Single crystals of chitosan Noi~l Cartier, Alain Domard* and Henri Chanzy~" Centre de Recherches sur les Macromol~cules Vbgbtales ( CNRS), BP 53X, 38041 Grenoble Cedex, France~
(Received 27 January 1990; 27 April 1990) Lamellar single crystals o f chitosan were prepared at 125°C by adding ammonia to a low DP fraction o f ehitosan dissolved in water. The crystals gave sharp electron diffraction diagrams which could be indexed in an orthorhombic P212121 unit cell with a = 8.07 •, b = 8.44 1~, c = 10.34 ]~. The unit cell contained two anti-parallel chitosan chains and no water molecules. It was found that cellulose microfibrils from Valonia ventricosa could act as nuclei for inducing the crystallization o f chitosan on cellulose. This produced a shish-kebab morphology. Keywords: Chitosan; electron diffraction;single crystals; shish-kebab morphology
Introduction Chitosan, the N-deacylated chitin is a versatile biopolymer which displays a series of unique properties as a consequence of a free amino group in its repeating unit. Because of this, chitosan is able to complex efficiently with a series of substances ranging from negatively charged proteins to heavy atom ions or strong acids x-5. Such interesting properties explain why so many investigations are being pursued in order to understand and utilize the complexing possibilities of chitosan. It is along this line that one can mention recent industrial developments where chitosan has demonstrated its efficiency in the clarification of waste water as well as of industrial effluents 6'7. The geometry of the chitosan molecule and, in particular, its conformation and the environment of its amino group are of great interest in understanding the specific properties of this biopolymer. Several crystallographic studies have been achieved on chitosan fibres and films 3'8-15. These studies have revealed that in most of the cases, chitosan crystallized with a fibre repeat comprising between 10.1 and 10.5 ,~, indicating a twofold helical structure for the chitosan molecule. In a few instances however, a fibre repeat of the order of 40 A was also observed but so far not explained 3. An unusual number of crystalline polymorphs seems to exist for chitosan. These polymorphs correspond to either chitosan alone or chitosan complexed with small molecules such as water, acids or salts. So far, only a small number of these polymorphs has been resolved by X-ray analysis and structure refinement to the point where atomic coordinates are given 14'15. Thus most of the conformational possibilities of chitosan remain to be unravelled. In particular, there is a need to know better the environment of the amino moiety that confers its specific properties to chitosan. This work was undertaken to study in detail the crystalline arrangement of chitosan in a number of its * Present address: Laboratoire d'Etude des Mati6res Plastiques et des Biomat6riaux, Universit6 Claude Bernard, 43 Bvd du 11 Novembre 1918, 69622 Villeurbanne Cedex, France. "t"To whom all correspondenceshould be addressed. :~Affiliatedwith the Joseph Fourier Universityof Grenoble. 0141-8130/90/050289~36 © 1990 Butterworth-HeinemannLimited
polymorphs. For this, a method was devised to prepare fractionated low D P chitosan, susceptible of crystallization as a suspension of micrometer-sized single crystals that could be analysed either by X-ray or electron diffraction. This report deals with a high temperature polymorph of chitosan where the unit cell does not contain any water. This crystalline chitosan bears some similarities to the high temperature polymorph reported by Ogawa et al. 13. However, our diffraction data analysis indicates that the present single crystals of chitosan are crystallized with an unit cell that is substantially different from that given by Ogawa et al.
Experimental Preparation o f low molecular weight chitosan
Chitosan from Protan (lot number 075 291 02) with a D P V of 4100 (measured according to Roberts and Domszy 16) was used for this study. As its degree of substitution in NAc was still 0.17, further deacetylation was required. This was achieved according to the process previously described 17 but without the addition of thiophenol. The acetyl-free chitosan was dissolved in dilute HC1 and stored as an HC1 chitosan salt after freeze drying; 100 mg of this chitosan were dispersed in 5 ml of 12 N HC1 and the suspension heated to 72°C. The chitosan dissolved and this temperature was kept for 75min during which substantial depolymerization occurred. The suspension was then cooled and evaporated to dryness with a rotavapor. The product was then redissolved with 10 ml of water and the pH of the solution adjusted to 8 by adding aqueous ammonia. The chitosan reprecipitated and the precipitate was washed twice with isopropanol. It was then redissolved in dilute HC1, freeze dried and stored for further use. It had a D P of 35 (DPw/DP . = 1.14). Crystallization
Chitosan, 3 mg, was dissolved in 10 ml of distilled water. The solution was poured into an autoclave into which was also positioned a sealed thin wall glass ampoule containing 2 ml of a 0.15% (w/w) mixture of ammonia and water. The autoclave was sealed and
Int. J. Biol. Macromol., 1990, Vol. 12, October
289
Single crystals of chitosan." N. Cartier et al. heated to 125°C in an oil bath. After 15 min the glass ampoule was broken by shaking and the crystallization started immediately. The autoclave was kept at 125°C for 1 h and then cooled by quenching in cold water. The crystalline suspension was isolated and washed by successive centrifugation, always using aqueous ammonia solution at pH 13. The crystals were then washed with methanol and stored as suspensions in this alcohol. For shish-kebab preparation, fragments of purified Valonia ventricosa cell walls were purified according to the method of Gardner and Blackwel118. They were then added to the chitosan solutions prior to pouring into the autoclave. At the end of the experiment, all the chitosan crystals were attached to the Valonia fragments. These were washed by successive immersion in aqueous ammonia solution at pH 13 and stored in methanol.
Characterization of the crystals Electron microscopy and electron diffraction experiments. Drops of the crystal suspensions were deposited
,'
f
Figure 1 Typical preparation of single crystals of chitosan obtained with a chitosan fraction of DP 23, shadowed with W/Ta
on carbon-coated grids. After drying, these grids were examined with a Philips EM 400T electron microscope. For imaging purpose, these grids were either observed in low dose at 120 kV or after shadowing with W/Ta at 80 kV. Electron diffractograms were obtained on never observed crystals. Calibration of the electron diffractograms was achieved with a gold standard.
Density measurement. The crystals suspended in methanol were equilibrated by slowly adding CC14. The density of the mixture was measured at the equilibration point. X-ray analysis. Suspensions of the crystals were allowed to dry and their powder inserted into thin wal! capillaries. X-ray powder diagrams were recorded using a vacuum fiat film Wahrus X-ray camera mounted on a Siemens Kristalloflex generator operated with CuK~ radiation.
Results and discussion
Figure 2 Low dose electron micrograph recorded on chit0san single crystals obtained with unfractionated chitosan. Inset: electron diffraction diagram with proper orientation of the crystal located at the centre of the figure
The crystals and their diffraction features When low molecular weight chitosan was crystallized according to the above procedure, all the sample precipitated as a suspension of micrometre-sized platelet crystals. Each platelet had a square-like shape with diagonals ranging from 0.2 to 0.5 pm and a thickness of around 120 A. Some crystal preparations consisted of monolamellar platelets (Figure 1) whereas in others (e.g. in Figure 2) the crystals were made of a stack of several superimposed square lamellae. In Figure 1, it can be seen that the surface of each crystal is not smooth, but bears striations originating from one of the two diagonals of the crystal and running perpendicular to the crystals edges. Preparations of the chitosan crystals were subjected to X-ray and electron diffraction analysis. With X-ray, a sharp powder diagram was obtained displaying 11 independent diffraction lines (Figure 3 and Table 1). When studied by electron diffraction, an isolated crystal yielded a well resolved spot diagram (Figure 2) that could be indexed along a rectangular net. This diagram contains around 100 spots symmetrically distributed in four quadrants with symmetry axes a* and b* aligned along the orthogonal diagonals of the square crystal. In
290
Int. J. Biol. Macromol., 1990, Vol. 12, October
Figure 3 X-ray powder diagram recorded on a preparation of single crystals of chitosan
Single crystals of chitosan: N. Cartier et al. Figure 2, one sees that even and odd reflections are present along a* and b*. However, when tilting the crystal around a* and b*, the odd reflections disappeared and only the even reflections remained. This indicates that the odd reflections correspond to double diffraction and that only the even reflections have to be considered. These observations lead us to propose the P99 symmetry for the electron diffractogram and therefore for the projection of the corresponding crystalline structure. The symmetry of this projection as well as the overall three-dimensional symmetry was further established by rotating a chitosan crystal about its two diagonals a* and b*. This is illustrated in Figures 4 and 5 which show significant tilted patterns when a crystal was rotated about a* by 22 °, 38 ° or 57 ° and about b* by 17 ° and 38 °. With each of these angles, a crystal rotated clockwise gave a pattern identical
to the one recorded with the same crystal rotated by the same angle but anticlockwise. Thus, the third reciprocal axis c* is perpendicular to the plane defined by a* and b*. In Figures 4 and 5, one sees that in the tilted diagrams the diffraction spots are also symmetrically distributed in four quadrants. Overall, by adding the diffraction spots from the tilted and the untilted diagrams, more than 30 independent hk0 and 100 hkl reflections can be listed and their intensities evaluated. Out of these 130 reflections, 54 have intensities above the background level. The resolution of the diagrams extent to 1.4 A in the hk0 patterns and 1.5 A in the tilted patterns. The list of the observed reflections and their indexation is presented in Tables 2 and 3. When analysed, the diffraction information collected on the chitosan crystals indicates that the different intensities are related by the general rules: IF(hkl)[ = IF(K~T)I= IF(akl)l = IF(h~l)l = ]Fthki)l
Table 1 Indexation and observed intensities of a powder X-ray diffraction pattern recorded on a preparation of single crystals of chitosan hk 1
d spacings (A)
Intensities a
110 012, 102, 020 200 210, 120, 121 211 212, 220 221 130, 222, 301, 131,310 230 320, 223, 321 040, 232, 033, 322, 140, 400
5.76 4.37 4.03 3.64 3.41 2.93 2.76 2.61 2.35 2.23 2.09
VS S S S M M W M W W W
aVS: very strong; S: strong; M: medium; W: weak
IFtaoo~l = 0 if h is odd
IF~oko~l = 0
if k is odd
Thus, two space groups are possible for this crystalline chitosan, namely P21212~ or P2~212, the former being the most likely. U p o n calibration, the following cell parameters can be refined: a = 8.07 A, b = 8.44 A, and c = 10.34 A ~ = fl = y = 90 ° These parameters, together with the observed density of 1.47 (calculated density = 1.56) indicate that there are four glucosamine residues per cell connected by the symmetry elements of the P212121 space group. One can assume that the chitosan chain axis is perpendicular to the base of the platelet crystals. Thus, the distance of c = 10.34 A corresponds to the fibre repeat of chitosan and there are two antiparallel chains per unit cell, each of these being located on one of the crystal screw axes.
a°
a°
Figure 4 Electron diffraction patterns obtained from chitosan crystals rotated by + 22 °, + 38°, + 57° around a*, in comparison with an untilted diagram (tilt 0°). For each setting, the same crystal was used for the positive and negative orientations and the resulting diagrams were identical in each case. Different crystals were used for each setting
Int. J. Biol. Macromol., 1990, Vol. 12, October
291
Single crystals of chitosan." N. Cartier et al.
Figure 5
As in Figure 4 but the crystals were rotated about b* by + 17 and + 38 °
Table 2 List of the observed hk0 intensities in the electron diffractograms recorded on un-tilted chitosan single crystals
Table 3 List of the observed hkl intensities in the electron diffractograms recorded on tilted chitosan single crystals a
hk0
d spacings (A)
Intensities a
hkl
d spacings (A)
Intensitiest
110 020 200 120 210 220 130 230 320 040 140 400 410 240 420 510 250 530
5.81 4.27 4.04 3.75 3.64 2.90 2.67 2.31 2.27 2.11 2.04 2.02 1.95 1.87 1.82 1.59 1.56 1.40
VS VS M VS M M S W VW M M S W W W VW VW VW
011 012 102 112 201 121 211 022 202 212 221 301 131 222 311 223 321 024 033 124 232 322 141 224 411 233 034 142 402 134 412 421 044 151 512 521
6.53 4.42 4.34 3.86 3.75 3.51 3.43 3.26 3.18 2.97 2.80 2.60 2.57 2.53 2.49 2.22 2.21 2.19 2.17 2.12 2.10 2.07 2.00 1.93 1.92 1.91 1.90 1.89 1.87 1.85 1.83 1.79 1.63 1.63 1.51 1.49
W S W VW W M VW W W W M M W VW W W M W W VW W VW W VW M W VW VW W VW M W W W W W
a VS: very strong; S: strong; M: medium; W: weak; VW: very weak; the reflections having intensities lower than the background or belonging to systematic absences have not been listed
N o water molecule seems to be present within the unit cell. This is supported by the fact that the resolution of the above diffraction diagrams is insensitive to the vacuum of the electron microscope: in particular identical diagrams were recorded either under normal or frozen hydrated conditions. In addition, if water had been present within this unit cell, the density of the crystals would probably have been greater than the observed value of 1.47. The crystalline chitosan observed here bears some similarity not only to most of the other chitosan 3'8-15 structures reported so far but also to related polysaccharides. In particular, the repeat distance of 10.34 A matches that of m a n y (1-4)fl-linked crystalline h o m o polysaccharides such as cellulose 18, chitin 8, m a n n a n j9 etc. In all cases, the chains of these polysaccharides a d o p t a twofold helical structure similar to the one which we think is occurring here. A m o n g all the chitosan structures that have been described, the above chitosan p o l y m o r p h seems to correspond quite closely to the one reported by O g a w a et al. 13, where the crystallization was induced by subjecting chitosan films to a specific hydrothermal treatment. These treated films gave sharp X-ray fibre diagrams with diffraction data very similar to ours.
292
Int. J. Biol. Macromol., 1990, Vol. 12, October
The list of intensities corresponds to those that are observed when the crystals are rotated from + to - 60 ° about a* and b* t VS: very strong; S: strong; M: medium; W: weak; VW: very weak; the reflections having intensities lower than the background have not been listed
a
Remarkably, all the reflections of medium and strong intensities in their X-ray diagrams are observed in the present electron diffraction patterns. O g a w a et al.13 have selected a unit cell where their b parameter was roughly twice as large as ours. The doubling of their b parameter is based on the sole occurrence of a very weak reflection located at 3.40 A, on the third layer line of their pattern.
Single crystals of chitosan." N. Cartier et al. In our case, such reflection is inaccessible as it would require the tilting of a crystal about a* by an angle of 78 °, typically an angle that our goniometer cannot reach. In our diagrams, all the 60 visible diffraction spots to which we have access by electron diffraction, can be accounted for if b has only the value of 8.44 A as opposed to 17.88/~ proposed by Ogawa et al. 13. Therefore, we believe that this value is more realistic for the present chitosan polymorph which appears to be identical to that described by Ogawa et al. 13. Being of the P212t.2 ~ symmetry and with a and b parameters close to 8 A for c (chain axis) close to 10.4 A, the unit cell of the above chitosan is very reminiscent of those of either cellulose IVn 2° or mannan I19. In the case of mannan I, the resemblance goes even further as the hk0 electron diffractograms recorded on mannan I single crystals are very similar to the one presented above in Figure 2. Therefore, it is likely that the structure of the present chitosan will be closely related to that obtained for mannan 12~. This will be confirmed in a report to be published where the crystal and molecular structure of chitosan will be presented 22. In contrast to synthetic polymers such as high molecular weight polyethylene 23, it is difficult to determine with certainty whether the chitosan chains can fold within their crystals. In the present experiments, chitosan of D P 35 gave lamellar crystals where each lamella was around 120 A in thickness. This value is of the same order as the length of the chains used for the crystal preparation. Therefore, these crystals do not contain any regular chain folding. When chitosan samples of higher D P were used, a polycrystalline precipitate was obtained devoid of any regular platelet morphology. These observations indicate that lamellar crystals of chitosan with regular chain folding are not favoured and that it is only with low molecular weight material that lamellar and, in particular, monolamellar crystals can be obtained. Thus chitosan resembles other (1-4) fl-linked homopolysaccharides such as xylan 24, cellulose 25 and mannan 26 which give single crystals only with low molecular weight polymer fractions.
Figure 6 Shish-kebab morphology prepared by crystallization of chitosan in the presence of Valonia ventricosa microfibrils. Inset: electron diffraction diagram of the circled area showing the common fibre axis of the chitosan and the cellulose crystals
Figure 7 As in Figure 6 but prepared with an amount of cellulose twice as large
Oriented crystallization When the crystallization solutions of chitosan were seeded by various microfibrillar specimens, shish-kebab structures of great regularity were sometimes found. Among these, the best case was obtained when the seeds consisted of cellulose microfibrils from Valonia ventricosa cell wall fragments. Typical preparations of such shish-kebabs are presented in Figures 6-8 where one sees that all the initially smooth cellulose microfibrils (the shish) have been decorated by regular arrays of perpendicular chitosan lamellae (the kebab). All the chitosan is attached to the cellulose microfibrils and there are hardly any loose crystals in the preparation. In Figures 6-8, one sees that almost all the chitosan crystals (the kebabs) have dimensions that are close to one another in a given preparation. In Figure 6, the crystals have lateral sizes that are close to 0.1/~m. This value can be halved (Figure 7) or increased four times (Figure 8) simply by increasing or decreasing the quantity of cellulose that is added to the solution prior to crystallization. In Figure 6 is also presented an electron diffractogram that was recorded from a well-oriented area of the
Figure 8 As in Figure 6 but prepared with an amount of cellulose microfibrils twice as low specimen. The diagram shows the superposition of a sharp fibre diagram of cellulose and a less oriented chitosan fibre pattern whose projected fibre axis is parallel to that of ceUulose. This is indicated by the arcing of the four main equatorial reflections of chitosan, namely 110, 020 and 120 and 130. In samples such as those presented in Figures 6-8, the co-orientation of chitosan on cellulose must be correlated with similar cases studied earlier in our laboratory. Indeed, it was shown that native
Int. J. Biol. Macromol., 1990, Vol. 12, October
293
Single crystals of chitosan." N. Cartier et al. microfibrillar cellulose acted as a good nucleating substrate for the crystallization of m a n n a n 126, cellulose 1127 and cellulose IV. 2°. In these three cases, as in the present one, the initial onset of the nucleation must proceed through an epitaxy mechanism where the initial crystallizing entity must adsorb on the cellulose surface. The specificity of this adsorption will be enhanced if a close match can be found between the conformation of the cellulose chains and that of the crystallizing polysaccharide. Chitosan, having the same repeat distance as cellulose will fall into this category. The affinity of chitosan for cellulose is clearly demonstrated in this study which has shown that fractionated chitosan adsorbs specifically on Valonia cellulose. Such affinity must be of the same type as what is occurring when chitosan is added to cellulose pulp in the papermaking operation. In that case, a substantial increase in the wet strength properties of paper has been demonstrated 2s'29. These additions however are achieved with industrial chitosan of high molecular weight and at much lower temperatures than those that were selected here. In the case of chitosan-treated paper, it would be quite interesting to see whether the chitosan adsorption is disorganized or whether it consists also of structures as regular as those observed above.
Polysaccharides: Genetics Engineering, Structure/Properties Relations and Applications', (Ed. M. Yalpani), Elsevier Science, 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
References 1 2 3 4 5 6
294
Muzzarelli, R. A. A. In 'Chitin', Pergamon Press, Oxford, 1977 Ogawa, K., Oka, K., Miyanishi, T. and Hirano, S. in 'Chitin, Chitosan and Related Enzymes', (Ed. J. P. Zikakis), Academic Press, 1984, p 327 Ogawa, K. and Inukai, S. Carbohydr. Res. 1987, 160, 425 Park, J. W., Park, M. O. and Park, K. K. Bull. Korean Chem. Soc. 1984, 5, 108 Domard, A. Int. J. Biol. Macromol. 1987, 9, 98 Sandford, P. A. and Hutchings, G. P. in 'Industrial
I n t . J. Biol. M a c r o m o l . ,
1990, Vol. 12, O c t o b e r
24 25 26 27 28 29
B.V. Amsterdam, 1987 Hirano, S. in 'Ullmann's Encyclopedia of Industrial Chemistry', VCH Verlagsgesellschaft, Weinheim, 1986, Vol A6, p 231 Clark, G. L. and Smith, A. F. J. Phys. Chem. 1937, 40, 863 Samuels, R. J. J. Polym. Sci., Polym. Phys. Ed. 1981, 19, 1081 Sakurai, K., Fujimoto, J., Shibano, T. and Takahashi, T. Sen-i Gakkaishi 1983, 39, T-493 Sakurai, K., Takagi, M. and Takahashi, T. Sen-i Gakkaishi 1984, 40, T 246 Genin, Ya, V., Sklyar, A. M., Tsvankin, D. Ya, Gamzazade, A. I., Rogozhin, S. V. and Pavlova, S. S. A. Polym. Sci. USSR 1084, 26, 2701 Ogawa, K., Hirano, S., Miyanishi, T., Yui, T. and Watanabe, T. Macromolecules 1984, 17, 973 Sakurai, K., Shibano, T. and Takahashi, T. Mere. Fac. Eng Fukui University 1985, 33, 71 Sakurai, K., Shibano, T., Kimura, K. and Takahashi, T. Sen-i Gakkaishi 1985, 41, T 361 Roberts, G. A. F. and Domszy, J. G. Int. J. Biol. Macromol. 1982, 4, 374 Domard, A.andRinuado, M.Int. J. Biol. MacromoL1983,5,49 Gardner, K. H. and Blackwell, J. Biopolymers 1974, 13, 1975 Nieduszynski, I. and Marchessault, R. H. Can. J. Chem. 1972, 50, 2130 Bul6on, A. and Chanzy, H. J. Polym. Sci., Polym. Phys. Ed. 1980, 18, 1209 Chanzy, H., P6rez, S., Miller, D. P., Paradossi, G. and Winter, W. T. Macromolecules 1987, 20, 2407 Mazeau, K., Lavaitte, S., Chanzy, H. and Winter, W. T. (to be published) Wunderlich, B. 'Macromolecular Physics', Vol. 1, Academic Press, New York, 1973, p 232 Yundt, A. P. Tappi 1951, 34, 89 Bul6on, A., Chanzy, H. and Froment, P. J. Polym. Sci., Polym. Phys. Ed. 1982, 20, 1081 Chanzy, H., Dub6, M., Marchessault, R. H. and Revol, J. F. Biopolymers 1979, 18, 887 Bul6on, A., Chanzy, H. and Roche, E. Polym. Lett. 1977,15,265 Allan, G. G., Fox, J. R., Grosby, G. D. and Sarkanen, K. V. in 'Fiber water interactions in paper making', Transaction of the Symposium held at Oxford, Sept. 1977, Vol. II, p 765 Terrassin, C. Doctoral Thesis, University of Grenoble, 1986
(Received 27 January 1990; 27 April 1990) Lamellar single crystals o f chitosan were prepared at 125°C by adding ammonia to a low DP fraction o f ehitosan dissolved in water. The crystals gave sharp electron diffraction diagrams which could be indexed in an orthorhombic P212121 unit cell with a = 8.07 •, b = 8.44 1~, c = 10.34 ]~. The unit cell contained two anti-parallel chitosan chains and no water molecules. It was found that cellulose microfibrils from Valonia ventricosa could act as nuclei for inducing the crystallization o f chitosan on cellulose. This produced a shish-kebab morphology. Keywords: Chitosan; electron diffraction;single crystals; shish-kebab morphology
Introduction Chitosan, the N-deacylated chitin is a versatile biopolymer which displays a series of unique properties as a consequence of a free amino group in its repeating unit. Because of this, chitosan is able to complex efficiently with a series of substances ranging from negatively charged proteins to heavy atom ions or strong acids x-5. Such interesting properties explain why so many investigations are being pursued in order to understand and utilize the complexing possibilities of chitosan. It is along this line that one can mention recent industrial developments where chitosan has demonstrated its efficiency in the clarification of waste water as well as of industrial effluents 6'7. The geometry of the chitosan molecule and, in particular, its conformation and the environment of its amino group are of great interest in understanding the specific properties of this biopolymer. Several crystallographic studies have been achieved on chitosan fibres and films 3'8-15. These studies have revealed that in most of the cases, chitosan crystallized with a fibre repeat comprising between 10.1 and 10.5 ,~, indicating a twofold helical structure for the chitosan molecule. In a few instances however, a fibre repeat of the order of 40 A was also observed but so far not explained 3. An unusual number of crystalline polymorphs seems to exist for chitosan. These polymorphs correspond to either chitosan alone or chitosan complexed with small molecules such as water, acids or salts. So far, only a small number of these polymorphs has been resolved by X-ray analysis and structure refinement to the point where atomic coordinates are given 14'15. Thus most of the conformational possibilities of chitosan remain to be unravelled. In particular, there is a need to know better the environment of the amino moiety that confers its specific properties to chitosan. This work was undertaken to study in detail the crystalline arrangement of chitosan in a number of its * Present address: Laboratoire d'Etude des Mati6res Plastiques et des Biomat6riaux, Universit6 Claude Bernard, 43 Bvd du 11 Novembre 1918, 69622 Villeurbanne Cedex, France. "t"To whom all correspondenceshould be addressed. :~Affiliatedwith the Joseph Fourier Universityof Grenoble. 0141-8130/90/050289~36 © 1990 Butterworth-HeinemannLimited
polymorphs. For this, a method was devised to prepare fractionated low D P chitosan, susceptible of crystallization as a suspension of micrometer-sized single crystals that could be analysed either by X-ray or electron diffraction. This report deals with a high temperature polymorph of chitosan where the unit cell does not contain any water. This crystalline chitosan bears some similarities to the high temperature polymorph reported by Ogawa et al. 13. However, our diffraction data analysis indicates that the present single crystals of chitosan are crystallized with an unit cell that is substantially different from that given by Ogawa et al.
Experimental Preparation o f low molecular weight chitosan
Chitosan from Protan (lot number 075 291 02) with a D P V of 4100 (measured according to Roberts and Domszy 16) was used for this study. As its degree of substitution in NAc was still 0.17, further deacetylation was required. This was achieved according to the process previously described 17 but without the addition of thiophenol. The acetyl-free chitosan was dissolved in dilute HC1 and stored as an HC1 chitosan salt after freeze drying; 100 mg of this chitosan were dispersed in 5 ml of 12 N HC1 and the suspension heated to 72°C. The chitosan dissolved and this temperature was kept for 75min during which substantial depolymerization occurred. The suspension was then cooled and evaporated to dryness with a rotavapor. The product was then redissolved with 10 ml of water and the pH of the solution adjusted to 8 by adding aqueous ammonia. The chitosan reprecipitated and the precipitate was washed twice with isopropanol. It was then redissolved in dilute HC1, freeze dried and stored for further use. It had a D P of 35 (DPw/DP . = 1.14). Crystallization
Chitosan, 3 mg, was dissolved in 10 ml of distilled water. The solution was poured into an autoclave into which was also positioned a sealed thin wall glass ampoule containing 2 ml of a 0.15% (w/w) mixture of ammonia and water. The autoclave was sealed and
Int. J. Biol. Macromol., 1990, Vol. 12, October
289
Single crystals of chitosan." N. Cartier et al. heated to 125°C in an oil bath. After 15 min the glass ampoule was broken by shaking and the crystallization started immediately. The autoclave was kept at 125°C for 1 h and then cooled by quenching in cold water. The crystalline suspension was isolated and washed by successive centrifugation, always using aqueous ammonia solution at pH 13. The crystals were then washed with methanol and stored as suspensions in this alcohol. For shish-kebab preparation, fragments of purified Valonia ventricosa cell walls were purified according to the method of Gardner and Blackwel118. They were then added to the chitosan solutions prior to pouring into the autoclave. At the end of the experiment, all the chitosan crystals were attached to the Valonia fragments. These were washed by successive immersion in aqueous ammonia solution at pH 13 and stored in methanol.
Characterization of the crystals Electron microscopy and electron diffraction experiments. Drops of the crystal suspensions were deposited
,'
f
Figure 1 Typical preparation of single crystals of chitosan obtained with a chitosan fraction of DP 23, shadowed with W/Ta
on carbon-coated grids. After drying, these grids were examined with a Philips EM 400T electron microscope. For imaging purpose, these grids were either observed in low dose at 120 kV or after shadowing with W/Ta at 80 kV. Electron diffractograms were obtained on never observed crystals. Calibration of the electron diffractograms was achieved with a gold standard.
Density measurement. The crystals suspended in methanol were equilibrated by slowly adding CC14. The density of the mixture was measured at the equilibration point. X-ray analysis. Suspensions of the crystals were allowed to dry and their powder inserted into thin wal! capillaries. X-ray powder diagrams were recorded using a vacuum fiat film Wahrus X-ray camera mounted on a Siemens Kristalloflex generator operated with CuK~ radiation.
Results and discussion
Figure 2 Low dose electron micrograph recorded on chit0san single crystals obtained with unfractionated chitosan. Inset: electron diffraction diagram with proper orientation of the crystal located at the centre of the figure
The crystals and their diffraction features When low molecular weight chitosan was crystallized according to the above procedure, all the sample precipitated as a suspension of micrometre-sized platelet crystals. Each platelet had a square-like shape with diagonals ranging from 0.2 to 0.5 pm and a thickness of around 120 A. Some crystal preparations consisted of monolamellar platelets (Figure 1) whereas in others (e.g. in Figure 2) the crystals were made of a stack of several superimposed square lamellae. In Figure 1, it can be seen that the surface of each crystal is not smooth, but bears striations originating from one of the two diagonals of the crystal and running perpendicular to the crystals edges. Preparations of the chitosan crystals were subjected to X-ray and electron diffraction analysis. With X-ray, a sharp powder diagram was obtained displaying 11 independent diffraction lines (Figure 3 and Table 1). When studied by electron diffraction, an isolated crystal yielded a well resolved spot diagram (Figure 2) that could be indexed along a rectangular net. This diagram contains around 100 spots symmetrically distributed in four quadrants with symmetry axes a* and b* aligned along the orthogonal diagonals of the square crystal. In
290
Int. J. Biol. Macromol., 1990, Vol. 12, October
Figure 3 X-ray powder diagram recorded on a preparation of single crystals of chitosan
Single crystals of chitosan: N. Cartier et al. Figure 2, one sees that even and odd reflections are present along a* and b*. However, when tilting the crystal around a* and b*, the odd reflections disappeared and only the even reflections remained. This indicates that the odd reflections correspond to double diffraction and that only the even reflections have to be considered. These observations lead us to propose the P99 symmetry for the electron diffractogram and therefore for the projection of the corresponding crystalline structure. The symmetry of this projection as well as the overall three-dimensional symmetry was further established by rotating a chitosan crystal about its two diagonals a* and b*. This is illustrated in Figures 4 and 5 which show significant tilted patterns when a crystal was rotated about a* by 22 °, 38 ° or 57 ° and about b* by 17 ° and 38 °. With each of these angles, a crystal rotated clockwise gave a pattern identical
to the one recorded with the same crystal rotated by the same angle but anticlockwise. Thus, the third reciprocal axis c* is perpendicular to the plane defined by a* and b*. In Figures 4 and 5, one sees that in the tilted diagrams the diffraction spots are also symmetrically distributed in four quadrants. Overall, by adding the diffraction spots from the tilted and the untilted diagrams, more than 30 independent hk0 and 100 hkl reflections can be listed and their intensities evaluated. Out of these 130 reflections, 54 have intensities above the background level. The resolution of the diagrams extent to 1.4 A in the hk0 patterns and 1.5 A in the tilted patterns. The list of the observed reflections and their indexation is presented in Tables 2 and 3. When analysed, the diffraction information collected on the chitosan crystals indicates that the different intensities are related by the general rules: IF(hkl)[ = IF(K~T)I= IF(akl)l = IF(h~l)l = ]Fthki)l
Table 1 Indexation and observed intensities of a powder X-ray diffraction pattern recorded on a preparation of single crystals of chitosan hk 1
d spacings (A)
Intensities a
110 012, 102, 020 200 210, 120, 121 211 212, 220 221 130, 222, 301, 131,310 230 320, 223, 321 040, 232, 033, 322, 140, 400
5.76 4.37 4.03 3.64 3.41 2.93 2.76 2.61 2.35 2.23 2.09
VS S S S M M W M W W W
aVS: very strong; S: strong; M: medium; W: weak
IFtaoo~l = 0 if h is odd
IF~oko~l = 0
if k is odd
Thus, two space groups are possible for this crystalline chitosan, namely P21212~ or P2~212, the former being the most likely. U p o n calibration, the following cell parameters can be refined: a = 8.07 A, b = 8.44 A, and c = 10.34 A ~ = fl = y = 90 ° These parameters, together with the observed density of 1.47 (calculated density = 1.56) indicate that there are four glucosamine residues per cell connected by the symmetry elements of the P212121 space group. One can assume that the chitosan chain axis is perpendicular to the base of the platelet crystals. Thus, the distance of c = 10.34 A corresponds to the fibre repeat of chitosan and there are two antiparallel chains per unit cell, each of these being located on one of the crystal screw axes.
a°
a°
Figure 4 Electron diffraction patterns obtained from chitosan crystals rotated by + 22 °, + 38°, + 57° around a*, in comparison with an untilted diagram (tilt 0°). For each setting, the same crystal was used for the positive and negative orientations and the resulting diagrams were identical in each case. Different crystals were used for each setting
Int. J. Biol. Macromol., 1990, Vol. 12, October
291
Single crystals of chitosan." N. Cartier et al.
Figure 5
As in Figure 4 but the crystals were rotated about b* by + 17 and + 38 °
Table 2 List of the observed hk0 intensities in the electron diffractograms recorded on un-tilted chitosan single crystals
Table 3 List of the observed hkl intensities in the electron diffractograms recorded on tilted chitosan single crystals a
hk0
d spacings (A)
Intensities a
hkl
d spacings (A)
Intensitiest
110 020 200 120 210 220 130 230 320 040 140 400 410 240 420 510 250 530
5.81 4.27 4.04 3.75 3.64 2.90 2.67 2.31 2.27 2.11 2.04 2.02 1.95 1.87 1.82 1.59 1.56 1.40
VS VS M VS M M S W VW M M S W W W VW VW VW
011 012 102 112 201 121 211 022 202 212 221 301 131 222 311 223 321 024 033 124 232 322 141 224 411 233 034 142 402 134 412 421 044 151 512 521
6.53 4.42 4.34 3.86 3.75 3.51 3.43 3.26 3.18 2.97 2.80 2.60 2.57 2.53 2.49 2.22 2.21 2.19 2.17 2.12 2.10 2.07 2.00 1.93 1.92 1.91 1.90 1.89 1.87 1.85 1.83 1.79 1.63 1.63 1.51 1.49
W S W VW W M VW W W W M M W VW W W M W W VW W VW W VW M W VW VW W VW M W W W W W
a VS: very strong; S: strong; M: medium; W: weak; VW: very weak; the reflections having intensities lower than the background or belonging to systematic absences have not been listed
N o water molecule seems to be present within the unit cell. This is supported by the fact that the resolution of the above diffraction diagrams is insensitive to the vacuum of the electron microscope: in particular identical diagrams were recorded either under normal or frozen hydrated conditions. In addition, if water had been present within this unit cell, the density of the crystals would probably have been greater than the observed value of 1.47. The crystalline chitosan observed here bears some similarity not only to most of the other chitosan 3'8-15 structures reported so far but also to related polysaccharides. In particular, the repeat distance of 10.34 A matches that of m a n y (1-4)fl-linked crystalline h o m o polysaccharides such as cellulose 18, chitin 8, m a n n a n j9 etc. In all cases, the chains of these polysaccharides a d o p t a twofold helical structure similar to the one which we think is occurring here. A m o n g all the chitosan structures that have been described, the above chitosan p o l y m o r p h seems to correspond quite closely to the one reported by O g a w a et al. 13, where the crystallization was induced by subjecting chitosan films to a specific hydrothermal treatment. These treated films gave sharp X-ray fibre diagrams with diffraction data very similar to ours.
292
Int. J. Biol. Macromol., 1990, Vol. 12, October
The list of intensities corresponds to those that are observed when the crystals are rotated from + to - 60 ° about a* and b* t VS: very strong; S: strong; M: medium; W: weak; VW: very weak; the reflections having intensities lower than the background have not been listed
a
Remarkably, all the reflections of medium and strong intensities in their X-ray diagrams are observed in the present electron diffraction patterns. O g a w a et al.13 have selected a unit cell where their b parameter was roughly twice as large as ours. The doubling of their b parameter is based on the sole occurrence of a very weak reflection located at 3.40 A, on the third layer line of their pattern.
Single crystals of chitosan." N. Cartier et al. In our case, such reflection is inaccessible as it would require the tilting of a crystal about a* by an angle of 78 °, typically an angle that our goniometer cannot reach. In our diagrams, all the 60 visible diffraction spots to which we have access by electron diffraction, can be accounted for if b has only the value of 8.44 A as opposed to 17.88/~ proposed by Ogawa et al. 13. Therefore, we believe that this value is more realistic for the present chitosan polymorph which appears to be identical to that described by Ogawa et al. 13. Being of the P212t.2 ~ symmetry and with a and b parameters close to 8 A for c (chain axis) close to 10.4 A, the unit cell of the above chitosan is very reminiscent of those of either cellulose IVn 2° or mannan I19. In the case of mannan I, the resemblance goes even further as the hk0 electron diffractograms recorded on mannan I single crystals are very similar to the one presented above in Figure 2. Therefore, it is likely that the structure of the present chitosan will be closely related to that obtained for mannan 12~. This will be confirmed in a report to be published where the crystal and molecular structure of chitosan will be presented 22. In contrast to synthetic polymers such as high molecular weight polyethylene 23, it is difficult to determine with certainty whether the chitosan chains can fold within their crystals. In the present experiments, chitosan of D P 35 gave lamellar crystals where each lamella was around 120 A in thickness. This value is of the same order as the length of the chains used for the crystal preparation. Therefore, these crystals do not contain any regular chain folding. When chitosan samples of higher D P were used, a polycrystalline precipitate was obtained devoid of any regular platelet morphology. These observations indicate that lamellar crystals of chitosan with regular chain folding are not favoured and that it is only with low molecular weight material that lamellar and, in particular, monolamellar crystals can be obtained. Thus chitosan resembles other (1-4) fl-linked homopolysaccharides such as xylan 24, cellulose 25 and mannan 26 which give single crystals only with low molecular weight polymer fractions.
Figure 6 Shish-kebab morphology prepared by crystallization of chitosan in the presence of Valonia ventricosa microfibrils. Inset: electron diffraction diagram of the circled area showing the common fibre axis of the chitosan and the cellulose crystals
Figure 7 As in Figure 6 but prepared with an amount of cellulose twice as large
Oriented crystallization When the crystallization solutions of chitosan were seeded by various microfibrillar specimens, shish-kebab structures of great regularity were sometimes found. Among these, the best case was obtained when the seeds consisted of cellulose microfibrils from Valonia ventricosa cell wall fragments. Typical preparations of such shish-kebabs are presented in Figures 6-8 where one sees that all the initially smooth cellulose microfibrils (the shish) have been decorated by regular arrays of perpendicular chitosan lamellae (the kebab). All the chitosan is attached to the cellulose microfibrils and there are hardly any loose crystals in the preparation. In Figures 6-8, one sees that almost all the chitosan crystals (the kebabs) have dimensions that are close to one another in a given preparation. In Figure 6, the crystals have lateral sizes that are close to 0.1/~m. This value can be halved (Figure 7) or increased four times (Figure 8) simply by increasing or decreasing the quantity of cellulose that is added to the solution prior to crystallization. In Figure 6 is also presented an electron diffractogram that was recorded from a well-oriented area of the
Figure 8 As in Figure 6 but prepared with an amount of cellulose microfibrils twice as low specimen. The diagram shows the superposition of a sharp fibre diagram of cellulose and a less oriented chitosan fibre pattern whose projected fibre axis is parallel to that of ceUulose. This is indicated by the arcing of the four main equatorial reflections of chitosan, namely 110, 020 and 120 and 130. In samples such as those presented in Figures 6-8, the co-orientation of chitosan on cellulose must be correlated with similar cases studied earlier in our laboratory. Indeed, it was shown that native
Int. J. Biol. Macromol., 1990, Vol. 12, October
293
Single crystals of chitosan." N. Cartier et al. microfibrillar cellulose acted as a good nucleating substrate for the crystallization of m a n n a n 126, cellulose 1127 and cellulose IV. 2°. In these three cases, as in the present one, the initial onset of the nucleation must proceed through an epitaxy mechanism where the initial crystallizing entity must adsorb on the cellulose surface. The specificity of this adsorption will be enhanced if a close match can be found between the conformation of the cellulose chains and that of the crystallizing polysaccharide. Chitosan, having the same repeat distance as cellulose will fall into this category. The affinity of chitosan for cellulose is clearly demonstrated in this study which has shown that fractionated chitosan adsorbs specifically on Valonia cellulose. Such affinity must be of the same type as what is occurring when chitosan is added to cellulose pulp in the papermaking operation. In that case, a substantial increase in the wet strength properties of paper has been demonstrated 2s'29. These additions however are achieved with industrial chitosan of high molecular weight and at much lower temperatures than those that were selected here. In the case of chitosan-treated paper, it would be quite interesting to see whether the chitosan adsorption is disorganized or whether it consists also of structures as regular as those observed above.
Polysaccharides: Genetics Engineering, Structure/Properties Relations and Applications', (Ed. M. Yalpani), Elsevier Science, 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
References 1 2 3 4 5 6
294
Muzzarelli, R. A. A. In 'Chitin', Pergamon Press, Oxford, 1977 Ogawa, K., Oka, K., Miyanishi, T. and Hirano, S. in 'Chitin, Chitosan and Related Enzymes', (Ed. J. P. Zikakis), Academic Press, 1984, p 327 Ogawa, K. and Inukai, S. Carbohydr. Res. 1987, 160, 425 Park, J. W., Park, M. O. and Park, K. K. Bull. Korean Chem. Soc. 1984, 5, 108 Domard, A. Int. J. Biol. Macromol. 1987, 9, 98 Sandford, P. A. and Hutchings, G. P. in 'Industrial
I n t . J. Biol. M a c r o m o l . ,
1990, Vol. 12, O c t o b e r
24 25 26 27 28 29
B.V. Amsterdam, 1987 Hirano, S. in 'Ullmann's Encyclopedia of Industrial Chemistry', VCH Verlagsgesellschaft, Weinheim, 1986, Vol A6, p 231 Clark, G. L. and Smith, A. F. J. Phys. Chem. 1937, 40, 863 Samuels, R. J. J. Polym. Sci., Polym. Phys. Ed. 1981, 19, 1081 Sakurai, K., Fujimoto, J., Shibano, T. and Takahashi, T. Sen-i Gakkaishi 1983, 39, T-493 Sakurai, K., Takagi, M. and Takahashi, T. Sen-i Gakkaishi 1984, 40, T 246 Genin, Ya, V., Sklyar, A. M., Tsvankin, D. Ya, Gamzazade, A. I., Rogozhin, S. V. and Pavlova, S. S. A. Polym. Sci. USSR 1084, 26, 2701 Ogawa, K., Hirano, S., Miyanishi, T., Yui, T. and Watanabe, T. Macromolecules 1984, 17, 973 Sakurai, K., Shibano, T. and Takahashi, T. Mere. Fac. Eng Fukui University 1985, 33, 71 Sakurai, K., Shibano, T., Kimura, K. and Takahashi, T. Sen-i Gakkaishi 1985, 41, T 361 Roberts, G. A. F. and Domszy, J. G. Int. J. Biol. Macromol. 1982, 4, 374 Domard, A.andRinuado, M.Int. J. Biol. MacromoL1983,5,49 Gardner, K. H. and Blackwell, J. Biopolymers 1974, 13, 1975 Nieduszynski, I. and Marchessault, R. H. Can. J. Chem. 1972, 50, 2130 Bul6on, A. and Chanzy, H. J. Polym. Sci., Polym. Phys. Ed. 1980, 18, 1209 Chanzy, H., P6rez, S., Miller, D. P., Paradossi, G. and Winter, W. T. Macromolecules 1987, 20, 2407 Mazeau, K., Lavaitte, S., Chanzy, H. and Winter, W. T. (to be published) Wunderlich, B. 'Macromolecular Physics', Vol. 1, Academic Press, New York, 1973, p 232 Yundt, A. P. Tappi 1951, 34, 89 Bul6on, A., Chanzy, H. and Froment, P. J. Polym. Sci., Polym. Phys. Ed. 1982, 20, 1081 Chanzy, H., Dub6, M., Marchessault, R. H. and Revol, J. F. Biopolymers 1979, 18, 887 Bul6on, A., Chanzy, H. and Roche, E. Polym. Lett. 1977,15,265 Allan, G. G., Fox, J. R., Grosby, G. D. and Sarkanen, K. V. in 'Fiber water interactions in paper making', Transaction of the Symposium held at Oxford, Sept. 1977, Vol. II, p 765 Terrassin, C. Doctoral Thesis, University of Grenoble, 1986
