Silyl radicals from disilanes

Hybrids of the type sp3 are unjustified for disilane. An important conclusion from the above hybridization statement No. 4 is concerned with the contrasting structures of the radicals SiH3 and CH3. The planar geometry of the methyl radical can readily be explained by the (bond-strengthening) sp2-hy-bridization, while the pyramidal silyl radical is thought to be stabilized (with respect to the planar arrangement) through the s-admixture to the lone electron orbital. 

The much studied photochemistry of aryldisilanes carried out in earlier years has been reviewed51,52. Cleavage of the silicon-silicon bond of the disilyl moiety is always involved, but various other reactions have been observed depending on the structure of the disilane and the conditions employed. Thus cleavage to a pair of silyl radicals, path a of Scheme 15, is normally observed, and their subsequent disproportionation to a silene and silane, path b, is often observed. There is evidence that the formation of this latter pair of compounds may also occur by a concerted process directly from the photoex-cited aryldisilane (path c). Probably the most common photoreaction is a 1,3-silyl shift onto the aromatic ring to form a silatriene, 105, path d, which may proceed via radical recombination52. A very minor process, observed occasionally, is the extrusion of a silylene from the molecule (path e), as shown in Scheme 15. 

Similar products were produced from copolymers such as poly-(cyclohexylmethylsilane-co-dimethylsilane), and materials characteristic of the copolymer composition were obtained. Three disilanes (i.e., 1,2-dicy-clohexyl-1,2-dimethyl- 1,1,2,2-tetramethyl- and 1-cyclohexyl-1,2,2-tri-methyldisilane), as well as the two substituted silylene adducts, were produced from the predominantly random copolymer. The disilanes, which are taken to be diagnostic of silyl radical abstraction, were assumed to accumulate in the mixture, because they no longer absorbed light significantly at 254 nm. The photoinstability of a model trisilane, 2-n-butyl-l, 1,1,2,3,3,3-hepta-methyltrisilane, was demonstrated under the reaction conditions even though its absorption maximum occurred at 215 nm. 

Several dimerization rates of alkyl-substituted silyl radicals were measured earlier [8]. However, the dimerization rates of silyl-substituted silicon-centered radicals have not previously been determined. In this study we have measured, using EPR spectroscopy, the rate constants for the recombination of four silyl radicals (lb, 2b, 3b, and 4b), to produce the corresponding disilane dimers of type a (i.e. la, 2a, 3a, and 4a respectively). This dimerization reaction is shown as the backward reaction of Eq. 2 in Scheme 1. Radicals lb, 2b, and 3b were generated photochemically fiom the corresponding disilane dimers of type a (Scheme 1, Eq. 2), while radical 4b was generated photochemically from the corresponding silylmercury compound 4c (Scheme 1, Eq. 1). 

The enthalpies, KH, of the homolytic cleavage of the central Si-Si bond in disilane dimers a can be calculated from temperature-dependent ESR experiments, using Eq. 3, in which C is the concentration of the radical, T is the absolute temperature and A is a. constant. The change in the concentration of the radicals as a function of temperature could be followed by EPR as shown in Fig. 6a for radical 3b. The concentrations of the thermally generated radicals (2b and 3b) were determined by calibration of the height of the EPR signal of the silyl radicals in comparison with a 3 X 10 M toluene solution of TEMPO (2,2,6,6-tetramethyl-piperidinooxy). 

Similarly, the attempted synthesis of diadamantylsilylene by extrusion from a trisilane 267 did not give the wanted product ". The predominant photoreactions are silicon-silicon bond homolysis to give the radicals 268 and 269 (equation 67). Disproportionation of 268 and 269 results in the formation of silene 271 and silane 272. TTie silene is identified by isolation of its head-to-tail dimer 273. In the presence of scavenger reagents like 2,3-dimethylbuta-l,3-diene radical trapping products like 270 could be detected in low yields. Secondary photoprocesses involving the disilane 272 take place. Formation of silyl radicals 269 and 274 with subsequent disproportionation of the radicals explain the formation of diadamantylsilane 275. 

The 1,3-silyl shift in aryl disilanes is suppressed when the aromatic ring is ortho-substituted144. An attempted silylene synthesis from 1,3-dimesitylhexamethyltrisilane 259, however, led to low yields of silylene trapping products (ca 30% generation of Me2S ). The major pathway is the homolytic cleavage of the trisilane, followed by disproportionation of the radicals 260 and 261 to the silene 262 and the disilane 263 (equation 65). 

---