ANALYST, FEBRUARY 1986, VOL. 111 133 Stability-indicating Assay for Oxyphenbutazone Part It.* High-performance Liquid Chromatographic Determination of Oxyphenbutazone and Its Degradation Products Huguette Fabre,t Andrianandrasana Ramiaramana, Marie-Dominique Blancslin and Bernadette Mandrou Laboratoire de Chimie Analytique, Faculte de Pharmacie, 34060 Montpellier Cedex, France A high-performance liquid chromatographic method is proposed for the simultaneous determination of oxyphenbutazone and six potential decomposition products, using a reversed-phase column and ultraviolet detection. The method is more sensitive than thin-layer chromatography and allows the determination of 0.1% of each degradation product (with respect to oxyphenbutazone). It has been applied to the analysis of commercial tablets, capsules and ointments.Keywords: Oxyphenbutazone determination; degradation products determination; high-performance liquid chromatography; stability-indicating assay; reversed-p hase chromatography In Part I,1 we outlined the difficulties relating to the establishment of a thin-layer chromatographic (TLC) method to be used as a stability-indicating assay of oxyphenbutazone. We proposed a quantitative thin-layer chromatographic pro- cedure that prevents air oxidation on the plate of oxyphen- butazone. As reversed-phase high-performance liquid chromato- graphy (HPLC) is particularly suitable for the determination of easily oxidised compounds, we propose here a reversed- phase HPLC procedure for separating and determining oxyphenbutazone and six potential decomposition products.Method Development Oxyphenbutazone and its potential decomposition products were chromatographed under different conditions in order to optimise the separation. The decomposition products, the formulae of which were given in Part I,1 were 4-hydroxy-4- butyl-l-phenyl-2-(4-hydroxy)phenylpyrazolidine-3,5-dione (I), 2-butyl-N-(4-hydroxyphenyl)-N’-phenylpropanediamide (11), 2-[1-phenyl-2-(4-hydroxy)phenylhydrazino]-3-0~0-2- butylpropionic acid (111), 4-hydroperoxy-4-butyl-l-phenyl-2- (4-hydroxyphenyl)pyrazolidine-3,5-dione (IV) , 2-butyl-N- { 3- [4-butyl-3,5-dioxo- 1-(4-hydroxyphenyl)-2-phenylpyrazolidin- 4-yl]-4-hydroxyphenyl}-N’-phenylpropanediamide (V) and 2-oxo-3-butyl-3-phenylcarbamoyl-5-hydroxyindole (VI). The starting solvent system was the mobile phase we previously used in the stability-indicating assay of phenylbutazone, viz., 0.1 M tromethamine (THAM) citrate buffer (pH 5.25) - acetonitrile (60 + 40).2 As this mobile phase could not achieve the separation of the compounds, the influence of pH and acetonitrile content of the mobile phase on the capacity factor was investigated.The method was developed using a mixed standard solution (5 pg ml-1) of each compound (Fig. 1). The capacity factor, k’, was calculated using the equation k’ = (u/L) (fR + I), where tR is the retention time of the compound, u the linear flow-rate of the mobile phase given by the supplier and L the length of the column. As a stability-indicating assay of oxyphenbutazone in drugs involves the determination of less than 1% of each decompo- sition product (relative to the drug), the resolution factor (R.F.) between oxyphenbutazone and VI (pH 4.1) and oxyphenbutazone and I (pH 5.25 and 6.4) was calculated using * For Part I of this series, see reference 1.t To whom correspondence should be addressed. the equation R.F. = 2(tR2 - tR1)/(W2 + w1), where t~~ and fR2 are the retention times and w1 and w2 the peak widths. The influence of the acetonitrile content of the mobile phase on the resolution factor is shown in Fig. 2. The pH values 5.25 and 6.4 gave a satisfactory resolution (R.F. > 2 with equal concentrations of each compound) but pH 6.4 was discarded because of the elution of compound I11 [Fig. l(a)] in the solvent peak (the mobile phase cannot be used as the solvent because of the instability of the compounds in the solvent system).Therefore, the mobile phase 0.1 M THAM citrate buffer (pH 5.25) - acetonitrile (65 + 35) was selected in further development of the method. In order to increase the resolution, the effect of the addition of tetrahy- drofuran (THF) (2-6%) on this mobile phase was investi- gated. The addition of 6% of THF increased the selectivity and gave a resolution factor R.F. = 1.32 between oxyphenbu- tazone and I using a concentration of 500 pg ml-1 of oxyphenbutazone and 5 pg ml-1 of I. The mobile phase A finally selected was 0.1 M THAM citrate buffer (pH5.25) - acetonitrile - THF (65 + 29 + 6). For the determination of V a more strongly eluting mobile phase B, 0.1 M THAM citrate buffer (pH 5.25) - acetonitrile (45 + 59, was selected from Fig.l(b). Experimental Apparatus A high-performance liquid chromatograph (Merck LMC System), equipped with a variable-wavelength ultraviolet detector (LC 313), a 10-p1 loop injector and fitted with a 25 x 0.4 cm i.d. stainless-steel cartridge, packed with 7 pm LiChrosorb RP-18 (Merck), was used. The column was equilibrated with the mobile phase for 30 min before use. Reagents and Materials Tromethamine and citric acid were of analytical-reagent grade. Oxyphenbutazone and its degradation products (I-VI) were the same as used previously.’ Tanderil ointment (5% oxyphenbutazone), Tanderil tablets (100 mg of oxyphenbutazone per tablet) were commercial formulations (Ciba-Geigy Laboratories). Kymalzone capsules (75 mg of oxyphenbutazone per tablet) were commercial formulations (Biocodex Laboratories).ANALYST, FEBRUARY 1986, VOL. 111 'bj 35 40 45 50 55 I C) \ 35 40 45 50 - 55 Ec------p-- Acetonitri le,% Fig. 1.Influence of the acetonitrile content of the mobile phase (0.1 M THAM citrate buffer - acetonitrile) on the capacity factors (k') of (0) oxyphenbutazone and its degradation products: (0) I, (A) 11, (D) 111, (A), IV, (+) V and (0) VI. (a) Buffer pH 4.10; ( b ) buffer pH 5.25; and ( c ) buffer pH 6.40 12 10 8 L w 0 m C + .s 6 - :: a 4 1 fn C 0 Q al U - (a L 0 Fig. 3. I 1 8 12 16 20 24 Time/m in Separation of oxyphenbutazone and potential degradation products under the optimised conditions. (a) Mobile phase A, 0.1 M THAM citrate buffer (pH 5.25) - acetonitrile - THF (65 + 29 + 6); flow-rate, 1.3 ml min-1; pressure, 98 bar; detector sensitivity, 0.02 a.u.f.s.; chart recorder speed, 0.5 cm min-'; wavelength, 239 nm; oxyphenbutazone, 1.15 pg ml-1; I, 1.30 pg ml-l; 11, 1.75 pg ml-I; 111, 1.37 pg ml-1; IV, 1.47 pg ml-1 and VI, 1.30 pg ml-1.( b ) Mobile phase - _ _ Fig. 2. Resolution factor versm the acetonitrile content of the mobile phase (0.1 M THAM citrate buffer - acetonitrile): (A) between oxvDhenbutazone and VI, buffer pH 4.10; (H) between B, 0.1 M THAM &rate buffer (PH 5.25) - acetonitrile (45 + 55); flow-rate, 1.8 ml min-l; pressure, 105 bar; detector sensitivity, 0.02 a.u.f.s.; chart recorder speed, 0.5 cm min-1; wavelength, 239 nm; oxyphenbutazone and 1, butter pH 5.25; and (0) between oxyphen- butazone and I, buffer pH 6.40 Table 1.Chromatographic parameters for oxyphenbutazone and its decomposition products oxypnenoutazone 1.17 pg ml-1; IV, 1.22 pg ml-1; V, 1.37 pg ml-I; VI, 1.67 pg ml-1 pg mIPL; 1, 1.w pg mlr L; 11, 1.J.L pg 1111 *; 111, Mobile phase A Mobile phase B Compound k' Hlmm As k' Hlmm As 1.16 0.35 0.067 - I . . . . . . 4.11 0.077 1 .oo 0.43 0.062 1 .oo I1 . . . . . . 11.32 0.040 1.16 1.08 0.042 1.20 I11 . . . . . . 0.74 0.166 1 .oo 0.00 IV . . . . . . 6.79 0.052 1.25 0.64 0.046 1 .oo v * . . . . . 3.26 0.043 1 .oo VI . . . . . . 8.79 0.047 1.40 0.88 0.043 1.30 Ox.* . . . . 2.63 0.460 - * OxyphenbutazoneANALYST, FEBRUARY 1986, VOL. 111 135 Table 2. Repeatability, sensitivity and detectability data for oxyphenbutazone and its decomposition products Repeatability, Sensitivity/ Detectability*/ Compound a,% mm pg-1 Ox.? .. _. . 0.72 2 435 0.0040 I . . . . . . 0.75 6 615 0.0015 I1 . . . . . . 1.26 3 314 0.0030 I11 . . . . . . 0.68 11 387 0.0010 IV 1.18 2 041 0.0050 V . . . . . . 1.59 7 299 0.0010 VI . . . . . . 0.89 3 231 0.0030 * Determined in the presence of oxyphenbutazone (5 pg). t Oxyphenbutazone. . . . . . . Table 3. Recovery of oxyphenbutazone and its degradation products from pharmaceutical formulations Formula- Initial Amount Amount tion Compound contentjmg added/mg found/mg Ointment. . . .Ox.* I I1 I11 IV V VI I I1 I11 IV V VI I I1 111 IV V VI * Oxyphenbutazone Tablet . . Ox.* Capsule . . Ox.* 50.200 0.034 0.009 Traces - - - 94.000 0.394 0.014 Traces Traces 78.750 0.065 - - - - Traces 0.051 - - 0.131 0.176 0.138 0.148 0.138 0.131 0.259 0.385 0.274 0.294 0.326 0.259 0.195 0.263 0.206 0.221 0.210 0.195 - - - 0.161 0.172 0.142 0.003 0.141 0.132 0.678 0.375 0.296 0.363 0.325 0.268 0.264 0.263 0.212 0.216 0.218 0.244 - - t s O n 0) a I X U 1 I I 4 8 12 16 20 24 Methanol, acetonitrile and THF were of HPLC solvent grade.Distilled water was filtered through a 0.45-pm filter (Millipore). Standard Solutions A mixed stock standard solution containing 250 pg ml-1 each of oxyphenbutazone and I-VI was prepared in methanol. This solution was suitably diluted with methanol to give a concen- tration range from 1 to 10 pg ml-1. Test Solutions Ointment An accurately weighed amount of about 250 mg of ointment was sonicated for 5 min with 25 ml of methanol in a 50-ml centrifuge tube. The emulsion was rotated at 4000 rev min-1 for 10 min.Tablets The coatings of ten tablets were carefully removed with a cutter and the average mass of one core was determined. The ten cores were combined and powdered in a mortar. A core mass of about 20 mg was accurately weighed into a 50-mi centrifuge tube, then sonicated for 5 min with 25 ml of methanol. The suspension was rotated at 4000 rev min-1 for 10 min. Capsules An accurately weighed amount of about 34 mg of capsule powder was sonicated for 5 min with 25 ml of methanol in a 50-ml centrifuge tube. The suspension was rotated at 4000 rev min-1 for 10 min. After centrifugation, the supernatant from the ointment, tablets or capsules was injected on to the chromatograph without dilution for the determination of the degradation products, then diluted (1 + 399) in methanol for the determination of oxyphenbutazone.Chromatography Duplicate injections (10 pl) of each standard solution and test solution were injected under the following isocratic condi- tions: mobile phase A, 0.1 M THAM citrate buffer (pH 5.25) - acetonitrile - THF (65 + 29 + 6); flow-rate 1.3 ml min-1; Time/min Fig. 4. Chromatograms of a test solution from a tablet formulation. ( a ) Without addition of degradation products; mobile phase A. ( b ) Spiked with I-IV and VI; mobile phase A. ( c ) Without addition of degradation products; mobile phase B. ( d ) Spiked with I-VI; mobile phase B. U, unidentified decomposition product136 ANALYST, FEBRUARY 1986, VOL. 111 t m c 0 Q Q, U 0 w 4 8 12 16 20 24 I I , I I I 4 8 12 16 20 24 Tirnehnin Fig.5. Chromatograms of a test solution from capsule formulation. Chromatograms (a)-(d) as in Fig. 4 )X U I I I 1 I I 4 8 12 16 20 24 ox I V I 4 8 0 4 8 Ti rn e/rn in Fig. 6. Chromatograms of a test solution from an ointment. Chromatograms (a)-(d) as Fig. 4 pressure, 95 5 0.1 bar; chart recorder speed, 0.5 cm min-l; detector sensitivity, 0.04 or 0.02 a.u.f.s.; detection wavelength, 239 nm; mobile phase B, 0.1 M THAM citrate buffer (pH 5.25) - acetonitrile (45 + 55); conditions as above except flow-rate, 1.8 ml min-1; pressure, 105 k 0.1 bar. Results and Discussion Specimen chromatograms of a standard solution recorded at 239 nm (a suitable wavelength for the simultaneous determi- nation of all the compounds), using mobile phases A and B, are given in Fig.3(a) and (b), respectively. Table 1 gives the chromatographic data for a mixed standard solution of oxyphenbutazone and its decomposition products (10 pg ml-1) expressed as the capacity factor, k ’ , the theoretical plate height, H , and the asymmetry factor, As, under the conditions used in this study. As, was calculated using the equation As = b/a, where b , is the distance after the peak maximum and a the distance before the peak maximum, both being measured at 10% of the total peak height. The stability of a solution of oxyphenbutazone and its decomposition products was tested by separately injecting on to the chromatograph, at different time intervals, a solution of oxyphenbutazone (500 pg ml-1) and I-VI (10 pg ml-l) in methanol.After 4 h at ambient temperature under diffused light, no decomposition was observed for oxyphenbutazoneANALYST, FEBRUARY 1986, VOL. 111 137 and I-VI within the limit of sensitivity of the method (at 0.02 a.u.f.s.). The stability of test solutions of the ointment, tablet core and capsule formulation (equivalent to 500 yg ml-1) was also investigated. These solutions can be kept for 6 h without any detectable decomposition. Validation of the HPLC Procedure The linearity of the response was examined by plotting the peak-height measurement for each solute against solute concentration in the range &lo0 yg ml-1 for each compound. The calibration graph was rectilinear and passed through the origin in all instances. The correlation coefficient of the linear regression analysis was higher than 0.999 for each compound.The repeatability, assessed by five replicate analyses of a mixed standard solution (10 pg ml-1) and expressed as the coefficient of variation, is shown in Table 2. The sensitivity, defined as the change in the peak height (mm) measured at the maximum detector sensitivity resulting from a concentration change of one unit (pg), is also given in Table 2, together with the detectability, defined as the amount of compound that yields a signal to noise ratio of 2. The limit of determination can be evaluated as about three times the detectability. Commercial formulations were spiked with known amounts of I-VI (0.25% of each with respect to the theoretical oxyphenbutazone content) and these spiked formulations were treated as indicated in the method. The chromatograms obtained are shown in Figs.4-6 and average results of duplicate injections are given in Table 3. Satisfactory results were obtained except for IV in the ointment, probably because a physical and/or a chemical interaction took place. Attempts to solve this problem using different extraction solvents in the procedure (acetonitrile, the mobile phase solvent system, aqueous alkaline solutions) were unsuccessful. In addition , oxyphenbutazone was very unstable in these solvents. Inert interference corresponding to a check on the placebo effect was carried out by treating a placebo of tablets, capsule formulation and the ointment as required in the method. No interference was observed. Conclusions The proposed procedure allows the detection and determi- nation of the potential degradation products of oxyphen- butazone at trace levels. The method is more sensitive than TLC,1 as less than 0.1% of the decomposition products (with respect to oxyphenbutazone) can be quantified. In addition, the sample preparation is simple and rapid and allows the method to be used easily for routine control purposes. References 1. Fabre, H., Ramiaramanana, A., Blanchin, M. D., and Mandrou, B., Analyst, 1985, 110, 1289. 2. Fabre, H., Mandrou, B., and Eddine, H.,J. Pharm. Sci., 1982, 71, 120. Note-Reference 1 is to Part I of this series. Paper A51267 Received July 22nd, 1985 Accepted September loth, 1985