Analyst October 1983 Vol. 108 pp. 1221-1226 1221 Electrochemical Studies on Minoxidil and its Determination in Tablets by Differential-pulse Pola rog rap hy Lawrence Amankwa Leslie G. Chatten and Stanley Pons* Faculty of Pharmacy and Pharmaceutical Sciences University of Alberta Edmonton Alberta T6G 2N8, Canada A simple differential-pulse polarographic method has been developed for the determination of minoxidil in pharmaceutical dosage forms. The extracting solvent was methanol and the supporting electrolyte was 1.0 N sulphuric acid. An excellent linear relationship was obtained between the concentration and current peak height with a correlation coefficient of 0.999 9. Good agreement was obtained between results with the differential-pulse polarography method and those by the manufacturer’s method of assay.There was no interference by the tablet excipients. In acid solution a mechanism for reduction a t - 1.2 V is proposed which involves the transfer of four electrons as well as dehydration and deamination steps. Keywords Minoxidil determination ; diflerential-pulse polarography ; con-trolled-potential coulometry ; cyclic voltammetry Minoxidil (2,4-diamino-6-piperidinopyrimidine-3-oxide) (I) is a recently marketed orally ad-ministered peripheral vasodialator which is useful for the treatment of patients with refractory hypertension.1 The detection and determination of this substance continues to be of interest, particularly because of its associated numerous side effects.2 As yet no official method for the determination of minoxidil has been listed in the US Pharmacopeia3 and the few reported methods of assay involve radioimmunoassay,4 thin-layer chromatography5p6 or radiochromato-gram scanning.’ Investigation with a gas - liquid chromatographic procedure revealed that substantial amounts of some derivatives were lost on the c01umn.~ H2NTyNH2 In addition to its basic properties minoxidil has two structural features that could be used for analytical purposes (earlier reports8-1° have shown that both the N-0 bond and the carbon-nitrogen double bond are electroreducible) .Accordingly in this paper we present a diff erential-pulse polarographic procedure for the determination of minoxidil in tablets which involves only a single extraction prior to the electroreduction. The method is sensitive, accurate and easy to perform for routine analysis.Experimental Apparatus and Conditions for Polarographic Analysis to measure the pH values of the solutions. A Fisher Model 320 pH meter fitted with a glass - calomel electrode system was employed * Department of Chemistry 1222 Analyst VoZ. 108 A PAR Model 174 polarographic analyser equipped with a drop timer Model 172A and a Houston Ominigraphic Model 2000 recorder were used in the investigations. A three-electrode combination was employed which consisted of a saturated calomel electrode a dropping-mercury electrode and a platinum wire as the auxiliary electrode. A conventional H-type cell was maintained at 25 & 1 "C and all sweeps utilised a scan rate of 2 mV s-l and a drop time of 2 s. In 1.0 N sulphuric acid (pH rn 0.5) the instrumental parameters were applied potential range -0.6 to -2.1 V; current 100 pA full scale; height of mercury column 75 cm; flow-rate of mercury 1.176 mg s-l; modulation amplitude set at 50 mV; and low pass filter set at a time constant of 1 s.Controlled-potential Coulometry A PAR Model 173 potentiostat - galvanostat equipped with a PAR Model 377A three-component coulometric cell system was connected to a Hi-Tek digital integrator and digital voltmeter. A 19-ml volume of 1.0 N sulphuric acid was placed in the coulometric cell on top of a 5-ml layer of triply distilled mercury and 1 ml of M solution of minoxidil in methanol was added. The system was purged for 10 min with purified nitrogen. The applied potential was set at -1.2 V with a current range of 10 pA full scale and the solution was electrolysed until the digital readout indicated a constant but small count.One hour was required to complete the electrolysis. The process was repeated with a blank consisting of 19 ml of 1 .O N sulphuric acid and 1 ml of methanol. Cyclic Voltammetry Cyclic voltammetric experiments at a hanging mercury drop electrode were performed with a four-component system consisting of a PAR EG and G Model 175 Universal Programmer a PAR Model 173 potentiostat - galvanostat a Houston Model 2000 Omnigraphic recorder and a PAR Model 9323 hanging mercury drop electrode fitted with a polarographic cell. Two supporting electrolyte systems were employed. In 1 .O N sulphuric acid the instrumental parameters were as follows potential range -0.8 to -1.3 V; current range 10 PA; and scan rate varied from 10 to 200 mV s-1.In a dimethylformamide - tetraethylammonium bromide system the following settings were used potential range -1.2 to -2.2 V; current range, 10 pA; and scan rate as in the previous system. Reagents The following reagents were used all of analytical-reagent grade barbitone boric acid, citric acid potassium dihydrogen orthophosphate dimethylformamide (DMF) anhydrous methanol tetraethylammonium bromide (TEAB) 0.2 N sodium hydroxide solution 1 .O N sulphuric acid and 1% tetraethylammonium bromide in DMF. Britton - Robinson buffers were prepared with distilled and de-ionised water at intervals of 0.5 pH unit over the pH range AMANKWA et al. ELECTROCHEMICAL STUDIES ON The instrument was operated in the differential-pulse mode.2.6-7 .O. Reference Standard further purification. Dependence Studies Preparation of Calibration Graph Five test solutions of varying concentrations from 1 to 5 x 1 0 - 5 ~ were prepared by appropriately diluting the stock solution with 1.0 N sulphuric acid while in the total sample volume of exactly 20 ml the amount of methanol was always maintained at 1 ml. All samples were purged with oxygen-free nitrogen for 10 min prior to each run and a stream of nitrogen was allowed to flow gently over the surface of the solution during the electro-reduction. Samples of each of five concentrations were run five times and resulted in a corre-lation coefficient for the graph of 0.9999. Minoxidil (100.4y0) was obtained from Upjohn Company of Canada Ltd.and used without These studies were carried out in Britton - Robinson buffer over the pH range 2.6-7.0. A stock solution of minoxidil (10-3 M) was prepared in anhydrous methanol October 1983 Diffusion Dependence Studies These studies were carried out in the aforementioned sulphuric acid - methanol system on a 5 x lo-* M solution of minoxidil. The applied potential was from -0.6 to -2.1 V and the height of the mercury column ranged from 60 to 80 cm. The mercury flow-rate was measured at each of five heights over that range. Analysis of Pharmaceutical Dosage Forms MINOXIDIL AND ITS DETERMINATION IN TABLETS BY DPP 1223 Two dosage forms 2.5 and 10.0 mg tablets were available from the manufacturer. Twenty tablets were weighed and finely powdered and an amount of powder was taken that, according to the label would result in an approximately M solution of minoxidil in 50 ml of solvent.The accurately weighed sample was stirred magnetically for 20 min in 20 ml of methanol. The mixture was quantitatively transferred into a 50-ml calibrated flask diluted to volume with methanol and then filtered through a Whatman No. 1 filter-paper discarding the first 5 ml of the filtrate. A 0.6-ml aliquot of the filtrate was transferred into the polaro-graphic cell and 19 ml of 1.0 N sulphuric acid and 0.4 ml of methanol were added. As previ-ously described the solution was purged for 10 min with purified nitrogen prior to recording the polarogram. The amount of minoxidil was determined from a calibration graph.Content Uniformity Test Ten tablets were randomly selected from the sample. Each tablet was placed in an individ-ual 150-ml beaker 20 ml of methanol were added and the system was allowed to stand for 5 min in order to promote disintegration of the tablets. The remaining larger lumps of tablet mass were crushed with a glass rod and the mixture was stirred magnetically for 20 min. After transferring the mixture quantitatively into a 50-ml calibrated flask the determination was continued as described in the previous section except that 1 ml of the filtrate and 19 ml of 1.0 N sulphuric acid were used. The amount of minoxidil in each tablet was calculated by the direct comparison method using reference standard solutions of 0.238 6 x and 0.9546 x Macro-scale Electrochemical Synthesis of 2-Amino-6-piperidinopyrimidine (X) from Minoxidil The procedure was similar to that used for the controlled-potential coulometry with the exception that the cell contained 150 mg of minoxidil in 25 ml of 20% V/V methanal in 1.0 N sulphuric acid.The applied potential was held at -1.2 V and the reduction time was 6 h. On completion of the reduction the product together with the supporting electrolyte was M for the 2.5- and 10-mg tablets respectively. I I I I -0.8 -1.0 -1.2 -1.4 -1.6 Applied potentialN Fig. 1. Effect of pH on the differential-pulse polaro-graphic waves of minoxidil (5 x M) in A 1.0 N H,S04; B C and D Britton - Robinson buffer a t pH 3.0 5.0 and 7.0 respectively 1224 AMANKWA et aZ. ELECTROCHEMICAL STUDIES ON AnaEyst VOZ.108 separated from the mercury the pH adjusted to 7.0 with ammonia solution and the resulting solution extracted with chloroform. The organic layer was separated dried over magnesium sulphate and concentrated to 1 ml. The concentrated solution was applied to a 1-mm thin-layer silica gel plate and then developed with a methanol - ammonia solution (100 + 1.5) solvent system. Two spots were observed. The RF value of one corresponded to that of minoxidil while the other component with an RF value of 0.60 was scraped off the plate leached out with methanol and the methanolic solution was evaporated to dryness. This isolation yielded X a yellow crystalline powder that decomposed between 80 and 100 "C. Other experimental data were as follows NMR (200 MHz CDCl,) 6 1.6 (m 6H), 3.5 (m 4H) 5.9 (m 3H) and 7.9 (d 1H) (see Discussion); M+ m/e 178; IR (KBr disc) 930, 1060 1090 1400 1680 and 3000 cm-l.I I -1.6 -1.8 -2.0 -2.2 Applied potentialN Fig. 2. Cyclic voltammogram of minoxidil (5 x ~ O - * M ) in a solution of TEAB in DMF. -0.8 -1.0 -1.2 Applied potentialN Fig. 3. Effect of scan rate on the cyclic voltammogram of minoxidil (5 x 10-4 M) in 1.0 N H,SO,. Scan rates A 10 mV s-1; B 100 mV s-l; and C 200 mV s-l. Results and Discussion Minoxidil exhibits two d.c. and d.p. polarographic waves in 1.0 N sulphuric acid and in Britton - Robinson buffer (pH 3.0-7.0). In sulphuric acid the first wave is intense and well resolved from the second wave which is partially overlapped by the supporting electrolyte discharge current (Fig.1). At that pH the first wave had an Et (half-wave potential) value of 0.95 V. In the Britton - Robinson system however the E moves cathodically while the i d decreases with increasing pH. The second wave which has an E value at approximately -1.20 V has a diffusion current that is very much higher than that of the first. This unexpected height increase is the result of interference by reduction of the supporting electrolyte. Both the i d (diffusion current) and E of this wave vary with pH in the same manner as those of the first. The first wave is attributed to the reduction of the fully protonated N-oxide while the second probably results from the reduction of the 3,4 carbon-nitrogen double bond. The latter functional group is more difficult to reduce owing to the presence of the amino group at position 4 of the pyrimidine moiety.l October 1983 MINOXIDIL AND ITS DETERMINATION IN TABLETS BY DPP 1225 The first step in the process would then involve protonaton of the N-oxide to an N-hydroxy intermediate which undergoes a two-electron reduction to 2,4-diamino-6-piperidinopyrimidine intermediate.At higher potentials the 3,4 carbon-nitrogen double bond can be reduced by a two-electron process to an unstable intermediate 3,4-dihydro-2,4-diamino-6-piperidinopyrimi-dine. This intermediate undergoes deamination to give the final product 2-amino-6-piperi-dinopyrimidine which was isolated in this work. The coulometric analysis of minoxidil in 1.0 N sulphuric acid indicates that four electrons per molecule were involved in the electro-reduction process.The graph of limiting current versus the square root of corrected height of the mercury col-umn is a straight line that does not pass through the origin. It is possible under certain conditions to delineate the processes in the first wave and con-sequently voltammetry of 5 x M minoxidil in the aprotic solvent DMF is shown in Fig. 2. Stabilisation of the initially formed anion radical of minoxidil is clearly indicated by the quasi-reversible one-electron voltammogram. A slow following chemical reaction of the anion radical is indicated by the non-linearity of a plot of ip (cathodic peak current) with scan rate. Strong multiple adsorption peaks as illustrated in Fig. 3 were observed in the cyclic voltam-mogram of minoxidil in 1 .O N' sulphuric acid.The peak centre occurs at about - 1.0 V and the multiplicity of the peaks decreases with increasing scan rate. The major product extracted from the macro-scale experiment emits a pinkish fluorescence under short-wave ultraviolet light. It also produces a negative result with the iron(II1) chloride test indicating the absence of an N-oxide group. The strong IR N-oxide absorption peak between 1250 and 1300 nm is absent in the IR spectrum of the product. The NMR (200 MHz) spectrum of the product exhibits four distinguishable peaks in CDC1,. In the CDC1 - D,O system however the peak at 6 5.9 is reduced to a doublet with an integration value corresponding to one proton. No other change in the spectrum was observed. From all of the foregoing observations the following pathway is proposed as minoxidil pro-ceeds by electrochemical reduction in aqueous media to the final product 2-amino-6-piperi-dinopyrimidine (Scheme I).0 H t I 2e 2H+ ___) - - H20 X Scheme 1 Resolution provided by the differential-pulse wave at an E value of -0.95 V was better than that of the d.c. wave; consequently the former mode was utilised for the analysis of the dosage forms. The peak height varied linearly with the concentration of the drug over the range 1 x 10-5-5 x M. Table I provides the results of the assay for each of the minoxi-dil dosage forms. Values obtained by the manufacturer's quality control laboratory are TABLE I ASSAY OF MINOXIDIL TABLETS BY DIFFERENTIAL-PULSE POLAROGRAPHY IN' A 1.0 N SULPHURIC ACID - METHANOL SYSTEM Recovery by Recovery by Recovery by Recovery by Tablet Label manufacturer/ manufacturer differential-pulse differential-pulse lot No.claimlmg mg % polarography /mg polarography % H-651 2.6 2.49 99.6 2.46f0.09 98.2 H-710 10 9.69 96.9 9.66f0.12 96.6 * Each value is the average of five determinations 1226 AMANKWA CHATTEN AND PONS presented for comparative purposes and excellent agreement is observed between the two results. Table I1 lists the average of the results obtained for each dosage form when ten single tablets were analysed. Common excipients do not interfere with the electrochemical method. TABLE I1 AVERAGE VALUE FOR THE ANALYSIS OF TEN INDIVIDUAL MINOXIDIL TABLETS IN A 1.0 N SULPHURIC ACID - METHANOL SYSTEM Recovery by Recovery by Tablet Labelled diff erential-pulse diff erential-pulse lot No.claimlmg polarographylmg polarography % H-651 2.5 2.49fO.l 99.9 H-710 10 10.08f0.7 100.8 The proposed method has the advantages of simplicity high sensitivity and rapidity. It can distinguish between minoxidil and those degradation products that do not contain the N-oxide group and consequently the method can be applied in purity and stability studies on minoxidil. The authors gratefully acknowledge the samples of pure minoxidil and tablets as well as the useful information that they received from Upjohn Company of Canada Ltd. In addition, they thank B. Speiser Department of Chemistry of the University of Alberta for his helpful suggestions and assistance with the reduction products. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Miller D. D. and Love D. W. Am. J . Hosp. Pharm. 1980 37 808. Gilmore E. Weil J. and Chidsey C. N . En@. J . Med. 1970 282 521. “United States Pharmacopeia,” Twentieth Revision Mack Easton PA 1980. Royer M. E. KO H. Gilbertson T. J. McCall M. J. Johnston K. T. and Stryd R. J . Pharm. The Upjohn Company of Canada Ltd. personal communication. Thomas R. C. Hsi R. S. P. Harpootlian H. and Judy R. W. J . Pharm. Sci. 1975 64 1360. Thomas R. C. and Harpootlian H. J . Pharm. Sci. 1975 64 1367. O’Reilly J. E. and Elving P. J. J . EZectroanaZ. Chem. 1969 21 169. Tachibana M. Sawaki S. and Kawazoe Y. Chem. Pharm. BUZZ. 1967 15 1112. Brooks M. A. DeSilva J. A. F. and D’Arconte L. J . Pharm. Sci. 1973 62 1395. Smith D. L. and Elving P. J. J . Am. Chem. SOC. 1962 84 2741. Sci. 1977 66 1266. Received March 22nd 1983 Accepted May 4th 198