Analyst, September, 1974, Vol. 99, $9. 555-564 555 Oxidation Procedures in the Assay of Some Drugs Containing a Diphenylmethylene Ether or Diphenylmethyleneamino Group BY B. CADDY, F. FISH AND J. TRANTER" (Division of Pharntacognosy and Forensic Science, School of Pharmaceutical Sciences, University of Strathclyde, Glasgow, G1 1XW) Methods €or the oxidation of diphenylmethylene ethers and diphenyl- methylenearnines have been compared ; oxidation with aqueous acidic di- chromate is advocated for the former and with alkaline permanganate for the latter in the assay of drugs containing these groups. The possible mech- anisms involved are proposed. THE use of oxidation procedures for enhancing the ultraviolet absorbance of compounds, thereby lowering their limits of detection, has been described by several authors.1-3 In order to obtain a rational approach to this method of analysis, we have studied several oxidation procedures applied to various groups of compounds. In addition to using oxidation methods in quantitative analysis we were also interested in assessing their routine use in toxicological screening for drugs (mainly basic drugs). This report concerns the oxidation of drugs con- taining the diphenylmethylene ether (Table I) or diphenylmethyleneamino group (Table 11) with aqueous acidic dichroinate and alkaline permanganate. EXPERIMENTAL REAGENTS- Hexane. Potassitdm dichtromate. Potassium permanganate. Sodium hydroxide, pellets. Sul9huric acid, concentrated. Diethyl ether. All of the above materials were of analytical-reagent grade.APPARATUS- A Unicam SP800A, ultraviolet - visible spectrophotometer and a Varian Aerograph 1400 gas chromatograph with a flame-ionisation detector were used. The conditions used were as follows: 1, a 6 foot x g inch i.d. glass column packed with 10 per cent. Apiezon L on Chromo- sorb G (60 to 80 mesh), operated at 250 "C with a nitrogen flow-rate of 50 ml min-1, injection block and detector block being maintained at 270 "C; 2, a G foot x Q inch i.d. glass column packed with 1 per cent. OV-25 on Chromosorb G (60 to 80 mesh), operated at 160 "C with a nitrogen flow-rate of 50 ml min-l, injection block and detector block being maintained at 200 "C. OXIDATION PROCEDURES- glass-coated magnetic follower (Note 2). Place a suitable amount of the drug (Note 1) in a 250-ml conical flask containing a 1-inch Add 20 ml of oxidant (Note 3) and 5 ml of hexane NOTES- The drug was usually introduced as an aqueous solution of one of its salts (e.g., 1 ml of a 0-01 per cent, In/ V solution prepared by diluting a 0.05 per cent.m/ V stock solution). When the solubility was insufficient, a 0.01 per cent. vn/V solution of the base in a suitable organic solvent (usually chloro- form or ethanol) was prepared. In such instances 1 ml of the solution was introduced into a 250-ml conical flask and evaporated to dryness at 50 to 60 "C under reduced pressure. It is imperative that all the organic solvent is removed because even a trace amount can drastically reduce the effective- ness of the oxidant, particularly with the permanganate reagent.* Present address : Government Laboratory, Oil Street, North Point, Hong Kong. Q SAC and the anthors. 1.556 CADDY, FISH AND TRANTER: OXIDATION PROCEDURES [AYZalyd, VOl. 99 Tt was neceiwry to use glass-coxtetl magnetic followers as both P T F F and other plastic- coated followers appeared to adsorb some of the oxidation product. The alkaline permanganate reagent should be prepared immediately before use by mixing 5 nil of 4 per cent. aqueous potassium permanganatc solution with 5 ml of 8 N aqueous sodium hydroxide 3olution. The acidic dichrornate reagent was prepared by mixing 5 ml of 4 per cent. potassium dichrornate solution with 15 ml of 12 II' sulphuric acid. 2 . 3. to the flask and heat the mixture at 60 to 70 "C under reflux, with constant stirring, for approxi- mately 40 minutes.Cool the flask and, with the condenser still in position, rinse them with approximately 5 ml of water. Transfer the contents of the conical flask into a 40-ml glass- stoppered test-tube. Place an aliquot of the hexane layer in a cell of l-cm path length and determine the ultraviolet spectrum between 225 and 400 nm, using hexane in the reference cell. Measure the absorption maxima at the appropriate wavelengths (Tables I and 11) and calculate concentrations by reference to a standard graph. TABLE I OXIDATION PRODUCTS, THEIR ABSORPTION MAXIMA AND YIELDS FROM CONTAINING A DIPHENYLMETHYLENE ETHER GROUP Compound Benztropine Bromodiphenh ydramine ChIorphenoxaminc Di phen ylpyral ine Diphenh ydramine Embramine Orphenadrine Benzh ydrol Deptropine Q R' R" H ANCH3 H CH,CH,N(CH,), CH, CK,CH,N(CH,) , H O C H .H CH,CH,N(CH,) CH, CH,CH,N(CHJ H CH,CH,N(CH,), H H ACH3 go R"' H 4-Br 4-C1 H I< 4-Br 2-C€I, H COMPOUNDS Oxidation yield, per cent. * - Oxidation product A,,./nm A Benzophenone 247 90 4-Eromobenzophenone 257 96 4-C.hlorobenzophenone 254 89 Benzophenone 247 96 13enzophenone 247 9Q 4Bromobenzophenone 257 71 2-Methylbenzophenone 247 90 Benzophenone 247 97 .4 :I)iHDBCHS B :Anthraquinone I3 73 66 + 64 70 .I. 63 100 _ _ * Alean values from at least two determinatiom. A: Potassium dichromate, 1 per cent. m/V in 9 N sulphuric acid a t about 60 "C. 13 : Potassium permanganate, 1 per cent. m/ I' in 6 N sodium hydroxide solution a t about 60 "C. t Insignificant. 2 DiHDBCH : 10,l l-Dihydro-5H-dibenzo[a,d]cyclohepten-5-one.Preparation of standard calibration graphs-Calibration graphs were prepared from the results of triplicate determinations by each of the oxidation procedures described above. The drug concentrations were varied so as to give between 2.0 and 20.0 pg ml-l of the benzo- phenone in the final hexane solution (0.1 to 1.0 ml of 0.01 per cent. m/V aqueous solutions of drug salts were used). HYDROLYSIS OF DIPHENHYDRAMINE- hydramine was hydrolysed under both acidic and alkaline conditions. In order to determine the possible reaction route with ethers, the compound diphen-September, 19741 IN THE ASSAY OF SOME DRUGS TABLE I1 OXIDATION PRODUCTS, THEIR ABSORPTION MAXIMA AND YIELDS FROM COMPOUNDS CONTAINING A DIPHENYLMETHYLENEAMINO GROUP 557 Compound R' Oxidation yield, per cent.* - R" Oxidation product Amax./nm A B A ~ C ( C " 3 ) ~ 4-Chlorobenzophenotie 254 32 '32 -NwN-c"2 Buclizine n I \ Chlorc Cyclizine yclizi tie E} -N\_/NCHs 4-Chlorobenzophenone 254 4 89 Benzophenone 247 4 04 4-Chlorobenzophenone 264 9 88 n Meclozine -NwN-cH2Q --.I CH3 * Mean values from at least two determinations. A: Potassium dichromate, 1 per cent. m/V in 9 h' sulphuric acid at about 60 "C. B: Potassium permanganate, 1 per cent. m/V in 6 N sodium hydroxide solution at about 60 "C. Acid hyd.p.oZysis-A solution of about 100 mg of diphenhydramine hydrochloride in 50 ml of 9 N sulphuric acid was heated at 60 "C under reflux for about 2 hours. After cooling, the solution was extracted twice with equal volumes of hexane. The ultraviolet spectrum of the hexane phase was recorded and compared with that obtained with authentic benzhydrol and diphenhydramine (Table 111). About 10 ml of the hexane solution were evaporated to small volume (about 0.2 ml) on a water-bath and a sample was submitted to gas-chromatographic analysis (condition 1).AZkaZine hyllrolysis-A solution of about 100 mg of diphenhydramine hydrochloride in 50 ml of 6 N sodium hydroxide solution was heated at 60 "C under reflux. After cooling, the solution was adjusted to pH 4 with sulphuric acid and quickly extracted (shaking vigorously by hand for 3 minutes) with two 50-ml volumes of hexane. The organic phase was then subjected to ultraviolet spectrophotometric and gas-chromatographic examination as described above. TABLE I11 ULTRAVIOLET SPECTROPHOTOMETRIC AND GAS - LIQUID CHROMATOGRAPHIC RESULTS FOB BENZHYDROL, DIPHENHYDRAMINE AND THE HYDROLYSIS PRODUCTS OF DIPHENHYDRAMINE Gas - liquid chromatographic retention value Amax.in hexanelnm relative to Material \ benzhydrol* f A Benzh ydrol 252-5 258 265 268(S) 1.oot Diphenhydraniine 252 257.5 264(S) 268(S) 1.77 Diphenhydramine hydrolysis pyoducts- Acid hydrolysis, acidic fraction 252-5 258 265 268(S) 1.00 Alkaline hydrolysis, acidic fraction No spectrum obtained I Alkaline hydrolysis, basic fraction 252 257.5 264(S) 268(S) 1.77 S = Shoulder. * Benzhydrol had a retention time of 2.6 minutes under the conditions used. t Condition 1 (see under Experimental).568 CADDY, FISH AIiD TRANTER: OXIDATION PROCEDURES [A%a&St, VOl. 99 The aqueous phase was made alkaline with 6 N sodium hydroxide solution and re- extracted with 50 ml of hexane, the final hexane layer being examined as before. The retention data for the products of hydrolysis are given in Table 111. EFFECT OF ACID CONCENTRATION ON THE RATE OF HYDROLYSIS OF DIPHENHYDRAMINE- About 100 mg of diphenhydramine hydrochloride were dissolved in sulphuric acid of various concentrations and the solutions were maintained at 40 "C for 45 minutes. After 25 and 45 minutes, aliquots of the solutions were removed, cooled, extracted quickly with an equal volume of hexane in order to remove liberated benzhydrol and the ultraviolet spectrum o f the aqueous phase was recorded in each instance (Table IV).TABLE IV ACID CONCENTRATIONS USED FOR THE HYDROLYSIS OF DIPHENHYDRAMINE AND THE ABSORBANCE OF EACH AQUEOUS PHASE AFTER 25 AND 45 MINUTES Absorbance at 257.5 nm after Concentration I A 'L of acid/N 25 minutes 45 minutes 6.0 0.06 * 0.01 * 4.8 0.12 0.06" 3-6 0.83 0.2 1 2.4 0.95 0.56 1.2 1.22 1.06 0.3 1.34 1-31 0-ot 1.40 1 *40 * No distinct maxima.t Aqueous drug solution. <;AS CHROMATOGRAPHY OF OXIDATION PRODVCTS- In order to confirm the identity of the various oxidation products listed in Tables I and I1 gas - liquid chromatography was employed. Hexane solutions from the oxidation of each particular drug were bulked and reduced to small volume (about 0.2 ml) at GO "C under reduced pressure. Samples of these solutions (2 to 3 pl) were injected on to the chromatographic column (condition 2) and retention values recorded relative to a suitable standard.Examination was also made of solutions of authentic benzophenones prepared at a concentration of approximately 0.05 per cent. m/V in hexane, which were diluted sufficiently to give detector responses similar to those obtained with the oxidation products. RESULTS AND DISCUSSION PRECISION OF THE OXIDATION PROCEDURES- For the study of precision of the methods used, diphenhydramine and cyclizirie were selected as, apart from being important in the context of toxicology, they represented the different chemical groups being examined. These drugs were subjected to fifteen replicate determinations by the procedure that gave the highest oxidation yield in each instance, and for diphenhydramine the alternative potassium permanganate procedure was also examined.The results (Table V) indicated that the selected oxidation procedures were acceptable for the assay of dipheihydramine and cyclizine and there is no reason to believe that the other drugs which gave a high oxidation yield could not be assayed with the same levels of precision. The increase in coefficient of variation that accompanies a decrease in oxidation yield is a trend exhibited in the oxidation of other compounds, which will be the subject of a further paper. TABLE V REPRODIJCIBILITY OF YIELDS ON OXIDATION FOR REPRESENTATIVE COMPOUNDS Mean absorbance Coefficient Oxidation Drug procedure* (247 nm) deviation variation per cent. Oxidation at Anl$*. Standard of yield, Diphenhydramine A 1-30 0.034 2.6 97 Cyclizine B 1-13 0.034 3.0 84 (20 pg ml-l) B 0.97 0.033 3.4 70 (20 p g ml-1) *A: Potassium dichromate, 1 per cent.m/V solution in 9 N sulphuric acid a t about 60 "C. B: Potassium permanganate, 1 per cent. m/V solution in 6 N sodium hydroxide solution a t about 60 "C.September, 19741 I N THE ASSAY OF SOME DRUGS 559 YIELDS OETAINED ox OXIDATION OF DIPHENYLMETHYLENE ETHERS- Table I. The drugs studied, their oxidation products and their oxidation yields are given in Yields were calculated from the equation 60 x x A!!d A b x Mb Percentage oxidation = where Z is the experinlentally observed absorbance following an oxidation equivalent to 20 pg nil-l of drug (or drug salt) in the final hexane solution; f i f d is the relative molecular mass of the drug (cr drug salt); .4b is the enhanced absorbance value of 10 pg ml-l of the authentic benzophenone in hexane (Note 4); and M b is the relative molecular mass of the benzophenone.Note 4. In calculating the yields on oxidation it was necessary to take into account the enhance- ment of the ultraviolet absorption of the benzophenones themselves brought about by the oxidation procedure and considered t o be attributable t o evaporation of the hexane.4 The enhancement for ben- zophenone was 7.7 per cent. and the assumption was made that all benzophenones showed a similar enhancement. Oxidation with acidic dichromate-It is difficult to comment on variations in yield of 90 to 98 per cent. for those compounds which possess two aliphatic hydrogen atoms but this effect must be due, at least partly, to the variation in purity of the commercial samples tested.No attempt was made to assay their purity but some of them were known to be less than 100 per cent. pure (e.g., benztropine mesylate, 95.7 per cent.), whereas the yields on oxidation were calculated on the basis of 100 per cent. purity for each drug. With the exception of chlorphenoxamine (89 per cent.) and embramine (71 per cent.), oxidation of the studied compounds with hot acidic dichromate solution gave yields of 90 per cent. or more of the expected benzophenone (Table I). The lower yields from the former must arise from the presence of a methyl substituent instead of hydrogen on the a-carbon atom, as this is a feature present in the two drugs but absent from the remainder. Vessman, Hartvig and Stromberg2 found, by using a non-aqueous solution of chromic acid as oxidant, that yields in excess of 90 per cent.were obtained from ethers that possess a hydrogen atom on the a-carbon atom. Replacement of this a-hydrogen atom by a methyl group resulted in a lower yield. These authors pointed out that the yields parallel those obtained from the oxidation of benzhydrol and its a-methyl analogue. The same authors further agreed with the view of Sasakis and of Wallace, Biggs and Dahll that the oxidation of diphenhydramine proceeds via hydrolysis to benzhydrol. In the present work, hydrolysis with sulphuric acid was found to yield benzhydrol (con- firmed by gas - liquid chromatography and ultraviolet spectrophotometry, Table 111) in agreement with the report of Kikkawa, Sasaki, Iwasaki and Ueda.6 The rate of hydrolysis was acid dependent (Table IV), which is consistent with the report of Staude and Patat.' These authors, while pointing out that no valid information on the kinetics of hydrolysis of aliphatic ethers is available, suggested two possible reaction routes following protonation of the ether (Scheme 1).The second reaction route would appear to be particularly acceptable in the present instance as the carbonium ion W O U J ~ be stabilised by resonance. Scheme 1 1 2 slow fast R '6" ROH C R6Hz - 2ROH + H30' R' H20 R .- fast R --+I- slow ROH + R' 's 2ROH i- "3"560 CADDY, FISH AND TRANTER: OXIDATION PROCEDURES [Analyst, \'Ol. 99 The intermediates with chlorphenoxamine and embramine would possess the extra stability of a tertiary cx-carbon atom and a halogen atom, which coulcl enter into resonance, consequently resulting in their lower yields on oxidation.The oxidation of a secondary alcohol with aqueous dichromate is usually quantitativc8 Vessman et aZ.2 reported the quantitative oxidation of heiizhydrol using non-aqueous condi- tions and this finding has been confirmed in the present work with aqueoxs acidic dichromate. The oxidation rate law,9 which applies to all secondary alcohols, is complex but readily indicates that the rate is dependent on both acid and oxidant concentration. The reaction scheme proposed9 to fit the kinetic data is shown in Scheme 2. Scheme 2 R2CHOH + H C r q i- H' - R2CHOCr03H + H20 Wibergs suggested three possible mechanisms for the decomposition of the chromate ester (Scheme 3).Whichever mechanism is in operation, it immediately becomes apparent that none is suitable for chlorphenoxamine and embramine because the alcohols arising from their hydrolysis will not possess cc-hydrogen atoms. The oxidation rate for tertiary alcohols has been found to be independent of the oxidant concentration and to correspond to the rate of acid dehydration of alcohols.lOrll On this basis, the oxidation of embramine and chlorphenoxamine can be rationalised by the scheme illustrated in Fig. 1. Oxidation with alkaliite permaizganate-Oxidation with alkaline permanganate of ethers of the type under discussion led to lower yields of benzophenones than oxidation with acidic dichromate, with chlorphenoxamine and embramine giving no detectable yields.Scheme 3 1 2 3September, 19741 IN THE ASSAY OF SOME DRUGS 561 X Q That diphenhydramine was inert to alkaline hydrolysis became evident from both ultra- violet spectrophotometric and gas-chromatographic examination of the hydrolysis mixture (Table 111), which is not surprising as alkyl ethers are reported to be stable to alkaline oxida- tion.' Obviously, the oxidation proceeds via a totally different route from that of oxidation with dichromate and this conclusion is further supported by the fact that benzhydrol gives a quantitative yield of benzophenone following oxidation with permanganate, whereas the best yield from the ether is only 73 per cent. 8 I H+ Q Q CH~-C-OCH~CH~N(CH~)Z - CH,-C-OH 0 0 X Q X X 1- H20 CxCH2 [Ol 4--- Embramine X 5 H Chlorphenoxamine X = CI Fig.1. Proposed oxidation route of embramine and chlorphenoxamine While no report on the mechanism of the oxidation of alkyl ethers under alkaline condi- tions could be found, the reaction probably proceeds via hydrate abstraction from the a- carbon atom. This view is supported by the fact that the two compounds chlorphenoxamine and embramine, which do not possess an a-hydrogen atom, appear to be inert to oxidation with permanganate. In addition, the carbonium ion produced by hydride abstraction would be stabilised by resonance. Wallace et aZ.l and Caddy, Fish, Mullen and Tranter3 found that the oxidation of diphenhydramine with alkaline permanganate gave a good yield of benzophenone and the latter group of workers also reported that the oxidation of orphenadrine gave a good yield.Deptropine gives a 55 per cent. yield of anthraquinone, which is typical of l0,ll-dihydro- 5N-dibenzo[a,d]cyclohepten-5-yl compounds. Oxidation of these and related compounds will be discussed in a separate paper. YIELDS OBTAINED ON OXIDATION OF DIPHENYLMETHYLENEAMINES- Compounds of this type were found to be readily oxidised with hot alkaline permanganate to the expected benzo- phenone in good yield. The amines examined in the present work are listed in Table 11. Conversely, hot acidic dichromate gave poor yields.562 CADDY, FISH AND TRANTER: OXIDATION PROCEDURES [Analyst, Vol. 99 Oxidation with alkaline permanganate-Under the basic conditions employed, solutions of permanganate might be expected to abstract a-hydrogen atoms12 rather than substitute on the nitrogen atom.A pertinent example of tertiary amine oxidation with permanganate is that of NN-dimethylbenzylamine to benzaldehyde.8 If one of the a-hydrogen atoms is replaced by a phenyl group the resulting amine will be similar to those under discussion, the corresponding oxidation product being benzophenone. The permanganate oxidation of the present drugs would be more facile than that of NN-dimethylbenzylamine as hydrogen atoms on tertiary carbon atoms are well known for their lability. The probable reaction route, which bears a superficial resemblance to that proposed for permanganate oxidation of ethers (as indicated above), is illustrated in Scheme 4. Scheme 4 I CH-NR2 Q x -H’ d Q +C--/\NR2 0 X + C=NR2 0 X 8 6 x X =p CI or H The slightly lower relative yield of the oxidation product from cyclizine may possibly be attributed to lack of purity in the commercial product examined, whereas the still lower yield from hydroxyzine may reflect the presence of an additional centre for oxidation, namely the primary alcohol group.Oxidation of this group to a carboxylate ion might either inhibit the abstraction of a hydride ion or result in stabilisation of the initially formed carbonium ion. Oxidation with acidic &chromate-Neumann and Gould, l3 in their studies on the oxidation of arylalkylamines, favoured the first of their proposed routes (Scheme 5) involving a free- Scheme 6 1 Ar R~CH-N’ a ‘R Ar H+ 2 R2CH-N’ - ‘R Ar Ar +/ - R2C-N-R - H+ -q + / H radical mechanism. However, in view of the fact that some yields were obtained (Table VI), and because of possible delocalisation of the negative charge resulting from the removal of a proton (Fig.2), the second route seems more likely.September, 19741 I N THE ASSAY OF SOME DRUGS 563 rl o C-N+ H R~ 0 C-N+ H R2 Fig. 2. Delocalisation of the anion formed by oxida- tion of diphenylmethyleneamine with acidic dichromate solution Other workers have commented on the oxidation with acidic dichromate or alkaline per- manganate of some of the amines studied in the present work. In assays for amitriptyline14 and chlorprothixene, 15 using oxidation with buffered alkaline pennanganate, cyclizine has been described as an interfering substance. Caddy et aL3 reported good yields of benzo- phenones following the oxidation of cyclizine and chlorcyclizine with a similar reagent.Vessman et aZ.2 reported low yields of benzophenones from those compounds following their oxidation procedure with non-aqueous dichromate, while Vessman and Hartvig16 reported yields of 69.0 to 72.5 per cent. for cyclizine, chlorcyclizine, hydroxyzine and meclozine, using a similar reagent but with a reaction time exceeding 24 hours. The latter authors also reported that doubling the oxidant concentration doubled the oxidation rate for cyclizine, but in the present work (with aqueous media) changing the dichromate concentration from 1 to 4 per cent. had only a minimal effect on the yield following a reaction time of 40 minutes although yields were significantly increased by increasing the acid concentration (Table VI).TABLE VI VARIATION I N YIELDS ON OXIDATION FOR CYCLIZINE AND CHLORCYCLIZINE WITH CHANGES I N ACID AND DICHROMATE CONCENTRATION Sulphuric Dichromate acid/n concentration, per cent. 9 1 4 12 1 4 16 1 4 20 1 4 Oxidation yield, per cent. Cyclizine Chlorcyclizine 4 4 6 6 4 4 6 6 9 20 11 32 24 41 28 42 r A > CONCLUSIONS Oxidation of drugs containing the diphenylmethyleneamino group with alkaline per- manganate gave good yields of benzophenones, which were shown to obey Beer - Lambert’s law over the concentration range studied. Oxidation with acid dichromate was found to be inadequate. With the exception of the oxidation of embramine and chlorphenoxamine with alkaline permanganate both reagents appear satisfactory for the oxidation of most compounds con- taining the diphenylmethylene ether group. Satisfactory calibration graphs were obtained for all other compounds over the concentration range studied but the acidic potassium di- chromate method was the better of the two as it invariably gave higher yields of benzophenone. REFERENCES 1. 2. 3. 4. 5. 6. 7. Wallace, J. E., Biggs, J. D., and Dahl, E. V., Analyt. Chem., 1966, 38, 831. Vessman, J., Hartvig, P., and Stromberg, S., Acta Pharm. Suec., 1970, 7, 373. Caddy, B., Fish, F., Mullen, P. W., and Tranter, J., J . Forens. Sci. Soc., 1973, 13, 127. Tranter, J ., Ph.D. Thesis, University of Strathclyde, Glasgow, 1973. Sasaki, D., J. Osaka Cy Med. Cent., 1954, 3, 207. Kikkawa, M., Sasaki, D., Iwasaki, T., and Ueda, J., Ibid., 1956, 3, 69. Staude, S., and Patat, I?., in Patai, S., Editor, “The Chemistry of the Ether Linkage,” Interscience Publishers Inc., New York, 1967, pp. 21-80.564 8. 9. 10. 1 1 . 12. 13. 14. 15. 16. CADDY, FISH AND TRANTER Wiberg, K. B., “Oxidation in Organic Chemistry, Part A,” Academic Press, New York and Westheimer, I?. H., and Novick, A., J . Chem. Phys., 1943, 11, 506. Sager, W. F., J . Amev. Chem. Soc., 1956, 78, 4970. Rocek, J., Chemicke‘ Listy, 1957, 51, 1838. Schechter, H., and Rawalay, S. S., J . Amer. Chem. Soc., 1964, 86, 1706. Neumann, F. W., and Gould, C. W., Analyt. Chem., 1963, 25, 751. Wallace, J. E., and Dahl, E. V., J . Forens. Sci., 1967, 12, 484. Wallace, J . E., J . Pharnt. Sci., 1967, 56, 1437. Vessman, J., and Hartvig, P., A d a Pharm. Sztec., 1971, 8, 229. London, 1965. Received January 24th, 1974 Accepted March 22nd, 1974