2422 J.C.S. DaltonVapour-density Determinations of Group 5 PentafluoridesBy John Fawcett, Alan J. Hewitt, John H. Holloway,' and Michael A. Stephen, Department of Chemistry,Vapour-density determinations on the saturated vapours of NbF5, TaF,, and SbF, a t temperatures above theirboiling points have been made by a modified Dumas method. The average molecular weights of the vapour-phasespecies near the boiling points are close to those for the respective trimers. Approaching 400 "C (or 300 "C forSbFs), however, the major constituent of the vapour is the monomeric pentafluoride.The University, Leicester LE1 7RHFOR VF,, PF,, and AsF, vibrational spectroscopicresults l-l0 indicate the presence of monomeric gaseousphases containing molecules of D3h symmetry, con-sistent with the interpretation of electron-diffractionstudies on VF5,l1 PF5,12 and AsF5.l3 By contrast,NbF,, TaF,, and SbF, are known to be a s s ~ c i a t e d .l l ~ ~ ~ - ~ ~Comparison of theoretical with experimental radialdistribution curves from electron-diffraction studies ofNbF, and TaF, has led to the conclusion that tetramerspredominate in the gas phase,21 as in the solids.22 Onthe other hand mass-spectrometric studies on the neutralspecies sampled from the saturated vapours of NbF, 23924and TaF, 23 have been interpreted to suggest an abundantconcentration of monomeric, dimeric, and trimeric ions,whereas ions with four or more metal atoms are scarce orabsent. In the SbF, case,23 the observation of pentamericions has been explained by the suggestion that thevapour of this compound is not composed predominantlyof tetrameric rings.Thus it is clear that, although theheavier Group 5 pentafluorides are largely polymerisedin the vapour, uncertainties about the quantitativecomposition of the vapours complicate the interpretationof vibrational and mass-spectrometric data.The aim of the present study was to determine theaverage molecular weight of molecules in the saturatedvapours of niobium, tantalum, and antimony penta-fluorides at a range of temperatures above their boilingpoints.EXPERIMENTALReagents.-Niobium pentafluoride and TaF, were pre-pared by the direct fluorination of the hydrogen-reducedmetal powders a t ca. 300 "C in dynamic systems. Purific-ation was achieved by slow sublimation under high vacuumand purity was monitored by determination of meltingR.G. Cave11 and H. C. Clark, Inorg. Chem., 1964, 3, 1789.2 H. H. Claassen and H. Selig, J . Chem. Phys., 1966, 44, 4039.3 J. E. Griffiths, R. D. Carter, jun., and R. R. Holmes, J .4 L. C . Hoskins and R. C . Lord, J . Chem. Phys., 1967,46,2402.5 R. M. Dieters and R. R. Holmes, J . Chem. Phys., 1968, 48,6 I. R. Beattie, K. M. S. Livingston, and D. J. Reynolds, J .7 H. Selig, J. H. Holloway, J . Tyson, and H. H. Claassen, J .L. C. Hoskins and C. N. Perng, J . Chem. Phys., 1971, 55,I. W. Levin, J . Mol. Spectroscofiy, 1970,33, 61.lo F. A. Miller and R. A. Capwell, Spectrochim. A d a , 1971,A27, 125.11 G. V. Romanov and V. P. Spiridonov, Zhur.strztkt. Khim.,1966, 7, 882; Izvest. sibirsk. Otdel. Akad. Nauk S.S.S.R., Ser.khim. Nauk, 1968, 126.1* K. W. Hansen and L. S. Bartell, Inovg. Chem., 1965, 4, 1775.l3 F. B. Clippard, jun., and L. S. Bartell, Inovg. Chem., 1970, 9.Chma. Phys., 1964, 41, 863.4996.Chem. Phys., 1969, 51, 4269.Chem. Phys., 1970, 53, 2559.5063.805.points 25 and from X-ray powder patterns.ee Antimonypentafluoride was prepared by direct fluorination ofantimony in a sloping reactor and was purified by repeatedsublimation under high vacuum until a highly viscousproduct was obtained with a conductivity of 6 x lo-* S tin-1.Vapour-density Detevuninati0n.s.-Samples of the penta-fluorides were sealed under vacuum in predried and pre-fluorinated, silica, modified Dumas bulbs of ca.35 Inm indiameter (Figure 1). The bulbs were scored with a glassFIGURE 1 Silica Dumas bulb (8 scale)knife a t points close to the top and close to the bottom ofthe stem. Bulbs were lowered into thermostatted baths(concentrated H,SO, was used at temperatures up to200 "C and 1.C.1.'~ ' Cassel ' TS 150 a t >200 "C) a t pre-determined temperatures above the boiling points of thepentafluorides. The tips of the bulbs were removed whenit was estimated that the pressure in the bulb had reachedatmospheric. (Successful opening was signified by asmooth continuous emission of vapour. If the bulb wasl4 L. E. Alexander, Inoirg. Nuclear Chem. Letters, 1971, '7, 1053.15 M. J. Vasile, G. R. Jones, and W. E. Falconer, Chm.Comm.,16 L. E. Alexander, I. R. Beattie, and P. J. Jones, J.C.S. Dalton,l7 B. Philips and Ivf. H. Rand, quoted in ref. 16.N. Acquista and S. Abramowitz, J . Chem. Phys., 1972, 56,1s I. R. Beattie, K. M. S. Livingston, G. A. Ozin, and D. J.2O L. E. Alexander and I. R. Beattie, J . Chem. Phys., 1972, 56,21 G. V. Romanov and V. P. Spiridonov, Izvest. sibirsk Otdel.z2 A. J. Edwards, J . Chem. SOC., 1964, 3714.Z3 W. E. Falconer, G. R. Jones, W. A. Sunder, M. J. Vasile,A. A. Muenter, T. R. Dyke, and W. Klemperer, J , Fluorine Chevt.,1974, 4, 213.z4 I. S. Gotkis, A. V. Gasarov, and L. N. Gorokhov. Russ. J .Inorg. Chem., 1975, 20, 702.25 E. L. Muetterties and C . W. Tullock, in ' Preparative In-organic Reactions,' vol. 2, ed. W. L. Jolly, Interscience, NewYork, 1965, p.237; N. Bartlett, ibid., p. 301.1971, 1355.1972, 210.522 1.Reynolds, J . Chem. SOC. ( A ) , 1969,958.5329.Akad. Nauk S.S.S.R., Ser. khim. Nauk, 1968, 1, 1051976 2423opened too early no pentafluoride vapour emerged artd iftoo late the emission was pulsatory.) The bulb was main-tained at the constant bath temperature until emission hadceased and was then sealed rapidlv above the second score-mark. The bulb was cooled, washed, and allowed toequilibrate with the surroundings before it was weighed(see Table). The bulb was broken under degassed distilledMeasurements and calcuhtion of the degree of associationRoom temperature (r.t. in K)Temperature of thermostat bath (K)Atmospheric pressure (mmHg)Weight of sealed bulb + vapour (g)Weight of bulb + water + residual air (g)Weight of bulb full of water (g)1Veight of bulb empty (g)= h f - - Rdensity of water at r.t.Bulb capacity (cm3) =Vapour capacity (cm3) = e - gdensitv of water at r.t.:Apparent weight of vapour (g) .= ji x 1.296* x 273 x c ~= 103* x u x 760- Buoyancy of sealed bulb (g) =:. True weight of vapour (g) = (d - g) + k = IThe weight of i cm3 of hydrogen at b K and c mmHg =-= 1 I Li x 273 x c x 0.09 -____-____103 x b x 7601I I L:= - weight of vapourweight of equal volume of hydrogen :. Vapour density = -0 , i s Ln3t :. Molecular weight of vapour = - a t b I<:_ Degree of association =___ .__- 2(llm) at b K* Densities of air and hydrogen were taken from ' Handbookof Chemistry and Physics,' 53rd edn., ed.R. C. Mreast, ChemicalRubber Company, Cleveland, 1972.molecular weight of MF, monomerwater at room temperature and the bulb, water, and anyresidual air-lock were weighed. The bulb was filled withwater from a syringe, reweighed, and then, finally, weighedempty. The measured values were tabulated and thedegree of association calculated by a traditional Dumasmethod, making corrections for buoyancy and for in-complete filling of the bulbs with vapour as outlined in theTable.RESULTS AND DISCUSSIOSThe results of measurements of the degree of associ-ation against temperature are represented in Figures2 4 . For each compound measurements were madefrom just above the boiling point to ca. 300 "C for SbF,and 400 "C for NbF, and TaF,.The results for NbF,and TaF, are very similar. The average degree ofassociation varied from cn. 3, close to the boiling point,to ca. 1 approaching 400 "C, the average number of MF,units associated at a given temperature being slightlyhigher in the tantalum fluoride case. Values obtainedby Beattie and his co-workers l6 for NbF, are alsoshown in Figure 2. The degrees of association theyobserved at a given temperature are generally lower thanour values. This is almost certainly due to the fact thattheir experiments were made with unsaturated vapours.The overall pictures are closely similar and clearlyestablish that at temperatures close to 400 "C monomericNbF, species are the most abundant, and at lower andlower tempera.tures there is increasing polymerisationuntil, at the boiling point, the average molecular weightis close to that of Nb3FI5.This rather high averagevalue suggests a virtual absence of monomer and thepossibility of a significant proportion of tetramer orlarger oligomer in the saturated vapour phase near theboiling point. The interpretation of the mass-spectro-metric results of Gotkis et aZ.24 is in agreement with this,aOOa go400/a go0% 3001 I '1. 2.0 3.0 2000 1.0Degree of associationFIGURE 2 Plot of temperature against degree of association forNbF,(O) ; (a), values obtained by Beattie and his co-workers l6008O O00003001 Y.2001 L00 00 fP1000 1.0 2.0 3.0Degree of associationFIGURE 4 Plot of temperature against degree of associationfor SbF,but is counter to the suggestion by Falconer et aZ.23that species with four or more metal atoms are scarceor absent under these conditions.This underlines thedifficulty in interpreting mass-spectrometric resultswhere electron-beam fragmentation of species and ion-molecule reactions occur. For TaF, higher tem-peratures are required to produce monomers than in theNbF, case, but again the mean value of the degree ofassociation at the boiling point is 3 3 (Figure 3)J.C.S. DaltonThe results for SbF, (Figure 4) are totally in accordwith the earlier, but less detailed, experiments of Huband RobinsonJZ6 who also used a Dumas method. Ourvalues are closely related to those of niobium andtantalum pentduorides ; the average degree of associ-ation at the boiling point is 3 whilst, at 300 "C, the mainz6 D. R. Hub and P. L. Robinson, J . Chem. Soc., 1954,2640.constituent is the monomer. This close correlationsuggests that the nature of the vapour of SbF, is notgreatly different from those of the other two compoundsand implies that, in all cases, a considerable proportionof tetrarners (or possibly larger oligomers) occurs in thevapours at temperatures close to the boiling point.[6/902 Received, 11th May, 1976