The Journal of
Physical Chemistry
0 Copyright, 1982, by the American Chemical Society
VOLUME 86, NUMBER 3
FEBRUARY 4,1982
LETTERS Isolated MH Stretching Frequencies, Anharmonicities, and Dissociation Energies in Silanes, Germanes, Phosphines, and Arsines D. C. McKean,. I. Torto, and A. R. Morrisson Chemistry Department, Universtty of Aberdeen, Aberdeen AB9 2UE, Scotlend (Received: October 1, 198 1; In Final Form: December 1. 1981)
The fundamental and first overtone MH stretching bands have been recorded on a Nicolet 7199 FT IR spectrometer from the following species: MHD,, MeMHD,, Me2MHD,Me3MH (M = Si, Ge); MHD,, MeMHD, MezMH (M = P, As);SiHF3,SiHCl,, Si2HD5;GeI-IDa (X =-F, Cl, Br, I). The fundamental vss band of PhSiHD2 was also observed. Treating the MH bond as a diatomic molecule, values of oe,xe, Dlin,and Doo are deduced. The spectroscopicdata clearly indicate that methyl substitution weakens the MH bond, in conflict with evidence from recent iodination kinetics studies. Trends in MH bond anharmonicity are observed and discussed.
Introduction Infrared studies of “isolated” CH stretching frequencies, vcHie,measured in partially deuterated compounds, show that u c ~ mis both a good quantitative measure of what may be termed the equilibrium strength of the CH bond, as measured by the stretching force constant, and also correlates well with experimental values of bond dissociation energies Dom, for hydrocarbons and fluorocarbons, except where stabilization (“relaxation”) energy is present to a significant extent in the free radical formed.’v2 The present study extends these earlier ones to MH bonds, where M = Si, Ge, P, and As. The exptl. data on DoB8available for comparison in these molecules is either lacking, as in the case of As compounds, scanty (P, Ge), or controversial (Si). For the latter, data exist for eight compounds, derived from iodination kinetics studies, which (1) D. C. McKean, Chem. SOC.Rev., 7,399 (1978). (2)Recent reviaions of a number of Dom valuesS have markedly improved this correlation. 0022-3654/82/2086-0307$01.25/0
show an invariance of D(SiH) to methyl ~ubstitution.~ This is surprising in view of the well-known weakening effect of methyl on CH bonds.’ Similar work on GeH45 and Me3GeHealso indicates equal Do298values for these compounds. Other kinetic studies of methylsilanes’ and -germanes,* however, indicate a progressive lowering of D(M-H) with a-methyl substitution, in the case of SiH bonds, about 2-3 kcal mol-l per methyl group. Since for three of the M elements concerned, the data are lacking for a v ~ vs.H Doms ~ MH ~ correlation, we adopt here an alternative route to estimates of dissociation energy, which utilizes in addition the first overtone of the isolated MH (3)R. R. Baldwin, R. W. Walker, and R. W. Walker, J. Chem. Soc., Faraday Trans. 1, 76,825 (1980). (4)R. Walsh, Acc. Chem. Res. 14,246 (1981). (5)GeH,: R. Walsh, private communication. (6)MesGeH: A. M. Doncaster and R. Walsh, J. Phys. Chem., 83,578, (1979). (7)E.R.Austin and F. W. Lampe, J. Phys. Chem., 81,1134 (1977). (8)E.R.Austin and F. W. Lampe, J. Phys. Chem., 81,1546 (1977).
0 1982 American Chemical Society
308
The Journal of Physical Chemistry, Vol. 86, No. 3, 1982
stretching motion. If this can be legitimately treated as for a diatomic molecule, values of uMHiS and 2VMHis yield the parameters we and wexe, from which dissociation energies may be calculated. Where we and w g e are obtained from observations of levels near the potential minimum, the resulting Doo value, known as Dying,is well-known to exceed the true value of Doo. However, the extent of the deviation in diatomic molecules has been explored throughout the periodic tablelo and for SiH and ClH, Doo = O.87Dli,. We shall assume this to hold also for polyatomic molecules containing SiH and PH bonds, and use for GeH and ASH bonds the single experimental value of 0.81 established for HBr.'O Since the value of Dh is very sensitive to the measured values of vMHis and 2YMHiS, it is vital (a) that the latter be measured w t h the highest possible precision (f0.l cm in vMHiS corresponds to f0.2 kcal mol-l in Dyin),(b) that the diatomic approximation for the MH stretching vibration, for example, in MHD3, be very good. Sample calculations show in fact that the influence of MD stretching on uMH' affects its value by less than 0.1 cm-l.
Experimental Section The VMH and 2vMH IR bands were recorded by use of a Nicolet 7199 FT IR spectrometer with a resolution normally of 0.25 cm-', but occasionally of 0.12 or 0.06 cm-', the frequency accuracy attainable being fO.O1 cm-'. Where the K structure of the perpendicular component of hybrid bands, or the J structure of parallel bands of symmetric tops, could be resolved for both fundamental and overtone, appropriate analyses were made to derive band centers. Elsewhere, only the frequencies of central maxima could be employed. In two cases, those of MePHD and MeAsHD, we, we%,, and Dlinvalues derived by either method agreed excellently. The following partially deuterated species were observed as impurities in the fully deuterated materials: MezSiHD, SiHD#iD3, PhSiHDz, Me2GeHD,PHD2,MePHD, A s H D 2 , MeAsHD. Standard methods of synthesis were used, advantage being taken of the deuterium exchange between (CD3)zS0and GeH, PH, and ASH bonds." Results The values of YMH~' and 2VMHis obtained are listed in Table I, together with those of we, 2xe,Db,and Dooderived from them, as well as the experimental DoB8 value where this is known. The data for three of the bands concerned are very uncertain. v1 of GeHD3is a highly perturbed band not susceptible to analysis and the 2usa bands in MezSiHD and PhSiHDz have irregular or uncertain contours. The calculations for these three species have been completed by using estimated anharmonicity values. Discussion urnis. Inspection of the values of vmiS in Table I shows that successive methyl substitution lowers the frequency progressively by 17-24 cm-l. This compares with the somewhat larger effects (42-26 cm-') on CH bonds.' It is of interest that, for Si and Ge, the effect diminishes slightly as the number of methyl groups increases whereas for PH
Letters
,/,'
,. Y ^:,.#
x
i
zco
31 i s
3
m '
Flgure 1. Plots of vCHk vs. Do,es(CH) for simple hydrocarbons and fluorocarbons. Data from ref 1 and 3 and references therein. Omission of the CH. plot could lead to a much improved fit.
.-if
2'00
20
*o
60
80
2200
vi:,
20
LO
60
80
2300
20
I cm-1
Figure 2. Plots of vSwkvs. D02e8(SiH)for substituted silanes. Data from this work and ref 4.
and ASH there is a small increase. It is very plain that the equilibrium strength falls with methyl substitution. Figures 1and 2 show correlation plots of vCHk and uSMk against DoB8. The former is clearly satisfactory, the latter unsatisfactory, especially for the methylsilanes. Anharmonicity. We feel that the quantity most appropriate for comparisons of anharmonicity is the parameter x (= 2xe) in the equation u = w(l
-x)
The values of x in Table I show several trends. The first is a consistent increase with the number of methyl groups. The second is that x is significantly larger for PH and ASH bonds than it is for SiH and GeH. All four bonds are significantly less anharmonic than CH ones, for which x = 0.04.14 Thirdly, in the germy1 halides there is a steady increase in x with decreasing electronegativity of the halogen. This could be combined with the effect of Me substitution to relate x to electron donor/acceptor functions; alternatively x might correlate with the weight of the substituents. It is of course possible that these effects derive in part from breakdown of the diatomic approximation; but from the consistency of the observed trends, and the absence of any effect of the weight of the element
(9)A. G. Gaydon, "Dissociation Energies and Spectra of Diatomic Molecules", 3rd ed, Chapman and Hd,London, 1968. DI, = W S ~ / ~ O & ~ - ws/2 + w 4 4 (we include the last, trivial, term),Doo,Dom are the true dissociation energies or AHo values for the dissociation reaction in the gas phase at 0 and 298 K, respectively. Dh = Dooif the potential energy curve is exactly reproduced by a Morse function. (10)C. L. Beckel, M. Shafi, and R. Engelke, J.Mol. Spectrosc. 40,519 (12)S.Cradock, D.C. McKean, and M. W. Mackenzie, J.Mol. Struct. (1971). 74, 265 (1981). (13)T. McAlliater and F. P. Lossing, J.Phys. Chem., 73,2996(1969). (11)D. C. McKean and G. P. McQuillan, J. Mol. Struct. 49, 275 (14)This laboratory, unpublished work. (1978).
The Journal of Physical Chemistry, Voi. 86, No.
Letters
3, 1982 309
TABLE I: Observed VMH' and 2VMHis Values, Anharmonicities, and Dissociation Energies for M H Bonds
SiHD, ( J ) MeSiHD, ( K ) Me,SiHD ( Q ) Me,SiH ( Q ) PhSiHD, ( Q ) Si,HD, (Q) SiHF, (Q) SiHCl, ( Q )
2187.19 2166.56 2146.5 2128.75 2166.33 2162.49 23 1 6 .76g 2 2 6 0 . 25g
4306.96 4264.75 4226 .t 5 4189.2 4255.84 4567.6 4450.6
2254.61 2234.93 (2214~3)~ 2 19 7 . 05 ( 2235.63)f 2231.63 2382.7 2330.2
GeHD, (Q) MeGeHD, (Q) Me,GeHD ( Q ) Me,GeH (8) GeHD,F (Q) GeHD,Cl (K)' GeHD,Br (K)' GeHD,I (K)' PHD, (Q) MePHD ( K ) Me,PH (Q)
(2108.7)h 2084.61 206 2.3' 2041.2 2128.1 2125.12" 2122.24" 2116.02 2323.81 2306.91 2288.27
4150.8 4102.0 4056.4 4014.0 4189.9 4182.77 4176.35 4163.02 4563.26 4527.16 4489.03
(2175. 3)h 2151.8 2130.5 2109.6 2194.4 2192.59 2190.37 2185.04 2408.17 2393.57 2375.78
2122.27 2101.24 2079.4
4168.0 4124.91 4080.1
2198.81 2178.81 2158.1
AsHD, (Q) MeAsHD ( K ) Me,AsH (Q)
2xe I cm- '
Dlin I (kcal mol- )
Doo/ (kcal mol- ) b
( kcal mol- )
0.029 90 0.030 59 [0.030 841" 0.031 08 [0.031 001" 0.030 98 0.027 67 0.030 00
104.6 101.3 (99.5)d 97.9 (99.9)f 99.8 119.7 107.7
91.0 88.1 (86.6)d 85.2 (86.94 86.8 104.1 93.7
90.3' 89.6' 89.4' 90.3' 88.2' 86.3' 100.1c 91.3'
[0.030 621" 0.031 23 0.032 01 0.032 42 0.030 21 0.030 77 0.031 10 0.031 59 0.035 03 0.036 21 0.036 83
(98.5)h 94.5 92.1 90.0 100.7 98.7 97.5 95.8 94.8 91.1 88.8
(79.8)h 76.5 74.6 72.9 81.6 80.0 79.0 77.6 82.5 79.3 77.3
0.034 81 0.035 60 0.036 47
87.2 84.4 81.5
D0z,81
81 2 2' 81'
83.9 .t 3"
70.6 68.3 66.0
Q = central maximum, K = analysis of K structure (I band), J = analysis of J structure ( / I band). Doo= fDljn, where f = 0.87 for SiH and PH and 0.81 for GeH and AsH; (DO,,, - Doo)calcd=1 . 4 8 kcal mol-'. ' Data from iodination kinetics studies, probable error 21-2 kcal m01-I.~ Assuming 2xe = 0.03084. e Estimated value. f Assuming 2xe = 0.03100. Assignment of the overtone band is doubtful. p The very weak satellites at 2302.1 ( F ) and 2248.0 (Cl) we assign to v I + v 4 v 4 on the basis of isotope structure and intensities. Assuming 2x, = 0.03062. The fundamental band is heavily perturbed. Reference 5. M . W. Mackenzie, Ph.D. Thesis, Aberdeen University, 1980. Reference 6. Data for 74Ge. " Data from ref 12. Reference 13.
'
M on the change in anharmonicity, we feel this is unlikely. Dissociation Energies. The values of Doopredicted for SiH4 and GeH4with the Do0/Dlhratios of 0.87 and 0.81, respectively, agree quite well with the iodination values of Dom, remembering that these two quantities will differ by about 1.5 kcal. The disilane and phenylsilane values also agree within the likely experimental error. However, the substitution of three methyl groups is predicted to produce a lowering of 6-7 kcal in Me3SiH and Me3GeH, in marked contrast to the iodination data. In PH3, the somewhat inaccurate Do298value from electron-impact experiments is in good agreement with the spectroscopic value. Throughout all four types of bond, there is a lowering of Dooof 1.5-3 kcal per methyl group substituted. By contrast with the methylsilanes, the spectroscopic Doo values for SiHF, and SiHC13are larger than the experimental ones. The spectroscopic Doo values are of course sensitive to the validity of the diatomic approximation and to the constancy of the DoO/Dlinfactor. If, for instance, x were constant for the series SiH4,..., Me3SiH,the change in dissociation energy would be roughly halved. However, the change of 6 kcal mol-' predicted here is close to the value of 8 kcal mol-' deduced by Austin and Lampe from other kinetic studies,'J5 and is reasonably less than the corresponding change of 11.5 kcal mol-' from methane to i~obutane.~ It remains to consider the implications if the iodination data are in fact correct. These are that potential energy curves for MH bond stretching must vary significantly in (15) In the s i m i i study by these authors of the methyl germanest the ~ GeH, to MeaGeH (26 kcal mol-') seems too overall change in D o from high. This may be due to fitting of the data to too high an aasumed value
of Dozssfor GeH4.
shape, with methyl substitution. Formally, this would be described by a varying DoO/Dlinfactor, corresponding perhaps tp varying amounts of radical relaxation. The only clearcut cases of shape variation known to us are in the field of CH bonds where the radicals produced are resonance stabilized, e.g., PhCH2., -CH2N02. This results in a CH dissociation energy lower than expected from the equilibrium strength, as measured by the stretching frequency. For the silanes and germanes, we would need a physical explanation of the opposite effect, why the dissociation energies are higher than the stretching frequencies predict, and why the heats of formation of the radicals produced are unusually high, and for methyl substitution only. So far, no cause has suggested itself to us. Conclusions The dissociation energies predicted for SiH, GeH, PH, and ASH bonds in some silanes, germanes, phosphines, and arsines from spectroscopic data, using empirical factors obtained from diatomic molecules, are of sensible absolute magnitudes and show progressive weakening of bonds with methyl substitution to the extent of 1.5-3 kcal mol-' per methyl group. This conflicts with the constancy evident in the Dozsavalues from iodination kinetics studies. Either the latter are in error, or there is a systematic change in shape of the potential energy curves. In any case the "equilibrium stren$h" of the MH bond diminishes with methylation.
Acknowledgment. We thank the Science Research Council for research assistantships for I.T. and A.R.M. and for the FT IR facility, and Dr. R. Walsh for communicating his review and data in advance of publication.