Carbohydrate

Polymers
ELSEVIER Carbohydrate Polymers 36 (1998) 157-165

Preparation and NMR characterization of highly substituted IV-trimethyl

chitosan chloride
A.B. Sieval’“, M. Thanoual*, A.F. Kotzkb, J.C. Verhoefa, J. BrusseeC, H.E. Junginger”
“Department of Pharmaceutical Technology, Leiden/Amsterdam Center for Drug Research, L&den University, PO Box 9502, 2300 RA Leiden,
The Netherlands
bDepartment of Pharmaceutics, Potchefstroom University for CHE, Potchefstroom 2520, South Africa
‘Department of Organic Chemistry, Leiden University PO Box 9502, 2300 RA Leiden, The Netherlands

Received 18 August 1997; revised 8 January 1998; accepted 16 January 1998

Abstract

N,N,N-Trimethyl chitosan chloride (TMC) is a chemically modified chitosan with improved aqueous solubility, compared with the native
chitosan. It is essential to follow a synthesis procedure in which the degree of substitution of the final product can be controlled by means of
the number of reaction steps, the duration of each reaction step and the amount of methyl iodide as reagent. A two-step reaction yields
products with high degrees of substitution (40-80%). Comparison of the NMR spectra of the product TMC, after a two-step reaction,
indicates that there is a peak assigned to the substituted amino group that shifts from 2.5 to 3.1 ppm upon acidification. This peak must be
assigned to N(CH3)* and not to N(CH3):. A three-step reaction procedure yields products with a degree of substitution > 80%, but with
substantially decreased water-solubility. 0 1998 Elsevier Science Ltd. All rights reserved

Keywords: N,N,N-Trimethyl chitosan chloride; NMR; Synthesis; Chitosan; Absorption enhancers

1. Introduction especially at the physiological pH value (7.4) where chito-

san is insoluble and consequently less effective. In past
Chitosan is a polymer which attracts high interest in the years, several derivatives of chitosan have been synthesized
pharmaceutical research area as a polymeric carrier of which are water-soluble over a wider pH range. Among
drugs. Recently, it has been shown that chitosan enhances these are N-carboxymethylated chitosan (Muzzarelli et al.,
the absorption of peptide and protein drug across nasal 1982; Le Dung et al., 1994) and N&methyl chitosan chlor-
(Illum et al., 1994) and intestinal epithelia (Artursson et ide (Le Dung et al., 1994; Domard et al., 1986; Muzzarelli &
al., 1994). Chitosan glutamate does not only enhance the Tanfani, 1985). N-Trimethyl chitosan chloride (TMC) is of
in vitro transport of small hydrophilic compounds like 14C- particular interest because of its well-defined structure,
mannitol (Borchard et al., 1996), but also the transport of improved solubility and easy preparation as described by
high-molecular-weight peptide drugs like 9-desglycina- Le Dung et al. (1994). They obtained 53% of quatemization
mide, 8-arginine vasopressin (LueBen et al., 1997). Chitosan of the amino group in just one reaction step, without any
hydrochloride has been validated as an intestinal absorption observable 0-methylation. This method was used as a basis
enhancer in vivo of the peptide drug buserelin when it was to prepare TMC with different degrees of substitution from
coadministered in rats (Luel3en et al., 1996). chitosan, in order to study the potential of these derivatives
However, chitosan has an apparent pK, of 5.6 (as mea- as absorption enhancers for hydrophilic and macromolecu-
sured by potentiometric titration) and is only soluble in lar drugs such as peptides and proteins.
acidic solutions with pH values lower than 6.0. This inter-
feres with the biomedical application of this polymer,
2. Experimental
* Corresponding author. Tel.: 0031715274219; fax: 0031 71 5274565; e-
mail: [email protected] Chitosan (93% deacetylated) was a gift from Pronova
’ Present address: Department of Organic Chemistry, Wageningen
Agricultural University, Dreijenplein 8, 6703 HB Wageningen, The Biopolymer A.S. (Drammen, Norway). The material was
Netherlands. crushed in a mortar with a pestle and sieved to obtain the

0144~8617/98/$19.00 0 1998 Elsevier Science Ltd. All rights reserved

PI1 SO144-8617(98)00009-5
1.58 A. B. Sieval et al./Carbohydrate Polymers 36 (1998) 157-165

fraction less than 500 pm. Only material 0’ small particle spectrum reported by Le Dung et al. (1994), except for the
size showed sufficient solubility. Methyl iodide and l- smaller acetyl peak at 2.0 ppm. It clearly shows a peak at
methyl-2-pyrrolidinone were obtained from Acres (Geel, 3.1 ppm assigned to N(CH&, together with a smaller peak
Belgium). All other chemicals are commercially available at 3.4 ppm assigned to N(CH&. According to the peak
and were used as received. assignment by Le Dung et al. (1994), we calculated primar-
One-step synthesis. To prepare TMC, 2 g of chitosan ily the degree of quaternization as 35%.
(sieved, < 500 pm) and 4.8 g of sodium iodide were dis- To investigate the effect of degree of substitution for
solved in 80 ml of 1-methyl-2-pyrrolidinone on a water bath pharmaceutical experiments, we tried to prepare TMC
at 60°C with stirring. After the chitosan was dissolved, with different and preferably higher substitution degrees.
11 ml of a 15% aqueous sodium hydroxide solution were After several attempts, varying the reaction times (45 to
added, and then 11.5 ml of methyl iodide, both with stirring. 360 min) and the amount of sodium hydroxide (aqueous
The mixture was stirred for 1 h (Le Dung et al., 1994). The solutions from 15 to 25%), we concluded that the optimal
product was precipitated using ethanol and subsequently way to synthesize highly substituted N-trimetbyl chitosan is
isolated by centrifugation. After washing with ethanol and the two-step synthesis as described above. After the first
centrifugation, the material was dissolved in 40 ml of water, step complete removal of the ether turned out to be proble-
to which 250 ml of 1 M HCl in ethanol (96%) were added matic, but the residual amount could be used to our advan-
carefully, thus exchanging the iodide for chloride. Centrifu- tage, because it refluxes together with the methyl iodide,
gation and washing with ethanol and subsequently with thus increasing the amount of methyl iodide in solution.
ether yielded a white, water-soluble powder, which was The addition of 2 ml of methyl iodide and 0.6 g of NaOH
dried in vacua at 40°C. pellets, at the end of the second step of the reaction, was
Two-step synthesis. A mixture of 2 g of sieved chitosan necessary since without this lower degrees of quatemization
(93% deacetylated), 4.8 g of sodium iodide, 11 ml of a 15% were obtained.
aqueous sodium hydroxide solution and 11.5 ml of methyl A ‘H NMR spectrum of the product obtained via the two-
iodide in 80 ml of 1-methyl-2-pyrrolidinone was stirred on a step synthesis showed unexpected results when compared to
water bath of 60°C for 1 h (Le Dung et al., 1994). Special the intermediate material. The peak at 3.0 ppm assigned by
care was taken to keep the methyl iodide in the reaction Le Dung et al. (1994) to the quatemized amino group has
mixture by using a Liebig condenser. The product was disappeared, the peak at 3.35 ppm assigned to the tertiary
precipitated using ethanol and thereafter isolated by centri- amino group has increased substantially, and a new peak
fugation. The N-trimethyl chitosan iodide obtained after this appeared at 2.5 ppm (Fig. 2). Upon addition of a drop of
first step was washed twice with ether on a glass filter to DCl, this new peak shifted to the quatemized position at
remove the ethanol. It was dissolved in 80 ml of 1-methyl-2- 3.1 ppm (Fig. 3). It is very unlikely that the NMR signal
pyrrolidinone and heated to 6O”C,thus removing most of the of a quatemized amino group will shift upon acidification of
absorbed ether. Subsequently, 4.8 g of NaI, 11 ml of 15% the solution. Therefore, the signal at 2.5 ppm in the neutral
NaOH solution and 7 ml of methyl iodide were added with solution must be from the dimethylated amino group, which,
rapid stirring and the mixture was heated on a water bath at upon acidification, will shift to a lower field. This leaves the
60°C for 30 min. An additional 2 ml of methyl iodide and peak at 3.4 ppm, which must be assigned to the quaternary
0.6 g of NaOH pellets were added and the stirring was con- amino group. The results described above indicate that Le
tinued for 1 h. Dung et al. (1994) misinterpreted their spectra and did not
The product, prepared as described above, was dissolved prepare N-trimethyl chitosan, but a mainly dimethylated
in 40 ml of a 10% NaCl aqueous solution, instead of HCI, to polymer with only lo-15% of quatemization. Using the
exchange the iodide. The polymer was precipitated with two-step method, degrees of quatemization of at least
ethanol, isolated by centrifugation and thoroughly washed 60% were obtained. This also explains our observation
with ethanol and ether. In vacua drying yielded a white, that a poorly water-soluble polymer was obtained after the
water-soluble powder. first step (15 min for the dissolution of a 2% (w/v) product
‘H NMR spectra were measured in D20 at 80°C using a upon stirring), when the iodide was exchanged for chloride
300 or 600 MHz spectrometer (Bruker, Switzerland). 13C using NaCl instead of hydrochloric acid. Upon acidification,
NMR spectra were measured in DzO at 80°C at 150 MHz. this material instantly dissolved in water. The product after
No attempts were made to remove the residual water from the two-step reaction rapidly dissolves in water in concen-
the NMR sample, because at 80°C this peak does not inter- trations up to 5% (w/v).
fere with the spectrum of the polymer. Further evidence for our hypothesis was obtained from
13CNMR spectroscopy. The spectra for the low- and high-
substituted material are presented in Fig. 4 and Fig. 5, respec-
3. Results and discussion tively. Also here the disappearance of the dimethylated sig-
nal, at 43.7 ppm (Muzzarelli & Tanfani, 1985), and the
The ‘H NMR spectrum of the TMC obtained after one- appearance of the trimethylated signal, at 55.1 ppm, are
step synthesis is depicted in Fig. 1. It is similar to the clearly visible. As expected, the highly substituted polymer
I I 1 I I 1 I I
PPm 5.0 4.5 4.0 3.5 3.0 2.5 2.0

Fig. 1. ‘H NMR spectrum of N-trimethyl chitosan chloride after the one-step reaction, according to &he proposed peak assignment; degree of substitution 35%.
160 A. B. Sieval et aLCarbohydrate Polymers 36 (1998) 157-165
 N(CH3)z

Fig. 3. ‘H NMR spectrum of N-trimethyl chitosan chloride after a two-step synthesis upon the addition of DC1 (one drop). The peak which had appeared at 2.5 ppm shifted after acidification to the referred
quatemized position 3.1 ppm. Therefore, the signal at 2.5 ppm in Fig. 2 must be from the dimethylated group.
 N(CHQ2

+N(CH+

‘....,.........,““““‘l”“““‘l”“~~~~~~=~~~’~~~~~~‘~~‘~~~~l”~~’~~~’l~~~~~~~~~~

wm 100 90 80 70 60 50 40 30

Fig. 4. 13C NMR spectrum of N-trimethyl chitosan chloride after the one-step synthesis.
 +N(CH+

Fig. 5. ‘k NhIR spectrum of N-trimethylchitosan chloride after a two-step synthesis.

164 A. B. Sieval et aNCarbohydrate Polymers 36 (1998) 157-165
 A. B. Sieval et al/Carbohydrate Polymers 36 (1998) 157-165 165

still shows a residual signal from the dimethylated part at Chemical Society. A. W. M. Lefeber is gratefully acknowl-
43.1 ppm. edged for recording the NMR spectra.
A three-step reaction procedure resulted in an even higher
substituted polymer (degree of quatemization > 85%).
However, it also resulted in complete 0-methylation, References
which strongly decreases the solubility. The ‘H NMR spec-
trum of this N,O-permethyl chitosan in D20 is depicted in Artursson, P., Lindmark, T., Davis, S.S., & Illum, L. (1994). Effect of
chitosan on the permeability of monolayers of intestinal epithelial
Fig. 6. It shows the same N(CH& peak as in Fig. 3, together
cells (Caco-2). Pbarm. Rex, 11, 1358-1361.
with those for 3-OCHs and 6-OCHs at 3.55 and 3.45 ppm, Borchard, G., LueRen, H.L., De Boer, A.G., Verhoef, J.C., Lehr, C.-M., &
respectively. This spectrum is similar to the one previously Junginger, H.E. (1996). The potential of mucoadhesive polymers in
published (Domard et al., 1987). enhancing intestinal peptide drug absorption. III. Effects of chitosan-
glutamate and carbomer on epithelial tight junctions in vitro. J. Control.
Rel., 39, 131-138.
4. Conclusions Domard, A., Rinaudo, M., & Tetrassin, C. (1986). New method for the
quatemization of chitosan. ht. J. Biol. Macromol., 8, 105-107.
Domard. A., Gey, C., Rinaudo, M., & Termssin, C. (1987). ‘? and ‘H
Our results show that N&methyl chitosan chloride can-
n.m.r. spectroscopy of chitosan and N-trimethyl chloride derivatives.
not be prepared in high degrees of substitution by the one- ht. J. Biol. Macromol., 9, 233-237.
step method as described by Le Dung et al. (1994). It results Illurn, L., Farray, N.F., & Davis, S.S. (1994). Chitosan as a novel nasal
mainly in the formation of the HCl salt of N-dimethyl chit- delivery system for peptide drugs. Pharm. Res., 11, 1186- 1189.
osan. The synthesis of N&methyl chitosan chloride with Le Dung, P., Milas, M., Rinaudo, M., & Desbriires, J. (1994). Water
soluble derivatives obtained by controlled chemical modifications of
higher degrees of quaternization requires a two-step reac-
chitosan. Carbohydrate Polymers, 24, 209-214.
tion. With this procedure high degrees of substitution up to LueBen, H.L., De Leeuw, B.J., Langemeijer, M.W.E., De Boer, A.G.,
70% can be obtained. Verhoef, J.C., & Junginger, H.E. (1996). Mucoadhesive polymers in
Furthermore, our results show that great care has to be peroral peptide drug delivery. VI. Carbomer and chitosan improve the
taken when preparing derivatives of chitosan. The presence intestinal absorption of the peptide drug buserelin in vivo. Pharm. Res.,
13, 1668-1672.
of even a small amount of acid, e.g. when used to exchange
LueBen, H.L., Rentel, C.-O., Kotzt, A.F., Lehr, C.-M., De Boer, A.G.,
the counterion or for the dissolution of chitosan, can give Verhoef, J.C., & Junginger, H.E. (1997). Mucoadhesive polymers in
rise to misleading results. peroral peptide drug delivery. IV. Polycarbophil and chitosan are potent
enhancers of peptide transport across intestinal mucosae in vitro. J.
Control. Rel., 45, 15-23.
Acknowledgements Muzzarelli, R.A.A., & Tanfani, F. (1985). The N-permethylation of
chitosan and the preparation of N&methyl chitosan iodide.
Carbohydrate Polymers, 5, 297-307.
This work was financed in part by the State Scholarship
Muzzarelli, R.A.A., Tanfani, F., Emanuelli, M., & Mariotti, S. (1982). N-
Foundation of Greece, Pharmacare Ltd (South Africa), the (Carboxymethylidene)chitosans and N-(carboxymethyl)chitosans:
Foundation for Pharmaceutical Education of the Pharma- novel chelating polyampholytes obtained from chitosan glyoxylate.
ceutical Society of South Africa, and by the Royal Dutch Carbohydrate Res., 107, 199-214.