-
Cellulose 10: 283296, 2003. 2003 Kluwer Academic Publishers.
Printed in the Netherlands. 283
Unconventional cellulose esters: synthesis, characterizationand
structureproperty relations
Thomas Heinze1,2,3, Tim F. Liebert1, Katy S. Pfeiffer1 &
Muhammad A. Hussain21Institut fr Organische Chemie und
Makromolekulare Chemie, Friedrich-Schiller-Universitt Jena,
Lessing-strasse 8, D-07743 Jena, Germany2Fachbereich 9 (Chemie),
Bergische Universitt Wuppertal, Gau Strasse 20, D-42097 Wuppertal,
Germany3Author for correspondence (E-mail:
[email protected])Received 26 November 2002; accepted 30
March 2003
Key words: Cellulose ester, Cellulose solvents, In situ
activation, Structure, Synthesis
Abstract
This paper summarizes selected results obtained during a
two-year research project in the framework of thefocus program
Cellulose and cellulose derivatives (SPP 1011), sponsored by the
German Science Foundation(DFG). New synthesis paths for the
preparation of the most important cellulose ester, cellulose
acetate, wereinvestigated. In contrast to conventional methods,
cellulose was converted in a homogeneous phase reaction withacetyl
chloride in the presence of different bases, including polyvinyl
pyridine and cross-linked polyvinyl pyridine.Moreover, results of
the conversion in the new solvent dimethyl
sulfoxide/tetrabutylammonium fluoride trihydrateare discussed. The
structures obtained were analyzed both on the level of the
anhydroglucose unit (AGU) andalong the polymer chain. It was found
that the addition of a base can significantly change the
selectivity of thereaction and thereby the properties of the
products (e.g., solubility). No signs of a non-statistical
distribution ofthe acetyl groups along the polymer chains were
observed. Furthermore, reactivity and selectivity of the
acylationreactions, using in situ activation with p-toluenesulfonyl
chloride (Tos-Cl), were studied for different long-chaincarboxylic
acids (capric-, caprylic-, decanoic-, lauric-, palmitic-, stearic
acid). The thermogravimetric analysisof these derivatives showed
that the decomposition temperature increased with an increasing
number of carbonatoms, starting from 292 C (cellulose caprate) to
318 C (cellulose stearate). New cellulose derivatives
weresynthesized, for example, cellulose adamantoyl ester. For this
purpose cellulose was converted homogeneouslyin
N,N-dimethylacetamide/LiCl with free acids in the presence of
activating reagents, for example, Tos-Cl
or1,1-carbonyldiimidazol.
Introduction
The development of new reaction paths for polymeranalogous
modification is one of the most impor-tant tools for the design of
cellulosics with tailoredproperties. In recent years we have
investigated al-ternative paths for the carboxymethylation of
cellulose(Liebert et al. 1996; Liebert and Heinze 1998a,b;Heinze et
al. 1999). A new concept was established,including the conversion
of cellulose dissolved in N,N-dimethylacetamide (DMA)/LiCl with
sodium mono-chloroacetate in the presence of solid NaOH
particles(Liebert and Heinze 1998a,b). This path yielded de-
rivatives with high degrees of substitution (DS) anda completely
new distribution of substituents on thelevel of the repeating units
and along the polymerchain, compared with carboxymethyl cellulose
(CMC)prepared in the industrially applied slurry process.Thus, it
was found that the alternative CMCs showeda preferred
functionalization at position 6 compared tocommercial samples. SEC
analysis after polymer frag-mentation with endoglucanase indicated
a block-likedistribution of the carboxymethyl functions along
thepolymer backbone (Saake et al. 2000). Atomic forcemicroscopy
(AFM) revealed a new superstructure.The alternatively prepared CMC
forms a network-like
-
284
system in solution, while commercially preparedsamples show
fringed micelles (Liebert and Heinze2001). These molecular and
supermolecular featuresresulted in a number of amazing new
macroscopicproperties, for example, different rheological and
col-loidal behavior (Ktz et al. 2001).
In one of our basic research projects, interest isfocused on the
search for new tools for the preparationof cellulose esters,
including the application of cellu-lose solvents, the in situ
activation of carboxylic acidsand the use of polymeric bases. The
esterificationof cellulose in DMA/LiCl has been extensively
stud-ied during the last decade (Dawsey 1994; El Seoudet al. 2000).
The conversion of the polymer withfree acids after in situ
activation was applied besidesacetylation with acid chlorides and
acid anhydrides.In situ activation of the carboxylic acids is
possiblewith p-toluenesulfonyl chloride (Tos-Cl). It was
firstapplied for the preparation of cellulose acetates(Shimizu and
Hayashi 1988). The extension of thispath on the homogeneous
derivatization of cellu-lose with waxy carboxylic acids was
studied. It wasshown that cellulose esters, having alkyl
substitu-ents in the range from C12 (laurylic acid) to
C20(eicosanoic acid), could be obtained with almostcomplete
functionalization of the OH groups (DSvalues 2.82.9; Sealey et al.
1996). It was exten-ded to the preparation of water-soluble
oxocarboxylicacid esters of cellulose (Heinze and Schaller
2000).Moreover, the very powerful condensation agent
N,N-dicyclohexylcarbodiimide (DCC) in combination
with4-pyrrolidinopyridine (PP) was exploited for the syn-thesis of
cellulose esters, starting from free carboxylicacids (Samaranayake
and Glasser 1993a,b). This ap-proach can be used to efficiently
prepare derivativeswith low DS.
This paper deals with results concerning differ-ent new paths
for the esterification of cellulose, in-cluding the application of
both alternative cellulosesolvents and polymeric bases, as well as
the exploi-tation of 1,1-carbonyldiimidazole (CDA) as
activatingagent. The influence of the esterification path on
theproperties of the esters obtained was also studied.
Experimental
Materials
Avicel (Fluka, Avicel PH-101, degree of poly-merization DP =
260) was used as a starting poly-mer. Non-crosslinked polyvinyl
pyridine had a Mw
of 200,000 g/mol. LiCl was dried for 6 h at 105 Cin vacuum prior
to use. Cross-linked polyvinylpyri-dine, tetrabutylammonium
fluoride trihydrate (TBAF),acetyl chloride, adamantoyl chloride
(AdCl), CDA,Tos-Cl, dimethyl sulfoxide (DMSO), DMA, and
thecarboxylic acids, supplied by Fluka, were used asreceived.
Methods
Dissolution of cellulose in DMA/LiCl (solution S1)For a typical
preparation, 1.0 g (6.2 mmol) of driedcellulose and 40 mL DMA were
kept at 130 C for 2 hunder stirring. After the slurry had been
allowed tocool to 100 C, 3 g of anhydrous LiCl were added.
Thecellulose was completely dissolved by cooling downto room
temperature under stirring.
Acetylation of cellulose with acetyl chlorideA solution S1 (see
above) was kept in an ice bath for15 min. To this cooled solution
was carefully added2.2 mL acetyl chloride (5 mol/mol AGU). The
systemwas heated to 80 C for 2 h and kept at room tempera-ture for
24 h. Isolation was carried out by precipitationinto 200 mL
ethanol, washing with ethanol and dryingin vacuum at 50 C (sample
A4).Yield: 1.5 g (84.9%).DSAcetate = 2.96 (determined by means of
1H NMRspectroscopy after perpropionylation).FTIR (KBr): 3502 (OH),
2890 (CH), 1750(C==OEster) cm1.13C NMR (DMSO-d6): 169.2169.9ppm
(C==O),60.3102.5ppm (cellulose backbone).If a base was applied, it
was added before the addi-tion of the acetyl chloride. If polyvinyl
pyridine wasused as a base the products were reprecipitated
fromDMSO.
Acetylation of cellulose in DMSO/TBAFFor a typical conversion, a
solution of 1 g (6.2 mmol)cellulose in 33 mL DMSO and 6.6 mL TBAF
wastreated with 1.14 mL (14.2 mmol) of vinyl acetate for70 h at 40
C (for other reagents see Table 1). Theproduct was isolated by
precipitation into 200 mLisopropyl alcohol, adding 50 mL water
(removal ofinorganic impurities, no signals for TBAF in NMRspectra)
and filtration. After washing with 200 mL iso-propyl alcohol, the
product was dried in vacuum at50 C (sample B2).Yield: 1.0 g
(80.3%).
-
285Table 1. Summary of reaction conditions and results of
acetylation of cellulose dissolved in DMA/LiCl withacetyl
chloride.
No. Molar ratio Partial DSAc a in position " Solubilityb(acetyl
chloride/AGU)
6 2,3 DMSOc Acetone CHCl3
A1 1.0 0.77 0.44 1.21d + A2 3.0 0.90 1.95 2.85 + +A3 4.5 1.00
1.94 2.94 + eA4 5.0 1.00 1.94 2.96 + +
a DS of the ester obtained, determined via 1H NMR spectroscopy
after perpropionylation.b+ Soluble; insoluble.c Dimethylsulfoxide.d
This stoichiometrically impossible value may result from
fractionation during work up.e The insolubility cannot be explained
by structural features.
DSAcetat = 1.04 (determined by means of 1H NMRspectroscopy after
perpropionylation).FTIR (KBr): 3490 (OH), 2905 (CH),
1752(C==OEster) cm1.13C NMR (DMSO-d6): 169.1169.9ppm
(C==O),60.3102.5 ppm (cellulose backbone).
Reaction of cellulose with AdCl in DMA/LiClAdCl (3.7 g, 18.6
mmol) and 1.8 mL (22.3 mmol)pyridine were added to a solution S1
and stirred for24 h at 80 C. The homogeneous reaction mixture
waspoured into 250 mL of ethanol. After filtration, thepolymer was
washed with ethanol and dried in vacuumat room temperature, product
D13.Yield: 2.3 g (78.6%).DSAd= 1.92 (determined by means of 1H NMR
spec-troscopy after perpropionylation).FTIR (KBr): 3457 (OH), 2909,
2854 (CH), 1720(C==OEster) cm1.13C NMR (CDCl3): = 176.5 (CO), 103.0
(C-1),100.9 (C-1), 81.3 (C-2,3s, C-4), 77.0 (C-3, C-5), 73.6(C-2),
61.2 (C-6s), 40.9 (-C), 39.0 (-CH2), 36.4(-CH2), 27.9 ( -CH)
ppm.
Reaction of cellulose with adamantane carboxylicacid
(AdOH)/CDAAdOH (3.4 g, 18.6 mmol) was dissolved in 20 mLDMA and 3.0
g (18.6 mmol) CDA was added. Thismixture was combined with a
solution S1 and stirredfor 24 h at 80 C. The mixture was
precipitated in300 mL of ethanol, filtered off, washed with
ethanoland dried in vacuum at room temperature (productD26).Yield:
1.8 g (76.7%).DSAd = 1.31 (determined by means of 1H NMR
spec-troscopy after perpropionylation).
FTIR (KBr): 3458 (OH), 2910, 2855 (CH), 1728(C==OEster) cm1.13C
NMR (DMSO-d6): = 176.4 (CO), 102.6(C-1), 99.5 (C-1), 78.8 (C-4),
73.4 (C-3, C-5, C-2),62.9 (C-6s), 61.6 (C-6), 40.1 (-C), 38.8
(-CH2),36.4 (-CH2), 27.8 ( -CH) ppm.
Esterification of cellulose with lauric acid/Tos-ClTos-Cl (35 g,
12.5 mmol) was added to a solution S1followed by 2.47 g (12.5 mmol)
of lauric acid understirring. The reaction mixture was stirred for
24 hat 80 C under N2. The homogeneous reaction mix-ture was
precipitated in 800 mL buffer solution (7.14 gK2HPO4 and 3.54 g
KH2PO4 per liter of H2O) andthe polymer was collected by
filtration. After washingthe polymer with 800 mL water three times,
Soxhletextraction with ethanol was carried out for 24 h. Thepolymer
was dried at 50 C under vacuum to yieldproduct C4.Yield: 2.1 g
(77.0%).DSLaur= 1.55 (determined by means of 1H NMRspectroscopy
after peracetylation).FTIR (KBr): 3486 (OH), 2925, 2855 (CH),
1238(COCEster), 1753 (=COEster) cm1.13C NMR (CDCl3): = 173.8 (CO),
104.0 (C-1),102.6 (C-1), 72.3 (C-2), 73.3 (C-3), 82.0 (C-4),
75.1(C-5), 20.634.0 (CMethylene), 13.9 (CMethyl) ppm.
Typical example for perpropionylationof a cellulose ester for DS
determinationA mixture of 6 mL pyridine, 6 mL propionic
acidanhydride and 50 mg 4-(dimethylamino)pyridine wasadded to 0.3 g
of the adamantoyl cellulose D13. After24 h at 80 C, the reaction
mixture was cooled toroom temperature and precipitated in 50 mL
ethanol.
-
286
For purification, the isolated product was reprecipi-tated from
chloroform into 50 mL ethanol, filtered off,washed with ethanol and
dried in vacuum at room tem-perature.Yield: 0.7 g (67%).DSAd= 1.92,
DSProp= 1.08 (both determined bymeans of 1H NMR spectroscopy).FTIR
(KBr): no (OH), 2910, 2854 (CH), 1758,1737 (C==OEster) cm1.13C NMR
(DMSO-d6): =177.0173.1 (CO), 100.262.6 (C atoms of the modified
anhydroglucose unit(AGU)), 41.1 (CH2-propionate), 39.4 (-C),
39.0(-CH2), 36.8 (-CH2), 28.2 ( -CH), 9.4 (CH3-propionate) ppm.1H
NMR (CDCl3): = 5.10 (H-3), 4.68 (H-2), 4.38(H-1, 6), 3.99 (H-6),
3.56 (H-4, 5), 2.18 (CH2-2,3-propionate), 2.03, 1.95, 1.88, 1.73
(H-adamantane),1.03 (CH3-2, 3-propionate) ppm.
Measurements
13C NMR spectra were acquired on a Bruker AMX400 MHz
spectrometer. The cellulose esters weremeasured in DMSO-d6, CDCl3
and THF-d8 at 40 and70 C, respectively. The number of scans was in
therange from 5000 to 20,000.
1H NMR spectra of the esters were acquiredin CDCl3 after
perpropionylation of the unmodifiedhydroxyl groups (Heinze and
Schaller 2000) to de-termine the DS-values. FTIR spectra were
measuredon a Bio-Rad FTS 25 PC, using the KBr pellettechnique.
Thermal decomposition temperatures (Td) of thecellulose esters
were determined by thermogravimet-ric analysis (TGA) on a Mettler
Toledo TC 15 MettlerTG 50 Thermo balance. The Td was reported as
theonset of significant weight loss from the heated sample(Sealey
et al. 1996). Samples (10 mg) were measuredunder air with a
temperature increase of 10 C/minfrom 35 C up to 600 C.
Elemental analyses were performed by CHNS 932Analyzer
(Leco).
For GPC analysis, JASCO equipment was usedincluding degasser
(DG-980-50), pump (PU-980), RI-detector (RI-930) and UV-detector
(UV-975) workingat 254 nm. THF was used as eluent (30 C, 1
mL/min).The separation was carried out using columns frompolymer
standards service (Mainz, Germany) with1000, 10,000 and 1,000,000 .
Polystyrene standardswere used for calibration.
The HPLC analysis of the Sisal cellulose sampleswas carried out
as described for cellulose derivatives(Liebert and Heinze
2001).
For the methylation, 0.5 g of the starting cellu-lose ester was
dissolved in 30 mL trimethylphosphate.Methyl trifluoromethane
sulfonate (4 mol/mol re-maining hydroxyl group) and
2,6-di-tert-butylpyridine(3 mol/mol hydroxyl group) were added.
This mixturewas stirred for 4 h at 60 C and 16 h at room
tempera-ture using argon as protective gas. Isolation was car-ried
out by precipitation into ethanol. For completedepolymerization the
methyl cellulose ester was treat-ed with 2 N TFA for 4 h at 120 C.
The acid and thewater were removed by distillation. The HPLC
experi-ments were carried out as described (Erler et al. 1992).
The Karl Fischer titration was carried out with aMettler-Toledo
Coulometer DL 37 using Hydranal Aand Hydranal C (Sigma-Aldrich) as
reagents.
Results and discussion
Cellulose acetate influence of baseson the reaction
Different paths for the homogeneous synthesis of cel-lulose
acetates are known. Thus, cellulose was acet-ylated in DMA/LiCl
using acetic anhydride (Marsonet al. 1999; El Seoud et al. 2000).
We studied theacetylation of cellulose dissolved in DMA/LiCl
withacetyl chloride without an additional base and in thepresence
of different pyridine derivatives. In a pre-liminary set of
experiments, cellulose dissolved inDMA/LiCl was converted
homogenously with acetylchloride. The experimental details and the
values ofthe DS of the products are summarized in Table 1.The
reaction succeeds with almost complete conver-sion of the reagent,
that is, it can be controlled bystoichiometry. 1H NMR experiments
of the perpropi-onylated samples show a preferred
functionalizationof the primary hydroxyl group.
In addition to the NMR spectroscopic experiments,the structure
of sample A1 was studied by HPLC afterpermethylation and
depolymerization. For this pur-pose, the product A1 was
permethylated with methyltrifluoromethane sulfonate in trimethyl
phosphate inthe presence of 2,6-di-tert-butylpyridine, to
convertthe pattern of substitution of the acetate into an in-verse
methyl ether pattern (Figure 1). After completesaponification of
the ester functions and degradationof the polymer with aqueous
trifluoroacetic acid, the
-
287
Figure 1. Analytical path for the determination of the
functionalization pattern of cellulose esters by means of HPLC
after permethylation anddegradation.
Table 2. Summary of reaction conditions and results of the
acetylation of cellulose dissolved in DMA/LiCl with acetylchloride
in the presence of pyridine.
No. Molar ratio Partial DSa in position " Solubilityb
Acetyl chloride/AGU Pyridine/AGU 6 2,3 DMSOc Acetone CHCl3
A5 1.0 1.2 0.63 0.37 1.00 + A6 3.0 3.6 0.94 1.62 2.56 + +A7 5.0
6.0 0.71 2.0 2.71 + + +A8 5.0 10.0 0.46 2.0 2.46 + + +
a DS of the ester obtained determined via 1H NMR spectroscopy
after perpropionylation.b + Soluble; insoluble.c
Dimethylsulfoxide.
mixture of methyl glucoses obtained can be separatedby means of
HPLC (Erler et al. 1992). A DSAcetate of1.16 was calculated from
the chromatogram, which isin good agreement with the DS obtained by
1H-NMRspectroscopy (DS= 1.21). No evidence for ester groupmigration
during the procedure was found, but it can-not be completely
excluded. A comparison of theresults obtained using this analytical
strategy with sta-tistical calculations was performed in the same
way asfor the analysis of CMC (Heinze et al. 1999). No
sig-nificantly increased amounts of glucose or trimethylglucose
were found by means of HPLC. Thus, glu-cose was determined to be 3%
(calculated 5.7%) and21% trimethyl glucose was found (calculated
23.1%).Consequently, cellulose acetates prepared via this pathhave
a statistically even distribution of substituentsalong the polymer
chain.
In another set of experiments the influence of abase on the
course of the reaction and on the distri-bution of substituents was
studied. An amazing resultwas that the application of pyridine as
base leads toproducts of a decreased DS (Table 2). Comparisonof
samples A2 and A6 or samples A4 and A7 showsa decrease of DS of
about 0.3. It is even more pro-nounced if the amount of base is
increased (see sampleA8). Moreover, 1H NMR spectroscopy of the
productsreveals less preferred substitution in position 6.
Thisselectivity is diminished by an increased concentra-
tion of the base. Thus, sample A8 shows a partialDSO-6 of 0.46
versus an overall DS of 2.46, thatis, all the secondary OH groups
are acetylated. Thiscould be a first hint for a preferred
deacetylation at the6-O-position.
GPC was applied to investigate hydrolytic deg-radation of the
polymer chain during the reaction.It was found that the
depolymerization was rathersmall without a base. All derivatives
were preparedwith Avicel as starting polymer, having a DP of
260.Product A7 possesses a DP of 256. However, the DPdecreases to
103 during the reaction under comparableconditions but using
pyridine as base (sample A8).
One possible explanation for the degradation mightbe the
formation of the acidic pyridinium hydrochlo-ride in the case of
the base-catalyzed reaction. Mostof the HCl formed is liberated
from the system if noadditional base is applied. It needs to be
mentionedthat the influence of the acidic pyridinium hydrochlo-ride
yields a product with a different solubility. Thus,sample A8
dissolves completely in acetone in contrastto sample A4 (prepared
with no base, see Table 1).Permethylation, degradation and HPLC as
describedabove did not show any hints for a non-statistical
dis-tribution of the substituents along the polymer
chain.Consequently, the different solubility is only due to
thedifferent distribution of substituents on the level of
theAGU.
-
288
Table 3. Conditions and results of the acetylation of cellulose
dissolved in DMA/LiCl with acetyl chloride in the presenceof
crosslinked polyvinyl pyridine.
No. Molar ratio Partial DSa in position " Solubilityb
Acetyl chloride/AGU Base/AGU 6 2,3 DMSOc Acetone CHCl3
A9 1.0 1.2 0.35 0.13 0.48 + A10 2.0 2.4 0.82 0.51 1.33 + A11 3.0
3.6 0.91 0.65 1.56 + A12 4.5 4.5 1.0 1.24 2.24 + + A13 5.0 6.0 1.0
1.62 2.62 + A14d 5.0 10.0 A15e 5.0 10.0 0.97 1.31 2.28 + +
a DS of the ester obtained determined via 1H NMR spectroscopy
after perpropionylation.b + Soluble; insoluble.c
Dimethylsulfoxide.d Product isolation not possible because
insoluble polymer is fixed on base surface.e Non-crosslinked base
was used.
For the first time, polymer-bound bases like cross-linked
polyvinyl pyridine were applied for the pre-paration of cellulose
carboxylic acid esters. Table 3summarizes the reaction conditions
and results. Again,a significant decrease of overall DS values can
be rec-ognized in comparison to reactions applying no base.Thus,
for sample A2 (see Table 1) an almost completesubstitution (DS =
2.85) was found if a molar ratio of3 mol acetyl chloride per mol
AGU is used. The DSreached was 1.56 in a comparable experiment
(sampleA11) applying polyvinyl pyridine. High selectivity ofthe
acetylation at position 6 was observed, in contrastto the
acetylation reactions with pyridine as base. Adrastic decrease of
both the DS values and the yield aswell as a different solubility
of the product is observedif a large surplus of polymer-bound base
is used.Thus, sample A13 is soluble in DMSO only, evenwith a high
DS of 2.62. If the molar ratio base/AGUis in the range >10,
product isolation is almost im-possible (sample A14). Extraction of
the precipitate,which consists mainly of polyvinylpyridine and
cel-lulose acetate, as can be confirmed by FTIR (signalsat 1595 and
3060 cm1 for the polyvinylpyridine andsignals at 1019 and 1740 cm1
for the cellulose ace-tate), yields only traces of the product, in
the rangeof 1%. If the extraction was carried out with DMSO,9g
cellulose acetate was recovered for an exper-iment with 1 g of
cellulose as starting material. IfTHF was used, 11g were isolated.
1H-NMR spec-troscopy was applied for structure determination butno
DS calculation was possible because of the pooryield. The
alternative solubility of these cellulose ace-tates is comparable
to p-toluenesulfonic acid esters
of cellulose (cellulose tosylate), with a
non-statisticaldistribution of substituents along the polymer
chainprepared in a reactive microstructure, that is, by con-version
of cellulose regenerated from solution on solidNaOH particles
(Einfeldt et al. 2002).
An acetylation experiment was carried out usingsoluble,
non-cross-linked polyvinyl pyridine (sampleA15; Table 3) to obtain
a cellulose acetate that canbe isolated from the polymeric base. No
regenerationor precipitation of the polymers occurred during
thecompletely homogeneous reaction. A cellulose acetatewas obtained
with a DS of 2.28, determined by 1HNMR spectroscopy, which is
easily soluble in acetone.HPLC analysis gave a DS of 2.30 and
showed no in-creased values for non- or fully-substituted
repeatingunits. It may be assumed that during the conversionthe
cellulose is not permanently fixed to the dissolvedpolymeric base
and an even distribution of substituentsresulted from an
equilibrium reaction.
It should be mentioned that in the framework ofthis study,
acetylation experiments were carried outusing in situ activation of
acetic acid with CDA (forthe mechanism see below). DSAcetate values
of 0.7, 1.5and 2.1, respectively, were achieved if molar ratios
of1:2:2, 1:5:5 and 1:10:10 (AGU/acid/CDA) were ap-plied. However,
this new path did not yield polymerswith a new pattern of
functionalization.
Acylation in the new cellulose solventDMSO/TBAF
A mixture of DMSO/TBAF represents an efficient cel-lulose
solvent. It dissolves cellulose completely with a
-
289
Table 4. Esterification of different types of cellulose in
DMSO/TBAF. Summary of reaction conditions and results.
No. Cellulose Acetylating agent Molar %TBAF in Time (h) Temp.
(C) DSb Solubilitytype ratioa DMSO
B1 Avicel Acetic anhydride 1:2.3 16 70 40 0.83 InsolubleB2
Avicel Vinyl acetate 1:2.3 16 70 40 1.04 DMSOB3 Avicel Vinyl
acetate 1:10.0 16 70 40 2.72 DMSOB4 Avicel Vinyl butyrate 1:2.3 16
70 40 0.86 InsolubleB5 Avicel Vinyl laurate 1:10.0 16 70 40 2.60
Pyridine, THF, CHCl3B6 Avicel Vinyl benzoate 1:2.3 16 70 40 0.95
InsolubleB7 Sisal Vinyl laurate 1:2.3 11 3 60 1.24 InsolubleB8
Sisal Acetic anhydride 1:2.3 11 3 60 0.3 InsolubleB9 Sisal Acetic
anhydride 1:2.3 8 3 60 0.96 DMSO, pyridineB10 Sisal Acetic
anhydride 1:2.3 7 3 60 1.07 DMSO, pyridineB11 Sisal Acetic
anhydride 1:2.3 6 3 60 1.29 DMSO, DMF, pyridine
a Mol AGU/mol acylation reagent.b DS of the ester obtained
determined via 1H-NMR.
DP of up to 650 without pretreatment within 15 min.In 13C NMR
spectra signals appear only at 102.7(C-1), 78.4 (C-4), 75.6 (C-5),
75 (C-3), 73.5 (C-2)and 59.9 ppm (C-6) (Heinze et al. 2000). This
clearlyshows that the cellulose is dissolved without
covalentinteractions, as can be concluded from the compar-ison of
the chemical shifts with values of cellulosedissolved in DMA/LiCl,
which is a typical so-callednon-derivatizing cellulose solvent.
This new solventwas exploited for a number of acylation reactions.
Asummary of reaction conditions and DS values of theproducts
obtained is given in Table 4.
The dissolved cellulose was treated with acetic an-hydride for
70 h at 40 C. A cellulose acetate with aDS value of 0.83 was
obtained if a molar ratio of2.3:1.0 (acylation reagent/AGU) was
applied. Com-parable conditions were used for the reaction
withvinyl acetate as acylating reagent. In case of the samemolar
ratio, a DS of 1.04 can be achieved, which isdue to the formation
of acetaldehyde during this con-version, shifting the equilibrium
towards the productside. On the other hand, the lower DS in the
case of theapplication of acetic anhydride is caused by the
com-parably fast hydrolysis of the reagent, due to the watercontent
of the solvent. A variety of vinyl carboxylicacid esters can be
exploited for this type of conversion(see Table 4). The DS values
can be controlled via theamount of reagent added. A remarkable
result was aDS as high as 2.6 for cellulose laurate, indicating
thatthis homogeneous esterification path is highly efficientfor the
preparation of fatty acid esters of cellulose.
Experiments were carried out with Sisal cellulose,which
represents fast-growing lignocellulosic material
comparable to sugarcane bagasse and linters (El Seoudet al.
2000; Marson et al. 2000; Ass and Frollini 2001;Sun et al. 2001).
The starting material had a DP of 650,a crystallinity index (Ic) of
77%, and contained about14% hemicellulose, as confirmed by 13C NMR
spec-troscopy (Figure 2) and HPLC analysis after
completedepolymerization (Figure 3). The conditions suitablefor
dissolution of cellulose materials (Avicel, woodpulp) in DMSO/TBAF
discussed above did not yieldoptically clear solutions in the case
of Sisal cellulose.This is obviously due to the presence of
hemicelluloseand the fibrous structure of Sisal cellulose. However,
itwas found that Sisal cellulose dissolves completely inthe mixture
DMSO/TBAF after 30 min at room tem-perature and 60 min at 60 C
(Ciacco et al. 2000).Nevertheless, static light-scattering
experiments of thesolution showed a fairly high amount of
aggregation.The values of the molecular weights determined werein
the range of 20 to 50 106 g/mol (Figure 4).
Sisal celluloses were esterified homogeneously inDMSO/TBAF using
acetic anhydride and vinyl laurateas acylating reagents. Reaction
conditions, results andsolubility of the esters obtained are listed
in Table 4.Transesterification with vinyl laurate yields the
corre-sponding ester with a DS of 1.24, while conversionwith acetic
anhydride gave an acetate with DS of 0.30.The concentration of TBAF
in the solution was variedfrom 6 to 11% to study this influence on
the dissolu-tion and the product features. All mixtures gave
clearsolutions. The DS values of the cellulose acetates pre-pared
in these different solvent mixtures decrease withincreasing TBAF
concentration (see Table 4). As thesalt is hydrated, the amount of
water in the medium
-
290
Figure 2. 13C-NMR spectrum of Sisal cellulose (cell) in
DMSO/TBAF, also showing peaks due to the presence of xylose
(xyl).
Figure 3. HPL-chromatogram of a completely depolymerized
Sisalcellulose sample. (A chiral detector signal, B refraction
indexdetector (RI) signal, dr detector response, RT retention time,
1 inorganic salts, 2 glucose, 3 xylose).
increases with the salt concentration. This in turn in-creases
the rate of hydrolysis both of the anhydride andprobably of the
ester moieties formed as well. Further-more, the interactions of
the water with the cellulosicOH groups may hinder the access of
acetic anhydride,resulting in a lower DS.
Figure 4. Berry plot of Sisal cellulose and alkali treated
Sisalcellulose in DMSO/TBAF.
Experiments directed towards removal of the waterin the solvent
were carried out. The water content wasanalyzed by means of Karl
Fischer titration. The addi-tion of molecular sieves did not
significantly influencethe water content. Because the treatment of
the solventwith strong dehydrating reagents like sodium
hydridewould produce the undesired dimsyl ions as a by-product, we
studied the dewatering of DMSO/TBAF,DMSO/TBAF/Sisal and
DMSO/TBAF/Avicel by va-cuum distillation.
A mixture of 60 ml DMSO and 6.6 g TBAF andcomparable mixtures
containing cellulose (between1.4 and 2.6%, w/w) were distilled
stepwise (stepsof 0.6 mL). The first sample obtained for pure
-
291
DMSO/TBAF contained 55% water. In the case ofa solution of
Avicel/DMSO/TBAF 22% water wasfound and 5% for a solution of
Sisal/DMSO/TBAF,after the first distillation step. These data lead
to theassumption that the water is strongly involved in thesolution
complex of cellulose. After distillation of atotal amount of about
6 mL of water (in the case of thecellulose-containing solution) a
drastic increase in theviscosity occurred. Therefore, it is useful
to preparea mixture of DMSO/TBAF, distill the majority of thewater
by removing about 30% (v/v) of the mixturein vacuum and dissolve
the cellulose in the resultingsolvent mixture. Solutions prepared
in this manner stillgave optically clear systems after the heat
treatmentdescribed above, which was also used for the acetyla-tion
of cellulose. The reactions in the solvent with areduced water
content lead to products with a signifi-cantly higher DS under
comparable conditions. TheDSAcetate increases from 0.30 (see Table
4, sample B8)to 1.15. Thus, this path is more efficient for the
acet-ylation of Sisal, applying acetic anhydride. 1H NMRanalysis of
the perpropionylated product shows a dis-tribution of the acetyl
groups at the reactive sites in theorder C-6>C-2>C-3. No
hints for a non-statisticaldistribution of substituents along the
polymer chainwere found. A reaction of cellulose dissolved in
an-hydrous DMA/LiCl, applying the same molar ratio ofacetic
anhydride, yields a product with a DS of 1.0(Ciacco et al.
2000).
Studies on the acylation of cellulose with carboxylicacids in
situ activated with Tos-Cl
An interesting new path for cellulose ester preparationis
homogeneous acylation after in situ activation ofcarboxylic acids
with Tos-Cl (Figure 5). It was shownthat cellulose esters, having
alkyl substituents in the
Figure 5. Schematic plot of the conversion of cellulose
withcarboxylic acid applying in situ activation with Tos-Cl.
range from C12 to C20, could be obtained with al-most complete
functionalization of the accessible OHgroups (Sealey et al. 1996).
A variety of differentcellulose esters was successfully synthesized
via thispath, however, without the use of an additional
base(Koschella et al. 1997; Heinze and Schaller 2000).
Considering these results, the question arises if thereaction
conditions (time, molar ratio of the reagents)and the application
of an additional base, for example,pyridine, influence the DS, the
molecular weight andother structural features of the products.
These studieswere performed with long-chain fatty acids becausethe
efficiency of this particular system for the prep-aration of the
corresponding esters had been shown(Heinze and Liebert 2001).
Thus, cellulose dissolved in DMA/LiCl was al-lowed to react with
two equivalents, carboxylic acid(capric-, caprylic-, decanoic-,
lauric-, palmitic- andstearic acid, respectively) and Tos-Cl
without an addi-tional base for 24 h at 80 C. The corresponding
esters(polymers C16, Table 5) show two characteristicpeaks in FTIR
spectra typical for the ester moieties atabout 1240 cm1 (COCEster)
and about 1750 cm1(C==OEster). Elemental analysis reveals the
absence ofsulfur in the samples, showing that there is no
remark-able introduction of tosylate groups, either
covalentlybounded or as impurity.
The 13C NMR spectrum of C4, for example, re-corded in CDCl3,
shows the characteristic signals at= 173.8 (CO), 104.0 (C-1), 102.6
(C-1), 72.3 (C-2),73.3 (C-3), 82.0 (C-4), 75.1 (C-5), 62.5 (C-6),
13.9(CH3) ppm. The signals of the methylene groups ofthe lauric
acid appear in the range of 22.634.0 ppm(Figure 6). The peak for
C-6 bearing an ester group ap-pears at = 62.5 ppm. The acylated
primary OH groupexhibits a downfield shift of about 3 ppm
comparedwith the corresponding carbon of the CH2OH func-tion. Again
DS values were determined by means of1H-NMR spectroscopy after
peracetylation of the re-maining OH groups. A representative 1H-NMR
spec-trum of cellulose acetate laurate (synthesized fromsample C4)
recorded in CDCl3 is shown in Figure 7.The protons of the laurate
moiety appear at 2.3 (H-8),1.21.6 (H-10-17) and 0.8 (H-18) ppm. The
acetatemethyl group leads to the signal at 1.9 (H-20) ppm.These
results are in very good agreement with valuesreported for a
cellulose acetate laurate synthesized inthe new solvent DMSO/TBAF,
applying vinyl laurateand acetic anhydride (see above). It was
found thatthe DS increased with the increasing carbon numberof the
carboxylic acid. Thus, a DS of 0.6 was found
-
292
Table 5. Conditions and results of esterification of cellulose
dissolved in DMA/LiCl mediated with Tos-Cl with different
carboxylic acids.
No. Carboxylic acid Molar ratioa Time (h) DSb Solubility
C1 Capric 1:2:2:0 24 1.31 DMFC2 Caprylic 1:2:2:0 24 1.40 DMSOc,
DMF, CHCl3C3 Decanoic 1:2:2:0 24 1.48 DMF, CHCl3, tolueneC4 Lauric
1:2:2:0 24 1.55 Toluene, CHCl3C5 Palmitic 1:2:2:0 24 1.60 Toluene,
CHCl3C6 Stearic 1:2:2:0 24 1.76 Toluene, CHCl3C7 Caprylic 1:2:2:4
24 1.76 DMF, CHCl3C8 Lauric 1:2:2:4 24 1.79 CHCl3C9 Palmitic
1:2:2:4 24 1.71 CHCl3C10 Stearic 1:2:2:4 24 1.92 CHCl3C11 Caprylic
1:1:1:0 24 0.63 DMSO, DMFC12 Lauric 1:1:1:0 24 0.36 InsolubleC13
Palmitic 1:1:1:0 24 0.46 InsolubleC14 Caprylic 1:4:4:0 24 2.56
Toluene, CHCl3C15 Lauric 1:4:4:0 24 2.56 Toluene, CHCl3C16 Palmitic
1:4:4:0 24 2.54 Toluene, CHCl3C17 Caprylic 1:2:2:0 4 1.27 DMF,
CHCl3C18 Lauric 1:2:2:0 4 1.55 CHCl3C19 Palmitic 1:2:2:0 4 1.50
CHCl3C20 Caprylic 1:2:2:0 1 1.25 DMSO, DMFC21 Lauric 1:2:2:0 1 1.36
InsolubleC22 Palmitic 1:2:2:0 1 1.36 CHCl3
a Mole AGU/mol carboxylic acid/mol Tos-Cl/mol pyridine.b DS
calculated by 1H NMR spectroscopy after peracetylation.c
Dimethylsulfoxide.
Figure 6. 13C-NMR spectrum of cellulose laurate C4 (DS = 1.55)
recorded in CDCl3 at 40 C, index means influenced by a
functionalizationof the neighbor position (number of scans
11,000).
for the cellulose caprate C1, while cellulose caprylateC2
possesses a DS of 1.4. Under comparable condi-tions, a cellulose
stearate C6 with a DS of 2.0 wasaccessible.
The cellulose esters possess a different solubil-ity depending
on their DS and the chain length ofthe carboxylic acid (Table 5).
In general, cellulosefatty acid esters having DS values higher than
1.4 are
-
293
Figure 7. 1H-NMR spectrum of cellulose acetate laurate (starting
polymer C4) recorded in CDCl3 at 40 C (16 scans were
accumulated).
soluble in CHCl3, independently of the chain length ofthe
carboxylic acid. Polymers with DS values higherthan 2.3 are soluble
in toluene.
In another series of experiments the influenceof an additional
base was investigated. Cellulosewas reacted with two equivalents of
carboxylic acidand Tos-Cl and four equivalents with pyridine
asbase. Thus, polymers C710 were obtained bearingcaprylic- (C7),
lauric- (C8), palmitic- (C9) and stearicester (C10) functions. It
was found that the DS valueswere higher than the samples prepared
without base(C16). For instance, a DS of 1.55 was found for
thecellulose laurate C4, synthesized without base. Theaddition of
base increases the DS to 1.79 (C8) (seeFigure 8). Elemental
analysis revealed the absence ofsulfur. Therefore, it can be
concluded that Tos-Cl actsonly as activating reagent. No tosylation
occurs.
GPC was applied to investigate hydrolytic de-gradation of the
polymer chain during the reaction.Cellulose palmitate C5,
synthesized in the absence ofbase, yielded a polymer with DP 41,
whereas cellulosepalmitate C9 obtained in the presence of base,
yiel-ded a DP value of 69. Similar results were obtainedfor
cellulose stearate C6 (without base, DP= 45) andC10 (with base, DP
= 61). Compared with the DP ofthe starting cellulose Avicel (DP
260), a fairly drasticdegradation occurred in every case.
Td were obtained by TGA for cellulose caprate(292 C), caprylate
(300 C), decanoate (301 C),laurate (302 C), palmitate (306 C) and
stearate(318 C). Cellulose esters C16 showed increasing
Figure 8. DS of cellulose esters synthesized in N,N-dimethyl
ace-tamide/LiCl using in situ activation with the Tos-Cl dependent
onthe carboxylic acid and the addition of pyridine (!) and
withoutpyridine (").
stability with an increase in chain length from C-6 toC-18. The
effect of the chain length on that behavioris more pronounced than
the influence of the rathersmall increase in DS of the samples
discussed. Theminimum Td value of cellulose laurate C4 was 292
C.The maximum Td value for cellulose stearate C6 was318 C. The
results of TGA were comparable with thereported behavior of
long-chain fatty acid esters ofcellulose (Sealey et al. 1996).
Synthesis and characterization of adamantoylcellulose prepared
via different paths
The esterification of cellulose with AdOH was stud-ied because
this ester moiety has found considerable
-
294
Table 6. Summary of reaction conditions and results of the
homogeneous reaction of cellulose with AdCl in DMA/LiCl.
No. Molar ratioa Time (h) Temperature (C) DSAdb Solubility
D2 1.0 24 20 0.24 InsolubleD3 1.3 24 20 0.51 DMSOc, pyridineD4
1.5 24 20 0.65 DMSO, pyridineD5 1.7 24 20 0.76 DMSO, pyridineD6 2.0
5 20 0.08 DMSO, pyridineD7 2.0 24 20 0.90 DMSO, pyridineD8 3.0 24
20 1.37 DMSO, pyridineD9 1.0 24 80 0.51 DMSO, pyridineD10 1.5 24 80
0.87 DMSO, pyridineD11 2.0 5 80 1.21 DMSO, pyridine, THFd
D12 2.0 24 80 1.71 Pyridine, THF, CHCl3D13 3.0 24 80 1.92
Pyridine, THF, CHCl3D14 4.0 24 80 1.94 Pyridine, THF, CHCl3D15 5.0
24 80 2.12 Pyridine, THF, CHCl3D16 2.0 24 35 1.19 DMSO, pyridine,
THFD17 2.0 24 50 1.38 DMSO, pyridine, THFD18 2.0 24 65 1.57 DMSO,
pyridine, THF
a Mole AdCl per mol AGU.b DS of adamantoyl functions determined
after perpropionylation by 1H NMR spectroscopy.c
Dimethylsulfoxide.d Tetrahydrofurane.
interest as a base-sensitive protecting group
indeoxyribonucleoside chemistry (Greene and Wuts1991). On the other
hand, incorporation of adamantoylfunctions leads to products with
interesting biolog-ical activities, such as antimicrobial and
antibacterialactivity (Orzeszko et al. 2000; Perrakis et al. 1999)
aswell as antitumor activity (Gerzon and Kau 1967).
For the activation of the carboxylic acid, differentmethods were
studied, namely acid chloride as wellas in situ activation with
Tos-Cl and CDA were ap-plied. All reactions were carried out
homogeneouslyin DMA/LiCl. Thus, cellulose dissolved in DMA/LiClwas
allowed to react with AdCl in the presence ofpyridine (Grbner et
al. 2002).
The conversion of Avicel with AdCl leads to theadamantoylated
products D218 (Table 6), whichshow the typical IR spectra.
Absorption bands ofthe cellulose backbone were found and,
additionally,a signal at 1758 cm1, (C==OEster), indicating
thepresence of the ester moiety. As can be seen fromTable 6, DS
values up to 2.12 were accessible via thispath. With regard to the
reaction efficiency, that is,the amount of carboxylic acid bound to
the polymerrelated to the molar ratio, the use of 2 mol AdCl permol
AGU is most effective. About 85% of the AdClreacts with the
cellulose backbone (sample D12).
The reaction of the dissolved cellulose with AdOHin the presence
of Tos-Cl leads to correspondingcarboxylic acid esters with a
rather high DS of 1.50and 1.75 at a molar ratio of AGU/AdOH/TosCl
of1/2/2 and 1/3/3, respectively within 24 h at 80 C re-action
temperature (samples D20 and D21, Table 7).Surprisingly, conversion
at a molar ratio of 1/1/1(sample D19) at 80 C as well as at room
temperaturedoes not yield sufficient DS.
Comparable results were obtained by the conver-sion of cellulose
with AdOH in the presence of CDA.This method is especially suitable
for cellulose modi-fication because the pH is not drastically
changedduring the conversion and consequently diminishedchain
degradation can be guaranteed. The by-productsformed during the
reaction (Figure 9) are only CO2and imidazole, and none of the
reagents and by-products are toxic. At room temperature, even at
amolar ratio of 1/3/3 (AGU/AdOH/CDA), no celluloseester was formed.
On the other hand, at the compa-rable molar ratio of 1/3/3, however
at 80 C, productD26 with a DS of 1.31 is obtained, which is rather
lowcompared to product D13 (DS = 1.92), synthesizedwith AdCl under
the same conditions. It is interestingthat the increase in the
amount of cellulose from 1 to10 g of starting material results in
higher DS values
-
295Table 7. Conditions and results of the reaction of cellulose
dissolved in DMA/LiCl with AdOH after in situactivation with Tos-Cl
(method A) or CDA (method B).
No. Method Molar ratioa Time (h) Temperature (C) DSb
D20 A 1:2:2 24 80 1.50D21 A 1:3:3 24 80 1.75D22 Bc 1:3:3 6 80
0.62D23 Bc 1:3:3 8 80 0.90D24 B 1:1:1 24 80 0.54D25 B 1:2:2 24 80
0.98D26 B 1:3:3 24 80 1.31D27 Bc 1:3:3 24 80 1.42
a Mole AGU/mol AdOH/mol Tos-Cl or CDA.b DS of adamantoyl
functions determined after perpropionylation by 1H NMR
spectroscopy.c Amount of cellulose 10 g.
Figure 9. Schematic plot of the conversion of cellulose
withcarboxylic acid applying in situ activation with CDA.
under comparable conditions. Thus, a product with aDS of 1.42
was synthesized (sample D27, Table 7).
The structures of the synthesized adamantoyl cel-luloses were
confirmed by means of 13C NMR- and 1HNMR spectroscopy, including
two-dimensional meth-ods (Grbner et al. 2002). Adamantoyl
celluloses D7,D10, D11, D17 and D21, of different DSAd, were
se-lected for screening for biological activity. It is knownthat
adamantoylated nucleosides are biologically act-ive because the
adamantoyl moiety binds precisely toa complementary hydrophobic
receptor region of pro-tein molecules (Gerzon and Kau 1967). Thus,
the an-timicrobial, antioxidant and anti-inflammatory effectsof the
cellulose derivative were tested. All samples in-vestigated did not
show antioxidant or anti-microbialeffects. However, polymers D7,
D11 and D21 pos-sessed a strong anti-inflammatory effect while
D10and D17 showed no effect. Obviously, the effect doesnot only
depend on the DSAd, because samples D7and D10 or D11 and D17 have
similar DS, however,
polymer D10 and D17 are not active at all. Furtherstudies about
the molecular weight, its distributionas well as about the
biological dependence of theadamantoyl cellulose activity on the
physico-chemicalproperties are needed to gain clear
structurepropertyunderstanding.
Conclusions and outlook
The results of our work show that polymeric bases,such as
polyvinyl pyridine and the new cellulosesolvent DMSO/TBAF, are new
tools for the prepa-ration of cellulose esters. Furthermore, the
efficiencyof the in situ activation of the carboxylic acids withCDA
and Tos-Cl, respectively, has been shown. Thus,a broad variety of
cellulose esters with tailored prop-erties, for example,
solubility, biological activity orthermal behavior can be
synthesized. These featuresare adjustable by the type of
substituent introducedand by the pattern of substitution, which can
be modi-fied by application of one of the methods described.
Unfortunately, the synthesis of cellulose esterswith an
unconventional (non-statistical) distributionof substituents along
the polymer chain has still notbeen achieved. Therefore, we
continue our work onthe application of polymeric reagents for ester
syn-thesis. On the other hand, the analysis of the patternof
substitution on this level is not yet satisfactory.The HPLC
analysis, via methylation and depolymer-ization, is connected to a
number of errors, mainlyundermethylation, migration of the
substituents duringthe second substitution and incomplete
depolymeriza-tion. Thus, at present, a comparable method is
under
-
296
investigation, applying percarbanilation of the cellu-lose ester
as the second derivatization step.
Acknowledgements
Financial support for this study (project HE2054/5-3)was
provided as part of the focus research program(Schwerpunktprogramm)
on Cellulose and cellulosederivatives molecular and supramolecular
structur-al design by the Deutsche Forschungsgemeinschaft(DFG)
[described in the editorial commentary ofProf G. Wenz in Cellulose
10-1] and by the Fondsder Chemischen Industrie. The authors would
like tothank former co-worker D. Grbner, and PhD studentG.T. Ciacco
(Instituto de Quimica des Sao Carlos,Universidade de Sao Paulo,
Brazil) having stayedfor six months in my group; they were included
inthe basic research program on the acylation of cel-lulose.
Furthermore, we thank Dr W. Radosta andDr W. Vorwerg (Fraunhofer
Institut fr AngewandtePolymerforschung, Golm, Germany) for GPC
studies.Moreover, the authors wish to thank M. Ktteritzschfor
technical assistance.
References
Ass B.A.P. and Frollini E. 2001. Aggregation of cellulose
duringdissolution and acetylation in
N,N-dimethylacetamide/lithiumchloride: An introductory study. An.
Assoc. Bras. Qum. 50:7682.
Ciacco G.T., Ass B.A.P., Ramos L.A. and Frollini E. 2000.
Acet-ylation of cellulose under homogeneous reaction conditions.In:
Mattoso L.H.C., Leo A.L. and Frollini E. (eds), NaturalPolymers and
Composites. So Carlos, pp. 139145.
Dawsey T.R. 1994. Applications and limitations of
LiCl/N,N-dimethylacetamide in the homogeneous derivatization of
cellu-lose. In: Gilbert R.D. (ed), Cellulosic Polymers, Blends
andComposites. Hanser Publishers, Mnchen, Wien, New York,
pp.157171.
Einfeldt J., Heinze T., Liebert T. and Kwasniewski A. 2002.
In-fluence of the p-toluenesulfonylation of cellulose on the
poly-mer dynamics investigated by dielectric spectroscopy.
Carbohyd.Polym. 49: 357365.
El Seoud O.A., Regiani A. and Frollini E. 2000. Derivatiza-tion
of cellulose in homogeneous conditions. A Brief Review.In: Frollini
E., Leo A.L. and Mattoso L.H.C. (eds), NaturalPolymers and
Agrofibers Composites. So Carlos, pp. 7390.
Erler U., Mischnick P., Stein A. and Klemm D. 1992.
Determinationof the substitution pattern of cellulose methyl ethers
by HPLCand GLC comparison of methods. Polym. Bull. 29: 349356.
Gerzon K. and Kau D. 1967. The adamantoyl group in medi-cinal
agents. III. Nucleoside 5-adamantoates. The adamantoylfunction as a
protecting group. J. Med. Chem. 10: 189198.
Grbner D., Liebert T. and Heinze Th. 2002. Synthesis of
noveladamantoyl cellulose using differently activated carboxylic
acidderivatives. Cellulose 9: 193201.
Greene T.W. and Wuts P.G.M. 1991. Protective Groups in
OrganicSynthesis. John Wiley & Sons, Inc., New York, p.
100.
Heinze Th. and Liebert T. 2001. Unconventional methods in
cellu-lose functionalization. Prog. Polym. Sci. 26: 16891762.
Heinze Th., Dicke R., Koschella A., Kull A.H., Klohr E.-A.
andKoch W. 2000. Effective preparation of cellulose derivatives ina
new simple cellulose solvent. Macromol. Chem. Phys. 201:627631.
Heinze T. and Schaller J. 2000. New water soluble cellulose
es-ter synthesized by an effective acylation procedure.
Macromol.Chem. Phys. 201: 12141218.
Heinze T., Liebert T., Klfers P. and Meister F. 1999.
Carboxy-methylation of cellulose in unconventional media. Cellulose
6:153165.
Koschella A., Haucke G. and Heinze Th. 1997. New
fluorescenceactive cellulosics prepared by a convenient acylation
procedure.Polym. Bull. 39: 597604.
Ktz J., Bogen I., Heinze Th., Heinze U., Kulicke W.-M. andLange
S. 2001. Peculiarities in the physico-chemical behaviorof
non-statistically substituted carboxymethylcelluloses.
ColloidSurface. A 183185: 621633.
Liebert T. and Heinze Th. 1998a. Induced phase separation: Anew
synthesis concept in cellulose chemistry. In: Heinze Th.J.and
Glasser W.G. (eds), Cellulose Derivatives: Synthesis,
Char-acterization and Nanostructures. ACS Symposium Series
688,American Chemical Society, Washington, DC, p. 61.
Liebert T. and Heinze T. 1998b. Synthesis path versus
distributionof functional groups in cellulose ethers. Macromol.
Symp. 130:271283.
Liebert T. and Heinze T. 2001. Exploitation of reactivity
andselectivity in cellulose functionalization using
unconventionalmedia for the design of products showing new
superstructures.Biomacromolecules 2: 11241132.
Liebert T., Heinze Th. and Klemm D. 1996. Synthesis and
carboxy-methylation of organo-soluble trifluoroacetates and
formates ofcellulose. J. Macromol. Sci. Pure A33: 613626.
Marson G., Regiani A., Frollini E. and El Seoud O.A. 1999.
Cellu-lose esterification in homogeneous medium. J. Polym. Sci.
Pol.Chem. 37: 13571363.
Marson G., Ciacco G.T., Frollini E. and El Seoud O.A. 2000.An
efficient one pot acylation of cellulose under homogeneousreaction
conditions. Macromol. Chem. Phys. 201: 882889.
Orzeszko A., Gralewska R., Starosciak B.J. and Kazimierczuk
Z.2000. Synthesis and antimicrobial activity of new
adamantanederivatives I. Acta Biochim. Pol. 47: 8794.
Perrakis A., Antoniadou-Vyza E., Tsitsa P., Lamzin V.S.,
WilsonK.S. and Hamodrakas S.J. 1999. Molecular, crystal and
solutionstructure of a beta-cyclodextrin complex with the bromide
saltof 2-(3-dimethylaminopropyl)tricyclo[3.3.1.1(3,7)]decan-2-ol,
apotent antimicrobial drug. Carbohyd. Res. 317: 1928.
Saake B., Horner S., Kruse Th., Puls J., Liebert T. and
HeinzeTh. 2000. Detailed investigation on the molecular structure
ofcarboxymethyl cellulose with unusual substitution pattern bymeans
of enzyme-supported analysis. Macromol. Chem. Phys.201:
19962002.
Samaranayake G. and Glasser W.G. 1993a. Cellulose
derivativeswith low DS. 1. A novel acylation system. Carbohyd.
Polym. 22:17.
Samaranayake G. and Glasser W.G. 1993b. Cellulose
derivativeswith low DS. 2. Analysis of alkanoates. Carbohyd. Polym.
22:7986.
Sealey J., Samaranayake G., Todd J. and Glasser W.G. 1996.
Novelcellulose derivatives. IV. Preparation and thermal analysis
ofwaxy ester of cellulose. J. Polym. Sci. 34: 16131620.
Shimizu Y. and Hayashi J. 1988. A new method for
celluloseacetylation with acetic acid. Sen-I Gakkaishi 44:
451456.
Sun R.C., Fang J.M., Tomkinson J., Geng Z.C. and Liu J.C.
2001.Fractional isolation, physico-chemical characterization and
ho-mogeneous esterification of hemicelluloses from
fast-growingpoplar wood. Carbohyd. Polym. 44: 2939.