摘要:

ANALYST. JANUARY 1984 VOL. 109 47 Spectrophotometric Investigation of Palladium( II) = 2-Thiobarbituric Acid Complexes in Aqueous and Aqueous Ethanolic Media Basilio Morelli Universita degli Studi di Bari Dipartimento di Chimica Via Amendola 173 70126-Bari Italy Palladium(ll) forms 2 3 and 1 3 complexes with 2-thiobarbituric acid in water and in 30% V/V ethanol, respectively. The molar composition of the complexes has been determined by continuous variation and molar-ratio methods and confirmed by elemental analysis. The system conforms to Beer's law up to 10 pg ml-1 of palladium [molar absorptivity (E) 1 .O x lo4 I mol-1 cm-1 at 374 nm] in water and up to 6 pg ml-1 (E 7.5 x lo3 I mol-1 cm-1 at 386 nm) in 30% V/Vethanol; the detection limits are 0.13 and 0.18 pg ml-' and the relevant Sandell's sensitivities are 0.0106 and 0.014 1 pg cm-2 per 0.001 absorbance unit respectively.Infrared spectroscopy has been used in order to establish the type of coordination between 2-thiobarbituric acid and palladium. The tolerance limits for platinum metals and other cations are given. The experimental results have been critically analysed and a comparison with the main most recent spectrophotometric reagents for palladium is presented. Keywords Palladium determination; 2-thiobarbituric acid; spectrophotometry In this paper we present a spectrophotometric study of the Pd(I1) - TBA system in both aqueous and aqueous ethanolic media. Experimental and Results Reagents All reagents were of analytical-reagent grade. Palladium(I1) standard solution.Working solutions of 6.1 x M palladium were prepared by dissolving pallad-ium(I1) chloride in hydrochloric acid (solutions were standard-ised using the dimethylglyoxime method). 2-Thiobarbituric acid solution 3 x M. Prepared in the usual way.'-4 Foreign ion solutions. These contained 2-10 pg ml-1 of the ion.2.4 Apparatus The apparatus was the same as that used previously.1-4 of pH above 1.5 showed a progressively lower absorbance in both media; moreover turbidity was observed in the more concentrated samples. All further measurements were made in a strongly acidic medium. Absorption Spectra Fig. 1 shows the absorption spectra of Pd(I1) - TBA complexes against a reagent blank in ( a ) aqueous medium and (b) aqueous ethanolic medium. The complexes show an absorp-tion maximum at 374 and 386 nm respectively.Calibration Graphs The system conforms to Beer's law up to 10 and 6 pg ml-1 of palladium in aqueous and aqueous ethanolic media respec-tively. The molar absorptivities are 1 .O x 104 1 mol- 1 cm-1 at 374 nm (aqueous medium) and 7.5 X 1031 mol-1 cm-1 at 386 nm (aqueous ethanolic medium); the Sandell's sensitivi-ties of the reaction for palladium were calculated to be 0.010 6 and 0.014 1 pg cm-2 per 0.00 1 absorbance unit respectively. Development of Colour and Analytical Procedure Full colour was developed by heating on a water-bath at 50 "C for 1 min in aqueous medium and at 80 "C for 10 min in aqueous ethanolic medium (30% V/V ethanol). Colour-developed solutions measured at 25-80 "C between 30 s and 30 rnin showed no change in the spectral curv.e.The recommended procedure for the determination of palladium was as follows. In a 5-ml calibrated flask mix a few microlitres of 6.1 x 10-3 M palladium standard solution with 2ml of 3 ~ 1 0 - 2 ~ TBA solution. Add 3 0 ~ 1 of concentrated hydrochloric acid (only for the experiments in aqueous medium). Dilute to the mark with distilled water (in aqueous medium) or with a 1 1 mixture of concentrated hydrochloric acid and absolute ethanol (in aqueous ethanolic medium). Heat the mixture on a water-bath at 50 "C for 1 min or at 80 "C for 10 min respectively. Cool and measure the absorbance at 374 and 386 nm respectively against a reagent blank. 1-4 Wavelengthlnm Effect of Acidity A constant and maximum absorbance at 374 and 386nm, respectively was attained in solutions of pH S1.5.Solutions Fig. 1. Absorption spectra of Pd(I1) - TBA complexes. (A) Aqueous medium; (B) 30% V/V ethanol. Pd concentration 23 pg per 5 ml; reference reagent blan 48 ANALYST JANUARY 1984 VOL. 109 Composition of the Complexes Aqueous solution Fig. 2 shows (a) typical continuous variations and (b) molar-ratio graphs. An extrapolation of ( a ) at the initial and final portion of the graph gives an intersection at a molar fraction of 0.4 of palladium corresponding to a ratio of palladium to TBA of 2 3; ( b ) shows a break at a molar ratio of 1.5 of ligand to metal by providing further evidence for the 2:3 composi-tion of the complex (this stoicheiometry is infrequent; however other 2 3 complexes of palladium with sulphur containing ligands are known5.6).Aqueous ethanolic solution ( a ) Continuous variation and (6) molar-ratio graphs are shown in Fig. 3. Both methods indicate palladium to TBA reaction ratios of 1 3. The composition was also determined by elemental analy-sis. Elemental analysis of the complex isolated from aqueous solution presumed to be Pd2( C4H4N202S)3 was as follows: calculated Pd 32.98 C 22.34 H 1.87 N 13.02 S 14.91 and C1 0%; found Pd 32.95 C 22.40 H 1.90 N 13.05 S 15.01 and C1 0%. 0.8 a, (0 0.6 e 1 0.4 Q: 0.2 0 0.2 0.4 0.6 0.8 1 2 3 4 5 [Pd2+1/t[Pd2+1 + [TBAI) Molar ratio Pd(ll) TBA Fig. 2. Continuous variations graph; Pd + TBA concentration 6.4 X (6) Molar-ratio graph; Pd concentration.2.68 x 374 nm; reference water Molar ratio of Pd(I1) - TBA complex in aqueous medium. (a) M. M. Wavelength, 0.6 0.5 0.4 a, c m a L 0.3 0, a a 0.2 0.1 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 [Pd2+l/llPd2+l + ITBAI) Molar ratio Pd(ll) :TBA Fig. 3. Molar ratio of Pd(I1) - TBA complex in 30% V/V ethanol. ( a ) Continuous variations graph; Pd + TBA concentration 1.2 X 10-3 M. ( b ) Molar-ratio EraDh TBA concentration. 1.08 X M. Wave-Elemental analysis of the compound isolated from aqueous ethanolic solution postulated to be Pd(C4H4N202S)3C12 was as follows calculated Pd 17.44 C 23.64 H 1.98 N 13.78, S 15.77 and C1 11.63%; found Pd 17.40 C 23.68 H 2.0, N 13.8 S 15.75 and C1 12.0%. The results seem to be in better agreement with the proposed structures than any other reasonable stoicheio-metry confirming respectively the 2 3 and 1 3 molar ratios of palladium to TBA.Infrared Spectroscopic Investigation An attempt to establish the manner in which TBA molecules coordinate with palladium atoms was made by means of infrared spectroscopy. Fig. 4 shows the infrared spectra of ( a ) TBA and ( b ) 2 3 and (c) 1 3 Pd -TBA complexes isolated and dried under vacuum. Infrared spectra were obtained by preparing potassium bromide discs of the solid substances. Evident differences between the spectra of the two complexes confirm the formation of different compounds in the two media. A tentative assignment to the main absorption bands was made by analogy with infrared studies on similar compounds viz.thiourea and metal - thiourea complexes,7 Cu - TBA8 and Ru - TBA2 complexes. A difference between the spectrum of TBA and the spectra of its palladium complexes appears in the 1140-1 150cm-1 region. A strong absorption of TBA (a doublet) in this region assigned to C=S stretching vibration disappears on complex formation. This observation seems to indicate that a coordinate bond is formed between sulphur and the central metal atom. The formation of the S+M bond increases the electron demand by the donor sulphur atom and favours the resonance with the nitrogen-carbon group increasing the contribution of the highly polar structure HNC(N+H)S-; the band at 1525cm-l in the spectrum of TBA assigned to an N-C-N stretchingvibration of 80 40 0 8 $ 80 0 m c-' 4- .-E 40 E 0 80 40 C Wavelengthhm 3.0 4.0 6.0 8.0 10 16 25 I I 1 I 1 I 1 (a) I I I 1 4000 3000 2000 1600 1200 800 400 Wavenum ber/cm - 1 Fig.4. Infrared spectra of ( a ) TBA (6) Pd(I1I - TBA corndex in length 386 nm; GfLrence 30% V/V ethanol aqueous medium;'and (c) Pd(11) - TBA cornpiex in 30% WV Gthano ANALYST JANUARY 1984 VOL. 109 49 the B1 type corresponds to the 1530 and 1 570 cm-1 bands of 2 3 and 1 3 complexes respectively. This frequency increase observed for the Pd-TBA complexes is in accord with the higher contribution of the double bond character of the carbon-nitrogen bond on coordination of TBA with the sulphur atom. Another important difference appears in the 6-pm region. The strong absorption band at 1 700 cm-1 in the spectrum of free TBA which can be assigned to carbonyl vibrations is shifted to 1 630 cm-1 on complex formation.The low position of the C=O stretching bands indicates a higher contribution of single bond character of this link and it would suggest that TBA acts as hybrid ligand with a soft donor atom (sulphur) and a hard one (oxygen); tentatively it is also possible to hypothesise that TBA acts as a monodentate ligand with the sulphur atom (in accordance with analogous complexes with copper* and ruthenium2) with convenient arrangement of its donor groups in the complexes and increase in the enol content of the carbon-oxygen bond. Effect of Foreign Ions The extent of interference of foreign ions was determined in the usual ~ a y . 2 9 ~ The substances tested and the tolerances, defined as in previous w0rks,2.~ are listed in Table 1.In general the tolerance to diverse ions is a little higher in aqueous medium than in aqueous ethanolic medium (espe-cially for the platinum group metals). Table 1. Tolerance of palladium(I1) - 2-thiobarbituric acid system to diverse ions. All solutions contained 23pg of Pd per 5ml. The tolerance to a foreign ion was taken as the largest amount that gives an absorbance not more than 1% absolute different from that of palladium alone Foreign ion Li(1) . . K(I) . . Tl(1) . . Ca(I1) . . Sr(I1) . . Ba(I1) . . La(1) . . Cr(II1) Mn(I1) Fe( 111) Ni(I1) . . Cs(1) . . Mg(I1) Co(I1) . . Cu(I1) . . * Zn(I1) . . Cd(I1) . . Hg(I1) . . A1 ( 111) Sn(I1) .. Pb(I1) . . As( 111) S b( 111) Bi(II1) . . Ru(II1) R h (111) 0 s (VI 11) Ir(II1) . . Pt(I1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amount tolerated relative to Pd Yo Aqueous medium 30% ethanol medium . . . . . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 140 190 67 8 30 74 2 500 5 38 3 6 125 64 14 220 21 12 175 52 25 8 45 4 4 10 18 29 10 0.7 140 135 141 47 13 26 147 2 400 6 30 2 3 17 30 58 220 15 17 29 46 33 7 53 25 1 7 18 8 3 0.8 Statistical Analysis of Results and Conclusions A critical evaluation of the proposed methods can be made by statistical analysis of the experimental results.The high values of the correlation coefficients of the calibration graphs (calculated from the best-fit graph by linear regression) i.e. 0.999 in aqueous and 0.997 in aqueous ethanolic media and the relevant values very close to zero of intercepts on the ordinate of 0.001 5 and 0.003 5 respectively, illustrated the linearity of calibration graphs aad the confor-mity of the system to Beer's law.Detection limits (D.L.) and variances (S;) at the p = 0.01 level of significance were calculated2-4 to be D.L. = 0.13 pg ml-1 and S; = 1.84 X 10-5 in aqueous medium and D.L. = 0.18 pg ml-1 and S; = 1.96 x lO-5in aqueous ethanolic medium. These values indicate the high sensitivity of the method and the negligible scatter of the points with respect to the line of regression. The variation of the confidence limits in the form of percentage uncertainty on concentration i. e. as AC/C%2-4 for both p = 0.05 andp = 0.01 levels of significance is shown in Fig. 5 which offers an immediate indication of the relative uncertainty on concentration over the full range of concentra-tions tested. From the analysis of the results the determination of palladium in water seems more advisable than in 30% ethanol for the reasons previously seen higher sensitivity wider useful concentration range greater precision and tolerance to foreign ions.However both methods allow the determination of palladium with very high accuracy. From a comparison with the main spectrophotometric methods for palladium devel-oped in recent years (Table 2) TBA is seen to rank among the more sensitive reagents; moreover the present method is faster and simpler than many other sensitive methods which often require prior solvent extraction of the complex and/or are utilisable in narrower concentration ranges. 30 25 20 8 0 Q 15 10 5 0 2 4 6 8 10 [Pdz+]/pg ml-' Fig. 5. Variation of confidence limits a t p = 0.01 andp = 0.05 levels of significance in the form of uncertainty (YO) on the concentration.Continuous lines in aqueous medium; broken lines in 30% V/V ethano 50 ANALYST. JANUARY 1983. VOL . 109 Table 2 . Comparison of spectrophotometric reagents for palladium(I1) Range of concentration/ Molar absorptivity/ pg ml-1 Imol-lcni-~ Reagent 2-Thiobarbituric acid . . . . . . . . . . . . . . 0.13-10 aqueous 1 . 0 X 10-V374 nm 7.5 X 103/386 nm medium; 0.18-6,, aqueous ethanolic medium a-Benzoin oxime . . . . . . . . . . . . . . . . l.R.4.4.6.trimethy I.( lH,4H).pyrimidine. 2.thiol. R = H . . . . . . . . . . . . . . . . . . R = C2HS . . . . . . . . . . . . . . . . . . R=n-C4Hg . . . . . . . . . . . . . . . . R = C6H5 . . . . . . . . . . . . . . . . . . R = C6H4OCH3 .. . . . . . . . . . . . . R=p.C6H4 NO2 . . . . . . . . . . . . . . 2-Chlorophenethiazine . . . . . . . . . . . . . . Pyridine 2-aldoxime . . . . . . . . . . . . . . Picolinaldehyde 2-hydroxy-5-phenylanil and . . . . . . chloride ion . . . . . . . . . . . . . . . . Chroma1 blue and cetyltrimethylammonium chloride . . Prochlorperazine maleate . . . . . . . . . . . . 5-Chloro-2-hydroxy-4-methylacetophenone oxime . . . . Mepazine hydrochloride . . . . . . . . . . . . Glycine thymol blue . . . . . . . . . . . . . . 2-Ethylthioisonicotinamide . . . . . . . . . . . . 1-(4-Bromophenylazo)naphth-2-ol . . . . . . . . 1-(3-Chlorophenylazo)naphth-2-ol . . . . . . . . Monothiotrifluoroacetylacetone . . . . . . . . . . Ethyl 3-phenyl-5-isoxazolone-3-carboxylate .. . . . . 5-(Chloromethyl)-4-selenohexahydropyrimidine-2- thione Phenylazobenzaldehyde oxime . . . . . . . . . . Chromotrope 2B . . . . . . . . . . . . . . . . Nitrosodibenzylaniline . . . . . . . . . . . . . . 2-Pyridyl-2-thienyl-fLketoxime . . . . . . . . . . Phenylmethylamidothionophosphonate . . . . . . . . l.R.4.4.6.Trimethy l.( 1 Hq4H)-pyrirnidine-2-thiol: R = phenacyl . . . . . . . . . . . . . . . . R = cyclohexyl . . . . . . . . . . . . . . . . R = propyl . . . . . . . . . . . . . . . . R = 2'-amino-4'-tolyl . . . . . . . . . . . . R = 6'-amino-5'-hydroxy-3'-mercaptopyrimidyl . . . . Sodium ethyl trithiocarbonate . . . . . . . . . . Bis(diarylthiophosphory1)disulphide . . . . . . . . 3.(4,5.Dimethyl.2.thiazolylazo). 2.niethylresorcinol . . 4.8.Diamino.l.5.dihydroxyanthraquinone.2.6. disulphonate . . . . . . . . . . . . . . . . Furoin thiosemicarbazone . . . . . . . . . . . . Isonitrosobenzoylacetone . . . . . . . . . . . . 2'-Hydroxy-3'-bromo-4-methoxy-5-methylchalcone oxime Picolinaldehydep-nitrophenylhydrazone . . . . . . Quinoline-2-aldehyde thiosemicarbazone . . . . . . 2-Pyridyl-2-thienyl Z-ketoxime . . . . . . . . . . 2.Hydroxy.3. 5.dimethylacetophenone oxime . . . . . . 2.2 '.Diquinolyl ketone 2-pyridylhydrazone . . . . . . Ethylenedithiodiacetic acid . . . . . . . . . . . . trans.Cyclohexane.1 2.diamine tetraacetate . . . . . . Ethyl benzoylisonitrosoacetate . . . . . . . . . . Solochrome Red B . . . . . . . . . . . . . . Sodium-2-[4-amino-3-( 1.2.4.triazolylazo) 1. napht h- 1 -01-4-sulphonate .. . . . . . . . . . . Methiomeprazine hydrochloride . . . . . . . . . . Thiocyanate and methylene blue . . . . . . . . . . 3-Bromo-2-hydroxy-5-methylacetophenone hydrazone . . Biacetyl monoxime glycinimine . . . . . . . . . . 20-2oc 5.3-20 5.3-16 5.3-16 5.1-15 3.6-15 4.0-14.8 0.366.4 3.0-3.5 10-200 Fg/lO ml 0.08-1.4 0.2-12 d 18 s 1 2 620 1-15 20.97 21.48 <6 3-30 aqueous medium; 2.7-27.4-methyl-pentan-2-one solution 1-7.7 2-1 6 G5.85 0.005-1 0.15-2.1 1-200 4.2-1 3.3 6.0-34.7 a20.5 <23 <1? s 6 . 2 4-30 0 .O& 0.76 4.8-18.2 10-130 pg/lO ml 1.4-42 3-9 2.5-20 0.2-2 2-70 0.25-5 0.4-4 10-80 10-110 Ug'10 ml 0.1-1 2-4 1.06-8.52, strongly acidic medium; 1.06-7.44, neutral and and alkaline media 0.4-21 0.03-0.5 3.77-27.32 2-9 3.9 x 10*/390 nm" 3.7 x 103/330 nm 4.2 x 104'430 nm 4.6 X 10-;/43Unrn 3.8 x 10V430 nm 4.4 x 103/430nm 5 x 103/430nm 4.63 x 1OV525 nm 2.34 x 10Y420 nm 4.9 x 103/62? nm* 1.01 x 10V670 nm 4.63 X 193i480 nm 2.8 X 103/400 nm" 4.46 x 10.Y510 nm 7.8 x 1O350O nm 5 x 103/470 nm 1.29 x 10%510 nm 1.25 x 104/500 nm* 6.3 x 10U28 nm* 3.5 x 10'/370nm* 3.9 x 103/370 nm* 1 x 104/340nm 3.6 x lO3/55O nm 1.23 x 104/610 nm 1.1 x IOV522 nm* 9.18 X 1031433 nmx 4 x 10j/400nm* 4 X 103/420 nm* 3.1 x 10V420 nm* 4.2 x 10V430 nm 5.3 x 103/440nm 5.6 x 103/380 nm 1.2 x 10Y370 nm 3.25 x 104/315 nm" 1.46 x 1OJ/56O nm 4.37 X 103/720 nm 1.98 x 1O4i36O nm 1.02 x 104!405 nm* 3.2 x 1031390 nm* 9.5 x l03/48Onm 2.6 X 103/510 nm* 9.18 x 103/433 nm' 1.08 X 103/400 nm* 1.95 x 104/624 nm* 2.01 x 104/290nm 1.05 x 10V340 nm 1.21 x 104/410 nm* 1.2 x 104/488 nm 7 x 10%70 nm 1.2 x 104540 nm 3.6 x 103/480 nm 1.7 x 10V660 nm? 4.5 x 10~/400 nm* 2.1 x 1oj/380 nm* Molar ratio 2 3 1 .3 1 2 1 4 1 3 1 3 1 3 1 4 1 :4 1 l 1 2 1 l l 1 3 1 2 1:2 1 l 1 1 1 l 1 2 1 2 1 l 1 4 1 4 1 2 1 l 1 2 1 3 1 2 1 l 1 2 1 2 1 4 1 :J 2 3 1 2 1 3 1 l I 1 1:: 1 2 1 1 1 l 1 2 1 l 1 2 1 l 1 2 1 l 1:2 1 l 1 1 1 2 1 l 1:2 1 2 1 2 Ref . This work This work 9 10 10 10 10 10 10 11 12 13 14 15 16 17 18 19 20 20 21 22 22 23 24 25 26 27 28 29 29 5 5 5 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 46 47 48 49 50 * By solvent extraction .t Flotation . spectrophotometric method ANALYST. JANUARY 1984. VOL. 109 51 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. References Morelli. B Analysr 1982 107 282. Morelli B Analyst 1983 108 386. Morelli B. Analyst 1983 108. 870. Morelli B. Analyst 1983 108 959. Nath D. Singh A. K. Katyal. M. and Singh R. P. Indian J. Chem. Sect. A 1978. 16 457. Balatron M. G. Pavon J. M. C. and Pino F. Talanta 1979, 26 71.Yamaguki A. Penland R. B. Mizushima S . Lane T. J., Curran. C. and Quagliano J. V J. Am. Chem. SOC. 1958.80, 527. Murphy R. J. and Svehla G. Anal. Chim. Acta 1978 99. 115. Paria P. K. and Majumdar S. K. Fresenius 2. A n d . Chem., 1976 279 207. Singh. A. K. Katyal M. Bhatti A. M. and Ralhan N. K., Talanra 1976 23. 337. Sanke Gowda H. and Thimmaiah K. N. Fresenius Z . Anal. Chem. 1976 279 208. Manku G. S . Curr. Sci. 1976 45 256. Otomo M. and Kodama K. Anal. Chim. Acta 1976,83,275. Uesugi K. A n d . Chim. Acta 1976 84 377. Sanke Gowda H. and Ramappa P. G. Indian J . Chem. Sect. A 1976 14 454. Lal K. and Gupta S. P. Curr. Sci. 1976 45 722. Sanke Gowda H. and Keshavan B. J. Indian Chem. SOC., 1976 53 688. Babkina T. A. and Rigin V. I. Zh.Anal. Khim. 1976 31, 2265. Sikorska-Tomicka H. and Lewicka M. Zesz. Nauk. Poli-tech Bialostockiej Mat. Fir. Chem. 1976 2 89 (pub. 1977). Balyuta I. G. Smirnov P. P. and Khvatkova 2. M. Zh. Anal. Khim. 1977 32,2398. Tarafdar S. A. Nucl. Sci. Appl. Ser. B 1977 10 102. Corigliano F. Di Pasquale. S. and Ranieri A. Analyst 1977, 102 25. Apostolescu M and Golgotiu T. Rev. Chim. (Bucharest), 1978 29 1077. Mahgoub A. E. Darwish N. A. and Skoukry M. M., Analyst 1978 103 879. Khalifa H. and Issa Y. M. Egypt. J. Chem. 1975 18 1057 (pub. 1978). Kothny E. L. Mikrochim. Acra 1978 1 425. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. Beaupre P. W. and Holland W. J. Mikrochim.Acta 1978,2, 327. Pilipenko A. T. Danilova V. N. Suleimanova M. G. and Shokol V. A . Zavod. Lab. 1978 44 927. Jain P. Katyal M. Singh R. P. and Ralhan N. K J. Inst. Chem. Calcutta 1978 50 217. Rao A. L. J. and Shekhar C. Indian J. Chem. Sect. A 1978, 16 463. Shishkov A. N. and Malakova Hr. G. Talanta 1978 25, 533. Sanchez-Pedreno C. Gonzalez Diaz V. and Polo Conde F., An. Quim. 1978 74 1502. Navas A . and Garcia Sanchez F. Afinidad 1979 36 161. Bhaskare C. K. and Devi S. Tulanta 1979,25 544. Desai B. J. and Shinde V. M. Fresenius 2. Anal. Chem., 1979 295 412. Deshmukh B. K. and Kharat R. B. J. Indian Chem. SOC., 1979,56 213. Carrillo J. and Guzman M. An. Quim. 1979,75,550. Khasnis D. V and Shinde V. M. Talanta 1079 26 593. Beaupre P. W. and Holland W. J. Mikrochim. Acta 1979,1, 279. Jetley U. K. Singh J. and Rastogi S. N. Chem. Era 1979, 15 23. Beaupre P. W. Holland W. J. and Sieler R. A. Mikrochim. Acta 1979 2 479. Napoli A. Ann. Chim. (Rome) 1979 69 399. Ezerskaya N. A. Solovykh T. P. Bochkova L. P. and Shubochkin L. K. Zh. Anal. Khim. 1980 35 81. Jagdale M. H. Desai B. J. and Shinde V. M. Mikrochim. Acta 1980 1 353. Elsirafy A. A. Analyst 1980 105 912. Mukherjee S . Garg B. S. and Singh R. P. Ann. Chim. (Rome) 1980 70,481. Sanke Gowda H. Padmaji K. A. and Thimmaiah K. N., Analyst 1981 106 198. Marczenko Z. and Jarosz M. Analyst 1981 106,751. Lal K. and Malhotra S. R. Ann. Chim. (Rome) 1981 71, 479. Riyazuddin P. Talanta 1982 29 1122. Paper A31177 Received June 20th I983 Accepted August 22nd I98

ISSN:0003-2654    

DOI:10.1039/AN9840900047

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984, VOL. 109 53 T h a I I i m et ri c Oxid at i o n s Part 111." Determination of Glyoxylic Acid and Binary Mixtures of Glyoxylic, Oxalic and Formic Acids by Photochemical Methods S. R. Sagi, K. Appa Rao and M. S. Prasada Rao lnorgan ic Chemistry Laboratories, Andhra University, Wa Ita ir 530003, India Glyoxylic acid is oxidised photochemically to carbon dioxide and water by thallium(lll) in the presence of Br- as catalyst. Based on this, a convenient photochemical redox method is described for the determination of glyoxylic acid by titrating the thallium(1) formed bromimetrically in 1.5-2.0 M hydrochloric acid in the presence of Methyl Orange as indicator. Based on the fact that the photochemical oxidations of glyoxylic acid, oxalic acid and formic acid proceed differently in the presence of Br-, CI- and Mn2+ and in combination, methodsare described for the analysis of binary mixtures of these acids by titrating the thallium(1) formed under different conditions.Keywords: Glyoxylic acid determination; glyoxylic, oxalic and formic acid analysis; photochemical thallimetric oxidations; bromimetric titration Glyoxylic acid (oxoethanoic acid) is generally determined by spectrophotometric methods at the microgram level with an accuracy of about 5% ,1-4 but few methods have been reported for its determination at the millimole level. Salzers deter- mined glyoxylic acid by treating it with a known excess of semicarbazide hydrochloride, treating the excess of unreacted semicarbazide with potassium iodate and titrating the liber- ated iodine with sodium thiosulphate.Ronzio6 mentioned that glyoxylic acid can be determined by oxidation with Tollen's reagent. We have already utilised thallium(II1) as a potential photochemical oxidimetric reagent and described convenient methods for the determination of formic acid (methanoic acid) and oxalic acid7 (ethanedioic acid). The reaction between thallium( 111) and glyoxylic acid has now been investigated and some convenient titrimetric methods are described for the determination of pure glyoxylic acid and the analysis of binary mixtures of glyoxylic, oxalic and formic acids. These methods should be of use in the analysis of the products of the synthesis of glyoxylic acid from oxalic acid and in the assay of biological degradation products.Experimental and Results Reagents Thallium( 111) hydroxide was prepared as reported earlier* and dissolved in a suitable amount of either perchloric or sulphuric acid. The thallium content was determined iodimetrically9.10 and verified by other redox methods using oxalic acid8 and tin(I1)" as reducing agents. Glyoxylic acid was prepared by dissolving a suitable amount of the BDH product in water and determined by the method of Salzer.5 All other reagents were of analytical-reagent grade. Apparatus Although the photochemical reaction described here is much faster under direct sunlight. as the intensity of sunlight is variable all the work was carried out with a Philips high- pressure mercury vapvur lamp (200-250V, 125 W) as a daylight source. The solutions were irradiated in colourless glass containers by placing them at a distance of 0.15 m from the lamp. * For Part I1 of this series, see reference list, p.55. Photochemical Reaction Between Thallium(II1) and Glyoxylic Acid Separate studies have shown that glyoxylic acid does not interfere in the titration of thallium(1) with bromate12-14 using Brilliant Ponceaux 5R or Methyl Orange as indicator. This reaction was therefore used for monitoring the progress of all the reactions in this investigation. In 1~ perchloric acid medium the oxidation of glyoxylic acid is complete in 6 h under light from a mercury vapour lamp when thallium(II1) is present at an 8-fold excess or more of the amount required for quantitative oxidation. Any decrease in the concentration of thallium(II1) or increase in the acid concentration in the reaction mixture increases the time required for the complete oxidation of glyoxylic acid.The stoicheiometry of the reaction may be represented as Unless the thallium(II1) concentration is very low it is hydrolysed in media that contain less than 1 M perchloric acid. By maintaining the acid concentration at I M and the thallium(II1) concentration at eight times that required for complete oxidation, the effects of halide ions on the photo- chemical reaction were studied. These studies confirmed that fluoride ions have no effect whereas chloride ion increases the time required for complete oxidation. The results obtained on the effect of bromide ions on the reaction are shown in Fig. 1. It is clear that when the bromide concentration is half that of thallium(II1) the reaction is the fastest, and it is also reasonably fast up to a ratio of thallium(II1) to bromide of 1 : 1.An increase in acid concen- tration has a negative effect on the bromide ion catalysis. The effects of perchloric acid and thallium( 111) concentra- tions on the reaction were studied by maintaining the bromide concentration at half that of thallium(II1); the results are given in Table 1. It is evident that the reaction is complete in 30 min in a medium containing about 1 M perchloric acid even in the presence of double the amount of thallium(II1) required for complete oxidation. HOOC-CHO + 2 Tl(II1) + H20 --+ 2 Tl(1) + 2C02 + 4H+ Recommended Procedure for the Determination of Glyoxylic Acid To an aliquot containing 0.025-0.25 mmol of glyoxylic acid in a 200-ml beaker add 20 ml of 1.0 M perchloric acid, 1.0 mmol of bromide and 2 mmol of thallium(II1) and dilute to 100ml with water.Stir the solution and expose it to light from a54 ANALYST, JAKUARY 1984, VOL. 109 Table 1. Effect of thallium(II1) and perchloric acid concentration on the rate of photochemical oxidation of glyoxylic acid (0.05 mmol) with thallium(II1). Volume of the reaction mixture = 100ml. Bromide ion concentration is half that of thallium(II1). Results given are times for completion of the reaction (minutes) Amount of thallium(II1) taken/mmol [HC104)/ M 0.2 0.3 0.4 0.5 0.6 0.8 1 .0 1.2 1.4 1.5 0.2 15 12 12 10 10 10 10 10 10 10 0.5 20 16 15 14 14 14 14 14 14 13 1 .o 30 25 25 22 20 20 20 20 19 19 2.0 40 35 32 30 30 30 28 26 25 25 3.0 150 120 110 100 90 75 75 75 70 70 / H 4 0.2 0.5 1 .o 2.0 3.0 Concentration of perchloric acid/M Fig.1. Effect of [H+] and [Br-] on the rate of photochemical oxidation of glyoxylic acid by thallium(II1). [Tl(III)]: [Br-1: A. 80 : 1; B, 40:l; C, 20:l; D, 1O:l; E , 2:l; F, 1:l; G, 1:2; and H, 1:3. Volume of reaction mixture. 100 ml; amount of thallium(III), 0.8 mmol; amount of glyoxylic acid, 0.05 mmol high-pressure mercury vapour lamp for 30 min (or keep it in bright sunlight for 5 min). Add 25 ml of concentrated hydrochloric acid and 0.1 ml of 0.1% Methyl Orange solution, heat to 60°C and titrate with 0 . 0 5 ~ potassium bromate solution (1.392 g 1-1) until the indicator is destroyed by bromination.15 The blank correction is negligible.This method is also applicable in dilute sulphuric acid medium. Typical results obtained by adopting this method showed that glyoxylic acid in the range 0.025-0.25mmol .can be determined satisfactorily with an accuracy of 0.3%. Analysis of Mixtures Bromide ions in small concentrations catalyse the photochem- ical oxidation of formic acid16 and oxalic acid7 by thal- lium(II1). We have already reported16 that Mn2+ ions drastically and chloride ions moderately inhibit the photo- chemical reaction between formic acid and thallium(II1). Whereas Mn2+ has very little effect, chloride ions catalyse the photochemical reaction between thallium(II1) and oxalic acid.’ Hence both formic acid and oxalic acid are expected to be photochemically oxidised by thallium( 111) in the presence of bromide.In the presence of sufficient Mn*+ and bromide, oxalic acid is expected to be oxidised, leaving formic acid unaltered. Further investigations have shown that under these conditions when the mixture is exposed t o light from a mercury vapour lamp for about 60 min, in addition t o oxalic acid about 2% of formic acid is also oxidised, which shows that Mn2+ when present alone does not completely prevent the photochemical oxidation of formic acid by thallium(II1) in the presence of bromide. Therefore. the reaction is carried out in the presence of chloride and Mn2+ ions, which completely stops the oxidation of formic acid and results in the quantita- tive oxidation of oxalic acid only. Oxidation of glyoxylic acid to carbon dioxide may proceed through either the formic acid or the oxalic acid stage.” If the reaction proceeds through the formic acid stage the presence of Mn2+ and chloride ions in the reaction mixture should decrease the rate of oxidation of glyoxylic acid to carbon dioxide and water, as these ions inhibit the oxidation of formic acid formed as an intermediate.The results of typical experiments showed that in 1 . 0 ~ perchloric acid and in the presence of an 8-fold excess of thallium(II1) over that required for complete oxidation of glyoxylic acid, the reaction proceeds to the formic acid* stage in 2 h and further exposure to the light from a mercury vapour lamp for a further 2 h leads to only about 10% of the oxidation of formic acid when the Mn2+ level is 1/16th that of thallium(II1) and 5% of the oxidation when the Mn2+ level is five times that of thallium(II1).A further increase in Mn2+ concentration retards the oxidation of glyoxylic acid to the formic acid stage. Because chloride also retards the reaction between thal- lium(II1) and formic acid, further studies were carried out on the oxidation of glyoxylic acid by thallium(II1) with a view to completely stopping the reaction at the formic acid stage. These studies revealed that the reaction completely stops at the formic acid stage when a minimum of one equivalent of chloride and a 3-fold excess of Mn2+ with respect to thallium(II1) are present. Further, under these conditions the oxidation of glyoxylic acid to the formic acid stage is catalysed by chloride ions. Based on these facts, the following conclusions can be drawn: (1) glyoxylic acid is oxidised only to the formic acid stage by thallium(II1) in the presence of Mn2+ and C1-; (2) oxalic acid can be photochemically oxidised by thallium(II1) in the presence of chloride7 and the effect of Mn2+ on this reaction is negligible; and (3) glyoxylic acid, formic acid16 and oxalic acid7 are oxidised to carbon dioxide and water photochemically by thallium(II1) in the presence of bromide.*The chromotro ic acid test is used for the detection of formic acid,’s but thalliumflI1) interferes. Hence, thallium(II1) is reduced to thallium(1) photochemically19 in the presence of excess of oxalate. Thallium(1) and oxalate do not interfere in the spot test. Such a modified test has shown the presence of formic acid.ANALYST.JANUARY 1984. VOL. 109 55 Table 2. Conditions for the analysis of binary mixtures of glyoxylic. oxalic and formic acids. Total volume of reaction mixture = 100ml. Concentration of perchloric acid = 1 M. The thallium(1) formed is determined bromimetrically as described earlier Composition of Catalyst/mmol Time of the mixture, 0.05- Tl(II1); exposure/ 0.25 mmol each mmol CI- Br- Mn,+ min Oxidations effected Amount of formed/ mmol* TI(I) 1. Glyoxylic + Glyoxylic acid to formic acid and Both glyoxylic and formic acids to Glyoxylic acid to formic acid and Both glyoxylic and oxalic acids to Oxalic acid to C02 + H20 and . Both oxalic acid and formic acid to formic . . . . . . 2.0 4 - 6 120 formic acid unaffected a 2.0 - 1 - 30 C0,+H20 b oxalic .. . . . . 2.0 4 - 6 120 oxalic acid to C02 + H20 C 2.0 - 1 - 30 CO2+H,O d formic . . . . . . 1.5 3 - 4.5 30 formic acid unaffected e 2. Glyoxylic + 3 . Oxalic + 30 CO,+H,O f 1.5 - 0.75 - * Calculations: Mixture 1: amount of glyoxylic acid = a mmol; amount of formic acid = ( b - 2a) mmol. Mixture 2: amount of glyoxylic acid = (d - c ) mmol; amount of oxalic acid = (2c - d ) mmol. Mixture 3: amount of oxalic acid = e mmol; amount of formic acid = (f - e) mmol. Table 3. Analysis of binary mixtures of glyoxylic, formic and oxalic acids Amount of acid (A)/ mmol Amount of acid (B)/ mmol Constituents of the mixture (A + B) Glyoxylic acid (A) and formic acid (B) . . Taken 0.0500 0.2500 0.1000 0.2500 0.0500 0.2500 0.1000 Glyoxylic acid (A) and oxalic acid (B) .. . . 0.0500 0.2500 0.1000 0.2500 0.1000 0.1500 0.2500 Formic acid (A) and oxalic acid (B) . . . . 0.0645 0.0645 0.0994 0.1290 0.1290 0.1780 0.2500 . . Found 0.0500 0.2495 0.0995 0.2505 0.0500 0.2490 0.1002 0.0500 0.2498 0.0998 0.2505 0.1003 0.1498 0.2495 0.0640 0.0640 0.0995 0.1285 0.1285 0.177” 0.2505 Taken 0.2500 0.2500 o.isoo 0.1500 0.1000 0.1000 0.0500 0.0500 0.1500 0.1000 0.2500 0.2500 0.2000 0.1750 0.1350 0.0525 0.1000 0.1975 0.0525 0.2500 0.2500 Found 0.2505 0.2505 0.149s 0. 149s 0.1000 0.1005 0.0500 0.0500 0.1495 0.100“ 0.2495 0.2505 0.1950 0.1745 0.1360 0.0526 0.0995 0.1975 0.052” 0. 24g5 0.2495 The conditions for the analysis of binary mixtures of glyoxylic, oxalic and formic acids were then investigated and typical synthetic mixtures were analysed. The results are summarised in Tables 2 and 3.1. 2. 3. 4. 5. 6 . 7. 8. 9. 10. 11. References Dickens, F., and Williamson, D. H., Biochem. J . , 1958,68,84. Kramer, D. N.. Klein, N.. and Baselice. R. A., Anal. Chem., 1959,31,250. Pesez, M., and Bartos, J., Bull. Soc. Chim. Fr., 1960, 481. Josimovic, L., and Gal, O., Anal. Chim. Acta, 1966, 36, 12. Salzer, F., 2. Anal. Chem., 1955, 146, 260. Ronzio, A. R., Microchem. J . , 1957,1,59;Anal. Abstr., 1958, 5, 874. Sagi, S. R., Appa Rao, K., and Prasada Rao, M. S . , Can. J. Chem., 1983, 61, in the press. Sagi, S. R., and Ramana, K. V., Talanta, 1969, 16, 1217. Proszt, J., 2. Anal. Chem., 1928,73, 401. Kolthoff, I. M., and Belcher, R., “Volumetric Analysis,” Volume 111, Interscience, New York, 1957, p. 370. Sagi, S. R., and Prasada Rao, M. S . , Talanta, 1979, 26,52. 12. 13. 14. 15. 16. 17. 18. 19. Kolthoff, I. M., Red. Trav. Chim. Pay-Bas, 1922,41, 172. Zinti, E., and Rienacker, G., 2. Anorg. Allg. Chem., 1926, 153, 276. Rienacker, G., and Knanel, G., 2. Anal. Chem., 1947, 128, 459. Metters-Tuladhar, C. H., and Ottaway, J. M., Anal. Chim. Acta, 1973, 66, 291. Sagi, S. R., Appa Rao, K., and Prasada Rao, M. S . , Talanta, 1983, 30,282. Sengupta, K. K., and Bhattacharya, S . D., 2. Phys. Chem. (Leipzig) 1969, 240, 279. Feigl, F., “Organic Analysis,” Volume 11, 4th Comprehensive Revised Edition, Elsevier, Amsterdam, 1954, p. 241. Sagi, S. R., Prakasa Raju, G. S., Appa Rao, K., and Prasada Rao, M. S., Talanta, 1982, 29, 413. Note-References 7 and 16 are to Parts I1 and I of this series, respectively. Paper A311 75 Received June 20th, 1983 Accepted August lst, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900053

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984, VOL. 109 57 Electrochemical Reduction of Ketoprofen and its Determination in Pharmaceutical Dosage Forms by Differential-pulse Polarography Lawrence Amankwa and Leslie G. Chatten 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 ketoprofen in dosage forms. Methanol was employed as the solvent for extracting the drug from the formulations, and Britton - Robinson buffer (pH 6.0) containing 5% V/Vmethanol was used as the supporting electrolyte. The EP occurs at -1.15 V. Results obtained by the proposed method are in excellent agreement with those provided by the manufacturer‘s laboratory. Commonly used tablet excipients and suppository bases were found not to interfere.A mechanism for the electrochemical reduction of ketoprofen at -1.25 V is proposed that involves the transfer of two electrons. Keywords: Ketoprofen determination; differential-pulse polarograph y; controlled-potential coulometry; cyclic voltammetry Ketoprofen, 2-(3-benzoylphenyl)propionic acid, possesses analgesic, anti-inflammatory and antipyretic activities that make it valuable in the treatment of rheumatoid arthritis and osteoarthritis.1.2 It is formulated either as a 50-mg capsule or as a 100-mg suppository. 0 Ketoprofen The manufacturer’s method for the determination of the substance in dosage forms involves extraction of the active ingredient from the tablet mass into methanol, followed by UV spectrophotometric measurement at 254 nm.3 Reported methods for its determination in biological samples include HPLC,4-6 TLC or paper chromatography,7.s GLC after derivatisation,”10 spectrophotometry,7.11 radioisotope trac- ings and colorimetry or polarography.9 Most of these methods are either tedious or too cumbersome to be adapted for the routine analysis of dosage forms.As yet, no official method has appeared in the USPI2 for the determination of ketopro- fen. In this paper, a differential-pulse polarographic method is presented for the assay of the active constituent in dosage forms. Only a single extraction is required prior to the polarographic analysis and the method is accurate, sensitive and easy to apply to routine usage. Experimental Apparatus and Conditions for Polarographic Analysis A Fisher Model 320pH meter fitted with a glass-calomel electrode system was employed to measure the pH values of the solutions. A PAR Model 174 polarographic analyser equipped with a drop timer (Model 172A) and a Houston Model 2000 Omnigraphic recorder were used.A three-electrode combina- tion was employed, which consisted of a saturated calomel electrode, a dropping-mercury electrode (DME) and a platinum wire as the auxiliary electrode. A conventional H-type cell was maintained at 25 f 1 “C and all sweeps utilised a scan rate of 2 mV s-1 and a drop time of 2 s. In Britton - Robinson buffer (pH 6.0) the instrumental parameters were as follows: applied potential range, -0.8 to -2.3 V; current, 50 pA full-scale; height of mercury column, 75 cm; flow-rate of mercury, 1.29 mg s-1; modulation ampli- tude, set at 50 mV; and low pass filter, set at a time constant of 1 s.The instrument was operated in the differential-pulse mode. 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 Britton - Robinson buffer, pH 6.0. was placed in the coulometric cell on top of a 5-ml layer of triply distilled mercury and 1 ml of a 10-2 M solution of ketoprofen in methanol was added. The system was purged for 10 min with purified nitrogen. The applied potential was set at - 1.25 V with a current range of 1 mA full-scale and the solution was electrolysed until the digital readout indicated a constant but small current count.The electrolysis time was 40min. The process was repeated with a blank consisting of 19ml of Britton - Robinson buffer 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 & G Model 175 universal program- mer, 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 Britton - Robinson buffer (pH6.0) the parameters were as follows: potential range, -0.8 to -1.4V; current range, 10pA; and scan rate, varied from 10 to 200mVs-1.In a dimethylformamide - tetraethylammonium bromide system, the potential range was -0.5 to -2.2 V and the current range and scan rates were as in the previous system. Reagents The following reagents were used, all of analytical-reagent grade: barbital, boric acid, citric acid, tetraethylammonium bromide, potassium hydrogen phosphate, dimethylformam- ide, anhydrous methanol, 0.2 N sodium hydroxide solution and 1% tetraethylammonium bromide solution in dimethyl- formamide. Britton - Robinson buffers were prepared with58 ANALYST. JANUARY 1984, VOL. 109 distilled, de-ionised water at intervals of 1 pH unit over the pH range 5.0-11 .O. Reference Standard Ketoprofen (99.8%) was available from Rhone-Poulenc Pharma Inc..Canada, Ltd., and was used without further purification. pH Dependence Studies These studies were carried out in Britton - Robinson buffer over the pH range 5.0-11.0. Preparation of Calibration Graphs A stock solution of ketoprofen (113-2~) was prepared in anhydrous methanol. Five test solutions of varying concentra- tions, from 1 to 5 x 1 0 - 4 ~ . were prepared by appropriately diluting the stock solution with Britton - Robinson buffer (pH6.0). In the total sample volume of exactly 20m1, the amount of methanol was always maintained at 1 ml. All samples were purged with purified nitrogen for 10 min prior to each scan and a stream of nitrogen was allowed to flow gently over the surface of the solution during the electroreduc- tion. Samples of each of five concentrations were run five times and resulted in a correlation coefficient for the graph of 0.999 4.Diffusion Dependence Studies These studies were carried out at pH 6.0 on a 5 x 1 0 - 4 ~ solution of ketoprofen. The applied potential was set at - 1.15 V and the height of the mercury column ranged from 60 to 80cm. The mass of mercury was also obtained at each of five heights over that range. Analysis of Pharmaceutical Dosage Forms Two dosage forms, 50-mg capsules and 100-mg suppositories, were available from the manufacturer. Analysis of Capsules The contents of 20 capsules were transferred quantitatively into a tared 100-ml beaker and weighed. The powder was mixed thoroughly and an amount of powder corresponding to the mass of one capsule content was accurately weighed into a 100-ml beaker.Methanol (20ml) was added and the sample stirred magnetically for 15 min. The mixture was transferred quantitatively into a 50-ml calibrated flask, diluted to volume with methanol and then filtered through a Whatman No. 1 paper, discarding the first 5 nil of the filtrate. A 1-ml volume of the filtrate was transferred into the polarographic cell, 19 ml of Britton - Robinson buffer (pH 6.0) were added and the solution was purged for 10 min with nitrogen prior to recording the polarogram. The amount of ketoprofen was calculated by the direct comparison method. using a reference standard solution of ketoprofen (0.393 7 x 10-2 M). Content Uniformity Test Ten capsules were randomly selected from the sample. The content of each capsule was quantitatively transferred into an individual 100-ml beaker, 20ml of methanol were added and the mixture was stirred magnetically for 20min.After transferring the mixture quantitatively into a 50-nil calibrated flask, the assay was continued as described in the previous section. Analysis of Suppositories Ten suppositories were randomly selected and weighed. Each was placed in an individual 150-rnl beaker, 50 ml of methanol were added and the mixture was warmed until complete dissolution was achieved. The mixture was stirred magnetic- ally for 10 rnin while still warm and then cooled in a refrigerator for 40 min. The mixture was allowed to stand for 10 min at room temperature, stirred with a glass rod to form a slurry, transferred quantitatively into a 100-ml calibrated flask and diluted to volume with methanol.After filtering through a Whatman No. 1 paper, a 1-ml aliquot of the solution was treated in the manner described under Analysis of Capsules. Macro-scale Electrolysis and Isolation of the Reduced Product The procedure for electrolysis was similar to that for controlled-potential coulometry except that the cell contained 180mg of ketoprofen in 50ml of methanol to which were added 20ml of 3 N sodium chloride solution as supporting electrolyte. The pH was adjusted to 6.0 by the dropwise addition of 0.2 N hydrochloric acid. The applied potential was set at -1.25 V and the reduction time required 4 h. The solution was acidified with 0.1 N sulphuric acid and then extracted with chloroform. The chloroform layer was washed with distilled water and dried over anhydrous magnesium sulphate.The final solution, which was concentrated to about 0.5 ml. was cooled in a refrigerator overnight and yielded 97 mg of compound X, a white crystalline powder that melted over the range 113-115°C. UV (MeOH): h,,,, 212nm. log E = 4.37: 252 nm, log& = 4.03. IR (KBr): 3 420 cm-1, 3 000, 2700, 1720, 1610, 1460, 1240, 1020, 900cm-1. NMR (60 mHz, perdeuterated dimethyl sulphoxide, DMSO-d6): 6, 7.3 (m, 9H). 5.6 (m, 2H), 3.6 (9, 1H) and 1.3 (d, 3H) (see Results and Discussion). Results and Discussion Ketoprofen . exhibits three polarographic waves in Britton -Robinson buffer over the pH range 5.0-11.0. At pH 6.0 a single, well resolved. pH-dependent wave is observed with an EA value at -1.15 V (curve A , Fig.1). The E+ shifts cathodically while the peak current decreases with increasing pH. The second wave is independent of pH and is observed within the pH range 6.5-9.0 (curve B, Fig. 1). The Eb value occurs at -1.39 V. At pH lU.0 the two waves merge to form a I T 17.0 LLA I 1 -0.9 -1.0 -1.1 -1.2 -1.3 -1.4 -1.5 -1.6 Applied potential, NV Fig. 1. Effect of pH on the d.p. polarographic wave of ketoprofen ( 5 X M) in Britton - Robinson buffer. A, pH 6.0; B, pH 7.0; and C, pH 11.0ANALYST, JANUARY 1984. VOL. 109 59 well resolved third wave, which is also pH dependent as evidenced by the E4 value. which shifts cathodically with increasing pH. At pH 11.0 the E3 value is -1.44 V (curve C, Fig. 1). The electrochemical processes at the DME giving rise to these waves have been discussed earlier.13 In addition, Zuman14 and Rifi and Covitzl5 have offered comments on the electrochemical reduction of carbonyl-containing compounds. The graph of diffusion current versus the square root of corrected height of mercury column is a straight line but does not pass through the origin. In Britton - Robinson buffer (pH 6.0) only one cyclic voltammetric reduction peak was ob- served, exhibiting an Epc at - 1.23 V (Fig.2). A peak potential shift of 60 mV was observed when the scan rate (V) was varied from 200 to 10 mV s-1. The graph of i/?& vs. 9 (where i is the current) is independent of scan rate, indicating a non- complicated homogeneous reaction that is also predominantly diffusion controlled. In dimethylformamide - tetraethylammonium bromide solu- tion, ketoprofen exhibits two cathodic and one reverse anodic cyclic voltammetric peaks.The Epc values occur at -1.66 and - 1.92 V and the EPa value is - 1.76 V (Fig. 3). The heights of all peaks decrease in the same proportion with decreasing scan rate. On continuous scanning at 200mVs-1, however, the A I \ I -0.9 -1.0 -1.1 -1.2 -1.3 -1.4 Applied potential, N V Fig. 2. Cyclic voltammogram of ketoprofen (5 X 1 0 - 4 ~ ) in Britton -Robinson buffer, pH 6.0 -0.3 -1.1 -1.3 -1.5 -1.7 -1.9 -2.1 -2.3 Applied potential, NV Fig. 3. Cyclic voltammogram of ketoprofen ( 5 X lo-") DMF -TEAB I first cathodic peak height decreases more markedly than that of the second cathodic or the anodic peaks. Coulometric analysis indicated that two electrons per molecule were transferred in the electrochemical process.The polarographic behaviour of ketoprofen at the DME has been found to be consistent with the reduction of a carbonyl group. In a DMSO-d6 - D20 system, the NMR peak at 6 S.6 is reduced to a singlet with an integration value corresponding to one proton. The strong IR peak at 3 420 cm-1 is indicative of the presence of a hydroxyl group in the reduced species. The absence of the carbonyl peak in the IR spectrum and the decrease in the UV absorbance peak at 252 nm indicate that the carbonyl group is the site of the electrochemical reduction. From all the foregoing observations, the reduced product of ketoprofen , compound X is probably 2-(3-benzhydrolyl)- propionic acid. X Both the differential-pulse polarographic peaks resulting from waves 1 and 3 are well resolved and may be utilised for the analysis of dosage forms; however, that of wave 1 is highly reproducible and was therefore chosen for the analysis of ketoprofen dosage forms.The peak current varies linearly with the concentration of the drug over the range 1 x 10-5 to 5 X M. Therefore, using the described cell and instrumen- tal settings. the method is sensitive to about 2.5 ug ml-1, which is well below blood concentrations reported for therapeutic dosages. Tables 1 and 2 provide the assay results for the two dosage forms investigated. Values obtained by the manufact- urer's quality control laboratory are also presented for comparison purposes. and excellent agreement is observed between the two results. Common tablet excipients and suppository bases do not interfere in the method, which is simple, sensitive and rapid. The method has the potential for application to stability studies in those instances where the decomposition process involves cleavage of the benzoyl carbonyl group, as the electrochemical process occurs specifically at that moiety.The method, or very simple variants of it, shows promise for general application to a series of similar arylalkanoic acid compounds associated with anti-inflammatory therapy. The authors gratefully acknowledge the ketoprofen sample , dosage forms and valuable information supplied by Rhone- Poulenc Pharm Inc., Canada, Ltd. Table 1. Assay of ketoprofen capsules by differential-pulse polarography in Britton - Robinson buffer (pH 6.0) - methanol Recovery stated by manufacturer Recovery by d.p.p.Capsule Label LotNo. claim/mg mg Y O mg * O/O * 247 50 50.07 100.1 49.7 99.4 5 0.1 * Average of five determinations. Table 2. Average value for the analysis of ten individual ketoprofen capsules and suppositories in Britton - Robinson buffer (pH 6.0) - methanol Label Recovery stated by Recovery by Capsule,Lot247 . . 50 97.6 99.7 2 1.3 Suppository, Lot 169 . . 100 99.8 101.5 k 3.6 Dosage form claim/rng manufacturers. YO d.p.p., YOANALYST, JANUARY 1984, VOL. 109 1. 2. 3. 4. 5. 6. 7. 8. References Reynolds, J. E. F., Editor, “Martindale, The Extra Pharmaco- poeia,” Twenty-eighth Edition, Pharmaceutical Press, Lon- don, 1982, p. 261. Brogden, R. N., Speight, T. M., and Avery, G. C., Drugs. 1974,8,168. Rhone-Poulenc Pharm Inc., Canada, Ltd., personal communi- cation. Jefferies, T. M., Thomas, W. 0. A., and Parfitt, R. T., J. Chrornatogr., 1979,162,122. Bannier, A., Brazier, J. I . , and Ribon, B., J . Chromatogr., 1978,155,371. Wpton, R. A., Buskin, J , N., Guentert, T. W., Williams, R. L., arid Riegelman, S., J. Chromatogr., 1980,190, 119. Ballerini, R., Cambi, A . , and Soldato, P. D., J. Phurm. Sci., 1977,66,281. Delbarre, F., Roucayrol. J. C., Amor, B., Ingrand, J . , Bourat, G., Fournel, J . , and Courjarat, J . , Scand. J . Rheumatol., Suppl., 1976,14,45. 9. Populaire. P., Terlain, B.. Pascal. S., Decouvelaere. B . , Lebreton, G., Renard, A . , and Thomas. P. J . , Ann. Pharm. Fr., 1973,31,679. Desager, P. J., Vanderbist, M.. and Harvengt, C.. J. Clin. Pharmacol., 1976,16, 189. Lott,R. B., Chim. Farm., 1975,114,351. “United States Pharmacopeia,” Twentieth Revision. Mack, Easton, PA, 1980. Chatten, L. G., Pons, S., and Amankwa, L., Analyst, 1983, 108,997. Zuman, P., “Topics in Organic Polarography,” Plenum Press, London, New York, 1970. Kifi, R. M., and Covitz, F. H., “Introduction to Organic Electrochemistry,” Marcel Dekker, New York, 1973, p. 167. 10. 11. 12. 13. 14. 15. Paper A311 17 Received April 25th, 1983 Accepted August 25th, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900057

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984. VOL. 109 61 Determination of Urea in Whole Blood Using a Urea Electrode with an lmmobilised Urease Membrane Kenji Yasuda, H. Miyagi, Y. Hamada and Y. Takata Branch Laboratory at Hitachi Research Laboratory, Central Research laboratory, Hitachi Ltd., 4026 Kuji-machi, Hitachi-shi, Ibaraki-ken, 3 19- 12, Japan A urea electrode using an ammonium ion-selective electrode with an immobilised urease membrane is described. The measurement of blood urea was carried out accurately without any pre-treatment of the samples. A conversion equation related to haematocrit was derived, which made it possible to determine plasma urea concentration from whole blood urea concentration. Rapid determinations of plasma urea concentrations were also made possible. Keywords: Urea electrode; immobilised urease membrane; unsegmented continuous flow analysis; whole blood urea determination Urea is the final product of human protein metabolism and its concentration in blood is an important index of renal function.Blood urea is determined frequently by the urease- indophenol colorimetric method.' However, the utilisation of a urea sensor with immobilised urease is becoming increas- ingly popular.2 Although there have been numerous reports on the determination of urea in sera or plasma, those in whole blood are relatively rare, e.g. references 3 and 4. Urea electrodes have an advantage over colorimetric methods that whole blood can be analysed directly. In an emergency, it is desirable for doctors to be able to determine blood urea by using whole blood samples.In this paper, a urea electrode that consists of an ammonium ion-selective electrode (neutral carrier type) and an immobili- sed urease membrane is described. An equation for calculat- ing the urea concentration in plasma from that in whole blood is derived and applied to whole blood samples from renal- failure patients. Experimental Apparatus The apparatus was assembled as shown schematically in Fig. 1. The electrodes were placed in a plastic cell with zig-zag flow-through tubing. The electrodes and the mixing coil were thermostated in a dry-air oven with a temperature controller. The buffer solution used as a carrier was stored in a water-bath adjusted to 37°C during measurement. All the tubing was made of Tygon. Colorimetric determination of urea was carried out with a Hitachi-400 Autoanalyser.r 10 11 I I waste Fig. 1. Schematic diagram of apparatus. (1) Buffer carrier; peristaltic pump; (3) depressurising bottle; (4) injection port; mixing coil; ( 6 ) electrode cell; (7) NH4+ electrode for background detection; (8) urea electrode; (9) reference electrode; (10) amplifier; and (11) recorder. 1.d. of the tubing, 1.0mm; distance between injection port and urea electrode, 1.5 m Preparation of Urea Electrode The ammonium ion-selective membrane was prepared by mixing 10 mg of nonactin, 300 mg of dioctyl adipate and 160mg of poly(viny1 chloride) (PVC) powder in 3ml of tetrahydrofuran and drying the mixture at room temperature for 24h.5 This membrane was cut into small disks (3mm in diameter). A disk was applied to the PVC electrode body with 3 yl of tetrahydrofuran.Ammonium chloride solution (0.1 M) was used as the internal filling solution and a silver-silver chloride wire as the internal electrode. The urease membrane was prepared by dipping aminosilan- ised unwoven cloth (ca. 30 ym thick, polyester) into the mixed solution of bovine serum albumin, urease and glutaraldehyde and drying it at room temperature. The mixed solution was prepared by adding 150 yl of 2.5% mlV glutaraldehyde solution to 500~1 of enzyme solution containing 50mg of urease and 15 mg of bovine serum albumin. In order to slow the gel formation, the mixture was cooled with ice-water. Urea electrodes were prepared by covering an ammonium ion-selective electrode with the urease membrane.Chemicals A powdered urease (Toyobo Co. , 17.2 U mg-1) was used for the preparation of the enzyme membrane. Other reagents for the enzyme membrane were bovine serum albumin (Sigma Chemicals, A4378) and glutaraldehyde (Wako Chemical Co., 25% aqueous solution). Nonactin was obtained from Sigma Chemicals. All other chemicals were of analytical-reagent grade. Urea standard solutions were prepared by diluting a stock solution (0.167 M urea) with phosphate buffer solution (20 mM Na2HP04 - NaH2P04, 1 mM EDTA disodium salt, pH 7.0). This buffer solution was also used as a carrier. As reference sera, Calibrate and Validate-A (General Diagnostics, Divi- sion of Warner-Lambert Co.) were used. Blood Samples Whole blood samples were collected from renal-failure patients. Heparin was used as an anticoagulant. Urea concen- trations in the blood from these patients were several times higher and the haematocrits were slightly lower than those of normal adults.Method of Measurement Undiluted, 20-yl samples were introduced into the unsegmen- ted carrier stream with a gas-tight microsyringe. Then the62 ANALYST, JANUARY 1984, VOL. 109 ammonium ions produced in the urease membrane were detected using a millivoltmeter. The signals showed sharp peaks and the peak height was measured. A calibration graph for urea was prepared in the same manner. This method employed unsegmented flow injection analysis6 and a urea electrode7 as basic techniques. Results and Discussion Optimum Conditions for Measurement The optimum conditions for urea determination were investi- gated with the present apparatus.Firstly, the effect of concentration of sodium phosphate buffer solution on sensi- tivity and linearity of the calibration graph was tested. Fig. 2 shows the increase in the e.m.f. response with decrease in concentration of buffer solution. However, the upper limit of linearity was at 10-'M urea. The lower limit of linearity increased with increase in buffer concentration. We chose a concentration of 20 mM sodium phosphate buffer solution to secure a wide linear range. The upper limit of linearity is probably due to a decrease in urease activity caused by an increase in pH in the urease membrane. In fact, the pH in the vicinity of the urease membrane surface increases consider- ably when urea is being hydrolysed in the membrane.Secondly, the effect of the pH of the carrier buffer solution on sensitivity was examined. The results are given in Fig. 3. The largest e.m.f. was obtained at pH 7.0, irrespective of urea concentration. I I I 1 - 1 1 5 10 30 50 100 [ U rea]/rn M Fig. 2. Effect of phosphate buffer (Na2HP04 - Na2H2P04, pH 7) concentration on calibration graph. (A) 6 mM; (B) 20 mM; (C) 60mM 100 80 60 > y: ui 40 E 20 a 5 6 7 8 PH Fig. 3. Effect of pH of the buffer carrier on the e.m.f. response of the urea electrode. Urea concentrations: (A) 107 mM; (B) 36 mM; (C) 11 mM; (D) 4 mM; (E) 2 mM. Carrier, 20 mM sodium phosphate buffer solution; flow-rate, 1 rnl min-I Thirdly, the effect of the flow-rate of the carrier stream was investigated. Fig. 4 illustrates the results.The dispersion (D) as defined by RfiiiEka and Hansen8 was 15.9 when 20yl of 10mM ammonium chloride solution was injected into the stream (flow-rate 1.0 ml min-1). This can be classified as large dispersion according to their definition. Urea in samples will also be diluted in the same proportion. However, with urea, both the residence time of urea in the membrane and the rate of conversion of urea into ammonium ions by urease must be considered. As the flow-rate increases, the e.m.f. correspond- ing to urea decreases and the difference between peaks corresponding to NH4+ and urea increases. Therefore, the diverse ions influence the e.m.f. of the urea electrode much more at higher flow-rates. Although the recovery time became longer, a flow-rate of 1 ml min-1 was chosen for the reason given above.Fourthly, the optimum oven temperature was investigated. The results are shown in Fig. 5. However, there was no clear optimum between 25 and 52"C, although a slight decrease in peak height above 40°C was observed with 11-54mM urea. The oven temperature was adjusted to 37 k 0.5 "C to match normal body temperature. An increase in temperature led to lower responses and shorter recovery times. Fig. 4. Effect of flow-rate on the e.m.f. response of (a) the urea electrode (4 mM) and ( b ) the NH4+ electrode (5 mM). Flow-rates: (A) 1 ml min-1; (B) 2 ml min-1; (C) 4 ml min-I 60 > E '? 40 E ui 1. 20 20 40 60 Oven temperature/"C Fig. 5. Effect of oven temperature on the e.m.f. response of the urea electrode. Urea concentrations: (A) 54 mM; (B) 25 mM; (C) 11 n M ; (D) 4 mhi.Carrier, 20 mM sodium phosphate buffer solution; flow- rate, 1 ml min-'ANALYST, JANUARY 1984, VOL. 109 63 effect of diverse ions was detected by using an additional ammonium ion-selective electrode to measure the background e.m.f. signal. The background signal detected can be conver- ted into NH4+ concentration by using a calibration graph for NH4+. Then the apparent concentration of NH4+ correspond- ing to the diverse cations is converted into a “urea” concentra- tion. This background “urea” concentration is subtracted from the apparent urea concentration detected by the urea electrode, and the true concentration is obtained. This correction technique proved to be effective, as shown in Table 2. The background levels were 1.0-1.1 mM urea in plasma and serum and cu.0.5 mM urea in whole blood. The background correction is indispensable for low-level determinations. Calibration Graph The measurement of blood urea was carried out under the optimum conditions: carrier, 20 mM Na2HP04 - NaH2PO4 buffer solution containing 1 mM EDTA; flow-rate, 1 ml min-1; pH, 7.0; and temperature, 37°C. The linear range of the calibration graph was 3.6-107mM (10-300 mg of urea nitrogen per 100 ml). The working range of the calibration graph was 1.8-54m~ (5-150mg of urea nitrogen per 100mlj. The coefficient of variation for 10 measurements of a 10.7 mM urea standard solution was 1.05%. To calculate the accuracy of the method, the reference sera Calibrate and Validate-A were used. The results are given in Table 1.Calculation of Urea Concentration in Plasma Blood samples can be analysed by colorimetric methods after the separation of blood cells. For an ”on-the-spot” or bedside measurement, speed has first priority. Therefore, the urea electrode, which can measure without any pre-treatment, is useful for this purpose. However, because of the compiled data on sera and plasma, it is desirable to express the data in terms of plasma concentration instead of whole blood concentration from the viewpoint of diagnosis. Hence an equation that readily converts urea concentrations in whole blood into those in plasma is needed. In this conversion, the presence of erythrocytes is troublesome because of the low water content. When a whole blood sample (volume V,) is taken for measurement, the urea concentration in plasma of the sample (C2) can be calculated from that in whole blood (Clj using haematocrit.In the first place, the presence of leucocytes and platelets is negligible because of their small volumes compared with erythrocytes. Then, suppose equilibrium between the urea concentrations in the aqueous phase of plasma (C,) and in the aqueous phase of erythrocytes (C,) is attained because of almost free diffusion of urea molecules through erythrocyte membrane: Use of Isotonic Buffer Carrier Haemoiysis tends to occur under hypotonic conditions, which causes an extremely high background signal owing to the diffusion of potassium ion from the inside of erythrocytes. Thus, an isotonic buffer solution was employed to prevent whole blood from haemolysing.‘The osmotic pressure of the carrier was adjusted to cu. 260mOsmol 1-1 by adding D-sorbitol as an inert, neutral additive at a concentration of 2 0 0 m ~ . The use of the isotonic buffer carrier caused no problems in the urea determination. Background Correction The signal peaks due to urea become apparently higher than the true peaks when samples contain diverse cations, because the ammonium ion-selective electrode has poor selectivity coefficients for them. In particular, the selectivity coefficient for NH4+ over potassium ion is approximately 10-1. Guilbault et al.9 reported that the interference level could be corrected using a buffer - potassium chloride standard. In this study, the Table 1. Results for reference sera by urea electrode (expressed with the background corrected) c, = CJ (1) Assuming that the solid phases of plasma and erythrocytes do not contain any trace of urea, then Urea concentration/mM Sample Urea electrode* Reference value’: Calibrate level1 .. . . . . 3.1 11 . . . . . . 13.2 I11 . . . . . . 29.4 Validate-A . . . . . . 17.4 * Single determinations. 100 b c, = c4 =-c, 3.2 14.6 28.3 16.9 where b is the percentage of water in the plasma. The haematocl’it (Hi, YO) of the sample is given by (3) v2 Ht= - 100 VI where V2 is the total volume of erythrocytes in the sample. the following car1 be written: Considering the total amount of urea existing in the sample. Table 2. Necessity for background correction in urea determinations in blood samples Urea concentrationimM (4) Urea electrode (urea + Sample* background) Serum A .. . . 4.6 B . . . . 7.0 c . . . . 5.s D . . . . 8.6 E . . . . 8.2 F . . . . 6.3 D . . . . 7.5 E . . . . 6.8 F . . . . 4.8 Plasma Whole blood Colorimetry (Urease - indophenol) NH,‘ electrode (background) where a is the average percentage of water in erythrocytes. By dividing both sides of equation (4) by V1 and substituting equation (3j into (4), we obtain 1.1 1.1 1.1 3.6 5.7 4.6 Ht c4u Ht c, = (1 - -) c2 +-.- 100 100 100 1 .0 1.1 1 .o 7.1 0.8 5.0 Substituting equation (2) into (5): (7. I)? (6.8)t (5.Oj: 0.6 0.5 0.4 Finally, n (7) CI c?_ = Ht U * Samples A-C were collected from normal adults and D-F from patients who had no problems with urea excretion. t These values are of concentrations in plasma. 1 -- ( I --) b 10064 ANALYST, JANUARY 1984, VOL.109 0 10 20 30 40 Haernatocrit, % Fig. 6. Correlation between haematocrit concentration and C2/CI ratio. Blood samples were collected from 38 rena.1-failure patients. The broken line represents the calculated CJC, ratio using equation 7 50 5 E =;. 40 -0 5 E ‘g 30 It! lJY - - m 3 Y 20 20 30 40 50 [Urea] (colorimetry)/rnM Fig. 7. Correlation of two methods for urea determination in plasma from 38 renal-failure patients. The concentrations in plasma deter- mined by the urea electrode were originally obtained as those in whole blood; y = -2.752 + 1.086x, r = 0.991 is obtained. Unless the dilution of blood after its introduction into the carrier stream alters the equilibrium of equation (l), then equation (7) still holds.Fig. 6 shows the relationship between haematocrit and C2/CI of blood samples from renal-failure patients undergoing dialyses. The values a = 65 and 6 = 93 were obtained from the literature.“) The results indicate that equation (7) using haematocrit did not fit so well, but it is useful for the determination of plasma urea concentration. The observed ratio C2/C1 deviated from the calculated value at higher haematocrits, probably because of the non-equilibrated dilution of whole blood along the carrier stream. The urea concentrations obtained by the proposed method were compared with those obtained by colorimetry (Fig. 7). The practical effectiveness of the conversion equation was confirmed by the correlation coefficient of the two methods ( r = 0.991) and the slope of the regression line.Conclusions The urea concentrations in whole blocd and plasma can be determined using a urease membrane electrode. The results obtained were in good accord with those obtained by conventional colorimetry. The urea concentration in plasma could be calculated from that in whole blood by an equation utilising haematocrit. This method permits “on-the-spot” or bedside measurements of blood urea. The authors thank Dr. Y. Yamagata, Mr. T. Tsukui and the staff of Hitachi General Hospital, Hitachi, Ibaraki, for their cooperation. They also acknowledge the helpful advice of Dr. T. Kobayashi of Hitachi Central Research Laboratory. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Wylenga, D. R., Clin. Chem., 1971, 17, 891. Carr, P. W., and Bowers, L. D., “Immobilized Enzymes in Analytical and Clinical Chemistry,” John Wiley, New York, 1980. Hansen, E. H . , and RGiiCka, J., Anal. Chim. Acta, 1974, 72, 353. Papastathopoulos, D. S . , and Rechnitz, G. A., Anal. Chim. Acta, 1975, 79, 17. Camman, K., translated by Schroeder, A. H . , “Working with Ion-selective Electrodes,” Springer-Verlag, Berlin, 1979, pp. 79-89. Rocks, B., and Riley, C., Clin. Chem., 1982, 28, 409. Guilbault, G. G., and Nagy, G., Anal. Chem., 1973,45,417. RfiiiEka, J., and Hansen, E. H., Anal. Chim. Acta, 1978, 99, 37. Guilbault, G. G., Nagy, G., and Kuan. S . S . , Anal. Chim. Acta, 1973, 67, 195. Terada, H., Suzuki, H., Niitani, K., Tadano, J . , Suganuma, K., and Aiga, S . , “Kensashitsu Ketsueki-gaku (Haematology for the Clinical Laboratory),” Kodansha, Tokyo, 1977 (in Japanese). Paper A31148 Received May 23rd, 1983 Accepted August 16th, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900061

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984. VOL. 109 65 Electroanalytical Studies of Beta-adrenergic Blocking Agents; N-lsopropylethanolam ine Derivatives; Procainam ide E. Bishop and W. Hussein" Chemistry Department, University of Exeter, Stocker Road, Exeter, EX4 4QD, UK The beta-adrenergic blocking agents metoprolol, oxprenolol, pindolol, practolol, propranolol and sotalol, and also the related procainamide, have been examined by anodic rotating disc electrode voltammetry in acidic media and in buffer media over the range of pH from 0 to 11.5. The best medium is 0.1 mol 1-1 sulphuric acid, in which most of the compounds show obedience to the Levich relationship at platinum and gold electrodes. Rapid determinations of oxprenolol (up to 5 x 10-3 mol I-'), practolol, propranolol and procainamide give very satisfactory results.Electrode kinetic parameters for mass and charge transfer have been determined for the latter four compounds. Reaction mechanisms have been elucidated and differ for each compound. Cyclic voltammetry has not proved to be particularly informative, and is illustrated for propranolol. Keywords: Beta-adrenergic blocking agents; procainamide; rotating disc electrode voltammetry; cyclic voltammetry; electrode kinetics and reaction mechanisms The beta-adrenergic blocking agents are competitive inhibi- tors of the effect of catecholamines at beta-adrenergic sites. Propranolol is the most prescribed for its antihypertensive , antianxiety , anticonvulsant and antianginal effects,172 and has been proposed for disfunctional labour3 and migraine.4 The other antagonists differ in potency, beta-receptor selectivity, agonist activity and membrane stabilisation .S Despite their considerable history, and their known participation in biolog- ical electron-transfer systems, no electrochemical investi- gation has yet been reported.Fluorimetry6~7 is commonly used for the determination of propranolol and its metabolites; spectrophotometry,s GLC,9J" HPLC,'l7l2 TLC,l3 radioim- munoassay14 and GC - MS15 have also been used. Procain- amide has a related use in cases of cardiac arrythmias,16 and may conveniently be included. It has been determined polarographically,17~1* as well as by other methods, but the common method of assay is by HPLC.19.20 Procainamide apart, the beta blocking agents are all N-isopropylethanolamine derivatives linked directly for sotalol, and via a methyl ether bridge in the rest, to a ring system which usually has another side chain, or second ring, as shown in Table 1.All have been investigated by direct anodic rotating disc electrode (RDE) voltammetry in aqueous media for analytical purposes and for reaction mechanisms and kinetics. Experimental The apparatus, instrumentation, electrode activation, solu- tion manipulation and procedures have been described.21 Oxprenolol , propranolol and procainamide were supplied as hydrochlorides, and were converted to sulphates.21 Samples were of Drug Standard grade, and supplied by the manufac- turers listed in Table 1. Normal scan speeds were 5 mV s-1. Results and Discussion Voltammetry of Propranolol in Acidic Media Although each compound displays a unique behaviour, propranolol, being the most frequently encountered, will alone be discussed in some detail.In 0.1 mol 1-1 sulphuric acid at platinum, propranolol gives two waves, as in Fig. l(a), the first being indistinct and merging with the second, which is well formed and obeys the Levich relationship with respect to square root of rotation speed and concentration. At the gold electrode a single wave appears [Fig. l(b)J, which also gives linear graphs of limiting current versus square root of rotation speed and versus concentration. In 1.0 moll-1 sulphuric acid, the first wave is better defined at platinum [Fig. 2(a)] but does not obey the Levich relationship. The second wave does conform to the Levich relationship.At the gold electrode, Fig. 2(b), the first wave peaks at higher rotation speeds, and both waves show Levich dependence on rotation speed and concentration. An orange colour develops in the solution during yoltammetry . / / \I 4A8- 08 10 1 2 1.60.6 "' ' 0.8 1.0 I 1.2 ' 1.4 ' 1.6 ' 1.8 ' EIV vs. S.C.E. Fig. 1. Voltammetry of 4.999 8 x 10-3 rnol 1-1 propranolol in 0.1 moll-1 sulphuric acid at (a) platinum and (b) old RDE. Conditions: nominal rotation speeds, (1) 10 Hz, (2) 20 Hz, f3) 30 Hz, (4) 40 Hz and (9, 50 Hz; scan speed, 5.0 mV s-1; electrode area, 0.503 cm2; temperature, 25 "C 0.6 0.8 1.0 1.2 1.4 1.6 0.8 1.0 1.2 1.4 1.6 1.8 EIV vs. S.C.E. Fig. 2. Voltammetry of 9.999 6 x 10-4 rnol 1-1 propranolol in 1.0 moll-' sulphuric acid at (a) platinum and (b) gold RDE.Conditions as in Fig. 1 * Present address: Department of Pharmaceutics, Faculty of Pharmacy, University of Karachi, Karachi-32, Pakistan.66 ANALYST, JANUARY 1984, VOL. 109 Table 1. Compounds examined C.A. Generic name number Structure CH3 Batch number 78R730 Supplier Proprietary name H H H H l 37350-58-6 Metoprolol c ~ ~ 0 - c ~ ~ - C H ~ I OH CH3 Geigy Lopresor Pharmaceuticals 6452-71-71 Oxprenolol 71178 Ciba Laboratories Trasicor H H H 0-c-c- H I OH 13523-86-9 Pindolol A/NO 81A313 Sandoz Products Ltd. Visken ADM 1072/73A ICI Eraldin 6673-35-4 Practolol 526-66-6 Propranolol 2685 1179A ICI Inderal 3930-20-9 Sotalol PO 1491022 Bristol Laboratories Sotacor 0 51-06-9 Procainamide C2H5, ,N-C-C-C-N H H 11 e N H 2 . H C I C2H5 H H H 02-655-4252 E. R. Squibb Pronestyl and Sons Ltd.Voltammetry of Propranolol in Buffer Media There is a small but significant change in the half-wave potential, and a greater change in limiting current, which reaches a maximum in 0.1 moll-' sulphuric acid, as the pH of the medium is changed (Fig. 3). The first wave disappears at platinum at pH >1 and at gold at pH >2. The quality of the waves degrades with increasing pH at both activated and deactivated electrodes and the wave vanishes for deactivated electrodes at pH >4. The optimum condition is 0.1 mol 1-1 sulphuric acid. The electrodes are not electrochemically deactivated in acidic media, but product adsorption at higher pH values leads to deactivation. root of frequency and on concentration. Pindolol (Fig. 6) gave two overlapping waves at platinum and two waves plus a peak at gold; the peak limiting currents did obey the Levich relationship. For sotalol (Fig.6) the waves merged with the background at platinum and became peaks at gold, which did show Levich dependence. Metoprolol (Fig. 6) revealed similar non-Levich behaviour at both platinum and gold. Voltammetry of Procainamide Procainamide gave a normal behaviour [Fig. 7(a)] at an activated platinum electrode with linear Levich dependences, but non-zero intercepts, and at an unactivated electrode the limiting current was diminished. At gold [Fig. 7(b)] the wave became a flat peak with a trough in it and the peak current gave linear graphs against square root of rotation speed and against concentration, but again with non-zero intercepts; unactivated electrodes revealed no deactivation effects.Despite the non-zero intercepts, the calibration graphs can be used in quantitative analysis. Voltammetry of Other Compounds The remaining beta blockers have been examined in 0.1 mol 1-1 sulphuric acid only. Practolol (Fig. 4) and oxprenolol (Fig. 5) both gave a single good wave at both platinum and gold electrodes; a light orange colour developed in the solutions. The limiting currents showed Levich dependence on squareANALYST, JANUARY 1984, VOL. 109 67 1.2 (a’ t h Oa2 t - PH 0 2.0 4.0 6.0 PH Fig. 3. Variation of (a) limiting current and ( b ) half-wave potential with pH for anodic voltammetry of 9.999 6 x 10-4 moll-’ propranolol at platinum (solid line) and gold (broken line) RDE. Rotation speed, 50 Hz 0 -0.8 -1.6 a E .2 -2.4 - 3.2 -4.0 I I I I \\\ 1 I I 1 I 0.8 1.0 1.2 1.4 1.6 0.8 1.0 1.2 1.4 1.6 1.8 E N vs. S.C.E. Fig. 4. Anodic voltammograms of 5 x 10-3 mol I- * practolol in 0.1 moll-’ sulphuric acid at (a) platinum and ( b ) gold RDE. Conditions as in Fig. 1 The influence of hydrogen ion concentration on the procainamide waves is illustrated in Fig. 8, over the pH range 0-1 1.5, maximum wave height is attained at both electrodes in 0.1 mol 1-1 sulphuric acid. At platinum a single, well defined, wave appears from pH 0 to 5 at an activated electrode and also at an increasingly lower limiting current at an unactivated electrode. At the latter, the wave vanishes in more alkaline media, while at an activated electrode the wave is eclipsed, only to reappear, at a much more positive potential at pH >lo.At an activated gold electrode the wave peaks at pH 7 and then becomes indistinct at pH >8 but does not reappear in more alkaline media. The wave becomes indistinct at an unactivated gold electrode at pH >2 and vanishes at pH >6. Analytical Validity Examples of calibration results for gold and platinum elec- trodes are given in Table 2. There is little to choose between the two electrode materials. Calibrations for oxoprenolol become non-linear (curvature towards the concentration axis) above 5 x 10-3 mol 1-1. To appraise the reliability of rapid determination, a series of four solutions of each of these drug standards was prepared and a single measurement of the limiting current was made at each of five rotation speeds.The concentration was calculated from slope and intercept, and the percentage relative standard deviations for each group of five determinations are given in Table 3. This involves the propagation of errors from five calibrations and small errors in rapid setting of rotation speed; the duration of each group of -0.8 c \ I- \ I -1.6 - a E -2.4 - . -3.2 - -4.0 -4.8 t 1 I I I L 0.8 1.0 1.2 1.4 1.6 0.8 1.0 1.2 1.4 1.6 1.8 E N vs.S.C.E. Fig. 5. Anodic voltammo rams of 4.999 8 mol 1-1 oxprenolol in 0.1 moll-’ sulphuric acid at (a7 platinum and (b) gold RDE. Conditions as in Fig. 1 five measurements is less than 10 min. Pindolol and sotalol can be determined at gold electrodes. Excipients in dosage forms encountered were electrochemically inert. Kinetics Conditional potentials are not accessible for these systems, so kinetic parameters are calculated with reference to the half-wave potentials.The half-wave potentials are similar for gold and platinum electrodes and show a small increase with increasing mass-transfer rate, that is with increasing rotation speed and concentration. The range for propranolol and oxprenolol is 1.10-1.26 V versus S.C.E.; practolol gives potentials about 100 mV smaller, while procainamide is intermediate. The electrode reactions are of moderate22 speed. The charge-transfer rate constants range from 2.83 to 7.68 X 10-6 1 cm-2 s-1 for practolol and about half of these values for the other compounds, showing modest increases with increasing rotation speeds and decreases with increasing concentration.The charge-transfer coefficients, p, lie in the range 0.07-0.22 for propranolol, oxprenolol and procain- amide and 0.22-0.62 for practolol, decreasing with increasing mass-transfer rate. Tabulations of the results of definitive sets of measurements for the four compounds, calculated by pattern theory,22 are available from the authors. The small differences in half-wave potential mean that the compounds cannot be resolved in mixtures by normal voltammetry, but practolol can be resolved from the others by differential-pulse volt amme try. Coulometry and Reaction Mechanisms An attempt to examine the first, indistinct, wave of proprano- lo1 in 0.1 mol 1-1 sulphuric acid by potentiostatic coulometry yielded no useful information. Amperostatic coulometry engaged both waves and showed that the over-all reaction involved four electrons. For example, passage of a charge equal to one half of an electron equivalent caused a decrease in limiting current of 13%, and a further similar charge led to a decrease of 25%, which accords with four electrons.Con- tinued scanning of the latter solution showed that the limiting current gradually increased until it regained the 13% level, when subsequent scans superimposed. Thus the product of the four-electron oxidation is unstable, and undergoes chemical reaction to regenerate the stable product of a two-electron process corresponding to the first wave. The solution became dark orange and a fine precipitate settled out. Practolol, however, showed a simple two-electron step, with no further oxidation or regeneration.Passage of, for example, charge equivalent to half an electron gave a reduction of limiting68 ANALYST, JANUARY 1984, VOL. 109 0 -0.8 -1.6 -2.4 -3.2 -2.4 - -3.2 - ..- 0.8 1.0 1.2 1.4 1.6 1.8 0.4 0.6 0.8 1.0 1.2 1.4 1.6 EN vs. S.C.E. 0 -0.4 -0.8 -1.2 -1.6 -2.0 0 -0.4 -0.8 -1.2 -1.6 -2.0 0.8 1.0 1.2 1.4 1.6 Fig. 6. Anodic voltammograms in 0.1 mol I-' sulphuric acid of (a and b) metoprolol, (c and d ) pindolol and (e and f) sotalol at (a, c and e ) platinum and (b, d a n d 8 gold RDE. Conditions as in Fig. 1. Detached line in a, b and e , unactivated electrode 0 -0.8 4 -1.6 E -!! . -2.4 -3.2 -4.0 1 I 1 I 0.7 0.9 1.1 1.3 1.5 1.7 0.7 0.9 1.1 1.3 1.5 1.7 EN vs. S.C.E. Fig. 7. Anodic voltammograms of 5.0002 x 10-3 mol 1-1 procainamide in 0.1 mol 1-1 sulphuric acid at (a) platinum and (b) gold RDE.Conditions as in Fig. 1 current of 25%, an additional similar charge decreased the limiting current by 50%, and repetitive scans of these solutions simply superimposed. The primary reaction product is there- fore stable. Oxprenolol, on the other hand, displayed an n-value of six electrons, with regeneration to the product of a four-electron step: thereafter subsequent scans superimposed, indicating that this product was stable. Amperostatic coulo- metry of procainamide revealed a simple single-step four- electron oxidation, producing a red colour turning brown, and a stable product that neither decomposed nor underwent further oxidation. That a group of compounds so closely related in functional structure should display such a diversity of behaviour is intriguing and puzzling.Sufficient information about the three members that underwent orderly quantitative anodic oxida- tion has been gleaned to state the mechanisms with clarity. The N-alkylethanolamine group is the common active site (Fig. 9). The hydroxyl function is oxidised by two sequential one-electron steps, I and 11, of which the second is rate determining, via the radical to the ketone, which is the stable product in all instances, and is the origin of the first wave of propranolol, which is more marked in 1 moll-' sulphuric acid (Fig. 2). The practolol molecule does not accommodate any further reaction. Propranol, by virtue of current sinking in the naphthol residue, permits a further two-electron step to the dication diradical, 111, which is not stable, and reacts quickly with the solvent to regenerate the stable ketone by step IV.For oxprenolol, the side chain in the 2-position is oxidised to the dieneoxy group by a two-electron step, so transferring four electrons and giving the stable diene ketone. The N-alkyl- ethanal side chain can also undergo oxidation to the dication diradical, 111, so giving a total of six electrons. But again the radical is unstable, and reacts by step IV, regenerating the stable ketone, so that the product of the primary amperostatic process decays to the stable four-electron stage. Procainamide reacts through the 4-amino group to give straightforward oxidation to the nitroso compound, VI. This resists further oxidation to the nitro compound, a situation for which there is precedent,23 although the survey of the effect of pH variation suggests that in 0.1 mol 1-1 sodium carbonate the nitro compound may be formed at platinum electrodes.ANALYST, JANUARY 1984, VOL.109 69 0.8 1 .o 1.2 1.4 . 1.6 u! c! ? 9 > G 0.8 1 .o 1.2 1.4 1.6 pH 11.5 pH10 p H 9 p H 8 pH7 pH5 p H 4 pH3 pH2 Current -+ Fig. 8. Effect of pH on the anodic wave of 10-3 mol 1-1 procainamide at (a) platinum and ( 6 ) gold RDE. Rotation speed, 50 Hz; media, sulphuric acid (pH = 0, l), citrate - phosphate buffer (pH = 2-8), sodium carbonate (pH = 9-11.5) adjusted to exact pH value at 25°C Curves: A, activated electrode; B, unactivated electrode Table 2. Example calibration results for RDE voltammetry.Electrode area, 0.503 cm2; medium, 0.1 moll-' sulphuric acid; temperature, 25 "C; and n = 4 Nominal Sample frequency/Hz Oxprenolol(0-5 x 10-3moll-1, gold) . . . , . . . . . . 10 20 30 40 50 Practolol(0-10-2 moll-1, platinum) . . . . . . . . 10 20 30 40 50 Propranolol(0-10-2 mol I-', platinum) . . . . . . . . 10 20 30 40 50 Procainamide (0-12-2 mol I - l , gold) . . . . . . . . . . 10 20 30 40 50 Slope/mA 1 mmol-1 0.471 58 0.619 74 0.719 74 0.789 47 0.846 84 0.275 00 0.370 00 0.495 79 0.543 68 0.597 89 0.366 68 0.471 68 0.548 85 0.610 00 0.660 02 0.269 77 0.335 77 0.39058 0.423 65 0.448 65 Intercept/ mA 0.006 32 0.003 95 0.003 95 0.007 89 0.017 37 0.095 00 0.123 33 0.161 05 -0.09 158 -0.130 53 0.146 67 0.198 33 0.248 33 0.250 00 0.246 67 0.249 17 0.351 54 0.386 16 0.412 31 0.426 31 Correlation coefficient 0.999 94 0.999 99 0.999 99 0.999 97 0.999 86 0.999 89 0.999 32 0.999 99 0.999 97 1 .ooo 00 0.999 01 0.999 30 0.999 56 0.999 19 0.999 46 0.999 97 0.999 94 0.999 96 1 .ooo 00 l*.OOO 00 S.d.of residuals/ mA 0.009 18 0.005 74 0.005 74 0.011 47 0.025 24 0.008 66 0.028 87 0.004 59 0.006 88 0.002 30 0.034 64 0.037 53 0.026 36 0.059 16 0.046 19 0.004 33 0.011 09 0.009 71 0.001 39 0.001 39 S.d. of slopelmA 1 mmol-1 0.002 58 0.001 61 0.001 61 0.003 22 0.007 09 0.002 04 0.006 80 0.001 29 0.001 93 0.000 65 0.008 16 0.008 83 0.007 41 0.012 23 0.010 89 0.001 02 0.001 88 0.001 65 0.000 24 0.000 24 R.s.d. of slope, % 0.55 0.26 0.22 0.41 0.84 0.74 1.84 0.26 0.36 0.11 2.23 1.88 1.49 2.08 1.65 0.34 0.56 0.42 0.06 0.0570 ANALYST, JANUARY 1984, VOL.109 Table 3. Precision of determinations by RDE voltammetry. Each result arises from measurement of the limiting current at each of five rotation speeds. Electrode area, 0.503 cm2; medium, 0.1 moll-' sulphuric acid; temperature, 25 "C Oxprenolol" Practolol Propranolol Procainamide Electrode Concentrationl R.s.d., Concentration/ R.s.d., Concentrationl R.s.d., Concentration/ R.s.d., mmoll-1 Yo mmol I- I % mmoll-1 Y O mmoll-1 O/O type Platinum . . . . 1.99991 2.22 1.999 70 3.61 1.999 94 0.86 2.000 08 3.41 4.999 79 0.67 4.999 26 0.45 4.999 86 0.76 5.000 20 3.60 7.999 66 0.47 7.998 81 0.31 7.999 77 0.98 8.000 32 1.28 9.999 58 4.23 9.998 52 1.39 9.999 72 1.43 10.00040 2.72 Gold . . . . . . 1.999 91 1.58 1.999 70 2.26 1.999 94 0.92 2.000 08 4.64 4.999 79 2.50 4.999 26 0.17 4.999 86 3.08 5.000 20 0.25 7.999 66 0.75 7.998 81 0.51 7.999 77 1.76 8.000 32 1.37 9.999 58 3.13 9.998 52 0.98 9.999 72 0.16 10.000 40 0.56 * Calibration graphs non-linear above 5 x 10-3 moll-'.Table 4. Cyclic voltammetry of propranolol. Concentration, 5 x 10-3 mol 1-1 in 0.1 mol I-' sulphuric acid; gold electrode, 0.503 cm2; temperature, 25 "C; potentials versus S.C.E. Anodic Cathodic ipcl Epcl mA V Scan rate (Y)/ ipA &A b t / EPA1I b , I E d mVs-1 mA mA mA V V V' 200 400 588 769 1000 1176 1369 1579 1765 2 000 2 105 2 384 0.575 1 .oo 1.35 1.75 2.075 2.2 2.35 2.65 2.85 3.02 4.45 4.65 0.675 1 .oo 1.25 1.525 1.75 1.95 2.1 2.425 2.80 3.125 3.75 4.25 0.575 1 .oo 1.275 1.575 1.775 2.025 2.225 2.55 -0.10 -0.08 -0.072 -0.060 -0.055 -0.04 -0.025 -0.020 -0.09 0.06 0.10 0.125 0.10 0.25 0.28 0.45 0.45 0.50 0.46 0.48 0.50 0.55 0.70 0.75 1 .o 0.91 0.95 0.95 0.95 0.95 0.95 0.95 1.25 1.25 1.75 1.75 1.75 1 s o 2.25 1.65 1.75 2.50 2.65 2.75 0.475 0.475 0.450 0.450 0.375 0.375 0.350 0.350 0.350 0.30 0.32 0.29 H H H H H H H H V RA-O-C-C=C ---+RA-O-C=C=C + 2H' + 2 e VI H2N-b + H20 __+ ON V R B + 4H+ + 4e Fig.9. Reaction mechanisms: RA in I-V = residue of molecule of beta blocking agent; RB in VI = residue of procainamide side chain, cf. Table 1ANALYST, JANUARY 1984, VOL. 109 71 I 1 1 -0.4 0 0.4 0.8 1.2 6.0 a 4.0 E -!! -. 2.0 0 a -2.0 E . 2 -4.0 -6.0 I 1 I I I ~~ -0.4 0 0.4 0.8 1.2 E N vs. S.C.E. Fig. 10. Scan speeds: ( a ) 1579 mV s-I; ( b ) 2384 mV s-l. Al, A2 and A3, anodic peaks; 1 , 2 and 3, first, second and third cycle Cyclic voltammograms of 5 X mol I-’ propranolol in 0.1 mol 1-1 sulphuric acid at a stationary 0.503 cm2 gold electrode.Cyclic Voltammetry Cyclic voltammograms of propranolol, practolol and ox- prenolol were examined for information about adsorption and fast reverse radical reactions, but were not fruitful. The behaviour of propranolol is exemplified in Fig. 10 and Table 4. The first anodic half cycle is alone informative, and gave essentially a single broad peak broken up into a peak (Al), a plateau (A2) and a small after-peak (A3). The peak A3 disappears at higher scan rates. There is a single cathodic wave. Peak currents and potentials are scan rate dependent, but do not accurately conform to the conventional relation- ships. In the example illustrated, an unactivated gold elec- trode gave a similar result to an activated electrode, suggesting that reaction and product species did not affect electrode activity . We thank Messrs.Bristol Laboratories, Ciba Laboratories, Geigy Pharmaceuticals, Imperial Chemical Industries, Sandoz Products Ltd. and E. R. Squibb and Sons Ltd. for the gift of the materials listed in Table 1, and the Royal Society for the SEL transfer standard DVM. W. H. thanks the Government of Pakistan for the award of a Scholarship, and the University of Karachi for the grant of leave of absence. References 1. Shand, D. G., N. Engl. J. Med., 1975, 293,280. 2. Harner, J., Grandjean, T., Melendey, L., and Gowton, G., Br. Med. J . , 1965,2,720. 3. Mitrani, A., Oettinger, M., Alunader, E.G., Sharf, M., and Klein, A., Br. J. Obstet. Gynaecol., 1975, 82, 651. 4. Weber, R. B., and Reinmuth, 0. M., Neurology, 1972,22,366. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Pendleton, R. G., Newman, D. J., Sherman, S. S . , Brawn, E. G., and Maya, W. E., J. Pharmacol. Exp. Ther., 1972,180, 647. Black, J. W., Duncan, W. A. M., and Shanks, R. G., Br. J . Pharmacol., 1965,25,577. Kraml, M., and Robinson, W. T., Clin. Chim. Acta, 1978,24, 171. Shaw, R. F., in Sunshine, I . , Editor, “Methodology for Analytical Toxicology,” CRC Press, Cleveland, Ohio, 1975. Disale, E., Baker, K. M., Bareggi, S. R., Watkins, W. D., Cidsey, C. A., Frigero, A., and Morselli, P. L., J. Chrom- atogr., !973,24,347. Hackett, L. P., andDusci, L. J., Clin. Toxicol., 1979,15,63. Pritchard, J. F., Schneck, D. W., and Hayes, A. H., J. Chromatogr., 1979,162,47. Walk, T., Pharmacologist, 1975,17,262. Kawashima, K., Levy, A . , and Specter, S . , J. Pharmacol. Exp. Ther., 1976,196,517. Garteiz, D. W., and Aqwalle, T., J. Pharm. Sci., 1972, 61, 1720. Ehrsson, H., J. Pharmacol., 1976,28,662. Kayden, H. J., Steele, J . M., and Mark, L. C., Circulation, 1951,4,13. Burghardt, H., Deut. Apoth.-Ztg., 1968,108,115. Kiseleva, L. A., and Orlov, Yu. E., Farmatsiya, 1977, 26, 38. Schmidt, G., Vandemark, F. L., and Adam, R. F., Chrom- atogr. Newsl., 1976,4,32. Griffiths, W. C., Dextraze, P., Hayes, M., and Dimond, T., Clin. Toxicol., 1980,16,51. Bishop, E., and Hussein, W., Analyst, 1984, 109, in the press. Bishop, E.,Analyst, 1972,97,761. Sidgwick, N. V., Taylor, T. W. J., and Nevil, I . , “Organic Chemistry of Nitrogen,” Third Edition, Oxford University Press, London, 1966, p. 339. Paper A31246 Received August 8th, 1983 Accepted September I9th, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900065

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984. VOL. 109 73 Electroanalytical Study of Tricyclic Antidepressants E. Bishop and W. Hussein" University of Exeter, Chemistry Department, Stocker Road, Exeter, EX4 4QQ UK ~~~ Anodic electroactivity, shown by clomipramine, desipramine, dibenzepin, dothiepin, imipramine, opipramol and trimipramine has been studied at rotating disc electrodes of platinum and gold in sulphuric acid and in buffer media by single scan voltammetry, and at stationary disc electrodes by cyclic voltammetry. Potentiostatic and amperostatic coulometry have been used to determine the number of electrons involved in the reactions. Rapid voltammetric determinations of the active drugs have been evaluated in the best medium, 0.1 mol I-' sulphuric acid. Adsorption of the oxidation product on the electrodes prevents determination of clomipramine and restricts the concentration range for imipramine and trimipramine.The kinetic parameters, half-wave potentials, mass and charge-transfer rate constants and charge-transfer coefficients have been determined under a wide range of conditions. Electrode reaction mechanisms have been elucidated. Dothiepin is oxidised by a single four-electron step to the sulphone. The azepines are oxidised to a radical that dimerises rapidly before a further one-electron step produces the dimer diradical; opipramol can undergo a further four-electron oxidation of the piperazine ring in the side chain to the 1,4-dihydropyrazine derivative. Keywords: Tricyclic antidepressants; rotating disc electrode voltammetry; cyclic voltammetry; coulometry; electrode kinetics and reaction mechanisms The tricyclic compounds in Table 1 are most effective for the treatment of physiological retardation depression.Imipram- ine and desipramine often succeed when other treatments fail, but are not well tolerated when hysterical traits are present. Elderly patients usually respond best to amitriptyline and young adults to nortriptyline. Clomipramine and dibenzepin appear between imipramine and amitriptyline in effective- ness. Trimipramine exerts anxiolytic activity and opipramol shows relatively weak thymoleptic activity but also has neuroleptic tranquillising activity, Dothiepin may provoke epilepsy in the elderly. Although tetracyclic, maprotiline shows pharmacological and toxic similarities to the tricyclics. Side effects arise from anticholinergic action.The tricyclics are extensively metabolised by demethylation and oxidation. Thus, imipramine and clomipramine after demethylation yield desipramine, and oxidation in aqueous media by primary electrode processes is of interest. Common structural features include propanamine and dibenzazepine (or thiepine) moie- ties. Polarographic reduction of clomipramine suggests a two- electron reduction with a non-linear concentration depen- dence ascribed to adsorption. 1 Anodic oxidation of dothiepin in non-aqueous media is reported2 as a one-electron step with the formation of a cation radical, while cyclic voltammetry of imipramine in acetonitrile indicated transfer of more than two electrons in the anodic process.3 Examples from the literature on the determination of these drugs are as follows: spectro- metry,435 fluorimetry,6?7 GLC,G13 liquid - solid chromato- graphy,l4 HPLC,15-17 radiochemistry,'s radioimmunoassay,19 mass fragmentometry,20 chemical ionisation21 and isotopic labelling.22 In this paper is reported a fundamental and analytical electrochemical examination of tricyclic antidepressants by high precision rotating disc electrode (RDE) voltammetry at platinum and gold electrodes in 0.1 mol 1-1 sulphuric acid.Experimental The glassware, rotating electrode assembly, its operation and electronics, electrode activation procedure, reagents, method of preparation of chloride-free samples and voltammetric and coulometric procedures have been described.23 The samples * Present address: Department of Pharmaceutics, Faculty of Phar- macy, University of Karachi, Karachi-32, Pakistan.mainly in their hydrochloride forms were of Drug Standard grade and were kindly donated, along with dosage forms, by the manufacturers named in Table 1. The hydrochlorides were converted to nitrates by the method described,23 and 0.01 moll-' stock solutions were prepared in 0.1 moll-' sulphuric acid, which has repeatedly proved to be the best medi~m.2"~~ The normal voltammetric scan rate was 5 mV s-1 and the geometric area of the disc electrodes was 0.503 cm2. Results and Discussion Of the compounds listed in Table 1 the dibenzohomocyclo- heptenes, I , X and XII, the heterocycloheptene with a fully oxidised hetero atom, doxepin, and the tetracyclic maprotiline were cathodically and anodically inactive in 0.1 mol 1-1 sulphuric acid at both platinum and gold electrodes.The sulphur and nitrogen heterocyclics were active in different ways, except for the carboxamide, 11, with a partly oxidised ring and an electron-withdrawing carboxyl group on the nitrogen. Opipramol and trimipramine showed behavioural similarities to imipramine, although opipramol has a further reaction step; while desipramine, dibenzepin and dothiepin grouped in a similar way, despite the different reaction of dothiepin. Clomipramine resembles imipramine and desi- pramine, but the electrode process is obscured by adsorption. Voltammetry Platinum electrodes in 0.1 mol 1-1 sulphuric acid Imipramine, trimipramine and, to a certain extent, opipramol (first wave) showed a sharp division of behaviour dependent on concentration, Below 2 X 10-3 moll-' imipramine gave a single, well formed wave, Fig.l(a), and graphs of limiting current against square root of rotation speed and against concentration were linear, but the intercepts, although small, were non-zero; the concentration graphs for opipramol, Fig. 2(g), are, however, acceptable. Opipramol is unique in showing two waves, Fig. 3, the second tending to peak at higher mass transport rates. At concentrations above 10-3 moll-1, imipramine behaves normally at first, the wave height increasing with increasing rotation speed and concentration, but with further increase in mass transport rate the wave first becomes a flat peak, which then sharpens and, above 5 X mol 1-1 at higher rotation speeds, finally recesses, the peak height dropping below that of earlier scans.74 ANALYST, JANUARY 1984, VOL.109 Table 1. Compounds examined C.A. No. number Generic name I 50-48-6 Amitriptyline I1 298-46-4 Carbamazepine I11 303-49-1 Clomipramine IV 50-47-5 Desipramine V 4498-32-2 Dibenzepin VI 113-53-1 Dothiepin VII 1668-19-5 Doxepin VIII 50-49-7 Imipramine IX 10262-69-8 Maprotiline X 72-69-5 Nortriptyline Structure WaHC1 \ a \ / .HCI 6 H - CH 2- C H .- N ( C H 3) 2 iH2-CH2--CH2-N/H 'CH3 Q N n . H C I / m \ HCI XI 315-72-0 Opipramol a Batch number COOA085/ 272 16 78R538 79R498 80R564 81B904 H-2937 3-3760 79R556 706t76 81A030 79R45 1 Proprietary Manufacturer name William R. Lentizol Warner Geigy Tegretol Pharmaceuticals Geigy Anafranil Pharmaceuticals Geigy Pertofran Pharmaceuticals Sandoz Products Ltd.Noveril Boots Co. Ltd. Prothiaden Pfizer Ltd. Sinequan Geigy Tofranil Pharmaceuticals Ciba Ludiomil Laboratories Eli Lilly Aventyl Geigy Insidon PharmaceuticalsANALYST, JANUARY 1984, VOL. 109 75 Table 1 (continued) No. number Generic name C.A. XI1 438-60-8 Protriptyline Structure Batch Proprietary number Manufacturer name XI11 739-71-9 Trimipramine I CH3 CH -CH2-CH 2- RP7162 May and Surmontil Q maleate I Baker Ltd. CH2-CH-CH2--N(CH3)2 I CH3 0 0.4 0.8 Q E 1.2 2 I 1.6 2.0 2.4 I / ‘5 PI ‘ 0.5 0.7 0.3 1.1 1.3 1.5 E N vs. I 1 0.4 !- ‘h o.8 1.2 LW 2.0 , S.C.E. 0.5 0.7 0.9 1.1 1.3 1.5 1.7 Fig. 1. Anodic voltammograms at (a) platinum and ( b ) gold rotating disc electrodes of 10-3 mol 1-1 imipramine in 0.1 mol 1-1 sulphuric acid.Electrode area, 0.503 cm2; scan speed, 5 mV s-l. Nominal rotation speeds for curves 1-5 are 10, 20, 30, 40 and 50 Hz, respectively; offset curves are for a repeat scan at 50 Hz without intervening activation of the electrode 4 - I , . I I I 2 4 6 0 2 4 6 0 2 4 6 a flJHzlh . E O 0 5 10 0 5 10 0 5 10 0 5 10 Concentration/lO-” mol I-’ Fig. 2. Levich plots for (a and d ) desipramine, ( c and e ) dibenzepin, (b and fj dothiepin and (g) opipramol at (a, c, d and e ) platinum and (b, f and g) gold RDE. In order of increasing current, the lines for the square root of frequency plots are for nominal concentrations of 2, 5, 8 and 10 x 10-3 mol 1-1 and the lines for concentration plots are for nominal frequencies of 10, 20, 30, 40 and 50 Hz A repeat scan without intervening activation of the elec- trode then shows total suppression of the wave, whereas with the more dilute solutions the wave persists at an unactivated electrode but with a lower limiting current.Levich depen- dence is destroyed when the wave peaks. Desipramine, dibenzepin and dothiepin give excellent voltammograms, Fig. 3, and good Levich plots up to at least 10-2 moll-1, Fig. 2, but despite the good concentration graph, Fig. (2e), dibenzepin gives grossly non-zero intercepts in square root of frequency graphs, Fig. 2(c). Clomipramine gives a peak and a trough,76 3.2 2.4 ' ANALYST, JANUARY 1984. VOL. 109 0 0.4 0.8 1.2 4 1.6 E . - I 0.5 0.7 0.9 1.1 1.3 1.5 1.7 0 0.8 I .6 2.4 3.2 4.0 1 I I 0 0.8 1.6 2.4 3.2 4.0 I-- ' I 4.8 0.8 1.0 1.2 1.4 1.6 1.8 E N vs.S.C.E. 0 0.8 1.6 2.4 3.2 4.0 0 0.8 1.6 2.4 3.2 4.0 0.5 0.7 0.9 1.1 1.3 1.5 1.7 Fig. 3. Anodic voltammograms of (a and 6) nominal 5 x mol 1-1 clomipramine, ( c and d ) dothiepin and ( e and f) opipramol in 0.1 mol I-' sulphuric acid at (a, c and e) platinum and (b, d and fi gold RDE. Nominal rotation speeds for curves 1-5 are 10, 20, 30, 40 and 50 Hz, respectively; scan speed, 5 mV s-I. Offset curves are for a repeat scan at 50 Hz without intervening activation of the electrode Fig. 3(a), with a second wave discernible in the trough region at gold. Levich dependence is displayed at low mass transport rates, but is destroyed by product adsorption as mass transfer rates increase; voltammetry is not useful in the determination of this drug.The dothiepin wave, Fig. 3(c), is twice the height of that of other compounds. Gold electrodes in 0.1 mol 1-1 sulphuric acid At concentrations up to at least 10-2 moll-' of imipramine, a single, well defined wave appears in the form of a plateau followed by a trough and the limiting currents bear a linear relationship to both square root of rotation speed and concentration, but again with non-zero intercepts. Similar results were obtained for opipramol [Fig. 3 0 1 and trimipram- ine. Desipramine and dothiepin [Fig. 2(b) and (f), Fig. 3(d)] gave near zero and dibenzepin non-zero intercepts for Levich plots of the well defined limiting currents. At low concentra- tions imipramine gives three waves as shown in Fig. l(b). Desipramine, trimipramine and opipramol formed three waves at any concentration; dibenzepin gave two plateaux and dothiepin a single wave.The limiting currents obeyed the Levich relationship and loss of wave height occurred on rescanning without intervening reactivation of the electrode. Clomipramine, Fig. 3(b), gives two waves obscured by adsorption. During electrolysis at both types of electrode of desipram- ine and trimipramine a green colour appeared, increasing in intensity with progressive oxidation, followed by a dark precipitate. After settlement of the precipitate, the super- natant liquor was colourless. Opipramol is itself yellow in colour, and gave a green colour but no precipitate. No colour or precipitate was observed with imipramine, dibenzepin or dothiepin, although a blue colour has been reported,3 and ascribed to dimer formation, on oxidation of imipramine in non-aqueous media.Prolonged electrolysis in aqueous media eventually produces a loose purple film on the electrode surface. Influence of p H The active compounds were examined in sulphuric acid at pHs 0 and 1, and in citrate - phosphate buffers at unit intervals from pH 2 to pH 7. For both platinum and gold electrodes the best performance was achieved in 0.1 mol 1-1 sulphuric acid. Variation in half-wave potential, wave height and wave suppression differed from compound to compound, but no fresh wave appeared and no benefit accrued on increasing the pH. Examples of the two principal electrode processes are shown in Fig. 4. In general, the quality of the voltammograms deteriorated on increasing the pH; opipramol, however, showed conflation of its two waves at pH 4 and enhanced mass transfer by adsorption gave an increase in wave height up to pH 7 at activated electrodes and the wave peaked at pH L 5.In all instances a second scan without intervening activation of the electrode gave a decrease in wave height with total suppression at pH > 3. Analytical validity The concentration graphs in Fig. 2 illustrate the calibrations, and Table 2 contains examples of the regression analyses of calibrations at gold electrodes. Assay of various dosage forms presented no difficulty, but this does not offer a proper evaluation of the precision of the method. Instead, four solutions of each of the Drug Standard samples were prepared, and the limiting current of the first wave was measured at the appropriate plateau potential for each of five rotation speeds. The concentration was calculated from theANALYST.JANUARY 1984. VOL. 109 77 Table 2. Example calibration results for RDE voltammetry at gold electrodes. Electrode area, 0.503 cm2; medium, 0.1 mol I - ' sulphuric acid; temperature, 25 "C; and n = 4 Nominal Sad. of S.d. of R.s.d. of Sample Hz mA 1 mmol- I mA coefficient mA mA 1 mmol- I slope, YO slopel frequency/ Slopel I n t e rce p t / Co r re1 a t i on residuals/ Desipramine (0-10-2 moll-') . . 10 0.243 65 0.21 1 54 0.999 95 0.006 93 0.001 18 0.48 20 0.361 10 0.102 63 0.999 84 0.01 1 47 0.003 23 0.89 30 0.425 37 0.198 16 0.999 90 0.018 24 0.003 01 0.71 40 0.477 22 0.212 I I 0.999 82 0.027 75 0.004 58 0.96 50 0.560 10 -0.103 95 0.999 85 0.017 21 0.004 84 0.87 (0-10 - 3 moll-1) , .10 0.400 00 0.100 00 1 .000 00 0.000 00 0.000 00 0.00 Imipramine 20 0.550 00 0.090 00 1 .000 00 0.000 01 0.000 01 0.002 30 0.666 67 0.086 67 1 . 000 00 0.000 002 0.000 01 0.001 40 0.736 54 0.092 3 1 0.999 98 0.001 39 0.002 36 0.32 50 0.800 00 0.100 00 1 .000 00 0.000 00 0.000 00 0.00 Trimipramine (,0-10-3mmoll-1) . . 10 0.364 97 0.087 69 0.999 92 0.001 39 0.002 36 0.65 20 0.506 73 0.083 85 0.999 99 0.000 69 0.001 18 0.23 30 0.573 08 0.094 62 0.999 86 0.002 77 0.004 71 0.82 40 0.633 33 0.01 1 33 1 .000 00 0.000 00 0.000 00 0.00 50 0.710 20 0.11612 0.999 80 0.004 29 0.007 07 1 .00 Table 3. Precision of rapid determination by RDE voltammetry. Each result arises from the measurement of the limiting current at the plateau voltage of the first wave at each of five rotation speeds.Electrode area, 0.503 cm2; and medium, 0.1 mol I--' sulphuric acid Desipramine Dibenzepin Dothiepin Imipramine Opipramol Trimipramine Concen- tration/ Electrode mmol R.s.d., Platinum . . . . 1.99684 2.26 4.9221 0.10 7.98736 2.51 9.9842 1.17 type I-' YO Concen- tration/ mmol R.s.d., 1-1 YO 1.99913 1.41 4.99782 2.45 7.99652 0.60 9.99565 2.32 Concen- tration/ mmol R.s.d., 1-1 Yo 2.00026 3.21 5.00065 1.90 8.00104 0.70 10.001 30 0.97 Concen- tration/ mmol R.s.d., I-' % 0.200016 3.19 0.50004 2.02 0.800 064 0.94 1.00008 1.27 Concen- tration/ mmol R.s.d., I-' YO 1.67034 1.86 4.17585 1.21 6.681 36 1.21 8.351 7 3.18 Concen- tration/ mmol R.s.d., I-' Yo 0.199 994 3.77 0.499 986 0.05 0.799 977 0.70 0.999 972 1.62 Gold .. 1.99684 1.07 1.99913 6.77 2.00026 1.21 0.200016 3.88 1.67034 6.62 0.199994 0.74 4.9921 2.24 4.99782 3.02 5.00065 2.83 0.50004 2.34 4.17585 2.64 0.499986 0.40 7.98736 1.58 7.99652 5.00 8.001 04 2.31 0.800064 0.69 6.68136 4.56 0.799977 0.53 9.9842 1.33 9.99565 1.20 10.00130 1.03 1.00008 1.36 8.3517 0.31 0.999972 1.15 slope and intercept (cf. Table 2) of the calibration graph and the relative standard deviation of each set of five measure- ments is recorded in Table 3. Each result involves the propagation of errors from five calibrations, yet the mean relative standard deviation is just under 2%, less than 10 min being required for each group of measurements. Cyclic Voltammetry Cyclic voltammetry of imipramine in acetonitrile has been reported.3 Cycles at potential scan rates from 200 to 2 384 mV s-1 in the potential range -0.6 to +1.2 V versus S.C.E.of 10-3 mol 1-1 imipramine in 0.1 mol 1-1 sulphuric acid were performed and those for two speeds at an activated gold electrode are shown in Fig. 5. Two anodic peaks, the first ill-defined at low scan rates (6400 mV s-1) and the two amalgamated at the highest scan rates (>2 000 mV s-I), were obtained on the first cycle only, and disappeared on subse- quent cycles at low scan rates or were vestigial at higher scan rates. Three cathodic peaks appeared on all cycles. The peak currents are scan rate dependent; their relationship to square root of scan rate is shown in Fig. 6(b) and the graphs of peak potential against the logarithm of scan rate are shown in Fig.6(c) and (d). The relationships indicate a complex reaction that is difficult to interpret. The cathodic peak currents relate to square root of scan rate in a fashion similar to the anodic peak currents. A scan at an unactivated electrode is shown in Fig. 6(a) and reveals a single peak at a higher current than that in Fig. 5(b), with the peak broadened and shifted to a more positive potential. Two or three fast scans suffice to leave an easily removed purple film on the electrode surface. Coulometric Determination of n-values Potentiostatic coulometry23 at 1.05 V versus S.C.E. to termi- nate at the background residual current in 0.1 mol 1-l sulphuric acid showed a charge transfer of 2.00 electron equivalents per mole for imipramine. Amperostatic coulo- metry at a current density one fifth of the initial limiting current density to extents between 0.1 and 1 electron equiv- alent yielded a solution in which the anodic voltammetric wave was completely suppressed. This behaviour is difficult to explain other than that by accumulation of radical species, particularly polymer radicals, on extended electrolysis an activated electrode is quickly deactivated on immersion in the solution.Amperostatic coulometry of dothiepin in 0.1 mol 1-l sulphuric acid was, however, successful; passage, for example, of a charge equivalent to half an electron equivalent produced a decrease in limiting current of 12.2% and a further equal charge gave a total decrease of 25.0% , identifying the number of electrons as four. There was no regeneration of the wave.2378 ANALYST, JANUARY 1984, VOL. 109 0.7 1.1 1.5 0.7 1 .I I .5 4 c! in vj 0.5 $ 0.9 1.3 0.: 0.: 1 .: 1 : - ~ _ _ :a) p H 7 p H 6 pH 5 p H 4 pH 3 pH2 pH 1 1 M H ~ S O ~ Fig.4. Anodic voltammograms of dothiepin at ( a ) platinum and ( b ) gold RDE and of imipramine at (c) platinum and ( d ) gold RDE. Curves: A, at a freshly activated electrode; B, a repeat scan without intervening activation. Concentrations, 0.001 mol 1-1; rotation speed, 50Hz; scan speed, 5 mVs-l; media, sulphuric acid (pH = 0, 1) or citrate - phosphate buffer, adjusted to exact pH unit 2.0 4 E e . 0 - 2.0 a -4.0 E . - -6.0 -8.0 -0.4 0 0.4 0.8 1.2 -0.4 0 0.4 0.8 1.2 E N vs. S.C.E. Fig. 5. Cyclic voltammograms of nominal 10-3 mol 1-1 imipranine in 0.1 rnol 1-1 sul huric acid at a stationary activated gold electrode, 0.503 cm2.Scan speeds: (a) 1 000 mV s-1; (b) 2 384 mV s-1. Successive cycles markecfl-3, anodic peaks 1 and 2 and cathodic peaks 3, 4 and 5ANALYST, JANUARY 1984, VOL. 109 79 a E 2.0 . -? 0 -2.0 3 -4.0 -!! . -6.0 -8.0 a E . - 5 2 :E 1 10 20 30 40 -0.4 0 0.4 0.8 1.2 EIVvs. S.C.E. Fig. 6 . mol I-’ imipramine in 0.1 mol I-’ sulphuric acid at an unactivated gold electrode at a scan speed of 2384 mV s-1. ( b ) Plot of peak current versus square root of scan speed for anodic peaks 1 and 2 of Fig. 5(a). ( c and d ) Relationship of peak potential for anodic peaks 1 (c) and 2 ( d ) of Fig. 4(a) versus the natural logarithm of scan speed (a) Cyclic voltammetry of nominal Mechanisms The cycloheptene fragment is not itself oxidisable, as shown by the non-reactivity of the dibenzocycloheptenes amitripty- line, nortriptyline and protriptyline, nor is the ethanoanthrac- ene of maprotiline active.Introduction of a heteroatom, such as oxygen in doxepin, which does not present attackable lone pairs does not confer activity, but a heteroatom capable of yielding an electron in cation or radical formation, such as sulphur in the thiepin or nitrogen in the azepines, does permit electrooxidation. The dibenzazepines are the most easily oxidised. The two-electron mechanism, in analogy to that for 5-methyliminobibenzy1725 is illustrated for imipramine in Fig. 7. The first electron is removed from the monomer nitrogen, and the radical can then exist in a number of resonance forms, three of which from the canonical set are shown.The monocation rapidly25 dimerises or reacts with an unoxidised molecule (dimerisation is the predominant pathway for unstable cation radicals). The dimerisation is accompanied by the loss of two protons per dimer, but it is uncertain whether the proton loss occurs before or after coupling. The dimer is more easily oxidised than the monomer and the product dication results with the loss of two electrons per dimer, an ECE mechanism. It is also possible to have a chemical oxidation of the dimer by the monomer radica1726J7 an ECC process, with regeneration of the unoxidised azepine. In non-aqueous media it has been claimed that for structurally related compounds a mixture of ECE and ECC mechanisms prevails.3 For the diazepine, dibenzepin, the influence of the ketone group is seen in the increased oxidation potential. The piperazine ring in the side chain of opipramol is also oxidisable.23 The first wave accords with the mechanism of Fig.7, and the potential is similar to the first wave potentials of the other azepines, but there is a second wave, as shown in Fig. 3, the potential of which is in agreement with the potential of the third wave of piperazineethanol phenothiazine deriva- tives.23 This arises from oxidation, possibly in two successive two-electron steps, which are not anodically separated, to the in-chain 1,6dihydropyrazine derivative. In the thiepin ring of dothiepin, the heterosulphur is oxidised by a different mechanism to the heterosulphur oxidation in the phen~thiazines,~~ because no evidence of unstable radicals has been found.Should they be formed as intermediates, their chemical disproportionation and reaction with solvent molecules must be extremely rapid, and, if the dication diradical sulphoxide is an intermediate it does not oxidise water, but reacts therewith very rapidly to form the sulphone, which is not formed by the phenothiazines. The four-electron oxidation of dothiepin goes right through to the sulphone without revealing any intermediate steps, although these may be involved in fast processes. The over-all reaction is that in Fig. 8, only stable products being formed. Kinetic Parameters Half-wave potentials, mass- and charge-transfer rate constants and charge-transfer coefficients (p) have been determined for dothiepin and the amenable azepines; desipramine, dibenze- pin, imipramine, opipramol and trimipramine in 120 sets of definitive measurements in concentration ranges up to 10-2 mol 1-1 and at rotation speeds up to 50 Hz at platinum and gold electrodes in a medium of 0.1 moll-’ sulphuric acid at 25 “C.Tabulations of the results are available from the authors. The half-wave potentials fall into two groups, dibenzepin and dothiepin in the range 1.03-1.19 V versus S.C.E. and the remainder in the range 0.78-0.86 V versus S.C.E., offering sufficient separation to permit the voltammetric determination of the sum of each group in mixtures. Clomipramine would interfere by tenacious adsorption. The half-wave potentials are slightly higher at gold than at platinum.All show an increase with increasing rotation speed and generally a slight increase with increasing concentration. The charge-transfer rate constants fall within the extreme range 1.3-15.5 x 10-6 1 cm-2 s-l, and increase with increasing rotation speed but vary little with concentration. The charge-transfer coefficients lie within the extremes of 0.2-0.5, and have a small tendency to decrease with increasing rotation speed , but little significant change occurs with change of concentration. Conditional80 ANALYST. JANUARY 1984, VOL. 109 I R I R I R / \ + QQJ I R I R I R I R Fig. 7. Reaction mechanism for azepines. R = (CH2)3N(CH3)2 for imipramine Fig. 8. Anodic reaction of dothiepin potentials are not accessible, so the charge-transfer para- meters are referenced to half-wave potentials and are com- puted by pattern theory.28 We thank Messrs.Boots Co. Ltd., Ciba Laboratories, Geigy Pharmaceuticals, Eli Lilly, May and Baker Ltd., Merck, Sharp and Dohme, Sandoz Products Ltd. and William R. Warner for the gift of the materials listed in Table 1, and the Royal Society for the SEL Transfer Standard DVM. W. H. thanks the Government of Pakistan for the award of a Scholarship and the University of Karachi for the grant of leave of absence. 1. 2. 3. 4. 5. References Brunt, K., Anal. Chim. Acta, 1978,98, 93. Elliathy, M. M., and Volke, J., Collect. Czech. Chem. Commun., 1978,43,812. Butkiewics, K., J . Electroanal. Chem., 1972, 39, 407. Wallace, J . E., and Biggs, J . D., J . Forensic Sci., 1969, 14,528. Heck, H. d’A., Flynn, N.W., and Anbar, M., Biomed. Mass Spectrom., 1978,5,250. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Moody, J. P., Tait, A. C., and Todrick, A., Br. J . Psychiatry, 1967, 113, 183. Sutfin, T. A., and Jusko, W. J., J. Pharm. Sci., 1979,68,703. Wedder, H. J . , and Bickel, M. H., J. Chromatogr., 1962, 37, 181. Pierce, W. O., Lamoreaux, T. C . , and Finkle, B. S . , J. Anal. Toxicol.., 1978, 2 , 26. Roseel, M. T., Bogaert, M. G., and Clacy, M., J . Pharm. Sci., 1978, 67, 802. Mausze, R., Dev. Neurosci., 1980, 7, 125. Gifford, L. A., Turner, P., and Pare, C. M. B., J. Chromat- ogr., 1975, 105, 107. Bertrand, M., Dupuis, C., Gagnon, M. A., and Dugal, R., Clin. Biochem., 1978, 11, 117. Chan, T. L., and Gershon, S . , Mikrochim. Acta, 1973,435. Westenberg, H. G., Drenth, B. F., and Korf, J., J. Chromat- ogr., 1977,142,725. Vandemark, F. L., Adam, R. F., and Schmidt, G. J., Clin. Chem., 1981, 24, 87. Suckow, R. F., and Cooper, T. B., J. Pharm. Sci., 1981, 70, 257. Loh, A., Zuleski, F. R., and Dicarlo, F. J., J. Pharm. Sci., 1977,66, 1056. Read, G. F., and Riad-fahmy, D., Clin. Chem., 1978, 24, 36. Hammer, C. G., Alexanderson, B . , and Holmstedt, B . , Clin. Pharmacol. Ther., 1971, 12,596. Crampton, E. L., Glass, R. C., Marchant, B., and Res, J. A., J . Chromatogr., 1980, 183, 141. Hammer, W. M., and Brodie, B. B . , J . Pharmacol. Exp. Ther., 1967,157,503. Bishop, E., and Hussein, W., Analyst, 1984,109, in the press. Bishop, E., and Hussein, W., Analyst, 1984, 109,65. Frank, S. N., Bard, A. J., and Ledwith, A., J. Electrochem. Soc., 1975, 122, 898. Nicholson, R. S., and Shain, I., Anal. Chem., 1964, 36, 706. Seo, E. T., Nelson, R. F., Fritsch, J. M., and Marcoux, L. S . , J . Am. Chem. SOC., 1966,88, 3498. Bishop, E., Analyst, 1972,97,761. Paper A31280 Received August 22nd, 1983 Accepted September 13th, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900073

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST. JANUARY 1984. VOL. 109 81 Determination of Residues of Dithianon in Apples by High-performance Liquid Chromatography Paul G. Baker and Paul G. Clarke Department of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SEI 9NQ, UK A method for the determination of dithianon in apples is presented. After extraction with acetone - hydrochloric acid, the dithianon is separated from co-extracted compounds by a partition process and chromatography on a silica-gel column before quantitative determination by high-performance liquid chromatography using an ultraviolet spectrophotometric detector. Confirmation of residues may be achieved by high-performance liquid chromatography using post-column derivatisation. Keywords: Dithianon residue determination; high-performance liquid chromatography; apples Dithianon [5,10-dihydro-5,10-dioxonaphtho-(2,3-b)-1,4- dithiin-2,3-dicarbonitrile] is a non-systemic fungicide effective against foliar diseases of pome fruit and is included in the list of Approved Products for Farmers and Growers1 for use in the control of scab on apples and pears.There are few published methods for the determination of dithianon residues. Sieper and Pies2 developed a spectro- photometric method for determining dithianon in apples, based on the formation of a coloured complex with morphol- ine. Amodori and Heupt3 described a similar spectrophoto- metric method for various fruits but using a Florisil column clean-up before reaction with morpholine and spectrophoto- metric determination of the product at 510 nm.More recently, Kojima et al.4 presented a high-performance liquid chromato- graphic method using post-column colorimetric derivatisation for the determination of dithianon in various fruits and vegetables and Burchberger and Winsauers examined the use of liquid chromatography with an electrochemical detector. This paper presents a rapid and simple method based on a modified version of the extraction procedure of Kojima et al. with the direct determination of dithianon by conventional liquid chromatography using an ultraviolet spectrophoto- metric detector. Method Apparatus The following apparatus was used: chromatographic columns, glass columns 200 mm in length x 15 mm i.d. fitted with a PTFE stopcock; an MSE homogeniser with vortex beaker (catalogue number 7700); and a rotary evaporator.Liquid chromatography System 1. A Waters Associates Model 6000 constant- volume solvent-delivery system was used with a Cecil Instru- ments Model CE 212 variable-wavelength ultraviolet detector set at 250 nm and fitted with a 10-p1 flow cell. The stainless- steel column (250 x 4.6 mm i.d.) was packed with 5-pm Spherisorb ODS (Phase Separations Ltd.) by use of a stirred slurry technique using propan-2-01 as the solvent. Samples were injected by means of a sample injection valve (Rheodyne Inc. Model 7125), fitted with a 2 0 4 loop. The mobile phase was methanol - water (4 + 1) using a flow-rate of 1.5 ml min-1. System 2. The same solvent delivery and detection system was used as above. The stainless-steel column (250 x 4.6 mm i.d.) was packed with 5-pm Spherisorb silica (Phase Separa- tions Ltd.) by use of a stirred slurry technique using propan-2-01 as the solvent.Samples were injected by a stop-flow technique. The mobile phase was propan-2-01 in Crown Copyright. 2,2,4-trimethylpentane (1 + 9) using a flow-rate of 2.0 ml min-1. System 3. As system 1, except the column eluent was mixed with 3% mlVsodium sulphide solution pumped by air pressure (10 p.s.i.) and the mixture was passed through a stainless-steel reaction tube (80 x 4.6 mm i.d.) packed with glass beads (80 mesh) before entering the detector flow cell. The detector wavelength was set at 375 nm. Reagents Analytical-reagent grade materials should be used unless indicated otherwise. Silica gel. Merck Kieselgel 60, 0.063-0.200 mm (70-230 mesh ASTM) treated as follows: heat overnight (15 h) at 130 "C and cool in a dessicator.Transfer a weighed amount into a tightly stoppered flask and add 1.0 M hydrochloric acid in the ratio' 1 ml of acid to 10 g of silica gel. Stopper the flask and shake for 60 min. Sodium sulphate. Anhydrous, granular. Acerone. General-purpose reagent grade. Dichloromethane. Methanol - acetic mid solution. Dilute 1 ml of glacial acetic Dilute hydrochloric acid. Add 1 volume of concentrated Sodium sulphate solution, 3% mlV. Dithianon standard solutions. Prepare a solution containing 100 pg ml-1 in methanol - acetic acid solution and dilute as required. acid to 100 ml with methanol (HPLC grade). hydrochloric acid to 5 volumes of water and mix.Procedure Extraction Transfer 10 g of apple peel into a 150-ml vortex beaker (or equivalent homogeniser beaker) containing 10 g of anhydrous sodium sulphate. Add 80 ml of acetone and 20 ml of dilute hydrochloric acid and homogenise the mixture for 5 min. Decant the solvent through a Buchner funnel under suction using a Whatman No. 3 filter-paper or equivalent. Add a further 40 ml of acetone and 10 ml of dilute hydrochloric acid and homogenise for a further 5 min. Filter the mixture through the Buchner funnel. Wash the vortex beaker with two 10-ml portions of acetone and use the washings to rinse the residue in the Buchner funnel. Transfer the filtrate into a 500-ml separating funnel and wash the Buchner flask with two 20-ml portions of acetone, adding the washings to the filtrate each time.Add 150 ml of sodium sulphate solution and 100 ml of dichloromethane to the separating funnel and shake theANALYST, JANUARY 1984, VOL. 109 mixture gently to avoid emulsion formation. Allow the phases to separate, and run the lower dichloromethane layer into a 500-ml rotary evaporator flask through a short column of anhydrous sodium sulphate. Repeat the extraction with a further 100 ml of dichloromethane. Concentrate the combined extracts just to dryness using a rotary evaporator with the water-bath at 40 "C. Dissolve the residue in dichloromethane and transfer the extract quantitatively into a graduated tube. Adjust the volume to 2.0 ml. Analysis Prepare a chromatographic column by slurry packing 5 g of the treated silica gel in dichloromethane.When the silica has settled in the column, overlay with a 1-cm depth of anhydrous sodium sulphate. Transfer the extract on to the column and allow the solution to pass through it until the liquid level reaches the top of the column. Rinse the sample tube with 1 .O ml of dichloromethane and transfer this into the top of the column. Allow the solution to pass through the column until the liquid level reaches the top of the column. Elute the column with 20 ml of dichloromethane, collecting all the eluates in a 50-ml pear-shaped flask. Reduce the volume of the eluate to about 1 ml using a rotary evaporator at 40 "C. Transfer the extract quantitatively, using dichloromethane, into a graduated tube and reduce the volume just to dryness with a gentle stream of air.Dissolve the residue in 1.0 ml of Table 1. Recovery of dithianon. Recoveries were measured using system 1 Recovery, % Fortification level/ mg kg-1 Standard solution Apple 84 0.05 - 0.25 106 - 81 1 .o 87 2.0 - - 13.5 95 * 82 * *System 3 was used. methanol - acetic acid solution. Inject 10 pl of this solution on to the liquid chromatograph (system 1) using a flow-rate of 1.5 ml min-1 for the mobile phase and compare the peak height with those obtained for similar injections of standard solu- tions. If there is any problem with interference from the sample or further confirmation of identity is required, then the extract should be examined using either system 2 or 3. Results and Discussion The recovery of dithianon from apples was checked by adding known volumes of a dithianon standard solution, in methanol - acetic acid, to 10-g portions of apples shown not to contain dithianon residues. The samples were treated as described under Procedure.The results obtained using reversed-phase liquid chromatography are shown in Table 1. The limit of determination of the method was found to be 0.05 mg kg-1. A small-scale survey of apple samples was carried out using the above method. Apple varieties that are susceptible to apple scab were selected, Twenty three samples, obtained in different parts of the United Kingdom, were examined. Fourteen of these were found to contain dithianon residues at levels between 0.05 and 3.6 mg kg-1 on the peel. This corresponds to <0.01-0.3 mg kg-1 when calculated on a whole fruit basis.Details of the residues found are shown in Table 2. Typical chromatograms obtained from a standard solution and apple peel containing dithianon are shown in Fig. l(a) and (b). Dithianon standards, made up in pure methanol, were found to break down rapidly. These solutions were also difficult to chromatograph and solutions of 20 pg ml-1 or less, gave no peak at all on the HPLC systems, presumably due to on-column breakdown. These problems were overcome by making up the standards with 1% acetic acid in methanol. This gave standard solutions that were stable over a period of months. Solutions of less than 1 pg ml-1 could be put on to the column without any breakdown occurring. It was also found necessary to keep the dithianon in acidic solution throughout the extraction and clean-up procedure, to avoid breakdown.For maximum recoveries it was necessary to analyse and examine the extracts on the same day. Table 2. Results of analysis of apple peel samples Dithianodmg kg- 1 *N.d., Country of origin Newzealand . . . . Italy . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . UK . . . . . . France . . . . . . France . . . . . . France . . . . . . France . . . . . . France . . . . . . France . . . . . . not detected. . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . . . Apple variety Sturmer Pippin Unknown Bramley Bramley Bramley Bramley Bramley Bramley Bramley Cox's Pippin Cox's Pippin Cox's Pippin Cox's Pippin Worcester Worcester Worcester Golden Russet Granny Smith Granny Smith Granny Smith Granny Smith Golden Delicious Golden Delicious ~ System 1 N.d.* Interference 3.6 Interference 0.05 0.1 Interference N.d.0.72 1.5 0.89 0.39 0.62 0.2 N.d. 0.46 0.23 N.d. N.d. N.d. 0.11 N.d. 0.47 System 2 N.d. 0.09 - - - - N.d. - - 1.3 Interference - - - - - - - N.d. - - - -ANALYST, JANUARY 1984, VOL. 109 E i Time __+ Fig. 1. Typical chromato rams for (a) a standard solution containing 100 ng of dithianon; and (b5 a 10-pl sample of apple extract containing 0.89 mg kg-1 of dithianon. HPLC system 1 was used 83 Three HPLC systems were tried. System 1, reversed phase, was found to be the most suitable for routine analysis. System 2, normal phase, could be used for confirmation purposes; however, at low levels, the dithianon peak was occasionally subject to interference from co-extracted material. System 3, post-column derivatisation, was subject to the least interfer- ence and is the best system for confirmation. It was, however, subject to base-line drift and took the longest of the three systems to stabilise. Also, as can be seen from Table 2, the number of occasions on which there was interference with either systems 1 or 2 was small and therefore the extra work involved in setting up system 3 may not be justified. 1. 2. 3. 4. 5. References Ministry of Agriculture Fisheries and Food, “1981 List of Approved Products and their uses for Farmers and Growers,” HM Stationery Office, London, 1983. Sieper, H., and Pies, H., Fresenius Z. Anal. Chem., 1968,242, 234. Amadori, E., and Heupt, W., Anal. Methods Pestic. Plant Growth Regul., 1978, 10, 181. Kojima, M., Sekigawa, N., Matano, O., and Goto, S., Bunseki Kagaku, 1980,29,738. Burchberger, W., and Winsauer, K., Mikrochim. Acta, 1980, 257. Paper A31268 Received August 17th, 1983 Accepted September 29th, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900081

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984. VOL. 109 85 Multi-residue Determination of Organochlorine, Organophosphorus and Synthetic Pyrethroid Pesticides in Grain by Gas - Liquid and High-performance Liquid Chromatography Peter Bottomley and Paul G. Baker Department of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SE7 9NQ, UK A multi-residue method for the determination of organochlorine, organophosphorus and synthetic pyrethroid pesticides and carbaryl in grain is presented. After extraction with acetone - methanol, the pesticides are separated from co-extractives by a partition process with dichloromethane and chromato- graphy on an acidic aluminium oxide column. Quantitative determinations are made by packed column gas - liquid chromatography using an electron-capture detector for the organochlorine pesticides and a flame-photometric detector for the organophosphorus pesticides.High-performance liquid chromatography using an ultraviolet spectrophotometric detector is used for the determination of synthetic pyrethroids and carbaryl. Capillary column gas - liquid chromatography is used for confirmation of identity of suspected residues of organochlorine, organophosphorus and synthetic pyrethroid pesticides. Keywords: Synthetic p yrethroids; organochlorine and organophosphorus pesticides; grain; gas - liquid chromatography; high-performance liquid chromatography A large number of pesticides are in common use as grain protectants. Some, such as malathion, have been in world- wide use for over 20 years whilst others, such as bioresmeth- rin, bromophos, dichlorvos, fenitrothion and pirimiphos- methyl, have been in use for 6-12 years.' The emergence of malathion-resistant strains of insect pests has led to the development of alternative grain protectants.Pyrethrins, which have been widely used for over two decades, are now being superseded by synthetic pyrethroids because of their greater photostability and enhanced insecticidal activity. However, control of some pest species may be made difficult by the presence of strains that are resistant to pyrethroids.2-5 The use of a combination of pesticides with a broad spectrum of activity has been evaluated in commercial grain The Food and Agriculture Organization and World Health Organization have recommended residue limits for biores- methrin, bromophos, carbaryl, chlorpyrifos-methyl, delta- methrin, dichlorvos, etrimfos, fenitrothion, fenvalerate, malathion, methacrifos, permethrin, phenothrin, pirimiphos- methyl and pyrethins used for the protection of grain.' A proposal for a Directive on the fixing of maximum levels for pesticide residues on cereals intended for human consumption has been published by the Commission of the European Communities,8 in which maximum residual levels are given for 26 compounds.The maximum residue limits for pesticides on grain are listed in Table 1. Only a few multi-residue methods for the determination of pesticide residues in grain have been published and none of these covers the wide range of pesticides listed in Table 1. Sahag developed a method for the detection of heptachlor, aldrin, dieldrin and endrin in wheat using a Soxhlet extraction followed by partition into light petroleum and clean-up on a magnesium oxide - Celite column.The residues were determi- ned by gas chromatography using electron-capture detection. Levi et aZ.10 reported a rapid screening method using gas chromatography with electron-capture detection for the deter- mination of organochlorine pesticides in wheat. This method was later modified11 to permit the simultaneous screening of organochlorine and organophosphorus residues using electron-capture and flame-photometric detectors. Mestres et al. 12 developed an elaborate extraction and clean-up proce- dure for the determination of organochlorine, organophos- phorus, natural pyrethrin and synthetic pyrethroid pesticides Crown Copyright.in cereal products. Gas chromatography using selective detectors was employed. The Committee for Analytical Methods of the Ministry of Agriculture, Fisheries and Food Table 1. Maximum residue limits for pesticide residues in grain and analytical limits of determination Pesticide Aldrin Dieldrin } ' ' * * Bioresmethrin . . Bromophos . . . . Carbaryl . . . . Chlordane . . . . Chlorpyrifos-methyl DDT(tota1) . . . . Deltamethrin . . . . Diazinon . . . . Dichlorvos . . . . Endosulfan . . . . Endrin . . . , . . Etrimfos . . . . Fenitrothion . . . . Fenvalerate . . . . Heptachlor/heptachlor epoxide . . . . Hexachlorobenzene wHCH . . . . . . P-HCH . . . . . . y-HCH . . . . . . Malathion . . . . Methoxychlor .. Methacrifos . . . . Permethrin . . . . Phenothrin , , . . Phosphamidon . . Pirimiphos-methyl . . Pyrethrins . . . . Piperonyl butoxide . . Trichlorphon . . . . FAOI Proposed EC WHO residue residue limit/ limit/ mg kg- I mg kg- 1 . . 0.02 0.02 . . . . . . 0.5 . . 0.05 . . 0.05 . . 0.1 . . 2 . . 0.2 (maize) 0.1 (other cereals) * . 0.02 . . . . * . . . . . . . 0.02 . . 0.01 . . 0.1 . . 0.02 . * 0.5 . . 8 . . 2 . . * . . . . . 0.1 . . 3 . . 20 . . 0.1 . . 5 10 5 0.05 10 0.2 2 0.1 2 0.02 10 10 5 0.02 0.05 0.5 8 10 2 5 10 3 20 0.1 Limit of determination/ mg kg- 1 0.01 0.05 0.05 0.05 0.05 0.02 0.05 0.05 0.05 0.05 0.05 0.05 0.002 0.003 0.006 0.05 0.05 0.1 0.05 0.05 0.0586 ANALYST, JANUARY 1984, VOL. 109 has collaboratively studied their recommended method for the determination of malathion and dichlorvos in grain for a wider range of organophosphorus pesticides.l 3 In the method described here samples of grain can be screened for residues of ten synthetic pyrethroids, ten organophosphorus pesticides, nine organochlorine pesticides and carbaryl using packed column or capillary column gas - liquid chromatography (GLC) and high-performance liquid chromatography (HPLC). Experimental For speed and simplicity, a common extraction procedure is employed for all the pesticides. This consists of extraction with a mixture of acetone and methanol followed by partitioning into dichloromethane. Carbaryl is determined in this extract by HPLC and organophosphorus pesticides can be deter- mined using GLC with a flame-photometric detector.It is necessary to clean up the extract further for the determination of organochlorine and synthetic pyrethroid pesticides by the use of column chromatography on aluminium oxide. This additional clean-up is optional for organophosphorus pesti- cides but carbaryl is not eluted from the aluminium oxide column under the conditions used. The eluate from the aluminium oxide column is examined by GLC with electron- capture detection for organochlorine pesticides and by HPLC for synthetic pyrethroids. The extract is then injected on to the capillary column gas chromatograph for confirmation of identity of suspected residues. HPLC system (a) takes only 30 rnin to screen for all ten pyrethroids, but suffers from the disadvantage that complete resolution of all the components is not achieved.14 The ten organophosphorus compounds all elute from the system before the synthetic pyrethroids. Chlorpyrifos-methyl, with a retention time of 8.5 min, is the last of the organophosphorus compounds to elute.Carbaryl has a retention time of only 4 rnin on this system. An injection of a mixture of nine organochlorine compounds gave rise to only three peaks, all of which had a retention time of less than 10 min. Thus there is no interference to the pyrethroids from other pesticides in the scheme. Of the pesticides examined, only carbaryl, with a retention time of 10 min, was eluted from HPLC system (b). The GLC system (b) takes only 13 rnin to screen for all ten organophosphorus pesticides (Fig. 3). No peaks were obser- ved on this system for any of the other pesticides examined.This system does, however, suffer from the disadvantage that phosphamidon and fenitrothion are not resolved and also that trichlorphon and dichlorvos are not resolved. It is known that trichlorphon decomposes to dichlorvos at elevated tempera- tures and it is possible that this occurs on the GLC column. The 25-m capillary column GLC system (a) gave the best resolution for all compounds, although two separate temper- ature programmes were required. One programme was required for organophosphorus and organochlorine pesticides and another for synthetic pyrethroids which have much longer retention times. The two groups of compounds do not overlap on either of the temperature programmes and therefore there is no cross-interference.All ten synthetic pyrethroids may be determined on this system, with each group of isomers being clearly separated (see Fig. 1). The organochlorine compounds are separated into two distinct groups [see Fig. 2(a)], a-HCH, P-HCH, y-HCH and HCB elute between 24.5 and 27.5 min, while dieldrin, p,p’-DDE, o,p-DDT, p,p’-TDE and p , p ’ - DDT elute between 41.5 and 48 min. Eight of the ten organophosphorus pesticides elute in a single distinct group between phosphamidon (29 min) and bromophos (37 min) [see Fig. 2(b)]. Dichlorvos elutes after 5.5 min and trichlor- phon at 15 min. It is worth noting that these two compounds are well resolved on this system. Phosphamidon gave two peaks on the capillary column GLC system, which probably correspond to the 2- and E-isomers.The GLC system (c) takes approximately 20 rnin to screen for all nine organochlorine compounds. Although carbaryl and pyrethroids do not interfere in this analysis, it was found that the organochlorine and organophosphorus compounds elute from the column simultaneously. This means that a peak on this system corresponding to an organochlorine or organo- phosphorus pesticide must be confirmed on one of the other systems. The method was checked by using fortified samples of 1 ,? d 9 10 16 0 30 Timelmin 60 Fig. 1. Separation of ten syntheticpyrethroids (10 ngof each) on a25-m OV-101 WCOTcapillary column [temperature programme (i)]. Peaks: 1, cismethrin; 2, bioresmethrin; 1 and 2, resmethrin; 3 and 4, tetramethrin; 5 , fenpropathrin; 6 and 7. phenothrin; 8.cis-permethrin; 9, trans-permethrin; 10, 11, 12 and 13, cypermethrin; 14 and 15, fenvalerate; and 16, deltamethrinANALYST, JANUARY 1984, VOL. 109 87 0 30 Time/min 60 Fig. 2. Separation of nine organochlorine compounds (1 ng of each) and ten organophosphorus compounds (10 ng of each) on a 25-m OV-101 WCOT capillary column [temperature programme ($1. ( a ) Peaks; 1, a-HCH; 2, HCB; 3, y-HCH; 4, P-HCH; 5, $eldrin;.6, p,p'-DDE; 7 , 0 , -DDT; 8, p,p'-TDE; and 9, p,p'-DDT. , Impurity from hexane. (fi Peaks: 1, dichlorvos; 2, trichlorphon; 3 and 6, phosphamidon; 4, diazinon; 5 , etrimphos; 7, chlorpyrifos-methyl; 8, fenitrothion; 9, pirimiphos-methyl; 10, malathion; and 11, bromophos barley and wheat. These were prepared by adding known volumes of mixed standard solutions of the pesticides in hexane to 10-g portions of finely ground grain shown not to contain residues of pesticides.The flask containing the fortified grain was shaken to ensure even distribution of the pesticides and, after allowing the hexane to evaporate with the aid of a gentle stream of air, the grain was left at room temperature for at least 2 h before extraction, in order for the pesticides to be absorbed on to the grain and correspond more closely to a treated commercial grain. Unfortified samples gave no observed interference on the HPLC systems ( a ) and (b) or on the GLC systems ( a ) and ( b ) . However, extracts of wheat gave an interfering peak corre- sponding to y-HCH and a similar extract of barley gave interfering peaks corresponding to y-HCH and p,p'-DDT on GLC system (c).This means that the capillary column system should be used for confirmatory purposes for these two compounds. Method Apparatus Centrifuge. Chromatographic columns. Length 250 mm, i.d. 15 mm, Grinder. A domestic coffee grinder is suitable. Homogeniser. Kuderna - Danish evaporator. Capacity 500 ml. fitted with a PTFE stopcock. Liquid chromatographs ( a ) Synthetic pyrethroids. A Waters Associates Model 6000A constant-volume solvent-delivery system was used. A variable-wavelength ultraviolet detector (Cecil Instruments Model CE 212) fitted with a 10-pl flow cell and set at 206nm was employed. The stainless-steel column (250 x 4.6 mm i.d.) was packed with a 5-pm Spherisorb ODS (Phase Separations) by use of a stirred slurry technique using propan-2-01 as the solvent.Samples were injected by means of a sample injection valve (Rheodyne Model 7125) fitted with a 20-pl loop. (b) Carbaryl. As above, except that the stainless-steel column (250 x 4.6 mm i.d.) was packed with 5-pm Spherisorb silica (Phase Separations). A fine-mesh wire disc was fitted on top of the column packing material and was covered with a 10-mm layer of 80-mesh silanised glass beads. Samples were injected with a 10-pl syringe via a needle guide on to the centre of the disc using a stop-flow technique. The ultraviolet detector was set at 224 nm. Gas chromatographs ( a ) Carlo Erba Fractovap 4200. This was equipped with a split - splitless injector, a temperature programmer (LT 400) and a nickel-63 electron-capture detector. The electron- capture detector was operated at a temperature of 275 "C in the constant-period mode with a pulse width of 0.1 ps.The chromatographic column was a 25 m x 0.23 mm i.d. wall-coated open-tubular (WCOT) fused-silica capillary with an OV-101 stationary phase of film thickness 0.12 pm. The column carrier gas was argon-methane (9 + 1) with a flow-rate of 1.0 ml min-1. Two different temperature programmes were used, as follows. (i) For synthetic pyrethroids, the injector tempera- ture was 215 "C and the column temperature was isothermal at 60 "C for 2 min followed by temperature programming to 215 "C at 30 "C min-1, the final temperature being maintained for 55 min. (ii) For organophosphorus and organochlorine compounds, the injector temperature was 215 "C and the column temperature was isothermal at 100 "C for 10 min followed by temperature programming to 195 "C at 3 "C min-1, the final temperature being maintained for 15 min.Splitless injection was used, opening the splitter 30s after injection. Additional "make-up" gas [argon - methane (9 + l)] was added to the capillary effluent at a flow-rate of 20 ml min-1. The septum purge was 1.Oml min-1 with argon - methane (4 + 1) and the splitter valve flow-rate was 20 ml min-l, also with argon - methane (9 + 1). The chromatograms were recorded using a Shimadzu C-RIA computing integrator. ( 6 ) For organophosphorus compounds. A Varian Model 3700 chromatograph equipped with an auto linear tempera- ture programmer and a dual flame-photometric detector was used. The detector was operated at 250 "C with a hydrogen flow-rate of 140 ml min-1 and air flow-rates of 80 (Air 1) and 170 ml min-1 (Air 2).The stationary phase was 3% OV-225 on Gas-Chrom Q (80-100 mesh), packed into a glass column (1 m X 4 mm i.d.). The column carrier gas was nitrogen with a flow-rate of 50 ml min-1. The standardised chromatographic conditions were as follows: injector temperature, 240 "C; column temperature, isothermal at 130 "C for 2 min followed by temperature programming to 220 "C at 10 "C min-1, the final temperature being maintained for 2 min. ( c ) For organochlorine compounds. A Pye 104 instrument fitted with a nickel-63 electron-capture detector was used isothermally. The detector was operated in the pulsed mode at a fixed frequency using the standard Pye 104 supply unit (pulse interval set to 150 ps) and amplifier.The detector was used at 250 "C with a flow-rate of nitrogen purge gas of 10 ml min-1. The stationary phase was a mixture of 0.4% OV-17 and 2.6% OV-210 on Gas-Chrom Q (80-100 mesh), packed into a glass column (2 m x 4 mm i.d.). The column was operated at 175 "C with a nitrogen carrier gas flow-rate of 40 ml min-1. The temperature of the injection port was maintained at 175 "C. Reagents Analytical-reagent grade materials were used unless otherwise indicated.88 ANALYST, JANUARY 1984, VOL. 109 Cotton-wool. Wash with hexane - acetone (1 + 1). Anti-bumping granules. Wash with hexane - acetone (1 + Aluminium oxide. Hopkin and Williams CAMAG M.F.C., Sodium sulphate, anhydrous, granular. Acetone.Rathburns glass-distilled grade. Dichloromethane. Hexane. Rathburns HPLC grade. Methanol. Fisons HPLC grade. Propan-2-01. 2,2,4-Trimethylpentane. Rathburns HPLC grade. Water, glass-distilled. Mobile phases for liquid chromatography. The mobile phases used were methanol - water (4 + 1) and 2,2,4- trimethylpentane - propan-2-01 (9 + l), both degassed. 1)- pH 4.5, Brockmann activity 1. Pesticide stock standard solutions Prepare 100 mg ml-l solutions of the pesticides in hexane (pyrethroids, organochlorine pesticides and carbaryl) or acetone (organophosphorus pesticides). These solutions are all stable for at least 1 month if stored in the dark at 4 “C. Pesticide working standard solutions Prepare 2.0 mg ml-1 solutions by diluting 1 ml of the pesticide stock solution to 50 ml with hexane or acetone.Procedure Extraction Transfer 10 g of finely ground grain into a 200-ml centrifuge bottle. Add 100 ml of acetone - methanol (1 + 1) and homogenise the mixture for 1 min. Centrifuge the mixture at 2 500 rev min-1 for 3 min and then decant the solvent through a filter funnel fitted with a glass sinter into a 1000-ml separating funnel. Add a further 75 ml of acetone - methanol (1 + I) to the centrifuge bottle and homogenise for 1 min. Centrifuge the mixture for 3 min and decant the solvent through the filter funnel into the separating funnel. Add 400 ml of sodium sulphate solution (2.5 g per 100 ml) to the separating funnel, followed by 50 ml of dichloromethane, and shake vigorously for 1 min. Allow the phases to separate and run the lower organic phase through 25 g of anhydrous sodium sulphate in a 15 mm diameter glass column.Collect the dry solvent in a Kuderna - Danish evaporator fitted with a 10-ml graduated test-tube containing two anti-bumping granules. Repeat the extraction with two further 50-ml portions of dichloromethane, adding each portion to the evaporator. Finally, wash the sodium sulphate column with 10 ml of dichloromethane and concentrate the combined extracts and washings to 1 ml on a steam-bath. Column clean-up Using a Pasteur pipette, transfer the dichloromethane extract on to a 15-g aluminium oxide column that has been slurry- packed in dichloromethane. Allow the solution to pass through the column until the liquid level reaches the top of the column. Rinse the test-tube with two 5-ml portions of dichloromethane and transfer the washings to the column. Allow the washings to pass through the column.Elute the column with 160 ml of dichloromethane, collecting the washings and eluate in a Kuderna - Danish evaporator fitted with a 10-ml graduated test-tube containing two anti-bumping granules. Concentrate the combined washings and eluate to 1 ml on a steam-bath. Remove the remaining solvent using a warm water-bath and a gentle stream of air and dissolve the residue in an appropriate volume of hexane or acetone. Analysis Organophosphorus pesticides and carbaryl. For carbaryl the column clean-up is omitted and for organophosphorus com- pounds analysed using a flame-photometric detector the column clean-up is optional. Thus the analysis of carbaryl and organophosphorus compounds is carried out before the column clean-up step.Remove any dichloromethane remaining from the extrac- tion step using a warm water-bath and a gentle stream of air and dissolve the residue in 2 ml of hexane or acetone. Inject 10 pl of this solution on to the HPLC system ( 6 ) using a flow-rate of 1.0 ml min-1 for the mobile phase and compare the peak height of carbaryl with that obtained for a similar injection of standard solution. Inject 10 pl of this solution on to the packed column GLC system (b) with flame-photometric detection and compare the peak heights with those obtained for similar injections of organophosphorus pesticide standard solutions. Organochlorine pesticides and synthetic pyrethroids. For organochlorine compounds and synthetic pyrethroids the aluminium oxide column clean-up step is required in order to remove possible interference from co-extracted material.Remove any solvent remaining from the solution analysed above using a warm water-bath and a gentle stream of air and dissolve the residue in 2 ml of dichloromethane. Transfer this solution on to a 15-g aluminium oxide column and follow the procedure described above. Dissolve the residue in an appropriate volume of hexane or acetone. Inject 10 1.11 of this solution on to the HPLC system ( a ) using a flow-rate of 1.0 ml min-l for the mobile phase and compare the peak heights of the synthetic pyrethroids with those obtained for similar injections of standard solutions. Inject 10 p1 of this solution on to the packed column GLC system ( c ) with electron-capture detection and compare the peak heights with those obtained for similar injections of organochlorine pesticide standard solutions. Inject 1 .O pl of this solution into the capillary column gas chromatograph and compare peak heights with those obtained for similar injec- tions of standard solutions.Results The recovery of organochlorine, organophosphorus and synthetic pyrethroid pesticides and carbaryl from barley and wheat was checked by adding known volumes of mixed standard solutions in hexane to 10-g portions of finely ground grain shown not to contain residues of the pesticides. The samples were then treated as described under Procedure. The results obtained are shown in Tables 2-4 and a typical chromatogram obtained from fortified wheat is shown in Fig.3. It was difficult to quantify the results obtained by capillary column GLC and the main use of this technique is for additional confirmation of identity of a residue. To ascertain the practical limit of determination for each pesticide, 10-g portions of wheat and barley were fortified at the 0.1 mg kg-1 level for carbaryl and organophosphorus pesticides, 0.05 mg kg-1 for organochlorine pesticides and 1 mg kg-1 for synthetic pyrethroids. The limits of determination are shown in Table 1. In previous studies on the quantitative determination of organophosphorus pesticides in grain,13 problems caused by co-extractives necessitated repeated injections of the sample extract to “condition” the GLC column before the preparation of calibration graphs and the quantitative examination of grain extracts.This “grain effect” causes some compounds to exhibit an increased response to the flame-photometric detector when they are in a grain extract rather than in a standard solution. In order to verify the quantitative results obtained for the recoveries of organophosphorus pesticides and to check whether any enhancement effect was occurring,ANALYST, JANUARY 1984, VOL. 109 89 Table 2. Recovery of organochlorine pesticides from fortified grain samples using gas - liquid chromatography Table 4. Recovery of carbaryl and synthetic pyrethroids from fortified grain samples using high-performance liquid chromatography Pesticide p,p’-DDE . . . . o,p’-DDT . . . . p,p’-DDT . . . . Dieldrin(HE0D) .. Hexachlorobenzene . . CX-HCH . . . . . . P-HCH . . . . . . Y-HCH . . . . . . p,p’-TDE . . . . * Interference. . . . . . . . . . . . . . . . . . . Fortifica- tion level/ mg kg- 1 . . 0.04 0.05 . . 0.04 0.05 . . 0.04 0.05 . . 0.04 0.05 . . 0.01 0.05 . . 0.01 0.05 . . 0.04 0.05 . . 0.01 0.05 . . 0.04 0.05 Recovery, YO Barley 121,97 123,109 - - - * - 104,91 93,94 87,84 107,101 - - - - * - 89,87 Wheat 92 128,107 81 90,94 100 99,100 92 130,120 83 91,96 100 85,88 124 80,85 * * 100 77,80 Table 3. Recovery of organophosphorus pesticides from fortified grain samples using gas - liquid chromatography Pesticide B romophos . . . . . . . . Chlorpyrifos-methyl . . . . Diazinon . . . . . . . . Dichlorvos + trichlorphon* . . Etrimfos . . . . . . . . Malathion . . . .. . . . Phosphamidon + fenitrothion” Pirimiphos-methyl . . . . Fortifica- tion level/ mg kg- 1 0.1 0.5 7.0 0.1 0.5 3.5 0.1 0.5 2.0 0.1 + 0.1 0.5 + 0.5 2.0 + 5.5 0.1 0.5 10.0 0.1 0.5 8.0 0.1 + 0.1 0.5 + 0.5 2.0 + 4.0 0.1 0.5 2.0 Recovery, YO Barley 103,116 - 99 - 104,114 96 92,94 95 87,94 72 94,93 99 102,110 97 77,81 - - - - - 101 109,120 - 99 Wheat 106 95,97 105 80 92,91 99 70 92,91 100 88 86,84 77 73 92,92 103 129 106,104 107 54 76,75 106 84 98,98 99 * Dichlorvos and trichlorphon, and phosphamidon and fenitrothion are not resolved. samples of wheat and barley were extracted as described under Procedure without the aluminium oxide column clean- up step. The extracts were evaporated to dryness and the residues dissolved in 2 ml of a solution containing the ten organophosphorus pesticides.The solution was left overnight and then injected on to the GLC system (b) under normal conditions without any column pre-treatment. No detectable enhancement was observed in the grain extracts when compared with a standard solution. The possible interference , with the determination of synthetic pyrethroids by natural pyrethrins (pyrethrum Fortifica- level/ tion Recovery, % Pesticide mg kg- 1 Barley Wheat Carbaryl . . Bioresmethrin Cismethrin . . Cypermethrin Deltamethrin Fenpropathrin Fenvalerate . . Permethrin . . Phenothrin . . Te tramet hrin * Interferences. . . 0.1 0.2 0.5 . . 1.0 . . 1.0 . . 1.0 . . 1.0 . . 1.0 . . 1.0 . . 1.0 . . 1.0 . . 1.0 85 85 93,78 104,100 97 88 85,66,67 108,72,66 89,69,69 11 1,77,63 70,85,79 83,97,79 95,100,94 92,100,91 146,* 102,107 108,106,79 99,113,105 82,98,98 106,92,92 114,89,76 128,111,89 176,* 89,89 113,98,99 119,105,107 I I 1 I 0 5 10 15 Ti rne/rni n Fig.3. Separation of ten organophosphorus compounds in wheat spiked at 0.1 mg kg-1 on a 1 m x 4mm i.d. OV-225 packed column. Attenuation, 32. Peaks: 1, dichlorvos and trichlorphon; 2, diazinon; 3, etrimphos; 4, chlorpyrifos-methyl; 5, pyrimiphos-methyl; 6, bromo- phos; 7, malathion; and 8, fenitrothion and phosphamidon extract, WHO World Standard, 1972) on the capillary column GLC and the HPLC system (a) has been previously exami- ned.14 Natural pyrethrins could be separated into their six main constituent components on both systems, although none of the peaks coincided with any of those from synthetic pyrethroids.A number of other substrates have been examined by this method, including oranges, apples, grapes, bananas and tomatoes.15 A detailed discussion of the results is beyond the scope of this paper, but preliminary investigations suggest that the recoveries of pesticides are satisfactory and little co- extractive interference occurs. References 1. “Pesticide Residues in Food 1981. Report of the Joint Meeting of the FA0 Panel of Experts on Pesticide Residues in Food and the Environment and the WHO Expert Group on Pesticide Residues,” FA0 Plant Production and Protection Paper, No. 37, FAO, Rome, 1982.90 ANALYST, JANUARY 1984, VOL. 109 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Fine, B. C., Pyrethrum Post, 1963, 7, 18. Farnham, A. W., Pestic. Sci., 1973, 4, 513 Farnham, A. W., and Sawicki, R. M., Pestic. Sci., 1976,7,278. Keiding, J., Pestic. Sci., 1976, 7, 283. Desmarchelier, J. M., J. Stored Prod. Res., 1977, 13, 159. Desmarchelier, J . M., Bengston, M., Connell, M., Henning, R., Ridley, E., Ripp, E., Sierakowski, C., Sticka, R., Snelson, J., and Wilson, A., Pestic. Sci., 1981, 12, 365. Proposal for Council Directives, Off. J. Eur. Commun., 1980, No. C56, 14. Saha, J. G., J . Assoc. Off. Anal. Chem., 1966,49,768. Levi, I., Mazur, P. B., and Nowicki, T. W., J . Assoc. Off. Anal. Chem., 1972,55,794. Levi, I., and Nowicki, T. W., J . Assoc. Off. Anal. Chem., 1974, 57, 924. 12. 13. Mestres, R., Atmawijaya, S., and Chevallier, C . , Ann. Falsif. Expert. Chim., 1979, 72, 577. Working Group of the Committee for Analytical Methods for Residues of Pesticides and Veterinary Products in Foodstuffs and Working Party on Pesticide Residues of the Ministry of Agriculture, Fisheries and Food, Analyst, 1980, 105, 515. Baker, P. G., and Bottomley, P., Analyst, 1982, 107, 206. Bottomley, P., and Baker, P. G., unpublished results. 14. 15. Paper A31243 Received August 4th, 1983 Accepted September 26th, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900085

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984, VOL. 109 91 Development of a Flow Injection Analyser for the Post-column Detection of Sugars Separated by High-performance Liquid Chromatography D. Betteridge," N. G. Courtney and T. J. Sly Department of Chemistry, University College Swansea, Singleton Park, Swansea, SA2 8PP, UK D. G. Porter Department of Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SE7 9NQ, UK A system has been developed for the determination of reducing sugars by flow injection analysis, following separation by high-performance liquid chromatography. The system is described in detail, and possible methods for optimising and extending it are considered. Keywords: Sugar determination; flow injection analysis; high-performance liquid chromatography Over the past 5 years, flow injection analysis (FIA) has become established as a simple, accurate and rapid analytical technique.1-4 The work described in this paper was under- taken in an attempt to apply FIA principles to the selective detection of compounds separated by high-performance liquid chromatography (HPLC), with the eventual aim of producing a multiple detection system consisting of several FIA systems running in parallel and linked to a microcomputer. Such a system would subject samples of eluate from the column to several different chemistries, and thus would be capable of both detecting and characterising chromatographic peaks (Fig.1). As the first stage of this project, the system shown in Fig. 2 was developed and used with considerable success on samples of confectionery, each containing a mixture of saccharides.This consisted of a peristaltic pump used to pump a solution containing a colour-forming reagent to a sampling valve, which injected samples of effluent from the HPLC column at regular intervals into the reagent stream. The solution then passed through a heated reaction coil and into a flow-through photometer to monitor the absorbance of the solution. Colour-forming Reaction The choice of a suitable colour-forming reaction was influen- ced by a number of factors. It was important that the chosen reaction should be quantitative at the microgram scale and that it should proceed fairly rapidly, as the length of reaction coil required for a slow reaction would disperse the peaks to an unacceptable extent. It was also important that the reaction should proceed below 81 "C, the boiling-point of acetonitrile (used as one component of the HPLC eluent mixture), and that no side-reaction with the acetonitrile should take place.Finally, preference was given to reactions that give a visible colour change as these could be monitored using a much simpler and cheaper detector than a reaction that yielded a fluorescent product. Keeping these criteria in mind, the reaction chosen was that of 2,3,5-triphenyltetrazolium chloride (TTC), which reacts with reducing sugars under alkaline conditions to form triphenylformazan, a red compound (Amax. = 564 nm5) that precipitates.6.7 This reaction was unaffected by acetonitrile, and proceeded to completion in about 10 min at 75 "C.It was found, however, that a reaction time of about 3 min gave an adequate colour for detection using a simple photometer at the sample concentration levels of interest. This also made it possible to measure the absorbance of the product before the triphenylformazan had precipitated from solution. * Present address: BP Research Centre, Sunbury-on-Thames, TW16 7LN, UK. Experimental Apparatus A schematic diagram of the apparatus is shown in Fig. 2. The equipment used consisted of an HPLC pump (ACS Model 750/03; Applied Chromatography Systems Ltd., Luton), which pumped eluent at a rate of 1 ml min-1 to a 4 mm i.d. column packed with 5-pm amino-bonded silica (Spherisorb S5-NH,; Phase Separations Ltd., Queensferry, Clwyd) via a modified high-pressure valve that served as a port for direct syringe injection of samples on to the column.Two alternative columns, of length 150 and 250mm, were used. Eluate from the column flowed into the sample loop of a sampling valve, and thence to waste. Reagent solution was pumped at 1.2ml min-1 by a peristaltic pump (Masterflex drive 7544-20 with pump head 7013; CP Laboratories Ltd., Bishop's Stortford) through a pulse suppressor (300 X 2.3 mm i.d. thin-walled PVC tube) into the sampling valve. This consisted of an eight-port Altex slider valve operated by two pneumatic actuators (Anachem Ltd., Luton) fitted with a 12-@ sample loop. The compressed air supply to the actuators was controlled by a solenoid valve operated from an electronic timer, which allowed the sam- pling rate and period to be adjusted.HPLC column ll I! Pump 1 Multi-Dort v v a a i G Pump 2 sampling valve (bulk memory) Fig. 1. column detection system Schematic diagram of the microcomputer-controlled post-92 SamDle ANALYST, JANUARY 1984. VOL. 109 Table 1. Effect of sampling period on peak height HPLC column Reagei l t + Waste Fig. 2. Schematic arrangement of the apparatus Solution passed from the valve through a reaction coil of stainless-steel tubing (2 m x 0.8 mm i.d.) wound around a metal cylinder fitted with a thermostatically controlled electric heater and mounted inside an insulated casing. For these experiments a temperature of 70°C was used. Solution flowed from the reaction coil to the photometer and thence to waste via a coil of PTFE tubing (5 m x 0.3 mm i.d.), which served to increase the back-pressure in the system and thus inhibit the formation of bubbles of acetonitrile vapour, which could give rise to spurious peaks from the detector.A similar coil was fitted after the sampling valve in order to prevent bubbles being formed in the sample loop. The flow-through phototransducer used was developed from that of Betteridge et a1.8 Improvements were made to the electronics and a new flow cell was developed in which the light path was normal to the direction of flow. These changes resulted in greatly improved reliability and reduced base-line noise, which extended the usable sensitivity of the unit. These developments are described more fully elsewhere.9 Signals from the detector were monitored with a chart recorder (Servoscri be RE5 1 1.20).All tubing used in the FIA system was 0.8mm i.d. PTFE, except where stated otherwise. The reagent solution was photosensitive3; therefore, all tubes carrying reagent solution were wrapped with black adhesive tape to exclude light. Reagents HPLC eluent. HPLC grade acetonitrile (Fisons) and dis- tilled water were used in the proportions indicated in the text. Colour-forming reagents, 1 % m1V 2,3,5-triphenyl- tetrazolium chloride in HPLC eluent solution 0.02 M in sodium hydroxide (see text). The solutions were prepared daily and were stored in amber-glass bottles until required. Sugar samples, 10% mlV. Stock solutions were prepared by dissolving the sugars (AnalaR grade; BDH Chemicals, Poole) in acetonitrile-water (75 + 25). Test samples were prepared by mixing and diluting the stock solutions with acetonitrile - water as required.All solutions were de-gassed in an ultrasonic bath before use. Procedure After switching on, the apparatus was allowed to equilibrate for 10 min before use. Samples of reducing sugars (5 p.1) were injected on to the HYLC column using a direct syringe injection technique. Aliquots of 12p1 of eluate from the column were injected automatically into the FIA detection system at fixed intervals. The sampling interval and injection period were adjusted to give the conditions of best sensitivity. Valve timing/s Peak height/mm 0.30 141 0.45 173 0.60 160 0.75 150 0.90 135 1 .oo 125 1 d c 0 Q, .- w .- J 1 I I I 30 20 10 0 Ti me/m in Fig. 3. Chromatogram obtained from the tes solution containing four sugars.1, Fruccose; 2, glucose; 3, maltose; and 4, lactose. Samplk size, 5 11; concentration, 0.05 M for each sugar Conditions for the chromatography were investigated by varying the relative proportions of acetonitrile and water in the eluent solution and comparing the separations obtained using a sample containing 1% mlV of fructose and glucose, and 2% mlV of maltose and lactose. Calibration graphs were obtained for each of these sugars in the range 0.1-1% mlV for monosaccharides and 0.1-2% mlV for disaccharides. Results and Discussion The valve sampling interval and injection period were kept equal and were adjusted over the range 0.3-1 s. The results obtained are given in Table 1. Optimum sensitivity was obtained for a setting of 0.45 s sampling and 0.45 s injection, but as the electronic timers used could only be set accurately to the nearest 0.2s, a slightly sub-optimum setting of 0.6s sampling and 0.6s injection was adopted for the remaining experiments. Use of the slower setting also served to reduce the rate of wear of the sampling valve.Experiments were performed using acetonitrile - water proportions of 70 + 30, 75 + 25 and 80 + 20 for the HPLC eluent. The optimum proportions for the 250-mm column were found to be 75 + 25 acetonitrile - water, giving a good separation, as shown in Fig. 3. Using 80 + 20 acetonitrile - water the separation was slightly improved, but the analysis time was increased by approximately 80%, and with 70 + 30 acetonitrile - water the fructose and glucose peaks were not adequately resolved.Linear calibration graphs were obtained for the four test sugars in the range 0.02-0.2% mlV for glucose and fructose and O.1-lo/0 mlV for maltose and lactose. These results are shown in Fig. 4.ANALYST. JANUARY 1984, VOL. 109 93 560 480 % 400 E 320 9 240 160 . m Y LL 80 0 10 20 30 40 Concentration of sugar/mg per 5 pl Fig. 4. glucose; 3, maltose; and 4, lactose Calibration graphs obtained for four sugars. 1, Fructose, 2, The results obtained show that the system is capable of good performance, and the fact that it is a selective technique may make sample pre-treatment unnecessary in some instances. However, there is obviously considerable scope for optimisa- tion of the technique with the aim of improving sensitivity and reducing peak broadening by careful and systematic adjust- ment of the various chemical and physical parameters of the system.One of the most important of these is the composition of the reagent solution. The colour-forming reaction favours strongly alkaline conditions, and in early experiments the reagent was made up in 0.2 M sodium hydroxide solution. However, the resultant peaks- were so small as to be virtually indistinguishable from base-line noise, owing to the very slow diffusion between the sample and reagent zones in different solvents. In order to promote rapid diffusion , and hence reaction, between sample and reagent it was judged necessary for the same solvent to be used for both, i.e. , for the reagent to be prepared in the HPLC eluent solution in place of the purely aqueous solution used originally.However, the solubility of sodium hydroxide in the acetonitrile - water mixture used as the eluent is extremely limited and attempts to obtain large concentrations of hydroxide in acetonitrile -water mixtures merely resulted in separation into the aqueous and organic phases. For this reason, subsequent experiments were carried out using a sodium hydroxide concentration of only 0.02 M. Some further tentative experiments were performed in which the concentrations of both acetonitrile and sodium hydroxide in the reagent soluion were varied according to a simplex optimisation procedure. This is an empirical mathe- matical technique for performing multivariate optimisation that is significantly faster and more reliable than the tradi- tional univariate methods of optimisation, especially where a large number of parameters are to be optimised.10Jl The results obtained, although not conclusive, show a tendency towards improved sensitivity with higher hydroxide concentrations in the region of 0 .0 5 ~ with no apparent insolubility problems. Work currently in progress is aimed at extending the use of the modified simplex optimisation technique to include many other parameters in the system, including flow-rates, reaction coil length and valve timing, in addition to the parameters mentioned above. This is currently being performed using a computer-controlled FIA system, with the intention of maximising the sensitivity of the system and minimising peak broadening. Such an approach, albeit using a manual FIA system, has already been applied to another determination in our laboratory with excellent results.12 Refractive index detection is for many applications the most sensitive means of detecting carbohydrate compounds, but can be extremely expensive. This system offers a sensitivity comparable to that obtainable with refractive index detectors of the type used routinely in HPLC detection, but at a much lower cost (ca. E250). Where only a single detection channel is required, the sampling valve may be replaced by a T-piece, simplifying the system and further reducing the cost. Conclusion The use of FIA as a post-column detection system for the HPLC of sugars offers a viable low-cost alternative to refractive index or ultraviolet detection. It is hoped to enhance the performance of the system further by the use of simplex optimisation techniques. The authors are grateful to the Department of Industry for the provision of studentships for N. G .C. and T. J. S. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Betteridge, D.,Anal. Chem., 1978,50,832A. RSiiEka, J . , and Hansen, E. H., Anal. Chim. Acta, 1975, 78, 17. Stewart, K. K., Beecher, G. R., and Hare, P. E . , Fed. Proc. Fed. Am. Soc. Exp. Biol., 1974,33,1439. Ranger, C. B.,Anal. Chem., 1981,53,20A. Courtney, N. G., MSc Thesis, University of Wales, 1980, p. 31. Mattson, A. N., and Tenson, C. O., Anal. Chem., 1950, 22, 183. Feigl, F., “Spot Tests in Organic Analysis,” Elsevier, Amster- dam, 1956, p. 389. Betteridge, D., Dagless, E. L., Fields, B., and Graves, N. F., Analyst, 1978,103,897. Sly, T . J . , Betteridge, D., Wibberley, D., and Porter, D. G. , J . Autom. Chem., 1982,4,186. Nelder, J. A. , and Mead, R., Comput. J . , 1965,7,308. Morgan, S. L., and Deming, S . N., Anal. Chem., 1974, 46, 1170. Betteridge, D., Sly, T. J., Wade, A. P., and Tillman, J . E. W., Anal. Chem., 1983,55,1292. Paper A31154 Received May 26th, 1983 Accepted August 22nd 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900091

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, JANUARY 1984, VOL. 109 SHORT PAPERS Determination of Deuterium Using a Membrane Polarographic Detector Andrew Mills Department of Chemistry, University College of Swansea, Singleton Park, Swansea, SA2 8PP, UK 95 Keywords: Deuterium determination; membrane electrode; polarographic detector; Clark-type electrode Deuterium is in current use in a variety of areas,’ e.g. as a moderator for nuclear reactions and as a tag or tracer in the establishment of reaction kinetics and rate-determining steps in chemical reactions, Determination of deuterium, however, appears2 limited mainly to mass spectrometry (MS) or gas chromatography (GC) (using a thermal conductivity detector) and both techniques necessitate removal of samples from the system, which can present problems, especially if several samples have to be removed.As an alternative to the above techniques a membrane polarographic detector for deuterium (D2-MPD) would offer a sensitive, selective and cheap method for the quantitative and continuous determination of deuterium in a closed system as well as in an open one. Although MPDs have been in use since 1959,3 only in the last decade has progress been made in the development of MPDs for gases other than oxygen, such as carbon dioxide,4 dinitrogen oxide,5 chlorine6 and iodine.7 In a recent paper,7 we described how a two-electrode 02-MPD (better known as a “Clark electrode”) could be modified to determine hydrogen and the results agreed with those obtained by other research groups.8.9 In this paper we report on what we believe is the first application of an MPD for the determination of deu- terium.Results and Discussion Polarograms The electrochemical characteristics for the oxidation reaction are similar to those for the reaction when carried out on a platinum electrode; for example, in the literature13 E” (2D+/D2) and E” (2H+/H2) are usually quoted as -0.044 and -0.000 V vs. S.H.E. and, in addition, Bockris and Kochl4 have reported exchange current densities of 2.6 2 0.8 X 10-4 and 5.1 k 1.7 x 10-4 A cm-2, respectively. It is hardly surprising, therefore, that the current versus voltage graphs (Fig. 1) recorded for the D2-MPD after saturation of the test solution with deuterium (curve B) and hydrogen (curve C) are similar in many respects. Probably the most important of these similarities are the “plateau regions” (i.e., regions where the current is diffusion controlled), which occur over the same voltage region (viz., +0.2 to +0.6 V vs.Ag - AgCl). ‘The polarograms (Fig. 1) indicate that a polaris- ing voltage of ca. +0.4 V vs. Ag - AgCl would be suitable for D2 - 2e+2D+ (1) H2 - 2e+ 2H+ (2) Experimental Apparatus and Reagents The Perspex MPD used (supplied by Rank Bros., Cambridge) was of a similar design to that described by Delieu and Walker,lo with a platinum working electrode (1.5 mm diam- eter), a silver - silver chloride counter - reference electrode and a 12.5-pm PTFE membrane. Constant polarising voltages were applied using a Metrohm E611 potentiostat. Gas chromatography was performed using a Perkin-Elmer F33 gas chromatograph with a thermistor-type thermal conductivity detector.The 2 m x 3.1 mm i.d. stainless-steel column was packed with activated (250”C, 12h) molecular sieve 5A (Jones Chromatography) according to the procedure outlined by Purnell and Laub.11 Argon was used as the carrier gas. The deuterium gas was supplied by BDH Chemicals and all other gases were supplied by BOC. Procedure The electrode preparation and setting-up procedure for the D2-MPD was identical with that previously reported for the H2-MPD .7 Polarograms were produced using the equipment and procedure described in an earlier paper.7 Calibration of the D2-MPD was achieved by saturating the test solution in contact with the MPD with a variety of deuterium - nitrogen gas mixtures, produced using a tangential gas mixer.’ In all instances the test solution was distilled water, which, together with the detector, required thermostating at 25 k 0.5 “C as both the permeability coefficient of deuterium in the mem- brane and its solubility in water are temperature dependent.’* L 1 I I 1 -0.2 0 0.2 0.4 0.6 Polarising voltageN versus Ag - AgCl Fig.1. Polarograms recorded for the D,-MPD using test solutions saturated with (A) nitrogen, (B) deuterium and (C) hydrogen. Voltage sweep rate, 70 mV min-I96 future work with the D2-MPD because, at this voltage, any currents produced (id) will depend on the rate of diffusion of deuterium to the electrode, and this can be related directly to the concentration of deuterium (cD2) dissolved in the test solution. Indeed, it can be shown12J5 that (3) id nFPm A - b CD, - - -.w iere id = diffusion-controlled current; n F = Faraday’s constant; Pm = permeability coefficient of the membrane; b = membrane thickness; A = electrode area. = number of electrons transferred in the electrochemical reaction; Ideally, we would expect the current versus voltage graphs to be independent of the direction of the voltage scan. Although this was found to be so for the plateau regions (i.e., +0.2 to 0.6 V vs. Ag - AgCl), the peak at the cathodic end of the positive scan (Fig. 1) was not shown for a scan in the reverse direction and is probably due to the oxidation of hydrogen evolved and then adsorbed on to the electrode at lower potentials. We found that the D2-MPD exhibited a similar insensitivity towards oxygen as reported for an H2-MPD7; however, the polarograms (Fig.1) indicate that it responds readily to both deuterium and hydrogen. One possible solution to this limitation in the number of possible applications of a D2-MPD is discussed later. Calibration of the D2-MPD Using a similar procedure to that described for an H2- MPD,7JS deuterium - nitrogen gas mixtures of known compo- sition (determined by GC) were used to saturate the test solution above the D2-MPD. The diffusion-controlled cur- rents (id) produced by the D2-MPD, after correction for residual current,7 were found to obey equation (3) over two orders of magnitude (i.e., from 100% down to <1% deu- terium saturation). Using the deuterium and hydrogen solu- bility data of Muccitelli and Wen16 (see Table l), the sensitivity of the D2-MPD was determined to be ca.6.6 Table 1. Useful data for the D,-MPD Gas under detection Property Deuterium Hydrogen SensitivityhA pmol-1 1 . . 6.6 7.7 ResidualcurrenthA . . . . 16 16 Linearity of response . . . . Over 2 orders Over 3 orders of magnitude of magnitude Solubility16 in water*/ Calculated*.$ solubility Calculated*.t permeability mol 1-1 . . . . . . . . 8.35 x 7.82 x 10-4 in PTFE membrane/moll-l 1.57 x 10-3 1.4 x 10-3 coefficient (Pm) in PTFE membrane/m*s-1 , . . , 7.88 x 10-11 7.59 x 10-11 * At 1 atm and 25 “C. T Using the procedure outlined by Van Krevelen and Hoftyzer.” ANALYST, JANUARY 1984, VOL. 109 nA pmol-1 1. Although the residual current was found to be fairly high (ca. 16 nA), the compensation current facility on the potentiostat enabled us to look at deuterium levels down to ca.2 pmol 1-1. Some useful data concerning the D2-MPD are collected in Table 1, including a comparison of its response towards deuterium and hydrogen. The problem of using an MPD to analyse a mixture of two (or more) gases that have similar electrochemical properties (e.g., hydrogen and deuterium or chlorine, bromine and oxygen) is not new.12 However, one possible solution might be to separate the gases in the mixture before detection, by GC for example, and indeed initial work carried out by us has shown that gas mixtures of hydrogen and oxygen or deuterium and oxygen can be analysed by GC using an MPD. Further work is now in progress to ascertain if such a simple detector offers a realistic alternative to a thermal conductivity detector in gas-mixture analysis by GC, as has been found for metallised-membrane electrodes.’* We thank the SERC for financial support of this work.We are indebted to Mr. J. Langley for his assistance with some of the work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Katz, J . J., in Mark, F. H., Othmer, D. F., Overberger, C. G., and Seaborg, G. T., Editors, “Encyclopedia of Chemical Technology,” Third Edition, Volume 7, Wiley-Interscience, New York, 1979, p. 539. Mackay, K. M., and Dove, M. F. A., in Bailar, J. C., Emeleus, H. J . , Nyholm, R., and Trotman-Dickinson, A. F., Editors, “Comprehensive Inorganic Chemistry,” Volume 1, Pergamon Press, Oxford, 1973, p. 77. Clark, L. C., U.S. Pat., 2 913 386, 1959.Albery, W. J . , and Barron, P., J . Electroanal. Chem., 1982, 138, 79. Albery, W. J . , Brooks, W. N., Gibson, S. P., Heslop, M. W., and Hahn, C. E. W., Electrochim. Acta, 1979,24, 107. Ben-Yaakov, S . , J . Electroanal. Chem., 1979,98, 15. Mills, A., Harriman, A., and Porter, G., Anal. Chem., 1981, 53, 1254. Gruniger, H. R., Sulzberger, B . , and Calzaferri, G., Helv. Chim. Acta, 1979, 62, 2434. Srinivasan, V. S . , and Tarcy, G. P., Anal. Chem., 1981, 53, 928. Delieu, T., and Walker, D. A., New Phytol., 1972, 71, 201. Purnell, J. H., and Laub, R. J., J . High. Resolut. Chromatogr. Chromatogr. Commun., 1980, 3, 195. Hitchman, M. L., “Measurement of Dissolved Oxygen,” Wiley-Interscience, New York, 1978, Chapter 5. Hunsberger, J. F., in Weast, R. C., Editor, “Handbook of Chemistry and Physics,” Sixty-first Edition, CRC Press, Boca Raton, FL, 1980, p. D-155. Bockris, J. O’M., and Kock, D. F. A., J . Phys. Chem., 1961, 65, 1941. Mills, A., in Harriman, A., and West, M., Editors, “Photo- generation of Hydrogen,” Academic Press, London, 1982, p. 1. Muccitelli, J., and Wen, W.-Y., J . Solution Chem., 1978, 7, 257. Van Krevelen, D. W., and Hoftyzer, P. J . , “Properties of Polymers,” Second Edition, Elsevier, New York, 1976, p. 403. Bergman, I., Coleman, J . E., and Evans, D., Chromato- graphia, 1975, 8, 581. Paper A31160 Received June 2nd, 1983 Accepted August 23rd, 1983

ISSN:0003-2654    

DOI:10.1039/AN9840900095

出版商:RSC    

年代:1984

数据来源: RSC