Studies of the series of cellooligosaccharide peracetates as a model for cellulose triacetate by 13C CP/MAS NMR spectroscopy and X-ray analyses

Article abstract of DOI:10.1016/S0008-6215(99)00218-9

The series of crystalline oligomers from α-cellobiose octaacetate through α-cellohexaose eicosaacetate, listed as below, was prepared from homogeneous acetylation of the corresponding cellooligosaccharides and characterized by ¹³C CP/MAS NMR spectroscopy and X-ray analysis in order to obtain the structural models of cellulose triacetate (CTA) in solid state. Progressing toward the hexamer, the NMR spectral feature of the oligomers, in comparison with two allomorphs of CTA I and CTA II, gradually approached that of CTA I. Specifically, chemical shifts of both the hexamer and the pentamer were in considerable respective agreement with those of CTA I. In addition, X-ray diffraction patterns of the oligomers established that the crystalline pentamer and hexamer have a CTA I lattice in spite of recrystallization from ethylacetate-hexane. We therefore concluded that the pentamer and hexamer would be useful models for the CTA I structure. Copyright (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(99)00218-9

www.elsevier.nl/locate/carres
Carbohydrate Research 322 (1999) 256–263
Studies of the series of cellooligosaccharide peracetates
13
as a model for cellulose triacetate by C CP/MAS
NMR spectroscopy and X-ray analyses
Hiroyuki Kono *, Yukari Numata, Nobuhiro Nagai, Tomoki Erata, Mitsuo Takai
Di6ision of Molecular Chemistry, Graduate School of Engineering, Hokkaido Uni6ersity, Sapporo,
Hokkaido 060-8628, Japan
Received 10 May 1999; accepted 6 August 1999
Abstract
The series of crystalline oligomers from a-cellobiose octaacetate through a-cellohexaose eicosaacetate, listed as
below, was prepared from homogeneous acetylation of the corresponding cellooligosaccharides and characterized by
13C CP/MAS NMR spectroscopy and X-ray analysis in order to obtain the structural models of cellulose triacetate
(CTA) in solid state. Progressing toward the hexamer, the NMR spectral feature of the oligomers, in comparison
with two allomorphs of CTA I and CTA II, gradually approached that of CTA I. Specifically, chemical shifts of both
the hexamer and the pentamer were in considerable respective agreement with those of CTA I. In addition, X-ray
diffraction patterns of the oligomers established that the crystalline pentamer and hexamer have a CTA I lattice in
spite of recrystallization from ethylacetate–hexane. We therefore concluded that the pentamer and hexamer would
be useful models for the CTA I structure. © 1999 Elsevier Science Ltd. All rights reserved.
Keywords: Cellulose triacetate; Crystal structure; 13C CP/MAS NMR spectroscopy; X-ray diffraction; Oligomeric compound
packing [8], and that of CTA II has an anti-
parallel structure [9,10]. In a study by Sprague
et al. [6], CTA I is only produced by heteroge-
neous acetylation from cellulose I, while the
crystalline state of CTA can be produced by
homogeneous acetylation from cellulose or by
heterogeneous acetylation from cellulose II in
CTA II. The authors also demonstrated that
cellulose I and cellulose II were respectively
obtained from the corresponding polymorphs
of CTA by saponification; thus, they con-
cluded that the polymorphism of the CTAs
was in absolute relation to that of cellulose [6].
However, recent opinions have differed over
the polymorphism relationships between cellu-
lose and CTA suggested by Sprague et al. In
previous reports [7,11] by authors in our
group, it was shown that typical CTA I was
1. Introduction
When compared with the other synthetic
polymers, the fundamental properties [1–4] of
cellulose triacetate (CTA), such as the struc-
ture–property relationships, crystallization be-
havior, and molecular dynamics, are poorly
understood due to the existence of two allo-
morphs, denoted CTA I and CTA II. In X-ray
crystallographic studies of the CTAs, diffrac-
tion patterns of CTA I and CTA II can be
distinguished from each other [5–7]; the struc-
tural model of CTA I has a parallel-chain
* Corresponding author. Tel./fax: +81-11-706-6568.
E-mail address: kohno@dove-mc.eng.hokudai.ac.jp (H.
Kono)
0008-6215/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(99)00218-9

H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
257
obtained from prehydrolyzed cellulose II by
heterogeneous acetylation without any dissolv-
ing process. In addition, CTA I derived from
heterogeneously acetylated cotton is reported
[12] to transform CTA II in the presence of
formic acid, and the resultant CTA II product
is saponified into cellulose I. Roche et al. [13]
reported that CTA I was produced from a
CTA in a solution state in trifluoroacetic acid.
Discussion regarding the crystal structures of
allomorphs of both CTA and cellulose are
currently in progress, though the most impor-
tant question remains whether reversals of
chain direction on areas of neighboring chains
are required for the transformation of one
allomorph of CTA into another.
of 13C cross polarization–magic angle sample
spinning (CP/MAS) NMR spectroscopy and
X-ray diffraction.
Comparisons of the crystal structures of the
CTA oligomers with those of the two allo-
morphs of CTA were made in order to obtain
a simplified oligomer model of CTA, which is
also described herein.
2. Results and discussion
Characterization of CTAs and CTA
oligomers.—CTA I and CTA II were prepared
by heterogeneous acetylation [17] of cellulose
powder and pre-mercerized cellulose, respec-
tively. The titration method [18] gave a degree
of substitution (DS) of 2.95 acetyls/glucose
repeat unit for the CTA I, and one DS of 2.93
for CTA II. The CTAs were subsequently
exposed to heat treatment at 210 °C for 15 min
in order to increase their crystallinities, then
air-dried. The products showed typical X-ray
diffraction patterns and CP/MAS NMR spec-
tra of the corresponding allomorphs of CTA,
which will be described later.
In order to investigate crystal structures of
CTAs, it is important to have determined the
structures of cellooligosaccharides acetates in
the solid state; examination of the homologous
series of oligomeric compounds, which asymp-
totically approach a polymer structure, has
revealed some polymer properties [14]. Regard-
ing CTA oligomers, crystal structures of b
anomer of the CTA dimer and trimer were
discussed by Pe´rez and Brisse [15], who con-
cluded that glycosidic torsion angles of the
middle and nonreducing end units in the trimer
were close to those of the CTA I and CTA II,
respectively. Buchanan et al. [16] compared
physical properties of the cellooligosaccharide
acetates from a degree of polymerization (DP)
of 2–9 with those of CTA by differential
scanning calorimetric analysis. However, thus
far, surprisingly few studies have examined the
relationship between the crystal structure of
CTAs and that of CTA oligomers.
The CTA oligomers were all obtained by
acetylation of the corresponding respective cel-
looligosaccharide and purified by recrystalliza-
tion
twice
from
ethylacetate–hexane.
Complete substitution of the acetyl group for
the hydroxyl group on the oligomers was esti-
mated by the disappearance of OH-group ad-
sorption at about 3200 cm−1 as determined by
FTIR spectroscopy and compound purity was
confirmed in each case by thin-layer chro-
matography (TLC) analysis. As shown in
Table 1, elemental and mass spectroscopic
analyses data of the CTA oligomers are com-
pletely coincident with the theoretical values,
which indicates that the initial hydroxyl groups
of the oligomers have been completely acetyl-
Herein, we describe preparation of the series
of crystalline a- -cellooligosaccharide acetates
D
from the dimer through the hexamer, where
n=0–4, by homogeneous acetylation and
characterize their crystal structures by analyses
13
ated. The complete C NMR spectral assign-
ments for the oligomers, as well as those for
CTA I in chloroform-d are summarized in
13
Table 2. The C NMR chemical shifts of the
CTA I and CTA oligomers are in agreement
with the previous report [16]. Regarding C-1
resonances of reducing end units, chemical
shifts of a anomeric carbons are in the region
88–90 ppm [19], and those of b anomeric
carbons are downfield in the region 90–92 ppm

258
H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
[19]. As shown in Table 2, the spectral data of
the oligomers gave no signals at the b
anomeric C-1 region, while C-1 resonances of
reducing end units of all the oligomers were
observed in the a anomeric carbon region.
This finding established that these products
were a-cellooligosaccharide peracetates. As
can be seen in Table 2, the chemical shifts of
Table 1
Yields and analyses of the cellooligosaccharide peracetates (DP=2–6)
Compound
DP
Yield (%)
Formula
Mr: Calcd (Found)
Calcd (Found) (%)
C
H
N
Dimer
Trimer
Tetramer
Pentamer
Hexamer
2
3
4
5
6
81
76
74
69
73
C28H38O19
C40H54O27
C52H70O35
C64H86O43
C76H102O51
678.6 (679)
966.8 (967)
1255.1 (1255)
1543.3 (1543)
1831.6 (1831)
49.56 (49.52)
49.69 (49.42)
49.76 (49.54)
49.81 (49.60)
49.84 (49.69)
5.64 (5.64)
5.63 (5.58)
5.62 (5.63)
5.62 (5.59)
5.61 (5.58)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
Table 2
13C NMR spectral data (l, ppm) for CTA I and cellooligosaccharide peracetates in CDCl3
Compound
DP
CTA I
Dimer
2
Trimer
3
Tetramer
4
Pentamer
Hexamer
6
5
Reducing end unit
C-1
C-2
C-3
C-4
C-5
C-6
89.2
70.7
70.7
76.0
71.8
89.1
69.6
69.6
76.0
70.8
89.1
69.6
69.6
76.0
70.9
89.1
69.6
69.6
76.2
70.8
61.8
89.1
69.6
69.6
76.2
70.8
61.8
61.5
61.8
61.7
C-1 Ac
C-2 Ac
C-3 Ac
C-6 Ac
168.8
169.6
169.5
170.2
168.8
169.7
169.6
170.1
168.8
169.7
169.6
170.2
168.8
169.7
169.7
170.2
168.8
169.7
169.6
170.2
Internal unit(s)
C-1
C-2
100.5
72.0
100.6
72.5
100.5
72.1
100.5
72.1
100.5
72.1
C-3
C-4
72.7
76.2
72.9
76.2
72.5
76.2
72.4
76.2
72.7
76.2
C-5
C-6
73.1
62.2
72.8
62.4
73.0
62.2
73.0
62.2
73.1
62.2
C-2 Ac
C-3 Ac
C-6 Ac
169.3
169.7
170.2
169.3
169.7
170.1
169.3
169.7
170.2
169.3
169.7
170.2
169.3
169.7
170.2
Nonreducing end unit
C-1
C-2
100.7
72.2
100.8
71.8
100.8
71.8
100.8
71.8
100.8
71.8
C-3
C-4
73.1
68.2
72.9
68.1
72.8
68.1
72.7
68.1
73.1
68.1
C-5
C-6
72.6
61.8
72.2
61.8
72.2
61.8
72.1
61.8
72.1
61.7
C-2 Ac
C-3 Ac
C-4 Ac
C-6 Ac
169.0
170.2
169.3
170.4
169.0
170.1
169.3
170.4
169.1
170.2
169.3
170.4
169.1
170.2
169.3
170.4
169.1
170.2
169.3
170.5
C-Me
20.5
20.7
20.9
20.5
20.7
20.5
20.7
20.5
20.8
20.5
20.8
20.5
20.8

H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
259
internal units of the oligomer are in reason-
able agreement with those of CTA. This find-
ing suggests that the chemical circumstances
of internal residues of the CTA oligomers
(DP\3) are approximately identical to those
of CTA, which could be applied to conforma-
tional analyses of CTA in solution state.
Solid-state NMR spectra.—The 13C CP/
MAS NMR spectra of the CTA oligomers are
shown in Fig. 1, arranged from the dimer to
the hexamer and indicating the assignments of
resonances to various carbons. Assignments
are based on the chemical shift data for the
oligomers in chloroform-d (Table 2). Progress-
ing toward the hexamer, the spectra of the
oligomers simplify because the signal intensi-
ties of the internal residues of the oligomers
gradually increase with the increase of their
DP relative to those of the reducing end and
nonreducing end units. The number of 13C
resonances in the ring carbon region (61–104
ppm) converged to five when the DP of the
oligomers reached a value equal to that of the
pentamer. Strong similarities were observed
between the pentamer and the hexamer in the
chemical shifts of the predominant five reso-
nances. The C-1 region (88–106 ppm) for the
oligomers shows a remarkable tendency for
the C-1 resonance(s) of internal residue at
about 103 ppm to become dominant, while
two resonances, at 90 and 99 ppm, derived
from a reducing and nonreducing end, respec-
tively, become diminished, then finally negligi-
ble. Because carbonyl carbon and methyl
carbon regions of the oligomers show a more
Fig. 1. 13C CP/MAS spectra of CTA I, CTA II, and cel-
looligosaccharide peracetates (DP=2–6). Peak intensities of
the methyl region are shown at 40% reduction.
the ring carbons and of the carbonyl carbons
of the CTA oligomers show little variation
with increasing DP. Specifically, when the DP
of the oligomers reach a value higher than
that of tetramer, the 13C chemical shifts of
Table 3
Solid-state 13C NMR spectral data (l, ppm) for CTA I, CTA II, and a-
D
-cellohexaose eicosaacetate
Compound
Carbonyl carbon
Ring carbon
C-1
Me
C-2ꢀC-5
C-6
CTA I
170.9
172.2
103.2
80.6
76.3
72.9
62.7
23.2
22.3
CTA II
169.8
170.6
172.6
172.9
101.8
103.2
80.8
78.2
76.0
75.0
73.2
65.8
61.9
21.8
Hexamer
170.9
173.0
80.6
76.3
73.0
62.6
23.2
22.3

260
H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
resolved quartet. Significant multiplicity of
CTA I and the hexamer was also confirmed in
the spectral features of the C-2ꢀC-5 (71–82
ppm) and the C-6 (66–61 ppm) region. In the
C-2–C-5 region of these spectra, the predomi-
nant signals numbered three for both CTA I
and the hexamer, in contrast to five for CTA
II. In the C-6 region of the spectra, both CTA
I and the hexamer show a single peak. Dou-
blet splitting for the C-6 carbon of CTA II
was also detected. Table 3 shows the close
correspondence of chemical shifts of ring car-
bon between the hexamer and CTA I in the
crystalline phase. From these results, it was
suggested that the crystalline pentamer and
hexamer both have the crystal CTA I lattice,
which was confirmed by X-ray diffraction
analyses.
X-ray diffraction.—In order to firmly estab-
lish the crystal structures of the oligomers
suggested by solid-state NMR analysis, we
measured X-ray powder diffractograms for
the two allomorphs of CTA and for the CTA
oligomers by the refraction method, the re-
sults of which are shown in Fig. 2. Heat
treatment of the CTAs results in sharp and
typical X-ray patterns [7] for CTA I and CTA
II, respectively. CTA I gave three sharp equa-
torial diffractions at 11.6 (2q=7.6°), 5.6
Fig. 2. X-ray diffractograms of CTA I, CTA II, and cel-
looligosaccharide peracetates (DP=2–6).
complex version of variation, trends in the
spectral features of the methyl regions were
not established.
,
(15.9°), and 4.37 A (20.3°), while CTA II gave
strong diffractions at 10.48 (2q=8.4°), 8.5
The 13C CP/MAS spectra of CTA I and
CTA II are also shown in Fig. 1. Chemical
shifts of the CTAs and the hexamer are sum-
marized in Table 3. The two allomorphs of
CTA could be distinguished from each other
by the chemical shifts and crystallographic
splittings, as previously reported [11,20]. In a
comparison of the CP/MAS spectrum of the
hexamer with the spectra of CTAs, conspicu-
ous similarities between the orientation of
acetyl groups in the hexamer and of those in
CTA I were observed. A sharper doublet for
Me (21–24 ppm) of the hexamer and of CTA
I, with splitting of 0.91 and 0.89 ppm, respec-
tively, was observed, while that for the CTA II
spectrum has a broad single peak (21.8 ppm).
In the carbonyl carbon resonance regions
(169–173 ppm), there is a doublet splitting for
both CTA I and the hexamer. Carbonyl car-
bons of CTA II, on the other hand, gave
shoulder peaks characteristic of a partially
(10.4°), 6.7 (13.4°), 5.4 (16.3°), 5.3 (16.7°), 4.8
(18.6°), and 3.8 A (23.4°). These diffraction
,
spots of both crystalline CTAs were in com-
plete agreement with those previously re-
ported [7]. On the other hand, although the
diffractograms of both the CTA dimer and
the trimer show a number of sharp peaks, the
number of diffractions of CTA oligomers de-
creased with an increase of their DP. The
spots and relative intensities of diffractions
detected on the crystalline CTA and on the
oligomers (DP\3) are summarized in Table 4
with their Miller indexes. In contrast to find-
ings on the X-ray diffraction behavior of the
dimer and trimer, many similarities were ob-
served between the CTA oligomer (DP\3)
and CTA I. Three remarkably strong diffrac-
tions appear in the CTA oligomers (DP\3)
at approximately 11.6 (2q=7.6°), 5.5 (16.1°),
,
and 4.3 A (20.3°), which correspond to the

H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
261
(001), (202), and (500) plane for the crystal
CTA I [7]. Although very weak diffractions
are observed in some spots in the diffrac-
tograms of the tetramer and pentamer, there
are no diffractions at these spots in CTA I or
CTA II. Differences between the weak diffrac-
tions of the oligomers and those of CTAs may
arise from an effect of the end-unit forms of
these samples. CTAs are composed of a num-
dent that the crystal pentamer and hexamer
have a crystalline CTA I lattice.
The assumption that CTA I is not at all
obtained from the homogeneous acetylation
of cellulose I, as well as from cellulose II, is
now widely accepted [6], whereas our findings
here prove clearly that crystal CTA oligomers
(DP\4) exhibit a crystalline CTA I structure
despite the homogeneous acetylation of the
corresponding cellooligosaccharides. To clar-
ify any apparent contradiction with our previ-
ous study employing X-ray analysis [7] in
which a mixed pattern of CTA I and CTA II
was obtained from CTA prepared by hetero-
geneous acetylation of a previously acid-hy-
drolyzed cellulose II, one must consider that
the proportions of the CTA I and CTA II
were dependent on cellulose sources as well as
on conditions for prehydrolysis of cellulose II
[7]. Although the distinct relationship between
the crystal structure of CTA and its DS has
not yet been proved, our findings herein sug-
gest that the crystal structure of oligomeric
ber of 2,3,6-tri-O-acetyl-b-
while the reducing and nonreducing end of the
oligomers are 1,2,3,6-tetra-O-acetyl-a- -glu-
cose residues and 2,3,4,6-tetra-O-acetyl-b-
D-glucose residues,
D
D-
glucose residues, respectively. However, the
pentamer and the hexamer gave approxi-
mately similar diffractograms for CTA I, and,
notably, the hexamer was in complete agree-
ment with CTA I in terms of interplanar
spacing values and diffraction intensities, ex-
cept in the case of a very weak diffraction at
,
10.1 A (8.8°). On the basis of the X-ray analy-
sis data considered in conjunction with the
results of the 13C CP/MAS spectra, it is evi-
Table 4
Interplanar spacing values (l) and relative intensity of equatorial diffractions for CTA I, CTA II, and cellooligosaccharide
peracetates (DP=4–6)
,
,
Compound
CTA I
2q (°)
l (A)
Intensity a
Miller index b
Compound
pentamer
2q (°)
l (A)
Intensity a
7.64
14.58
15.90
17.80
20.30
22.37
26.50
11.56
6.07
5.57
4.95
4.37
3.97
3.36
v.s.
w.
v.s.
w.
s.
w.
w.
(001)
(102)
(202)
7.64
8.75
11.56
10.09
6.05
5.55
4.97
4.33
3.90
3.51
v.s.
w.
w.
v.s
w.
s.
14.60
15.92
17.76
20.40
22.65
25.17
(500)
(103)
w.
w.
CTA II
8.43
10.42
13.14
16.28
16.74
18.59
21.39
23.39
26.50
10.48
8.48
6.73
5.44
5.29
4.77
4.15
3.80
3.36
v.s.
v.s.
v.s.
s.
s.
s.
w.
s.
w.
(110)
(210)
(310)
(410)
(220)
(320)
(600)
(130)
tetramer
7.49
9.56
11.78
9.23
6.17
5.55
5.01
4.85
4.33
3.91
3.74
v.s.
w.
w.
v.s.
w.
w.
s.
14.31
15.91
17.61
18.19
20.40
22.60
23.61
w.
w.
Hexamer
7.64
8.75
11.56
10.09
6.03
5.55
4.97
4.34
3.92
3.34
v.s.
w.
w.
v.s.
w.
s.
24.12
25.31
28.70
3.66
3.49
3.08
s.
s.
w.
14.63
15.91
17.76
20.35
22.48
26.42
w.
w.
a The abbreviations used are as follows: v.s., very strong; s., strong; w., weak.
b Previously reported value [7].

262
H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
CTA, prepared by homogeneous acetylation,
may be same as that of CTA I. We therefore
conclude that the crystalline state of the CTA
hexamer and pentamer is identified with the
crystalline CTA I lattice. Further, the crystal
hexamer would be a useful model for CTA I.
Provided that the oligomers could be grown to
appropriate crystal size, this single crystal
could facilitate complete elucidation of the
molecular order of CTA I.
many) with toluene–ethyl acetate in appropri-
ate proportions.
General
methods.—Fast-atom-bombard-
ment mass spectra (FABMS) were obtained
on a Joel JMX-HX110 mass spectrometer op-
erated at 1.0 keV. The spectrometer was cali-
brated with cesium iodide, and samples were
prepared by wetting the FABMS probe tip
with 3-nitrobenzyl alcohol. Elemental analyses
were carried out with a Hewlett–Packard
model 185 analyzer. FTIR spectra were
recorded on samples on KBr pellets on a
Jasco FTIR-350 spectrometer at a resolution
of 4 cm−1. NMR spectral data were obtained
on a Bruker MSL 400 spectrometer at 23 °C
operating at 400 MHz for proton and at 100.6
MHz for carbon. The sample concentrations
were ca. 10 mg/mL of CDCl3. Double-quan-
tum filtered 1H–1H-correlated spectroscopy
(DQFCOSY) and total correlation spec-
troscopy (TOCSY) with a spin-lock time of 60
3. Experimental
Preparation of CTAs and CTA oligomers.—
CTA I and CTA II were prepared from CF-11
cellulose powder (Whatman International
Ltd., UK) by mercerization of the cellulose
powder, respectively, by heterogeneous acety-
lation using the method of Tanghe et al. [17].
The mercerization of the cellulose was carried
out as follows: a 2 g sample of the cellulose
powder was mixed at 20 °C in a solution of
20% (w/v) NaOH in water. After standing for
20 min, the powder was filtered, washed in a
continuous stream of water for 1 h, and dried
in vacuum overnight. The two allomorphs of
CTAs were recrystallized by heat treatment
under nitrogen at 210 °C for 15 min as previ-
ously described [7]. The DS of the CTAs were
determined according to the usual method
[18]. A 0.2 g sample was transferred to a flask
containing 0.2 mol/L potassium hydroxide in
30 mL of 50% (v/v) aq MeOH at 23 °C. After
24 h 30 mL of 0.2 mol/L HCl was added, and
after a further 30 min the solution was titrated
with aq 0.1 mol/L NaOH, phenolphthalein
being used an indicator.
a-Cellooligosaccharide acetates (DP=2–6)
were prepared from the corresponding cel-
looligosaccharides (Seikagaku Co., Japan) by
homogeneous acetylation as follows: a 200 mg
sample of the cellooligosaccharide was added
to a mixture of 12 mL of pyridine and 8 mL
of Ac2O. After stirring at 100 °C for 3 h, the
mixture was poured into ice water and filtered,
and the water was evaporated. The oligomer
acetate was obtained by recrystallization twice
from 1:1 ethylacetate–hexane. All the prod-
ucts were assayed for homogeneity by TLC
using Silica Gel 60 plates (E. Merck, Ger-
1
ms were used to assign H resonances. Het-
eronuclear single quantum coherence (HSQC)
and 1H-decoupled multiple-bond heteronu-
clear multiple quantum coherence (HMBC)
were used to assign 13C resonances. All the
NMR experiments were performed according
to standard pulse sequences. Chemical shifts
were reported in ppm from tetramethylsilane
(TMS), with CHCl3 as an internal reference.
1
For H NMR spectra, residual CHCl3 was
13
taken as 7.24 ppm. For C NMR spectra, the
center peak of the triplet resonance of CDCl3
was taken as 77.0 ppm.
Solid-state 13C NMR spectra.—Solid-state
NMR spectra were recorded on the Bruker
MSL 400 spectrometer in the usual way [21]
by combining the techniques of CP, high-
power proton decoupling, and MAS. All of
those CP/MAS spectra were taken at 23 °C
with a uniform set of parameters: 2 ms CP
time and 3460–3800 Hz MAS frequency. The
spectra were usually accumulated 2000–3000
times to achieve a reasonable signal-to-noise
ratio. Chemical shifts were calibrated through
the carbonyl carbon resonance of glycine as
an external reference at 176.03 ppm and con-
verted to values downfield from TMS (ppm).
X-ray diffraction analyses.—X-ray diffrac-
tograms were measured on a Rigaku Rint-
2000 diffractometer by the refraction method

H. Kono et al. / Carbohydrate Research 322 (1999) 256–263
263
[4] L.C. Wadsworth, D. Daponte, in T.P. Nevell, S.H. Zero-
nian (Eds.), Cellulose Chemistry and Its Applications,
Ellis Horwood, Chichester, 1985, pp. 344–362.
[5] K. Hess, K. Trogos, Z. Phys. Chem. (Leipzig), B5 (1929)
161–176.
[6] B.S. Sprague, J.L. Riley, H.D. Noether, Text. Res. J., 28
(1958) 275–287.
using nickel-filtered Cu Ka radiation operated
in ꢀ−2q scanning mode between 5 and 30°
(2q). Slit systems were set at 1° for divergence,
0.15 mm for receiving, and 1° for scatter.
[7] S. Watanabe, M. Takai, J. Hayashi, J. Polym. Sci., Part
C, 23 (1968) 825–835.
Acknowledgements
[8] A.J. Stipanovic, A. Sarko, Polymer, 19 (1978) 3–8.
[9] W.J. Dulmage, J. Polym. Sci., 26 (1957) 277–288.
[10] E. Roche, H. Chanzy, M. Boudeulle, R.H. Marchessault,
P. Sundararajan, Macromolecules, 11 (1978) 86–94.
[11] M. Takai, K. Fukuda, J. Hayashi, J. Polym. Sci., Part
C: Polym. Lett., 25 (1987) 121–126.
[12] J.J. Creely, C.M. Conrad, Text. Res. J., 35 (1965) 184–
187.
[13] E.J. Roche, J.P. O’Brien, S.R. Allen, Polym. Commun.,
27 (1986) 138–140.
This work was supported by a Grant-in-Aid
for Scientific Research from the Ministry of
Education, Science and Culture of Japan. The
authors are grateful to Mr E. Yamada for
help with NMR measurements.
[14] A.J. Benesi, D.A. Brant, Macromolecules, 18 (1985)
1109–1116.
[15] S. Pe´rez, F. Brisse, Biopolymers, 17 (1978) 2083–2096.
[16] C.M. Buchanan, J.A. Hyatt, S.S. Kelly, J.L. Little,
Macromolecules, 23 (1990) 3747–3755.
[17] L.J. Tanghe, L.B. Genung, J.W. Mench, Methods Carbo-
hydr. Chem., 3 (1963) 193–212.
References
[1] D.W. Lowman, in T.J. Heinze, W.G. Glasser (Eds.),
Cellulose Deri6ati6es: Modification, Characterization, and
Nanostructures, ACS. Symp. Ser. 688, Washington, DC,
1998, pp. 131–162.
[18] T.P. Nevell, S.H. Zeronian, Polymer, 3 (1962) 187–194.
[19] K. Bock, C. Pedersen, Ad6. Carbohydr. Chem. Biochem.,
41 (1983) 27–66.
[20] M. Hoshino, M. Takai, K. Fukuda, K. Imura, J.
Hayashi, J. Polym. Sci., Part A: Polym. Chem., 27 (1989)
2083–2092.
[2] D. Klemm, B. Philipp, T. Heinze, U. Heinze, W. Wa-
genknecht, Systematics of Cellulose Deri6atization, Com-
prehensi6e Cellulose Chemistry, Vol. 2: Functionalization
of Cellulose, Wiley–VCH, Weinheim, 1998, pp. 1–308.
[3] M. Fukuda, K. Kohata, K. Fukuda, H. Kawai, in J.F.
Kennedy, G.O. Phillips, P.A. Williams (Eds.), Cellulose:
Structural and Functional Aspects, Ellis Horwood,
Chichester, 1989, pp. 317–322.
[21] J. Schaefer, E.O. Stejskal, R. Buchdahl, Macromolecules,
8 (1975) 291–296.
.