1490 Analyst, December, 1983, Vol. 108, pp. 1490-1494 Cholate Liquid Membrane lon-selective Electrode for Drug Analysis* Luigi Campanella, Lorenzo Sorrentino and Mauro Tomassetti Institute of Analytical Chemistry, University of Rome, 001 85-Rome, Italy A cholate liquid membrane electrode employing benzyldimethylcetyl- ammonium cholate as sensor was prepared, characterised and applied to the analysis of commercially available drugs containing cholanic acids. The results are comparable to those obtained using a benzoate electrode. Keywords : Cholate ; liquid membrane ; ion-selective electrode ; benzyldimethyl- cetylammonium cholate ; drug analysis The determination of cholic acids is becoming increasingly important for drug control and their quantitative determination is currently performed by high-performance liquid chroma- tography (HPLC) ,l thin-layer chromatography (TLC)2 and gas chromatography (GC)3 and by en~ymatic,~ spectrophotometric,5 radiochemical6 and calorimetric' methods.Each of these methods presents some problems, such as toxicity of the reagents, high cost and complexity of the apparatus or the procedure. In a previous study8 we proposed a potentiometric method based on the use of a liquid membrane electrode indicator containing tributylcetyl- phosphonium benzoate dissolved in nitrobenzene as sensor. The results obtained were of a To the electrometer t PTFE rinas AgCl PT F E housing Internal solution Liquid exchanger 6 PTFE discs Fig. 1. Liquid membrane electrode assembly. lower precision (but almost the same accuracy) compared with those obtained by traditional methods, but the method is both simpler and cheaper.In this paper we present a new liquid membrane electrode indicator, containing a quaternary ammonium cholate salt, benzyl- dimethylcetylammonium cholate, as sensor. This new sensor has been characterised and employed for the determination of the cholic acids in some commercial drugs. Results are compared with those obtained by a benzoate liquid membrane electrode and by the enzymatic - spectrophotometric method of T a l a l a ~ . ~ * Pap": presented at the Symposium on "Electroanalysis in Biomedical, Environmental and Industrial Sciences, Cardiff, 6 8 t h April, 1983.CAMPANELLA, SORRENTINO AND TOMASSETTI 1491 Experimental Reagents Cholic acid, sodium cholate, deoxycholic acid and chenodeoxycholic acid were supplied by Merck, ursodeoxycholic acid by Giuliani and lithocholic acid and benzyldimethylcetylammonium chloride (BDMCACl) by Fluka.All reagents for the enzymatic - spectrophotometric (ultraviolet) tests for bile salts, using a previously reported pro~edure,~ were provided by Nyegaard, Oslo. All reagents were of analytical-reagent grade. Apparatus a recorder (Varian G-14 A2) and an automatic burette (Radiometer ABU-11). calomel electrode was employed as reference electrode. were performed with a Perkin-Elmer 320 spectrophotometer. Potentiometric measurements were carried out using an electrometer (Radiometer PHM64), A saturated Spectrophotometric measurements Fig. 2. Comparison between the response of (a) the cholate electrode with that of (b) the benzoate electrode in standard solutions of sodium cholate with changing cholate concentration (C).A and A': C (initial) = 2.0 x M ; C (after dilu- tion) = 1.9 x 10-SM. B and B': C (initial) = 3.8 x 1 0 - 4 ~ ; C (after addition) = 9.0 x 1 0 - 4 ~ ; C (after dilution) = 3.8 x M. C and C': C (initial) = 4.0 x M ; C (after addition) = 6.8 x M ; C (after dilution) = 4.4 x 1 0 - 5 ~ . D and D': responses of the electrodes for successive increases of cholate concentra- tion, D, C = 6.7 x - 2.2 x 1 0 - 3 ~ ; D', C = 8.0 x 10-4 - 2.5 x 10-3 M. M ; C (after addition) = 4.5 x Procedure Benzyldimethylcetylammonium cholate (BDMCACh) is prepared by the reaction between commercially available BDMCACl in chloroform and an aqueous solution of cholic acid at pH 9. To the chloroform phase diethyl ether is added with stirring and the BDMCACh is1492 CAMPANELLA et al.: CHOLATE LIQUID MEMBRANE Analyst, VoZ. 108 precipitated. The product obtained is purified by recrystallisation and characterised by melting-point determination (104-106 "C) , elemental analysis, thermal analysis, TLC on silica gel, infrared spectroscopy, NMR spectroscopy and X-ray powder diffraction. The electrode assembly characteristics are as follows : electrode body, PTFE ; sensors, BDMCACh (C49H,505N.H20) ; membrane solvent, decan-1-01 (dielectric constant 8.1 relative to vacuum; viscosity 12.5 cP) ; membrane solution concentration, 0.01 M in BDMCACh; internal solution, sodium cholate 0.01 M, potassium chloride 0.01 M ; internal reference electrode, Ag - AgCl- C1-; PTFE discs; and Millipore supports, 1.3 x lop2 m diameter, 1 x m thickness and 2 x lO-'rn pore size.The electrochemical cell is operated under the following conditions: Ag - AgCl- 0.01 M KCl, 0.01 M sodium cholate I] 0.01 M BDMCACh in decan-1-01 11 solution under test 11 saturated calomel electrode. The arrangement details of the electrode are shown in Fig. 1. TABLE I SELECTIVITY CONSTANTS ACCORDING TO THE MOODY - THOMAS METHOD Background level of jn- Kchol- interference/M Benzoate . . . . 0.11 1 x 10-3 Acetate .. . . 8.30 x 1 x 10-2 Nicotinate . . . . 1.92 x 1 x 10-1 Citrate.. .. . . 3.80 x 10-4 1 x 10-2 Oxalate . . .. 4.00 x 10-4 1 x 10-1 Nitrate .. . . 1.37 x 10-2 1 x 10-2 Sulphate . . .. 1.77 x 10-4 1 x 10-1 Chloride .. .. 8.50 x 1.2 x 10-2 Phosphate . . . . 8.70 x 1 x 10-5 Hydroxyl . . . . 0.10 1.6 x 10-3 Results The electrode was checked for electrochemical and analytical characteristics using standard sodium cholate solutions. The response time was 10 s maximum. Fig. 2 shows the response of the cholate electrode compared with that of the benzoate electrode with varying cholate concentration. The linearity range was 4.00 x 10-5-0.01 M; the slope of the calibration graph was -0.0577 (& 0.0006) volts per decade of mean activity, at 16 "C; and the relative standard deviationlo was 0.5% over the same range. The accuracy, for sodium cholate solutions of concentrations ranging between 1 x and 0.01 M, differed according to the technique employed (i.e. titration method, direct potentio- metry, standard additions and Gran's plotllSl2) and the best results were obtained by a Gran's plot.The error was not higher than 3% compared with about 5.5% for the standard additions method and 6% for the other two methods. The electrode was also checked for selectivity constants relative to several common anions; the values, obtained according to the method of Moody and Thomas,l33l4 are shown in Table I. Values of the selectivity constants, K i j , were obtained by measuring the e.m.f. in a solution containing a fixed amount (see Table I) of the interferent j and a varied activity of the primary ion i for which the electrode is selec- tive. The value of Kij is calculated from K i j = ar/afv where z and y are the charges of i and j ; ai and a j are the values that correspond to the intersection of the part of the calibration TABLE I1 LINEARITY RANGE AND SLOPES OF THE CALIBRATION GRAPHS FOR CHOLIC ACIDS DETERMINATION Slope, volts per Acid decade of concentration Linearity range/M Cholic acid .. . . .. -0.0567 4.00 x 10-5-1.00 x 10-2 Deoxycholic acid . . . . . . -0.0598 3.98 x 10-4-3.98 x 10-3 Chenodeoxycholic acid . . .. -0.0593 2.00 x 10-4-3.16 x Ursodeoxycholic acid . . .. -0.0521 1.00 x 10-4-5.01 x 10-3 Lithocholic acid , . .. .. - 0.2000 2.61 x 10-5-7.94 xDecember, 1983 ION-SELECTIVE ELECTRODE FOR DRUG ANALYSIS 1493 TABLE I11 COMPARISON BETWEEN POTENTIOMETRIC DETERMINATIONS WITH A BENZOATE ELECTRODE, A CHOLATE ELECTRODE AND BY ENZYMATIC DETERMINATION OF CHOLANIC ACIDS I N COMMERCIAL DRUGS Each value is the mean of at least five determinations; values in parentheses are standard deviations (yo).Cholanic acid found, '$!, Cholanic Drug acid 1 Chenodeoxy- 2 3 cholic acid 4 Ursodeoxy- cholic acid 5 , Benzoate Choiate electrode electrode Cholate Enzymatic - Nominal (standard (standard electrode spectro- Differences between found and nominal values, yo value, addition addition (Gran's plot photometric & \ % method) (1) method) (2) method) (3) method (4) Method 1 Method 2 Method 3 Method4 71.4 62.1 (9.1) 75.0 (0.0) 76.4 (2.1) 76.4 (2.1) -13.0 +5.0 +7.0 +7.0 50.0 55.0 (8.1) 51.0 (7.1) 51.0 (8.2) 53.2 (3.0) +10.0 +2.O +2.0 +6.4 94.3 108.4 (9.8) 96.8 (9.7) 96.4 (9.9) 97.1 (2.9) +15.0 +2.7 +2.2 +3.0 83.3 82.1 (6.2) 85.8 (0.0) 85.0 (1.5) 82.0 (2.1) -1.4 +3.0 4-2.0 -1.8 56.6 53.8 (6.5) 57.4 (7.0) 58.3 (0.7) 56.6 (1.0) -4.9 +1.4 +3.0 0.0 graph that has an approximately zero slope.This part corresponds to the complete inter- ference by ion j with the Nernstian (or approximately Nernstian) part and corresponds to the electrode function for the primary ion i. For K i j < 1 the electrode responds preferentially to ion i and for Kij > 1 the electrode responds preferentially to ion j . In order to evaluate the possible uses of the electrode in the analysis of aqueous solutions containing the other cholanic acids (deoxy-, chenodeoxy-, ursodeoxy- and lithocholic acid) , all of biological and pharmaceutical interest , linearity concentration ranges and slopes were determined for these unconjugated acids (Table 11). Finally, the electrode was applied to the determination of two cholanic acids (chenodeoxy- cholic and ursodeoxycholic acids) , both of which are contained in a commercial drug used for dissolving biliar gallstones, using the following procedure. A weighed amount of each of five examined drugs was dissolved in water at pH 11 and after sedimentation of a little insoluble matter the solution was filtered and appropriately diluted.Each was then analysed TABLE IV COMPOSITIONS OF THE EXAMINED DRUGS Drug Component Content, yo mlm 1 Chenodeoxycholic acid 71.4 Corn starch 26.6 Aerosil 1.4 Magnesium stearate 0.6 2 Chenodeoxycholic acid 50.0 Lactose 38.4 Starch 2.0 Talc 2.0 Starch - sodium glycolate 6.0 Magnesium stearate 0.8 Precipitated silica 0.8 3 Chenodeoxycholic acid 94.3 Polyvin ylp yrrolidone 3.8 Colloidal silica 1.1 Magnesium stearate 0.8 4 Ursodeoxycnolic acid 83.3 Starch 10.0 Precipitated silica 3.3 Magnesium stearate 3.3 5 Ursodeoxycholic acid 56.6 Lactose 37.7 Pol y vin y lp yrrolidone 3.8 Magnesium stearate 1.1 Colloidal silica 0.81494 CAMPANELLA, SORRENTINO AND TOMASSETTI by an enzymatic - spectrophotometric and by a potentiometric method with a cholate electrode and a benzoate electrode.The better results for precision and accuracy were obtained by standard additions and Gran’s plot methods, using the cholate electrode. In Table I11 experi- mental data, obtained by this sensor, are reported and compared with those obtained by both a benzoate electrodea and enzymatic method.9 In Table IV the nominal percentage compo- sitions of all the examined drugs are reported.Conclusions From the results several conclusions were drawn, evidencing the superiority of cholate electrodes over benzoate electrodes. In general, faster response times are obtained using a cholate electrode-10 s was the maxi- mum time obtained. The accuracy in cholate standard solutions is almost the same as for a benzoate electrode and the best results are obtained by a Gran’s plot. The precision is higher for a cholate electrode than for a benzoate electrode and the linearity range is wider and the minimum detection limits are lower; also the values of the selectivity constants for anions are low enough to prevent any common interference when using the cholate electrode. Considera- tion of the slopes of the calibration graph for the five cholanic acids examined (Table 11) leads to the conclusion that the value of the slope increases as the number of hydroxyl groups in the steroid ring decreases or if one of these hydroxyl groups is in the p-position towards the plane of the ring.The results of drugs analysis are, however, less precise using a cholate electrode compared with an enzymatic - spectrophotometric method but comparably accurate, and the analysis with the cholate electrode is faster and cheaper than that by the enzymatic method; moreover it has no problems of reagent availability and storage. Finally, in comparison with other possible methods such as calorimetry’ and chromato- graphy,3 potentiometry is perhaps less precise but does not require expensive and complicated apparatus or need any pre-treatment of the sample.This work received financial support from the Italian C.N.R. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References Sian, M. S., and Harding Rains, A. J., Clin. Chim. Acta, 1979, 98, 243. Brusgaard, A., Clin. Chim. Acta, 1970, 28, 495. Setchell, K. D. R., and Matsui, A., Clin. Chim. Acta, 1983, 127, 1. Talalay, P., Methods Biochern. Anal., 1960, 8, 119. Biader Ceipidor, U., Curini, R., D’Ascenzo, G., and Tomassetti, M., Thermochim. Acta, 1981, 46, 269. Minder, E., Karlaganis, G., Schmied, U., Vitins, P., and Gustav, P., Clin. Chim. Acta, 1979, 92, 177. Biader Ceipidor, U., Curini, R., D’Ascenzo, G., and Tomassetti, M., Thermochim. Ada, 1981, 46, 279. Campanella, L., Sorrentino, L., and Tomassetti, M., Anal. Lett., 1982, 15, 1515. Biader Ceipidor, U., Curini, R., D’Ascenzo, G., Tomassetti, M., Alessandrini, A., and Montesani, C., Irving, H. M. N. H., Freiser, H., and West, T. S., Editors, IUPAC, “Compendium of Analytical Mascini, M., Ion-Scl. Electrode Rev., 1980, 2, 17. Moody, G. J., and Thomas, J. D. R., Sel. Rev. Anal. Sci., 1973, 3, 59. Moody, G. J., and Thomas, J. D. R. , Editors, “Ion-selective Electrodes,’’ Merrow Technical Library, Moody, G. J., and Thomas. J . D. R., in Pungor, E., Editor, “Ion-selective Electrodes,” Symposium Received May 6th, 1983 Accepted July 27th, 1983 G . Ital. Chim. Clin., 1980, 5, 127. Nomenclature,” Pergamon Press, Oxford, 1978. Watford, 197 1. held a t MBtrafured, Hungary, 23-25 October 1972, Akademiai Kiado, Budapest, 1973, p. 97.