Analyst, January, 1973, Vol. 98, $p. 9-18 9 The Analysis of Carbamate and Urea Herbicides by Fluorimetry of Their Dansylated Amine Moieties BY R. W. FREI, J. F. LAWRENCE AND D. S. LEGAY (Trace Analysis Research Centre, Department of Chemistry, Dalhousie University, Halifax, Nova Scotia, Canada) Carbamate and urea herbicides have been analysed by thin-layer chro- matography and fluorimetry of the 1-dimethylaminonaphthalene-5-sulphonyl (dansyl) derivatives of their aniline hydrolysis products. The herbicides were hydrolysed with 1 M sodium hydroxide solution at 80 "C for 40 minutes and then the liberated anilines and amines were extracted into hexane. Aliquots of the hexane solution were spotted on to a chromatographic plate and made to react in situ with a t least a ten-fold excess of dansyl chloride, which was spotted over them.Next, the plate was sprayed until moist with a 20 per cent. solution of triethanolamine in propan-2-01 and the spots were analysed fluorimetrically in situ. The hydrolysis and coupling reaction of the herbicides, fluorescence phenomena and chromatographic properties of the derivatives were investigated. Detection limits of 1 ng per spot taken at a 3: 1 signal to noise ratio and a reproducibility of 3 to 6 per cent. relative standard deviation can be expected and the calibration graphs arc linear up to 300 ng per spot. Natural water samples can be analysed to determine concentrations of herbicides of a few parts per billion (log) with high recoveries and without the necessity of a clean-up step. UREA and N-phenylcarbamate herbicides have been of considerable interest recently because of the advantages of their rapid degradation in the environment.Some of the more important discussions on their mode of action are li~ted.l-~ While the analysis of these compounds by direct gas - liquid chromatographic methods is the procedures require relatively mild gas - liquid chromatographic conditions that are not always practical when carrying out sample analyses. The clean-up problem in gas - liquid chromatography, which often consumes much time and effort, is also an unfavourable factor with this technique. Thin- layer chromatographic methods, which usually involve the formation of coloured derivatives by means of chromogenic spray reagent^,^ have also found wide use in the analysis of these types of corn pound^.^-^^ However, none of these methods is genuinely quantitative at the level of detection (a few nanograms) that is normally required for residue analysis.The combination of thin-layer chromatography with fluorescence offers a considerable advantage as it increases the sensitivity by 10 to 100 times. The determination of Sevin (carbaryl, 1-naphthyl N-methylcarbamate) as its naturally fluorescent hydrolysis product, 1-naphthol, by an in situ fluorimetric thin-layer chromatographic technique13 demonstrated the applicability of such an approach to quantitative pesticide analysis. This work has been extended to include the analysis of non-fluorescent pesticides and fluorigenic spray reagents were developed for the analysis of a number of organophosphate and carbamate insecti- cides.14-16 Although the spray techniques are rapid and simple, they are not selective, and are therefore more susceptible to interferences and give a sensitivity only about ten times greater than for the chromogenic spray procedures.An alternative to these spray methods is fluorigenic labelling, which involves the formation of fluorescent derivatives that are separated by thin-layer chromatography and quantitatively determined by use of fluori- metry in situ. This technique has the advantage that it does not require a spray reaction for the formation of the fluorescent spots, thus avoiding background irregularities caused by uneven spraying. The fluorigenic labelling of non-fluorescent compounds has found a great deal of use in biological and pharmaceutical analyses.Amino-acids, peptides and biogenic amines have @ SAC and the authors.10 FREI, LAWRENCE AND LEGAY: ANALYSIS OF CARBAMATE AND UREA [Analyst, Vol. 98 been fluorimetrically determined for more than a decade as their dansyl (1-dimethylamino- naphthalene-5-sulphonyl) derivatives following separation by thin-layer chromatography or high-voltage electrophoresis. Seiler and Wiechmannl' have recently reviewed progress in this field with a detailed discussion of the reactions of dansyl chloride with amines and phenols. Preliminary work on the analysis of carbamate insecticides by this fluorigenic labelling technique has been carried out by Frei and Lawrence,leJ9 who used dansyl chloride as the labelling reagent.This compound reacts with both amine and phenol hydrolysis products of N-methylcarbamates, resulting in two derivatives suitable for the determination of a few nanograms of carbamate insecticide residues. The reaction procedure has since been investigated20 and results indicate that complete formation of the dansyl derivatives of the carbamates takes place in less than 1 hour. The fluorescence behaviour of these derivatives was examined21 in order to determine the stability of the derivatives with time and to investigate the optimum conditions for their quantitative determination. The thin-layer chromatographic properties of the dansyl derivatives of a number of carbamates have also been studied for both one and two-dimensional separations.22 The application of such a method to the quantitative analysis of N-methylcarbarnates in natural water samples without using clean-up procedures has recently been reported.23 The use of dansyl chloride as a fluorigenic labelling reagent for herbicides in the carbamate and urea classes seems particularly promising and worthwhile because upon hydrolysis these compounds yield amines or anilines that can react with dansyl chloride to form highly fluorescent derivatives.The investigation of the dansyl labelling and quantitative analysis of these compounds is the subject of the present work. EXPERIMENTAL REAGENTS- The common and systematic names of the herbicides used in this work are given in Table I. Stock solutions of these compounds were prepared a t a concentration of 1 mg ml-1 in redistilled analytical-reagent grade dichloromethane. The labelling reagent consisted of a 0.2 per cent.solution of analytical-reagent grade dansyl chloride (1-dimethylaminonaphtha- lene-5-sulphonyl chloride, obtained from Mann Research Laboratories, New York, U.S.A.) in redistilled reagent grade acetone. Spray solutions consisting of 20 per cent. of triethanol- amine in propan-2-01 and 20 per cent. of paraffin oil in toluene were prepared. A 1 M solution of sodium hydroxide was used to hydrolyse the herbicides. Dansyl derivatives of aniline, 3-chloroaniline (obtained from Matheson Coleman and Bell, Norwood, Ohio, U.S.A.) and 3,4-dichloroaniline [obtained from BDH (Chemicals) Ltd., Poole, Dorset] were used as chromatographic standards for derivative identification.All other materials were of analytical-reagent grade quality. TABLE I COMMON AND SYSTEMATIC NAMES OF THE HERBICIDES USED Compound Systematic name Carbamates- IPC (propham) . . .. . . Isopropyl N-phenylcarbamate CIPC (chlorpropham) . . . . Isopropyl N-(3-chloro) phenylcarbamate Swep .. .. .. . . Methyl N-(3,4-dichloro)phenylcarbamate Barban .. .. .. . . 4-Chlorobut-2-yn-1-y1 N-(3-~hlorophenyl)carbamate Linuron . . .. .. . . N-Methyl-N-methoxy-N'- (3,4-dichloro) phcnylurea Diuron . . .. .. . . NN-Dimethyl-N'-(3,4-dichloro)phenylurea Maloran (chlorbromuron) . . N-Methyl-N-methoxy-N'-( 3-chloro-4-bromo)phenylurea Fluometuron . . .. . . NN-Dimethyl-"-( 3-trifluoromethy1)phenylurea Urea- REACTION PROCEDURE- Ten microlitres of a stock solution of the pesticide were placed in a 2-ml centrifuge tube and the solvent was evaporated in a water-bath at 40 "C.To the dry residue, 0.2 ml of sodium hydroxide solution was added and the tube was loosely stoppered and heated in a water-bath at 80 "C for 30 to 40 minutes. The tube was then cooled to room temperature,January, 19731 HERBICIDES BY FLUORIMETRY OF THEIR DANSYLATED AMINE MOIETIES 11 0.2 ml of hexane added and the tube shaken. The phases were permitted to separate and 10 p1 of the organic layer were spotted on to a 20 x 20-cm thin-layer plate at a point 2 cm from the bottom, 4p1 of the dansyl chloride solution being spotted 011 top of the sample spot, The plate was then kept covered with a clean glass sheet in the dark at room temperature for 60 minutes before elution with the chromatographic solvent.CHROMATOGRAPHY- The thin-layer plates were prepared with a Desaga (Heidelberg) thin-layer applicator by applying a slurry consisting of 30 g of silica gel G with calcium sulphate binder (obtained from Machery, Nagel and Co., Duren) and 60 ml of distilled water to the plates at a thickness of 250 pm. The plates were dried at 110 "C for 10 minutes prior to spotting, reaction and elution by the ascending-solvent technique at room temperature in the dark. The eluting solvent mixture was benzene - triethylamine - acetone (75 + 24 + l ) , freshly prepared before use. The plates were developed to a 13-cm solvent front, dried in a stream of cool air, sprayed until moist with one of the spray solutions mentioned above (to enhance and stabilise the fluorescence of the derivatives) and then dried again to remove the propan-2-01 or toluene from the plate.INSTRUMENTAL ANALYSIS- A Zeiss Chromatogram Spectrophotometer PMQ I1 was used in the fluorescence mode for all quantitative measurements. A 365-nm Zeiss filter was used with a mercury lamp as the excitation source and the fluorescence was monitored with an emission slit width of 0.3 mm for all work. Spectra and scanning peaks were recorded on a Honeywell Electronik 194 strip-chart recorder and the peak areas were integrated electronically with an Autolab Vidar 6300 digital integrator. Fluorescence spectra were also recorded in sita with an Aminco-Bowman Spectrophotofluorometer equipped with the thin-layer scanning attach- ment. For the qualitative, visual examination of the developed chromatographic plates, a Camag Universal ultraviolet lamp was used at 350 nm.ANALYSIS OF WATER SAMPLES- Three 500-ml samples of local lake water were spiked to give 10 p.p.b. (parts per 109) each of swep and IPC. The samples and a blank were extracted twice with 50-ml portions of dichloromethane. The combined extracts for each sample were reduced to 1 ml by rotary vacuum evaporation a t room temperature and subsequently transferred to 2-ml centrifuge tubes and evaporated to dryness in a water-bath at 40 "C under a gentle stream of air. To rinse down the sides of the tube and to re-dissolve the residue, 1 or 2 drops of acetone were added followed by 0 6 m l of sodium hydroxide solution. The contents were then treated as described under Reaction procedure.Each sample was spotted on to the thin-layer chromatographic plates in triplicate and compared with standards run on the same plate. Time/minutes Fig. 1. Rates of formation of aniline in 1 M NaOH solution for A, swep; B, linuron; and C, IPC a t 200ng each12 FREI, LAWRENCE AND LEGAY: ANALYSIS OF CARBAMATE AND UREA [AIZUhSt, VOl. 98 RESULTS AND DISCUSSION REACTION CONDITIONS- A 1 M solution of sodium hydroxide was found to be most suitable for the liberation of the anilines from the herbicides. The relative rates of aniline formation of a representative group of herbicides with a 1 M solution of sodium hydroxide can be seen in Fig. 1. These were determined by forming the dansyl derivative of the free anilines in the hydrolysis mixtures and fluorimetrically determining the amount of aniline liberated as a function of time.The hydrolysis is essentially complete in 60 minutes for most of the herbicides studied. A reaction scheme is shown in Fig. 2 for the formation of the derivatives of a typical urea and an N-phenylcarbamate. CA RBAM ATES (Cl PC) 1. NH-C-OCH ‘CH3 20H- ___t UREAS (LINURON) 1. CIL I c1 S02CI NH2 I + (CH3I2CHOH + Cog- CI NH2 I 9 CHsNHOCH3 + COi- CI CI Fig. 2. Dansylation of CIPC and linuron: 1, hydrolysis of the herbicides to give the free chloroaniline; and 2, reaction of the dansyl chloride onsilica gel G layer with the aniline to form the dansyl derivativeJanuary, 19731 HERBICIDES BY FLUORIMETRY OF THEIR DANSYLATED AMINE MOIETIES 13 = A = B I 80 120 Time/minutes Amount of dansyl chloride/pg Fig. 3.Rate of reaction of the free anilines of swep (A) and IPC (B) with a ten-fold excess of dansyl chloride on the thin- layer plates a t room temperature in the dark Fig. 4. Effect of the amount of dansyl chloride on the forma- tion of the aniline derivative of 260 ng of IPC after reaction for 60 minutes on the plate The coupling of the anilines with dansyl chloride takes place more rapidly directly on the thin-layer plates than in solution. In addition, when effecting the reaction in solution it is necessary to reduce the pH of the hydrolysis solution to 9 to 10 before the dansylation reaction will form derivatives in any significant yield. The rate of reaction to form anilines directly on the plate is shown in Fig. 3. About 40 to 50 minutes are required for complete formation of the derivatives; the N-phenylcarbamates form only the aniline derivative. For the N-phenylureas that contain a methoxy substituent on the methylamine moiety of the molecules, only the aniline derivative was found in the chromatogram.The ureas that contain a methylamine or dimethylamine moiety as well as an aniline group (ie., diuron and fluometuron in Table I) form dansyl derivatives with both hydrolysis products. The amount of dansyl chloride spotted over the hydrolysed herbicides was found to influence the results. Fig. 4 shows the effect of varying the dansyl chloride concentration on the yield of dansyl derivative for IPC (200 ng). An excess (about ten-fold) of reagent is required for complete conversion into the dansyl derivative.The dansyl chloride is in solution in acetone, which is an extremely polar liquid and washes the aniline and reagent into the 0 Second - Fig. 6 . Two-dimensional chromato- gram of the dansylation spray reaction with 200ng of hydrolysed swep. The first dimension was eluted with benzene - acetone (96 + 4) and the second with benzene - triethylamine - acetone (75 + 24 + 1). The cross-hatched area is the dansyl spray14 [Analyst, vol. 98 same zone, providing close contact between the two reactants. At the 200ng of herbicide level, the diameter of the spot is not important. However, at low concentrations it is preferable to keep the spots as small as possible for maximum contact between the labelling reagent and aniline. Also, small spots are much more desirable for chromatography because better separations can be achieved.An interesting use of the in situ reaction between the anilines and dansyl chloride is in the elimination, by two-dimensional chromatography, of any interferences that may be encountered. The anilines can be eluted in one direction before dansylation and then sprayed with dansyl chloride solution, as shown in Fig. 5, and eluted in the second dimension after dansylation is complete. To avoid any interference from excess of dansyl chloride, which appears as a broad band at or near the solvent front in the second elution, the plate can be sprayed with dilute base (0.1 M sodium hydroxide solution, for example) before the second elution. The dansyl-OH (1-dimethylaminonaphthalene-5-sulphonic acid) formed does not move in the solvent systems used.As a more polar solvent system is required for elution of the anilines the first dimensional run will remove ‘many interferences that would cause problems in the analysis of the dansyl derivatives in the second elution. In Fig. 5 a two- dimensional separation of a sample of swep is shown. FLUORESCENCE PHENOMENA- The spectra for the dansylated aniline derivatives from IPC and linuron are shown in Fig. 6, while the excitation and emission maxima for the dansylated aniline derivatives of all the herbicides studied by use of the two sprays are given in Table 11. The paraffin oil spray caused a bathochromic shift in the emission maxima of the derivatives as compared with the triethanolamine spray. The spectra of all derivatives are similar for a given spray.This last property is useful if more than one derivative is present as all the derivatives can be scanned under the same instrumental conditions. On the other hand, the fluorescence spectrum is consequently not very useful as a qualitative indication of an aniline. FREI, LAWRENCE AND LEGAY: ANALYSIS OF CARBAMATE AND UREA Excitation Emission 300 400 500 600 700 Wave I en g t h I n rn Fig. 6. Fluorescence spectra recorded on the hminco-Bowman instrument with a 0.2-mm excita- tion slit of the dansylated anilines of IPC and linuron on silica gel G : broken line, IPC; and solid line, linuron The effects of constant irradiation by the ultraviolet excitation light of the Zeiss and Aminco-Bowman instruments on the fluorescence intensities of the aniline derivatives are shown in Fig.7, with use of triethanolamine as the stabilising spray. A greater degree of decomposition is observed with the Zeiss than with the Aminco-Bowman ultraviolet source. This occurrence was also noticed in previous work with N-methylcarbamate derivatives.21 The triethanolamine spray is recommended (Fig. 8) for increased stability of the spots under ultraviolet light and daylight conditions. The results of multiple scanning of the dansylated aniline derivative of IPC are shown in Fig. 8 and they indicate that scanning itself causesJanuary, 19731 HERBICIDES BY FLUORIMETRY OF THEIR DANSYLATED AMINE MOIETIES 15 TABLE I1 FLUORESCENCE MAXIMA OF THE DANSYLATED ANILINE DERIVATIVES OF THE HERBICIDES STUDIED Compound Spray* Barban, CIPC .. .. TEA IPC .. .. .. TEA Swep, linuron, diuron . . TEA Fluometuron . . .. TEA Maloran . . .. .. TEA PO PO PO PO PO Aminco-Bowman equipment$ excitation and emission Zeiss equipment emission? wavelength/nm wavelengthslnm 626 360, 612 632 362, 632 638 366, 636 635 366, 636 626 360,610 636 370, 636 620 360, 606 632 365, 628 610 366, 606 536 366, 636 * TEA = 20 per cent. V/V triethanolamine in propan-2-01; t A 0.3-mm emission slit width and a 365-nm excitation filter were used. $ A 0.2-mm excitation slit width was used. PO = 20 per cent. V / V paraffin oil in toluene. most of the degradation of the spots. However, under normal conditions only one scan per spot is required for a quantitative determination. Storage of the plates covered and in the dark increases the stability of the derivatives by more than 4 weeks with no detrimental effects21 A 15 30 45 Time/m i nu tes Fig.7. Effect of ultraviolet excita- tion light on the fluorescence intensities of the aniline derivatives. Curves A to D were obtained by using the Aminco- Bowman instrument and curves E to H were recorded on the Zeiss instrument. Triethanolamine was used as the spray. Curves A and E,-linuron, diuron, maloran and swep; B and F, CIPC and barban; C and HI fluometuron; and D and G, IPC CHROMATOGRAPHY- A typical thin-layer chromatogram of the dansyl derivatives of a number of herbicides is shown in Fig. 9. The R, values decrease with the increase in number of halogen sub- stituents on the anilines (see also Table I). The aniline derivatives are much more polar than the phenol derivatives of N-methylcarbamates,22 therefore separation of these two classes16 FREI, LAWRENCE AND LEGAY: ANALYSIS OF CARBAMATE AND UREA [AndySt, VOl.98 Number of scans Fig. 8. Influence of scanning on the fluores- cence intensity of the dansylated aniline derivative of IPC; curve A, triethanolamine spray; and B, paraffin oil spray as dansyl derivatives is easy. The dansyl-OH remains at the origin in the solvent systems investigated and hence does not interfere in quantitative analysis. The excess of dansyl chloride that may still be present travels well above the dansylated aniline derivatives and does not interfere. Solvent front 1 1 ; . ? t , 3 4 5 Dimethylamine - derivatives -Start Fig. 9. Thin-layer chromatogram of the dansyl derivatives of five herbi- cides on silica gel G developed with benzene - triethylamine - acetone (76 + meturon; 4, diuron; and 5, linuron.Dimethylamine derivatives are also formed in the reactions with fluo- meturon and diuron 24 + 1 ) ; 1, IPC; 2, CIPC; 3, flue- QUANTITATIVE ANALYSIS- The reproducibility of the method was investigated for linuron determinations and found to be 4.6 per cent. average relative standard deviation with a range of standard deviations of 2.3 to 6-8 per cent., determined on concentrations ranging from 40 to 400ng per spot, with at least four plates for each concentration and a minimum of seven spots per plate (Table 111). The results obtained for linuron can be assumed to be repre- sentative of the herbicides studied because they are structurally similar and the reaction and chromatography stages are the same.January, 19731 HERBICIDES BY FLUORIMETRY OF THEIR DANSYLATED AMINI: MOIETIES 17 Visual detection limits under the ultraviolet lamp were about 5 to long per spot for all compounds studied, while 1 ng per spot or less can be determined with the Zeiss equipment at a 3: 1 signal to noise ratio under optimum conditions. The calibration graphs pass through the origin and are linear up to 200 to 300ng per spot with essentially no change in slope over a 30-minute period when using either spray.TABLE I11 REPRODUCIBILITY STUDIES ON THE DANSYL LABELLING OF LINURON Excitation wavelength 350 nm; emission wavelength 510 nm; emission slit width 0.3 mm Concentration, Relative standard deviation, Average, Signal to noise ng per spot per cent.* per cent.ratio 400 5-3, 4.6, 5.3, 3-8 4.7 240 200 2.3, 4.8, 5.2, 3.2 3.9 200 80 4.5, 6.2, 4.6, 5.3 6.1 05 40 4.1, 6.8, 4.1, 5.1 5-0 60 * Four plates were analysed for each concentration with a minimum of 7 spots per plate. The analysis of local lake water samples spiked with swep and IPC was carried out with little interference from co-extractives. A sample containing only swep at a concentration of 10 p.p.b. gave 102 & 5 per cent. recovery, while samples containing both swep and IPC gave recoveries of 98 In Fig. 10, a thin-layer chromatogram and fluorescence scan of a swep determination are shown. 5 and 84 & 4 per cent., respectively. 0 0 2 t t i I i l Scan- Fig. 10. Chromatogram and fluorescence scan for a 10 p.p.b.swep-containing water sample extract after dansyl chloride labelling ; 1, dansyl-OH and impurity at the origin; 2, dansylated aniline (swep); 3, excess of dansyl chloride; and 4, solvent front and impurity. A Zeiss 365-nm excitation filter was used and the fluorescence monitored at 525 nm with a 0.3-mm slit width. Scanning speed 15 cm min-l; and chart speed 40 cm min-l CONCLUSIONS The dansyl labelling reaction of the anilines resulting from the hydrolysis of the herbicides on thin-layer plates's simple, rapid and reproducible. The method is suitable as an alternative technique to gas - liquid chromatography or as a check method for gas - liquid chromato- graphic results as similar detection limits and precisions can be obtained.24 The quantitative determination of two herbicides (at concentrations of 10 pg 1-l) in natural water samples was successfully carried out without necessitating a clean-up step.18 FREI, LAWRENCE AND LEGAY The labelling reaction with dansyl chloride carried out directly on thin-layer chromato- graphic plates has been found to be applicable to other non-volatile amines, including amino-acids (Lawrence, J.F., and Frei, R. W., unpublished work). This work was supported by funds provided by a Public Health Research Grant (Project No. 602-7-141) of the National Health Grants Program. Additional assistance was provided by the National Research Council of Canada. D. S. LeGay is grateful for the award of a Summer Research Scholarship of the Atlantic Provinces Inter-University Committee on the Sciences (A.P.I.C.S.). 1.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. REFERENCES Moreland, D. E., A . Rev. PI. Physiol., 1965, 18, 365. Bleidner, W. E., J . Agric. Fd CJzem., 1954, 2, 682. Hilton, H. W., and Nomura, N., Weed.Res., 1964, 4, 216. Sheets, T. J., and Harris, C. I., Residue Rev., 1966, 11, 119. Cohen, I. C., and Wheals, B. B., J . Chromat., 1969, 43, 233. McKone, C. E., Ibid., 1969, 44, 60. Yip, G., J . Ass. Of. Analyt. Chem., 1971, 54, 327. Katz, S. E., and Strusz, R. E., J . Agric. Fd Chem., 1969, 17, 1409. Abbott, D. C., Blake, K. W., Tarrant, K. R., and Thomson, J., J . Chromat., 1967, 30, 130. &ha, F., and Kljajic, R., Ibid., 1969, 40, 304. Skrinde, R. T., J . Wat. Pollut. Control Fed., 1970, 42, 863. Hance, R. J., J . Chromat., 1969, 44, 419. Frei, R. W. , Lawrence, J. F., and Belliveau, P. E. , 2. analyt. Chem. , 1971, 254, 271. Mallet, V., and Frei, R. W., J . Chromat., 1971, 54, 251. Belliveau, P. E., and Frei, R. W., Chromatographia, 1971, 4, 189. Belliveau, P. E., Mallet, V., and Frei, R. W., J . Chromat., 1970, 48, 478. Seiler, N., and Wiechmann, M., in Niederwieser, A., and Paraki, G., Editors, “Progress in Thin-layer Chromatography and Related Methods,” Volume 1, Ann Arbor Science Publishers, Ann Arbor, Mich., 1970, p. 94. Frei, R. W., and Lawrence, J. F., in “I.U.P.A.C. Pesticide Congress, Tel Aviv, Israel, February 1971, Congress Proceedings,” Gordon and Breach, London and New York, 1971, p. 193. Lawrence, J. F., and Frei, R. W., Int. J . Envir. Analyt. Chem., 1972, 1, 317. Frei, R. W., and Lawrence, J. F., J . Chromat., 1971, 61, 174. Lawrence, J. F., and Frei, R. W., Ibid., 1972, 66, 93. Williams, I. H., Residue Rev., 1971, 38, 1. , J , Ibid., 1972, 66, 295. , Ibid., 1972, 67, 87. -- -- Received January 3rd. 1972 Accepted July 24th, 1972