Allenes 1-silaallenes

Malrieu predicted qualitatively that the l-silaallene framework would be nonlinear. but how much the moiety would deviate from 180 was unclear. The first quantitative values were seen with the isolation and structural determination of the two 1-silaallenes 56 and 59a, whieh have very similar Si=C=C angles of 173.5 and 172.0, respectively (Table III). This is an average deviation of 7.3 from linearity—significant, but relatively small compared to the deviations shown by the germanium and tin substituted allenes. 

Reiterating the idea that the -complexes 132a and 86a are not truly heteroallenes, one must consider the central carbon C NMR chemical shifts. Silaallene complex 132a starts to approach allene status with a chemical shift of 175.5 ppm, but the most deshielded carbon for alkylidenetelluragermirane 86a is only 153.04 ppm. 

Like the C NMR chemical shifts for the central carbon of heteroallenes. changing the composition of the allene fragment can have as much of an effect (if not more) on the Si chemical shift as does changing a substituent. Switching the carbon at position 3 to a phosphorus (silaphosphaallene 122) moves the " Si chemical shift downheld to 75.7 ppm, which is the most deshiclded " Si shift fora silaallene to date, even though it has two aryl groups on silicon, which should keep the shift uplield. 

How might this held develop in the future At this time, the chemistry of 1-silaallenes is fairly well understood, and that of 1-germaallenes is at least partially explored. Perhaps the series can be extended to 1 -stannaallenes, but these are predicted to have quite limited stability. The isolation of the 1,2.3-tristannaallcne shows that any combination of heteroatoms may be combined to form allenes... 

If doubly-bonded compounds >M = X (M = Si, Ge, Sn) are now well known, this is not yet the case for the heavy allenic derivatives >M = C = X. Transient silaallenes >Si = C = C 160-167 or silaazaallene >Si = C = N—168 were postulated some years ago, but the first stable derivative of this type, a formal 3-stanna-l-azaallene >Sn = C = N —, was isolated only in 1992 by Griitzmacher et al169 The compound Tbt(Mes)Si = C = NR170 has also been obtained, but according to the authors it is closer to a silylene Lewis base complex than to a silacumulene. The first stable silaallene was only reported in 1993 by West et al.111 and the first metastable >Si = C = P— was characterized still more recently.172... 

The first synthesis of stable 1-silaallenes follows closely an established route developed in carbon chemistry292. An overall nucleophilic 1,3 substitution at propargylic halides gives allenes in good yields (equations 206 and 207). In organosilicon chemistry, special attention must be paid to avoid direct substitution at the silicon which is the preferred site for nucleophilic attack. 

The first stable 1-silaallenes (R2Si=C=CR2) and 1-germaallenes (R2Ge=C=CR2) were synthesized recently by West and coworkers338a d. Stable 2-metallaallenes, R2C=M=CR2, are not yet known. In contrast to allene (61), which has a... 

The structures of some stable silaallenes have been revealed by X-ray crystallographic analysis (Figure 20). The geometries of their Si=C=C skeletons are nearly the same with shorter Si=C double bonds (ca. 1.70 A) than the typical Si=C bond lengths (see Section 3.2). Such a short distance is conceivable to be affected by the sp-hybridized central carbon atom. The C=C bond lengths (ca. 1.325 A) in these Si=C=C allenes are in normal range of C=C double bonds. It is worthy of note that the Si=C=C bond angle is not completely 180°... 

The successful isolation of 1-silaallenes and 1-germaallenes raises the expectation of synthesis of a novel class of allenic compounds having an sp-hybridized silicon unit, that is, 2-silaallene derivatives. Although the generation of 2-silaallene (f-Bu)-CH2-CH=Si=CH-CH2-(t-Bu) was postulated in the reaction of (H2C=CH)2SiCl2 with 2-equivalent of t-BuLi, no distinct evidence on the generation of such a compound has been obtained so far. 

In alkynyldisilanes, the 1,3-shift should lead to 1-silaallene derivatives. No such product has been isolated so far, but there is ample evidence for their intermediacy. In a series of investigations145"150 it was shown that the irradiation of alkynyldisilanes of the type R C=CSiR2SiMe3 produces intermediates that dimerize in the absence of a trapping agent146,149, add methanol145,146 and dimesitylsilylene149,15°, and in the presence of acetone yield the allene and the oligomers of dimethylsilanone that would be expected from an intermediate silaoxetane (equation 42)145,147. 

Silaallenes are of recent vintage. While 2-silaallenes and silazaallenes have only been encountered as intermediates, stable 1-silaallenes with 2,4,6-triisopropylphenyl substituents on silicon have been synthesized recently While 1-silaallenes with highly sterically hindered substituents are stable at room temperature, Me2Si=C=C(SiMe3)Ph exhibits a lifetime of about 25 ps at room temperature in nitrogen-saturated hexane solution. Also, the first stable allenes with double-bonded phosphorous with one 2,4,6-triisopropylbenzene substiment on silicon have been synthesized. In view of the instability of silaallenes, their dimerization reactions are mainly observed. 

Also, silaallenes, generated in the photolysis of ethynyldisilanes, are trapped with acetone to give a transient [2+2] cycloadduct, which decomposes to polysiloxanes and the corresponding allenes . 

---