Journal of Molecular Structure 554 (2000) 251±269 www.elsevier.nl/locate/molstruc Conformational equilibrium in dimethyl vinyl ¯uorosilane studied by infrared and Raman spectroscopy A. Horn a, P. Klaeboe a, V. Aleksa a,b, A. Gruodis a,b, C.J. Nielsen a, Y.E. Nashed c, G.A. Guirgis c,1, J.R. Durig c,* a Department of Chemistry, University of Oslo, P.O. Box 1033, 0315 Oslo, Norway Department of General Physics and Spectroscopy, Vilnius University, Vilnius 2734, Lithuania c Department of Chemistry, University of Missouri-Kansas City, 5100 Rockhill Road, Kansas City, MO 64110-2499, USA b Received 12 April 2000; accepted 30 May 2000 Abstract The Raman spectra (3500±20 cm 21) of liquid with depolarization values and solid, as well as the infrared spectra of the gas, the sample isolated in argon and nitrogen matrices at ca. 5 K and solid dimethyl vinyl ¯uorosilane, CH2yCHSi(CH3)2F, have been recorded. Both gauche and syn rotamers have been identi®ed in the ¯uid phases but only the syn conformer remains in the solid. Variable temperature (255 to 21508C) studies of the infrared spectra (4000 and 400 cm 21) of dimethyl vinyl ¯uorosilane dissolved in liquid xenon and krypton have been recorded. From the xenon and krypton data, the enthalpy differences have been determined to be 53 ^ 9 cm 21 (0.64 ^ 0.10 kJ/mol) and 44 ^ 7 cm 21 (0.53 ^ 0.09 kJ/mol), respectively, with the gauche conformer being the more stable form. The intensity variations with temperature of the Raman spectrum of the liquid gave an enthalpy difference of 25 ^ 15 cm 21 (0.30 ^ 0.18 kJ/mol) also with the gauche conformer being the more stable form. Vibrational assignments are provided for both conformers. Complete equilibrium geometries have been determined for both rotamers using ab initio calculations employing the 6-31G(d), 6-3111G(d,p) and 6-3111G(2d,2p) basis sets at the levels of restricted Hartree±Fock (RHF) and/or with full electron correlation by the perturbation method, Moller±Plesset (MP), to second order. The syn conformer is predicted to be the more stable conformer from all ab initio calculations except those of MP2/6-31(d) which predict the gauche form being the more stable conformer by 54 cm 21 (0.65 kJ/mol) although the values favoring the syn form are all very small. These results are compared to the corresponding quantities of some similar molecules. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Conformational stability; FT-IR spectra; Ab initio calculations; Dimethyl vinyl ¯uorosilane 1. Introduction A number of silanes in which the silicon atom is * Corresponding author. Tel.: 11-816-235-1136; fax: 11-816235-5191. E-mail address: durigj@umkc.edu (J.R. Durig). 1 Permanent address: Analytical Research and Development Department, Bayer Corp., P.O. Box 118088, Charleston, SC 29423, USA. attached to a sp 2 hybridized carbon atom have been investigated by infrared and Raman spectroscopy. When the silicon atom has different substituents attached, these molecules will have possibilities for conformational equilibria. The vinyl silanes CH2yCHSiX2Y in which X and Y are different groups will, like the corresponding propenes, exist in a syn conformer with a plane of symmetry and in two equivalent gauche conformers. Thus, from the 0022-2860/00/$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0022-286 0(00)00677-3 252 A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Fig. 1. Raman spectra of dimethyl vinyl ¯uorosilane in two polarization directions, 1700±800 cm 21 with ordinate scale 0±3000 counts, 700± 100 cm 21 with 0±30,000 counts. vibrational spectral data [1,2] and microwave investigations [3] it was reported that vinylsilylchloride is present in syn and gauche conformers in the vapor and liquid states, whereas the gauche conformer is the stable form in the crystal. However, for the vapor state these investigators ®rst reported [1] the gauche as the more stable form but from a later investigation [2], Si-d2 isotopomer, the syn rotamer was reported as the more stable conformer. Although the gauche rotamer was calculated to be the more stable conformer from ab initio RHF/6-31G(d) calculations, the syn conformer was determined to be more stable in the liquid by variable temperature Raman studies [4,5]. However, it should be noted that the conformer that is the most stable form in the liquid, may not be the most stable rotamer in the gas. Therefore, we [6] carried out a variable temperature FT-IR investigation of rare gas solutions of vinylsilyl chloride. These studies indicated that the gauche was more stable by 78 ^ 11 cm 21 (0.93 ^ 0.13 kJ/mol). Infrared and Raman studies combined with ab initio calculations have also been carried out for dimethyl vinyl chlorosilane [7] and for methyl vinyl dichlorosilane [8]. Since the methyl group and the chlorine have approximately the same size, the conformational preference is by no means obvious for these molecules. In the former molecule the gauche conformer was more stable and was present in the crystal [7] whereas in the latter the syn conformer had lower energy and was the sole conformer present in the crystal [8]. These results indicate a preference of the methyl group eclipsing the double bond over the chlorine atom. When these studies are extended to ¯uorine substituents it was observed that in methyl vinyl di¯uorosilane [9,10] the gauche conformer has a lower energy in the gas and liquid whereas the syn form is present in the crystal. In order to obtain more information on the relative stability of the conformers of dimethyl vinyl halosilanes we have recorded the infrared and Raman spectra to determine the conformational stability of dimethyl vinyl ¯uorosilane, CH2yCHSi(CH3)2F, (DVFS), and the results of this study will be described in the present paper. 2. Experimental The sample was prepared by the reaction between the chloro derivative CH2yCHSi(CH3)2Cl, and freshly sublimed antimony tri¯uoride at room temperature without solvent for 1 h. The sample was puri®ed by a low temperature, low pressure fractionation column and the purity was checked by mass spectrometry. The Raman spectra of the liquid, amorphous solid and crystal were obtained at different temperatures in a capillary tube of 2 mm inner diameter, surrounded by a Dewar, cooled by gaseous nitrogen evaporated A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 253 Fig. 2. Raman spectra (1050±50 cm 21) of amorphous (solid line) and annealed (dotted line), crystalline solids of dimethyl vinyl ¯uorosilane. from a reservoir [11]. These spectra were employed for calculating the enthalpy difference DH between the conformers in the liquid. DVFS has a pronounced hysteresis (undercooling) and it was possible to study the liquid far below the freezing point. The crystallization often occurred spontaneously at ca. 21238C and the anisotropic crystal containing only one conformer was obtained. Independently, the vapor of DVFS was condensed on a copper ®nger at 21968C, and the Raman spectrum of the amorphous phase was recorded. Subsequently, the amorphous solid was annealed to temperatures slightly below the melting point, the sample turned crystalline from Fig. 3. Infrared spectrum of a gas of dimethyl vinyl ¯uorosilane. visual inspection and was recooled to 21968C before the spectrum was obtained. The Raman spectra were recorded digitally using a Dilor RTI-30 spectrometer (triple monochromator, with a Peltier cooled detector) coupled to a PC. An argon ion laser from Spectra Fig. 4. Far infrared spectra of dimethyl vinyl ¯uorosilane: (A) gas; (B) amorphous (solid line) and annealed solid (dotted line). 254 A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Fig. 5. Infrared spectra (1100±400 cm 21) of amorphous (solid line) and crystalline (dotted line) solids of dimethyl vinyl ¯uorosilane. Physics (model 2000) was employed with perpendicular illumination using the 514.5 nm line for excitation. The Raman spectra are shown in Figs. 1 and 2. The infrared spectra (Figs. 3±8) were recorded with Fig. 6. Infrared spectra (1350±800 and 860±500 cm 21) of dimethyl vinyl ¯uorosilane in an argon matrix unannealed (solid line) and annealed (dashed line). various Fourier transform spectrometers: Bruker models IFS-88 and IFS-66 (4000±450 cm 21), a Nicolet model 800 (4000±450 cm 21), a Perkin± Elmer model 2000 (4000±450 cm 21) and on two vacuum benches: Bruker IFS-113v spectrometer (600±50 cm 21) and Bomem model DA 3.002 (600±50 cm 21). The latter instrument had Fig. 7. Infrared spectra (1025±950 cm 21) of dimethyl vinyl ¯uorosilane in a nitrogen matrix unannealed (solid line) and annealed. A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Fig. 8. Infrared spectra of dimethyl vinyl ¯uorosilane: (A) experimental spectrum of the liquid Krypton (21258C); (B) calculated spectrum of the mixture (DH 44 cm 21); (C) calculated spectrum of the syn conformer; (D) calculated spectrum of the gauche conformer. a helium cooled Bolometer as detector, the other instruments had detectors of DTGS. Beamsplitters of Ge substrate on KBr were used in the midinfrared regions (MIR) whereas beamsplitters of Mylar of thickness 3.5 and 12 m as well as one of a metal mesh were employed in the far infrared (FIR) region. The spectrum of the vapor was recorded with the sample contained in cells with KBr windows and path length 10 cm in MIR and in cells of 20 cm and 1 m path lengths with polyethylene windows in the FIR region. The spectra of the amorphous and crystalline solids were obtained by depositing the vapor on a CsI window and on a wedge shaped window of silicon, cooled with boiling liquid nitrogen, for the MIR and FIR regions, respectively. The sample was diluted with argon and nitrogen (1:500 and 1:1000) and deposited on a CsI window of a three stage Displex cryostat from APD (model HS-4) at either 2268 or 22588C. The matrices were 255 subsequently annealed to various temperatures from 2253 to 22368C (argon) and from 2253 to 22398C (nitrogen) in periods from 10 min to 1 h and the window was recooled to 22688C and the spectra recorded (Figs. 6 and 7). The mid-infrared spectra (Fig. 8) of the samples dissolved in lique®ed xenon (255 to 21008C) and krypton (2105 to 21508C) as a function of temperature were recorded on a Bruker model IFS 66 Fourier transform interferometer equipped with a Globar source, a Ge/KBr beamsplitter and a TGS detector. In all cases, 100 interferograms were collected at 1.0 cm 21 resolution, averaged and transformed with a boxcar truncation function. For these studies, a specially designed cryostat cell was used. It consisted of a copper cell with a path length of 4 cm with wedged silicon windows sealed to the cell with indium gaskets. It was cooled by boiling liquid nitrogen to 21968C. The temperature was monitored with two Pt thermoresistors. The complete cell was connected to a pressure manifold, allowing the ®lling and evacuation of the cell. After the cell had cooled to the desired temperature, a small amount of the compound was condensed into the cell. Next the pressure manifold and the cell were pressurized with the noble gas, which immediately started to condense in the cell, allowing the compounds to dissolve. All observed infrared and Raman bands with signi®cant intensities are listed in Table 1. 3. Ab initio calculations The LCAO±MO±SCF calculations were performed with the gaussian-94 program [12] with Gaussian-type basis functions. The energy minima with respect to the nuclear coordinates were obtained by the simultaneous relaxation of all of the geometric parameters, except for the symmetry restrictions for the gauche and cis conformers, using the gradient method of Pulay [13]. The structural parameters were determined from RHF/6-31G(d) (restricted Hartree±Fock), MP2/6-31G(d) (full electron correlation by the perturbation method to the second order), MP2/6-3111G(d,p) and MP2/6-3111G(2d,2p) calculations and the results are given in Table 2. The energy difference that resulted from these various calculations ranged from 46 cm 21 (0.55 kJ/mol) from Infrared 256 Table 1 Infrared and Raman spectral data (abbreviations used: s, strong, m, moderate; w, weak; v, very; bd, broad; sh, shoulder; p, polarized; d, depolarized. A, B and C denote vapor contours; asterisks denote band vanishing in the crystal spectra; arrows pointing upwards and downwards signify matrix bands which increase and decrease in intensities after annealing; and P, Q, and R refer to the rotational±vibrational branches) for dimethyl vinyl ¯uorosilane (CH2yCHSi(CH3)2F) Raman Solid Vapor m 3025 w 2986 R 2973 Q,C 2961 P 2915 max 1937 R 1928 Q,A 1919 P 1608 R 1601 Q,A 1595 P m w Amorphous (80 K) Crystalline (80 K) Liquid 3074 w 3063 m 3027 w 3018W 3071 w 3062 m 3026 w 3019 w 3059 m 3059 m 3064 m,br,P n10 n1 3018 w/m 3022 w,D? n2 3000 w 3000 w 2988 vw 3024 w 3018 w 2993 w,sh 2986 w/m,sh 2974 m/w,sh 2963 m 2954 m 2976 m 2970 w 2957 w 2916 vw 1941 vw w m 1929 vw 1923 vw 1605 w 1603 m 1598 w o " 1446 w 1442 w 1423 w 1421 vw " 1420 Q m 1414 Q 1411 Q 1405 sh m m m 1413 m 1408 m 1297 sh w 1272 R 1266 Q m 1263 Q m 1254 sh 1398 m # 1293 w 1277 w m m m o 1262 vs 1257 vs " 2982 vw 2975 m 2963 w 2963 s 2957 w 2916 vw 2888 w 2954 s 2907 w 2876 vw 1939 w 1930 w 1601 m # 1599 m " 1597 m/s 1596 w/m 1596 w/m 1447 m 1446 w 2907 w 2882 vw 1930w 1917 Amorphous (80 K) Crystalline (80 K) Interpretation n3 2981 vs,P 2960 m,sh,D n4 2911 vs,P n 5,n 6 n7 n 8,n 9 o 1601 vs,P 1601 vs,P 1597 m 1597 m 1601 m 1595 m n11, n12 1425 vw 1421 w # n13 n14 1415 w 1413 s 1409 m 1403 m " 1401 m # 1294 vw 1277 w # 1263 s 1258 vs 1253 s " 1250 s n10 n10 0 o 1409 s 1412 m 1410 m 1398 m 1299 w 1397 m 1305 w 1275 w 1271 w 1279vs,P 1272m 1254 vs,br 1257 m 1252 m 1260 vw 1258 vw 1416 vs,P 1412 s 1415 m 1413 m 1402 w,sh 1405 w,sh 1274 m 1270 m 1263 w 1258 w n15 n15 0 n 16 n 17 A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 3071 R 3065 Q,C 3058 P Ar matrix (5 K) N2 matrix (5 K) Table 1 (continued) Infrared Raman Solid Vapor Ar matrix (5 K) N2 matrix (5 K) 1013 max m 1007 max m 904 R 895 Q 886 sh 855 R 850 Q,C 848 Q,C 841 P 806 R 799 Q 789 P 772 Q m 1244 w 1018 w " 1015 m # 1011 m Crystalline (80 K) Liquid Amorphous (80 K) Crystalline (80 K) Interpretation 1244 m 1252 vw 1250 vw n 18 1008 w 1009 w n 19 1004w 1002 w n 20 n 20 0 n 21 0 n 21 1017 vw 1012 s 1012 s 1008 s # 1006 w 1005 s 1002 s 969 vs " 970 s 959 s 967 w,D 971 w 957 w 967 w,D 963 w p 883 w,D 874 vw 967 m " 966 s,sh " 964 s # 963 s 961 vs 958 m 890w 966 s # p 890 m " 893 w,sh vs vs 866 vs 884 vs 876 w # 883 vs # 879 vs # 873 m # 876 s 865 s 862 s,sh vs vs 847 vs " 845 vs # 850 vs " 848 vs # 845 vvs 846 vs m 1014 w,D p 839 m " 805 m 796 m 793 s 788 m/w 769 m 765 m 741 vw p n 22 n 22 0 864 w 855 vw 850 w 848 w n 23 n 23 0 799 m,D 806 w 795 w n 24 770 m 767 w 805 w 800 w 792w 769 vw V25 743 vw 742 vw n 26 801 s " 796 vs # 799 vs 797 vs 772 m # 766 w " 736 vw # 766 s 739w 711 s " 711 s p 712 m,p 714 m p n 27 m 770 M # 768 M " 733 W # 714 W " 712 vs " 710 m " 692 m # 694 m # 695 m 696 m,P 695 m 695 m n 28,n 27 0 616 R 607 Q vw 607 w 606 w # 601 w # 607 m 696 m 692 w 613 vs,P 621 m,sh 613 vs n 29 0 521 max m 611 vs 530 w m 527 m " 525 m # 522 w # 526 w,D 515 max 528 s " 526 s # 523 s # 716 R 710 Q,C 703 P 694 sh vs m m 524 s 519 w n 29 n 30 523 vw n 30 0 257 532 s p A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 974 R 969 Q,C 966 P 966 R 964 Q,C 955 1249 w # 1016 m " 1014 m # 1011 m # 1008 m 1007 s 1005 w,sh Amorphous (80 K) 258 Table 1 (continued) Raman Solid Vapor 400 R 394 Q, A/C 388 P 365 max 273 max m m s 267 sh 218 vw,sh 180 175 170 164 w vw w vw Ar matrix (5 K) N2 matrix (5 K) Amorphous (80 K) Crystalline (80 K) Liquid 520 m " 519 s 519 m 394 s 394 s 396 m,P 397 m p 275 s 371 m,P 275 m,D 261 s 265 m,P 368 286 w 277 w 263 m p 225 w,sh 209 w,sh 195 s,D 187 w,sh 178 w 230 w 214 m 201 m 190 w,sh 178 w 101 w 95 vw 104 m,br 515 m " 366 m 275 s 272 s 263 s 228 w 208 w 201 w 179 vw 151 vw 209 w 178 vw 178 vw 151 vw Amorphous (80 K) Crystalline (80 K) Interpretation 517 vw 395 m n 31 0 p n 31 n 32,n 33 0 285 w 264 m 258 m n 33,n 32 0 p n 34 209 m 187 w 175 w n 35,n 34 0 n 36,n 35 0 n 37 n 36 0 n 39 110 m 102m 64 m 47 m lattice A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Infrared Table 2 Ê , bond angles in (8), rotational constants in MHz, and dipole moments in Debye), rotational constants, dipole moments and energy for Structural parameters (bond distances in A dimethyl vinyl ¯uorosilane RHF/6-31G(d) MP2/6-31G(d) MP2/6-3111G(d,p) MP2/6-3111G(2d,2p) syn gauche syn gauche syn gauche syn rC1 ±C2 rSi±C2 rF±Si3 rC1 ±H5 rC1 ±H6 rC2 ±H7 rSi±C8 rSi±C9 rC8 ±H10 rC8 ±H11 rC8 ±H12 rC9 ±H13 rC9 ±H14 rC9 ±H15 /SiC2C1 /FSiC2 /H5ClC2 /H6ClC2 /H7C2C1 /C8SiF /C9SiF /H10C8Si /H11C8Si /H12C8Si /H13C9Si /H14C9Si /H15C9Si t FSiC2C1 t H5ClC2Si t H6C1C2H5 t H7C2C1Si t C8SiFC2 t C9SiFC2 t H10C8SiF t H11C8SiH10 t H12C8SiH10 t H13C9SiF t H14C9SiH13 1.325 1.869 1.609 1.078 1.077 1.081 1.876 1.876 1.087 1.086 1.087 1.087 1.087 1.087 124.5 107.4 122.3 122.4 117.4 107.5 106.7 111.7 111.8 110.3 111.2 111.3 111.3 117.9 180.0 180.0 180.0 120.9 2119.0 178.7 121.1 2119.4 2179.3 2119.9 1.325 1.868 1.608 1.078 1.076 1.082 1.876 1.876 1.088 1.086 1.087 1.088 1.086 1.087 123.3 106.4 122.3 121.9 117.8 107.9 107.9 111.1 111.2 111.4 111.1 111.2 111.4 0.0 180.0 180.0 180.0 119.9 2119.9 181.8 119.8 2119.9 2181.8 2119.8 1.344 1.859 1.634 1.088 1.087 1.091 1.866 1.866 1.094 1.093 1.094 1.094 1.093 1.094 123.4 107.7 122.6 121.9 117.5 107.7 107.2 111.3 111.8 110.2 110.7 111.3 111.3 115.0 180.0 180.0 180.0 120.5 2119.2 178.8 121.0 2119.4 2179.3 2119.8 1.344 1.859 1.632 1.087 1.086 1.091 1.867 1.867 1.094 1.093 1.094 1.094 1.093 1.094 122.5 106.2 122.6 121.3 117.8 108.1 108.1 110.8 111.0 111.4 110.8 111.0 111.4 0.0 180.0 180.0 180.0 119.6 2119.6 182.1 119.7 2119.9 2182.1 2119.7 1.347 1.859 1.637 1.087 1.088 1.091 1.861 1.861 1.094 1.093 1.094 1.094 1.094 1.094 123.4 107.3 122.1 121.7 117.2 107.4 106.7 111.4 111.4 110.1 110.7 111.0 111.1 120.7 180.0 180.0 180.0 119.9 2119.4 178.4 121.0 2119.5 2180.6 2119.7 1.347 1.858 1.636 1.087 1.087 1.091 1.861 1.861 1.095 1.093 1.094 1.095 1.093 1.094 122.9 106.2 122.0 121.3 117.4 107.6 107.6 110.6 110.8 111.2 110.6 110.8 111.2 0.0 180.0 180.0 180.0 119.7 2119.7 181.7 119.6 2119.8 2181.7 2119.6 1.340 1.857 1.621 1.080 1.081 1.084 1.859 1.859 1.087 1.086 1.087 1.087 1.086 1.087 123.2 107.5 122.1 121.5 117.0 107.8 106.8 111.2 111.4 110.1 110.6 110.9 111.1 121.6 180.0 180.0 180.0 119.8 2119.4 178.2 120.9 2119.4 2179.7 2119.8 1.340 1.856 1.621 1.081 1.080 1.084 1.859 1.859 1.087 1.086 1.087 1.087 1.086 1.087 122.7 106.3 122.0 121.1 117.3 107.8 107.8 110.6 110.8 111.1 110.6 110.8 111.1 0.0 180.0 180.0 180.0 119.7 2119.7 181.9 119.7 2119.9 2181.9 2119.7 259 gauche A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Parameter 119.9 3407.7 2069.9 1997.3 0.152 0.000 1.832 1.838 1.450232 120.2 3352.0 2064.1 1973.3 1.071 0.283 1.845 2.152 1.450188 10 119.8 3381.3 2055.2 1985.9 0.128 0.000 2.030 2.034 1.348714 120.2 3327.1 2048.2 958.4 1.205 0.278 2.062 2.404 1.348615 22 t H15C9SiH13 A B C um a u u m bu um c u um t u 2(E 1 545) DE (cm 21) 120.0 3314.7 2048.5 1959.5 1.114 0.209 1.833 2.155 0.943956 119.9 3389.8 2055.0 1971.9 0.004 0.000 1.698 1.698 0.142580 120.1 3320.0 2052.6 1945.2 1.033 0.213 1.716 2.014 0.142371 46 119.9 3371.9 2054.9 1987.6 0.041 0.000 1.796 1.796 0.943711 54 gauche gauche syn gauche Parameter MP2/6-31G(d) RHF/6-31G(d) Table 2 (continued) syn MP2/6-3111G(d,p) syn gauche syn A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 MP2/6-3111G(2d,2p) 260 RHF/6-31G(d), 254 cm 21 (20.64 kJ/mol) from MP2/6-31G(d) and 22 cm 21 (0.26 kJ/mol) from MP2/6-3111G(d,p) and 10 cm 21 (0.12 kJ/mol) from MP2/6-3111G(2d,2p) calculations with the syn rotamer the more stable conformer from each calculation except from the MP2/6-31G(d) calculation which predicted the gauche rotamer being more stable. For the normal coordinate analysis, the force ®eld in Cartesian coordinates was obtained with the gaussian-94 program [12] from the MP2/6-31G(d) calculations. Internal coordinates were de®ned as shown in Fig. 9, which were used to form the symmetry coordinates listed in Table 3. The Cartesian coordinates obtained from the optimized geometry were used to calculate the B-matrix elements with the G matrix program of Schachtschneider [14]. These B-matrix elements were used to convert the ab initio force ®eld in Cartesian coordinates to a force ®eld in the desired internal coordinates. The resulting force ®elds for the gauche and cis conformers are available from the authors. These force ®elds were used in a massweighted Cartesian coordinate calculation to reproduce the ab initio vibrational frequencies and to determine the potential energy distribution (PED) which is given in Table 4 for the two conformers. All the elements of the force ®eld in internal coordinates from the MP2/6-31G(d) calculation were then assigned scaling factors of 0.9 for the stretches and bends and 1.0 for the torsions and the calculation repeated to obtain the ®xed scaled force ®eld and scaled vibrational frequencies. To aid in the vibrational assignment for the CH2CHSi(CH3)2F molecule, the infrared and Raman spectra were calculated using frequencies, Raman scattering activities (RHF/6-31G(d)), and infrared intensities (MP2/6-31G(d)) determined from the ab initio calculations. The evaluation of the Raman activity by using the analytical gradient method has been developed [15,16]. The activity Sj can be expressed as: Sj gj 45a2j 1 7b2j where gj is the degeneracy of the vibrational mode j, a j the derivative of the isotropic polarizability and b j the derivative of the anisotropic polarizability. The Raman scattering cross sections, 2s j =2V; which are A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 261 Fig. 9. Internal coordinates of dimethyl vinyl ¯uorosilane. proportional to the Raman intensities, can be calculated from the scattering activities and the predicted wavenumbers for each normal mode using the relationship [17,18]: 2s j 2V 0 ! 2 4 p4 B B B 45 @ 1 ! C n0 2 nj 4 h C Sj C A 2hcnj 8p2 cnj 1 2 exp kT where n 0 is the exciting frequency, n j the vibrational frequency of the jth normal mode and Sj the corresponding Raman scattering activity. To obtain the polarized Raman scattering crosssections, the polarizabilities are incorporated into Sj by Sj 12 j= 11 j where r j is the depolarization ration of the jth normal mode. The Raman scattering crosssections and calculated frequencies were used together with a Lorentzian function to obtain the calculated spectrum. The experimental and predicted Raman spectra of dimethyl vinyl ¯uorosilane are shown in Fig. 10. The predicted spectra are compared to the experimental Raman spectrum of the liquid, which is shown in Fig. 10A. These spectra were very useful for making the vibrational assignments to the correct bands for the two conformers. Infrared intensities were also calculated based on the dipole moment derivatives with respect to the Cartesian coordinates. The derivatives were taken from the ab initio calculations MP2/6-31G(d) and transformed to normal coordinates by ! X 2m u 2mu Lij 2Qi 2Xj j where Qi is the ith normal coordinate, Xj the jth Cartesian displacement coordinate and Lji the transformation matrix between the Cartesian displacement coordinates and normal coordinates. The infrared intensities were then calculated by " # 2my 2 Np 2mx 2 2m z 2 1 1 Ii 2 2Qi 2Qi 2Qi 3c In Fig. 8, the predicted infrared spectra are shown for the pure gauche (Fig. 8D), pure syn (Fig. 8C) and the mixture (Fig. 8B). The experimental infrared spectrum of the normal species dissolved in liquid krypton at 21258C is also shown for comparison in Fig. 8A. Excluding the overtones or combination bands, the calculated spectra have some differences from the experimental one especially in comparing the relative intensities of the bands in the 1300 cm 21 region. Nevertheless, they provide support for the assignments of the observed bands to the indicated fundamentals for each conformer. 4. Conformational stability There are a few fundamentals which show conformer doublets in the infrared and Raman spectra 262 A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Table 3 Symmetry coordinates (not normalized) for dimethyl vinyl ¯uorosilane Description CH2 CH3 CH3 CH3 antisymmetric antisymmetric antisymmetric antisymmetric Internal coordinate stretch stretch stretch stretch CH2 symmetric stretch CH3 antisymmetric stretch CH stretch CH3 symmetric stretch CH3 symmetric stretch C C stretch CH2 deformation CH3 antisymmetric deformation CH3 antisymmetric deformation CH3 antisymmetric deformation CH3 antisymmetric deformation CH3 symmetric deformation CH3 symmetric deformation CH in-plane bend CH2 twist CH2 wag CH2 rock SiF stretch CH3 rock CH3 rock CH3 rock CH3 rock SiC stretch SiC2 antisymmetric stretch SiC2 symmetric stretch C±H out-of-plane bend CCSi bend SiC2 wag SiC2 rock SiC2 deformation methyl torsion CSiF bend SiC2 twist methyl torsion Asymmetric torsion S1 r1 2 r2 S2 r5 2 r6 2 r8 1 r9 S3 r1 1 r2 S4 2r4 2 r5 2 r6 1 2r7 2 r8 2 r9 S5 r1 1 r2 S6 2r4 2 r5 2 r6 2 2r7 1 r8 1 r9 S7 r3 S8 r4 1 r5 1 r6 1 r7 1 r8 1 r9 S9 r4 1 r5 1 r6 2 r7 2 r8 2 r9 S10 Q S11 2a 2 b1 2 b2 S12 2d1 2 d2 2 d3 2 2d4 2 d5 2 d6 S13 d1 2 d3 2 d4 1 d6 S14 2d1 2 d2 2 d3 2 2d4 1 d5 1 d6 S15 d1 2 d3 1 d4 2 d6 S16 d1 1 d2 1 d3 2 f1 2 f2 2 f3 1 d4 1 d5 1 d6 2 f4 2 f5 2 f6 S17 d1 1 d2 1 d3 2 f1 2 f2 2 f3 2 d4 2 d5 2 d6 1 f4 1 f5 1 f6 S18 y1 2 y2 S19 g S20 b1 2 b2 S21 e S22 S S23 f2 2 f3 1 f5 2 f6 S24 f2 2 f3 2 f5 1 f6 S25 2f1 2 f2 2 f3 1 2f4 2 f5 2 f6 S26 2f1 2 f2 2 f3 2 2f4 1 f5 1 f6 S27 R S28 X1 2 X2 S29 X1 1 X2 S30 g 0 S31 2p 2 y1 2 y2 S32 v1 1 v2 2 u1 2 u2 S33 up 2 u2 1 v1 2 v2 1 Sp 34 6 1 2S2 6 2 2v 2 u 1 2 u2 2 v1 2 v2 S35 tp 2 t3 2 Sp 36 6 2 2S2 6 1 2v 1 u 1 1 u2 1 v1 1 v2 S37 u1 2 u2 2 v1 1 v2 S38 t2 1 t3 S39 t1 Fig. 10. Raman spectra of dimethyl vinyl ¯uorosilane: (A) experimental spectrum of the liquid; (B) calculated spectrum of the mixture (DH 44 cm 21); (C) calculated spectrum of the syn conformer; (D) calculated spectrum of the gauche conformer. of the ¯uid phases. The two bands at 521 and 513 cm 21 in the infrared spectrum of the gas which are assigned as the C±H out-of-plane bending modes, demonstrate the presence of conformers, where by repeated annealing of the amorphous solid only the lower frequency band remains. Other bands which are observed in the infrared and Raman spectra of the ¯uid phases and amorphous solid but not in the spectrum of the annealed solid are observed at 962, 710, 607, 366, 272, 228 and 201 cm 21. These data clearly indicate that there are two conformers present in the ¯uid phases at ambient temperatures but only one rotamer remains in the polycrystalline solid. The band at 710 cm 21 is the only band predicted by ab initio calculations in this region (703 cm 21) and it can de®nitely be assigned to the Si±C stretch of the gauche conformer. Since this band is drastically decreased in the intensity in the infrared spectra of the argon and nitrogen matrices and annealed solid, it can be concluded that the syn Table 4 Observed and calculated wavenumbers for gauche and syn conformers of dimethyl vinyl ¯uorosilane Description A0 n1 n2 n4 n5 n6 n7 n8 n9 n 10 n 11 n 12 n 13 n 14 n 15 n 16 n 17 n 18 Ab initio a Fixed IR scaled b int. c Raman dp Obs e d d act ratio PED Ab Fixed IR initio a scaled b int. c Raman dp Obs e d d act ratio PED 3282 3113 15.5 134.4 0.14 3062 99S1 3289 3120 10.4 134.0 0.14 3071 99S1 3206 3042 7.1 70.6 0.67 3019 40S2,41S3,10S6 3207 3042 3.4 68.2 0.73 3019 94S2 3203 3039 8.4 76.1 0.57 3000 54S3,44S1 3207 3042 10.2 64.8 0.63 3000 96S3 3201 3037 14.6 81.1 0.75 2975 53S4,20S5,10S6 3197 3033 14.4 140.6 0.75 2975 95S4 3200 3036 2.1 139.9 0.73 2970 1 54S5,19S4,18S7 3201 3037 11.7 51.0 0.75 2970 1 85S5,14S7 3198 3034 3.9 74.2 0.72 2970 1 77S6,18S4 3195 3031 0.4 146.8 0.68 2970 1 94S6 3187 3108 3023 2948 9.4 0.4 33.7 146.2 0.72 0.01 2957 2916 75S7,24S5 61S8,39S9 3182 3105 3019 2946 7.3 1.3 14.6 217.5 0.75 0.01 2957 2916 86S7,13S5 100S8 3105 2946 0.6 68.2 0.01 2916 61S9,39S8 3104 2945 0.4 0.4 0.75 2916 100S9 1680 1534 1526 1594 1455 1447 11.2 9.3 4.1 24.8 3.0 26.8 0.14 0.47 0.74 1601 1446 1 1442 1 61S10,33S15 49S11,45S12 51S12,44S11 1677 1525 1531 1591 1446 1453 11.8 1.9 11.8 30.9 2.3 32.2 0.16 0.75 0.75 1599 1442 1 1446 1 1521 1443 1.9 17.5 0.75 1425 79S13,17S14 1519 1441 1.2 13.3 0.75 1425 43S13 1517 1439 0.7 8.9 0.71 1421 76S14,17S13 1515 1438 0.9 16.2 0.75 1421 93S14 1489 1413 23.1 31.5 0.41 1413 65S15,24S10 1486 1410 18.3 30.8 0.34 1409 68S15,24S10 1379 1308 33.5 0.8 0.59 1263 1 97S16 1377 1307 30.6 0.8 0.43 1262 1 98S16 1374 1303 58.0 1.2 0.75 1257 1 96S17 1372 1302 56.4 1.0 0.75 1257 1 97S17 1324 1256 2.9 13.3 0.37 1244 59S18,25S20 1323 1255 1.3 14.5 0.35 1244 58S18,26S20 1062 1053 995 948 916 1008 999 944 900 869 20.5 15.6 38.7 174.7 163.0 0.5 1.1 3.1 0.5 1.8 0.74 0.75 0.64 0.74 0.72 1013 p 1008 966 883 848 58S19,37S30 60S20,29S18 98S21 45S22,36S25 70S23 1063 1050 999 947 919 1007 996 947 898 872 27.0 20.1 127.5 129.8 178.0 0.4 1.3 3.3 0.9 1.7 0.75 0.75 0.64 0.64 0.69 1013 p 1006 969 879 850 62S19 57S20,30S18 99S21 45S22,30S25 71S23 62S10,31S15 55S11,40S16 55S12,40S18 263 n 19 n 20 n 21 n 22 n 23 Syn A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 n3 CH2 antisymmetric stretch CH3 antisymmetric stretch CH3 antisymmetric stretch CH3 antisymmetric stretch CH2 symmetric stretch CH3 antisymmetric stretch CH stretch CH3 symmetric stretch CH3 symmetric stretch CyC stretch CH2 deformation CH3 antisymmetric deformation CH3 antisymmetric deformation CH3 antisymmetric deformation CH3 antisymmetric deformation CH3 symmetric deformation CH3 symmetric deformation CH in-plane bend CH2 twist CH2 wag CH2 rock SiF stretch CH3 rock Gauche 264 Table 4 (continued) Description Gauche Ab initio a Fixed IR scaled b int. c Raman dp Obs act d ratio d 842 799 114.4 0.9 0.67 799 n 25 n 26 n 27 n 28 822 797 741 727 779 756 703 689 11.6 0.2 37.9 2.5 2.1 2.1 3.6 4.8 0.74 0.75 0.52 0.74 620 588 0.7 16.6 Raman dp Obs e d d act ratio PED 846 802 147.4 1.4 0.75 799 34S24,38S28, 13S33 766 736 711 694 823 797 723 725 781 757 686 688 13.9 1.0 22.9 3.1 2.1 1.7 5.4 5.4 0.75 0.75 0.75 0.43 772 736 694 694 45S25,43S22 61S26,25S24 41S27,36S29,11S23 50S28,22S26,26S24 0.06 601 56S29,30S27 626 594 0.0 14.3 0.01 606 45S29,34S27 25S30,12S36, 32S19,21S38 46S31,27S33 43S32,33S36 40S33,11S31,25S37 28S34,33S32 51S35,21S34,14S36 18S36,29S34,34S35 40S37,18S31,13S35, 13S33,14S38 76S38,14S37 68S39,28S30 530 507 21.8 5.4 0.75 515 404 255 282 199 177 155 175 383 241 269 189 174 149 167 19.1 7.0 11.4 0.4 0.1 0.5 0.1 4.4 1.7 1.1 1.3 4.1 0.5 0.3 0.25 0.75 0.43 0.73 0.75 0.75 0.72 394 p 267 p 273 p 1958 175 p 151 X 170 p 33S30,30S19, 16S38 42S31,41S36 71S32 68S33,10S37,14S38 82S34 76S35,10S36 36S36,23S35,27S31 74S37 152 78 152 78 0.0 0.1 0.0 8.4 0.75 0.75 ± 101 # 92S38 62S39,28S30 540 518 21.2 6.2 0.68 527 373 271 257 221 189 174 166 353 257 244 210 189 172 159 10.3 13.6 6.5 0.6 0.4 0.0 0.2 3.6 1.0 1.4 1.6 2.6 1.4 0.4 0.48 0.67 0.71 0.52 0.75 0.69 0.66 365 p 273 p 267 p 218 1958 175 p 164 p n 38 methyl torsion n 39 asymmetric torsion 158 70 158 70 0.0 0.1 0.1 7.7 0.69 0.75 ± n 31 n 32 n 33 n 34 n 35 n 36 n 37 Ab Fixed IR initio a scaled b int. c 22S24,41S28, 15S26, 11S33 46S25,45S22 57S26,30S24 41S27,13S23,29S29 46S28,34S24,17S26 CH3 rock CH3 rock SiC stretch SiC2 antisymmetric stretch SiC2 symmetric stretch C±H out-of-plane bend CCSi bend SiC2 wag SiC2 rock SiC2 deformation methyl torsion CSiF bend SiC2 twist n 30 PED 95 # Calculated with the MP2/6-31G(d) basis set. Scaling factors 0.9 for stretching and bending coordinates and 1.0 for torsional coordinates. c Calculated infrared intensities in km/mol from the MP2/6-31G(d) calculation. d Ê 4/amu and dp ratios, from RHF/6-31G(d) calculation. Calculated Raman activities in A e Frequencies are obtained from the annealed spectrum of the nitrogen matrix except for the ones with ( p), ( X), ( 1), (8), and ( #) signs are taken from infrared gas, infrared solid, argon matrix, Raman liquid and Raman solid, respectively. b A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 n 24 CH3 rock n 29 a Syn e A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Table 5 Temperature and intensity ratios for conformational study of dimethyl vinyl ¯uorosilane dissolved in liquid xenon T (8C) 1000/T (K) I710/I695 2ln K 255 260 265 270 275 280 285 290 295 2100 DH a 4.58 4.69 4.80 4.92 5.05 5.18 5.31 5.46 5.61 5.78 2.8438 2.9173 2.8899 2.9631 3.0203 3.0587 3.0263 3.0630 3.1828 3.0814 21.1647 21.1651 21.1774 21.1725 21.1398 21.1794 21.1999 21.12601 21.2459 21.2813 53 ^ 9 a DH 53 ^ 9 cm 1 (0.64 ^ 0.10 kJ/mol) with conformer the more stable form. the gauche conformer remains in the annealed solid with a minor amount of the gauche conformer (Figs. 5±7). Further support for this conclusion is found from the assignments for the CH2 rock and CCSi bend which are observed at 966 and 370 cm 21 for the gauche conformer. Similarly, the other listed bands, which disappear upon solidi®cation and annealing, are all assigned to the gauche conformer and they will be discussed latter. Therefore, all the spectral data indicate that the syn form is the stable conformer in the annealed solid. Variable temperature studies of the infrared spectra of DVFS dissolved in liquid xenon (Table 5) and krypton (Table 6) were conducted to determine the 265 enthalpy difference between the two stable conformers. An important advantage to this temperature study is that the conformer peaks are better resolved and the area under them is more easily measured than bands observed in the infrared spectrum of the gas. Infrared spectral data from 4000 to 400 cm 21 were obtained at different temperatures between 255 to 21008C for the xenon solution and between 2105 and 21508C for the krypton solution. The spectral changes in lique®ed krypton of the pairs at 962/965, 710/695 and 522/513 cm 21 are shown in Fig. 11. From all spectral data from lique®ed xenon and krypton solutions we observed increases in the intensity of the infrared bands assigned to the gauche conformer as the temperature decreases. This clearly con®rms the stability of the gauche rotamer over the syn conformer in these rare gas solutions. In order to obtain the enthalpy difference, ten spectral data points were obtained over the temperature range 255 to 21008C for the xenon solution and 2105 to 21508C for the krypton solution. The intensities of each conformer pair were ®t to the equation 2ln K DH=RT 2 DS=R where K is the intensity ratio (Ig/Ic) and it is assumed that DH is not a function of temperature. Using a least squares ®t of the slope of the line, a DH value of 53 ^ 9 cm 21 was obtained from the 710/695 cm 21 bands from the xenon data. The pair at 962/965 cm 21 was not resolved suf®ciently to be measured in the temperature range of the xenon measurements and, therefore, not utilized in the calculation. Since the signal-to-noise ratio was relatively low in the range of 600±500 cm 21, we were Table 6 Temperature and intensity ratios for conformational study of dimethylvinyl ¯uorosilane dissolved in liquid krypton T(8C) 1000/T (K) I962/I965 2ln k I695/I710 2ln k I51/I513 2ln k 2105 2110 2115 2120 2125 2130 2135 2140 2145 2150 DH a (cm 21) 5.95 6.13 6.32 6.53 6.75 6.99 7.23 7.51 7.80 8.12 2.5385 2.7410 2.7051 2.8240 2.7941 2.8300 2.7710 2.8501 2.8876 2.9420 20.9316 21.0083 20.9951 21.0382 21.0275 21.0403 21.0192 21.0473 21.0604 21.0791 32 ^ 7 3.2050 3.2062 3.2460 3.2300 3.1262 3.2525 3.3200 3.5256 3.4760 3.6014 21.1647 21.1651 21.1774 21.1725 21.1398 21.1794 21.1999 21.12601 21.2459 21.2813 39 ^ 8 1.9853 2.0637 2.1420 2.3272 2.4425 2.4469 2.4657 2.3350 2.4022 20.6858 20.7245 20.7617 20.8447 20.8930 20.8948 20.9025 20.8480 20.8764 a Average DH 44 ^ 7 cm 21 (0.53 ^ 0.09 kJ/mol) with the gauche conformer the more stable form. 69 ^ 21 266 A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 Fig. 11. Temperature dependent infrared spectra of dimethyl vinyl ¯uorosilane. unable to measure DH from the pair at 522/513 cm 21. Utilizing the krypton data for the above three conformer pairs (Fig. 11), DH values of 32 ^ 7, 39 ^ 21 and 69 ^ 21 cm 21 were obtained with an average value of 44 ^ 7 cm 21 (0.53 ^ 0.09 kJ/mol) with the gauche form the more stable conformer. Additional infrared spectra of DVFS were recorded in argon and nitrogen matrices (1:500 and 1:1000) deposited at 5 and 15 K; the spectra in the argon matrices are given in Fig. 6, whereas detailed spectra in a nitrogen matrix are presented in Fig. 7. Supposedly, the conformational equilibrium of the vapor phase is maintained when the gas mixture is shock frozen on the CsI window at 5 or 15 K, provided the barrier to conformational equilibrium is above 3 and 5 kJ mol 21, respectively. When the matrices were annealed below 20 K some small spectral changes occurred, which were interpreted as a relaxation of DVFS in the matrix lattice. The samples were subsequently annealed for 10 min at every 3 K between 20 and 37 K for argon and 20 and 34 K for the nitrogen matrices before being recooled to 5 K and the spectra recorded. Prominent changes were observed when the samples were annealed to ca. 34 K for argon and 32 K for nitrogen. Certain bands were enhanced, others diminished in intensities after annealing which are interpreted as a displacement of the conformational equilibrium. Qualitatively, the same intensity changes occurred in the spectra of both matrices, but they were more prominent in the spectra obtained from the argon matrix. As we shall see, the infrared bands which vanished or were reduced in intensity in the spectra of the crystal were enhanced in the matrices after annealing, whereas those present in the crystal spectra were reduced in intensities. These bands are indicated with arrows pointing upwards or downwards, respectively, in Table 1. The observed annealing temperatures (34 and 32 K) suggested that the conformational barrier is ca. 9 kJ mol 21 from the curves given by Barnes [19]. However, it can be seen from the spectral data shown in Figs. 5±7 that no infrared bands disappeared completely in the spectra of the matrices after annealing. Therefore the enthalpy difference between the conformers must be quite low in both matrices since an equilibrium was maintained at the annealing temperatures at 32±34 K. A rough estimate using a simple Boltzmann distribution suggests an enthalpy difference of 17±33 cm 21 in the matrices which is in agreement with the results from the xenon and krypton solutions in con®rming that the gauche conformer is the more stable form. Upon crystallization, bands at 969, 876, 711, 607, 366, 228 and 201 cm 21 presented in the Raman spectra of the liquid and amorphous phases signi®cantly diminish and/or disappeared. These bands are due to the second conformer and seven spectra data points of the liquid were recorded between 19 and 21208C in order to obtain the enthalpy difference. Band pairs at 712/696 and 371/396 cm 21 were A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 21 selected to obtain DH. The band at 696 cm might have a contribution from the other conformer. A series of van't Hoff plots based on measured peak heights were obtained giving DH values of 17 and 33 cm 21, respectively, for the pairs mentioned above with the average value of 25 ^ 15 cm 21 (0.30 ^ 0.18 kJ/mol) also with the gauche form being the more stable conformer. 5. Vibrational assignment The conformational analysis of DVFS shows that the molecule exists in two stable conformations in the ¯uid phase. The syn conformer has Cs symmetry and the 39 fundamentals will span the irreducible representation 23 A 0 and 16 A 00 , whereas for the gauche conformer with C1 symmetry all the fundamentals belong to species A. Since most of the observed bands in the infrared spectrum of the vapor and the Raman bands of the liquid are common to both the syn and gauche conformers, the band contours and the polarization ratios are of limited help in the spectral interpretation. For the sake of similarity, the fundamentals of both the gauche and syn conformers have been numbered consecutively, instead of the conventional numbering of the modes belonging to species A 0 before those of A 00 in the syn conformer. Guided by the assignments of the similar normal modes for dimethyl vinyl chlorosilane [7] and also by the calculated spectral intensities and predicted wavenumbers from ab initio calculations, we propose the vibrational assignments listed in Table 1. The assignments of the carbon±hydrogen modes have been previously reported [7] for dimethyl vinyl chlorosilane, and with only minor wavenumber shifts, they remain essentially the same for DVFS. The spectra of these two compounds look similar down to about 1000 cm 21. In the region below 1000 cm 21, a few features were observed in the infrared and Raman spectra of the gas or liquid which disappear upon crystallization. The C-type Q-branches located at 969 and 964 cm 21 in the infrared spectrum of the gas are assigned to the CH2 rocks with the syn conformer having the higher wavenumber. The intensity of this band increases with decreasing the temperature of the krypton (Fig. 11) solution 267 con®rming that the gauche form is the more stable conformer in the gas phase. The Si±C stretch is assigned to the C-type Q-branch at 710 cm 21 for the gauche form and to the shoulder at 694 cm 21 for the syn conformer where the former has almost disappeared from the spectrum of the crystalline solid. The later one is also assigned to the SiC2 antisymmetric stretch for the both conformers. It should be noted that the gauche conformer has a minor contribution to this band but we used it as a conformer band (Fig. 11) for the enthalpy determination. The CH outof-plane bending conformer pair is observed in the infrared spectrum of the liquid krypton solution at 521 and 513 cm 21 for the gauche and syn conformers, respectively. The CCSi bending mode observed at 365 cm 21 for the gauche form is evident in the far infrared spectra of the gas and amorphous solid but disappears from the spectrum upon annealing (Fig. 6) the sample. In the amorphous and crystalline states the infrared and Raman bands from 258 to 286 cm 21 are assigned to SiC2 wag and SiC2 rock. The ab initio calculations predict very weak infrared bands as well as weak Raman lines for the normal modes below 225 cm 21 and the assignments of these remaining ®ve fundamentals n 34 ± n 38, are less certain as shown in Table 1. The Raman bands at 225 and 209 cm 21, appearing as shoulders and their corresponding very weak infrared counterparts, are assigned to the SiC2 deformation for the gauche. However, the intense peak at 195 cm 21 in the Raman spectrum of the liquid is assigned to both the SiC2 deformation for the syn conformer and the methyl torsion for the gauche form. The n 36 and n 35 0 modes are observed as a very weak infrared band at 175 cm 21 in the spectrum of the vapor for the gauche and syn conformers, respectively. The weak infrared band of the vapor at 170 cm 21 corresponds to the infrared and Raman bands around 178 cm 21 in the condensed phases and is attributed to n 37 of the syn rotamer and the very weak infrared band at 164 cm 21 (spectrum of the gas) is assigned to the gauche for the same normal mode. There are no bands observed for n 38, the methyl torsion. The asymmetric torsional mode is assigned to the two bands at 101 and 95 cm 21, which were observed in the Raman spectrum of the liquid and calculated at 78 and 70 cm 21 for the syn and gauche conformers, respectively. 268 A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 6. Discussion In the present study of DVFS, neither the infrared band conformer of the gas nor the depolarization ratio of the liquid was helpful in the determination of the conformational stability of this molecule. Additionally, the calculated energies for the syn and gauche rotamers derived from the ab initio calculations have large uncertainties, but the force constants and the wavenumbers after appropriate scaling usually give good agreement with those of the observed fundamentals. Within the group frequency regions for the CH3 and CH2 stretching and deformation vibrations, the calculated wavenumbers for the syn and gauche fundamentals are usually separated by less than 5 cm 21 (Table 2). Below 1450 cm 21 there are 12 instances in which the fundamentals of the syn and gauche conformers are separated by more than 5 cm 21. The largest shifts are calculated for the six fundamentals n 27, n 28, n 30, n 31, n 32 and n 37 in which they should be larger than 10 cm 21. As expected, most of the observed syn/gauche conformer separations predicted pairs are located approximately with the wavenumber. The following observed band pairs from the Raman spectrum of the liquid: 971/963, 883/864, 712/696, 621/613, 371/396, 275/265, 225/209 and 195/ 187 cm 21 in which the high frequency bands vanished (or were reduced in intensity) in the infrared and Raman spectra of the annealed solid are correlated with the calculated scaled wavenumbers of the syn and gauche conformers, respectively (Table 4). It was found from all eight of these pairs that the high frequency bands can qualitatively be ®tted with the predicted wavenumbers of the gauche form and the remaining bands to those of the syn conformer whereas the opposite interpretation is not feasible. The experimental and calculated frequencies are also in reasonable agreement for the n 21, n 27, n 28, n 31, n 33, n 34 and n 35 fundamentals. Two band pairs have less convincing assignments where for n 22 the observed difference is 19 cm 21 and that calculated only 1 cm 21 and n 30 which was assigned to overlapping bands at 515 cm 21 although a difference of 11 cm 21 was predicted. However, we feel that there are compelling reasons to attribute the vanishing bands to the gauche conformer, which means that the syn conformer remains in the crystal even though the gauche conformer has lower energy than the syn rotamer in the rare gas solutions, the liquid state and also in the matrices. In addition to the band pairs assigned to the syn and gauche conformers on the basis of spectral changes on crystallization, described above, the infrared spectra in argon and nitrogen matrices can frequently give clues to close lying conformer pairs. Thus, the band pairs attributed to n 10, n 15, n 17, n 19, n 23, n 25, n 29 and n 30 observed at 1601, 1411, 1254, 1013, 848, 772, 521 and 515 cm 21, respectively, can all be tentatively assigned to separate conformer bands. As apparent, each of these pairs of bands are characterized by intensity changes in one or both matrices after annealing and they are indicated with arrows in Table 1. The bands, which are enhanced after annealing, are assigned to the gauche form, whereas those that diminish in intensity to the syn conformer. It is characteristic that in each of these pairs the wavenumber difference in the matrices is small (between 2 and 8 cm 21) leading to overlapping bands in the ¯uid phases. Since the bandwidths are much lower in the matrices the separate bands due to the syn and gauche conformers can be detected. Because of matrix, effects frequently encountered during annealing some of these assignments may be erroneous. However, if the same general features are observed in both matrices we can be fairly con®dent in the experimental results. It should be emphasized that for the additional band pairs n 21 around 964 cm 21 and n 23 at 710 and 694 cm 21 the bands changed both during crystallization and on annealing the matrices and the conclusions should be de®nitive. The conformational energy difference in DVFS can be compared with the corresponding value for related silanes. In dimethyl vinyl chlorosilane [7] CH2y CHSi(CH3)2Cl, the gauche conformer also has the lower energy, with DH (syn±gauche) equal to 0.5 ^ 0.1 kJ mol 21 (equal to that of DVFS within the experimental uncertainty); however, in the present study the syn form is present in the crystal which is in contrast to the chloro analogue. The vibrational spectra of the two conformers are strikingly similar for DVFS and the corresponding chloro analogue. In vinylsilyl chloride [5] (CH2yCHSiH2Cl) the syn conformer was more stable by 1.2 kJ mol 21, in the liquid state revealing increased stability of the syn conformer when the two methyl groups are absent. A. Horn et al. / Journal of Molecular Structure 554 (2000) 251±269 However, the gauche conformer was present in the crystal [5], as well as in the rare gas solution [6] which makes this molecule quite different to the situation in DVFS. A comparison of the structural parameters predicted by ab initio calculations (Table 2) with all basis sets indicates little difference in the parameters upon conformer changes from the gauche to the syn form. With a given basis sets, for example 6-31G(d) and 6-3111G(d,p) with full electron correlations, the Ê and bond angles bond distances agree within 0.005A within 18 for the corresponding parameters of the two rotamers. Even with the larger 6-3111G(2d,2p) basis set, the differences become insigni®cant for most of the parameters (Table 2). Two halogens (F or Cl) and one methyl group attached to the Si atom leads to methyl vinyl di¯uorosilane (CH2yCHSi(CH3)F2) and methyl vinyl dichlorosilane (CHyCHSi(CH3)Cl2) both of which the conformational stabilities have been investigated. In methyl vinyl di¯uorosilane the gauche conformer is more stable in the rare gas solution [10], but the syn form is present in the crystal [9]. The synconformer was the more stable form in the rare gas solution [20] and this form was also present in the crystal of vinyl dichlorosilane [8], demonstrating the effect of the larger chlorine atom compared to the ¯uorine atom for these similar molecules. It should be noted that in the latter molecules the syn conformer has both halogens in the gauche positions compared to the CyC bond, whereas in the gauche conformer, one of the halogens is situated in the syn position of the CyC bond. Finally, in methyl vinyl silane (CH2yCHSi(CH3)H2) the single methyl group prefers the gauche conformation [21] with a DH equal to 1.59 ^ 0.13 kJ mol 21 but the syn conformer is present in the crystal [22]. If the energy difference between the conformers is low as observed for DVFS, the high energy conformer, which frequently has the larger dipole moment, can be preferred in the crystal because of larger crystallization energies for that conformer which can overcome the higher conformational energy. Acknowledgements J.R.D. acknowledges the University of Kansas City 269 Trustees for a Faculty Fellowship award for partial ®nancial support of this research. References [1] V. Govern, L. Khristenko, Yu.A. Pentin, Vopr. Stereokhim. 2 (1972) 57. [2] L.V. Khristenko, Yu.A. Pentin, Vestn. Mosk. Univ. Ser. 2 Khim. 31 (1976) 304. [3] M. Imachi, J. Sci. Hiroshima Univ. Ser. A. 42 (1978) 43. [4] J. Sullivan, M. Qtaitat, J. Durig, J. Mol. Struct. (Theochem) 202 (1989) 159. [5] J. Durig, J.F. Sullivan, G.A. Guirgis, M.A. Qtaitat, J. Phys. Chem. 95 (1991) 1563. [6] J.R. Durig, Y.E. Nashed, M.A. Qtaitat, G.A. Guirgis, J. Mol. Struct. 525 (2000) 191. [7] G. Guirgis, Z. Shen, M. Qtaitat, J. Durig, J. Mol. Struct. 403 (1997) 57. [8] J. Durig, G. Guirgis, Y. Kim, W. Yan, M. Qtaitat, J. Mol. Struct. 382 (1996) 111. [9] J.R. Durig, G.A. Guirgis, M.A. Qtaitat, J. Raman Spectrosc. 26 (1995) 413. [10] Y.E. Nashed, G.A. Guirgis, J.R. Durig, unpublished data. [11] F. Miller, B. Harney, Appl. Spectrosc. 24 (1970) 291. [12] M.J. Frisch, G.W. Trucks, H.B. Schlegel, P.M.W. Gill, B.G. Johnson, M.A. Robb, J.R. Cheeseman, T. Keith, G.A. Petersson, J.A. Montgomery, K. Raghavachari, M.A. AlLaham, V.G. Zakrzewski, J.V. Ortiz, J.B. Foresman, J. Cioslowski, B.B. Stefanov, A. Nanayakkara, M. Challacombe, C.Y. Peng, P.Y. Ayala, W. Chen, M.W. Wong, J.L. Andres, E.S. Replogle, R. Gomperts, R.L. Martin, D.J. Fox, J.S. Binkley, D.J. Defrees, J. Baker, J.P. Stewart, M. HeadGordon, C. Gonzalez, J.A. Pople, gaussian 94, Revision D. 2, Gaussian, Inc., Pittsburgh PA, 1995. [13] P. Pulay, Mol. Phys. 17 (1969) 197. [14] J.H. Schachtschneider, Vibrational Analysis of Polyatomic Molecules, Parts V and VI, Technical Report Nos. 231 and 57, Shell Development Co. Houston, TX, 1964 and 1965. [15] M. Frisch, Y. Yamaguchi, J. Gaw, H. Schaefer III, J. Binkley, J. Chem. Phys. 84 (1986) 531. [16] R. Amos, Chem. Phys. Lett. 124 (1986) 376. [17] A. Anderson (Ed.), The Raman Effect, vol. 1, Marcel Dekker, New York, 1971. [18] P. Polavarapu, J. Phys. Chem. 94 (1990) 8106. [19] A. Barnes, J. Mol. Struct. 113 (1984) 161. [20] G.A. Guirgis, P. Zhen, J.R. Durig, Spectrochim. Acta (2000) (in press). [21] Y. Jin, G.A. Guirgis, J.R. Durig, submitted for publication. [22] J. Durig, J. Sullivan, M. Qtaitat, J. Mol. Struct. 243 (1991) 239.