Alkhamis KA, Obeidat WM, Najib NM. Adsorption of Allopurinol and Ketotifen by Chitosan. AAPS PharmSciTech. 2001; 2(1): article 3.

Adsorption of Allopurinol and Ketotifen by Chitosan
Khouloud A. Alkhamis,¹  Wasfi M. Obeidat,¹  and Naji M. Najib¹ 

¹Jordan University of Science and Technology, Irbid, 22110, Jordan

Correspondence to:
Khouloud A. Alkhamis
Tel: 011-962-2-7095111-23437
Fax: 011-962-2-7095019
Email: khou@just.edu.jo

Submitted: November 28, 2000; Accepted: January 12, 2001; Published: January 26, 2001

Keywords:  Adsorption, Chitosan, Allopurinol, Ketotifen, Particle Size

Abstract

The experimental work of studying the adsorption of ketotifen and allopurinol by chitosan focused on determining the solubilities and the adsorption isotherms of the adsorbates employed in this study. The adsorption of the aforementioned compounds by chitosan was studied using the rotating bottle method. The concentrations, both before and after the attainment of equilibrium, were determined with the aid of a reversed-phase high-performance liquid chromatography column. The results of these studies demonstrated that ketotifen and allopurinol are both adsorbed by chitosan. The nonlinear Langmuir-like and the Freundlich models both were applied to the experimental data. The correlation coefficients obtained from the nonlinear Langmuir-like model were better than those obtained from Freundlich model, suggesting that allopurinol and ketotifen interacted with certain specific binding sites on the chitosan surface. The allopurinol adsorption experiments indicated that the particle size of chitosan and therefore the surface area can significantly affect the Langmuir capacity constant, while the affinity constants are statistically the same. As expected from the solubility studies, the ketotifen adsorption experiments at 2 different pHs (7 and 10) showed that the adsorption affinity at pH 10 was much higher than at pH 7. What was not expected was that the capacity constants were significantly different, suggesting that further studies are needed using common ion buffers and multicomponent adsorption for the proper mechanism to be determined.

Introduction

Recently, natural polymers such as polysaccharides and proteins have received much attention in the pharmaceutical field owing to their good biocompatibility and biodegradability.¹ Among polysaccharides, chitosan, the deacetylated product of chitin, is thought to be one of the most useful natural polymers from the viewpoint of effective utilization of natural resources.²

Chitosan is insoluble at neutral and alkaline pH values, but forms salts with inorganic and organic acids such as hydrochloric acid. Upon dissolution, the amine groups of chitosan are protonated and the resultant polymer is positively charged. Because chitosan exhibits a positive charge, it has been recently introduced to the market as a weight loss aid and cholesterol-lowering agent. The mechanism behind chitosan may be its effect on lipid transport in the gut, where the positively charged chitosan can bind to the free fatty acids and bile salt components and hence disrupt lipid absorption.³

The effect of chitosan has been considered mainly because of its positive charge; however, the adsorption process could also be the result of other forces that might exist between molecules, such as hydrogen bonding or van der Waals forces.⁴ These interactions might have a strong impact on the absorption and bioavailability of pharmaceutical compounds, especially for drugs that are potent and have low water solubility. Therefore, it was of interest to study the interaction of certain pharmaceutical compounds with chitosan at neutral and alkaline pH values. At neutral and alkaline pH values, the amine groups of chitosan are not expected to be protonated, and the role of other forces that might exist between molecules on the solid surface can be investigated.

Allopurinol and ketotifen were selected as model compounds for the following reasons. Firstly, allopurinol is a weak acid that has an acidity constant (pKa) of 10.2,⁵ whereas ketotifen is a weak base that has a pKa of 6.7.⁶ Therefore, allopurinol and ketotifen are expected to be in the unionized form at neutral and alkaline media, respectively. Second, allopurinol and ketotifen exhibit low aqueous solubility at the aforementioned conditions. Last, the presence of a hydroxyl group in allopurinol and a carbonyl group in ketotifen are likely to be important for hydrogen bonding with chitosan.

Materials and Methods

Chitosan polymer (lot numbers F971201-200 and F971201-198, JPM, Amman, Jordan) was washed (for 24 hours each time) with deionized distilled water. After each wash, the chitosan was filtered and dried, and then washed again for a total of 6 times for 24 hours each time. After washing, the samples were analyzed using a spectrophotometer and high-performance liquid chromatography (HPLC); the results showed that the washed chitosan was pure enough to be used. The washed chitosan was then dried and used in the adsorption experiments. The characteristics of the chitosan employed in this study are presented in Table 1.

Chitosan was vacuum dried at 60°C for 24 hours before use using a vacuum oven (Lab-line, Squaroid-Duo-Vac-Oven, Melrose Park, IL). The vacuum oven was connected to an oilless vacuum pump (KNF, model 035 AN.18).

A stock solution was prepared by dissolving ketotifen dihydrogen fumarate (lot number 2790399, Hikma Pharmaceuticals, Amman, Jordan) or allopurinol (lot number 4790498, Hikma Pharmaceuticals, Amman, Jordan) in 500 mL of buffer (phosphate or borate). Aliquots were then removed from the stock solution and diluted to 100 mL using the same buffer. Five milliliters from each dilution were removed and were used as standards for further analysis.

Preliminary studies showed that the extent of adsorption of allopurinol and ketotifen by chitosan increased significantly when the chitosan was hydrated with deionized water. This indicated that chitosan expands in the presence of water molecules; therefore, the dried chitosan was hydrated before use. The time to full hydration was previously determined experimentally in our lab (6 hours), and the chitosan used in the adsorption experiments was fully hydrated.

Chitosan samples (500 mg each) were weighed quickly after removal from the vacuum oven. Each sample was then placed in a screw cap bottle and 25 mL of buffer was added to each sample. The bottles were rotated in a VanKel sustained release apparatus (VanKel Technology Group, VK 7500, model 65-3100, serial no. 6-0411-0399, Cary, NC) for 6 hours (15 rpm and 26°C). Rotation was then stopped and 25 mL of the serial drug dilutions (ketotifen or allopurinol) were added to each bottle. The filled bottles were wrapped with parafilm and the caps were screwed on. The closed bottles were then rotated in the sustained release dissolution apparatus (15 rpm, 26°C) for 1 hour for ketotifen (pH 10) and for 3 hours for allopurinol and ketotifen (pH 7). The time of the adsorption experiment was enough for the equilibrium to be reached. Rotation of the bottles was then stopped with the bottles in an upright position in the water bath; the chitosan was allowed to settle to the bottom of the bottles for 2 hours at 26°C. Aliquots were removed for subsequent dilutions and analysis. The experiments were done at least in triplicate, and statistical analysis was also performed using curve-fitting program (TableCurve 2D V3, Jandel Scientific, San Rafael, CA).

An excess amount of each drug (ketotifen or allopurinol) was added to 10 mL of the specified buffer. The samples were rotated in the sustained release dissolution apparatus for 24 hours (30 rpm, 26°C). Rotation was then stopped and the supernatant was taken. The supernatant was filtered using stainless steel filter holders and Teflon membranes, then was diluted and analyzed using HPLC. The time of the experiment allowed equilibrium to be reached, and the experiments were done in triplicate.

Samples were analyzed using HPLC and employing a reversed-phase system. The HPLC (10A VP) system consisted of a pump, a UV-VIS detector connected to a personal computer, and a system controller (all from Shimadzu Co, Tokyo, Japan). The HPLC conditions that were used in this study are given in Table 2.

Solubility Studies

The solubility of the adsorbate can significantly affect the extent of adsorption; therefore, it was necessary to determine the solubilities of the adsorbates employed in this study, namely allopurinol and ketotifen. The solubilities, the experimental conditions, and the pKa are presented in Table 3.

The results in Table 3 clearly show that the solubility of allopurinol at pH 7 was low. This is expected because allopurinol is a weak acid that is not ionized at pH 7. The results also show that the solubility of ketotifen at pH 7 was much higher than the solubility at pH 10. This was also expected because ketotifen is a weak base (ionization decreases by increasing the pH).

Adsorption of Allopurinol and Ketotifen by Chitosan

Adsorption of allopurinol and ketotifen by chitosan was studied. The concentrations, both before the addition of chitosan and after the attainment of adsorption equilibrium, were determined with the aid of an HPLC system employing a reversed-phase column. The nonlinear Langmuir-like equation was applied to the experimental data. The Langmuir⁷ treatment is summarized by the equation

(1)

Where x is the amount of solute adsorbed, m is the mass of the adsorbent, C_eq is the concentration of the unadsorbed solute at equilibrium, k₁ is the capacity constant, and k₂ is the affinity constant. The derivation of the Langmuir-like equation is dependent upon the following assumptions: the heat of adsorption is independent of surface coverage (all of the sites available for adsorption are energetically equivalent), the adsorbed phase is confined to a monolayer, there are no lateral interactions between adsorbate molecules, the adsorbate solution is very dilute, and there is no mixed film formation at maximum solute adsorption.

The parameters of the Langmuir-like equation for the adsorption of allopurinol by chitosan are presented in Table 4. A typical nonlinear equilibrium adsorption isotherm for allopurinol is presented in Figure 1; the adsorption parameters are presented in Table 4.

The results in Table 4 show that the Langmuir capacity constants for the 2 batches are quite different. This is expected because the particle sizes and therefore the surface areas are not the same. The difference in the adsorption capacity resulting from a difference in particle size indicates that the interaction is on the surface (adsorption) and not absorption into the chitosan matrix. The results in Table 4 also show that the Langmuir affinity constants are quite similar. This is also expected because the strength of the interaction and the equilibrium process should not be affected by the difference in particle sizes.

The good correlation that was obtained using the Langmuir-like equation suggests that allopurinol interacts with certain specific binding sites on the chitosan surface; however, for this result to be confirmed, further studies are needed using solution calorimetry.

The ability of chitosan to act as an adsorbent, in most of the cases, was considered to be mainly the result of ionic interactions; however, the adsorption of allopurinol by chitosan shows that this is not always true. The adsorption of allopurinol by chitosan is likely to be the result of hydrogen bonding between the adsorbate and the adsorbent.

Adsorption of ketotifen by chitosan was also studied at 2 different pHs (7 and 10). The nonlinear Langmuir-like equation was applied to the experimental data. The parameters of the Langmuir-like equation are presented in Table 5. A typical nonlinear equilibrium adsorption isotherm is presented in Figure 2. The results clearly show that the adsorption affinity at pH 10 is much higher than the adsorption affinity at pH 7. This was expected because the fraction ionized and the solubility of ketotifen at pH 10 are much lower than those of ketotifen at pH 7; the higher the solubility, the lower the extent of adsorption and the lower the affinity constant (assuming that the binding sites are the same). The capacity constants were significantly different, however. This was not expected because the binding sites on the chitosan surface at the beginning were assumed to be constant.

The difference in the capacity constants might be the result of several factors. First, the difference in adsorption capacities could be due to a change in the orientation of the ketotifen molecule on the chitosan surface at different pHs (change in the binding site). Second, the buffer that was used at pH 10 might be competing for the same binding sites (boric acid) as the ketotifen molecules. It is not clear at this moment which mechanism is responsible for the difference in the adsorption capacity at different pHs; however, the second explanation is preferred because previous investigators studied the adsorption of boron by chitosan.⁸ Further studies are needed using common ion buffers and multicomponent adsorption in order for the proper mechanism to be determined.

The adsorption of ketotifen (weak base) by chitosan indicates that the interaction cannot be the result of ionic interaction, and that it is likely to be a result of other physical interaction such as hydrogen bonding between the carbonyl group of ketotifen and the hydroxyl group of the chitosan. This conclusion is based on the fact that ketotifen is a weak base, whereas chitosan is either positively charged or neutral at the aforementioned experimental conditions.

The ability of allopurinol and ketotifen to interact with more than 1 binding site was not excluded; therefore, the Freundlich model was also applied to the experimental data. The Freundlich^7,9 equation is

(2)

Where K is a constant related to the capacity of the adsorbent for the adsorbate and P is a constant related to the affinity of the adsorbent for the adsorbate. Although this equation was first employed empirically, it can be derived with the assumption of a continuously varying heat of adsorption. There is no assurance that the derivation of the Freundlich equation is unique; consequently, if data fit the equation, it is only likely, but not proven, that the surface is heterogeneous. The Freundlich model unfortunately predicts both infinite adsorption at infinite concentration and an infinite heat of adsorption at zero coverage.

The parameters of the Freundlich model for the adsorption of allopurinol and ketotifen by chitosan are presented in Tables 4 and 5, respectively; the nonlinear adsorption isotherms are presented in Figures 3 and 4. The correlation obtained using the Freundlich model was less satisfactory than that obtained from the nonlinear Langmuir-like model. This might indicate that chitosan is interacting with certain specific binding sites. However, the most appropriate model should never be selected solely on the basis of statistical¹⁰ comparisons. Selection must await further studies on heats of adsorption because Langmuir-like and Freundlich treatments differ fundamentally with respect to the requirements placed on the differential heat of adsorption as a function of the extent of adsorbent surface coverage.¹¹

First, allopurinol and ketotifen are both adsorbed by chitosan, and the interaction is likely to be the result of hydrogen bonding and not ionic interactions. Second, the particle size and surface area of chitosan can significantly affect the adsorption capacity of chitosan, while the adsorption affinity remains statistically the same. Last, the pH and the ionization of ketotifen can significantly affect the capacity and the affinity constants of the Langmuir-like equation.

Acknowledgements

The authors would like to thank Professor Dale Eric Wurster for his help.

1.   Henriksen I, Skaugrud Ø, Karlsen J. Use of chitosan and chitosan malate as an excipient in wet granulation of three water soluble drugs. Int J Pharmaceutics. 1993;98:181-188.
 

2.   Imai T, Shiraishi S, Saito H, Otagiri M. Interaction of indomethacin with low molecular weight chitosan, and improvements of some pharmaceutical properties of indomethacin by low molecular weight chitosans. Int J Pharmaceutics. 1991;67:11-20.
 

3.   Illum L. Chitosan and its use as a pharmaceutical excipient. Pharm Res. 1998;15(9):1326-1331.
PubMed  

4.   Parfitt GD, Rochester CH. "Adsorption of Small Molecules" In: Adsorption from Solution at the Solid/Liquid Interface. Academic Press, Inc.: Orlando, FL; 1983:9-13.
 

5.   The Merck Index. 12th ed. Whitehouse Station, NJ: Merck and Co., Inc.; 1996.
 

6.   Lee Y, Chiang C, Chen J. In vitro and in vivo percutaneous absorption studies of ketotifen patches. Drug Dev Ind Pharm. 1994;20(19):2965-2976.
 

7.   Kipling JJ. "Adsorption of Solids from Solution" and "Adsorption at the Liquid-solid Interface" In: Adsorption from Solutions of Non-electrolytes. Academic Press, Inc.: London; 1965.
 

8.   Matsumoto M, Matsui T, Kondo K. Adsorption mechanism of boric acid on chitosan resin modified by saccharides. J Chem Eng Japan. 1999; 32(2):190-196.
 

9.   Adamson AW. "The Solid-Liquid Interface" In: Physical Chemistry of Surfaces. edited by: ed. Adamson AW and Gast AP; 6th ed. John Wiley and Sons, Inc.: New York; 1997:394.
 

10.   Wurster DE, Burke GM, Berg MJ, Veng-Pedersen P, Schottelius DD. Phenobarbital adsorption from simulated intestinal fluid, USP, and simulated gastric fluid, USP, by 2 activated charcoals. Pharm Res. 1988;5(3):183-186.
PubMed  

11.   Burke GM, Wurster DE, Buraphacheep V, Berg MJ, Veng-Pedersen P, Schottelius DD. Model selection for the adsorption of phenobarbital by activated carbon. Pharm Res. 1991;8(2):228-231.
PubMed