH was precipitated by
neutralizing the solution with hydrobromic acid and dialyzed against distilled water
for 3 days. The content of the dialysis tube was freeze-dried.

Analysis. The analysis of saccharide mixtures was carried out by gas chromato­
August 31, 2012 | http://pubs.acs.org

graphy (GC) with a gas chromatograph 4BPMF (Shimadzu Corp.) after trifluoro-
acetylation (3). Cellulose derivatives were hydrolyzed in sulfuric acid. Gas
chromatography-mass spectrometry (GC-MS) was applied to support the assignments
obtained by the GC analysis with Shimadzu GC-MS L K B 9000S gas
chromatograph-mass spectrometer (Shimadzu Corp.). Details of the conditions for
GC and GC-MS analyses were described previously (3). NMR spectra were recorded
with a spectrophotometer JNM-A500 (JEOL, Ltd.).

Halogen Exchange Reactions of Model Saccharides. Halogen exchange reactions

were carried out in DMAc. In order to terminate the reaction, the lithium chloride
was precipitated by pouring the reaction solution into butyl acetate and the bromide
ion was precipitated as silver bromide by adding silver lactate to the solution. After
filtration, the filtrate was evaporated, the residue was trifluoroacetylated, and
subjected to the GC analysis.

Results and Discussion

Evaluation of Reactivity of Halodeoxycellulose using Model Saccharides. There

is a difference in the configuration of repeating units in chlorodeoxycellulose (Cell-
Cl) and that in bromodeoxycellulose (Cell-Br) with high DS (Scheme 1). Cell-Cl with
DS over 0.8 includes 3,6-dichloro-3,6-dideoxyallose units (3) while Cell-Br with high
DS contains two different dibromodideoxysaccharide units, that is, 3,6-dibromo-3,6-
dideoxyglucose and 3,6-dibromo-3,6-dideoxyallose units (6). The mechanism of

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 85

CH OH
2 CH OH
2 CH C1
2 CH C12

OH OH OH α OH

OH OH OH
CH OH
2 CH Br
2 CH Br2 CH Br
2
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

LiBr-DMAcT \ ^ _ ^ \ \ ^ ^ \ \ ^ ^ \

OH OH OH Br OH

NCS, Λ^-Chlorosuccinimide: NBS, TV-Bronwsuccinimide: ΒΓ3Ι111, Tribromoimidazole

Scheme 1
these halogenations of cellulose is considered to be the same; formation of
triphenylphosphonium ester followed by Sn2 attack of halide ion (12). We have found
that the high reactivity of Cell-Br is the reason for this difference. Three types of
model experiments using halogenated methyl glycosides were carried out to examine
the reactivities of halodeoxysaccharides.
Type I is the halogen exchange reaction at C-3 of a methyl 3,6-dideoxy-3,6-
dihaloglycoside using the ion of the same kind of halogen as included in the starting
saccharide (70, 75);
August 31, 2012 | http://pubs.acs.org

(1)

X OH OH

where X is CI or Br. This type of reaction can be monitored b y GC analysis of the

reaction mixture after trifluoroacetylation. The reversible reaction shown above is
considered to be a model for the inversion of configuration at C-3 positions of
dideoxydihalo-units which may occur during the substitution of hydroxyl groups of
cellulose with halogen atoms in lithium halide-DMA. The rate equation for this
reaction is relatively simple because the concentration of halide ion is constant during
the reaction.

y^yeil-expi-k.tBli/y.)} (2)

where y and y are the mole fractions of product saccharide at time t and at
t e

equilibrium, respectively, and [B] is the concentration of the halide ion (75).
Type II is the halogen exchange reaction of a methyl 6-deoxy-6-haloglucoside

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 86

using the ion of halogen not contained in the starting saccharide. In this type of
reaction, the exchange occurs only at C-6 and the configuration is not changed. It is
possible to follow the exchange at C-6 most conveniently by GC analysis;

OH OH
1
where X is CI or Br and X is Br or CI, respectively. The molar response of the FID
detector in GC analysis is different for the saccharides. We obtained following values
relative to that of methyl ct-D-glucoside (internal standard); 0.76, 0.82, 0.89 and 0.85
for methyl 6-chloro-6-deoxy-a-D-glucoside (Me 6-Cl-a-Glc), methyl 6-bromo-6-
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

deoxy-a-D-glucoside (Me 6-Br-a-Glc), methyl 6-chloro-6-deoxy-P-D-glucoside (Me

6-Cl-P-Glc) and methyl 6-bromo-6-deoxy-P-D-glucoside (Me 6-Br-P-Glc),
respectively. The rate equation for these reactions is more complicated than that for
Type I reactions;

y = y {l-exp(-a/)}/{l+pexp(-ag}
t e (4)

where a = k (2[B] /y -[A] -[B] ), β = l-y ([A] +[B] )/[B] , y and y are the mole
I 0 e 0 0 e 0 0 0 t e

fractions of product saccharide at time t and at equilibrium, respectively, and [A] and
0

[B] are the initial concentrations of the starting saccharide and halide ion,
0

respectively. In the cases where the equilibrium points are shifted largely to one side,
the rate constants are determined from the initial slopes in order to avoid possible
errors in iterative calculations.
Type III is the halogen exchange reaction of a methyl 3,6-dideoxy-3,6-
dihaloglycoside using the ion of halogen not contained in the starting saccharide. This
will give a complex saccharide mixture as shown in Scheme 2 for the reaction of
August 31, 2012 | http://pubs.acs.org

methyl 3,6-dichloro-3,6-dideoxy-P-D-alloside (Me 3,6-Cl -All) with LiBr as an

2

example. It is very difficult to analyze this mathematically exactly. Eight saccharides

can theoretically be present in the solution and 24 rate constants would be necessary
to effectively describe the total reaction. For simplicity, the substitution at C-3 is
assumed to be independent of that at C-6. Equations (2) and (4) are applied to obtain
the average rate constants for the substitution at C-3 and C-6, respectively. The mole
fraction of each saccharide is calculated as the product of the substitution ratios of
constituent halogen atoms at C-3 and C-6.
A typical time course of Type I reaction between Me 3,6-Cl -All and LiCl (13) is
2

shown in Figure 1. The symbols show the experimental data and the theoretical
curves were calculated with the obtained parameters. The curves coincide very well
with the data. The equilibrium is on the glucoside side at all temperatures studied.
The kinetic parameters for Type I and Type II reactions are summarized in Table I.
The rates of interconversion between dibromodideoxyglycosides are about 1000 times
higher than those between dichlorodideoxyglycosides. The difference in the
configurations at C-3 of repeating units between Cell-Cl and Cell-Br can be ascribed
to that in the rates of nucleophilic substitution of halogen atoms in dideoxydihalo-
units with halide ions in lithium halide-DMAc (10,13). Most of the 3,6-dibromo-3,6-
dideoxyallose units originally formed in Cell-Br are converted quickly to
thermodynamically more stable 3,6-dibromo-3,6-dideoxyglucose units in
LiBr-DMAc.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

Table I Kinetic parameters for Type I and Type II reactions

3
Starting Temp. [Saccharide] [LiX] fc.xlO 1ί χΙ(?
Λ Activation energy (kJ/mol)
1 1
compound (°Q (g/L) (mol/L) (Lmol 'min ) (Lmol 'min )

80 1.65 2.14 0.19 0.03

Me 3,6-α -β-Α11
2
90 1.58 2.04 0.77 0.13 175 177
100 1.68 0.99 4.67 0.80

60 1.00 0.0826 17.5 2.02

70 1.03 0.042 101 17.2
Me 3,6-ΒΓ -β-Α112
80 1.00 0.040 264 36.1 143 145
90 0.93 0.00415 1470 192

70 3.07 2.32 0.34 (401)

Me ό - α - β - G l c 80 2.68 2.03 0.93 (191) 85
90 2.33 2.03 1.74 (286)

63 2.78 0.00993 257 (0.49)

Me ό-ΒΓ-β-Glc 70 2.69 0.00993 373 (0.34) 100
80 2.93 0.00993 1370 (0.27)

In Cellulose Derivatives; Heinze, T., et al.;

70 2.67 2.31 0.50 (67)
Me 6-Cl-a-Glc 80 2.88 2.09 0.95 (144) 59
90 2.73 2.21 1.53 (132)

60 2.37 0.00993 152 (15.8)

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

Me 6-Br-a-Glc 70 2.34 0.00993 455 (89.4) 102
80 2.33 0.00993 1230 (57.3)
 88

CHgCI CHaBr
yLqOCH3

[— HOVT "*CF HOVY

CI O H

A
CI O H
ci- . I*ci-
Br
Br
CH CI 2 CHgBr
J-OOCH3 J—0ÇCH3
Br
HO
ci-
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

OH OH
A A
ci- CI- Br Br Br Br CI" ci-

CH CI 2 CH Br 2

J-COCH3 ^COCH
Br ^ /^y
3

HOW ^Cl- HOW

Br O H Br O H

A
ci- Br CI- Br
August 31, 2012 | http://pubs.acs.org

CH CI
2 CHaBr

HO
Cl-
OH OH

Scheme 2
The kinetic parameters are considered to be affected by the leaving group,
nucleophile, anomeric structure, position of halogen substitution, and if halogen is on
C-3, configuration at C-3. For example, it is possible to estimate the effect of
anomeric structure on the rate of substitution at C-6 from data shown in Table I; the
difference in rates is very small between the anomers. The halogen exchange between
Me 3,6-Cl -All and LiBr (Type III) was carried out to estimate the effects of other
2

factors. Only five dideoxydihalosaccharides were detected on the chromatograms.

An additional assumption is introduced to simplify the analysis, that is, molar
response values of these saccharides are equal. Figure 2 shows the time course of the
Type III reaction. The calculated curves fit with the experimental data very well
except two at 48 h. Table II compares the rate constants obtained with those for Type
I and Type II reactions. Although the values for Type III may contain some errors, it

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 100

Time (h)
Figure 1. Time course of reaction between Me 3,6-θ2-β-Α11 and l i d
Reaction temperature: · , 80°C; A , 90°C; • , 100°C.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

100

0 10 20 30 40 50
Time (h)
Figure 2. Time course of reaction between Me 3,6-01 -β-Α11 and LiBr. O, Me
2

3,6-α -β-Α11; • , methyl 6-bromo-3-chloro-3,6-dideoxy^-alloside; • , methyl

2
August 31, 2012 | http://pubs.acs.org

3-bromc*^Worc-3,6-dideoxy^-glucoside; Δ , Me 3,6-Br ^-Glc; • , Me 3,6- 2

ΒΓ,-β-Glc.

Table II Kinetic parameters of halogen exchange reactions at 90°C

3 ! l
Type of exchange* k χ 10 (Lmol min )
Position Configuration Halogen Type IIP Type I and II
C-6 — CI — B r 2.97 1.74
— B r - * CI 116 1370 (80°C)
C-3 alio — gluco CI — B r 0.476 —
CI — C I — 0.77
Br — B r — 1470
gluco — alio Br — C I 85.4 —
Br — B r 192
CI — C I 0.13
a
Changing from the left-side one to the right-side one.
b
The differences in relative molar responses were not taken into account.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 90

is clearly shown which factor affects the reaction most strongly; when bromine is the
leaving group, the rate constants are about 1000 times larger than those for the case
where chlorine is the leaving group. The other factors, kind of nucleophile, position
of halogen substitution and configuration at C-3, affect much less than the kind of
halogen. From these findings, it can be concluded that the reactivity of Cell-Br for
nucleophilic substitution is much higher than that of Cell-Cl.

Reaction of Cell-Br with Thiols in a Homogeneous System. Above results show

that Cell-Br, whose hydroxyl groups are regioselectively substituted with bromine, is
potentially very useful for the regioselective modification of cellulose derivatives
through nucleophilic substitution. Only few studies have been reported in this area.
We reported the reactions of Cell-Br with amines (14, 15), and with thiols under
heterogeneous (76) and homogeneous (77) conditions. The reactivities of amines in
nucleophilic substitution are relatively low while the reactions with thiols under
homogeneous conditions are convenient for the introduction of a variety of functional
groups into cellulose, especially under alkaline conditions. The LiBr-DMAc solvent
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

system was used where no pretreatment was necessary for the dissolution of Cell-Br
(77). Structures of reaction products of Cell-Br (bromine substitution at C-6) with
thiols were studied with GC-MS and NMR analyses which confirmed that bromine
atoms were substituted with thiol moieties. The amino groups of cysteine and 2-
aminoethanethiol did not react We tried the reaction of Cell-Cl but the substitution
did not occur under the reaction conditions studied.
August 31, 2012 | http://pubs.acs.org

7 7.5 8 8.5 9 9.5 10 10.5 11

p/Caof S H

Figure 3. Relation between p£a of mercapto group and conversions (filled

squares show the data obtained at higher concentrations of triethylamine
and/or thiols). Thiol: 4ABT, 4-aminobenzenetiol; M B A , 2-mercaptobenzoic
acid; AET, 2-aminoethanethiol; CYS cysteine; MET, 2-mercaptoethanol;
MPA, 3-mercaptopropionic acid; MBDA, 2-mercaptobutanedioic acid.

Figure 3 shows the relation between p£a values of mercapto groups of used
thiols and achieved conversions of Cell-Br to 5-substituted deoxymercaptocellulose
derivatives (77). The data plotted with filled circles were obtained under the same
reaction conditions. It is clear that the reactivity of thiol depends strongly on the pATa
value of mercapto group. This relation could not be observed in the case of the
heterogeneous reactions in aqueous alkali (76). The pKa. values of mercapto groups
for thiols having carboxyl group(s) are low and the conversions are low as shown in
the figure. Even these thiols give higher conversions of Cell-Br when the
concentrations of triethylamine and/or thiols are increased (shown with filled
squares).
Table III compares the solubilities of the products (77). Some of the products
obtained under homogeneous conditions are soluble in aqueous solutions while all the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 91

Table III Solubilities of cellulose derivatives obtained from Cell-Br and thiols
HÔ NaOH
Thiol used DS IN pH5.0 H0
2 pH9.0 IN
Aminoethanthiol 0.72 - - - - -
Mercaptopropionic acid 0.92 - - - - ±

Meracaptobutandioic acid 0.47 - + + + +

Cysteine 0.55 + - - - +
4-Aminobenzenethiol 0.84 - - - - -
2-Mercaptobenzoic acid 0.64 - + + + +
+, Soluble. ±, Slightly soluble or swallen. - , Insoluble
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

products obtained under heterogeneous conditions were insoluble. The carboxyl

group is effective in improving the solubilities of samples. We used four carboxyl-
bearing thiols and three of them gave samples soluble in alkaline solutions. The most
part of sample having cysteine moiety (Cell-CYS) is soluble both in acidic and
alkaline solutions, and the dissolved Cell-CYS is precipitated by neutralization of the
filtered solution. This cycle of dissolution and precipitation can be repeated many
times.

Synthesis of Deoxymercaptocellulose (Cell-SH). As another example of the

applications of Cell-Br, we studied the introduction of mercapto groups at C-6.
Mercapto-bearing cellulose derivatives are interesting materials. They are expected to
+
be useful, for example, as a specific sorbent for Hg and Ag* and also as starting
materials for the introduction of functional groups through reactions with various
organic halides. The syntheses of Cell-SH from Cell-Cl (77) or cellulose tosylate (18)
were reported previously.
August 31, 2012 | http://pubs.acs.org

OH OH OH
Cell-Br Cell-TU Cell-SH

Scheme 3

In this paper, Cell-Br was converted to Cell-SH through the reaction with
thiourea (Scheme 3) where DMSO was used as a solvent Other organic solvents such
as DMAc with or without lithium halide were examined but DMSO gave the best
result. The DS by bromine (DS*) of starting Cell-Br ranged from 0.85 to 0.97. The
time course of the formation of deoxyisothiouroniumcellulose bromide (Cell-TU)
from Cell-Br and thiourea is shown in Figure 4. The degree of substitution by
isothiouronium moiety (DS ) and DS for Cell-TU were calculated with nitrogen and
TO Br

bromine contents. The figure shows that DS™ gradually increases with increasing
time (DS decreases at a corresponding rate). The maximum conversion of bromine
Br

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 92

1.0

Time / h
Figure 4. Time course of reaction between Cell-Br and thiourea.
into isothiouronium moiety achieved is higher than 70%. Approximately 20% of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

bromine in the original Cell-Br was lost within 24 h, probably by

dehydrobromination. It is impossible to compare these results with those reported for
Cell-Cl (77) because the chlorine contents of Cell-SH samples from Cell-Cl were not
given. Cell-TU samples obtained in this study are soluble in water while those from
Cell-Cl and thiourea were only swollen in water (77). The increased solubility can be
explained in terms of the homogeneous distribution of isothiouronium residues which
is due to the fact that both the synthesis of Cell-Br and reaction with thiourea
proceeded under homogeneous conditions.
Cell-TU was treated next with alkali to convert isothiouronium groups into
mercapto groups. Cell-TU was dissolved in water and Cell-SH was precipitated by
neutralizing the solution after alkali treatment. Two conditions were examined for the
August 31, 2012 | http://pubs.acs.org

ι I • I ι ι ι ι I ι ι ι ι I ι ι ι ι
4000 3000 2000 1500 1000 500

Wavenumber / c m - 1

Figure 5. IR spectra of Cell-TU and Cell-SH. a, Cell-Br (DS* 0.85); b,

cellulose; c, Cell-TU (DS™ 0.60); d, Cell-SH (DSs„ 0.57) obtained under
condition B.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 93

alkali treatment in water; treating with a Na C0 solution (pH 11.0) in a dialysis tube
2 3

for 15 h (condition A) and stirring in 2% NaOH solution for 1.5 h (condition B). With
regard to the conversion and reaction time, condition Β is better than condition A.
The conversion of isothiouronium moieties to mercapto groups reached 95% under
condition B. Much harsher conditions, for example, boiling in an alkaline solution for
several hours, were applied to the synthesis of mercapto-bearing polymers (79). Both
of our alkali treatment conditions are much milder but effective enough to obtain Cell-
SH. This may be ascribed to the higher hydrophilicity of cellulose chains.
Figure 5 shows IR spectra of Cell-TU and Cell SH together with those of Cell-Br
and cellulose. A weak peak due to S-H stretching vibration is clearly observed at
1
2560 cm in the spectrum of the Cell-SH sample. This peak was not reported before
for the cellulose derivatives containing mercapto groups.
Cell-SH and other deoxymercaptocellulose derivatives were stored under reduced
pressure in the dark. Their IR spectra showed no detectable change after storage for
several months indicating that no oxidation by air had occurred under the storage
conditions.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch006

We showed here the reactions between 6-bromo-6-deoxycelllulose and some

nucleophiles under homogeneous conditions. We believe that this method is generally
applicable to other nucleophiles and is useful for the regioselective introduction of
functional groups to cellulose.

Literature Cited.

1. Ishizu, A., In Wood and CellulosicChemistry;Hon, D.N.-S.; Shiraishi, Ν.,

Eds.; Marcel Dekker: New York, 1991, chap. 16.
2. Furubeppu, S.; Kondo, T.; Ishizu, A. Sen'i Gakkaishi, 1991, 47, 592.
3. Furuhata, K.; Chang, H.-S.; Aoki, N.; Sakamoto, M. Carbohydr. Res., 1992,
230, 151.
4. Dawsey, T. R.; McCormick, C. L. JMS-Rev. Macromol. Chem. Phys., 1990,
C30, 405.
5. Furuhata, K.; Koganei, K.; Chang, H.-S.; Aoki, N.; Sakamoto, M. Carbohydr.
August 31, 2012 | http://pubs.acs.org

Res., 1992, 230, 165.

6. Furuhata, K.; Aoki, N.; Suzuki, S.; Sakamoto, M.; Saegusa, Y.; Nakamura, S.
Carbohydr. Polym., 1995, 26, 25.
7. Dean, D.M.; Azarek, W.A.; Jones, J.K.N. Carbohydr. Res., 1974, 33, 383.
8. Evans, M.E.; Long, Jr., L.; Parrish, F.W. J. Org. Chem. 1968, 33, 1074.
9. Hannesian, S.; Plessas, N.R. J. Org. Chem., 1969, 34, 1035.
10. Furuhata, Κ.; Aoki, N.; Suzuki, S.; Arai, N.; Ishida, H.; Saegusa, Y.;
Nakamura, N.; Sakamoto, M. Carbohydr. Res., 1995, 275, 17.
11. Tashiro, T.; Shimura, Y. J. Appl. Polym. Sci., 1982, 27, 747.
12 Hodosi, G.; Podányi, Β.; Kuszmann, J. Carbohydr. Res., 1992, 230, 327.
13. Furuhata, K.; Aoki, N.; Suzuki, S.; Arai, N.; Sakamoto, M.; Saegusa, Y.;
Nakamura, N. Carbohydr. Res., 1994, 258, 169.
14 Aoki, N.; Taniguchi, T.; Arai, N.; Furuhata, K.; Sakamoto, M. Sen'i Gakkaishi,
1993, 49, 563.
15. Saad, G.R.; Sakamoto, M.; Furuhata, K. Polym. Intern., 1996, 41, 293.
16. Aoki, N.; Koganei, K.; Chang, H.-S.; Furuhata, K.; Sakamoto, M. Carbohydr.
Polym., 1995, 27, 13.
17. Αοki, N.; Furuhata, K.; Saegusa, Y.; Nakamura, S.; Sakamoto, M. J. Appl.
Polym. Sci., 1996, 61, 1173.
18. Sakamoto, M.; Yamada, Y.; Ojima, N.; Tonami, H. J. Appl. Polym. Sci., 1972,
16, 1495.
19. Chanda, M.; O'Druscill, K. F.; Rempel, G. L. React. Polym., 1986, 4, 11.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 7

Cationization of Cellulose Fibers in View

of Applications in the Paper Industry

E. Gruber, C. Granzow, and Th. Ott

Institute of Macromolecular Chemistry, University of Technology Darmstadt,

Alexanderstrasse 10, D-64283 Darmstadt, Germany
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

Three methods were presented to render pulps cationic:

• direct reaction of epichlorohydrin and a tertiary amine

• coupling of oligo-ionomers
• grafting of cationic monomers

Advantages of such cationic pulps are:

• more effective than soluble cationic celluloses

August 31, 2012 | http://pubs.acs.org

• not sensitive towards pH-changes

• very versatile
• more economic

The direct reaction of epichlorohydrin and a tertiary amine is catalyzed by

hindered tertiary amines (possible auto catalysis). A wide range of different
products carrying various chemical groups (different polarity, accessibility,
charge density) can be achieved by this method. Disadvantages of this reaction
are, that the reaction proceeds also within the fiber and that it causes cross
linking.
The method of coupling ionomers to the fiber yields higher charge densities,
but surface selectivity is still poor. Surprisingly a higher surface charge has an
adverse effect on retention of anionic filler particles.
For radical grafting besides charged monomers neutral comonomers have to
be used (e.g. acrylamide). This method exhibits the best surface selectivity.
As paper aids cationic pulps excel at

• high total retention effectiveness

• good strength properties
• good drainability

94 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 95

Cellulose fibers normally are negatively charged. This has a strong bearing on
properties of a paper stock, which predominately consists of anionic species like pulp
fibers, fillers, and additives. Repelling interactions between such negatively charged
particles hamper the process of flock formation, which may cause problems for water
drainage and particle retention.
It is common practice in paper making to add cationic polymers to improve flock
formation and particle retention. However such soluble polyelectrolytes may also
exhibit some disadvantages. One draw back is the viscosity contribution of soluble
polymers, which in most cases also depends on pH and salt concentration in the
aqueous medium. Polymeric aids are also very sensitive to dosage. They should be
applied in stoichiometric concentrations, over dosage leads to a change from a
negative to a positive net charge for all species present, resulting in another repelling
interaction.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

This is demonstrated on a soluble cationic cellulose. Figure 1 shows, that the

effectiveness of soluble ammonium alkyl cellulose as a retention aid passes a
1
maximum as a function of dosage . The maximum shifts to higher dosages, when the
DS increases.
23
A similar effect is found with most other cationic polymers used as paper additives .
The reason is, that a flexible polyelectrolyte may easily approach anionic charges on
the surface of either a fiber or filler particle, thus effectively neutralizing their
electrical charge. A stiff polyelectrolyte on the other hand will only form a minor
fraction of ion pairs, such aggregates will still carry naked anions as well as „unused"
cations. Figure 2 demonstrates schematically this different situation. If there are only
stiff charge carriers flock formation is more efficient, as there are always free ions
available.
Bridging capacity of cationic compounds will also depend on the accessibility of the
August 31, 2012 | http://pubs.acs.org

cationic groups. Evidently they should sit on the surfece and have some mobility to be
able to form contact ion pairs. The chemical nature of the cationic groups will also be
of importance as it will control the degree of hydration.
Based on such considerations there should be a potential for cationic fibers for paper
making. Such fibers could offer some benefits over soluble polyelectrolytes:

• improving filler retention

• support drainage
• adsorb anionic trash
• do not disperse into white and waste water
• are more easily biologically degradable
4
Stone and Rutherford have already described cationization of cellulose fibers by
using glycidyl ammonium salts in 1969. Krause and Kâufer suggested to use such
5
cationic pulps as paper making additives and investigated their basic properties .
This paper describes some investigations based on such previous work extending it by
using different chemical routes to obtain cationic pulp fibers.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

ammoniumalkyl cellulose
[ mmol/kg]

Figure 1 : Retention effect of a soluble ammonium - alkyl - cellulose

(trimethyl glycidyl ammonium cellulose)
August 31, 2012 | http://pubs.acs.org

soluble poly cations flock formation

retention retention
flocculation
drainage

cationic fibers
Figure 2: Interactions among electrically charged fibers, pigments, and polymers

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 97

Different types of cationic pulp fibers

In order to study morphological effects we prepared three different kinds of
grafted fibers illustrated schematically in Figure 3. Direct addition reaction yields
glycidyl ammonium derivatives of cellulose, where charged groups are tightly bond
to the cellulose material. By coupling oligomers short side chains carrying cations
can be introduced. Finally by grafting cationic polymer chains somewhat longer
and more accessible side chains of ion carriers could be attached to the cellulose
fibers.

Chemical routes to cationic pulp fibers

Simultaneous reaction of epichlorohydrin and a tertiary amine
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

Normal cationization uses glycidyl trimethyl ammoniumchloride as a reactant. To

obtain a wider range of different cationic functionalities a one step method can be
applied, which starts from epichlorohydrin and any tertiary amine (Equation 1).

Equation 1: One step cationization of a polysaccharide

Cf Rl
Rl ^v. ^—N— R 2
Posac—Ο i
+
X
Posac—OH +P>^ C1
( Yl—R 2

R/
Polysaccharide Epichlorohydrin Cationic Polysaccharide

As there are also some carboxylic functions present in natural polysaccharides,

August 31, 2012 | http://pubs.acs.org

besides ethers, esters and salts may be formed (Equation 2).

Equation 2: Possible side reactions with carboxylic groups

(Ester linkage)
Ο Ο cr
Posac—C* Posac—C*
X)H x
o-
Polysaccharide

Tertiary Ammonium salt

Amine

To get some insight in the mechanism of these reactions, methyl-ct-D-

glucopyranoside and glucuronic acid were used as model substances. These were
reacted with epichlorohydrin and triethyl amine and the resulting products were
analyzed by N M R and M A L D I mass spectroscopy. The conclusions, drawn from
these investigations are listed in Table 1.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 98

Table J: Features of one step cationization

Feature Method Result
yield MALDI if stoichiometrically applied,
epichlorohydrin is consumed
quantitatively
formation of ethers NMR C2, C3, C6 rather similar in
reactivity
formation of uremic esters NMR insignificant
formation of uremic NMR positive
ammonium salts
formation of advancement MALDI few oligomers
products
crosslinking reaction in solution of some crosslinking occurs
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

dimethyl acetamide / LiCl

We should expect, that the reaction proceeds predominately on the surface of the
fiber. However the kinetics of the reaction (see Figure 4) show that both, a surface
reaction and penetration into the depth of the fiber occurs. A slow diffusion controlled
reaction of accessible areas within the fibers follows a swift reaction on the surface.
The ratio between surface and bulk reaction depends on catalysts applied and the size
of the ligands of the amine. The degree of surface cationization can be determined by
polyelectrolyte titration (see Figure 5).

Cationization bv 0U20 - ionomers

By coupling a oligo-ionomer to the fiber higher charge densities can be achieved.

August 31, 2012 | http://pubs.acs.org

Such oligomers are synthesized according to Equation 3:

Equation 3: Synthetic route for preparing oligo - ionomers

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

Figure 3: Types of cationic pulp fibers, prepared by different methods

August 31, 2012 | http://pubs.acs.org

0 2 4 6 8 10 12

Figure 4: Kinetics of amine/epichlorohydrin cationization

(ECH = epichlorhydrin; T E A = triethyl amine; DABCO = 1,4-
diazabicyclo[2.2.2]octane; G M A C = glycidyl trimethyl ammonnium chloride;
IMIZ = 4-methylimidazole)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 100
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

Figure 5: Comparison of some different degrees of cationization obtained

6
(colorimetric titration according to reference )

Figure 5 shows, that cationization by oligo-ionomers is most effective, the surface

selectivity however is still poor (app. 10%).

As an example of technical features of such pulps the retention of filler (CaC0 ) was
3

studied. The fiber stock used as model of paper stock contained unmodified pulp,
modified pulp, and filler. Figure 6 shows the results obtained with pulp slurries, which
August 31, 2012 | http://pubs.acs.org

contained the same amount of cationic charge equivalent but different charge densities
of the cationic pulps added.

It can be seen, that filler retention capacity decreases with the charge density applied.
Oligomer grafted pulps had the highest surface charges but were specifically least
effective. To achieve optimal filler retention a rather high feed of charged fibers is
needed, each of which should carry only a moderate surface charge. Among
differently substituted ammonium ions trimethyl compounds are most effective.

Grafted cationic pulp

Heterogeneous grafting
Another way to render cellulose cationic was pursued by grafting cationic monomers
to fibers. The reaction consists of a radical polymerization of an unsaturated monomer
(Equation 4) starting at a radical, generated on the fiber surface.
Equation 4: Grafting on cellulose (Cell- = cellulose radical)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 101

Table 2: Substances usedfor grafting

Name Short Formula

Diallyl dimethyl
ammoniumchloride
DADMAC
Μ
\·Λ·
Me' " ^ ^ ^

Ο
[3-(Methacryloylamino)-
propyl]-trimethyl- MAPTAC
ammonium chloride
TA
0
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

Acrylamide AAM

The cellulose radical may be generated by chemical activated initiation of grafting:

• redox systems (e.g. Ce IV or Fe II)
• chemical oxidation (e.g. ozone, peroxide)
In these investigations Ce (IV) was used as a matter of convenience.
The substances used for grafting are listed in Table 2.
As fiber material bleached beech sulfite pulp was applied.
Trials to graft charged monomers alone to cellulose however were not successful.
With D A D M A C practically no grafting reaction was observed. But even with the
neutral monomer acrylamide only very low grafting yields could be achieved.
Surprisingly both monomers together reacted vigorously to a graft copolymer with
August 31, 2012 | http://pubs.acs.org

copolymeric side chains. The reaction is summarized in Equation 5.

Equation 5: Cationic grafting onto cellulose

Pure (meth)acryl ammonium compounds like M A P T A C also do not graft to cellulose

but contrary to allyl ammonium salts, they do not copolymerize easily with acrylamide
(Figure 7).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 102
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

8 10 12 14
beating time [min]

Figure 6: Filler retention as a function of beating time at various degrees of

cationization
(ECH = epichlorhydrin; G M A C = glycidyl trimethyl ammonnium chloride)

1 1 1
August 31, 2012 | http://pubs.acs.org

N-content [mmol/kg] • AAM+DADMAC

I1 5+1 mol/kg
1 11
Ο A A M 6 mol/kg
Δ AAM+MAPTAC
5+1 mol/kg
-
—
<

>
s .--^
ο DADMAC 6 mol/kg
1
•T • MAPTAC 6 mol/kg
ι r »
0 2 4 6 8 10 12 14 16 18
reaction time [h]

Figure 7: Influence of monomer type and reaction time on grafting yield

( A A M = acrylamide; D A D M A C = diallyldimetyl ammoniumchloride; M A P T A C =
[3-(methacryloylamino) propyl] trimethylarnmoniumchloride)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 103

Technical features of grafted cationic pulp fibers

Retention effects

As expected cationic pulps operate as retention aids to negatively charged filler

particles. The grafted cationic pulp was tested in a fiber stock containing calcium
carbonate. The effect on retention of such filler particles was compared with a soluble
cationic polymer (PDAMAC), which contained the same charged groups like the
grafted pulp. As seen in Figure 8 at lower dosage of the polyelectrolyte the soluble
polymer (PDADMAC) is more effective than the grafted pulp (BuSiKat). At higher
dosages however grafted fibers are more active and do not show any sign of
saturation and over dosage.
From this experiment we may conclude, that cations fixed to fibers are specifically less
effective than chain bond ones, but there is no self inhibition and a higher total amount
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

of filler can be fixed.

Mechanical properties
Each chemical modification may damage the pulp fibers leading to losses of
mechanical strength. In addition to that, the modified surface charge will have an
influence on the fiber to fiber interactions and on the formation of the sheet.
To test the influence of modified pulps, sheets were made from a mixture of modified
and unmodified pulp fibers. As shown in Figure 9 the strength of the paper is
augmented by the content of grafted pulp. This may be caused by stronger fiber
flocks formed and may be also an indication, that the strength of the fibers as such is
not hampered.
In practice however this positive effect may be camouflaged by the retention effect,
which will lead to a higher filler content in papers containing cationic pulp. To
August 31, 2012 | http://pubs.acs.org

evaluate this issue, filled papers were made using either grafted pulp or a synthetic
polyelectrolyte (PDADMC) as a retention aid. Figure 10 shows, that the mechanical
strength decreases as normal, when the filler content is increased. However this
decline is comparable with fibers and soluble polyelectrolytes.
Drainage effects
As drainage of a paper slurry depends among others on flock formation, it has to be
expected, that grafted pulp fibers will influence drainability. As a general information
on drainage properties Schopper-Rieglerfreenesswas measured as a function of feed
of grafted pulp and beating time. The results are summarized in Figure 11.
As can be seen, cationic pulps drain more readily and exhibit slightly higher beating
resistance. This may suggest, that flocks of higher densities are formed by fibers of
opposite charge. This leads to slightly more substantial inhomogeneities in the paper.
In fact this can be confirmed by image analysis of microphotographs. Opacity and, as
shown, mechanical strength however does not suffer by these stronger variations of
fiber density.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 104

80
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

0 5 10 15 20

Figure 8: Retention of C a C 0 by cationic graft pulps compared to a soluble

3

cationic polymer
(PDADMAC = poly diallyldimetyl ammoniumchloride; BuSiKat = beech sulphite
pulp grafted by PDADMAC)
August 31, 2012 | http://pubs.acs.org

Figure 9: Mechanical strength of lab sheets without filler

(made from grafted + ungrafted beech sulphite pulp)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 105

breaking length [m
5000
4500
4000
3500
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

3000
• Bu Si Kat
2500 ° PDADMAC
retention [%]
2000
ί­ 10 _30_ J|0_ .m.
ο 10 15 %CaCOj

Figure 10: Breaking length as a function of filler concentration

(PDADMAC = poly diallyldimetyl ammoniumchloride; BuSiKat = beech sulphite
pulp grafted by PDADMAC)
August 31, 2012 | http://pubs.acs.org

5 10 15 20 25
cone, of cations / dry fiber [mmol/kg]

Figure 11 : Influence of cationic pulps on the drainability andfreenessoffiberstock

suspensions

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 106

Acknowledgment

This work was sponsored by a grant AIF 10210 Ν by the German „Arbeitsgemeinschaft
Industrieller Forschungseinrichtungen" and by the German Papermakers Association
(„Verband der Deutschen Papierfabriken").

Literature Cited

1
Ott, G., Thesis Darmstadt 1992
2
Swerin, A.; Sjödin, U.; Ödberg, L.; Nordic Pulp Paper Res. J. 8 (1993) No. 4,
389 - 396
3
Klix,J.,Thesis Darmstadt 1991
4
Stone, F.W., Rutherford, J.M.; US-Patent 3 472 840 (1969)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch007

5
Käufer, M.; Thesis, Darmstadt 1982;
Käufer, M., Krause, Th.; Schempp, W.; Das Papier 34 (1980), Nr. 12, 575-
579
Käufer, M., Krause, Th.; Schempp, W.; Das Papier 35 (1981) Nr. 10A, V33-
V38
Käufer, M., Krause, Th.; Das Papier 37 (1983), Nr. 5, 181-185
Käufer, M., Das Papier 35 (1981) Nr.12, 555-562
6
Gruber, E.; Ott, Th.: Das Papier 49 (1995), Nr. 6, 289 - 296
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 8

A Preliminary Study of Carbamoylethylated Ramie

1 1,3 1 2 2
C. C. L. Poon , Y. S. Szeto , W. K. Lee , W. L. Chan , and C. W. Yip

1 2
Institute of Textiles and Clothing and Department of Applied Biology and
Chemical Technology, Hong Kong Polytechnic University, Hong Kong
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

Ramie fibres were treated with acrylamide in the presence of alkali

acting as a catalyst; the add-on ranged from 0.8% to 11.0%. Both the
amount of char produced during thermal degradation and the flame
resistance in terms of Limiting Oxygen Index were found to have
increased from 18 to 24. The properties of the resulting textile were
compared with those of the intact ramie fibre, and the relationship
between the yield of reaction and the application parameters -
temperature, catalyst concentration, and duration of the reaction - is
described.
August 31, 2012 | http://pubs.acs.org

The chemical modification of cellulose and methods for modifying cotton have
been studied extensively in recent decades. One method for modifying cotton, the
Michael addition reaction, uses a vinyl compound to react with the cellulose chain.
Frick et al. (7-2) reported that acrylamide readily reacted with cotton under base-
catalysed conditions. The carbamoylethyl ether derivative of cellulose fibres prepared
by this reaction has good fabric properties and modified dyeing characteristics. This
carbamoylethylated cotton can be further modified; the resulting fabric can be dyed
with different classes of dyestuffs that have very little or no affinity to cellulose (5).
Ramie, which is also cellulosic in nature, is a bast fibre obtained from the
stems of the plants Boehmeria nivea or Boehmeria tenacisseama. The fibre of ramie
possesses many superior properties, which make it popular in North American and
European countries - it is strong, white, lustrous, and durable (4). Nevertheless, there
are few academic reports on the chemical modification of ramie as a means of further
improving its properties.
Based on previous work on cotton, this study aimed to determine some of the
factors involved in the application of the Michael addition to the preparation of
carbamoylethylated ramie. The physical and thermal properties of the treated and

Corresponding author.

©1998 American Chemical Society 107

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 108

the untreated ramie were compared. The thermal behavior of the treated samples was
improved, and their textile properties, including tensile strength, were retained.

Experimental

Materials. The ramie fabric used was plain weave that had been desized and scoured.
The ramie fibre had 96-98% α-cellulose on dry basis with a small amount of lignin.
Commercial acrylamide (97%) was used as received. Reagent grade sodium
hydroxide was used as catalyst.

Carbamoylethylation. Ramie fabrics weighing 5.0 grams were carbamoylethylated

by immersing them in a 200 ml aqueous solution containing acrylamide and sodium
hydroxide at specified temperatures and durations (Table 1). Thorough water washing
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

was followed by Soxhlet extraction, using water as a solvent to remove the unreacted
chemicals.

Measurements

Weight Gain. The weight gain of the fabric after carbamoylethylation was
calculated using the following equation:

Wf- Wi
Weight Gain (%)= χ 100
Wi

where W and Wj are the weights of the treated and the untreated fabrics, respectively.
f

A l l the fabrics were held at 21°C, 65% relative humidity, overnight before
August 31, 2012 | http://pubs.acs.org

measurement.

Moisture Regain. The moisture regain of sample fibres was evaluated

according to the standard procedure, A S T M D2654-89a.

Thermogravimetric Analysis. The thermal properties of the reacted fabrics

were studied by a Mettler TA2000 thermal analysis system; scanning ranged from
3
100°C to 600°C at 30K/min in nitrogen atmosphere with a flow rate of 200 cm /min.
The onset temperature and the residual amount were evaluated.

Limiting Oxygen Index. The flammability of the fabrics was determined as

Limiting Oxygen Index according to the procedures stated in A S T M D2863-87.

Tensile Properties. Tensile strength of the fabrics was measured by the

standard strip test method (ASTM D 5035) using the Instron tensile tester.

Results A n d Discussion

Influence of Reaction Conditions. Figure 1 shows that increasing the acrylamide

concentration under 5% sodium hydroxide was accompanied by an increase in weight

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 109

Table 1. Experimental Conditions of Carbamoylethylation of Ramie

Composition of Solution Reaction Parameters Weight
Experiment Acrylamide NaOH Temp. Duration Gain
Number % % °C Hrs (%>
CAA001 5 5 40 4 2.60%
CAA002 10 5 40 4 4.80%
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

CAA003 15 5 40 4 7.40%
CAA004 20 5 40 4 6.60%
CAA005 25 5 0 4 0.80%
CAA006 25 5 10 4 1.00%
CAA007 25 5 20 4 4.00%
CAA008 25 5 30 4 7.00%
CAA009 25 5 40 4 8.40%
CAA010 25 5 50 4 5.80%
CAA011 25 5 60 4 4.60%
CAA012 25 5 40 0.25 2.00%
CAA013 25 5 40 0.5 3.60%
CAA014 25 5 40 0.75 4.60%
August 31, 2012 | http://pubs.acs.org

CAA015 25 5 40 1 5.00%
CAA016 25 5 40 1.5 5.60%
CAA017 25 5 40 2 7.00%
CAA018 25 5 40 3 8.40%
CAA019 25 5 40 4 8.80%
CAA020 25 5 40 6 8.00%
CAA021 25 5 40 8.33 7.00%
CAA022 25 5 40 10 7.20%
CAA023 25 0.5 40 4 0.60%
CAA024 25 2.5 40 4 3.40%
CAA025 25 5 40 4 8.60%
CAA026 25 7.5 40 4 11.00%

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 110

gain. At lower concentration, 0-10% of acrylamide, the weight gain increased almost
linearly with the increase in concentration. At higher concentrations, as the molar
ratio of catalyst to acrylamide became smaller, the increase in weight gain became
non-linear and less significant.

10.00 -
9.00 -
^ 8.00 -
g 7.00 -

Ο 5.00 -
% 4.00 -
'I
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

3.00 ;
2.00 -
1.00
0.00 ^
0 5 10 15 20 25
Cone, of Acrylamide (%)
Figure 1. Effect of Concentration of Acrylamide on Weight Gain of Fabric.

Khalil et al. (5) reported the following reactions when starch was
carbamoylethylated in a mixture of cellulose, acrylamide, sodium hydroxide, and
water:
NaOH
August 31, 2012 | http://pubs.acs.org

CELL—OH + CH =CH ^ 2 CELL—O—CH —CH 2 2 (1)

I I
CONH 2 CONH 2

NaOH
CELL—O—CH —CH 2 2 + H 0 • CELL—O—CH —CH
2 2 2 + N H (2)
3

CONH 2 COONa

N a H
CH =CH2 + H 0 2 ° » CH =CH
2 + NH 3 (3)
CONH 2 COONa
NaOH
CELL—OH + CH =CH 2 • CELL—O—CH —CH 2 2 (4)
COONa COONa

Khalil et al. (5) also noted that the extent of the reactions depended upon the
temperature of the reaction, the concentration of the catalyst, and the duration of the
reaction. Our experiments were designed to study the effect of these same three
parameters on the carbamoylethylation of ramie in terms of weight gain of fabric after
reaction.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Ill

Figure 2 shows the effect of reaction temperature on the weight gain of ramie
fabrics treated with 25% acrylamide and 5% sodium hydroxide for 4 hours. It is
apparent that an optimum reaction temperature of 40°C gave the maximum fabric
weight gain. In the range of 0 to 40°C, the higher the temperature employed, the faster
the reaction rate, and the greater the weight gain.
10.00 -[-
9.00 ·
^ 8.00 I
£ 7.00 J
Ο 5.00 \
1
f> 4.00
I
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

3.00
2.00 r
1.00 Î *
0.00 4 f λ i 1 1 1

0 10 20 30 40 50 60
Reaction Temperature (°C)

Figure 2. Effect of Reaction Temperature on Weight Gain of Fabric.

When the temperature was higher than 40°C, the weight gain was lower
because of hydrolysis of the ether linkages of carbamoylethylated ramie (6) and the
loss of beta and gamma cellulose of ramie in the presence of alkali. The activation
August 31, 2012 | http://pubs.acs.org

energy of the reaction was found by plotting the logarithm value of percentage weight
gain versus 1/T (Figure 3).
2.50 r

•
2.00
s—\

a
•
8 1-50 J

Î 1.00 y = -5.7005X + 20.549

& 0.50
0.00
3J0 3.20 3.30 3.40 3.50 3.60 * 3.70
-0.50
1/Τ(χ10Κ)

Figure 3. A Plot of In (% Weight Gain) vs 1/T.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 112

The activation energy, E , was calculated by the equation:

A

δ In (% Weight Gain)
E A = -R

The activation energy of the carbamoylethylation of acrylamide on ramie

cellulose was 47.40 kJmol-1.
In order to develop efficient reaction conditions, the duration of reaction time
was studied (Figure 4). As reaction time increased to 4 hours, weight increased
significantly, obviously because of the increase in the reaction and interaction of
acrylamide molecules with cellulose hydroxyl groups.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

10.00

8.00

.9 6.00 f
J
03

Ο ψ
f> 4.00 J *
2.00

0.00
August 31, 2012 | http://pubs.acs.org

0 2 4 6 8 10
Reaction Time (Hrs)
Figure 4. Effect of Reaction Time on Weight Gain of Fabric.

As reaction time increased beyond 4 hours, weight gain decreased, probably

because of the hydrolysis of acrylamide in the solution and the partial hydrolysis of
ether linkages of carbamoylethylated ramie, i.e. de-etherification, as suggested by
Ibrahim etal. (7)
Figure 5 shows the relationship between the concentration of the catalyst used
and the percentage weight gain. The higher the concentration of NaOH, the greater the
weight gain, until the NaOH concentration reached 7.5%, after which gelling occurred
on the surface of the fabric. The gelling is due to the extensive crosslinking of
acrylamide with cellulose, as well as to self-polymerization under highly alkaline
conditions (8-9). Thus, in this study the highest possible concentration of catalyst was
5.0%, offering the least significant side reactions.

Moisture Regain. Figure 6 shows the relationship between the weight gain and the
moisture regain of carbamoylethylated ramie. Although the increase of moisture
regain was not very significant when compared to the untreated ramie, the
carbamoylethylated ramie generally showed an enhancement of moisture regain.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 113

Thus, the hydrophilic properties of cellulose increased with the addition of

amide groups, whereas the addition of other vinyl monomers caused a hydrophobic
reaction (10).

12.00

10.00

8.00
α
•a
ο 6.00
-g
4.00
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

2.00

0.00
0 2 4 6 8
Concentration of NaOH (%)
Figure 5. Effect of Concentration of Sodium Hydroxide on Weight Gain of
Fabric.

8.00

7 50
gr -
August 31, 2012 | http://pubs.acs.org

I 7.00
00

i
(2 6.50
Ο
6.00

5.50 \

5.00 - I —
0.00 2.00 4.00 6.00 8.00 10.00
Weight Gain (%)

Figure 6. Relation of Moisture Regain of Carbamoylethylated Ramie with

Weight Gain.

Thermal Behavior of Carbamoylethylated Ramie. The results of the

thermogravimetric analysis of thermal behavior are given in Table 2.
As the weight gain of the fabric increased, the onset temperature decreased.
This resulted in early decomposition of the fibre and delay of the decomposition rate,
so the residual amount of the treated ramie was higher than that of the untreated ramie.
The addition of amide groups to the cellulose chain thus improved the thermal

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 114

behavior of ramie, as was reported by Shimada and Nakamura (77) for the graft
copolymerization of acrylamide for cotton and for the carbamoylethylation of cotton.

Evaluation of Flammability by Limiting Oxygen Index (LOI). The LOI test

determines the minimum concentration of oxygen in a flowing mixture of oxygen and
nitrogen that will just support a flaming combustion of the material. The limiting
oxygen index, expressed as volume percentage, is calculated as follows:

100 χ 0~
«o/. — L·

where n% = the limiting oxygen index, 0 = volumetric flow rate of oxygen in

2
3 3
mm /s, N = corresponding volumetric flow rate of nitrogen in mm /s.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

As the weight of the fabric increased, its flammability became correspondingly

higher, as shown in Table 2. These results confirm the results of the
thermogravimetric analysis, which showed improved thermal stability for
carbamoylethylated ramie.

Table II. Thermal Behavior of Carbamoylethylated Ramie

Weight Onset Residue
Experiment Gain Temperature Amount LOI
No. (%) (°C) (%) (%)
Control 0 360.10 14.51 18
CAA005 0.80 357.90 16.43 19
August 31, 2012 | http://pubs.acs.org

CAA013 3.60 357.20 15.07 20

CAA011 4.60 341.10 17.08 21
CAA002 4.80 336.30 18.22 21
CAA022 7.20 337.70 19.41 22
CAA009 8.40 336.80 22.86 22
CAA019 8.80 330.30 21.19 22
CAA026 11.00 325.40 30.04 24

During the pyrolytic degradation of carbamoylethylated ramie, the amide

group will decompose to ammonia. The released ammonia, which is an energy-poor
fuel, exhibits flame retardant effects by several modes (72-75). The flame inhibitory
action is a result of the concentration dilution of the released volatile combustible
material from the decomposed cellulose, i.e. the levoglucosan, during the combustion.

Tensile Properties of Carbamoylethylated Ramie Fabrics. Tensile strength of the

fabrics was measured by standard method A S T M D 5035. The breaking strength of
carbamoylethylated fabrics gradually decreased with increasing weight gain. The

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 115

strength loss ranged from 0 to a maximum of 23% (Figure 7). This result was similar
to that of the carbamoylethylated cotton (14).

100.00 ^ •
90.00
80.00 4- •
Ο
70.00 -
•fi 60.00
50.00 - -
40.00
30.00
20.00
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

10.00 ι
0.00 1

0.00 2.00 4.00 6.00 8.00 10.00

Weight Gain (%)
Figure 7. Strength Retention of Carbamoylethylated Ramie.

The loss of tensile strength is due to the addition of amide groups, which are
bulkier than the hydroxyl groups, thus reducing the extent of hydrogen bonding
between cellulose chains. Nevertheless, the loss of strength is not very significant, as
amide groups can also form hydrogen bonds between cellulose chains.

Conclusions
August 31, 2012 | http://pubs.acs.org

The application of Michael addition was useful in modifying ramie fibre by using the
all-in exhaustion method. The best yield of the reaction between ramie cellulose and
acrylamide was obtained at 40°C, a 4-hour reaction, in the presence of 5% NaOH. The
carbamoylethylated ramie had better hydrophilicity and thermal properties without
significant loss of fabric strength after the reaction.

Acknowledgment

The authors thank The Hong Kong Polytechnic University for the financial support of
this project.

References

1. Frick, J.W.; Reeves, W.A.; Guthrie, J.D. Text. Res. J. 1957, 27, No.2, pp.92-99.
2. Frick, J.W.; Reeves, W.A.; Guthrie, J.D. Text. Res. J. 1957, 27, No.4, pp.294-
299.
3. Abou-zeid, N.Y.; Anwar, W.; Hebeish, A . Cell. Chem. Technol. 1981, 15,
pp.321-330.
4. How, Y.L.; Cheng, K.P.; Lau, M.P. Textile Asia 1991, 22, No.5, pp.74-78.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 116

5. Khalil, M.I.; Bayazeed, Α.; Farag, S.; Hebeish, A. Starch/Stärke 1987, 39, No.9,
pp.311-318.
6. Feit, B.A.; Zilkha, H. J. Org. Chem. 1963, 28, p.406.
7. Ibrahim, N.A.; Haggag, K.; Abo-Shosha, M . H . Am. Dyestuff Reptr. 1988, 77,
No.7, pp.34-42.
8. Thomas, W . M . and Wang, D.W. In Encyclopedia of Polymer Science and
Engineering, Herman F. Mark. Ed., 2nd Edition, John Wiley & Sons: New
York, 1985, p.185.
9. Kurenkov, V.F. and Myagchenkov, V . A . In Polymeric Materials Encyclopedia,
Joseph C. Salamone. Ed., CRC Press, Inc., 1996, Vol.1, pp.47-54.
10. Hebeish, Α.; Guthrie, J.T. The Chemistry and Technology of Cellulosic
Copolymers, Springer-Verlag, Berlin-Heidelberg-New York, 1981, p.295.
11. Machiko Shimada and Yoshio Nakamura In Inititation ofpolymerization, ACS
symposium series 212, Frederick E. Bailey, Jr. Ed., Bailey, March/April, 1982,
pp.237-248.
12. Miller, D.R.; Evans, R.L.; Skinner, G.B. Comb. Flame 1963, 7, pp.137-142.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch008

13. Haynes, B.S.; Jander, H.; Mätzing, H.; Wagner, H.G. 19th Symp. (Intl.) on
Combustion, The Combustion Institute, 1982, pp.1379-1385.
14. Grant, J.N.; Greathouse, L.H.; Reid, J.D.; Weaver, J.W. Text. Res. J 1955, 25,
N o . l , pp.76-83.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 9

Characterization of Polysaccharide Derivatives

with Respect to Substituent Distribution
in the Monomer Unit and the Polymer Chain

Petra Mischnick

Department of Organic Chemistry, University of Hamburg, Martin-Luther-King­

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

-Platz 6, D-20146 Hamburg, Germany

Besides other structural parameters the distribution of substituents

greatly influences the properties of polysaccharide derivatives. To
characterise the substitution pattern in the monomer unit and in the
polymer chain, different approaches can be used. Strategies of
monomer analysis by gas chromatography and mass spectrometry
after chemical degradation will be presented. Examples of alternative
separation methods will be given. The distribution along the poly­
saccharide chain was studied on methyl amyloses as model
compounds. Oligomeric mixtures obtained after partial degradation
were analysed by mass spectrometry. The results were interpreted by
comparison with a homogeneous distributrion calculated from the
August 31, 2012 | http://pubs.acs.org

monomer composition. First applications on cellulose sulfates and

acetates are reported.

Polysaccharides are very well suited for their functions in nature, e.g. to build up
mechanically stable cell walls in plants. However, to get new materials with new
properties cellulose and starch as aie main sources of renewable polymers have been
modified in different ways since more than 150 years now. The primary structure of
those derivatives is responsible for the formation of higher molecular architectures as
helices, aggregates in solution, liquid crystalline phases or monomolecular layers.
Properties as thickening, gelation, film building or flocculation are mainly influenced
by the distribution of substituents in the polymer chain, while effects involving
molecular recognition also show a dependence on the regioselectivity of substitution
in the monomer unit. Therefore, analysis in this field has to aim at a determination of
all structural features for a better understanding of the relationship between the
reaction conditions, the primary structure and the resulting properties.

Monomer Analysis of Cellulose Derivatives. Cellulose is a linear homopolymer of

1,4-linked β-D-ghicose residues. By a partial derivatisation of the O H functionalities it
becomes a "copolymer" of up to eight different monomers, if one type of substituent is
introduced. Due to the high DP the ends of the chains can be neglected. Then, 3
positions (C2, C3, and C6, m = 3) per anhydro glucose unit (AGU), which can be

118 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 119

m 3
either O H or OR (n = 2), result in n = 2 = 8 different monomer patterns. If two
types of substituents are present (n = 3) as in mixed esters as acetates/butyrates
3^ = 27 different monomer units are possible.
The distribution in the A G U , especially the distribution on the positions 2, 3
1 3
and 6, have been determined for a number of different cellulose derivatives by C
N M R spectroscopy (see chapter x). The alternative approach includes degradation of
the polymer to monomers without any discrimination, to yield a mixture which is
representative for the original copolymer. The constituents of this mixture in an
appropriate form are then separated by an efficient chromatographic method. After
peak assignment by combined mass spectrometry or comparision with authentic
standard compounds, the relative molar ratios of all components are calculated from
their peak areas after correction for their individual detector response (Figure 1).
How can this be achieved in practice? The polysaccharide derivative is usually
permethylated in the first step to protect and to sign all free hydroxy groups.
Furthermore, by this step all hydrogen bonds are broken, the derivative can be really
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

dissolved, and all A G U should be accessible. Then, this mixed derivative is submitted
to subsequent hydrolysis, reduction and acetylation to get partially methylated,
partially substituted glucitol acetates (Figure 2, standard methylation analysis). An
alternative type of methylation analysis is shown in Figure 3. In the reductive-
cleavage-method, introduced by Rolf and Gray (7), the glucosidic linkages are cleaved
under promotion of a Lewis-acid. The cyclic carboxonium ions formed are in situ
reduced to yield the corresponding 1,5-anhydroglucitols which are usually acetylated
in this one-pot-reaction sequence. A mixture of volatile compounds is obtained,
appropriate for capillary gas chromatography with its high separation efficiency. A
further advantage of GC is the combination with a flame ionization detector (FID).
The response of this detector in principle corresponds to the number of C-atoms
present in the compound. Electronegative atoms reduce the response of a carbon for a
certain amount. Therefore, the response values of the individual components can be
calculated by an experimentally established increment system [effective-carbon-
concept, (2-4)]. In addition, the combination with a mass spectrometer (GC/MS)
allows the elucidation of the structure and the differentiation of regioisomers.
August 31, 2012 | http://pubs.acs.org

Therefore, the tedious synthesis of standard compounds is not essential. These

procedures have been applied to alkyl ethers (5,6) hydroxyethyl- and hydroxyproypl
derivatives (7-10).
If two or more types of substituents are present in the cellulose derivative, the
situation becomes much more complex, e.g. for ethylhydroxyethylcellulose [ErIEC,
(77)] or hydroxypropylmethylcelhilose (HPMC). Without tandem substitution (oligo-
ether formation of the hydroxyalkyl groups) 4 different types of OR are possible in the
3
A G U : OH, OMe, OHP, and O M P (methoxypropyl). That means that 4 = 64
monomer patterns can theoretically occur. Due to the asymmetric C in the HP residue,
HP derivatives occur in two diastereomeric forms, di-O-HP (or MP-) substituted A G U
give four diastereomers and so on. At higher MS values, tandem-substitution will
cause additional products of theoretical unlimited number. We have investigated a
series of HPMC with a D S of 2 and a MSHP of about 0,2.
M e

To avoid that small amounts of O-HP and O-MP ethers, which are spread over
glucose derivatives with all different methylation patterns, do not escape detection, we
used a twofold approach (Mischnick, P.; Dônnecke, J., unpublished results). To
analyse the methyl pattern, which was responsible for about 90% of the total DS, the
cellulose ethers were directly hydrolysed, reduced and acetylated. The partially
methylated glucitol acetates are eluted first in the gas chromatogram (Figure 4a) and
could easily be assigned and calculated. The main peaks of glucitols bearing hydroxy-
propyl groups and methyl groups could be identified by GC-MS. For the analysis of
the HP pattern the sample was permethylated prior to degradation to focus all
monomers with a HP/MP group in a certain position in one peak: for example 1,4,5-
tri-0-acetyl-2-0-(2-methoxy)propyl-3,6-di-0-methyl-D-^ucitol includes the 2-O-HP

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 120
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

ι 1
August 31, 2012 | http://pubs.acs.org

Figure 1
General Approach for the determination of the substituent distribution in the
monomer unit of polysaccharide derivatives

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 121
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009
August 31, 2012 | http://pubs.acs.org

Figure 2
Standard methylation analysis of cellulose derivatives

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 122
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009
August 31, 2012 | http://pubs.acs.org

Figure 3
Reductive-cleavage of cellulose derivatives

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 123
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009
August 31, 2012 | http://pubs.acs.org

Figure 4
(a) GC of HPMC after hydrolysis, reduction and acetylation and (b) after
permethylation, hydrolysis, reduction and acetarytion according to Figure 2.
Peaks are assigned according to the substituted position in the A G U

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 124

ethers with eight different methyl patterns (Figure 4b). So the sensitivity is significantly
enhanced.
Figure 5 shows the results obtained for the methyl pattern and the distribution
of the HP groups summerized for the positions 2, 3 and 6. The complete monomer
compositon only slightly deviates from the model of Reuben (12-13) (Figure 6). This
kinetical model already includes the enhanced reactivity of 3-OH in 2-O-methylated
A G U . Prefered 2-Omethylation is observed, which is favoured by low alkali
concentration during the reaction. Then the 2-OH as the most acidic one is the most
reactive. In the competition with the methyl iodide the HP groups are directed to the
steric less hindered primary 6-OH.
The procedures just outlined require, that the substituents are stable under the
reaction conditions applied. If this is not the case, as for trialkylsuyl ethers, acetates or
sulfates, an indirect determination of the substitution pattern by the analysis of the
complementary methyl pattern is performed. Trialkylsuyl ethers of cellulose as
TBDMS- or THxDMS-derivatives are important intermediates in the synthesis of
regioselectrvely substituted cellulose derivatives (see chapter 1). During the hydrolysis
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

or the reductive-cleavage of the permethylated samples the saryl groups are cleaved
and partially methylated (anhydro)ghicitol acetates are obtained. However, during the
analysis of a series of cellulose-, cyclodextrin-, and amylose derivatives it turned out,
that the siryl groups nearly quantitatively migrate from 0-2 to 0-3 in ct-ghicans under
the alkaline methylation conditions, while only 5% or less rearrangement was observed
for the β-linked cellulose (14-15) This different behaviour may be caused by stereo-
electronic effects (a versus β) and conformational differences. Methyl 2-0-TBDMS-
β-D-ghicopyranoside shows about 60% rearrangement of the siryl group. With methyl
inflate as the methylating agent no migration occured as expected.
Sulfation of several polysaccharides as laminarin, schizzophyllan or cellulose
plays an important role in the research for heparinoidic materials or drugs with
antiviral properties. The ionic sulfate groups shall also help to get soluble products.
Nehls et al. reported, that a DS of 0,3 was efficient to achieve water solubility for
cellulose sulfates from a homogeneous reaction, while a DS of 1,8 was necessery for
samples produced under heterogeneous conditions (16). Methylation anaryis (Figure 2)
August 31, 2012 | http://pubs.acs.org

can be applied (Gohdes, M . ; Mischnick, P. Carbohydr. Polym. in press). However,

problems arise from the residues sulfated in position 2, which can be displaced by an
intramolecular nucleophilic attack of the vicinal 3-0" anion under the alkaline
methylation conditions. In addition, these residues are preferably hydrolysed and
partially lost during long hydrolysis times. For such samples the method of Stevenson
and Furneaux reported for polysaccharides from algae (17) is the method of choice.
Acyl groups are labile under both alkaline methylation conditions and acid
hydrolysis. Methylation can be achieved with methyl triflate or trimethyloxonium
tetrafhioroborate and 2,6-w-ter/.-butyl-pyridine (18). In a similar etherification
reaction with benzyl triflate neither migration nor cleavage of acyl groups was
observed (19-20). In a subsequent alkaline alkylation step the acyl residues are
exchanged against ethyl or higher alkyl groups. The resulting mixed ether can be
analysed as described above. Lee and Gray have reported a direct analysis of the
methylated cellulose acetates by reductive-cleavage (21). With a high excess of the
reagents, the acyl groups are reduced to the corresponding alkyl ethers (OAc
* OEt) (22). This method also allows the analysis of mixed esters (Gray, G. R ,
University of Minneapolis, unpublished data).
Besides gas chromatography HPLC has been used for the determination of the
molar fractions of un-, mono-, di- and trisubstituted glucose residues (23), which
complements the data for 2 : 3 : 6 substitution obtained bei N M R spectroscopy.
Anionic deriviates as carboxymethyl cellulose (CMC) and sulfoethyl cellulose (SEC)
have been analysed by High-pH anion exchange chromatography (HPAEC) after
hydrolysis (24). We recently succeeded in the separation of partially methylated
glucoses, which are still soluble in water (25). In contrast to gas chromatography all

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 125
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

Figure 5
Distribution of methyl groups and hydroxypropyl groups in five H P M C
August 31, 2012 | http://pubs.acs.org

Figure 6
Comparison of the monomer composition (methyl pattern) of one H P M C with the
calculated ratios according to the model of Reuben (73)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 126

standard compounds had to be synthesized for peak assignment and determination of

the response in the pulsed amperometric detector (PAD). With every blocked hydroxy
function the response dramatically decreases.
Cationic 0(2-hydroxy-3-trimethyl ammonium)propyl ether derivatives could
be transformed to neutral 0-(2-methoxy-2-propenyl) ethers by Hofinann-eliinination,
which could be analysed by GC and GC/MS (26).

Substitution Pattern in the Polymer Chain. In contrast to the analysis of the

monomer pattern only a few studies on the substituent distribution in the polymer
chain have been reported in the literature. Gelman (27) and Wirick (28) treated
hydroxyethylcelhilose (HEC) and C M C with celhilases and determined the number of
chain cleavages/1000 A G U . These data were compared with theoretical values
calculated from the amount of unsubstituted A G U ( * average block length of
unsubstituted A G U ) and the assumed enzyme specifity. Steeneken et al. reported on
different enzyme accessibility of methyl starches produced under homogeneous or
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

heterogeneous conditions (29) and on topochemical effects caused by the layered

structure of the starch granules (30). Recently, a new approach involving partial
hydrolysis, FAB-MS analysis of the oligomeric mixtures obtained after perdeutero-
methylation, and statistical evaluation was published by Arisz et al. for methyl
celluloses (37). Furthermore, pyrolysis chemical ionisation M S and pricipal component
analysis was applied to HEC (32).
We performed model studies on methyl amyloses, which were prepared under
homogeneous or heterogeneous reaction conditions in protic or aprotic solvent
systems (33). The analytical strategy applied included (a) monomer analysis and
comparison of the experimental data with the kinetical model of Reuben as described
above, (b) perdeuteromethylation of the free O H groups to get a chemical uniform
product and prove a random degradation process, (c) partial depolymerisation by mild
methanolysis or mild reductive cleavage, (d) mass spectrometric analyis by FAB-MS
or MALDI-TOF-MS, and (e) comparison of the substitution patterns in the oligomeric
mixtures with a calculated statistical arrangement of the monomers. The average DS
of each oligomeric mixture has to agree with the average DS of the sample, to
August 31, 2012 | http://pubs.acs.org

confirm, that its substituent pattern is representative for the whole sample. For a
methyl amylose, which was prepared in a homogeneous solution in water, the
calculated statistical composition and the experimental data fitted very well. Samples
from a heterogeneous reaction showed a broader distribution. When amylose was
dissolved in DMSO and treated with powdered NaOH and methyl iodide a bimodal
distribution pattern was observed as the result of two competing methylation
processes. The amylose molecules are adsorbed from the solution on the solid surface
of the NaOH particles according to the model of Fleer and Scheutjens (34). There, the
deprotonation and subsequent methylation occured very fast while the not adsorbed
molecules are nearly randomly methylated with a lower rate. Filtration and titration of
the NaOH suspensions in DMSO showed that the NaOH is not really dissolved in
DMSO. The smallest particles were found to have a diameter of about 280 nm by light
scattering experiments (Mischnick, P., Burchard, W., unpublished data). Larger
aggregates of these particles were also present. While amylose was completely
recoverd from a solution in DMSO after filtration (16 μπι), it was strongly retained in
the presence of powdered NaOH. The extent depended on the ratio of amylose and
NaOH and the particle size of the latter. Liebert and Heinze (see chapter y) also
reported on an enhanced ratio of the trisubstituted fraction in C M C from such reaction
systems, which indicates the fast etherification process of the adsorbed polymer.
For a quantitative M S analysis the oligomers of a certain DP should be as
similar as possible with respect to the molecular mass and the polarity, since these
parameters mainly influence the desorption and ionisation properties. This requirement
is best met by mixed C H / C D ethers. As already described above acetates, sulfates or
3 3

cellulose silyl ethers can in principle be transformed to their complementary methyl

ethers and subsequent to the oligomeric mixed ethers. Figure 7 shows the results for a

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 127

DP3

40 τ

calculated

experimental
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

0 1 2 3 4 5 6 7 8 9

calculated

experimental
August 31, 2012 | http://pubs.acs.org

0 1 2 3 4 5 6 7 8 9 101112

n(CD) 3

Figure 7
Distribution of C D groups (corresponding to the original sulfate groups) in the
3

tri- and tetrameric traction of a cellulose sulfate from a homogeneous reaction

cellulose sulfate with preferred 6-sulfation from a homogeneous reaction in the

N 0 / p M F system (35). The average DS of the oligomers deviated about 10%,
2 4

indicating that the partial hydrolysis with simultaneous cleavage of the sulfate group
was not really random Therefore, desulfation and perdeuteromethylation should be
performed prior to partial degradation. However, a relative good agreement with the
calculated homogeneous statistical distribution pattern along the chain can already be
realised. These investigations are in progress. First results for a commercial cellulose
acetate (DS 2) show a heterogeneous distribution of the type Β in Figure 8.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 128

Distribution pattern Type

26 · A

»"t / Χ Random distribution in the polymer chain

0 1 2 3 4 5 6 7 8 9 10 11 12

26 · Β
More heterogeneous distribution in the
polymer chain compared to a random
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

distribution: regions with high and low DS

enhanced, oligomers of the average DS
diminished
0 1 2 3 4 6 0 7 8 9 10 11 12

ΙΑ.
More regular distribution in the polymer
chain compared to a random distribution:
regions with high and low DS diminished,
oligomers of the average DS enhanced
highest regularity is reached for repeating
0 1 2 3 4 6 8 7 8 9 10 11 12 units

26 :
D
August 31, 2012 | http://pubs.acs.org

Distorted heterogeneous distribution,

indicating a DS-gradient in the sample.

Ε
»I A
Bimodal distribution resultingfromtwo
parallel reaction processes

0 1 2 3 4 5 8 7 8 9 1 0 11 12

Figure 8
Classification of substitution patterns demonstrated for the tetrameric fraction of
a theoretical polysaccharide derivative with a DS 1,5. The plane graph shows the
random distribution as a reference. 0, 1, 2, 3,.... 12 is the number of substituents in
2-, 3-, and 6-position. (Reproduced with permission from ref. 33, Copyright
1996, Elsevier Science Ltd)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 129

The random substituent distribution in the polysaccharide chain, which is

calculated for a certain monomer composition, always serves as a defined reference
structure. The typical deviations have been classified in a qualitative sense (Figure 8).
Mathematical models have to be developed, which include the positive or negative
intermonomeric effects, e.g. the enhanced rate of the reaction due to a local enhanced
solubility and consequently accessibility after the introduction of the first substituents,
or a real cooperative effect on the vicinal AGUs, which propagates along the chain,
when a higher structure of the polymer in the solution has been disturbed.

Conclusion

While the monomer analysis of polysaccharide derivatives is well established, the

determination of the substitution pattern on higher structural levels and therefore a
molecular understanding of the macroscopic properties is still in the beginning.
Analytical strategies for the determination of the monomer compositon of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

cellulose derivatives have been outlined for alkyl and hydroxyalkyl ethers, chemically
less stable substituents as siryl ethers, sulfates or organic esters, and also for anionic
and cationic species. With respect to the substituent distribution in the polymer chain
model studies with methyl amyloses have given promising results and a first insight in
the influence of the reaction system on the primary structure of the derivatives
obtained. The analytical approach applied to the methyl ethers is also appropriate for
acetates sulfates or siryl ethers.

Literature Cited

(1) Rolf, D.; Gray, G. R. J. Am. Chem. Soc. 1982, 104, 3539-3541.
(2) Sweet, D. P.; Shapiro, R H . ; Albersheim, P. Carbohydr. Res. 1975, 40, 217-
225.
(3) Scaidon, J. T.; Willis, d. E. J. Chromatogr. Sci. 1985, 23, 333-339.
(4) Jorgensen, A. D.; Picel, K. C.; Stamoudis, V. C. Anal. Chem. 1990, 62, 683-
August 31, 2012 | http://pubs.acs.org

689.
(5) D'Ambra, A . J.; Rice, M . J.; Zeller, S. G.; Gruber, P. R.; Gray, G. R. Carbohydr.
Res. 1988, 177, 111-116.
(6) Mischnick-Lübbecke, P.; Krebber, R Carbohydr. Res. 1989, 187, 197-202.
(7) Lindberg, B.; Lindquist, U . ; Stenberg, O. Carbohydr. Res. 1987, 170, 207-214.
(8) Mischnick, P. Carbohydr. Res. 1989, 192, 233-241.
(9) Steeneken, P. A . M.; Woortman, A. J. J.; Tas, A. C.; Venekamp, J. C.
Carbohydr. Netherlands 1993, 9, 31-34.
(10) Arisz, P.W.; Lomax, J.Α.; Boon, J.J. Carbohydr. Res. 1993, 243, 99-114.
(11) Lindberg, B.; Lindquist, U . ; Stenberg, O. Carbohydr. Res. 1988, 176, 137-144.
(12) Reuben, J. Macromolecules 1984, 17, 156-161.
(13) Reuben, J.; Casti, T.E. Carbohydr. Res. 1987; 163, 91-98.
(14) Mischnick, P.; Lange, M.; Gohdes, M.; Stein, Α.; Petzold, Κ. Carbohydr. Res.
1995, 277, 179-187.
(15) Icheln, D.; Gehrcke, B.; Piprek, Y.; Mischnick, P.; König, W.A.; Dessoy, M . A . ;
Morel, A. F. Carbohydr. Res. 1996, 280, 237-250.
(16) Nehls, I.; Phihpp, B.; Wagenknecht, W.; Klemm, D.; Schnabelrauch, M.; Stein,
Α.; Heinze, T. Das Papier 1990, 44, 633-640.
(17) Stevenson, T.T.; Furneaux, R.H. Carbohydr. Res. 1991, 210, 277-298.
(18) Mischnick, P. J. Carbohydr. Chem. 1991, 10, 711-722.
(19) Lemieux, R. U . ; Kondo, T. Carbohydr. Res. 1974, 35, C4-C6.
(20) Berry, J. M.; Hall, L. D. Carbohydr. Res. 1976, 47, 307-310.
(27) Lee, C. K.; Gray, G. R. Carbohydr. Res. 1995, 269, 167-174.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 130

(22) Mischnick, P. Min. 5th Intern. Symp. Cyclodex. Ed. de Santé, Paris 1990, 90-
94.
(23) Erler, U.; Mischnick, P.; Stein, Α.; Klemm, D. Polym. Bull. 1992, 29, 349-356.
(24) Kragten, Ε. Α.; Kamerling, J. P.; Vliegenthart, J. F. G. J. Chromatogr., 1992,
623, 49-53.
(25) Heinrich,J.;Mischnick, P. J. Chromatogr. A 1996, 749, 41-45.
(26) Wilke, O.; Mischnick, P. Carbohydr. Res. 1995, 275, 309-318.
(27) Gelman, R. A. J. Appl. Polym. Sci. 1982, 27, 2597-2964.
(28) Wirick, M. G. J. Polym. Sci., Part A 1968, 6, 1705-1718 and 1965-1974.
(29) Steeneken, P. A. M.; Woortman, A. J. J. Carbohydr. Res. 1994, 258, 207-221.
(30) Steeneken, P. A. M.; Smith, E. Carbohydr. Res. 1991, 209, 239-249.
(31) Arisz, P.W.; Kauw, H. J.J.;Boon, J.J.Carbohydr. Res. 1995, 271, 1-14.
(32) Arisz, P. W. Thesis: Mass spectrometry analysis of cellulose ethers,
Amsterdam, 1995.
(33) Mischnick, P.; Kühn, G. Carbohydr. Res. 1996, 290, 199-207.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch009

(34) Scheutjens, J. M. H. M . ; Fleer, G. J. J. Phys. Chem. 1980, 84, 178-190.

(35) Wagenknecht, W.; Nehls, I.; Philipp, B. Carbohydr. Res. 1993, 240, 245-252.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 10

Characterization of Cellulose Esters by Solution-State

and Solid-State NMR Spectroscopy

Douglas W. Lowman

Research Laboratories, Eastman Chemical Company, Kingsport, TN 37662-5150

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

The characterization of organic cellulose esters, including acetyl,

propionyl, and butyryl esters, mixed cellulose esters, and acylated
cellulose ethers by modern solution-state and solid-state N M R
techniques over the past 10 - 15 years is reviewed. Modern 1D and 2D
NMR techniques enable detailed structural elucidation of these
heteropolymers. The importance of molecular characteristics
determined by N M R , such as total and site-specific degree of
substitution, solution conformation, and molecular dynamics as well as
crystalline and amorphous content, are discussed in terms of structure­
-property relationships for these cellulosic polymers.
August 31, 2012 | http://pubs.acs.org

The major organic cellulose ester derivatives, including acetate, propionate, and
butyrate, have been important commercial products for many years. They find
applicability in plastics as well as in biodegradable polymers. Study of these acylated
polymers in relation to various properties has not been possible until recently due to
the lack of detailed structural information on these very complex, heterogeneously
substituted homopolymers and heteropolymers. With the development of new
structure elucidation techniques in nuclear magnetic resonance spectroscopy (NMR),
these detailed analyses are becoming possible.
Cellulose esters are polymers resulting from acylation of cellulose. Cellulose (1)
is a linear 1,4-p-D-glucan with three hydroxyl groups per anhydroglucose unit (AGU).
Each AGU contains hydroxyl functions at the 2-, 3- and 6-positions. Acylation can
occur at none of the hydroxyl positions, at any one of the three hydroxyl positions, at
any two of the three hydroxyl positions, and at all three hydroxyl positions, resulting

1 R = H, Cellulose r 6^0R 0 R η
2 R = Acetyl, CTA ^ - 0 Λ ^ ° \ ΐ η 7^7^--
u /
3 R = Propionyl, CTP R 0 - A ^ ^ ^ ^ 0 η
4 R = Butyryl, CTB L 3 υκ \ ^
© 1998 American Chemical Society 131

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 132

in the possible formation of 8 different AGU's in a cellulose ester polymer (Figure 1).
If all three hydroxyl positions are acylated, the triacylated cellulose ester
homopolymer has a degree of substitution (DS) of 3. Depending on how the acylation
is accomplished, either by direct acylation or back-hydrolysis after acylation from a
DS = 3 homopolymer, the possible range of DS in the heteropolymer containing any
combination of the 8 possible AGU's is between 0 and 3. The complexity of the
chemical structure of the cellulose ester heteropolymer, and thus its properties,
conformation and dynamics, is related to the polymer DS.
N M R techniques have been successfully developed and applied to the analysis of
structure-property relationships, sequence determinations, molecular dynamics, and
conformations in biological heteropolymers, such as peptides, proteins and enzymes.
These same techniques are now being used for similar analyses of cellulose esters.
Structural details available from these modern N M R techniques have proven very
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

useful in aiding our understanding of structure-property relationships, conformational

properties, and the molecular dynamics of these polymers.
It is the intent of this chapter to discuss the characterization of organic cellulose
esters, including acetyl, propionyl, and butyryl esters, by modern solution-state and
solid-state N M R techniques in terms of and chemical shifts, total and site-
specific DS, solution conformation, molecular dynamics, and crystalline allomorphs.
In this chapter, inorganic cellulose esters, such as nitrate, sulfate, and phosphate, will
not be considered. The application of N M R to related studies on acylated cellulose
August 31, 2012 | http://pubs.acs.org

Figure 1. Eight A G U ' s present in Cellulose Acetate

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 133

ethers will be discussed. The chapter will begin with an analysis of applications of ID
and 2D solution-state N M R techniques to C T A (2), CTP (3), CTB (4), cellulose mixed
esters, and cellulose ethers. The chapter will conclude with an analysis of solid-state
N M R techniques applied to cellulose esters. A discussion of the N M R analysis of all
known cellulose esters is outside the scope of this chapter. However, a review
appeared recently discussing the characterization of cellulose and cellulose derivatives,
including esters, by N M R (7). Kamide and Saito (2) recently reviewed the
research from the authors' laboratories between 1985 and 1993 relative to the
molecular and supramolecular characterization of cellulose and cellulose derivatives,
including cellulose acetates.

Solution-State NMR
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

Cellulose acetate (CA) is the most common commercial cellulose ester. The largest
amount of detailed information is available about this polymer's structure, DS,
conformation, and dynamics based on N M R analysis relative to the other cellulose
derivatives discussed in this chapter. The classic work of Goodlett and coworkers (5)
in 1971 provided the first method for direct determination of the acetyl distribution in
C A . They reacted the unacetylated hydroxyl groups in C A with DS less than 3 with
acetyl-d3 chloride. The proton N M R spectrum in CD2CI2 of this fully acetylated C A
presented 3 acetyl methyl proton resonances with chemical shifts of 2.09, 1.99, and
1.94 ppm assigned to substitution at the 6-, 2- and 3-positions, respectively. These
assignments were confirmed by Shiraishi and coworkers (4). Using a digital computer
method, Goodlett and coworkers (5) determined the relative DS at each site to a
standard deviation of about 0.03. This simple measurement allowed the correlation of
the site-specific DS with methods of preparation for C A .
August 31, 2012 | http://pubs.acs.org

In 1984, Shibata and coworkers (5) used ^ C N M R to measure site-specific

acetylation based on the ring carbons of C A in DMSO-d6- Acetylation at C-6 results
in deshielding of C-6. Resonances for C-2 and C-3 are not separated enough from the
C-5 resonance to allow direct observation of the impact of acetylation on these carbon
resonances. Instead, substitution at C-2 and C-3 results in shielding of the C - l and C-
4 resonances, respectively. Using the assignments of Goodlett and coworkers (5) and
comparing signal intensities, Shibata and coworkers (5) assigned the carbonyl carbon
resonances of the acetyl groups also. The three carbonyl carbon resonances at 169.9,
169.5 and 168.8 ppm were assigned to acetyl groups substituted at the 6-, 3- and 2-
positions, respectively. Later, Kowsaka, Okajima and Kamide (<5) made similar
assignments, correcting previous assignments by Kamide and Okajima (7).

Chemical Shift Assignments. In order to correlate structure and properties, chemical

shift assignments must be known in detail. Proton and ^ C N M R assignments depend
on a 2D homonuclear correlated spectrum (COSY) to assign the ring protons, then a
1 3
2D C - ! H direct-detected (HETCOR) or inverse-detected (HMQC) heteronuclear
correlated spectrum to assign the protonated ring carbons, followed by a long-range
heteronuclear correlated spectrum (COLOC) or multiple-bond HETCOR or a multiple-
bond inverse-detected heteronuclear correlated spectrum (HMBC) to assign acyl
carbon and proton resonances including carbonyl carbons.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 134

a b
Table I. Proton N M R Chemical Shifts and Coupling Constants
for CTA, CTP, CTB and Cellulose
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

CTA CTP CTB Cellulose

DMSO-d6 DMSO-d6 CDCI3 CDCI3 CDCI3 DMSO-d6

25°C (8) 80°C (8) 25°C (8) 25°C (8) 25°C (8) 80°C (26) c

H-l 4.65 (7.9) 4.65 (7.9) 4.42 (7.9) 4.35 (7.9) 4.34 (7.9) 4.35
H-2 4.52 (7.3) 4.55 (8.6) 4.79 (8.6) 4.77 (8.6) 4.76 (8.6) 3.10
H-3 5.06 (9.2) 5.04 (9.2) 5.07 (9.0) 5.07 (9.1) 5.06 (9.2) 3.38
H-4 3.65 (9.2) 3.68 (9.2) 3.71 (9.2) 3.66 (9.1) 3.61 (9.2) 3.38
H-5 3.81 3.77 3.53 3.47 3.48 3.38
H
" S 6
4.22 (10d) 4.26 (lid) e e e 3.78
August 31, 2012 | http://pubs.acs.org

H-6R 3.98 4.04 4.06 4.03 4.03 3.60

a
In ppm relative to CDCI3 at 7.24 ppm or DMSO-d6 at 2.49 ppm
b Shown in parentheses, in Hz; digital resolution 0.2-0.26 Hz
c
Unsubstituted A G U from a low DS C A
d2
J6S,6R
e
H-6s and H - l overlap

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 135

a 5
Table II. Carbon-13 N M R Chemical Shifts and Coupling Constants*
for CTA, CTP, and CTB

CTA CTP CTB

DMSO-d6 CDCI3 CDCI3 CDCI3
90°C C
25°CC 25°C 25°CC

C-l 99.8 (167) 100.4(165) 100.3 (163) c

100.37 d
100.1 (163)
C-2 72.2 (152) 71.7 (153) 71.7(150) 71.81 71.4(150)
C-3 72.9(151) 72.5 (148) 72.2 (148) 72.25 71.8 (147)
C-4 76.4(151) 75.8 (153) 75.90
C-5 72.5 (146)
76.0e
72.7 (139) 73.0 (138) 73.08
75.8 e
73.1 (143)
C-6 62.8 (151) 61.9(151) 61.9 (147) 62.04 61.9(145)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

a
In ppm relative to CDCI3 at 77.0 ppm or DMSO-d6 at 39.5 ppm.
b Shown in parentheses, in Hz; digital resolution 0.52 Hz.
c
Reference 8.
d Reference 19.
e
The coupled resonance overlaps with the solvent resonance.

The first detailed, unambiguous chemical shift assignments for the ring protons
(Table I) and carbons (Table II) of C T A , CTP and CTB dissolved in DMSO-d6 or
CDCI3 using 2D N M R techniques were provided by Buchanan and coworkers (8).
Proton assignments were accomplished with COSY (9,10) (CTA, Figure 2) while ^ C
August 31, 2012 | http://pubs.acs.org

assignments were accomplished with HETCOR (11,12) (CTA, Figure 3) for all three
cellulose esters. Two- and 3-bond correlations (CTA, Figure 4) from HETCOR (73),
COLOC (13,14), and INAPT (75) experiments extended these assignments from the
ring protons across the ester oxygen to the acetyl carbonyl carbons (in CDCI3: C6:
170.1 ppm, C3: 169.6 ppm, C2: 169.1 ppm) (75) and methyl protons (in 1,1,2,2-
tetrachloroethane-d2: H6: 2.07 ppm, H2: 1.97 ppm, H3: 1.94 ppm) (14). Acetyl
methyl carbon assignments are taken from the HETCOR spectrum. Ring proton 3-
bond coupling constants are presented in Tables I and III from the homonuclear 2D J-
Resolved experiment (16) and ID proton N M R (77) using resolution enhancement
post-processing (8). Kowsaka and coworkers (18) employed similar 2D correlation
techniques to accomplish these assignments in DMSO-d6 and CDCI3 for C T A . Shuto
and coworkers (19) accomplished these assignments by a similar protocol for CTP in
CDCI3 at 25°C. Tezuka and Tsuchiya (20) determined carbonyl carbon assignments
(C6: 173.6 ppm, C3: 173.1 ppm, C2: 172.7 ppm) for CTP in CDCI3 by INAPT.
For a series of peracetylated oligomers from dimer (degree of polymerization, DP
= 2) to nonamer (DP = 9), the ring protons and carbons as well as carbonyl carbons
were assigned (27). The ring carbons of oligomers DP = 2 through DP = 5 were also
assigned by Capon and coworkers (22). Chemical shift assignments were
accomplished by COSY, 2D homonuclear Hartman-Hahn correlated (HOHAHA)
spectroscopy (23-25), HETCOR, and INAPT. Subspectra of the reducing terminus,

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 136
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010
August 31, 2012 | http://pubs.acs.org

' ι I ι ι ι I J I I I I J ι ι 1 ι ι—
5.D Η.5 Η.Ο 3.5
Figure 2. COSY spectrum of CTA in CDCI3 at 25°C. (Adapted from ref. 8.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

τ 1 1—ι 1 ι ι 1 1 1 ι ι ι ι ι ι ι 1 Γ

5.0 N.5 N.D 3.5

Figure 3. C-H Heteronuclear Correlated 2D N M R spectrum of CTA in CDCI3 at

25°C. (Reproduced with permission from ref. 8. Copyright 1987 American
August 31, 2012 | http://pubs.acs.org

Chemical Society.)

C6Ac C3Ac

—1—•—•—•—•—1—'
17β.Θ 169.S
ppn

Figure 4. 2-Bond (A) and 3-Bond (B) Heteronuclear Correlated 2D N M R

Spectrum of CTA. (Reproduced with permission from ref. 13. Copyright 1990
John Wiley and Sons, Inc.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 138

Table III. 3-Bond Coupling Constants (in Hz) for C T A in CDCI3

50°C 50°C 25°C

(77) (16) (8)
H1.H2 7,6 7.6 7.9
H2.H3 9.0 9.7 8.6
H3.H4 9.0 9.0
H4,H5 9.5 9.5 9.2
H6S.H5 2.0
H6R,H5 5.5
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

nonreducing terminus and internal AGU's were extracted from the H O H A H A spectra
of the series (Figure 5). The A G U adjacent to the nonreducing terminus was also
demonstrated to be unique for DP = 3 through 6 oligomers. Therefore, the termini of
these lower oligomers must be viewed as being composed of three monomer units ~
the reducing terminus, the nonreducing terminus, and the monomer unit adjacent to the
nonreducing terminus.
A high level of chain rigidity is demonstrated by comparing T i values for these
oligomers and C T A (27). At a DP of about 7, the critical DP is reached since T i
values from the internal monomer ring (Figure 6) and carbonyl (Figure 7) carbons in
the oligomers are similar to that of CTA.
For cellulose esters with DS < 3, interpretation of individual resonances for each of
the 8 AGU's (Figure 1) displaying 12 magnetically different acetyl groups becomes
much more complicated. Buchanan and coworkers (26) synthesized a series of
August 31, 2012 | http://pubs.acs.org

carbonyl carbon ^C-enriched cellulose acetates with DS = 2.54, 2.03,1.99,1.61,1.06

and 0.42. A total of 16 carbonyl carbon resonances (Table IV) were observed in the
series of cellulose acetates including CTA. A l l of these resonances were correlated
with acetyl groups in either the 2-, 3- or 6-position using INAPT. Specific monomers
in the heteropolymer backbone were identified by 2D COSY (Figure 8). Cellulose
acetates with DS < 2.85 are not soluble in chloroform. DMSO-d6 exhibits a wide
solubility range for cellulose acetates with DS between 3.0 and 0.4, making it the
solvent of choice for these studies.
To assign the ring protons of the individual monomers, several observations from
the proton N M R spectra of cellulose esters were used (8): 1) H3 protons attached to
carbons bearing an acetyl group are the most deshielded, 2) ring protons attached to
carbons that do not bear an acetyl group resonate between 3.8 and 2.9 ppm, and 3) H2,
H3 and H6 attached to carbons bearing acetyl groups should resonate between 5.1 and
3.9 ppm. On the basis of proton N M R chemical shift assignments and these
observations, coupling network A (Figure 8) was assigned to CTA, the triacetylated
monomer. Assuming preferential hydrolysis of the C6 acetyl and that loss of this
acetyl causes only a minor perturbation in the rest of the monomer structure, network
Β is assigned to the 2,3-diacetyl monomer. Since H3 in network C connects to an H2
above 3.9 ppm, network C was assigned to overlapping 3,6-diacetyl and 3-monoacetyl
monomers. Buchanan and coworkers (26) suggest that network D results from either

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 139

(a) ο

eft Ε
α
00 α
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

in

0 ©
5.2 4 8 4.4 4. 0

(b)

Nonreducing End

!lljLlL
August 31, 2012 | http://pubs.acs.org

Internal

Reducing End

56
ULuJ!
4.8 4.0 32

Figure 5. (a) 2D HOHAHA spectrum of cellotriose hendecaacetate. (b)

Subspectrum for each monomer is obtained by taking slices through the Fi
dimension. (Reproduced with permission from ref. 21. Copyright 1990
American Chemical Society.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 140

T1 (sec)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

Degree of Polymerization (DP)

Figure 6. Plot of ring carbon T] values versus degree of polymerization for C T A

oligomers with DP = 2 through 9. (Reproduced with permission from ref. 21.
Copyright 1990 American Chemical Society.)

5.C
August 31, 2012 | http://pubs.acs.org

4.5

4.0

3.5

T1 (sec) 3
-°

2.5"

2. Ο­

Ι . 5-

1.0-
1
0 2 4 6 8 10 CTA

Degree of Polymerization (DP)

Figure 7. Plot of internal carbonyl carbon T] values versus degree of

polymerization for CTA oligomers with DP = 2 through 9. (Reproduced with
permission from ref. 21. Copyright 1990 American Chemical Society.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 141

ΓΜ
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

Ο
CL
CL

œ
SX

ι Γ
August 31, 2012 | http://pubs.acs.org

4 8 4. 0 3.2
ΡΡ (ΤΙ

Α Β C D
Η1 4.66 Η1 4.71 Η1 4.33 Η1 3.58
Η2 4.56 Η2 4.71 Η2 3.10 Η2 4.49
Η3 5.05 Η3 5.05 Η3 4.82 Η3 4.97
Η4 3.71 Η4 3.71 Η4 3.50 Η4 3.72
Η5 3.77 Η5 3.57 Η5 3.67 Η5 3.43
H6R 4.05 H6R H6R .... H6R
H6S 4.28 H6S — H6S .... H6S — -

Figure 8. COSY spectrum of cellulose acetate with DS = 2.54. Four coupling

networks, found in this spectrum, are identified by the solid and dashed lines and
by their chemical shifts. (Reproduced with permission from ref. 26. Copyright
1990 American Chemical Society.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 142

intra-chain or inter-chain hydrogen bonding between triacetyl monomer carbonyl

carbons and a hydroxyl group either on an adjacent A G U or on a neighboring chain.
By similar analyses, other coupling networks were assigned. From these coupling
network assignments and the INAPT experiment, 15 of the 16 observed carbonyls
were assigned to monomers (Table IV). The remaining carbonyl carbon resonance at
169.86 ppm could only be assigned in general terms to a C6 carbonyl carbon. This
assignment resulted from the observation that C3 carbonyl carbons resonate between
169.56 and 169.05 ppm. A l l C6 carbonyl carbon resonances are deshielded relative to
C3 carbonyl carbon resonances. A l l C2 carbonyl carbon resonances are shielded
relative to C3 carbonyl carbon resonances. The limitation of this approach to a
detailed analysis of monomer content in heteropolymeric cellulose esters is spectral
resolution due to peak overlap and the absence of correlation for H5 and H6.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

Table IV. Carbon-13 N M R Chemical Shifts for the Carbonyl Resonances

from C T A and the Series of Cellulose Acetates with DS < 3 (26)

Chemical Shift, ppm Monomer

170.01 (0.01,0.02)* 6-mono or diacetyl

169.92 (0.01,0.02,0.03) 6-triacetyl
169.86 C6 carbonyl carbon**
169.83 (0.01,0.02, 0.03) 6-mono or diacetyl
169.78 (0.01) 6-mono or diacetyl
August 31, 2012 | http://pubs.acs.org

169.56 3-monoacetyl c

169.48 (0.01,0.02, 0.04) 3-monoacetyl

169.44 2,3-diacetyl c

169.36(0.01,0.02, 0.03) 2,3-diacetyl

169.25 ^o-diacetyl 0

169.14 3-triacetyl c

169.13(0.06) J^-diacetyl
169.05 (0.02, 0.03,0.05) 3-triacetyl
168.83 (0.01,0.02) 2,3-diacetyl
168.72 (0.01) 2-triacetyl
168.63 (0.01) 2-monoacetyl, 2,6-diacetyl

a
The numbers in parentheses represent the downfield shift for that resonance with
increasing hydroxyl level (decreasing DS).
b
See text.
c
This resonance has been shifted by hydrogen bonding.

Source: Adapted from ref. 26.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 143

From Figure 1, it is clear that only 4 carbonyl carbon resonances should be

observed for C3 acetylated cellulose esters. In Table IV however, there are 8 carbonyl
carbons assigned to C3 acetylated monomers. Buchanan and coworkers (26)
suggested that the extra 4 monomers result from hydrogen bonding similar to that
described for network D. Also, the chemical shift of the observed carbonyl carbon
resonance is impacted by DS. As the hydroxyl content increases, i.e., DS goes down,
carbonyl carbon resonances assigned in spectra from higher DS samples shift
downfield by the amounts indicated in parentheses (Table IV). This downfield shift is
attributed to hydrogen bonding between AGU's or chains.
Kowsaka and coworkers (27) also attempted to assign the individual carbonyl
carbon resonances in a series of CA's with DS ranging from 2.92 to 0.43. The three
resonances for C T A (DS = 2.92) were easily assigned. For the three resonances
observed in the DS = 0.43 sample, it was assumed that the major composition would
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

contain a distribution of monoacetylated AGU's with substitution at the 6-, 3- and 2-

positions. Assignments for the monosubstituted AGU's in DS = 0.43 C A were based
on the same order of chemical shifts as observed for CTA. Assignments for the
remaining closely spaced carbonyl carbon resonances in the diacetylated AGU's were
postulated based on substituent additivity relationships only.
Hikichi and coworkers (28) analyzed proton COSY and relayed COSY (29,30) 2D
N M R spectra of cellulose diacetate with DS = 2.46 to assign 9 different spin systems.
These spin systems were assigned to individual acetylated AGU's. The acetylated
AGU's identified were four different 2,3,6-triacetylated AGU's, two 2,3-diacetylated
AGU's, a 2,6-diacetylated A G U , a 3,6-diacetylated A G U , and a 6-monoacetylated
A G U . The authors suggest that the four 2,3,6-triacetylated AGU's result from
chemical shift variation due to the nature of the acetylated AGU's on either side of the
2,3,6-triacetylated A G U . Since the 2,3,6-triacetylated A G U can be flanked by any of
August 31, 2012 | http://pubs.acs.org

7 different acetylated AGU's or 1 unacetylated A G U , there are 64 different species

possible. The four spectra identified by the authors result from the four A G U triads
centered on a 2,3,6-triacetylated A G U with the greatest population. Confirmation of
these assignments is provided by comparing the spectrum of the ring proton spectral
region to a spectrum resulting from the weighted combination of simulated spectra
from each of the 9 assigned acetylated AGU's.

Conformation and Molecular Dynamics. Buchanan and coworkers (57) used ^ C

T i measurements to probe molecular dynamics for CTA, CTP and CTB. The authors
also used absorption-mode phase-sensitive 2D nuclear Overhauser enhancement
(nOe) exchange spectroscopy (NOESY) (32,33) to probe the solution conformation of
these polymers. A plot of T i values as a function of temperature (Figure 9) for the
carbonyl carbons of C T A in DMSO-d6 and CDCI3 indicates a transition near 53°C. A
similar transition has been observed (34) for C T A in acetic acid on the basis of
intrinsic viscosity, specific volume, optical rotation, absolute viscosity, and sorption of
solvent measurements. The occurrence of this transition has been attributed to a
conformational change in the macromolecular structure of C T A by several groups
(17,31,34). Ogura and coworkers (35) observed a similar transition near 50-55°C in
variable temperature lineshape studies on CTA in CDCI3 which they attributed to a
chair-boat conformational change in the A G U . A transition within the temperature

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 144

range of 40-60°C was not observed in similar N M R studies on CTP and C T B by

Buchanan and coworkers (57).
The cause of this transition at 53°C is difficult to assign since it is difficult to
differentiate between microscopic conformational changes and supramolecular
structure changes based on these results. Microscopic conformational changes can
result from changes in either monomer conformation or the virtual angle about the
glycosidic bond between repeat units. Changes in interactions between polymer
chains result in supramolecular structural changes. Chair-boat conformational changes
and rotational changes about the glycosidic bond would result in changes in the
internuclear distances between ring protons and between protons on neighboring
AGU's. The N O E S Y experiment provides an excellent means of probing internuclear
distances between protons and solution conformational changes.
Buchanan and coworkers (57) used the NOESY experiment to examine the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

solution conformation of CTA, CTP and CTB in CDCI3. Several N O E S Y spectra

were collected at different mixing times between 40 and 200 msec at 25 and 60°C for
C T A concentrations of 0.1M and 0.35M in CDCI3. Similar data were collected for
CTP (0.09 M) and CTB (0.08 M) in CDCI3 at 25°C. A typical NOESY spectrum for
0.35 M C T A in CDCI3 is shown in Figure 10 with all resonances labeled. From peak
volume measurements at each mixing time, the initial portion of the nOe build-up
curve was used to measure individual nOe build-up rates. These individual rates were
then used to calculate internuclear distances by comparison of these rates to the build­
up rate for the nOe between the geminal protons on H6.
The average distance calculated for each N M R observable interaction is presented
in Table V . Internuclear distances from x-ray analysis (36,37) of C T A crystalline
forms I and II in fiber are shown in the second column of Table V . There are no
significant differences in the N M R data for the five systems. Keeping in mind the
August 31, 2012 | http://pubs.acs.org

small energy differences between chair conformations, the similarity of internuclear

distances, and the consistency of the proton coupling constants (Tables I and III), the
data are consistent with retention of the ^ C i conformation for the A G U . Substituent,
concentration, and solvent do not perturb this chair conformation. Vicinal coupling
constants around 9 Hz (Tables I and III) are consistent with the ^ C i conformation for
a β-D-glucopyranosyl residue. However, there may be some flattening of the A G U
ring based on the smaller size of the H1-H2 coupling constant (7.9 Hz) (8).
The virtual angle between HI and H4' (Figure 11) can be estimated from the
N O E S Y spectra. C T A in crystalline forms I and II can be represented as a 2/1 helix (a
1
syn planar conformation) where HI and H4 are aligned with a virtual angle of 0°.
Using models, the measured distance between HI and H4' in a 2/1 helix is 1.8 Â. By
rotating one A G U until the distance between HI and H4' equals the distance measured
in the N O E S Y experiments, the virtual angle between HI and H4' in solution for these
cellulose esters is estimated to be 30 - 34°. This virtual angle is consistent with a 5/4
helix suggesting that 5 glucose monomers are required to rotate through 360°. This
virtual angle is consistent with studies on helical conformation in cellulosic chains
from circular dichroism (38,39).
The conformation about the C5-C6 exocyclic bond has also been examined (8).
Based on the observed coupling constant between H6R and H5 as well as the smaller
coupling constant between H6S and H5 (Table III), the preferred rotamer population

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

1.8

—ι 1 1 1 1 1 1 Γ
35 40 45 50 55 60 65 70
Temperature, °C
Figure 9. CTA (0.35 M) carbonyl carbon T i values versus temperature in
CDCI3 (solid lines) and DMSO-d6 (dashed lines) for acetyl groups substituted at
C6 (squares), C3 (triangles) and C2 (diamonds). (Adapted from ref. 31.)
August 31, 2012 | http://pubs.acs.org

4.8 4.0
p p m

Figure 10. Stack plot of a NOES Y spectrum for C T A (0.35 M , 25°C) at a mixing
time of 100 ms. The F i and F2 axes have equal .dimensions. (Reproduced with
permission from ref. 31. Copyright 1989 American Chemical Society.)
In Cellulose Derivatives; Heinze, T., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 146

Table V . Average Internuclear Distances (in Angstroms) Obtained by the N O E S Y

Experiment for CTA, CTP and CTB Compared to X-Ray Crystal Structure Distances

Interactions X-Ray CTB CTP CTA CTA CTA

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

25°C 25°C 25°C 60°C 25°C

0.08 M 0.09 M 0.1 M 0.1 M 0.35 M
H1-H4' a 2.15 2.11 2.17 2.16 2.08
H1-H3 2.76 2.22 2.25 2.39 2.28 2.24
H3-H5 2.53 2.58 2.73 2.76 2.85 2.57
H2-H4 3.13 2.63 2.79 2.88 3.05 2.63
H1-H5 2.15 2.11 2.12 2.26 2.17 2.09
H3-H4 3.16 3.13 3.35 3.46 b 3.01
H4-H5 2.95 2.22 2.80 2.68 b 2.66
H1-H2 3.00 3.34 3.70 3.83 b 3.48
H3-H2 2.99 3.07 3.51 3.47 b 3.27
H2-H5 3.89 b 3.89 b b 3.62
H6R-H5 2.45 c
2.69 3.07 3.14 b 3.04
August 31, 2012 | http://pubs.acs.org

H6S-H5 2.29 c
2.69 3.08 3.15 b 3.17
H6R-H4 2.79C 3.08 3.50 3.84 b 3.45
H6S-H4 3.58 c
3.08 3.52 3.87 b 3.40
H6R-H6S 1.70

a
See the text.
b Cross-peaks for these interactions were not observed.
c
Assignments of R and S for the X-ray data may be reversed relative to the N M R data.

Source: Adapted from ref. 31.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 147

about the C5-C6 bond is the RCS rotamer. The ROS rotamer contributes minimally
(Figure 12) in agreement with Perlin and coworkers (77).
Tezuka (40) examined dipolar interactions between the acetyl methyl protons of
C T A in DMSO-d6 and CDCI3. Using a mixing time of 500 msec in the NOES Y
experiment, Tezuka observed different interactions in the two solvents. In CDCI3,
methyl protons on the acetyl group attached to C3 (Ac3, labeled H3 in Figure 13)
exhibited a through-space interaction with Ac6 (Figure 13). In DMSO-d6, both Ac2
and Ac3 exhibited a through-space interaction with Ac6 (Figure 14). Tezuka
concluded that the solution conformations for C T A in these two solvents are different.
A 2/1 helical conformation suggests interactions between Ac6 and both Ac2 and Ac3
could be observed which would be consistent with the results in DMSO-d6- The 5/4
helix suggested earlier (57) in CDCI3 would result in increased through-space
separation between Ac2 and Ac6 while the distance between Ac3 and Ac6 would
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

become closer. This analysis is consistent with the conformation observed in CDCI3
by the N O E S Y experiment (57).
Morat and Taravel (41,42) provide another view of the solution conformation of
C T A (DP = 30) in CDCI3. The selective heteronuclear 2D J-Resolved experiment
(43) was used to determine dihedral angles across the glycosidic bond. The dihedral
angles across the fragments Cl'-0-C4-H4 and across the fragment H l ' - C r - 0 - C 4
follow a Karplus relationship. By measuring the long-range heteronuclear coupling
constants across these dihedral angles, the torsional angles are calculated to be 0°,
leading to a 2/1 helix in CDCI3 solution. This result is inconsistent with results
presented above unless there is a DP-dependence to solution conformation about the
glycosidic bond.

Structure-Property Relationships. A careful monomer compositional analysis

August 31, 2012 | http://pubs.acs.org

enables an analysis of composition versus the properties of water solubility and water
absorbency (44) for a series of cellulose monoacetates (CMA's). This structure-
property relationship study is the first case for which a detailed analysis has been
accomplished. The relative degree of substitution (RDS) at C6, C3 and C2 (45,46)
and 3-monoacetyl content (45) have been suggested as important parameters for water
solubility. It has also been suggested (46) that there is no evidence for a correlation
between RDS and water solubility or absorbency.
Several clear differences in water solubility and water absorbency properties can
be correlated to monomer composition and molecular weight for CMA's in the DS
range of 0.5-0.9 (44). In general, for water absorbent CMA's, the monomer ratio of 3-
monoacetyl-to-2,3-diacetyl is less than 0.5, the monomer ratio of 3-monoacetyl-to-3-
triacetyl is less than 1.1, the monomer ratio of 3-monoacetyl-to-total 3-acetyl content
is less than 0.25, and the ratio of the DS for 3-monoacetyl monomer to the total DS is
generally less than 0.11. For water-soluble CMA's, each of the monomer ratios above
is greater than the number shown. In general, a C M A with a low molecular weight
(IV < 1.4; Μ ψ < 2.0 X 10^) that contains more than 10-11 3-monoacetyl monomers
per 100 repeat units will be water soluble. The importance of 3-monoacetyl monomer
content and molecular weight in determining water solubility and water absorbency for
CMA's is clearly illustrated.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 148
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

Figure 11. Cellobiose subunit where the virtual angle, Θ, between HI and H4'
1
represents the deviation of H4 from a syn planar relationship with H I .
(Reproduced with permission from ref. 31. Copyright 1989 American Chemical
Society.)
August 31, 2012 | http://pubs.acs.org

H 5 H 5 H 5

RHS RCS ROS

Figure 12. Schematic representation of possible rotamers about the exocyclic
C5-C6 bond. (Adapted from ref. 8.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010
August 31, 2012 | http://pubs.acs.org

Figure 13. NOES Y spectrum of the acetyl methyl region of CTA obtained with
mixing time of 500 ms in CDCI3 at 40°C. (Reproduced with permission from
ref. 40. Copyright 1994 John Wiley and Sons, Inc.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010
August 31, 2012 | http://pubs.acs.org

Figure 14. NOES Y spectrum of the acetyl methyl region of CTA obtained with a
mixing time of 500 ms in DMSO-d6 at 40°C. (Reproduced with permission from
ref. 40. Copyright 1994 John Wiley and Sons, Inc.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 151

Hydroxyl content is also important for these properties. For the water absorbent
CMA's, the 3-acetyl of the 2,3-diacetyl monomer was assigned to a non-hydrogen
bonding monomer. For three of five water-soluble CMA's, the same resonance was
assigned to a hydrogen bonding monomer. These data suggest that a minimum level
of 3-acetyl substitution must be present to disrupt intra- and/or intermolecular
hydrogen bonding. The delicate balance of C2 hydroxyl content must be maintained
to ensure hydrophilicity of the polymer backbone. The C6 hydroxyl is too remote
from the polymer backbone to have a significant impact on hydrophilicity.

Cellulose Ethers. As with cellulose esters, knowledge of the specific location of

etherification on the A G U is important for understanding properties of the ether
derivatives. Unreacted hydroxyl groups in cellulose ethers have been acetylated so
that the acetyl distribution can be examined as a means of determining the ether
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

distribution indirectly. Tezuka and coworkers have used this approach extensively for
studying ether substituent distribution in 0-(methyl)cellulose (47), 0-(2-hydroxy-
ethyl)cellulose (48), 0-(2-hydroxypropyl)cellulose (49), 0-methyl-0-(2-hydroxy-
ethyl)cellulose (50), 0-methyl-0-(2-hydroxypropyl)cellulose (50,51), 0-(carboxy-
methyl)cellulose (52), and 0-(2-hydroxypropyl)-0-methylcellulose acetate succinate
(53). Guo and Gray (54) also examined acetyl-derivatized O-ethylcellulose and O-
methylcellulose. Generally esterification provides a sensitive indirect probe of ether
substitution patterns by examination of the chemical shifts of the ester carbonyl carbon
resonances. Assignments are made for carbonyl carbon resonances at 6-, 3- and 2-
positions. No attempt is made in these studies to determine detailed esterifi-
cation/etherification patterns for individual monomers from the carbonyl carbon
resonance patterns.
For methylcellulose (47), high ether DS (above 2.0) polymers are difficult to
August 31, 2012 | http://pubs.acs.org

analyze in DMSO by the acetyl carbonyl carbon resonance approach due to poor
solubility. However acetylated methylcellulose is soluble in CDCI3 over the entire
range of DS values. The methyl ether carbon resonance splits into multiple resonances
with sensitivity to the nature of the substituent at each location as well as neighboring
substituents. Assignment of these methyl ether carbon resonances was accomplished
by comparison of the N M R results with results from G L C analysis of the polymer
hydrolyzed back to the individual monomer units. For example, in Figure 15, C2(3)
labels the resonance of the methoxy carbon at the C2 position with a methoxy group at
C3 and an acetoxy group at C6.
For acetylated hydroxyalkyl-celluloses, four carbonyl carbon resonances were
observed corresponding to acetyl carbonyl carbons in the 6-, 3- and 2-positions as well
as the carbonyl carbon from the acetyl group attached to the hydroxyalkyl group. For
hydroxyethylcellulose, these four resonances overlap in DMSO-d6 while they are well
separated in CDCI3 (48). The four resonances also overlap for 0-methyl-0-(2-
hydroxyethyl)cellulose in DMSO-d6 while they are a broad, unresolved resonance in
CDCI3 (50).
In hydroxypropylcellulose (49), the site-specific degree of substitution at C2 can
be determined. The CI ring carbon chemical shift is sensitive to the nature of the
substituent at the 2-position. For 2-O-hydroxypropyl substitution, the CI chemical
shift is 104.8 ppm while, for 2-O-acetyl substitution, this chemical shift is 102.8 ppm.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010
August 31, 2012 | http://pubs.acs.org

Figure 15. Proposed methoxy methyl resonance assignments of acetylated

methylcellulose in CDCI3 at 30°C. (Reproduced with permission from ref. 47.
Copyright 1987 American Chemical Society.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 153

Mixed Esters of Cellulose. The two cellulose mixed esters that have received
detailed N M R analysis by modern N M R techniques are cellulose acetate propionate
(CAP) and cellulose acetate butyrate (CAB). Knowledge of the relative and total acyl
concentration for each acyl group and the site-specific distribution of the various acyl
groups might allow optimization of the final properties.
Iwata and coworkers (55) assigned all proton and ^ C resonances for the
stereoregular CAP's 6-0-acetyl-2,3-di-0-propanoylcellulose (CADP) and 2,3-di-O-
acetyl-6-O-propanoylcellulose (CPDA). By using homonuclear COSY and HETCOR
for short-range correlations, all of the ring protons and carbons as well as acyl carbons
were identified in the two CAP's. The J-coupling networks could not provide any
information about specific sites of acylation. Correlations over 3-bonds between ring
protons and acyl carbonyl carbons were accomplished by H M B C (23-25,56,57). This
work was the first published application of the H M B C experiment for a cellulose
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

derivative. Figure 16 shows the result of the H M B C experiment for CADP. Even
though correlations to H5 and H6 were not clearly observed, assignments to the other
ring protons could easily be made by the H M B C experiment to provide site-specific
assignments. The H M B C experiment provides a rapid and reliable method for
unambiguous site-specific assignments for cellulose esters.
In order to determine acetyl distribution in CA's with DS < 3, Tezuka and
Tsuchiya (20) derivatized unreacted hydroxyl groups in the CA's making mixed
acetate propionate esters. Using the derivatized mixed ester, they observed the acetyl
and propionyl ester carbonyl carbon resonances to analyze the acyl distribution. Proof
that acetyl ester exchange during the derivatization step did not occur was provided by
monitoring the total acetyl content before and after the derivatization and by failure to
exchange any acetyl groups in C T A with propionyl groups.
Tezuka (58) assigned the backbone and sidechain protons of cellulose acetate
August 31, 2012 | http://pubs.acs.org

butyrate (CAB) by COSY. The acetyl carbonyl carbon assignments agreed with
previous assignments (14,15). Butyryl carbonyl carbon assignments (C6: 172.6 ppm,
C3: 172.1 ppm, C2: 171.6 ppm) were accomplished by the INAPT experiment.

Solid-State NMR

Solid-state ^ C N M R spectra of several cellulose esters, including CA's at various DS

levels, in different crystalline forms and with different crystalline-to-amorphous ratios,
CTP in different crystalline forms and CTB, are reported. N M R chemical shifts for
these different cellulose esters are collected in Table VI. In these studies, solid-state
N M R was used to address questions relative to hydrogen-bonding, spectral
differences due to DS, determination of the composition of crystalline and amorphous
components, ester identification, and conformation.
As discussed earlier (5), acylation of the C2 and C3 positions on cellulose results
in an upfield shift for adjacent CI and C4 (γ effect) ring carbons, respectively, while
acylation at the C6 position causes a downfield shift for C6 (β effect) in solution-state
13c N M R spectra. Interestingly, all three carbons ( C l , C4, and C6) are shifted upfield
on acylation in solid-state l ^ C N M R spectra (compare Tables II and VI).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

ο
χ

ο
χ

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
Ti J — ι — ι — ι — ι — r — I
165 no 60 30 20 10 0
,3
C Chemical s h i f t s (δ. p.p.m.)

Figure 16. ^H-detected i H - ^ C heteronuclear multiple-bond correlation

(HMBC) spectrum of cellulose acetate dipropionate (CADP). (Reproduced with
permission from ref. 55. Copyright 1992 Elsevier Science Publishers)
 155

Cellulose acetates with 0.48 to 2.92 (CTA) DS have received the greatest amount
of attention compared to cellulose propionates and butyrates. VanderHart and
coworkers (59) observed that ^ C N M R spectra of C T A oligomers with DP between 5
and 9 have the spectral characteristics of high DP C T A in crystalline form I, C T A (I),
by noting the consistency of the dominant backbone resonances from these oligomers
(Figure 17). ^ C N M R spectra of C T A crystalline forms I and II are different (Figure
18) (59-61). The N M R spectrum of CTA (I) is simpler than the corresponding
spectrum of C T A (II) in agreement with crystal structure effects observed in the ^ C
N M R spectra of the two cellulose crystalline allomorphs I and II. In the C T A (I) unit
cell, all C T A residues are magnetically equivalent and symmetry, by implication, is
high. In C T A (II), there are two magnetically inequivalent sites. In the methyl carbon
spectral region, the presence of a unique chemical shift for one of the six methyl
carbons is indicative of this magnetic inequivalence as well as the much larger number
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

of resonances observed in the spectrum of C T A (II) relative to the spectrum of C T A

(I) .
Takai and coworkers (61) prepared CTA (I) from cellulose (I), designated C T A
(II) , and from cellulose (II), designated CTA (In), for comparison of these polymers to
C T A (II). From these samples, the authors were able to differentiate crystal structures
for the three C T A samples, C T A (Ii), C T A (In) and C T A (II).
Subtraction of the spectrum of a related amorphous polymer provided the
subspectrum for the crystalline component of the nonamer. By this approach, the
nonamer was estimated to contain 47% crystallinity (Figure 18). Subtraction of the
13c N M R spectrum of amorphous C T A from spectra of C T A prepared by different
synthetic routes was used to estimate the C T A (I) and C T A (II) crystalline content
resulting from these preparations (59).
1 3 1 3
While C N M R spectra of C T A (I) and C T A (II) are quite different, C N M R
August 31, 2012 | http://pubs.acs.org

spectra of CTP prepared from either cellulose (I) or cellulose (II), both under low
swelling heterogeneous acylation conditions, are similar (60).
Doyle and coworkers (62) also observed amorphous and crystalline CTA. Proton
spin-lattice relaxation time, T i , measurements in the laboratory frame ranged from 0.3
to 1.5 s with a single exponential decay. Proton spin-lattice relaxation time, Tip,
measurements in the rotating frame exhibited a biexponential decay. The short Tip
values ranged from 3-5 ms and were assigned to amorphous C T A while the long Tip
values ranged from 15-25 ms and were assigned to crystalline CTA.
At DS = 2.5, the CI resonance of C T A splits into three resonances. These three
resonances have been assigned to resonances from the two crystalline forms Ια and
Ιβ (62), as observed in cellulose (63).
Buchanan and coworkers (26) discussed the importance of hydrogen bonding
1 3
between A G U ' s based on solution-state C N M R results. Differences in C2 and C3
carbonyl carbon resonance intensities in the solution- and solid-state NMR
spectra of C T A have been suggested to result from hydrogen bonding (64). Similar
effects due to hydrogen bonding were not observed for CTP.
Pines and coworkers (65) suggested that l ^ C shielding tensors should provide a
3
sensitive tool for studying hydrogen-bonding schemes in the solid-state. The ^ C
shielding anisotropy in the carbonyl group is strongly dependent on hydrogen bonding
to the oxygen atom. For C T A and CTP, similar chemical shift anisotropics are

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 156

Table VI. Solid-State Carbon-13 N M R Chemical Shifts for Cellulose Esters

Sample C=0 Cl C4 C2-C5 C6 C-α,β,γ Ref

Cellulose 104.8 88.5 74.4- 64.9 66

71.6
Cellulose 105.1 88.8 74.3 64.6 62
104.2 (+low 73.7 (+low
103.6 frequency 71.8 frequency
shoulder 70.8 shoulder
at about at about
84) 62)
C T A (I) 169.4 102.2 71.5 61.6 a: 20.9 59
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

171.9 75.0 a: 22.0

79.2
CTA © 169.00 101.37 71.10 61.01 a: 20.44 61
Crystalline 170.00 74.36 a: 21.42
171.45 78.51
C T A α> 168.87 101.56 71.01 60.90 a: 20.41 61
Crystalline 169.97 74.39 a: 21.56
171.41 78.52
CTA 170.35 100.69 72.85 63.48 a: 20.51 62
Amorphous 63.36
DS = 2.90 61.40
60.44
C T A (I) 169.83 101.34 71.11 61.12 a: 20.47 60
August 31, 2012 | http://pubs.acs.org

DS = 2.92 171.42 74.47

78.60
C T A (II) 170.0 100.0 71.1 63.5 a: 20.8 59
172.2 72.1 66.4 a: 23.4
173.2 73.3
74.5
74.9
77.9
78.7
CTA ai) 169.87 99.31 71.70 60.64 a: 20.50 61
170.62 72.99 63.05 a: 22.86
171.08 73.84 65.23
181.54 76.73 66.90
172.89 78.08
C T A 01) 169.79 99.06 70.58 63.17 a: 20.45 60
DS = 2.90 171.43 71.41 65.15 a: 22.70
172.81 72.91
73.66
76.46
78.02

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 157

Table V I . Continued

Sample C=0 CI C4 C2-C5 ' C6 C-α,β,γ Ref

CTA 169.8 99.8 72.4 62.8 a: 19.9 62
63.8
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

DS = 2.93 170.3 a: 20.4

CA 169.7 100.0 87.8 71.1 64.5 a: 19.9 62
DS = 1.5 100.6 (+low 72.2 (+low a: 20.3
103.7 frequency 73.6 frequency
shoulder) shoulder)
CA 169.8 100.3 87.8 71.1 64.6 a: 20.1 62
DS = 0.97 170.5 103.5 88.4 72.1 (+low
104.4 (+low 73.0 frequency
105.2 frequency 73.9 shoulder)
shoulder) 74.6
CA 170.2 103.4 88.7 70.7 65.2 a: 20.3 62
DS = 0.48 104.1 (+low 71.8 (+low
105.0 frequency 73.5 frequency
August 31, 2012 | http://pubs.acs.org

shoulder) 74.2 shoulder)

CTP (Cell I) 172.68 98.61 72.03 63.57 a: 27.53 60
DS = 2.92 101.74 73.83 β: 9.13
CTP (Cell II) 171.98 98.47 72.06 62.27 a: 26.93 60
DS = 2.80 101.22 73.83 β: 8.83
104.32
106.53
CTB 170.92 101.11 72.26 61.42 a: 36.21 60
DS = 2.80 172.54 76.51 62.95 β: 18.42
γ: 14.07

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 158

C2-C5
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

π—ι—Γ-
180.0 80.0
100.0
ppm
1 3
Figure 17. C CP/MAS spectra (25.2 MHz) of the indicated oligomers of CTA.
(Reproduced with permission from ref. 59. Copyright 1996 American Chemical
Society.)
August 31, 2012 | http://pubs.acs.org

ι J ι—ι—ι—ι—ι π—ι—ι—ι—ι—ι—ι—ι—ι—ι—ι—ι—ι—ι -
π—ι—ι I I

180.0 100.0 80.0 60.0 20.0

ppm
1 3
Figure 18. Solid-state C CP/MAS N M R spectra of the nonamer of CTA, the
crystalline component of the nonamer, and the two C T A allomorphs.
(Reproduced with permission from ref. 59. Copyright 1996 American Chemical
Society.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 159

observed for the aliphatic and ring carbons while a much larger chemical shift
anisotropy is observed for the carbonyl carbons based on examination of spinning
sidebands for ordered, layered polymer film. Based on the measurement of chemical
shift tensors a n , σ22 and <*33 of the carbonyl resonances in several C T A and CTP
films with different physical properties, a correlation between shielding anisotropy and
film transparency was observed (64). The largest shielding anisotropy and asymmetry
factor for the shielding tensors was found for the film with the least transparency.
These observations of tensor anisotropy and reduced transparency are consistent with a
more ordered polymer. This study demonstrated the usefulness of probing polymeric
ordered domains with carbonyl shielding tensor measurements.

Future Directions
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

Over the past 10-15 years, considerable progress has been made in understanding the
importance of detailed knowledge relating properties, structures, molecular
conformations and dynamics for ester- and ether-derivatized cellulosic polymers. The
application of new N M R techniques for elucidating detailed polymeric structures has
made the analysis of structure-property relationships possible. In this chapter we have
brought together relevant solution- and solid-state N M R literature from the past 10-15
years on the structure of acylated cellulosic polymers, including acetate, propionate,
and butyrate, as well as acylated cellulose ethers. These studies provide a rich
knowledge-base on site-specific substitution as well as specific monomer composition
for individual esterified or etherified anhydroglucose units. Also, they begin to make
possible an understanding of the impact of structural modifications as they relate to
properties, possibly enabling fine-tuning of these structural modifications to approach
desired properties.
August 31, 2012 | http://pubs.acs.org

As other powerful N M R techniques are applied to questions related to structure in

cellulosic polymers, more structural information will emerge to aid in our
understanding of important property relationships. Determination of the site-specific
degree of substitution will undoubtedly benefit from the application of higher field
N M R spectrometers as well as application of 3D N M R techniques.
The development of methods for sequencing esterified and etherified
anhydroglucose units is an important problem that needs to be solved. Understanding
the heterogeneous nature of substitution along the A G U backbone of cellulose
derivatives can give useful insight into the solution conformation and molecular
dynamics of these derivatives. Some minimal information is available now based on
hydrogen bonding, but much more detail is needed to enable further analysis of the
structure of these complex heteropolymers.
In studies presented here, specific acetylated monomer assignments from the 8
possible monomers have provided a detailed look at the impact on structure of
acetylation at C2 and C3. Only general assignments could be made for C6 acetylated
monomers. A monomer-specific analysis of acetylation at C6 is needed and may
become possible by application of 3D N M R techniques, zero-quantum coherence 2D
N M R techniques, or higher field N M R spectrometers.
Analyses of the solution conformation of cellulose acetates have provided
information about the helical nature of the polymer. Some of these studies suggest

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 160

differing views on the solution conformation about the glycosidic bond. The solution
conformation about this bond possibly changes with increasing degree of
polymerization. A study of solution-state conformation as a function of degree of
polymerization in CDCI3 as well as DMSO-d6 with emphasis on the virtual angle
across the glycosidic bond might resolve this question. In addition, the transition
between 50 - 55°C measured for several properties needs to be examined in sufficient
detail to determine its source.
Monomer-specific assignments for the cellulose acetate carbonyl carbon
resonances have provided detailed knowledge about structure-property relationships
pertaining to water solubility and absorbency for this polymer. Similar knowledge
based on ^C-enrichment in ester-derivatized cellulose ethers should improve our
understanding of important structure-liquid-crystalline property relationships in
hydroxyalkylcellulose and methylcellulose derivatives that form thermotropic
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

cholesteric liquid-crystalline phases.

The future looks very bright for applying N M R to solve these tough structure-
property relationship problems in cellulose ester and ether chemistry. The results of
such studies should aid in the design of cellulose esters and ethers with predictable
properties.

Acknowledgments

The author thanks John A. Hyatt, Eastman Chemical Company, for helpful discussions
and Yasuyuki Tezuka, Tokyo Institute of Technology, for his assistance.

Literature Cited
August 31, 2012 | http://pubs.acs.org

1. Nehls, I.; Wagenknecht, W.; Philipp, B.; Stscherbina, D. Prog. Polym. Sci.,
1994, 19, 29.
2. Kamide, K.; Saito, M. Macromol. Symp., 1994, 83, 233.
3. Goodlett, V. W.; Dougherty, J. T.; Patton, H. W. J. Polym. Sci., Pt. A-1, 1971, 9,
155.
4. Shiraishi, N.; Katayama, T.; Yokota, T. Cellulose Chem. Technol., 1978, 12,
429.
5. Miyamoto, T.; Sato, Y.; Shibata, T.; Inagaki, H.; Tanahashi, M. J. Polym. Sci.,
Polym. Chem. Ed., 1984, 22, 2363.
6. Kowsaka, K.; Okajima, K.; Kamide, K. Polym.J.,1986, 18, 843.
7. Kamide, K.; Okajima, K. Polym.J.,1981, 13, 127.
8. Buchanan, C. M.; Hyatt, J. Α.; Lowman, D. W. Macromolecules, 1987, 20, 2750.
9. Nagayama, K.; Kumar, Α.; Wuthrich, K.; Ernst, R. R. J. Magn. Reson., 1980, 40,
321.
10. Bax, Α.; Freeman, R. J. Magn. Reson., 1981, 44, 542.
11. Maudsley, Α. Α.; Muller, L.; Ernst, R. R. J. Magn. Reson., 1977, 28, 463.
12. Bax, Α.; Morris, G. A. J. Magn. Reson., 1981, 42, 501.
13. Uhrinova, S.; Petrakova, E.; Ruppeldt,J.;Uhrin, D. Magn. Reson. Chem., 1990,
28, 979.
14. Dais, P.; Perlin, A. S. Carbohyd. Res., 1988, 181, 233.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 161

15. Buchanan, C. M.; Hyatt, J. Α.; Lowman, D. W. Carbohyd. Res., 1988, 177, 228.
16. Gagnaire, D. Y.; Taravel, F. R.; Vignon, M. R. Macromolecules, 1982,15,126.
17. Rao, V. S.; Saurio, F.; Perlin, A. S.; Viet, M. T. P. Can. J. Chem., 1985, 63,
2507.
18. Kowsaka, K.; Okajima, K.; Kamide, K. Polym.J.,1988, 20, 1091.
19. Shuto, Y.; Murayama, M.; Azuma,J.;Okamura, K. Bull. Inst. Chem. Res., Kyoto
Univ., 1988, 66, 128.
20. Tezuka, Y.; Tsuchiya, Y. Carbohyd. Res., 1995, 273, 83.
21. Buchanan, C. M.; Hyatt, J. Α.; Kelley, S. S.; Little, J. L. Macromolecules, 1990,
23, 3747.
22. Capon, B.; Rycroft, D. S.; Thomson, J. W. Carbohyd Res., 1979, 70, 145.
23. Bax, Α.; Davis, D. G. J. Amer. Chem. Soc., 1985, 107, 2820.
24. Bax, Α.; Davis, D. G. J. Magn. Reson., 1985, 65, 355.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

25. Summers, M. F.; Marzilli, L.J.;Bax, A. J. Amer. Chem. Soc., 1986, 108, 4285.
26. Buchanan, C. M.; Edgar, K.J.;Hyatt, J. Α.; Wilson, A. K. Macromolecules,
1991, 24, 3050.
27. Kowsaka, K.; Okajima, K.; Kamide, K. Polym.J.,1988, 20, 827.
28. Hikichi, K.; Kakuta, Y.; Katoh, T. Polym.J.,1995, 27, 659.
29. Wagner, G. J. Magn. Reson., 1983, 55, 151.
30. Bax, Α.; Drobny, G. J. Magn. Reson., 1985, 61, 306.
31. Buchanan, C. M.; Hyatt, J. Α.; Lowman, D. W. J. Amer. Chem. Soc, 1989, 111,
7312.
32. States, D.J.;Haberkorn, R. Α.; Ruben, D. J. J. Magn. Reson., 1982, 48, 286.
33. Olejniczak, E. T.; Hoch, J. C.; Dobson, C. M.; Poulsen, F. M. J. Magn. Reson.,
1985,64,199.
34. Ryskina,I.I.;Vakilenko, N. A. Polym.Sci.USSR, 1987, 29, 340.
August 31, 2012 | http://pubs.acs.org

35. Ogura, K.; Sobue, H.; Kasuga, M. Polym. Letters Edit., 1973,11,421.
36. Stipanovic, A.J.;Sarko, A. Polymer, 1978, 19, 3.
37. Steinmeier, H.; Zugenmaier, P. Carbohyd. Res., 1987, 164, 97.
38. Ritcey, A. M.; Gray, D. G. Biopolymers, 1988, 27, 479.
39. Stipanovic, A.J.;Stevens, E. S. J. Appl. Polym. Sci.,Appl.Polym. Symp., 1983,
37, 277.
40. Tezuka, Y. Biopolymers, 1994, 34, 1477.
41. Morat, C.; Taravel, F. R. Bull. Magn. Reson., 1989,11,321.
42. Morat, C.; Taravel, F. R. Tetrahedron Letters, 1990, 31, 1413.
43. Bax, Α.; Freeman, R. J. Amer. Chem. Soc., 1982, 104, 1099.
44. Buchanan, C. M.; Edgar, K.J.;Wilson, A. K. Macromolecules, 1991, 24, 3060.
45. Miyamoto, T.; Sato, Y.; Shibata, T.; Tanahashi, M.; Inagaki, H. J. Polym. Sci.,
Polym. Chem. Ed., 1985, 23, 1373.
46. Kamide, K.; Okajima, K.; Kowsaka, K.; Matsui, T. Polym.J.,1987, 19, 1405.
47. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Macromolecules, 1987, 20, 2413.
48. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Polymer, 1989, 30, 2288.
49. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Carbohyd. Res., 1990, 196, 1.
50. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Makromol. Chem., 1990,191,681.
51. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Polym.J.,1991, 23, 189.
52. Tezuka, Y.; Tsuchiya, Y.; Shiomi, T. Carbohyd. Res., 1996, 291, 99.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 162

53. Tezuka, Y.; Imai, K.; Oshima, M.; Ito, Κ. Carbohyd. Res., 1991, 222, 255.
54. Guo, J.-X.; Gray, D. G. J. Polym. Sci.: Pt. A: Polym. Chem., 1994, 32, 889.
55. Iwata, T.; Azuma, J.-I.; Okamura, K.; Muramoto, M.; Chun, B. Carbohyd. Res.,
1992, 224, 277.
56. Frey, M. H.; Leupin, W.; Sorensen, O. W.; Denny, W. Α.; Ernst, R. R.;
Wuthrich, K. Biopolymers, 1985, 24, 2371.
57. Bax, Α.; Summers, M. F. J. Amer. Chem. Soc., 1986, 108, 2093.
58. Tezuka, Y. Carbohyd. Res., 1993, 241, 285.
59. VanderHart, D. L.; Hyatt, J. Α.; Atalla, R. H.; Tirumalai, V. C. Macromolecules,
1996, 29, 730.
60. Hoshino, M.; Takai, M.; Fukuda, K.; Imura, K.; Hayashi, J. J. Polym. Sci: Pt. A:
Polym. Chem., 1989, 27, 2083.
61. Takai, M.; Fukuda, K.; Hayashi, J. J. Polym. Sci.: Pt. C: Polym. Lett., 1987, 25,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch010

121.
62. Doyle, S.; Pethrick, R. Α.; Harris, R. K.; Lane, J. M.; Packer, K.J.;Heatley, F.
Polymer, 1986, 27, 19.
63. VanderHart, D. L.; Atalla, R. H. In The Structures of Cellulose--
Characterization of the SolidState;Atalla, R. H., Ed.; ACS Symposium Series
340; American Chemical Society: Washington, DC, 1987, 88-118.
64. Nunes, T.; Burrows, H. D.; Bastos, M.; Feio, G.; Gil, M. H. Polymer, 1995, 36,
479.
65. Pines, Α.; Chang, J.J.;Griffin, R. G. J. Chem. Phys., 1974, 61, 1021.
66. Pawlowski, W. P.; Sankar, S. S.; Gilbert, R. D. J. Polym. Sci.: Pt. A: Polym.
Chem., 1987, 25, 3355.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 11

13
C NMR Structural Study on Cellulose
Derivatives with Carbonyl Groups as a Sensitive Probe

Y. Tezuka

Department of Organic and Polymeric Materials, Faculty of Engineering, Tokyo

Institute of Technology, O-okayama, Meguro-ku, Tokyo 152, Japan
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

A new versatile analytical method for cellulose ester and ether

derivatives has been developed, to determine the degree of
substitution (DS) at the individual positions (2, 3 and 6) on the glucose
residue. Thus unsubstituted hydroxyl groups in the starting cellulose
derivatives are first converted to appropriate acyl groups. The
obtained peresterified derivatives become soluble in common NMR
solvents regardless of the DS of the samples, in contrast to the original
cellulose derivatives. The subsequent 13-C NMR analysis with the
acyl carbonyl carbon as a structural probe allows to determine the
August 31, 2012 | http://pubs.acs.org

detailed distribution pattern of the unsubstituted hydroxyl groups, and

consequently that of the ester or ether substituents in the starting
cellulose derivatives. In addition, supramolecular structures of
cellulose triacetate in different solutions have been studied by means
of 2D-NOESY technique by the detection of the through-space
interaction of acetyl protons.

Cellulosic materials have gained a renewed interest along with the recent
developments of both "high-tech" and "bio-tech" applications (1). Since most
cellulose derivatives are produced through the reaction of the hydroxyl groups
either at 2, 3 or 6 positions in an anhydroglucose residue, the precise control and
deterimination of the distribution pattern of the substituents are of primary
importance not only for the elucidation of structure-property relationships to
achieve optimal performance of the final product, but also for the quality control of
the product to address the product-liability requirements.

©1998 American Chemical Society 163

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 164

We have so far proposed a new analytical technique applicable for cellulose

derivatives, in particular such cellulose ethers as methylcellulose,
hydroxyalkylcelluloses and those having both methyl and hydroxyalkyl
substituents, in which their peracetylated derivatives are subjected to 13-C N M R
measurement (2-6). The following advantages are noted for this procedure (7,8).
1. The acetylation of cellulose ethers is an easy and simple pretreatment to
provide products readily soluble in common N M R solvents over a wide DS range
of the starting derivative, facilitating the spectral comparison with those of model
polymers.
2. The acetyl carbonyl carbon signal can act as a remarkably sensitive structural
probe, reflecting its location on the anhydroglucose residue.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

3. The acetylation of the hydroxyl groups can eliminate spectral complication

arisen from intra- and intermolecular hydrogen bonds.
4. The polymeric form in cellulose derivatives can be maintained during the
analytical procedure, to avoid the cumbersome hydrolysis pretreatment inevitable
for chromatographic techniques such as GLC and H P L C .
As an extension of the preceding studies, the present paper describes first on
the application of this novel technique to industrially important cellulose ester and
ether derivatives, namely cellulose acetate (CA) (9) and carboxymethylcellulose
(CMC) (10), respectively. Secondly, 2D-NOESY results on cellulose triacetate
are presented with acetyl protons as structural probe, to detect the through-space
interaction of acetyl groups to provide supramolecular structures of cellulose
August 31, 2012 | http://pubs.acs.org

triacetate in different solutions (11).

Results and Discussion

Substituent Distribution in Cellulose Acetate (CA). Cellulose acetate is

commercially produced with appropriate degrees of substitutions, and is provided
for wide applications such as fibers, plastics, films and coatings. Although the total
acetyl content in CA may be determined by a standard titration technique, the DS
at the individual positions in the anhydroglucose residue is not readily obtainable
by chromatographic techniques since the hydrolysis pretreatment of C A samples
is accompanied by the deacetylation. Thus a 1-H N M R technique has been
applied by using costly perdeuterioacetyl derivative of CA, since acetyl proton
signal in cellulose triacetate appears as a resolved triplet reflecting its location on
the glucose residue (12).
An alternative and versatile 13-C N M R technique has been developed by

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 165

using perpropanoated derivative of CA. Quantitative propagation of the hydroxyl

groups in C A samples without the concurrence of ester exchange reaction is a
prerequisite for the precise determination of the DS by use of the propanoated
derivative. Such a reaction condition has been realized with propanoic
anhydride/pyridine/4-(dimethylamino)pyridine system. Indeed, the possibility of
ester exchange reactions was excluded through observing constant total acetyl
content during the propanoation treatment, and from the fact that attempted
propanoation of cellulose triacetate under the present condition failed to introduce
propanoate groups.
The carbonyl region spectrum of a propanoated derivative of CA is shown in
Figure 1, together with that of the starting CA. In the latter, carbonyl carbon
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

signal appears as overlapped multiplet peaks, which reflect the substitution

patterns of acetyl groups and the hydrogen-bond interactions between acetyl and
hydroxyl groups. On the contrary, a remarkably simple spectrum is obtained for
the propanoated derivative, where both acetyl and propanoyl carbonyl carbon
signals appear separately and are resolved into the two sets of three peaks
corresponding to their substitution positions on the anhydroglucose residue. The
three peaks are assigned as that on the 2, 3 and 6 positions, respectively, from up
field.
Thus a quantitative-mode 13-C N M R measurement for a series of
propanoated CA derivatives allows one to determine the individual DS of the
starting C A samples. In addition, the propanoyl group distribution, which is
August 31, 2012 | http://pubs.acs.org

complementary to that of acetyl groups, facilitates the determination of the DS of

C A samples of low acetyl contents with high precision. Consequently, the present
N M R technique is considered as a significantly improved analytical means to
determine the distribution pattern of acetyl groups in C A samples of wide range of
DS. In the 1-H N M R technique with perdeuterioacetyl derivatives of CA, in
contrast, the absolute signal intensity decreases along with the decrease in the
acetyl content.
This novel technique has been successfully applied also for a commercially
important cellulose ester derivative having two different ester substituents,
namely cellulose acetate butyrate (CAB) (13). 13-C N M R carbonyl region spectra
listed in Figure 2 demonstrate that both acetyl and butanoyl carbonyl carbon
signals appear separately and are resolved into two-sets of three peaks as in the
case of propanoated C A samples. In consequence, the relative content of acetyl
and butanoyl groups for a series of C A B samples can be determined with a
quantitative-mode measurement. The absolute distribution of the two
subsutituents is subsequently determined by taking into account the total degree
of substitution by the two substituents.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 166
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011
August 31, 2012 | http://pubs.acs.org

Ί
I ι ι ι ι I ι ι—ι—τη—ι—ι—ι—τη—ι—ι—ι—ι—|—ι—ι—ι ι |—ι—ι—ι—ι—|—ι—ι—ι—τη—ι—ι—ι—τη
175 174 173 172 171 170 169 168 167

Figure 1. 100MHz 13-C N M R carbonyl region spectra of cellulose acetate

t0
having the total DS of 1.43 in DMSO-de ( P> and of its propanoated derivative
in CDCI3 (bottom). (Reproduced with the permission from ref. 9. Copyright
1995 Elsevier Science Ltd.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 167

Ά) i > )
R :C O C H 3 or C O C H C H C H
2 2 3
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011
August 31, 2012 | http://pubs.acs.org

1 11

175 ~r - 1 I165' —I
170 160
Figure 2. 67.8 M H z 13-C N M R carbonyl region spectra of cellulose acetate
butyrate having different substitution patterns in CDCI3. (Reproduced with
the permission from ref. 13. Copyright 1993 Elsevier Science Ltd.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 168

Substituent Distribution in Carboxymethylcellulose (CMC). Sodium

carboxymethylcellulose, commonly termed carboxymethylcellulose or C M C , is an
important water-soluble cellulose ether derivative applied in such human-contact
uses as food, pharmaceutics and cosmetics as well as in such ecologically-
sensitive uses as soil treatment, oil recovery and paper sizing processes.
Although the total carboxymethyl content in C M C samples may be
determined by a standard titration technique, the individual DS at the 2, 3 and 6
positions on the anhydroglucose residue is obtainable only after the hydrolysis
treatment. The hydrolytic pretreatment of C M C , however, requires precautions
to avoid the loss of the hydrolysate during both reaction and recovery processes.
The side reactions, in particular intra- and intermolecular lactonization between
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

carboxylic acid and hydroxyl groups under acidic condition, should also be
carefully eliminated. The presence of anomeric isomers in the hydrolysate may
also cause spectral complication in the chromatographic and spectroscopic
analysis.
An alternative 13-C N M R technique has been developed by using
peresterified derivative of CMC, thus without the hydrolysis pretreatment. Since
C M C possesses carboxylate salt substituent, it is soluble only in water. By this
characteristic property, the acylation reaction of the unsubstituted hydroxyl
groups is circumvented to give a product soluble in any of common N M R solvents.
Hence, a two-step derivatization of C M C has been employed, where sodium
carboxymethyl groups are first converted into methyl ester groups by the reaction
August 31, 2012 | http://pubs.acs.org

with dimethyl sulfate. Although this methylation treatment was occasionally

accompanied by undesirable side reactions to result in partly insoluble product,
subsequent propanoation treatment of the residual hydroxyl groups could produce
a series of peresterified C M C samples soluble in DMSO-de- The acetylation in
place of the propanoation of the hydroxyl groups proceeds as well, while the acetyl
carbonyl carbon signal appears in the coincided region with that of carboxymethyl
carbonyl signal.
13-C N M R carbonyl region spectra of a series of peresterified C M C samples
are listed in Figure 3. The propanoyl carbonyl signal appears as three resolved
peaks regardless of the DS of the sample, and corresponding to its substitution
positions on the anhydroglucose units. In addition, the carboxymethyl carbonyl
carbon signal, being resolved also as three peaks in case of high total DS, provides
the complementary information on the distribution of the substituents. Minor
unassignable signals occasionally visible at 169-170ppm may be arisen from side

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 169
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011
August 31, 2012 | http://pubs.acs.org

ι 1 1 1 1
—I
185 180 175 170 165

Figure 3. 100MHz 13-C N M R carbonyl region spectra of peresterified

carboxymethylcellulose having different DS values in DMSO-dg.
(Reproduced with the permission from ref. 10. Copyright 1996 Elsevier
Science Ltd.)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 170

reactions during the methylation process. Finally, a quantitative-mode 13-C N M R

measurement was performed to determine the individual DS of the starting C M C
samples having a wide range of degrees of substitution.

Interaction between Substituents in Cellulose Triacetate (CTA). The

control of the substitution pattern in cellulose derivatives may lead to the design
of supramolecular structures of cellulosics in bulk and in solution. The detection
of the interaction between the substituents is a key to elucidate the chain
conformation and the resulting supramolecular structures. In the preceding study
(2) , the through-space interaction between the substituents at the 3 and 6
positions in the anhydroglucose residue on acetylated methylcellulose has been
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

postulated from the chemical shift change in the methoxy methyl carbon signals.
The detection of the through-space interaction between acetyl groups in
cellulose triacetate dissolved in different solvens is now achieved by means of a
2D-NOESY technique with acetyl proton as N M R probe, since acetyl proton
signal in cellulose triacetate appears as a resolved triplet reflecting its location on
the glucose residue. Figure 4 shows NOESY spectra of CTA in CDCI3 and in
DMSO-de solutions, in which off-diagonal cross peaks are generated besides a
large diagonal signal. In CDCI3, the cross-peaks correlate the acetyl proton signal
at the 3 and 6 positions on the glucose residue. In DMSO-d6, on the other hand,
the off-diagonal cross-peaks appear at the 2 and 6 positions in addition to the 3 and
6 positions. These results demonstrates that the NOESY technique is a powerful
August 31, 2012 | http://pubs.acs.org

means to observe the interaction between substituents, to provide the solution

dynamics of three-dimensional structures of cellulose derivatives.

Experimental

Cellulose acetate (CA) samples of different DS values were prepared by acid

hydrolysis of cellulose triacetate (Aldrich or Daicel). Propanoation of a series of
CA samples was carried out by propanoic anhydride/pyridine/4-
(dimethylamino)pyridine at 100°C for l h . Cellulose acetate butyrate (CAB)
samples of different distribution patterns were supplied from Eastman Chemical
Japan, Ltd. Carboxymethylcellulose (CMC) samples of different DS values were
obtained from Dai-ichi Kogyo Seiyaku Co. Methyl esterification of sodium
carboxymethyl groups in a series of C M C samples was performed by dimethyl
sulfate in DMSO at 45°C for 24h. Subsequent propanoation of methyl-esterified
CMC samples was carried out by propanoic anhydride/pyridine/4-
(dimethylamino)pyridine in DMAc/LiCl at 100°C for 6h.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

3
H

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
Figure 4. NOESY spectra of acetyl methyl region of cellulose triacetate in
CDCI3 (left) and in DMSO-de (right). ( Reproduced with the permission from
ref. 11. Copyright 1994 John Wiley & Sons, Inc.)
 172

13-C N M R measurements were performed with either JEOL GX-270 or E X -

400 apparatus. Chemical shift values were referenced from the solvent signal of
either CDCI3 (77.0) or DMSO-d6 (43.5). Quantitative-mode 13-C N M R
measurements were conducted by a non-NOE gated decoupling technique with a
pulse repetition time of 30sec. 2D-NOESY measurements were carried out with a
mixing time of 500msec, and 16 transients were acquired for each t\ value.

Acknowledgments. The author thanks to Eastman Chemical Japan, Ltd., Shin-

Etsu Chemical Co., and Dai-ichi Kogyo Seiyaku Co., for the gift of cellulose
derivatives. The financial supports from The Agricultural, Chemical Research
Foundation, Nestle Science Promotion Committee, and a Grant from the Ministry
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch011

of Education, Science and Culture are gratefully achnowledged. The author is also
grateful to the collaboration of M . Oshima, K. Ito of Shin-Etsu Chemical Co., and
of Y. Tsuchiya and T. Shiomi of Nagaoka University of Technology throughout the
present study.

Literature Cited
1. Cellulosics Utilization, Research and Reward in Cellulosics; Inagaki, H.;
Phillips, G.O., Eds.; Elsevier, London, 1989.
2. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Macromolecules, 1987, 20, 2413.
3. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Carbohydr. Res., 1990, 196, 1.
4. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Polymer, 1989, 30, 2288.
August 31, 2012 | http://pubs.acs.org

5. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Makromol. Chem., 1990, 191,
681.
6. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. Polym.J,1991, 23, 189.
7. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. In Cellulose and Wood,
Chemistry and Technology, Schuerch, C. Ed.; John Wiley & Sons, New York,
1989, p1011.
8. Tezuka, Y.; Imai, K.; Oshima, M.; Chiba, T. In Cellulose: Structural and
Functional Aspects, Kennedy, J.F.; Phillips, G.O.; Williams, P.A. Eds.; Ellis
Horwood, Chichester, 1990, p251.
9. Tezuka, Y.; Tsuchiya, Y. Carbohydr. Res., 1995, 273, 83.
10. Tezuka, Y.; Tsuchiya, Y.; Shiomi, T. Carbohydr. Res., 1996, 291, 99.
11. Tezuka, Y. Biopolymers, 1994, 34, 1477.
12. Goodlett, V.W.; Dougherty, J.T.; Patton, H.W. J. Polym. Sci. Part-A, 1971, 9,
155.
13. Tezuka, Y. Carbohydr. Res., 1993, 241, 2853.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 12

Novel Approaches Using FTIR Spectroscopy To Study

the Structure of Crystalline and Noncrystalline
Cellulose

T. Kondo, Y. Kataoka, and Y. Hishikawa

Forestry and Forest Products Research Institute (FFPRI),

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

P.O. Box 16, Tsukuba Norin Kenkyu, Tsukuba, Ibaraki 305, Japan

This paper deals with introduction of two methods with

combination of FTIR which are very powerful to investigate
cellulose supermolecular structures. One is using a microscopic
accessary and the other is using a special reaction cell for
deuteration. The former method depends on interpretation of
change of the characteristic IR bands for cellulose Ια and Iβ
crystalline phases to determine crystalline structures during
coniferous wood cell wall formation, while the latter needs kinetic
approaches to characterize amorphous structures by monitoring
changes of OH bands during the deuteration process with D O as a2

probe for amorphous 3 film samples, cellulose, 2,3-di-O­

-methylcellulose (23MC) and 6-O-methylcellulose (6MC).
August 31, 2012 | http://pubs.acs.org

There are lots of powerful tools to investigate cellulosic structures such as X-ray,
electron diffraction and microscope, FTIR, Raman spectroscopy and solid-state
NMR. Among them we have been using FTIR to study cellulose structures as well as
hydrogen bonding formation (1-3). In particular not only FTIR but also FTIR in
combination with suitable attachments could provide us with further information on
cellulose supermolecular structure.
In this paper we will introduce the following two IR methods and their applications
(Figure 1) to study the supermolecular structures for cellulose and its derivatives.
The methods are i) FTIR with a microscopic accessary (4) and ii) FTIR with a special
reaction cell for deuteration. In the following, the former was applied to change of
crystalline form and crystallinity of wood cell wall cellulose during cell wall
formation, whereas the latter was employed for the study of amorphous regions using
deuteration process. The advantages of the two combination will be described here
briefly. For FTIR with a microscopic attachment, FTIR spectra for such a small
amount of sample less than 1 mg can be obtained from small area (minimum : 50 χ
50μπι2). Using the special reaction cell for deuteration, FTIR spectra can be obtained
throughout a deuteration process of cellulosics and thus this method enables us to
perform a kinetic analysis of the reaction.

©1998 American Chemical Society 173

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 174

Experimental
1. FTIR with a microscopic accessary.
Materials. To investigate each stage of formation for coniferous wood tracheid cell
wall, a sectinonal sample was prepared from radial and radial face (perpendicular to
radial) direction of Cryptomeria japonica D. Don (Japanese cedar) and
Chamaecyparis obtusa Endl (Japanese cypress) with 30 μπι thickness as illustrated in
Figure 3. In this section, each stage from cambial zone to mature xylem including the
primary wall (P) and secondary walls ( S i , S2 and S3) was lined up in order of
maturity. Purification was thoroughly performed according to previous manners (5,6)
to remove pectin, hemicellulose and lignin. After purification it was freeze-dried with
terr-butyl alcohol and then provided for FTIR measurements.
F T I R measurements. FTIR attached with a microscopic accessary was Nicolet
Magna 550 in combination with a Nicplan microscopic attachment. The spectra from
4000-650 cm"l were the average of 64 scans recorded at a resolution of 4 c m ' l with a
M C T detector. As shown in Figure 3, IR beam was irradiated from the radial face
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

direction with an area of 50 χ 300 μπι 2 for each stage of the cell wall formation.
When the beam was irradiated from radial direction, we could not obtain any spectra.
2. FTIR with a special reaction cell for deuteration of cellulosics.
Materials. Three kinds of amorphous cellulosic films were prepared by casting
from their Ν,Ν-dimethyl acetamide (DMAc) solutions. Namely, they were
regioselectively methylated amorphous 2,3-di-O-methylcellulose (23MC)(7) and
amorphous 6-O-methylcellulose (6MC)(8), and pure amorphous cellulose prepared
from a D M A c - L i C l cellulose solution by casting and the subsequent washing with
ethanol. The films were sufficiently thin with a thickness from 5 to 10 μπι to obey the
Beer-Lambert law (9). The 3 amorphous samples were confirmed by the very
diffused patterns in their wide angle X-ray diffractgrams.
F T I R measurements. The film samples above mentioned were fixed in a reaction
cell as shown in Figure 2. Inside of the cell there was a pool for heavy water and it
was filled and saturated with D2O vapor (10,11). IR beam passed through calcium
fluoride windows and irradiated the sample and then was detected by a DTGS
August 31, 2012 | http://pubs.acs.org

detector. The IR spectra were obtained every 5 minutes for the first 5 hours and then
every 15 minutes for the next 2 hours and finally every an hour until the end of the
deuteration using a Nicolet Magna 550 spectrophotometer. The wavenumber range
scanned was 4000-400 c m ' l ; scan rate was 28 scans per minute; 32 scans of 2 cm"l
resolution were signal averaged and stored.

Results and Discussion

1. FTIR with a microscopic accessary. The main advantages of the this method is
that FTIR spectra for such a small amount of sample less than 1 mg can be obtained
from small area (minimum : 50 χ 50μπι2). Therefore this can be applied to the in situ
biological system. For this application we have to remove H2O to get completely
dried state. Under such condition, we have investigated the crystalline form and
crystallinity in both primary and secondary walls during coniferous wood cell wall
formation
The process of biosynthesis and subsequent crystallization of cellulose is the
initial stage of plant cell wall formation. Although many in vitro investigations on
this process have been carried out, the in vivo mechanism for cellulose crystallization
has still not been clearly resolved. Several years ago, Attala and VanderHart (12)
found that the cellulose I crystalline structure is really a composite of two allomorphs,
thereby providing a new method by which to investigate the mechanism of cellulose
crystallization. Why are two cellulose allomorphs, Ια and Ιβ, crystallized at the
surface of plant cells? A n earlier study (13) pointed out that cellulose from enlarging

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 175

Cellulose structure has been examined

through means of FT-IR, Raman, NMR, X-ray, etc.

FT-IR method, if combined with suitable attachments, can

provide more important information on cellulose
supermolecular structure.

0
<2=J>
Microscopic special
attachment reaction cell
We could examine small amount of |/Ve could examine change of cellulose
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

cellulose focused on a small area. by diffusion of a gas (D2O etc.).

Ο
Cell wall cellulose was successfully
. o
Deuteration processes were
examined during cell wall formation. completely followed.

Figure 1. Two IR methods introduced in this present paper.

August 31, 2012 | http://pubs.acs.org

External View Sectional View

Figure 2. The IR sample cell used for monitoring the deuteration process.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 176

algae was rich in the Ια phase, whereas the I β phase was the dominant cellulose
component in higher plants which form thick secondary walls once the primary
geometry of the cells has been established.
Here, we want to explore the idea that plant cellular forces exerted during growth
have a direct influence on crystallization of newly biosynthesized cellulose which can
be gel or liquid crystal phase. This paper provide an in vivo evidence that the major
crystalline component of cellulose changes from Ι α to Iβ when cellular enlarging
growth ceases.
The purified sections were analyzed on small areas (300 X 50 μπι) by a FTIR
spectrometer using a microscopic attachment which was focussed on the radial face of
the sections as already described. After FTIR analysis these sections were observed
by a scanning electron microscope (JEOL JXA840A) to examine tracheid cell wall
formation. This procedure allowed us to examine the developing cellulose crystalline
structure in four distinct stages of the forming cell wall (P, P+Sl, P+S1+S2,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

P+S1+S2+S3) as illustrated in Figure 3.

Cellulose Ι α shows characteristic IR bands at 3240 and at 750 cm-1, while the
characteristic bands for the I β phase appear at 3270 and at 710 cm"l (14). Significant
differences in the FT-IR spectra appeared in 1000 - 650 cm-1 region obtained for both
wood samples as shown in Figure 4. The IR absorption band at 750 cm-1 attributed to
the Ια cellulose form appeared only in the spectra from the primary wall (P). This
band became weaker with the subsequent deposition of the secondary wall layers on
the pre-formed wall layers (P+Sl, P+S1+S2, P+S1+S2+S3). On the other hand, the
band at 710 cm-1 assigned to Iβ cellulose was not detected in the spectra for the
primary wall, whereas after the start of the secondary wall deposition, it began to
appear as a significant shoulder peak. Because the secondary wall is much thicker
than the primary wall, the deposition of the secondary wall containing the Ιβ phase
results in the baseline intensity increasing and it begins to overlap the Ια absorption.
Thus, in the region of the spectra having both walls, the Ι α absorption band is not
August 31, 2012 | http://pubs.acs.org

clearly distinguishable. These results show that the cellulose in the primary wall is
rich in the Ια phase, whereas the cellulose in the secondary wall is rich in the Ιβ
phase. The second derivatives of IR spectra obtained from the O H stretching region
after deuteration (Figure 5) also supported the above suggestion. Although the
absorption due to Ια phase at 3240 cm-1 appeared unclear in the FT-IR spectra even
for the deuterated primary wall (a in Figure 5), this absorption was proved clearly in
the second derivative spectrum (b in Figure 5). On the other hand, it was not detected
in the spectra for the deuterated mature wall (P+S1+S2+S3) in the same Figure.
According to Fengel et.al (15, 16), the band position due to Ια phase may change
after the deuteration while the band due to Iβ phase at 3472 cm-1 is hardly influenced
by milling or deuteration. In our case for the comparison between Ρ and
P+S1+S2+S3, the second derivative spectra after the deuteration showed the
difference significantly as Michel reported previously (77).
There are two contradictory reports that wood cellulose is rich in the Ια phase (18)
while at the same time the I β phase is dominant in it (6 ). Our IR results can explain
this apparent contradiction because the mixture of Ια-rich primary and Ιβ-rich
secondary walls was analyzed concurrently in the samples in those reports.
In addition, primary wall cellulose has been described as being disorganized
cellulose I (5) or totally amorphous (19). However, we have obtained a result that
crystallinity of tracheid primary wall cellulose can be higher than that of secondary
wall in the following. Figure 6 shows the change of crystallinity determined by a IR
method for each stage of wood cell cellulose formation. At first stage for the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 177

Developing
tracheid walls

Radially
sectioninc

Differentiating
xylem

Phloem Mature
xylem
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

Wood trunk

Figure 3. The sectioning method for obtaining IR spectra for each stage of
tracheid wall formation. P; primary wall, and S1-3; secondary walls. In the
developing tracheid walls S i layer deposits on the primary wall from inside
(cell side), and in the same manner S layer deposits on the S - 1 -
n n

Japanese cedar Japanese cypress

ο
Φ
wall foi
August 31, 2012 | http://pubs.acs.org

pci δ"

ΙΟΟΟΘΟΟβόΟ 700 1000 900 800 700

Wavenumbers (cm ) -1

Figure 4. Comparison of IR characteristic absorption bands for the Ι α forms

1 1
of cellulose (at 750 cm" ) with that for I β forms (at 710 c m " ) in each stage
of two coniferous developing tracheid walls; Japanese ceder (left) and
Japanese cypress (right). Spectra for Valonia and Halocynthia represent
standard spectra of cellulose rich in Ια and pure Ιβ allomorphs, respectively.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 178

3500 3250 3000 2750 2500

1
Wavenumbers(cm )
|3240(Ια)

3
[D 0]
2

£±31+S2+S3
P+S1+S2+S3[D 0]
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

Valonia (la-rich)
Halocynthia (Ιβ-rich)
3400 3350 3300 3250 3200
1
Wavenumbers (cm* )

Figure 5. Changes in FT-IR spectra for deuterated cellulose in the region

1
from 3400 to 2800 cm" (A) and their second derivatives (B) from Japanese
cypress tracheid walls composed of Ρ and P+S1+S2+S3, compared with
standard second derivative spectra of non-deuterated cellulose rich in Ια
(Valonia) and pure Iβ (Halocynthia) allomorphs. IR absorption bands at 3240
1
cm" is characteristic for the Ια form of cellulose.
August 31, 2012 | http://pubs.acs.org

Cell Wall Formation

Radial enlarging

Hi
^ ^P+Si]^Si+^[^Si+S2+S3

X c
ο c
•σ ο
.£ ω
> oo
IN
= CM

% -

ce η 0 100 2 0 0 3 0 0 4 0 0 500 6 0 0
CO
Distance from the boundary
between xylem and phloem (μπι)
1
Figure 6. Change of IR crystallinity index (1427 / 895 cm" ) for cellulose
during tracheid cell wall formation. Results are averaged of 6 analyses.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 179

formation of primary wall the crystallinity of the cellulose increased, and then with
the deposition of secondary wall the crystallinity started to decrease gradually, and
finally it became saturated at 2.1 as IR crystallinity index which corresponds to
approximately 60 % for the crystallinity index determined by X-ray. The value
coincides with the crystallinity for ordinary wood. These results clearly suggest that
crystallinity for the primary wall cellulose is higher than that for the secondary wall
cellulose. As described above, using FTIR equipped with a microscopic attachment
enabled us to obtain in situ changes of both cellulose crystalline form and crystallinity
during cell wall formation.
2. FTIR with a special reaction cell for dueteration of cellulosics. In previous
papers (1-3), we believed that a characterization of the hydrogen bonds found in
amorphous cellulose would be of fundamental value. Furthermore, we proposed that a
structural study of amorphous cellulose in light of hydrogen bonding might be a first
step in uncovering details of how molecules rearrange in going from the liquid to the
crystalline state (3). For the above objectives, it is important to get the information
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

about "amorphous regions" for cellulose. Currently there is a shortage of structural

information about the amorphous regions, perhaps partly because terminology
suggests and impression persists that molecular chains in these regions are completely
without structure and partly because methodology has been limited to W A X D and
1 3
CP-MAS C - N M R for measuring order in the presence of substantial amount of
disorder. It is not uncommon for substrates that are not identifically crystalline by a
method such as X-ray diffraction to be labeled "amorphous", but definition of
amorphous goes beyond noncrystalline to unorganized and having no pattern of
structure (20). At present FTIR is one of the best tools available to derive structural
information including hydrogen bonding. Therefore, we particularly wanted to
examine details of the amorphous regions for the amorphous cellulose film samples,
regioselectively methylated 23MC and 6 M C amorphous film samples, as model
components of amorphous cellulose using FTIR methods (3).
Applications of deuteration methods to IR have been so far focussed on the
separation of IR spectra for cellulose structure into two parts of crystalline and
amorphous regions, respectively, and then only the discriminated crystalline regions
August 31, 2012 | http://pubs.acs.org

have been studied (21-23). In this paper we will focus only on the amorphous regions
for our cellulose model compounds using the FTIR monitoring of the deuteration
process. The main advantages of this method is that the use of our reaction cell has
enabled us to obtain not only change of reproducible and correct IR spectra for
amorphous regions in cellulose, but also the rate constant for the deuteration reaction
which is assumed to correspond to the nature of amorphous regions. Thus we have
attempted to characterize the morphology for amorphous regions by analysing the
diffusion behavior of D2O as a probe. Amorphous cellulosic samples 23MC and 6MC
we used are shown in Figure 7. As already reported in a previous paper (2), each
polymer is considered to have different types of inter- and intramolecular hydrogen
bonds. Namely, 6MC may have two intramolecular hydrogen bonds, one between the
OH at the C-3 position and an adjacent ether oxygen of the glucose ring, and the
second between the ether oxygen at the substituted C-6 position and an adjacent OH
at the C-2 position while 23MC may have both inter- and intramolecular hydrogen
bonds at the C-6 position.
Figure 8 shows change of IR spectra for 6MC accompanied by the deuteration
process. When the deuteration proceeded, OH stretching vibration around 3470 cm-1
decrease, and instead the bands at 2559 cm"l due to OD at the C-2 and C-3 positions
appeared and then increased. The similar behaviors were observed for the OH and OD
bands in the amorphous 23MC and cellulose. However, the deuteration was not
completed regardless of the amorphous samples, giving unreacted O H groups. The
amount of the unexchangable O H was 14.3, 10, and 13.3 % for 6MC, 23MC and
amorphous cellulose, respectively. We assume that these unexchangable hydroxyl

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 180

( i ) 2,3-di-O -methylcellulose (23MC):

Intra; OH at C - 6 and OCH at C - 2 .
3

Inter; OH at C - 6 .

( ii ) 6-0 -methylcellulose (6MC):

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

Intra; OH at C - 3 and Ο in the neighboring ring,

OH at C - 2 and OCH at C - 6 .
3

Inter; Nothing.

Figure 7. Hydrogen bonding formation proposed for 23MC and 6MC.

August 31, 2012 | http://pubs.acs.org

Figure 8. Changes of OH and OD IR absorption bands for monitoring the

deuteration of amorphous 6MC.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 181

bands may indicate the presence of domains involving the intermolecular hydrogen
bonds which is already proposed as an amorphous model (5).
Here we will postulate that the OH-OD exchange reaction obeys pseudo-first
order kinetics because of the large amount of D2O to the amount of OH in each film
sample. The general equation, -d[OH]/dt=k[OH], can be employed. Thus the derived
relationship, ln{[OH]t/[OH]o}=-kt, was drawn in Figure 9 for the 3 amorphous
samples, 23MC, 6MC and cellulose, respectively. Each of 3 samples showed the
similar kinetic behaviors which were devided by three single reactions. In other
words, three kinds of the exchange reaction (1, 2, and 3 in Figure 9) can be
competitively coexisted. Table I shows the rate constants for each reaction.

Table I Rate constants (k) from OH to OD for every single reaction

in the deuteration processes of the amorphous 3 film samples.

Time course 1
k(rf ) R 2
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

Reaction 1
23MC 0.08 - 5 hr 0.136 0.98
6MC 0.08 - 4 hr 0.247 0.99
Cellulose 0.08 - 2 hr 0.264 1.00
Reaction 2
23MC 6-12hr 0.085 0.98
6MC 5-13hr 0.057 0.99
Cellulose 3-12hr 0.075 0.98
R: Correlation coefficient.

The rate constant for reaction 1 was relatively different among the 3 samples. On
the contrary the rate constant for 2 reaction seems to be similar among them. Of
course, this is still speculative. In general adsorption and diffusion in crystalline
polymers is considered to occur in amorphous regions. However, it is very complex
August 31, 2012 | http://pubs.acs.org

because the structure for the crystalline regions may have somehow influences on it.
In fact the morphological effects has not sufficiently studied yet. In this present case
we have only to investigate the adsorption and diffusion of D2O since the film
samples exhibited amorphous. In another paper, details of this interpretation will be
reported.
Considering that the samples exhibited amorphous for the X-ray measurements,
amorphous regions in these cellulosics may be discriminated into 3 phases i)
deuteration is fast, corresponding to reacion 1, ii) deuteration is slower, corresponding
to reaction 2, and iii) almost undeuterated, corresponding to reaction 3. Interestingly
the second rate constants among the three samples were very similar although the first
rate constants for them differed. This indicates that the second one may correlate with
the deuteration for the intramolecular hydrogen bonds (see Figure 7) in amorphous
phases which are common interactions among the above three samples. Further
investigation will be needed because the type of intermolecular hydrogen bonds found
at the C-2, C-3 and C-6 hydroxyl positions is participating, to some extent, in
determining the structure of amorphous cellulose (3).

Acknowledgments
We thank Dr. A . Isogai, Mr. E . Tsushima, Mr. R. Nakata, Dr. K . Takabe and Ms. N .
Hayashi for providing the native cellulose samples.

Literature Cited
1. Kondo, T.; Sawatari, C.; Manley, R. St.J.;Gray, D. G. Macromolecules 1994,
27, 210.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 i
Ο

Ο
JE

Ο 10 20
T i m e c o u r s e (hr)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012
August 31, 2012 | http://pubs.acs.org

0 10 20 30
T i m e c o u r s e (hr)

0 10 20 30
T i m e c o u r s e (hr)

Figure 9. The relationships between decreasing ratio of OH groups,

ln{[OH]t/[OH]o} and time, t, during the deuteration process for the three
amorphous samples, 23MC, 6MC and pure cellulose.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 183

2. Kondo, T. J. Polym. Sci.B: Polym. Phys. 1994, 32, 1229.

3. Kondo, T.; Sawatari, C. Polymer 1996, 37, 393.
4. Kataoka, Y.; Kondo, T. Macromolecules 1996, 29, 6356.
5. Chanzy, H.; Imada, K.; Vuong, R.; Barnoud, F. Protoplasma 1979, 100, 303.
6. Wada, M.; Sugiyama, J.; Okano, T. Mokuzai Gakkaishi, 1994, 40, 50.
7. Kondo, T.; Gray, D. G. Carbohydr. Res. 1991, 220, 173.
8. Kondo, T. Carbohydr. Res. 1993, 238, 231.
9. Coleman, M. M.; Painter, P. C. J. Macromol. Sci., Rev. Macromol. Chem. 1978,
C16, 197.
10. Tsuboi, M. J. Polym. Sci. part C 1964, 7, 125.
11. Smith, J. K. J. Polym. Sci. part C 1963, 2, 499.
12. Atalla, R. H.; VanderHart, D. L. Science 1984, 223, 283.
13. Horii, F.; Yamamoto, H.; Kitamaru, R.; Tanahashi, M.; Higuchi, T.
Macromolecules 1987, 20, 2946.; Sugiyama, J.; Okano, T.; Yamamoto, H.; Horii,
F. Macromolecules 1990, 23, 3196.
14. Sugiyama, J.; Persson, J.; Chanzy, H. Macromolecules 1991, 24, 2461.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch012

15. Fengel,D.Holzforschung 1993, 47, 103.

16. Fengel, D.; Strobel, C. Acta Polymer. 1994, 45, 319.
17. Michell, A. J. Carbohydr. Res. 1993, 241, 47.
18. Tanahashi, M.; Goto, T.; Horii, F.; Hirai, Α.; Higuchi, T. Mokuzai Gakkaishi
1989, 35, 654.
19. Nowak-Ossorio, M.; Gruber, E.; Schurz, J. Protoplasma 1976, 88, 255.
20. Rowland, S. P.; Howley, P. S. Text. Res. J. 1988, 58, 96.
21. Marrinan, H. J.; Mann, J. J. Polym. Sci. 1958, 32, 357.
22. Jeffries, R. Polymer 1963, 4, 375.
23. Taniguchi, T.; Harada, H.; Nakato, Κ Mokuzai Gakkaishi 1966, 12, 215.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 13

Molecular Weights and Molecular-Weight Distributions

of Cellulose and Cellulose Nitrates During Ultrasonic
and Mechanical Degradation

Marianne Marx-Figini

Polymer Division, INIFTA, National University, La Plata, Argentina

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

Cellulose and cellulose nitrate were subjected to

depolymerization by ultrasonication and high-speed mechanical
agitation (stirring). Both treatments resulted in significant, but
different, depolymerization with the resulting degradation products
having considerably narrower molecular weight distributions than the
parent substrates. In all cases a level-off degree of polymerization
was reached which was different for unsubstituted cellulose and the
derivatives. The results obtained with unsubstituted cellulose are
August 31, 2012 | http://pubs.acs.org

explained with the existence of "weak links" in cellulose backbones.

The treatment of polymers with ultrasound has received considerable

attention in the past (1-6). Cellulose depolymerization by ultrasonification, by
contrast, has been largely overlooked (7-8); additionally, this work predates the
advent of modern and reliable cellulose molecular weight determination
methodology. Since one of the unusual features of polymer degradation with
ultrasound is the narrowing of the molecular weight distribution in addition to the
apparent existence of a limiting level-off DP, it was of interest to reexamine
cellulose (derivative) degradation by ultrasound in view of the potential of
generating cellulose preparations with narrow molecular weight distributions.
Alternatively, this task requires time consuming fractional precipitation for
preparative size exclusion chromatography (SEC) experiments.
Likewise, cellulose degradation studies using high speed mechanical agitation
(9-12) were performed at a time when reliable molecular weight methodology did
not exist. Since an understanding of the molecular weight behavior of cellulose
under mechanical stress can provide insight into the effect of industrial processing

184 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 185

on macromolecular properties of cellulose, cellulose depolymerization during

mechanical high speed stirring was investigated as well.

Experimental Section

L Determination of Molecular Weights and Molecular Weight

Distributions: Molecular weights were determined viscosimetrically as well as by
size exclusion chromatography (SEC). The resulting molecular weight data were
interpreted in terms of degrees of polymerization (DPs) so as to reach a universally
applicable polymeric size parameter that is dependent only on the number of repeat
units and not on the average weight or substitution pattern of repeat units. DP is
independent of the nature and degree of substitution of the anhydroglucose repeat
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

unit. Viscosimetric measurements were carried out in ethyl acetate and

cupriethylenediamine (cuen) solution for cellulose nitrate and unsubstituted
cellulose, respectively, according to Marx-Figini and Schulz (13). All measurements
were performed with solutions whose concentration was adjusted so that η was 8 ρ

between 0.3 and 0.6. All intrinsic viscosity measurements were expressed as D P
using Equations 1 to 4.

076 0 7 6
ME. =5.70XDP (1) and [ η ] 0ϋβη =2.29xDP (2) forDP>950

fo] a
E =1.06xDP (3) and h ] c u e n =0.42xDP (4)forDP<950
August 31, 2012 | http://pubs.acs.org

where Ea stands for ethylacetate. These equations were recently established using
more than 60 representative light scattering and ultracentrifugation data (14).
Adjusting the pertinent values of [η] to standard conditions, these equations were
found to produce considerable agreement for data from many different authors. In
addition, the fact that the ratios of intrinsic viscosities of cellulose nitrate in two
different solvents as well as those for cellulose nitrate and cellulose remain constant
over the entire DP range serves as an indication that the exponent α in the Mark
Houwink equation is invariable for all different solvents investigated (15).
Molecular weight determinations by SEC were performed using a Waters
6
H P L C unit with UV-detector (model 441) and a set of ultrastyragel (10 A) and
5 4 3
microstyragel (10 , ΙΟ , 10 Â) columns. THF served as solvent, and known
cellulose nitrate samples were used as calibration standards. DP was derived by

logDP = A - B V c (5)

where V is elution volume, and A and Β are three different pairs of parameters for
e

DP ranges between 100 and 8,000. Different constants for A and Β were used for

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 186

the DP ranges of 100 to 500; 500 to 1000; and 1000 to 10,000, and their
mathematical derivation is described elsewhere (17, 18). The resulting DP values
agreed well with D P and D P determined by viscosimetry and osmometry,
n n

respectively. (See reference 16 for more information.)

Π. Molecular Degradation: Ultrasonic Degradation - Five nitrated cotton

cellulose samples with D P between 400 and 7,500 and a degree of substitution of
n

2.90 + 0.02 were subjected to sonication in ethylacetate solution at a concentration

between 0.045 and 0.15%. The sonication apparatus was a Bandlin-Sonopuls H D -
60 apparatus with a frequency of 20 kHz and a delivered power of 22 W.
Temperature was maintained constant at 20°C. (Concentration was found to play
an insignificant role in the molecular degradation). Samples taken after different
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

sonication times were analyzed by viscosimetry and SEC. D P , D P , and D P

n n W

values were calculated. In each case ϋ Ρ -values obtained by viscosimetry agreed

η

well with those obtained by SEC.

High SpeedStirring - Six cotton cellulose samples with DP-values between
300 and 7,500 in N-butanol suspension (150 mg cellulose per 100 mL butanol) were
subjected to a high speed rotary homogenizer operating at 13,500 rpm. Degradation
experiments were performed at a constant temperature of -30°C, and the molecular
weights of the filtered and dried degradation samples were measured in cuen as well
as by SEC following nitration.

Results and Discussion

August 31, 2012 | http://pubs.acs.org

The macromolecular degradation of cellulose nitrate samples during exposure

to ultrasound reveals that the degradation rate increases with increasing DP (Figure
1). As D P declines, the effect of sonication time (t) diminishes. All cases seem to
0

result in a limiting DP (DP..) below which no further degradation takes place.

Similar results were reported for other polymers, principally low dispersity
polystyrenes (5). Despite the large differences between the DP -values, all D P - -
0

values lie in a narrow range, between D P 500 and 700. Samples having a D P
n 0

lower than the lowest D P - do not degrade. In agreement with observations made
with synthetic polymers (5), cellulose depolymerization proceeded faster, and to a
lower level-off DP», if delivered sonication power was raised from 22 to 37 W.
Corresponding observations were made with the depolymerization of
suspended cellulose using high speed mechanical stirring (Figure 2). A limiting,
level-off DP is reached following an exponential degradation rate. However, in
contrast to degradation by sonication, mechanical degradation results in (a) a higher
DP--value; (b) D P - is independent of D P ; and (c) the decrease of DP is
0

independent of the rate of stirring (8,000 or 13,500 rpm). The results suggest that

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 187

10 20 30 40 50 60 70 ÔO 90
t (min.)-
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

Figure 1. Behaviour of ΟΡ \νΰη sonication time. Key: χ is Ό? (0) = 7500; Ο is Ό? (0)

η Ά Ά

= 1650; <is ϋ Ρ (0) = 392; · is ΌΡ (0) = 3430; and • is ΌΡ (0) = 900.

η Ά Ά

10

Ο
August 31, 2012 | http://pubs.acs.org

Figure 2. Behaviour of the ϋΡ \νίιη high-speed stirring time (rpm = revolutions per min­
η

ute). Key: O , O, • , • , χ, Δ indicates rpm = 13,500; « i s rpm = 8000; O, · is ϋ Ρ (0) = η

7300; • is Ό Ρ (0) = 2207; χ is ΌΡ (0) = 915; Ο is ϋ Ρ (0) = 4150; • is ϋ Ρ (0) =

η Ά η η

1560; Δ is ΌΡ (0) = 330.

Ά

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 188

1}
Table I: Cellulose Nitrate Polydispersity-Values (U) as Related to
Sonication Time.
Sonication Time (min) Total
D P
n(0)
0 5 10 20 45 90 Decrease * 2

(%)

7500 1.66 1.33 1.32 1.32 3

1.34 > 1.30 54

3700 1.89 1.86 1.39 1.44 1.43 1.38 57

1700 1.92 1.53 1.53 1.45 1.41 1.41 55

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

937 1.92 1.75 1.70 1.57 1.56 1.49 47

455 1.96 2.08 1.99 2.11 2.05 2.02 -

1)
U = DP /DP w n

2 )
OfU
3 )
Sonication time of 40 min.

cellulose degradation by stirring, like cellulose nitrate degradation by

ultrasonication, does not proceed in a random scission process but by selective
bond rupture. This is suggested by the behavior of the molecular weight
distributions.
Polydispersity, U , (DP /DP ) reveals a significant narrowing of the
August 31, 2012 | http://pubs.acs.org

w n

molecular weight distribution with sonication time (Figure 3). U was found to
depend on both sonication and D P (Table I). This suggests that molecular
0

degradation proceeds more rapidly with high DP preparations than with lower
ones.
The degradation behavior of macromolecules during ultrasonication is
generally assumed to proceed by cleavage of the molecules in their center, or
according to a Gaussian distribution on both sides adjacent to the center. For this
case, bimodal DP distributions were predicted (on mathematical grounds) and also
experimentally observed (2, 4). The molecular weight distribution curves obtained
with cellulose in this study, however, are uniform rather than bimodal. Bimodal
distributions were obtained only from nearly monodisperse polystyrenes (U <1.2)
(2,4). The present cellulose nitrate samples showed initial distributions having a U
of 1.6 to 2.0. Macromolecules with an initial polydispersity of U >1.5 are also
predicted to produce degraded polymers with uniform distributions (19). This
agrees with the current findings. It can therefore be concluded that ultrasonic
degradation of cellulose nitrate proceeds via central scission or scission near the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 189

17,5 τ
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013
August 31, 2012 | http://pubs.acs.org

DP

Figure 3. Distribution curves of a representative sample (DP^o) = 1,650,

D P n ( O ) 943 at different times of sonication.
=

Starting sample (....); degradation time 5 minutes (-.-.-);

degradation time 10 minutes ( ); degradation time 20 minutes (- );
degradation time 45 minutes (-..-); degradation time 90 minutes (—).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 190

center of the molecule. This is in contrast to synthetic polymers which

depolymerize following a Gaussian distribution (2, 3, 19).
Native (with respect to DP) celluloses, derived by biosynthesis are
monodisperse (20). Unprocessed cotton cellulose, however, is available only with a
polydispersity of about 1.5 or 1.6 due to unavoidable degradation reactions during
storage. This U still increases with industrial processing, and it eventually
approaches a value of 2.0. Molecular degradation by ultrasonication produces
uniform materials with reduced U-values. High speed stirring of cellulose
suspensions produces DP» values of nearly 3,200 regardless of D P . This chain
0

length corresponds approximately to the number of anhydroglucose units between

two "weak links" in a cellulose molecule of high DP (21). Such "weak links" have
recently been revealed in studies on the acid catalyzed hydrolysis of cellulose (22).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

According to Marx-Figini and Coun-Matus (22), each native cellulose molecule

(with D P of approximately 13,000 (20)) has three to four "weak links." The
0

distance between two weak bonds would therefore be approximately 3,200

6
monomer units. Under acid hydrolysis conditions, "weak links" break 10 times
faster than regular glycosidic bonds. "Weak link" hydrolysis has a lower activation
energy than that found for glycoside hydrolysis (22), but it is high enough to permit
the assumption that there are native modifications in the chemical structure of the
cellulose molecule which cause the rapid splitting. The alterations in the glucose
unit may, for example, involve the occurrence of 2-deoxyglucosides or fructosides,
3 5
or xylan-units, the hydrolysis rate constants of which are known to be 10 - 10
times larger than the corresponding β-1,4 glucoside. A n analytical proof of the
chemical nature of the alterations, however, appears impossible because of the very
low fraction of the non-glucosidic units. Assuming that mechanical force only
August 31, 2012 | http://pubs.acs.org

breaks "weak links," a narrowing of molecular weight distribution can be expected,

and this was experimentally confirmed (21) (Figure 4).
Assuming that only "weak links" were cleaved during cellulose degradation,
polydispersities should decline as depolymerization progresses. A comparison of
experimental with theoretical data provides compliance with this model (Figure 5).
The mathematically derived data (21) were derived in accordance with an equation
system previously developed by Schulz and Husemann. (23). The data shown in
Figure 5 are based on the starting values of DP and U obtained by SEC on the initial
material. The data of Figure 5 also reveal that high speed mechanical stirring also
produces reduction of U .
Since there is no definitive evidence as yet in support of a relationship
between crystallite size and DP, no attempt was made to explain the mechanical
degradation behavior with crystallite dimensions. Since former detailed
investigations on the heterogeneous hydrolysis of cellulose (22) have shown that
heterogeneous state fails to influence degradation behavior (by, for example,
diffusion control), no effect of crystallite size on polydispersity index was
expected.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 191

2.5
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

Figure 4. Distribution curves of a representative sample ( ϋ Ρ η ( 0 ) = 7,500,

DPn(O) = 4,630) after different times of high-speed-stirring.

Starting sample (...); degradation time 15 minutes ( );

degradation time 30 minutes (—); degradation time 2 hr (-.-.-).
August 31, 2012 | http://pubs.acs.org

0 1
1 1.2 1.4 1.6 1.8 2 2.2 2f*
DP /OP -
w(o)/ w(t)
Figure 5. Polydispersity (U) as a function of degree of degradation. Continuous
line: mathematical model.
Experimentally determined data points: - Ο - , -€)-, - · -, and - © -
represent degradation times of 5 minutes (-Ο-), 15 minutes (-•-), 30 minutes
( - © - ) , 2 hr ( - · - ) , and 3 hr. (-Θ-).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 192

Conclusions

Treatment of cellulose nitrate in solution with ultrasound causes a rapid

decline of DP which is accompanied by a significant narrowing in polydispersity.
This is limited, however, to cellulose derivatives having a DP^o) « 7 0 0 . Native
celluloses with ϋ Ρη ( 0 )>10,000 and U <1.4 already are narrowly dispersed as a
consequence of biosynthesis. Sonication of these materials would indeed further
reduce their polydispersity, but according to results obtained with polystyrene (3,
4), the degradation would provoke a bimodal distribution. This means that only
polymers with an initial polydispersity of between 1.6 and 2.0, corresponding to
cellulose derivatives with aDP^o) between 900 and 8,000, are capable of generating
degraded polymer solutions with narrow distribution values.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013

Suspended cellulose is degraded by mechanical action only if DP >3,000. 0

Degradation is independent of applied shear forces. The level-off DP-value is

reached rapidly. Once reached, the limiting DP, which is independent from DPQ, no
longer declines. Both observations, the existence of a level-off DP and the
narrowing of U during degradation, suggest that mechanical shear forces degrade
cellulose by cleaving "weak links" present in native cellulose. The results observed
in this study are consistent with this hypothesis (21). The results therefore suggest
that mechanical treatments of cellulose during industrial processing, including
storage, are responsible for significant molecular degradation.

Literature Cited
August 31, 2012 | http://pubs.acs.org

1. Jellinek, H. H. G. and G. White. J. Polym. Sci. VI, Ν 6, (1951) 757.

2. Glynn, P. A. R. Β. M. E. Van der Hoff. J. Macromol. Chem. A7/8(1973)
1695.
3. Van der Hoff, Β. M. E. and P. A. R. Glynn. J. Macromol. Sci.-Chem., A8/2
(1974) 429.
4. Niezette, J. and A. Linkens. Polymer 19 (1978) 939.
5. Price, G. J. and P. F. Smith. Polymer International 24 (1991) 159.
6. Price, G. J. and P. F. Smith. Polymer 34 (1993) 4111.
7. Thomas, Β. B. and W. J. Alexander. J. Polym. Sci. 15 (1955) 361.
8. Thomas, Β. B. and W. J. Alexander. J. Polym. Sci. 25 (1957) 285.
9. Grohn, H. J. Polym. Sci. 30 (1958) 551.
10. Grohn, H. and W. Deters. Faserforsch. und Textiltechn. 13 (1962) 544.
11. Deters, W. and H. Grohn. Faserforsch. und Textiltechn. 14 (1963) 58.
12. Ott, K. L. J. Polym. Sci. Part. A2 (1964) 973.
13. Marx-Figini, M. and G. V. Schulz. Makromol. Chem. 31 (1959) 140.
14. Marx-Figini, M. and G. V. Schulz. Makromol. Chem. 54 (1962) 102.
15. Marx-Figini, M. Angew. Makromol. Chem. 72 (1978) 161.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 193

16. Soubelet, O., M. A. Presta and M. Marx-Figini Appl. Macromol. Chem.

175 (1990) 117.
17. Vribergen, R. K., A. A. Soeteman, G. A. M. Smit. J. Appl. Polym. Sci. 22
(1978) 1267.
18. Andreetta, A. H. and R. V. Figini. Appl. Macromol. Chem. 93 (1981) 143.
19. Ballauf, M. and B. A. Wolf. Macromolecules 14 (1981) 654.
20. Marx-Figini, M. J. Polym. Sci. C', Ν 28 (1969) 57.
21. Marx-Figini, M . and R. V. Figini. Appl. Macromolec. Chem. 224
(1995)179.
22. Marx-Figini, M. M. Coun-Matus, Makromol. Chem. 182 (1981)3603.
23. Schulz, G. V., E. Husemann, Z. Phys. Chem., Abt. B, 52 (1942) 23.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch013
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 14

Depolymerization of Cellulose and Cellulose Triacetate

in Conventional Acetylation System

Shu Shimamoto, Takayuki Kohmoto, and Tohru Shibata

Research Center, Daicel Chemical Industries, Ltd., 1239 Shinzaike, Aboshi-ku,

Himeji, Hyogo, 671-12 Japan
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

In a conventional system for the acetylation of cellulose, acetic acid,

acetic anhydride and sulfuric acid act as a diluent, an acetylation reagent
and a catalyst, respectively; the sulfuric acid acts as a catalyst not only
for acetylation but also for depolymerization. The depolymerization
behaviors of cellulose and cellulose triacetate were studied so that the
degree of polymerization of the final product could be predicted. Model
experiments in which the acetylation reagent was absent revealed that
the depolymerization of cellulose in earlier stages of the reaction
proceeds considerably faster than that of cellulose triacetate, and
depolymerization of cellulose triacetate is random whereas that of
cellulose is not. Simulations of the final degree of polymerization of the
product were carried out using the activation energies obtained by the
August 31, 2012 | http://pubs.acs.org

model experiments. The estimated degree of polymerizations showed

good agreements with the experimentally obtained ones when the
depolymerization rate of cellulose was assumed to be proportional to the
amount of sorbed sulfuric acid.

Among known systems for the acetylation of cellulose, one comprised of acetic acid,
acetic anhydride and sulfuric acid has been widely employed in commercial
productions of cellulose triacetate (CTA) and cellulose diacetate because of its
efficiency (7). Sulfuric acid catalyzes depolymerization as well as acetylation in the
system. Although there are some studies on the depolymerization during the
pre treatment (2) and acetylation (J), a quantitative prediction of degree of
polymerization (DP) of the final product based on depolymerization kinetics of
cellulose and the intermediate product has not been reported as far as we know. In an
attempt to accomplish this, depolymerization behaviors of cellulose and C T A were
studied in the system in the absence of the acetylation reagent in this paper.

Result and Discussion

Evaluation of DP of cellulose. Although there are several known methods to

evaluate DP of cellulose, we established an novel method in which cellulose was
converted to C T A without remarkable degradation in order to apply the same Mark-
Houwink-Sakurada equation to the evaluation of depolymerization of cellulose and

194 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 195

C T A . That was to avoid a systematic error caused by applying different Mark-

Houwink-Sakurada equations to cellulose and CTA.
In order to examine the degradation during the acetylation using a lithium
chloride / dimethyl acetamide mixture, cellulose regenerated from C T A with hydrazine
was converted to CTA again by the method described in the experimental section. The
original and the resulting CTA were analyzed by a GPC low angle laser light scattering
( G P C - L A L L S ) technic. The results are shown in figure 1 as molecular weight
distribution curves. The distribution curves were almost identical although there was a
slight difference at the higher molecular weight region, and the difference in weight
average molecular weight was quite small, only 8 %. From these results, it was
concluded that there is no remarkable degradation of cellulose during the acetylation
using a lithium chloride / dimethyl acetamide mixture.

Depolymerization of cellulose and C T A in acetic acid / sulfuric

acid system. Depolymerization of cellulose and C T A in an acetic acid / sulfuric acid
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

system in the absence of an acetylation reagent were studied separately. Cellulose

degraded much faster than CTA in earlier stages of the reaction as shown in figure 2.
It is well known that number average DP at reaction time t (DP ) in a random
mt

depolymerization is given by the following kinetic equation (5,4).

1 1
-+ k t (1)
DPn,t DPn,0
D P , o is the initial D P n and k is the rate constant of depolymerization. Weight average
n

D P ( D P W ) at reaction time t can be expressed by the similar manner by assuming a

Schulz - Zimm type distribution of D P and the polydispersity factor of 2 (5).

1 1
- I-k.«
+ (2)
DPw,t DPw,o 2
August 31, 2012 | http://pubs.acs.org

Equation 2 means that a plot of D P W versus time t gives a linear relationship in a

random depolymerization. C T A gave an almost straight line in the plots whereas
cellulose did not, as shown in figure 3. These results suggest that depolymerization of
C T A in the system is random but that of cellulose is not.
In figure 3a, temperature dependence of CTA depolymerization is clearly seen ;
the higher the temperature is the faster the depolymerization rate is, as expected. The
activation energy of the C T A depolymerization obtained by a plot of the logarithmic
rate constant versus the reciprocal temperature was 13.7 kcal/mol. On the other hand,
cellulose depolymerization showed much smaller dependence on temperature (figure
3b). It is known that sulfuric acid is sorbed on cellulose in an acetic acid / sulfuric acid
system (<5). Malm et. al. showed that the depolymerization rate of cellulose correlates
with the quantity of sulfuric acid sorbed on cellulose (<5). Temperature dependence of
sulfuric acid sorption is shown in figure 4. The amount of sulfuric acid sorbed on
cellulose decreased remarkably with increasing temperature. The small temperature
dependence of cellulose depolymerization could be explained by the temperature
dependence of sulfuric acid sorption.

Simulation of DP of the final product in actual acetylation system.

Simulations of the D P of the final product in an actual acetylation system comprised of
acetic acid, acetic anhydride and sulfuric acid were attempted by using the above
mentioned results.
In order to obtain a kinetic equation for the non-random depolymerization of
cellulose, it was assumed that cellulose was composed of two regions both of which
were depolymerized by random manners. We tentatively call the faster depolymerizing

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 196

Original CTA
Mw 185,000

C
Ο

s Regenerated -
reacetylated CTA
Mw 172,000
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

10" 10" 10° 1(f

molecular weight

Figure 1. Molecular weight distribution of original and regenerated

reacetylated CTAs.
August 31, 2012 | http://pubs.acs.org

60 120 180 240

time (min)

Figure 2. Plots of DP of cellulose and CTA vs. time in acetic acid / sulfuric
W
e
acid system at 50 C.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 197
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

• 50*C
A 40'C
30'C
•
6 0 120 180 240 300 360
time (min)

0.003
Δ
•
August 31, 2012 | http://pubs.acs.org

0.002

Q.
Û
• ο
0.001
ο • 50 C e

A 40'C
0 30'C
o.oooL
0 6 0 120 1 80 240 300 360
time (min)

Figure 3. Plots of 1 / D P vs. time in acetic acid / sulfuric acid system,

W

a : CTA. b : cellulose.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 198

0.5

Ό CD
Ο w 0.4
CO ο
ο 3
' = ô>

a υ 0.3
w τ-

ω ο
€ ε
ο Ê
Ε 0.2 I
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

(Λ

0.1
10 20 30 40 50
temperature f C )

Figure 4. Effect of temperature on amount of sulfuric acid sorbed on cellulose

in cellulose / acetic acid / sulfuric acid system (1 g / 43 g / 0.45 mmol).

Table I. Comparison of experimental DPw and calculated ones for

acetylation of celulose (DPw 2,100) in acetic acid / acetic anhydride / sulfuric
acid system.
August 31, 2012 | http://pubs.acs.org

No. Reaction conditions DPw

Liquid to Amount of Temperature,
solid ratio sulfuric acid, Experimental Cale. 1 Calc. 2
g / 100g cellulose

1 83.8 9.5 40 324 303 326

2 7.0 8.0 40 715 114 780

3 6.0 4.2 60 609 141 631

4 6.0 2.1 60 625 262 757

Calc. 1 : assuming that the rate constants of depolymerization of cellulose were

proportional to the amount of total sulfuric acid.
Calc. 2 : assuming that the rate constants of depolymerization of cellulose were
proportional to the amount of sorbed sulfuric acid.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 199

region amorphous and the slower one crystalline in this paper. In this case, D P at W

time t is given in the following equation.

α 1-α
DPw,t = (3)
DP W j, Ο " 1 + 0.5 · k c r y » · t DPw,0"' + 0.5 · k a m o · t
Where, α is the weight fraction of the crystalline region and kcrys and kamo are the rate
constants of depolymerization of the crystalline and amorphous regions. The rate
constants of depolymerization were determined by a multi-variable fitting method using
the experimental results. In the fitting procedure, α was adjusted at 0.3 to obtain the
best fitting. The activation energies of depolymerization, obtained by plots of the
logarithmic rate constant versus the reciprocal temperature without considering the
variation of the sulfuric acid sorption with temperature, were 0.8 kcal/mol and 6.7
kcal/mol for the crystalline and amorphous region, respectively. However, they were
calculated to be 8.3 kcal/mol and 14.2 kcal/mol for the crystalline and amorphous
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

regions by assuming that the depolymerization of cellulose was propotional to the

amount of sorbed sulfuric acid. While the latter set of activation energies seemed more
reasonable, both of the sets of activation energies were used in the following
simulation.
The following equation was employed to simulate the D P of the final
acetylation product

(4)

In equation 4, the function F(t,x) represents the D P , of the molecules converted from
w t

cellulose to C T A at reaction time τ. The function G(x) stands for the weight fraction of
the molecules converted from cellulose to CTA at reaction time τ.
August 31, 2012 | http://pubs.acs.org

The simulation results obtained under various reaction conditions are listed in
Table I with experimental results. The values in the column of Calc. 1 were obtained
by assuming that the rate constants of depolymerization of cellulose and C T A were
proportional to the amount of total sulfuric acid. The values in the column of Calc. 2
were obtained by assuming that the depolymerization rate of cellulose was proportional
to the amount of sorbed sulfuric acid. The latter simulation gave much closer values to
the experimental results than the former did.

Conclusion

Depolymerization of cellulose is much faster than that of C T A in earlier stages of the

reaction. Depolymerization of CTA in the acetylation system is random, whereas that
of cellulose is not. The estimated degree of polymerizations agreed well with the
experimentally obtained ones when the depolymerization rate of cellulose was assumed
to be proportional to the amount of sorbed sulfuric acid.

Experimental

A prehydrolyzed kraft pulp and C T A with degree of substitution of 2.9 were used
throughout this study. D P of these materials evaluated by the method described
W

below were 2,100 and 1,000 for prehydrolyzed kraft pulp (cellulose) and C T A ,
respectively. The polydispersity factors evaluated by a G P C - L A L L S technic were 2.3
for cellulose and 1.9 for CTA.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 200

The general procedure of depolymerization experiment of cellulose was as

follows. To a suspension comprised of 0.5 g of cellulose and 21 g of acetic acid was
added 0.4 g of acetic acid containing 0.25 mmol of sulfuric acid and it was retained at
an isothermal condition for a required period. The reaction was terminated by an
addition of sodium acetate dissolved in an acetic acid / water mixture. The degraded
cellulose was recovered by filtration and washed with water, saturated aqueous sodium
hydrogencarbonate and water, successively. The depolymerization experiment for
C T A was carried out at the same conditions with a solution of C T A in acetic acid.
DP of the degraded cellulose was evaluated by means of viscometry after it was
acetylated by the following procedure. A degraded cellulose sample was washed with
dimethyl acetamide by using a sintered glass filter to remove water. It was dissolved in
e
a lithium chloride / dimethyl acetamide mixture at 100 C for 3 hours then acetylated at
e
60 C for 5 hours by adding pyridine and acetic anhydride to the solution. C T A as the
reaction product was recovered by pouring the reaction mixture into an excess amount
of water, then it was washed carefully with water several times and dried in vacuo.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch014

The limiting viscosity number ([η]) of the resulting C T A in dimethyl acetamide was
determined by a usual extrapolation method. The viscosity average DP was calculated
from [η] by the equation established by Kamide et. al. (7). The D P was assumed to
W

be the same as the viscosity average DP.

Literature Cited

1. Malm, C. J.; Hiatt, D. G. Cellulose and Cellulose Derivatives ; Interscience

Publishers, 1954; 763 - 824.
2. Malm, C. J.; Barkey, K. T.; Lefferts, E. B.; Gielow, R. T. Ind. Eng.
Chem. 1958, 50, 103.
3. Frith, W. C. Tappi 1963, 46, 739.
4. af Ekenstam, A. Bericht 1936, 69, 553.
5. Saito, O. Statistical Properties of Polymers; Publisher of Chuoh University,
1992; 149 - 158.
6. Malm, C. J.; Barkey, K. T.; May, D. C.; Lefferts, Ε., B. Ind. Eng.
August 31, 2012 | http://pubs.acs.org

Chem. 1952, 44, 2904.

7. Kamide, K.; Miyazaki, Y.; Abe, T. Polym. J. 1979, 11, 523.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 15

Progress in the Enzymatic Hydrolysis of Cellulose

Derivatives

B. Saake, St. Horner, and J. Puls

Institute of Wood Chemistry and Chemical Technology of Wood Federal Research

Centre of Forestry and Forest Products, 21031 Hamburg, Germany
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

Single enzymes of the cellulase complex are efficent tools for

understanding the biodegradation of cellulose derivatives. Sensitive size
exclusion chromatography makes it possible to detect minor
fragmentations. The accessibility of cellulose derivatives is clearly a
function of the degree of substitution (DS). In addition, charge and size
of the substituents and the substituent distribution play major roles for
enzymatic attack. Whereas carboxymethyl cellulose of a certain DS was
fully resistant against endoglucanase action, a methylcellulose of the
same DS was markedly fragmented. Cellulose acetate was taken as an
example for demonstration of the impact of esterases besides cellulases
in biodegradation. The presence of acetyl esterase enabled the
August 31, 2012 | http://pubs.acs.org

endoglucanase to degrade cellulose acetate much faster than when this

type of enzyme was absent.

There is a continuous interest in the enzymatic hydrolysis of cellulose derivatives,

starting with the pioneering work of Husemann (7) and Reese (2). Whereas Husemann
was mainly interested in the location of the hydrolysis, Reese's main focus was to
prevent a microbial attack to cellulose. One of the means to prevent an enzymatic
degradation was identified to consist of an exchange of the free O H groups of the
anhydroglucose units by bulky substituents. In contrast to the early research, today's
concern is at least partly the conservation of the biodegradability of cellulose
derivatives, i.e., cellulose should be substituted by certain derivatives only to the extent
that the biological attack is not impaired. However, early papers as well as recent
publications discuss the action of cellulases against cellulose derivatives in order to
draw conclusions about the location of the substituents within the anhydroglucose
units, as well as about the substituent distribution along the cellulose chain.
Unfortunately, the early work was exclusively performed using enzyme cocktails of
different compositions and origins; for this reason this overview was supplemented by
additional practical experiments conducted in the authors' laboratory. In these
experiments a mono-component cellulase, namely a fungal endoglucanase. was used.

©1998 American Chemical Society 201

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 202

Analytical Methods for Measuring the Enzymatic Degradation of Cellulose

Derivatives. The analytical methods mainly used to follow the enzymatic hydrolysis of
water-soluble cellulose derivatives are based on measuring changes in viscosity, the
amounts of glucose, and reducing sugars. From these data only indirect conclusions
about the length of cleaved chains and the relative amount of derivatized cellulose
fragments can be drawn.
In more recent investigations the enzymatic action was visualized by changes in
the molecular weight distribution, as revealed by size-exclusion chromatography
(SEC). The main problem with SEC-characterization of polyelectrolytes such as C M C
and the separation of their enzymatic degradation products was the selection of an
appropriate eluent. The nature of the buffer and the ionic strength were essential
factors in avoiding unwanted interaction with the stationary phase material (3). In
order to differentiate the DS, ^ C - N M R spectroscopy was used (4).
Demeester et al. could demonstrate the usefulness of light scattering for the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

detection of the enzymatic action (5). The cellulolytic degradation of cellulose was also
considered as a sample pretreatment prior to the structural characterization of cellulose
derivatives (6). This method, however, could not be applied for high DS samples.

The Cellulose Degrading Enzyme System. The complete hydrolysis of natural

cellulose demands the action of exoglucanases (also called cellobiohydrolases, CBH),
endoglucanases (EG), and β-glucosidases. Only a few microorganisms produce the
complete set of cellulases for efficient degradation of insoluble cellulose. The general
term cellulase usually refers to this complete set of cellulolytic enzymes. According to
the current hypothesis, EGs initialize the attack on cellulose by randomly hydrolyzing
internal bonds in amorphous regions of the cellulose. The action of E G thus produces
new chain ends, which then become available for CBHs, which are believed to
hydrolyze the chains from both the reducing (CBH I) and the non-reducing end (CBH
August 31, 2012 | http://pubs.acs.org

II), whereby cellobiose is liberated (7,8). Finally, β-glucosidases hydrolyze cellobiose

into glucose. The different cellulases have been shown to hydrolyze cellulose
synergistically.
The proposed degradation pathway, as well as the synergy, has been the subject of
extensive discussion during recent years. Thus the present classification of specific
cellulases as being either C B H or E G seems dubious. Today cellulases are classified for
their amino acid sequence, followed by a classification by means of hydrophobic cluster
analysis, HCA, in addition to classification according to their action against specific
substrates (9). The H C A considers the different charges of amino acids and indentifies
hydrophobic clusters along the two-dimensional sequences. Based on this information a
three-dimensional structure can be predicted. This method is especially useful for the
detection of similar folds in different enzymes with low sequence identity. Of the 11
cellulase families currently classified (10), fungal cellulases are found in six.
The only true exoglucanase structures solved are the two CBHs of the fungus
Trichoderma reesei. The three-dimensional structures of the catalytic domain of T.
reesei C B H II (family 6) and C B H I (family 7) revealed that the active site of both
enzymes was identified to be situated in a tunnel ranging through the whole domain
and formed by stable surface loops (11,12). It therefore seems that the active site
tunnel is a general feature of exoglucanases. Indeed, exoglucanases are believed to be
unable to attack cellulose derivatives, and the existence of the active site inside a tunnel
could explain why especially bulky cellulose derivatives could get stuck inside the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 203

tunnel. Thus, EGs should have similar overall folds, but the active site should be more
open, allowing a random hydrolysis of the cellulose (77). The three-dimensional
structure of the Thermomonospora fusca E G E2 (75), belonging to the same family 6
as C B H II of T. reesei, confirmed this hypothesis. A similar structure was found for
E G I (family 7) from Humicoia insolens (14). E G V from Humicola insolens belongs
to family 45 and cleaves 1,4-B-glucosidic linkages with inversion of configuration,
whereas E G I catalyses cleavages with retention of configuration. As a general feature,
the active sites of endoglucanases were found to be located in long open grooves.
The structure of E G V has been solved, and it was found that this protein has 7
subsites (A to G) for binding with its substrate, the cleavage taking place between
subsites D and E. On the basis of mutation experiments of the active site of E G V ,
there was strong evidence that an aspartate near subsite Ε, sitting in a predominantly
hydrophobic environment, acts as the proton donor in the hydrolysis mechanism,
whereas another aspartate functions as the base, activating the nucleophile.The active
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

site residues of the H. insolens endoglucanase I are two glutamates and an aspartate,
the latter functioning as the proton donor (14). This enzyme has the negative charge of
its active site in common with E G V of Humicola insolens. In all cases investigated so
far, the major role of the carboxylic amino acids glutamate and aspartate for scisson of
the polysaccharide has been identified (75).

Accessibility of Cellulose Derivatives to Enzymes. It is well known that the

accessibility of cellulose to enzymatic hydrolysis is dependent upon physical and
structural features of the polymer. The relevant features are crystallinity, degree of
swelling, solubility, and presence of other structural components (lignin and
hemicelluloses). Cellulose derivatives are prepared by exchanging hydrogen atoms on
primary and secondary hydroxyl groups with various functional groups such as methyl,
carboxymethyl, diethylaminoethyl. Modification of cellulose usually makes it non­
August 31, 2012 | http://pubs.acs.org

crystalline and in many cases soluble. Both factors increase the susceptibility of the
cellulose to enzymatic hydrolysis. The susceptibility of cellulose derivatives to
enzymatic hydrolysis increases as the substrate becomes less crystalline and more
water-soluble. Water-solubility is dependent upon the degree of substitution (DS). The
DS value at which complete water-solubility is achieved ranges from DS 0.4 to 0.7,
depending on the solvation capacity and on the pattern of substitution (16). However,
when on the average more than one substituent occurs per each anhydroglucose unit,
the enzymatic hydrolysis rate decreases, and higher DS values result in a complete
inertness of the polymer (16).
When the degree of substitution approaches an average of one substituent per
glucose unit, steric factors become important, and enzymatic hydrolysis is retarded.
The International Union of Pure and Applied Chemistry has published
recommendations on the measurement of endoglucanase activities. This commission
recommended substrates such as carboxymethyl cellulose (CMC) of DS 0.7 and
hydroxyethyl cellulose (HEC) of DS 0.9 - 1.0 (77). In connection with the use of
C M C as a test substrate for endoglucanase activity measurement, it has been stated
that besides the DS, the same origin (i.e., the same method for the preparation of the
derivative) is absolutely necessary in order to ensure a reproducible distribution of the
substituents (18).
Reasons why HEC is preferable to C M C or to other water-soluble substrates have
been discussed (19). According to earlier reports (20) two or more contiguous non-

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 204

substituted anhydroglucose units in C M C are necessary for an enzymatic scission of the

polysaccharide chain. In contrast Ach claims that cellulose acetate of DS 2.5 is
biodegradable, although degradation proceeds extremely slowly (27). For these reasons
it seems advisable to differentiate the accessibility of cellulose ethers and cellulose
esters.

Enzymatic Degradation of Insoluble Cellulose Derivatives. Only a few papers deal

with the heterogeneous enzymatic hydrolysis of cellulose derivatives. Highly
substituted and water-insoluble cellulose derivatives (i.e., cellulose triacetate) are
regarded to be completely resistant to enzymatic attack, because of a combination of
factors - a lack of hydrophilicity, reduced swelling, and the hindrance caused by spatial
substituents (22).

Enzymatic Degradation of Cellulose Ethers. Most of the work published on the

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

enzymatic degradation of cellulose derivatives was performed with HEC and C M C (20,
23-26). Some results have also been obtained with M C (1). From Husemann, it may be
assumed that the small methyl substituents inhibit degradation to a lesser extent than do
the more bulky carboxymethyl groups (7). Intensive studies on the enzymatic
degradation of cellulose ethers have established that the degree and the uniformity of
substitution are decisive factors for their accessibility (2, 20, 26, 27).
According to Reese (2), a scission of monosubstituted M C is possible. This finding
was confirmed by Schuseil (28), whereas C M C requires a sequence of three non-
substituted anhydroglucose units (24). Wirick (20, 26) came to the conclusion that
glycosidic bonds within cellulose ethers are only enzymatically cleaved in the case that
two neighbouring non-substituted anhydroglucose units or that eventually a non-
substituted and a C-6 substituted anhydroglucose unit are present. This result is in
accordance with Bhattacharjee and Perlin (25), who pointed out that a scission
August 31, 2012 | http://pubs.acs.org

between non-substituted units as well as between a substituted and a non-substituted

anhydroglucose unit seems possible depending on the position of the substituent.
According to Kasulke et al. (29), three non-substituted glucose units were
required in order to achieve the liberation of glucose. In this case, both hydrolysis
products carried a non-substituted anhydroglucose unit at the end.

Enzymatic Degradation of Cellulose Esters. In the hydrolysis of water-soluble

cellulose acetates Kamide et al. (30) found this polymer to be degradable up to DS 1.
The authors observed a gradual precipitation of high DS cellulose acetate in the course
of enzyme incubation, indicating that short highly-substituted cellulose acetate blocks
can exist in water-soluble cellulose acetate chains (4). According to Buchanan et al.
(57), the biodegradable DS range of cellulose acetates is even wider. The authors
found cellulose acetates with DS values between 1.7 and 2.5 to biodegrade by mixed
culture systems, at least partly following deacetylation. A series of other esters besides
acetates was synthesized by Glasser et al. (32). The biodegradability of cellulose esters
using cellulolytic enzymes was found to depend on two factors: degree of substitution
and substituent size. The cellulose esters had acyl substituents ranging in size from
propionyl to myristyl and DS values between 0.1 and nearly 3.0. The maximum degree
of acylation, that the enzyme could tolerate before the polymer became undegradable
and that resulted in degradation in excess of 10 % by weight, ranged from DS 0.5 to at
least 1.0, depending on the ester type. It became evident that the larger the substituent

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 205

of the cellulose ester, the harder it became for the enzyme to recognize the
macromolecule and degrade it.

Influence of Substituents and Degree of Substitution. A set of C M C of DS,

ranging from 0.6 to 2.1 and different molecular weights were incubated with three
cellulase preparations of different origin and purity (33). A decrease in enzymatic
hydrolysis with increasing DS was observed for all enzyme activities. This decrease
could be explained by the difficult access of the enzyme to the 1,4-B-glycosidic linkage
due to the steric hindrance from the carboxyl groups, located mainly at C-2. The action
of cellulase enzyme preparations on C M C samples of DS 0.7 and varying M W ranging
6
from 0.55 to 8.71 χ 10 g/mol was not affected by the large difference in molecular
size. Complete hydrolysis was obtained with a purified cellulase at hydrolyzing CMCs
with DS up to and including 0.9. Philipp et al. (18) reported on two different phases in
the enzymatic degradation of C M C , both of them first-order reactions. The first (quick)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

reaction was explained by the combined action of endo- and exo-enzymes, which was
favoured by the presence of longer sequences of non-substituted anhydroglucose units.
The second (slower) phase was explained by sporadic EG-mediated "hits", which were
made possible by shorter (possibly two) non-substituted units.

Influence of Derivatization Procedure. Cellulose derivatives are obtained mainly by

heterogenous derivatization, which is normally considered to yield a statistic
distribution of substituents. Cellulase hydrolysis yields large fragments from the more
highly substituted regions, whereas small fragments are obtained from those regions
which had a lower degree of substitution (16). Philipp et al. (34) could clearly show
that the enzymatic hydrolysis of C M C in the DS range of 0.5 to 0.8 depends not only
on the DS, but also on the procedure applied for preparing samples, with different
distributions of substituents along and between the polymer chain. A C M C sample,
August 31, 2012 | http://pubs.acs.org

which has been prepared after solubilization of the cellulose, excels by low values for
liberated glucose in comparison to a C M C sample prepared in a heterogenous reaction
system. Beyond a DS of 0.7 this value seems to be the determining factor for the
liberation of reducing sugars viz. glucose (in case sufficient β-glucosidase activity is
present). In this higher DS range factors derived from the procedure applied for the
production of the derivatives seem to be of minor importance. Philipp et al. (34) draw
the conclusion that longer sequences of non-splitable linkages were present in higher
DS CMCs. For the DS region 0.5 to 0.7, average sequences of derivatized
anhydroglucose units of 4 to 6 were given, whereas in the DS region >1 averages of
up to 40 carboxymethyl-glucose blocks were calculated.
Recently Heinze at al. stated that C M C samples, which were synthesized via an
induced phase separation, contained a significantly higher amount of both
tricarboxymethylated and unsubstituted units than those obtained in a slurry of
cellulose in isopropanol/water at comparable DS values (35). This finding pointed out
that it is not adequate to automatically assume a homogeneous substituent distribution
from a homogenous reaction mixture nor a heterogenous distribution from a
heterogenous system. Since the substituent distribution along the polymer chain is
difficult to access, detailed insight on its impact on enzymatic degradation is still
missing.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 206

Influence of the Position of the Substituents. Little is known about a possible

toleration of endoglucanase activity for a certain position carrying a substituent. The
three possible positions for derivatization in cellulose are the hydroxyl groups at C-2,
C-3, and C-6. The experimentally determined molar distribution of carboxymethyl
groups in C M C has been reported to be 2:1:2.5 for C-2, C-3, and C-6 hydroxyls (36).
While the C-6 primary hydroxyl is generally thought to be the most reactive in this
particular system, this is not always so. There is considerable evidence that the C-2
hydroxyl group is the most acidic in cellulose (37). As a result, many equilibria and
rate-controlled reactions that involve cellulosic alkoxide ions appear to favour this site.
N M R studies of substituent distribution indicate C-2 > C-6 > C-3 (6). The finding that
reducing end residues liberated during enzymatic hydrolysis of cellulose were never
substituted at the 2-position (6) is a strong indication that this position seems to be
needed for binding of the enzyme protein to the substrate.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

The Influence of the Type of Substituent. The susceptibility of cellulose derivatives

to cellulose hydrolysis is also dependent on the type of the substituent (25). The
derivatization will certainly modify the charge of the polysaccharide and thus influence
the interaction of the enzyme's active site with its substrate. Wirick (20, 26)
investigated the enzymatic hydrolysis of C M C , M C , H E C , and hydroxypropyl
cellulose. In a comparison of the above-mentioned cellulose derivatives, and in
accordance with Husemann, C M C was the most resistant substrate.
Interestingly, Philipp and Stscherbina (22) reported a shift in the pH-optimum
from pH 4 (DS 0.7) to pH 7.5 (DS 1.0) in the degradation of C M C using the enzyme
system of Pénicillium citrioviride. The nature of C M C to be a polyelectrolyte was also
pointed out by Nicholson and Merritt (38). According to Philipp and Stscherbina (22)
only the reaction velocity was reduced compared to the degradation of non-ionic
cellulose derivatives. The extent of hydrolysis was not impaired by the charge of the
August 31, 2012 | http://pubs.acs.org

substrate.
An increase in the negative charge resulted in a reduction of the hydrolysis rate,
whereas the presence of positive charges markedly increased the enzyme activity (39).
This finding could be verified and explained by recent investigations on the reaction
mechanism (75) and the fine structure of endoglucanases (14).

Experimental

Substrates. All CMCs were gifts of Wolff-Walsrode (Walsrode/ Germany). The M C

samples came from Kalle, a subsidary of Hoechst A G (Wiesbaden / Germany). The DS
0.7 C A was a gift of Hoechst-Celanese (Charlotte, N.C. / USA). The DS 0.9 to DS 2.9
CAs were gifts of Rhône-Poulenc Rhodia (Freiburg / Germany). The DS-values were
determined by the manufacturers.

Enzymes. The mono-component endoglucanase preparation was an experimental

product from a genetically modified Aspergillus strain, obtained from Novo Nordisk
(Bagsvaerd / Denmark) and used after removal of low-molecular weight components by
ultrafiltration. The Aspergillus enzyme mix was obtained from Nagase & Co (Tokyo /
Japan) and used after ammonium sulfate precipitation and desalting.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 207

Enzyme Treatments. 0.2 % solutions of C M C , M C , and the water-soluble DS 0.7

C A in 0.1 M sodium nitrite were incubated with 10000 nkat per mg substrate. The
incubations were performed in an Eppendorf Thermomixer at 45 °C and 1200 rpm for
3 or 6 days. After completion, the samples were boiled for 3 minutes for protein
denaturation. The precipitated enzyme protein was removed by centrifugation at 2000
g for 10 minutes. Substrate blanks were incubated for 3 days and treated
correspondingly. The incubation of the water-soluble DS 0.7 C A with the enzyme mix
was performed as a 1 % solution in bi-destilled water at 40 °C and 1200 rpm The
enzyme dosage was 500 nkat/mg. Enzyme protein was removed according to the
procedure mentioned above. For SEC analysis, samples were diluted to a 0.2 %
solution using sodium nitrite. Liberated acetic acid was determined using the
Boehringer Test Combination Kit (Cat. No. 148 261 Boehringer, Mannheim /
Germany).
For endoglucanase treatment of cellulose acetate higher in DS than 0.7, 1 %
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

solutions or suspensions in bi-destilled water were incubated with 10000 nkat

Aspergillus endoglucanase / mg substrate. The incubation was finalized by ammonium
hydroxyde addition and incubation at 30 °C overnight for protein denaturation and
acetyl saponification. The solutions were freeze-dried and afterwards stored in vacuum
over P 0 . Blanks underwent the same procedure.
2 5

Carbanilation. 100 mg of saponified samples were suspended in 100 ml of pyridine

and derivatized at 80 °C using 7 ml phenylisocyanate. After 48 h the reaction was
stopped. A refinement procedure by direct evaporation of the pyridine (46) was applied
as published previously (40).

SEC of Water-Soluble Derivatives. SEC was performed using sample

concentrations of 0.2 % and sample volumes of 100 μΐ injected into three SEC
August 31, 2012 | http://pubs.acs.org

columns, coupled in line (TSK G5000PWXL, G4000PWXL, G3000PWXL, 300X7.8 mm

each, and a G2500PWXL guard-column, 40x6 mm, TosoHaas, Stuttgart / Germany).
The column temperature was kept constant at 40 °C. The mobile phase (0.4 ml/min)
was 0.1 M sodium nitrite in water. The elution profiles were detected by changes in
refractive index. The WINGPC 3.0 software (Polymer Standard Service, Mainz /
Germany) was used for data aquisition.

SEC of cellulose carbanilates. 0.5 mg cellulose derivative per ml of stabilized

tetrahydrofurane (THF) was shaken for 3 days for dissolution. Samples were then
centrifuged for 2 h at 20 °C (16000 to 22000 g), and the supernatant was used for
further analysis. SEC was performed using a Waters 510 pump, a Kontron 360
3 4
autosarnpler (100 μΐ), Waters Ultrastyragel 10 Â and 10 Â columns (300x7.8 mm
each), a Spectra Physics SP8400 U V detector, a Shodex RI-71 detector, and WINGPC
3.0 software. The eluent was THF with a flow rate of 1 ml/min at 20 °C. Polystyrene
standards were monitored at 254 nm and cellulose tricarbanilates at 235 ran. A
calibration curve was obtained by a broad fit of two cellulose tricarbanilate samples
with a molar mass of 210000 g/mol and 1000000 g/mol (40) to a polystyrene standard
3
curve. The Mark-Houwink constants obtained by this procedure were K: 1.756 +10"
ml/g and a: 0.890.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 208

Results and Discussion

Endoglucanase Degradation of C M C . CMCs in a DS range from 0.6 to 2.4 were

incubated for 3 and 6 days with an overdosage of a mono-component Aspergillus
endoglucanase (10000 nkat/mg substrate), in order to reach the final possible stage of
fragmentation. Substrate blanks were incubated for 3 days. Figure 1 demonstrates the
hydrodynamic volume of the blank samples to be too high for a perfect separation in
the chromatographic system. This is especially the case for the DS 0.6 and DS 0.8
samples, starting at an elution volume of 13.75 ml, with a pronounced steep shoulder.
While most blanks had a narrow elution curve, the DS 1.2 and 1.6 samples showed a
tailing in the lower molar mass region.
The intensity of endoglucanase degradation was strongly dependent on the DS.
DS 0.6 and 0.8 samples gave a strong shift of the elution profile to the low-molecular
region and displayed distinct signals for oligomeric compounds. The fragments which
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

were washed from the columns at the elution volume of around 29.6 ml can be
attributed to monomelic degradation products. It becomes evident from Figure 1 that
the maximal possible fragmentation of the polymer was completed after 3 days of
incubation.
DS 0.9 and 1.2 CMCs were also markedly fragmented by the endoglucanase,
which is visualized by pronounced shifts of the elution profile to longer elution volumes
(Figure 1). However, the degradation of these derivatives was less intense in both
regions, the higher molar mass region and the oligomeric range. In the degradation of
these samples, a prolongation of the reaction time gave rise to an additional
fragmentation. Possibly the prolonged and slower phase can be explained by sporadic
cuts between shorter non-substituted or less substituted regions, according to the
explanation given by Kasulke et al. (27). In addition, tertiary conformation effects of
the charged polymer could limit or slow down the accessibility of the enzyme to
August 31, 2012 | http://pubs.acs.org

positions which principally can be cleaved.

The DS 1.6 sample shows only a minor shift in the elution curve and a small
bimodality in the lower molar mass range after endoglucanase treatment. For the DS
2.4 C M C sample, the effect of the treatment on the elution curve was even more
reduced. However, a small but significant bimodality demonstrates the availability of
some positions along the C M C chain to the endoglucanase. This minor change had no
effect on the elution curve of the main polymer peak. For both the DS 1.6 and DS 2.4
samples only minor changes occurred even after prolonged incubation times. These
results are in accordance with Philipp et al. (34), who could demonstrate a reduction in
viscosity for DS 1.7 C M C after incubation with a Gliocladium culture filtrate. Due to
the close involvement of the carboxylates of two amino acids in its active site, the
reduced activity of endoglucanase against the negatively charged C M C becomes
explainable.
It is noteworthy that the intensity of the elution curve of the enzyme treated DS
2.4 sample was slightly higher than the substrate blank. It could be excluded that this
phenomenom was derived from experimental errors. However, it was noted that the
number of visible gel particles in enzyme-incubated samples was significantly reduced;
this effect explains the higher yield of enzyme-treated samples. It was even more
pronounced for the treatment of both M C and water soluble C A , discussed in the
following paragraphs.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 209
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015
August 31, 2012 | http://pubs.acs.org

12 17 22 27 32 12 17 22 27 32

Elution volume [ml] Elution volume [ml]

Figure 1. Influence of DS on the fragmentation of carboxymethyl-cellulose
by Aspergillus endoglucanase, after 3 days and 6 days incubation, monitored
by aqueous SEC.
blank: 3 days: 6 days:

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 210

Endoglucanase Degradation of MC. It could be verified that endoglucanase

fragmentation of M C in the range from DS 0.5 to 2.1 was much more effective then
that of C M C . The DS 0.5 sample was more or less completely degraded into
oligomeric products (profile not shown). After 3 days of degradation, the
fragmentation of the DS 1.5 M C in Figure 2 was comparable to the DS 0.6 C M C in
Figure 1. The M C curve had an even more intensive shift to low molar mass and
oligomeric products. At 31.2 to 31.3 ml, the elution curve falls almost rapidly, since
monomelic M C products elute into the separation limits of the column.
In another comparison, the DS 2.1 M C was significantly fragmented, whereas the
elution curve of the DS 2.0 C M C was only slightly modified, compared to the substrate
blank.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

12 17 22 27 32 12 17 22 27 32
Elution volume [ml] Elution volume [ml]
August 31, 2012 | http://pubs.acs.org

Figure 2. Influence of DS on thefragmentationof methyl cellulose by

Aspergillus endoglucanase (3 days incubated) monitored by aqueous SEC.
blank: enzyme treated:

All endoglucanase-treated and centrifiiged M C samples have in common that the

intensities of the elution curves were by far higher than the substrate blanks. This
phenomenon underlines the possibility of endoglucanase-mediated improvements in
solubility.

Endoglucanase Degradation of Cellulose Acetate. Cellulose acetate fragmentation

was performed with DS 0.7 to DS 2.9 samples, which included heterogeneous
hydrolysis of water-insoluble cellulose acetate powder of DS > 1 and homogeneous
hydrolysis of water-soluble samples DS < 1. The substrates have been prepared by acid
saponification of cellulose triacetate. This is the reason that the molecular weight of the
starting material was reduced with decreasing DS (Table I). Due to the heterogeneous
hydrolysis of most of the cellulose acetate substrates, the material could not be directly
analyzed by gel permeation chromatography, but had to be gently saponified using
ammonium hydroxide prior to tricarbanilation and SEC analysis in tetrahydrofurane
(40).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 211

Table I. Homogeneous and heterogeneous endoglucanase hydrolysis of cellulose

acetate samples as revealed by SEC of the corresponding carbanilates

DS Starting material Endoglucanase treated material

M w DP M w
DP
0.9 16 000 1.5 31 1 860 1.1 4
1.2 44 000 1.8 85 2 500 1.2 5
1.6 72 000 1.7 138 14 000 5.1 27
1.7 48 000 1.7 92 26 000 1.2 50
1.9 98 000 2.0 189 52 000 11.0 100
2.5 164 000 3.2 316 159 000 3.0 306
2.9 201 000 4.1 387 204 000 3.3 394
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

Due to the surplus of endoglucanase activity used in this study, the DS 0.9 sample
was extensively degraded. Caused by acid saponification prior to the hydrolysis step,
the partly deacetylated material was already considerably reduced in degree of
polymerization (DP = 31); the endoglucanase treatment, however, caused a further
degradation by homogeneous hydrolysis to DP = 4. A similar result was obtained for
the DS 1.2 sample, which was degraded from molecular weight ( M ) 44000 to Mw w

2500, equivalent to an average DP of 5.5, indicating a reasonable accessibility of both

cellulose acetate samples to endoglucanase action. The accessibility was a clear
function of the DS. From DS 1.6 on, the degradation by endoglucanase action was
considerably retarded, which was certainly caused by two facts: the water-insolubility
of the material and the shielding by increasing amounts of acetyl substituents. In spite
of the unfavourable conditions, the material was still markedly degraded up to a DS of
1.9 (from M 98000 to M 52000). Cellulose acetate of DS 2.5, which is normally
w w

used for technical applications, was more or less resistant.

August 31, 2012 | http://pubs.acs.org

On first glance, these results might be not fully consistent with the results given by
Buchanan et al. (57), who found cellulose acetate up to DS 2.5 to be biodegradable.
However, these authors used a mixed culture system instead of a single, cell-free
enzyme. It was noted in their investigations that cellulose acetate degradation was
followed by deacetylation. In our own experiments, no acetyl was released by the
Aspergillus endoglucanase.

Acetyl Esterase Involvement in Cellulose Acetate Fragmentation

In the search for suitable enzyme preparations, it was found that most commercial
ceilulase preparations had the capability of releasing varying amounts of acetyl groups.
Due to the fact that the same endoglucanase dosage (tested with HEC as substrate)
was used, it could be concluded that acetyl release is not a common feature of
endoglucanase activity, but must be deduced to a separate enzyme. It is known from
earlier studies that acetyl xylan esterase (41) and acetyl mannan esterase (42) are
common features in hemicellulolytic enzyme systems. From more detailed work it is
known that some esterases are highly specific, whereas others are not (43). Due to the
man-made origin of cellulose acetate, nature will not have provided microorganisms
with specific cellulose acetate esterases, but will probably have provided non-specific

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 212
esterases capable of deacetylating cellulose as well as other polymers such as pectin
and hemicelluloses.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

20 24 28 32 29 30 31 32
Elution volume [ml] Elution volume [ml]

Figure 3. Elution profiles of cellulose-acetate (DS 0.7) monitored by aqueous SEC,

fragmented by Aspergillus endoglucanase (a) and an Aspergillus enzyme mix (b).

In order to learn more about this phenomenon, a DS 0.7 cellulose acetate sample
was comparatively incubated with the mono-component Aspergillus endoglucanase
and another commercial Aspergillus enzyme mix (Celluzyme), which had previously
been used in the authors' laboratory as a source for the isolation of mannanase activity
(44) and acetyl mannan esterase activity (42). The incubated samples were directly
August 31, 2012 | http://pubs.acs.org

analyzed by aqueous SEC. The time course of the degradation is illustrated by the
elution curves from 0 to 72 hours incubation (Figure 3 a and b). The improving RI
intensity, in conjunction with the increasing degradation rate, especially as compared to
the substrate blank (0 h) with the sample after 1 hour incubation, can be explained by
the improved water-solubility of the material. The shift to smaller fragments equivalent
with longer elution times is particularly obvious within the first 24 hours of incubation.
There was almost no change between 48 hours and 72 hours of incubation. The
substrate blank was not included in Figure 3b in order to allow an improved resolution
of the low molecular weight fragment (note the different ml-scale in Figures 3a and
3b). Indeed the polymeric material was already drastically reduced in chain length after
one hour of incubation. After 24 hours only minor additional changes occurred.
This difference in speed and extent of degradation has been made possible by the
presence of an acetyl esterase, in addition to the endoglucanase activity.
Unambiguously, the presence of this enzyme was established by the release of acetic
acid into solution (Figure 4). The speed of acetic acid liberation was going hand in
hand with the fragmentation of the polysaccharide and became slower after 24 hours of
incubation. About 50 % of the total acetyl-content of the material was cleaved off after
72 hours. There could be various reasons for the limited release of acetic acid: 1. The
acetyl esterase could be a specific enzyme acting exclusively on intact polysaccharides.
With increased endoglucanase action, the enzyme could have lost its specificity due to

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 213

the presence of shorter fragments. 2. The acetyl esterase could be restricted in its
catalytic capacity to certain positions within the anhydroglucose unit. 3. The enzyme
could be restricted in its action due to the presence of a non-accessible, highly
substituted region within the cellulose chain.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

0 > Γ
0 20 40 60 80

Time [h]

Figure 4. Time course of acetic acid release from cellulose acetate

(DS 0.7) during incubation with the Aspergillus enzyme mix.

There is some probability for the third possibility, because no acetic acid was
released from the DS 2.5 cellulose acetate, incubated with the enzyme mix. However,
a similar phenomenon was found for the acetyl xylan esterase, which acted
August 31, 2012 | http://pubs.acs.org

preferentially on acetylated xylan and not on xylan fragments (45). The question of the
existence of a specific acetyl cellulose esterase or a rather unspecific acetyl esterase
could not be solved in this study, but will be the subject of further investigation.

Conclusion

The accessibility of cellulose derivatives clearly was a function of the degree of

substitution (DS), in that the material became less degradable with increasing DS. In
addition to this factor, charge and size of the substituents play major roles for
enzymatic attack. Whereas C M C of DS 1.6 was almost resistant against
endoglucanase action, a M C of DS 2.1 was markedly fragmented into shorter chains.
Although low DS cellulose ethers were quite markedly degraded by the mono-
component endoglucanase, they could not be completely converted into very short
fragments. This is a strong indication for the occurrence of non-degradable, highly
substituted regions. Consequently, the allocation of the substituents within the
anhydroglucose unit and along the cellulose molecule are additional impacts on the
availability of the polysaccharide chain for enzymatic fragmentation.
Cellulose acetate was more degradable than could be anticipated from the
literature. Cellulose acetate was taken to demonstrate of the impact of esterases,
besides endoglucanases, in biodégradation. The presence of acetyl esterase enabled the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 214

endoglucanase to degrade cellulose acetate much faster and intensively, and with less
enzyme protein. Until now this class of enzymes has been neglected in the
consideration of biodegradability of cellulose esters. The specificity of this enzyme is
not yet understood and will take our attention in the near future.
The comparison of different authors' work was complicated by the fact that most
researchers have drawn their conclusions from viscosity measurements as well as from
the determination of nonsubstituted anhydroglucose values and reducing sugars, using
culture filtrates containing different ratios of endoglucanases, cellobiohydrolases, β-
glucosidases, and possibly acetyl esterases. It is clear that these catalysts contribute to
a different extent of fragmentation, leading to the increase in reducing sugars and
liberated anhydroglucose from cellulose derivatives.

Acknowledgment:
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

This study was supported by DFG-project Pu 81/5-1 and EU-project BI02-CT94-

3030. We thank the suppliers of enzymes (NOVO-NORDISK A/S and N A G A S E
CORP.) and cellulose derivatives (HOECHST A G , HOECHST-CELANESE CORP.,
RHONE-POULENC-RHODIA A G , WOLFF-WALSRODE AG) for their generous
support.

Literature Cited

(1) Husemann, E. Das Papier 1954, 8, 157-162.

(2) Reese, Ε. T. Ind. Eng. Chem. 1957, 49, 89-93.
(3) Hamacher, Κ.; Sahm, H. Carbohydr. Polym. 1985, 5, 319-327.
August 31, 2012 | http://pubs.acs.org

(4) Iijima, H.; Kowsaka, K.; Kamide, K. Polym. J. (Tokyo) 1992, 24, 1077-1097.
(5) Demeester,J.;Eigner, W.-D.; Huber, Α.; Glatter, O. J. Wood Chem. Technol.
1988, 8, 135-153.
(6) Parfondry, Α.; Perlin A.S. Carbohydr. Res. 1977, 57, 39-49.
(7) Ståhlberg,J.;Divne, C.; Koivula Α.; Piens K.; Claeyssens,M.;Teeri T. T.;
Jones, T. A. J. Mol. Biol. 1996, 264, 337-349.
(8) Harjunpää, V.; Teleman Α.; Koivula Α.; Ruohonen L.; Teeri T.T.; Teleman O.;
Drakenberg T. Eur. J. Biochem. 1996, 240, 584-591.
(9) Henrissat, B. Biochem. J. 1991, 280, 309-316.
(10) Tomme, P.; Warren, R. A.J.;Miller Jr., R. C; Kilburn, D. G.; Gilkes, N. R. In:
Enzymatic Degradation of Insoluble Carbohydrates; Saddler, J. N., Ed.; ACS:
Washington, DC, 1995, Vol. 618; pp. 142-163.
(11) Rouvinen,J.;Bergfors, T.; Teeri, T.; Knowles, J. K. C.; Jones, T. A. Science
1990, 249, 380-386.
(12) Divne, C.; Ståhlberg,J.;Reinikainen, T.; Ruohonen, L.; Pettersson, G.;
Knowles, J. K. C.; Teeri, T. T.; Jones, T. A. Science 1994, 265, 524-528.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 215

(13) Spezio, M.; Wilson, D. Β.; Karplus, P. A. Biochemistry 1993, 32, 9906-9916.
(14) Davies, G.J.;Schülein, M. In: Carbohydrate Bioengineering (Progress in
Biotechnology, Vol. 10); Peterson, S. B.; Svensson, B.; Pedersen, S., Eds.;
Elsevier: Amsterdam, Netherlands 1995, pp. 225-237.
(15) Withers, S. G. In: Carbohydrate Bioengineering (Progress in Biotechnology,
Vol. 10); Peterson, S. B.; Svensson, B.; Pedersen, S., Eds.; Elsevier:
Amsterdam, Netherlands, 1995, pp. 97-124.
(16) Focher, B.; Marzetti, Α.; Beltrame, P. L.; Carniti, P. In: Biosynthesis and
Biodegradation of Cellulose; Haigler, C.H., Ed.; Marcel Dekker Inc.: New
York, NY, 1991, pp. 293-310.
(17) Ghose, T. K. Pure & Appl. Chem. 1987, 59, 257-268.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

(18) Philipp, B.; Kasulke U.; Lukanoff B.; Jacopian V.; Polter, E. Acta Polymerica
1982, 33, 714-718.
(19) Deemeester,J.;Bracke M.; Lauwers A. In: Hydrolysis of Cellulose: Mechanisms
of Enzymatic and Acid Catalysis; Brown Jr., R. D.; Jurasek, L., Eds.; ACS:
Washington, DC, 1979, Vol. 181; pp. 91-125.
(20) Wirick, M. G. J. Polymer Sci.: Part A-1 1968, 6, 1965-1974.
(21) Ach, A. J. Macromol. Sci., Pure Appl. Chem. 1993, A30, 733-740.
(22) Philipp, B.; Stscherbina, D. Das Papier 1992, 46, 710-722.
(23) Almin, Κ. E.; Eriksson K.-E. Arch. Biochem. Biophys. 1968, 124, 129-134.
(24) Eriksson, K.-E.; Hollmark B.H. Arch. Biochem. Biophys. 1969, 133, 233-237.
August 31, 2012 | http://pubs.acs.org

(25) Bhattacharjee, S. S.; Perlin A. S. J. Polymer Sci.: Part C 1971, 36, 509-521.
(26) Wirick, M. G. J. Polymer Sci.: Part A-1, 1968, 6, 1705-1718.
(27) Kasulke, U.; Dautzenberg, H.; Polter, E.; Philipp, B. Cell. Chem. Technol.
1983, 17, 423-432.
(28) Schuseil J. Charakterisierung von Methylcellulosen: Verteilung der
Substituenten innerhalb der monomeren Einheiten und entlang der
Polymerketten; Thesis; University of Hamburg, Germany, 1988, 75 pp..
(29) Kasulke, U.; Linow K.-J.; Philipp B.; Dautzenberg H. Acta Polymerica 1988,
39, 127-130.
(30) Kamide, K.; Iijima, H.; Kowsaka, K. In: Cellul. Sources Exploit; Kennedy, J. F.;
Phillips, G. O.; Williams, P. Α., Eds.; Ellis Horwood, Chichester, UK, 1990, pp.
365-370.
(31) Buchanan, C. M.; Gardner, R. M.; Komarek, R.J. J. Appl. Polym. Sci. 1993,
47, 1709-1719.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 216

(32) Glasser, W. G.; McCartney, Β. K.; Samaranayake, G. Biotechnol. Prog. 1994,

10, 214-219.
(33) Melo, Ε. H. M.; Kennedy, J. F. Carbohydr. Polym. 1993, 22, 233-237.
(34) Philipp, B.; Kasulke, U.; Dautzenberg, H.; Polter, E.; Hubert, S. Acta
Polymerica 1983, 34, 651-656.
(35) Heinze, T.; Erler, U.; Nehls, I.; Klemm, D. Angew. Makromol. Chem. 1994,
215, 93-106.
(36) Croon, I.; Purves, C.B. Svensk Papperstidn. 1959, 62, 876-882.
(37) Lenz, R.W. J. Amer. Chem. Soc. 1960, 82, 182.
(38) Nicholson, M. D.; Merritt, F.M. In: Cellulose Chemistry and its Applications;
Zeroninan, S. H.; Nevell, T. P., Eds.; Ellis Horwood: Chichester, UK, 1985, pp.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch015

363-383.
(39) Boyer, R. F.; Redmond, M. A. Biotechnol.Bioeng.1983, 25, 1311-1319.
(40) Saake, B.; Patt, R.; Puls,J.;Linow, K.J.;Philipp, B. Makromol. Chem.,
Macromol. Symp. 1992, 61, 219-238.
(41) Biely, P.; Puls,J.;Schneider, H. FEBS Lett. 1985, 186, 80-84.
(42) Puls,J.;Schorn, B.; Schuseil, J. In: Biotechnology in Pulp and Paper
Manufacture; Kuwahara, M.; Shimada, M., Eds.; Uni Publishers Co.: Tokyo,
Japan, 1992, Vol. 57; pp. 357-363.
(43) Tenkanen, M.; Schuseil,J.;Puls,J.;Poutanen, K. J. Biotechnol. 1991, 18, 69-
84.
August 31, 2012 | http://pubs.acs.org

(44) Yamazaki, N.; Dietrichs, H. H. Holzforschung 1979, 33, 36-42.

(45) Tenkanen, M.; Poutanen, K. In: Xylans and Xylanases (Progress in
Biotechnology Vol. 7); VisserJ.;Beldman, G.; Kusters-van Someren, Μ. Α.;
Voragen, A. G. J., Eds.; Elsevier: Amsterdam, Netherlands, 1992, pp. 203-212.
(46) Wood, B. F.; Conner, A. H.; Hill Jr., C. G. J. Appl. Polymer Sci. 1986, 32,
3703-3712.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 16

Evidence of Supramolecular Structures of Cellulose

Derivatives in Solution

1 1,3 2
Liane Schulz , Walther Burchard , and Reinhard Dönges

1
Institute of Macromolecular Chemistry, University of Freiburg, D-79104
Freiburg, Germany
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

2
Hoechst AG/Kalle-Albert, Forschung Alkylose, D-65174 Wiesbaden, Germany

Recent studies on the solution properties of cellulose derivatives

disclosed that two classes have to be distinguished, i.e. the fully and
partially substituted chains. Molecularly dispersed solutions are
obtained when all OH groups are substituted, while in the other case
free OH groups can undergo hydrogen-bonding resulting in association
or aggregation. The measurements gave strong evidence for a non
random aggregation that might result from uneven derivatization along
the chain leading to blocks of low-substituted chain segments. A
fringed micellar structure is concluded from the global properties of the
aggregates. The model is confirmed and further specified by the
August 31, 2012 | http://pubs.acs.org

angular dependence of the scattered light. The fringed micelle consists

of a hard core of laterally aligned segments from which flexible chains
emerge. The anisotropic hard core was confirmed by TEM and flow
birefringence measurements. The flexibility of the dangling chains
could be detected by dynamic light scattering. However, the flexibility
became strongly reduced when the concentration was increased beyond
the overlap concentration. This behavior resultsfromthe fact that the
dangling chains cannot penetrate the hard core of thefringedmicelle.

Cellulose derivatives are semisynthetic polymers composed of a natural backbone and

synthetic side groups. As a consequence of their chemical structure they own a broad
range of application comprising chemical, pharmaceutical and food industries, where
they are mainly used as thickeners for solutions. Recent measurements have disclosed
(1-6) that depending on the degree of substitution, two classes of cellulose derivatives
have to be distinguished, e.g. fully and partially substituted chains. The main reason for
such differentiation results from non substituted hydroxyl groups that interact in a very
specific manner by intra- and /fttermolecular hydrogen bonding. In general, specific
Corresponding author.

218 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 219

interaction leads to the formation of aggregates or associates that can be significantly

influenced by the solvent. In contrast to these aggregated structures, molecularly
dispersed polymer solutions can be obtained i f the cellulose backbone is fully
derivatized. Typical examples are cellulose tricarbanilate (CTC) (7), cellulose trinitrate
(CTN) (8) and cellulose in complexing solvents such as cuoxam (CUOXAM-Cell) (?)
which prevents intramolecular hydrogen bonding. In spite of being molecular dispersed
these three systems have special disadvantages. These are:
CTC: Difficulties in the preparation of high molecular weight polymers and reduced
solubility of derivatives with degree of polymerization DP > 6000 based on hydrogen
bonds that have to be broken up cooperatively.
CTN: Tendency to degradation.
CUOXAM-Cellulose: Decreasing solubility for macromolecules with D P > 6000
because of intermolecular crosslinking as a consequence of the complexation process
(9).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

This contribution deals with the solution structure of partially substituted

cellulose derivatives. The aim of this contribution is a presentation of characteristic
features of partially substituted methylhydroxypropyl- or methylhydroxyethyl-
celluloses (MHPC, M H E C ) in water and trifluoroethanol, and of partially substituted
carboxymethyl celluloses (CMC) in 0.1 M NaCl solution (Scheme 1). Surprisingly
polyelectrolytes as well as hydrophobically and hydrophilically modified chains display
very similar behavior. These features are striking and call for an interpretation of the
mechanism of structure formation. The reasons for this behavior will be discussed in
detail. Most interesting was the question whether these aggregates are random in
nature or have already supramolecular characteristics. The present contribution is
essentially a report on the main features and is not considered as a detailed
communication on all experimental details. Parts have been already submitted (1-4,30)
other details will follow.
The paper is divided in three main sections. One is concerned with the global
August 31, 2012 | http://pubs.acs.org

structure, e.g. the hydrodynamic radii and radii of gyration Rg, the intrinsic
viscosities [η] and the second virial coefficients A . Another one deals with the shape
2

of the particles determined by the angular dependence of the scattered light in static
light scattering (SLS) experiments. Finally, in further sections we discuss particu­
larities of the internal or segmental mobility determined by dynamic light scattering
measurements (DLS), flow bfrefringence, rheo-optics and electron microscopy.

Experimental

Materials. The cellulose ethers were laboratory samples from Hoechst-Kalle A G , and
the cellulose acetates were from Rhône-Poulenc. The cellulose tricarbanilates were
prepared in the Freiburg laboratory by treating cellulose in hot pyridin with
phenylisocyanate as described previously (49). The cellulose ethers, ionic and non-
ionic samples, were dialysed for one to two weeks against distilled water and the result
was controlled by conductivity measurements. This was done to assure equilibrium
solution states and to free the materials from included salts, respectively. The samples
were then freeze dried and kept under dry atmosphere until use. Only in a few cases
also the non dialysed samples were chosen. These are indicated by a suffix u. The
cellulose 2.5-acetates were purified by centrifugation of 1% solutions in acetone
followed by precipitation in water and dried in a vacuum oven at about 50°C. The
degree of substitution is given in the legends to the corresponding Figures.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 220
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016
August 31, 2012 | http://pubs.acs.org

< PQ

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
Scheme 1: Schematic representation of the chemical structure of various cellulose
derivatives considered in this contribution: A - M H P C ; Β - M H E C ;
- C M C ; D - CTC.
C
 222

Instrumentation Static light scattering measurements were made with a computer

driven modified SOFICA instrument that was equipped with a new light detection
system (50).The instrument used for combined static and dynamic L S was an A L V
3000 instrument in combination with the A L V correlator/structurator. The set-up was
described previously (51), where also details of the instrument calibration are given.
The digital pick-up of data with both instruments allowed an immediate evaluation of
molecular parameters by common Zimm-, Berry-, Guinier- and Kratky-plots, respec­
tively.
Rheo-optic A common Bohlin constant stress (CS) rheometer was used for
rheological measurements. The instrument was equipped with a home made device for
detection of the birefringences induced by the shear. The recently developed set up
was made by Richtering et al. (30).
TEM One drop of a 0.2% aqueous MHPC Π solution was picked up by a copper grid
and negatve-stained with uranyl acetate as usual.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

Global Properties

The global properties of polymer solutions are defined by parameters which describe
averages over the entire particle including thermodynamic or hydrodynamic
interactions among them A l l these properties are based on three quantities, i.e. radius,
volume and molar mass of the particles. Typical examples are: (i) The two molecular
radii Rg and R , where Rg, the radius of gyration, is defined geometrically, while the
h

hydrodynamic Rj, includes the effect of hydrodynamic interactions which in addition

depend on the segment density (10). (ii) Tlie second virial coefficient A . It expresses
2

the thermodynamic interactions between two polymer coils in dilute solution. Its value
is proportional to the volume of the molecule (~Rg) divided by the square of the molar
mass M (77). (iii) The intrinsic viscosity [η] which is determined by the ratio of the
w

hydrodynamic volumes to their molar mass (77,72). (fv) Another quantity of interest is
August 31, 2012 | http://pubs.acs.org

the ratio of the two radii Rg/R s ρ; this p-parameter allows a rough estimation of the
n

segmental density but depends also on the molecular size non uniformity (73).

Experimental Findings. Partially substituted derivatives have been characterized in

different solvent systems over a broad range of concentration (1,2,6). Because of the
large number of experimental data we confine ourselves to only a few examples which
emphasize the particularities in behavior of the solutions from M H P C , M H E C and
CMC. Figures 1 and 2 give a preliminary impression on the deviations of the cellulose
derivatives from the behavior of common synthetic polymers. Because of the basic
similarity, (the same backbone, fully substituted derivative), CTC has been chosen as
reference in the following comparison. According to their global properties C T C
dissolved in dioxane behaves like a synthetic polymer (7). In particular the exponents
for Rg, Rh, A and [η], obtained from the slopes in Figures 1 to 4, obey the predicted
2

scaling properties (14).

a
R a R h =v
a
A 2
3v-2 (1)
3v-l

Their values are listed in Table I.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 223

4 . MM.) I . . M U . , . . . 1l l l l | I . . Mlll| . . . Mill

1000 Γ partially substituted ~

Ay ;
-
Ε a R =0.22
c

100
01
-
\= · 0 61
Γ/i -

ν MHPC ( h s ) !
fully substituted Δ M H P C (Is) -
• MHEC
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

• CTC
1 I 1 I 1 III 1 1 1 1 1 1 III I
10
10" 10° 10° 10 /
10 a
10 a

-1
M w / g mol

Figure 1 : Rg as a function of particle weight M for three cellulose derivatives, (hs): w

high HP-substituted, (Is): low HP-substituted. (Is): MHPC 1-5; DS = 1.84,

MSi-OCaH^OH) = 0.2; η = 19-30 Pa at (2%); (hs): M H P C I-V;
DS = 2.07, M S ( - O C H O H ) = 0.78; η = 241-96120 Pa at 2% . M E H C :
3 6

DS between (Is) and (hs), (estimation). CTC in dioxan; DS = 3.0.

"ΓΤΤΤΤΤη 1—1 1 I l l l l | 1 I I I llll| 1—ΓΤΤΤΤΤΤ] 1—I I I 1—I 1 1 1 III

August 31, 2012 | http://pubs.acs.org

1000 Γ a R =0.15
: h
I ^ — * — t r 7 F

partially substituted

: -Vfully substituted

I • MHEC :
- JS* QB =0.64 Δ M H P C (Is) :
h
• ν MHPC (hs) *
_ • CTC
ii . 1 11 lllll 1 ' ι •ιm l 1 1 11Hill 1 1 1 1 1 III
ι '
10 3
10 4
10 5
10 6
10 7
10 8
10

M w / gmol

Figure 2 : R as a function of particle weight M for three cellulose derivatives.

h w

Meaning of (hs), (Is) and the other notation as in Figure 1.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 224

ι ι ι nm| ' 1 1 1 III!

10 il » I Ml..,, , , I I,IN, , , Μ..Μ, , . .τ τ τ π ρ - τ I

CM fully substituted
I
10" 3
-·
a A =-0.19 \
Ε —I
ο
4
10"
Δ
mol

:
; α Α =-0.53

5
10"
CN
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

partially substituted [
< • Δ

ίο" 6 il ι ι ι mill ι ι ι mill ι ι ι ι mil ι ι ι ι ι• •• ι *

3 4 5 6 7 8 9
10 10 10 10 10 10 10 10

Figure 3 : Molar mass dependence of the second virial coefficient A for CTC in 2

dioxane (fully substituted) and the partially substituted M H P C and M H E C

ethers.
August 31, 2012 | http://pubs.acs.org

y0=1.92 (Δ)
0)

j. sphere
Ε
σ AV •
D Δ°νν ^
Δ V microgel
Ο­ v

0. 1 I 1—ι ι ι I III! 1 1 I I Mill I I I I I 1111 1 1 I I I 1111 1 1 I I I I III

4 5 6 7 8 9
10 10 10 10 10 10
. -1
M w / g mol
Figure 4: Molar mass dependence of the p-parameter for the same derivatives as
shown in Figure 3.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 225

Table I
Predicted (columns 2 and 3) and experimentally observed exponents in the molar
mass dependencies of the radius of gyration, the hydrodynamic radius, the
second virial coefficient and the intrinsic viscosity for three cellulose derivatives

Exponent good solvent θ-solvent CTC MHPC/MHË CMC

dioxan C water 0.1 MNaCl
a
R g
= v
0.60 0.50 0.61 ±0,01 0.22 ±0.01 0.22 ±0.02
a
R h 0.59 0.50 0.64 ±0.01 0.15 ±0.02 0.86 ±0.03
a
A 2 -0.20 - -0.19 ±0.04 -0.53 ±0.06 -0.68 ±0.07
a
[n] 0.70-0.80 0.50 0.89 ±0.04 not measured —*
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

* almost no molar mass dependence, large scatter of data.

Strikingly, the exponents of the partially substituted cellulose derivatives deviate from
this behavior while those for the fully substituted CTC fulfil these conditions.

Discussion of the Global Properties. Comparison of the fully and partially

substituted derivatives reveals two deviations from the properties of molecularry
dispersed flexible chains. A much weaker increase in Rg and R^ is found for the
MHPC and M H E C samples than expected and observed for linear chains. In other
words, the molar mass increases far more pronounced than the dimensions that only
increase weakly. Evidently, the arrangement of the polymer segments becomes more
densely packed with growing molar mass. In this context it is relevant to stress that the
actually measured molar mass is at least by a factor ten larger than 130 - 200 kmol/g
of common cellulose pulp with a DP of 800 - 1200. These two findings are obvious
August 31, 2012 | http://pubs.acs.org

indications for the presence of aggregates.

In order to find an explanation for the uncommon slopes expressed by a = ν R g

the concept of fractal dimensions dj- may be consulted according to which df = 1/v
(75). The fractal dimension is a quantity that usually indicates self similarity of the
polymer structures. In principal dj- can vary between 1 (rodlike structures) and 3
(compact structures). Values higher than 3 are physically irrelevant and have to be
considered as meaningless. The values one would obtain from the slopes observed
with the cellulose derivatives are in the range of 4.5 to 5 and cannot be sensible
dimensions; they need another explanation. One possibility would be the following:
Because of the high molar mass aggregation has occurred, and the explanation may be
sought in this fact. Indeed, the observed phenomena are in agreement with a model of
fringed micelles (7(5,77) (Figure 6). This model consists of laterally aligned chains
which form a rather compact and probably geometrically anisotropic core from which
dangling chain sections emerge. Due to the lateral aggregation the radius of gyration
changes only little while the mass increases strongly.
The question arises whether there exist other indications for the formation of the
above introduced aggregation model. Plots of A and ρ reveal behavior that again is
2

consistent with a dense packing of segments (see Figures 3, 4 and 5). Indeed, the
stronger negative exponent o is typical for crosslinked or associated structures, and
A2

the low value of the p-parameter is characteristic of a high structural density that is to
be expected with fringed micelles.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 226
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016
August 31, 2012 | http://pubs.acs.org

Figure 5 : (a) Cellulose 2.5-acetates in acetone: Rg, and p-parameter as a

function of molar mass, (b) C M C in 0.1M NaCl solution: p-parameter.
DS = 0.93; η = 118-66900 Pa at 2%.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 227

In fact, the model reconciles the main, apparently contradicting properties:

• Very high molar masses.
• Increasing segment density with molar mass.
• A weak increase of the dimension.
On the other hand, the global parameters are not sufficient for distinguishing more
detailed features of fringed micelles. Various shapes are still compatible with the above
findings; a few of them are shown in Figures 6a, b and c together with two other
modifications of laterally aligned chain sections.

Shape of Aggregates

General Remarks and Results. For further information on the structure it is useful
to measure the angular dependence of the scattered light. There exist several
techniques of presenting the data in suitable graphs (J8). For the present study the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

Kratky presentation (19) proved to be most instructive. In such plots the normalized
scattering intensity Re is multiplied by q and plotted against u=qRg, where Re is the
2

Rayleigh ratio at the scattering angle θ and q = (4π/λ)8ΐη(θ/2) is related to the

scattering angle. As usual, the radius of gyration Rg can be determined from the initial
2
slope of the inverse scattered intensity as a function of q (Zimm plot) (20). The
Kratky representation has the advantage that the asymptotic part of the scattering
2
function at qRg > 1 is amplified by (qRg) , which makes even small changes clearly
detectable. This asymptotic regime is sensitive to the internal architecture of the
particles. Therefore, one can easily distinguish rods from random coils or hard spheres.
Branched materials and compact aggregates are expected to develop curves in
between of those for hard spheres and random coils. A qualitative estimation can be
made already on a first sight. This facilitates the qunatitative evaluation by fits to
special models. Figure 7 exhibits some examples from the same sample measured
under different conditions. For comparison the curves for random coils and hard
August 31, 2012 | http://pubs.acs.org

spheres are added. Stiff rods exhibit a linear increase with qRg in such a plot.
Another, commonly used way of plotting the data is the double logarithmic
graph of the scattered intensity against q. The plot, that is not presented here, displays
a bent curve at low q-values but develops a power law decay at large q. The slope
gives the ensemble fractal dimension <dj>. Their values deviate strongly from
df = 4 - 5 , that was found with the data from the molar mass dependence of the
radius of gyration when using the samples of different degree of polymerization of the
individual nacromolecules. The values obtained for the individual samples are all
around <df> = 2.0 and 2.9, and he fully in the range of physically meaningful values.
Actually the true fractal dimension is not obtained from the asymptotic slope of
the scattering curve. It still contains an influence of the molar mass distribution which
results from the fact that the mean square radius of gyration is a z-average while the
molar mass is a weight average (21). However, as long as the ratio M / M does not
z w

significantly change with M the correction is small and can be neglected. For the
w

present cellulose derivatives the molar mass distribution is not known and could not be
determined by size exclusion chromatography since the materials adhered at the gel
matrix. The width of the size distribution is probably not large as will be demonstrated
below.

Attempt of Interpretation. The fractal dimensions varied significantly for different

samples from the same homologous series and even more when the solvent was
changed and other derivatives were compared. These data for df correlate well with

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 228
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

Figure 6 : Three different forms of fringed micelles (above) and two other models of
laterally aligned chain sections (below). Only in the first three cases a weak
increase of the radius of gyration with the molar mass is obtained. The
inserts at the upper left corner indicate the alignment of the chains in the
core. A more realistic picture exhibits the electron micrograph by Fink et
August 31, 2012 | http://pubs.acs.org

al. (48).

3.0

2.5

2.0

CL" 1.5
CN
D f=8,7
1.0

0.5 f = 38,6
f = 1 1 2.6
0.0

0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 20

6
Figure 7 : Kratky plots of a MHPC 2 ( M = 7.98xl0 ) in three solvents. The curves
w

are fits to the model of star-branched macromolecules, / denotes the

number of arms in a star molecule, see Eq.(2).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 229
the differences in the scattering curves of the Kratky representation (Figure 7 presents
only a small selection of measurements).
The observed curves resemble those of star shaped macromolecules with /
polydisperse arms. As was shown previously (22,23) the angular dependence of such
star molecules with flexible chains could be calculated. It is given in the Kratky
representation by the equation

2 2
l+u (f+l)/6f]

which for u = qRg » 1 approaches an asymptotic plateau of

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

2
u P ( q ) - > 1 2 — ( 3 )
(f + 1)
Thus the plateau height is a sensitive measure of the number of arms / Most of the
experimental scattering curves could be well fitted by the model of star-branched
macromolecules, but in a few cases, when / < 10, the fit became poor and not
applicable at all. Figure 8 demonstrates the effect.
Instead of approaching a plateau a strong increase is now observed at large q-
values. Such behavior is characteristic of stiff chains. Now the plot of q P ( q ) M w

approaches a constant plateau with a height of (25-29)

qP(q)M -^M w L (4)

Here M is the linear mass density of rod like chain sections of length
L and molar
mass M where the subscript k stands for Kuhn segment. The measured linear mass
August 31, 2012 | http://pubs.acs.org

density can be compared with the calculated one for a single strand MJl , where M Q Q

and l are the molar mass and bond length of the repeating unit. From the ratio of the
Q

measured to the calculated values the number of laterally aggregated chains can be
estimated. Before reaching the asymptotic plateau the curve for semiflexible chains
passes through a maximum Its position should appear at qRg = 1.41 i f chains of
uniform length are present and at qRg = 1.70 if M / M = 2.00 (5(5). For even broader
v v n

size distributions the maximum is shifted towards larger values. In the present
examples a value of 1.71 was in no case exceeded.
As a next point we checked whether the fractal dimension df is correlated to the
number of arms / Figure 9 shows the result. A strong correlation between the two
quantities can be clearly stated, though not a strict functional dependence. These
results from the angular dependence of the scattered light confirm the conclusion
drawn previously from the global properties and specify further details of the model.
The fact that dj- increases with the number of aggregated chains makes clear that the
aggregates are not randomly constructed because in this case the fractal dimension
would change between 2.0 and 2.5 (15,21). Hence the two models in Figure 6d and 6e
can be excluded. The increase beyond a value of 2.5 up to approximately 3.0 indicates
an increasingly dense packing. The objects must still have a large number of dangling
chains since otherwise the curves could not be fitted by the model of star-branched
molecules.
Moreover, for f < 10 we observed low fractal dimensions (df < 1.7) and clear
stiff chain behavior. Hence the densely packed core must have a remarkable rigidity

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 230
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016
August 31, 2012 | http://pubs.acs.org

u
Figure 8 : Kratky plots of some MHP-celluloses which cannot be fitted to the model
of star molecules. The strong increase at large q-values indicates chain
stiflhess. These curves could be fitted with the model of wormlike chains
(24-27).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 231

which could be quantified by the fit of the scattering curves with the model of worm-
like chains. Applying Koyama's theory of worm-like chains (24) the Kuhn segment
length could be determined (27) which ranged from l = 65 nm up to about 300 nm
k

(For comparison the Kuhn segment length of polystyrene is about 2 nm and that of
D N A molecules around 80 nm)
We are aware of the approximate character when describing fringed micelles by
star shaped objects which include dense and geometrically anisotropic cores, but
nonetheless this approach will enable refinements in the derivation of mathematical
models which can further be examined by small angle X-ray (SAXS) or small angle
neutron scattering (SANS) measurements.

Evidence for Fringed Micelles from other Experiments

The geometric anisotropy could be expected to be sufficiently large such that the core
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

might be visualized by electron microscopy. The geometric anisotropy will be very

likely connected with an optical anisotropy which should develop a noticeable
birefringence when the material is sheared. Both conjectures could indeed be realized.
Figure 10 shows the shear rate dependence of the viscosity and the development
of flow birefringence with increasing shear rate (30). The decrease in viscosity is a
common feature of entangled polymer chains, but in the present example it is also
accompanied by orientation of the rigid core. This conclusion becomes apparent by the
observation that the birefringence just started when the shear thinning exhibited a
significant effect.
Figure 11 gives a picture obtained by transmission electron microscopy (TEM).
Only the dense, slightly anisotropic cores are seen; the individual dangling chain
sections are too small in their cross-section. The cores appear to be isolated from each
other. This effect is caused by the dangling chains which keep the cores apart and
sterically stabilize the micelles.
August 31, 2012 | http://pubs.acs.org

Local Dynamics

The dynamics of polymers has been studied so far mostly by oscillatory rheology.
Although in the past much progress has been achieved in the development of sensitive
fluid rheometers, these instruments work satisfactorily only in semidilute solutions
where the viscosity is sufficiently high. Rheology is a mechanical technique that probes
the dynamics in a macroscopic manner, and conclusions on the local dynamics can be
drawn only via appropriate theories. For about 15 years the chain dynamics can be
studied also by dynamic light scattering, which gives a molecular response without
disturbing the systems by external forces. Information on the local dynamics can be
obtained when the particles have large dimensions, i.e. in the order of the wave length
of the light used. The method has recently been applied with success to branched
macromolecules (31-33) with the following result:
Much smaller distances than the radius of gyration are examined i f an angular
region of qRg > 3 is chosen. The influence of the translational diffusion of the center of
mass (that dominates at qRg < 2) has there already fully decayed (34\ and only the
segmental motions are seen. According to Zimm (35) and the refined theories on
dynamic light scattering from flexible linear chains by Pecora (36) , de Germes (37)
and Akcasu (38-40) the first cumulant T(q) of the time correlation function (TCF)
3
reaches a q dependence (the first cumulant represents the initial slope of the
logarithmic TCF against the delay time).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 232

3.6
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

100 150

Figure 9 : Plot of the fractal dimensions dj- obtained with various fractions and
derivatives of cellulose against the number of arms / The systems are
indicated by the insert. C A : cellulose 2.5-acetate in acetone;
Polyelectrolytes: C M C and alginates (not discussed here).

40
August 31, 2012 | http://pubs.acs.org

30

Δ Χ
10"
C/3 Δ viscosity ΔΧ 20 £
03
Ο­ Χ birefringencsl Δ
ο

Δ J 10

ΙΟ 1
h χ„ ,, χ

• Mill
; ι ι ι ι ιιI
10'· 10* 10 υ
ΙΟ 1

1
shear rate /s'
Figure 10: Shear rate dependencies of flow birefringence and shear viscosity of the
sample MHPC II in water (30).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 233
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016
August 31, 2012 | http://pubs.acs.org

Figure 11 : T E M picture from a 0.2% aqueous solution of the sample M H P C Π

7
( M = 2.07xl0 ). Negative staining with uranyl acetate.
w

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 234

It is convenient to define a reduced (dimensionless) first cumulant as

(5)

that should approach a constant plateau. The plateau height has been calculated for
linear chains in good and Θ-sorvents, repectively (40). In the pre-average
approximation the result is
Γ*(οο) = 0.071 good solvent
Γ*(οο) = 0.053 Θ-solvent
For branched materials the T*(q) values reach no constant plateau; instead, it
continuously decreases. This behavior is approximately described by a power law
(3 J,33,4J)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

r*(q) ~q-0.17±0.03 (6)

The curves of the different branched materials lie in the region between random coil
and hard sphere behavior.
At a value of qRg = 3 the T*(q) reached values of 0.040 and 0.030 for randomly
crosslinked polyesters and degraded amylopectin molecules, respectively. These
observations led us to the conclusion that the height of T*(q) at large qRg > 3 is a
measure of the internal flexibility. This conclusion is supported by the value for hard
spheres that for large qRg approaches T*(q)->0. The decrease in the flexibility is
evidently due to branching and the resulting high segment density (number of segments
per volume). Theories on the dynamics of branched macromolecules are presently still
missing.
So far, all theories and experiments were confined to infinitely dilute solutions.
In the present study we now extended the region of measurement to semidilute
August 31, 2012 | http://pubs.acs.org

solutions and performed measurements with the MHPC1 sample in the three solvents
(i) water, (ii) 2 M guanidine HCL/water and (iii) trifluoroethanol (TFE). The T*(q)
curves exhibited in some cases a m i n i m u m and finally increased again, but most of the
curves assymptotically approached a constant value. We compared the values found at
qRg = 3 which are shown in Figure 12.
A very pronounced decay of the T*(qRg=3) values with increasing concentration
was found. For a better display of the data we thus plotted *(qR = 3) against the
g

concentration. Even at c = 0 the highest value (0.033) was much lower than 0.071 for
linear chains, and this is considered being a result of the aggregated state. The
reduction of the T*(qRg=3) values depended slightly on the nature of the solvent used,
and the lowest value was found for the sample in 2 M guanidin HCl/water where also
the highest aggregation number was observed. The drastic decrease of the T*(qR^=3)
values with increasing concentration, (that is not found with polystyrene), gives
evidence for a remarkable inhibition of the segmental motion. In fact, the dangling
chains of the fringed micelles can interpenetrate only partly to form a transient
network of entangled chains. Beyond a certain concentration, however, the chains hit
the dense core where they cannot penetrate any further. Hence the chains become
more and more compressed, and this finally causes a complete loss of mobility.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 235

0.20
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

0.15

0.10

0.05

0.00
August 31, 2012 | http://pubs.acs.org

C / %

Figure 12 : r*(qRg=3)) as a function of polymer concentration for measurements of

6
M H P C 1 ( M = 6.47xl0 ) in water, 2 M Guanidine HCl/water and
w

Trifluoroethanol (TFE). The overlap concentration c* = A M c lies in all

2 w

cases around 0.3%.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 236

Conclusions

The following 5 facts may be summarized: (i) The very weak increase of the
dimensions that is accompanied by a very pronounced increase in the molar mass
indicates the formation of compact aggregates similar to that in micelle formation, (ii)
The analysis of the angular dependence of static light scattering measurements
revealed similarities to star branched macromolecules which posses a remarkable chain
stiffness (averaged over the whole structure). These two observations give strong
evidence for a fringed micellar structure as depicted in Figure 6a. (in) The anisotropy
of the hard core that is expected from laterally aligned chain segments could be
visualized by T E M micrographs and was confirmed by flow birefringence, (iv) The
flexibility of dangling chains was proven by dynamic light scattering, (v) The loss of
internal mobility on increasing the polymer concentration is in agreement with the
existence of a hard core that cannot be penetrated by the dangling chains. A l l these
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

facts are in close agreement with the fairly regular fringed micelle structure of Figure
6a. Since this structure is observed essentially with all not fully substituted cellulose
chains it can be considered as a rather universal feature of a special super molecular
structure. In searching for reasons of this uncommon aggregation structure two
explanations can be offered. Both are based on an uneven substitution along the
chains.
(i) The uneven derivatization is very likely caused by the morphology of the
native or regenerated cellulose in the solid state. These fibers show a structure
consisting of crystalline laterally aligned chain sections interrupted by amorphous but
still highly oriented domains (42,48). This particular structure was given the name
fringed micellar crystals. Most of the derivatives are prepared heterogenousry, and
thus the easier accessible amorphous chain sections will be preferably derivatised
leaving the crystal regions widely unaffected. The highly substituted chain sections
dissolve in the various solvents while the crystalline part remains stabilized by the
August 31, 2012 | http://pubs.acs.org

many hydrogen-bonds. Solubility of the colloidal particles (based on macromolecules)

results from the entropy of mixing of dangling chains with the solvent (43). The
mechanism is the same as predicted by Wegner (44,45) and Ballauff (46) for then-
hairy rod systems. Thus the first explanation is based on an incomplete dissolution
process.
(ii) The chains in the crystalline domains may be less derivatized than in the
amorphous domains but still may contain a number of substituents. In this case special
solvents may be found which can break up the hydrogen-bonds in the core. In fact the
aggregation number could be changed by different solvents. The guaninidin HCl/water
solvent was expected preferably to break the hydrogen bonds, but opposite behavior
was found. This might be the result of complexities in the hydrophobic interactions.
TFE decreased the number down to 4-10 laterally aligned chains. Application of large
shear rates may possibly allow full disruption of the structure, but the aggregated
fringed micellar structure is expected to reconstitute under equilibrium conditions.
Observations with an arabinoxylan (47) gave clear indications for a strong increase of
chain stiffening as chains start to align laterally, which is probably the result of a
cooperative formation of an hydrogen bond system Once 6-7 chains have been aligned
further lateral aggregation took place in which almost no change in the radius of
gyration was found.
The final result in both explanations is very similar, and further investigations
with other derivatives are needed for distinguishing between the two indicated
possibilities of dissolution. Depart of this uncertainty on the mechanism, it may be

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 237

noted that similarities in the morphology of different cellulose sources and

particularities in the chain topology of the individual chains are the mean reason for the
surprisingly universal supermolecular structure. This fact offers new chemical routes to
transforming these structures into functional elements.

Acknowledgement The work was partially supported by the Ministry of Research

and Development of the Federal Republic of Germany in Bonn.

Literature Cited

1 Burchard, W.; Schulz, L. Dos Papier 1989, 43, 665.

2 Schulz, L.; Burchard, W. Dos Papier 1993, 47, 1.
3 Burchard, W.; Lang, P.; Schulz, L.; Coviello, T. Makromol. Chem., Macromol.
Symp. 1992, 58, 21.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

4 Burchard, W.; Schulz, L. Macromol. Symp. 1995, 99, 57.

5 Lang, P.; Burchard, W. Makromol. Chem. 1993, 194, 3157.
6 Schulz, L. Ph.D. Thesis, University of Freiburg 1996.
7 Wenzel, M.; Burchard, W. Polymer 1986, 27, 195.
8 Benoit, H.; Holtzer, A.M.; Doty, P. J. Phys. Chem. 1954, 58, 635.
9 Seger, B.; Burchard, W. Macromol. Symp. 1994, 83, 291.
10 Kirkwood, J.G. J. Polym. Sci. 1954, 12, 1.
11 Yamakawa, H. Modern Theory of Polymer Solutions, Harper & Roe,
New York 1971.
12 Flory, P.J.; Fox, T.G. J. Am. Chem. Soc. 1951, 73, 1904.
13 Burchard, W.; Schmidt, M.; Stockmayer,W.H.Macromolecules 1980, 13,
1265.
14 De Gennes, P.-G. Scaling Concepts in Polymer Physics, Cornell University
Press, Ithaca, NY, 1979.
August 31, 2012 | http://pubs.acs.org

15 Daoud, M.; Martin, J.E. In: The Fractal Approach to Heterogeneous Chemistry,
Ed. Avnir, D., Wiley & Sons, New York 1992, p 109-130.
16 Hermann, K.; Gerngross, O. Kautschuk 1932, 8, 565.
17 Burchard, W. TRIP 1993, 1, 192.
18 Burchard, W. Adv. Polym. Sci. 1983, 48, 1.
19 Kratky, O.; Porod, G. J. Collod Sci. 1949, 4, 35.
20 Zimm, B.H. J. Chem. Phys. 1948, 16, 1093.
21 Stauffer, D. Introduction to Percolation Theory, Taylor & Francis,
San Francisco 1985.
22 Burchard, W. Macromolecules 1974, 7, 841.
23 Burchard, W. Macromolecules 1977, 10, 919.
24 Koyama, R. J. Phys. Soc. Japan 1973, 34, 1029.
25 Schmidt, M.; Paradossi, G.; Burchard, W. Makromol. Chem., Rapid Commun.
1985, 6, 161.
26 Denkinger, P.; Burchard, W. J. Polym. Sci. 1991, 20, 589.
27 Dolega, R. Fitprogram KOYFIT, Freiburg 1992.
28 Casassa, E.F. J. Chem. Phys. 1955, 23, 596.
29 Holtzer, A.M. J. Polym. Sci. 1955, 17, 432.
30 Schmidt, J.; Richtering, W.; Weigel, R.; Burchard, W. Macromol. Symp. 1997,
submitted.
31 Trappe, V.; Bauer, J.; Weissmueller, M.; Burchard, W. Macromolecules,
submitted

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 238

32 Galinsky, G.; Burchard, W. Macromolecules, submitted.

33 Trappe, V.; Burchard, W. Proceedings of the Krakow Symposium 1996,
Light Scattering and Photon Correlation Spectroscopy, submitted.
34 Berne, B.J.; Pecora, R. Dynamic Light Scattering, Wiley & Sons, New York
1976.
35 Zimm, B.H. J. Chem. Phys. 1956, 24, 269.
36 Pecora, R. J. Chem. Phys. 1968, 49, 1038.
37 Dubois-Violette, Ε.; de Gennes, P.-G. Physics 1967, 3, 181.
38 Akcasu, A.Z.; Benmouna, M.; Han, C.C. Polymer 1980, 21, 866.
39 Benmouna, M.; Akcasu, A.Z. Macromolecules 1978, 11, 1187.
40 Benmouna, M.; Akcasu, A.Z. Macromolecules 1980, 13, 409.
41 Burchard, W. Adv. Colloid Interface Sci. 1996, 64, 45.
42 Fengel, D.; Wegener, G. Wood, de Gruyter, Berlin 1989.
43 Flory, P.J. Principles of Polymer Chemistry, Cornell University Press, Ithaca
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch016

NY, 1953.
44 Seufert, M.; Fakirov, C.;Wegner, G. Adv. Mater. 1995, 7, 52.
45 Seufert, M.; Schaub, M.; Wenz, G. Wegner, G. Angew. Chem. Int. Ed. Engl.
1995, 34, 340.
46 Ballauff, M. Fluid Phase Equilibria 1993, 83, 349.
47 Ebringerova, A.; Hromadkova, Z.; Burchard, W.; Dolega, R.; Vorwerg, W.
Carbohydr. Polym. 1994, 24, 161.
48 Fink, H.-P.; Purz, H.J.; Bohn, Α.; Kunze, J. Macromol. Sympos. submitted
49 Wenzel, M.; Burchard, W.; Schätzel, K. Polymer. 1986, 27, 195.
50 Baur, G., 1988, Baur Instrumentenbau, Hausen, Germany.
51 Bantle. S.; Schmidt, M.; Burchard, W. Macromolecules 1982, 15, 1604.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 17

Phase Behavior, Structure, and Properties

of Regioselectively Substituted Cellulose Derivatives
in the Liquid-Crystalline State

Peter Zugenmaier and Christina Derleth

Institute of Physical Chemistry, Technical University of Clausthal, D-38678

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

Clausthal, Zellerfeld, Germany

The helical twisting power, value and handedness, of the super-

molecular helicoidal structure of liquid crystalline (lc) cellulose
derivative / solvent systems strongly depends on the substituents
introduced. Investigations on lc cellulosetrisphenylcarbamate and
cellulosetris-3-chlorophenylcarbamate reveal different sign of the
twisting power in the same solvent triethylene glycol monomethylether.
The parameters which influence such a behavior have been studied.
These are the site of phenylcarbamate substitution at the anhydroglucose
unit (2, 3, 6), the site of substitution at the phenyl ring (3 or 4) including
different substituents (hydrogen, chloro, methyl, fluoro) on regio-
selectively substituted chains in ethylene glycol monomethylether
August 31, 2012 | http://pubs.acs.org

acetate and, for one case, a random substitution of two groups along the
cellulose chain. A strong polymer solvent effect has been detected which
is predominantly influenced by the substitution site at the
anhydroglucose unit and at the phenyl ring. A study of the phase
diagram supports the idea that clusters with bound solvent are formed.
These have to be regarded as the structural units for the lyotropic
cholesteric phase.

Lyotropic liquid crystalline cellulose derivatives formed by highly concentrated

solutions belong to a special class of chiral materials (/). Right- and left-handed super-
molecular hélicoïdal structures of chiral nematic, also termed cholesteric mesophases,
are observed with positive or negative temperature and concentration gradients for the
twisting power. Studies also show that not all cellulose derivatives exhibit lyotropic
liquid crystals, rather some lead from the semi-dilute state, where already microgels
appear, directly to the gel state, although the chain backbone of these structures has
similar stiffness as compared to those derivatives which produce liquid crystalline (lc)
phases. This behavior might depend on the chain length that is the molecular mass.
Cellulose derivatives with high molecular mass normally omit the lc state and form gels

©1998 American Chemical Society 239

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 240

only. This behavior might also occur with short chain molecules depending on the
substituents. Considering these observations, it seems that for the lyotropic lc cellulose
derivative systems, the polymer solvent interaction plays an important role. Little is
known about this interaction in cellulosic systems. From solvent built-in crystals, which
are border line cases of liquid crystals, a fiber structure analysis shows that, depending
on the solvent, different conformations of the cellulose backbone may occur, or only
the side group conformation changes. In most cellulosics the molecular structure does
not change at all, rather the packing of the chains accommodates for the solvent. The
other border line case, the semi-dilute state, clearly exhibits for the fully substituted
cellulose molecule (degree of substitution DS=3 for the anhydroglucose monomer unit)
that reversible aggregates rather than molecular dispersed single molecules form the
basic building blocks. This statement should also hold for the liquid crystalline state,
since no dissolution of these blocks is observed when going to the liquid crystalline
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

state. It has also been found that the solvent in these lyotropic systems, semi-dilute or
liquid crystalline, is rather tightly bound to the polymer chain and not freely available
for a crystallization process of the solvent by lowering the temperature. In the dilute
state molecularly dispersed single molecules are observed for fully substituted cellulose
derivatives not capable of hydrogen bonding.
The driving force of the transition from the isotropic to the anisotropic phase of
lyotropic systems for stiff chains is believed to lie in the structuring of the solvent.
Studying semi-dilute solutions of cellulose derivatives, we have been able to show that
most of the solvent is bound to the polymer that means little free solvent is left and that
a structuring of the cellulose chains has to be considered as well forming reversible
clusters (2, 3).
Lyotropic liquid crystalline cellulose derivatives exhibit chirality at three levels.
The chirality caused by the configuration of the molecule (here chiral centers), by the
August 31, 2012 | http://pubs.acs.org

conformation of the macromolecules (here normally left-handed helices) and by super-

molecular structures as the cholesteric hélicoïdal structure of various handedness.
Chirality at these different levels should be reflected, e.g., by the twisting power of the
cholesteric phase. At the present little is known about the correlation of chirality and
twisting power except for the rare case of a thermotropic liquid crystal where two
chiral centers are placed far apart in a molecule, and the conformation of the molecules
with different configurations seems to be very similar. For this example additivity of the
twisting power for the two centers was proven. The twisting power by mixing various
configurations of these compounds has been described by the weighed twisting power
of the different configurations with the molar fractions as weights (4).
The ultimate goal of our investigations is to establish a relationship between
chirality and the twisting power of cellulose derivatives knowing the twisting power of
the various substituents at the different position of the anhydroglucose unit and the
conformational helix. This would enable to predict the helix conformation in the
lyotropic liquid crystalline state with the experimentally determined twisting power.
Little is known about this correlation at the present time. There are two pathways to
start such an investigation: Firstly, to substitute statistically with two different groups
and changing the overall composition along the chain, preferential at the different
possible sites forming a copolymer. Secondly, to substitute regio-selectively at the
various positions 2, 3, 6 at the anhydroglucose unit uniformly along the chain. In this
paper we will report results for various but very similar cellulose trisphenylcarbamates

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 241

also termed cellulosetricarbanilates (CTC cf. Scheme 1) belonging to both groups that
?

will, on the other hand, establish equally well the importance of the polymer solvent
interaction.

Experimental

The cellulosetrisphenylcarbamate and cellulosetris(3-chlorophenylcarbamate) have been

synthesized by well-established procedures as was the statistical copolymer with
random distributed phenylcarbamate and 3-chlorophenylcarbamate side groups. These
derivatives were characterized to establish their chemical constitution by elemental
analysis, IR, N M R , etc. (5). The synthesis and characterization of regio-selective
derivatives, always trisubstituted, have been described elsewhere (<5). An overview of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

regio-selective materials available for this investigation is summarized in Table I with

abbreviations and chemical constitution of the molecules given in Scheme 2.

Table L Overview of Cellulose Derivatives Synthesized and their Corres­

ponding Polymer Code for the Position of Substitution 2, 3, 6 at the
Anhydroglucose Unit (C: Chloro, M : Methyl, H: Hydrogen, F: Fluoro).
Cellulose Derivative Polymer Code
2, 3, 6
cellulosetris(3 -chlorocarbanilate) CCC
cellulose-2,3-bis-(3-chlorocarbanilate)-6-(carbanilate) CCH
cellulose-2,6-bis-(3-cWorocarbanilate)-3-(carbanilate) CHC
cellulose-2,3-bis-(carbanilate)-6-(3-chlorocarbanilate) HHC
cellulosetricarbanilate HHH
August 31, 2012 | http://pubs.acs.org

cellulose-2,3-bis-(carbanilate)-6-(4-fluorocarbanilate) HHF
cellulose-2,3-bis-(4-fluorocarbariilate)-6-(carbanilate) FFH
cellulose-2,3-bis-(3-chlorocari)ardlate)-6-(3-methylcarbanilate) CCM
cellulose-2,3-bis-(3-methylcaii)anilate)-6-(3-cWorocait)anilate) MMC
cellulosetris(3-methylcarbanilate) MMM

The highly concentrated solutions with triethylene glycol monomethyl ether

(TRIMM) for the copolymer and with ethylene glycol monomethylether acetate
(EMMAc) for the regio-selectively substituted ones were prepared by mixing an
appropriate amount of the derivative with the necessary volume of solvent in an
Eppendorf vessel under stirring. Homogeneous textures were obtained by placing the
highly concentrated solution between two cover glass plates kept 50 μιη apart by a
spacer cut out of a commercially available foil of this thickness.
Most of the samples showed selective reflections of one color in the visible light
range, proving that a cholesteric hélicoïdal structure with the helix axis perpendicular to
the glass plates was present as depicted in Figure 1. The pitch of the super-molecular
structure was then deterrriined by spectroscopic means, by ORD and UV-VIS, taking
the zero optical rotation in the anomalous dispersion range or the peak of the transmis-
sion as selective reflection λ ο and calculating the pitch Ρ according to de Vries (7) with
Ρ = λ ο / η , η being a mean refractive index. If the selective reflection was not situated

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017
August 31, 2012 | http://pubs.acs.org

X1 = -H, - Cl, -CH3 Χ 2 = -H, - F

Scheme 2. Schematic representation of substituted CTC.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 243
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

left-handed
hélicoïdal
licoidal s t r u c t u r e ^ ^ / / / / ^ ^ ^ ^ ^
/
August 31, 2012 | http://pubs.acs.org

Figure 1. Schematic representation of cholesteric liquid crystalline structures.

Right-handed hélicoïdal structures are assigned a positive, left-handed ones a
1
negative value of the pitch Ρ or twisting power P" , respectively.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 244

in the visible spectrum range, the Cano-Grandjean (8) method was applied with
additional measurement of the handedness of the hélicoïdal structure by ORD.

Results and Discussion

1
Statistical Cellulose Copolymer. The twisting power P" of the lc state as a function
of composition (percentage of 3-chlorophenyl versus phenyl groups) in T R I M M is
plotted in Figure 2. Pure cellulosetris(3-chlorotricarbamate) (CCC) exhibits a right-
handed hélicoïdal structure, pure cellulosetricarbanilate (HHH) a left-handed one. It is
1
clear from this graph that the twisting power P" of a known composition χ of 3-
chlorophenyl side groups may be represented as additive in a first approximation of
1 1
average twisting powers PH" and Pc" of the corresponding pure trisubstituted
compounds H H H and CCC, respectively:
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

1
F ^ X P C ^ + O-XÎPH" (1)
1
At a composition χ of about 0.5 a compensated structure is obtained with P" =
0 or pitch infinity that resembles a nematic phase. It is also known from solid state fiber
structure analysis that both derivatives, H H H and CCC, form similar left-handed 3-fold
helices with a repeat of 15 Â as preferred conformations of the cellulosic chains. If the
idea is correct that a right-handed chain conformation leads to a right-handed super-
molecular lc structure (9), the conformation of a single C C C chain must be right-
handed. Since the preferred conformation in the solid state is left-handed, it has to be
concluded that the polymer solvent interaction causes a reversal of the twist sense.
Such a reversal can be easily visualized, if the solvent causes two adjacent monomer
units to become the building block of the chain conformation. A 3-fold left-handed
August 31, 2012 | http://pubs.acs.org

helix of a polymer chain may then turn to a 3-fold right-handed one with the pitch twice
the size of the left-handed helix. Assuming a similar overall conformation of the
backbone, a different placement in the side group of two adjacent residues in e.g. the 6
position or in the solvent attached to the two adjacent units leads to a dimer as
conformational unit and a twist inversion of the conformation.
It is amazing how much a small amount of a second side group with a statistical
distribution of both side groups along the chain causes a change in the twisting power
although a full helix pitch of one component, right- or left-handed, can not be
established. Also the different substitution sites 2, 3, 6 at the anhydroglucose unit
differently influences the twisting power. From our investigations on semi-dilute
solutions cluster formation was deduced. These clusters have to be considered in the lc
state to exist and make an interpretation of the results on the twisting power of a
copolymer, as depicted in Figure 2, even more difficult.

Regio-Selectively Substituted Cellulose Derivatives. The optical rotatory dispersion

and the transmission curves for right-handed super-molecular lc structures are depicted
in Figures 3 and 4 for the system C C H / E M M A c at various temperatures and a fixed
concentration of 0.70 g/ml, respectively. A decrease of the peak height at elevated
temperatures is observed in the anomalous dispersion range of the ORD curves. This
effect is mirrored in the UV-VIS spectra. The transmission Τ should amount to 50 %

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 245
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017
August 31, 2012 | http://pubs.acs.org

J
-4000
1
Figure 2. Twisting power P" as a function of composition for a copolymeric
phenylcarbamate- / 3-chtorophenylcarbamate cellulose; random distribution along
the chain in triethylene glycol monomethylether (TRIMM); c = 0.8 g/ml, room
temperature. Percentage for the degree of substitution for chlorophenyl groups
DS(C1) corresponds to x=DS/100. (Adapted from ref. 5).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 246

300
200
V fff\

II
V JM Ί (h

100
σ>
0)
Ό

-100
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

-200 Η
-300
200 300 400 600 700 800
500

λ / nm
Figure 3. Optical rotatory dispersion (ORD) curves for the right-handed
cholesteric mesophase C C H / E M M A c ; c = 0.70 g/ml; sample thickness 50μπι.
The temperature for the various curves rises from left to right: 301 K , 305 K , 309
K,313K,317K.
August 31, 2012 | http://pubs.acs.org

100

750
Figure 4. Transmission curves in the visible spectral range (UV-VIS) for lc C C H /
E M M A c ; c = 0.70 g/ml; sample thickness 50 μπι. The temperature for the various
curves rises from left to right: 297 K , 301 K , 305 K , 309 K , 313 K .

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 247

for linearly polarized light according to theoretical considerations but is found to be

only about 35 % or less at higher temperatures. The ORD curves can only be described
with the general de Vries's equation when two additional factors are introduced. A
dispersion term for the chromophores of the side groups that accounts for the rapid
increase of the optical rotation at smaller wavelengths and a damping factor that
reduces the optical rotation in the dispersion range. A perfect hélicoïdal structure
should depict a singularity at the selective reflection wavelength λο. For a real structure
1
the pitch Ρ or the twisting power P" are deduced from zero optical rotation in the
anomalous dispersion region or the peak in the UV-VIS curves, which represent the
selective reflection λο. With the knowledge of the mean refractive index η of a nematic
sheet of the cholesteric structure, the pitch is calculated by Ρ = λ ο / η . The mean
refractive index η was obtained by measuring the ordinary and extraordinary refractive
index of the samples with an Abbé refractometer at the desired temperature.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

1
The temperature dependent measurements of the twisting power P" are
represented for regio-selective lc cellulose derivative systems in Figure 5, and the pitch
Ρ listed at T= 303 Κ in Table II. For some of the derivatives the selective reflection lies
outside the instrumental spectral range. The pitch was then determined by the
Grandjean-Cano technique.

Table IL Pitch Ρ and Handedness of Various Cellulose Urethane / Solvent

Systems at Τ = 303 K, c = 0.7 g/ml and Slightly Varying Degree of
Polymerization DP; Sample Thickness 50 μιη.

Polymer/Solvent System Pitch DP°>

P/nm
August 31, 2012 | http://pubs.acs.org

CCC/EMMAc + 318 280

C C M / EMMAc > b
+ 280 250
MMC / EMMAc + 315 290
MMM/EMMAc + 302 285

CCH/EMMAc + 413 210

CHC/EMMAc +1140 270

HHC / E M M A c -1015 280

HHH/EMMAc -670 245

HHH / EMMAc c)
-522 245
HHF / E M M A c c)
-488 110
FFH / E M M A c c)
-517 100

a)
cf. réf. 6
b)
c = 0.8 g/ml
c)
c = 0.9 g/ml

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 248

0,0045

0,0030
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

0,0015 -

Έ
c
^ 0,0000 -
û.

-0,0015 Η

-0,0030 -
August 31, 2012 | http://pubs.acs.org

-0,0045 -J , , , , , 1
270 280 290 300 310 320 330 340
Τ/Κ
1
Figure 5. Temperature dependence of the twisting power P" of various left- and
right-handed hélicoïdal structures of lc cellulose derivatives in E M M A c . C C M 0;
M M M • , M M C V , C C C • , C C H A , H H H · , H H F Δ, HHF Ο (c = 0.80 g/ml);
the concentrations for the other lc states are listed in Table Π.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 249

A negative helical twisting power, signifying a left-handed super-molecular

helical structure of the lyotropic cholesteric phase, is found for cellulosetricarbanilate
(HHH) and the 4-fluoro derivatives of the phenyl residue, HHF and FFH, as well as for
HHC, the derivative of the tricarbanilate for which a 3-chloro substituted phenyl ring is
placed at the 6 position of the anhydroglucose unit. All the other derivatives exhibit
right-handed hélicoïdal super-molecular structures. The temperature gradients of the
twisting power for all lc systems shown in Figure 5 are very similar in their absolute
values. Nevertheless, they are positive for left-handed structures and negative for right-
handed ones.
From a broad study on lc behavior in ref. 3, it can be concluded that all meta
substituted cellulosetricarbanilates (CTC) with F, CI, C H 0 , C F in a variety of
3 3

solvents led to right-handed super-molecular structures as did all bis-substituted in 3, 4

position at the phenyl ring. All para substituted C T C (Cl, Br, F, C H 0 ) exhibit left-
3
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

handed ones. Only pure lc CTC changes handedness depending on solvent.

Discussing the hélicoïdal structure in more detail, it is clear from Table II that
replacing a 3-chlorophenyl by a 3-methylphenyl group at all sites of the anhydroglucose
unit does not change the pitch. However, exchanging a 3-chloro or 3-methyl group by
a hydrogen at the phenyl residue and placing this substituent in 2 or 3 position of the
anhydroglucose unit drastically alters the pitch in size for CHC and in sign for HHC
and HHH. Replacing a 3-chloro group by hydrogen at the phenyl substituent in 6
position slightly increases the pitch only. A compensated chiral nematic phase may be
obtained for CHC and HHC by adjusting either the external parameters as temperature
and concentration and/or by mixing the two compounds physically or statistically
distributing the various groups along the molecular chain. These investigations of the
helical twisting power reveal the most sensitive substitution sites for the super-
molecular structure of lyotropic lc cellulose derivatives to be the 2 and 3 position at the
August 31, 2012 | http://pubs.acs.org

anhydroglucose unit. Similar conclusions can be drawn for the helix conformation and
packing of a single chain in the solid state for which the structure is predominantly
determined by the substitution in 2 and 3 position at the anhydroglucose unit (10). The
electronegativity of the substituent at the phenyl residue is of minor importance as
compared with the substitution sites 3 or 4. The 4-fluoro derivatives HHF and F F H all
lead to left-handed structures with similar pitch as H H H in contrast to the 3-chloro
derivative C C H with a right-handed one. Since the pitch was found to depend on the
molecular mass (77), the degree of polymerization DP is also listed in Table II. From
former studies it can be concluded that a DP of > 150 is beyond any influence on the
size of the pitch, but a small effect may be expected for the fluoro derivatives in
comparison with all the other regio-selective cellulose derivatives.
1
The temperature dependence of the twisting power P" of chiral nematic phases
was investigated by Kimura et al. (72) and Equation 2 derived:

l
V = Q(JJT-\) (2)

Q is a factor that includes geometry and volume fraction of the polymer in solution; Q
> 0 for right-handed super-molecular structures and Q < 0 for left-handed ones.
T represents the inversion temperature for which the super-molecular structure
n

changes the sign of the twisting power. T lies above the clearing temperature T for
n c

the systems considered.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 250

Two plots of the pitch for positive and negative Q values as a function of temperature
are shown in Figure 6. Comparing the results for the regio-selectively substituted
cellulose derivatives with those in Figure 6 leads to the following conclusions: Curve b
(left plot) with an increasing pitch or decreasing twisting power best describes the
experiments for right-handed hélicoïdal super-molecular structures (Q > 0). For such a
behavior predominant polar and steric effects are responsible. The same argument holds
for the left-handed structures. The decreasing size of twisting power with temperature
for both types of structures also supports the idea that left-handed conformational
helices produce a left-handed super-molecular structure and a right-handed
conformational helix a right handed cholesteric structure.

Phase Behavior. The study of the phase behavior and properties at the phase
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

transition represents a crucial test for theoretical models that have been developed for
various kinds of liquid crystals. For lyotropic le CTC with predominant polar
interactions, none of the existing models can explain the phase diagram exactly (2).
Although the volume fraction for which the anisotropic-isotropic transition occurs
might be predicted to some accuracy, the actual small biphasic region for a broad
molecular mass distribution and the bending of the curves at higher temperatures
cannot be described by any model. Figure 7 depicts the phase diagram for C C C /
E M M A c taken from texture observations in the polarization microscope. A small
biphasic region is detected and a bending of the curves occurs at higher temperatures,
the same features as described above, although only steric mixed with polar interactions
are detected, and instead of a left-handed super-molecular structure as above, a right-
handed one was established. The same difficulties arise for the description of the CCC /
E M M A c system adjusting to the theoretical models as for lc CTC, and only fair
August 31, 2012 | http://pubs.acs.org

agreement is obtained with the model of Warner and Flory (73), although anisotropic
interactions are taken into account. The bending of the curves may be explained by a
variation of persistence length with temperature but the small biphasic region cannot be
accounted for. It is questionable from the current knowledge of the structure of these
systems that the models discussed may be suitable for a description of the isotropic-
anisotropic transition in cellulosics. Clearly cluster formation occurs and bound solvent
is present in these systems. These effects may play an important role in considering
phase transitions. Also a phase separation between lower and higher molecular masses
may occur, since different kinds of clusters are observed depending on the size of the
cellulosic molecules.
At higher concentrations below the cholesteric phase, a columnar phase appears
as in the CTC / diethylene glycol monoethylether system (14). In this case an oriented
fiber could be produced, and the X-ray analysis revealed hexagonal packing of the
molecules.

Acknowledgments. Part of the work reported was supported by a grant from

Deutsche Forschungsgemeinschaft.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 251
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

Figure 6. Pitch P versus temperature Τ according to Equation 2 for right-handed

conformational helices Q > 0 (left plot) and left-handed ones Q < 0 (right plot):
(a) predominant polar interactions; (b) polar and steric effects; (c) predominant
steric effects. (Adapted from ref. 2).

380
August 31, 2012 | http://pubs.acs.org

0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3

c/(g/ml)

Figure 7. Phase diagram for the system CCC / E M M A c evaluated by texture

observations.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 252

Literature Cited

1. Gilbert, R. D. In Polymeric Materials Encyclopedia; Salamone, J. C., Ed.; CRC

Press: Boca Raton, Florida, 1996, pp 118.
2. Haurand, P.; Zugenmaier, P. Polymer 1991, 32, 3026.
3. Klohr, E.; Zugenmaier, P. Cellulose 1995, 1, 259; Klohr, E. Dissertation, TU
Clausthal, D-38678 Clausthal-Zellerfeld, 1995.
4. Dierking, I.; Gießelmann, F.; Zugenmaier, P. Mol. Cryst. Liq. Cryst. 1996, 281,
79.
5. San-Torcuato, A. Diploma Thesis, Institut für Physikalische Chemie der TU
Clausthal, D-38678 Clausthal-Zellerfeld, 1989; Zugenmaier, P. Das Papier 1989,
43, 658.
6. Aust, N.; Derleth, C.; Zugenmaier, P. Macromol. Chem. Phys. 1997, 196, in press.
7. De Vries, H. Acta Cryst. 1951, 4, 219.
8. Grandjean, F. C. R. Acad. Sci. Fr. 1921, 172, 91; Cano, R. Bull. Soc. Fr.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch017

Minéral. Cristallogr. 1968, 91, 20.

9. Hartshorne, Ν. H. The Microscopy of Liquid Crystals, Microscope Publications
Ltd., London, England, 1974, p 80.
10. Iwata, T.; Okamura, K.; Azuma, J.; Tanaka, F. Cellulose 1996, 3, 91 and 107;
Möller, R. Diploma Thesis, Institut für Physikalische Chemie der TU Clausthal, D-
-38678 Clausthal-Zellerfeld, 1982.
11. Siekmeyer, M.; Zugenmaier, P. Makromol. Chem. Rapid Commun. 1987, 8, 511.
12. Kimura, H.; Hosino, M.; Nakano, H. J. Phys. (France) 1979, 40, C3-174 and J.
Phys. Jpn. 1982, 51, 1584.
13. Warner, M.; Flory, P. J. J. Chem. Phys. 1980, 73, 6327.
14. Hildebrandt, F.-I. Diploma Thesis, Institut für Physikalische Chemie der TU
Clausthal, D-38678 Clausthal-Zellerfeld, 1991; Zugenmaier, P. In Cellulosics:
Chemical Biochemical and Material Aspects; Kennedy, J. F.; Phillips, G. O.;
Williams, P. Α., Eds.; Ellis Horwood: New York, NY, 1993, pp 105.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 18

Blends of Cellulose and Synthetic Polymers

R. St. J. Manley

Department of Chemistry, McGill University, Montreal, Quebec H3A 2A7, Canada

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

This article reviews and summarizes a series of experiments designed to

explore the possibility of forming miscible blends of cellulose with
synthetic polymers. It is emphasized that for this purpose it is important
to choose synthetic polymers containing functional groups that can
interact strongly with the hydroxyl groups of the cellulose chains. The
results support the view that cellulose/synthetic polymer blends are not
as intractable as they were once thought to be, and that a very intimate
level of mixing can be attained in these systems.

During the past decade intense interest has been focussed on polymer blends, partly
August 31, 2012 | http://pubs.acs.org

because of their intrinsic scientific interest and partly because of practical considerations,
because it can be anticipated that such materials will show new desirable physical and/or
physico-chemical properties not to be expected in conventional homopolymers. Binary
blends in which the two components are synthetic polymers have been extensively
investigated (1,2). Somewhat surprisingly, however, very little work has been done on
blends in which one component is unmodified cellulose. This is primarily because of
certain difficulties inherent in the preparation of blends involving cellulose. There are
basically two ways in which a blend can be made. One is by mixing the components in
the softened or molten state and the other is to blend them in solution. But cellulose
cannot be melted (the thermal decomposition temperature lies well below the melting
point) and until relatively recently no convenient organic solvent was known. In recent
years, however, a variety of new solvent systems for the dissolution of cellulose have
been described (3-6), and as it happens these systems will also dissolve many of the
synthetic polymers that are of interest. Thus the way is now open for the systematic
study of cellulose/synthetic polymer blends and the field has become a subject of
increasing interest (7-10).
Polymer blends can be subdivided into different categories. The most important
distinction is between the so-called incompatible and compatible blends. Incompatible
blends in which the two components consist of separate well-defined phases or domains

©1998 American Chemical Society 253

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 254

represent the large majority of all polymer blends. Compatible or miscible blends which
consist of a single phase are in the minority. One of the main objectives of our work was
to characterize the state of miscibility of a variety of cellulose/synthetic polymer blends.
The state of miscibility is an important property because immiscible blends generally
have a coarse structure which is reflected in poor mechanical properties. On the other
hand miscible blends may combine the properties of the miscible components and hence
the mechanical properties may be superior to those of the component polymers.
For blending with cellulose, it is important to choose synthetic polymers
containing functional groups that can interact strongly with the hydroxyl groups of the
cellulose chains. Such intermolecular interaction is recognized as providing the driving
force for the attainment of thermodynamic miscibility (i.e., miscibility down to the
molecular level) in polymer blend system. Many important synthetic polymers satisfy
this condition, for example, polyamides, polyesters and vinyl polymers such as
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

polyvinyl alcohol), poly(acrylonitrile), polyvinyl pyrrolidone), and poly(4-vinyl

pyridine). Figure 1 shows examples of the kind of interactions that are involved.
The purpose of this article is to discuss the principles associated with the
formation of miscible cellulose/synthetic polymer blends using four examples drawn
from recent research in our laboratory, namely (cellulose/poly(acrylonitrile) (11),
cellulose/poly (vinyl alcohol) (12), cellulose/poly (vinyl pyrrolidone) (13), and
cellulose/poly(4-vinyl pyridine) (14).

Examples of the Results Obtained with Various Blends

It is appropriate to begin by making brief reference to the methods of sample

preparation. The cellulose sample was a wood pulp with a degree of polymerization of
~ 930, corresponding to a molecular weight of about 160,000. The solvents were N , N -
August 31, 2012 | http://pubs.acs.org

dimethylacetamide/lithium chloride or dimethyl sulfoxide/paraformaldehyde. For the

preparation of the blends two solutions were made at a concentration of about 1.5% in
the same solvent. One was the solution of cellulose and the other of the synthetic
polymer. The two solutions thus separately prepared were mixed at room temperature
in the desired proportions, so that the relative composition of the two polymers in the
mixed solutions ranged from 10/90 to 90/10, in a ratio of weight percent, the first
numeral referring to cellulose content. Each blend solution was then coagulated with a
non-solvent to form a film or else solution cast directly to form a film.
A commonly used method for investigating miscibility in polymer blends is to
measure the glass transition temperature (T ) of blends of various compositions. It is
g

well known that a miscible blend must show a single glass transition temperature
intermediate between the values for the pure components, over the whole range of blend
compositions. Figure 2 shows an example of the results of such measurements for the
blend system cellulose/poly(acrylonitrile). Here T was measured by DSC and dynamic
g

mechanical analysis. The lower curve is the result from DSC measurements, while the
two upper curves correspond to the T obtained from tan δ and from the loss modulus
g

E " measurements. The DSC method is much less sensitive than the dynamic mechanical
analysis and is unable to detect T for compositions higher than 50%. The data show a
g

single T for the whole range of compositions as is expected for a miscible system.
g

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 255

CELLULOSE AND POLYVINYL PYRROLIDONE)

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

CELLULOSE AND POLYVINYL ALCOHOL

August 31, 2012 | http://pubs.acs.org

Figure 1. Examples of hydrogen bonding interactions in cellulose/synthetic

polymer blends.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 256

ι 1 1
260 - -
ι
ι
L
ι;
11
>:
220 - '-
r
ι'
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

II
'/ *
t _

180
August 31, 2012 | http://pubs.acs.org

100

I 1 • • ι
0
25 50 75 100
Cellulose (wt%)

Figure 2. Plot of glass transition temperature against cellulose content in the

cellulose/poly(acrylonitrile) blends. Filled circles correspond to data
from tan δ measurements, solid triangles are data from E"
measurements, and squares are data from d.s.c. measurements.
Extrapolation of the data to 100% cellulose content indicates that the
T of cellulose lies in the range 240-260°C. Reproduced with
g

permission from reference 11.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 257

Below 50 wt. % cellulose there is only a modest change in T with composition

g

indicating a relatively low degree of miscibility, but above 50 wt. % cellulose, T g

increases dramatically indicating a good state of miscibility. By extrapolating the data

to 100% cellulose we obtain a T of pure cellulose of 240 - 260 °C, which corresponds
g

very well with the T of cellulose obtained by other methods. It is quite likely that the
g

good state of miscibility at higher cellulose contents is driven by hydrogen bond

formation between the C N functionality of the P A N and the OH groups of the cellulose
chains. As the cellulose content decreases, the total number of nitrile groups eventually
exceeds the number of hydroxyl groups (available for hydrogen bonding) and dipole-
dipole association between pairs of nitrile groups becomes progressively dominant.
Next we consider the T composition behavior for blends of cellulose with poly
g

(4-vinyl pyridine). In this blend system we expect that there should be hydrogen
bonding interaction between the cellulose hydroxyls and the nitrogen of the pyridine
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

rings. The blends were prepared by dissolving the two polymers separately in the
DMSO/paraformaldehyde solvent. The two solutions thus separately prepared were
mixed in appropriate ratios to give blends with different compositions. Blend films were
cast from the blend solutions at room temperature over a period of about 4 hrs. In a first
series of blends, the cast films were dried at 125 °C overnight in vacuum. In the second
series of samples the cast films were steeped in ammonium hydroxide solution, washed
in water, and dried in vacuum at 125°C. (Both series of blend films were optically clear,
showing no sign of phase separation.) It should be noted that the cast heat treated film
is really a methylol cellulose/DMSO complex with a low degree of methylol substitution
(0.07), while the other is a true cellulose/poly (4-vinyl pyridine) blend.
Figure 3 shows the Tg/composition data for the two sets of blends as determined
from dynamic mechanical measurements of E " . The upper data points (squares)
correspond to the cellulose/poly(4-vinyl pyridine) blends (CELL/P VPy) while the lower
August 31, 2012 | http://pubs.acs.org

points are from the methylol cellulose/poly(4-vinyl pyridine) (MC/P VPy) blends. The
4

solid lines are the T /composition dependence predicted by certain semi-empirical

g

equations from a knowledge of the T 's of the components and their weight fractions.
g

It can be seen that the theoretically predicted T 's fit the data very well. As seen in the
g

figure, the T results for the MC/Ç V P y blend pair are different from those of the
g

CELL/P VPy pair. For MC/P VPy, the Τ o f the blends falls below the calculated weight
4 4 g

average values of the T 's of the components (negative deviation), while for
g

CELL/P VPy the T of the blends is higher than the corresponding weight average values
4 g

(positive deviation). Several authors have shown that deviations of the Tg/composition
curve can be related to the strength of the interaction between the blend components.
Large negative deviations are associated with weak interactions, while a positive
deviation has been interpreted as an indication of very strong interactions. Thus the
results in the present case suggest that the components of the CELL/P VPy pair interact
4

more strongly than they do in the MC/P VPy pair. 4

The next case of interest is that of blends of cellulose with poly(vinyl alcohol)
P V A . This is an example of a crystalline/amorphous system because the blend films
were regenerated in non aqueous media, with the result that the cellulose component in
the blend is predominantly amorphous (12). It is expected that hydrogen bonding
between the OH groups of the two polymers should drive the system to thermodynamic

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 258

260
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018
August 31, 2012 | http://pubs.acs.org

0.0 0.2 0.4 0.6 0.8 1.0

Polysaccharide Weight Fraction

Figure 3. Theoretical T of MC/P VPy and CELL/P VPy blends as a function

g 4 4

of composition (full line), calculated to give the best fit to the E"
data points. The broken line is the tie line representing the weight-
average values. Note that the T of C E L L was not actually measured;
g

the estimated value of 250°C was used for the calculations.

Reproduced with permission from reference 14.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 259

miscibility. If such interaction occurs it is anticipated that it should be reflected in the

melting temperature of the P V A in the blends. To this end, the fusion and crystallization
behavior of the P V A in the blends was studied by DSC as a function of blend
composition. Figure 4 shows the fusion behavior of the specimens. The pure P V A
sample gives a large and sharp melting endotherm with a peak maximum at around
230 °C. As cellulose is blended with P V A up to 60 wt. %, the endothermic peak of P V A
tends to lose its prominence with an accompanying depression in the T values. In m

blends containing more than 70 wt. % of cellulose, it becomes difficult to detect the
melting endotherm of P V A in the DSC curves. Table I shows the actual melting
temperature T of the P V A in the blends, and the heats of fusion per gm of sample as
m

quantitatively assessed by measuring the area under the peaks. It is seen that in going
from 0 to 60 wt. % cellulose there is a melting point depression of more than 30°C,
which is in fact very large. Simultaneously there is a rapid decrease in the heat of fusion
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

of the P V A , which implies that there is a strong decrease in the degree of crystallinity
of the P V A component due to blending with cellulose.

Table I. Melting Temperature, T , and Heat of Fusion, AHf, of Cellulose/PVA

m

Blends Measured by DSC. Reproduced with permission from ref. 12.

Cellulose/PVA Τ °c A H , cal/g
f

w/w
0/100 230.1 18.7
10/90 226.8 14.7
20/80 224.5 12.4
30/70 220.3 9.4
40/60 212.9 6.2
August 31, 2012 | http://pubs.acs.org

50/50 205.6 4.0

60/40 197.0 2.0
70/30 -189 -
80/20 - -0

The phenomenon of the strong depression of the melting temperature of P V A in

blends with cellulose can be explained in terms of thermodynamic mixing accompanied
by an exothermic interaction between a crystalline polymer and an amorphous polymer.
The well-known Flory-Huggins equation for the melting point depression in
crystalline/amorphous polymer blends can be written as

e 0 2
AT = T n . T = -T (V /AH )Bv
n I B 2 l 2 u I

B = RT °(x /V )
m 12 lu

where R is the gas constant. Here the subscripts 1 and 2 are used to designate the
amorphous and crystalline components respectively, T ° is the melting point of pure m

crystalline polymer 2, T is the melting point of the mixture, V is the volume fraction
m

of the amorphous component, V is the molar volume of the repeating units, A H is the
u U

enthalpy of fusion per mole of repeating unit, Β is the interaction energy density of the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 260

-ι 1 1 Γ
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018
August 31, 2012 | http://pubs.acs.org

Temperature ro

Figure 4. DSC thermograms for a series of cellulose/PVA blends. The broken-

line curve is a thermogram obtained for a mechanical blend (80+20)
of fine powders of both polymers. The sensitivity of the scans for the
samples containing more than 40 wt % cellulose is twice that for the
others. Reproduced with permission from reference 12.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 261

two polymers, which is related to the thermodynamic interaction parameter χ . This Ι2

2
equation indicates that a plot of AT versus v, should be linear with a zero intercept. As
m

seen in Figure 5, the straight line fits the observed values well and yields a slope of
98.2°C and an intercept of 2.9°C. The deviation of 2.9°C is attributed to a residual
entropie effect which was neglected in deriving the equation. The interaction parameter
Xi2 derived from these results assumes the large negative value of -0.985 (at 513 K)
which is 2 to 5 times larger than values specified in the literature for other polymer pairs.
This leads to the conclusion that there is a high degree of interaction between the P V A
and the cellulose chains.
Finally, reference is made to some results with the blend system
cellulose/poly(vinyl pyrrolidone) (CELL/PVP). In this case a single T was observed at
g

every composition, strongly suggesting that the system is miscible. Furthermore, FTIR
analysis indicated that the interaction between the two polymers involves the carbonyl
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

functionality of PVP. N M R measurements have provided two types of information.

, 3
Firstly, as shown in Figure 6, solid state C N M R spectra were obtained for the pure
homopolymers and for blends of various compositions. The spectra for the blends are
not a superposition of the spectra of the constituent homopolymers, because shifts in
resonance frequencies are observed. This indicates that the two polymers are interacting
with one another. The largest spectral shift is observed for the carbon of the carbonyl
functionality of the PVP which occurs at about 175 ppm. This immediately indicates that
it is this functionality that is involved in the intermolecular interaction, probably from
hydrogen bond formation with the hydroxyl groups of cellulose. The second piece of
information provided by the N M R measurements relates to the scale of mixing. When
the state of miscibility of a blend is discussed, the question of the scale of mixing has to
be taken into consideration. A particular blend may be characterized as miscible with
one technique and immiscible with another. For example, visual determination of optical
August 31, 2012 | http://pubs.acs.org

clarity establishes the absence of domains exceeding 200-300 nm. The T measured by
g

DSC is sensitive to domain sizes of 35-40 nm, while dynamic mechanical analysis is
more sensitive with an upper limit of about 15 nm. From N M R measurements it is
possible to estimate domain sizes of a few angstroms to a few tens of nanometers. Thus,
using the solid state N M R apparatus it has been possible to measure the so-called spin
lattice relaxation time in the rotating frame (T ) for the pure homopolymers and for the
lp

homopolymer components in the blends. The results are shown in Table II. It is seen that
the two components have identical relaxation times which are significantly different
from those of the pure unblended homopolymers. This also means that there is extensive
mixing of the two polymers. The scale of mixing can be estimated from the relaxation
times. It turns out that in the blends the size of the domains is about 2.5 nm. This small
domain size means that a very intimate level of mixing has been achieved in these
blends. Similar N M R studies on cellulose/poly(acrylonitirile) and cellulose/poly(4-
vinylpyridine) indicate that they are also mixed on a very fine scale (i.e., with average
minimum domain dimensions in the 3-15 nm range (19). In a sense, this is somewhat
surprising because cellulose is a semirigid polymer while the synthetic polymers with
which it has been blended so far are flexible. Flory has predicted that in such a mixture
the ordering of the semirigid chains would reject the flexible coil molecules rendering
the mixture incompatible. The attainment of a very intimate level of mixing in the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 262

0 I • L_
Ο 0.1 0.2 0.3 0.4
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

Figure 5. Depression of melting point of P V A in cellulose/PVA blends as a

function of volume fraction of cellulose, plotted according to eq. 1.
Reproduced with permission from reference 12.
August 31, 2012 | http://pubs.acs.org

200 150 100 50 0

Figure 6. CP-MAS spectra of PVP, cellulose, and three blends. The labelled
peaks correspond to the carbons identified in the chemical structure.
The peak identified as S in the cellulose spectrum corresponds to
bound DMSO. Note that the methylol adduct (C OH) can also be on
7

0 H or 0 H . Reproduced with permission from reference 13.

2 3

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 263

cellulose/synthetic polymer blends is probably due to the introduction of strong specific

interactions between the components (21,11).

Table II. Proton T for Solid Films of Cellulose and PVP in their Blended and
l p

Unblended States. Reproduced with permission from reference 13.

a
T ms lp

Cellulose/PVP Blend Cellulose PVP

0/100 10.8
30/70 7.3 7.3
50/50 6.2 6.3
80/20 4.2 4.2
100/0 4.2
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

a
±5%

Conclusions
Cellulose/synthetic polymer blends are not as intractable as they were once
thought to be. Perfectly respectable blends can be obtained, and their behavior is entirely
analogous to that of the purely synthetic polymer blends. The presence of cellulose in
the blends does not seem to lead to any unusual or unexpected behavior. To date several
synthetic polymers that form good miscible blends with cellulose have been found, and
there seems little doubt that others will be found if we look for them. However, it now
seems important to focus on commercially interesting systems. For this purpose we have
to recognize that the method of blending used so far is totally unsuitable for industrial
applications. The solvent systems (DMSO/paraformaldehyde) or (DMAc/LiCl) are far
August 31, 2012 | http://pubs.acs.org

too expensive and not easily recycled. For commercial applications a more suitable
solvent for mixing would be N M M O . Possible applications could be the areas of textile
fibers and packaging materials. For example garments made from polyesters have poor
comfort properties but good wrinkle resistance, whereas the opposite is true for
cellulose. By using a blend it might be possible to obtain the best characteristics of each
polymer in a single fiber. This is where the challenge now lies.

Literature Cited

1. Paul, D.R. and Newman, S. (Eds.) "Polymer Blends", Academic press, New
York, 1978.
2. Olabisi, O., Robeson, L.M., and Shaw, M.T., "Polymer-Polymer Miscibility",
Academic Press, New York, 1979.
3. Hudson, S.M., and Cuculo, J.Α., J. Macromol. Sci. - Rev.Macromol.Chem.
1980, C18, 1.
4. Gagnaire, D., Mancier, D., and Vincendon, M., J. Polym. Sci. Polym. Chem.
Edn. 1980, 18, 13.
5. Turbak, A.F., Hammer, R.B., Davies, R.E., and Hergert, H.L., Chemtech, 1980,
51.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 264

6. McCormick, C.L., and Shen, T.S., "Macromolecular Solutions" (Eds. R.B.

Seymour and G.A. Stahl), Pergamon Press, New York, 1982, pp. 101-107.
7. Jolan, A . H . , and Prud'homme, R.E., J. Appl. Polym. Sci. 1978, 22, 2533.
8. Seymour, R.B., Johnson, E.L., and Stahl, G.A., "Macromolecular Solutions",
(Eds. R.B. Seymour and G.A. Stahl), Pergamon Press, New York, 1982, pp. 90-
100.
9. Field, N.D., and Song, S.S., J. Polym. Sci., Polym. Phys. Edn. 1984, 22, 101.
10. Field, N.D., and Chien, M.-C., J. Appl. Polym. Sci. 1985, 30, 2105.
11. Nishio, Y., Roy, S.K., and Manley, R.St.J., Polymer 1987, 28, 1385.
12. Nishio, Y., and Manley, R.St.J., Macromolecules 1988, 21, 1270.
13. Masson, J-F., and Manley, R.St.J., Macromolecules 1991, 24, 6670.
14. Masson, J-F., and Manley, R.St.J., Macromolecules 1991, 24, 5914.
15. Bélorgey, G., Prud'homme, R.E., J. Polym. Sci., Polym. Phys. Ed.1982,20,191.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch018

16. Bélorgey, G., Aubin, M . , and Prud'homme, R.E., Polymer 1982, 23, 1053.
17. Kwei, T.K., J. Polym. Sci., Polym. Lett. 1984, 22, 307.
18. Pennacchia, J.R., Pearce, E . L . , Kwei,T.K., Bulkin, B.J., and Chen, J.-P.,
Macromolecules 1986, 19, 973.
19. Flory, P.J., Macromolecules 1978, 11, 1138.
20. Wang, L.F., Pearce, E . M . , and Kwei, T.K., Polymer 1991, 32 (2), 249.
21. Painter, P.C., Tang, W.-L., Graf, J..F., Thomson, B., and Coleman, M . M . ,
Macromolecules 1991, 24, 3929.
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 19

Studies of the Molecular Interaction Between Cellulose

and Lignin as a Model for the Hierarchical Structure
of Wood

1 2 3
Wolfgang G. Glasser, Timothy G. Rials , Stephen S. Kelley , and Vipul Davé
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

Biobased Materials/Recycling Center and Department of Wood Science and Forest

Products, Virginia Polytechnic Institute and State University,
Blacksburg, VA 24061

Wood and dietaryfiberproducts all belong to a class of biomolecular

composites that are rich in cellulose and lignin. The interaction
between cellulose and lignin determines such properties as
mechanical strength (wood); creep, durability and aging; cellulose
purity (pulp); and digestibility (nutrients). The understanding of the
interaction between cellulose and lignin can be approached from
various types of analyses involving the natural biocomposites, or it
August 31, 2012 | http://pubs.acs.org

can be explored by studying the physical mixtures of the two types

of macromolecules. The latter can be prepared by mixing the
respective polymers in solid, solution or melt form within the
constraints of solubility and melt-flowability. Such mixtures have
been examined, and the results suggest that cellulose and its
derivatives form two distinct phases with lignin and its derivatives; a
crystalline polysaccharide-phase and a continuous amorphous phase
that provides evidence for strong intermolecular interaction between
the two components. In addition, results suggest that lignin and/or
its derivatives are capable of contributing to the supermolecular
organization of cellulose (derivatives). The interaction between
lignin and cellulose varies in relation to chemical differences as well
as molecular parameters. The results are consistent with the view
that the hierarchical structure of the natural biocomposite wood is
not only the consequence of a sequence of biochemical events, but

1
Southern Forest Experiment Station, U.S. Forest Service, Pineville, LA 71359
2
National Renewable Energy Laboratory, Golden, CO 80401
3
Johnson and Johnson, Skillman, NJ 08558

©1998 American Chemical Society 265

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 266

that it is the result of various thermodynamic driving forces that are

independent of the biosynthetic origin.

Hierarchical structures "are assemblages of molecular units or their aggregates that

are embedded or intertwined with other phases, which in turn are similarly
organized at increasing size levels" (1). It is the multimolecular combination of
virtually all biological materials that is responsible for the multilevel architectures
that confer the unique properties to the composites of nature. Wood (or more
generically, "lignocellulose") is a complex material on all dimensional levels of the
structural hierarchy, from the nano- to the millimeter-scale. Lignocelluloses are
mixtures of crystalline and non-crystalline polysaccharides with lignin that are
assembled into a structural architecture in which the interaction between the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

biomacromolecules is dictated by the specific sequence of biochemical events that

take place during biosynthesis (i.e., plant growth). The resulting multilevel
architecture determines all properties, regardless of whether these are mechanical,
chemical, sorptive, nutritive, Theological, or degradative in nature. The
understanding of hierarchical molecular structures in biological systems is beginning
to be taken as a guide for the development of new, man-made materials (1). Several
models have been advanced that describe wood (and lignocellulose) as a multiphase
material that achieves its remarkable fracture toughness on the basis of the need to
create an almost infinitesimal new surface area during fracture (2). It is the creation
of interfibrillar cracks during mechanical failure which prevents fiber pullout at all
levels of moisture sorption or temperature (2).
Single-phase materials, uniform polymers, often suffer from low impact
strength and low dimensional stability when heated. New material properties are
August 31, 2012 | http://pubs.acs.org

achieved when two or more types of molecules are blended or mixed. The resulting
morphology of these mixtures is a direct result of the method of blending and the
specific chemical and molecular interactions (3). Impact strength and resistance to
deformation at elevated temperature rise when mixtures of macromolecules with
distinct phases remain molecularly intertwined.
This paper reviews and summarizes a series of experiments designed to
explore the specific molecular interactions between cellulose and lignin and their
respective derivatives. Man-made blends of these biopolymers are to be compared
with the natural biocomposite with a view towards determining the nature of the
interaction between the two components. While it is evident that these interactions
are also operative during the process of wood formation (i.e., lignification), between
lignin precursors and the polysaccharide matrix, this paper is limited to polymer-
polymer interaction arguments, exclusively. However, the reader is referred to the
vast body of literature dealing with the biosynthetic aspects of the creation of
molecular interactions between polysaccharides and lignin, such as the recent book
by Jung et al. (4).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 267

Experimental Section

Materials: A l l cellulose and cellulose derivatives were obtained as chemically

pure, commercially available materials. Lignin and lignin derivatives were obtained
from Aldrich Chemical Company except for derivatives described in the primary
literature as indicated. Solvents were used as provided from chemical suppliers.

Methods: 1. Blends. Blends of cellulose with lignin were prepared by mixing

cellulose solutions in D M A c / L i C l with lignin dissolved in D M A c in accordance
with earlier work (5-10). Cellulose derivatives, hydroxypropyl cellulose, ethyl
cellulose and cellulose mixed esters (CAB), were blended with lignin by using both
melt and solution mixing. Common solvents were pyridine, dioxan and acetone.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

This work has been described in detail elsewhere (5-7). Melt-blended specimens
were produced by injection molding.
2. Thermal Analysis: Thermal analysis was conducted using differential
scanning calorimetry (DSC) and/or dynamic mechanical thermal analysis ( D M T A )
with either thin films (from solvent casting) or melt processed test specimens (dog
bones from injection molding).
3. Other Characterization Methods: Ultimate strength was determined by
tensile tests using an Instron tensile tester. Dynamic viscosity was determined on a
Rheometrics Mechanical Spectrometer (RMS 800) using 20% (w/w) solutions in a
parallel-disk geometry (10). Transmission electron microscopy was conducted on a
Jeol SEM-100CX-II electron microscope. All methods have been described before
where indicated.
August 31, 2012 | http://pubs.acs.org

Results and Discussion

L Wood

The non-crystalline component of wood, which is responsible for the

viscoelastic nature of this natural material, gives rise to a variety of responses
during heating (11-13). Whereas only a single, broad transition can be detected for
native wood at low moisture content, a distinctly trimodal distribution of damping
transitions (tan ô-peaks) can be detected at elevated moisture content (Figure 1).
Whereas one transition (β) has no discernible impact on the storage modulus, and
can safely be attributed to local site exchange of moisture, the occurrence of two
distinct glass transition temperatures at which mechanical damping occurs suggests
the existence of two different non-crystalline molecular entities. These undergo
separate and independent glass to rubber transitions. Specifically, the existence of
distinct tan ô-transitions in moist wood, at 30% moisture content, at -10°C (a ) 2

and 60°C (a,) (11), suggests the presence of two different molecular components
that each reside in their independent phase on the molecular (nano-) level. B y
investigating the moisture response of these two glass transitions, and by

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 0.1
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

t a n ft

0.05

-150 •50 50
August 31, 2012 | http://pubs.acs.org

150
TEMPERATURE CO

1. D M T A spectra (tan ô-transitions) of a section of solid spruce wood

recorded at moisture contents rising from 5% (bottom) to 10, 20, and 30%
(top). Peaks a , ct , and β reflect large-scale segmental motion and a
x 2

secondary dispersion (β-transition) characteristic of glass transitions and

site exchange of water, respectively. The T s were assigned to
g

hemicelluloses (a ) and lignin (αϊ). According to ref. 11.

2

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 269

comparing them to isolated lignin preparations using the model of Kwei (14), a, and
a were attributed to the presence of the molecular phases representing lignin and
2

hemicelluloses, respectively (11). However, it needs to be pointed out that the

existence of separate phases as indicated by D M T A is not inconsistent with the
existence of primary or secondary bonds between those phases. Block copolymers
between thermoplastic cellulose derivatives and lignin were also found to exhibit
3
phase distinctions with the individual blocks having molecular weights as low as 10
daltons (15, 16). There is strong evidence suggesting that the two non-crystalline
components of wood, hemicellulose and lignin, are covalently linked in block
copolymer fashion (17) and each of these two polymeric phases undergo an
independent glass to rubber transition at a different temperature. Due to the
hydrophilic nature of these phases, their transition temperatures depend on
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

moisture content; and so does our ability to observe these transitions.

IL Cellulose Derivative/Lignin Blends

Any attempt at recreating wood's native structure by solvent or melt

processes is complicated by the variability of the chemical structure of the matrix
and the intractability of the cellulose in terms of solubility and melt properties.
This limitation may be overcome in part by chemical modification in the form of
derivatives. While it is recognized that this severely constrains the realism of the
model, it does provide insight into the chemical and physical interaction that can be
formed in a binary blend of lignin and cellulose derivatives.
The state of miscibility of a polymer pair is commonly evaluated by
studying T -behavior in relation to the volume fraction of the respective polymers.
August 31, 2012 | http://pubs.acs.org

Phase-separated mixtures exhibit the T s of the individual parent homopolymers

g

while a single transition intermediate to the values of the individual homopolymers

indicates miscibility. Partial miscibility is indicated by the T s migrating towards a
g

single, common transition in relation to fractional mixing. The DSC-thermograms of

a series of hydroxypropyl cellulose (HPC)/lignin (L) blends (prepared by injection
molding) provide evidence for strong intermolecular interaction (Fig. 2) (5). A
single T is observed for blends with a lignin content of up to 55%, and this T rises
g g

with lignin content.

Dynamic mechanical thermal analysis (DMTA) of this same series of
HPC/L blends (Fig. 3) reveals a similar elevation in temperature of damping
transitions with lignin content rising. The two tan ô-transitions of the pure H P C -
spectrum (at 30 and 85°C) have been assigned to the T s of an amorphous phase
g

and that of an organized, liquid crystalline (LC) mesophase, respectively (18). The
effect of lignin causes a significant reduction of the temperature range over which
the tan ô-transition occurs, with an apparent L-association with the L C mesophase
to the exclusion of the lower-temperature amorphous phase (5). This behavior is in
conflict with the normal effect of multiphase materials on thermal transitions (19).
Normally, the glass to rubber transitions of a mixture of two non-crystalline

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019
August 31, 2012 | http://pubs.acs.org

10 40 70 100 130 160 190

TEMPERATURE (°C)

Fig. 2. DSC thermograms of solvent-cast films consisting of mixtures of

hydroxypropyl cellulose (HPC) and organosolv lignin (L). Lignin content as
indicated by the numbers of each tracing (0 to 100%). Both T and T g m

show variation in relation to blend composition. T has been attributed to a

2

liquid crystalline mesophase present in HPC (18). According to ref. 5.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 271

polymers broaden on the temperature-scale, and they may stretch over the entire
region in which the constitutive components undergo thermal transitions in pure
state if they are partially miscible. In the 55% L-content blend, a material with
more highly ordered morphology is indicated (Fig. 3). The addition of lignin
apparently contributes to the enhancement of HPC's L C mesophase at the expense
of an amorphous phase.
A similar rise in relaxation intensity and simultaneous decrease in breadth of
the tan 5-transition was observed in blends of HPC with a partially ethylated lignin
(EL) (Fig. 4) (6). However, in this case the improved uniformity of phase response
results in the complete disruption of the supermolecular structure of the HPC and
the formation of a seemingly continuous, miscible, amorphous blend of H P C and
EL.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

Different observations are made with blends of ethyl cellulose (EC) and L
where the addition of a second, immiscible molecular component produces the
expected broadening of the tan 5-transition (Fig. 5) (7). Since the resulting two T - g

transitions, however, are found at temperatures below and above those of the
respective parent (pure) components, the appearance of two separate phases is
explained with the formation of a supermolecularly ordered (discrete) phase with a
T above that of either parent constituent in addition to a lower-T (uniform)
g g

amorphous phase. The creation of an L C mesophase architecture by the addition of

lignin is also supported by an increase in storage modulus upon passage through T g

(not shown, refer to ref. 7).

An examination of the melting point of the HPC component in relation to
lignin content revealed (5) that T migrates in relation to the volume fraction of
m
August 31, 2012 | http://pubs.acs.org

lignin. The degree of melting point depression was approximately dependent on

lignin volume fraction and a polymer-polymer interaction parameter, Β (20). For
blends with components whose molecular weight is >2,000, the melting point
depression is approximately related to Β according to

0 -BV2u ο 2
Τ ml - Tm2 - — — T <Pl
ώ

m2
DLH2U
(1)

where the subscript 2 refers to the crystallizable component (i.e., HPC), Tii is its
equilibrium melting temperature, A H / V is its heat of fusion per unit volume of
2u 2u

repeat unit for 100% crystalline material, V is its molar volume, and φ is its
2 2

volume fraction in the blend. Β can be directly evaluated from the slope of a plot of
(Til - Tmi ) vs. Φ\ (Fig. 6) (21). Working with lignin derivatives in which hydroxyl
groups were selectively removed (by acetylation or ethylation), a relationship
between Β and the phenolic hydroxyl content was established (6). The interaction
parameter was lowest (and intermolecular interaction most favorable) when the
phenolic hydroxyl content was approximately 0.25 per phenylpropane repeat unit

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 272

0.40

tan δ fh
Π it
I :
0.24
LIGNIN CONTENT
INCREASING

0.08
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

-50 0 50 100
TEMPERATURE ( C) e

Fig. 3. D M T A Spectra (tan ô-transitions) of melt-processed HPC/L blends

containing, 0 (-), 5 (—), 20 ( ), 40 (-.—), and 55% ( ) L-content.
The narrowing of the tan δ-peak with rising L-content suggests an increasing
degree of molecular interaction (According to ref. 5).
August 31, 2012 | http://pubs.acs.org

Fig. 4. D M T A spectra (tan ô-transitions) of solvent-cast blend films (dioxan) of

HPC with ethyl lignin (EL). EL-content is indicated by the numbers of each
tracing (5 to 20%). The progressive narrowing of the tan ô-peak with rising
EL-content suggests enhanced molecular interaction with L-content rising.
According to ref. 6.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 273

0.5

0.4
/ι
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

0.3h
00
ζ LIGNIN CONTENT
< INCREASING
0.2

0.1
August 31, 2012 | http://pubs.acs.org

100 125
TEMPERATURE t°C)

Fig. 5. D M T A spectra (tan ô-transitions) of solvent-cast blend films (dioxan) of

ethyl cellulose (EC) with organosolv lignin (L). L-content rises from 0 (—)
to 10 (--), 20 (-.-.), and 40% (—). The separation of the tan ô-peak into
two distinct transitions is as expected for a molecularly immiscible blend;
however, since the higher-temperature transition (at ca. 147°C) is higher
than the two parent polymer components, this is attributed to a
supermolecularly-ordered phase that is separated from a uniform
(continuous) amorphous phase. (Lignin-T is at 95°C). According to ref. 7.
g

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 274

50 Γ
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019
August 31, 2012 | http://pubs.acs.org

10 12 14 16
φ (χΙΟΟ)
I
2
Fig. 6. Relationship between ( Τ ^ - Τ,^) and φ from eq. 1 for HPC/L blends
prepared from dioxan (O), pyridine ( · ) , and from melt ( • ) . According to
ref. 5.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 275
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

-1

- Β
August 31, 2012 | http://pubs.acs.org

0-1 0-3 0-5

Phenolic OH (per C ) 9

Fig. 7. Relationship between polymer-polymer interaction parameter, B , and a

lignin structural feature (i.e., phenolic OH-content per C -repeat unit). The
9

lowest value for Β denotes greatest interaction between polymers.

According to ref. 6.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 276

in lignin (Fig. 7). This indicates that the lignin which has a structure (in terms of
functionality) similar to the one in native wood, is the one that provides for
conditions most favorable to strong intermolecular interaction between cellulose
(derivatives) and lignin, and this favors component miscibility (5-7).
In an attempt to further increase the phase compatibility between lignin and
cellulose (derivatives), block copolymers were synthesized which consisted of
covalently-linked lignin and cellulose ester segments (or "blocks") (15-16). It was
revealed that copolymer architecture, which normally enhances phase
compatibility, was unable to provide further improvements in lignin/cellulose
(derivative) blend compatibility (16). Glass transition temperatures were shifted
towards an intermediate temperature for low molecular weight copolymers, but
they did not migrate for high molecular weight components, when blended with
cellulose propionate (CP) regardless of whether the lignin was the component of a
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

lignin-CP block copolymer or not (not shown, refer to ref. 16). No improvement in
miscibility resulted from the modification of lignin by copolymerization with CP
segments.
The blend experiments involving cellulose derivatives and lignin suggest that
lignin disrupts both the ordered and the non-ordered forms of cellulose derivative
morphology by favoring the formation of an amorphous or liquid crystalline
mesophase structure through strong interactive association between lignin and the
polysaccharide component.

IIL Regenerated Fibers

The spinning of cellulose from an ordered solution state has become
August 31, 2012 | http://pubs.acs.org

commercial practice with the introduction of an N-methyl morpholine-N-oxide-

based solvent process (22). Cellulose ester derivatives can also be converted into
fibers by spinning from an anisotropic, liquid crystalline solution-state in a variety
of solvents (10, 23-26). The formation of anisotropic solutions of cellulose esters
in various solvents has been studied in detail (27). Continuous cellulose ester fibers
were found to exhibit the expected behavior in terms of mechanical (tensile)
properties (i.e., increased strength with increasing orientation) when spun from
biphasic or anisotropic solutions (24). The addition of lignin to cellulose and
cellulose ester solutions was found to impact the dynamic elastic modulus of
concentrated solutions differently: whereas the addition of lignin reduced the
dynamic elastic modulus of cellulose ester solutions at all levels of lignin content,
cellulose solutions (in DMAc/LiCl) became more viscous (Fig. 8). Considering that
isolated lignin has a molecular weight of only ca. l/100th that of cellulose and
cellulose esters, and lignin is usually considered to be a highly compact or spherical
molecule, its impact was expected to be one of viscosity-reduction. The fact that
the dynamic elastic modulus of cellulose, but not of cellulose ester, solutions
increased instead to decline at all shear frequencies is explained with strong
secondary interactions of lignin with cellulose in the D M A c / L i C l solvent system.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

10000
10000
LIGNIN CONTENT
CO CO
(L INCREASING
(/>
£L 1000
3 <
/
>
3
3
ΌΟ 3
ΌΟ
.ί£ 1000 ~ 100
ce I
S
LU

ε • É
A 0% L i g n i n 10
CO LIGNIN CONTENT
c
>» • 4% L i g n i n ο
Ο INCREASING

100
0.1 1 10 100 1000 0.1 1 10 100 1000
Frequency (rad/sec) Frequency (rad/sec)

In Cellulose Derivatives; Heinze, T., et al.;

Fig. 8. The dynamic elastic modulus of cellulose (left) and cellulose ester (CAB)
(right) solutions with lignin in D M A c (with LiCl added in case of cellulose),
in relation to lignin content at different frequencies. Whereas the presence
of lignin in the solution raises the modulus of the cellulose/lignin mixture at
all frequencies, lignin contributes to a reduction of solution modulus at all

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

frequencies in case of cellulose esters. According to ref. 10.

-a
 278

However, even when cellulose ester (cellulose acetate butyrate, CAB)/L mixtures
were spun into continuous fibers from D M A c solution, both fiber tensile strength
and modulus increased significantly (10). The strength and modulus (stiffhess)-
enhancing effect of lignin on cellulose ester fibers was limited to the initial 4%;
beyond 4% lignin content, no further positive effect of lignin on fiber strength was
noted (Figure 9). It is surprising that the addition of small amounts of (low
molecular weight, isolated) lignin neither interfered with the formation of
anisotropic solutions nor with the ultimate strength of the resulting fibers (10). A
tenacity-increasing effect of small amounts of lignin on cellulose ester fibers can be
explained only with a positive effect by lignin on the molecular order of the
cellulose derivative in solution and in solid state (10).
The propensity of cellulose esters to form liquid crystalline morphologies
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

has been observed previously (27). A recent study has indicated that the addition
of lignin enhances the formation of ordered structures in cellulose acetate butyrate
(CAB) (28). Experimental evidence suggests that, as solvent evaporates and both
constituents solidify, the surface of phase-separated lignin particles serves to create
cholesteric liquid crystalline order by nucleation (Figure 10) (28). The resulting
structure provides evidence that lignin phase-separates from cellulose ester
derivatives and becomes an integral part of a two-phase architecture in which the
degree of organization in the polysaccharide matrix is substantially increased at the
apparent expense of an amorphous phase. The addition of lignin was consistently
found to enhance the liquid crystalline mesophase order in non-crystalline cellulose
derivatives, and this order is often responsible for increased strength properties. A
similar phase-separated morphology also was found in blends of cellulose with
lignin (Fig. 11). This morphology reveals heterogeneity at the nano-level, and this
August 31, 2012 | http://pubs.acs.org

provides the basis for a structural hierarchy that has become the trademark of
biological materials, such as wood (1).

CONCLUSIONS

Results with blends of cellulose and cellulose derivatives with lignin suggest
that the two biopolymers are immiscible.
Experimental evidence supports the hypothesis that lignin enhances the
organization of non-crystalline cellulosic structures and gels, and that liquid
crystalline mesophase, ordered structures are created that result in multiphase
architectures. This enhanced heterogeneity often produces materials with higher
modulus and higher strength.
The effect of lignin in blends with cellulose esters is found to disrupt both
crystalline order and non-crystallinity by contributing to the formation of a
mesophase liquid crystalline morphology that produces an architecture on the
dimension of nanometers. This secondary order is held responsible for observed
strength gains in lignin/cellulose derivative blends.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 279

1.20

ω 1.00

0.20 1
' ' 1 1

0 5 10 15 20
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

LIGNIN CONTENT, %

Fig. 9 Relationship between CAB-fiber tensile strength and lignin content at

different draw ratios. (Draw ratios increase from 0.8, - · - , to 1.5, for
all lignin contents except 20%, where they ranged between 0.26 and 0.5,
respectively.) Tensile strength is seen to increase with lignin content rising
to 20%. This is unexpected since lignin has a molecular weight of only ca.
1/100th of that of cellulose derivative and is expected to reduce tensile
strength in relation to degree of dilution. According to ref 10.
August 31, 2012 | http://pubs.acs.org

Fig. 10. Transmission electron micrograph (TEM) of a cellulose ester (CAB) film
containing 20% lignin. The unstained film shows a phase-separated particle
that provides a nucleating surface for cellulose ester liquid crystals. The
well-ordered cholesteric arrangement was found to be distinctly more
pronounced in the presence of lignin. The periodicity between striation
lines was smallest on the surface of the lignin particle indicating lignin's
contribution to the cellulose ester's organization. According to ref. 28.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 280
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019
August 31, 2012 | http://pubs.acs.org

Fig. 11. Transmission electron micrograph (TEM) of the stained cross-section of a

cellulose/lignin blend fiber containing 4% (w/w) lignin. There is evidence for
an even dispersion of lignin particles in the size-range of 10-20 nm in
addition to much larger aggregates.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 281

The results suggest that the natural composite structure of lignified plant
materials is not only a consequence of a sequence of well-defined biochemical
events, but that it is also a consequence of the thermodynamic driving forces that
regulate the interaction between the most prevalent polymer constituents present in
lignocellulose, lignin and (cellulosic) polysaccharides.

ACKNOWLEDGMENTS

This report is based on studies financially supported by the National

Science Foundation (Washington, D.C.), the Center for Innovative Technology
(Herndon, VA), and the U S D A (Washington, DC).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

LITERATURE CITED

1. National Research Council, "Hierarchical Structures in Biology as a Guide

for New Materials Technology," National Materials Advisory Board, ed.,
National Academy Press, Washington, DC, 1994; pg. 1.
2. G. Jeronimidis, "Wood, One of Nature's Challenging Composites," in "The
Mechanical Properties of Biological Materials," SEB Symposium No. 34,
Cambridge Univ. Press, Cambridge, England, 1980, 169-182.
3. T. A. Oswald and G. Menges, "Materials Science of Polymers for
Engineers," Hanser Publishers, Munich Vienna New York, 1995, 475 pg.
4. H. G. Jung, D. R. Buxton, R. D. Hatfield, and J. Ralph, editors. "Forage
Cell Wall Structure and Digestibility," American Society of Agronomy, Inc.,
Madison, Wisc, 1993; 794 pg.
August 31, 2012 | http://pubs.acs.org

5. T. G. Rials, W. G. Glasser, J. Appl. Polym. Sci. 37, 2399-2415 (1989).

6. T. G. Rials, W. G. Glasser. Polymer 31, 1333-1338 (1990).
7. T. G. Rials, W. G. Glasser, Wood and Fiber Sci. 21(1),80-90 (1989).
8. V. J. H. Sewalt, W. de Oliveira, W. G. Glasser. J. Sci. Food Agric. 71, 204-
208 (1996).
9. G. Gamier, W. G. Glasser, Polymer Eng. Sci. 36, 885-894 (1996).
10. V. Davé, W. G. Glasser, Polymer 38, 2121-2126 (1997).
11. S. S. Kelley, T. G. Rials, W. G. Glasser. J. Mater. Sci. 22, 617 (1987).
12. L. Salmen, J. Mater. Sci. 19, 3090 (1984).
13. A.-M. Olsson, L-Salmén, Chapter 9 in "Viscoelasticity of Biomaterials," W.
Glasser and H. Hatakeyama, eds., ACS Symp. Ser. 489, 133-143 (1992).
14. T. K. Kwei, J. Polym. Sci.: Polym. Lett. 22, 307 (1984).
15. W. de Oliveira, W. G. Glasser. Macromolecules 27, 5 (1994).
16. W. de Oliveira, W. G. Glasser. Polymer 35(9), 1977-1985 (1994).
17. N. Terashima, K. Fukushima, L.-F. He, K. Takabe. Chapter 10
"Comprehensive Model of the Lignified Plant Cell Wall," in "Forage Cell
Wall Structure and Digestibility," H. G. Sung, D. R. Buxton, R. D. Hatfield,

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 282

J. Ralph, Eds., Amer. Soc. Agronomy, Inc., Madison, Wise, 247-270

(1993).
18. T. G. Rials, W. G. Glasser, J. Appl. Polym. Sci. 36, 749-758 (1988).
19. E. A. Turi, ed. "Thermal Characterization of Polymeric Materials,"
Academic Press, New York, 1981, 972 pg.
20. P. J. Flory. J. Chem. Phys. 17, 223 (1949).
21. J. E. Harris, D. R. Paul, J. W. Barlow. In "Polymer Blends and Composites
in Multiphase Systems," ACS Adv. Chem. Ser., No. 206, C. D. Han, ed.,
1984, 17.
22. S. A. Mortimer, A. A. Peguy, Cellulose Chem. Technol. 30, 117-132
(1996).
23. V. Davé, W. G. Glasser. In "Viscoelasticity of Biomaterials," W. G. Glasser
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch019

and Hatakeyama, eds., ACS Symp. Ser. No. 489, 144-165 (1992).
24. V. Davé, W. G. Glasser, G. L. Wilkes. J. Polymer Sci., Pt. B: Physics, 31,
1145 (1993).
25. V. Davé, W. G. Glasser. J. Appl. Polym. Sci. 48, 683 (1993).
26. V. Davé, J. Wang, W. G. Glasser, D. Dillard. J. Polym. Sci., Pt. B: Physics
32, 1105 (1994).
27. P. Zugenmaier, J. Appl. Polym. Sci.: Appl. Polym. Symp. 37, 223-238
(1983).
28. V. Davé, W. G. Glasser, G. L. Wilkes. Polymer Bulletin 29, 565-570
(1992).
29. J. R. Penacchia, E. M. Pearce, T. K. Kwei; B. J. Bulkia, J.-P. Chen,
Macromolecules 19, 973 (1986).
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 20

A Study of the Molecular Interactions Occurring

in Blends of Cellulose Esters and Phenolic Polymers

1 1 1 2 1
M . F. Davis , X. M . Wang , M . D. Myers , J. H. Iwamiya , and S. S. Kelley

1
Center for Renewable Chemical Technologies and Materials, National Renewable
Energy Laboratory, Golden, CO 80401
2
Advanced Technology Center, Lockheed Martin Missiles and Space,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

Palo Alto, CA 94304

Relaxation and spin diffusion measurements were performed on a series

of blends containing cellulose esters and phenolic polymers such as
poly(vinyl phenol) (PVP) and Novolac resins. These experiments were
used to determine the extent of molecular-level mixing between the
cellulose esters and phenolic polymers. Changes in the peak positions of
the phenolic hydroxyl carbon of PVP or Novolac and the carbonyl peak
of the cellulose ester were used to study the formation of hydrogen bonds
13 1
between the two different polymers. C CP/MAS and H CRAMPS
NMR spectroscopy was used to measure the degree of polymer mixing
August 31, 2012 | http://pubs.acs.org

on the molecular scale. Blends of cellulose acetate and phenolic

13
polymers were determined to be mixed on the molecular level. C
CP/MAS NMR spectroscopy suggested that in the case of cellulose
acetate butyrate/phenolic polymers blends that one (or more) butyryl side
group attached to a specific position along the anhydroglucose ring was
more efficient at forming hydrogen bonds with the phenolic polymers.

Cellulose esters are a class of commercially-important bio-based polymers used

for production of fibers, plastics and films. In many of these applications it would be
desirable to increase the mechanical strength and improve the melt processibility of the
cellulose ester while maintaining optical clarity. There are a number of reports that
suggest that the mechanical properties of polymers can be improved and clarity
maintained through the preparation of miscible or homogeneous polymer blends (/).
Blends of cellulose polymers with synthetic polymers have been studied by
several researchers. This work includes mixtures of unmodified cellulose with synthetic
polymers (2) and blends of synthetic polymers with various cellulose derivatives,
including cellulose esters (3-5) and cellulose ethers (6). Several of these blends appear

©1998 American Chemical Society 283

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 284

to be uniformly mixed at the molecular level (2,6), while others appear to be somewhat
heterogeneous on a molecular scale (3-5). In the case of the uniformly mixed or miscible
blends the mixing seems to be driven by hydrogen-bonding interactions between the
cellulosic component and the synthetic polymer.
Miscible polymer blends can be created by selecting polymers that can exploit
specific interactions between the different polymers. In order for two or more polymers
to form a miscible blend, the free energy of mixing, A G , must be negative and the
m

second derivative of A G with respect to the composition must be positive (7). A

m

reduction in the A G for specific combination of polymers is generally driven by

m

secondary interactions, such as a dipole-dipole interaction or a hydrogen-bonding

interaction, which can cause the enthalpy of mixing to be negative or exothermic. The
entropie contribution to the A G will also be slightly negative resulting in the formation
m

of a miscible blend. In the case of blends of cellulose and synthetic polymers the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

secondary interaction appears to be between the hydroxyl groups on the cellulose and
hydroxyl or amide groups on the synthetic polymer. The hydroxyl groups of the cellulose
acts as a hydrogen-bonding donor in these blends. However, in the case of cellulose
ethers, the hydrogen bonding appears to be between the ether linkages within or between
the anhydroglucose rings and hydroxyl groups on the synthetic polymer. The cellulose
ether then acts as a hydrogen-bonding acceptor.
Phenolic polymers such as polyvinylphenol (PVP) or Novolac resins are also
known to form miscible blends with several polymers, including ester-containing
polymers, e.g., acrylates and methacrylates (7-/7), and ether-containing polymers, e.g.,
polyvinyl methyl ether) and polyethylene glycols (12,13). The miscibility of
PVP/methacrylate blends are very sensitive to changes in the chemical structure of the
methacrylate polymers. For example, poly(methyl methacrylate) and poly(methyl
ethylacrylate are miscible with PVP while poly(methyl butylacrylate) may or may not be
August 31, 2012 | http://pubs.acs.org

miscible, depending on its tacticity. Based on these results, it appears likely that miscible
blends of a variety of cellulose esters and phenolic polymers can be created by exploiting
the hydrogen-bonding interaction between the phenolic hydroxyl group and either the
hydroxyl groups of the cellulose backbone, the ring oxygen, the glycosidic oxygen, or
the carbonyl of the substituted ester group. These possible interactions are shown below
in Scheme 1.
Recently, in our laboratory, blends of cellulose esters and phenolic polymers have
been examined using Differential Scanning Calorimetry (DSC) and Fourier Transform
Infrared Spectroscopy (FTIR). Cellulose acetates and cellulose mixed esters blended
with either PVP or Novolac resins show a single glass transition temperature (Tg)
intermediate between the two pure components. The single Tg varies with the
composition of the blend. This observation is consistent with a well-mixed blend,
although inhomogeneities on the 10-20 nanometer scale are not easily detected with
DSC. Analysis with FTIR spectroscopy shows changes in the both the hydroxyl stretch
l
(3000-3800 cm "') and the carbonyl stretch (1700-1800 cm ). These results are
consistent with secondary interactions between the phenolic hydroxyl hydrogen and the
carbonyl oxygen. Smaller, more subtle changes were also observed in the ether
stretching region. As the acetate side groups were substituted with butyrate side
groups, it was observed that the hydrogen bonding interaction became weaker as
evidenced by smaller shifts observed in the FTIR spectra.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 285
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

Scheme 1. Possible hydrogen bonding interactions (represented by

dashed lines) that could occur in cellulose acetate/poly(vinyl phenol)
blends.

Experimental

All of the cellulose esters were commercial samples obtained from Eastman
Chemical Co., Kingsport, TN. The cellulose acetate (CA) was C A 398-30 which has a
degree of substitution (D.S.) of acetyl groups of 2.45. The cellulose acetate butyrate
(CAB) was CAB-381-20 with a acetyl D.S. of 0.4 and a butyryl D.S. of 2.2. The PVP
and polystyrene were obtained from Polymer Scientific and had nominal molecular
weights of 15,000 daltons. The Novolac polymer was obtained from Plastics Engineering
Co., Sheborgan, WI. Solution N M R was used to confirm the chemical composition of
August 31, 2012 | http://pubs.acs.org

these materials.
The blends were created by dissolving both components in acetone at 5% solids.
All of the solutions were transparent. The solutions were cast onto glass plates and the
solvent allowed to evaporate. A l l of the resulting films were clear except the
CA/poly(styrene) blend. Residual acetone was removed by heating under vacuum at 60
C for 1 hour.
, 3
C cross-polarization/magic-angle spinning (CP/MAS) N M R spectra were
obtained using a Bruker DSX-200 spectrometer (4.7T field strength), 1 ms contact
time, and a *H π/2 pulse width of 4ps. Chemical shifts are externally referenced to the
aromatic resonance of hexamethylbenzene (132.3 ppm). Proton combined rotation and
multiple pulse (Ή CRAMPS) spectra were obtained on a Bruker MSL-200 spectrometer
(4.7T field strength) using a tau value of 3.0 μ s and π/2 pulse width of 1.5 μ s. The Ή
spectra were externally referenced to tetramethylsilane (TMS).

NMR Experiments Exploring the Extent of Polymer Mixing

Nuclear magnetic resonance can be used to study the interactions that are
important when different polymers form miscible blends and, by observing the rate of

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 286

proton spin diffusion, the intimacy of mixing over a range of tens of nanometers (14-20).
Spin diffusion is a process in which magnetization is transferred throughout the proton
spin reservoir using the dipolar couplings of the nuclear spins. The measurement of spin
diffusion can be made directly if the different polymers within the sample can be
distinguished by either differences in chemical shifts or can be resolved because of
differences in relaxation times. These experiments rely on the application of a suitable
filter to create a non-equilibrium state that can be monitored as it returns to equilibrium.
Spin diffusion can also be monitored indirectly by measuring relaxation times, such as
T , the proton spin-lattice relaxation time, or T , the proton spin- lattice relaxation
1H l p H

time in the rotating frame, and observing the effect of the polymer blending on these
relaxation times. T values have been used to determine a upper limit for the domain
1H

size in polymer blends. Generally, T values are at least an order of magnitude larger
l H

than values for T and as such can be used to probe larger distances.
l p H
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

A series of T experiments were used to probe the extent of mixing in CA/PVP

l p H

and CA/PS blends and the results are shown in Table I. The T values were measured
l p H

at 2 locations in the spectrum; 74 ppm, assigned to the C2, C3, and C5 carbons of the
C A anhydroglucose unit, and 40 ppm, assigned to the methylene carbons of the PVP
backbone. The T values measured for the blends were the same (within experimental
l p H

error) for both components and were intermediate when compared to the T values of l p H

either the pure CA or PVP. A lower limit to the domain size can be estimated from the
T l p Hdata using the relationship (79)

2
x =jDt (1)
August 31, 2012 | http://pubs.acs.org

where χ is the distance that spin diffusion can occur over during a time, t. The diffusion
1
constant, D, is estimated to be ~10" crrrY. Using a value of ~10 msec (an average T
12
l p H

value) for t, the domains are estimated to be less than 3 nm

Table I. T values measured for cellulose acetate, poly(vinyl phenol),

l p H

poly(styrene) and various blends.

Sample PVP, PS T l p H (msec) CA T l p H (msec)

cellulose acetate 17.1

poly(vinyl phenol) 7.7
25/75 (w/w) CA/PVP 8.8 9.3
50/50 (w/w) CA/PVP 10.1 11.3
75/25 (w/w) CA/PVP 12.8 13.0
50/50 (w/w) CA/PS 6.1 11.4

To contrast to the CA/PVP results, T l p H results from a 50/50 mixture of CA/PS are

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 287

shown in Table I. The T value measured for CA is different (and longer) than the T
l p H l p H

value measured for the PS for the 50/50 blend. This indicates that the C A is not
intimately mixed with the PS suggesting the importance of the hydrogen-bonding
interactions between the phenolic hydroxyls and the cellulose esters in the formation of
a well-mixed polymer blend.
T values measured for pure C A and PVP were the same (within experimental
1H

error) and, as such, were not useful for determining the extent of polymer mixing. It is
interesting to note that the T of the 50/50 CA/PVP blend (5 seconds) was longer than
l H

the T of either the C A or PVP polymers (-1.5 seconds). The relaxation time data
1 H

could be fit to a single exponential indicating that the polymer blend consisted of a single
phase. The increase in the T for the blend with respect to the constituents suggests that
1H

the molecular-level processes that are responsible for the T in the blend are occurring
1 H

on a different time scale. This indicates that the mobility of one or both of the polymers
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

that make up the polymer blend has been effected by the mixing of the polymers. Proton
spin-lattice relaxation times, T , were also measured for the 50/50 (w/w) CA/PS blend.
1H

Different T rates were measured for each of the different polymers in the PS/CA blend;
1H

3.0 s for the CA polymer and 6.4 s for the polystyrene. Using equation 1 and the longer
PS T value, the maximum domain sizes for the 50/50 CA/PS blend are estimated to be
1H

larger than 70 nm.

The relaxation results agree with the DSC results (data not shown) that show a
single T for the CA/PVP blends as opposed to two T 's, one corresponding to C A and
g g

the other corresponding to PS, for the CA/PS blend. The DSC results indicate the
polymer separation in the CA/PVP blends to be less than 20 nm ( the limit of the DSC
sensitivity) and the NMR T results extend this limit to less than 3 nm.
l p H

Spin diffusion between the CA/PVP and CA/PS blends was also probed directly
by using a chemical shift filter to select only the proton magnetization from aromatic
August 31, 2012 | http://pubs.acs.org

protons. The extent of polymer mixing is determined by observing how quickly the
magnetization is transferred between the different polymers comprising the polymer
blend. The chemical shift filter used was a proton analog of the S E L D O M (selectivity by
destruction of magnetization) sequence used to select magnetization from specific
13
resonances that are observed using a dilute spin such as C (21). The S E L D O M
strategy uses a train of π/2 pulses spaced with an appropriate delay to select the
resonance of interest. In proton spin systems, the proposed selection scheme based on
a SELDOM-type strategy can be diagrammed as follows:

{prepy - [multipulse cycle] - prep. - τ } - x

0 y k ddk - multipulse detection (2)

The initial pulse, prepy, is used to place the proton magnetization into the toggling frame
appropriate for the multiple pulse sequence chosen, i.e., a π/2 pulse is chosen for
MREV-8, a π/4 pulse is chosen for BR-24. The proton magnetization is then allowed
to evolve due to chemical shift interactions while homonuclear interactions are removed
by the multipulse sequence. In our experiments, we found it convenient to place the
frequency of the transmitter off resonance and adjust n, the number of multiple pulse
cycles, such that the magnetization to be selected made one complete revolution in the
toggling frame, taking into account the appropriate scaling factor of the multiple pulse
sequence used. The cycle was repeated k times until the desired selectivity is achieved.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 288

Generally, transmitter offsets of ~5 to 7 KHz were required for the selection experiments
compared to typical multiple pulse offsets of 2-3 KHz. The larger transmitter offsets
used in the selection experiment do degrade the resolution of the spectra to a small
degree. The spinning speed is also adjusted such that an integer number of rotor cycles
occurs during the multiple pulse evolution period. This sequence was applied to 50/50
(w/w) blends of CA/PVP and CA/PS.
Typical Ή CRAMPS spectra (without the chemical shift filter) of CA, PS, PVP,
50/50 CA/PS and 50/50 CA/PVP using a BR-24 pulse sequence are shown in Figure 1.
The CA spectrum consists of a peak centered at ~2 ppm, assigned to the acetate methyl
groups and at least two broader peaks centered around ~4.5 ppm assigned to cellulose
hydroxyls and protons attached to the cellulose backbone. The PVP and PS spectra are
characterized by 2 peaks centered around -2 ppm, assigned to the methylene protons,
and 7 ppm, assigned to aromatic protons. The phenolic hydroxyl protons of PVP should
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

also resonate at about 7 ppm..

August 31, 2012 | http://pubs.acs.org

70 5 0 ppm
Figure 1. Typical CRAMPS spectra of cellulose acetate (CA), poly(vinyl
phenol) (PVP), Polystyrene) (PS), a 50/50 (w/w) blend of C A and PVP (50/50
CA/PVP) and a 50/50 (w/w) blend of CA and PS (50/50 CA/PS).

The extent of polymer mixing is probed using the pulse sequence diagrammed in
(2) and incrementing the delay, x , after the aromatic magnetization from the PVP or
dd

PS has been selected. During the time period, x , spin diffusion will transfer
dd

magnetization to spatially nearby protons. The time required for PVP proton magnetiza­
tion to be transferred to the protons of the C A is determined by the extent of mixing in
the PVP/CA polymers blend. Figure 2a shows the spectrum of the 50/50 (w/w) CA/PVP
blend after the application of the chemical shift filter to select the PVP aromatic protons
centered around 7 ppm. The small upfield shoulder at about 5 ppm indicates that a small
amount of residual magnetization from the C A remains due to inefficiencies in the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 289
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

—ι 1 1
10 5 Ο ppm
Figure 2. *Η spectra of 50/50 (w/w) CA/PVP as a function of the spin diffusion
mixing time,^. a) τ^ρΐ ps b)x =200 ps c)x =750 psec and d) *H CRAMPS
dd dd

spectrum without application of the chemical shift filter.

chemical shift filter. Magnetization is observed to grow in rapidly between 0 and 5 ppm
and after a spin diffusion mixing time of 750 ps, the spectrum (Figure 2c) looks similar
to the typical *H CRAMPS spectrum of the CA/PVP blend (Figure 2d). The clear
observation of spin diffusion between the PVP and C A polymers at 200 ρ s (Figure 2b)
August 31, 2012 | http://pubs.acs.org

indicates that the polymers must be mixed at the molecular level because for spin
diffusion to occur on this time scale the protons have to be within 2-5 angstroms (79).
In contrast are the results of the spin diffusion experiment applied to the 50/50
(w/w) CA/PS blend shown in Figure 3. Figure 3a shows the spectrum obtained
immediately after the application of the chemical shift filter to select the aromatic protons
of the PS. Again there is a small upfield shoulder due to residual C A magnetization. At
100 ρ s (Figure 3b), the residual C A magnetization has equilibrated throughout the C A
spin system, but it appears that no magnetization is being transferred from the PS
aromatic protons to the CA polymer. At 1000 ρ s (Figure 3c), the spectrum looks similar
to the polystyrene *H CRAMPS spectrum (Figure 1(PS)) indicating that magnetization
has been transferred throughout the polystyrene spin system. However, the lack of any
observable magnetization that can be assigned to cellulose acetate in the x=1000ps
spectrum indicates that the polystyrene must not be in intimate contact with the cellulose
acetate.

N M R Experiments Exploring Interactions Between Functional Groups

Changes in N M R chemical shifts can be used to probe sites of specific interactions in

solids although not with the predictability that is associated with infrared spectroscopy.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 290

Belfiore and coworkers have demonstrated that the chemical shift of the phenolic carbon
of P V P is sensitive to nearest neighbor interactions that can distort the electronic
environment of the phenolic carbon (22-24). Hydrogen-bonding interactions between
PVP and polymers such as poly(ethylene succinate), PES, and poly(vinyl methyl ketone),
PVMK, have been studied previously. The chemical shift of the phenolic carbon of PVP
and the carbonyl carbons of PES and P V M K were sensitive probes of hydrogen bonds
formed during mixing of the polymers. However, the magnitude of the change in the
chemical shift of the phenolic carbon cannot be used to determine the strength of the
hydrogen bond because other factors can cause changes in the carbon chemical shifts
(24).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020
August 31, 2012 | http://pubs.acs.org

π 1 1
10 5 0 PPm
l
Figure 3. H spectra of 50/50 (w/w) CA/PS as a function of the spin diffusion
!
mixing time,-^. a) τ^ρΐ ps b ^ ^ l O O us c)x =100 psec and d) H CRAMPS
dd

spectrum without application of the chemical shift filter.

13
An expansion of the C CP/MAS spectrum showing the carbonyl and phenolic
hydroxyl carbon region for a series of CA/PVP blends differing in CA/PVP composition
are shown in Figure 4. The peak appearing at 153.5 ppm in the 0/100 (w/w) spectrum
(pure PVP) is assigned to the carbon in the aromatic ring bearing the phenolic hydroxyl
and the peak at 171.5 ppm in the 100/0 (w/w) spectrum (pure CA) is assigned to the
carbonyl carbons of the acetate groups attached to the cellulose backbone.
As the amount of C A is increased in the CA/PVP blends, differences can be
clearly seen in the spectra of the different blends. The first observation is that the
hydroxyl carbon of PVP is seen to shift from 153.5 ppm in the 0/100 CA/PVP spectrum
to 155.1 ppm in the 75/25 CA/PVP spectrum. The carbonyl peak of the C A broadens
upon the addition of the PVP with the FWHH (full width at half height) increasing from

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 291

CA/PVP
0/100

25/75

50/50

75/25
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

100/0

185 180
A
175 |l70 185 160 155; 150 145 ppm

13
Figure 4 Expansion of the carbonyl and phenolic carbon region of the C CP/MAS
spectrum of blends of cellulose acetate (CA) with polyvinyl phenol) (PVP). The
peak heights are scaled to arbitrary intensity. The dotted lines indicate the chemical
shifts of undiluted C A and PVP.

4.3 ppm in the 100/0 spectrum to 7.0 ppm in the 25/75 spectrum. The broadening of
August 31, 2012 | http://pubs.acs.org

the carbonyl peak appears to be due to a shift of intensity from 171.5 ppm to about 174
ppm. The changes in chemical shift observed in the NMR spectra are consistent with
changes observed in FTIR spectra obtained from the same samples (data not shown).
Therefore, we conclude that the peak shifts observed in the N M R spectra reflect
hydrogen-bonding interactions of different strengths similar to the conclusion made by
Belfiore and coworkers (22-24).
The 90/10 CA/PVP spectrum ( Figure 5) shows the appearance of a new peak
at 161 ppm (this peak is also present in the 75/25 CA/PVP spectrum but cannot be
clearly seen in Figure 4). The appearance of the 161 ppm peak in the 90/10 and, to a
lesser extent, 75/25 CA/PVP blend spectra indicates that there may be phenolic hydroxyls
that form stronger hydrogen bonds at higher CA concentrations than is observed at lower
CA concentrations. At the higher CA concentrations, the phenolic hydroxyls may be able
to form hydrogen bonds with the unreacted CA hydroxyl.
A second possible explanation of the 161 ppm peak may be that more than one
C A carbonyl (or other hydrogen-bonding site) is interacting with a single phenolic
hydroxyl and the multiple hydrogen bonds cause the larger change in the PVP phenolic
carbon chemical shift. The disappearance of the 161 ppm peak at higher PVP
concentrations then indicates that these multiple-hydrogen bonds (or stronger hydrogen
bonds) are not present at lower C A concentrations as more phenolic hydroxyls become
available for hydrogen bonding.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 292

The FWHH of the phenolic carbon decreases from 4.0 ppm to 3.0 ppm as the
concentration of C A increases from 0% to 90% . The trend in decreasing linewidth
follows a trend in decreasing T as one lowers the PVP concentration. The measured Tg
g

of the 25/75 CA/PVP blend is 175 °C which decreases to 155 °C in the 90/10 CA/PVP
blend. A narrowing of the phenolic carbon in PVP blends with decreasing T has been g

observed previously by Qin and coworkers in their study of PVP/poly(vinyl methyl

ketone) blends (24).
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

13
Figure 5. Expansion of the carbonyl and phenolic carbon region of the C
August 31, 2012 | http://pubs.acs.org

CP/MAS spectrum of a 90/10 blend of cellulose acetate (CA) with poly(vinyl

phenol) (PVP).

Interactions between different functional groups were also explored by changing

the nature of the polymers constituting the blend. One modification of the polymers
studied was replacement of the acetate side groups with bulkier, more hydrophobic,
butyryl sidechains that could hinder the formation of hydrogen bonds between the
carbonyls on the cellulose ester and the phenolic hydroxyls on the PVP. The CAB/PVP
blends form a single phase as indicated by a single T and FTIR spectra indicate the
g

presence of hydrogen bonding between the phenolic hydroxyls and the carbonyls. The
13
C CP/MAS spectra of the CAB/PVP blends (Figure 6) show changes that are consistent
with the formation of weaker hydrogen bonds between the C A B and the PVP when
compared to the CA/PVP blends.
There is a smaller shift in the phenolic carbon of PVP from about 153.5 ppm to
154.2 ppm that would suggest a weaker hydrogen bonding interaction, although Belfiore
and coworkers have suggested that factors other than the strength of the hydrogen
bonding may also effect the phenolic carbon chemical shift. The unresolved group of
peaks centered around 172.5 ppm in Figure 6 is assigned to butyryl carbonyls attached
at different positions on the cellulose backbone. This assignment is consistent with
liquid-state spectra of C A B that show the carbonyls bonded to different carbons on the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 293

CAB/PVP
0/100

25/75

50/50

75/25

A
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

100/0

185 180
A 175170 165 160 155! 150
ppm 1 3
Figure 6. Expansion of the carbonyl and phenolic carbon region of the C
CP/MAS spectrum of blends of cellulose acetate butyrate (CAB) with
poly(vinyl phenol) (PVP). The peak heights are scaled to arbitrary intensity.
The dotted lines indicate the chemical shifts of undiluted C A B and PVP.

cellulose backbone have different chemical shifts (25,26). To our knowledge, the
assignment of the unresolved peaks have not been made in the solid state. The smaller
shoulder centered at - 169 ppm may be due to residual acetyl carbonyls present in the
polymer. Formation of a broad downfield shoulder on the carbonyl peak centered
August 31, 2012 | http://pubs.acs.org

around 172.5 ppm indicates that the carbonyls of the butyryl side groups are involved in
hydrogen bonding with the phenolic carbons. It is apparent from the changes in the
relative intensities of the different unresolved carbonyl peaks as the PVP content
increases that a butyryl group attached to at least one position may be favored in the
formation of hydrogen bonds although the exact substitution position can not be
determined without further assignment of the carbonyl region in the solid state spectra.
The change in the relative peak heights could also be caused by a change in molecular
motion of the butyryl sidechains due to the hydrogen bonding. A change in molecular
motion of the butyryl sidechain could alter the carbonyl linewidths and intensities.
Further evidence of a possible favored carbonyl position in the hydrogen-bonding
formation can be observed in blends of the C A B with a Novolac resin. The phenolic
hydroxyls in this Novolac resin are relatively hindered compared to PVP in terms of their
ability to hydrogen bond because of the presence of ortho methylene bridges adjacent to
the phenolic hydroxyl group. These methylene bridges also restrict the number of
confirmations that the aromatic rings can assume, further hindering the formation of
13
hydrogen bonds. Changes in the C CP/MAS spectra of CAB/Novolac blends, shown
in Figure 7, are consistent with the formation of hydrogen bonding interactions between
the CAB and Novolac resin. There is a shift in the phenolic carbon from 150.6 ppm to
152.3 ppm and a broad downfield shoulder appears in the carbonyl region centered
around 172.5 ppm suggesting that the phenolic hydroxyls and butyryl carbonyls are

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 294
forming new hydrogen bonds. The unresolved butyryl carbonyl peaks clearly show
changes in the relative intensities as the Novolac content is increased. These changes in
intensity again indicate that one or more of the butyryl positions may be favored. It is
interesting that the changes in the relative intensities of the unresolved carbonyl peaks
appear to be greater in the CAB/Novolac blends than the CAB/PVP blends.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

1Θ5 180 175 170 165 160 155 150

ppm 13
Figure 7. Expansion of the carbonyl and phenolic carbon region of the C
CP/MAS spectrum of blends of cellulose acetate butyrate (CAB) with an uncured
August 31, 2012 | http://pubs.acs.org

Novolac resin. The peak heights are scaled to arbitrary intensity. The dotted lines
indicate the chemical shifts of undiluted C A B and Novolac resin.

Conclusions

Based on relaxation and spin diffusion measurements the results of this study clearly
show that cellulose ester and phenolic polymers are intimately mixed at the molecular
level with interproton separations of the two polymers of about 5 angstroms. The same
experiments performed on blends of cellulose acetate and poly(styrene) showed no
evidence of molecular-level mixing between the polymers and that the domains must be
larger than 70 nanometers. Changes in the peak positions of the carbonyl peak of the
cellulose esters and the phenolic hydroxyl carbon of the phenolic polymers were
consistent with the formation of hydrogen bonds between the two different polymers.
13
C CP/MAS NMR spectroscopy of blends of CAB/PVP and CAB/Novolac suggested
one or more butyryl side groups attached to a specific position along the cellulose
backbone may have been more efficient at forming hydrogen bonds.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 295

Literature Cited

(1) Krause, S. In Polymer Blends; Paul, D. R.; Newman, S., Eds.; Adademic Press,
Inc., New York, New York, 1978, Vol. 1; 16-106.
(2) Masson,J.;Manley, R St. J.. Macromolecules 1992, 25, 589.
(3) White, A. W.; Buchanan, C. M.; Pearcy, B.G.; Wood, M. D. J. Appl. Polym. Sci.
1994, 52, 525.
(4) Sun,J.;Cabasso, I. Macromolecules, 1991, 24, 3603.

(5) Aptel, P.; Cabasso, I. J. Appl. Polym Sci. 1980, 25, 1969.
(6) Kondo, T.; Sawatari, C.; Manley, R St. J. .; Gray, D. G. Macromolecules 1994,
27, 210.
(7) Goh., S.H.; Slow, K.S. Polym. Bull. 1987, 17, 453.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch020

(8) Serman, C.J.;Xu, Y.; Painter, P.C.; Coleman, M. M. Macromolecules 1989, 22,
2019.
(9) Jong, L.; Pearce, E. M.; Kwai, T. K. Polymer, 1993, 34, 48.
(10) Fahrenholtz, S. R.; Kwei, T. K. Macromolecules 1981, 14, 1076.
(11) Pennacchia, J. R.; Pearce, E. M.; Kwei, T. K.; Bulkin, B. J.; Chen, J-P.
Macromolecules 1986, 19, 973.
(12) Serman, C.J.;Xu Y.; Painter, P. C.; Coleman, M. M. Polymer 1991, 32, 516.
(13) Moskala, E.J.;Varneli, D. F.; Coleman, M. M. Polymer 1985, 26, 228.
(14) Caravatti, P.; Neuenschwander, P.; Ernst, R.R. Macromolecules 1985, 18, 119.
(15) Caravatti, P.; Neuenschwander, P.; Ernst, R.R. Macromolecules 1986, 19, 1889.
(16) Campbell, G. C.; VanderHart, D. L. J. Magn. Reson. 1992, 96, 69.
(17) Clauss,J.;Schmidt-Rohr K.; Spiess H. W. Acta Polymer 1993, 44, 1.
(18) Yang, H.; Kwei, T. K.; Dai, Y. Macromolecules 1993, 26, 842.
August 31, 2012 | http://pubs.acs.org

(19) VanderHart, D. L. Macromolecules 1994, 27, 2837.

(20) VanderHart, D. L. Macromolecules 1994, 27, 2826.
(21) Tekely, P.; Brondeau,J.;Elbayed, K.; Retournard, Α.; Canet, D. J. Magn. Reson.
1988, 80, 509.
(22) Belfiore, L. Α.; Qin. Q; Pires, Α.; Ueda, Ε. Polym. Prep. (Am. Chem. Soc., Div.
Polym. Chem.) 1990, 31, 170.
(23) Belfiore, L. Α.; Lutz, T.J.;Cheng, C.; Bronnimann, C. E. J. Polym. Sci., Polym.
Phys. Ed. 1990, 28, 1261.
(24) Qin, C.; Pires, Τ. Ν.; Belfiore, L. A. Macromolecules 1991, 24, 666.
(25) Buchanan, C. M.; Hyatt, J. Α.; Lowman, D. W. Macromolecules 1987, 20, 2750.
(26) Tezuka, Y. Carbohydrate Res. 1993, 241, 285.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 21

Interchain Hydrogen Bonds in Cellulose-Poly(vinyl

alcohol) Characterized by Differential Scanning
Calorimetry and Solid-State N M R Analyses Using
Cellulose Model Compounds

1 2
Tetsuo Kondo and Chie Sawatari
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

1
Forestry and Forest Products Research Institute (FFPRI), P.O. Box 16, Tsukuba
Norin Kenkyu, Ibaraki 305, Japan
2
Faculty of Education, Shizuoka University, Shizuoka 422, Japan

To characterize the nature of the interchain hydrogen bonds

involved in cellulose/PVA blends, D S C and solid-state N M R
analyses were carried out on mixtures of two cellulose model
compounds (2,3-di-O-methycellulose; 23MC and 6-O-methyl­
-cellulose;6MC) and P V A . The conclusions reached, based on our
experimental data, are as follows: first, the thermodynamic data
confirms that the interchain hydrogen bonds between the cellulose
ring oxygen and the O H group of P V A are favored over bonds
formed between the hydroxyl groups on each homopolymer.
Second, the proton spin-lattice relaxation time in rotating frames,
August 31, 2012 | http://pubs.acs.org

T1ρH, showed that both M C / P V A blends examined were

homogeneous down to a scale of 17.5 nm or less at most
concentrations, suggesting that blends of pure cellulose and P V A
may also be miscible on a molecular scale. Finally, we comment on
the correlation between the interaction parameter, x, assessed from
thermodynamic data obtained using DSC and the domain size for
the hydrogen bonding pairs in the blends.

Since the late 1970's cellulose/synthetic polymer blends have been extensively
studied while more recently attention has been focussed on biodegradable polymer
blends(7-75). The polymer-polymer interactions believed to be reponsible for the
miscibility are mainly attributed to hydrogen bonds formed between the three
hydroxyl groups on the anhydroglucose units of cellulose and the functional groups
present on the synthetic polymers. More recently, the domain size or good-mixing
scale was estimated by measuring the T i and T i p proton with solid state CP-MAS
1 3
C - N M R (16-18). However, in cellulose/PVA blends it has been difficult to
accurately characterize the hydrogen bonds involved in the process because there are
three hydroxyl groups in each repeating unit of the cellulose. In a previous paper
(79), the authors reported on the FT-IR differences of the hydrogen bonds formed
depending on the regiochemistry with regard to the miscibility in certain cellulosic
blends. The work was carried out using cellulose model compounds with
regioselectively methylated hydroxyl groups, namely 2,3-di-O-methylcellulose
(23MC)(20)and 6-O-methylcellulose (6MC)(2i) (Figure 1). From the FT-IR study of

296 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 297

the M C / P V A blend samples, it was revealed that two types of hydrogen bonds can be
formed in cellulose / P V A blends: i) hydrogen bonds between the OH groups at either
the C-2 or the C-3 position of cellulose and the side chain O H groups of P V A (these
interactions are postulated to form mainly at the C-2 position of cellulose), and ii)
hydrogen bonds between the ring oxygen (0-5) of cellulose and the O H groups in
P V A (Figure 1). In the present study we want to clarify further the regiochemical
effects on the pure cellulose/PVA blend previously investigated (19). We also
estimated quantitatively the thermodynamic interaction of the hydrogen bond
formation between the two regioselectively methylated cellulosics and P V A using
DSC analysis and we extrapolated these results to the case of pure cellulose/PVA
blends. Further, the domain size was also evaluated by spin diffusion from solid-state
N M R measurements.

Experimental
Materials. The two cellulose model samples, 2,3-di-O-methylcellulose (23MC)(20 )
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

and 6-0-methylcellulose (6MC)(2i)> were prepared according to refi9. Each

polymer had a uniform structure having every structural unit substituted
regioselectively. Polyvinyl alcohol) (PVA) was purchased from Polyscience Inc. and
4
it had a nominal molecular weight of 2.5 χ 1 0 . HPLC-grade N,N-dimethylacetamide
(DMAc) (Aldrich Chemical Co., Inc.) was used without further purification.
Preparation of film specimens. M C / P V A blends were prepared from mixed
polymer solutions in D M A c according to the film casting method described in an
earlier paper 19. The two film blend systems subjected to DSC measurements with a
Perkin-Elmer DSC-7 apparatus had relative compositions of the two polymers (MC /
PVA) of 80/20, 65/35, 50/50, and 30/70 by weight. For the solid-state N M R
measurements, the compositions were 75/25,50/50 and 25/75 by weight, respectively
for both the 23MC/PVA and 6 M C / P V A mixtures. Wide-angle X-ray diffraction
patterns (79) verified that the blended films were predominantly amorphous. The
densities for the two highly amorphous cellulosic homopolymers (23MC and 6MC)
and for the P V A films were measured by pycnometry in a mixed medium of /?-xylene
August 31, 2012 | http://pubs.acs.org

and carbon tetrachloride. The characterization data for the three films studied are
listed in Table I.
DSC Measurements. Film specimens weighing from 5 to 17 mg were placed in
aluminum sample pans which were heated to 242° C at the heating rate of 10°C/min
and maintained at this temperature for 5 minutes in an atmosphere of nitrogen to
eliminate P V A crystalline residues. The samples were quenched at the selected
isothermal crystallization temperature T i , were then held at T i for 7 hours to allow
c c

complete crystallization and then were cooled to 20° C. After each sample was
isothermally crystallized, the melting point T was then measured using a heating
m

rate of 10°C/min. Subsequently by using Hoffman-Weeks plots (22) of the Tm's thus
e
obtained, the equilibrium melting point (Tm Q) was determined for each film.
Solid-state NMR Measurements. Before N M R measurements the samples were
completely dried for more than a week under vacuum at 80° C. N M R spectra were
obtained on a JEOL JNM-GSX 400 instrument equipped with a dedicated solids
accessory. The CP/MAS measurements were obtained at 9.4T, which corresponds to
frequencies of 100.4 MHz for l ^ C and 400 MHz for protons at room temperature.
Spinning rates were 5.0 to 6.0 kHz, and the Hartmann-Hahn match was adjusted
prior to each accumulation. Cross-polarization times were typically 1 to 2 ms. The
recycle time between pulses was 10s. The spectra were accumulated ca. 500 times.
The C H signal of adamantane was used as an external reference to determine
chemical shifts. The pulse sequence for ΤχρΗ measurements was done as described
in a previous paper (18). Measurements of proton spin-lattice relaxation times in the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 298

OHCH -)2 n

23MC/PVA
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

(Β) -(-CH ÇH-) -

2 n
August 31, 2012 | http://pubs.acs.org

-(-CH CH-) -
2 n

6MC/PVA
Figure 1. Proposed hydrogen bonding schemes for the two cellulosics blended
with PVA, (A) 23MC/PVA and (B) 6MC/PVA.

Table I Characterization data for the homopolymer

components.
Sample Source Molecular Density
weight (gem' 3 )

23MC Synthesized 4.0x10 4 1.30

6MC Synthesized 3.5x10 4 1.30
PVA Polyscience 2.5x10 4 1.27

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 299

rotating frame (TipH) were obtained by using a computer-generated best fit for the
intensity of the 13C-NMR spectra to the single-exponential equation M(t)=M(0)exp(-
τ/ΓίρΗ). The proton spin lock time (delay times), τ, ranged from 1 to 35 ms
depending on the blend under investigation and eight τ values were employed to
determine each relaxation time. Each plot used to obtain ΤχρΗ was a single
exponential.

Results and Discussion

DSC Characterization. Since both amorphous 6MC and 23MC homopolymers
showed diluent effects in our blend systems, we have assumed the change in the
chemical potential with increasing amount of the M C components (4,6,9, 23-29 )
reflected itself as a depression in the melting point of the P V A crystallized
isothermally in the M C / P V A blends. The thermodynamic mixing of the two
polymers has been dealt with by Scott (30) using the Flory-Huggins approximation
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

(31). The conventional equation for the thermodynamic depression of the melting
point caused by a diluent is as follows:
0
l/rm-lArm = - R ( V / A H ) 2 u 2 u

X {lnv2 / V2 + (1/ V - 1/ V i )vi + Bvi^/RTm} 2 (1)

where T ^ is the melting point of P V A and T is the observed melting point of the
m m

blended P V A . In this equation, 1 and 2 refer to M C and P V A , respectively. v\ and v 2

are the polymer volume fractions while V j and V are molar volumes. V is the 2 2 u

molar volume of the repeating units of 2, Δ Η is the enthalpy of fusion per mol of 2 υ

the repeating units of 2, and is the square of the volume fraction of non-
crystallizable component 1 (MC).
4
Since V j and V are on the order of 1 0 for 23MC, 6MC and P V A the entropy
2

term in eq 1 can be entirely neglected 24. Eq 1 can then be rearranged into the
August 31, 2012 | http://pubs.acs.org

following form (eq 2) where the enthalpic contribution to the melting point
depression can be evaluated:
0 2
ΔΤη^Τπι - T = - T ° ( V / A H ) B m m 2 u 2 u (2) V l

where A T is the melting point depression of the P V A component and R is the

m

universal gas constant, 1.986 cal-deg'l-mol'l. Eq 3 relates the Β parameter, or

interaction energy density characteristic of the two polymers, to the Flory-Huggins
interaction parameter, χ ^ , which describes the enthalpy of mixing:
2

B=RT( /V )X l 2 l u (3)
However, there are some morphological effects which must be considered;
specifically the effects are known to be mainly due to the degree of perfection and the
finite size of the crystals (22). To cancel these effects, the equilibrium melting points,
ec
Ύτη^Φ and T m l , obtained from the Hoffman-Weeks plots (22) were used instead of
TmO and Tm respectively in eq.2. In order to use the Hoffman-Weeks plots, it is
necessary that the P V A component be completely crystallized isothermally. Thus, the
e (
TmO and T m l were determined from the Hoffman-Weeks plots using the melting
temperatures, Tm's. In Figure 2, the experimental data for the melting point
depression, A T , for each system were plotted versus the square of the volume
m

fraction of the cellulosic component, v ^ o we could use eq. 2. The solid line was
S

drawn by using a least-square fitting method assuming there is a linear relationship

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 300

between A T and v ^ . The volume fraction was calculated using density data for the
m

cellulosics and P V A as noted in Table I. Both straight lines yielded positive

intercepts. Residual entropie effects might be responsible for the non-zero intercept
(23-29). However, in the present cases the low magnitude of the intercepts makes
their contribution negligible. From the slopes of the two A T versus v ^ plots, we
m

can assess values for the Β parameter and χ ^ by using eq. 2 and 3 in combination
with other known necessary quantities as follows; the heat of fusion per unit volume
(ΔΗ2ι1/ν2ι1) of P V A is 45.4 cal-cm" and V =146.15 and 135.38 (cmS-mol" ) for
6 3
lu
1

23MC and 6MC, respectively. V j values are calculated from the molar masses (190
u

and 176 for 23MC and 6MC) and densities (Table I) for each cellulosic. The slopes
of the two straight lines were 155.5 deg and 110.0 deg for the 23MC/PVA and
6MC/PVA, respectively.
From these calculations, the thermodynamic interaction parameters were
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

evaluated (Table II). The fairly large negative values for the Β and parameters for
both of the M C / P V A blends can be explained by the two components interacting
thermodynamically in the mixture. The values were also greater than that for a pure
cellulose/PVA blend, indicating that the two polymer pairs present were more
favorably miscible than a pair consisting of pure cellulose and P V A . In other words,
the two model systems showed enhanced miscibility. The large negative values of Β
for the model systems were estimated to be 2 to 6 times larger than those reported for
other polymer pairs (23-29) while the Β value of a cellulose/PVA system was only
1.5 to 4 times (6 ) the Β values for other polymers. This result strongly suggests the
presence of favorable interactions presumably due to the formation of interchain
hydrogen bonding. This conclusion is also supported by our previous FT-IR results
(19).
The difference in Β values between 23MC/PVA and 6 M C / P V A is assumed to
reflect mainly the difference in strength of the interchain interactions involved in
hydrogen bonding in the blends. In a previous paper using FTIR (19), we reported
August 31, 2012 | http://pubs.acs.org

that in 23MC/PVA hydrogen bonding takes place between an ether oxygen in the
anhydroglucose ring and a hydroxyl group of P V A , while the hydrogen bonds in
6MC/PVA were formed mainly between the hydroxyl groups of each component.
Thus, in cellulosic blends the formation of hydrogen bonds between ring oxygens and
OH groups is considered to be more favorable than those between component
hydroxyl groups. In addition, since the former type of hydrogen bond (O-HO) is
considered to be very similar to one forming intramolecular hydrogen bonds between
ring oxygens and hydroxyls at the C-3 position in cellulose molecules, this result also
indicates that the intramolecular hydrogen bonds formed should be strong as is well-
known in cellulose chemistry.
Proton T i and T i p Measurements. The relaxation of proton spins in the spin-
lattice (TiH) and the rotating frame (ΤχρΗ) is observed by its effect on the N M R of
the carbon nucleus to which the proton is bound. If the different protons are not in
contact, each has its own characteristic relaxation time. At the other extreme, if the
protons are closely coupled by rapid spin-spin exchange dipole-dipole interactions,
i.e. spin diffusion, then these protons share the same ΤχρΗ. In miscible polymer
blends on the scale characterized by the relaxation time, the measured proton
relaxation rate is an average of the proton relaxation rates for the constituent
polymers (16-18,38). The relaxation measurement method is complementary to CP-
M A S since the relaxation times are sensitive to homogeneity scales which differ from
the scale to which CP-MAS is sensitive. It has been previously reported that the

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 301
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021
August 31, 2012 | http://pubs.acs.org

Table II Values of the interaction energy density, B , and

of the interaction parameter, χ ^ , for binary blends with
components compatible in the melt.

Cellulose/PVA 6
23MC/PVA 6MC/PVA

Β -9.38 -13.65 -9.65

(Cal. cm-3)
χ 1 2 -0.985 -1.94 -1.32
(at 517 Κ) (513K)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 302

ΤίρΗ values for cellulose/PVA blends could not be used to calculate a domain size,
because the T i p H value for both blend components were virtually identical within the
experimental error. In contrast, in the present study such a problem did not exist since
both 23MC and 6MC had significantly different T i p H values from P V A .
The measured T i p H values for the protons attached to the characteristic carbon
nuclei at 107, 104 and 45 ppm for 23MC, 6 M C and P V A , respectively, are listed in
Table III for each blend studied. As denoted by the underlined values in Table III, the
compositions which gave identical component relaxation times for 23MC/PVA and
6MC/PVA were 75/25 and 25/75, respectively. Thus, at these compositions both
blends were thoroughly-mixed at this measurement level. Since the 75/25
composition of 23MC/PVA is the closest to a molar ratio of 1/1 among the studied
b l e n d
compositions, the miscible 23MC/PVA was considered to be equivalently
mixed in the two components. On the other hand, 6 M C and P V A were well-mixed
only when an excess of P V A was present. This behavioral difference might be due to
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

t n e
distinctive type of hydrogen bond interactions, previously described, in the two
blends. In the above cases when the protons for the two components in the blend have
the same T i p H values, then eq. 4 (39) can be used if the molecule are in contact at an
intimate molecular distance, L, over which the protons can effectively diffuse in a
given time, t.
2 2
<L >~(t/T2)-<l0 > (4)

where lo is the distance between protons, typically 0.25nm and T is the proton spin-
2

spin relaxation time, οζΛΟμς below Tg (39). The available time for spin diffusion, t,
is usually equal to T i p H . Using the average T i p H values of t=30 ms for 23MC/PVA
and 49ms for 6MC/PVA, we calculated the observation diameter for spin diffusion to
be about 13.7 nm (23MC/PVA) and 17.5 nm (6MC/PVA). Hence, each pair in
23MC/PVA and 6 M C / P V A is homogeneous on a scale of 13.7 and 17.5 nm or less,
August 31, 2012 | http://pubs.acs.org

respectively. These values are significantly different considering the error range of ±
5%.
Despite being highly amorphous, as described in the experimental section, the
PVA showed a much longer T i p H time of 15.9 ms, than reported previously by
Masson and Manley (18) (4.3 ms), 9.7 ms by Zhang et al. (34) (9.7 ms)„ and by Horii
et al. (36) (8.2 ms). In observing P V A by CP-MAS spectroscopy, the characteristic
methine signal I due to the mm triad exhibited a relatively higher and sharper peak
when compared to the same signals in the above previous reports (18,34,36). This
signal has already been attributed (35) to the formation of intramolecular hydrogen
bonds by the mm triad. Thus the longer T i p H can be attributed to the formation of a
larger amount of intramolecular hydrogen bonds. Furthermore, for the 25/75
composition of 23MC/PVA, the T i p H value for the proton bound to the C - l carbon
of 23MC is shorter than that of the P V A which has the shortest T i p H of the
homopolymers, suggesting that a conformational change in the skeleton of the 23MC
may occur (38). As our proposed interaction for the 23MC/PVA blend was a
hydrogen bond between an ether ring oxygen of 23MC and an OH of P V A (Figure
1A), the conformation of the 23MC glucose ring can reasonably be changed by the
interaction.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 303
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

Table III Proton 7Ί and Tip relaxation time

for 23MC, 6MC, PVA and their blends.
Tip , * ms
H

blend 23MC PVA 6MC PVA

Composition (107ppm) (45ppm) (104 ppm) (45ppm)

0/100 15.9 15.9

25/75 10.0 31.2 50.5 47.1

50/50 40.2 25.2 16.7 25.2

August 31, 2012 | http://pubs.acs.org

75/25 30.2 29,6. 27.4 67.9

100/0 76.4 125.9

* A c c u r a c y is ± 5 % .

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 304

Conclusions
We have attempted to characterize the interchain interactions engaged in hydrogen
bonding in a cellulose/PVA blend systems using two cellulose model compounds,
23MC and 6MC, to represent the cellulosic component.
Thermodynamic DSC data allowed us to quantify the interchain interaction in the
blends and mixtures by using interaction parameters, χ. The χ values were all
negative and very similar for the M C / P V A blends and for the cellulose/PVA systems,
indicating that the cellulosic components in both systems interact favorably with
PVA. In light of our previous FTIR results (19), these interactions can be assumed to
be unique for each blend. Furthermore, an evaluation of the interaction parameters, χ,
for each of the two blends suggests that hydrogen bond interactions between the ring
oxygen and the O H groups of the P V A are more favorable than those between the
hydroxyl groups on each component homopolymer. Thus the large negative value of
χ values in the cellulose/PVA system implies that the system is themodynamically
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

miscible, which means down to the molecular level.

To evaluate in more detail the domain sizes for the two M C / P V A blends, we
observed changes in the ΤχρΗ relaxation times with blend composition. These results
gave domain sizes of 13.7 nm for 23MC/PVA and 17.5 nm for 6MC/PVA. Masson et
al. (18) reported that the domain of the cellulose and P V A pair is on a scale of 36 nm
or less. Since their cellulose and P V A showed equal T i p H values within the
experimental error, they could not use these T i p H values in their calculations for the
blends. However, fortunately for our model compounds the T i p H measurements
permitted us to analyze the blend domain size on a molecular level scale. Therefore
the above calculated domain sizes for the M C / P V A blend prove that the
cellulose/PVA blend is miscible on a molecular level as already suggested by FTIR.
Finally in addition to the investigation on a pure cellulose/PVA blend, some
interesting aspects of hydrogen bond interactions in the two model systems
themselves were revealed. From the DSC analysis, we quantitatively propose that the
August 31, 2012 | http://pubs.acs.org

more favorable formation of interchain hydrogen bonds occurs between the ether ring
oxygen and the O H groups rather than between the two hydroxyls in the blends. In
interpretating the T i p H values in light of the above proposal, the stronger interaction
in 23MC/PVA seems to make the miscible domain of the blend smaller, while the
weaker interaction in 6MC/PVA can be attributed to the larger domain size. Namely,
a strong interaction makes the molecular assembly more compact and hence the
density may be higher, which appears reasonable. In addition, when considering that
w a s
the interaction parameter, X23MC/PVA almost 50 % higher than X6MC/PVA »
whereas the domain size for 23MC/PVA was almost 20 % smaller than that for
6MC/PVA, it is concluded that the interaction strength in the blend is the dominant
factor directly contributing to the domain size.

Acknowledgment. The authors wish to thank Dr. Rita S. Werbowyj of the Pulp
and Paper Research Institute of Canada for editing the text.

Literature Cited
1. Jolan, A. H.; Prud'homme, R. E. J. Appl. Polym. Sci. 1978, 22, 2533.
2. Field, N. D.; Song, S. S. J. Polym. Sci. Polym. Phys. Ed. 1984, 22, 101.
3. Nishio, Y.; Roy, S. K.; Manley, R. St. J. Polymer 1987, 28, 1385.
4. Nishio, Y.; Manley, R. St. J. Macromolecules 1988, 21, 1270.
5. Morgensten, B.; Kammer, H. M. Polymer Bull. 1989, 22, 265.
6. Nishio, Y.; Haratani, H.; Takahashi, T.; Manley, R. St. J. Macromolecules 1989,
22, 2547.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 305

7. Nishio, Y.; Roy, S. K.; Manley, R. St. J. Polym. Eng. and Sci. 1990, 30, 71.
8. Jutier, J.-J.; Lemieux, E.; Prud'homme, R. E. J. Polym. Sci. Polym. Phys. Ed.
1988, 26, 1313.
9. Nishio, Y.; Hirose, N.; Takahashi, T. Polym. J. 1989, 21, 347.
10. Nishio, Y.; Hirose, N.; Takahashi, T. Sen-i Gakkaishi 1990, 46, 441.
11. Sakellariou, P.; Hassen, Α.; Rowe, R. C. Polymer 1993, 34, 1240.
12. Scandola, M.; Ceccorulli, G.; Pizzoli, M. Macromolecules 1992, 25, 6441.
13. Ceccorulli, G.; Pizzoli, M.; Scandola, M. Macromolecules 1993, 26, 6722.
14. Buchanan, C. M.; Gedon, S. C.; White, A. W.; Wood, M. D. Macromolecules
1992, 25, 7373.
15. Buchanan, C. M.; Gedon, S. C.; Pearcy, B. G.; White, A. W. Macromolecules
1993, 26, 5704.
16. Masson, J.-F.; Manley, R. St. J. Macromolecules 1991, 24, 5914.
17. Masson, J.-F.; Manley, R. St. J. Macromolecules 1991, 24, 6670.
18. Masson, J.-F.; Manley, R. St. J. Macromolecules 1992, 25, 589.
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch021

19. Kondo, T.; Sawatari, C.; Manley, R. St. J.; Gray, D. G. Macromolecules 1994,
27, 210.
20. Kondo, T.; Gray, D. G. Carbohydr. Res. 1991, 220, 173.
21. Kondo, T. Carbohydr. Res. 1993, 238, 231.
22. Hoffman, J. D.; Weeks, J. J. J. Res. Natl. Bur. Stand., Sect. A 1962, 66, 13.
23. Nishi, T.; Wang, T. T. Macromolecules 1975, 8, 909.
24. Imken, R. L.; Paul, D. R.; Barlow, J. W. Polym. Eng. Sci. 1976, 16, 593.
25. Kwei, T. K.; Patterson, G. D.; Wang, T. T. Macromolecules 1976, 9, 780.
26. Paul D. R.; Barlow, J. W.; Bernstein, R. E.; Wahrmund, D. C. Polym. Eng. Sci.
1978, 18, 1225.
27. Ziska, J.J.;Barlow, J. W.; Paul, D. R. Polymer 1981, 22, 918.
28. Martuscelli, E.; Pracella, M.; Yue, W. P. Polymer 1984, 25, 1097.
29. Martuscelli, E. Polym. Eng. Sci. 1984, 24, 563.
30. Scott, R. L. J. Chem. Phys. 1949, 17, 279.
31. Flory, P. J. "Principles of Polymer Chemistry" 1953, Cornell University Press,
Ithaca.
August 31, 2012 | http://pubs.acs.org

32. Kondo, T.; Sawatari, C. Polymer 1994, 35, 4423.

33. Grobelny,J.;Rice, D. M.; Karasz, F. E.; MacKnight, W. J. (a)
Macromolecules 1990, 23, 2139; (b) Polym. Commun. 1990, 31, 86.
34. Zhang, X.; Takegoshi, K.; Hikichi, K. Polym.J.1991, 23, 87
35. Terao, T.; Maeda, S.; Saika, A. Macromolecules 1983, 16, 1535.
36. Horii, F.; Hu, S.; Ito, T.; Odani, H.; Kitamaru, R.; Matsuzawa, S.; Yamaura, K.
Polymer 1992, 53, 2299.
37. Kondo, T. J. Polym. Sci., B: Polym. Phys. 1994, 32, 1229.
38. Dickinson,L. C.; Yang, H.; Chu, C.-W.; Stein, R. S.; Chien, J. C. W.
Macromolecules 1987, 20, 1757.
39. McBrierty, V.J.;Douglass, D. C. J. Polym. Sci., Macromol. Rev. 1981, 16, 295.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 22

Supramolecular Architectures of Cellulose Derivatives

1
M . Schulze , M . Seufert, C. Fakirov, H. Tebbe, V. Buchholz, and G. Wegner

Max Planck Institute for Polymer Research, P.O. Box 3148, D-55021 Mainz,
Germany
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

Abstract: Several cellulose derivatives belong to a special class of

polymers called hairy-rod macromolecules that are used to generate well-
-defined supramolecular architectures by the Langmuir-Blodgett (LB)
technique. In particular, trimethylsilyl-cellulose (TMSC) forms
monomolecular films on the Langmuir-trough and is transferred onto
hydrophobic substrates with a constant transfer ratio, as it does not
undergo chemical changes in the film-building process. Silylated
celluloses were regenerated, a convenient method for the generation of
homogenous ultrathin films with hydrophilic surfaces. The adsorption of
polymers and dyes as well as biomolecules onto regenerated and modified
cellulose L B films have been studied. In addition, chemical reactions -
August 31, 2012 | http://pubs.acs.org

such as cycloaddition, desilylation, and crosslinking reactions within single

monolayers - have been performed.

Hydrophobic ethers of cellulose can be spread on a water surface to form

monomolecular layers (1, 2). These layers can be repeatedly transferred onto solid
hydrophobic substrates by the Langmuir-Blodgett (LB) technique. Within the so
formed multilayer systems, the hairy rod macromolecules are oriented in one direction
and distributed homogeneously over a large area in the layer plane (3, 4). The layer
distance, which is in the range of 10-20 À, is also well-defined by the side-chain
interactions. Because the alkyl side-chains of the cellulose derivatives are in a fluid
state, diffusion of small molecules is still possible. Small molecules are able to enter
the layered systems from a gas or liquid phase, swell it, and react with functions at the
side-chains. The present paper will briefly review (i) chemical reactions performed
between adjacent monolayers such as [2+2] cycloaddition, Diels-Alder, crosslinking,
and desilylation reactions; (ii) investigations of crosslinked multilayers as liquid
separation and transport membranes; and (iii) adsorption studies on ultrathin films of
regenerated cellulose derivatives.

'Current address: Fraunhofer Institute of Applied Materials Research, Kantstrasse 55, D-14513 Feltow,
Germany

306 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 307

Layered cellulose assemblies as media of chemical reactions (5).

Reactions in layered assemblies of organic molecules as obtained by the L B -

technique are frequently of a topochemical nature (6, 7), that is, the course of the
reaction and the direction in which the product is formed are completely controlled
by the packing of the starting material in the lattice (8). The monolayers of which
L B assemblies are composed extend in the x y plane; the assembly direction defines
f

the ζ axis (Figure la). All reactions occurring in such systems can be characterized
by a dimensionality, where dimensionality of the reaction should be differentiated
from product dimensionality (5). The latter refers to the spatial structure of the
product, that is, whether a linear macromolecule or a two- or three-dimensional
network is the reaction product. In contrast, the dimensionality of the reaction
describes the tensorial character of the reaction. The polymerization of amphiphilic
monomers within LB-layered assemblies is a reaction that proceeds in one direction
and gives polymeric chains as one-dimensional products. Reactions with zero-
dimensionality, the type of reaction that takes place at the site of individual
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

molecules only, are exemplified, for example, by cis-trans isomerizations.

Layered assemblies of cellulose derivatives, which may be regarded as
copolymers composed of anhydroglucose units randomly substituted by ether and
ester groups are investigated. In particular, isopentylcellulose 1 with a degree of
substitution (DS) of 2.9, [5-(9-anthrylmethoxy)pentyl]isopentyl cellulose 2, (DS -
2.8 with respect to the isopentyl and 0.1 with respect to the anthryl residues, and a
partially modified polymer containing fumarate 3 (DS - 2.5 with respect to fumarate
groups) have been used (Figure 2). The average degree of polymerization P of the n

cellulose derivatives 1, 2 and 3 was 87. The anthryl residues undergo 4,+4,
cycloaddition when irradiated, a process that leads to a network (9). This reaction
can be monitored by the decrease of the anthracene absorption at 250 nm. The
reaction in the ζ direction is prevented by embedding single layers of 2 between
layers of 1. This lack of reaction partners in ζ direction restricts the reactivity of the
anthryl groups to the x,y plane and results in the formation of a two-dimensional
network. Since the self-diffusion coefficient of the macromolecules is immeasurably
August 31, 2012 | http://pubs.acs.org

small in the layered assemblies (JO), a photoreaction involving lateral diffusion of

the macromolecules can be excluded.
A further strategy to create two-dimensional products involves selective cross-
linking of adjacent layers. This reaction, which ought to proceed one-dimensionally
in the ζ direction, nevertheless results in a two-dimensional network, as can be seen
in Figure lc. Reactions occurring in the x,y plane can be prevented by
functionalization of adjacent layers. Single layers of 2 and 3 were each separated by
two layers of 1. If this periodic (112113) multilayer was annealed at 120-150 °C for
5
l

more than five hours, the reaction stopped occuring. If, however, an analogous
layered assembly was prepared in which adjacent layers of 2 and 3 were each
embedded between four layers of 1, a substantial decrease of the intensity of
anthracene absorption could be observed. The layered assembly exhibits a thickness
of D - 817 A immediately after preparation, as determined by X-ray reflection; the
thickness shrinks by about 3% to 790 A in the course of the reaction. The stability of
the two-dimensional network obtained by the Diels-Alder reaction was checked by
washing the whole sample with solvent. It is important to note that a periodic layer
structure remains intact even after removal of the cellulose derivative 1. Thus, a
formation of extended 2-d-networh has been achieved.

Ultrathin membranes of cellulose assemblies on porous substrates (11).

Hairy rod macromolecules based on cellulose alkylethers can be used to construct

membranes if transferred by the LB-technique to a porous Celgard 2400® film as the
substrate. Celgard 2400 polypropylene film is a well-characterized commercial

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 308

a) I

polyrmr backbone
I

substrate fluid
siécfaniratrix

c)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

:ooo react ί on z>ne"

Figure 1. a) General architecture of a L B layered assembly; b) layered

assemblies composed of hairy rod macromolecules; c) definition of the
reaction zones within a hexagonally dense packing relevant for
intermolecular reactions.

Mi
August 31, 2012 | http://pubs.acs.org

Figure 2. Chemical structure of the used cellulose derivatives: isopentyl

cellulose 1, [5-(9-anthrylmethoxy)pentyl]isopentyl cellulose 2, and cellulose
fumarate 3.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 309

product with nearly uniform pore size distribution that can be used as a substrate
without further treatment (12). Because of the small thickness (less than 10 Â per
layer) and the possibility of introducing functional groups into the cellulose
molecules, cellulose ethers seem to be promising molecules for the design of
membranes via L B technique (Figure 3).
Sites for photo-crosslinking are introduced by attaching a few cinnamyl residues
to the cellulose backbone. A mixed hydrophobic (isopentyl/cinnamyl) cellulose
ether is prepared and spread from a dilute chloroform solution. L B layers are
formed with transfer ratios of 100%. Membrane assemblies consisting of different
numbers of monolayers on porous Celgard 2400 polypropylene film were prepared
and the multilayer assembly was crosslinked by photoirradiation according to the
reaction shown in Figure 4 (13).
The morphology and quality of the films on the porous substrate were
investigated by T E M (Zeiss 902, 80 kV) using a simple one-stage replication
technique. A defect-free membrane surface was obtained after deposition of 40
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

monolayers. The pores in the substrate film were completely coated by the LB-film.
Photocrosslinked and thus insoluble L B assemblies of cellulose derivatives can be
reversibly swelled by organic solvents, which are good solvents for the
uncrosslinked materials (13). The osmotic pressure exerted on the membranes
exposed to swelling solvents does not change the morphology or damage the
membranes. The transport properties of the composite membranes were tested in a
conventional membrane osmometer based on the time dependence of the osmotic
pressure of toluene solutions of polystyrene molecular weight standards. Samples of
degree of polymerization 8 < Ν < 1632 were used. The decay of osmotic pressure
with time could be represented by a monoexponential function in a first approach to
this effect (14). The half-life time t of the initial pressure was determined for each
1/2

polymer sample, where the concentration of the initial polymer solution was fixed to
10 g/1. The results are plotted in Figure 5 in a double logarithmic scale, together
with data describing the case of uncoated Celgard as the separation of the osmotic
cell (lower solid line).
Coating of the porous substrate by 60 or 70 cellulose monolayers strongly
August 31, 2012 | http://pubs.acs.org

changes the permeation behavior. The porous substrate alone shows an approximate
6
dependence of t «N° for good solvents. For a freely diffusing polymer in solution,
I/2
0 5 0 6
the behavior expected is between t « N for θ-conditions and t « N for good
m m

solvents.
The membrane covered by 60 monolayers shows an approximate relationship of
1 0 3
t « N , while the one covered by 70 monolayers shows an unexpected dependence
1/?

with strongly nonlinear behavior over the range of polystyrene samples tested for
!57
permeation. A dependence of t «N is found for the higher molecular weight
1/2

samples.
2
Whether the data found for the migration of polymer of P 80 through thisn

particular membrane is caused by residual micro- or other defects is unknown at

present. In any case, it is significant that a membrane built by deposition of 50
1 7 8
layers poly(glutamate) on top of Celgard exhibits t « N behavior. It should be
}/2

mentioned that the thickness of an individual monolayer of the cellulose derivatives

is 0.96 nm, as compared to 1.74 nm for the poly(glutamate). Thus, the absolute
value of t is reduced by a factor of 10 comparing the copoly(glutamate) with the
1/2

cellulose system of 70 monolayers.

The last point in each curve corresponds to the molecular-weight-cut-off
(MWCO) of each particular membrane. In other words, the membrane was
impermeable for polymers of a degree of polymerization Ν > N , ^ ^ . The transport of
macromolecules through ultrathin layers of the type discussed here (Figure 1) is also
interesting from a theoretical point of view. The solvent-swollen multilayer part of
the membranes, which obviously controls the transport kinetics, has a thickness d
that is ca. 10-100 times larger than the radius of gyration of the macromolecules

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

Figure 3. Schematic presentation of the crosslinked layer system of hairy

rods on top of the Celgard 2400® polypropylene membrane surface.
August 31, 2012 | http://pubs.acs.org

Figure 4. Crosslinking reaction of the isopentyl cinnamyl cellulose ether 1;

only one of several possible types of crosslinks is shown.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 311

migration. However, adjacent backbones within the layers form a grid with a
distance between the backbones rather rigidly fixed by the side chains and crosslinks
to a value of ca. 05.-1 nm in the swollen state. Thus, transport of the polymer chains
should only be possible by reptation (75). In what way the structure of the layered
assemblies enforces distortions of the unperturbed dimensions of the
macromolecules while they are migrating is not known at present. It will be
particularly interesting to study cases in which the total layer thickness d becomes
smaller or comparable to the radius of gyration, as one could then probably observe
a crossover in the transport behavior.

Adsorption on regenerated cellulose LB films (16-18).

Silylated cellulose derivatives such as trimethylsilyl cellulose (TMSC) can be
regenerated to cellulose by exposure to aqueous acids and thus allow in situ
conversion of T M S C hydrophobic films to hydrophilic films of regenerated cellulose
(Figure 6a) (16, 17). The regeneration is quantitative after 30 s and leads to a well-
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

defined ultrathin cellulose film. Earlier studies could show that (i) in the T M S C L B -
film, as well as in the regenerated cellulose film, the polymer backbones are
preferentially oriented parallel to the dipping direction, and (ii) both films contain
regular and homogeneous structures [10].
Figure 7 shows the regeneration process for a film of 100 layers T M S C . In
contrast to the hydrophobic L B films of silylated cellulose, the regenerated cellulose
films have hydrophilic surface properties, as evident from the static contact angle
with water (78° for hydrophobic TMSC and 23° for hydrophilic cellulose). In
addition, the cellulose multilayer systems are insoluble in most common organic
solvents and water, and are stable against oxidation and thermal degradation. The
obtained regenerated cellulose films have been used as substrates for adsorption of
monomers (dyes) and polymers (18).
In addition, modified cellulose materials have been prepared in order to
introduce carboxyl functionalities. Cellulose with a higher content of carboxyl
groups should show a different adsorption behavior from native analogues. Thus,
August 31, 2012 | http://pubs.acs.org

cellulose succinates have been prepared using ring-opening esterification with

succinic anhydride, as shown in Figure 6b, (18). The adsorption properties of
unmodified and modified films were investigated using high molecular weight
cationic polyacrylamide, (poly(acrylamido propyl)trimethylammonium chloride
A P T A C - C , see Figure 6c), with a content of 6% randomly distributed cationic side-
groups.
The polyelectrolyte A P T A C - C is a typical polymeric additive used in
papermaking processes; it works as a flocculating agent helping to retain filler
particles and fiber fragments in the sheet being formed. Analysis of the surface
plasmon resonance (SPR) spectra of APTAC-C adsorbed on cellulose and modified
cellulose films reveals that the latter films adsorb much more A P T A C - C than does
the unmodified cellulose, and the adsorbed amount increases as the reaction time
with S A is prolonged. The amount of adsorbed polyelectrolyte was doubled when a
film of 20 layers was treated with SA for 3 hours. Films modified with S A for 13.5
h adsorb more than four times the amount of A P T A C - C compared to the unmodified
films (Figure 7). It is noteworthy that the time for reaching the plateau region
increases with the length of the reaction time. When the number of cellulose
monolayers is increased to 100 and the film is treated with S A for 13.5 h, the system
2
absorbs more than 18.7 mg/m and reaches the plateau level after two hours. In the
case of thicker films, one should consider that the modification could also occur in
deeper layers and that the polyelectrolyte could not only adsorb on the surface but
also penetrate into the modified multilayer system. Preliminary investigations on the
influence of salts and an additional polyelectrolyte have shown that it is possible to
detect double-adsorption processes. Following this concept, a simulation of a two-

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 312

+ uncoated Celgard 2400

6 A 60 layer rsopentylcinrtamylceJIulose
• 70 layer isopentykcinnamylcettulose
• 50 layer porygkjtamale

1.0 1.5 2.0 2.5 3.0

log Ν
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

Figure 5. Double logarithmic plot of the half-life time tm of osmotic

pressure versus degree of polymerization Ν of the polystyrene molecular
weight standards.
August 31, 2012 | http://pubs.acs.org

Figure 6. a) Desilylation reaction to obtain regenerated cellulose; b)

treatment of regenerated cellulose with succinic anhydride (SA) giving
cellulose succinate derivatives; c) chemical structure of the polyelectrolyte
APTAC-C.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 313

1,0
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022
August 31, 2012 | http://pubs.acs.org

""20 25 30 35 40 45 50 55 60 65 70 75
Angle Θ [°]

Figure 7. Surface plasmon resonance spectra of (a) the substrate; (b)

regenerated cellulose; (c) TMSC vs. air; (d) cellulose modified with S A for
13.5 h against buffer (1 m M Na (C0 ) and (e) the corresponding
2 3

measurement after adsorption of APTAC-C.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 314
component retention system consisting of a combination of cationic and anionic
polymers should be possible.

Literature Cited

(1) Wegner, G. Mol. Cryst. Liq. Cryst.1993,235, 1.

(2) Wegner, G.; Schulze, M.; Seufert, M.; Schaub, M.; Vahlenkamp, T. Polym.
Prepr. 1995, 598.
(3) Schwiegk, S.; Vahlenkamp, T.; Xu, Y.; Wegner, G. Macromolecules 1992,
25, 2513.
(4) Ferencz, Α.; Armstrong, N. R.; Wegner, G. Macromolecules1994,27, 1517.
(5) Seufert, M.; Schaub, M.; Wenz, G.; Wegner, G. Angew. Chem. Int. Ed. Engl.
1995, 34, 340.
(6) Whitten, D. G. Angew. Chem. Int. Ed. Engl. 1979, 18, 440.
(7) Eggl, P.; Pink, D.; Quinn, B.; Ringsdorf, H.; Sackmann, E. Macromolecules
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch022

1990, 23, 3472.

(8) Tieke, B. Adv. Polym. Sci. 1985, 71, 79.
(9) Müllen, K.; Böhm, Α.; Fiesser, G.; Garay, R. O.; Mauermann, H.; Stein. S.
Polym. Prepr. 1993, 34, 195.
(10) Schaub, M.; Mathauer, K.; Schwiegk, S.; Albouy, P.-A.; Wenz, G.; Wegner,
G. Thin Solid Films 1992, 210/211, 397.
(11) Seufert, M.; Fakirov, C.; Wegner, G. Adv. Mater. 1995, 7, 52.
(12) Sarada, T.; Sawyer, L.; Ostler, M. J. Membr. Sci. 1983, 15, 97.
(13) Iida, S.; Schaub, M.; Schulze, M.; Wegner, G. Adv. Mater. 1993, 5, 564.
(14) Coll, H. Makromol. Chem. 1967, 109, 38.
(15) de Gennes, P. G. Scaling Concepts in Polymer Physics, Cornell University
Press, Ithaca 1979.
(16) Schaub, M.; Wenz, G.; Wegner, G.; Stein, Α.; Klemm, D. Adv. Mater. 1993,
5, 919.
(17) Schaub, M.; Fakirov, C.; Lieser, G.; Wenz, G.; Wegner, G.; Albouy, P.-A.;
August 31, 2012 | http://pubs.acs.org

Schmidt, Α.; Majrkzak, C.; Satija, S.; Wu, H.; Foster, M. D. Macromolecules
1995, 28, 1221.
(18) Buchholz, V.; Wegner, G.; Stemme, S.; Ödberg, L. Adv. Mater. 1996, 8, 399.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 23

Surface Segregation Phenomena in Blends

of Cellulose Esters

Ulrike Becker, Jason Todd, and Wolfgang G. Glasser

Biobased Materials/Recycling Center and Department of Wood Science and Forest

Products, Virginia Polytechnic Institute and State University,
Blacksburg, VA 24061
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

The principle of surface segregation and subsequent self-assembly of

mixtures of molecules has been utilized to create cellulosic films that
have properties characteristic of smart materials. This has been
achieved by solvent-casting a blend of cellulose propionate and an F-
-containing cellulose ester derivative. Evidence suggests that surface
segregation of the F-containing species takes place spontaneously in
cellulosic systems. Cellulose propionate films containing a small
amount of F-containing cellulose derivative exhibit a much lower
surface free energy (i.e. are more hydrophobic) than predicted by the
rule of mixing, and this indicates surface segregation. The current
research examined the influence of the chemistry of the fluorine-
August 31, 2012 | http://pubs.acs.org

-containing component (type of derivative and chemistry of the F-

-containing group) on the segregation process.

A smart material is defined as a structure "...having the following attributes: it senses

the environment, it responds in a purposeful way through design or by means of a
logic controller element" (1). The surface segregation and subsequent self-assembly
observed in some multicomponent films can be classified as a smart process, and thus
the resulting film is a smart material.
The principle of surface segregation (see below) allows constructing films with
surface energetics distinctly different from the bulk phase. Natural polymers with
hydrophobic surfaces can be created in this manner, although the bulk retains the usual
hydrophilic nature which is due to a high oxygen content. This creates new
opportunities for applications of natural polymers. For example, cellulose and its
derivatives have achieved prominence in the biomedical field as kidney dialysis
membranes. However, advancements in this field have been limited because of the
limited surface energetics of cellulosic materials. As polymers of natural origin,
cellulose and its derivatives are prime candidates as raw materials for biomedical
implants because of their general biological compatibility and biodegradability.

©1998 American Chemical Society 315

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 316

However, for an implant material to function at the interface with living cells and body
fluids, characteristic surface requirements must be met which prevent the spontaneous
formation of blood clots. The general ability of foreign materials to cause thrombosis
constitutes one of the principal limitations of biomedical implant materials. While it is
recognized that the process of blood clotting is initiated by the adsorption of proteins
on the surface of the implants, consensus has not yet been reached over whether this
process is aggravated by excessively hydrophilic or hydrophobic surfaces (2,3).
General agreement has, however, been reached over the issue of the importance of
surface engineering in terms of polarity. Thus, the ability to engineer the critical
surface free energy (y ) of implant materials is pivotal to the development of
c

biomedical implants (4). Most studies indicate that thrombogeneity increases with
increasing y (5), and that simultaneously, clottingtimedecreases (6). This suggests
c

that a more hydrophobic surface is less prone to thrombosis. It is therefore of interest

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

to examine methods of engineering the surface characteristics of cellulose-based

materials in terms of their surfacefreeenergy.
While fluorine plasma treatment represents the easiest and most widely
practiced method of surface modification, negative toxicological effects have been
noted from this method which result from the partial molecular breakdown and
leaching of toxic degradation products into the surrounding medium (7). Utilizing the
smart process of self-assembly, it is possible to create cellulose-based materials with
hydrophobic surfaces for potential use as biomedical implant materials.

The Principle of Spontaneous Self-Assembly by Surface Migration

The principle of spontaneous segregation of blend components by migration of one

component to the surface on the nano-scale has been employed for engineering bi-
August 31, 2012 | http://pubs.acs.org

layered materials in the past. By mixing two incompatible polymers with each other in
a blend, molecular composites with a distinct surface layer, which is differentfromthe
bulk phase, can be obtained.
Surface migration is a thermodynamically-driven process. Fundamental
thermodynamics dictates that a system adopt the state of lowest possible total free
energy under equilibrium conditions. In the case of a film consisting of at least two
polymers, three quantities have to be taken into account: 1. entropy of mixing; 2.
interaction energy between polymers; and 3. surfacefreeenergy (8). Quantities 1 and
2 are given by the composition of the system. At a given film composition, however,
the surfacefreeenergy, which is the interfacial energy between thefilmand air (which
is considered to be hydrophobic) can be subjected to changes. Duringfilmformation
the system rearranges itself so that the lowest free energy component resides
preferentially at the surface, and therefore the lowestfreeenergy of the entire system
is gained. Therefore, surface migration can be used to create a system with a low
surfacefreeenergy, i.e. a hydrophobic surface.
Sincefluoro-carbonmolecules are known to have low surface free energies,
they are likely to migrate to the surface in a mixture of F-free and F-containing
molecules. The goals of this study were to examine the surface migration effects of
two types of novel F-containing cellulose derivatives in a blend with cellulose

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 317

propionate. Thefirsttype of F-containing component was a mixed ester of cellulose

propionate (CP) with fluoroalkoxy acetate groups whereby the F-containing
substituents were randomly distributed, and the second type of component was a
fluorine-terminated cellulose propionate oligomer with exactly one F-containing
endgroup (ie, with a blocky type of architecture). Both F-containing species were
expected to surface-segregate, but differences in migration behavior were expected
depending on the difference in the chemistry of the derivative and the type and the
distribution of F-containing group. Two different F-containing groups were used,
trifluoroethyl (with CF terminus) and octafluoropentyl (with CF2H terminus)
3

substituents.
Energetic effects that stem from the presence of atoms other than hydrogen
(X) in hydrocarbon containing molecules are the cause for the surface migration of F-
containing molecules. These effects are described with nearest neighbor interactions,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

χ, is defined as the difference between the adsorption energies (U ) of the segments

a

of the components (a = 1 or 2 in a two component system),

χ, = Ui - U2 (equation 1),
where
U = u« /k T
a B (equation 2),
and
ιΐα Ξ θα· -0.5cDaa (equation 3)
with ke as the Boltzmann constant, Τ the temperature, ow as the intracomponent
interaction energy, and ω» the surface interaction energy. u« represents the difference
in adsorption energies between the surface and the bulk (9). Thus χ. is a measure for
the adsorption energy difference between the segments of the two constituent chains,
and this is a measure for the preference of one component to be at the surface over the
other (10). In the case of a mixture at an energetically neutral surface (like air) the
August 31, 2012 | http://pubs.acs.org

term simplifies to
χ. = - (on - Q>22)/kBT (equation 4)
with (On and 022 as the interaction energy between the monomers in the bulk where
O n is larger than 022. A positive χ. is an expression for the fact that the system
partitions the component with the relatively less favorable monomer interaction in the
bulk, i.e. less attraction forces between the monomers, to the surface (11). This
species experiences less energy loss due to the smaller number of neighbors at the
surface than a species with a large interaction energy. In systems like C-H and C-F,
possible interactions arisefromLondon forces. The C-F bond has a high polarity, but
at the same time, due to the high electronegativity of the fluorine atom, a low
polarizability. London forces are proportional to polarizability. The C-F containing
species has therefore weaker London forces and thus less interaction energy when
compared to a C-H containing species. This results in a positive χ,-parameter and it
drives the partitioning of thefluorinatedcomponent to the surface. Altogether, the
result is a lower surface free energy. The more F-atoms present, the weaker are the
London forces. As a result, the χ, parameter becomes more positive. This means that
the driving force for surface migration increases.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 318

The F-containing, randomly-substituted cellulose esters have more F-

containing substituents per molecule than an F-terminated oligomer, which has exactly
one F-containing substituent (Figure 1). Thus the driving force for the random
copolymer is larger on a molecular basis. But at the same time, the F-containing
groups of the terminated oligomers are independent from each other, whereas in the
random copolymer many F-substituents are connected through the cellulose backbone.
In addition the oligomers are more mobile and encounter fewer entanglements due to
their smaller size. Thus it is expected that the surface migration of the oligomers is
more efficient when compared to the random copolymers on the basis of number of F-
containing groups in the system. Consequently, blends containing F-terminated
oligomers are expected to yield materials with lower surface free energy than blends
with randomly substituted F-containing copolymers.
Comparing the F-containing substituents alone shows two effects depending
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

on the chemistry of the substituent. First, the χ. parameter increases with increasing
number of F-atoms in the group. In the case of an energetically neutral surface, the
value of χ, can be deduced from the cohesive energy density (CED) since the CED is
proportional to ωαο (12). For polymers, the CED can be calculated through group
contribution as the sum of the molar attraction constants, F, for the group (13). For
every proton that is replaced by a fluorine atom, the interaction energy decreases.
Therefore, the more Η-atoms are replaced by F-atoms, the lower is the interaction
energy of the species. Consequently, χ. increases. As χ. represents the driving force
for the surface segregation, a larger value of χ, indicates a stronger "pulling force" for
the surface segregation and the equilibrium state is expected to be reached in a shorter
time. Accordingly, derivatives containing the octafluoropentyl group are expected to
show a larger degree of surface segregation and therefore a more hydrophobic surface
when compared to the trifluoroethyl-containing derivative of the same type.
August 31, 2012 | http://pubs.acs.org

The overall objectives of this study were to synthesize novel types of F-

containing cellulosic derivatives and to examine their surface segregation properties in
relation to fluorine content and polymer architecture. The F-containing components
were either randomly-substituted cellulose esters with variable F-content and with
either CF2H- or CF -terminal groups, or they were F-terminated cellulose propionate
3

oligomers.

Experimental

L Materials. Micro-crystalline CF-11 cellulose was purchased from Whatman

Chemicals. Cellulose propionate was obtained from Eastman Chemical Co. A l l
solvents were obtained from Fisher Scientific and used as received, except
tetrahydrofuran for the synthesis of the F-terminated oligomers, which was used
freshly distilled over sodium. P-toluene-sulfonyl chloride (TsCl) , toluene diisocyanate
(TDI), and other reagents were obtained from Aldrich Chemical Co. and used as
received.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

F-terminated oligomer
Random Copolymer ("blocky" structure)

high molecular weight low molecular weight

August 31, 2012 | http://pubs.acs.org

high molar F-density low molar F-density

Figure 1. Schematic illustration of F-containing cellulose ester options.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 320

IL Methods

L Copolymer synthesis. Cellulose mixed esters with fluorine-containing

substituents randomly-distributed along the cellulose backbone were synthesized via
homogeneous phase reaction of cellulose solution in dimethylacetamide / lithium
chloride (DMAc/LiCl). This reaction is described in detail elsewhere (14). Briefly, a
fluorinated acid was reacted with cellulose in solution via a mixed anhydride
intermediate formed with p-toluene sulfonyl chloride (TsCl). Thefluorinatedacids
used were trifluoroethoxy acetic acid and octafluoropentoxy acetic acid. The acids
were synthesized by reaction of either 2,2,2-trifluoroethanol or 2,2,3,3,4,4,5,5-
octafluoropentanol, respectively, with chloroacetic acid in water under reflux with
strong alkali. Pyridine, followed byfluorinatedacid, andfinallyTsCl, were added to
cellulose in DMAc/LiCl solution. The reaction mixture was stirred overnight at 40-
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

60°C. The resulting randomly-substitutedfluorinatedcellulose esters had a degree of

substitution offluorinatedsubstituent (DSF) of approximately 1.5 (by NMR and/or
elemental analysis). The remaining cellulose hydroxyl grops were subsequently
acylated via reaction with acetic or propionic anhydrides, resulting in cellulose mixed
esters with a random distribution offluorinatedand unfluorinated ester substituents.
Molecular weight distributions of the random mixed esters were determined by GPC.
Fluorine-terminated cellulose propionate oligomers were prepared using
monofunctional (OH-terminated) cellulose esters of variable degree of polymerization
and coupled withfluorine-containingalcohols. In brief, the procedure involvesfirstthe
hydrolytic degradation of cellulose propionate into CP oligomers of various degrees of
polymerization (DP) as described elsewhere (15). The resulting oligomers were
endcapped with toluene diisocyanate according to deOliveira and Glasser (16). The
isocyanate terminated oligomers were immediately coupled with afluorinatedalcohol.
Typically, 3 grams of the predried endcapped oligomer were placed in a 250 ml three
August 31, 2012 | http://pubs.acs.org

neck round bottomflaskwith 0.01 ml of stannous octanoate as a catalyst. Two necks

of the flask were closed with rubber septa, then the flask was placed in a
polycarbonate desiccator with rubberseal and connected to a vacuum pump overnight.
The flask was then connected to a condenser andflushedwith prepurified nitrogen
gas. Subsequently, 60 ml offreshlydistilled THF were introduced into theflaskusing
syringe and needle. Theflaskwas put into an oilbath and heated to 50°C. At that time,
a 5 molar excess of the F-containing alcohol was added via syringe. Theflaskwas
stirred at 50°C for 24 hours under nitrogen. After that, it was allowed to cool to room
temperature. The product was precipitated in petroleum ether under vigorous stirring.
The precipitate was filtered and washed several times with petroleum ether. The
product was dried and kept until further use in a small polypropylene container.

2. Copolymer characterization. Copolymer characterization involved *H and

19
F-NMR spectroscopy, gel permeation chromatography (GPC), differential scanning
calorimetry (DSC), and solubility studies. The molecular weight and the distribution of
hydroxyl-terminated CP oligomers were determined by GPC with a differential
viscosity detector (Viscotek Model No. 100) and a differential refractive index
(concentration) detector (Waters 410) in sequence. The system was controlled by

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 321

Viscotek software (Unical GPC software, Version 3.02). The CP oligomers were
dissolved in THF and analyzed using a high pressure liquid chromatography system.
The calculations were based on an universal calibration curve using polystyrene
standards.
19
F-NMR spectra were recorded on a Varian 400 MHz spectrometer. The
samples were dissolved in protonated THF and run without lock. In order to
accurately determine peak shifts, all samples were run with the addition of 3-
(trifluoro)methyl benzophenone as an internal standard.
'H-NMR was used to determine the degree of polymerization through
endgroup determination. The analysis was conducted on a Varian 400 MHz
instrument with deuterated chloroform as the solvent. In order to increase the
detection sensitivity of the hydroxyl proton, the oligomers were silylated at the lone
terminal hydroxyl group as described elsewhere (17). The DP was calculated as a ratio
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

of the peak areas for the silyl protons to the cellulose backbone protons.
DSC measurements were conducted on a Perkin Elmer model DSC4 with a
Perkin Elmer Thermal Analysis Data Station. The temperature was scanned between
-30°C and +270°C at a heating rate of 10°C/min. The samples were subjected to three
heating and three cooling cycles. The glass transition temperatures (Tg) were taken as
the mid-point of the step-function change in slope of the baseline and the melting
transition was taken as the temperature corresponding to the maximum point of the
endothermic peak.
The solubility of the F-containing samples was tested in various solvents.
Typically a small amount of the sample was transferred into a 4 ml glass vial and
solvent was added. If completely dissolved, the resulting solution had a concentration
of about 0.5-1%. The vial was equipped with a magnetic stir bar and stirred overnight.
At that point solubility was determined visually.
August 31, 2012 | http://pubs.acs.org

3. Blend Preparation. Cellulose propionate/F-containing cellulose propionate

(CP/CP-F) blends were prepared volumetrically. This involved initially the preparation
of stock solutions of CP and CP-F in THF. Blends were prepared by mixing these
solutions to the desired content of F-containing species. The total solids content of all
solutions was 5% w/v.

4. Contact Angle Measurements. Cellulose ester and cellulose ester blend

films were characterized by contact angle (CA) measurements. This was based on a
modified Wilhelmy Plate method. Microscope cover slides were cleaned thoroughly in
hexanes, dried and subsequently dip-coated with the respective solutions. The solvent
was evaporated at 4°C over night and allfilmswere stored for three days before the
contact angle measurements. Great care was taken to ensure that all films were
prepared in the same fashion. The contact angle measurements were conducted with
deionized water as the wetting medium using a CAHN Analyzer, controlled both
manually and by CAHN control software DCA2d Version 2.0. The instrument
recorded the force depending on the immersion depth, and the water contact angle
was calculated using the force and the surfacefreeenergy of water. Six measurements
were performed per sample.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 322

Results and Discussion

1. Synthesis. The synthesis of F-containing random copolymers and block

copolymers is schematically illustrated in Figure 2. Monofunctional, OH-terminated
cellulose propionate oligomers with variable degree of polymerization were prepared
by partial hydrolytic degradation with HBr under esterification conditions.
The molecular weights of the cellulose propionate segments were determined
by GPC and N M R spectroscopy (Table I). The two methods are mutually supportive.
For N M R , the signal of the one O H end group was enhanced by reaction with chloro-
triethylsilane. The 15 ethylsilane protons show a characteristic and easily distinguished
proton peak at 0.55 ppm. Integration of this peak and the peaks associated with the
cellulose backbone protons is used to calculate the degree of polymerization. At high
block sizes (large DP), the enhanced ethylsilane signal becomes less distinct and block
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

size determination by N M R becomes less accurate. A comparison of DP

determinations by N M R and GPC (Figure 3) reveals that N M R and GPC produce
almost identical results at low DP-values whereas at high block sizes Qsige DP), the
determination by N M R yields a higher DP value. At higher molecular weights, a
complete reaction of the chloro-triethylsilane becomes more difficult. This accounts
for the fact that the DP values obtained by N M R are higher than those found by GPC.
The OH-terminated oligomers were subsequently end-capped with toluene
diisocyanate resulting in a monofunctional terminated oligomer. The resulting N C O -
terminated cellulose propionate oligomer was immediately (to avoid degradation of
the remaining isocyanate groups) reacted with F-containing alcohols in which both F-
content and terminal functionality (CF2H versus CF ) vary (Figurel). The alcohols
3

were trifluoroethanol (CF -terminus) and octafluoropentanol (CFiH-terminus). The

3

reaction of the F-containing alcohols and the cellulose propionate oligomer was
19
confirmed by F - N M R spectroscopy. As a result of the formation of a terminated
August 31, 2012 | http://pubs.acs.org

Table L Molecular weight characteristics of cellulose propionate oligomers

Oligomer Molecular Weight (Mn) Molecular Weight

Designation byGPC 2)
by H - N M R 3)
Distribution (MWD)
1}
B-7 2375 2150 1.4
B-14 4480 5000 1.6
B-28 9380 11250 1.5
B-70 24100 26300 1.4

The number behind B - denotes the DP of the segment.;

Using THF as solvent and a universal calibration curve.
Following OH-modification with chloro-triethyl silane (see Experimental Section).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 323

.OH
Ο
OH
HO-
propionic acid/ HO DMAc/LiCl
anhydride
CH 2

Ο
I
OPr
Pr+ Ο CH 2

PrO. ,OPr
T>rO I
Η - Ο - γ ^ρ - Ο
-•η
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

HBr
C=0
propionic anhydride I
CH 2

I
Ο
0 P r
I
P rι + O ^ X
/
^ O Iι
CH 2

PrO-
R

propionic acid/

TDI Î anhydride

R
ΌΡΓ Ο !
CH
-o^X^o \\ 6
2
PH
August 31, 2012 | http://pubs.acs.org

0
? & \ ^ r ^ \ ' ^ ^NH—C H*-NCO
7
I
CH 2
1
I
c=o
I
ρ
F-OH PrfO Ο
PrO
,OPr
Ο
I
c=o
P r - O ^ C ^ O (J I
0
CH 2

ProV-^-\' ^ ^NH-Cy^-NH' O-R I

Ο
I
CH 2

R=CF 3
R
R=CF CF CF CF H
2 2 2 2

Figure 2. Reaction schemes for blocky oligomer (left side) and random
copolymer (right side); η indicated the high degree of polymerization of
cellulose propionate, which is maintained in the random copolymer, whereas m
indicated the reduced degree of polymerization in the blocky oligomer which is
achieved by hydrolytic degradation.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 324
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023
August 31, 2012 | http://pubs.acs.org

10 15 20 25 30 35 45
DP by GPC (direct determination)
Figure 3. Comparison of DP determination by GPC and by NMR; GPC
determination represents direct determination, while NMR is an indirect
determination by means of determining the endgroup concentration.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 325

19
oligomer, the CF -peak in the F - N M R spectrum shifted downfield from -14.4 ppm
3

19
to -11.5 ppm (Figure 4). A l l F - N M R samples were run with an internal standard (3-
19
(trifluoro)methyl benzophenone). This was necessary because F-resonances often
shift, consequently, the shifts can only be accurately measured when the CF -peak 3

position is observed in relation to a standard peak. The F-containing cellulose

derivatives used in this study are summarized in Table Π.

2. Characterization of F-containing cellulose derivatives. The characteristics of

random and block copolymers containing fluorinated substituents involve solubility
testing and thermal analysis. The solution characteristics of copolymers in various
solvents are shown in Table ΙΠ. The results reveal significant differences between the
random copolymers and the blocky oligomers. Among the random copolymers,
differences in solution behavior depended on the nature of the F-containing
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

substituent. Copolymers containing a substituent with a CF -terminus were soluble in

3

non-polar organic solvents like chloroform and dichloromethane, as well as in acetone.

In contrast, copolymers with an F-containing group with a CF H-terminus are not
2

only soluble in these solvents but also in such alcohols as methanol and ethanol.
This behavior is attributed to the lone proton of the CF H-terminus. This
2

proton is covalently linked to a carbon to which two F-atoms are attached. Due to the
high electronegativity of the F-atoms, the carbon has a strong positive partial charge.
This charge increases the negative inductive effect of the carbon onto the hydrogen
atom. Altogether, the proton becomes very electron deficient and as such is able to
engage in secondary interaction, like for example hydrogen bonding, with the
hydroxyl-groups of the alcohols. These interactions lead to the dissolution of the
copolymer in alcoholic solvents.
The end-terminated oligomers, on the other hand, do not exhibit a solubility
August 31, 2012 | http://pubs.acs.org

behavior that depends on the chemistry of the F-containing endgroup. A l l oligomeric

derivatives were formed turbid solutions in all tested solvents and at all
concentrations, and did not settle out in several weeks. This behavior also did not
depend on the molecular weight of the cellulose propionate segments or the type of F-
containing endgroup.

Table Π. Identification of F-containing derivatives used in this study

Designation Description
B-14-CF 3
oligomer with DP = 14, terminated with
trifluoroethyl group
B-H-CF2H oligomer with DP = 14, terminated with
octafiuoropentyl group
R-IOO-CF3 random copolymer (DP = 100) with
trifluoroethoxy acetate substituent (DSF
= 1.5), peracetylated
R-IOO-CF2H random copolymer (DP = 100) with
octafluoropentoxy acetate substituent
(DSF = 1.5), perpropionated

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 326

CF3-signal in terminated
standard oligomer
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

CF3-signal in alcohol

/ 1
oligomer/alcohol mixture
August 31, 2012 | http://pubs.acs.org

******* κ . *******

Chemical Shift in F-NMR Spectrum

Figure 4. 19F-NMR spectra of B-14-CF (top); trifluoro ethanol (middle); and

3

B-H-CF3 spiked with trifluoro ethanol. The standard (3-(trifluoro)methyl

benzophenone) is arbitrarily set to 0 ppm.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 327

Table ΙΠ. Solution behavior offluorinatedcellulose derivatives; sample

designation is as outlined in Table Π.

sample solvent
acetone dichloromethane chloroform ethanol methanol
B-14-CF3 Τ Τ Τ Τ Τ
B-14-CF2H Τ Τ Τ Τ Τ
R-100-CF2H ++ + + ++ ++
R-100-CF3 ++ + + - -

Τ = turbid solution
+ = soluble with some difficulty
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

++ = rapidly soluble
- = insoluble

Consistent with the solution behavior of other amphophilic substances, it is at this

point hypothesized that this behavior reflects the formation of micelles where the F-
containing endgroups form the core, and the corona consists of cellulose propionate
residues (Figure 5).
The examination of thermal properties of copolymers by DSC produces results
consistent with the generic architecture of the different derivatives (Figure 6).
Whereas the random copolymers revealed a gradual, transitionless decline in T and g

T with increasing content of fluorinated substituents, the block copolymers revealed

m

thermal transitions that were identical to those of the parent cellulose propionate. The
fluorinated end groups in the block copolymers, constituting only a minor part of the
August 31, 2012 | http://pubs.acs.org

molecule, proved to have an insignificant influence on the thermal copolymer

properties.

3. Contact angle measurements of blended films. Films prepared from blends of

cellulose propionate and fluorinated cellulose derivatives were characterized by
contact angle measurements. Preliminary results revealed that water contact angles
varied between approximately 75 and 110°, and the degree of variation from the CP
control depended on fluorine content, substituent type, and polymer architecture in a
significant way (Figure 7). The pure random copolymer film with a CF2H-terminus
achieves a maximum advancing water contact angle of 110°. This angle is lower than
values found in the literature for fluorinated surfaces (18). However, the literature
values describe surfaces with CF -groups, and the present case examines a surface
3

containing CF2H-groups. The lower contact angle is again explained by the presence
of the lone proton at the CF2H-terminus which can engage in hydrogen bonding with
the water molecules in the same fashion as outlined above.
The copolymer shows an exponential increase in contact angle at low blend
contents of F-containing component. This behavior represents a deviation from the
rule of mixing which validates the segregation phenomenon of the F-containing
species at the surface.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 328
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

Figure 5. Schematic representation of micelles believed to be formed by F-

terminated cellulose propionate oligomers. The circles represent the endgroups
which from the core of the micelle and the lines represent the cellulose
propionate chains in the corona.

300

250
August 31, 2012 | http://pubs.acs.org

R-100-CF3 R-100-CF2H CP B-14-CF3 B-14-CF2H

Figure 6. Melting transitions of cellulose propionate (CP) compared to blocky

and randomly fluorinated derivatives: thermal behavior of the blocky
oligomers is similar to CP and independent of the chemistry of the F-
containing endgroup but random copolymers show a clear dependence of
thermal transitions on the chemistry of the F-containing substituent. No
melting transition for R-IOO-CF2H was observed. Sample abbreviation is as
outlined in Table Π.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 329

120 j
115
110 --
105 --
I
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

< 100
3 95 -n

δ 90 -f
Ui

a 85 f
4
Wa

80
75 *

70 4-
20 40 60 80 100

Weightfractionof F-species in Blend (%)

F-content
in blend (%):
August 31, 2012 | http://pubs.acs.org

-B-CF,H 0 • 1.2
-B-CF3
-R-CF H 2 0 • 14.9

Figure 7. Experimental water contact angle for blends with random copolymer
R-CF2H (with octafluoropentyl substituent) (•); and for two blends with
blocky oligomers: B-CF3 (trifluoroethyl-terminated) (Δ) and B-CF2H
(octafluoropentyl-terminated) (o). The dashed line for the B-CF3 indicates
extrapolated values. Pure B-CF could not be cast into a stable film. Note that
3

the F-content (in %) is higher in the random copolymer than in the blocky
oligomers

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 330

The terminated oligomers, however, show a linear increase of contact angle

with increasing amount of CP-F in the blend. In this case the rule of mixing is
followed. The apparent lack of surface enrichment is explained by micelle formation.
The F-containing endgroups are "locked" into the core of the micelles and are
therefore prevented from orienting themselves to the surface. This also explains the
fact that the oligomers are less effective in increasing the water contact angle when
compared to the random copolymer. The pure blocky oligomer with a CFiH-terminus
achieves a water contact angle of only 93°.
The effect of the nature of the F-containing group is examined in two blocky
oligomers having an octafluoropentyl and a trifluoroethyl endgroup, respectively. The
octafluoropendyl group is comparatively F-rich and is expected therefore to result in a
higher contact angle; but preliminary results show that the oligomer with the trifluoro
ethyl endgroup has a higher contact angle. This behavior is again explained by the
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

secondary interaction of the lone proton with water. The expected increase in water
contact angle in the F-rich octafluoropentyl terminated oligomer is off-set by the
interactions of the proton of the CF2H-terminus.

Conclusion

Surface engineering of cellulose ester derivatives by blending with fluorine containing

cellulose ester copolymers appears to follow expectations within the following
framework.
The nature of the terminus on the F-containing substituent, CF2H versus C F , 3

substantially determines surface characteristics. The observed differences are

explained with the presence of an electron deficient proton in the CF H-terminus
2

which readily engages in secondary interactions, like for example hydrogen bonding.
August 31, 2012 | http://pubs.acs.org

This overshadows any effects that stem from the difference in atomic F-density per
substituent.
The type of copolymer architecture, random versus block design, has a
significant impact on both, solubility and surface character. The observations are
consistent with the view that fluorinated block copolymers can form micelles of the
type shown in Figure 5. The micelles are thought to consist of a fluorine-rich core
with a cellulose propionate corona. They are virtually insoluble in all common
solvents, and they form stable colloids in solution and in solid state. This accounts for
the observed modest influence of the blocky copolymers on contact angle. Random
copolymers cannot form micelles because of the delocalized distribution of the F-
containing groups. Therefore they are both soluble in appropriate solvents and more
effective in increasing the contact angle.

Acknowledgment

This study was financially supported by a grant from the U . S. Department of

Agriculture, CSREES Contract #96-35103-3835. This support is acknowledged with
gratitude. Thanks is also given to Ms. Jody Jervis, Dep. WSFP, for the help with the
molecular weight determination.

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 331

This paper represents part I of a publication series entitled "Smart Surfaces of

Biobased Materials".

Literature Cited

1. Simmons, W.C. In Smart Materials Technologies. Simmons, W.C., Aksay, I.Α.,

Huston, D.R., eds., Proceedings SPIE, SPIE: Bellingham, WA, 1997, Vol. 3040,
pp 2-7
2. Lyman, D.J.,Knutson K. In Polymeric materials andpharmaceuticalsfor
biomedical use. Goldberg, E.P., Nakajima, A., eds., Academic Press, NY, 1980
3. Perez-Luna, V.H., Horbett, T.A., Ratner B.D. J.Biomed.Mat. Res., 1994, 28,
1111
4. Biomaterials and interfacial approach. Hench, L.L.,. Ethridge, E.C., eds..
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch023

Biophysics and Bioengineering Series. Academic Press, NY, 1982, Vol. 4

5. Harrison J.H. Am. Jt. Surg, 1958, 95, 3
6. Lyman, D.J., Muir, W.M., Lee, I.J. Trans. Am. Soc. Artif. Intern. Organs. 1965,
11, 301,
7. Poncin-Epaillard, F., Legeay, G., Brosse, J. J. Appl. Polym. Sci., 1992, 44, 1513,
8. Pan, D.H.,. Prest, W.M. Jr. J. Appl. Phys. 1985, 58 (8), 2861,
9. Hariharan, Α., Kumar, S.K., Russell, T.P. J. chem. Phys. 1993, 98 (8), 6516,
10. Hariharan, Α., Kumar, S.K., Russell, T.P. MacromoIecules, 1991, 24, 4909,
11. Hariharan, Α., Kumar, S.K., Russell, T.P. J. chem. Phys. 1993, 98 (5), 4163,
12. Hiemenz, P.C. Polymer Chemistry. The basic concepts. Marcel Dekker, Inc., NY,
1984
13. Cowie, J.M.G. Polymers: Chemistry andphysics of modern materials. Blackie
Academic and Professional, London, 1991
August 31, 2012 | http://pubs.acs.org

14. Sealey, J.E., Samaranayake, G., Todd, J.G.,. Glasser, W.G. J. appl. Polym. Sci. B:
Polym. Phys., 1996, 34, 1613,
15. Mezger, T.,. Cantow, H.-J. Polym. Photochem., 1984, 5, 49,
16. deOliveira, W., Glasser, W.G. Polymer, 1994, 35 (9), 1977,
17. deOliveira, W., Glasser, W.G. Cellulose, 1994, 1, 77,
18. Wang, J.-H., Claesson, P.M., , Parker, J.L., Yasuda, H. Langmuir, 1994, 10, 3887

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Chapter 24

Relation Between the Conditions of Modification

and the Properties of Cellulose Derivatives:
Thermogelation of Methylcellulose

J. Desbrieres, M . Hirrien, and M . Rinaudo

Centre de Recherches sur les Macromolécules Végétales (CERMAV-CNRS),

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

affiliated with Joseph Fourier University, BP 53, 38401 Grenoble Cedex 9, France

The most important natural polysaccharides consist of cellulose,

amylose and amylopectin produced by plants, and chitin from
crustaceous shells. Because they are rich in hydrogen bonds, they
are insoluble or have low solubility in water and have high
cohesion. Usually their derivatization is performed in heterogeneous
conditions leading to a heterogeneous distribution of the substituents
along the macromolecular chain and non-reproducible properties
(depending on their origin). In this work conditions for
homogeneous chemical modification of cellulose are proposed. A
specific substitution on C-2 and C-3 positions is also carried out. A
wide spectrum of samples with different degrees of substitution and
August 31, 2012 | http://pubs.acs.org

different distributions of substituted units is obtained. These results

lead to a better understanding of the methylcellulose (MC) gelation
mechanism and of the role of the chemical structure on the properties
of MC solutions (either in dilute or semi-dilute regime). The
experimental data presented allow discrimination among the
proposed mechanisms. In particular, the gelation phenomenom is
initiated by hydrophobic interactions involving zones of
trisubstituted units.

Naturally occurring polymers such as cellulose, starch components, or chitin are

produced each year on an enormous scale. Each offers advantages, such as being
renewable and biodegradable, properties that are more and more valuable for
environmental protection. Nevertheless, in their native forms they have a limited range of
applications. Therefore, derivatives are needed in which both the nature of the substituent
and the degree of substitution can be varied, in order to obtain material adequate for the
considered application. For these important polysaccharides, derivatization is able to
cover the range of solubilities in organic solvents and water or to produce thermoplastic
materials.
Due to their structure and organization in plants or animals, polysaccharides lack
solubility and usually result in heterogeneous chemical modifications. Such reactions lead
to non uniform substituent distribution on repeating units. Heterogeneity of the
substitution gives a poor predictability of the derivatives properties and may explain, for

332 ©1998 American Chemical Society

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 333

example, the variety and variability of the coefficients relating a physical property to the
molecular weight found in the literature. To compete with synthetic polymers,
polysaccharides must have a regular distribution of substituents and must produce
samples with reproducible characteristics. In order to control the conditions of reaction,
the reactions must be carried out in a homogeneous phase. In this paper the homogeneous
conditions to prepare methylcellulose (MC) and the relationships between the structure of
such derivatives and their physicochemical properties are presented. They are compared
with commercial heterogeneously prepared samples, and the mechanism of the
thermogelation of aqueous solutions is discussed.

Results and discussion

Reaction conditions. Generally the chemical modification of polysaccharides begins

in heterogeneous conditions producing progressively soluble derivatized molecules which
separate from the substrate when the synthesis is performed in solubilizing conditions for
the derivatives. To perform a homogeneous substitution in all the range of degrees of
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

substitution (DS), it is necessary to first solubilize the polysaccharide. This is, in our
opinion, the most important step. Very few direct solvents are available especially for
cellulose and chitin. The swelling and/or the dissolution in different experimental
conditions depend on the morphology of the native substrate. The accessibility of the
reactive groups will depend on this supermolecular structure in addition to the degree of
crystallinity (diffusion control of the reactants...) and the intra- or interchain hydrogen
bonds giving a modulation of the reactivity of the hydroxyl groups in 2,3 or 6 positions.
Only complete destruction, passing through a solution state, will enable the memory
effect of the original state of organization to be avoided. Along the macromolecular chain,
statistic substitution will be obtained, giving a well-defined derivative with predictable
physical properties, especially in relation to the molecular weight and the degree of
substitution.
For cellulose, the solvent DMAc/LiCl proposed from 1979 seems to be the most
promising (7,2). The advantage of this mixture is to be a direct solvent without
derivatization nor depolymerization. Usually the L i C l content adopted is between 5 and
August 31, 2012 | http://pubs.acs.org

9% wt/wt related to D M A c solvent, and the reactions are often carried out under mild
conditions (5). Due to its polar aprotic character, a range of organic reactions is possible
and in particular the modifications of hydroxyl groups. Many reactions are listed in the
papers from McCormick and Dawsey (7-5). The role of the solvent is first to disrupt the
hydrogen bond network and then to be a solvent of the polyhydroxylated polymers (i.e.,
to be polar). Another method usually applied for etherification uses alkaline conditions.
The advantage of alkaline conditions (depending on the temperature and alkaline
concentration) is not only to disrupt the supermolecular structure but also to form the
alcoolate (Cell-O") in concentrate conditions (pH>13) (4), which is the first step in
etherification. These alkaline conditions remain powerful for etherification
(carboxymethylcellulose production). Some time ago sodium in liquid ammonia was used
to produce C M C with very high DS (up to 3). These conditions were adopted to obtain
intercrystalline and crystalline -OH accessibility (5).
Industrial preparation of commercial methylcellulose (DS from 1.3 to 1.7)
involves the use of a heterogeneous slurry of cellulose in a so-called solvent that swells
but does not dissolve the polymer. It leads to a non-uniform substituent distribution on
glucose repeat units along the chain and the presence of blocks of tri-, di-, mono, or
unsubstituted units, especially for intermediate degrees of substitution due to different
accessibilities of cellulose areas to the reagents (6). Hence, there is a poor control of the
reaction and poor predictability of product properties.
It is possible, at a laboratory scale, to prepare methylcellulose under a
homogeneous process allowing a uniform substituent distribution on cellulose units. The
conditions adopted are presented in Tables I and Π. The solvent used is D M A c / L i C l and
dimethylsulfinyl anion as reactive intermediate (7). Moreover, specific methylation may
be performed by intermediate specific tritylation on C-6 position (7,8).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

Table I : Conditions for methylcellulose preparation and consequences on

substituent distribution

Industrial process Laboratory process

H E T E R O G E N E O U S phase H O M O G E N E O U S phase Specific methylation

synthesis synthesis

Swelling in
caustic medium Cellulose dissolved in Tritylation : selective on
+ D M A c / LiCI 6% C-6 position
CH3I or CH3CI + sulfinyl anion
+ CH3I Methylation with CH3I

Detritylation
Limited - O H accessibility
(surface of crystalline Better -OH accessibility Deacetylation
zones, amorphous zones)

In Cellulose Derivatives; Heinze, T., et al.;

Substituent distribution by R E G U L A R distribution of Methyl g r o u p s o n C-2 and
ZONES substituents C-3 positions

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

Sample A4C Samples M22, M15, M12, Sample S23
M29, M18
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

Table II : Experimental conditions for homogeneous synthesis of methylcelluloses

Degree of Cellulose / DMAc-LiCI NaH/DMSO Methyl Iodide Reaction Time

substitution (DS) (g/g) (g / mL) (mL) (h)

1.2 5.8 / 800 15 / 80 20 24

1.7 6 / 740 20 / 100 30 36

In Cellulose Derivatives; Heinze, T., et al.;

2.1 6 / 700 10 / 50 20 7 days

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

 336

The methylcellulose samples investigated have the substituent distributions given

in Table ΙΠ, where the fraction of the different modified monomelic units are listed.

Solubility properties (role of the substitution pattern). The role of the

distribution of the methyl groups on solubility was previously mentioned. Miyamoto and
co-workers demonstrated clearly that the degree of substitution necessary for water
solubility is lower for homogeneously prepared materials than for similar materials
prepared heterogeneously (9). This better solubility is probably caused by the more
uniform distribution of the substituents of the homogeneously-derivatized products along
the chain. In our work homogeneous water-soluble M C was prepared with DS between
0.9 and 2.2, but commercial samples need a degree of substitution larger than 1.3 to be
water soluble. A water-soluble M C sample with a DS value of 0.9 is the minimum
substitution that will suppress the intermolecular hydrogen bonds in cellulose. The better
regularity of the substituent distribution is clearly demonstrated when we compare two
products with the same average degree of substitution (1.7), but prepared under
heterogeneous (A4C) or homogeneous (M29) conditions : in M29 there are fewer glucose
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

units but a larger content of monomethyl derivatives with the same average amount of
trimethylated units (Table ΙΠ).
Methylcellulose solution properties were studied either in the dilute (the polymer
concentration below the overlap concentration c*) or the semi-dilute regime.

Properties in the dilute regime. In dynamic light scattering experiments, we

observed the size of the molecules in aqueous solution, expressed as the hydrodynamic
radius RH, as a function of the temperature (Figure 1). At 20°C and up to 50°C the
2
correlation function exhibits a single relaxation, and the inverse relaxation time is q -
dependent, indicating a diffusive relaxation. Only one size of molecule is present in
solution, with RH around 20 nm, of the same order of magnitude as the radius of
gyration of similar cellulosic derivatives. At 60°C the 20 nm molecules are still present
but in smaller quantity, and a second population appears (RH around 200 nm)
corresponding to the formation of aggregates.
The evolution of the refractive index signal of the material eluted in Steric
August 31, 2012 | http://pubs.acs.org

Exclusion Chromatography for different M C samples is examined as a function of

temperature (70). For the lowest DS (smaller than 1.5) the chromatograms do not change
much when temperature increases; on the contrary, for A4C and higher DS samples new
peaks appear for lower elution volumes corresponding to molecules with higher
dimensions. Each sample was filtered off before the experiments and maintained one
night at the chosen temperature. The quantity of the polymer actually eluted through the
columns is indeed a good indication of the formation (or not) of aggregates retained at the
front of the column when the temperature increases. The ratio between me, the eluted
weight calculated from the area of the refractive index peak and the refractive index
increment, and mi, the injected weight of sample, is a good indication of the ratio of
polymer which is eluted (Figure 2). At low temperature this ratio is close to 1, indicating
all the polymer is eluted and confirming that the methylcellulose solution is a true
solution. Two behaviours were obtained according to the DS values. For DS smaller than
1.5, this ratio does not vary significantly while, that for DS larger than 1.5, this ratio
decreases greatly above 45°C. For temperatures higher than this value, interactions occur
leading to the formation of aggregates.
The presence of interactions has also been observed using fluorescence
spectroscopy. Pyrene was used as a probe because the vibrational structure of its
fluorescence emission spectrum is sensitive to the polarity of its environment (77). The
ratio between I i and I3, respectively, the intensities of the first (at 373 nm) and the third
(at 383 nm) peak of the fluorescence emission spectrum, is equal to 0.6 in hexane and

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

Table III : Characterization of some methylcelluloses prepared.

Sample A4C M22 M15 M12 M29 M18 S23

DSa 1.7 1.2 1.3 1.5 1.7 2.2 1.3

% Non S 10 9 9 5 4 9 16

% MonoS 29 68 63 51 45 12 47

%DiS 39 19 21 29 32 36 36

In Cellulose Derivatives; Heinze, T., et al.;

% TriS 22 4 7 15 19 43 -

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

a determination by 13c n.m.r. in DMSO-d6 (353K)
 Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024
August 31, 2012 | http://pubs.acs.org

Figure 1. Evolution of the hydrodynamic radius (RH)of A 4 C in water with

temperature (DS=1.7)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 339

(a) DS < 1.5

1 j I I I i j I I . I II I I I Ii Ί i ι j I I I I Iι ι ι ι ι ι ι • ι

s~ :
5
I M22
: — α - M23
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

0.7 - - * - M14

Q_g I . . . . ι . . . . ι . . . . ι . . . . ι . . . . ι . . . . ι ... ' ι

25 30 35 40 45 50 55 60
T(°C)

(b) DS > 1.5

August 31, 2012 | http://pubs.acs.org

25 30 35 40 45 50 55 60
T(°C)

Figure 2. Evolution of the mass loss of the polymer eluted in steric exclusion
chromatography of methylcellulose samples in water with temperature (a) DS <
1.5, (b) DS > 1.5

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 340

1.9 in water. Experiments with aqueous M C solutions were performed up to a

temperature at which turbidity is observed. For A4C in water this ratio is equal to 1.8 at
10°C, signifying that pyrene is in a polar environment, and decreases quickly to a value
of 1.5 for temperatures higher than 50°C, indicating the formation of hydrophobic
domains in which pyrene stays (Figure 3). A hysteresis is formed on cooling as a proof
of interactions established at higher temperature. The relative viscosity indicates the same
mechanism; the interactions are demonstrated from the large increase in the viscosity of
the solution (Figure 3).
From the dilute regime experiments one can conclude :
* methylcellulose in water forms, at low temperature, a true solution which can be filtered
and contains no undissolved matter,
* for samples with DS larger than 1.5, observable aggregates are formed when the
temperature is greater than 45°C; for lower degrees of substitution, no formation of
aggregates is observed up to 55-60°C within the experimental time scale,
* as interactions occur when the temperature increases, they are essentially of
hydrophobic nature, as is confirmed by the role of electrolytes such as sodium chloride
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

which decreases the gelation temperature (72).

Properties in semi-dilute regime. Oscillatory shear experiments were carried out; a

typical curve for A4C sample is presented in Figure 4. For a concentration larger than 2.5
g/L (in the range of the overlap concentration), the evolution of the storage modulus G '
shows two distinct waves. At low temperature the solution is clear, and in the range 30-
50°C a weak gel appears; then the solution becomes turbid above a temperature which
depends on the concentration. An elastic turbid gel with a large increase of viscosity is
observed above 60°C within the experimental time scale. This second wave is more
pronounced with the higher polymer concentrations. The concentration 2.5 g/L may be
considered for the A4C sample as the critical polymeric concentration needed for gelation.
It is surprising that the wave at low temperature is apparently not dependent on the
polymer concentration.
The gelation phenomenom is time dependent and reversible (10). The influence of
the substitution distribution and of the DS values was studied further using the same
August 31, 2012 | http://pubs.acs.org

procedure as for the A4C sample (Figure 5). With the low DS samples (DS < 1.5) only
one wave was observed, with a slight increase of G' for temperatures higher than 60°C.
For the sample without trisubstituted units (S23), G' remains constant with temperature;
G" decreases following the viscosity of the solvent. Hysteresis was also observed in
relation to the presence of cooperative interchain interactions, except for the S23 sample
(70). Furthermore, with low DS samples the temperature for which G ' is equal to G" is
shifted toward higher temperatures. For the sample with the highest DS value (2.2), a
behaviour similar to that of A4C was observed in the presence of two waves in the
temperature domain studied, a finding opposite to that of several authors (75). With this
degree of substitution, the proportion of trisubstituted units is relatively high and the
presence of blocks of such units may be suspected. Moreover, a sol-gel transition may be
observed even with methylcellulose samples with low DS (= 1.3) prepared from an
homogeneous process. The evolution of G ' and G " with frequency for a solution of a
sample with DS of 1.3 at 70°C is typical of a weak gel state rheogram (Figure 6). At low
frequency the G ' value is nearly constant and higher than G". The obtained gel is not
turbid at this temperature.
Calorimetric experiments were carried out on methylcellulose solutions as were
performed on other polysaccharide gels (74,75). With the A4C sample, during heating an
endothermic peak is observed but two exothermic peaks are present during cooling
(Figure 7). Only the energy of interaction associated with the phase separation (and the
large increase in viscosity) is observed on heating. It is necessary to heat up to 50°C and
65°C to observe peaks at 30°C and 40°C respectively on cooling (72). These two peaks
seem to correspond to interactions of a different nature : the peak at 30°C seems to be

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 341

ι ιιιI,• ιιIι ιιιIιι ι ι Iιιι ιIιι ιιI I I I I

Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

0 10 20 30 40 50 60 70
T(°C)

Figure 3. Influence of temperature on fluorescence (I1/I3 ratio) and relative

viscosity for A4C methylcellulose (DS=1.7, solvent water, c=2g/L)

—J ! 1 1111. 1, ,
1er
1
ι 1 1 1 J 1 1 I I
--
I Τ Ί J 1 1 1
. Ξ

• C = 15g/L · •" ~~ - z
August 31, 2012 | http://pubs.acs.org

10' Δ C=10g/L //
03 \ + C = 5g/L • * ****** :

S 10 L • C = 2.5g/L : · ·* :
ID ι
0
10
r X :

10" 1

^ * ***
" ι ι" ι 4
i 1 1 1 1 1 1. 1 1 1 111 11 1 1 1 1 1 11 1 1 1

20 30 40 50 60 70 80
Temperature (°C)
Figure 4. Influence of polymer concentration and temperature on the elastic
1
modulus of A4C solutions (G (Pa) modulus at 1 Hz, DS=1.7, solvent water),
reproduced, by permission, from Journal of Chimie Physique, Elsevier, ref. 10

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 342

a) c)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

o.i ι •••• ι . . . · • • . . . ' • ,, > •, · • , ι , , · , I ο.ι I · · • · ' • • · • ' , • , , · · • ' • • • τ ' • · , - '

20 30 40 50 60 70 80 20 30 40 50 60 70 80
ΤCO TCO

Figure 5. Rheological moduli of homogeneous methylcellulose solutions in water

1
as a function of temperature (G and G " moduli at 1 Hz) : (a) M22, DS=1.0,
August 31, 2012 | http://pubs.acs.org

c=56.5 g/L, (b) M12, DS=1.5, c=31.1 g/L, (c) M18, DS=2.2, c=14.6 g/L, (d)
S23, DS=1.3, c=36.1 g/L

100

(Pa)
10

Ο Ο Λ Λ ^
0 0
οοοο ο ο ο ο °

• G'
ο G"

Q ·| I ' t ι ι • ...I ι 1 I ιuni I » I 1 11 til 1 1 lllllJ

1 1 2 3
10" 10° 10 10 10
Freq.(Hz)

Figure 6. Influence of frequency on G' and G" rheological moduli (M15, DS=1.3,
c=19.8 g/L , T=70°C, solvent water), reproduced, by permission, from Journal of
Chimie Physique, Elsevier, ref. 10

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 343
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

0.10
A4C, DS = 1.7, c = 14.75 g/L, 0.5 K/min

-0.10

-0.15

I
August 31, 2012 | http://pubs.acs.org

I I I l l l I I I I I I » I l I I !
Temperature (°C)

Figure 7. Calorimetric thermograms of aqueous solutions of A4C methylcellulose

(DS=1.7, c=14.75 g/L, temperature rate 0.5 deg/min)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 344

related to interactions involving blocks of trisubstituted units, and the 40° peak to di- and
monosubstituted units. This assumption is proposed from the observation of
thermograms performed with M C samples containing low trisubstituted units (72).
For homogeneously prepared methylcelluloses with a degree of substitution of
2.2 (i.e., for the same rheological behaviour as A4C), we observe a very sharp peak on
heating and only one peak on cooling (Figure 8). The temperature difference of these
peaks is much smaller than those characterizing hysteresis with the A4C sample. The
presence of one peak indicates an homogeneous distribution of substituent along the
chain, but with commercial samples the presence of blocks of different natures of units is
revealed by the different peaks on cooling. The position of the M l 8 peak on cooling also
indicates a smaller energy of interaction compared with the A4C sample.
From fluorescence spectroscopy M18 and A4C have a similar behaviour (Figure
9) . A n increase of interactions was observed related to the decrease of the I1/I3 ratio at a
temperature very close to the temperature for which a modification of behaviour occurs in
calorimetry and rheology. But this technique can be carried further. Even on
methylcellulose substituted on only 2- and 3- positions, a decrease of the ratio is
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

obtained; it is the only technique that indicates the presence of loose hydrophobic
interactions without gelation, as demonstrated from the rheology experiments (Figure
10) .
Nevertheless, when solutions of S23 or methylcellulose with low DS at high
concentration (60 g/L) are placed in an oven at 90°C for a long time (up to weeks) phase
separation is observed. This phenomenon, which is shared by most polymers, means that
phase separation related to substitution is promoted by temperature increase but with a
low kinetics.

Mechanism of gelation. For many years mechanisms of thermogelation of aqueous

methylcellulose solutions were proposed. The major discussions concern the nature of
the zones responsible for gelation. Savage et al. (16) express the ability to gel from the
presence of zones coming from the original cellulosic structure; they were refuted by
Heyman (77) who has studied highly substituted methylcelluloses. Rees (18) speaks
about micellar interactions and Sarkar (79) postulates that gelation is due to hydrophobic
August 31, 2012 | http://pubs.acs.org

or micellar interactions. Khomutov et al. (20) proposed the gelation is due to

crystallization, Haque and Morris (27) imply crystalline zones within the gelation process
while Kato et al. conclude that the "crosslinking loci" of methylcellulose gels consist of
crystalline sequences of trimethylglucose units (22).
From all these observations and from comparisons between the different
methylcellulose samples (prepared from heterogeneous or homogeneous processes) with
different degrees of substitution and different structural characteristics, a mechanism of
themogelation of aqueous methylcellulose solutions is proposed :
* we start at low temperature from a true solution which can be filtered and contains no
undissolved matter,
* when temperature increases, whatever the sample, hydrophobic interactions occur due
to methyl groups. According to the structure (distribution of substituents and presence or
not of blocks of highly substituted units) these interactions lead to an increase of viscosity
and sometimes to the presence of small hydrophobic domains. In this case these are due
to the presence of zones of predominant trisubstituted units leading to a clear gel.
* increasing the temperature more leads to phase separation and turbidity : polymer-
concentrated domains connect and a turbid gel is obtained. The temperature at which
phase separation or "gelation" occurs depends upon the molecular weight of the
macromolecular chain and the polymer concentration of the solution besides the structure.
Whatever the sample, the same steps are observed but with modified
characteristics depending on the substitution : the critical temperatures for physical
changes are displaced to higher values when the substitution is more regular even if the
average substituent content is the same (due to the presence of less highly substituted
zones). But in any case the presence of trisubstituted units is compulsory for observable
gelation (high viscosity, turbidity, and elastic character).

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 August 31, 2012 | http://pubs.acs.org
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

Heat Flow (mW)

Exo
f
0.15

0.10

0.05

Lo.OO

U0.0S

Up.io

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
J
20 40
1
60 80 100 ΤθΠφθΓβΙυΓΘ (°C)
1 1 I .
Figure 8. Calorimetric thermograms of M18 sample (DS=2.2, c=14.23 g/L,
temperature rate 0.25 deg/min)
 346

1.85

Temperature (°C)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

Temperature (°C)
August 31, 2012 | http://pubs.acs.org

Figure 9. Fluorescence I1/I3 ratio of aqueous solutions of methylcellulose : (a)

A4C, DS=1.7, c=7.4 g/L, (b) M18, DS=2.2, c=4.9 g/L

1.65
11/13 ;
S23-6
1.6
; DS = 1.3
1.55
'_
1.5

1.45

1.4 -*— heating

: —cooling
1.35 1 . . . . I . . . . 1 .... 1 1 . 1 1 1 1 1 . 1 I 1 1 1
10 20 30 40 50 60 70 80
Temperature (°C)

Figure 10. Fluorescence I1/I3 ratio of aqueous solutions of S23 methylcellulose

(DS=1.3, c=7.4 g/L)

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 347

Conclusion

T h e distribution of the substituents clearly controls the m a i n properties o f

methylcellulose. This distribution depends on the conditions for chemical substitution,
i.e. whether conditions are homogeneous or heterogeneous. The solubility needs less
methyl substituent regularly disposed along the chain avoiding the packing of cellulosic
chains by cooperative hydrogen bonds. The stability of the junctions, a consequence of
hydrophobic interactions, i n the clear gel phase is directly related to the existence of
highly substituted blocks and mainly exist in the commercial samples obtained by
heterogeneous chemical modification. The rheology depends directly not only on the
average degree of substitution but on the substituent ctistribution.

Experimental
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

The commercial methylcellulose sample (Methocel A 4 C ) was kindly provided by

D o w Chemical.
The synthesis of homogeously prepared methylcellulose samples is carried out
according to the procedure previously described (7).
1 3 c n.m.r. measurements were realized on a Bruker A C 3 0 0 spectrometer at
353K using D M S O - d 6 as solvent
Total hydrolysis of methylcellulose was carried out with 2 N trifluoroacetic acid
within 4 hours at 120°C. After evaporation of the acid in vacuo, water was distilled from
the samples ten times to remove traces of the acid before h.p.l. chromatography.
H.p.l.c. analyses were performed with a W a t e r s - M i l l i p o r e chromatograph
equipped with a differential refractometer detector (Waters 410). A reversed-phase
column Nucleosil C I 8 5μ (250*4.6 mm) with water as eluent at ambient temperature was
used. This technique gives the content in unsubstituted and substituted monomers.
T h e aqueous solutions were prepared by d i s s o l u t i o n o f freeze dried
methylcellulose in water at 5°C over 24 hours to assure a complete solubilization.
August 31, 2012 | http://pubs.acs.org

The quasi-elastic scattering experiments were performed using the A L V

apparatus. The scattered light of a vertically polarized λο = 488 n m argon laser (Spectra
Physics model 2020, 3 W , operating around 0.3W) was measured at different scattering
angles. The autocorrelation function of the scattered intensity was obtained using the
A L V - 5 0 0 0 autocorrelator and was analyzed by the constrained regularization method
( C O N T I N ) . The results presented are obtained for a scattering angle of 130°.
Steric Exclusion Chromatography ( S E C ) experiments were carried out using
water as eluent and the multidetection equipment described previously (23). The
refractive index increment dn/dc for the M C is taken to equal 0.136. The experiments
were performed on grafted silica gel columns (Shodex OHpak B-804, B-805).
Fluorescence experiments were performed using a L S 5 0 B luminescence
spectrometer from Perkin Elmer.
Oscillatory shear measurements were carried out using a C S 5 0 rheometer from
C a n i - M e d in the linear regime at a frequency of 1 H z (25% deformation) during heating
and cooling scans at 0.5 deg/min.
Calorimetric experiments were performed using a micro D S C ΠΙ calorimeter from
Setaram. The temperature rate was 0.5 or 0.25 deg/min.

Literature Cited

1 Dawsey, T.R. In Cellulosic polymers, blends and composites; Gilbert, R.D. Ed;
Hanser : Munich, 1994, 157
2 Dawsey, T.R. Rev. Macromol. Chem. Phys. 1990, C30 (34), 405

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 348

3 Mc Cormick, C.L.; Callais, P. Polymer 1987, 28, 2317

4 Elmgren, H; Norrby, S. Die Makromol. Chem. 1969, 123, 265
5 Hudry Clergeon, G.; Rinaudo, M J. Chim. Phys. 1967, 64, 1746
6 Nevell, T.P.; Zeronian, S.H. In Cellulose chemistry and its applications; Nevell,
T.P., Zeronian, S.H., Edts, John Wiley : New York, 1985, 15
7 Hirrien, M.; Desbrieres, J.; Rinaudo, M Carbohydr. Polym. 1996, 31, 243
8 Kondo, T.; Gray, D.G. Carbohydr. Res. 1991, 220, 173
9 Takahashi, S.I.; Fujimoto, T.; Miyamoto, T.; Inagaki, H. J. Polym. Sci. Part A :
Polym. Chem 1987, 25, 987
10 Vigouret, M.; Rinaudo, M.; Desbrieres, J. J. Chim. Phys. 1996, 93, 858
11 Kalyanasundaram, K.; Thomas, J.K. J. Amer. Chem. Soc. 1977, 99, 2039
12 Hirrien, M.Thesis, Grenoble, 1996
13 Savage, A.B. Ind. Eng. Chem. 1957, 49, 99
14 Watase, M.; Nishinari, W.; Clark, A.H.; Ross Murphy, S.B. Macromolecules
1989, 22, 1196
15 Watase, M.; Nishinari, W. Makromol. Chem. 1987, 188, 1177
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ch024

16 Savage, A.B.; Young, A.E.; Maasberg in Cellulose and cellulose derivatives. Part II;
Ott, E., Spurlin, H.M., Grafflin, M.W., Ed; Interscience : New York, 1963; 904
17 Heymann, E. Trans. Faraday Soc. 1935, 31, 846
18 Rees, D.A. Chem. Ind. London 1972, 630
19 Sarkar, N. J. Appl. Polym. Sci. 1979, 24, 1073
20 Khomutov, L.I.; Ryskina, I.I.; Panina, N.I.; Dubina, L.G.; Timofeeva, G.N.
Polym. Sci. 1993, 35, 320
21 Haque, Α.; Morris, E.R. Carbohydr. Polym. 1993, 22, 161
22 Kato, T.; Yokoyama, M.; Takahashi, A. Colloid and Polym. Sci. 1978, 256, 15
23 Tinland, B.; Mazet, J.; Rinaudo, M. Makromol. Chem., Rapid Comm. 1988, 9, 69
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Author Index
Liebert, T., 61
Aoki, N., 83
Lowman, Douglas W., 131
Becker, Ulrike,315
Manley, R. St. J., 253
Besemer, A. C , 73
Marx-Figini, Marianne, 184
Buchholz, V., 306
Mischnick, Petra, 118
Burchard, Walther218
Myers, M . D., 283
Chan, W. L., 107
Ott, Th., 94
Davé, Vipul, 265
Pecorini, Thomas J., 38
Davis, M . R, 283
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix001

Poon, C. C. L., 107

de Nooy, A. E. J., 73
Puis, J., 201
Derleth, Christina, 239
Rials, Timothy G., 265
Desbrieres, J., 332
Rinaudo, M., 332
Donges, Reinhard, 218
Saake, B., 201
Edgar, Kevin J., 38
Sakamoto, M., 83
Fakirov, C , 306
Sawatari, Chie, 296
Furuhata, K., 83
Schulz, Liane, 218
Glasser, Wolfgang G., 2, 38, 265, 315 Schulze, M., 306
Granzow, C , 94 Seufert, M., 306
Gruber, E., 94 Shibata, Tohru, 194
Heinze, Thomas, 2, 61 Shimamoto, Shu, 194
Hirrien, M., 332 Szeto, Y. S., 107
Hishikawa, Y., 173 Tebbe, H., 306
Horner, St., 201 Tezuka, Y., 163
Iwamiya, J. H, 283 Todd, Jason, 315
August 31, 2012 | http://pubs.acs.org

Kataoka, Y., 173 van Bekkum, H., 73

Kelley, Stephen S., 265, 283 Wang, X. M., 283
Klemm, D. O., 19 Wegner, G., 306
Kohmoto, Takayuki, 194 Yip, C. W., 107
Kondo, Tetsuo, 173, 296 Zugenmaier, Peter, 239
Lee, W. K., 107

350

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 Subject Index

A time course of reaction between Cell-Br and

thiourea, 91,92/
Acetylation of cellulose. See Cellulose and See also Chlorodeoxycellulose (Cell-Cl)
cellulose triacetate depolymerization in Bunte salts of cellulose and derivatives
acetylation system crosslinking in the presence of oxidizing
Amorphous cellulose reagents, 29,31/
depolymerization rate comparison with subsequent reactions of low-molecular weight
crystalline form, 195, 199 Bunte salts, 29, 30/"
See also Fourier transform infrared (FTIR) synthesis via cellulose p-toluenesulfonates, 25,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

spectroscopy of crystalline and non-crystalline 28/

cellulose
Anhydroglucose repeat unit (AGU), general
approach for substituent distribution
13
determination, 118-119, 120/ C CP/MAS NMR spectroscopy. See Nuclear
Aspergillus endoglucanase, enzymatic hydrolysis magnetic resonance (NMR) spectroscopy
of cellulose derivatives, 206-207 Carbamoylethylated ramie
activation energy calculation, 112
application of Michael addition reaction, 107-
Β 108,115
carbamoylethylation procedure and conditions,
Biomedical implants 108, 109r
cellulose candidates, 315-316 experimental materials, 108
critical surface free energy, 316 influence of reaction conditions, 108, 110-112,
Boehmeria nivea and Boehmeria tenacisseama, 113/
ramie fibers, 107 limiting oxygen index (LOI) procedure, 108
August 31, 2012 | http://pubs.acs.org

Bromodeoxycellulose (Cell-Br) LOI for flammability evaluation, 114

analytical methods, 84 moisture regain enhancement with amide groups,
degree of substitution (DS) of 112-113
halodeoxycellulose in homogeneous moisture regain measurement, 108
DMAc/LiCl system, 83 tensile properties of carbamoylethylated ramie
derivative for further reactions, 83 fabrics, 114-115
evaluation of reactivity of halodeoxycellulose tensile property measurement, 108
using model saccharides, 84-90 thermal behavior, 113-114
experimental syntheses, 84 thermogravimetric analysis procedure, 108
halogen exchange reactions of model weight gain measurement, 108
saccharides, 84 6-Carboxycellulose preparation. See Selective
IR spectra of derivatives from Cell-Br, 92/, 93 oxidation method for cellulose
kinetic parameters of model reactions, 86-90 Carboxymethylcellulose (CMC)
13
mechanism of cellulose halogenation, 84-85 C NMR carbonyl region spectra, 169/
model experiments using halogenated methyl conventional slurry synthesis method in
glycosides, 85-86 isopropanol/aqueous NaOH, 70
product solubility comparisons, 90-91 degree of substitution (DS) dependence on
reaction of Cell-Br with thiols in homogeneous NaOH concentration, 62-63
system, 90-91 experimental procedures, 170, 172
solubilities of cellulose derivatives from Cell-Br functional analysis by HPLC method, 62,64/
and thiols, 91; induced phase separation synthetic method, 63,
syntheses of deoxymercaptocellulose, 91-93 65-70

351

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 352

large-scale production in slurry process, 62 carbonyl region spectrum of CA and its

partially substituted cellulose derivative, 219— propanoated derivative, 165, 166/
220 13
substituent distribution by C NMR, 164-165
reaction condition dependence, 5, 7 Cellulose and cellulose nitrate degradation
13
substituent distribution by C NMR, 168, 170 behavior of DP (degree of polymerization) with
n

substituent distribution in hydrolyzed CMC from high-speed stirring time, 186, 187/ 188
conventional slurry process, 63, 64/ behavior of DP with sonication time, 186, 187/
n

substituent distribution in hydrolyzed CMC from cellulose degradation by high-speed stirring, 186
induced phase separation, 65, 66/ cellulose nitrate polydispersity values as related
See also Induced phase separation synthetic to sonication time, 188f
method; Supramolecular structures of mechanical force breaking weak links of
cellulose derivatives cellulose, 190, 191/
Canonization of cellulose fibers molecular weight and molecular weight
cationic grafting onto cellulose, 101, 102/ distribution measurement, 185-186
cationic pulp fiber types prepared by different narrowing polydispersity with sonication, 188-
methods, 99/ 190
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

chemical routes to cationic pulp fibers, 97-102 polydispersity comparison of ultrasonication and
considerations for paper making, 95-96 high-speed stirring, 190, 192
coupling oligo-ionomers, 94, 98-100 ultrasonic degradation of nitrated cotton
drainage effects, 103, 105/ cellulose, 186
equations for side reactions possible with Cellulose and cellulose triacetate
carboxylic groups, 97 depolymerization in acetylation system
grafted cationic pulp, 94, 100-102 analysis by gel permeation chromatography-low
grafting on cellulose equation, 100 angle laser light scattering (GPC-LALLS),
heterogeneous grafting, 100-102 195, 196/
interactions among electrically charged fibers, degree of polymerization (DP) evaluation, 194-
pigments, and polymers, 95, 96/ 195
kinetics of amine/epichlorohydrin cationization, depolymerization in acetic acid/sulfuric acid
99/ system, 195
mechanical properties, 103, 104/, 105/ DP versus time in acetic acid/sulfuric acid, 195,
W

one-step cationization equation for 196/

polysaccharides, 97
August 31, 2012 | http://pubs.acs.org

experimental materials, 199-200

retention effect of soluble ammonium alkyl general depolymerization procedure, 200
cellulose, 95, 96/ number- and weight-average DP equations, 195,
retention effects, 103, 104/ 199
retention of filler (CaC0 ) as function of beating
3 simulation and experimental DP result
W

time at various degrees of cationization, 100, comparisons, 198f, 199

102/ simulation of final product DP in actual
simultaneous reaction of epichlorohydrin and acetylation system, 195, 199
tertiary amine, 94, 97-98 temperature dependence of depolymerization,
substances for grafting, 10If 195, 197/
synthetic route for preparing oligo-ionomers, 98 temperature dependence of sulfuric acid
technical features of grafted cationic pulp fibers, sorption, 195, 198/
103 viscometry for DP measurement of degraded
types of cationic pulp fibers, 97 cellulose after acetylation, 200
See also Paper industry Cellulose and lignin as model for hierarchical
Cellulose wood structure
capabilities for regiocontrolled introduction of cellulose derivative/lignin blends to recreate
functional groups, 61-62 native wood structure, 269, 271, 276
differential scanning calorimetry (DSC)
controlled distribution of functional groups thermograms of hydroxypropyl
within AGU, 62 cellulose/lignin blends, 210f
homopolymer of D-anhydroglucopyranose units dynamic elastic modulus of cellulose and
(AGU), 61 cellulose ester solutions with lignin, 277/
Cellulose acetate (CA) dynamic mechanical thermal analysis (DMTA)
acetyl and butanoyl carbonyl carbon signals, 165, of hydroxypropyl cellulose/lignin blends, 269,
167/ 272/ 273/

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 353

enhancement of liquid crystalline mesophase (CP/MAS) NMR spectroscopy, 285

order in non-crystalline cellulose derivatives cellulose acetate(CA)/poly(vinyl phenol) and
by lignin addition, 278, 279/, 280/ CA/poly(styrene) blends, 286-289
experimental materials and methods, 267 chemical shift filter, proton analog of SELDOM
melting point of hydroxypropyl cellulose (selectivity by destruction of magnetization)
component in relation to lignin content, 271, sequence, 287-288
274/ 13
expansion of C CP/MAS spectrum in carbonyl
non-crystalline components of wood, 267-269 and phenolic hydroxyl carbon region, 290-292
polymer-polymer interaction parameter (B) experimental procedures, 285
relation to phenolic hydroxyl content of lignin, exploring extent of polymer mixing by NMR,
271,275/276 285-289
regenerated fibers, 276-278 exploring interactions between functional groups
Cellulose chemical modification. See Chemical by NMR, 289-294
modification of cellulose hydrogen bonding interactions in cellulose
Cellulose degradation. See Cellulose and cellulose acetate butyrate (CAB) with Novolac resin,
nitrate degradation; Enzymatic hydrolysis of 293-294
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

cellulose derivatives nature of blend with bulkier side-chains of CAB,

13
Cellulose derivative structural study by C NMR 292-293
advantages of analytical technique, 163-164 phenolic polymers miscible with several
experimental procedures, 170,172 polymers, 283-284
interaction between substituents in cellulose possible hydrogen bonding interactions for
triacetate, 170, 171/ cellulose acetate/poly(vinyl phenol) blends,
substituent distribution of 285
carboxy methylcellulose, 168-170 proton combined rotation and multiple pulse
!
substituent distribution of cellulose acetate, 164- spectroscopy ( H CRAMPS), 285
167 pulse sequence experiment for extent of mixing,
Cellulose derivatives. See Highly engineered 288-289,290/
cellulose derivatives by homogeneous phase spin diffusion experiments, 286-289
L
modification; Supramolecular structures of typical U CRAMPS spectra of polymers and
cellulose derivatives blends, 288
Cellulose ester characterization Cellulose ethers
August 31, 2012 | http://pubs.acs.org

13
C NMR chemical shifts for carbonyl characterization by solution-state NMR, 151,
resonances for series of cellulose acetates, 152/
142f non-derivatizing DMAc/LiCl solvent with
future directions, 159-160 triethylamine and pyridine catalysts, 5, 7
sequencing methods desirable for esterified and regioselective synthesis and subsequent reactions
etherified AGU, 159 of cellulose ethers, 23-25
solid-state NMR, 153, 155-159 Cellulose/poly(vinyl alcohol) blends
solution-state NMR cellulose model compounds 2,3-di-O-
chemical shift assignments, 133-143 methylcellulose and 6-O-methylcellulose,
conformation and molecular dynamics, 143- 296-297, 298/
147 DSC measurement procedure, 297
mixed esters of cellulose, 153, 154/ experimental materials, 297
structure-property relationships, 147, 151-152 interchain interaction findings of model
See also Nuclear magnetic resonance (NMR) compounds, 304
spectroscopy melting point depression versus square of
Cellulose esters volume fraction cellulosics, 299-300, 301/
acylation of cellulose, 131-132 NMR measurements of proton spin relaxations in
classical preparation and processing, 38-39 spin-lattice (ΤΊΗ) and rotating frame (T H),
lp

NMR analysis technique, 132-133 300, 302, 303/

See also Long-chain cellulose esters (LCCE) preparation and characterization data of film
Cellulose esters, carbamates, and lactones, specimens, 297, 298r
synthesis in DMAc/LiCl non-derivatizing results of DSC characterization, 299-300
solvent system, 5, 6 solid-state NMR measurement procedure, 297,
Cellulose esters/phenolic polymer blends 299
13
C cross-polarization/magic-angle spinning thermodynamic interaction parameters, 300, 30If

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 354

types of hydrogen bonds in blends, 296-297 fluorescence spectroscopy with pyrene probe
Cellulose structural analysis. See Fourier showing interactions at high temperature, 336,
transform infrared (FTIR) spectroscopy of 340, 341/
crystalline and non-crystalline cellulose; heterogeneous chemical modifications typical,
Supramolecular structures of cellulose 332-333
derivatives; Supramolecular architectures of homogeneous process at laboratory scale, 333,
cellulose derivatives 334f, 335f, 336
Cellulose/synthetic polymer blends hydrodynamic radius as function of temperature,
cellulose/poly(acrylonitrile) blend, 254, 256/, 336, 338/
257 influence of polymer concentration and
cellulose/poly(vinyl alcohol) (PVA) blend temperature on elastic modulus, 340, 341/
(crystalline/amorphous example), 257, 259, mechanism of gelation, 344
261 properties in dilute regime, 336, 340
cellulose/poly(4-vinyl pyridine) blend, 257, 258/ properties in semi-dilute regime, 340, 344
cellulose/poly (vinyl pyrrolidone) blend, 261-263 reaction conditions, 333, 336
differential scanning calorimetry (DSC) for T , Theological moduli of homogeneous
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

254, 256/, 257 methylcellulose solutions as function of

DSC thermograms for series of cellulose/PVA temperature, 340, 342/
blends, 260/ solubility properties and role of substitution
dynamic mechanical analysis for T , 254, 256/
g
pattern, 336
257, 258/ steric exclusion chromatography as function of
examples of hydrogen bonding interactions, 255/ temperature, 336, 339/
Flory-Huggins equation for melting point See also Highly engineered cellulose derivatives
depression, 259, 261 by homogeneous phase modification
method of sample preparation, 254 Chiral nematic or cholesteric mesophase. See
miscibility of polymer blends by Tg
Liquid crystalline cellulose derivatives
measurement, 254, 256/ 257 Chlorodeoxycellulose (Cell-Cl)
NMR measurements for domain size estimates, synthesis of cellulose derivatives, 83
261,263r See also Bromodeoxycellulose (Cell-Br)
polymers with functional groups for interaction COLOC (long-range heteronuclear correlated
with cellulose hydroxyl groups, 253-254 spectrum), chemical shift assignments in
solution-state NMR, 133
August 31, 2012 | http://pubs.acs.org

See also Cellulose/poly(vinyl alcohol) blends

Cellulose triacetate COSY (2D homonuclear correlated spectrum),
experimental procedures, 170, 172 chemical shift assignments in solution-state
interaction between substituents, 170 NMR, 133-135, 136/ 143
NOESY (2D) spectra, 171/ Crystalline cellulose. See Fourier transform
Cellulose tricarbanilate infrared (FΉR) spectroscopy of crystalline and
fully derivatized cellulose derivative, 218-219 non-crystalline cellulose
See also Enzymatic hydrolysis of cellulose
derivatives; Liquid crystalline cellulose
derivatives; Supramolecular structures of D
cellulose derivatives
Cellulose trinitrate, fully derivatized cellulose Degradation. See Cellulose and cellulose nitrate
derivative, 218-219 degradation; Enzymatic hydrolysis of cellulose
Cellulosetrisphenylcarbamate and cellulosetris(3- derivatives
chlorophenylcarbamate) Depolymerization of cellulose
structures for liquid crystalline study, 241-242 ultrasonication and high speed mechanical
See also Liquid crystalline cellulose derivatives agitation, 184-185
Chemical modification of cellulose See also Cellulose and cellulose nitrate
calorimetric thermograms of methylcellulose degradation; Cellulose and cellulose triacetate
solutions, 340, 343/, 344, 345/ depolymerization in acetylation system
characteristics of methylcelluloses prepared at 2,3-Dicarboxycellulose preparation. See Selective
laboratory scale, 337/ oxidation method for cellulose
experimental procedures, 347 MN-Dimethylacetamide with lithium chloride
fluorescence spectroscopy of loose hydrophobic (DMAc/LiCl), non-derivatizing solvent system,
interactions without gelation, 344, 346/ 5,7

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 355

2,3-Di-0-methylcellulose, model for Esterification. See Regiocontrol in cellulose

cellulose/poly(vinyl alcohol) blends, 297, 298/ chemistry
M^-Dimethylformamide (DMF)/N 0 solvent
2 4
Etherification. See Regiocontrol in cellulose
system, cellulose nitrate intermediate, 7-9 chemistry
Dimethyl sulfoxide (DMSO) with S 0 and 2

diethylamine
non-derivatizing solvent system, 5, 7
superior reaction rates and yields in comparison
19
with DMF/N 0 and DMAc/LiCl, 7
2 4
F NMR. See Nuclear magnetic resonance
Dynamic mechanical thermal analysis (DMTA) (NMR) spectroscopy
spectra of melt-processed and solvent-cast Flory-Huggins equation, melting point depression
hydroxypropyl cellulose/lignin blends, 272/, in polymer blends, 259, 261
273/ Fluorescence spectroscopy
spectra of spruce wood at various moisture experimental procedure, 347
contents, 268/ influence of temperature for methylcellulose,
See also Cellulose and lignin as model for 336, 340, 341/
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

hierarchical wood structure Fluorine-containing cellulose materials. See

Surface segregation in cellulose ester blends
Fourier transform infrared spectroscopy (hllR) of
Ε crystalline and non-crystalline cellulose
advantages of novel methods, 173
Enzymatic hydrolysis of cellulose derivatives comparison of absorption bands for Ια and Ιβ
accessibility of cellulose derivatives to enzymes, forms of cellulose, 176,177/
203-204 experimental materials and measurement of
acetyl esterase involvement in cellulose acetate FΉR with deuteration reaction cell, 174,175/
fragmentation, 211-213 experimental materials and measurement of
analytical methods for measuring enzymatic FΉR with microscopic accessory, 174
degradation, 202 FÏIR spectral changes for deuterated cellulose
carbanilation procedure, 207 and second derivativesfromtracheid walls,
cellulose degrading enzyme system for complete 176, 178/
hydrolysis, 202-203 ITIR with deuteration reaction cell for
August 31, 2012 | http://pubs.acs.org

cellulose ester degradation, 204-205 amorphous cellulose study, 179-181,182/

cellulose ethers degradation, 204 FΉR with microscopic accessory for cellulose
endoglucanase degradation of carboxymethyl crystallization study, 174, 176-179
cellulose, 208, 209/ IR crystallinity index change during tracheid cell
endoglucanase degradation of cellulose acetate, wall formation, 176,178/, 179
210-211 kinetic study of deuteration of amorphous model
endoglucanase degradation of methyl cellulose, samples and cellulose, 181,182/
210 monitoring deuteration of amorphous cellulosic
enzyme and mixfromAspergillus strain, 206 samples, 179,180f
enzyme treatment procedure, 207 schematic of novel methods, 175/
exoglucanases, endoglucanases, and β- sectioning method for obtaining IR spectra at
glucosidases for complete hydrolysis of each stage of tracheid wall formation, 176,
natural cellulose, 202 177/
experimental substrates for testing, 206 Fringed micelle structure. See Supramolecular
influence of derivatization procedure, 205 structures of cellulose derivatives
influence of substituent position, 206 ITIR. See Fourier transform infrared spectroscopy
influence of substituent type, 206 (hTIK) of crystalline and non-crystalline
influence of substituents and degree of cellulose
substitution, 205
insoluble cellulose derivatives degradation, 204
limited substitution to conserve biodegradability, G
201
proposed degradation pathway, 202 Gel permeation chromatography (GPC)
size exclusion chromatography of cellulose degree of polymerization (DP) determination
carbanilates and water-soluble derivatives, 207 method, 320-321

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 356

DP results for F-containing oligomers compared cellulose derivatives by homogeneous phase

with NMR results, 322, 324/ modification
Gel permeation chromatography low angle laser HPAEC. See High-pH anion exchange
light scattering (GPC-LALLS), cellulose chromatography (HPAEC)
triacetate analysis after acetylation, 195, 196/ HPLC. See High performance liquid
chromatography (HPLC)
Humicola insolens, enzymes for hydrolysis of
H cellulose derivatives, 203

*H CRAMPS NMR spectroscopy. See Nuclear

magnetic resonance (NMR) spectroscopy
Halodeoxycellulose. See Bromodeoxycellulose
(Cell-Br); Chlorodeoxycellulose (Cell-Cl) Impeller method, long-chain cellulose ester
I3 !
HETCOR ( C- H direct-detected heteronuclear preparation, 39-40
correlated spectrum), chemical shifts in Implant materials, cellulose derivatives as prime
solution-state NMR, 133-135, 137/ candidates, 315-316
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

Hierarchical wood structure. See Cellulose and Induced phase separation synthetic method
lignin as model for hierarchical wood structure application to other etherification reactions, 67,
Highly engineered cellulose derivatives by 69
homogeneous phase modification carboxylation experimental procedures, 70
dissolution by derivatization with leaving groups, conditions and results of carboxymethylcellulose
12-14 (CMC) via induced phase separation, 7If
dissolution by derivatization with protective distribution of carboxymethyl groups by
groups, 12, 13 deuterated sulfuric acid degradation and Ή
dissolution with derivatizing solvents, 7-12 NMR spectroscopy, 67
dissolution with non-derivatizing solvents, 3-7 experimental materials, 69
examples of cellulose dissolution with partial experimental measurements, 70
derivatization, 8r influence of NaOH particle size, 67, 68/
examples of subsequent reactions on isolated mole fractions of repeating units in CMC by
cellulose intermediates, 10/ induced phase separation, 65, 66/
reactive intermediates by derivatizing solvents, system behavior if alkali becomes mobile by
7-12 addition of water, 65, 66/ 67
August 31, 2012 | http://pubs.acs.org

typical non-aqueous cellulose solvents, At See also Carboxymethylcellulose (CMC)

High performance liquid chromatography (HPLC)
determination of molar fractions of un-, mono-,
di-, and trisubstituted glucose residues, 124, L
126
statistical pattern of functionalization, 62, 64/ Langmuir-Blodgett (LB) films. See
High-pH anion exchange chromatography Supramolecular architectures of cellulose
(HPAEC), analysis of anionic derivatives of derivatives
carboxymethyl cellulose (CMC) and sulfoethyl Leaving groups in cellulose derivatization
cellulose after hydrolysis, 124, 126 cellulose sulfonates for nucleophilic
HMBC (multiple-bond inverse-detected substitutions, 12-13
heteronuclear correlated spectrum) halodeoxycelluloses via nucleophilic substitution
chemical shift assignments in solution-state of tosylates by halides, 13-14
NMR, 133 Lignin and lignocelluloses. See Cellulose and
mixed esters of cellulose, 153, 154/ lignin as model for hierchical wood structure
HMQC (^C-'H inverse-detected heteronuclear Liquid crystalline cellulose derivatives
correlated spectrum), chemical shift exhibiting chirality at three levels, 240
assignments in solution-state NMR, 133 experimental cellulose derivatives and analytical
HOHAHA (2D homonuclear Hartman-Hahn procedures, 241, 244
correlated spectroscopy), chemical shift hélicoïdal structure and pitch changes with
assignments in solution-state NMR, 135, 138, substituents, 249
139/ optical rotatory dispersion and transmission
Homogeneous phase modification. See Chemical curves for right-handed super-molecular
modification of cellulose; Highly engineered structures, 244, 246/

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 357

phase behavior and properties at phase transition degree of substitution (DS), 53, 55/
crucial test for theoretical models, 250 solubility parameters of selected cellulose esters,
phase diagram for cellulosetris(3- 56r
chlorocarbanilate), 25 1/ T , melt viscosity, and modulus correlation with
g

pitch and handedness of various cellulose solubility parameter, 53, 54/

urethane/solvent systems, 247r T values as function of degree of substitution
g

regio-selectively substituted cellulose, 244, 247- (DS) and chain length, 48, 52/
250 thermal transitions of waxy phase as function of
schematic of cellulosetricarbanilate (CTC) and number of carbons in acyl substituent, 47-48,
substituted CTC, 242 49/
schematic of cholesteric liquid crystalline tosyl chloride in impeller method, 42-43
structures, 243/ Lyotropic liquid crystalline cellulose derivatives.
seeking relationship between chirality and See Liquid crystalline cellulose derivatives
twisting power, 240-241
solvent interactions and side-group contributions,
239-240 M
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

statistical cellulose copolymer, 244

temperature dependence of twisting power of Mass spectroscopy, analysis for substitution
chiral nematic phases, 249-250, 251/ pattern in polymer chain, 126
temperature dependence of twisting power of Mechanical degradation. See Cellulose and
regio-selective liquid crystalline cellulose cellulose nitrate degradation
derivatives, 247-249 Methylcellulose derivatives, models for
twisting power of liquid crystalline state as cellulose/poly(vinyl alcohol) blends, 297, 298/
function of composition, 244, 245/ Methylhydroxyethylcellulose (MHEC) and
Long-chain cellulose esters (LCCE) methylhydroxypropylcellulose (MHPC)
acid chlorides with pyridine as acid acceptor to partially substituted cellulose derivatives, 219—
overcome solubility problems, 43 220
cellulase enzyme biodegradability decline with See also Supramolecular structures of cellulose
substituent content and size, 56, 57/ derivatives
cellulose acetate nonanoates by titanium(IV) N-Methyl morpholine-N-oxide-based solvent
isopropoxide-catalyzed esterification in process, spinning of cellulose, 276
August 31, 2012 | http://pubs.acs.org

DM Ac, 44,46f Micelle (fringed) structure. See Supramolecular

chemical structure of typical LCCE with structures of cellulose derivatives
covalently bonded internal plasticizer, 47-48 Michael addition reaction
direct method with acid chloride at elevated carbamoylethylation of ramie fibers, 107-108
temperatures, 39-40 cotton modification, 107
early synthetic attempts, 39 See also Carbamoylethylated ramie
experimental evaluation of mechanical Molecular weight and molecular weight
properties, 58-59 distribution
heterogeneous acylation with titanium catalyst in intrinsic viscosity measurement for cellulose and
amide solvent, 43-47 derivatives, 185-186
LCCEs by solution esterification in DMAc/LiCl, size exclusion chromatography (SEC) of
41f degraded cellulose and derivatives, 185-186
LCCEs by titanium(IV) isopropoxide-catalyzed See also Cellulose and cellulose nitrate
esterification in DMAc, 44, 45r degradation
long side-chains effective flow aids, 48, 52/ 53
mechanical and Theological properties, 47-53
mechanism of esterification in Ν
dicyclohexylcarbodiimide/4-
pyrrolidinopyridine (DCC/PP), 40, 42 NOES Y. See Nuclear Overhauser enhancement
melt viscosity and flexural modulus exchange spectroscopy (NOESY)
extrapolations to zero plasticization, 60/ Novolac resin. See Cellulose esters/phenolic
normalized viscosities of plasticized and non- polymer blends
plasticized cellulose esters, 48, 52/ 53 Nuclear magnetic resonance (NMR) spectroscopy
13
preparation by solution acylation, 40-43 C cross-polarization/magic-angle spinning
relationship between thermal transitions and (CP/MAS) spectroscopy, 285

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 358

13
C NMR technique using carbonyls of cellulose carboxymethyl cellulose degradation, 206
esters as structural probe, 163-164 Phenolic polymers. See Cellulose esters/phenolic
19
F NMR method for F-containing cellulose polymer blends
derivatives, 320-321 Poly(4-vinyl pyridine). See Cellulose/synthetic
19
F NMR spectra for CF -terminated oligomer,
3 polymer blends
322, 325, 326/ Poly(acrylonitrile). See Cellulose/synthetic
x l
H H COSY spectra of high molecular cellulose polymer blends
and starch derivatives, 29, 33/ Polydispersity. See Cellulose and cellulose nitrate
proton combined rotation and multiple pulse degradation
spectroscopy (*H CRAMPS), 285 Polymer blends. See Cellulose esters/phenolic
solid-state NMR, 153, 155-159 polymer blends; Cellulose/poly (vinyl alcohol)
solution-state NMR blends; Cellulose/synthetic polymer blends;
cellulose ethers, 151-152 Surface segregation in cellulose ester blends
chemical shifts, 133-143 Polysaccharide derivative characterization. See
conformation and molecular dynamics, 143-147 Substituent distribution in polysaccharides
mixed esters of cellulose, 153, 154/ Poly(styrene) blends. See Cellulose
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

structure-property relationships, 147, 151 esters/phenolic polymer blends

Nuclear Overhauser enhancement exchange Poly(vinyl alcohol). See Cellulose/synthetic
spectroscopy (NOESY) polymer blends
average internuclear distance compared to X-ray Polyvinyl phenol). See Cellulose esters/phenolic
crystal structure distances for cellulose esters, polymer blends
144, 146r Poly(vinyl pyrrolidone). See Cellulose/synthetic
solution conformation of cellulose esters, 143- polymer blends
144, 145/ Protective groups in cellulose derivatization
6-0-thexyldimethylchlorosilane (TDMS) with
great regio-selectivity, 12
Ο triphenylmethyl (trityl) group, 12

Optical rotatory dispersion (ORD), handedness

measurement of hélicoïdal structure, 244, 246/, R
247
Oxidation of cellulose. See Selective oxidation
August 31, 2012 | http://pubs.acs.org

Ramie fibers. See Carbamoylethylated ramie

method for cellulose Reactive cellulose intermediates
derivatizing solvents, 7-12
examples of subsequent reactions on isolated
Ρ intermediates, 10/
trifluoroacetylation at C-6, 9,11/
Paper industry trimethylsilyl (TMS) celluloses, 9
breaking strength as function of filler Regiocontrol in cellulose chemistry
concentration, 105/ analysis of molecular structure, 29
comparison of CaC0 retention by cationic graft
3 Bunte salts of cellulose and derivatives, 25, 29
pulps and soluble cationic polymer, 104/ crosslinking of Bunte salts of cellulose in
considerations for paper making, 95-96 presence of oxidizing reagents, 29, 31/
drainage effects, 103 β-cyclodextrin as model compound, 21, 23
influence of cationic pulps on drainability and four principal synthesis pathways to
freeness of fiber stock suspensions, 105/ functionalized celluloses, 20f
mechanical properties, 103 functionalization via ester intermediates of
mechanical strength of lab sheets without filler, controlled stability and reactivity, 25, 29
104/ general principles of regioselectivity, 19-23
retention effects, 103 W H COSY NMR spectrum of 6-0-acetyl-2,3-
retention of filler (CaC03> as function of beating di-O-methylcellulose prepared from 6-0-
time at various degrees of cationization, 100, thexyldimethylsilyl cellulose, 29, 33/
102/ high 6-0-regioselectivity for
technical features on grafted cationic pulp fibers, thexydimethylsilylation, 23
103 methoxytrityl ethers as 6-O-protective group, 27/
Pénicillium citrioviride, enzyme system for preferential reaction site in AGU, 21

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 359

reasons to study regiocontrol, 21 spin diffusion experiments in polymer blends,

regiocontrol by selective protection and silyl 287-288
group migration, 34, 36/ Size exclusion chromatography (SEC), analysis of
regioselective synthesis and subsequent reactions degraded cellulose and derivatives, 185-186
of cellulose ethers, 23-25 Smart materials, surface segregation and self-
regioselectively functionalized cellulose assembly smart process, 315
examples, 22/ Solid-state nuclear magnetic resonance (NMR)
silyl group migration in alkylation of 2,6-di-O- spectroscopy
13
thexyldimethylsilyl starch, 26/ C chemical shifts for cellulose esters, 156f,
silylethers as suitable intermediates in ether and 157i
13
ester preparation, 34, 35/ C NMR spectra of cellulose acetate oligomers,
subsequent reactions of low-molecular weight 155,158/
Bunte salts, 30/ cellulose ester characterization, 153-159
subsequent reactions of 6-O-silylated celluloses, See also Nuclear magnetic resonance (NMR)
spectroscopy
synthesis of acetyl-methyl celluloses for H H Solution-state nuclear magnetic resonance (NMR)
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

COSY NMR spectroscopy, 29, 32/ spectroscopy

See also Liquid crystalline cellulose derivatives cellulose acetate analysis, 133
cellulose ethers, 151-152
chemical shift assignments, 133-143
conformation and molecular dynamics, 143-147
mixed esters of cellulose, 153,154/
SEC (size exclusion chromatography), analysis of structure-property relationships, 147, 151
degraded cellulose and derivatives, 185-186 See also Nuclear magnetic resonance (NMR)
SELDOM (selectivity by destruction of spectroscopy; Solid-state nuclear magnetic
magnetization) sequence, chemical shift filter resonance (NMR) spectroscopy
for spin diffusion experiments in polymer Solvents
blends, 287-288 Ν,Ν-dimethylacetamide with lithium chloride
Selective oxidation method for cellulose (DMAc/LiCl), 5,7
6-carboxy cellulose by phosphoric acid/sodium N-methyl morpholine-N-oxide-based solvent
nitrate/nitrite procedure, 81 process for spinning cellulose, 276
August 31, 2012 | http://pubs.acs.org

6-carboxy cellulose by sodium non-derivatizing solvents and solvent complexes

hypochlorite/bromide/TEMPO procedure, 81- for homogeneous cellulose modification, 3-7
82 See also Highly engineered cellulose derivatives
catalytic effect of sodium nitrite, 75-76,77/ by homogeneous phase modification
classical method with SPR (surface plasmon resonance spectroscopy),
periodate/chlorite/hydrogen peroxide, 75 analysis of regenerated cellulose Langmuir-
experimental materials, 80 Blodgett (LB) films, 311,313/
kinetic measurement procedure, 81 Steric exclusion chromatography
molecular weight distribution for pullulan by experimental procedure, 347
TEMPO/hypochlorite method, 79/ formation of aggregates in methylcellulose
molecular weight retention by TEMPO-method, solutions, 336, 339/
78-79 Substituent distribution in polysaccharides
oxidation in phosphoric acid with sodium advantages of gas chromatography/mass
nitrate/nitrite selective at 6-position, 75-76, spectrometry, 119
77/ analysis by mass spectroscopy, 126
13
oxidation method procedures, 80-81 C NMR technique for cellulose acetate, 164-
oxidation of β-cyclodextrin using phosphoric 167
acid/nitrate/nitrite method, 76,78r classification of substitution patterns for
oxidation of polysaccharides, 73-74 tetrameric fraction, 128/
TEMPO-oxidation of starch, pullulan, and complexity of two or more substituent types,
cellulose, 76, 78, 79f 119, 123/ 124, 125/
Selectivity by destruction of magnetization example of hydroxypropylmethylcellulose
(SELDOM) sequence, chemical shift filter for (HPMC), 119, 124

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 360
gas chromatography of HPMC after treatment for global properties of partially substituted cellulose
substituent analysis, 123/ derivatives in solution, 222-227
general approach by NMR for determination in instrumentation methods, 222
monomer unit, 119, 120/ interpretation attempt using fractal dimensions,
high-pH anion exchange chromatography for 227, 229-231
anionic derivatives of carboxymethyl cellulose Kratky presentation of data, 227, 228/
and sulfoethyl cellulose, 124, 126 local dynamics of macromolecules, 231, 234,
model studies on methyl amyloses, 126 235/
monomer analysis of cellulose derivatives, 118- possible explanations for uncommon aggregation
126 structure, 236
monomer substitution analyses by HPLC Theological measurements, 222
complement NMR, 124 shape of aggregates, 227-231
reductive-cleavage method of methylation, 119, transmission electron microscopy (TEM), 222
122/ See also Supramolecular architectures of
standard methylation analysis of cellulose cellulose derivatives
derivatives, 119, 121/ Surface plasmon resonance (SPR) spectroscopy,
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002

substitution pattern in polymer chain, 126-129 analysis of regenerated cellulose Langmuir-

Supramolecular architectures of cellulose Blodgett (LB) films, 311,313/
derivatives Surface segregation in cellulose ester blends
adsorption on regenerated cellulose LB films, blend preparation method, 321
311,314 characterization of fluorine(F)-containing
APTAC-C (poly(acrylamido cellulose derivatives, 325, 327
propyl)trimethylammonium chloride) additive comparison of DP determination by GPC and
in papermaking process, 311 NMR, 324/
chemical reactions of regeneration process, 312/ contact angle measurement method, 321
chemical structures of derivatives studied, 308/ contact angle measurements of blended films,
crosslinking reaction example for isopenyl 327, 330
cinnamyl cellulose ether, 309, 310/ copolymerization characterization methods, 320-
Langmuir-Blodgett (LB) technique, 306 321
layered cellulose assemblies as media of experimental materials, 318
chemical reactions, 307, 308/ experimental water contact angles for blends,
August 31, 2012 | http://pubs.acs.org

poly electrolyte (APTAC-C) structure, 312/ 329/

schematic of crosslinked layer system via LB F-containing cellulose esters, 318, 319/
technique, 310/ F-containing copolymer synthesis method, 320
surface plasmon resonance (SPR) spectra of identification of F-containing cellulose
APTAC-C adsorbed on cellulose and modified derivatives, 325/
films, 311, 313/ molecular weight characteristics of cellulose
transport properties by conventional membrane propionate oligomers, 322/
osmometer, 309,312/ principle of spontaneous self-assembly by
ultrathin membranes of cellulose assemblies on surface migration, 316-318
porous substrates, 307, 309, 311 proposed micelle formation by F-terminated
Supramolecular structures of cellulose derivatives oligomers, 327, 328/
aggregate formation, 218-219 reaction schemes for F-containing block
cellulose classes of fully and partially substituted oligomer and random copolymer, 323/
chains, 218-219 solution behavior of fluorinated cellulose
chemical structures of various cellulose derivatives, 327/
derivatives, 220-221 synthetic results for F-containing random and
concept of fractal dimensions, 225, 228/ block copolymers, 322-325
deviations from properties of molecularly
dispersed flexible chains, 225
evidence for fringed micelles from other
experiments, 231, 232/, 233/ Τ
experimental materials, 219
experimental versus predicted thermodynamic Thermogelation of methylcellulose
and hydrodynamic interactions, 222-225 mechanism of gelation, 344
fractal dimension correlated to number of arms, time-dependent and reversible phenomenon, 340,
229, 231,232/ 344

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.
 361

See also Chemical modification of cellulose U

6-0-Thexyldimethylchlorosilane (TDMS), Ultrasonic degradation. See Cellulose and
protective group with great regio-selectivity, 12 cellulose nitrate degradation
Thermomonosporafitsca,enzymes for hydrolysis
of cellulose derivatives, 203 W
Trichoderma reesei, enzymes for hydrolysis of Wood structure
cellulose derivatives, 202-203 hierarchical structures, 266
Triphenylmethyl (trityl) protective group in See also Cellulose and lignin as model for
derivatization, 12 hierarchical wood structure
Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0688.ix002
August 31, 2012 | http://pubs.acs.org

In Cellulose Derivatives; Heinze, T., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1998.