the microbial degradation of cellulose acetate
TRANSCRIPT
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.
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
(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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter 1 - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter 1 - Introduction
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
Chapter 1 - Introduction
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
Chapter I - Introduction
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
COOH I
H2N -C-H I H
Glycine
COOH I
H2N- C- H I
H- C -OH I H
Serine
42
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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
Chapter 2 - Materials and Methods
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.
Chapter 3 - Results and Discussion
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).
Chapter 3 - Results and Discussion
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 10. Discarded cigarette filter fibres, 200x magnification.
Photograph 11. Discarded cigarette filter fibres, 21 Ox magnification.
Photograph 12. Discarded cigarette filter fibres, 210x magnification.
Photograph 13. Discarded cigru'ette filter fibres, IOOOx magnification.
Photograph 14. Discarded cigarette filter fibres, 1000x magnification.
Photograph 15. Discarded cigarette filter fibres, 1350x magnification.
Photograph 16. Discarded cigarette filter fibres, 1500x magnification.
Chapter 3 - Results and Discussion
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 .
Photograph 28. Algae that grew on the cigarette filters .
Photograph 29. Algae that grew on the cigarette filters .
Chapter 3 - Results and Discussion
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)
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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.
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
These two pieces of evidence proved that the CA with a DS of 2.5 would not
biodegrade even if additives were employed.
63
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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.
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
~
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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)
Chapter 3 - Results and Discussion
TABLE 3.13
Table of relative and specific viscosities vs. concentration for the CA with
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
Table of relative and specific viscosities vs. concentration for the CA with
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
CONCENTRATION E+3 (g/ml)
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)
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Table of relative and specific viscosities vs. concentration for the CA with
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
CONCENTRATION E+3 (g/ml)
Chapter 3 - Results and Discussion
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 .
Chapter 3 - Results and Discussion
TABLE 3.19
Table of relative and specific viscosities vs. concentration for the CA with
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
Table of relative and specific viscosities vs. concentration for the CA with
a DS of 1.7 (DMAlLiCI, wide tube, 60°C)
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)
Chapter 3 - Results and Discussion
TABLE 3.21
Table of relative and specific viscosities vs. concentration for the CA with
a DS of 1.5 (DMAlLiCI, wide tube, 60°C)
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)
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
TABLE 3.23
Table of relative and specific viscosities vs. concentration for the CA with
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
Table of relative and specific viscosities vs. concentration for the CA with
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
CONCENTRATION E+3 (g/ml)
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)
Chapter 3 - Results and Discussion
TABLE 3.25
Table of relative and specific viscosities vs. concentration for the CA with
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
CONCENTRATION E+3 (glml)
Chapter 3 - Results and Discussion
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
CONCENTRATION E+3 (glml)
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
CONCENTRATION E+3 (g/ml)
1.8 1.9
'''',
H.; . ,
I, ;., ,
•
2
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
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 3880 Hz
Gaussian broadening 0.008 sec FT size 32168 Ambient temperature
"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
Observe C13 Frequency 15.430 MHz Spectral width 30001.5 Hz Acquisition time 20.3 ms Relaxation delay 5.0 sec No. repetitions 80
Cross polarization Contact time 3.00 ms Spin-rate 3130 Hz
Gaussian broadening 0.008 sec FT size 32168 Ambient temperature
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
Cross polarization Contact time 3.00 ms Spin-rate 3140 Hz
Gaussian broadening 0.008 sec FT size 32166 Ambient temperature
'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
Observe C13 Frequency 75.430 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 4130 Hz
Gaussian broadening 0.008 sec FT size 32768 Ambient temperature
" 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
Gaussian broadening 0.008 sec FT size 32168 Ambient temperature
'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
Cross polarization COntact time 3.00 ms Spin-rate 2620 Hz
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
Cross polarization COntact time 3.00 ms Spin-rate 4090 Hz
Gaussian broadening 0.005 sec FT size 32768 Ambient temperature
2 0 240 220 200 180 160 140 1 0 100 80 60 40
o ~ ... o N
20 ppm
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 3 - Results and Discussion
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
Chapter 4 - Conclusions
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
Chapter 4 - Conclusions
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
Chapter 4 - Conclusions
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
Chapter 4 - Conclusions
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
Chapter 4 - Conclusions
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