ANALYST, DECEMBER 1986, VOL. 111 1367 Liquid Membrane Electrode for the Direct Determination of Ephedrine in Pharmaceutical Preparations* Saad S. M. Hassant and M. M. Saoudi Department of Chemistry, Faculty of Science, Ain Shams University, Cairo, Egypt The construction and performance characteristics of a liquid membrane electrode responsive to the ephedrine cation are described. The electrode is based on the use of the ephedrine-5-nitrobarbiturate ion-pair complex in nitrobenzene a s an ion-exchange site. The electrode shows a stable, near-Nernstian response for 10-2-10-5 M ephedrine over the pH range 4-7. The lower limit of detection is 4.5 x 10-6 M and the response time 20-90 s, and the selectivity coefficients for ephedrine relative to a number of interfering substances were investigated.Many organic and inorganic cations and pharmaceutical excipients and diluents commonly used in drug formulations do not interfere. The determination of 0.1-2000 pg ml-1 of ephedrine in aqueous solutions shows an average recovery of 99.2% and a mean relative standard deviation of 1.5%. The direct determination of ephedrine in some pharmaceutical preparations gives results that compare favourably with those obtained by the British Pharmaceutical Codex method. Keywords: Ephedrine electrode; liquid membrane; potentiometry; pharmaceutical analysis; 5-n itro ba rbitu ra te ion -pa ir complex Ephedrine is a sympathomimetic drug that stimulates both a- and P-adrenergic receptors. It is used in therapeutic doses at the level of 15-60 mg to produce peripheral vasoconstriction, to raise blood pressure, to prevent hypotension and to treat allergic states, catalepsy and myasthenia gravis.It is also utilised as an antidote for poisoning by central nervous system depressants. Official methods used for the determination of ephedrine in various pharmaceutical preparations are usually based on its extraction as a free base, followed by spectro- photometric determination at 241 nm.l Many organic com- pounds, drug excipients and various organic bases, however, absorb at the same wavelength and hence significantly interfere. The determination of ephedrine by spectrophotometric methods has been suggested based on oxidation with perio- date,2 hypohalite,3 chromate4 and hydrogen peroxide5 to give products that can be determined either directly in the ultraviolet region223 or in the visible region after condensation with semicarbazide or aminoantipyrine.2.4>5 Chromogenic reactions using l-fluoro-2,4-dinitrobenzene,6 picryl chloride ,7 ninhydrin,s bromothymol blue,9 tetrabromophenolphthalein ethyl esterlo and copper salts11 have also been described.These methods involve a time-consuming extraction step, require strictly controlled reaction conditions and suffer from severe interference from amines, amides and amino acids. Ephedrine in pharmaceutical preparations has been deter- mined by polarographic and potentiometric techniques. The polarographic methods are based on a prior conversion of ephedrine into polarographically active derivatives through brominationl2 or nitrosation13 reactions, followed by measur- ing the redox wave.Most of the potentiometric methods are based on the extraction of the free base, followed by titration with standard acids in non-aqueous media.14 Membrane electrodes incorporating ephedrine tetraphenylborate dis- persed in organic solvents or poly(viny1 chloride) have recently been used for direct potentiometric monitoring and potentiometric titration of ephedrine. 15-19 The presence of tertiary amines, various classes of alkaloids, arenediazonium salts, some inorganic cations and many of the drug excipients seriously interfere. * Presented at the 30th IUPAC Congress, Manchester, UK, 9-13 t To whom correspondence should be addressed. Current address: September 1985. Department of Chemistry, Quatar University, Doha, Quatar.In earlier papers, we have described liquid membrane electrode systems for the determination of strychnine,20 atropine ,21 caffeine ,22 lidocaine23 and some biogenic amines.24 These electrodes incorporate the picrolonate, reineckate, picrylsulphonate and flavianate ion-association complexes of these compounds as electroactive materials. In this study, a liquid membrane electrode with significantly improved characteristics for ephedrine was developed. It is based on the use of the ephedrine-5-nitrobarbiturate ion-pair complex in nitrobenzene as an electroactive material. The electrode has been successfully used for the determination of ephedrine in simple and complex matrices. Experimental Reagents All solutions were prepared with de-mineralised, doubly distilled water and all chemicals were of analytical-reagent grade unless stated otherwise.Standard 10-2-10-6 M aqueous ephedrine solutions were freshly prepared by accurate dilutions of a 10-1 M ephedrine hydrochloride stock solution (20.17 g 1-1). An aqueous 10-1 M solution of 5-nitrobarbituric acid was prepared by dissolving 0.207 g of 5-nitrobarbituric acid trihydrate in 100 ml of de-mineralised, doubly distilled water and filtering. Pharmaceutical preparations containing ephedrine were obtained from local drug stores. Equipment The potentiometric measurements were made with an Orion microprocessor Ionalyzer (Model 901) using the ephedrine-5- nitrobarbiturate liquid membrane electrode in conjunction with an Ag - AgCl double-junction reference electrode (Orion Model 90-02) containing a 10% mlV KN03 solution in the outer compartment.The cell potentials were measured for stirred solutions at 25 k 1 "C at pH 4-8 using the following electrochemical cell: Ag - AgClI 10-2 M ephedrine hydro- chloride - 10-2 M KC11 10-2 M ephedrine-5-nitrobarbiturate in nitrobenzenel !porous membrane1 Isample test solution1 Ag - AgCl double-junction reference electrode. An Orion glass - calomel combination electrode (Model 91-02) was used for pH adjustment.1368 ANALYST, DECEMBER 1986, VOL. 111 Procedure Membrane preparation The ephedrine-5-nitrobarbiturate ion-pair complex was pre- pared by mixing 15 ml of aqueous 10-2 M ephedrine hydro- chloride with 20 ml of aqueous 10-2 M 5-nitrobarbituric acid. The mixture was cooled in an ice - water mixture, the precipitate was filtered off by suction, washed with de- mineralised, doubly distilled water, dried at ca.70 "C for 15 min and then ground. Elemental analysis data of the product agreed with the composition CI4Hl8N4O6. The infrared spectrum of the product displays almost all the absorption bands that appear in the spectra of both reactants and also a stretching vibration band at 2460 cm-1 assigned to the imino group. The liquid ion-exchange membrane was prepared by making a 10-2 M solution of ephedrine-5-nitro- barbiturate in nitrobenzene. u OH Electrode preparation The body of an Orion Series 92 electrode, equipped with a microporous membrane (Orion 92-05-04), was used. The electrode was assembled and the internal reference and liquid ion-exchange solutions were injected into the appropriate ports.The internal reference solution was a mixture of an equal volume of KC1 and ephedrine hydrochloride solutions, 2 x 10-2 M each. The electrode was conditioned by soaking in a 10-3 M aqueous ephedrine hydrochloride solution for 2 d before use. When not in use, the electrode was immersed in the same solution. The selectivity coefficients were measured using the mixed solution method.25 The performance characteristics of the electrode were evaluated as previously described.2@24 Determination of ephedrine in pharmaceutical preparations The contents of five ephedrine vials were mixed and a volume equivalent to one vial was transferred into a 50-ml beaker, followed by dilution with 30 ml of de-mineralised, doubly distilled water.The solution was acidified with 1 ml of 0.1 M HCl, heated at ca. 70 "C for 5 rnin and cooled to room temperature. The pH of the solution was adjusted to 4-7 with 0.05 M NaOH. The solution was then transferred into a 50-ml calibrated flask, diluted to the mark with de-mineralised, doubly distilled water, shaken and transferred into a 100-ml beaker. The ephedrine-5-nitrobarbiturate liquid membrane electrode in conjunction with the reference electrode was immersed in the test solution. The potential reading was recorded when stable and compared with a calibration graph prepared from pure ephedrine hydrochloride solutions under identical conditions. For the determination of ephedrine in pharmaceutical tablets, ten tablets were pulverised and a weighed portion equivalent to one tablet was transferred into a 50-ml beaker and dissolved in 30 ml of demineralised, doubly distilled water.The procedure used for the determination of ephedrine in vials was then followed. OH CH3 CH3 OH Fig. 1. Ephedrine-5-nitrobarbiturate ion-pair complex I I I I L 6 5 4 3 2 -Log ([ephedrineli~) Fig. 2. ephedrine-5-nitrobarbiturate liquid membrane electrode Calibration graph for ephedrine in the pH range 4-7 using the 70 30 > E Gi -10 -50 Results and Discussion Membrane Material and Characteristics Some organic bases, including ephedrine, can be identified by examining their photomicrographs or by measuring the melting-points of their 5-nitrobarbiturate derivatives.26 In this study, the ephedrine-5-nitrobarbiturate ion-pair complex was prepared (Fig.l ) , characterised and tested as a novel ion-exchange site in a liquid membrane electrode responsive to ephedrine. Solutions of the ephedrine-5-nitrobarbiturate ion-pair com- plex (10-2-10-3 M) were prepared in lipophilic solvents (such as nitrobenzene, octan-1-01 and decan-1-01) and were tested as 3 6 9 Effect of pH on the potential of the ephedrine-5-nitrobarbitu- PH Fig. 3. rate liquid membrane electrode liquid ion-exchange membranes. Potentiometric measure- ments at 25 k 1 "C with the ephedrine-5nitrobarbiturate liquid membrane electrode incorporating nitrobenzene as a membrane solvent give a near-Nernstian response for the ephedrine cation over the concentration range 10-2-10-5 M (Fig. 2). The initial slope of the calibration graph is 55 mV per concentration decade.The least-squares equation obtained from the calibration data is E(mV) = (55 k 0.5) log C + (160 k 0.7).ANALYST, DECEMBER 1986, VOL. 111 1369 The standard deviation is 1.1 mV and the detection limit calculated according to the IUPAC recommendation27 is 4.5 X 10-6 M. The slope and limit of detection offered by this electrode system are better than those obtained using either octan-1-01 or decan-1-01 as membrane solvents; this is probably due to the ease of dissolution and dissociation of the complex in nitrobenzene. Response Time and Stability of the Membrane The average time required for the ephedrine-5-nitrobarbitu- rate liquid membrane electrode to reach a potential within k 1 mV of the final equilibrium value after successive immer- sion in a series of ephedrine solutions, each having a 10-fold difference in concentration, was measured.Stable responses were achieved almost instantaneously for concentrations 210-3 M, and within 60-90 s for concentrations < l o - 4 M. The electrode exhibits a day-to-day reproducibility of about 2 2 mV for the same solutions for 6 weeks after preparation, provided that the electrode is not used in the presence of high concentrations of strong interfering compounds or in highly alkaline or acidic solutions. The influence of pH on the electrode response to different ephedrine concentrations is shown in Fig. 3. The electrode potential is independent of pH in the range 4-7. Over this range the potential does not vary by more than -2.0 mV for any ephedrine concentration in the range 10-2-10-4 M.Prolonged immersion of the electrode in alkaline solutions (pH > 9) causes a deterioration in its response, probably because of the back-extraction of the 5-nitrobarbiturate anion from the organic membrane phase to the aqueous test solution and/or gradual precipitation of the ephedrine base in the test solution. The decrease of the potential below pH 4 may be due to interference from H+. Table 1. Selectivity coefficients for ephedrine-5-nitrobarbiturate liquid membrane electrode Selectivity coefficient, Interfering species (B) e;;, B Me thylurea . . . . . . . . Ace tamide . . . . . . . . Aminobenzoicacid . . . . Piperidine . . . . . . . . Diethylamine . . . . . . Triethanolamine . . . . . . Tetramethylammonium chloride G1 ycine .. . . . . . . . . Alanine . . . . . . . . Nicotine . . . . . . . . Nicotinicacid . . . . . . Strychnine . . . . . . . . Caffeine . . . . . . . . Quinine . . . . . . . . Li+ . . . . . . . . . . Ba2+ . . . . . . . . . . NH4+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 x 10-3 3.6 x 10-3 2.4 x 10-3 2.1 x 10-2 1.9 X 10-2 2.1 x 10-2 6.7 x 10-2 6.3 X 10-3 8.1 x 10-1 6.6 x 10-2 2.8 1.7 1.4 3.1 x 10-3 6.2 x 10-3 8.2 x 10-3 6.1 x 10-3 7.3 x 10-2 5.4 x 10-2 2.5 x 10-3 Table 2. Direct potentiometric determination of ephedrine using ephedrine-5-nitrobarbiturate liquid membrane electrode and the calibration graph method Ephedrine hydrochloride Standard deviation, Addedtyg ml-1 Recovery,* % Y O 2000.0 98.5 0.4 1500.0 98.4 0.4 1000.0 99.1 1.5 500.0 98.1 1.6 100.0 98.0 1.7 50.0 99.0 1.4 10.0 98.0 1.4 5.0 99.5 1.8 1 .o 101.5 2.0 0.10 101.8 2.8 * Average of three measurements.Selectivity Coefficients of the Membrane The potentiometric selectivity coefficients (K$&B) of the ephedrine-5-barbiturate liquid membrane electrode were experimentally determined by the mixed solution method recommended by IUPAC and were calculated from the modified Nernst equation .25327 The ephedrine concentrations were varied from 10-3 to 10-5 M while the concentration of the competing cation was kept constant at 10-3 M. The data in Table 1 show the selectivity of the ephedrine electrode over a series of potentially interfering organic and inorganic cations.Soluble drug excipients and diluents such as maltose, glucose, lactose, starch and gelatin binder that are present in some tablets do not interfere with the response of the electrode. Further, the electrode exhibits negligible interference from many nitrogenous compounds such as amines, amides and amino acids. The electrode is, however, not really selective for ephedrine over some other alkaloids such as strychnine, quinine and caffeine. Determination of Ephedrine Ephedrine solutions in the concentration range 0.1- 2000 pg ml-1 were prepared from pharmaceutical grade reagents and determined by the direct potentiometric method using the ephedrine-5-nitrobarbiturate liquid membrane elec- trode. The potentials recorded in these solutions were compared with a calibration graph. The results obtained (Table 2) for ten samples, each determined in triplicate, showed an average recovery of 99.2% and a mean standard deviation of 1.5%. Ephedrine was also determined in some pharmaceutical preparations.Injections and tablets were homogenised, treated with 0.1 M HC1, heated to effect complete solubilisa- tion and diluted with de-mineralised, doubly distilled water. The potential was then measured after the adjustment of pH to 4-7 and compared with a calibration graph. The results obtained (Table 3) show an average recovery of 99.1% of the nominal values and a mean standard deviation of 1.7%. No interference was caused by active or inactive ingredients and diluents commonly used in drug formulations.The British Table 3. Determination of ephedrine in some pharamaceutical preparations using ephedrine-5-nitrobarbiturate liquid membrane electrode Electrode method BPC method1 Preparation Labelled ephedrine Recovery,*% S.D.,% Recovery,*% S.D.,% Ephedrine sulphate (injection) . . . . 25 mg ml- 100.0 1.7 98.8 2.0 Ephedrine sulphate (injection) . . . . 50 mg ml- 99.4 1.5 98.7 1.8 Ephedrine hydrochloride (tablet) . . . . 25 mg per tablet 98.3 1.6 102.2 2.3 Ephedrine hydrochloride (tablet) . . . . 50 mg per tablet 98.6 1.8 101.4 2.0 * Average of five measurements.1370 ANALYST, DECEMBER 1986, VOL. 111 Pharmaceutical Codex method,l involving a prior extraction of the free base with cyclohexane followed by spectropho- tometric determination at 241 nm, was also used for compari- son.The results obtained by both methods are in good agreement (Table 3). The electrode method, however, offers several advantages in term of simplicity, selectivity and precision. Further, it eliminates the time-consuming extrac- tion step and is directly applicable to coloured and turbid solutions. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. References “British Pharmaceutical Codex 1973,” Pharmaceutical Press, London, 1973, p. 671. Wallace, J. E., J. Pharm. Sci., 1969, 58, 1489. Chafetz, L., Gosser, L. A . , Schriftman, H., and Daly, R. E . , Anal. Chim. Acta, 1970, 52, 374. Wallace, J. E., Anal. Chem., 1967, 39, 531. Jaakko, H., and Arja, K., Farm. Aikak, 1976, 85, 125. Tammilehto, S . , Farm. Aikak, 1975, 84, 53. SiirSunovi, M., and Chi, N. T. K., Cesk.Farm., 1966,15,474. Zobin, A., and Gracza, M., Acta Pharm. Hung., 1975,45,101. Kafedzhieva, P., and Vuleva, E., Farmatsiya (Sofia), 1977,27, 5. Sakai, T., Hara, I., and Tsubouchi, M., Chem. Pharm. Bull., 1976, 24, 1254. Morvay, J., and Sti’er, G., Acta Pharm. Hung., 1965,35, 199. DuSinsky, G . , and Aavabik, T., Cesk. Farm., 1959, 8, 205. DeMarco, A., and Mecarelli, E . , Farmaco, Ed. Prat., 1967,22, 795. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Sell, E., and Rajzer, D., Chem. Anal. (Wursaw), 1976,21,933. Selinger, K., and Staroscik, R., Pharmazie, 1978, 33, 208. Fukamachi, K., and Nakagawa, R., Morimoto, M., and Ishibashi, N., Bunseki Kagaku, 1975, 24, 428. Goina, T., Hobai, S., and Rosenberg, L., Farmacia (Bucharest), 1978, 26, 141. Zeng, J., Yaoxue Xuebao, 1982, 17, 841. Selinger, K., and Staroscik, R., Chem. Anal. (Warsaw), 1982, 27, 223. Hassan, S. S. M., and Elsayes, M. B., Anal. Chem., 1979, 51, 1651. Hassan, S. S. M., and Tadros, F. S., Anal. Chem., 1984, 56, 542. Hassan, S. S. M., Ahmed, M. A., and Saoudi, M. M., Anal. Chem., 1985,57, 1126. Hassan, S. S. M., and Ahmed, M. A., J. Assoc. Off. Anal. Chem., 1986, 69, 618. Hassan, S. S. M., and Rechnitz, G. A . , Anal. Chem., 1986,58, 1052. Ma, T. S . , and Hassan, S. S. M., “Organic Analysis Using Ion Selective Electrodes,” Volumes 1 and 2, Academic Press, London, 1982. Chatten, L. G., and Barry, P. J., Can. J. Pharm. Sci., 1968,3, 40. IUPAC Analytical Chemistry Division, Commission on Ana- lytical Nomenclature, “Recommendations for Nomenclature of Ion Selective Electrodes,” Pure Appl. Chem., 1976, 48, 127. Paper A61126 Received April 25th, 1986 Accepted July 3rd, 1986