Characterization of Cellulose Acetate Phthalate (CAP)

Drug Development and Industrial Pharmacy, 24(1 I), 1025-1041 (1998)

Drug Development and Industrial Pharmacy Downloaded from informahealthcare.com by Pfizer Ltd (Active) on 04/21/10 For personal use only.

RESEARCHPAPER

Characterization of Cellulose Acetate Phthalate (CAP) Pernilla Roxin,’ Anders Karlsson,’ Satish K. Singh2s*

’ Dept. of Pharmaceutical Analytical Chemistry, Pharmacia and Upjohn AB, S-751 82 Uppsala, Sweden 2Dept. of Pharmaceutical Technology, Pharmacia and Upjohn AB, 5-751 82 Uppsala, Sweden

ABSTRACT Cellulose acetate phthalute (CAP) is a commonly used enteric coating polymer. CAP powder has been studied by various methods to determine characteristics that have an injuence on its functionality. While some of the parameters are well known, such asfree-acid content and substituent composition, new methods have been developed to examine them. Other characteristics, such as the molecular mass distribution, have not been reported earlier. Fourier transform infrared spectroscopy (FTIR), nuclear magnetic resonance (NMR), and thermal analysis have also been performed on fresh samples, as well as samples stored under various temperature und humidity conditions. Humidity is by far a more critical storage parameter than temperature, although the two act in conjunction; high humidity is more deleterious to the functionality qf the polymer than high temperature. Functionality in this case is taken to be determined by the substituents and by the molecular mass distribution. Mass-average molecular nmss of a number of batches of the polymer hus been measured and ranges around 48 kg/mol with a degree of polydispersity of1.6. A method to perj?orma rough estimation of the molecular mass of CAP has also been suggested based on knowledge of the substituent content. It may be possible to use the values of and obtained herefiw any other batch of the same viscosity grade of CAP. NMR has been employed to determine the fraction substituents in the polymer. However, an attempt to obtain the pattern of substitution of the CAP molecule by NMR was unsuccessful. Glass transition temperatures of CAP samples were measured. However, this characteristic of the polymer is judged not as sensitive to the loss of substituents as the molecular mass. Thermal treatment of the polymer in oxygen and inert atmospheres gave slightly dtfferent degradation products.

To whom correspondence should be addressed. Fax: +46-18-16 63 36. E-mail: [email protected] 102s

Copyright 0 1998 by Marcel Dekker, Inc.

www.dekker.com

Roxin, Karlsson, and Singh

1026

Key Words: Cellulose acetate phthalate; Humidity; Free acid; Substituent content; Phthalyl content; Acetyl content; Molecular m s s distribution; SEC-MALLS; Thermal degradation; Glass transition temperature.


INTRODUCTION Cellulose acetate phthalate is a commonly used tablet coating material employed to produce so-called enteric films, which resist prolonged contact with the strongly acidic gastric fluid, but soften, swell, and finally dissolve in the mildly acidic or neutral intestinal environment. This function of the polymer depends on a number of factors, such as structure, chemistry, molecular mass distribution, thermal behavior, and stability, along with the process (concentration, solvent/dispersion system, temperature, flow rate, pressure, humidity, etc.) and formulation (plasticizer, other additives, active substance, etc.) variables (1). In this work, we report an examination of some of the polymer characteristics, including the effect of storage. While a number of these characteristics have been studied earlier, others (such as the molecular mass distribution) have not been reported. Some new methods have also been developed to enable a more rapid examination of these characteristics than that allowed by the pharmacopoeia methods, for instance. A search of the literature on CAP returns 977 listings in CAplus (1 967- 1997). However, the literature is very scattered, and it is difficult to get an overview of the behavior of this polymer. A number of studies have treated the film of the polymer (free film or as a tablet coating), but few studies can be found on the polymer itself. The most comprehensive study we could find is by Delporte (2) on the physicochemical properties of CAP films on storage.

MATERIALS CAP was obtained from Eastman Chemical Company (Kingsport, TN), batches 50103 (51062-Ol), 50105 (51063-Ol), 50106 (51064-Ol), 50104 (41238-Ol), and 40706 (4 1240-01 ). These batches were received over a period of 10 months, between January 1995 and October 1995. All other reagents were used as received.

METHODS Water Content The water content of the polymer is used to correct the calculated free acids, phthalic, and acetic content and will also influence the measured viscosity.

Loss on Drying

Drying to constant weight at 105°C was performed on one batch of CAP. Karl Fischer Titration Karl Fischer titration was performed with a Metrohm 665 Dosimat and a Metrohm 682 Titroprocessor (Metrohm Ltd., Merisau, Switzerland). Titrator KF solution A (Merck 9247) and KF solution B (Merck 9246) (1 : 1) were used as the reagent, and MeOH and pyridine (1 : 1) were the solvents. Samples of 10-20 mg were analyzed. Karl Fischer analysis was also performed on a Blending Volumetric KF Titrator, Turbo 2 from Orion (Boston, MA). The reagent used was Hydranal Composite 5 (Riedel de Haen).

Measurement of Free Acids Free-acid (as phthalic acid) content is included in the pharmacopoeia specifications for CAP. While the USP-NF specifies extraction in MeOH :H20 50 :SO for 2 hr, the Ph. Eur. specifies extraction in water for 5 min only. Total content of acids is determined by titration with 0.1 M NaOH to a phenolphthalein end point. Bauer (3) states that extraction in water alone is less efficient than in the MeOH/H20 mixture, and that the extraction requires at least 30 min to reach equilibrium. Extraction of Free Acids (Phthalic and Acetic Acid) A water-based extraction procedure was examined. Free acetic acid was included in the test. An exact weight of CAP (approximately 10 mg) was suspended in 5 rnl of C02-freewater and shaken for various lengths of time, ranging from 10 to 120 min at room temperature. Samples were then centrifuged (10 min, 3000 rpm) and filtered (0.45 pm) prior to analysis. Recovery was tested by extracting a spiked CAP sample (CAP + 500 p1 spiking solution = 18.1 pg/ml phthalic acid, 654.4 pg/ml acetic acid) and comparing with extraction from the same volume of the spiking solution itself. From the results of chromatographic analysis (data not shown), it was found that equilibrium was not reached at 120 min in all cases, in agreement with the comments by Bauer (3). Measurements showed not only high recov-

Characterization of CAP eries (-loo%), but also large standard deviations (-30% and 10% for phthalic and acetic acids, respectively) at an extraction time of 30 min. A new extraction procedure was developed based on CAP precipitation. An exact weight of CAP (approximately 10 mg) was dissolved in 250 p.1 THF with stirring. A total of 5 ml of water was added dropwise with vigorous stirring, allowing the polymer to precipitate out. The samples were centrifuged and the supernatant filtered before analysis by chromatography. Recovery was checked with spiked samples as described above. The results of the chromatographic analysis showed high recovery (-100%) and lower standard deviations ( < I % for phthalic acid and -10% for acetic acid at quantification limit). The relative standard deviation (RSD) of the process was 4.1% for phthalic acid and 4.4% for acetic acid. This extraction procedure was then used in all subsequent tests.


-

Chromatographic Analysis of Free Acids Bodmeier and Chen (4) have developed an isocratic reversed-phase high-performance liquid Chromatography (RP-HPLC) method to quantify the organic acids (acetic, propionic, butyric, and phthalic acids) formed as a result of ester hydrolysis in aqueous pseudolatexes of cellulosic esters. A Beckman-Ultrasphere CIS,5 pm, 250 X 4.6 mm column was used with a mobile phase consisting of 0.025 M phosphate buffer-methanol (80: 20 v/v, pH 3.0), with detection at 210 nm. Linear response was obtained over the studied range of 2- 10 mM for the aliphatic acids and 0.02-0.1 mM for phthalic acid. Minimum detectable concentrations were 0.02 mM, 0.05 mM, 0.1 mM, and 0.0005 mM for acetic, propionic, butyric, and phthalic acids, respectively. When applied in our laboratory, the best results were obtained with a Waters C18Nova-Pak@,150 X 3.9 mm column using 10%MeOH in 0.025 mM phosphate buffer of pH 3 at 1 ml/min, injection volume 20 pl. (Other columns tested were HiChrom C18200 X 4 mm; CISRP-18 SuperspheP 100, 125 X 4 mm; C,8 Endcapped RP-18 Superspher@100, 125 X 4 mm). Retention time for acetic acid was 1.7 min, compared to 8.4 min for phthalic acid. These correspond fairly well with retention times obtained by Bodmeier and Chen (4) with a 20% MeOH eluent. The acetic acid was eluting close to the void in our system, which could only be improved at the cost of total analysis time; increasing retention time for acetic acid slightly caused a large increase in phthalic acid elution time. While gradient elution could be used to get around the retention time problem, the need for equilibration times between runs made this solution im-

1027

practical. Instead of further optimizing the HPLC method, an ion exclusion chromatography (IEC) method was developed. The IEC method was run on a Dionex Ion Pac@ HPICE@AS1 column, 250 X 9 mm, using a 2% acetonitrile in 0.5 mM H2S04eluent. Phthalic acid gives a broad peak in this system. Increasing the acid concentration from 0.1 to 0.5 mM reversed the elution order of acetic and phthalic acid, with phthalic acid eluting later. Acetonitrile in the mobile phase helped to reduce the hydrophobic interactions between this acid and the column and reduce peak width. Best results were obtained with 2% CH,CN, above which resolution is affected negatively. Final retention times were 9.3 rnin for acetic acid and 10.6 min for phthalic acid. Formic acid eluted at 8.2 min in this system. 10% THF in water was also injected into the system to ensure that elution of THF did not interfere with any of the acids. This IEC method was then used to analyze the above precipitated and extracted samples. Limits of detection were 2 pg/ml (0.04 mM) and 0.05 pg/ml (0.0003 mM) for acetic and phthalic acid, respectively. The linear detection range studied extended to I . 1 mM for acetic acid and to 0.15 mM for phthalic acid. The relative standard deviation for acetic acid determination was 1 .O% and for phthalic acid was 1.O% at 50 pg/ ml.

Total Phthalic and Acetic Content The USP/NF method for phthalyl content specifies dissolution of CAP in a mixture of EtOH :CH2C12,addition of EtOH and phenolphthalein, and titration with 0.1 M NaOH. For acetyl content, the sample is hydrolyzed in a mixture of water and 0.5 N NaOH, refluxed for 60 min, and titrated with 0.5 M HCl. The Ph. Eur. method for phthalyl content involves dissolution of CAP in ethylene glycol monomethylether and subsequent titration with 0.1 M NaOH. The acetyl content method is essentially similar to the USP/NF method except for a shorter reflux time (30 min) and the use of 0.1 M NaOH in the hydrolysis. Bauer (3) recommends elimination of ethylene glycol monomethylether since it has a hazard rating similar to ethylene glycol monoethylether, which is a suspected carcinogen. Hydrolysis of Phthalic and Acetic Groups A method more suitable for rapid analysis of phthalic and acetic groups was developed based on the total hydrolytic release of these groups as free acids by NaOH. A factorial experimental design was used to optimize the hydrolysis procedure with respect to temperature, con-

Roxin, Karlsson, and Singh

I028


ccntration of NaOH, and time. The measured levels of acetic acid and phthalic acid were found to reach a steady level after 2 hr at 105°C in 2 M NaOH. The final method involved dissolution of an exact weight (approximately 10 mg) of CAP in 2 in1 of 2 M NaOH. The solution was then placed in an oven at 105°C for 2 hr, neutralized with 3 ml of 1 M H,SO, dropwise with stirring, subjected to filtration (0.45 pm), and analyzed by the IEC method above.

Molecular Mass Distribution CAP is supplied along with a viscosity specification. Viscosity blending is often used by manufacturers to obtain the desired in-specification viscosity. Since the same viscosity can be obtained using different blends, a better measure of the actual chain characteristics of the polymer is its molecular mass distribution (MMD). No reports on the molecular mass determination of CAP were found in the literature, although there are several articles dealing with MMD analysis of cellulose derivatives by size exclusion chromatography (SEC) (see, e.g., Refs. 5 and 6). We devcloped a system based on SEC-MALLS-RI to perform an absolute characterization and another SECUV-RI system more suitable for routine use. Normally, dual detection is not required when running traditional SEC-UV or SEC-RI, but one of the aims of this method dcvelopment was to compare and contrast light-scattering detection with traditional SEC and also compare UV (ultraviolet) detection with RI detection. Size Exclusion Chromatography Molecular-volume-based separation was carried out on two PLgel 5 pm Mixed-C 300 X 7.5 mm columns (Polymer Labs Ltd., Shropshire, UK) connected in series. The columns are rated for separation in the range 0.23000 kg/mol (polystyrene). The SEC-UV-RI system was equipped with a Waters System Interface Module, Waters 616pump( I ml/min), Waters712WISP(injectionvolume 50 pl), Waters 490 multiwavelength detector (280 nm), and a Waters 4 10 differential refractometer. The SECMALLS-RI system was set up similarly, except that the U V detector was replaced by a DAWN niultiangle laser light scattering (MALLS) detector (Wyatt Technology Corp., Santa Barbara, CA). Both systems were equipped with a Millenium 2010 Chromatography Data System. The columns were calibrated with polystyrene standards ( I 0 standards in the range 0.5 -3000 kg/mol; Polymer Labs Ltd., Shropshire, UK) and saturated with CAP using a 0.5% w/w solution of CAP in THF in order to prevent further unspecific adsorption of polymer to the

column during the analyses. Samples for measurement were also dissolved in THF at a concentration of 0.5% w/w. All samples were filtered through a 0.45-pm filter prior to injection. The number-average molecular mass , massaverage molecular mass