Synthesis and characterization of camphorsulfonyl acetate of cellulose

Article abstract of DOI:10.1016/j.carres.2004.06.009

Novel cellulose derivatives were prepared from reacting (1R)-(+)-camphor-10-sulfonic chloride (CSC) with cellulose acetate (CA) in acetone and triethylamine. The reaction conditions, including reaction time and reactant molar ratios, were optimized. The structure of the products was confirmed by means of ¹H NMR, ¹³C NMR, FT-IR and elementary analysis. The techniques were also used to determine the degree of the substitution of camphorsulfonyl groups (DS_CS). The data calculated from ¹H NMR, ¹³C NMR, percent grafting (G%) and elementary analysis coincided with those from chemical analysis. Compared to cellulose acetate, the cellulose derivatives exhibited decreased thermal stability, improved solubility in organic solvents and enhanced enantioselectivity towards tyrosine isomers. The solubility and enantioselectivity increased with increasing degrees of camphorsulfonyl substitution.

Full text of DOI:10.1016/j.carres.2004.06.009

Carbohydrate
RESEARCH
Carbohydrate Research 339 (2004) 1925–1931
Synthesis and characterization of camphorsulfonyl
acetate of cellulose
a
Dingshu Xiao, Jiwen Hu,
a,b,
b
Mingqiu Zhang, Mingwei Li,
a
*
a
Guozhi Wang and Haisong Yao
a
a
Key Laboratory of Polymer Materials for Electronics of Guangdong Province, Guangzhou Institute of Chemistry,
Polymer Science, Chinese Academy of Sciences,
Guangzhou Wushan Leyiju 1122, Guangzhou 510650, PR China
b
Key Laboratory for Polymeric Composite and Functional Materials of Ministry of Education,
Zhongshan University, Guangzhou 510275, PR China
Received 20 March 2004; received in revised form 23 May 2004; accepted 13 June 2004
Abstract—Novel cellulose derivatives were prepared from reacting (1R)-(+)-camphor-10-sulfonic chloride (CSC) with cellulose
acetate (CA) in acetone and triethylamine. The reaction conditions, including reaction time and reactant molar ratios, were opti-
1
13
mized. The structure of the products was confirmed by means of H NMR, C NMR, FT-IR and elementary analysis. The
techniques were also used to determine the degree of the substitution of camphorsulfonyl groups (DSCS). The data calculated from
1
13
H NMR, C NMR, percent grafting (G%) and elementary analysis coincided with those from chemical analysis. Compared to
cellulose acetate, the cellulose derivatives exhibited decreased thermal stability, improved solubility in organic solvents and enhanced
enantioselectivity towards tyrosine isomers. The solubility and enantioselectivity increased with increasing degrees of camp-
horsulfonyl substitution.
ꢀ
2004 Elsevier Ltd. All rights reserved.
Keywords: Cellulose acetate; Sulfonylation; Thermal stability; Enantioselectivity; Solubility
1
. Introduction
stationary phases in HPLC and membranes for enan-
4
;5
tioseparation. Considering that (1R)-(+)-camphor-10-
sulfonic acid (HCSA) has acted as chiral resolution
Techniques for preparative-scale enantiomer resolution
are urgently needed because of the growing interests of
the pharmaceutical industry in single-enantiomer
6
agent in crystallization and chromatography, and the
camphor structure proves to be a very useful chiral
probe in HPLC, NMR and X-ray crystallographic
1
drugs. It is known that an ideal enantioselective
7
polymer membrane material must possess all of the
following properties: enantioselectivity, permeability,
mechanical strength, self-membrane forming ability and
cost effectiveness. As an important and traditional
cellulose derivative, cellulose acetate (CA) has been
widely used in producing fibres, selective membrane,
bioactive and biocompatible materials, etc. Its deriva-
tives with chiral recognition sites are now used as chiral
determination of absolute configuration, we expected
that a material consisting of cellulose acetate and chiral
camphor sulfonic acid might well fulfil the above
requirements of enantioselective membranes owing to
the chiral environment generated by the chiral active
carbons on the ring backbone of cellulose and the
optically active bulky pendants. To the best of our
knowledge, the synthesis of camphorsulfonyl acetate of
cellulose (CA–O–CS) has not yet been reported. We
report in this paper the preparation of CA–O–CS and its
use to prepare novel enantioselective membranes.
Enantioselectivity of the material is demonstrated by the
2;3
*
0
Corresponding author. Tel./fax: +86-20-85232307; e-mail: hujiwen@
mail.gic.ac.cn
008-6215/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2004.06.009

1926
D. Xiao et al. / Carbohydrate Research 339 (2004) 1925–1931
selective adsorption
CA–O–CS power and the selective permeation of
L
-tyrosine over
-
tyrosine through the CA–O–CS membranes. The degree
D
-tyrosine by the
D
hydroxyl groups were sulfonylated, while only 8–9% free
hydroxyl groups remained on the anhydroglucose unit
(AGU). Although the degree of camphorsulfonyl sub-
stitution is not high, further sulfonylation of the residual
hydroxyl groups are rather difficult probably due to the
of substitution of camphorsulfonyl group (DS ) was
CS
determined by various approaches. The properties of the
modified polymer, including solubility and thermal sta-
bility are also characterized for practical application
8
steric effect. In this context, the camphorsulfonyl ace-
tate of cellulose with DSCS of 0.33 derived from CSC
and CA using 5 wt & catalyst of 4-dimethylaminopyr-
idine and a raw molar ratio of CSC:AGU-OH ¼ 1.5:1 in
acetone/triethylamine should be the optimistic product.
purposes as a function of DSCS
.
13
2
. Results and discussion
Figure 1 gives the typical C NMR spectrum of CA–
O–CS (in Me SO-d ). The assignments are made
2
6
CA–O–CS was obtained by reacting CA with CSC. The
reaction was performed under different conditions with
results shown in Table 1. The data in Table 1 indicated
that the percent grafting is greatly influenced by the
nature of the solvent and the organic base used as the
acidic scavenger. Pyridine has been reported to act as
the solvent and the organic base in further esterification
of cellulose acetate. however, the best yields were
achieved by us in the current work using acetone and
triethylamine as solvent and acid scavenger, respec-
tively. It might result from the fact that triethylamine
salt formed after neutralization precipitated from the
solvent phase. The dry and mild reaction conditions
used avoided chain degradation that was quite sensitive
to alkaline and acidic attacks in water.
according to the differences in intensity, line shape and
position of the signals. The signals at d 25.0 (C-7), 26.2
(C-5), 40.1 (C-4), 42.3 (C-6), 47.8 (C-10) and 57.6 (C-1)
are attributed to the carbon of camphorsulfonyl (CS)
groups, while the peak d 19.3 is assigned to the dimethyl
carbon of CS groups. The signals of carbonyl carbon of
CS and acetyl groups appear at d 213.8 and 169.4–170.4,
respectively. The well-resolved lines for C-1 (d 99.2), C-4
(74.8), C-6 (62.2) carbon atoms of the anhydroglucose
unit, as well as the poorly-resolved lines for C-2, -3, -5 (d
71.40–68.1) carbon atoms appearing at d 62.2–99.6 are
attributed to the signals of cellulose backbone. In
addition, the peak of C-3 of camphorsulfonyl groups
and that of the anhydroglucose unit partly overlap.
8
;9
The structures of the resultant derivatives were con-
1
Table 1 also shows that the percent grafting increases
with increasing reaction time up to 24 h. After that
percent grafting increases significantly with time. When
the reaction was completed, about half of the available
firmed by the H NMR spectrum in CDCl (Fig. 2). The
peak assignments were performed by comparing the
peak areas, and by comparing the spectrum with that of
3
10
cellulose. The proton signals of camphorsulfonyl
Table 1. Effects of reaction conditions on DSCS
DSCS
Cat. (&)
Concentration (mol/L) Reaction time (h)
CSC/AGU-OH (mol/mol)
Medium
Percent grafting (%)
0.04
0.05
0.08
0.19
0.23
0.25
0.32
0.33
5
5
1
5
5
5
5
5
0.38
0.24
0.38
0.24
0.24
0.24
0.21
0.21
20
24
15
20
20
24
24
48
2:1
1.5:1
1.5:1
2:1
Pyridine
Pyridine
3.16
4.01
Acetone/(Et)
Acetone/(Et)
Acetone/(Et)
Acetone/(Et)
Acetone/(Et)
Acetone/(Et)
3
3
3
3
3
3
N
N
N
N
N
N
6.29
15.35
17.23
20.12
25.67
26.13
1.5:1
1.5:1
1.5:1
1.5:1
13
Figure 1. Typical C NMR spectra of CA–O–CS.

D. Xiao et al. / Carbohydrate Research 339 (2004) 1925–1931
1927
ꢀ
1
of S@O and the deformations of C–H (1369 cm ), the
symmetry stretching of S@O and stretching of C–O–C
ꢀ
1
(1166 cm ), the skeletal stretching of C–O of pyranose
ring (1049 cm ) and the stretching of S–O (906 and
ꢀ
1
ꢀ
1
8
31 cm ). It is seen that the differences between the two
spectra in Figure 3 are relatively unobvious, mainly due
to the fact that the asymmetrical and symmetrical
ꢀ
1
stretching modes of S@O (maso2 1420–1330 cm , and mso2
ꢀ
1
1
200–1140 cm ) of camphorsulfonyl groups partly
conceal the peaks of C–H deformation and C–O anti-
symmetrical stretching of acetyl, respectively. In addi-
tion, the asymmetry stretching band of C@O of CS is
also hard to be distinguished because of the wide scat-
tering of C@O band of cellulose acetate (1700–
1
3
Figure 2. Typical H NMR spectra of CA–O–CS (in CDCl ).
ꢀ
1
1800 cm ). Nevertheless, the structural changes can still
be confirmed by the enhanced intensity of the peaks at
substitution mainly appearing at the high magnetic field
were assigned as follows: H-8 and -9 (d 0.833–1.232,
H), H-5 (1.119–1.459, 2H), H-4 (1.662, 1H) and H-6,
ꢀ1
2900, 1750, 1370 and 1170 cm , and the appearance of
the stretching modes of S–O at 982 and 831 cm .
6
ꢀ1
H-10 (2.384, 2H, 3.048, 2H), respectively. The proton
resonance of the AGU is shown in the d 3.355–5.050.
When the OH groups were substituted, the peaks
The degree of camphorsulfonyl substitution (DS )
CS
1
13
was estimated by chemical analysis, H NMR,
C
NMR, elementary analysis and percent grafting deter-
mination. Chemical analysis, which has been tradition-
ally used to estimate the degree of acetyl substitution in
cellulose acetate, works also for quantifying the content
of camphorsulfonyl groups in CA–O–CS. Since the
acetyl groups of the cellulose derivatives can be com-
pletely hydrolyzed without affecting the camphorsulfo-
nyl groups under the conditions of chemical analysis, the
camphorsulfonyl groups bonded to the AGU result in a
rise of the molecular weight of the cellulose units, and
hence the reduced apparent content of acetate acid.
Accordingly, the degree of camphorsulfonyl substitution
is obtained from the variations in the apparent per-
(
AGU-H-3 and H-6) shifted to lower magnetic field than
that of cellulose. The wide and strong signals in the d
.044–3.685 can be attributed to the resonance of free
3
hydroxyl groups remaining on AGU other than that of
H-5 and H-3 at AGU and camphorsulfonyl, respec-
tively. In contrast to the case of CSC, the signal of
camphorsulfonyl groups adjacent to CA skeletons shif-
ted towards lower magnetic field regime due to the
deshielding effect of cellulose backbone.
The spectra in Figure 3 reveal that the resultant
derivatives exhibit not only the characteristic bands of
skeletal vibration of cellulose but also the scattering of
the newly introduced sulfonyl ester. That is, the
1
centage of acetic acid. In H NMR spectra, estimation
of DSCS was carried out by measuring the peak area
ꢀ
1
stretching modes of O–H of AGU (3494 cm ) and C–H
ꢀ
1
of methyl and methylene (2990–2850 cm ), the asym-
metry and symmetry stretching of C@O of acetyl (1751
and 1232 cm , respectively), the asymmetry stretching
ratio of acetyl methyl (d 1.921–2.094) to sulfonyl methyl
13
of C-10 (d 0.833–1.232). For C NMR spectra, DSCS
was yielded from the intensity ratio of the acetyl car-
ꢀ
1
bonyl carbon peak (ꢁ170 ppm) to that of CS
(
ꢁ213 ppm). The values of DSCS determined by various
approaches are plotted against those from chemical
analysis in Figure 4. Evidently, the DSCS data are in
good agreement with each other. It is therefore believed
that the procedure based on chemical analysis described
in the present paper offers a reliable, economic and
simple way for quantification of the substituting degree
of partially sulfonylated acetate cellulose and other
analogous cellulose derivatives.
The curves of thermogravimetry (TG) and derivative
thermogravimetry (DTG) in Figure 5 and the corre-
sponding data summarized in Table 2 reveal that the
thermal stability and thermal degradation behaviour of
cellulose acetate have changed considerably after
camphorsulfonyl attachment. Three distinct weight loss
processes are observed for CA and CA–O–CS: dehy-
dration, 28–46% weight loss resulting from a one-step
Figure 3. Annotations on the bands in FT-IR spectra of CA and CA–
O–CS.

1928
D. Xiao et al. / Carbohydrate Research 339 (2004) 1925–1931
11
acids. The following issues might be responsible for the
decay. (1) The camphorsulfonyl groups act as autocat-
alysts deteriorating the thermal stability of CA–O–CS
and promoting char formation because the decomposi-
tion product (i.e., SO ) would accelerate dehydration of
2
12
hydroxyl groups. (2) The degradation processes of CA
and its derivatives are similar to that of cellulose,
13;14
including dehydration and depolymerization.
The
introduction of large camphorsulfonyl groups results in
partial destruction of inter- and inner-molecular
hydrogen bonds, and hence increases the distance be-
tween the macromolecular chains, which would intensify
decomposition of CA–O–CS. (3) The crystallinity of CA
decreases after the substitution of AGU free hydroxyls
with bulk camphorsulfonyl groups.
1
Figure 4. DSCS values determined by elementary analysis (E),
H
The introduction of camphorsulfonyl groups signifi-
cantly enhances the solubility of cellulose acetate in
many organic solvents, such as benzene and dichlo-
romethane (see Table 3). The higher DSCS is, the better
the solubility is. However, the derivatives of cellulose
acetate still cannot dissolve in water and methanol. The
solubility behaviours of the modified polymer can be
partly attributed to the variation in its molecular
structure. That is, the hydrophobic alkyl and bulky ring
moieties, and the hydrophilic AGU and sulfonic ester
fragments should take the responsibility. Furthermore,
the increased inter-molecular gaps, created by the
introduction of the bulky substituents, decrease inter-
chain forces and facilitate dissolution. The improvement
of the solubility has broadened the fabrication window
for membrane forming. The polymer can be easily
turned into membrane-like material by the solution
casting method.
13
NMR, C NMR and percent grafting (G%) in relation to those by
chemical analysis.
In the present study, enantioselectivity of the newly
synthesized products was characterized by their optical
Figure 5. TG and DTG curves of CA (curve A) and CA–O–CS (curve
B with DSCS of 0.08, and curve C with DSCS of 0.20).
rotation, selective adsorption and permeability of D-
tyrosine and -tyrosine, respectively (Table 4). Optical
L
pyrolysis, and complete oxidation of the residue. The
main decomposition process (from temperature T
activity is proportional to the total contributions from
all the stereocentres in the polymer. The optical rota-
tion, ½aꢂ , which is a function of DS , may thus be
1
2
to T )
of modified CA is shifted from 370–388 ꢁC (for CA) to
D
CS
2
with higher DSCS has more char residue at the final stage
30–260 ꢁC (for CA–O–CS). Besides, the CA–O–CS
regarded as a characteristic signature of the products.
The larger ½aꢂ is, the higher DS is. This demonstrates
CS
D
of pyrolysis (T ꢁ 550 ꢁC) but lower peak decomposition
the success of chiral modification and likelihood
enantioselectivity of the camphorsulfonyl acetate of
cellulose. Additionally, the higher stereospecific
adsorption and the lower enantioselective permeation
f
M
temperature (T ). It means that the thermal stability of
the resultant derivatives decreases with a increase in
DSCS, which is contrary to the case when cellulose hy-
droxyls are substituted by some long chain organic
for
L-tyrosine than D-tyrosine summarized in Table 4
Table 2. Weight loss and the characteristic temperatures of CA and its derivatives
DSCS
First process of weight loss
Second process of weight loss
Third process of weight loss
T
M
(ꢁC)
Residual weight
at 550 ꢁC (%)
T
1
(ꢁC)
Weight loss (%)
T
2
(ꢁC)
Weight loss (%)
T
3
(ꢁC)
Weight loss (%)
0
50–315
50–229
50–227
1.56
1.25
1.03
315–408
229–269
227–254
84.16
67.27
60.82
408–550
269–550
254–550
6.61
16.41
22.30
383.6
254.0
247.2
7.67
15.07
15.85
0
0
.08
.20
The definitions of the parameters are given in Figure 5.

D. Xiao et al. / Carbohydrate Research 339 (2004) 1925–1931
1929
Table 3. Influence of DSCS on solubility
a
DSCS
Solubility
H
2
O
C
6
H
6
Me
2
SO
CH
2
Cl
2
CHCl
3
CH
3
COOC
2
H
5
Acetone
0
0
0
0
0
––
––
––
––
––
––
±
+
+
+
+
+
––
±
––
±
––
±
+
+
+
+
+
.09
.15
.23
.32
±
+
+
+
+
+
+
+
+
+
+
+
–
–: Insoluble; +: soluble; ±: highly swollen.
a
Visually judged.
Table 4. Influence of DSCS on enantioselective recognition ability
2
D
0a
DSCS
½aꢂ
Adsorption (%)
Permeation rate
b
c
A
d-tyrosine
A
l-tyrosine
%ee
A
P
d-tyrosine
P
l-tyrosine
%ee
P
ꢀ5
ꢀ6
ꢀ6
ꢀ6
ꢀ6
ꢀ5
ꢀ6
ꢀ6
ꢀ7
ꢀ7
0
7.32
17.86
21.45
26.38
28.54
6.2
10.5
11.6
12.1
12.8
8.3
16.8
20.8
25.4
30.7
14.5
23.1
28.4
35.5
41.1
3.67 · 10
7.43 · 10
2.89 · 10
1.54 · 10
1.21 · 10
3.56 · 10
5.85 · 10
1.85 · 10
8.24 · 10
6.38 · 10
1.52
11.89
21.94
30.28
30.95
0.09
0.15
0.23
0.32
a
b
c
2
Measured in a 1–2 g/dL (w/v) Me SO solution.
%
ee
A
¼ ðAl-tyrosine ꢀ Ad-tyrosineÞ=ðAl-tyrosine þ Ad-tyrosineÞ ꢃ 100%.
¼ ðPd-tyrosine ꢀ Pl-tyrosineÞ=ðPd-tyrosine þ Pl-tyrosineÞ ꢃ 100%.
%
ee
P
reveal clearly the enantioselectability of resultant
derivatives. The introduction of chiral camphorsulfonyl
groups has improved the enantioselective adsorption of
cellulose acetate, which is evidenced by the higher ste-
0 ꢁC. The mixture was stirred for 30 min at 0 ꢁC and then
heated to 40–50 ꢁC for 2 h. SOCl was removed via re-
2
duced pressure distillation. The crude material, followed
by repeated recrystallization in ligroin, then dried under
vacuum, gave the desired product (56 g, 98%), colour-
reospecific tendency for L-tyrosine than D-tyrosine in all
the cases. The higher DSCS, the higher enantiomer excess
percent (%ee) of the adsorption. The best enantioselec-
tive adsorption is achieved at %ee of 41.1. Furthermore,
the presences of chiral camphorsulfonyl groups result in
reduce of the permeation rate, but the permeation rate
less, mp 65–66 ꢁC (Ref. 18: 63–66 ꢁC); ½aꢂ +32 (c 1.0,
D
ꢀ
1
CHCl ); IR (mulled in mineral oil): m 1739 cm (C@O),
1 1
3
ꢀ
ꢀ1
1376 and 1184 cm (O@S@O), 586 cm (S–Cl);
H
NMR (400 MHz, CDCl ): d 0.902 (s, 3H, H-9), 1.117 (s,
3
3H, H-8), 1.471 (m, 1H, H-4), 1.751 (m, 1H, H-4), 1.946
(d, 1H, J3;2 18.4 Hz, H-3), 2.139 (t, 2H, J4;5 4.4, H-5),
3.684 (d, 1H, J2;3 18.4, H-2), 4.265 (d, 1H, J2;3 18.4, H-2);
of
l-tyrosine) in all case. The reduced diffusion rate of
-tyrosine relative to that of -tyrosine should have
resulted from the increased affinity of the -isomer
D
-tyrosine (Pd-tyrosine) is higher than that of L-tyrosine
(
P
13
L
D
C NMR (400 MHz, CDCl
(C-7), 26.7 (C-4), 42.2 (C-3), 42.6 (C-5), 48.1 (C-10), 59.5
(C-6), 64.2 (C-3); Anal. Calcd for C10 Cl : C,
3
): d 19.5 (2C, 2CH
3
), 25.1
L
owards camphorsulfonyl acetate of cellulose, which is in
agreement with the result of static adsorption. The
strong interaction leads to increased adsorption capacity
but decreased permeability of the membrane. The
appropriate amphiphilic performance of the membrane
material might play a significant role in the enantio-
H
15
O
3
S
1
1
47.90; H, 5.99; S, 12.77. Found: C, 47.32; H, 5.84; S,
12.40.
3.2. Chemical modification of CA with CSC
15
selective recognition.
Commercially available cellulose acetate was used in the
present study. Its total degree of acetyl substitution
(DSAC) is 2.42 and the individual degree of substitution
3
. Experimental
(IDSAC) at )2, )3 and )6 positioned hydroxyls in the
anhydroglucose are 0.85, 0.84, 0.73, respectively. DSAC
and IDSAC were determined by chemical analysis and
propanoation method, respectively. Shown in Scheme 1,
the cellulose acetate typically modified with CSC
is processed as follows: CSC solution (3 g, 11 mmol
of CSC dissolved in 20 mL acetone) was dripped to
CA solution (5 g, 19 mmol of CA dissolved in 50 mL
dry acetone) in ice-bath accompanied by stirring and
3
(
.1. Synthesis of (1R)-(+)-camphor-10-sulfonyl chloride
CSC)
(1R)-(+)-camphor-10-sulfonyl chloride (CSC) acting as
the chiral auxiliary was prepared from (1R)-(+)-cam-
phor-10-sulfonic acid (HCSA). SOCl2 (70 mL,
962 mmol) was dripped into HCSA (50 g, 217 mmol) at
16

1930
D. Xiao et al. / Carbohydrate Research 339 (2004) 1925–1931
Scheme 1. Schematic drawing of the synthesis of CA–O–CS.
nitrogen protection. The CA solution containing
4-dimethylaminopyridine (0.025 g) as the catalyst and
Et N (1.5 mL, 11 mmol, with the same equivalent as that
additional 15 min. Finally, the sample was cooled to
room temperature and titrated with 0.5 M NaOH after
72 h. After 10 h the solution was back-titrated with
0.5 M HCl. Phenolphthalein was used as the indicator.
The DSCS values of the samples were deduced from the
following formula:
3
of AGU-OH) was stirred at 0–5 ꢁC for 3 h, and then at
room temperature for 2 days. Eventually, it was poured
into MeOH. The white fine precipitate was collected and
purified by repeatedly washing. Further purification of
the product was performed by Soxhlet extraction with
MeOH for the first 2 h, after which the solvent was
changed for fresh one and the extraction was conducted
for the additional 24 h. Then the end product was dried
under vacuum at 85 ꢁC and stored in desiccators. Yield:
6
000DSAC
Acetic acid ð%Þ ¼
162 þ DS þ 214DS
AC
CS
ðD ꢀ CÞNNaOH ꢀ ðA ꢀ BÞNHCl
¼
ꢃ 6:005
W
7
.29 g, 99.2%.
where D and B represent millilitres of HCl and NaOH
solutions consumed for titration of blank, respectively.
C and A denote millilitres of HCl and NaOH solutions
consumed for titration of the sample. W means grams of
the sample. DSAC ¼ 2.42 representing the total degree of
substitution of acetyl group in the sample.
3
.3. General methods
1
3
1
C NMR and H NMR spectra were recorded with
1
Bruker DRX-400 spectrometers (for 400 MHz for H,
13
18
1 or
deuterated acetone solution as indicated. FT-IR spectra
00 MHz for C) at 25 ꢁC for Me
2
SO-d
6
or CDCl
3
To estimate IDSAC values, propanoation processes
in which the free hydroxyl groups in CA would be fully
substituted by propanoyl groups were carried out. CA
(1.0 g) was dissolved in 10 mL pyridine, and 15 mL of
propionic anhydride and 0.5 g of 4-dimethylaminopyr-
idine as catalyst were added in the solution. The solution
were collected on Analect RFX-65A spectrometer (wave
ꢀ
1
number range: 400–4000 cm ). CA and CA–O–CS were
tested via pressing with KBr to pellets or polymeric
films. Elementary analyses were processed on Heraeus
CHN-O-RAPID. Optical rotations were determined at
was then stirred for 1 h at 100 ꢁC under N and then
2
2
5 ꢁC with a Model W22-1S automatic polarimeter.
poured into MeOH after cooling. The white fine pre-
cipitate was washed repeatedly with MeOH, filtrated
and dried for 24 h under vacuum. As previously re-
Melting points were inspected with a X-4 digital display
binocular microscope. TG data were recorded by a
NETZSCH TG 209. Temperature range: 23–550 ꢁC;
13
ported, IDSAC was estimated in the quantitative
C
Heating rate: 20 ꢁC/min; N flow: 60 mL/min. The frac-
NMR by the integrated intensity ratio of two triplet
signals assigned to C@O at C-6, -3 and -2 positions of
the propanoyl and acetyl C@O from down- to up-field,
respectively.
2
tional fixation of CS on the cellulose acetate, G, is de-
fined as follows:
W
1
ꢀ W
0
G ¼
W
0
3
.5. Enantioselective adsorption
where W and W refer to weights of CA and CA–O–CS,
0
1
respectively.
CA–O–CS powder was added to a 0.100 wt % solution
of -tyrosine and -tyrosine, respectively. The mixtures
D
L
3
.4. Chemical analysis and propanoation method
were stirred for 24 h, and then filtered, washed with
water for 24 h to desorb the tyrosine from the polymer.
The aqueous solution containing the desorbed tyrosine
was concentrated and its concentration was deter-
mined using a 7504 DC UV–visible spectrophotometer.
Enantioselectivity of adsorption was estimated by the
enantiomer excess percent (%ee) from the following
formula:
DSCS was determined by chemical analysis method as
17
described in the following. The sample (0.3 g) was
added to a flask containing 75% aqueous alcohol solu-
tion (40 mL) and then heated to 60 ꢁC for 30 min.
Afterwards, 40 mL of 0.5 M NaOH aqueous solution
was added to the system, which was further heated for

D. Xiao et al. / Carbohydrate Research 339 (2004) 1925–1931
1931
Al-tyrosine ꢀ A
References
d-tyrosine
%
ee
where A denotes the amount of the adsorbed tyrosine.
.6. Enantioselective permeation
A
¼ A
ꢃ 100%
l-tyrosine
þ Ad-tyrosine
1
. Norbert, M. M.; Pilar, F.; Wolfgang, L. J. Chromatogr. A.
001, 906, 3–33.
2
2
3
. Aoki, T. Prog. Polym. Sci. 1999, 24, 951–953.
. Van der Ent, E. M.; Van’t Riet, K.; Keurentjes, J. T. F.;
Van der Padt, A. J. Membr. Sci. 2001, 185, 207–221.
. Yashima, E.; Noguchi, J.; Okamoto, Y. Macromolecules
1995, 28, 8368–8374.
3
4
5
6
7
8
9
The nonporous membrane prepared by casting acetone
solution of CA–O–CS was placed between two 500 mL
cells. Then, 0.5 wt % aqueous solution of D-tyrosine or
. Setsuji, T.; Teruyuki, M.; Ken, E. J. Membr. Sci. 1996,
1
18, 31–40.
. Klaus, S.; Matthias, K.; Karlheinz, D. Tetrahedron:
Asymmetry 1997, 8(7), 979–982.
L
-tyrosine and water were filled in the cells called donor
cell and acceptor cell, respectively. After a while, the
amount of the tyrosine that permeated through the
membrane from the donor cell to the acceptor cell was
determined at the wavelength of 272 nm by a UV–visible
spectrometer. The permeation rate ðPÞ was estimated
from P ¼ QL=At, where Q is the quantity of the per-
meated tyrosine, t the permeation time, L and A the
thickness and area of the membrane, respectively.
Accordingly, the enantiomer excess of permeation
. Harada, N.; Soutome, T.; Murai, S.; Uda, H. Tetrahedron:
Asymmetry 1993, 4(8), 1755–1758.
. Ga ꢀe lle, C.; Alain, B.; Jean-Francois, T.; Luc, S. Carbo-
hydr. Res. 2003, 388, 743–750.
. Dobele, G.; Rossinskaja, G.; Telysheva, G.; Meier, D.;
Faix, O. J. Anal. Appl. Pyrol. 1999, 49, 307–317.
0. Friebolin, H.; Keilich, G.; Siefert, E. Angew. Chem., Int.
Ed. Engl. 1969, 8, 766–773.
1. Jandura, P.; Ricdl, B.; Kokta, B. V. Polym. Degrad. Stab.
1
1
1
1
2
000, 70, 387–394.
2. Suarez, F.; Martinez-Alonso, A.; Tascon, J. M. D. Polym.
Degrad. Stab. 2002, 75, 375–383.
3. Chun, L.; Hanno, B. Carbohydr. Res. 2002, 337, 1297–
307.
14. Soares, S.; Gamino, G.; Levchik, S. Polym. Degrad. Stab.
995, 49, 275–283.
(
%ee
p
) is obtained from the expression:
Pd-tyrosine ꢀ P
l-tyrosine
1
%
ee
P
¼ P
ꢃ 100%
d-tyrosine þ Pl-tyrosine
1
1
5. Atsuchi, M.; Noriyuki, A.; Takehisa, T.; Masanori, T.;
Kohei, S.; Naoya, O. Macromolecules 1990, 23, 2748–
2
752.
6. Vandewalle, M.; Van der, E. J. Tetrahedron 1986, 42(14),
035–4043.
7. ASTM Designation: D871-72 (Reapproved 1983), 175–
90.
1
1
1
Acknowledgements
4
The financial support by the Natural Science Founda-
tion of Guangdong Province (Grant No: 021471) is
gratefully acknowledged.
1
8. Tezuka, Y.; Tsuchiya, Y. Carbohydr. Res. 1995, 273(1),
83–91.