196 Analyst, March, 1979, Vol. 104, pp. 196-200 Analysis of Steroids Part XXXII." Polarographic and Gas-chromatographic Methods Sandor Gorog, Anna Lauk6 and Zsofia Sziklay Determination of Allyloestrenol by Titrimetric, Chemical Works Gedeon Richter Ltd., P.O. Box 27, H-1476 Budapest, Hungary A titrimetric method is described for the determination of allyloestrenol based on methoxymercuration of its double bonds and titration of the acetic acid formed with standard sodium hydroxide solution. The relative standard deviation of the method is 0.29%. The polarographic reduction of the mercury addition compound on the dropping-mercury electrode is used for the determination of allyloestrenol in a tablet formulation with a relative standard deviation of 3.1 %. A gas-chromatographic method with a relative standard deviation of 1.5% is also described.The applicability of these methods to the determination of the stability of allyloestrenol and of its dosage form is discussed. Keywords : A llyloestvenol determination ; titrimetry ; polarography ; gas chromatography The analysis of allyloestrenol (17~-allyloestr-4-en-17-01) and its tablet formulation sets serious problems for the analyst. The molecule of this synthetic progestogen is one of the least substituted of the steroid drugs, its functional groups being two isolated double bonds and a tertiary hydroxyl group. In addition the drug, both in bulk and in formulations, is very sensitive to atmospheric oxygen and, therefore, any reliable analytical method should enable the unchanged material to be determined in the presence of its oxidative degradation products.Little information is available in the literature on the determination of allyloestrenol. Fokkens and Polderman1 described its examination by thin-layer chromatography, including the detection of the degradation products. Ganshirt and Poldennan2 developed this method into a quantitative colorimetric procedure using aqueous sulphuric acid for the colour development after thin-layer chromatographic separation and spot elution. This method enabled an assay of the stability of solid dosage forms to be carried out. Another colori- metric method was described by Kato3 using aluminium chloride as the reagent. Cavina et aL4 reported on the column-chromatographic determination of the material using flame- ionisation detection for monitoring of the eluent.The gas-chromatographic separation and gas-chromatographic - mass-spectrometric investigation of allyloestrenol in biological fluids was described by Brooks and Middletich.5 The purpose of this study was to develop a reliable titrimetric method for the determina- tion without the use of standards and to develop rapid polarographic and gas-chromato- graphic methods for examining the stability of the drug when in tablet form. Experimental Reagents and Apparatus without further purification. All materials and the solvent (methanol) were of analytical-reagent grade and were used Titrimetric method is prepared. Mercury(II) acetate solution, 0.1 M in methwzol. Sodium nitrate solution, saturated solution ixb methanol.Sodium bromide solution, 4 M in water. Sodium hydroxide solation, 0.1 M in water. medium using phenolphthalein as indicator. This should be used the same day as it Standardised against oxalic acid in a methanolic * For Part XXXI of this series, see Analyst, 1978, 103, 346.GOROG, L A U K ~ AND SZIKLAY Polarographic method Radelkis (Budapest) OH-105 d.c. - a.c. polarograph. &lercury(II) acetate solution, 0.1 M in methanol containing 1 ml 1-1 of acetic acid. Thymol solution, 0.1% m/V in methanol. Sodium hydroxide solution, 1 M in water. 197 Gas chromatographic method integrator. Materials investigated Bulk allyloestrenol (crude and purified samples) and other steroids were products of Chemical Works Gedeon Richter Ltd., Budapest. Turinal (Richter) tablets contain 5 mg of allyloestrenol per tablet.Hewlett-Packard 7620 gas chromatograph with jame-ionisation detector and 3380 reporting Internal standard. Dehydroepiandrosterone (3/3-hydroxyandrost-5-en-17-one). Procedures Titrimetric method An accurately weighed sample of allyloestrenol (approximately 0.12 g) is dissolved in 10 ml of methanol, 20 ml of 0.1 M mercury(I1) acetate solution and 5 ml of saturated sodium nitrate solution are added and the solution is allowed to stand for 30 min. Then 2 ml of 4~ sodium bromide solution are added and the slightly turbid solution is titrated with 0.1 M sodium hydroxide solution to a phenolphthalein end-point. A blank titration is carried out in the same manner but omitting the allyloestrenol, and this titration value (less than 0.2 ml) is subtracted from the above titre.The equivalent mass is half of the relative molecular mass (150.2). Polarographic method Fifteen Turinal tablets are finely powdered and a portion of the powder equivalent to about 60 mg of allyloestrenol is accurately weighed. It is triturated with 20 ml of methanol for 30min and the solution is filtered through a filter-paper into a 25-ml calibrated flask, washed with small portions of methanol and the filtrate diluted to volume with methanol and mixed. A 10-ml volume of this solution is transferred into a 25-ml calibrated flask, 5 ml of 0.1 M mercury(I1) acetate solution and 2 ml of saturated sodium nitrate solution are added and the solution is allowed to stand for 30 min. Then 2.5 ml of 1 M sodium hydroxide solution are added and the solution is diluted to volume with methanol.A 10-ml portion of the yellow, turbid solution is transferred into a polarographic cell, 1 ml of 0.1% thymol solution is added and the cell is carefully de-aerated with oxygen-free nitrogen. The solution polarographed in the ax. mode from -1 V, using dropping-mercury and mercury-pool electrodes as the working and reference electrodes, respectively. A standard solution is prepared by dissolving an accurately weighed amount of standard allyloestrenol (about 25 mg) in 10 ml of methanol in a 25-ml calibrated flask and treating the solution as described above beginning at ". . . 5 ml of 0.1 M mercury(I1) acetate solution and 2 ml o f . . ." The allyloestrenol content of the tablets is calculated from the peak currents of sample and standard in the usual way.Gas-chromatographic method One accurately weighed tablet is placed in a 25-ml calibrated flask, 26ml of methanol are added and the tablet is disintegrated by vigorous shaking. The flask is shaken for a further 15 min, 2 ml of internal standard solution containing 10 mg of dehydroepiandro- sterone dissolved in methanol are added and the volume of the solution is adjusted to the mark with methanol. The solution is mixed thoroughly and allowed to stand until the tablet base has settled. A 2-pl aliquot of the clear supernatant liquid is injected into the gas chromatograph. A glass column, 6 f t x 3mm i.d., packed with Anakrom ABS, 90-100 mesh, coated with 1% of OV-101 was used. The column temperature was 220°C, the198 GOROG, L A U K ~ AND SZIKLAY: ANALYSIS A?aalyst, Vol.102 vaporiser zone temperature 250 "C and the detector temperature 250 "C. The carrier gas (nitrogen) flow-rate was 30 ml min-l. The retention times of allyloestrenol and dehydro- epiandrosterone were 5.3 and 6.6 min, respectively. Results and Discussion Titrimetric method Titrimetric methods are seldom used in the analytical chemistry of steroids6 The favourable results obtained in this laboratory for the titrimetric analysis of steroidal double b0nds,7-~ ethyny1,lO keto,ll phenolic hydroxyl12J3 and epoxide14 groups, 16, 17-diols15 and 21-acyloxycorticosteroid~~~ encouraged us to try to develop such a method for the deter- mination of allyloestrenol. Further, titrimet ric methods in general have the advantage that no standard sample is necessary for the analysis; this is particularly important in the instance of allyloestrenol as it is one of the 1es:j stable steroid drug materials.Our first experiments in which we aimed to base the determination on the titration of the double bonds by our earlier described catalytic bromination m e t h ~ d , ~ failed to give acceptable results. However, good results were obtained by the addition of mercury( 11) acetate and methanol to the double bonds, complexing the excess of mercury(I1) acetate with sodium bromide and titration of the liberated acetic acid. OH OH + 2Hg(OCOCH3)2 + 2CH30H 1 O-C-CHs ll + 2CH3COO H 0 The principle of this method was described by Martin17 and developed to its generally accepted form by Johnson and Fletcher.l* On the basis of the papers published on the application of this method and references in standard monographs,lS2l this method appears to be applicable mainly to olefins with three, or at least two, hydrogen atoms in the cis position attached to the double bond.As double bonds of this type are usually not present in steroids, no indication of the use of this method for steroids has been found in the literature. According to our measurements 2 mol of mercury(I1) acetate are consumed by 1 mol of allyloestrenol, indicating that both double bonds are involved in the reaction. If the reaction is carried out at 0 "C 1 mol is consumed instantaneously while the reaction with the second mole takes place at a measurable rate (1.15 mol after a reaction time of 2 min).The first mole of mercury(I1) acetate undoubtedly reacts with the allylic double bond. The reaction between the A4 double bond and.niercury(I1) acetate is sluggish even at room temperature and a reaction time of more than 1 h would be necessary to complete the reaction at 25 "C. The use of sodium nitrate as catalyst considerably decreases the reaction time. The use of longer reaction times does not influence the results. By using the described method a freshly prepared standard allyloestrenol sample was found to be 100.21 yo pure. The relative standard deviation (eight parallel determinations) was 0.29%. Similarly good results were obtained with oestr-4-en-l7p-01 and oestr-4-en-17- one where naturally only 1 mol of the reagent was consumed. Several other steroids were also tested in order to obtain data for determining the selectivity of the method.It has been found that the most frequently occurring double bonds in steroid hormones do not react at all (6-ene-3p-hydroxy, 4-ene-3-ket0, 1,4-diene- 3-keto, 16-ene-20-ket0, 9-ene or ll-ene derivatives) and hence the method can be regarded as being fairly specific. The most serious interference was caused by the 17a-ethynyl group and the free phenolic hydroxyl group of oestrogens. However, no stoicheiometric amountsMarch, 1979 OF STEROIDS. PART XXXII 199 of acetic acid were formed in these instances (about 2.5 equivalents in the first group and 1-2 equivalents, depending on the reaction time, in the second group). The results obtained suggest that the titrimetric method is accurate and precise enough for it to be used for the assay of bulk allyloestrenol.The relatively large sample size obviously precludes the possibility of its use for the assay of tablets. Polarographic method The polarographic determination of olefinic unsaturation based on methoxymercuration was described by Fleet and Jee.22s23 The present method is a slightly modified application of the above method to the determination of allyloestrenol in tablets. Rectilinear calibration graphs were obtained over the range of allyloestrenol concentration of 5 x The peak potential veys'szcs the mercury-pool electrode was found to be -1.47 V. Antioxidants such as tocopherol do not interfere in the assay. As the polarographic method is based on the same reaction scheme, the remarks made on the selectivity of the volumetric method apply also to this method.to 5 x 10-3 moll-l. Gas-chromatographic method As the molecule of allyloestrenol has few substituent and stable functional groups, the gas-chromatographic determination can be carried out without derivatisation and at not too high a temperature. Use of the internal standard technique affords a suitable method for the assay of single tablets containing allyloestrenol. A comparison of the results for the allyloestrenol content of one batch of tablets as deter- mined by the polarographic and gas-chromatographic methods is shown in Table I. TABLE I DETERMINATION OF ALLYLOESTRENOL IN TURINAL TABLETS Each tablet had a nominal allyloestrenol content of 5 mg.Allyloestrenol content per tabletlmg r I Polarographic Gas-chromatographic 4.79, 5.05, 4.86 5.15, 4.76, 4.96 method method 5.01, 5.13, 4.91 5.04, 5.01, 6.08 Mean .. .. . . .. . . 4.93 Standard deviation .. . . 0.153 Relative standard deviation . . 3.1% 5.03 0.076 1.6% Examination of the Stability of Allyloestrenol When allyloestrenol as such or its tablet dosage form is exposed to air, polar degradation products appear in its thin-layer chromatogram.lP2 We examined the three methods to see whether they could measure allyloestrenol selectively in the presence of the degradation products. The oxidative degradation product can easily be separated from a highly contaminated sample on the basis of its extremely low solubility in hexane. Analysis by thin-layer chromatography showed that it was a mixture of several components, presumably peroxides or even more highly oxidised derivatives.When tested by the titrimetric method an allyloestranol content of about 50% was found for this material, indicating that only one of the double bonds (presumably that in the A4 position) is attacked during the autoxidation. From these results we conclude that the excellent accuracy and precision of the titrimetric method make it suitable for the deter- mination of bulk allyloestrenol but it must be remembered that a completely decomposed material shows a virtual content of 50%. The polarographic assay based on the same reaction is suitable for the rapid testing of tablets but neither method is suitable for examining the stability of samples.The isolated degradation product does not give gas-chromatographic peaks under the recommended The gas-chromatographic procedure appears to be suitable for this purpose.200 GOROG, LAUKC) AND SZIKLAY conditions or at higher temperatures, indicating that its components are further decomposed or polymerised in the vaporiser zone. Gas chromatography is therefore suitable for deter- mining the active ingredient content of tablets that have been stored under various con- ditions and hence this method was used in testing the stability of Turinal tablets. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. References Fokkens, J., and Polderman, J., Pharm. Weekbl., 1961, 96, 657. Ganshirt, H. G., and Polderman, J., J . Chromat., 1954, 16, 510.Kato, K., Chem. Pharm. Bull., Tokyo, 1964, 12, 578 and 824. Cavina, G., Moretti, G., Mollica, A., and Antonini, R., J . Chromat., 1971, 60, 179. Brooks, C. J . W., and Middletich, B. S., Clinica Chim. Actu, 1971, 34, 145. Gorog, S., and SzAsz, Gy., “Analysis of Steroid Hormone Drugs,” Elsevier, Amsterdam, 1978, Gorog, S., Acta Chim. Hung., 1966, 47, 1. Gorog, S., Acta Chim. Hung., 1966, 48, 121. CsizCr, E., and Giirijg, S., J . Chromat., 1973, 76, 502. Gorog. S., Acta Chim. Hung., 1966, 47, 7. Gorog, S., and TomcsAnyi, L., Acta Chim. Hung., 1966, 47, 121. Gorog, S., and Foldes, V., Acta Chim. Hung., 1!366, 48, 249. Gorag, S., Acta Chim. Hung., 1967, 51, 221. CsizCr, E., Giirog, S., and Gyenes, I., Acta Chinz. Hung., 1972, 73, 175. CsizCr, E., Giirdg, S., and SzCn, T., Mikrochim. Acta, 1970, 966. Gorog, S., J. Pharm. Pharmac., 1969, 21, 46s. Martin, R. W., Analyt. Chem., 1949, 21, 921. Johnson, J. B., and Fletcher, J. P., Analyt. Chem., 1959, 31, 1563. Cheronis, N. D., and Ma, T. S., “Organic Fuiictional Group Analysis,” John Wiley, New York, Polgk, A., and Jungnickel, J. L., iiz Mitchell, J . , Kolthoff, I. M., Proskauer, E. S., and Weissberger, Gyenes, I., “Titrationen in nichtwassrigen Medien,” Enke Verlag, Stuttgart, 1970, pp. 592-594. Fleet, B., and Jee, R. D., Talanta, 1969, 16, 1561. Fleet, B., and Jee, R. D., J . Electroanalyt. Chern., 1970, 25, 397. pp. 261-262. 1964, pp. 374-375, 525-527. A., Editors, “Organic Analysis,” Volume 111, Interscience, New York, 1956, pp. 301-310. NOTE-References 7, 8, 9, 10, 11, 12, 13, 14, 15 and 16 are to Parts 1, IT‘, XXIII, 11, 111, v, VI, XIX, Received August 7th, 1978 Accepted September 27th, 1978 XV and XI1 of this series, respectively.