Reaction intermediates disilenes

Styrene and substituted styrenes react with tetramesityldisilene 1, tetra-tert-butyl-disilene 21, and tetrakis(tert-butyldimethylsilyl)disilene 22 to afford the corresponding disilacyclobutane derivatives.127,134 Similarly, [2 + 2] additions occur between the disilenes with a C = C double bond in an aromatic ring135 and acrylonitrile.136 Bains et al. have found that the reaction of disilene 1 with trans-styrene- provides a 7 3 diastereomeric mixture of [2 + 2] adducts, 201 and 202 [Eq. (95)] the ratio is changed, when czs-styrene-Ji is used.137 The formation of the two diastereomeric cyclic adducts is taken as the evidence for a stepwise mechanism via a diradical or dipolar intermediate for the addition, similar to the [2 + 2] cycloaddition of phenylacetylene to disilene ( )-3, which gives a 1 1 mixture of stereoiso-meric products.116,137... 

Trapping experiments with triethylsilane and diphenylacetylene, which lead to 20 and 21, proved the existence of the silylene 18 and the disilene 19 as reaction intermediates of the reaction in... 

In a related manner, disilacyclopropane (disilirane) derivatives can be synthesized by the reaction of disilene (322) with other reactive compounds. Thus, via diazomethane one can add a methylene component to 322 to form the corresponding 1,2-disilacyclopropanes (323). These are versatile intermediates which give rise to a large variety of compounds (vide infra). One type of application, the conversion of 323 with m-CPBA to a l,3-disila-2-oxacyclobutane (disilaoxetane) derivative (324), is shown in equation 147171. 

Although the mechanism of formation of 11-13 cannot be proven experimentally, the following proposal seems to be reasonable. In analogy to the reaction of disilenes, the reaction sequence could be initiated by a [2+1] cycloaddition of a chalcogen atom to one of the Si=Si double bonds, followed by a rearrangement of these intermediates into the less strained five-membered rings ll-13.f ... 

Since the anhydride 28 apparently exhibits a high reactivity towards silicon-silicon multiple bonds, we have also examined its reaction with disilene 23, from which the bicyclic compound 32 was obtained in high yield (Scheme 6). Compound 32 can formally be considered as a [2+3] cycloadduct of the starting materials after a spontaneous 1,2-hydrogen shift. The driving force for the formation of 32 is assumed to be the oxophilicity of silicon. This leads to the dipolar addition product 30 that, in turn, affords the intermediate 31 through 1,4-addition. The last step of the sequence would then be a 1,2-proton shift to furnish the isolated compound [16]. 

Evidence for a 7t-coordination was obtained through the reaction of various disilenes with Hg(OCOCF3)2, a reaction which leads regioselectively to bis(tri-fluoracetyl)disilanes. A disilene n-complex (79), which is stable up to — 50 °C, could be identified as an intermediate by spectroscopic methods. 

From X-ray crystal structures of the products, the reactions of these stereoisomeric disilenes with episulfides, and with sulfur, were shown to proceed with retention of configuration at silicon. These findings suggest that the reaction proceeds in a concerted fashion, through intermediates or transition states involving tetracoordination for the sulfur atom being transferred (Scheme 13). Similar intermediates are believed to occur in other sulfur-transfer reactions.86... 

The early stages in the oxidation of disilene have been treated theoretically for the parent molecule H2Si=SiH2.95 The first intermediate along the reaction coordinate is the open-chain trans diradical 64 (Scheme 16), which is in equilibrium with a gauche form, 65. From the latter, closure to the 1,2-dioxetane 66 would probably be rapid. The open-chain form can react with a second molecule of disilene to give the diradical 67, which could collapse into two molecules of the disilaoxirane 68. If similar steps are followed in the oxidation of 3, they must be quite rapid, since the relative configuration at the silicon atoms is maintained in both products, 59a and 61a.93... 

The first 1,2-disilacyclobutene (82) was prepared in 1973 by the gas phase reaction of dimethylsilylene and 2-butyne (73JOM(52)C21). It probably results through silylene insertion into the intermediate silacyclopropene (Section 1.20.3.4), but silylene dimerization followed by addition to the alkyne is also suggested (76JA7746), since (82) is formed in good yield if the disilene is generated directly (Scheme 127) (78JOM(162)C43). 

The formation of an intermediate difluorodisilene 25 (equation 8) was proposed by Jutzi and coworkers34 in the reaction of decamethylsilicocene with tetrafluoroboric acid. The disilene which was characterized by the 29Si NMR spectrum, then formed the isolable cyclotetrasilane by [2 + 2] cycloaddition. 

The results in Table 3 were explained as shown in Scheme 4. From the fact that no kinetic isotope effect was observed in the reaction of phenyl-substituted disilenes with alcohols (Table 1), it is assumed that the addition reactions of alcohols to phenyltrimethyl-disilene proceed by an initial attack of the alcoholic oxygen on silicon (nucleophilic attack at silicon), followed by fast proton transfer via a four-membered transition state. As shown in Scheme 4, the regioselectivity is explained in terms of the four-membered intermediate, where stabilization of the incipient silyl anion by the phenyl group is the major factor favoring the formation of 26 over 27. It is well known that a silyl anion is stabilized by aryl group(s)443. Thus, the product 26 predominates over 27. However, it should be mentioned that steric effects also favor attack at the less hindered SiMe2 end of the disilene, thus leading to 26. 

The high diastereoselectivity in the addition of i-PrOH, t-BuOH and EtOH (at low concentration) suggests that E Z photoisomerization of (E)- or (Z)-16 does not occur in solution at room temperature or that the trapping of (E)- or (Z)-16 by alcohols proceeds faster than the E Z isomerization. In addition, the results show that proton transfer in the intermediate adduct formed by the disilenes and alcohols occurs much faster than rotation around the Si—Si bond. However, in the reaction with ethanol, an appreciable amount of the anti addition product was formed. Thus, the diastereoselectivity remarkably depended on the concentration of ethanol. 

Some of the reactions given in the sections above are important routes to reactive organosilicon intermediates such as silenes and silylenes, and because of their tendency to dimerize readily, to disilenes, the latter being formed when matrix-isolated silylenes are warmed up. It appears worthwhile to summarize some of the more useful reactions leading to silenes and silylenes, and their subsequent behaviors. The topics of silylenes97,142 and disilenes97,142,143 have recently been reviewed. 

The first 1,2,4-thiadisiletane results from the reaction of carbon disulphide and the hindered silylene [2,4,6- (Me3Si)2CH 3C6H2]MesSi (TbtMesSi ) formed from the Z-disilene precursor. The mechanism is thought to involve a skeletal rearrangement of the 3,3 -spirobi(l,2-thiasilirane) intermediate formed by silylene addition to each carbon-sulphur double bond (equation 31)64. 

Further cycloaddition reactions of silylenes generated by the photolysis of cyclotrisilanes have been published since Weidenbruch and coworkers summarized these reactions in an excellent review. Different siliranes were prepared by [2+1]-cycloaddition of di-t-butylsilylene to various alkenes and dienes (Scheme 6)46. Quite interesting results are obtained from the photolysis of hexa-i-butylcyclotrisilane in the presence of unsaturated five-membered ring compounds47 (Scheme 7). With cyclopentadiene and furane, [4 + 2]-cycloaddition of the photolytically generated disilene occurs only as a side reaction. Furthermore, [2 + 1]-cycloaddition of the intermediately formed silylene is highly favored and siliranes are primarily obtained. A totally different course is observed for the reaction in the presence of thiophene. The disilene abstracts the sulfur atom with the formation of the 1,2-disilathiirane as the major product with an extremely short Si—Si distance of 230.49 pm. 

The first step of the retro-reaction involves loss of silylene 79, which could be trapped with 1-pentyne to give the known silirene 81 (equation 125). In the absence of a trapping agent, 79 recondenses to 77, probably by first dimerizing to the disilene Ar2Si=SiAr2 followed by 2 +1 cycloaddition to give 77 (equation 126). From the principle of microscopic reversibility, the fact that silylene is formed in the retro-reaction leads to the conclusion that 79 must also be an intermediate in the cycloaddition reaction. 

The 1,2,3-azadisiletidine 47 and 1,2,3-azadisiletine 48 are, in a formal sense, the products of a [2+2] cycloaddition reaction between nitriles and disilene (Scheme 20). It can be assumed that the latter is the crucial intermediate formed during the thermolysis of hexasubstituted cyclotrisilane <1995TL8187>. 

Disilenes react with ketones, aldehydes, esters and acid chlorides by formal [2 + 21-cycloaddition to yield the corresponding disiloxetanes (equation 73)8,16. The reaction is non-concerted and proceeds through the initial formation of a 1,4-biradical intermediate, as has been shown by the products of reaction of tetramesityldisilene (110) with the cyclopropyl aldehyde 117 (equation 90)163. The absolute rate constants listed in Table 19 indicate there to be a significant difference in reactivity between the monophenyl-substituted disilene 103 and the 1,2-diphenyl-substituted derivatives 104, consistent with a steric effect on the rate of formation of the biradical intermediate. As would be expected, no kinetic deuterium isotope effect is discernible from the relative rates of addition of acetone and acetone- to these compounds. 

---