Chapter 7
Possible Cause of Structural Irreversibility Between Cellulose I and Cellulose II Families 1
1
1
2
Jisuke Hayashi , Hiroshi Kon , Mitsuo Takai , Masahiro Hatano , and Tsunenori Nozawa 2
1
Department of Applied Chemistry, Faculty of Engineering, Hokkaido University, Sapporo 060, Japan ChemicaI Research Institute of Nonaqueous Solution, Tohoku University, Katahira-cho, Sendai, Japan
2
Transformations of allomorphs in the cellulose I fami ly to allomorphs in the cellulose II family are gene rally thought to be irreversible. We propose that the conformations of chain skeletons are the same within each family but different between the families, and that the difference in stabilities (through relief of a short contact between H1' and Η4) of the conforma tions results in the irreversibility. In contrast, Blackwell et a l . and Sarko et a l . have proposed that the crystal structures of allomorphs in the cellulose I and II families are based on parallel and antiparal l e l packing of chains, respectively, and the irrever s i b i l i t y results from the increase in entropy inherent in converting from parallel to antiparallel arrange ments. IR and solid state C NMR spectra of the allomorphs were studied to investigate the proposals. In both spectral methods, the bands or signals that are related to chain conformation split into two peaks in the II family (II, II , IV ). This indicated "bent and twisted" form chain conformations within the unit cell which we ascribe to asymmetry within chains rather than two different chains. In spectra from members of the I family (I, I I I , IV ), the same peak showed l i t t l e or no splitting, indicating a single, more symmetric chain conformation. Other bands that are related to chain packing could be similar between families but differed when the unit cell shapes were different. 13
II
II
I
I
The transformation from c e l l u l o s e I to II occurs e a s i l y , but the reverse transformation i s not r e a d i l y accomplished. I t had been thought that the i r r e v e r s i b i l i t y results from a large difference i n thermodynamic s t a b i l i t y between the c r y s t a l structures of c e l l u l o s e I and II u n t i l quite recently. In our previous papers (1, 2), we showed that the same i r r e v e r s i b i l i t y exists between allomorphs of the c e l l u l o s e I family ( I ,
0097-6156/87/0340-0135$06.00/0 © 1987 American Chemical Society
THE STRUCTURES OF CELLULOSE
136
I I I ! and IV!) and those of the c e l l u l o s e II family (II, I I I n and IVn). The trasformations among allomorphs within each family occurred reversibly, but transformations between families were i r r e v e r s i b l e . For example, IVj could be made either d i r e c t l y from I or through I I I i , and was transformed into I by hydrolysis. Also, IV can be converted into I H n and I V U through c e l l u l o s e I I . When I V U , derived from IVi or other members of the c e l l u l o s e I family, was transformed into II by hydrolysis, i t could not be transformed back into a member of the c e l l u l o s e I family. There i s strong s i m i l a r i t y between the x-ray patterns of I I I i and I H n and between those of IVj and I V , but the meridional i n t e n s i t i e s are quite d i f f e r e n t ( 3 ) . The IR and 13C NMR spectra are c l e a r l y d i f f e r e n t and i t i s clear that IIIj , Ι Π , IV and IV are d i s t i n c t allomorphs. When c e l l u l o s e t r i n i t r a t e Ι (TNC Ι ) and t r i a c e t a t e s Ι and ΙΙ (TAC Ij and I I ) were prepared from allomorphs of the I family i n the fibrous state under low-swelling conditions, they could be saponified into c e l l u l o s e I. On the other hand, TNC Ι , TAC Ι and TAC U n can be saponified into c e l l u l o s e II (4.-6). When the esters of the I family were r e c r y s t a l l i z e d by heat treatment, they were transformed i r r e v e r s i b l y into corresponded esters (for exam ple, TAC Ij to TAC I n ) . They crossed the b a r r i e r of i r r e v e r s i b i l i ty and were saponified into c e l l u l o s e I I . The same phenomena were observed for soda c e l l u l o s e s . Nac e l l u l o s e I j prepared from the I family under conditions of low swelling (high temperature or with stretching) could be converted to c e l l u l o s e I with hot water. Na-cellulose Ι prepared from cellulose II under the same conditions was converted to c e l l u l o s e II (7-10). In ordinary mercerization of native c e l l u l o s e a mixture of Na-cellulose Ι and Ι was obtained; both decomposed to c e l l u lose II with cold water. Na-cellulose Ι was transformed i r r e v e r s i bly into Na-cellulose I . The allomorphs and derivatives prepared from c e l l u l o s e I and II i n s o l i d state could be transformed into c e l l u l o s e I and I I , respectively. The memory phenomenon of the o r i g i n a l c r y s t a l struc ture should be due to a s t r u c t u r a l c h a r a c t e r i s t i c (chain conforma t i o n , chain p o l a r i t y or others) of an i n d i v i d u a l chain that i s common within each family and kept through the change of c r y s t a l structure. There were d i r e c t i r r e v e r s i b l e conversions between cor responding c e l l u l o s e esters, Na-cellulose and c e l l u l o s e IV prepared from c e l l u l o s e I and II just l i k e that between I and I I . According l y , the s t r u c t u r a l c h a r a c t e r i s t i c should be the cause of the struc t u r a l i r r e v e r s i b i l i t y between the I and II f a m i l i e s . In the 1970's ( 3 , Λ ) , we proposed that the chains i n the cellulose I family have a "bent" form, which has an approximate 2i screw axis. The chains i n the c e l l u l o s e II family were proposed to be "bent and twisted"; they deviate substantially from a 2i axis. By keeping the conformation of the o r i g i n a l member i n a family, the allomorphs and derivatives can return to the o r i g i n a l crystal structures. In the "bent" form the non-bonded distance between H1 and ΗΛ i s about 1.8 A for a wide range of rotations of g l y c o s i d i c units about the 01-04 v i r t u a l bond. (Figure 4). This i s a short distance compared to the sum of the r a d i i of two hydrogen atoms (2.Λ A). On the other hand, i n the "bend and twisted" form, the distance can be increased by the twisting. When the intramolecular hydrogen bonds that f i x the chain conformation are relaxed, the X
n
Π
T
IX
τ
τ
τ
τ
X
τ
τ
τ
ι
ι
τ
τ
τ
τ
τ
T I
f
7.
HAYASHI ET AL.
Irreversibility Between Cellulose I and II
137
twisting of the chain i s caused by a small drop i n energy and the "bent" form i s changed to the "bent and twisted" form. In addition, the i r r e v e r s i b i l i t y between families should a r i s e from the drop i n energy that drives the twisting of the chains. A t a l l a and Van der Hart (11, 12) concluded, based on t h e i r Raman and C NMR spectra, that the molecules i n cellulose I and II have d i f f e r e n t conformations. Based on x-ray analyses, Sarko et a l . (13, H ) and Blackwell et a l . (15, 16) both concluded that c r y s t a l structures of cellulose I and II were based on p a r a l l e l and a n t i p a r a l l e l packing, respectively, of chains that have similar backbone conformations. Sarko (17) concluded that the allomorphs i n the I and II families were based on p a r a l l e l and a n t i p a r a l l e l chains, respectively. The i r r e v e r s i b i l i t y may arise from the increase i n entropy when p a r a l l e l packing i s converted to a n t i p a r a l l e l packing. In the present work, we t r i e d to determine which proposal i s better using IR and s o l i d state l C NMR. There are many papers on the spectra of cellulose I and II (12, 18-24.), however, there are few on the other allomorphs. Mann and Marrinan (21) found d i f f e r e n ces i n the OH stretching bands of IR spectra of IIIj and I I I and of IV-j- and I V . Chidambareswaran et a l . (25) reported IR spectra for several allomorphs, but their spectra lacked enough resolution f o r detailed discussion of chain conformation. The 1 C NMR spectra of the other allomorphs have not been reported. 1 3
3
n
IT
3
Experimental Preparation of Cellulose Samples Native c e l l u l o s e . Celluloses from p u r i f i e d Valonia ventricosa, Acetobacter xylinium, ramie and cotton l i n t e r s were used. P e l l i c l e s of b a c t e r i a l cellulose were grown under the conditions described by Colvin (26) f o r 4-8-96 hours. The b a c t e r i a l medium was thoroughly washed away with d i s t i l l e d water and 1% NaOH aqueous solution. A membrane of b a c t e r i a l cellulose having u n i a x i a l orientation was prepared from a young p e l l i c l e by stretching to about twice the i n i t i a l length i n a wet state and drying under tension. Cellulose I I . Fortisan, ordinary viscose rayon, mercerized ramie, mercerized b a c t e r i a l c e l l u l o s e , and saponified t r i a c e t a t e f i l m were used. A f i l m having u n i a x i a l orientation was prepared from saponif i e d triacetate f i l m by stretching and drying under tension. Cellulose I I I i and I I I n · Native cellulose and cellulose I I , described above, were treated with l i q u i d ammonia f o r about 2 hours at about - 80°C. Gradual evaporation of ammonia was permitted i n dry a i r at about - 15°C. Cellulose IV and I V n . These allomorphs were prepared from I I I i and I I I n , respectively, by heating i n glycerol at 260°C f o r 30 min. T
Deuteration. The f i l m samples were dried under vacuum i n the presence of P 0 f o r 2 days, deuterated i n vapor of D 0 (99-75%) f o r 1 day, and then dried under vacuum i n the presence of P 0 f o r 1 day. 2
5
2
2
5
138
THE STRUCTURES OF CELLULOSE
X-ray D i f f r a c t i o n X-ray diffractograms were measured by r e f l e c t i o n method with CuKaSamples of membrane were measured laying surface of membrane paral l e l to the r e f l e c t i o n plane. Other samples were measured as powders smaller than 60 mesh. IR Spectra IR spectra of the allomorphs were obtained with an A 202 spectro meter manufactured by Japan Spectroscope Co. The infrared radiation was polarized with an AgCl plate p o l a r i z e r . Solid-state
13
C
NMR
Spectra
1 3
Solid-state C NMR spectra were observed using the cross-polariza tion/magic angle spinning technique on a Bruker CXP-300 spectro meter operating at 7 T. Contact time, r e p e t i t i o n time and high power decoupling observation time were 2 ms, 2s, and 20-30 ms, respectively. The magic-angle spinning frequency was i n the range 3460-3800 Hz. The spectra were also obtained with a JE0L FX60 spectrometer operating at 59·8 mHz. Results and
Discussion
Infrared Spectra of Cellulose Allomorphs In order to compare IR spectra of allomorphs i n d e t a i l , especially i n the 900-1500 cm" region, highly c r y s t a l l i n e samples are needed. Such samples of ΙΙΓτ, and IVi were prepared from Valonia cellulose. Attempts to transform Valonia cellulose I to II were d i f f i c u l t i n general and gave only samples of low c r y s t a l l i n i t y . Viscose and saponified films of cellulose esters had c r y s t a l l i n i t y that was too low to y i e l d high-resolution spectra. B a c t e r i a l cellulose membranes of suitable thickness were mercerized under stretching, giving r e l a t i v e highly c r y s t a l l i n e cellulose I I . From t h i s , sample of I I I n and I V were made. Figures 1 and 2 show x-ray diffractograms of members of the cellulose I and II f a m i l i e s , respectively. Diffractograms of each were t y p i c a l , and indicated complete transformation and uniplanar orientation of (110) r e l a t i v e to the membrane surface. It was remarkable to r e t a i n t h i s orientation of the mercerized b a c t e r i a l cellulose and of the I I I n I ^ I I prepared from i t . The c r y s t a l l i n i t y of members of the cellulose II family were not high. But t h e i r IR spectra showed enough resolution for detailed discussion. Figure 3 shows IR spectra of OH stretching bands of the allomorphs. Arrows show OH stretching bands having strong dichroism p a r a l l e l to the chain axes and correspond to 03 -05 intramolecular hydrogen bonds (27). The 0-H stretching band i n a l l allomorphs i n the II family was s p l i t into two. Figure 4 shows the chain conformations of cellulose I ("bent") and of cellulose II ("bent and twisted") proposed i n our previous work (28) on the c r y s t a l structure of cellulose I I . These models of cellulose I and II have one and two kinds of intrachain (03'-05) hydrogen bond, respectively. The number of 0-H stretching peaks i n 1
TI
a n d
f
7. HAYASHI ET AL.
139
Irreversibility Between Cellulose I and II
J
10
I
I
15
20 2
L
25
θ (deg.)
F i g . 1 X-ray diffractograms of the allomorphs i n cellulose I family by the r e f l e c t i o n method (the r e f l e c t i o n plane i s p a r a l l e l to the membrane surface). A: I, valonia c e l l u l o s e , B: IV prepared from A through I I I ( C ) , C: I I I ! prepared from valonia. X
I
2
F i g . 2 X-ray family by the c e l l u l o s e , Β: prepared from stretching.
θ (deg.)
diffractograms of the allomorphs i n cellulose i n II r e f l e c t i o n method. A: I I , mercerized b a c t e r i a l IV prepared from A through I I I n (C), C: u n A. The treatments were carried out under XI
I
I
3600
3200
ι ι ι ι ι ι ι
Wave Number (cm"')
3200
Fig. 3 IR spectra of the allomorphs i n OH stretching region. The samples used for IR spectra are same to the samples i n F i g . 1 and 2. Arrow shows OH stretching bands having p a r r a l l e l dichroism and assigned to i n t r a molecular hydrogen bonds. There are one and two o f the band i n the I and I I f a m i l i e s , respectively.
3600
ι ι ι ι ι ι ι
T
( Β)
F i g . 4 Chain conformations of c e l l u l o s e proposed by X-ray analysis 28). A: "bent form chain for c e l l ulose I, B: "bent and twisted form chain deviated from 2^ for c e l l u l o s e I I . There are two kinds o f the distance o f 03' - 05 Internal r o t a t i o n angle around g l u c o s i d i c linkage i s according to Jone's expression 32). 01 - 01' distance i s 5.16A, C - C, C - 0, and C - Η bond length are 1.54, 1.34, and 1.09 A r e s p e c t i v e l y . Bond angle i s 109.28'
( A )
ι
7.
141
Irreversibility Between Cellulose I and II
HAYASHI ET AL.
the IR spectra found with p a r a l l e l p o l a r i z a t i o n corresponds to the number of types of intrachain hydrogen bonds.
0-H Stretching with P a r a l l e l of Cellulose Allomorphs
Table I.
1
Allomorph
Frequency (cm- ) 3345 3475 3340
I IIIi IYi II
3480
IIIi ι
3475
3440
24Z5
3440
IVTJ:
3440
Mann and Marrinan ( 2 2 ) assigned the two bands of the OH stretching in II to the intrachain hydrogen bonds of 0 3 - 0 5 and 0 6 ' - 0 2 . The r e s u l t s of Liang and Marchessault ( 2 3 , 2 4 ) argued against 06'-02 hydrogen bonds. They found that the dichroism of CH symmetric and antisymmetric stretching bands indicated a p o s i t i o n (gt) f o r 0 6 that i s not suited f o r formation of t h i s hydrogen bond. Instead, an intermolecular hydrogen bond from 0 6 " to 0 3 was proposed. We ob served the following CH stretching bands. !
2
2
Table I I .
Allomorph I IIIi IVi
Symmetric CH Stretching ( P a r a l l e l dichroism) of Cellulose Allomorphs 2
Table I I I .
Allomorph I nii IVi
-1
Allomorph II IIIn IVn
1
Frequency ( c n r ) 2850 2850 2850
Frequency (cm ) 2850 2850 2860
Antisymmetric CH Stretching (Perpendicular dichroism) of Cellulose Allomorphs
Frequency 2940 2935 2940
2
-1
(cm )
1
Allomorph II IIIn IVn
Frequency (cm- ) 2930 2925 2935
These r e s u l t s are consistent with a gt p o s i t i o n f o r 06 i n a l l allomorphs, with l i t t l e p o s s i b i l i t y of the 06 -02 hydrogen bond. There i s a large difference between the dimensions of the ab plane of the unit c e l l s of II and I V . Therefore, intermolecular hydrogen bonds should be f a i r l y d i f f e r e n t between II and I V . However, OH stretching bands assigned to intrachain hydrogen bonds had almost the same frequency i n II and IV . Liang and Marches sault's reliance on an inter-chain i n t e r a c t i o n to achieve a second kind of 03'-05 bond i s not consistent with t h i s spectral informa t i o n , but i t can be explained by the two 03 -05 distances a r i s i n g from our "bent and twisted" conformation. T
I X
TI
i:t
!
142
THE STRUCTURES OF CELLULOSE
Sarko and Muggli (13) and Gardner and Blackwell (15) both concluded that cellulose I had two intrachain hydrogen bonds, 03 05 and 06 -02. Stipanovic and Sarko ( H ) and Kolpak and Blackwell (16) proposed structures of II i n which the corner chains i n the unit c e l l had 03'-05 and 06 -02 bonds and the center chain had only 03 -05 bonds. According to t h e i r c r y s t a l models, there are chains having two kind of hydrogen bonds, 03 -05 and 06'-02 i n both of c e l l u l o s e I and I I . I f the two kinds of bonds are the cause of the two p a r a l l e l OH stretching bands, the two bands should be observed i n cellulose I. But a l l allomorphs i n the cellulose I family showed only one of the bands. There i s the p o s s i b i l i t y that the s p l i t t i n g of the spectral bands might arise from two d i f f e r e n t chain confor mations i n the unit c e l l . But i f the chain conformation changes depend on the c r y s t a l structure, the frequencies of the two p a r a l l e l OH bands should change with c r y s t a l structures. The frequencies of allomorphs of the cellulose II family were almost the same i n spite of t h e i r quite d i f f e r e n t c r y s t a l structure. I f the chain conformation does not change with c r y s t a l structure, the question would arise whether the chain conformation of cellulose I can change into two kinds of chain conformation i n equal amounts by mercerization or other transformation processes. This i s a pending problem. At l e a s t , however, the OH stretching bands assigned to the intrachain bonds showed common characteristics among allomorphs i n each family. 1
T
!
!
f
Marrinan and Mann (22) reported that there were two types of OH stretching bands i n native c e l l u l o s e ; the difference appeared near 3250 cm" . However, no differences i n the main bands of OH stretching are noticed now. Our IR spectra of valonia d i d not show essential differences with that of ramie shown by Liang through a l l range. I t should not hinder the present discussion. Figure 5 shows IR spectra i n the 0D stretching region. Only c e l l u l o s e that i s amorphous or on surfaces of c r y s t a l l i t e s i s deuterated, so c r y s t a l packing should have s l i g h t influence on the 0D band. In the I family, a main band of 0D stretching was at 2^80 cm . The band showed dichroism p a r a l l e l to the chain axis and was assigned to the intramolecular bond 03 (D)-05- In a l l allomorphs of the II family, the main band was at 25Λ5 cm" and exhibited p a r a l l e l band dichroism. Another p a r a l l e l band was at 2560 cm"" i n II and I I I n although i t was not clear i n IVn . The results of the 0D bands suggested that the amorphous cellulose had conformations similar to the molecules i n the c r y s t a l l i t e s . The main band of I I I i was at the same frequency as the main band i n I or I V i , despite the corresponding band of OH being at higher frequency than that i n I or I V i . I t suggested that the s h i f t to higher frequency of the OH band of IIIi had been effected with molecular packing i n the cellulose II type of unit c e l l . Figure 6 shows the region from 900 to 1200 cm"". The spectra showed common c h a r a c t e r i s t i c s within each family. Two bands were assigned to stretching of the glucose r i n g f o r allomorphs i n the cellulose I family. The bands were sharp and showed a weak perpendicular dichroism. The corresponding bands of the II family were s p l i t into two each although they were not well-separated. They also d i f f e r e d from those f o r the I family by showing strong p a r a l l e l dichroism. One of the four bands was at 1090-1095 cm"" and obscured separation of the -1
-1
f
1
1
1
1
f
-
2500
2300
ι I • l•
cm ' F i g . 5 IR s p e c t r a o f t h e d e u t e r a t e d a l l o m o r p h s i n OD stretching region. A: I , v a l o n i a c e l l u l o s e , B: IV-jp r e p a r e d f r o m ΙΙΙχ ( C ) , C: ΙΙΙχ from A, D: I I , c e l l ophane, E: U n from D, F: ΐνχχ from E.
2700
_l
1200 -1
Wave Number (cm )
1000
Ring stret.
1000
F i g . 6 IR s p e c t r a o f t h e a l l o m o r p h s i n 900 t o 1200 cm~l r e g i o n . Arrow shows bands a s s i g n e d t o r i n g stretching. Dot shows main bands a s s i g n e d t o C - C and C - 0 s t r e t c h i n g . Samples a r e same t o them i n F i g . 4.
1200
Ring stret.
£ w
144
THE STRUCTURES OF CELLULOSE
main band a r i s i n g from r i n g stretching at about 1100 c m main band of C-C stretching at about 1065 cm"".
-1
and the
1
Table IV. Glucose Ring Stretching of Allomorphs i n Cellulose I Family Frequency (cm*-1) 1110 1120 1120 1105 1105 1115
Allomorph I ΙΠι IVi
Table V.
Glucose Ring Stretching of Allomorphs i n Cellulose II Family
Allomorph II IIIn IV
1120 1120 1120
TT
Frequency 1115 1115 1110
1
(cm"" ) 1110 1110 1105
1090 1090 1100
For each allomorph i n the I family, there were two sharp bands i n the region of 1020 - 1070 cm"" assigned to C-0 and C-C stretching. 1
Table VI.
C-0 and C-C Stretching of Allomorphs i n Cellulose I Family Frequency (cm" 1060 1075 1060
Allomorph I IIIi IVi
As before, the two corresponding bands were s p l i t into four for the II family.
Table VII. Allomorph II IIIn IVn
C-0 and C-C Stretching of Allomorphs i n Cellulose II Family Frequency (cm"" ) 1030 1060 1065 1030 1070 1065 1060 1035 1065
1
) 1035 1030 1035
bands
1
1
1020 1020 1020
Figure 7 shows the region from 1200 to 1500 cm"" for the a l l o morphs. The unit c e l l s of IV and I V n are similar to that of I, and those of I I I i and I I I n are similar that of to I I . E s p e c i a l l y i n this spectral region, there were common c h a r a c t e r i s t i c s within these groups based on s i m i l a r i t i e s of the c e l l as well as member ship i n the same family. Accordingly, there were s i m i l a r i t i e s between I and IVi and between II and I I I n , although there were distinguishable differences. The IR spectra of I I I i , II IHII X
a
n
d
7.
HAYASHI ET AL.
145
Irreversibility Between Cellulose I and II
F i g . 7 IR spectra of the allomorphs i n 1200 to 1500 region. Arrow shows bands assigned to CH bending.
cm
-1
146
THE STRUCTURES OF CELLULOSE 1
had a CH bending band at about 1370 cm- that showed strong absorp t i o n compared with neighboring bands. However, the band f o r I I I i was single, while the band f o r I I I n consisted of two bands. The IR spectra f o r IVn showed some of the c h a r a c t e r i s t i c s of I and IVi, but i t s CH bending band was s p l i t , similar to I I . Generally, bands related to the chain conformation were singu l a r i n the cellulose I family and doublets i n the cellulose II family. This i s consistent with two chain conformations, i e . "bent" (I family) and "bent and twisted" (II family), that have one and two types of glycosidic linkage. I f chain conformations were essen t i a l l y the same i n a l l allomorphs, the d i f f e r i n g sets of character i s t i c s common within the families would not be explicable. The differences i n the IR spectra should r e s u l t from differences i n chain conformation and from chain packing. Even though the chain p o l a r i t y i s the same when the unit c e l l types are as d i f f e r e n t as I and IIIi ι the chain packing must be quite d i f f e r e n t . By the difference of the chain p o l a r i t y between the families, the c h a r a c t e r i s t i c s common within the families could not be also explicable. The bands that showed common c h a r a c t e r i s t i c s based on the unit c e l l types and gave important information on the intermolecular hydrogen bonds w i l l be discussed i n another paper. High Resolution Solid-state
1 3
C NMR
Spectra of the Allomorphs
A t a l l a et a l . (12), E a r l et a l . (18) and H o r i i et a l . (19) reported that the signals of C1 and C4 i n s o l i d - s t a t e C NMR spectra consist of two, three or more peaks f o r cellulose I and I I . The chemical s h i f t s of C1 peaks were at 105.2 and 107.0 ppm f o r I and 106.0 and 108.3 ppm f o r II and the C4 peaks were 89·3 and 90.1 ppm f o r I and 88.7 and 89.9 ppm f o r II (12). Figures 8 and 9 show s o l i d state C NMR spectra of the allo morphs i n the I family prepared from ramie and those i n the II family prepared from Fortisan. The signal of C1 for III-,- and IV-,did not show enough resolution to s p l i t into peaks though there was a shoulder at about 105 ppm. For a l l allomorphs i n the I family, the chemical s h i f t s of the C1 signal ranged from 106.5 to 107.0 ppm and t h e i r half-widths were 250 Hz. The C1 signals f o r the II family were a l l s p l i t into two peaks at 106 and 108.5 ppm. The half-widths were 320, 330 and 290 Hz, f o r I I , I I I n and I V , respectively. Larger widths were observed f o r the cellulose II family. The values of the chemical s h i f t and half-width were averages of values mea sured f o r the samples with the various o r i g i n s , except f o r the Valonia and b a c t e r i a l c e l l u l o s e s , which had s u b s t a n t i a l l y d i f f e r e n t values. Recently, we have obtained better NMR data f o r the allomorphs with a Bruker instrument at 200 MHz for protons. The C1 signal of IIIj seemed to be a singlet, but i t s p r o f i l e was broadened unsymmet r i c a l l y to the lower ppm side and suggested a d d i t i o n a l weak peaks. The half-width was almost the same as that of cellulose I. The common c h a r a c t e r i s t i c s related to half-width of the C1 s i g n a l within each families were c l e a r l y evident. The same phenomena were observed f o r signals from 0 4 · The chemical s h i f t s f o r the I family were 89.3 to 90.0 ppm, with h a l f widths of 180 ( I ) , 200 ( I I I i ) and 200 ( I V i ) . The chemical s h i f t s of 13
1 3
T I
HAYASHI ET AL.
Irreversibility Between Cellulose I and II Β ο * u cd
CO 13
ο
»-ι
ο
cd Pu
u •P
α ·· ω m eu M
g co >—I
Ή •Η ^ cd cd 4-1
CO H \ Pli
U
ο
• Ο H i l 00 rH M U •H rH Μ Ο
h
American Chemical Society Library 1155 16th St., N.W. Washington, O.C. 20036
cd
h
h
148
THE STRUCTURES OF CELLULOSE
C4 for the II family were 88.0 to 89-5 ppm with half-widths of 310 ( I I ) , 320 ( I I I n ) and 320 ( I V n ) . The half-width was controlled b a s i c a l l y by the degree of separation of the two peaks s p l i t i n the C1 and C4 signals. The e f f e c t of the c r y s t a l l i n i t y of samples on the half-width was small. The uniformity of half-widths within the families suggests that the signals of C1 and 0>J+ consist of two peaks even though they are observed as single peaks. The signals of C1 and C4 should r e f l e c t the symmetry at the glycosidic linkage. The s p l i t t i n g into two peaks agrees with the "bent and twisted" conformation proposed f o r c e l l u l o s e I I . The s p l i t t i n g f o r the I family suggested that the conformation i s not the simple "bent" model i n the s t r i c t e s t sense, but, r e f e r r i n g to the IR spectra, the degree of twisting may not be as large as i n the II family. Dudley et a l . (29) suggested the observed doublets of C1 and C4 i n c e l l u l o s e II were due to independent chains i n the unit c e l l . Cael et a l . (30) explained the C NMR spectrum of c e l l u l o s e I with the eight-chain unit c e l l using Dudley's proposal. But the proposal could not explain the s i m i l a r i t y of the C1 and C4 signals of the allomorphs i n each of the f a m i l i e s . They should show s i m i l a r i t y between IIIi and I I I n or IVi and I V n because t h e i r chain packing should be s i m i l a r to each other. When the high resolution NMR spectra of these are measured the problem w i l l be more clear. The chemical s h i f t of the signal from C6 was influenced by the unit c e l l type as well as the family. This i s because 06 plays an important part i n intermolecular hydrogen bonding. The range was from 66.3 to 67 ppm f o r I and I V i . Cellulose I I I i i s i n the I family but i t s unit c e l l i s the II type; the s h i f t f o r 06 was 6^.0 ppm. For II and I I I n , "the s h i f t was 63.5 to 6Λ.0 ppm. The s h i f t of C6 i n I V n , which i s i n the c e l l u l o s e II family but has a type I unit c e l l , has a value of 6^.5. ppm, which i s rather high f o r a member of the c e l l u l o s e II family. The r e l a t i o n between chemical s h i f t of the peak from C6 and the intermolecular hydrogen bonds w i l l be further discussed i n another paper. The r e s u l t s were summarized i n Table VIII. 1 3
Table VIII.
Family I IIII IVÏ II
IIIn IVn
1 3
S i m i l a r i t y of C NMR Spectra of the Allomorphs Based on the Families or the Unit C e l l Types
Unit C e l l type
I
I
I
II
I
I
II
II
II
II
II
I
Half--width (Hz) C1 CA 180 250 200 250 200 250 310 320 320 330 320 290
1 3
Chemical s h i f t (ppm) C6 67.0 6Λ.0 66.3 6Λ.0 63.5 6Λ.5
By analogy with the IR spectra, the C NMR spectra showed that there were common differences based on the c e l l u l o s e families at signals related to the chain conformation. The r e s u l t s were given a more reasonable i n t e r p r e t a t i o n through our hypothesis than the proposal of p a r a l l e l and a n t i p a r a l l e l chain systems.
7.
HAYASHI ET AL.
Irreversibility Between Cellulose I and II
149
Kai et a l . (31) have reported that nascent m i c r o f i b r i l s of b a c t e r i a l cellulose were transformed into cellulose II with 10"" % NaOH aq. Solution or 86 v o l % acetone aq. solution. They examined them with electron microscopy and d i f f r a c t i o n . The microfibrils transformed into II with the acetone solution did not show any morphological change, retained the m i c r o f i b r i l l a r state of the o r i g i n a l material, and had uniformly t h i n width and long lengths. The r e s u l t i s not consistent with a mechanism based on change from " p a r a l l e l chain" to " a n t i p a r a l l e l chains" during the transformation from I to I I . In conclusion, we think that the cause of structural i r r e v e r s i b i l i t y between the cellulose I and II families i s an i r r e v e r s i b l e transformation between the s k e l e t a l chain conformations i n the f a m i l i e s . Although we expect that further studies of c e l l u l o s e w i l l provide clearer d e t a i l s of chain conformation, i t i s not l i k e l y that i t w i l l be possible to completely solve the structures on the basis of the limited amount of X-ray data available. 3
Acknowledgments The authors are indebted to Dr. R. H. Marchessault, Xerox Research Centre of Canada, f o r valuable suggestions, and to Dr. T. Yoshida, Government I n d u s t r i a l Development Laboratory, Hokkaido, f o r h i s a i d i n obtaining some of the C NMR spectra. The preliminary study of the IR spectra by Dr. A. Sueoka provided valuable assistance i n t h i s work. The work was supported by a grant-in-Aid f o r S c i e n t i f i c Research from the Ministry of Education, Japan. 1 3
Literature Cited 1.
Hayashi, J.; Sueoka, Α.; Watanabe, S. Nippon Kagaku Kaishi, 153(1973). 2. Sueoka, Α.; Hayashi, J.; Watanabe, S. Nippon Kagaku Kaishi, 594(1973), ibid, 1345(1973). 3. Hayashi, J.; Sueoka, Α.; Watanabe, S. J. Polym. Sci., Polym. Letters., 13, 23(1975). 4. Watanabe, S.; Imai, K.; Hayashi, J. Kogyo Kagaku Zasshi 74, 1427(971), 1470(1971). 5. Hayashi, J.; Sueoka, Α.; Watanabe, S. Nippon Kagaku Kaishi, 160(1973). 6. Hayashi, J.; Imai, K.; Hamazaki, T.; Watanabe, S. Nippon Kagaku Kaishi, 1587(1973). 7. Hayashi, J.; Yamada, T.; Watanabe S. Sen-i Gakkaishi, 30, 190(1974). 8. Hayashi, J.; Yamada, T. Sen-i Gakkaishi, 31, 544(1975). 9. Hayashi, J.; Yamada, T.; Kimura, K.; Watanabe, S. Nippon Kagaku Kaishi, 1967(1974). 10. Hayashi, J.; Yamada, T.; Kimura K. J. Appl. Polym. Sci., Symposium 28, 713(1976). 11. Atalla, R.H. J. Appl. Polym. Sci., Symposium 28, 659(1976). 12. VanderHart, D.L.; Atalla, R.H. Macromolecules, 17, 1465(1984). 13. Sarko, Α.; Muggli R. Macromolecules, 7, 486(1974). 14. Stipanovic, A.J.; Sarko A. Macromolecules, 9, 851(1976). 15. Gardner, K.H.; Blackwell, J. Biopolymers, 13, 1975(1974).
150 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
30. 31. 32.
THE STRUCTURES OF C E L L U L O S E
Kolpak, F.J.; Blackwell, J. Macromolecules, 9, 273(1976). Sarko, A. Tappi, 61, 59(1978). Atalla, R.H.; Gast, J.C.; Sindrot, D.W.; Bartuska, V.J.; Maciel, G.E. J. Am. Chem. Soc., 102, 3249(1980) Earl, W.L.; VanderHart, D.L. Macromolecules, 14, 570(1981), J. Am. Chem. Soc., 102, 3249(1980). Horii, F.; Hirai, Α.; Kitamaru, R. Ploym. Bulletin, 8, 163(1982). Marrinan, H.J.; Mann, J. J. Appl. Chem., 4, 204(1954), J. Polym. Sci., 21, 301(1956). Mann, J.; Marrinan, H.J. Trans. Faraday Soc., 52, 177(1956), J. Polym. Sci., 32, 125, 357(1958). Liang, C.Y.; Marchessault, R.H. J. Polym. Sci., 37, 358(1959), ibid, 39, 269(1959). Marchessault, R.H.; Liang, C.Y. J. Polym. Sci., 43., 71(1960). Chidambareswaran, P.K.; Srebnivasan; Patil, N.B. J. Appl. Polym. Sci., 27., (1982). Colvin, J.R. Can. J. Microbiol., 11, 641(1965). Hayashi, J.; Sueoka, Α.; Watanabe, S. Nippon Kagaku Kaishi, 1320(1974). Watanabe, S.; Hayashi, J. Kogyo Kagaku Zasshi, 73, 1890(1970). Dudley, R.L.; Fyfe, C.A.; Stephenson, P.J.; Deslandes, Y.; Hamer, G.Κ.; Marchessault, R.H. J. Am. Chem. Soc., 105, 2469(1983) Cael, J.J.; Kwoh, D.L.W; Bhattacharjee, S.S. Macromolecules, 18, 819, (1985). Kai, Α.; Koseki, T. Bull. Chem. Soc., Japan, 57, 1437(1984), Chem. Lett., No. 5, 607(1985), Polym. preprint, Japan, No. 9, 2477(1985) Jones, D.W. J. Polym. Sci., 32, 371(1958)
RECEIVED March 5, 1987
