Analyst, May, 1978, Vol. 103, $9. 497-508 497 Polarographic Method for the Identification of 1,4- Benzodiazepines W. Franklin Smyth, M. R. Smyth," J. A. Grovest and S. B. Tan Department of Chemistvy, Chelsea College, University of London, Manresa Road, London, S W3 6LX The polarographic behaviour of 12 therapeutically important 1,4-benzo- diazepines in Britton - Robinson universal buffers, pH 4.0 and pH 12.0, has been investigated. Differences in the polarographic peak potentials of these compounds in these media are explained. The rates of hydrolysis of certain benzodiazepines in acidic solution were investigated. Bromazepam and flunitrazepam, both of which possess a strongly electron-withdrawing substituent on the 5-o-phenyl group, were found to undergo rapid acid hydrolysis. On the basis of these findings, and taking into account the extraction profile of some of the compounds over a pH range, a scheme is devised for the identification of any one or more of 12 1,4-benzodiazepines.It is suggested that this procedure would be applicable to the analysis of unknown formulations or body fluids in forensic cases where the parent compound exists in relatively high concentrations compared with its meta- bolites. Keywords : 1,4-Benzodiazepine identification ; polarography The lJ4-benzodiazepines are one of the most frequently prescribed group of drugs in the UK and, with the ever increasing number of compounds in this series that are used for thera- peutic purposes, there is a need for the development of rapid methods of identification, particularly for use in forensic situations, e g ., identification of a particular 1,4-benzo- diazepine in unknown formulations or body fluids. The most commonly employed method of identification is thin-layer chromatography of the intact compounds or of their acid hydrolysis products, i.e., benz0phenones.l In the examination of body fluids, a separation step such as adsorption-column chromatography2 or solvent extraction3p4 is usually employed prior to thin-layer chromatography. Identification is usually effected by comparison of R, values, the colours produced by chromogenic spraying and spot patterns produced by the drug and its metabolite^.^^^^^ Gas - liquid chromatography using the nickel-63 electron- capture detector appears to be the method of choice for the determination of the 1,4- benzodia~epines.~-~ Smyth and co-workers10-l6 have investigated the acid - base properties and polarographic behaviour of many of the members of this group of drugs.Groves17 has made a detailed study of the mechanism of hydrolysis of 5-fluorophenyl-l,4-benzodiazepinesl7 and com- pared the rates of hydrolysis of other non-fluorinated 1,4-benzodia~epines.~s This paper shows how a knowledge of these physico-chemical properties can be used to devise a scheme for the identification of 12 lJ4-benzodiazepines : medazepam (I) , chlordiazepoxide (11) , potassium chlorazepate (111), bromazepam (IV), diazepam (V, X = C1, Y = CH,, 2 = H, W = H), oxazepam (V, X = Cl, Y = H, 2 = OH, W = H), prazepam (V, X = C1, Y = cyclopropyl, 2 = H, W = H), lorazepam (V, X = C1, Y = H, 2 = OH, W = Cl), nitrazepam (V, X = NO,, Y = H, 2 = H, W = H), clonazepam (V, X = NO,, Y = H, 2 = H, W = Cl), flunitrazepam (V, X = NO,, Y = CH,, 2 = H, W = F) and flurazepam (V, X = C1, Y = (CH,), NEt,, 2 = H, W = F).This scheme involves the combination of simple solvent-extraction procedures and polarographic examination of these extracts, in some instances following acid hydrolysis. It is not intended to be an absolute method of identifi- cation but it is suggested that it can be used as a rapid method for obtaining information complementary to that obtained by other identification techniques. The study has been confined to parent compounds so that it would be directly applicable to the identification of unknown fonnulations and in overdose cases where a relatively high concentration of the free parent drug still remained in the particular body fluid.* Present address : Chemistry Department, State University of Colorado, Fort Collins, Colo. 80523, USA. t Present address : Beecham Products, Block F2-F, Brentford, Middlesex.498 SMYTH et al. : POLAROGRAPHIC METHOD FOR Analyst, Vol. 103 1 I l l Br C=N I V v Experiment a1 Apparatus Instrumentation A PAR Model 174A Polarographic Analyser produced by Princeton Applied Research Corporation, N. J., USA, was used in conjunction with a Servoscribe 15 recorder throughout this study. A three-electrode operation was employed using a platinum counter electrode. The dropping-mercury electrode (D.M.E.) used had an outflow velocity of 2.571 mg s-1 and a drop time of 3.46 s at the potential of the calomel electrode and at a mercury pressure of 55 cm in 1 M potassium chloride solution.Polarographic cells For investigations on the polarographic behaviour in different buffer solutions, the cell shown in Fig. 1 was used. This was based on a 25-ml Quickfit flask and was suitable for volumes of solution from 2 to 20 ml, the optimum volume being about 5 ml. The platinum- wire counter electrode was sealed into the glass of the vessel, as were the inlet tubes for passing gas into and over the sample solution. A Radiometer-type saturated calomel electrode (S.C.E.) was used as the reference electrode. As it was intended to carry out and monitor the hydrolysis in the polarographic cell, a Socket( B 10) S.C.E. (25 ml) Solution Platinum counter electrode Fig. 1, Polarographic cell.May, 1978 THE IDENTIFICATION OF 1,4-BENZODIAZEPINES 499 vessel was required that could be maintained at constant temperature.The cell shown in Fig. 2 was designed and constructed specifically for this purpose. It is based on a 25-mI three-necked, pear-shaped flask that is enclosed in an outer glass jacket. Water at the required temperature could be pumped between the inner and outer vessels. A tap provided at the bottom of the cell allowed for emptying and cleaning purposes. Narrow-bore glass tubes were sealed into the apparatus to pass nitrogen through or over the solution. The cell functioned satisfactorily with solution volumes from 2 to 15 ml, but in practice 5-10 rnl was ideal. The reference electrode was again a Radiometer-type saturated calomel electrode.A platinum wire sealed into a glass tube served as the counter electrode. Water was circulated through the jacketed cell from a thermostatically controlled water-bath. A 10-ml volume of aqueous solution in the cell could be brought from room temperature to any temperature in the range 2545 "C in 10 min and maintained at that temperature within k0.05 "C. Throughout this study, the hydrolyses were performed at 25 "C. The passage of nitrogen saturated with water vapour either through or over the solution did not alter the temperature. Platinum counter. electrode Nitrogen Water jacket Water in __I.F_ out Fig. 2. Polarographic cell with water jacket used for hydrolysis studies. Reagents Sug5porting electrolyte A modified Britton - Robinson universal buffer was used to provide buffer solutions of pH 4.0 and 12.0.The stock solution of this buffer (pH 1.8) contained 0.04~ acetic acid, orthophosphoric acid and boric acid. The stock solution was adjusted to the required pH by adding 0.2 M sodium hydroxide solution until the desired pH had been reached, as monitored by using a pH meter. All reagents and solvents used were of analytical-reagent grade. Stock solutions of the benxodiazepines Stock solutions of concentration 1 X 1 0 - 3 ~ were freshly prepared in AnalaR methanol. The addition of 0.1 ml of solution to 9.9 ml of the buffer solution would produce a working concentration of 1 x M. Experimental Techniques Polarographic behaviour in p H 4.0 and 12.0 Britton - Robinson bufers For each run, 0.1 ml of the stock solution of the benzodiazepine was pipetted into a 10-ml calibrated flask and then diluted to the mark with the appropriate buffer solution.After thorough mixing, the solution was placed in the polarographic cell shown in Fig. 1, with the S.C.E. and the D.M.E. positioned in the cell, and nitrogen was passed through the solution for 10 min in order to remove oxygen from the solution. After de-gassing, nitrogen was500 SMYTH et al. : POLAROGRAPHIC METHOD FOR Analyst, Vol. 103 passed over the solution and the polarogram was recorded using the following polarographic conditions : mode, differential-pulse polarography (DPP) ; initial potential, 0.0 V veYsZts S.C.E. ; scan rate, 5 mV s-l; chart speed, 3 c;m min-l; modulation amplitude, 100 mV; low pass filter, 0.3 s; potential scan range, 3.0 V; scan direction, negative; current range, 2 or 5 PA; drop time, 1 s.Hydrolysis The addition of 0.1 ml of the stock solution to 9.9 ml of acid in the thermostatically controlled cell would produce a working concentration of 1 x l O A 5 ~ . The concentration of methanol in the solution was therefore only 1%. The hydrolysis of each compound included in this study was carried out in 0.1 M hydrochloric acid, which was prepared from a Volucon ampoule. The latter solution exhibited an acceptable polarographic background at the current ranges employed. A 9.9-ml volume of 0.1 M hydrochloric acid was pipetted into the thermostatically controlled cell and the oxygen removed by passing nitrogen through the solution for 10 min.During this time, the contents of the cell attained the required temperature. After de-gassing, nitrogen was passed over the solution and the S.C.E., platinum counter electrode and D.M.E. positioned in the cell. The polarogram of the acid alone was recorded using the polaro- graphic conditions described above for the buffer solution. Subsequently, 0.1 ml of the methanolic stock solution of the benzodiazepine was introduced into the cell by means of an automatic zero pipette. This addition was achieved by partially removing the platinum counter electrode and inserting the pipette through the neck into the cell. Timing was begun when approximately half of the 0.1-ml aliquot had been added. The solutions were thoroughly mixed by passing a stream of nitrogen through the cell for about 1 min.The polarogram of the solution was recorded after a convenient time. Preliminary experiments indicated that the method of recording the entire polarogram would be suitable for the present investigation. The use of a concentration of electroactive species of 1 x 1 0 - 5 ~ precluded the use of d.c. polarography as the polarograms at this concentration exhibit a certain degree of slope that makes measurement of the limiting current difficult. Also, the half-wave reduction potentials of some of the reactants and products are close (with a difference of less than 100mV) and so would not be resolved as individual waves when using d.c. polarography. The cathode-ray polarograph could have been employed but again, accurate measurement of the limiting current of two waves close to each other would be difficult.As a result of these considerations, differential-pulse polarography (DPP) was chosen as the means by which to record the polarograms. This technique has the required sensitivity and would allow a compromise between speed of recording and resolution of the polarographic waves. The shape of the peaks, without serrations, also allows for a more convenient way of measuring concentration, Results and Discussion Polarographic Behaviour The peak reduction potentials, E,, for each of the compounds were measured from consideration of the potential scan rate and the chart speed. Table I shows the E, values (volts veysus S.C.E.) obtained in the buffer solutions of pH 4.0 and 12.0.In pH 4.0 buffer, chlordiazepoxide shows three peaks, at -0.37, -0.73 and -1.17 V, corresponding to the reduction of the =N+Oi, C=N and N=C moieties, respectively.15 In pH 2.0 buffer it gives one peak at -1.24V. In pH 4.0 buffer, bromazepam shows two peaks, at -0.40 and -1.12 V (see Fig. 3). 'The electronegative influence of the 5-pyridyl moiety shifts the reduction of the azomethine C=N group from -0.7-O.80VJ as observed in other benzodiazepines, to -0.40 V. The wave at -1.12 V for bromazepam corresponds to a reduction process involving the 5-pyridyl group.ll In pH 12.0 buffer a well defined peak was observed at -0.99 V. At low concentrations, the waves at -1.17 and -1.12 V for chlordiazepoxide and bromazepam, respectively, (in pH 4.0 buffer) can be masked by the reduction of the supporting electrolyte or other impurities co-extracted from a body fluid. When in a mixture bromazepam and dhlordiazepoxide are best differentiated polaro-May, 1978 THE IDENTIFICATION OF 1,4-BENZODIAZEPINES 501 TABLE I E, VALUES OF SOME BENZODIAZEPINES IN pH 4.0 AND pH 12.0 BRITTON - ROBINSON BUFFERS E,/V ueYsus S.C.E.r I A 1 ,QBenzodiazepine pH 4.0 buffer pH 12.0 buffer Chlordiazepoxide . . .. -0.37, -0.73, -1.17 - 1.24 Nitrazepam .. . . -0.16, -0.78 -0.61, -1.23 Clonazepam . . . . -0.15, -0.73 -0.60, -1.22 Flunitrazepam . . . . -0.16, -0.73 -0.61, -1.20 Oxazepam . . . . . . -0.76 - 1.50 Lorazepam . . .. -0.74 - 1.45 Diazepam . . .. . . -0.74 -1.15 Bromazepam . . . . -0.40, -1.12 -0.99 Prazepam . . * . .. -0.73 -1.22 Flurazepam .... - 0.72 -1.10 Potassium chlorazepate . . -0.73 -1.15 Medazepam .. .. -0.82 - 1.23 graphically by the peaks observed in pH 12.0 buffer (Pig. 4). The peak a t -0.73 V at pH 4.0 for chlordiazepoxide is not used as it is also observed in other benzodiazepines, e.g., see Fig. 5, which shows the differential-pulse polarograms of a mixture of bromazepam, chlordiazepoxide and flunitrazepam in pH 4.0 and 12.0 buffers. Flunitrazepam is included in Figs. 3-5 as an example of a 1,4-benzodiazepine giving a typical C=N reduction in pH 4.0 and 12.0 buffers. Nitrazepam, clonazepam and flunitrazepam can be differentiated from all other 1,4- benzodiazepines included in this study by the peak observed at -0.16 V in pH 4.0 buffer, corresponding to the reduction of their NO, group.At pH 12.0 this reduction is observed at -0.60 to -0.61 V and this peak could also be used for identification. Figs. 6 and 7 show the differential-pulse polarograms of some other 1,4-benzodiazepines which have only one reducible group, i.e. , the 4,5-azomethine group. Oxazepam, lorazepam, diazepam, prazepam and medazepam give well defined peaks in pH 4.0 buffer, but the peak potentials do not differ significantly except for medazepam, which is reduced at a potential of the order of 75 The peak heights for the first two are approximately twice the size of the other peaks as they are reduced 15 mV more negative than the remainder. -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 - 1 . 4 Potential/V Fig. 3. Diff erential-pulse polarograms of : A, flunitrazepam ; B. chlordiazepoxide ; and C , brom- azepam, as 1 x M solutions in pH 4.0 buffer solution.3.0 2.0 Q f? 3 1.0 -0.4 -0.6 -0.8 - 1.0 - 1.2 - 1.4 - 1.6 - 1.8 Potential/V Fig. 4. Differential-pulse polarograms of: A, flunitrazepam ; B, bromazepam ; and C, chlordiaz- epoxide, as 1 X M solutions in pH 12.0 buffer solution.502 SMYTH et al. : POLARCIGRAPHIC METHOD FOR Analyst, Vol. 103 3.0 --. % +.’ c 2.0 L a 1 .o 0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 PotentialN Fig. 5. Differential-pulse polarograms of a mixture of bromazepam, chlordiazepoxi de and flunitrazepam as 1 x M solutions in: A, pH 4.0 buffer solution; and B. pH 12.0 buffer solution. in four-electron processes that involve reductive dehydroxylation of the C-3 position.14 Oxazepam and lorazepam show different polarographic behaviour in pH 12.0 buffer in that they give rise to relatively small broad peaks at substantially more negative potentials than diazepam, prazepam and medazepam.This effect is a polarographic manifestation of the acid - base equilibrium in which species VI is reduced at a more negative potential than the neutral molecule, and is also due to repulsion of the anion (VI) from the negatively NH-CO H \ / L 7 CI C=N OH C1 C=N 0- VI charged mercury surface. This behaviour can be used to differentiate oxazepam and lorazepam from diazepam, prazepam and medazepam and also to a certain extent from each other. Diazepam can be differentiated fr,om prazepam and medazepam in this buffer as their E, values differ by 70 mV. Flurazeparn and potassium chlorazepate give reduction potentials of -0.72 and -0.73 V in pH 4.0 buffer and -1.10 and -1.15 V in pH 12.0 buffer, but these can be separated prior to polarographic analysis by using their unique acid - base properties, as discussed below under Analytical Applications.Hydrolysis Studies The aim of these investigations was to differentiate between members of groups of 1,4- benzodiazepines which it was not possible to resolve by polarography and/or solvent extrac- tion. In 0.1 M hydrochloric acid at 25 “C, flunitrazepam was found to hydrolyse much faster than either clonazepam or nitrazepam, and has a half-life of about 50 min. Fig. S(a) shows the differential-pulse polarograms obtained for flunitrazepam at 1 x M in 0.1 M hydro- chloric acid at 25 “C. The times indicated are those that elapsed between the start of the The three nitro-containing substances are an example of such a group.May, 1978 THE IDENTIFICATION O F 1,4-BENZODIAZEPINES 503 0 - 0.50 PotentiaVV - 1.0 Fig.6. Differential-pulse polarograms of: A, lorazepam; B, oxazepam; C, medazepam; D, diazepam; and E, prazepam, as 1 x M solutions in pH 4.0 buffer solution. reaction and the start of the recording. Fig. 8(b) and (c) shows the polarograms obtained for the acid hydrolyses of clonazepam and nitrazepam, respectively. The decrease in the polarographic peak currents caused by the disappearance of the reactants in 0.1 M hydro- chloric acid, and the corresponding increase in the reaction products of these three compounds are shown in Fig. 9. These results indicate that flunitrazepam can be differentiated from the other two nitro- containing benzodiazepines on the basis of its rate of hydrolysis in 0.1 M hydrochloric acid.Clonazepam hydrolysed at a slightly faster rate than nitrazepam (Fig. 9) but this could not be used to differentiate between these two compounds. Provided that no other benzo- diazepines were present in the unknown sample, they could be differentiated by their peak potentials in pH 4.0 buffer (Table I). Groves18 has studied the mechanism of hydrolysis of 1,4-benzodiazepines in dilute acidic solutions (pH 0-2) and found that such reactions occur relatively rapidly when there is an electron-attracting group in the 5-phenyl ring, e.g., 5-o-fluorophenyl-l,4-benzodiazepines. In addition, these reactions are most rapid in the region of the pKa value corresponding to the azomethine group and fall off at pH < pK, and pH > pK,.This suggests that the mechanism involves acid catalysis of the non-protonated benzodiazepine and benzophenones of type VII are generally formed. In strong acids (pH < 0 ) , benzophenones of type VIII are formed and it is these species that are used in identification procedures involving thin- layer chromatography .504 SMYTH et al. : POLAROGRAPHIC METHOD FOR Analyst, Vol. 103 R R I VI I Vlll The half-lives (tJ of these reactions and the corresponding changes in potential are given in Table I1 for flunitrazepam, clonazepam and nitrazepam. The slower rate observed for clonazepam presumably occurs as a result of i:he interference of the large chlorine atom in the hydrolytic reactions that involve the azoniethine group.This interference could come TABLE I1 HALF-LIVES OF HYDROLYSIS REACTIONS AN;D CORRESPONDING CHANGES IN POTENTIAL E, values in volts veysus S.C.E. OF FLUNITRAZEPAM, CLONAZEPAM AND NITRAZEPAM Medium f A > 1 .O M hydrochloric acid 0.1 M hydrochloric acid f n 7 f A \ E P E P E P E P Compound td/s (product) (reactant) t t / s (product) (reactant) E,(NO,) Flunitrazepam . . 237 -0.547 - 0.675 25 -0.584 -0.723 -0.069 Clonazepam . . .. 1069 -0.520 -0.654 285 -0.562 -0.701 -0.069 Nitrazepam . . .. 672 -0.552 -0.665 652 -0.589 -0.755 -0.069 2.0 a -3 E 5 1.0 + C 1 I I I I I I I I I I rl - 0.6 - 0.8 - 1.0 - 1.2 - 1.4 - 1.6 - 1.8 PotentiaVV Fig. 7. Differential-pulse polarograms of: A, diazepam; B, prazepam; C, medazepam; D, oxazepam; and E, lorezapam as 1 X M solutions in pH 12.0 ;buffer solution.May, 1978 THE IDENTIFICATION OF lJ4-BENZODIAZEPINES 505 - 0.ov 10.5 35 65 Time/min - 0.ov I 1 I 1 20 40 60 Time/min - 0.ov ‘ I I J 35 65 85 Timehin Fig.8. Diff erential-pulse polarograms of : (u), flunitrazepam ; ( b ) , clonazepam ; and (c), nitrazepam, as 1 x 10-5 M solutions in 0.1 M hydrochloric acid a t 25 “C. about by direct shielding of the group by chlorine and/or twisting the phenyl ring a t position 5 out of plane with the azomethine group, causing a reduction in conjugative interaction. These results seem to indicate that the presence of a relatively positively charged nitrogen atom in the azomethine group results in a relatively rapid hydrolysis in dilute acid solutions. Bromazepam, which contains a 5-pyridyl substituent, was found to react about ten times faster than flunitrazepam in 0.1 M hydrochloric acid with a t, of 5 min. Its rate of reaction increases with decreasing pH at pH values less than the pK, value of the azomethine group and Smyth et aZ.have proposed an alternative hydrolytic mechanism in these dilute acid solutions .11 Analytical Applications The results of the polarographic studies on the l,4-benzodiazepines show that it is possible to identify bromazepam and chlordiazepoxide in an unknown sample if a peak is found in pH 4.0 buffer in the potential range -0.37 to -0.40V. These two compounds can be differentiated by polarography in pH 12.0 buffer, determination of the rate of acid hydrolysis in 0.1 M hydrochloric acid in which chlordiazepoxide is stable or formation of the copper - bromazepam complex, as described elsewhere.If a peak occurs at approximately -0.16 V, a nitro-containing 1,4-benzodiazepine is suspected. Rapid acid hydrolysis in 0.1 M hydro- chloric acid will confirm the presence of flunitrazepam (“rapid” being defined as a marked decrease in the peak current corresponding to the azomethine group occurring within 15 min). Nitrazepam and clonazepam can be differentiated by the peak potentials corresponding to their azomethine reductions (-0.78 and -0.73 V, respectively) in pH 4.0 buffer provided that no other benzodiazepines are present. In such a sample, the azomethine reduction peak (plus the hydroxylammonium reduction peak) will appear abnormally low in relation to the nitro reduction peak (in theory, they should be identical in pH 4.0 buffer, but in practice, the differential-pulse peak corresponding to the nitro reduction is higher than the506 SMYTH et al.: POLAROGRAPHIC METHOD 1.3 1.2 1 .I 1 .o 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 r: 3 0 \ c F -.--- 10 20 30 40 50 60 70 80 ! Tiimdmin FOR Analyst, Vol. 103 Fig. 9. Hydrolyijis of clonazepam, nitrazepam and flunitrazepam as 1 x M solution:; in 0.1 M hydro- chloric acid at 25 "C. A, Clonazepam; B, nitrazepam; C, flunitrazepam; D, reaction product of flunitranepam; E, reaction product of clonazepam; and I;, reaction product of nitrazepam. sum of the other two reduction peaks). This effect occurs because differential-pulse polaro- graphy is not particularly responsive to the reduction of aromatic hydroxylamines.If only one peak is obtained and it occurs between --0.70 and -0.76 V, the presence of a benzo- diazepine with only the electroreducible azomethine group present is suggested. Oxazepam and lorazepam can be differentiated from diazepam, medazepam and prazepam by polaro- graphy in pH 12.0 buffer. The same buffer can be used to differentiate oxazepam and lorazepam as their peak potentials differ by 50 mV but, because of the broad nature of these peaks, this difference is of limited value. Diazepam can be differentiated from medazepam and prazepam in pH 12.0 buffer as it is reduced 70 mV earlier. Xledazepam is reduced 90 mV more negatively than prazepam in pH 4.0 buffer. The acid - base equilibria that exist in aqueous solutions of f l u r a ~ e p a m l ~ ~ ~ ~ and potassium chlorazepate12 have been the subject of previous publications.Briefly, flurazepam is extractable at neutral and alkaline pH but inextractable at acid pH (of the order of 3.0) owing to the formation of an extractable ion pair (IX) in the former pH region. The other CH3COO-- HO OH \ / X IX7-7 I Sample I L-----T----i I Adjust pH to 9.0 Aqueous phase contains potassium chlorazepate I 1 Adjust pH to 4.0 and extract with ethyl acetate Evaporate and tak up extract in pH 4.0 buffer mono protonate Solution pH 4.0 Record differential- pulse polarogram pulse polarogram pulse polarogram Peak a t - 0.73-0.74 V corresponds Peak a t - 0.16 V indicates n.itro- containing benzo- 1 diazepine Re-extract drug with ethyl acetate, evaporate, dissolve in 0.1 M hydrochloric acid, hydrolyse Rapid hydrolysis indicates flunitrazepam indicates chlordiazepoxide or bromazepam Ch lordiazepoxide stable in 0.1 M hydrochloric acid Bromazepam rapidly hydro- lysed, confirm by wave a t I - 0.99 V a t 1 pH 12.0 1 Only one peak a t - 0.70 to - 0.76 V indicates an azomethine )C = N group and suggests the presence I of one or more of oxazepam, 1 lorazepam, diazepam and prazepam; identify these by their polarographic I waves in pH 4.0 and 12.0 buffers concentration greater than 10 pg mi- I.This is confirmed by a peak a t - 0.82 V a t Fig. 10. Scheme for identification of a 1,4-b enzodiazepine in an unknown sample.508 SMYTH, SMYTH, GROVES AND TAN 11 1,4-benzodiazepines are extracted at pH 3.0 with a high efficiency.Potassium chlor- azepate is the only 1,4-benzodiazepine of the 12 under study that possesses a carboxylic acid group. Because of the formation of an inextractable species (X) in solutions of neutral and alkaline pH this is the only 1,4-benzodiazepine of this group that is not extractable a t pH 9. The scheme shown in Fig. 10 is therefore suggested for the identification of any one of the above-mentioned 1,4-benzodiazepines in an unknown formulation or in body fluids in forensic cases where the parent compound exists in relatively high concentrations compared with its metabolites. It should be possible to identify the drugs at concentrations of 10-5-10-7 M. Reproducibility of results can be achieved only by strict control of the polarographic con- ditions given in the section on Experimental 'Techniques.It would be expected that polarograms of all 12 benzodiazepines included here would be recorded under appropriate conditions before an identification is attempted. This procedure would allow for the inevitable differences in peak potentials that occur from one laboratory to the next. However, the relative positions of these peaks would not be affected. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 1s. References Lafargue, P., Meunier, J., and Lemontey, Y., J . Chromat., 1971, 63, 423. Hetland, L. B., Knowlton, D. A., and Couri, I)., Clinica Chim. Acta, 1972, 36, 473. Berry, D. J., and Grove, J., J . Chvomat., 1973, 80, 205. Sine, H. E., McKenna, M. J., Law, M. R., and Murray, M. H., J. Chromat. Sci., 1972, 10, 297. Schultz, C., Post, D., Schewe, G., and Schultz, H., 2. Analyt. Chem., 1972, 262, 282. Zingales, I. A., J . Chromat., 1971, 61, 237. Knowles, J. A., and Ruelius, H. W., Arzneimiltel-Forsch., 1972, 22, 687. Marcucci, F., Mussini, E., Airoldi, L., Guaitani, A,, and Garattini, S., J . Pharm. Pharmac., 1972, De Silva, J. A. F., Puglisi, C. V., and Munno, .N., J . Pharm. Sci., 1974, 63, 520. Barrett, J., Smyth, W. F., and Davidson, I. E., J . Pharna. Pharmac., 1973, 25, 387. Smyth, M. R., Beng, T. S., and Smyth, W. F., Analytica Chim. Acta, 1977, 92, 129. Smyth, W. F., and Leo, B., Analytica Chim. Acta, 1975, 76, 289. Clifford, J. M., Smyth, M. R., and Smyth, W. F., 2. Analyt. Chem., 1974, 272, 198. Goldsmith, J. A., Jenkins, H. A., Grant, J., and Smyth, W. F., Analytica Chim. Acta, 1973, 66, 427. Barrett, J., Smyth, W. F., and Hart, J. P., J . Pharm. Pharmac., 1974, 26, 9, Clifford, J. M., and Smyth, W. F., 2. Analyt. Chem., 1973, 264, 149. Groves, J. A., PhD Thesis, University of London, 1976. Groves, J. A., and Smyth, W. F., to be published. 24, 63. Received September 23rd, 1977 Accepted October 19th, 1977