the microbial degradation of cellulose acetate

This item was submitted to Loughborough's Research Repository by the author. Items in Figshare are protected by copyright, with all rights reserved, unless otherwise indicated.

The microbial degradation of cellulose acetateThe microbial degradation of cellulose acetate

PLEASE CITE THE PUBLISHED VERSION

PUBLISHER

© E. Samios

LICENCE

CC BY-NC-ND 4.0

REPOSITORY RECORD

Samios, Eleftherios. 2019. “The Microbial Degradation of Cellulose Acetate”. figshare.https://hdl.handle.net/2134/11195.

https://lboro.figshare.com/

This item was submitted to Loughborough University as a PhD thesis by the author and is made available in the Institutional Repository

(https://dspace.lboro.ac.uk/) under the following Creative Commons Licence conditions.

For the full text of this licence, please go to: http://creativecommons.org/licenses/by-nc-nd/2.5/

LOUGHBOROUGH UNIVERSITY OF TECHNOLOGY

LIBRARY

=.

l' AUTHOR/FILING TITLE ,

$PlM,OS ~ -----------------------/----------------------

ACCESSION/COPY NO.

<54-0 1\ 6~ I 5 ----------------- ---- --- ---- --- ----------- --------VOL. NO.

28 JUN 199

27 JUN 1997

21 MAR 1997

CLASS MARK

II1 I 1111111 1IIIIn 11111

.-

THE MICROBIAL DEGRADATION OF CELLULOSE ACETATE

by

Eleftherios Samios, M.Sc., C.Chem. MRSC

A doctoral thesis submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of the L. U.T.

May 1995

Supervisors: R.K. Dart, B.Sc., Ph.D., C.Chem. FRSC, C.Biol. Mffiiol Prof. J.V. Dawkins, Ph.D., D.Sc., C.Chem. FRSC

Department of Chemistry

© E. Samios, 1995

-- --=f

,. ", ......... _,., ..... , ~,,_' ...... 4"

1o,~'gf1~}Dr~!!:~~~;~': Ur~'}~):"f:.~iXJ eT T,~.'· ,,' " . , ; . .'i,::';~:?y

, • ".',J, __ """,

~cc.-··-6'+ b \ \ b'l ( ~ "

ACKNOWLEDGEMENTS

I would like to thank Rothmans Research for sponsoring this work, and particularly Drs Neil Sinclair, Martin Duke, David Lindsay and Derek Mariner for their useful input throughout this project.

I would also like to thank Jill Thorley and John Brennan of the Microbiology Section and David Wilson of the Polymer Group for their help.

I would also like to thank the sometimes neglected stores team, Mike, Di and Margaret for always furnishing me with all the little things without which the research machine would undoubtedly grind to a halt!

Warm thanks must also be extended to John Kershaw for running some proton NMR spectra, to the University of Warwick for their proton NMR support, as well as to Dr David Apperley of the University of Durham for the solid state NMR spectra that he so quickly provided.

To Dr Elisabeth Meehan of Polymer Laboratories, Church Stretton, for the GPC results, I extend my warmest thanks.

I would also like to thank Helen Lowe, Sirnon Coe and Mike Richardson of Courtaulds Research for the long talks and their constructive comments, as well as for supplying the most enormous bag of cellulose acetate flake that I had ever seen!

I would also like to thank Drs I.G.Vlachonikolis and A.H. Osbaldestin of the Mathematics Department for their help with the statistical analysis undertaken in this work.

Finally, and above all, I would particularly like to thank my supervisors, Professor John Dawkins and Dr Kinsey Dart, for their active help and support throughout this rather difficult project. I feel particularly indebted to them, as they surpassed a merely supervisory role, and became my mentors in my quest to play my tiny part.in s'iientific research.

ORIGINALITY

All the work in this thesis has been carried out by the author except where acknowledged and has not previously been presented for a degree at this University or any other institution.

DEDICATION

This is the end of the line. A long, sometimes painjiil, yet always rewarding

line that saw a little boy with little more than a lunch box under his arms,

battle on, andjinally reach this ultimate challenge.

There have been people, however, who over the years have encouraged me,

helped me any way they could, laughed with my joy and cried with my sorrow.

People without whose help, 1 might have never reached this point.

1 would like to thank my grandfather for being always there for me and

playing so well the role of the father 1 never had. My grandmother, the

sweetest woman anyone could meet, for giving me her unconditional love and

for being the perfect substitute mother at home, while my mother. had to work,

so that my dreams could be jiiljilled.

But the biggest thank you and my complete gratitude goes to my mother, a

woman who battled alone, against all hardships, often neglecting herself, in

order to provide me with the best.

This thesis is for all of them, as well as for a few good people that 1 can call

my friends, for their understanding, their love, and their patience.

1 hope that 1 have done you all proud. ..

(

ABSTRACT

Cellulose acetate is a chemical of great industrial importance. Its uses range from the manufacture of textiles to cigarette filters.

Cellulose acetate is not biodegraded easily. The aim of this project was to identifY, micro-organisms that would attack cellulose acetate and to propose a possible mechanism for the biodegradation process.

Discarded cigarette filters were taken from the street and they were plated onto Sabouraud medium, a selective medium for fungi. The growth observed was on the outside of the filters. The middle portion of discarded cigarette filters was opened asceptically, and added in flasks containing nutrient broth. Eighty percent of the flasks showed no signs of bacteriological growth after 24 hours, showing that the inside of the filters was sterile. It would appear, that cigarette filters are a very effective barrier towards microbial penetration.

Cigarette filters were laid on potting compost, sand and tile surfaces, in order to monitor their progress over a period of 12 months. These experiments took place under moist and warm conditions, in order to enhance biological growth. The sand and tile experiments were abandoned after a relatively short period as no obvious changes could be seen.

The experiments on compost did not show any visible signs of biodegradation for 7 months. After that period, algal growth developed on the filters exposed to light, and a slight decrease in the degree of substitution (the average number of acetyl groups per anhydroglucose unit) was observed.

Cellulose acetates with varying degrees of substitution were synthesised and used as carbon source for the growth of the fungus Aspergillus jilmigatus, a common soil species. Previous experiments had shown that this species was the predominant one growing on the filters.

It was found that biodegradability varied with the degree of substitution. The higher the degree of substitution, the slower the biodegradation. Biodegradation could not be shown in cellulose acetate with a degree of substitution of2.5, thc material from which cigarette filters are made.

The degradation products were analysed by means of FTIR spectroscopy, IH and \3C NMR, solution viscosity and GPC. From the results obtained, it could be deduced that the biodegradation proceeded by a mild de-acetylation (esterase) prior to de-polymerisation (cellulase).

TABLE OF CONTENTS

Chapter I

1.1 Availability of cellulose .............................................................................. 1 1.2 Slruclure of cellulose .................................................................................. 2 1.3 Biosynthesis of cellulose ........................................................................... 14 1.4 Reactivity of cellulose ............................ ................................................... 15

1.4.1 Swelling ........................................................................................... 20 1.5 Synthesis of cellulose acetale .................................................................... 21

1.5.1 Pre-treatment of cellulose for acetylation ......................................... 21 1.5.2 Acetylation in solvent media ....................... ...................................... 22

1.5.2.1 Catalysts ............................................................................... 23 1.5.2.2 Diluents ................................................................................ 23 1.5.2.3 Acetylation processes ............................................................ 24

1.5.3 Saponification or "ripening" of cellulose acetate ............................ 24 1.5.4 Precipitation of cellulose acetate ..................................................... 25 1.5.5 Final aftertreatment of cellulose acetate .......................................... 26 1.5.6 Acetylation in non-solvent media:Preparation ofjibrous CA ........... 28

1.6 Industrial uses of cellulose acetate ......................................... .................. 28 1. 7 Breakdown of cellulose acetate ................................................................ 28 1.8 A review of some esterases and cellulases ................................................ 33

1.8.1 Esterases ............................................. ............................................ 33 1.8.2 Cellulases ........................................................................................ 33

Chapter 2

2.1 Studies on discarded cigarette jilters ..... .................................................. 35 2.1.1 Bacteriological activity of the jilters ............................................... 35 2.1.2 Studi.es on the jibres of the discardedjilters ................................... 39

2.2 Long term biodegradation studies on virginjilters .................................. 39 2.3 Study of additives on CA biodegradation ................................................ 41 2.4 Preparation of CA with given degree ofsubstitution ............................... 44 2.5 Determination of the DS ......................................................................... 45

2.5.1 Chemical determination ................................................................. 45 2.5.2 Spectroscopic determination ......................................... ................. 47

2.5.2.1 FT-JR spectroscopy ........................................................... .47 2.5.2.2 NMR spectroscopy ........................... ................................. .48

2.5.2.2.t 1C NMR ............................................................. .48 2.5.2.2.2 Solid state 13C NMR ........................................... .48

2.6 The biodegradation of CA of given DS using Aspergillus fumigatus ...... .49 2.6.1 The isolation of the fungus A~pergillus fumigatlfs ......................... .49 2.6.2 The inoculation of the CA-containing medium with the jimgus ..... .49

2.7 The characterisation of the CAs with given DS in their original and biodegradedform ....... 51

2.7.1 Molecular weight determination .................................................... 51 2.7.1.1 GPC method ...................................................................... 51 2.7.1.2·Solution viscosity method .................................................. 53

Chapter 3

3.1 Studies on discardedfilters .................................................................... .55 3.1.1 Bacteriological activity of the filters ........................................... ... 55

3.1.2 Studies on the fibres of the discardedfilters ................................... 57 3.2 Long term biodegradation studies on virginfilters .................................. 58 3.3 Study of additives on the CA biodegradation ....... .................................... 61 3.4 Preparation of CA with given DS ............................................................. 64 3.5 The biodegradation of CA of given DS using Aspergillus filmigatus ........ 76 3.6 The characterisation of the CAs with given DS in their original and

biodegradedform ........ 80 3.6.1 Molecular weight determination ................ ..................................... 80

3.6.1.1 GPC results ........................................................................ 80 3.6.1.1.1 THF system ......................................................... 80 3.6.1.1.2 DMAILiCI system ................................................ 85

3.6.1.2 Viscometry results .............................................................. 89 3.6.2 NMR spectroscopic determination ................................................ 1 05

3.6.2.1 Proton spectra .................................................................. l05 3.6.2.2 Carbon spectra ................................................................. 112

3.6.3 X-Ray diffraction analysis ............................................................ 117 3.7 General discussion ................................................................................ 118

Chapter 4

Conclusions ...... ............................................................................................. 127 Recommendations for future work .. ............................................................... 131

Bibliography ................................................................................................ 133

TABLE OF TABLES

Table 3.1 ......................................................................... 59 Table 3.2 ......................................................................... 60 Table 3.3 ......................................................................... 74 Table 3.4 ......................................................................... 78 Table 3.5 ......................................................................... 81 Table 3.6 ......................................................................... 82 Table 3.7 ......................................................................... 83

-Table 3.8 ......................................................................... 84 Table 3.9 ......................................................................... 86 Table3.10 ....................................................................... 86 Table 3.11 ....................................................................... 90 Table 3.12 ....................................................................... 90 Table 3.13 ....................................................................... 91 Table 3.14 ....................................................................... 91 Table 3.15 ....................................................................... 94 Table 3.16 ....................................................................... 95 Table 3.17 ....................................................................... 95 Table 3.18 ....................................................................... 96 Table 3.19 ....................................................................... 97 Table 3.20 ....................................................................... 97 Table 3.21 ....................................................................... 98 Table 3.22a ..................................................................... 99 Table 3.22b ..................................................................... 99 Table 3.23 ..................................................................... 100 Table 3.24 ..................................................................... 100 Table 3.25 ..................................................................... 101 Table 3.26 ..................................................................... 102 Table 3.27 ..................................................................... 102 Table 3.28 ..................................................................... 1 03 Table 3.29 ..................................................................... 107 Table 3.30 ..................................................................... 108 Table 3.31.. ................................................................... 114 Table 3.32 ..................................................................... 126

TABLE OF EQUATIONS

Equation 2.1 ......................................................... 46 Equation 2.2 ......................................................... 46 Equation 2.3 ......................................................... 53 Equation 2.4 ......................................................... 53 Equation 2.5 .... : .................................................... 53 Equation 2.6 .......................................................... 53 Equation 2.7 .......................................................... 53 Equation 2.8 .......................................................... 53 Equation 2.9 .......................................................... 53 Equation 2.1 0 ........................................................ 53 Equation 3.1 .......................................... ................ 70 Equation 3.2 .......................................................... 70 Equation 3.3 .......................................................... 70 Equation 3.4 .......................................................... 70 Equation 3.5 .......................................................... 71 Equation 3.6 .......................................................... 71 Equation 3.7 .......................................................... 71 Equation 3.8 .......................................................... 71 Equation 3.9 .......................................................... 72 Equation 3.10 ........................................................ 72 Equation 3.11... ..................................................... 72 Equation 3. 12 ........................................................ 73 Equation 3.13 ........................................................ 92 Equation 3.14 ........................................................ 93 Equation 3.15 ...................................................... 104 Equation 3.16 ...................................................... 104 Equation 3.17 ...................................................... 123 Equation 3.18 ...................................................... 124 Equation 3.19 ...................................................... 124 Equation 3.20 ...................................................... 124 Equation 3.21 ...................................................... 124 Equation 3.22 ...................................................... 124 Equation 3.23 ...................................................... 125

TABLE OF FIGURES

Figure 2.1 .......................................................... 42 Figure 2.2 .......................................................... 43 Figure 3.1.. ....................................................... 110

TABLE OF GRAPHS Note that the graph pages are not numbered. The page numbers below

correspond to the page number preceding the appropriate graph.

Graph 1 .............................................................. 78 Graph 2 .............................................................. 81 Graph 3 .............................................................. 90 Graph 4 .............................................................. 91 Graph 5 .............................................................. 91 Graph 6 .............................................................. 94 Graph6a ............................................................. 94 Graph6b ............................................................. 94 Graph 7 ............................................................... 95 Graph 8 ............................................................... 97 Graph 9 ............................................................... 97 Graph 10 ............................................................. 98 Graph 11 ........................................................... 100 Graph 12 ........................................................... 100 Graph 13 ........................................................... 101 Graph 14 ........................................................... 102 Graph 15 ........................................................... 102 Graph k = 0.5 .................................................... 126 Graph k = 1.0 .................................................... 126 Graph k = 1.5 ..................................................... 126 Graph k = 2.0 ..................................................... 126 Graph k = 2.5 ..................................................... 126 Graph k = 3.0 ..................................................... 126

TABLE OF SPECTRA Please note that the spectra pages are not numbered. The page numbers below

correspond to the page number preceding the appropriate spectrum.

FT-JR spectra ......................................................... 78 IH NMR spectra .................................................... 106 l3e NMR spectra ................................................... 113 X-Ray spectra ....................................................... 117

TABLE OF PHOTOGRAPHS Please note that the photograph pages are not numbered. The page numbers below correspond to the page number preceding the appropriate photograph.

Photographs 1-4 ................................................... 55 Photographs 5-6 .................................................. .56 Photographs 7-16 ................................................. 57 Photographs 17-29 ............................................... 58 Photographs 30-33 .............................................. .59 Photographs 34-37 ............................................... 60 Photograph 38 ...................................................... 61 Photographs 39-44 ................................................ 77

TABLE OF CHROMATOGRAMS Please note that the chromatogram pages are not numbered. The page numbers below correspond to the page number preceding the appropriate chromatogram.

Starting polymers (PEOIPEG calibration) ............. 85 Starting polymers (Polystyrene calibration) ........... 85 Biodegraded polymers (PEOIPEG calibration) ...... 85 Biodegraded polymers (Polystyrene calibration) .... 85

J

CHAPTERl

INTRODUCTION

Chapter 1 - Introduction

1.1 Availability of cellulose

There are several sources from which cellulose can be obtained, however, only

two of them are of significance in industryl. The most widely used source is

wood. There are two main classes of wood. The hardwoods or angiosperms,

such as oak, are generally composed of closely packed cells with thick walls;

the softwoods or gymnosperms, such as pine or cedar, are usually composed

of large cells with thin walls. The composition of wood varies with the

species and also with the part of the tree from which it is taken. In round

numbers, on a dry weight basis, wood contains 40 to 50% cellulose, 20 to 30%

lignin, and 10 to 30% hemicelluloses and polysaccharides other than cellulose.

Resins, gums, proteins and mineral compounds are also present.

Several kinds of wood are used in the preparation of wood pulp. The most

common are the pines (Pinus spp.), spruces (Picea spp.), fus (Abies spp.),

beeches (Fagus spp.), poplars and aspens (Populus spp.), and hemlocks

(Tsuga spp., especially Tsuga canadensis).

The second important source of cellulose is seed hairs. The most important of

these is cotton. The cotton fibre contains the lowest percentage of

noncellulosic material, (4 to 12%) of all the commercial raw materials for

cellulose. This makes purification simpler than for other cellulosic materials.

The cotton plant belongs to the genus Gossypium, which has been a difficult

genus to classify, so that confusion exists in the enumeration and

nomenclature of the species. It is probable that practically all the important

commercial varieties of cotton are hybrids.

There are no other seed hairs which have found application to the same extent

as cotton. The kapok fibre, from Ceiba pentandra is of. importance. Kapok

1

Chapter I - Introduction

fibre is widely used as a stuffmg material particularly for life preservers,

because of its buoyancy and resistance to moisture. It contains from 55 to

65% cellulose on a moisture-free basis, but is not used as a source of cellulose

pulp. Bombax cotton obtained from several species of Bombax is used for

wadding and upholstering. Seed hairs from certain species of Asclepias, such

as milkweed (A. syriaca), are also used in upholstery, as is the pulu fibre from

fern trees of the Cibotium genus.

The attraction of using wood and cotton as sources for cellulose is that they

are both bio-renewable if they are managed properly.

1.2 Structure of celIulose2-6

The nature of the building units and their linking, together with the average

molecular length and its range, are of primary importance in the establishment

of the structure of macromolecules. These aspects have been thoroughly

studied for cellulose (see also figure 3.1, Chapter 3).

Cellulose, on hydrolysis with inorganic acids, gives almost a quantitative yield

of glucose. Completely acetylated cotton on methanolysis gives an

equilibrium mixture of a- and ~-acetylated methylglucose, which accounts for

98.1 % of cellulose. The only products of the reaction are glucosides, and as

the reaction mixture fails to give the furfural test for pentosans, this is taken as

good evidence to show that cellulose consists only of glucose. The lower

than-theoretical yields are attributed to reversion and/or decomposition of

glucose by strong acids.

Purified cotton and cellulose acetate can be hydrolysed with 40% HCI to

glucose, which is then estimated as CO2 by catalytic oxidation with FeCh.

Results show 99.1% of glucose in the original cellulose in.both cases.

2


Sugar identification by chromatography shows that very pure forms of

cellulose, such as ramie, yield only glucose on hydrolysis. All the above tests

show that the basic monomeric unit of cellulose is glucose.

The reaction of one a- or ~-glucose form with a hydroxyl group of another

glucose molecule gives a- or ~- bonded dimers, trimers or higher polymers

bonded through the anomeric carbon atom. Maltose is the a(I-4)-linked dimer

of glucose. This is the repeating unit found in starch dextrins and amylose

(see also figure 2.2, Chapter 2).

Likewise, cellulose is the ~(1-4)-linked dimer of glucose and the polymeric

chain built up of cellobiose residues in cellulose. Cellobiose is one of the

major products of the hydrolysis of cellulose under acid conditions. Cellulose

itself has been shown to be poly-(1-4),~-D-glucopyranose.

Carefully isolated cellulose shows very little reducing power as theoretically

there is only one reducing group per chain, but it develops this property on

hydrolysis. This fact, as well as the production of a nearly theoretical yield of

cellobiose, indicates that the bond is glucosidic in nature. The bond involves

the anomeric carbon of one glucose molecule and a hydroxyl group of another.

Cellobiose hexaacetate, obtained by hydrolysis of fully acetylated cellulose,

has been shown to resist the yeast enzyme a-D-glucosidase (maltase) (EC

3.2.1.20), which readily hydrolyses the a-glucosidic bonds in the starch

degradation product, maltose. Cellobiose is hydrolysed to glucose by emulsin,

which establishes the disaccharide linkage in the ~ configuration. Work on the

infrared absorption of cellulose and starch compared with known oligomers

has reconfinned the ~ configuration as the only interglucosidic valence bond

in cellulose.

3


Cellotriose, cellotetraose, cellopentaose, cellohexaose and celloheptaose have

also been identified as products of hydrolysis of cellulose. From a comparison

of the physical and chemical data obtained from compounds of increasing

anhydroglucose (glucopyranose) content, and from the fact that they showed

no chemical differences, it became evident that their formulae should

extrapolate to cellulose when the number of anhydroglucose units was

assumed to be very large.

Methylated cellulose, obtained by treating sodium hydroxide-soaked cotton

with dimethyl sulphate, upon hydrolysis yields only the 2,3,6 trimethyl ether

of glucose. It was therefore established that the three free hydroxyl groups in

the cellulose occupy the 2,3 and 6 positions. The groups have decreasing

acidic properties in the order 2,3 and 6. The primary hydroxyl group at

position 6 is sterically the most unhindered.

Methylation of cellobiose gave the crystalline octamethyl derivative, which

upon acid hydrolysis produced 2,3,4,6-tetra-methyl-J3-D-glucopyranose and

1,2,3,6-tetra-methyl-J3-D-glucopyranose in equal amounts. Cellobiose could

therefore only be J3-D-glucopyranosyl-(1-4) or (1-5)-J3-D-glucopyranose. The

fact that no 5-methyl derivative was found in either of the two fractions

supported the formula already established theoretically, that the 5 position in

the molecule was inaccessible for chemical reactions. This was finally proven

when the D-glucose-2,3,5,6-tetramethyl ether was obtained with D-glucose-

2,3,4,6-tetramethyl ether from an octamethyl-cellobionic acid. This acid was

prepared by eliminating the cyclic structure in the reducing half of the

cellobiose by oxidation with bromine water. Cellobiose was thus established

to be J3-D-gluco-pyranosyl-(1-4)-J3-D-glucopyranose.

4


Four types of evidence for the uniformity of the linkage in cellulose have been

described. This is accurate up to an extent of about 99%. Chemical evidence

is derived from methylation studies of glucose, other oligomers, and cellulose.

Polarimetric evidence is based on optical rotation of methyl cellulose in

suitable solvents. The values show very good agreement with the theoretical

value based on considering the chain to be built up of only j3-glucopyranose

units. Studies on cellopentaose, cellohexaose and celloheptaose further

confIrm the results.

Kinetic evidence is obtained from the change in rate, optical rotation, and

reducing power during acid hydrolysis. These can be quantitatively accounted

for by assuming that all the hydrolysable links in a uniform chain are equal

and equivalent except for the reducing terminal unit which can be neglected

when the chains are inf'mitely long: otherwise, a correction can be made for

the faster rate of the cleavage of the bonds adjacent to the reducing end of the

chain molecules.

Quantitative evidence consisted of the assessment of the actual quantities of

oligosaccharides formed during hydrolysis. By assuming all bonds to be

equivalent, it is possible by a mathematical treatment to explain the low yields

of intermediate compounds. The higher yield of cellobiose was attributed to

its ready crystallisation, which prevented further breakdown.

As cellulose chain molecules are very long, it is very difficult to detect a small

number of bonds other than the j3-glucosidic, if present and even the best

evidence applies only to 99% of the bonds. Further work incorporating

advanced techniques and mathematics in the study has reduced the extent of

error in the proofto about 0.1%. The uncertainty about the nature of this 0.1%

of bonds in the cellulose itself has for some time r~sulted in a heated

5


controversy over the existence of weak bonds in cellulose. Certain studies

indicate the absence of such bonds in native cellulose, but it is possible that

there may be a few j3-glucosidic bonds which are sensitised to certain

reactions, such as acid hydrolysis by induction effects or physical strains in the

molecules.

As early as 1920, it was recognised that cellulose from such widely different

sources as cotton, ramie and wood gave identical X-Ray diagrams and

concluded that all these fibres had identical crystalline structures. Later work

has shown that this identity extends to all other natural cellulosic materials

including bacterial and animal cellulose. This crystal form has become

commonly known as the "native" cellulose form, but the more general term

"cellulose f' is now preferred. If cellulose is dissolved and precipitated from

solution, however, the molecules do not reassemble into the characteristic

cellulose I lattice but into an allotropic modification. This modification is

known as "regenerated" cellulose, or "cellulose II)16.

However, the structure of cellulose in general is subject to some speculation.

For example, new evidenceS for native cellulose using solid state l3C NMR

indicates that native cellulose has two distinct crystalline forms. One form is

dominant in bacterial and algal celluloses, whereas the other is dominant in

celluloses from higher plants. The resonance multiplicities reported in the

solid-state l3C NMR spectra of native celluloses have been examined at higher

resolution for a variety of native forms and for a sample of regenerated

cellulose 1. The pattern of variation among spectra of the native forms

suggests that they are all composites of two distinct crystalline forms of

cellulose. This observation provides a basis for re-assessing some of the

conflicting interpretations of data concerning the structures of native

celluloses. Significant questions remain with respect to t\;te structure of native

6


celluloses. These include the symmetry of the unit cell and its application in

the analysis of diffraction data as well as the confonnational differences, or

lack thereof, between native cellulose and its most common alternate

polymorph. The solid-state 13C NMR spectra represent an important source of

new infonnation that can help to resolve these questions.

Although spectra of pure samples of cellulose II could be rationalised in tenns

of non-equivalent sites within a unique unit cell, the spectra of native

celluloses reveal multiplicities that cannot be so explained. The spectra were

recorded by applying the cross polarisation-magic angle spirming technique in

a high-field instrument. This method involves cross polarisation to enhance

the l3C signal, high-power proton dipolar decoupling to eliminate dipolar line

broadening due to protons, and spinning of the sample about an axis at a

particular angle to the static field to eliminate chemical shift anisotropy. The

samples included a bacterial cellulose from Acetobacter xylinum, an algal

cellulose from Valonia ventricosa and two fibrous celluloses, cotton linters

and ramie. Finally, celluloses of the pure polymorphic fonns I and II were

regenerated. From the spectra obtained, the novel idea of multiple crystal

fonns was brought forward. At the same time, it confinned some reports that

Acetobacter and Valonia celluloses were structurally different from other

celluloses such as cotton and ramie.

New light on the native cellulose structure was also shed by R. Colvin4•

Native cellulose microfibrils from many sources have been considered

separate, autonomous, one-dimensional entities. Direct unions of the

microfibrils were assumed not to exist. Colvin suggests that the native,

minimally disturbed microfibrils in the pellicle of bacterial cellulose join to

fonn three-dimensional interconnected structures. It was suspected that the

ultrastructure of bacterial cellulose might be different to that of other sources,

and therefore, the fine morphology of cellulose from a representative green

7


plant was also established. Colvin thus investigated cellulose in mature cotton

as a representative type. Extensive observations with scanning electron

mlcroscopy and independent observations with transmission electron

Illicroscopy confIrm that the native, undisturbed structure of the cellulose

component from the pellicle of Acetobacter xylinllm is three-dimensional. The

microfIbrils are extensively cross-linked to form a coherent whole and are not

simply intertwined or superimposed.

The most recent observations by Colvin indicate that the native, undisturbed

microfIbrils of cotton have the same morphology as the bacterial cellulose. All

available means of study indicate extensive interaction and crosslinking

between microfIbrils, both in the native state and after disruption. The

conclusion Colvin has reached is that the native cotton cellulose and that from

the bacterium probably have the same structure. The microfIbril links that

produce the cellulose network may vary in strength from strong hydrogen

bonds to true covalent connections. The sum of these interconnections, Colvin

suggests, is responsible for the extensive, intact sheet structure that has been

frequently reported.

It may be premature to extend the conclusion to all celluloses. However, it is

clear that two of the purest sources, one bacterial and one plant in origin, do

have a coherent three-dimensional structure. If this structure is limited in one

direction, it may appear as a sheet, but the three-dimensional connections

remam. Colvin speculates that the reason this structure has been overlooked

may be attributed to undue emphasis on trying to fmd the fInal, physical,

structural unit in cellulose from fIbre to microfIbril to the so-called elementary

fIbril. This investigative path requires severe maceration and digestion, and

destroying the evidence for interconnections.

8


The structure proposed by Colvin does not eliminate the need for microfibrils.

The essential difference is simply that microfibrils are now considered

interconnected and are not autonomous threads. This structure is fully

consistent with the observations of other investigators. If both bacterial and

plant sources yield a three-dimensional structure, an interesting sidelight is

that this structure eliminates the spinneret as a means for forming the fibrils.

Spinnerets are capable of forming only one-dimensional fibres. Extrusion

processes are therefore unlikely.

The structure of cellulose II, is universally reported to have antiparallel chains.

That structure was supposed to result from some nearly instantaneous

conversion from the parallel chain structure present in native cellulose I on

treatment with strong alkali or on regeneration from solutions of cellulose or

cellulose derivatives. However, new evidence was brought forward3 which

contests those findings as they simply cannot take place in the time constraints

and conditions under which it is known to occur. The belief dates back to

Meyer and Misch7, whose initial calculations indicated that antiparallel chains

packed more comfortably than parallel chains into the cellulose II unit cell

dimensions.

Such a conversion from parallel to antiparallel packing requires that every

other chain having 1000 or more glucose units must sever its bonding

relationships with its neighbours, assume a new position that is opposite by

1800, and become energetically more comfortable almost instantaneously.

Perhaps even more confusing is that this result also requires molecules in an 8-

10% solution to somehow find alternate neighbours headed in opposite

directions while flowing through a spinnerette at 100 metres/minute,

coagulating and regenerating to make a fibre or film. Thi~ process supposedly

has been justified by the better fit of the antiparallel chains into the new

9


cellulose n unit cell dimensions. By usmg a custom made computer

programme the workers have been able to vary 252 parameters in a full matrix

minimisation of packing energies for each of the parallel up, parallel down,

and antiparallel configurations. The three coordinates of one atom (x, y and z

direction) were held fixed in order to establish the origin of the triclinic unit

cell. The computer evaluation of packing energy minimisation by allowing the

atoms and the chains to seek their lowest energy positions in the cellulose n unit cell led to some interesting results.

The interatomic contact and interactions in both the parallel-up and

antiparallel models are equally acceptable and have nearly equal van der

Waals energies of -45.7 and -43.8 kcallmole respectively. The lattice

parameters for the parallel-up and the antiparallel models are essentially the

same and differ by less than 0.904 A. This is well within the standard

deviation accuracy of X-ray diffraction measurements, so it is impossible to

differentiate between these two models on the basis of any expected lattice

parameter differences. All hydroxyl hydrogens are involved in hydrogen

bonding, and all the 0(5) ring oxygens are involved in intramolecular

hydrogen bonding with the H-0(3) of the adjacent ring for both the parallel

and the antiparallel models. All the 0(6)s are in the "gt" gauche-trans

position and therefore, are not involved in intramolecular hydrogen bonds with

H-0(2) of an adjacent glucose in the parallel-up model. This allows both the

H-0(2) and the H-0(6) freedom to be involved primarily in intermolecular

hydrogen bonds with neighbouring chains. This result is in line with modulus

of elasticity values predicted by molecular mechanic energy calculations and

further confirmed by experimentally observed elastic modulus values.

Therefore, a more logically acceptable view of the cellulose n structure can be

obtained. Starting with cellulose I having a parallel ch~in structure wherein

the O( 6)s are in a "tg" position. This allows considerable hydrogen bonding to

10


occur intermolecularly between adjacent (6) hydroxyls and (2) hydroxyls

along each chain. Such a result is certainly statistically favourable during the

cellulose polymerisation stage. Simultaneously the C(3) hydroxyl bonds to an

adjacent 0(5) ring oxygen, and all this fits properly into the unit cell

dimensions determined for cellulose I.

On treatment with strong alkali the intramolecular hydrogen bonds between

C(2) and C(6) hydroxyls are broken, the glucose residues twist, and the C(6)

hydroxyl can now turn to a more energetically favourable gt position as the

unit cell angle becomes more oblique. The C(6) and C(2) hydroxyls can form

new intermolecular hydrogen bonds with neighbouring chain hydroxyls that

are stronger by -45.7 kcals/mole. These changes occur with retention of

parallel packing and without invoking a need to have alternate chains changing

direction. If the cellulose is totally dissolved and regenerated from solution,

then the chains can energetically pack equally well into the cellulose IT unit

cell dimensions in a parallel or antiparallel manner (-45.7 kcallmole vs -43.8

kcallmole - essentially the same energy change) to give a mixture of crystals

that are indistinguishable by X-ray diffraction techniques. This does not

require an antiparallel structure but allows either parallel or antiparallel

structures to form, whichever is most probable for a given number of adjacent

chains at the time.

One further ramification of these new computer programme results is that the

data clearly indicate that the cellulose repeat unit is not glucose but the glucose

dimer cellobiose. Adjacent anhydroglucose units are not identical, and there is

a slight but significant difference in the C-H-O angles for H-C(I)-O compared

with H-C(4)-0 hydrogens at the glycosidal bridge (Atalla\ This could have

interesting implications on the mechanism of cellulose formation by plants and

bacterial cells.

11


The chain molecules in natural cellulose are not of the same length. The

number of glucose units in different chains varies. This is revealed by

different samples of cellulose of no detectable chemical difference giving

different alkali solubilities and viscosities. A given sample represents a

molecular homologous series in which there is no molecular heterogeneity.

One has to deal with averages, such as average molecular weight and average

chain length. The degree of polymerisation (OP) of unopened cotton has been

reported at 15,300; this value decreases rapidly to 8,100 on exposure to the

atmosphere. Bast fibres have an average OP of 9,900, while wood species

vary between 7,500 and 10,500.

The long cellulose molecule effectively camouflages the presence of the two

end groups in chemical analysis. Upon methylation the non-reducing end

group should give a O-glucose-2,3,4,6-tetramethyl ether; this has been

obtained under very careful reaction conditions. The reducing end, on the

other hand, could never be isolated, presumably because of the fast

demethylation of the Cl methyl group during acid hydrolysis.

Six-membered rings can assume either a boat or chair conformation according

to their energy level. The pyranose ring assumes a chair form in preference to

the boat form because of internal strain in the latter. Two possible chair

conformations will exist in the case of the pyran ring; in the fust, the oxygen

bearing substituents lie mainly in the same plane as the pyranose ring

(equatorial), and the hydrogen atoms stand away from the ring (axial); the

situation is reversed in the second case. O-glucopyranosides exist and react in

the chair conformation. The transition into a boat form., under certain strains

or activating-energy influences, might be the cause of differences in physical

structure and chemical reactivity of cellulose under these conditions. Similar

conclusions have been obtained from X-ray diffraction patterns and from

infrared spectroscopy.

12


Whenever the distance between the various oxygen and hydrogen atoms in the

cellulose molecule reaches 0.3nm or less, they interact with each other to form

intramolecular and intermolecular hydrogen bonds. Infrared spectroscopy has

verified the existence of these hydrogen bonds. The intramolecular hydrogen

bridges anchor the anhydroglucose units to a very limited region of free play

around the acetal linkage. Thus, they impart a certain stiffness to the cellulose

molecule. This and the fact that the (1-4)-13 bond demands a rotation of 1800

of each subsequent glucose unit to fit the l3-configuration of the connecting

hemiacetal linkage, gives the cellulose molecule a rod-like chain structure.

The l3-glucosidic linkage in cellulose and the resulting intramolecular

hydrogen bonds render the cellulose molecule straight and stiff. On the other

hand, in starch the glucose units can be arranged in a helix-like chain

molecule.

The involvement of the hydroxyl groups in hydrogen bonding, as well as

general dispersion forces, determined by the proximity of neighbouring atoms,

impart a different reactivity to the three hydroxyl groups available for

chemical reactions. Esterification and etherification studies have shown that

the C-6 group is esterified ten times faster than the other groups, whereas on

etherification the C-2 group is etherified twice as fast as the C-3. The primary

alcoholic group at C-6 is distinguished from the two secondary alcoholic

groups in that it has an axis of free rotation around the C-5 to C-6 bond, which

is, however, somewhat restricted by the hydrogen bonds. It has been

observed by infrared spectroscopy that rotational isomers must exist especially

in the alkali-swollen cellulose. The reactivity of the primary alcoholic groups

seems to be related to this isomerisation. Structural differences between

cotton, wood cellulose and mercerised cellulose appear to be due to this.

rotational isomerisation. Equally relevant are aspects of cellulose reactivity

(see section 1.4)

13


1.3 Biosynthesis of celIu)ose8-IO.

12

Cellulose is the most abundant carbohydrate in nature and the most abundant

compound of plant cell walls, although it has been suggested that chitin might

also be a possibility. It is the basic structural material of the cell walls of all

higher land plants and is also found in some algae. It is also synthesised by a

few bacteria. Cellulose is a 13(1-4)-linked D-glucan having a DP of about

10,000. The glucan chains line up in a parallel arrangement to form

elementary fibrils. The cellulose microfibril has a well defined crystalline

structure. Almost nothing is known about how the synthesis of cellulose

chains is initiated or terminated, nor whether the synthesis of primary cell wall

cellulose is mediated by the same enzyme system or mechanism as secondary

cell wall cellulose.

14


1.4 Reactivity of cellulose

CeJIulose reacts as a trihydric alcohol with one primary and two secondary

hydroxyl groups per glucose unit. The relative reactivity of the hydroxyl

groups of both low molecular mass carbohydrates and ceJIuIose has been

studied. In the former, the 2- and 6- hydroxyl groups are usuaJIy the most

reactive. With ceJIulose, certain data indicate the preferential reactivity of the

2- hydroxyl and others of the 6- hydroxyl group.

The reactions of ceJIulose may be conveniently divided into two main kinds:

those involving the hydroxyl groups and those comprising a degradation of the

chain molecules.

The former includes the following reactions:

I. Esterification: nitration, acetylation and xanthation.

2. Etherification: alkylation and benzylation.

3. Replacement of -OH by -NH2 and halogen.

4. Replacement of -H in -OH by Na.

5. Oxidation of -CH20H to -C02H.

6. Oxidation of secondary -OH groups to aldehyde and carbonyl.

7. Formation of addition compounds with acids, bases, and salts.

These reactions take place without breakdown of the chain and may have only

a local effect, e.g., causing change in the terminal groups or in individual

members of the chain, or they may affect aJI, or the majority of, the members

of the chain simultaneously. In the former case it is exceedingly difficult to

detect the changes analytically in high molecular weight· products, for which

15


reagents of the utmost sensitivity are required. Sometimes, however, these

reactions are manifested indirectly. Changes in the ceJlulose molecule

resulting from oxidation in an acid medium affect only a few members of the

chain and are scarcely detected by direct means; yet, later on they become

clearly noticeable in that the chain splits up at the affected parts upon

subsequent contact with alkaline liquids. There are many chemical reactions -

the esterification and etherification of the hydroxyl groups in particular -

which are liable to take place over the entire chain more or less uniformly,

with often little difference in reactivity of the -OH groups in positions 2,3, and

6, though occasionally possible distinctions have been made.

The degradative reactions of importance are those brought about by hydrolysis

of the glucosidic bonds and by oxidation. Hydrolytic breakdown takes place

in the presence of acids, while oxidation may occur in an alkaline, acid or

neutral medium. As hydrolytic degradation involves the scission of the

ceJlulose acetal bonds, (i.e. the f3 -glucosidic bonds) by acids, the resulting

increase in reducing power, the decrease in DP of ceJlulose, and the extent of

oligosaccharide formation provide methods to study the kinetics of this

reaction. Hydrolysis of cellulose can be either homogeneous or

heterogeneous, i.e., intracrystaJline or intercrystalline, being dependent upon

the sweJling capacity of the acid used. In the case of homogeneous hydrolysis,

under mild conditions the end products are the monomer (glucose) and some

reversion products of the gentiobiose 1-6 linked type. When heterogeneous

hydrolysis is carried out, the rate of the reaction decreases as it progresses to

regions of high crystallinity. There is an initial rapid rate which gradually

decreases to a leveJling-off DP value. The first effect of heterogeneous acid

hydrolysis is the formation of hydro cell uloses. When the ceJlulose is oxidised,

the chain usually breaks down, probably as the result of opening and cleavage

of the monomeric rings. Side by side with this, other reactions not interfering

with chain length may occur. These include oxidation of the primary hydroxyl

16


groups in the C-6 position to aldehyde or carboxyl groups, oxidation of the

secondary hydroxyl groups at C-2 and C-3 to ketone groups, oxidative opening

of rings to form two aldehyde or carboxyl groups, etc.

The main reactions of cellulose have been classified into four "possibilities"

which are derived from the two-phase crystalline-amorphous structural

concept and the physical and chemical reactivities of the hydroxyl groups.

These possibilities are:

(i) Reaction takes place exclusively with one of the two types of hydroxyl

groups (either primary or secondary), and is topochemical.

(ii) Reaction takes place preferentially with one of the two types of hydroxyl

groups and is permutoid (quasi-homogeneous). The difference in reactivity of

the hydroxyl types determines the rate, e.g. tritylation.

(iii) Reaction takes place equally with hydroxyls of both types, and the

reaction is topochemical.

(iv) Reaction takes place equally with both types of hydroxyl group, and the

reaction is permutoid. The rate of the reaction is uniform throughout, e.g.

nitration.

One might be tempted to regard cellulose as a trihydric alcohol similar to

sugars in the type of its reactions. This, however, has been definitely shown

not to be the case when reactions with cellulose are performed. Cellulose,

being a fibre-forming, high polymeric material behaves quite differently from

simple trihydric alcohols; The following division accounts for most cellulose

reactions, although it naturally represents a considerable simplification of the

. extremely complex state of the matter:

17


(A) Heterogeneous reactions [which include surface reactions,

macroheterogeneous reactions, micellar heterogeneous reactions and

permutoid (or quasi-homogeneous) reactions], and

(B) Homogeneous reactions.

Surface reactions are limited to the fibre surface, with the capillary system of

pores and cracks remaining closed throughout the reaction due to the absence

of any form of swelling. In other cases the molar volume of the reagent may

be too great to allow the penetration of the latter into the pores and interstices,

for example, in the case of isoalkyl halides.

Macroheterogeneous reactions are characterised by the fact that the conversion

starts on the substrate surface and then proceeds through the fibre from layer

to layer. Two different processes may occur:

(a) If the secondary wall is penetrable but the primary wall is only partially

swollen, then the reaction will start at the points of the fibre surface at which

the cuticle has been damaged. A typical topochemical type of reaction is thus

obtained. The capillaries which may be regarded as submicroscopical reaction

rooms, serve as supplying canals for the reagent. If they contain a liquid,

immiscible with the reagents, the penetration of the latter will naturally be

opposed, and a topochemical macroheterogeneous reaction will be obtained.

The acetylation of fibrous cellulose with acetic anhydride in benzene furnishes

a typical example. Another is the denitration of cellulose nitrate in alcohol

(b) If the primary wall is swollen but the secondary wall is only slightly

penetrable, the reaction will begin simultaneously all over the fibre surface

and then proceed from layer to layer in the secondary walL An example of

this type is the denitration of cellulose nitrate in aqueous media.

18


If both the primary and secondary wall are swollen in a cellulose reaction, the

reaction is of the microheterogeneous type. Microheterogeneous reactions in

turn consist of two very different types:

(i) micellar heterogeneous reactions, i.e. they involve a rapid and complete

conversion of the micellar surface, after which the micelle itself undergoes a

reaction layer after layer. An inner core of unchanged cellulose remains until

the very end,

(ii) permutoid reactions, which refer to the ability of the permutites to

exchange ions or molecules present in the lattice quantitatively and rapidly

with ions or molecules outside .the lattice, the crystal structure still being

retained.

Intercrystalline or intermicellar reactions take place in the amorphous regions

or on the surface of crystal lites, intracrystalline or intramicellar reactions

within the crystallites. The former involve no changes in the X-ray diagram of

the sample in question, whereas the latter lead to the formation of a new

pattern, the two types being distinguished in this way.

Reactions of cellulose in dispersion conditions produce a very uniform

distribution of the entering groups along the chain molecules, as indicated by

the good solubility properties of the product obtained. Several cases are

known in which cellulose is converted quite homogeneously. An example is

the acetylation of cellulose dispersed in phosphoric acid. The reaction of

cellulose derivatives, dissolved in pyridine, with tosyl chloride is a

homogeneous esterification reaction. All kinds of cellulose ethers can also be

prepared homogeneously with the help of quanternary ammonium bases.

19


1.4.1 Swellini,6

A solid is said to swell when it takes up a liquid while at the same time (a) it

does not lose its apparent homogeneity, (b) its dimensions are enlarged, and

(c) its cohesion is diminished. This swelling is not to be confused with

capillary imbibition which is due to the fme structure of cellulose with its

capillaries and interstices. As the dimensions of the latter are partially

submicroscopic, there is a continuous transition from swelling to capillary

imbibition. The swelling phenomena of cellulose are suitably subdivided

through intermicellar or intramicellar swelling, each with either limited or

unlimited swelling. Intermicellar swelling is restricted to the amorphous

domains of the fibre, and in X-ray diagrams the diffuse halo of the swelling

agent is therefore superimposed upon the original cellulose pattern. If only a

limited amount of reagent is absorbed, the swelling will be limited; otherwise

it would be unlimited. In the case of intramicellar swelling, the reagents

penetrate the crystallites. If the swelling is limited, formation of an addition

(swelling) compound takes place between the cellulose and the swelling agent.

This compound has a distinct X-ray pattern. In the case of unlimited swelling,

the uptake of the reagent continues until the crystallites are completely

dispersed.

Cellulose undergoes swelling in solutions of acids, bases, and salts as well as

in some organic solvents. Swelling generally involves breaking of

intermolecular bonding of cellulose and in many cases formation of new bonds

with the swelling agents to give swelling compounds. Such swelling is an

important feature of cellulose modification through graft polymerisation

whether to cellulose or one of its derivatives.

20


1.5 Synthesis of cellulose acetatel4•15

The commercial manufacture of cellulose acetate (CA) involves the following

operations:

I. Pre-treatment ofthe cellulose

2. The acetylation stage

3. Saponification or "ripening" stage

4. Precipitation

5. Aftertreatment

1.5.1 Pre-treatment of cellulose for acetylation

Early in the development of the commercial production of CA it was realised

that the esterification of cellulose proceeded more smoothly and more rapidly

if the raw material was first modified by some form of pre-treatment. In the

processes described in most of the early patent specifications, modified or

hydrocellulose was used, and this was usually prepared from the natural cotton

in an entirely separate process. When industrial requirements, as regards both

quality and price, became more stringent, it became necessary to make the

acetylation as speedy, yet as controllable, as possible and the pre-treatment of

the cellulose was introduced as an integral part of the acetylation process. The

pre-treatment of the raw cotton may effect a purification of the raw material;

but its main function is to open up the cellulosic structure so that the

acetylating agent may penetrate both rapidly and uniformly and not merely act

on the surface. .

The pre-treatment which requires little control, can be carried out on a large

scale and requires no special plant, so that the actual time in the costly

acetylisers is materially reduced. The most common treatment is using glacial

21


acetic acid, either with or without a small addition of an inorganic acid such as

sulphuric. Pre-treatment may be carried out at any temperature up to the

boiling point of the acetic acid, but it is usually effected at or below 50°C. In

the absence of a catalyst the molecular weight of the cellulose (as measured by

its solution viscosity) is not greatly affected by the pre-treatment, but in the

presence of a catalyst, such as sulphuric acid, the viscosity may be drastically

reduced. The pre-treatment is thus in effect a mild acetylation, rendering the

acetylation proper less violent and thus more controllable. The risk. of

degradation of the cellulose structure is reduced, and the probability of

uniform acetylation is increased. The varying reactivities after pre-treatment

of cellulose from various sources has been ascribed to moisture content

variation and to what is styled "moisture history" and in particular the lowest

water content to which it has been subjected.

The process based on this concept aims at overcoming differences due to such

effects, by bringing the water content of cellulose up to 15%, then reducing it

in a controlled manner so that at no time does it fall below 4-5%, nor is there

sufficient water present to form a continuous phase. This work tends to show

that moisture is not a fortuitous circumstance, but that it is likely to constitute

a significant part of the molecular structure of cellulose.

1.5.2 Acetylation in solvent media

As already mentioned, the methods available for the acetylation of cellulose

may be classified, according to the diluent used, as solvent or non-solvent

methods. In the former class the cellulose ester passes into solution as it is

formed, so that the reaction is completed in the presence of the liquid phase

only. In non-solvent methods, solid and liquid phases persist throughout the

reaction since the acetate produced is insoluble in the reaction medium.

22


Methods of acetylation, whether in homogeneous or heterogeneous media,

have much in common. Under the usual conditions of commercial operation,

acetic anhydride is almost universally employed as the acetylating principle.

Another acetylating agent which has been suggested is acetyl chloride. The

actual mechanism of the acetylation process has been a fruitful subject for

investigation and speculation, and even today is not completely understood.

1.5.2.1 Catalysts

It was realised at an early stage that acetylation by means of mixtures of acetic

anhydride and acetic acid could only be effected economically when a catalyst

was used. The catalyst was long regarded as the key to efficient production,

and for many years received considerable attention. In spite of the vast

amount of research carried out, concentrated sulphuric acid is almost

universally employed today.

1.5.2.2 Diluents

The function of the diluent is mainly to facilitate and render more

homogeneous the acetylation of the cellulose. The diluent is usually a non

solvent for cellulose, although it should be able to cause at least mild swelling

to facilitate the penetration of the acetylation mixture. It mayor may not be a

solvent for the resulting primary cellulose acetate, although the use of a

solvent diluent is the more usual. The presence of the diluent makes the

temperature of the acetylation more controllable; this is especially the case in

non-solvent methods where larger quantities of diluent are introduced. In

modem homogeneous methods, glacial acetic acid is universally used. This is

desirable from the viewpoint of recovery of the spent acid, if for no other

reason, since in the course of the acetylation acetic anhydride is converted to

acetic acid, so that ultimate recovery is much simpler when the diluent is also

acetic acid.

23


1.5.2.3 Acetylation processes

On the industrial scale 280 - 300Kg of acetic anhydride are required to

acetylate 100Kg of cotton. The amount of glacial acetic acid required is

naturally in excess of the anhydride used, and synthesis of 1,000Kg of

acetone-soluble cellulose acetate require 15,000Kg of acetic acid. The amount

of catalyst used varies widely according to the nature of the catalyst, the type

of cellulosic material used, and the nature and conditions of acetylation. In all

cases, but especially when concentrated sulphuric acid is used, the catalyst

content is maintained as low as possible; it rarely exceeds 10% on the weight

of cellulose and is usually much lower.

1.5.3 Saponification or "ripening" of cellulose acetate

Whilst it is more usual to think of ripening in connection with the solvent or

homogeneous method of acetylation, it may also be carried out in

heterogeneous systems; but the commercial saponification of fibrous cellulose

acetate has so far met with limited success. The ripening medium, which is

responsible for the change in solubility, is usually acetic acid either alone or in

the presence of water, or mineral acids with or without neutral salts. It has

been established that the nature of the final product will vary considerably

according to the amount of water added to the primary solution and is largely

governed by the duration and temperature of the ripening. A typical method of

ripening would be a follows: Water equal to 22.5% calculated on the weight of

cellulose used, is added to the primary acetylation mixture. The batch is

mixed well and after cooling to 21 QC, is poured into the ripening vessel and

transferred to a well ventilated ripening room maintained at 21 QC. The time of

ripening will vary from 65 to 75 hours, so that it is necessary to control the

progress of ripening in each individual batch after about 54 hours, by

examination of samples withdrawn from the ripening pan at frequent intervals . . -

A small quantity of the viscous solution is removed; the cellulose acetate is

24


precipitated by judicious addition of water, washed thoroughly, and dried

rapidly in an electric or vacuum oven. As ripening proceeds it will be found

that the product becomes progressively less soluble in hot anhydrous

chloroform until, when true acetone solubility is reached, the trial sample

yields a hard plastic mass in warm chloroform. The behaviour of the sample

in alcohol-benzene (50:50 v/v) gives a definite indication of the progress of the

ripening. About Sm1s of such a mixture are added to 2g of the sample of

cellulose acetate and the whole warmed in a water bath. Water is then added

drop wise from a burette, with constant stirring, and the number of drops

required to effect perfect solution is noted. This number decreases as ripening

proceeds, and when it fans to two or three, it is an indication that the acetone

soluble stage has been reached.

1.5.4 Precipitation of cellulose acetate

The precipitation of cellulose acetate was for many years the one stage in the

process which should be classed as an "art", and which depended for its

success almost entirely on the manipulative skill and judgement of the foreman

in charge of the process.

There is no intrinsic difficulty in precipitating cellulose acetate; but to separate

it in a form neither too lumpy nor too powdery so that it can be readily

purified without due loss, is a matter caning for expert judgement and long

expenence. Precipitation applies, of course, primarily in those processes

where the cotton has been acetylated in the presence of a non-volatile solvent

diluent such as glacial acetic acid. In those methods of homogeneous

acetylation where a volatile diluent only (such as methylene chloride or liquid

sulphur dioxide) is used, precipitation can be effected by evaporation of the

liquid diluent. On the process scale the operation is carried out as follows:

25


Individual batches of material, certified by the laboratory as having attained

the desired degree of acetone-solubility, are transferred to the precipitators.

These are usually cylindrical vessels, with a sluice outlet at the bottom and

provided with variable-speed paddles or beaters with overhead or, more

usually, base drive. The precipitators are usually located above the centrifuges

to minimise subsequent handling of the acetate. When the precipitator is

charged, the stirrers are set in motion at slow speed and water is added

gradually, with or without previous addition of sodium acetate or carbonate

sufficient to neutralise the sulphuric acid catalyst present in the mixture, care

being taken that at no period is the addition of water sufficient to cause more

than incipient separation of the acetate. As the precipitation proceeds, the rate

at which water is run in will have to be reduced if the formation of lumps is to

be avoided and a point is eventually reached where the solution assumes an

opalescent appearance with small white flakes floating in the mass. At this

point the rate of addition of water is increased and the rate of stirring is

. increased to a maximum of 300-35Orpm. When the required excess of water

has been added, the mass is allowed to stand for 15-20 minutes for the fibres

to harden, and the weak acid is then drained off to storage vats while the solid

cellulose acetate is deposited into the centrifuges below. Excess liquor is

separated by "whizzing" and indeed the initial stage of washing is also carried

out with the basket of the centrifuge in motion, this being continued until the

effluent has an acid content of 7-10%. The opaque fibrous secondary acetate

is then transferred to the washers for the Imal purification.

1.5.5 Final after-treatment of cellulose acetate

Cellulose acetate as it comes from the precipitator contains a smalI quantity of

free acid and also of mineral acid catalyst. Both have to be removed in the

fmal after-treatment. The removal of the former is a simple matter of water

washing, but the last residues of sulphuric acid are more. difficult to dislodge.

26


Some years ago no attempt was made to remove this, and it was not

unconunon to find that the sulphur content of material was as high as 0.1-0.2%

on the weight of the dry cellulose acetate. The presence of this combined

sulphur naturally rendered the product less stable, so that on prolonged storage

progressively increasing quantities of acetic acid were liberated. This

instability accounted for many early failures both in rayon and plastics.

The substantially complete elimination of sulphur is a matter of vital

importance in the production of cellulose acetate for rayon or plastics, and this

is now accomplished by means of a definite stabilisation treatment following

the washing. The washing and stabilisation is carried out in large wooden vats

provided with stirring gear. The desired quantity of cellulose acetate, from

which the acetic acid has been substantially removed in the centrifuge is

charged into the washer and washed with several changes of water, the initial

washings being moderately hot (circa 50-70°C), until the acidity does not

exceed 0.01%. When this has been accomplished, the actual stabilisation

commences. The water in the vat is raised to boiling point using live steam,

and sufficient sulphuric acid is added to bring the content of the bath up to

about 0.02%. This addition requires careful analytical control, since the

amount of acid to be added will depend on the initial acidity of individual

batches. Boiling is then continued with stirring until the acidity of the bath

reaches a constant maximum, usually after 1-2 hours, and here again careful

laboratory control is necessary. When this point is reached the vat is flooded

with cold water. The cellulose acetate is finally washed until free from acid,

after which it is transferred in convenient amounts to the drying plant. An

ordinary tray-dryer may be used, or hot air may be forced through the

cellulose acetate in suitable containers. The drying temperature should not

exceed 100°C. Drying is continued until the moisture content reaches the

desired level, which for most commercial purposes is usually between 2-3%.

27


1.5.6 Acetylation in non-solvent media: Preparation of fibrous CA

Non-solvents for cellulose acetate such as benzene, toluene, xylene or carbon

tetrachloride have been used to replace the usual acetic acid diluent, and it

was claimed that not only was the process more economical, but the

temperature of acetylation was more easily controlled. In spite of these very

definite advantages, however, it was found that the fibrous acetates, which

. were of course the primary products approximating to the triacetate, contained

a considerable amount of combined sulphur and were consequently of low

stability.

1.6 Industrial uses of cellulose acetate17

Cellulose acetate has varying uses. Firstly, the use that is of primary

importance in this work is the filter tow used in the tobacco industry. Other

uses include cellulose acetate yarn for the textile industry, biological filters,

photographic films, transparent and pigmented sheeting and plastic

compositions such as those used for compressIOn, extrusion, injection

moulding and to a lesser extent surface coatings.

1. 7 Breakdown of cellulose acetate

There have been conflicting publications about the biodegradation of cellulose

acetate. Many workers reported that cellulose acetate was not attacked by

fungi or bacteria, others claimed that it was. Much of the confusion may be

due to cellulose acetates of differing DP values being used. The first concrete

piece of evidence came from Courtaulds l8 Their work was confined to

studying the effects of soil burial and it fell into two parts. Firstly, whole

cigarette filter rods were buried and examined at three month intervals. After

six months there was very marked erosion of the surface of the fibres and after

9 months the filters had completely disintegrated. Although the pattern of the

erosion appeared to be biological, organisms found in the decayed filters

28


would not degrade fresh cellulose acetate and eventually the cultures died out.

The second part consisted of burying acetate yarns with and without triacetin

plasticiser and smoke condensate. A control sample was buried in sterile soil.

Loss of strength of the yarn was used as a measure of degradation as well as

visual appraisal. Again significant degradation occurred in a few weeks and

differences between normal and sterile soil indicated that the action was

largely biological. Similar experiments were undertaken by Rhodia in

Germanyl9. Their studies focused on the marine environment and again

degradation was observed. However, neither can offer as yet any mechanism

for the degradation.

Two methods of attack are possible. Firstly, the esterases of micro-organisms

could attack the ester groups. The other method of decomposition would be

cellulases attacking the glucosidic bonds on the backbone, hence breaking

down the chain itself.

Most of the work done to date has focused on cellulose acetate membranes.

Most workers report that when operational cellulose acetate reverse-osmosis

membranes were examined for evidence of biological degradation, numerous

fungi and bacteria were isolated both directly and by enrichment techniques.

When tested, most of the fungi were active cellulose degraders but none of the

bacteria were. Neither fungi nor bacteria were able to degrade cellulose

acetate membranes in vitro20, although many fungi were able to degrade

cellulose acetate membranes after they had been de-acetylated.

Organisms did not significantly degrade powdered cellulose acetate in pure or

mixed cultures as measured by reduction in acetyl content or intrinsic viscosity .

or production of reducing sugars. Organisms did not affect the performance of

cellulose triacetate fibres when incubated in them. The inability of the

organisms to degrade cellulose acetate was attributed to the high degree of

29


acetate substitution of the cellulose polymer. Microbial degradation of

operational cellulose acetate reverse osmosis membranes was thus unlikely.

On the other hand, there have been authors who have reported that cellulose

acetate membranes and textiles are biodegradable21.

22 Buchanan and his co

workers used two separate assay systems to evaluate the biodegradation

potential of cellulose acetate: an in vitro enrichment cultivation technique

(closed batch system) and a system in which cellulose diacetate films were

suspended in a wastewater treatment system (continuous feed system). The in

vitro assay employed a stable enrichment culture, which was initiated by

inoculating a basal salts medium containing cellulose acetate with 5% (v/v)

activated sludge. Microscopic examination revealed extensive degradation of

the cellulose acetate fibres with a degree of substitution of 2.5 after 2-3 weeks

of incubation. Subsequent characterisation of these fibres demonstrated a

lower average degree of substitution and a change in the molecular weight

profiles. In vitro enrichments with cellulose acetate with a degree of

substitution of 1.7 were able to degrade more than 80% of the films in 4-5

days. Films with a degree of substitution of 2.5 required 10-12 days for

extensive degradation. Films prepared from cellulose triacetate remained

essentially unchanged after 28 days in the in vitro assay. The wastewater

treatment assay was less active than the in vitro enrichment system, but the

same trends were observed. The authors also claimed that in the above

mentioned experiments the degree of substitution was lowered by an average

of 0.55, whilst the degree of substitution of the control sample was not

significantly affected. Comparison of the GPC data of the inoculated samples

with that of the control and starting material showed that Mz of the inoculated

samples was decreased while the Mn increased, which was reflected by the

narrowing of the polydispersities (M,viMn, MiMn). Hence this data suggested

that both deacetylation of the cellulose acetate and random cleavage of the

cellulose acetate to a smaller chain size were being observed. Finally

30


supporting evidence for the biodegradation potential of cellulose acetate was

obtained through the conversion of cellulose [1-14C]-acetate to 14C02 in the in

vitro assay.

Northrop and his co-workers22 also came to the conclusion that cellulose

acetate textiles were biodegradable in anaerobic conditions. Thin strips of

textile were buried in plastic garden pots in garden top soil (pH 7.5). The pots

were placed in a laboratory where the ambient temperature was always

between 20 and 28°C. The pots were watered once a week. The garden pots

had holes at the bottom for drainage so that the soil did not become

waterlogged. According to the authors, evidence of deterioration was

apparent after two months of burial and all of the cellulose acetate samples

were completely destroyed within four to nine months. The solubility

behaviour of the degraded cellulose acetate fibres was found to have changed

significantly. Whereas before burial the cellulose acetate samples were

completely soluble in formic acid, glacial acetic acid, acetonitrile and

hexafluoroisopropanol, after exhumation the buried samples were only

partially soluble in each of these solvents. In order to characterise the

insoluble residue, a portion of degraded cellulose acetate from a sample of

textile that had been buried for seven months was treated with glacial acetic

acid to remove the acetic acid-soluble fraction and an infrared spectrum was

recorded. The residue was found to have a significantly decreased carbonyl

stretching band at approximately 1750cm-1, as well as a decreased methyl

symmetric stretching band at approximately I350cm-1. These results were

consistent with the degraded cellulose acetate having lost acetate moieties due

to hydrolysis. These workers put forward a different degradation scheme from

Buchanan et apl. They claimed that cellulose acetate was fully hydrolysed to

cellulose which was then degraded by cellulase to oligosaccharides and

glucose.

31


Other workers claimed that cellulose acetate had to undergo some chemical

modification before biodegradation became viable23-25 Penn and co

workers23-24 suggested cellulose acetate graft copolymers (addition and

condensation type grafting), whereas Ach25 proposed the addition of specific

low molecular weight and oligomeric compounds to the cellulose acetate

chain, which, when taken alone, biodegraded rapidly. These additives were

also non toxic. In effect, therefore, he used these additives as an extra

incentive for various micro-organisms to attack cellulose acetate.

32


1.8 A review of some esterases and cellulases

1.8.1 Esterases

Esterases by and large have a broad specificity26 and this has created

difficulties in classification and nomenclature. Many of the enzymes

represented by a single entry in sub-group 3.1.1 (carboxylic ester hydrolases)

are groups of enzymes with closely related specificities; some have been

shown to exist in multiple forms in a single tissue with slight differences in

specificity, and there are often differences in specificity between the

corresponding enzymes from different species. There is also an overlap in

specificity with some enzymes listed in other groups.

The main factors influencing the specificity of simple esterases and lipases are

the lengths and shapes of the hydrophobic groups on either side of the ester

link.

1.8.2 Cellulases

As far as the cellulases are concerned, they are a very complex system of

enzymes and there are three main types of enzyme found in cellulase systems

that can degrade crystalline cellulose3o.31 :exo-cellobiohydrolase (EC 3.2.1.91),

endo-l,4-j3-D-glucanase (EC 3.2.1.6) and j3-D-glucosidase or cellobiase (EC

3.2.1.21). Exo-cellobiohydrolases are found as major components in some

cellulase systems, but are absent from most. All enzymes appear to exist in

multiple forms which differ in their relative activities on a variety of

substrates.

Endo-l,4-j3-glucanases: They hydrolyse cellulose chains at random to

produce a rapid change in the degree of polymerisation. Substrates include

carboxymethylcellulose and H3P04 or alkali-swollen (l!ID0rphous) cellulose.

33


Crystalline cellulose such as cotton fibre or A vicel is not attacked to any

significant extent. Hydrolysis of amorphous cellulose yields a mixture of

glucose, cellobiose, and soluble cello-oligosaccharides. The rate of hydrolysis

of the longer chain cello-oligosaccharides is high, and the rate increases with

the degree of polymerisation: glucose and cellobiose are the principal products

of the reaction. Some endo-I,4-j3-glucanases act in synergism with the

cellobiohydrolase isolated from fungal cellulases to solubilise crystalline

cellulose; some however, do not.

Exo-l,4-j3-glucaoases. Exo-I,4-j3-glucanases of the fungi act by removmg

glucose or cellobiose from the non-reducing end of the chain.

Cellobiohydrolase is the most common enzyme. Most cellobiohydrolases

appear to release small amounts of glucose from cellulose. Cotton fibre is not

attacked to a significant extent, but H3P04-swollen cellulose is hydrolysed

with a characteristic slow fall ill the degree of polymerisation.

Carboxymethylcellulose and cellobiose are not substrates but cellobiose and

longer chain cello-oligosaccharides are hydrolysed with the rate. increasing

with an increasing degree of polymerisation. A vicel is a substrate that has

proved to be very useful for isolating and measuring cellobiohydrolase.

13-D-glucosidases. These are not strictly speaking cellulases but they are,

nevertheless, very important components of the cellulase system. They

complete the hydrolysis of the short-chain cello-oligosaccharides and·

cellobiose which are released by the other enzymes to glucose. j3-D

glucosidases hydrolyse cellooligosaccharides at a rate that decreases with an

increasing degree of polymerisation, but cellulose is not attacked. Other

characteristics are that they are not specific for the 1,4-j3-linkage, and they

possess transferase activity that acts on glucose units to form other sugar

molecules such as trimers, and higher oligosaccharides.

34

CHAPTER 2

MATERIALS AND METHODS

Chapter 2 - Materials and Methods

2.1 STUDIES ON DISCARDED CIGARETTE FILTERS

Cigarette filters that were lying on the car park were collected as the first step

in identifying possible micro-organisms growing on them. The ones collected

had the appearance of having gone through various "wet-dry", (rain-sunshine)

cycles.

Various tests were performed on those filters as described below.

2.1.1 BACTERIOLOGICAL ACTIVITY OF THE FILTERS

PREPARATION OF THE BACTERIOLOGICAL MEDIA

Nutrient Agar Medium

"Oxoid" nutrient agar powder (28g) was suspended in 1 litre of distilled water.

The solution was brought to the boil to assist dissolution and the pH was

adjusted to 7.0 ± 0.2. The solution was then sterilised by autoclaving at 12I a C

for 15 minutes. The medium was poured onto petri dishes, allowing 20mls for

every petri dish, and allowed to solidify. Once set, the petri dishes were dried

in an inverted position at 37aC for 30 minutes.

Nutrient Broth Medium

"Oxoid" nutrient broth powder (13g) was added to I litre of distilled water.

The solution was mixed well and the pH adjusted to 7.4 ± 0.2. The solution

was then sterilised by autoclaving at 12Ia C for 15 minutes.

Sabouraud Maltose Agar Medium

"Oxoid" Sabouraud maltose agar powder (65g) was suspended in 1 litre of

distilled water. The solution was boiled and the pH adjusted to 5.6 ± 0.2. The

medium was then sterilised by autoclaving at 121 ac for 15 minutes.

35


Tributyrin Agar Medium

This medium was prepared by the addition of 1% of tributyrin (glyceryl

tributyrate) to nutrient agar and autoclaved at 121°C for IS minutes. The pH

was then adjusted to 7.S ± 0.2.

Bacillus Cereus Selective Agar

"Oxoid" medium powder (20.Sg) was suspended in 47Smls of distilled water

and brought gently to the boil to dissolve completely. The pH was adjusted to

7.2 ± 0.2 and the solution was sterilised by autoclaving at 121°C for IS

minutes. Despite the name, this medium also detects other Bacillus spp.

Pseudomonas Selective Medium

The agar base was prepared by suspending 24.2g of"Oxoid" agar base powder

in SOOmls of distilled water. Glycerol (Smls) was added and the solution was

brought to the boil to dissolve completely. The pH was adjusted to 7.1 ± 0.2

and the medium was sterilised by autoclaving at 121°C for IS minutes. The

medium was then allowed to cool to SO°C and the Pseudomonas C-F-C Agar

supplement was added. The supplement consists of the contents of 1 vial of

"Oxoid" Pseudomonas C-F-C supplement SR 103 rehydrated with 2mls of

sterile distilled water. This was added to SOOmls of agar base cooled to SO°c.

Supplement SR 103 is recommended for the selective isolation of

Pseudomonas spp.

Czapek Dox Agar

"Oxoid" powder (4S.4g) was suspended in I litre of distilled water. The

solution was brought to the boil to dissolve completely. The pH was adjusted'

to 6.8 ± 0.2. The medium was then sterilised by autoclaving at IISoC for 20

minutes.

36


METHODOLOGY

Discarded fllters were laid on plates containing nutrient agar medium,

Sabouraud maltose agar medium, tributyrin medium, Bacillus Cereus medium

and Pseudomonas selective medium. The plates were incubated at 37°C for

up to 3 days depending on the medium and the various organisms were

identified. (See also Section 3.1.1).

In order to aid identification, micro-organisms were examined as follows:

A clean grease free slide was removed from a jar containing alcohol using

forceps and the film was burned off in alcohol. A small drop of distilled

water was placed on the slide. A nichrome wire was flamed and used to

remove a small quantity of growth from the culture. This was emulsified in

the drop of distilled water.

The moist film was dried by holding above a small bunsen flame. The film

dried quickly and was just visible when dry. The smear was passed quickly

through the flame two or three times when dry and was then ready for

staining.

Staining methods.

The Gram Stain

This was a very important diagnostic technique and a fresh culture was always

used as old cultures may give false negative results.

The smear was stained for one minute with Hucker's ammoruum oxalate

crystal-violet.

The slide was washed rapidly with tap water and blotted dry.

The smear was covered with Gram's iodine for one minute.

The iodine was washed off rapidly with tap water and the slide blotted dry.

The slide was decolourised with 95% ethanol for 30 seconds and was blotted

dry.

The smear was then counterstained with Safranine for 10-15 seconds.

37


Finally, the smear was examined under an oil immersion lens (x 1000): Gram

positive organisms were stained blue and Gram negative organisms were

stained red.

The Spore Stain

A smear was prepared from an old (96 hour) nutrient agar culture. The smear

was covered with a 13mm A.A. disc.

The disc was saturated with 5% aqueous malachite green.

The slide was placed across the top of a flask of boiling water and left for 5

minutes. The water had to be kept boiling and the disc saturated with the

stain.

The A.A. disc was removed and washed rapidly with distilled water and was

blotted dry.

The smear was counterstained with safranine and was examined under an oil

immersion lens. The spores were stained green and the vegetative cells were

stained red.

Oxidase test

Small pieces of filter paper were soaked in freshly prepared 1 % aqueous tetra

methyl-p-phenylene diarnine dihydrochloride. Some filter papers might give a

blue colour and these are not to be used. The papers might be dried or used

wet. A small portion of a fresh young culture was scraped with a clean

platinum wire or a glass rod and was rubbed on the filter paper. A blue colour

within 10 seconds was a positive oxidase test. Old cultures gave unreliable

results.

Microbial penetration test

In order to test whether the micro-organisms penetrated the filters, the middle

part of 10 filters was cut out asceptically and laid on plate~ containing Nutrient

38


Agar Medium as well as in flasks containing Nutrient Broth Medium. These

plates and flasks were incubated at 37°C for 2 days.

2.1.2 STUDIES ON THE FIBRES OF THE DISCARDED FILTERS

Electron miscoscopy was used to study the fibres of the discarded filters in

detail and see whether any signs of biodegradation were visible on individual

fibres. Virgin smoked and unsmoked filters, as well as plasticised and

unplasticised ones were used to compare the results obtained from the

discarded filters. (See also Section 3.1.2).

2.2 LONG TERM BIODEGRADATION STUDIES ON VIRGIN

FILTERS

In order to monitor the possible biodegradation of cigarette filters from the

point that they were discarded, it was necessary to devise some long term

biodegradation studies taking into account all possible factors that could

enhance CA biodegradation, and hence create an ideal environment for

biodegradation to occur.

The various factors that were taken into consideration were the following:

• Possible surfaces on which cigarette filters might be discarded.

• Temperature.

• Humidity.

• Light conditions.

• Whether filters were plasticised or not.

• Whether filters were smoked or not.

• Whether filters were complete, sliced open or whether they were crushed.

Three types of surface were used. Potting compost bought from a gardening

centre, sand and roofmg tiles. Seed trays were used in order to contain the

above materials. Every tray consisted of six rows, each containing five filters.

39


The first three rows consisted of virgin ftIters that were complete (first row),

opened (second row) and crushed (third row) and the other three rows

contained smoked flIters arranged as before.

These experiments were repeated for all three surfaces, for plasticised and

unplasticised filters, in the light and in the dark, as well as at three

temperatures (room temperature, 30°C and 37°C).

All filters were inoculated with a soil suspension of 1 ml either on the surface,

or injected in the middle of each filter.

The humidity was kept high on all experiments and the tipping paper was

removed from all filters before use.

The sand and roofmg tiles did not retain moisture and the experiments

conducted on those two surfaces were terminated after 5 months. The potting

compost proved to be the best surface as it retained moisture adequately, and

those experiments were continued for up to 18 months.

Another set of seed tray experiments was set up using whole filter rods and

fIlter tow instead of filters. In these experiments, half of the fIlter rods and

half of the filter tow were buried in the potting compost/soil mixture and the

other half was left lying on the surface of the soil. This way the behaviour of

buried and unburied cellulose acetate in various forms could be studied

simultaneously under identical conditions. (See also Section 3.2).

40


2.3 STUDY OF ADDITIVES ON CA BIODEGRADATION

Simple sugars and aminoacids were considered as possible means of

enhancing the biodegradability of CA.

The additives used were the following (see also figure 2.1, overleaf):

• glucose

• xylose

• glycine

• senne

• glycerol

• starch •

'Starch exists in two fonns: amylose (the unbranched type of starch), and

amylopectin (the branched fonn). Amylose consists of glucose residues in a-

1,4 linkage. Amylopectin has about one a-I,6 linkage per thirty a-I,4

linkages (see also figure 2.2).

Virgin cigarette filters were laid on potting compost trays enriched with an

equal amount of garden soil. The filters were injected with 1 ml of the various

additives and their progress was monitored over 30 days. (See also Section

3.3).

41

Figure 2.1

Structures of the additives.

CHO I

HCOH I

HOCH I

HCOH I

HCOH I CH20H

D-glucose

H I

CHO I

HCOH I

HOCH I

HCOH I CH20H

D-xylose

HO- CH2-C-CH2- OH

I OH

Glycerol


COOH I

H2N -C-H I H

Glycine

COOH I

H2N- C- H I

H- C -OH I H

Serine

42


Figure 2.2

Structure of amylose and amylopectin.

HOCH2 HOCH2

o

-0 o 0-

OH OH

Amylose

-0 o

OH

-0 o 0-

OH

Amylopectin

43


2.4 PREPARATION OF CA WITH GIVEN DEGREE OF

SUBSTITUTION (DS)

In order to assess accurately and quantitatively the extent of biodegradation of

the CAs in the long tenn experiments mentioned above, it was important to

produce a set of chemically hydrolysed CAs with given DS. These had to be

fully characterised, in order to be able to compare them with the CA obtained

from the biodegradation experiments.

The starting material was the CA used in the filter manufacture, i.e. with a DS

of2.5 (2.5 acetyl groups per glucose unit) and a degree of polymerisation (DP)

ofabout 300.

CA with a DS of 2.5 (60g) was dissolved in 1200mls of glacial acetic acid in a

2 litre B 34 "quickfit" conical flask fitted with a magnetic stirrer and a

condenser. Concentrated sulphuric acid (24g) was added to the mixture

followed by 132mls of distilled water. The water was added slowly so as to

avoid precipitation. The temperature was then raised to 80°C for the required

amount of time.

Finally, 650.4g of a 21% aqueous solution of magnesium acetate were added

to the reaction as to neutralise the sulphuric acid. The solution was then

filtered so as to remove the precipitate of magnesium sulphate and the product

was precipitated either from water or from isopropanol depending on the final

DS (see also Table 3.3, Chapter 3). The product was then washed free from

glacial acetic acid using an appropriate solvent and dried in an electric oven at

55°C overnight. The product was fmally ground to a fme powder and then

fully characterised. (See also Sections 3.4 and 3.6).

44


2.5 DETERMINATION OF THE DS

2.5.1 CHEMICAL DETERMINATION

Approximately Ig of the dry CA sample was weighed accurately in a weighing

bottle. The sample was then transferred to a 250ml Erlenmeyer flask, and the

bottle was reweighed to detennine the exact sample weight. Ethanol (40mls of

75%) was then added to each sample, and reagent blanks were set up and

carried through the rest of the procedure.

The flasks, loosely stoppered, were heated for 30 minutes at 60°C. Then,

40mls of 0.5N sodium hydroxide solution were accurately measured with a

Jencons "Digitrate" digital dispenser and were added to each of the flasks

which were then heated again at 60°C for 15 minutes. The flasks were then

stoppered tightly and allowed to stand at room temperature (below 30°C) for

72 hours.

The excess alkali was then titrated with 0.5N hydrochloric acid usmg

phenolphthalein as indicator. An excess of Iml of acid was added, and the

alkali was allowed to diffuse from the regenerated cellulose overnight. The

disappearance of the pink colour indicated the complete neutralisation of the

alkali. The small excess of acid was then back titrated with 0.5N sodium

hydroxide to the phenolphthalein end-point. After the solution had acquired a

faint pink colour, the flask was stoppered and shaken vigorously. The colour

might fade because of acid diffusing from the cellulose. The addition of alkali

and the shaking were continued until the faint pink end-point persisted.

45


The DS was then calculated as follows:

% Acetyl = [(A - B)Nb - (C - D)N.J x 4.3 / W

and n = (3.86 x % acetyl) / (102.4 - % acetyl)

where:

A = mls of sodium hydroxide added to the sample

B = mls of sodium hydroxide added to the blank

Nb = normality of the sodium hydroxide solution

C = mls of hydrochloric acid added to the sample

D = mls of hydrochloric acid added to the blank

N. = normality of the hydrochloric acid solution

W = weight of the sample in grams

4.3 = factor to calculate the % acetyl

Equation 2.1

Equation 2.2

n = DS = average number of acetyl groups per anhydro-D-glucose unit of

cellulose

46


2.5.2 SPECTROSCOPIC DETERMINATION

2.5.2.1 FT-IR SPECTROSCOPY

FT-IR spectroscopy was also used for the detennination of the OS. The

instrument used was the Nicolet 200XC FT-IR spectrometer with the Omnic

software. The samples were run as thin films on NaCl discs. The films were

prepared as follows:

Dilute solutions of the CA samples in N,N-Dimethylacetarnide (DMA) were

prepared (2%). A few drops of the solution were placed on a clean, polished

NaCI disc and the disc was placed in a vacuum desiccator. The disc remained

under vacuum overnight. The disc was then removed from the desiccator and

placed in an electric oven (65°C) for a further 24 hours to dispel any

remaining moisture from the film. A transmission spectrum of the sample was

then obtained. Two peaks were of interest: the carbonyl peak at

approximately 1750 cm- l and the OH peak at approximately 3460 cm-I. The

peak area of the carbonyl peak was then recorded as well as the peak height

for the OH peak in absorbance units. For the CA samples whose OS was

already determined by the chemical method, a graph of the ratio of the height

of the OH peak over the area of the carbonyl versus the DS could be plotted.

A calibration curve was thus obtained, from which the DS of any CA could be

calculated once its spectrum was available (see also Graph 1, Chapter 3).

47

2.5.2.2 NMR SPECTROSCOPY

2.5.2.2.1 IH NMR


The samples (20mgs) were dissolved in 0.8ml deuterated DMSO in a 5mm

diameter NMR tube. After dissolution, they were run at 25°C in an SRC WH-

400 instrument.

2.5.2.2.2 SOLID STATE I3C NMR

The samples were ground to a fme powder and were placed on a 7mm

diameter rotor and were spun at the "magic angle" of 54.7°. The instrument

used was a Varian Unity Plus, 300MHz, with a Doty Scientific probe.

48


2.6 THE BIODEGRADATION OF CA OF GIVEN DS USING

ASPERGILLUS FUMIGATUS

2.6.1 THE ISOLATION OF THE FUNGUS ASPERGILLUS

FUMIGATUS

Nutrient agar plates containing 1% of CA with a OS of 1.7 were inoculated

with garden soil and they were allowed to grow for 48 hours. Colonies were

subcultured on plates containing the Czapek Oox medium which is selective

for fungi and were allowed to grow for 3 days at 37°C. The process of

subculturing and purifying the fungus colonies was repeated three times. A

strain of the purified fungus was then sent for identification to the

International Mycological Institute at Kew, where it was identified as

Aspergillus jumigafus, a common soil species.

2.6.2 THE INOCULATION OF THE CA-CONTAINING

MEDIUM WITH THE ASPERGILLUS FUMIGA TUS

Medium composition

KH2P04 O.ISg

MgS04 0.20g

CaC03 0.02Sg

Yeast extract 0.02Sg

(NH4hS04 1.00g

Tap water I litre

The pH of the medium was adjusted to pH 7.0 and then sterilised by

autoc1aving at 121°C for IS minutes. Magnesium sulphate solution was.

autoc1aved separately in order to avoid precipitation of magnesium phosphate

in the medium solution. Magnesium sulphate was then added to the medium

when cool to give the concentration shown above.

49


CA (1%) was added to the medium as the sole carbon source. The CA was

added to the medium asceptically after autoclaving in order to avoid any

thermal degradation of the CA.

The medium was then inoculated with Aspergillus filmigatus and was

incubated at 30°C in a rotary incubator for up to ten days. At the end of every

incubation period, the CA was isolated from the medium and was fully

characterised. (See also Sections 3.5 and 3.6).

50


2.7 THE CHARACTERISATION OF THE CAs WITH GIVEN DS

IN THEIR ORIGINAL AND BIODEGRADED FORM

The CAs prepared above were fully characterised. Tests included molecular

weight determination using Gel Permeation Chromatography (GPC) (M n),

viscometry (M,), as well as spectrophotometric techniques i.e. Fourier

transform infra-red (FT-IR) and proton and carbon nuclear magnetic resonance

eH NMR and I3C NMR). X-Ray Diffraction was also performed in order to

test the crystallinity of the samples. The DS of every sample was also checked

again after a period of 6 months in order to test whether any change had taken

place. The DS values were found to be constant at the end of the above

mentioned period.

2.7.1 MOLECULAR WEIGHT DETERMINATION

2.7.1.1 GPC METHOD

Two sets of GPC results were obtained. The first set only applied to the

starting material and was run on a tetrahydrofuran (THF) system at

Loughborough. The system was calibrated with 3 standards, each containing 3

polystyrene polymers of known molecular weight. A small amount of toluene

was added to each standard as a reference point. The distance of each peak

from the injection point was recorded, as well as the % equivalent of toluene

(distance of peak / distance of toluene). The calibration curve was a plot of

log Mp vs. % toluene (see also Table 3.5, Chapter 3).

Initially, 4mgs of sample were dissolved in 4mls of THF (HPLC grade,

unstabilised) and the solution was left for 5 hours to equilibrate. A very small

amount of toluene was added to the solution. The solution was filtered prior

to use and Iml was injected in the instrument. The elution was recorded with

the aid of a chart recorder. The flow was Imllmin.

51


The sample peak was then sectioned in 2mm stripes and the distance of each

stripe from the injection point, as well as its height in mm were recorded. The

MW was then worked out with the aid of the CALl computer programme (see

also Table 3.6, Chapter 3).

The second set of results was obtained at Polymer Laboratories, Church

Stretton, Shropshire. The system used was as follows:

Columns: 3 x PLgel lO).Im MIXED-B 300x7.5mm

Eluent: Dimethylacetamide with 0.5% w/v LiCl

Flow rate: Irnl/min

Temperature: 60°C

Detector: Differential Refractometer (PL GPC 110)

Data handling: PL Caliber GPC software

The samples were prepared as I - 2 mg/m! solutions in an aliquot of the eluent.

Gentle heat and stirring were used to aid dissolution. All solutions were

filtered over a 0.45J..UD membrane prior to injection. An injection volume of

100).11 was used and samples were analysed in duplicate.

The column set was calibrated using both narrow polydispersity polystyrene

and polyethylene oxide/glycol standards. Each sample was evaluated against

both calibrations (see also Tables 3.9 aild 3.10, Chapter 3).

52


2.7.1.2 SOLUTION VISCOSITY METHOD

The viscometer used was the Schott-Geriite AVS 310 model (Camlab UK)

with two capillary viscometer tubes - Schott-Geriite Type 531 0 I (0.53mm

capillary) and Type 531 13 (0.84mrn capillary). The former will be referred to

as the narrow tube and the latter as the wide tube.

[n viscometry, the following equations are used:

llr = t / to

where t = run time of the solution, and

to = run time of the solvent

llsp = llr - I = (t - to) / to

llsp / c = [ll] + k, [llfc Huggins' equation

where [ll] = the limiting viscosity

Equation 2.3

Equation 2.4

Equation 2.5

By plotting llsp / c against c, [ll] becomes the intercept and k, [ll]2 the slope.

Another useful equation is the Kraemer equation

In llr / c = [ll]- k2 [llfc

where k, + k2 = 0.5

Equation 2.6

Equation 2. 7

The Schultz - Blaschke equation is also valid: llsp / c = [ll] + k [ll] llsp.

Equation 2.8

When llsp / c is plotted vs. llsp, [ll] becomes the intercept and k [ll] becomes

the slope.

The following should also hold: 1.2::: llr::: 2, or 0.2::: llsp:::1. Equation 2.9

Finally, the Mark-Houwink-Sakurada (MHS) equation gives a molecular

weight expression: [ll] = KM" Equation 2.10

where K and a are constants specific to every solvent at a given temperature.

The first viscometry measurements were made in acetone (for DS2.5) and

70:30 acetone:water (for DS!.7 and 1.5) at 25°C. The temperature could be . monitored accurately as the viscometer tube was immersed in a tank of water

53


maintained at the required temperature. Although the K and a constants for

the MHS equation (Equation 2. 10) were not readily available for the DS 1. 7

and 1.5 polymers, it was hoped that some trend might be observed linking the

drop in the DS (and hence molecular weight), with the drop in the viscosity.

The wide tube viscometer was used and a number of dilutions were made. For

increased accuracy all results were repeated ten times and an average value

was calculated (See also Tables 3.12 - 3.14, Chapter 3).

N,N-Dimethylacetarnide was used next for two reasons. It was a solvent for

CA over a wide range of DS values and there were known K and a constants

for a few CAs at 25°C. From these constants, a calibration curve was created

in order to predict the values for any CA. The wide tube viscometer was used.

The next step involved adding LiCI to DMA (0.5g LiCI were added to 50 mls

DMA). These viscosities were run at 60°C. The wide tube viscometer was

used. This set was then repeated at 25°C, using the narrow tube viscometer.

(See also Tables 3.23 - 3.27).

In all the above cases, appropriate polymer concentrations were used so as to

satisfy equation 2.9.

54

CHAPTER 3

RESUL TS AND DISCUSSION

Chapter 3 - Results and Discussion

3.1 STUDIES ON DISCARDED FILTERS

3.1.1 BACTERIOLOGICAL ACTIVITY OF THE FILTERS

Discarded filters were placed in various bacteriological media in order to

establish which micro-organisms were growing on the filters.

Nutrient agar and nutrient broth were used for the cultivation of organisms

which were not exacting in their food requirements. The consortia of

organisms on the original nutrient agar plates, were isolated, purified and

identified using selective media.

Seven pure colonies were isolated. Colonies 1-4 gave a spore positive, Gram

positive stain and were identified as spore forming Bacillus spp. This was

verified by their growth on Bacillus selective medium plates (see photograph

1 ).

Colonies 5-7 gave a Gram negative, oxidase positive test and were identified

as Pseudomonas. This was verified by their growth on Pseudomonas selective

medium plates (see photograph 2).

All colonies were grown on tributyrin medium, and despite heavy growth on

the plates, no esterase activity was observed, as there was no clearing of the

medium around the colonies (see photograph 3).

Discarded filters were also grown on Sabouraud maltose agar which was

suitable for the cultivation of fungi. Growth was recorded after 3 days, but it

was concentrated on the outside and on the edges of the filters as can be seen

on photograph 4.

This observation led to the need to check the inside of the filters for any

microbiological activity. This was accomplished by tl!king discarded filters,

55

Photograph 1. Colony No. 4 on Bacilll/s selective medium.

Photograph 2. Colony No. 6 on Pseudomonas selective medium.

Photograph 3. Colony No. 4 on Tributyrin mediwn.

J

Photograph 4. Cigarette filter on Sabouraud maltose agar.


asceptically removing their tipping paper, dissecting out their middle portions

and laying them on plates containing nutrient agar as well as in flasks

containing nutrient broth. Nutrient broth was used because, being a liquid, it

would pick up any micro-organisms round the filter, while the nutrient agar

would only show any activity on the side of the filter in contact with the

medium.

Eighty percent of the nutrient agar plates and nutrient broth flasks were sterile

after 48 hours. The remaining 20% of the plates showed only slight growth

and the flasks showed only a slight turbidity and it was believed that it was

due to the handling of the filters rather than any growth coming from within

them. This experiment was repeated on numerous occasions using batches of

filters collected under different circumstances. The results were similar on

each occasion (see photographs 5 and 6).

This was a very surprising fact, as it was shown that biodegradation could not

occur as no micro-organisms were able to penetrate the filters. It was

therefore necessary to modify the filter technology in such a way as to allow

micro-organisms to penetrate and hence initiate biodegradation.

In order to achieve this, it was important to look closely into the way the filters

are manufactured. The filter consists of several layers all devised to keep it

tightly packed. The outer layer is the tipping paper, followed by the plug

wrap, followed by the filter itself. The tipping paper and the plug wrap had to

be removed, hence exposing the CA and making it available for any micro

organisms to penetrate.

The tipping paper is held into place by a glue, that was specifically designed to

keep the filter together on contact with water. The first bi.g change in the filter

manufacture was, therefore, the nature of the glue. It was suggested that a

56

Photograph 5. Centre of filter on nutrient agar after 48 hrs (note no growth).

Photograph 6. Centre of filter in nutrient broth after-48 hrs (note no growth).


water based glue should be introduced, which would allow the tipping paper to

separate from the rest of the filter once it got wet.

The removal of the tipping paper was not enough for optimum biodegradation.

Minimum volume and maximum surface area are two factors which greatly

enhance biodegradability. However, in the cigarette filters, the exact opposite

is observed, i.e. there is a maximum volume with a minimum surface area. In

order to maximise its surface area by separating the threads of CA,

mechanical force had to be applied from within the filter to make it swell and

ultimately disintegrate. Swelling agents were introduced that expanded in

contact with water and helped to force the filter fibres apart, hence making the

access to micro-organisms easier. This process has been patented and field

trials are being undertaken by Rothmans29.

3.1.2 STUDIES ON THE FIBRES OF THE DISCARDED FILTERS

The electron microscope used was an ISI-SS40 scanning electron microscope.

The study of the discarded filters at fibre level showed similar results. There

were no significant differences between the fibres of the virgin filters and

those that were discarded. The fibres were in both cases intact with only

discolouration due to dirt picked up from the street being observed on the

fibres of the discarded filters, and no signs of microbiological activity (see

photographs 7 to 16).

57

Photograph 7. Unplasticised un smoked filter fibres, 500x magnification.

Photograph 8. Plasticised unsmoked filter fibres, 500x magnification.

Photograph 9. Discarded cigarette filter fibres, 75x magnification.


Photograph 11. Discarded cigarette filter fibres, 21 Ox magnification.


Photograph 13. Discarded cigru'ette filter fibres, IOOOx magnification.



3.2 LONG TERM BIODEGRADATION STUDIES ON VIRGIN

FILTERS

In order to obtain a more accurate picture of the biodegradation process, the

seed tray experiments were devised. The seed trays in the 30°C and 37°C

incubators were kept in the dark because the incubators had no glass doors or

any means of letting sunlight in them. The ones in the greenhouse, however,

were split into two categories. The experiments to be run under light

conditions were covered with a transparent plastic cover, while the

experiments to be run under darkness, had the plastic covers covered with

aluminium foil. There was also a minimum - maximum thermometer installed

to monitor the temperature fluctuations in the greenhouse. The temperatures

recorded were in the range of _lOoC to +45°C.

At first it was thought that the high temperature in conjunction with the high

humidity (which was kept high on all seed tray experiments) would be the

predominant factor in the biodegradation of the cigarette filters. However,

after 12 months the seed trays in the 30 and 37°C incubators showed neither

visual signs of biodegradation nor a drop in their DS values (see photographs

17 to 21). On the other hand, the filters which were left in the greenhouse and

were subjected to light started to turn green after 9 months (see photographs

22 to 25). This was due to algae that grew on the surface of the cigarette

filters. Algae need light to grow and the high humidity also helped their

growth. The predominant genera of algae were Sfichococcus, Cryptomonas.

Chlamydomonas and Coelastrum (see photographs 26 to 29).

Some of those filters were washed with distilled water and an algal suspension·

was obtained. New filters were injected with hnl of the algal suspension and

run at room temperature in order to determine whether the algal growth had

any effect on the biodegradation of the cigarette filters, Control filters were

58

Photograph 17. Seed tray experiment, 0 days, 30°C.

Photograph 18. Seed tray expeliment, 60 days, 30°C.

Photograph 19. Seed tray expeliment, 0 days, 37°C.

Photograph 20. Seed tray experiment, 60 days, 37°C.

Photograph 21. Seed tray experiment, 11 months, 37°(',

&

Photograph 22. Seed tra) e pellment subjected to 1Ight 0 day_

Photograph 23. Seed tlay eXp~llll1cnt ubJcLtcd to 1Ight 12 lIIonth, .

Photograph 22. Seed Ira, expenment subjected to light, 0 da)s

Photoeraph 23. Seed tl a\ e pCl1ment ubJected to light 12 months

Photograph 24. Seed tray expenment subjected to light, 0 days.

Photograph 25. Seed tray epelllllent subj.:cted to light. 12 months.

Photograph 26. Algae that grew on the cigarette filters.

Photograph 27. Algae that grew on the cigarette filters .


also set up. These were injected with ImI of sterile water. Determination of

DS on filters over a number of months showed that the DS started to drop

when the algae were present. A possible explanation for this was that the cell

walls of the algae consisted of cellulose and therefore, as the cell walls of the

algae degraded, the cellulase responsible for this degradation started acting on

the CA as weU, hence causing some biodegradation32.33

,34 Light, therefore,

seemed to be an important factor in the CA biodegradation.

Table 3.1 below shows the relationship between the incubation time and the

DS.

TABLE 3.1

Tbe effect of the algae on the cigarette filters over a period of 1 months

Algae

Control

2.50 ± 0.02 2.30 ± 0.03

2.50 ± 0.02 2.48 ± 0.02

2.25±0.04 2.19±0.03 2.14 ±0.03

2.44 ±0.03 2.43 ± 0.02 2.40 ± 0.04

Other seed tray experiments were also set up in order to check the

biodegradability of other forms of CA apart from the cigarette fi lters. These

consisted of filter tow and cigarette filters half buried in the potting

compost/soil mixture, in order to have the experiment and the control samples

on the same tray, under identical conditions (see photographs 30 and 3\). The

filter tow was CA in the form of strands which were opened up and

unplasticised (see photographs 32 and 33). This way, the surface area was

greatly increased. However, these experiments produced similar results to the

ones carried out on filters. The other two surfaces which were used at the

beginning of the project, i.e. sand and tile were discontinued after 5 months

because they did not retain any moisture, hence the filters dried out very

59

Photograph 30. Half buried filter rods, 0 days .

• • •

•

r H 1

1'1 ) .. r

Photogral)h 31. Comparison of the buried and free part of the rod after 12 months (note no difference).

Photograph 32. Half buried filter tow, 0 days

Photograllh 33. Comparison of the bUried and fi'ee part of the tow after 12 months (note no dIfference)


quickly and minimised their chances of biodegradation (see photographs 34 to

37).

Table 3.2 shows the change in the OS of various types of CA (with an original

OS of2.5) after 12 months.

TABLE 3.2

The changes in the DS of various types of CA after 12 months

DS 2.00 ± 0.03 2.40 ± 0.0 I 2.38 ± 0.03 2.44 ± 0.03 2.43 ± 0.04

60

Pbotograpb 34. Tile experiment, 0 days.

Photograph JS. Tile experiment, 12 months

hP'!. ~ ... 13

Photograph 36. Sand experiment, 0 days

EXPERIMENT 13 ~.2 MONTIIS

Photograph 37. Sand experiment, 12 monlhs


3.3 STUDY OF ADDITIVES ON THE CA BIODEGRADATION

As the cigarette filters on their own proved to be vel)' resistant to

biodegradation, various additives were introduced which were highly

biodegradable (simple sugars and amino-acids). ]t was hoped that tbe micro

organisms would start by attacking the additive and once a significant number

of them were present -and the additive depleted- they would then attack the

CA instead of dying. This could happen, as the micro-organisms might

change the nature of the excreted cellulases and/or esterases to suit the CA

biodegradation.

In the case of CA, this did not happen. Heavy growth was observed on the

outside of the filters 10 days after inoculation (see also photograph 38), which

gradually disappeared over the next 10 days, once all the additive was used up.

The fact that additives did not encourage CA biodegradation was expected for

a number of reasons.

In the manufacture of cigarette filters, a plasticiser is sprayed on the CA tow

before it proceeds to the next steps of the filter manufacture, in order to harden

it. The tow is a flat arrangement of thousands of CA unplasticised fibres.

This is the form in which the CA is introduced in the filter machines. The

reason that plasticisation is so important, is that the fibres on their own are

vel)' soft and it is impossible to compact them in an adequate fashion to form

the perfectly cylindrical shape of the standard cigarette filter. The fact that the

fibres must be tightly packed is also vel)' important for the correct operation of

the filter. After plasticisation, however, the fibres do not only become rigid,

but also they stick together forming the filter, whicb as has been described

earlier, is vel)' effective not onJy against the smoking by-products, but also

against microbial penetration. The plasticiser used is triacetin, whicb is a very

good source of energy for the micro-organisms. In effect, the cigarette filters

have an excellent additive already incorporated in them. However, this does

6 1

GL~CERINE 5 D~yS

Photograph 38. Growth on the additive enriched filters after 10 days.


not seem to enhance the biodegradation of CA. Therefore, it is a reasonable

assumption, that any similar compounds that could have been used for the

same reason, would fail also.

The fact that triacetin was readily biodegradable was also demonstrated using

the Warburg apparatus. The principle involved is that if the volume of a gas is

held constant, at constant temperature, any changes in the quantity of gas may

be measured by changes in pressure. The Warburg instrument consists of a

flask attached to a manometer by means of a ground glass joint. The flask may

have one or more side bulbs which permit the addition of the substrate or

reagents at intervals as required. The manometer fluid is contained in a

reservoir and its level can be adjusted by means of a screw clamp. The tap

permits the flask to be opened to the air. When assembled, the apparatus is

fitted on a shaking device attached to a thermostatic water bath and so

arranged that the flask is completely submerged. Accurate temperature control

IS necessary. In operation the level of fluid in the closed limb of the

manometer is always adjusted to the zero mark and the level in the open limb

recorded. This observed pressure difference (in millimetres) when multiplied

by a constant, which must be determined for each flask and manometer, gives

the quantity of gas evolved or absorbed.

The respiration of most living cells, as opposed to many enzyme preparations,

results in the consumption of oxygen and the evolution of carbon dioxide.

Plasticised and unplasticised filters were examined using the above described

technique. Flasks with two side-arms were used, the first containing a

commercially available esterase and the other, sodium bicarbonate. Evolution

of CO2 was only observed from the plasticised filters. This meant that the

plasticiser was attacked but not the polymer itself. When the plasticiser came

in contact with the esterase, acetic acid was formed. Sodjum bicarbonate was

then released in the flask, releasing the CO2 which was measured.

62


These two pieces of evidence proved that the CA with a DS of 2.5 would not

biodegrade even if additives were employed.

63


3.4 PREPARATION OF CA WITH GIVEN DS

It was obvious from the evidence up to that point, that CA with a OS of 2.5

was highly resistant to biodegradation. It was therefore necessary to

chemically hydrolyse CA with a OS of 2.5 to lower DS products, fully

characterise these and check them for biodegradation. Although preparative

methods for cellulose acetates with lower DS values have been published, they

are of laboratory use only.

The first such method is the successive solution fractionation method, used by

Kamide et at2. This method gave a wide range of DS values by using an

appropriate non-solvent to successively precipitate the required polymer

fraction out of solution. The other method reported by Buchanan et at3, was

specifically designed to produce monoacetates. However, the method required

high temperatures and pressures, as well as making use of chemicals that

would not be feasible in an industrial environment.

The aim in this project was to devise a method that could be used in industry

with the already available means. The production of these CAs posed a great

number of problems. The original recipe was the one used by Courtaulds, but

extensive modifications had to be made before it could be used in the

laboratory on a reasonable scale (50-100g).

The Courtaulds recipe consists of a one stage acetylationlhydrolysis process.

Woodpulp, which is the starting material, is fully acetylated to the triacetate

and then hydrolysed back to a DS of 2.5. The woodpulp is initially pre-treated

with glacial acetic acid and the pre-treated pulp is placed in an oven at 50°C'

for 30 minutes. The reason for this pre-treatment stage is to "open up" the

cellulosic matter and make the acetylation proper more uniform, controllable

andless harsh (see also Chapter 1).

64


The acetylation mixture is a mixture of acetic acid and acetic anhydride with

sulphuric acid as the catalyst. The mixture is cooled to -16°C and it is added

to the woodpulp. The acetylation mixture is cooled in order to avoid the

premature acetylation of the woodpulp. The reaction is allowed to proceed at

about 30°C with constant stirring. The end product is a viscous triacetate

dope.

The reaction is then held at that temperature to allow the viscosity of the

polymer to fall. At the required viscosity (70-100 cP in a 6% solution of

polymer in a 95:5 acetone:water solvent mixture), a hydrolysis charge is added

to the mixture. This contains magnesium acetate and acetic acid and has to be

added slowly to avoid precipitation. This hydrolysis charge is used to "kill

oft" all the acetic anhydride and some of the sulphuric acid.

The resulting mixture is heated to 65°C and held for 30 minutes. At the end of

that time, a second hydrolysis charge is added, containing magnesium acetate,

acetic acid and water.

The mixture is then heated to 80°C until the required acetyl value is achieved.

At that point, a neutralisation charge containing magnesium acetate is added to

neutralise all the sulphuric acid present.

In this project, the starting material was already acetylated (CA with a DS of

2.5). Therefore, the acetylation was omitted. After initial consultation with

Courtaulds, it was suggested that the addition charges remained as they

contained acetic acid, despite the fact that they contained magnesium acetate,

which could neutralise part of the catalyst.

Small scale reactions (5g CA) were performed, in order to establish a working

formula under laboratory conditions. The parameters that were taken into

65


account were the amounts of catalyst, glacial acetic acid and water to be used

as well as the hydrolysis time. In the pilot plant fonnula, the amount of

catalyst was 13.75g per 100g of cellulose and the solvent was 345g per 100g

of cellulose. As with all industrial processes, the amount of solvent present

was kept to a minimum. It became obvious, however, that the quantity of

solvent would have to be greatly increased in order to obtain a mixture that

could be stirred under laboratory conditions. The solvent quantity was

therefore increased to 2000mls per IOOg of CA. The catalyst present was also

increased to 20g in 100g of CA. Water had also to be added to the mixture

and it was set at 10% of the total weight of the CA, catalyst and solvent added.

The two addition charges were scaled down accordingly for the laboratory

scale reaction. The hydrolysis time started when the reaction temperature

reached 80°C ± 5°C after the addition of the second charge and ended with the

addition of the neutralisation charge. For the fIrst reaction, the hydrolysis time

was set at one hour and thirty five minutes. This time was arbitrary, and it was

hoped that by trial and error, a set of CAs with appropriate DS values would

be prepared. The product was precipitated out from distilled water, but its DS

characterisation showed that no hydrolysis had taken place. The hydrolysis

time was increased in order to see what effect it would have on the end

product. The new time was set at two and a half hours. However, after

precipitation from distilled water, the product still did not display any drop in

its DS. Finally, the hydrolysis time was increased to four hours at 80°C ±

5°C. However, the product proved very difficult to precipitate out of solution.

Water was again used as the precipitant, but the amount isolated was very

small. The reaction mixture was then boiled down to a minimum volume and

reprecipitated again. The resulting product was brown, indicating that the

reaction conditions were too harsh, possibly causing thennal decomposition

of the product.

66


The next step was to increase the catalyst concentration. The new

concentration was 40g in 100g of CA and two polymers were synthesised with

hydrolysis times of two and three hours. It became evident that the OS of the

resulting products was significantly reduced when it was not possible to

precipitate the polymers from distilled water. Acetone was used as the

precipitant, but the precipitation time was long and the amount of recovered

polymer was minimal.. It became immediately apparent, that acetone was not

a good precipitant of low OS material and that an alternative solvent would

have to be found. The OS of both polymers was deduced to be zero. The

reason for using acetone as the precipitant was the following. The high OS

polymers were acetone soluble and were precipitated out from water. It was

therefore thought, that as the OS was evidently very low, then the polymers

would be water soluble and the reverse would hold also and acetone would

precipitate them out of solution.

Furthermore, the precipitation technique was modified further, to account for

any CA irrespective of its OS. The products with higher OS could be easily

precipitated from distilled water. But as the DS dropped, so did their

solubility in acetone. In general, CA molecules with a DS value between 2.5

and 1.8 were acetone soluble (and could be precipitated from water).

Cellulose acetates with OS between 1.7 and 1.1 were soluble in acetone/water

mixtures and CA with a DS lower than I were regarded as water soluble. At

the time of the hydrolysis, the OS of the CA was unknown and it was a matter

of careful planning as to which solvent to use in order to precipitate the

product out of solution. The method which was used, was described below.

A series of acetone/water mixtures was set up ranging from 100% acetone to

100% water, in 5% increments. Cellulose acetate solution (lrill) was added to

5rn1s of each of the solutions in a preweighed weighi~g bottle. The CA

precipitate in each bottle was filtered and the bottle was reweighed. The

67


solution which gave the largest weight of CA was the one used to precipitate

out the CA.

The final step was to decrease the hydrolysis temperature from 80°C ± 5°C to

65°C ± 5°C. The hydrolysis times ranged between one and three hours for this

new procedure. After one hour, the product was precipitated from distilled

water and was white in appearance, with its DS slightly decreased to 2.1.

After two hours, the DS had not changed significantly from before. After

three hours, however, the DS had dropped considerably as it was precipitated

from a 60:40 (v/v) acetone:water mixture. When the DS was worked out,

however, the method still gave a DS of zero.

By that time, it was evident that the method used for the characterisation of the

DS was flawed. As the polymer was soluble in a 40:60 acetone:water mixture,

its DS should be between 1.7 and 1. 1. The great discrepancy between the

solubility observations and the DS determination, showed that the DS results

were not reliable. The Courtaulds method for determining the DS and which

was used up to that point was as follows:

Approximately 1.5g of fmely ground CA sample were weighed in a glass

weighing bottle which was placed in an electric oven at 110°C for three hours.

At the end of this time, the weighing bottle was transferred to a desiccator,

allowed to cool and then stoppered. When cool, the bottle was weighed

accurately and the contents were poured carefully in a 500ml B34 glass

stoppered conical flask and the bottle was re-weighed. 10mls of distilled

water and 90mls of acetone were added.

A blank was carried out using potassium hydrogen phthalate (KHP). KHP

(2.6g) was weighed in a weighing bottle and then dried in an oven at 110°C

for three hours. After cooling in a desiccator, the KHP was weighed

68


accurately by difference into a B34 glass stoppered 500ml conical flask, and

60rnls of distilled water and 90mls of acetone were added.

The flask containing the sample and the flask containing the KHP were then

warmed on a hot plate to assist dissolution of the CA and the KHP, and to

expel dissolved carbon dioxide from the solvent. When solutioning of the

sample and the KHP was complete, the flasks were allowed to cool.

NaOH (IN, 25m1s) was added to each flask from an automatic pipette, the

contents being stirred continuously with a magnetic stirrer. In the case of the

sample, the continuous stirring ensured the formation of a fine precipitate.

The draining time from the automatic pipette should be identical for each

delivery.

The flasks were stoppered and allowed to stand for 30 minutes with occasional

stirring. At the end of this time, 50mls of distilled water were added to the

sample flask and then 25m1s of O.5N hydrochloric acid were added from an

automatic pipette, the draining times being identical.

After shaking, the contents of the flask were titrated with O.IN sodium

hydroxide using phenolphthalein as the indicator. It was desirable to carry out

the final O.IN titrations with sodium hydroxide and the weights of the sample

and the KHP were arranged to ensure this. In certain abnormal cases, with a

sample of very low acetic acid yield of CA, it might be necessary to carry out

the fmal titration with O. IN sulphuric acid.

The acetic acid yield was calculated as follows:

69


Blank figure (BF) = weight ofKHP * 4.896 - 0.1 * titration O.IN NaOH

Equation 3.1

Acetic acid yield = (BF + 0.1 * titration O.IN NaOH) * 6.005/wt of sample

OR

(BF - 0.1 * titration O.IN H2S04) * 6.005 / wt of sample Equation 3.2

The DS was converted from the acetic acid yield figure with the aid of tables.

In order to verify whether the method was flawed or not, the chemistry was

worked out from fust principles with the aid of results obtained by that

method:

Weight of CA : 1.4815g

Weight ofKHP: 2.5657g

Titration value for CA : 19.65cm3

Titration value for KHP: 7.32cm3

SAMPLE

CA + NaOH ..

302g (triacetate) yields 3 x 40g NaOH

265.5g (DS 2.5) X? X = l05.5g NaOH

265.5g

1.4815g

yields 105.5g NaOH

Y?

25mls of IN NaOH were used.

In 1000mls of solution

25rn1s

40g ofNaOH

Z?

Y = O.59g NaOH Equation 3.3

Z= IgNa.OH Equation 3.4

70


Equation 3.3 is the weight of NaOH needed to hydrolyse the CA of a OS of

2.5 to cellulose.

Equation 3.4 is the weight of NaOH actually added.

From equations 3.3 and 3.4: there is an excess of0.41g ofNaOH present.

Equation 3.5

Again from reaction (I] the weight of CH3C02Na can be determined.

265.5g of CA yield 205g ofCH3C02Na

1.4815g X? X = l.l4g CH3COzNa

X is the amount ofCH3C02Na yielded by equation l I].

CH3COzNa + HCI

82g 36.5g

1.14g Y? Y = O.51g HCI Equation 3.6

Equation 3.6 is the amount ofHCl needed for the reaction

25cm3 ofO.5N HCl were added

lOOOmls of solution contain 18.25g HCl

25nlis Z? Z = O.46g HCl Equation 3.7

From equations 3.6 and 3.7 there is a deficit of HCl of O.05g Equation 3.8

From equations 3.5 and 3.8 it can be deduced that there is a O.4lg excess of

NaOH and that reaction [ 11 ] does not go to completion.

From reaction ( 11 ]

82g of CH3C02Na yield

1.14g

60g ofCH3C02H

X? X = O.S3g CH3COzH

71


CH3C02H + NaOH

60g 40g

0.83g Y? Y = O.55g NaOH Equation 3.9

From the titration ofNaOH (O.IN NaOH) with acetic acid:

In 1000mIs of solution 4g ofNaOH

19.63mls Z? Z = 0.44g NaOH Equation 3.10

From equations 3.9 and 3.10 there is a deficit of O.llg NaOH which is just

taken up by the 0.41g excess ofNaOH.

BLANK

o

204.23g

2.5657g

COOK

COOH

X = O.50g NaOH

+ NaOH

40g

X?

Equation 3.11

236.23g

Y?

COOK

COONa

+ H20

[IV I

72


COOK

y= 2.84g o COONa

From equations 3.4 and 3.11 there is a O.SOg excess ofNaOH.

o

226.23g

2.84g

X=O.92g HCI

COOK

+ 2HCI---- 0 COONa

73g

X?

Equation 3.12

COOH

+ NaCI + KC!

COOH

[V I

From equations 3.6 and 3.12 there is a deficit of HCI and an excess of NaOH

which again made titration impossible.

As this method conclusively proved that it gave unreliable results, a new

method was used which was supplied from Eastrnan Kodak which is fully

described in Chapter 2.

After many small scale hydrolyses, a set of CAs with various DS values was

prepared. The method used was the one described in pages 66-68 and the DS

characterisation was undertaken with the new Eastrnan Kodak method.

73


The same reactions were repeated on a bigger scale (60g), but, unfortunately,

the scaling up changed the hydrolysis times completely. Therefore, the same

procedure had to be repeated, until new reaction times were established.

Furthermore, the precipitation procedure had to be modified for the lower DS

polymers. As previously mentioned the low DS polymers were difficult to

precipitate from solution. When acetone was used as the precipitant, the

amount recovered was minimal and therefore, a new non-solvent had to be

utilised. Many solvents were tried, and eventually, isopropanol was found to

be the best.

Finally, despite Courtaulds scepticism, the reactions were also speeded up

even further by eliminating the two addition charges that were initially used

and increasing the hydrolysis temperature back to 80°C. The elimination of

the addition charges prevented the partial neutralisation of the catalyst, and it

was found that although no more acetic acid was added during the reaction, it

gave very good results.

Table 3.3 shows the relationship between hydrolysis time, DS and the

precipitation medium.

TABLE 3.3

Table of the hydrolysis time and precipitation medium used to isolate the

various chemicaUy synthesised CA

water water water isopropanol isopropanol

74


The problems that were encountered with the chemical hydrolyses were

mostly methodology problems which took a very long time to resolve. The

fact that the initial OS method was unreliable made the procedure even

lengthier, as contradictory results were obtained. The fact that all hydrolyses

took place on a hotplate, meant that the temperature control was not very

accurate despite very careful monitoring of the temperature inside the flask

with the aid of a thermometer. It was very easy to exceed the 5°C margin and

that meant that many hydrolyses had to be repeated more than once in order to

get reproducible results. If the OS of any batch was outside 0.2 of the target,

it was discarded and a replacement batch was produced. Many batches were

also lost in the search for a good non-solvent for the low OS products.

75


3.5 THE BIODEGRADA nON OF CA OF GIVEN DS USING

ASPERGILLUS FUMIGATUS

The chemically hydrolysed cellulose acetates were subjected to biological

hydrolysis in order to establish a possible biodegradation pathway for CA.

The literature seems to be unclear as regards the mechanism of the

biodegradation of the CA., i.e. whether de-acetylation precedes de

polymerisation, or vice versa, or if both processes are happening

simultaneously. By following the biodegradation of the CA with the lower DS

values over a period of time, it was possible to study the biodegradation

mechanism in some detail. The medium used contained CA as the sole

carbon source, and it was inoculated with the fungus Aspergillus fumigatus, as

this was the fungus that was isolated and purified from a consortium of micro

organisms that were initially found on discarded cigarette filters. The

hydrolysis time was 10 days at 30°C.

The problem that was encountered initially, was the isolation of the CA at the

end of the biodegradation period. The fungus after 10 days had grown in such

a way that it formed a large number of globules, leaving a very small amount

of CA at the bottom of the flask. In addition, the medium had acquired a faint

yellow colour. When the CA was filtered and weighed, it represented a very

small amount of the initial 5g that were added to the flask at time zero.

This implied two things. The CA was completely hydrolysed to soluble

derivatives, carbon dioxide and water, or the CA was adhering to the fungus.

In order to test this, glacial acetic acid was added to the fungus and the

resulting solution was left overnight. If there was any CA adhering to the .

fungus, it would dissolve in the glacial acetic acid and, after filtering off the

fungus, the CA could then be isolated as described in the previous section.

76


Following this method, more CA was isolated, proving the association of the

polymer with the fungus.

This method again proved the resistance of CA with a OS of 2.5 to microbial

attack. After 10 days it was the only CA that did not associate with the

fungus, the liquid remained colourless, and the whole 5g were recovered,

allowing for a small quantity (0.3g) that was lost in the recovery of the

polymer. The amounts recovered for the OS 1.7, 1.5, 1.0 and 0.7 ranged from

3g for the OS 1.7 down to 0.7g for the OS 0.7. This experiment also proved

the point that the lower the OS the greater the chance of biodegradation (see

photograph 39). After the initial experiment, the contents of each flask were

transferred to measuring cylinders. It was immediately obvious that the

amount of fungus increased with decreasing OS (see photographs 40 to 44).

A problem that arose, however, was the determination of the OS value for the

biodegraded CA samples. The Eastman Kodak method was deployed, but the

results obtained were unreliable. The OS values for the degraded products

appeared to have significantly larger values than the equivalent starting

polymers. A possible explanation for this behaviour was that the Eastman

Kodak method could not be employed successfully below a certain degree of

polymerisation (DP), i.e., if the chain was cleaved to such an extent that the

number of monomer units decreased dramatically, the method broke down.

This was overcome by usmg a spectroscopic method for determining OS

values. By using the starting undegraded CA samples as the polymers with the

known OS values, a calibration curve of the ratio of the hydroxyl peak to the .

carbonyl peak versus the OS was constructed, from which any CA could be

assessed for its OS value as long as an FT-IR spectrum was available.

Table 3.4 and Graph 1 summarise the results.

77

Photograph 39. The difference in growth of the fungus with changing DS values (note the lower the DS the IIDoger the growth) o

Photograph 40. DS2. 5 after being in contact with the fungus for \0 days. (Note no growth) .

Photograph 41. OS 1. 7 after being in contact with the fungus for to days.

Pllotograph 42. DS 1.5 after being in contact with the fungus for J 0 days.

Photograph 43. OS 1.0 after being in contact with the fungus for 10 days.

PhotogralJh 44. DSO 7 after being 111 contact wIth the fungus for 10 days.


TABLE 3.4

Table of the ratio of OH (peak height) over C=O (peak area) for CA

with given DS values

2.5 42.375 0.099 2.34

1.7 18.471 0.106 5.75

1.5 20.893 0.129 6.18

1.0 8.674 0.06875 7.93

1.0 (dry) 13.491 0.108 8.00

10DS1.5 7.126 0.0607 8.518

Note, that IODS 1.5 is the CA that has been in contact with the micro

organisms for 10 days and with a starting DS of 1.5. Its final DS was

calculated by using the above mentioned calibration curve and was deduced to

be 0.9 (See also Graph 1).

It can be observed by looking at the attached FT-IR spectra (overleaf), that

some moisture is present in the spectra of the CA with DS values 1.7, 1.5 and

10DS 1.5. This fact could make the whole procedure unreliable, because some

of the contribution of the OH peak would be related to the atmospheric

moisture and not to the polymer itself. Therefore, it was imperative to assess

the extent of the atmospheric moisture contribution and fInd a way of

eliminating it altogether. This was done by drying the DS 1.0 fIlm very

thoroughly as described in the experimental section. After 24 hours in the

vacuum desiccator, and 24 hours in an electrical oven, this spectrum was run

again (in the table above it is recorded as DS 1.0 (dry» and the difference with

78

1.3 - DS2.5 FT-IR spectrum

1.2 -

1.1 -1755.8 cm-1 c=o

1.0 -

0.9 -

0.8 -A b s 0.7 -0

r b 0.6 -a n c

0.5 -e

, 0.4 -

0.3 -

, 0.2 -

0.1 - 3480.1 cm-1 OH

~ If 0.0 - U )v ~ -0.1

I I I I I , , 4000 3500 3000 2500 2000 1500 1000 500

Wavenumbens (cm-1) ~

0.50 -DS 1.7 FT-IR spectrum 1749.2 cm-1 C=O

0.45 -

0.40 -

0.35 -

0.30 -

0.25 -

0.20 -

A 0.15 - 3466.9 cm-1 OH b s 0 0.10 -r r b 0.05 -

rc~k~ N a

IV \ IrJ ~ n c 0.00 - ~ e vi -0.05 - ,

-0.10 -

-0.15 -

-0.20 - --0.25 -

-0.30 -

-0.35 -

I I I I I I

4000 3500 3000 2500 2000 1500 1000 500

Wavenumbers (cm-1) "

0.70

0.65 DS1.5 FT-IR spectrum

749.2 cm-I c=o

0.60

0.55

0.50

0.45 A b s 0.40 0

r b a 0.35

n c

• 0.30 cm-I OH

0.25

0.20

0.15

0.10

0.05

4000 3500 3000 2500 2000 1500 1000 500

Wavenumbers

0.40 -:

0.38 -: OS 1.0 FT-IR spectrum

0.36 - 1749.2 cm-I C=O

0.34 -

0.32 -:

0.30 -

0.28 -:

0.26 -

A 0.24 -b s 0.22 -0

r 0.20 -:

b 11 a

0.18-' n c 3466.9 cm-l OH e 0.16 -:

0.14 -: ,

~ 0.12 -

0.10 -:

0.08 - \rJ

0.08 -W Ill. 0.04 - W

0.02 -

0.00 -: If I I I I I I I

4000 3500 3000 2500 2000 1500 1000 500

Wavenumbers (cm-I) ,.

0.18 - IODS 1.5 FT-IR spectrum ~cm-l c=o ,

0.16 -

0.14 -

0.12 -

0.10 -

A b 0.08 -s 3466.9 cm-l OH 0

r b 0.08 -

a n c 0.04 -e

1

002 -

~~ ~~ '"' 000-r ~

~

~ -0.02 -

-0.04 - V -0.08 -

I I I I ' I I I

4000 3500 3000 2500 2000 1500 1000 500

Wavenumbers (cm-l) .

12

10

8

ffi o

11 6 ~ :c o

4

2

o. 0.5

GRAPH 1 - GRAPH OF OHlC=O vs.DS

1 1.5 2 2.5

DS


the old spectrum was noted. It can be seen that the difference is minimal, but

nevertheless significant, as this small difference shows the contribution of the

atmospheric moisture on the polymer film.

79


3.6 THE CHARACTERISATION OF THE CAs WITH GIVEN DS

IN THEIR ORIGINAL AND BIODEGRADED FORM

3.6.1 MOLECULAR WEIGHT DETERMINATION

3.6.1.1 GPC RESULTS

3.6.1.1.1 THF SYSTEM

This method was used only on the starting material, as THF was not a solvent

for the rest of the CAs. This set of results was obtained very early on in the

project, as a method of characterising the starting material.

A starting material characterisation is always important, but in this case it was

vital, as there was no reliable information about it. There were doubts as to its

molecular weight, DP and OS and various values were quoted from different

sources. Although the apparatus and especially the data collection system was

not particularly modem, this method would give the fIrst concrete evidence as

to the molecular weight of the starting CA.

Table 3.5 summarises the calibration procedure used for the THF GPC system.

80


TABLE 3.5

The molecular weight of the polystyrene standards vs. their % equivalent

oftoluene

8,900 3.95 149/192 = 0.78

1,850 3.27 161/192 = 0.84

106,000 5.03 129.5/193 = 0.67

22,000 4.34 143/193 = 0.74

2,855 3.46 158/193 = 0.82

~4' 198,000 5.30 121/190 = 0.64

32,500 4.51 137/190 = 0.72

5,000 3.70 152/190 = 0.80

Graph 2 overleaf shows the resulting calibration curve:

81

o

'"" ;, ~ ::-u

~ 0 z

0 -E-o

~ '"" Z

l:Q

'"" -

;;;J

....;!

'"" < U

0 Eo-

U Q. ~

I ?Je.

M

::t Q.

": ~ 0

~


Table 3.6 below shows the results of a typical run using 4mgs of sample in

4mgs of solvent and using toluene as the marker. As mentioned in the

experimental section, the sample peak was sectioned in 2mm stripes, and the

distance of each stripe from the injection point as well as its height in mm

were recorded. By feeding this information into a computer operating with the

CALl software, molecular weights could be calculated.

TABLE 3.6

Table of the distance of the peak stripes vs. their height

117 2

119 3.5

121 5

123 6.5

125 7.5

127 8

129 8.5

131 7.5

133 6.5

135 5

137 3.5

139 2.5

141 1

191 82 (Toluene)

82


The program evaluated the molecular weights as follows:

Mn = 83,463

Mw = 137,739

Mp = 107,220

Mwd= 1.65

The DP was therefore worked out as 318.

The experiment was repeated several times. Table 3.7 summarises the results:

TABLE 3.7

Table of the molecular weights of CA with a DS of 2.5 at various

concentrations

It can be seen that the lower concentration of 4mgs/ml gave higher results than

the 7mgs/ml concentration. Furthermore, there seems to be a split in the

results in the higher concentration. The first two results are in quite good

agreement with each other and so are the second two results which were

obtained later in the day. However, there is a big difference between the two

sets of results with the 0 P dropping from approximately 240 down to 177. It

was thought that with time, the polymer chains might "coagulate" and give

83


false readings. To test this, a new set of results was obtained for three

different concentrations at five different times.

The results are summarised in Table 3.8.

TABLE 3.8

Table of the molecular weights of CA with a DS of 2.5 at various time

intervals and concentrations

5h 55mins 66,707 149,942 100,011 2.25

6h 57mins 72,719 184,312 115,771 2.53

7h 47mins 65,182 170,034 105,277 2.61

13h 06mins 64,110 180,036 107,435 2.81

6 4h 45mins 75,724 184,093 118,069 2.43

6h 10mins 65,616 182,069 109,301 2.77

7h 05mins 63,565 164,355 102,212 2.59

7h 55mins 61,337 168,375 101;625 2.75

12h 20mins 63,491 173,912 105,080 2.74

8 5h 20mins 60,277 165,039 99,740 2.74

6h 55mins 59,079 156,535 96,166 2.65

7h 35mins 62,160 173,498 103,849 2,79

8h 30mins 59,402 171,419 100,909 2.89

13h 10mins 76,140 188,587 119,829 2.48

After having repeated the above experiment for 3 different concentrations and

for 5 different times, it can be seen that no "coagulation" effects are observed,

as there if no pattern in the molecular weights in each of the sets. These

deviations could therefore ouly be attributed to equipm~nt variations. As the

84


variations were so marked, an exact figure could not be determined, but it

could be estimated that the starting material had a molecular weight of about

65,000 and an average DP of about 240, which was roughly in the expected

range as informed by Courtaulds.

3.6.1.1.2 DMAlLiCI SYSTEM

DMA has been used by many workers as an appropriate solvent for CA

polymerlS-42 This system has certain advantages over the THF system.

Firstly, it can give information for a greater number of CA polymers as DMA

is a better solvent for a wider range of OS values. Secondly, due to its up-to

date data processing unit (using the PL Caliber GPC software), results can be

obtained from much smaller quantities than before. This is particularly useful,

as the solubility in DMA (and generally in all solvents that were tried) drops

quite significantly with decreasing OS. Two different standards were used,

but the polystyrene standards gave results that were closer to molecular

weights expected. The resulting chromatograms are overleaf.

Table 3.9 shows the relationship between Mo and the starting CAs and Table

3. 10 shows the relationship between the molecular weight and the CA after

having been in contact with the fungus for a period of 10 days.

85

.'; ".-

Polyllt!r Liiboratories GPC Overlay - dw/dlogM vs logH Plots 14:27 Fri Feb 04 1994 Overlaid Differential Molecular Weight Graphs ---DDMAC4.009 DDHAC4.010 DDMAC5.002 ---CW7C1O&RMAC4 .007

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

I

l ~ ~

I

l

J

3.0

OSl.O

OSO.7

4.0

---DDMAC5.003

OS 1.5

OSI.7

5.0

Overlaid chromatograms for the starting polymers PEOIPEG calibration

OS2.5

6.0

Polytller Laborator jes SPC Overlay - dw/dlogM vs logM Plots 1~ 15 Fri Feb 04 1994 Overlaid Differential Molecular Height Graphs - - -OOMAR4.009 DOMAR4.010 D0MAR4.011

dw/dlo!lllMAR5.003

OSLO

0.8

0.7

0.6 DSO.7

0.5

I 0.4 ...j

0.3 ~ 0.2

0.1

0.0

---DDMAR5.002

OS 1.5

OSl7

5.0

OS2.5

Overlaid chromatograms for the struting polymen Polystyrene calibration

t. t 100SI 0

t.O

0.9

0.8

0.7

0.8

0.5

0."

0.3

0.2

O.t

0.0 ".0

---1ISAII9.01" DS 1·1

lOOS 1.5

lOOS I 7

---IISAII9.007 VS f.c>

100S2.5

:0

Overlaid chromatograms for the biodegraded polymer~

PEOIPEG calibration

--

N}IIIr I ' wtrJ. liFe Onrlly 14: .. TuI SIp 20 llI84 Onrll1d Dlffll'lllt111 1ID11CU1 .. Ml1f1t IINp/II ---IIIM!I,007 D!WIS.Ol0

dll/dlogll os I.Q J)S 2 · 5

1.1

IODSI.O 1.0

0.1

0,8

0.7

0.8

0.5

0.4

0,3

0.2

0.1

. 0.0

---D!WIS,013 DS I. '!)

IODS1.5

--0IAll3,014 OS I.,

IODS1.7

Overlaid chromatograms for the biodegraded polymer Polystyrene calibration

IOD82.5

11,0


TABLE 3.9

The molecular weights (Mn) of the starting CA with varying OS and their

corresponding OP using the two sets of calibration standards.

DP

PObYST¥rulNE

UP

TABLE 3.10

167

98,500

393

137

96,000

336

130

78,800

326

42

28,500

130

32

18,900

91

The molecular weights (Mn) of the CA with varying OS after 10 days of

biodegradation using the two sets of calibration standards.

114,500 104,900 87,000 34,500

By looking at the results in Table 3.9, it can be seen that for the higher DS

polymers (DS 1. 7 and 1.5), there is some decrease in the degree of

polymerisation (DP) accompanying the decrease in the DS.

The drop in molecular weight in the last two polymers with DS values of 1.0

and 0.7 is much greater, indicating a very large drop in the DP (see also the

discussion on GPC accuracy on the next page). This trend illustrates a well

known problem in industry, i.e. that chemical hydrolysis under these

conditions, also promotes some chain scission. However, for the higher DS

polymers, the problem is not too marked.

86


By looking at the molecular weights of the biodegraded samples (Table 3.10),

the picture becomes more complex as these values are larger than the

molecular weights recorded for the starting materials for the polystyrene

standards and decrease only slightly in the case of the polyethylene glycol

standards, with the exception of the DS l.0 where the molecular weight is also

higher compared to the starting polymer.

This apparent discrepancy, highlights some of the disadvantages of the GPC

technique. Firstly, the results obtained, are relative to the standard used to

calibrate the instrument and are not absolute values. This in turn means that

should the calibration be incorrect in any way, then the results would be

incorrect also.

The other problem with GPC is that association between the column and the

polymer in solution can take place43. The very big drop in the low DS

polymers highlights the problem. As the DS drops, the solubility of the

polymer decreases. This in turn increases the adsorption of the polymer to the

column packing. This means that the polymer elutes later than expected,

giving lower MW values than expected28. This could, therefore, mean that the

low DS polymers may have a higher molecular weight than indicated in Table

3.9.

Somewhat surprising is the variation in the results between the DS 2.5 and

IODS 2.5 polymers. All the information up to this point indicates that there is

no difference between the two polymers. Even more surprising is the fact that

the polymers after biodegradation seem to have a higher MW than the original

ones. This apparent discrepancy is very closely related to the fIrst GPC

problem mentioned above, i.e., the errors arising due to some problem in the

column calibration. However, due to limited instrument time at Polymer

Laboratories, it was impossible to repeat the biodegraded set, therefore, no

87


direct comparisons can be made on the two sets of polymers. The trends in the

second set, however, do follow the trends for the starting polymers, i.e., there

is a drop in the molecular weight, becoming very marked with decreasing DS.

88


3.6.1.2 VISCOMETRY RESULTS

Viscosity was the second technique used to obtain molecular weights for the

various CA polymers. This, however, proved to be rather difficult because of

the following problems.

The viscosity can be converted to molecular weights usmg the Mark

Houwink-Sakurada (MHS) equation as described in equation 2.10, Chapter 2.

In order to obtain results from the above equation, however, two constants (K

and a) are needed, which are specific to a particular polymer/solvent system at

a specific temperature. These constants, however, have not been easy to

obtain for most of the systems under consideration in this project.

The first system used was an acetone or acetone/water (70/30 v/v) system.

The original problem with this experiment was that even when acetone/water

mixtures were used, they could not solubilise the whole range of the polymers

in this study. Furthermore, the K and a constants were not available for all the

acetone/water mixtures that were needed, as the literature available only

. . d th I DS al 37 38 40 44-47 mvestlgate e arger v ues . .. . Furthermore, another very

important parameter that was not covered in any of the above papers was the

possible inaccuracies that could arise by using such a volatile solvent as

acetone. With the passing of time, acetonew.ould evaporate from the

viscometer tube and the resulting dilutions would be inaccurate, giving

misleading results. However the MW value for the starting material (DS2.5)

has been calculated in order to compare it with the other MW methods used.

The original results are summarised in Table 3.11.

89


TABLE 3.11

Table of the polymer run times (in seconds) with varying concentrations

(wide viscometer tube, 25°C)

162.24

143.91

126.82

109.22

96.02

371.36

332.13

297.62

264.88

240.09

342.45

310.92

282.20

255.71

235.20

From the flow times above, the relative and specific viscosities can be

calculated for the three polymers. These values are presented in Tables 3.12 -

3.14 and in Graphs 3 - 5:

TABLE 3.12

Table of relative and specific viscosities vs. concentration for the CA with

aDS of 2.5 (acetone, wide tube, 25°C)

3

i

1.1

1.71

1.51

1.30

1.15

0.71

0.51

0.30

0.15

179.43

171.24

151.83

132.82

90

GRAPH 3 _ DS2.5 (acetone, wide tube, 25°C)

200

160

--.. ~ ~#

'" -Q" 160 '" =

140

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

CONCENTRATION E+3 (g/ml)


TABLE 3.13


a DS of 1.7 (acetonelwater, wide tube, 25°C)

TABLE 3.14

1.54

1.38

1.23

1.12

0.54

0.38

0.23

0.12

135.00

127.56

115.28

104.91


a DS of 1.5 (acetone/water, wide tube, 25°C)

1.44

1.31

1.19

1.09

0.44

0.31

0.19

0.09

111.11

103.68

93.98

84.26

It can be seen that in all three cases the value of Ttr has fallen below the value

of 1.2, hence making the more dilute results less reliable than the more

concentrated ones (see also Equation 2.9, Chapter 2).

91

GRAPH 4 _ DS1.7 (acetone/water, wide tube, 25°C)

150

145

140

135

~

-::S 130 ~

'" --=- 125 '" =

120

115

110

105

100 0.5 1 1.5


GRAPH 5 _ D81.5 (acetone/water, wide tube, 25°C)

120

115

110

95

90

85

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

CONCENTRATION E+3 (glml)


In order to compare the results with those from the other techniques used in

this work (viscosity measurements in different solvents, GPC), the MW was

calculated for the DS 2.5 polymer for which the K and a constants for the

MHS equation were known from the literature44.45.4

7 The values quoted in the

literature were empirical1y derived giving the following MHS expression:

[11] = 0.133 MwO.616 Equation 3.13

for CA with a DS of 2.5 at 25°C. The limiting viscosity, [T)], was the intercept

as determined in Graph 3.

The Mw for the DS 2.5 polymer was deduced to be 65,000, and in good

agreement with the results published by Kamide et af4. Below is the

comparison between the various Mw values for the starting material from the

different techniques:

Mw (viscosity, acetone solvent, 25°C) = 65,000

Mw (GPC THF system, polystyrene calibration, 25°C) = 138,000

Mw (GPC DMAlLiCI system, polystyrene calibration, 60°C) = 176,500

Mw (GPC DMAlLiCI system, PEOIPEG calibration, 60°C) = 95,100

Furthermore, the polymers radius of gyration (S2)lI2, which is the root-mean

square distance of the ends of the chain from its centre of gravity can also be

determined44. This gives an indication of how "good" a particular solvent can

be for a given polymer. In a thermodynamically "good" solvent, where

polymer-solvent contacts are highly favoured, the coils are relatively extended.

In a "poor" solvent they are relatively contracted.

92


The radius of gyration for the DS 2.5 polymer dissolved in acetone at 25°C is

expressed as follows:

(S2)1/2 = 7.39 X 10-8 Mw0

308 (cm) Equation 3_14

Substituting the value of Mw as calculated in equation 3.13 the radius of

gyration is deduced to be 2.24 x 10-6 cm which is in good agreement with the

literature value for the given [Tj].

93


The second system that was used was a DMA system. The reason for this was

that it was a better solvent for a wider range of CA polymers and that K and a

constants for a few polymers were quoted in the literature48 (see also Table

3.15 below). From these values, a calibration curve was created which could

help predict any K or a constant as long as the DS of the polymer in question

was known.

TABLE 3.15

Table ofthe MHS constants for given DS values (DMA, 25°C)

.2.50

3.00.

95.8

39.5

26.4

-1.02

-1.40

-1.58

Graphs 6, 6a and 6b illustrate the resulting calibration curves.

0.65

0.738

0.750

94

GRAPH 6 _ MHS CALl BRA TlON CURVE (DMA, 25°C)

-0.6

-0.7

-0.8

-0.9

-1

~ -1.1 ell

0 --1.2

-1.3

-1.4

-1.5

• -1.6

a

GRAPH 6a - GRAPH OF a CONSTANTS vs. DS

0.74

0.72

0.7

~ 0.68

0.66

0.64

0.62

0.6 ~~~~~~~hlill 1 1.5

DS

2

o 0.5

2.5 3

GRAPH 6b _ GRAPH OF K CONSTANTS vs. DS

200

180

160

140

,- 120 e.c ---e -- 100 rf')

+ \;I;l

:;:t:: 80

60

40

20

0 0 0.5 1 1.5 2 2.5 3

DS


The results for the CA with a OS of2.5 are summarised in Table 3.16:

TABLE 3.16

Table of the run times (in seconds) of the CA with a DS of 2.5 with varying

concentrations (DMA, wide tube, 25°C)

····Soly~nt 140.66

0:\ g in 20m1s 289.17 ,?' --'- ';

O,lg in 25mls 260.59

O.lg in 33mls 231.11

d.1g in 50mls 207.34

O.lg in 90mIs 204.52

Converting those results into viscosities, they give:

TABLE 3.17


a DS of 2.5 (DMA, wide tube, 25°C)

4

3

2

1.1

1.85

1.64

1.47

1.45

0.85

0.64

0.47

0.45

212.50

2\3.33

235.00

409.09

The results are graphically illustrated in Graph 7 overleaf.

95

GRAPH 7 - DS2.5 (DMA, wide tube, 25°C)

450

400

300

250

200+-------~~~----~~~

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5



From the results obtained, it is immediately evident that the behaviour of the

polymer in solution is polyelectrolytic, i.e. the mutual repulsion of its charges

causes particularly marked expansions of the chain, giving incorrect results.

This was evident by the attached graphs, as no linear relationship can be

obtained.

To remedy this problem, a small quantity of LiCl (0.5g in 50 mls DMA) was

added to the polymer solution. The addition of this low molecular weight salt

to the solution increased the ionic strength of the solution outside the polymer

coil relative to that inside. This made the chain contract, and therefore,

remedied the polyelectrolytic behaviour observed previously. These

viscosities were ran at 60°C in order to simulate the conditions used in the

GPC work.

Table 3.18, below summarises the results:

TABLE 3.18

Table of the polymer run times (in seconds) with varying concentrations

(DMAlLiCI, wide tube, 60°C)

5mgs;~ml·

4mgs/ml

3mgsIml

45.80

42.05

38.60

41.73

39.06

36.53

Converting these results to viscosities we get:

35.81

34.49

33.23

96 .


TABLE 3.19


a DS of 2.5 (DMAlLiCI, wide tube, 60°C)

5

4

3

See also Graph 8.

TABLE 3.20

1.55

1A2

1.31

0.55

OA2

0.31

109.77

105.51

101.79

87.65

87.66

9001



See also Graph 9.

1.41

1.32

1.24

OA1

0.32

0.24

82.25

80.23

78A6

68.72

69A1

71.70

97

GRAPH 8 - DS2.5 (DMA/LiCl, wide tube, 60°C)

114 , ,

", '

112

' :':

110

----~ 1'1

~ 108 0

"" Q. '" c ... ;.; ,",-

106 ;"., ,

"\;

104

102

1 00 ~ciCG1.:l~~~~~ilil! 3 3.5 4 4.5 5 5.5 6

CONCENTRA nON E+3 (glml)

GRAPH 9 - DS1.7 (DMAlLiCI, wide tube, 60°C)

84

83

82 --,.... --~ -a '"

81

= i . "'1 ~ ,

80

3.5 4 4.5 5 5.5 6

CONCENTRA TION E+3 (g/ml)


TABLE 3.21



5

.4}

See also Graph 10.

1.21

1.17

1.12

0.21

0.17

0.12

42.20

41.60

41.26

38.12

39.25

37.78

As it can be seen from the attached graphs, this last set worked very well. The

only problem was that there were no constants available in order to convert

these results into molecular weights. Therefore, these experiments were

repeated with a narrower viscometer tube and at 25°C, in order to increase the

flow times. The results are overleaf:

98

GRAPH 10 - DS1.5 (DMAlLiCl, wide tube, 60°C)

42.6

42.4

42.2

~ -~ 42 <J --Q.

'" = 41.8

41.6

41.4

3 3.5 4 4.5 5 5.5 6

CONCENTRA TION E+3 (glml)


TABLE 3.22a

Table of the starting polymer flow times (in seconds) with varying

concentrations (DMAlLiCI, narrow tube, 25°C)

6mgs/ml 486.24 450.30 429.23

5mgsl mI 445.50 415.39 398.45

4mgsl mI 402.67 382.46 367.56

3mgs/ml 363.38 345.68 338.36

Note that the IODS 2.5 polymer gave identical results to the starting material.

This was further proof that the CA with a DS of 2.5 did not biodegrade.

TABLE 3.22b

Table of the 10 day biodegraded polymer flow times (in seconds) with

varying concentrations (DMAlLiCI, narrow tube, 25°C)

300.17

287.06

274.57

261.70

279.37

272.01

265.92

258.36

The above results are converted into viscosities as follows:

99


TABLE 3.23


a OS of 2.5 (OMA!LiCI, narrow tube, 25°C)

6

5

4

3

See also Graph 11.

TABLE 3.24

1.91

1.75

1.58

1.43

0.91

0.75

0.58

0.43

151.61

149.93

145.36

142.38

107.85

111.92

114.36

119.22


a OS of 1. 7 (DMA!LiCI, narrow tube, 25°C)

See also Graph 12.

1.63

1.50

1.36

0.63

0.50

0.36

126.28

125.52

119.21

97.72

IOU7

102.49

100

150

148 .:: =-'" = 147

146

3

GRAPH 11 _ DS2.5 (DMAlLiCI, narrow tube, 25°C)

',\.\

.-., ;"

.,..T'

•

6


GRAPH 12 _ DS1.7 (DMA/LiCl, narrow tube, 25°C)

129

128

127

126

- .. ,) '-i-"~" --X 125 -=-' ~ Cl. 124 '" =

,- .,,;,'

123

122

121

120

6 119

3 3.5 CONCENTRATION E+3 (g/ml)


TABLE 3.25


a DS of 1.5 (DMAlLiCI, narrow tube, 25°C)

5

4

3

See also Graph 13.

1.56

1.44

1.33

0.56

0.44

0.33

112.98

110.89

109.63

88.94

91.16

95.06

101

GRAPH 13 _ DS1.5 (DMAlLiCI, narrow tube, 25°C)

114.5

114

113.5

113 ---. ----- 112.5 ~

-a 112 '" =

111.5

111

110.5

110

109.5 3 3.5 4 4.5 5 5.5 6



TABLE 3.26

Table of relative and specific viscosities vs. concentration for the 10 day

biodegraded CA with an original DS of 1.7 (DMAlLiCl, narrow tube,

25°C)

1

.0.5

See also Graph 14.

TABLE 3.27

1.08

1.03

0.08

0.03

78.35

55.61

76.96

59.12

Table of relative and specific viscosities vs. concentration for the 10 day

biodegraded CA with an original DS of 1.5 (DMAlLiCI, narrow tube,

25°C)

1

0.5

See also Graph 15.

1.04

1.01

0.04

0.01

44.38

29.38

39.22

19.90

102

GRAPH 14 - 10DS1.7 (DMAlLiCI, narrow tube, 25°C)

90

88

82

80

1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2


49

48.5

48

47.5 --. -~ ex, 47

~ c. 46.5 'f1

>="

46

45.5

45

44.5

44

GRAPH 15 - 10DS1.5 (DMAfLiCI, narrow tube, 25°C)

1 1,1 1.2 1.3 1.4 1.5 1.6 1.7


1.8 1.9

'''',

H.; . ,

I, ;., ,

•

2


The two biodegraded products due to their very poor solubility, have not given

significant results. For the sake of completion, the starting materials with a DS

of 2.5, 1.7 and 1.5 were also run at the lower concentration but also failed to

give significant results, as the flow times of the solvent and the polymer were

very close to each other.

As already mentioned, there are no MHS constants for the DMAlLiCl system

readily available, so absolute MW values carmot be calculated. However, by

using the calibration curves as detailed in charts 6a and 6b for DMA at 25°C,

relative comparisons can be made for the trends in the MW for the various

polymers. Table 3.28 summarises the results.

TABLE 3.28

Table of the MW values of some starting and degraded polymers

53,000

50,000

8,500

The trends seen above seem to be in agreement with the results obtained by the

GPC DMAlLiCI system in Table 3.9. There seems to be a relatively small

drop in the MW values of the high DS polymers consistent with a loss of

acetate groups. However, there is a dramatic drop in the biodegraded sample, .

which would be consistent with not only a loss of acetate groups (the DS drops

from 1.5 to 0.9 after 10 days), but also with significant chain scission.

103


Furthermore, from the data available in the literature39,42, the radii of gyration

could be calculated for the OS 2.5 and 1.7 polymers using the following

equations:

For the OS 2.5 polymer: (S2)1/2 = 0.68 X 10.8 Mwo.53 (cm) Equation 3.15

For the OS 1.7 polymer: (S2)1/2 = 0.38 X 10,8 Mw052 (cm) Equation 3.16

By substituting the MW values from Table 3.28, the radii of gyration were

deduced to be 2.32 x 10-6 cm for the OS 2.5 polymer and 1.09 x 10-6 cm. This

shows that DMA becomes progressively a "poor" solvent for CA. This is in

agreement with this work. As mentioned earlier in the Chapter, the OS 1.0

and 0.7 polymers were hardly soluble in DMA and no meaningful results were

obtained from their viscosity measurements.

104


3.6.2 NMR SPECTROSCOPIC DETERMINATION

3.6.2.1 PROTON SPECTRA

Goodlett et at9 were the first to develop a method for determining the

distribution of acetyl groups in CAs. The simplest CA from the NMR point of

view is the triacetate, and it is the one that was investigated first. The NMR

spectrum of cellulose triacetate (CTA) consists of two regions of absorption.

The protons associated with the anhydroglucose unit give peaks from 5.10 to

3.250. The acetyl protons give three peaks at 2.09, 1.99 and 1.948. The

observation of three peaks due to the three different kinds of acetyl groups in

CT A imply the potential application of NMR for determining the acetyl

distribution in partially acetylated celluloses. The spectra of partially

acetylated celluloses have in addition to the above mentioned peaks, additional

peaks and a general "filled-in" appearance. The reason for this is that the

partially acetylated CAs have an irregular structure compared to the CTA. The

CT A has the same structural unit repeated throughout the length of the

polymer chain, whereas the partially acetylated CAs do not.

Assigmnent of the three peaks to the three positions of acetyl substitution is

made as follows. Maim et alo reported methods of determining the degree of

substitution of the primary hydroxyl (6-position). By comparing samples with

various degrees of substitution at the 6-position, as determined by reaction

with triphenyl methyl chloride, it was possible to assign the peak at 2.090 to

acetyl groups in the 6-position. The other two peaks were assigned on the

basis of the different reactivities shown by the 2- and 3- position hydroxyl

groups. It has been reportedS1 that the hydroxyl group in the 2-position is

about twenty times more reactive to p-toluene-sulphonyl chloride than the

group in the 3-position. Three reactions were studied - regenerated cellulose

with acetyl chloride, low-acetyl CA with acetyl chloride and low-acetyl CA

with acetic anhydride. It was found that there was a considerably greater

105


reactivity shown at the position responsible for the peak at 1.990 than was

shown by the position responsible for the peak at 1.940. These experiments

also confIrmed the assignment of the peak due to substitution at the 6-position.

Thus, the three peaks at 2.09, 1.99 and 1.940 were assigned to the 6-position,

the 2-position and the 3-position respectively. These fIgures were in good

agreement with work published by Kamide et al52

The spectra obtained for the starting and hydrolysed materials showed a

similar picture to the above and are shown overleaf. They were obtained from

the SRC WH-400 instrument at the University of Warwick. Due to the fact

that all of them were partially acetylated, the spectra displayed a "filled-in"

appearance. Tables 3.29 and 3.30 show the NMR peaks of interest.

106

DS2.5 IH NMR spectrum

n,' ,. " ..

I' .,1\" :1 .. :1:11.m

H ,ll l! ,,,I,' U . t:: ;!:',:,:: IV

"",--"",,--~.""--"""<.~-~.>r----: . ·----~--~.r'--~'~r'--~---'~,--_uO--~'T---o'T'------~'~. --"r'--~'~'---n"------~,~,--~,~,---rr"--~.,,,'------

5 4 3 2

DSI.7 IH NMR spectrum

il , .. I,'

\ '---___ --'A_-""_

5 --.~·--~~---n"----~,,r_--,,·~~.--------_T __ --~.T'----~ .. ,----"'r, --------~,~,----~W---'~T'---------T'~'---,'T'----.'~T'----~--_T'~'--_,'T'--__,'~'T'_----~

2 1 4 3

DS 1.5 'H NMR spectrum

:i ,il Jl ", ~,'

----------------~-----~---~---i i

"-...'-----,~-"--" --~--_n,----~"_--T_--_"~~.--------~ .• ,----.. ~T___oT'--------_n,,_--~~r.---,.T'----~ .. r--------,,~.----~ .. ---1.T.----... 7--------~' •• --_,,~.--__,.'~T'----~

5 4 3 2 I

DS 1.0 IH NMR spectrum

It .... :.'

rr ~~; ;,:: j~i

-'~.--~'T'---T,~,--~,,~-"r---.'T.----~--~--__ .~'--__ .~'T,--------~"----n'.!---.'~'T'--------"~'----'T'----'.T'--__ '~.,----~--"~,----'IT'----rr---~,----~ 5 4 3 2 1

DSO.7 'H NMR spectrum

!I, .. !,'

~ _____ lL _"7.---T'---~--~'r'---c.~,.T ----5----~--~---,"~'T'----4----~"--~,~,---,"~'7'----3~--~,r, ---n"---'''~'T'--~2----~''----'~'---''''---'T'----1

IODS2.5 lH NMR spectrum

:t .il il ... j,'

I'

v i/

V

11 --------~~~~--~~~ _,., .. __ . __ ,..-_,~. ~'''------''''''---5- -",-~,,~- ,,-----,.,.,~--__rr_-_n_- - ____ ~....J '---"---

4 " " ,..----".,..., -;---n-d - ...... -.",------.,,,.--:-----n~~__T~~.,...-- ' 3 '" " 2 d " , , "

IODS!.7 IH NMR spectrum

",,", u"

n ,;i)!:a :j ,,''''~:

il,,,!,'

-=====::::::=::::::::=::::::::::=::::::::::""--..J '--___ --' ....... ---.-"",1 '-....... ....,...-___ JI-A.J

-v

1---------'/

~~,~.---.~,--~--~,~y------~~~.~'--~T--~.'------~,~,--,~~,--~,·~~,------~, .. --~ur--"~,---rr"------~,,r--"r,--~,r..---n,,------5 4 3 2

lODS1.5 IH NMR spectrum

!.l ... I,'

IODS1.0 IH NMR spectrum

','". ,. , ...

" :i ,I' "" !.'

.,.--....,.,,,--_-_'.,.'-~'.,___.T. --~-_,, __ ..,'r' -~,...--.-"~ .. --~-....,.~-_rr--,.,....-..,.r_----...._-_,,_-~,,",--,,,'--------,' ... '--n"-~ . .....---..'.,.. ---5 2

IODSO.7 IH NMR spectrum

11 .. ,1,'

1!;'1; ,::: ~l!

5 3 2


TABLE 3.29

The o-acetyl peaks of the starting polymers (0 days)

2 1.99 9.2

3 1.94 23.6 2.6

1.7 6 2.07 16.8 1.8 !.-:",i· I . ·'l*.:r_·:~jW~':: ~ 2 1.99 9.2 , I

3 1.94 30.4 3.3

6 2.07 9.3

2 1.99 13.3 1.4

3 1.94 18.1 1.9

6 2.07 5.1

2 2.01 12.0 2.4

3 1.94 10.9 2.1

6 ND ND 0

2 2.01 10.0 2.3

3 1.95 4.4

Note: ND is not detectable.

107


TABLE 3.30

The o-acetyl peaks of the biodegraded polymers (10 days)

2 1.99 11.6

3 1.94, 1.96 30.1 2.6

6 ND ND 0

2 1.99 5.9 1

3 1.94, 1.95 10.1 1.7

6 ND ND 0

2 2.01 5.4 1

3 1.94, 1.95 6.S 1.3

6 ND ND

2 ND ND

3 ND ND

6 ND ND

2 ND ND

3 ND ND

From the above, the following conclusions could be made.

Firstly, the resistance to biodegradation of the CA with a DS of 2.5 is obvious

as the spectra at zero time and ten days are identical (see also the relative peak

intensities in Tables 3.29 and 3.30). Furthermore, a statement about the

conformation of the polymer could be made. Frommer et al53 looked at the

NMR spectra of cellulose triacetate in deuterated chloroform at various

temperatures. At room temperature and below, the three acetyl peaks were

10S


clearly separated, and in good agreement with the assigrunents made by

Goodlett et af9. On heating, the peaks for the acetyl protons on the 2-position

and 3-position, merged to a single peak at a temperature of approximately

95°C. With subsequent lowering of the temperature, separation of the two

peaks was again observed. This could be explained by assuming that the

acetyl protons in the two positions became equivalent at above 95°C. At

lower temperatures, each anhydroglucose unit had, therefore, to exist in the

chair form with free rotations of the CrO and C3-O linkages being forbidden

due to steric hindrance. This would make the acetyl protons of the two

positions not equivalent. In order to allow the free rotations of the two

linkages, each anhydroglucose unit had to exist in the boat form at elevated

temperatures. It would be a reasonable assumption to equate this behaviour to

the starting polymer of this work (DS 2.5). From the spectra obtaiJ:!ed it could

be assumed that the anhydrog\ucose units of the CA adopted the chair

conformation.

Secondly, a statement could be made about the hydrolysis of the three ester

groups. In theory, the ester group in the 6-position (the primary site) should

be attacked first. The ester group in position 2 should be attacked second and

the 3-position should be attacked last, as the position 2 ester would be closer

to the (O)-link, and also more exposed than the 3-position (see also Figure 3.1

overleaf).

However, by looking at the intensity ratios in Tables 3.29 and 3.30, it can be

seen that the picture is more complicated and does not always follow the

theoretical trend.

As far as the starting polymers are concerned, position-6 should have been the

most vulnerable position, and therefore one would have expected that it would

109


have the lowest intensity ratio in the series. This was only true for the lower

DS polymers (1.5, 1.0 and 0.7, where in fact the signal disappears

Figure 3.1

Structure of cellulose triacetate.

o

H R

where R : OCOCH3- cellulose triacetate

OH - cellulose

o

H R

n

completely). The two highest DS polymers (2.5 and 1.7) favour the 2-

position. Similar anomalies are observed for the positions 2 and 3. Very

marked is the discrepancy in the DS 0.7 polymer where the peak intensity for

position-2 is nearly two and a half times greater than that of position-3.

As far as the degraded samples are concerned, the ones of interest are the

higher DS ones. As mentioned before, the biodegraded sample with an initial

DS value of 2.5 did not show any difference to its starting polymer. The

position-6 peak disappears from the IODS 1.7 and 10DS 1.5 polymers. As for

the positions 2 and 3 signals, the ratio of position-2 to position-3 is as

expected (i.e. smaller intensity for position-2 than for position-3), however, as

we go down this lODS series there is a further reduction in the position-3

signal, while the signal due to position-2 remains the same. This again is in

contradiction to the theory, as one would have expected that after the

110


disappearance of the position-6 peak, the position-2 peak would have greatly

reduced.

The other peaks on the spectra were due to the solvent (1.87 and 1.890) and ..

the cellulose (2.5 to 5.10).

To conclude, therefore, on both senes, overall the pnmary acetate IS

hydrolysed first, and is the first signal to disappear from the starting series, and

also from the 10DS 1.7 and 1.5 polymers. There seems to be some

discrepancy with the theory about the other two esters, and it is believed that

they are not hydrolysed in the order that was predicted by theory. As for the

biodegraded polymers with low OS (lOOS 1.0 and 10DS 0.7), they have

biodegraded to such an extent that their acetyl content has been completely

removed.

III


3.6.2.2 CARBON SPECTRA

As all the samples displayed a very reduced solubility in the common NMR

solvents, it was impossible to obtain any useful carbon spectra in solution.

Therefore, solid state BC NMR was performed. However, many authors

claimed solubility of CA with various DS values in NMR solvents54-59

. The

conflicting evidence proved once again that the relationship between the

crystallinity and solubility of CA was not fully understood or appreciated.

The crystallinity of CA polymers seems to be closely related to their method

of synthesis. At the beginning of this project, and taking into account the

material available, it was stated that the crystallinity was related to the DS.

The higher the acetyl content, the higher the crystallinity. This however, is not

necessarily correct. As will be shown in the next section, X-ray analysis on

CA with DS values of 2.5, 1.7 and 1.5, shows that these polymers are

amorphous. Therefore, the crystallinity was not merely associated with the

DS.

Ooyle et a('o studied the BC NMR spectra of various CAs in solution and in

the solid state. The synthesis used involved a heterogeneous acetylation of

cotton linters to the appropriate DS. For the solution spectra, DMSO was used

for the CAs with a OS greater than 0.5, whereas a mixture of DMSO and N

methylmorpholine-N-oxide was needed for the cellulose sample and the CA

with a DS of 0.5. The authors found that for CA with a DS value greater than

0.5, irrespective of the DS, the spectrum obtained was that of the triacetate,

and there was no evidence for partially substituted or unsubstituted cellulose,

despite the fact that chemical analysis clearly showed that acetylation was far·

from complete. By comparing their results with results obtained by other

workers on nitration of cellulose, Doyle and co-workers concluded that, using

their conditions, the rates of acetylation were not conn:olled by the reactivity

of the particular sites but rather by accessibility.

112


The authors also compared their DS 2.5 sample with a commercially available

sample of the same DS and found the two to be different. They further

concluded that the commercial sample was generated under such conditions

that the basic cellulose structure was destroyed and the observed DS of 2.5

was achieved by subsequent hydrolysis of the triacetate, which is indeed the

method used commercially. The commercial sample spectrum displayed both

substituted and unsubstituted cellulose features.

They finally concluded that the initial acetylation in their samples occurred in

the disordered accessible regions of the cellulose. Once this had been

completed, further acetylation occurred in the ordered regions without them

losing their integrity. This meant, therefore, that the solutions subsequently

obtained, contained dispersions of these ordered regions and hence the spectra

corresponded to the solubilised part of the cellulose structure only. An

important implication was, therefore, that DMSO and N-methyimorpholine-N

oxide might not be true solvents for cellulose, but rather were capable of

achieving dispersions of the ordered regions. It was thus possible that such

regions survived to a large extent in fibrous CA and that the solution spectra

corresponded to solubilised surface groups which occurred in amorphous

areas.

The solid state l3C NMR spectra were obtained from the EPSRC solid-state

NMR service at the University of Durham and are shown overleaf.

113

OS 2.5

Date File Pulse

Sep 23 94 dataOl/jvd23sep9403 sequence xpolar

Observe C13 Frequency 15.430 MHz Spectral width 30001.5 Hz Acquisition time 20.3 ms Relaxation delay 5.0 sec No. repetitions 120

Cross polarization COntact time 3.00 ms Spin-rate 3600 Hz

Gaussian broadening 0.008 sec FT size 32768 Ambient temperature

'2£0' ., i •• 240

, • I' , 220

• i I i i • I i , • i I i •

200 180 160 'i10' i • I " 140

-o -

i' L i i

100 • I i

80 , i ' 60

i I ' 40

'i' .20

, • I m

DS 1.1

Date File Pulse

Sep 23 94 data01/jvd23sep9409 sequence zpolar


Cross polarization Contact time 3.00 ms Spin-rate 3880 Hz


"I' , , , I " i i I i i

260 240 220 f i i I i i , 'I' ,

200 180

'" "' '"

i 'I i i

160

"' .... ": ~

0 ~ ~ ~

'" ..; ~

" , , I ppm

"I' , 120

i i I i, 100 "I" 140

'I i

80 , I ' 60

i I i

40 i I i

.20

DS 1.5

Date File Pulse

sep 23 94 data01/jvd23sep9401 sequence xpolar




iI I ' , o"n

,. I i i

o~n i i I i i

oon i I i i

onn , 'I i ,

,an " I ' , 160

., I" 140

"I' , 120

~

o

,. I" 100

• I i • I i

BO 60

\

m 'I i

40 • I ' , 'I

.20

OS 0.1

Date sep 23 94 File dataOl/jvd23sep9404 PUlse sequence xpolar

Observe C13 Frequency 15.430 MHz Spectral width 30001.5 Hz Acquisition time 20.3 ms Relazation delay 5.0 sec No. repetitions 240



'2t6 . •• I' • 220

• •• I I i i

240 " I • i

200 I· ,. I"

180 "I' • 160

N

'" "! N ~ N N ~

.; ~

i i' i I ppm

,. I" • I 140

i i I' i

120 •• I i i

100 " i I'

80 i I' 60

i I i

40 i I I

.20

10 os 2.5

Date FUe PUlse

sep 23 94 data01/jvd23sep9402 sequence xpolar




" I •• 260

i i I " 220

" I • , 240

i i i' i

180 '260'

'" o o

" i i ,

160 i' i'" 'i • i i i •

140 120

~

o

, • i ii • I

100 • I' 80

• I ' i I i

60 40

"' q

'" o N

• I ' 20

i i, i I ppm

10DS1.1 packed ln talc TOSS

Date Sep 23 94 File dataOl/jvd23sep9408 Pulse sequence xpolar

Observe C13 Frequency 15.430 MHz Spectral width 30001.5 Hz Acquisition time 20.3 MS Relaxation delay 5.0 sec No. repetitions 160

Cross polarization Contact time 3.00 MS spin-rate 2630 Hz


'2~O ' i'l I. 240

i • I i' 'I 220 "I" 200

I i I" 180

,i I I i

160 i i i" ,I 140 'i~o'

~ ~ ~

i i f I, 100 'I' 80

~ N C

~ ~

i I' 60

i I' 40

~ ~

'" C N

i I' 20

•••• I ppm

10 OS 1.5 pacJced in talc

Date sep 23 94 File dataOl/jvd23sep9406 Pulse sequence %pOlar

Observe C13 Frequency 75.430 MHz Spectral width 30007.5 Hz AcquiSition time 20.3 ms Relaxation delay 5.0 sec No. repetitions 240


Gaussian broadening O.OOB sec FT size 3276B Ambient temperature

., l' • 260

•• , f i'

240 i i l' i

220 •• I i •

200 i • 1 " 180

m -~ -... -

, {la " • i i i 1 i i •• I' , 140 120

N o -

i' I' i

100 I i

m ~ o N

"

• 1 i

80 • " , I ppm

• l' 60

i I i

40 i j i

20

10 DS 0.7

Date Sep 23 94 Flle dataOl/jvd23sep9401 PUlse sequence zpolar

Observe C13 Frequency 75.430 Milz Spectral wldth 30007.5 Hz Acquisltion time 20.3 ms Relaxation delay 5.0 sec No. repetitions 640



2 0 240 220 200 180 160 140 1 0 100 80 60 40

o ~ ... o N

20 ppm


Table 3.31 below shows the l3C chemical shifts for the starting CAs and the

biodegraded samples in the solid state.

TABLE 3.31

The I3C chemical shifts for the starting CAs (0 days) and the biodegraded

samples (10 days) (given in ppm using tetramethylsilane as the shift

reference compound)

(15) (--------- 62 --------- ---------) (25)

2.5 171.0 101.3 ND 73.5 62.6 20.9 "',""

(15) (--------- 61 ---------- ---------) (24)

171.3 102.0 ND 73.5 63.2 21.0

(14) (--------- 64 ---------- ---------) (22)

171.3 102.0 82.0 73.3 63.0 20.9

(12) (--------- 62 ---.------ ---------) (26)

171.3 102.0 ND 73.5 63.3 21.0

(13) (--------- 66 ---------- ---------) (21)

171.5 102.1 82.0 73.3 64.2 20.9

(12) (--------- 66 ---.------ ---------) (23)

171.7 104.5 82.9,81.0 73.3 64.6 21.0

(12) (--------- 68 ---------- ---------) (20) .

10ns 0.7 171.2 ND ND 72.9 ND 20.8

Note that the numbers in brackets represent the relative intensities of the

peaks.

Several points can be made about the results obtained..

114


Firstly, some spectra, and especially 10DS 0.7 were noisy. This was due to

the small quantities of sample that were available for analysis. Normally 0.5g

is needed for a good solid state l3C NMR spectrum. However, in some cases,

only 0.3g or less was available.

Some of the samples, despite having been ground prior to despatch to Durham,

were not fme enough. The fact that they were too hard to grind further was an

added problem. In order to obtain a spectrum for these polymers, they were

packed in talc to help spinning. In these cases the spinning was still relatively

slow (2600 Hz compared to 3600 Hz) so a sideband suppression sequence

(TOSS) had to be used to improve resolution ..

It is also important to note that the relative intensities of the peaks as detailed

in Table 3.31 can be compared sample to sample, but the intensities within one

spectrum do not necessarily have a 1: 1 correlation with the number of carbons

they represent. This is due to the cross-polarisation technique used to obtain

the spectra. The technique involves the magnetisation of the protons adjacent

to the carbons being transferred to the appropriate carbons. Therefore, if there

are more protons around a particular carbon atom, then the intensity will be

higher than if there are fewer or no protons close to the carbon atom.

By taking these factors into consideration, the DS 2.5 material does not show

any marked differences before or after biodegradation (see also Table 3.31).

This confirms the previous evidence as to its resistance to biodegradation. The

biodegraded DS 1.7 and 1.5 polymers, however, do show some changes from

their starting compounds. In both cases a C-4 peak appears. Taking into

account that crystalline cellulose displays a peak at about 88ppIn, this could

imply that the biodegraded polymers may have acquired some crystalline

content and that some de-acetylation had occurred. The fact that the DS 0.7

polymer features a peak at tlie same resonance would also emphasise the point.

115


The biodegraded polymer with an initial OS of 0.7 did not give any significant

results as the background noise was extremely marked due to the lack of

material present in the rotor.

Furthermore, the starting polymers show a decreasing trend in the intensities

of the carbonyl and methyl peaks. This is the expected trend, as there is a

decrease in the acetyl content with decreasing OS. Furthermore, there is an

increase in the C-l, C-2,3,5, C-6 and importantly, C-4 intensities. This again

shows that the crystalline content increases with decreasing OS.

As far as the biodegraded samples are concerned, the decreasing carbonyl

trend continues, as does the increasing crystalline content. There seems to be

a discrepancy in the methyl intensities, however, as there seems to be a rise for

the lOOS 1. 7 polymer before it decreases again for the IODS 1.5 polymer.

This might be due to the reasons noted before, as to the use of talc in order to

spin the samples and having to use a slower spin rate and also to the fact that

some corrective software had to be used (TOSS). These factors could have

made direct comparisons less accurate.

116


3.6.3 X-RAY DIFFRACTION ANALYSIS

The X-ray diffraction spectra were obtained from a Philips PW 1130

generator, using Cu a radiation (1.=1.5418 A), coupled with a Hiltonbrooks

motor drive.

The X-Ray spectrum for theSA with a DS of 1.5 was typical for the higher

DS polymers (DS 2.5, 1.7 and 1.5). These polymers were amorphous.

However, as the DS decreased, an increase in the crystallinity was observed.

This was due to the fact that as the acetyl content of these low DS polymers

decreased, a more "cellulose-like", semi-crystalline structure was adopted.

This behaviour was also indicated by the 13C NMR spectra of the lower DS

polymers as well as for the biodegraded ones. The X-Ray spectra are shown

overleaf.

The increased crystallinity would make biodegradation more difficult, i.e. the

time needed for these polymers to biodegrade would be longer than if no

crystalline content was present.

]]7

c o u n t s

Sample: c:\Sie122\datalds15 • 03/04/95

DS1.5 X-Ray spectrum

5 10 15 20 25 30 Degrees 2-Theta

c o u n t s

Sample: c:\Sie122\data\ds1 * 03/04/95

DS1.0 X-Ray spectrum

5 10 15 Degrees 2-Theta

4.37A

4.14A

4.01A

3.38A

20 25 30

Sample: c:lsie122ldatalds07 * 03/04/95 C 600-r~~~~~~~----~~~---------------------------------------------------------.

o u DSO.7 X-Ray spectrum n t s

600 4.36A

4.65

'3.69 '3.29

400 '5.

071 5.4 3.36

3.0

200 6.67A

O~'-,,-.-r.-'-,,-.-r.-'-,,-.-r'-'-,,-.-r'-"rT-.-..-,,-.-.-.'-.-,,-'-r'-'-,,-.-r.-~

5 10 15 20 25 30 Oeg rees 2-Theta

C 800 _~s~a~m~p~le~:~c:~~~ie~1~22~~~at=a~~~el~lu_IO~s __ ' __ 0~3~ro_4_~_5 __________________________________________________________________ -,

o u Cellulose X-Ray spectrum n t s

600

400

200

O~'-"-'-r.-'-,,-'-.'-'-,,-.-''-r-,,-.-''-,,-'-r.-.-,,-'-r.-.-,,-'-r.-.-,,,,-.-''-~

5 10 15 20 25 30 Degrees 2-Theta


3.7 GENERAL DISCUSSION

By summarising the results obtained from all the techniques used in this work,

the following observations can be made regarding the microbial degradation of

CA.

From the examination of discarded cigarette filters and the controlled seed tray

experiments, it was clear that the CA with a DS of 2.5 used in the cigarette

filter manufacture was very recalcitrant to biodegradation. Even under

optimum conditions of temperature, humidity and exposure to light, nearly a

year was needed for the filters to show a change in their appearance (they

turned green due to algal growth - see also photographs 23 and 25-29). By

using this algal growth on new filters, some drop in the DS was observed but

this was only relatively small (0.35 of a DS unit in 7 months, see also Table

3.1). After this time, no great change was registered in the DS values (see also

Table 3.2).

This seems to be in disagreement with the rest of the evidence and especially

the biodegradation studies involving the fungus Aspergillus jllmigatlls in

which no change in the DS was recorded after 10 days. This might be due to

the following reasons. Firstly, in the titration method which was used to

determine the DS values for the algal growth experiments, the cigarette filter

was used whole in the analysis, i.e. no re-precipitation from solution was

necessary. As the DS value is a mean value, the possibility could exist where

some small part of the filter with a slightly smaller initial DS value than 2.5,

could have been attacked by the algae, and shown in the titration. In the case

of the fungal biodegradation studies, the CA powder had to be re-dissolved in .

glacial acetic acid and then re-precipitated from the appropriate solvent. The

reason for this was two-fold. Firstly, the addition of the acid stopped the

action of the fungus. Secondly, it helped in the isolati~m of the CA from the

other mixture components (mainly the fungus), especially as it was shown in

118


Chapter 3 that the fungus coated the powder. In the re-precipitation phase, it

might have been possible that a very small amount of degraded polymer

remained in the liquid phase. This, however, would have been a very small

part, as from the 5g of CA initially used, less than O.3g were not recovered. It

is also important to note that by repeating the fungal experiment with a CA

with a DS of 2.5 for a longer period of time (I month), the polymer exhibited

the same kind of behaviour. One must also take into account that in the algal

experiments, the action of the algae seemed to stop after the initial drop in the

DS. This might be attributed to the fact that the enzymes in the algae were

only able to cause a mild de-acetylation but not proceed any further.

These observations lead to the thought that the only way to degrade

biologically CA polymers with a high DS content was via a consortium of . .

nucro-orgarnsms.

The recalcitrance towards biodegradation of the CA with a DS of 2.5 was

further demonstrated by the lack of any degradation after the addition of

simple sugars and aminoacids which are themselves very easily biodegraded.

This was to be expected, however, as the plasticiser used in the filter

manufacture, triacetin [(CH3C02CH2hCH(02CCH3)], should act in the same

way as the other additives used.

Finally, by looking at the biodegradation of chemically synthesised CA

polymers with lower DS values, a more accurate picture of the CA

biodegradation was obtained.

It became immediately obvious that the DS was a very significant factor in the

biodegradability of these polymers. The lower the DS the easier the

biodegradation (see also photographs 39 to 44).

119


There was one factor that caused surprise and that was that, contrary to

infonnation received by the tobacco industry, the higher DS polymers were

amorphous and the crystallinity increased with decreasing DS. This was

demonstrated by the X-Ray spectra as well as by the I3C NMR spectra with

the appearance of the C-4 peak. This meant that the predominant factor that

hindered biodegradation was the DS alone. It was originally thought that

crystallinity in the high DS polymers played an important role in the reduced

biodegradability .

The FT-IR technique made the DS detennination easier and faster than the

chemical titration method. The other problem with the titration method was

the fact that below a certain DP, the method broke down and gave incorrect

results. With the FT-IR technique, the DS of any CA polymer could be

determined with the aid of the calibration curve as shown in Graph 1, Chapter

3.

The NMR work also gave an insight in the biodegradation mechanism. From

the theory, it was expected that the primary 6-position would be more

vulnerable to hydrolysis, followed by the position-2 and finally by position-3.

The actual removal of acetates proved to be more complicated, but in general

tenns, the primary 6-position was more likely to be hydrolysed, and in fact, it

disappeared completely from the polymer with a DS of 0.7 as well as from the

biodegraded polymers with initial DS values of 1.7 and 1.5.

However, there seems to be an ongoing debate about the relationship between

synthesis conditions, DS, water solubility and chemical structure in the CA

molecule27• 43, 50. It is quite clear that the method of preparation of the CA

polymers is a very important factor as far as the chemical structure is

concerned. Although, for instance, the position-6, as already mentioned,

should be the easiest to hydrolyse, this is still dependent upon the amount of

120


water present during the hydrolysis. If the amount of water is below a certain

level, then an alternative position (2 or 3) will be favoured5o. Furthermore,

according to Kamide et ap7, the low DS polymers are only water soluble if the

three positions are roughly equally substituted. This, on the other hand, is

hotly disputed by Buchanan et a(l3, who cannot see such a behaviour in his

polymers. However, as the polymers prepared in this work have not shown

any water solubility throughout the DS range, and the substitution ratios

between the three positions, is not equal, then the behaviour according to

Karnide seems to be closer to the polymers synthesised in this work.

As far as the molecular weight work is concerned, solution viscosity was the

technique that gave better results. There were two main problems with the

GPC work.

Firstly, the two sets of polymers (starting polymers and biodegraded

equivalents) were not directly comparable. This was partly due to the fact that

the calibration curve for the biodegraded polymers was not as reliable as the

one for the starting ones. Therefore, only trends within a set were possible.

Even so, it became obvious that the chemical hydrolysis caused some de

polymerisation as well. This was not particularly marked in the higher DS

polymers but became very important in the lower DS polymers (DS 1.0 and

0.7).

This, however, might have been also partly attributable to a well-known

problem with the GPC technique, i.e., the column - polymer interactions

especially in the presence of "poor" solvents. As the DS decreased, DMA

became progressively a "poorer" solvent. This meant that the polymer might

associate with the column packing causing it to elute later than expected and

hence give lower molecular weights than expected. Even if this were the case,

however, the drop in the DP would still be very large.

121


This was also confirmed by the solution viscosity work. By looking at the

radius of gyration of some polymers in various solvents, it was confirmed that

the lower the DS the poorer the solubility.

It also became obvious by looking at the molecular weights that the

biodegraded polymer with an initial DS of 1.5 (which was investigated in

greater detail) had not only de-acetylated (final DS was 0.9 as determined by

FT-IR), but it had also significantly de-polymerised (initial MW was 50,000

compared to 8,500 after biodegradation).

After all the evidence was assessed, some statement about the mechanism of

biodegradation could be made. It was obvious that the starting polymer and

the biodegraded one were very different. It was also evident that a consortium

of micro-organisms was responsible for this biodegradation. An esterase was

needed to de-acetylate the polymer and a cellulase was present to de

polymerise it. From the evidence to date one cannot be absolutely certain as to

the precise mechanism. However, as the fungus does not biodegrade cellulose,

one can assume that the esterase must have acted first. Once the DS drops

below a certain level, then enough room is created around the chain for the

cellulase to attack. The fact that the CA with a DS of 2.5 does not biodegrade

can be attributed to some steric phenomenon, whereby there is not enough

room for the esterase to attack.

This is in agreement with common microbiological knowledge that in such

cases, there should be at least two neighbouring unsubstituted glucose

molecules before biodegradation can commence. This also agrees with the

fmdings in this work that biodegradation is more difficult for the high DS

polymers.

122


In order to quantify the probability of having two adjacent unsubstituted

glucose molecules in the CA chain, a mathematical model has been devised in

association with the Mathematics Department of the University and it is

described below:

Consider a sequence ofn molecules arranged in a straight line. Each molecule

has initially k substituents. We assume that as time progresses substituents

leave molecules at random. This implies that the exit time of each substituent

is an independently distributed random variable with a common function F(t).

Let YI, Y2, ....... ,Yk be the exit times of the k substituents in a molecule. Then,

the molecule is void of substituents at time t, if X = max {YI , Y2, ......... ,Yd ::; t.

Hence the probability of a molecule being void at time t is given as

P(t) = Pr(X::; t) = [F(t)t Equation 3.17

It may be argued that if XI, X2, ....... ,xn are the times at which the 1st,

2nd, ....... ,nth molecule become void, respectively, then these times are

independent random variables with distribution function P(t) as in equation

3.17.

Now let Pn(t) be the probability that up to time t there have not been two

adjacent void molecules. Clearly, P n(t) is the distribution function of the time

at which two adjacent molecules of a sequence of length n become void. The

problem of obtaining an expression of P,,(t) in terms of t, n and k is similar to

those studied by Mott et af'3 in a different context.

123


As the last molecule must be either void or not void, we defme

so that

A/I(t) = Pr (no adjacent voids and last molecule is void)

B/I(t) = Pr (no adjacent voids and last molecule is not void)

Equation 3.18

Now if the last molecule is void, the sequence of the first n-i molecules must

have no adjacent voids and the molecule at the (n-i) position must not be void.

Thus

Equation 3.19

To obtain an equation for Bn(t) we note that as the last molecule is not void the

first (n-i) molecules constitute a sequence of length (n-I) with no adjacent

voids and no restriction on the molecule at the (n-i) position. Thus we have

B,,(t) = [J-P(t)] P,,-1(t) Equation 3.20

Combining equations 3.18, 3.19 and 3.20 we fmd that

P,,(t) = P(t) Bn_1(/) + [I-P(t)] P,,-1(t) =

= P(t) [J-P(t)] Pn-2(t) + [I-P(t)] PII-dt) Equation 3.21

This is a second order difference equation with initial conditions .. ,.

Plt) = 1 - [p(t)f and P3(t) = [J-P(t) + [1-[P(t)]2 P(t)] Equation 3.22

124


The solution is as follows

( jn

(ffi+3P J I -2P+2p2 I - -1P! I -l+P l ~_1+3Pp_l+P_~_1+3P)

-l+P -l+P +----~====~======~------~====3_---

~_ 1+3P(~_ 1+3P P-l+P-~- 1+3PJ -l+P -l+P -l+P

Equation 3.23

Please note that in the above equation P(t) is denoted as P.

The exact dependency of Pn(t) on t will be taken into account according to the

conditions imposed on the distribution function F(t) of the substituent exit

times, while the dependency on k, the number of substituents in a molecule, is

a consequence of P(t) = [F(t)t (see also equation 3.17).

Under the assumption made above about the exit times being random, the

following relationship also holds

F(t) = 1- e(·/(!..)

which is an exponential distribution function of the parameter IvO. Empirical

evidence with the biodegraded polymer with starting DS of 1.5 suggests that

125


the exponential distribution model is valid, but there may not be a constant

value for 'A across all polymers with any DS.

The graphs overleaf depict the graphical representation of equation 3.23

implemented with the above exponential distribution function for n = 300, 'A =

5 and k = 0.5 to 3.0 in 0.5 increments. The x-axis is in arbitrary units of time

and the y-axis is the probability of two unsubstituted glucose units being

adjacent to each other. The probability axis runs from 0 (certainty) to 1 (no

unsubstituted glucose units adjacent to each other). As expected, the lower the

DS, the sharper the slope, and hence, the faster it is going to be for being

celtain that two unsubstituted glucose units will be adjacent.

In chemical terms, therefore, n depicts the DP, k depicts the DS and 'A an

arbitrary constant that is a measure of the time required for two adjacent

glucose units to be unsubstituted. These figures agree well with the empirical

evidence, particularly when k<2. For larger values of k it might be necessary

to use larger values of 'A.

Table 3.32 below illustrates graphically the time needed for two adjacent

glucose units to become void of substituents.

Table 3.32

Table of the probability of baving two unsubstituted adjacent glucose units

vs. DS.

It is obvious from the above table that a CA with a DS of 0.5 will degrade 37

times faster than the CA with a DS of 2.5.

126

Probability

1

0.8

0.6

0.4

0.2

k 0.5

-!--~--~~--~~~--~~~--~~------~-----'~-----'~4 time o 0.2 0.4 0.6 0.8 1 1.2 1.

Probability

1

0.8

0.6

0.4

0.2

k 1.

,

-±o------~0~.2~--~0~.4~--~0~.~6~~~0~.~8~----~1------~1~.~2----~1~.4 time

Probability lr--__

0.8

0.6

0.4

0.2

k = 1. 5

-b0------70L.=2----~0~.~4~~~0~.~6~~~0~.~8------71------~lL.~2==::==1~.4 time

Probability k = 2. lr-----------__ __

0.8

0.6

0.4

0.2

time o 0.2 0.4 0.6 0.8 1 1.2 1.4

Probability

lr-----------~k~~2~.~5 __ __

0.8

0.6

0.4

0.2

~0------nO~.2;---~0'.~4~--~0~.<6~--~0~.~8------~1------7l~.~2----l~~.4 time

Probability k = 3 .. lr-----------~~~--------__ -

0.8

0.6

0.4

0.2

CHAPTER 4

CONCLUSIONS

Chapter 4 - Conclusions

It is important to emphasise that when this project was started, there was no

reliable information about any of its aspects. There were references claiming

that biodegradation of CA with a DS of 2.5 was feasible and others claiming

the opposite (see also Chapter I).

The problem of the biodegradation of CA in the form of cigarette filters,

however, had never been properly addressed. It is also important to

emphasise that cigarette filters represent a system in which biodegradation will

be minimised because their geometry produces a maximum volume with a

minimum surface area.

Even the CA producing industries did not know much about their product. CA

is a compound that had been produced in a certain way from the beginning of

the century and few changes have taken place since. After extensive talks and

meetings with various manufacturing teams, it was evident that the only factor

addressed in the production of CA was its solubility in acetone. If the CA was

acetone soluble, it could be spun readily and, therefore, it could be used in the

filter manufacture. In addition, no attempt had been ever made by the tow

suppliers to try and produce CA with a lower DS than 2.5. As it has been

shown in Chapter 3, however, the biodegradability of CA mcreases

dramatically with the decrease in the DS.

It was therefore necessary to start at the most elementary level and work up to

the project at hand. The first step was to take discarded cigarette filters and

determine whether they showed any signs of biodegradation. This proved to

be a very important part of the project as it gave us useful pointers regarding.

the shortcomings of the filter manufacture and as to the ways where it might

be improved.

127


The fact that the filters proved to be sterile inside, was a very surprising fact.

These were cigarette filters that looked visibly distressed, and had been lying

on the ground for a rather long period of time. It was obvious, therefore, that

even if the CA with a DS of 2.5 was biodegradable, the properties of the filter

made it impossible for micro-organisms to penetrate it and to initiate

biodegradation.

As the project progressed, it became obvious, however, that the starting

material was recalcitrant to biodegradation, and every possible method of

initiating a biodegradation process gave negative results. This became

particularly clear when various additives were tried and all failed to improve

the biodegradation potential of the CA. These experiments, however, were

expected to fail, as the cigarette filters have already incorporated in them an

excellent additive, triacetin. This is the plasticiser which makes the CA fibres

hard. It is readily biodegradable, and, should this have been a viable route,

triacetin would have been an appropriate choice.

The long term biodegradation experiments (seed tray experiments) also proved

that the CA did not biodegrade easily. The conditions for these experiments

were chosen in such a way as to idealise the conditions under which any

microbial attack would take place. The humidity was kept high, the

temperature ranged from approximately 25°C to 37°C and for those that took

place in the greenhouse, although the temperature could not be controlled as

precisely as in the other incubators, there was also the presence of the

sunshine to aid the process.

Indeed, the fust thoughts were that the higher temperatures would give the

expected result. However, as it turned out, the sunshine was the important

factor for seeing some change in the filters. Algae started growing on the

filters and experiments were repeated with algae injected in them to see

128


whether some accelerated trials could take place. However, even then, the DS

of the filters did not drop to any significant level.

By monitoring also the surfaces on which filters could be discarded, a further

set of conditions was investigated. However, the findings showed that the

surfaces most likely to accommodate discarded filters (roof tiles-simulating the

road and sand-any soft but rather dry surface) did not favour biodegradation at

all. These experiments were discarded after 12 months as no signs of

biodegradation were observed irrespective of the level of light or humidity and

temperature.

It became then clear that in order to obtain more favourable biodegradation

conditions, lower DS samples would have to be investigated. This proved to

be a challenge, as all methods available for producing such materials were

only useful in the laboratory and would be quite useless in an industrial

environment, where even the smallest deviation from the current working

practice is sometimes impossible to implement. Furthermore, the DS

determination proved to be quite difficult, as the method originally used

(provided by the tobacco industry) proved to be incorrect, and did not make

sense chemically.

However, once these methodology problems were sorted out, vanous CA

samples with different DS values were prepared using the existing

manufacturing recipe as the basis.

In order to test these new CA samples for biodegradability, a naturally·

occurring fungus was used, which was isolated from the micro-organisms

growing on the outside of discarded filters. By a series of isolations and

purifications, a fungus (Aspergillus Jumigatus) was isolated and was used

throughout for the biodegradation experiments. These experiments proved

129


again that the starting material was very recalcitrant to biodegradation. As the

DS dropped, the biodegradability increased.

The fact that during the biodegradation, a considerable amount of polymer was

not recovered, should also be taken into account. This could mean that part of

the polymer chain is cleaved to such an extent that it is reduced to

oligosaccharides with various degrees of acetylation. However, for chain

scission to happen, it was generally thought that two unsubstituted glucose

groups adjacent to each other were needed. A mathematical model was

devised (see also Chapter 3) which quantified the probability for such a case to

be true with varying the DS. As it was expected, the probability of two

unsubstituted glucose units being adjacent to each other was significantly

increased with decreasing DS.

Bearing all these matters in mind, a possible mechanism for the biodegradation

of CA was proposed. An esterase and a cellulase would be both needed in

order to biodegrade CA. The esterase would act first in order to reduce the .

number of acetyl groups and once the adjacent unsubstituted glucose units

mentioned above were created, then the cellulase would cleave the chain. By

looking at the information obtained for the biodegraded sample with an initial

DS of 1.5, this mechanism certainly holds. Its final DS, as calculated from the

FT-IR data, is 0.9. Although the drop in the DS is not very large, its DP has

dropped quite dramatically as demonstrated in Table 3.28, Chapter 3. This is

also consistent with the fact that a chemical DS determination was impossible,

hinting to the fact that the DP must have dropped below a limit which makes

the method unreliable. However, one could still argue that degradation was .

also helped by the drop in the DP in the original polymers. Although this

possibility cannot be ruled out completely at this stage, it is thought as rather

improbable that a relatively small drop in the DP (as for instance in the case of

the polymer with a DS of 1.5) could have such an effect on biodegradability".

130


RECOMMENDATIONS FOR FUTURE WORK

This was a very exciting project and there are still many avenues open to

examination.

Further biodegradation studies should be undertaken, in order to monitor the

behaviour of CA polymers with various OS values over different lengths of

time. This could also give a better insight as to a more exact mechanism. It

would also be advantageous to look at a greater number of micro-organisms in

order to find an optimum system for biodegradation. The work should

concentrate at consortia rather than on single enzymes.

Furthermore, various methods of synthesising CA polymers with high OS

values should be investigated, as from some literature references it becomes

obvious that methods of synthesis and chemical structure are closely linked. It

might be the case that by starting from cellulose and acetylating directly to the

required OS value rather than going to the triacetate and then hydrolysing back

to the required OS, might improve the products biodegradability.

A solvent for the whole series of CA polymers would also be advantageous,

especially if the MHS constants were known. This could give a more absolute

molecular weight determination rather than relative figures. More work on the

GPC technique would also be advantageous (see also point below).

Finally, a coupled column chromatographic method giving information on

both the OS and the molecular weight of the polymer by using size exclusion

chromatography (SEC) and high performance liquid chromatography (HPLC) .

would be a distinct advantage. Such a method is currently under investigation

with other polymers and would be of tremendous value if used on the CA

131


polymers. The main problem again with such an investigation might be the

need to find a compatible solvent with the SEC and HPLC columns.

132

BIBLIOGRAPHY

1. Ward, K. Jr., Cellulose and cellulose derivatives (Volume V, Part I),

edited by Ott, E.; Spurlin, H.M. & Graffin, M.W., pp 15-17, Interscience,

New York (1954).

2. Hebeish, A. & Guthrie, J.T., The chemistry and technology of cellulose

copolymers, pp 3-11, Springer Verlag, Germany (1981).

3. Trubak, A. & Sakthivel, A., Solving the cellulose enigma, Chemtech, 20,

444-446, July 1990.

4. Colvin, J.R., Studies shed more light on cellulose structure, Chemical and

Engineering News, 63, 40, September 23, 1985.

5. Atalla, R.H. & Vander Hart, D.L., Native cellulose: a composite of two

distinct crystalline forms, SCience, 223,283-285, January 20, 1984.

6. Clark, J., New thoughts on cellulose bonding; , Tappi Journal, 67, 82-83,

December 1984.

7. Meyer, K.H. & Misch, L., Positions des atomes dans le nouveau modele

spatial de la cellulose, Helv. Chim. Acta" 20, 232 (1937).

8. Higuchi, T., Biosynthesis and biodegradation of wood components, pp

124-128, Academic Press Inc., New York (1985).

9. Delmer, D.P., Cellulose biosynthesis, Ann. Rev. Plant Physiol., 38, 259-

290 (1987).

10.Ross, P.; Mayer, R. & Benziman, M., Cellulose biosynthesis and function

in bacteria, Microbiological Reviews, 55(1), 35-58, (1991).

II.Hogetsu, T. & Shibaoka, H., Effects of Colchicine on Cell Shape and on

Microfibril Arrangement in the Cell Wall of Closterium acerosum,

Planta, 140, 15-18 (1978).

12.Aloni, Y. and Benziman, M., Cellulose and other natural polymer

systems, edited by R.M. Brown Jr., pp 341-361, Plenum, New York

( 1982).

13.Machlachlan, G.A., Cellulose and other natural polymer systems, edited

by R.M. Brown Jr., pp 327-339, Plenum, New York (1982) ..

133

14.Sookne, A.M. & Harris, M., Cellulose and cellulose derivatives (Volume

V, Part I), edited by Ott, E.; Spurlin, H.M. & Grafflin, M.W., pp 197-207,

Interscience, New York (1954).

15.Yarsley, V.E., Cellulosic plastics, cellulose acetate, cenulose ethers,

regenerated cellulose, cellulose nitrate, pp 11-51, lliffe, New York (1964).

l6.Howsmon, I.A. & Sisson, W.A., Cellulose and cellulose derivatives

(Volume V, Part I), edited by Ott, E.; Spurlin, H.M. and Grafflin, M.W.,

pp 236-240, Interscience, New York (1954).

17.ReverJey, A., Cellulose and its derivatives (Chapter 17), by Kennedy, I.F.,

pp 211-225, Ellis Harwood Ltd.

18.Lowe, H.M. & White, I.N.T, Degradation of cenulose acetate in the

environment, Courtauldsfilter tow publication, CORESTA (1991).

19.Rhone-Poulenc Rhodia filter tow, Degradability of cenulose acetate

filters, (1991).

20.Ho, L.C.W.; Martin D.D. & Lindemans, W.e., Ability of microorganisms

to degrade cellulose acetate reverse osmosis membranes, Appl.Environ.

Microbiol., 45(2),,418-427 (1983).

2 1. Buchanan, e.M.; Gardner, R.M. & Komarek., R.I., Aerobic

biodegradation of cellulose acetate, J. of App. Pol. Sci., 47, 1709-1719

(1993).

22.Northrop, D.M. & Rowe, W.F., Effect of the soil environment on the

biodeterioration of man-made textiles, Journal of forensic sciences, 30(3),

602-603 (1985).

23.Penn, B.G.; Stannett, V.T. & Gilbert, R.D., Biodegradable cellulose graft

copolymers. I. Condensation-type graft reactions, J. Macromol. Sci. -

Chem., A16(2), 473-479 (1981).

134

24.Penn, B.G.; Stannett, V.T. & Gilbert, R.D., Biodegradable cellulose graft

copolymers. 11. Vinyl addition-type graft reactions, J. Macromol. Sci. -

Chem., AI6(2), 481-486 (1981).

25.Ach, A., Biodegradable plastics based on cellulose acetate, . J.MS.-Pure

Appl. Chem., A30(9&10), 733-740 (1993).

26.Dixon, M. & Webb, E.C., Enzymes, 3rd Edition, p. 251,Longmans,

London (1979).

27.Kamide, K.; Okajima, K.; Kowsaka, K. & Matsui, T., Solubility of

Cellulose Acetate prepared by different methods and its

correlationships with average acetyl group distribution on

glucopyranose units, Polymer Journal, 19(12), 1405-1412 (1987).

28.Dawkins, J.V., Theory of Gel Permeation Chromatography. Mechanism

of separation and the influence of polymer-sorbent interaction, Pure and

AppliedChemistry, 51, 1473-1481 (1979).

29.UK patent No. 9300901.7 (1994) ..

30.Froede, H.c. & Wilson, I.B., The Enzymes, 3rd Edition, edited by Boyer,

P.D., Vo15, p. 87, Academic Press, New York (1971).

31.Wood, W.A. & Ke\logg, S.T., Methods in enzymology, v. 160, part A, pp.

86-90, Academic Press Inc., New York (1988).

32.Weiszfeiler, Gy.J., Urmossy, M. & Hadju, J., The action of cellulase

enzyme on the unicellar algae Chlorella pyrenoidosa and Scenedesmus

quadricauda, Proc. Microbiol. Res. Group Hung. Acad. SCi., 2, 113-20

(1968).

33.Brock, V., Crassostrea gigas (Thunberg) hepatopancreas-cellulase

kinetics and cellulolysis of living monocellular algae with cellulose walls,

Journal of experimental marine biology and ecology, 128(2), 157-164

(1989).

34.Burczyk, J., Grzybek, J., Banas, J. & Banas, E., Presence of cellulase in

the algae Scenedesmus, Experimental cell research, 63(2-3), 451-3 (1970).

135

35.Timpa, 1.0., Characterisation by size exclusion chromatography with

refractive index and viscometry: complex carbohydrates-cellulose,

starch and plant cell wall polymers, Abstracts of papers of the American

Chemical SOciety, 206(2), 134-PMSE (1993)

36.Kamide, K; Miyazaki, Y. & Abe, T., Dilute solution properties and

unperturbed chain dimension of cellulose triacetate, Polymer Journal,

11(7),523-538 (1979).

37.Ishida, S.; Komatsu, H.; Katoh, H.; Saito, M.; Miyazaki, Y. & Kamide, K,

Limiting viscosity numbers and sedimentation coefficient of solutions of

cellulose acetates, Makromol. Chem., 183,3075-3087 (1982).

38.Kamide, K & Saito, M., A method for estimating unperturbed chain

dimension from molecular weight dependence of sedimentation and

diffusion coefficients Eur. Polym. J., 18,661-665 (1982).

39.Saito, M., Viscometric and light scattering study on dilute solution

properties of cellulose acetate with a degree of substitution of 1.75,

Polymer Journal, 15(3), 249-253 (1983).

40.Kamide, K; Miyazaki, Y. & Abe, T., Mark-Houwink-Sakurada

equations of cellulose acetate in various solvents, Makromol. Chem., 180,

2801-2805 (1979).

4l.Kamide, K; Saito, M. & Abe, T., Dilute solution properties of water

soluble incompletely substituted cellulose acetate, Polymer Journal,

13(5),421-431 (1981).

42.Kamide, K & Saito, M., Thermodynamic and hydrodynamic properties

of cellulose diacetate-dimethylacetamide solutions, Polymer Journal,

14(7),517-526 (1982).

43.Buchanan, C.M.; Edgar, K.I. & Wilson, A.K., Preparation and

characterisation of cellulose monoacetates: The relationship between

structure and water solubility, Macromolecules, 24, 3060-3064 (1991).

136

44.Kamide, K; Terakawa, T. & Miyazaki, Y., The viscometric and light

scattering determination of dilute solution properties of cellulose

diacetate, Polymer Journal, 11(4),285-298 (1979).

45.Suzuki, H.; Miyazaki, Y. & Kamide, K., Temperature dependence of

limiting viscosity number and radius of gyration of cellulose diacetate in

acetone, European Polymer Journal, 16, 703-708 (1980).

46.Kamide, K & Saitoh, M., New viscosity method for estimating

unperturbed chain dimensions of macromolecules when draining effect

and/or non-gaussian nature in the unperturbed state are not negligible,

European Polymer Journal, 17, 1049-1055 (1981).

47.Ahrnad, N.; Ehsan, J. & Rashad, A., Viscometric determination of

various parameters of cellulose acetate in acetone solutions, PhYSical

Chemistry, 7(1),35-40 (1988).

48.Brandrup, J. & Immergut, E.H., Polymer Handbook (third edition), page

V/147, Table K2, Wiley-Interscience Publication, New York (1989).

49.Goodlett, V.W.; Dougherty, J.T. & Patton, H.W., Characterisation of

cellulose acetates by Nuclear Magnetic Resonance, 1. Polym. Sci., Part A-

1,9(1),155-61 (1971).

50.MaIm, C.J.; Tanghe, L.J. & Laird, B.C., Primary Hydroxyl Groups in

Hydrolysed Cellulose acetate, 1. Amer. Chem. Soc., 72,2674 (1950).

51.Brown, H.c., A convenient preparation of volatile acid chlorides,

1.Amer.Soc., 60, 1325 (1938).

52.Kamide, K; Okajima, K & Saito, M., Nuclear Magnetic Resonance

study of thermodynamic interaction between cellulose acetate and

solvent, Polymer Journal, 13(2),115-125 (1981).

53. Frommer, M.A., & Shporer, M., A conformational change of cellulose

triacetate in solution, Amer. Chem. Soc., Div. Org. Coatings Plast. Chem.,

Pap., 32(1),374-83 (1972).

137

54.Miyamoto, T.; Sato, Y. & Shibata, T., 13C NMR spectral studies on the

distribution of substituents in water-soluble cellulose acetate, J. Polym.

Sei. Chem. Ed., 23(5),1373-81 (1985).

55.Kowsaka, K.; Okajima, K. & Kamide, K., Further study on the

distribution of substituent group in cellulose acetate by I3C CH) NMR

analysis: assignment of carbonyl carbon peaks, Polymer Journal, 18(11),

843-849 (1986).

56.Sei, T.; Ishitani, K.; Suzuki, R. & Ikematsu, K., Distribution of acetyl

group in cellulose acetate as determined by NMR analysis, Polymer

Journal, 17(9), 1065-1069 (1985).

57.Kamide, K. & Okajima, K., Determination of distribution of O-Acetyl

group in trihydric alcohol units of cellulose acetate by I3C NMR

analysis, Polymer Journal, 13(2),127-133 (1981).

58.Miyamoto, T.; Sato, Y.; Shibata, T. & Inagaki, H., 13C NMR studies of

cellulose acetate, J. Polym. Sei., Polym. Chem. Ed., 22(10), 2363-70

(1984).

59.Kowsaka, K.; Okajima, K & Kamide, K., Determination of the

distribution of the substituent group in cellulose acetate by full

assignment of all carbonyl carbon peaks of I3C (IH) NMR spectra,

Polymer Journal, 20(10),827-836 (1988).

60.DoyJe, S.; Pethrick, R.A.; Harris, R.K.; Lane, I.M.; Packer K.J. & Heatley,

F., 13C NMR studies of cellulose acetate in solution and solid states,

Polymer, 27(1), 19-24 (1986). ~

61.Hoshino, M.; Takai, I.; Fukuda, K.; Imura, K. & Hayashi, I., I3C NMR

study of cellulose derivatives in the solid state, Drug Des. Delivery, 4(2),

93-95 (1989).

62.Takai M.; Fukuda, K. & Hayashi, I., I3C NMR study of cellulose

triacetate in the solid state, J. Polym. Sei. Part C: Polym. Lett., 25(3),

121-126 (1987).

138

63.Mott, R.F.; Kirkwood, T.B.L. & Cumow, R.N., A test for tbe statistical

significance of DNA sequence similarities for applications in databank

searcbes, CAB/OS, 5, 123-131 (1989).

139

the microbial degradation of cellulose acetate

Documents