摘要:
1624 J.C.S. Dalton Reactions of Tin(iv) Chloride with Silyl Compounds. Part 1. Reactions with Inorganic Silyl Compounds By Stephen Cradock E. A. V. Ebsworth and Narayan Hosmane Chemistry Department Edinburgh Uni-The reactions between SnCI and (SiH,),X have been investigated [n = 1 X = H F CI Br SiH, Mn(CO)5 or Co(CO),; n = 2 X = 0 or S ; n = 3 X = N or P] ; the reaction with (Me,SiH),O was also investigated. With SiH, Si,H6 SiH,F SiH,CI [SiH,Mn(C0)5] and (SiH,),N reaction led to monochlorination at silicon and spectroscopic parameters for the compounds Si H ,FCI [Si H ,CI M n (CO) 4 ClSi H,N (Si H ,) , (ClSi H *) ,N Si H and (CISiH,),N all prepared by this route are reported. For the amines 29Si and 15N chemical shifts were measured by heteronuclear double resonance. Reactions with other silyl compounds led to the formation of SiH,CI although some evidence was obtained for the formation of intermediate trichlorostannyl species.versity West Mains Road Edinburgh EH9 3JJ STANNIC chloride is a useful chlorinating agent for silanes and germanes. We have already published a prelimin-ary account of its reaction with SiH,. Here and in a subsequent paper we report the reaction of SnC1 with a variety of silyl and organosilyl compounds. EXPERIMENTAL Reactions were carried out using a conventional vacuum system; because of the solubility of SnCl in grease, reaction vessels and one trap-section were fitted with ‘ Sovirel ’ greaseless taps but the main part of the system was built using greased taps. Volatile reaction products were identified using measurements of i.r.and/or n.m.r. spectra and from their molecular weights (measured by exact mass spectroscopy). 1.r. spectra were obtained by means of Perkin-Elmer 457 and 225 double-beam spectro-meters Raman spectra using a Cary 83 instrument operat-ing with the argon-ion laser line a t 488 nm mass spectra using an A.E.I. MS902 and n.m.r. spectra using a Varian HA 100 spectrometer operating a t 100 MHz for protons or 94 MHz for fluorine whose probe was double-tuned to accept a second radiofrequency generated by a frequency synthe-siser. Reactions of SnCl,.-(a) With SiH,. Monosilane (0.47 mmol) and stannic chloride (0.50 mmol) were allowed to warm together to room temperature in a vessel of volume ca. 35 ml; a thin white film formed slowly on the walls of the reaction vessel.After 18 h SiH (0.32 mmol) was recovered, with HCI (0.15 mmol) SiH,Cl (0.11 mmol) SiH,CI (0.04 mmol) and SnC1 (0.35 mmol). The white solid residue was shown by X-ray powder photography and Raman spectroscopy to be SnCl (0.029 g 0.15 mmol). In a second experiment SiH (3 mmol) was allowed to react with SnCl (0.5 mmol) at room temperature (18 h) ; unchanged SiH (2-5 mmol) was recovered with HCl (0-5 J. E. Bentham S. Cradock and E. A. V. Ebsworth Inovg. Nucleav Chevn. Letters 1971 7 1077. mmol) SiH,Cl (0.5 mmol) and SnC1 (0.5 mmol). KO SiH,Cl was detected. (b) TVzth SiH,Cl. Monochlorosilane (0.47 inmol) was set aside with SnC1 (1.03 mmol) at room temperature (24 h) in a reaction vessel (35 ml) fitted with a Sovirel tap.A thin white film formed very slowly on the walls of the vessel. Fractional condensation of the volatile products gave HCl (0.13 mmol) SiH,Cl (0- 13 mmol) unchanged SiH,Cl (0.34 mmol) and unchanged SnC1 (0.88 mmol) ; SnCl (0.13 mmol) remained in the reaction vessel. Monofluorosilane (1.00 mmol) and SnCl, (1.17 mmol) were kept together at room temperature (18 11) in a reaction vessel (35 ml) fitted with a Sovirel tap. A thin white film formed slowly on the walls of the vessel; the reaction products consisted of HC1 (1.00 mmol) un-changed SnCl (0.11 mmol) and a mixture of SiH,Cl,, SiH,F, and SiH,ClF. No unchanged SiH,F was recovered. T t was impossible to separate quantitatively SiH,ClF from SiH,Cl and SiH,F, but careful fractional distillation and condensation gave SiH,F (0.3 mmol) SiH,Cl (0-3 mmol), and SiH,CIF (0.4 mmol) .Chlorofluorosilane was volatile in vacuo at - 120 “C but involatile a t - 130 “C; the mass spectra gave parent ions a t 83.959229 (Calc. for 28SiH,-35ClF 83.959833) and 85.956166 (Calc. for 28SiH,37C1F: 85.956884). The 1H spectra (Me,Si solvent and standard) consisted of a doublet [T = 4.79 p.p.m; ,J(HF) = 56 Hz], and the 19F spectra (CCl,F solvent and standard) of a triplet [6(F) = - 146.4 p.p.m; 2J(HF) = 56 Hz]. The i.r. spectrum of the vapour gave peaks at the following fre-quencies (cm-l; point group assumed to be C,; capital letters refer to type of band contour) 2240~s C(vSiH a”) 2200s A (uSiH a‘) 978m A/B hybrid (SSiH, a’) 952s A/B hybrid (GSiH, a’) 922vs A (vSiF a’) 735w C (SSiH, u”) 649m C (SSiH, a”) 570vs A (uSiC1 a’) and 235w A (GFSiC1 a’).The compound was found to disproportionate slowly, giving SiH,CI and SiH2F,. 2 D. E. J. Arnold J. S. Dryburgh E. -4. V. Ebsworth and D. W. H. Rankin J.C.S. Dalton 1972 2518. (c) With SiH,F 1975 1625 (d) With SiH,Br. Bromosilane (1.00 mmol) was allowed t o react with SnC1 (1.00 mmol) at room temperature (1 h) ; a white solid formed in the reaction vessel. The products consisted of SiH,Cl (1.00 mmol) and a mixture of chloro-bromo-derivatives of SnIV some of which were volatile and liquid a t room temperature. (e) With Si,H,. Disilane (1.00 mmol) and SnC1 (1-01 mmol) were allowed to react together at room temperature (24 h) in a reaction vessel (35 ml) fitted with a Sovirel tap. A white solid began to form almost immediately.The vola-tile products consisted of HCl (1.00 mmol) unchanged Si,H, (0.52 mmol) Si,H,Cl (0.25 mmol) and ClSiH,*SiH,Cl (0.23 mmol). The halogenodisilanes were identified by their i.r. and n.m.r. spectra. A residue of SnCl (1-01 mmol) re-mained. If SiH,Cl-SiH,Cl was allowed t o stand a t room temperature ( > 24 h) decomposition gave a small propor-tion of SiH,SiHCl,. In a second reaction Si,H (3.06 mmol) was allowed to react with SnC1 (0.50 mmol) a t room temperature (18 h ) ; unchanged Si,H (2.50 mmol) was recovered with HC1 (0.60 mmol) Si,H,Cl (0.5 mmol) and SnC1 (0.5 nimol). No Si,H4C1 was detected. (f) Will8 (SiH,),O. Disiloxane (0.50 mmol) and SnCl, (0.50 niniol) were kept a t room temperature (6 min) ; a thin film of white solid formed on the walls of the vessel and the liquid bubbled vigorously.The volatile products consisted of HC1 (0.50 mmol) and SiH,Cl (0-5 mmol). The involatile solid residue evolved no additional volatile material during a further 15 min; its i.r. spectrum contained bands at2160w, 1 l.iOsh 1 lOOs 1070s 975m 940m 920~7 875sh 830s 520w, and 410w cm-l those a t 2160 1100 520 and 410 cm-l im-plying the possible presence of Si-H Si-0 Si-C1 and Sn-C1 bonds in the material. In a second experiment disiloxane (0.50 mniol) was al-lowed to react with SnC1 (0.50 mmol) a t - 46 "C (2 h). The only volatile product obtained a t that temperature was SiH,C1 (0.50 mmol). When the reaction mixture was allowed to warm to room temperature (30 rnin) HCl (0.50 nimol) was evolved and a white involatile solid remained.JVhen disiloxane (0.25 mmol) and SnC1 (0.25 inmol) were sealed in an n.m.r. tube in cyclopentane solvent (- 46 "C 2 h), single peaks were observed at 7 5.41 (SiH,CI) and 5.29. The spectrum was unchanged after the tube had been allowed to stand a t 28 "C for 45 niin. (8) With (Me,SiH) 20. Tetrametliyldisiloxane (0.50 mmol) and SnCI were allowed to warm together for 3 min. A vigorous reaction took place and a thin film of white solid formed on the walls of the reaction vessel. The only vola-tile product recovered a t this stage was Me,SiHCl (0-50 mmol). The residue was kept a t room temperature for a further 10 min; HC1 (0.50 mmol) was evolved; the i.r. spectrum of the solid residue contained bands a t 1260s, 1085s 1025sh 810 720m 560w and 390w cm-l.In a second experiment (Me,SiH)10 (1-00 mmol) was allowed to react with SnC1 (1.00 mmol) a t -29 "C (1 h ) ; the only product volatile at that temperature was Me,SiHCl (1.00 mmol); when the residue was allowed to warm to room temperature (20 min) HC1 (1.00 mmol) was evolved. When (Me,SiH),O (0-25 mmol) was sealed with SnCl (0.25 niniol) in an n.m.r. tube in cyclopentane as solvent the spectrum observed initially (- 50 "C) consisted of peaks [T 9-85 (doublet) and 5.27 (heptet)] due to (Me,SiH),O. After 1 h a t -22 OC these peaks had disappeared; the J. E. Drake and N. Goddard Inorg. Nuclear Chem. Letters, 1968 4 386. spectrum of Me,SiHCl [Y 9-55 (doublet) and 5.11 (heptet)] was observed with an additional doublet (T 9.3) and heptet (T 5.29) due to some other Me,SiH derivative.After 30 rnin a t +2S OC the spectrum of the unidentified species had disappeared though that of Me,SiHCl was unaffected; a new singlet (7 9-91> was present instead. Disilyl sulphide (1.0 nimol) was al-lowed to react with stannic chloride (1.0 mmol) a t - 46 "C (2 h). No solid film was observed on the walls of the reac-tion vessel. The only material volatile a t -46 "C was SiH,Cl (1.0 mmol); this was removed. After 30 rnin a t room temperature additional SiH,Cl (1.0 mmol) was evolved ; a thin white film of involatile solid remained on the walls, and an involatile yellow residue at the bottom of the vessel. Neither the white nor the yellow solids showed strong i.r. bands in the region 4000-400 cni-l; all the SiH groups are accounted for as SiH,Cl and the involatile residues may be identified respectively as SnC1 and S.An n.ni.r. study of the reaction failed to reveal the formation of any inter-mediate although the evolution of 1 mmol of SiH,Cl a t -46 "C does imply that SiH,SSnCl may have been formed. Trisilylamine (1.0 mmol) was allowed to react with stannic chloride (1.00 mmol) a t room tempera-ture (15 min). A white solid formed immediately. Frac-tional distillation of the volatile products gave SiH,Cl (0.82 nimol) unchanged trisilylamine (0.27 mniol) and three chlorosilylamines ClSiH,W (SiH,) (0.2 mmol volatile at -64 "C) (ClSiH,),NSiH (0.15 mmol volatile a t -46 "C), and (ClSiH,),N (0.1 mmol involatile at - 46 "C). The com-pounds were identified by their mass and n.ni.r.spectra; for the n.m.r. spectra the operation was repeated using trisilyla-mine prepared from 957h IjNH,. The exact masses of the molecular ions are given in Table 1 and the n.m.r. para-meters are summarised in Table 2. It was possible by start-ing with an excess of either (SiH,),N or SnC1 to prepare (h) With (SiH,),S. (i) With (SiH,),N. TABLE 1 Exact masses of parent ion of chlorosilylamines 0 bserved 140.965004 142.962 144 174.926011 176.923180 178.920312 208.886433 2 10.883342 2 12.879640 2 1 4.8 7 7434 Calculated 140.965311 142.962361 174.926339 176.923389 178,920439 208-88 7 364 2 10.8844 18 2 12.881468 2 14.8785 18 pure samples of ClSiH,N(SiH,) and (ClSiH,),N but (ClSiH,) ,-NSiH was only obtained as a mixture with one of the other chlorosilylamines.The i.r. and Raman frequencies with tentative assignments are given in Ta.bles 3 and 4 and are discussed below. (j) WitJz (SiH,),P. Trisilylphosphine (0.2 mmol) and SnCl (0.6 mmol) were allowed to react together a t room temperature (3 min) . A yellowish solid formed immediately. The only volatile product was SiH,Cl (0.6 mmol); the yellowish solid was not identified. (k) With [SiH,Mn (CO) 5 ] .-Pentacarbonylsilylmanganese (0.99 mmol) and SnC1 (1.06 mmol) were allowed to warm to room temperature together. A white solid began to form immediately on the walls of the reaction vessel. After 25 min at room temperature the volatile products were found to consist of HCl (10.4 mmol) SiH,Cl (0.10 mmol), SiH,Cl (0.04 mmol) and [ClSiH,Mn(CO),] (0.82 mmol) 1626 J.C.S.Dalton Pentacarbonylchlorosilylmanganese was identified by its lH n.m.r. spectrum [7 = 4-60; lJ (,OSiH) = 214 Hz] its mass spectra (for parent ions exact masses 259.874050 and 261.871 100; 35ClSiH2MnC,05 requires 259.874102 ,'ClSiH,-MnC,O requires 261.871152) and its vibrational spectra. ducts were found to consist of SiH,CI (1.14 mmol) and un-changed SnCl (0.04 mmol) ; neither [HCo(CO)J nor [CISiH,-Co(CO),] were detected among the products. The solid residue (ca. 0.1 g) was extracted with benzene; the solution gave a single 13C resonance at 191.1 p.p.m.; from the solution TABLE 2 Parameters for the lH n.m.r. spectra of chlorosilylamines (SiH,) ,N CISiH,N(SiH,) (ClSiH,) ,NSiH (ClSiH,),N T(SiH,) 4.92 4.84 4.79 T( SiH,) 5-68 5.57 5.54 6(29SiH,) /p.p.m.- 17.7 - 17.2 - 20.2 6(,OSiH3) /p.p.m. - 39.9 - 39.8 - 47.3 6(15N)/p.p.m. b - 79.9 - 51.3 -22.1 + 1.2 ' J ( 29SiH2)/Hz 258.0 260.0 264.0 2J(15NSiH,)/Hz 6.2 6.1 6.0 1J(29SiH,) /Hz 212.0 212.0 224.0 ,J( lSNSiH,)/Hz 4.2 4.0 4-0 a Relative to Me,Si; +ve to high frequency. Relative to Me,N; +ve to high frequency. TABLE 3 Raman spectra of chlorinated trisilylamines (liquid) ClSiH,N(SiH,) (CISiH,) ,NSiH (CISiH,),N - c-Frequency (clcm-l) 218Ovvs br 998s 946vs br 880w 747 f 2m 702s 7 Pol. ratio P P dP dp ? dP dP Frequency POI: (P/cm-l) ratio 2216 & 2vs p 996w P 970w dP 943m dP 907w dP 746s dP 870mv br dp Frequency (s/cm-l) 2220vs 1 ooovw 945w 91ovw 878 f 2w 745m .Pol. ratio Assignment p vSiH p vSiN(asym) dp? GSiH and/or SiH, dp? SiH, dp ? dp pSiH and/or SiH, dP 536m dP 480vs P 680m dP 49ovvs P 285m P? 358w dP 230m dP 19Ow br dp 160s P 628vw 580s P ? 585s p vSiCl(sym) 550m dp vSiCl (asym) 482vs p vSiN(sym) 322s SiNSi bend ? 278m :F ClSiN bend 140s p? ClSiN bend Where v = very s = strong m = medium w = weak sh = shoulder br = broad p = polarised and dp = depolarised. TABLE 4 Infrared spectra of chlorinated trisilylaniines CISiH,14N(SiH,) CISiH,16N(SiH,)2 Vapour Frequency FFr$:7 Assignment (Flcm-l) Assignment 2195 f 1.0m} vSiH 2170s I m sw 922s vSiN(asym) 978s vSiN(asym) 941vs GSiH ::!$} GSiH, 887s SSiHt 883s GSiH, 740w br pSiH 738w br pSiH, 550w vSiCl 5 .5 0 ~ vSiCl 488vw vSil'N(syni) 480 lvw vSilW(sym) (ClSi H,) ,l'NSiH (Cl SiH,) WSiH (CISiH2),14N (ClSiH,),16N Vapour Vapour A 7 f L 'I Frequency Frequency Frequency Frequeuc y (P/cm-1) Assignmeqt (P/cm-l) Assignment (tlcrn-') Assignment (vlcm-l) Assignment vSiH :;;Eh} vSiH 2210s vSiH 2210s vSiH 1002s vSiN(asym) 9i9s vSiN(asym) 990.5s vSiN(asym) 969.5s vSiN(asym) N:;;} SSiH and/ W;;;} SSiH and/ 949s SSiH 949s GSiH, or SiH or SiHz 8i6.5vs} 8 76.5"s) ;:;} pSiH and/ 540 & l w pSiH and/ 741 &- l w pSiH 741 6 w pSiH, 572 lsh vSiCl 572 & lsh vSiCl 584w vSiCl 584w vSiCl 542m vSiCl 342x1 vSiCl 548m vSiCl 548m vSiCl 483vw vSi14N(sym) 483vw vSP6N(sym) 498vw vSilP(sym) 498vw vSiI6(sym) or SiH2 710 5 2sh) or SiH, Where v = very s = strong m = medium \v = wcak sh = shoulder and br = broad.The i.r. and Raman spectra do not seem to have been reported in detail before and are given in Table 4 with assignments. [Cl,SiHMn(CO) 5 ] among the products. mmol) was allowed to warm to room temperature with SnCl (1.20 mmol) ; a yellowish-white solid began to form a t once. After 5 min at room temperature] the volatile pro-an orange crystalline solid was obtained whose i.r. spectrum was the same as that reported for [ClCo(CO),]. There was no sign of [HMn(CO),] [ClMn(CO),] or DISCUSSION There appear to be two distinct modes of reaction be-(1) With [SiH&o(C0)4]. TetracarbonYlsilYlcobalt (1.14 tween SnCl and silyl compounds. One involves substi-tution: XYZSiH + SnC1,- XYZSiCl + HC1 + SnCl, B.J. Aylett and J. M. Campbell J. Chem. SOC. (A) 1969 s M. Pankowski and M. Bigorgne Corn& rend. 1967 2MC, 1916. 1382 1975 1627 Stannic chloride reacts in this manner with SiH, SiH,F, SiH,Cl (SiH,),N [SiH,Mn(CO),] and Si,H,. Reaction is much slower with the halogenosilanes than with the other compounds perhaps because the process involves electrophilic attack at silicon or SiH; this explains why SiH,Cl*SiH,Cl is formed readily in the reaction with Si,H,, whereas the more stable SiH,SiHCl is only formed in small amounts and probably as a rearrangement product of the symmetrical species. In all these reactions except that with (SiH,),N HC1 is evolved; with (SiH,),N it is presumed that any HC1 reacts with trisilylamine, forming SiH,Cl.From the absence of SiH,Cl it must be deduced that HCl reacts much more rapidly with (SiH,),N than with the chlorosilylamines. The other mode of reaction can be regarded as an exchange process of the form : SiH,X + SnCl + SiH,Cl + XSnCl, When X = Br this may be a reasonable representation of the intial reaction although there is probably halogen redistribution leading to the formation of a mixture of chlorobromostannanes. When X = Co(CO), the initial product may be [Cl,SnCo(CO),] but this decomposes a t least in part to give [ClCo(CO),]. Similarly with (SiH,),S and (SiH,),P all the SiH groups are evolved as SiH,Cl; in the case of the sulphide reaction with equimolar proportions gave no other volatile products, and it is probable that the residue consisted of a mixture of sulphur and SnCl,.With (SiH,),P the molar reacting ratio phosphine SnC1 was 1 3; despite this the sole volatile product was SiH,Cl and no PCl was detected, so the unidentified solid residue may have been a compound containing SnCl groups bound to P. With disiloxane both exchange and substitution seem to have occurred. Initially SiH,C1 was formed when the reactants were allowed to combine at low temperature; however when the residue was warmed to room tempera-ture HC1 was evolved implying that substitution had taken place. This is consistent with a reaction scheme like this: (SiH,),O + SnCl -m SiH,OSnCl + SiH,Cl SiH30SnC13 -m [SiH,ClOSnCl] + HC1 The intermediate trichlorostannyl compound was not detected even a t low temperatures.However in an n.m.r. study of the reaction between (Me,SiH),O and SnCl, resonances due to an unstable species containing Me,SiH groups were observed at low temperatures; this may well have been Me,SiHOSnCl,. Those compounds for which exchange occurs are in general those for which exchange would be expected on the basis of previous studies:6 it has been found that equilibria favour systems in which lighter and more electronegative elements are bound to silicon and heavier and less electronegative elements to germanium and tin. The results with the oxygen and nitrogen compounds are not consistent with this pattern; it is not clear why substitution occurs with the amine but exchange in the initial reaction with disiloxane. Spectroscopic Properties of the Chlorosily1amines.-The 1H n.m.r.parameters of ClSiH,N(SiH,), (ClSiH,),NSiH,, and (ClSiH,),N present no great surprises. The lH chemical shifts are similar those associated with the SiH, groups appearing some 0.6 p.p.m. to low field of those due to the SiH protons; 1J(H29Si) is substantially greater in the SiH than in the SiH groups. There is an unexpectedly large change both in 1J(29SiH) and in 6(29Si) of the SiH groups from ClSiH,N(SiH,) to (ClSiH,),NSiH,. The 29Si chemical shifts of the SiH, groups are less sensitive and the whole range is less than 2-5 p.p.m. The 15N chemical shifts move to high fre-quency by some 25 p.p.m. with the introduction of each successive chlorine atom. In the vibrational spectra it is not possible to make complete assignments.The point group of highest possible symmetry for both C1SiH,N(SiH3) and (ClSiH,), NSiH is Cs whereas in certain conformations (ClSiH,),N would belong to the point group C% or C3h. Bands ap-pear in regions associated with SiH stretching and defor-mation modes; furthermore the bands near 550 cm-l may well be assigned to Sic1 stretching modes. How-ever comparison with the vibrational spectrum of (SiH,),N reveals some interesting features. In trisilyla-mine7 there are two skeletal stretching modes the e’ mode at 990 cm-l and the a’ mode (giving a polarised Raman line) a t 496 cm-l. There are strong lines in the Raman spectra of all three chlorosilylamines at ca. 990 and 490 cm-l and in each compound the band near 990 cm-l shifts to lower frequency when 14N is replaced by 15N.It seems reasonable to assign these two lines in each spectrum to SIN stretching modes though the lower-frequency mode may be mixed with Sic1 stretching in the chlorosilylamines. The problem with this assign-ment is that for l-monochloro- and 1 1’-dichloro-trisilyla-mine three distinct SIN stretching modes would be ex-pected ; even for 1 l’ 1 “-trichlorotrisilylamine it is only in special conformations that the degeneracy of one of the skeletal modes is preserved. It would be necessary to suppose that the presence of the chlorine atoms did not significantly affect the pattern of SiN stretching fre-quencies for the Si,N skeletons. However as pointed out above the Ranian line at 990 cm-l in the spectrum of (SiH,),N though weak is apparently depolarised; in the Raman spectra of ClSiH,N(SiH,), (C1SiH2),NSiH, and (ClSiH,),N the lines near 990 cm-l are all strong and all strongly polarised. It is true that for the C forms of ClSiH,N(SiH,) and (ClSiH,),NSiH there should be at least two SIN stretching modes of symmetry class a’ (giving polarised Raman lines). However it is interest-ing to see that the chlorine atoms seem to have a marked E. A. V. Ebsworth J . R. Hall M. J . Mackillop D. C. McKean N. Sheppard and L. A. Woodward Spectrochim. Acta S. Cradock and E. A. V. Ebsworth J . Chem. SOC. ( A ) 1967, 1226. 1968 13 202 1628 J.C.S. Dalton effect on the state of polarisation of the Raman lines, even if they do not affect the pattern of frequencies. In mode appears to be polarised though less strongly so than in the spectra of (ClSiH,),N. order to see whether this effect could be detected in the spectrum of a similar compound we recorded the Raman spectrum of (CH,SiH,),N; the line at 990 cm-l which N. S. H. gratefully acknowledge the sward of a studentship by Edinburgh University. has been assigned to t i e antisymmetric SIN stretching [4/2294 Received 5th Nocetnber 1974
ISSN:1477-9226
DOI:10.1039/DT9750001624
出版商:RSC
年代:1975
数据来源: RSC