Analyst, September, 1973, Vol. 98, $9. 647-654 647 The Determination of Volatile Metal Chelates Using a Microwave-excited Emissive Detector BY R. M. DAGNALL, T. S. WEST AND P. WHITEHEAD (Biochemistry Division, Huntingdon Research Centre, Huntingdon, PE18 6ES) (Chemistry Department, Imperial College of Science and Technology, London, S W7 2A Y ) A microwave-excited emissive detector operated at atmospheric pressure has been used in conjunction with gas chromatography in order to achieve the separation and determination of various metal chelates of acetylacetone and trifluoroacetylacetone. The operating parameters have been optimised and the limits of detection and degrees of selectivity evaluated. The proposed method is both selective and very sensitive and can also confirm the presence of a particular metal.THE great advantages that gas chromatography possesses over many other separation techniques have raised considerable interest in its application to inorganic analysis. In particular, the gas chromatography of metal chelates has been extensively studied and has shown much greater potential than that of other metal derivatives. The principal chelates investigated have been those of acetylacetone and its fluorocarbon analogues, e g . , trifluoroacetylacetone and hexafluoroacetylacetone. All of these acetylacetone derivatives form stable chelates with a wide range of metals; Moshier and Sieversl have reviewed progress to 1965, although more recently the use of other ligands has been investi- gated. Fluoro-derivatives of dipivalylmet hane [ (CH ?) 3C .CO.CH,. CO. C (CH J 3] have been used in order to form fairly volatile chelates with alkali metals2 The chelates of monothio- p-diketones with a wide range of bivalent metals, including nickel, cobalt, zinc, palladium and platinum, are stable and volatile and have been chromatographed successfully.3-5 Amino-substituted /I-diketones also show potential as chelating agentsg In order to provide an acceptable method for the analysis of mixtures of metals, the gas- chromatographic determination of metal chelates requires good resolution of components, minimum interferences and, in particular, sensitivities at least as high as those of other methods in common use, e.g., atomic-absorption and emission spectroscopy. Hence the choice of detector is of major importance. The katharometer is non-selective and provides sensitivities in the range to 10-8 g.7~8 In addition, some decomposition of chelates occurs on the hot metal frlaments.l Flame- ionisation detectors have also been widely used but their response to metal chelates is less sensitive than that to organic compounds containing no metals.The presence of fluorine atoms and the metal atom itself in the chelate decreases the response of the detector and only .moderate sensitivities have been obtained.1° The electron-capture detector has been widely used for the determination of fluorinated chelates7,I1 and is highly sensitive.1°J2 The electron affinity of the chelates is a function of both the metal and any halogen atoms present ; however, the detector is much more sensitive to halogenated chelates.Under comparable conditions, the limit of detection for chromium( 111) acetylacetonate is 2.5 x 10-lo mol, for the trifluoroacetylacetonate 1.8 x mol and for the hexafluoroacetylacetonate 4.9 x mol.ll Often, the choice of chromatographic conditions is limited by the volatility and stability of the chelates used and hence complete resolution of mixtures is not always possible. Selective detection of individual metals would, therefore, be of considerable interest, but very little work has been reported on the application of selective detectors. Flame-photometric detection of both metal chelates and halides has been studied by Juvet and Durbin,l39l4 who viewed the emissions from an oxy-hydrogen flame by using a Beckrnann DU spectrophotometer. Although the sensitivity of the detector response was not high, good selectivity and a range @ SAC and the authors.648 DAGNALL, WEST AND WHITEHEAD : DETERMINATION OF VOLATILE [ A ?Za&!Lf, VOl.98 of linearity of lo4 were obtained. A similar system was used by Juvet and &do15 but with less sensitive results. It is only very recently that mass spectrometry has been applied in the determination of chromium and beryllium chelates a t the 10-l2,g level.16 Microwave-excited emissive detectors have proved highly sensitive and fairly selective in organic analysis,l7~l* but no application of microwave-excited emission to the determination of metal chelates has been reported. The potential usefulness of this type of emission for the determination of metals has been shown by the work of Bache and Lisk,lg who used emission a t the 253.7-nm atomic mercury line from a low-pressure helium plasma in order to measure the amounts of organomercury compounds present in fish, and by the work of Runnels and Gibson20 and Dagnall, Sharp and West.21 These two groups of workers used microwave plasmas in order to excite metal emissions from compounds that had been flash evaporated into the plasma from a platinum loop.This technique is highly sensitive, but is limited by the small volume (about 11.1) of solution that can be retained on the loop and to those compounds which can be volatilised under these conditions. Therefore, the response of the previously described22 microwave-excited emissive detector operated a t atmospheric pressure to some metal chelates was investigated.The chelating agents chosen were trifluoroacetylacetone and acetylacetone because they had been exten- sively used by other workers and comparison with other detectors would be possible. Pure metal chelates were bought or prepared, characterised, and their emission spectra in a micro- wave-excited plasma recorded. Solutions of the chelates in benzene were chromatographed and the major emission lines monitored. Samples of the eluates were collected and compared with the original compounds. EXPERIMENTAL PREPARATION OF METAL CHELATES- The acetylacetonates of aluminium(III), chromium(III), copper(I1) and iron(II1) were obtained commercially (Koch-Light Laboratories, Ltd., Colnbrook, Buckinghamshire) .Acetylacetone and trifluoroacetylacetone were also obtained from the same supplier and used to prepare the acetylacetonate* of scandium(II1) and the trifluoroacetylacetonates* of aluminium(III), chromium(III), copper(II), gallium(III), iron(III), scandium(II1) and vanadium( IV) . Trifluoroacetylacetone was found to deteriorate rapidly on exposure to air 'and was distilled immediately before use, the fraction boiling between 105 and 107 "C being collected. All the chelates, with the exception of aluminium(II1) and chromium(II1) trifluoro- acetylacetonates, were prepared by the method described by Berg and Tr~emper.~, In each instance, the pH of a suitable aqueous solution of the metal ion was adjusted to about 9.0 with 5 per cent. sodium acetate solution and an ethanolic solution of the chelating agent added.A precipitate formed on stirring, which was filtered, washed with water, air dried and either sublimed under vacuum or recrystallised from benzene or hexane. The preparation of Cr(tfa), was similar except that the pH was adjusted with dilute ammonia solution. Al(tfa), was prepared by the reaction of aluminium chloride in dry carbon tetrachloride with trifluoroacetylacetone. All the chelates used were characterised by means of their melting-points, and ultraviolet and infrared spectra. The melting-points were determined on a Kofler block, and agreement with values in the l i t e r a t u r e l ~ ~ ~ - ~ ~ was satisfactory. The infrared spectra of mulls of the chelates in Nujol were measured in the range 650 to 4000 cm-l by using a Perkin-Elmer 157 sodium chloride spectrophotometer. All the chelates of each ligand gave very similar spectra.The bands for the acetylacetonates were assigned by reference to the results of Nakamoto,28 whereas those for the trifluoroacetylacetonates were assigned from the results of Nakamoto, Morimoto and Marte11,29 and Morris, Moshier and S i e ~ e r s . ~ ~ The ultraviolet spectra of solutions of the complexes in ethanol, obtained by using a Perkin-Elmer 402 spectrophoto- meter, were relatively simple and contained, in most instances, one broad absorption band in the region 210 to 380 nm. The present results were very similar to those reported in the l i t e r a t ~ r e . ~ ~ ~ ~ l - ~ ~ * The abbreviation (acac) and (tfa) are used in the formulae of metal acetylacetonates and trifluoro- acetylacetonates, respectively.September, 19731 METAL CHELATES BY A MICROWAVE-EXCITED EMISSIVE DETECTOR 649 DETERMINATION OF THE EMISSION SPECTRA OF METAL CHELATES IN A MICROWAVE-EXCITED The emission spectrum of each metal chelate in the plasma was examined in order to determine the major emission lines that had the minimum of interferences from background, carbon and other metal emissions.An experimental system was set up so as to obtain an approximately constant, continuous level of vapour of the metal chelate in the argon entering the plasma, which enabled the entire spectrum to be scanned and more detailed studies to be made of the regions of interest. No attempt was made to obtain precise quantitative data on relative peak heights, as the concentration of the vapour of the chelates could not be maintained sufficiently constant.The microwave system was similar to that used in previous studies22 and consisted of a &wave Evenson-type microwave cavity, supplied from a microwave generator, and 1-mm i.d. silica tubing. When power was supplied to the cavity, a plasma could be maintained in the argon carrier gas flowing through the silica tube. A microwave power of 70 W was used throughout. The spectral emissions from the plasma were recorded by using a Beckmann DU monochromator and a d.c. read-out system. PLASMA- The apparatus used is shown in Fig. 1. PS = Power supply R = Read-out PM = Photomultiplier NI = Monochromator B = Sample boat HC = Heated column C = Cavity Q = Quartz tube S = Microwave power supply F = Furnace T = T-piece G = Pressure gauge P = Pressure controller Fig.1. Apparatus for the determination of the spectra of metal chelates It was found necessary to monitor each chelate as close as possible to the bottom of the plasma and to heat the connecting tubing. In the arrangement used, about 100 mg of the chelate under investigation were placed in a silica boat mounted in a brass T-piece just below the plasma. The silica tubing, which contained the plasma, passed into the middle of the T-piece just above the boat so that the vapour of the metal chelate could not easily come into contact with any hot metal surfaces. The carrier gas was pre-heated by passage through an empty steel column about 1 m in length and heated by means of heating tape to about 200 "C.A capillary restriction enabled the flow-rate of gas to be maintained at 3 1 h-1. The sample boat and silica tubing beneath the plasma were heated by means of a 12042 furnace, consisting of Nichrome wire round a 17 mm diameter silica tube, 90 mm long. The wire was covered with alumina cement. The furnace and the heating tape were controlled by using Variac variable transformers. For each set of spectral measurements, the chelate in the sample boat was introduced into the T-piece, the plasma initiated, the coil of tubing heated and finally the temperature650 DAGNALL, WEST AND WHITEHEAD : DETERMINATION OF VOLATILE [Analyst, Vol. 98 of the furnace raised slowly until carbon and metal emissions were observed. The variation of emission intensity with position in the plasma was measured as follows.The vertical height of the viewing slit of the monochromator was reduced to about 1 mm by a horizontal slit. Vapour of the sample was passed continuously into the plasma. The cavity, and hence the plasma, was moved vertically by means of a rack and pinion. With the plasma in the optimum position, the spectrum from 200 to 600 nm was scanned and when measurements for each chelate were completed, the silica tubing was replaced. Position of metal emissions in the plasma-The variation of the intensity of metal emissions with position in the plasma is shown in Fig. 2 for Al(acac), (396.2 nm) and Cu(acac), (327.4 nm). For comparison, the variation with position of the atomic carbon emission (247.9nm) for Al(acac), is included.In contrast to the carbon emission, metal emissions could be observed only near the beginning of the plasma. In addition, after the passage of relatively large amounts of a chelate into the plasma, a deposit with a metallic appearance also formed near the beginning of the plasma. It was concluded that the absence of metal emis- sions higher in the plasma was due to the metal atoms being removed from the gas phase on collision with the walls of the silica tubing. 0 20 40 Distance from bottom of plasma/mm Fig. 2. Variation of emission intensity with position in plasma. Graph A(+), 327.4 nm [Cu(acac),]; graph B ( x ), 396.2 nm [Al(acac),] ; and graph C( O), 247.9 nm [Al(acac)J The concentration of metal emissions in the first 10 mm of the plasma had two disadvan- tages.The extreme ends were the most unstable parts of a plasma and hence the variation in background emission was also greatest there. The metal deposit on the silica tubing gradually increased the background level of metal emission observed and after about milligram amounts of metal chelate had passed into the plasma the deposit significantly reduced the over-all intensity of light that reached the monochromator. Both of these effects were minimised by using small samples. Emission spectra of metal cdzelates-The major emissions observed for each met a1 species are shown in Table I. The recorded spectra also contained many less intense emissions that were considered to be less suitable for monitoring the passage of a gas-chromatographic eluate.However, they would be useful for providing a characteristic spectrum in order to confirm qualitatively the presence of a particular metal. In all instances when two compounds of a metal were studied, both of their spectra were found to be essentially identical. In order to confirm their identification, the emissions observed were compared with those obtained from electrodeless discharge lamps of each metal under the same experimental conditions and with those listed in the National Bureau of Standards Monograph No. 3236 obtained by using a 200-V d.c. arc. Owing to the relatively poor resolution of the mono- chromator used in the present study, many of the atomic lines listed could not be resolved.September, 19731 METAL CHELATES BY A MICROWAVE-EXCITED EMISSIVE DETECTOR 651 The major individual atomic lines are shown in parentheses after each composite emission peak observed.The spectra observed by using this experimental arrangement showed that suitable metal emissions could be obtained from a microwave-excited plasma for all the chelates used. Therefore, a gas-chromatographic system was set up and the major emissions were used to monitor the eluted chelates. TABLE I EMISSION CHARACTERISTICS OF METAL CHELATES Metal species Principal emissions/nm A1 396-2 Cr 357.9 425.4 520.5 (520-8, 520.6, 520-4) cu 324.7 327.4 Ga 287.4 294.4 (294.3, 294.4) 417.2 403.3 Metal species Principal emissions/nm Fe 344.1 357.0 373.5 s c 361.4 363-1 364.4 (364.5, 364.3) 357.5 (357-3, 357.6, 358.0) 424-7 vo 318.4 (318.3, 318.4, 318.5) 292.4 (292.4, 292-5) 438.1 249.0 (248.8, 249.0, 249.1) GAS~CHROMATOGRAPHY OF METAL CHELATES- The detector system was the same as that used for the above determination of spectra.The sample heater was replaced by a Pye Series 104 gas-chromatographic oven and column. Two columns were used- . . 0.6 m x 4.8 mm i.d. borosilicate glass packed with Universal B 0.6 m x 4.8 mm i.d. borosilicate glass packed with 0-5 per cent. Both columns were conditioned by heating them at 200 "C for 36 hours in a stream of argon. Poor peak shapes were obtained for Al(acac) 3r Cu(tfa), and VO(tfa), on column I and column I1 was therefore used in preference in these instances. In order to avoid decomposition or condensation of the metal chelates, the silica tubing containing the plasmas was connected directly to the column exit and the monochromator was positioned so as to view the plasma just above the roof of the oven.In order to reduce further the risk of condensation, the silica tubing between the column and the plasma was heated by means of a short coil of wire and a Variac variable transformer. The samples were injected directly on to a glass-wool pad in the top of the column, before the packing, and the top of the column was heated by means of the small electrical furnace described previously. The temperature of the top of the column could be maintained at about 30 "C above the oven temperature by applying a potential of 40 V across the furnace. Operating +arameters-The experimental parameters of the detector system were investi- gated by making 1-p1 injections of a 950 p.p.m.(m/V) solution of Ga(tfa), in benzene. Ga(tfa), was used because it was found to be conveniently eluted from column I at 125 "C after 3.1 minutes with a fairly good peak shape and the intense gallium atomic line at 294-4 nm lies in a region of very low background. Any variation of detector response at this wavelength could, therefore, be associated directly with changes in the metal emission. As expected from the studies with the continuous system, emissions characteristic of the metallic elements were obtained only from the first 10 mm of the plasma. In practice, in order to obtain a linear calibration graph that passes through the origin, it was found necessary to position the plasma so that emissions from the first 3 mm of the plasma could be monitored, and this position was used for all measurements of chelate emissions.The variation of the peak height at 294-4 nm with microwave power was measured by repeated injections of the Ga(tfa), solution. For each new power setting, the length of the plasma altered and the position of the plasma had to be re-optimised. The maximum peak height was obtained at a power of 70 W. The variation of noise with power was found to be very small and the maximum signal to noise ratio also occurred at 70 W. Column I .. Column I1 . . pre-coated with 10 per cent. Apiezon L. Apiezon L on glass micro-beads (0.2 mm diameter). . .652 DAGNALL, WEST AND WHITEHEAD : DETERMINATION OF VOLATILE [Analyst, Vol. 98 The variation of detector response with flow-rate of the carrier gas was also measured by repeated injections. The area response was found to be proportional to the reciprocal of the flow-rate, and the slowest convenient flow-rate (3.3 1 h-l) was used.In summary, the optimum experimental parameters were found to be the same as those for organic compounds.22 A microwave power of 70 W and a flow-rate of carrier gas of 3.3 1 h-l were used throughout the metal chelate determinations. Sensitivity and selectivity of emissions-The signal to noise ratios for the major emissions characteristic of the metallic elements observed previously were determined for solutions of each of the trifluoroacetylacetonates. In addition, the detector response at each wavelength to a 1000 p.p.m. solution of benzyl alcohol in benzene was measured in order to determine the selectivities of the metal emissions over carbon compounds.Benzyl alcohol was chosen because it has convenient retention times (1.0 to 3.5 minutes) on the columns used. The molar selectivity ratio was calculated in each instance as the ratio of the number of gram- atoms of carbon injected to the number of gram-atoms of the metal required to produce the same response. Benzene was used as the solvent for all the chelates. It was eluted after only 10 to 20 s and the plasma was not initiated until about 30 s after injection. The optimum experimental conditions used are shown in Table 11. All the trifluoro- acetylacetonates were chromatographed successfully, as were the acetylacetonates of alumin- ium, chromium and scandium. However, no peaks could be obtained for the copper and iron acetylacetonates. The most sensitive wavelength found for each metal, with the limit of detection and selectivity, are shown in Table 111.TABLE I1 OPTIMISED OPERATING PARAMETERS Chelate Column Al(acac), . . .. .. I1 Al(tfa), . . .. .. I Cr(acac), . . .. . . I Cr(tfa), . . .. .. I Cu(tfa), . . .. . . I1 Ga(tfa), , . .. . . I Fe(tfa), . . .. .. I Sc(acac), . . .. .. I Sc(tfa), . . .. .. I VO(tfa), . . .. .. I1 Temperaturel'C 175 100 190 130 140 125 135 135 160 160 Retention timelminutes 1.67 4.83 4-80 3.30 2.00 3.10 2.90 2.90 4-00 1.45 Interferences-Interference with the response of the detector to a metal chelate may be (1) spectral emissions from other metals or carbonaceous species at the wavelength being monitored; (2) suppression or enhancement of the detector response by another metal chelate that is eluted simultaneously; and (3) distortion of the plasma by a large excess of another compound.Type (3) is common to all compounds and methods of overcoming this problem, particu- larly in regard to the solvent, have been discussed previously.22 Types (1) and (2) will be considered separately. of three types- TABLE I11 LIMITS OF DETECTION AND SELECTIVITIES FOR SOME METAL CHELATES Chelate A1 (acac) , .. .. Al(tfa), . . .. .. Cr (acac), .. .. Cr(tfa), . . .. .. Cu(tfa), . . .. .. Ga(tfa), . . . . .. Fe(tfa), . . .. .. Sc(acac), .. .. Sc(tfa), . . .. .. VO(tfa), . . .. . . Wavelengthlnm Limit of detectionlg s-1 of metal 396.2 396.2 357.9 357.9 324.7 294.4 344.1 361-4 361.4 318.4 2.0 x 10-11 1.9 x 10-11 2-9 x 10-12 3.6 x 10-la 8.0 x 10-12 2.7 x 1.3 x 10-11 2.1 x 10-12 3.0 x 10-la 8.5 x 10-12 Selectivity 990 3930 2250 1170 1610 1620 1400 - - -September, 19731 METAL CHELATES BY A MICROWAVE-EXCITED EMISSIVE DETECTOR 653 The effect of emissions from benzyl alcohol, as shown by the selectivities in Table 111, was found to be, in general, very small.Possible spectral overlap of atomic metal lines was investigated by repeated injections of 1 pl of about 2 per cent. solutions of each chelate in benzene with the monochromator set in turn to each of wavelengtlis listed in Table 111. These concentrations are approximately the largest that could be injected without distortion of the plasma. Apart from the mutual interference of Cr(tfa), and Sc(tfa),, the only inter- ference observed was that of VO(tfa), on Cu(tfa),.Further investigations into spectral interferences were not carried out as the selectivity could easily be increased by reducing the slit width. If a particular interference presented a problem, alternative emissions (Table I) could often be used. The second type of interference was studied by preparing mixtures of Cr(tfa), and Ga(tfa) , in benzene and selecting gas-chromatographic conditions such that the peaks partially overlapped. Even in the presence of a 1000-fold excess of Ga(tfa),, the peak height of Cr(tfa), did not vary by more than 2 per cent. An example of the overlap of the peaks is shown in Fig. 3. Fig.. 3. Chromatorrrams of a mixturg of 985 p.p.m." of Ga(tfa), and 10 p.p.m.of Cr(tfa), in benzene. ( a ) , 357.8 nm (Cr); ( b ) , 294-4 nm (Ga); and ( G ) , 247-9 nm (C). R = re-ignition of plasma and I = injec- tion Characterisatiofi of the chromatographic ehates-The eluates were collected and charac- terised in order to confirm their identities and, in addition, to search for evidence of partial decomposition, which might be shown by the presence of an impure condensate. In order to collect a sufficient amount of each compound, 50p1 of about 1 per cent. solutions of each chelate in benzene were injected under the conditions used previously. The plasma was not initiated and the silica tubing not heated until the solvent was completely eluted. The tubing was cooled and the condensate collected. The melting-point and infrared and ultraviolet spectra of the condensate were measured in the same way as for the chelates.In all instances, the results were in close agreement with those obtained previously. CONCLUSIONS The microwave-excited emissive detector operated at atmospheric pressure responded to all the chelates used, both non-selectively by monitoring atomic carbon and selectively by using metal emissions. The detector response was linear and highly sensitive to the metals studied with limits of detection between 2 x 10-12 and 2 x 10-11 g s-1 (Table 111). The limits of detection of the acetylacetonates and trifluoroacetylacetonates of the same metals were approximately equal, indicating that the detector response may well prove to be independent of the chelating agent used.654 DAGNALL, WEST AND WHITEHEAD The sensitivity of this detector is considerably higher than those of other selective detec- tors used in the analysis of metal chelates, with the exception of mass spectrometry, and is of the same order as that of the electron-capture detector. Further, the dependence of the detector response on the metal, rather than on the chelate as a whole,‘ avoids the limitations of having a halogen atom or other electron-capturing species present in order to achieve high sensitivity. It is highly selective and, further, has the advantage that the pattern of atomic metal lines can be rapidly scanned in order to confirm the presence of a compound of a particular metal.The limitations of this detector for the analysis of mixtures of metals are similar to those for organic analysis.The principal problem is that of overloading, which imposes an upper limit on the working range of the detector. In addition, the formation of metallic deposits requires the tubing to be changed frequently when concentrated samples are used. 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. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. REFERENCES Moshier, R. W., and Sievers, R. E., “Gas Chromatography of Metal Chelates,” Pergamon Press, Belcher, R., Dudeney, A. W. L., and Stephen, W. I., J . Inorg. Nucl. Chem., 1969, 31, 625. Stephen, W. I., Thompson, I. J., and Uden, P. C., Chem. Communs, 1969, 269. Belcher, R., Stephen, W. I., Thompson, I. J., and Uden, P. C., J .Inorg. Nucl. Chem., 1971,33, 1851. Stephen, W. I., Proc. SOC. Analyt. Chem., 1972, 9, 137. Sievers, R. E., Ponder, B. W., Morris, M. L., and Moshier, R. W., Inorg. Chem., 1963, 2, 693. Sievers, R. E., Chem. Engng News, 1963, 41, 41. Hill, R. D., and Gesser, H., J . Gas Chromat., 1963, 1, 11. Albert, D. K., Analyt. Chem., 1964, 36, 2034. Ross, W. D., Ibid., 1963, 35, 1596. Barratt, R. S., Proc. SOC. Analyt. Chem., 1972, 9, 86. Juvet, R. S., and Durbin, R. P., J . Gas Chromat., 1963, 1, 14. -- , Analyt. Chem., 1966, 38, 565. Juve’t, R. S., and Zado, F. M., Ibid., 1966, 38, 569. Wolf, W. R., Taylor, M. L., Hughes, B. M., Tiernan, T. O., and Sievers, R. E., Ibid., 1972, 44, 616. McCormack, A. J., Tong, S. C., and Cooke, W. D., Ibid., 1965, 37, 1470. Bache, C. A., and Lisk, D. J., Ibid., 1966, 38, 1757. -- , Ibid., 1971, 43, 951. Runiels, J . K., and Gibson, J. H., Ibid., 1967, 39, 1398, Dagnall, R. M., Sharp, B. L., and West, T. S., Nature, Phys. Sci., 1972, 235, 65. Dagnall, R. M., West, T. S., and Whitehead, P., Analytica Chim. Acta, 1972, 60, 25. Berg, E. W., and Truemper, J. T., J . Phys. Chem., 1960, 64, 487. Fay, R. C., and Piper, T. S., J . Amer. Chem. Soc., 1963, 85, 500. Henne, A. L., Newman, M. S., Quill, L. L., and Staniforth, R. A., Ibid., 1947, 69, 1819. Reid, J. C., and Calvin, M., Ibid., 1950, 72, 2948. Stanforth, R. A., Dissertation, Ohio State University, 1943. Nakamoto, K., “Infrared Spectra of Inorganic and Coordination Compounds,” Second Edition Nakamoto, K., Morimoto, Y., and Martell, A. E., J . Phys. Chem., 1962, 66, 346. Morris, M. L., Moshier. R. W., and Sievers, R. E., Inorg. Chem., 1963, 2, 411. Charles, R. G., J . Inorg. Nucl. Chem., 1958, 6, 42. Holm, R. H., and Cotton, F. A., J . Amer. Chem. SOC., 1958, 80, 5658. Bamun, D. W., J . Inorg. Nucl. Chem., 1961, 21, 221. -, Ibid., 1962, 22, 283. Hazeldine, R. N., Musgrave, W. K. R., Smith, F., and Turton, L. M., J . Chem. SOC., 1951, 609. Meggers, W. F., Corliss, C. H., and Scribner, B. F., “Tables of Spectral Line Intensities,” Natn. Received January 24th, 1973 Accepted A p r i l loth, 1973 Oxford, 1965. 8 J J , Chem. Communs, 1970, 1019. ---__. Wiley-Interscience, New York, 1970. BUY. Stand. Monog., No. 32, Part I, 1961.