Pulsed flash pyrolysis in combination with matrix isolation of isotopomer. Upon irradiation of matrix-isolated silylsilacyclopropenylidene (1) a second C 2 H 4 Si 2 isomer, namely three precursor molecules (10,11,12) led to the formation of silylsilacyclopropenylidene (1), which was identified by (silylethynyl)silylene (4) was formed. However, no evidence could be found for the formation of disilatetrahedrane or 1,2-comparison of experimental and calculated (BLYP/6-31G*) IR spectra for both, unlabeled 1 and its perdeuterated or 1,3-disilacyclobutadiene.
Having discussed C 3 X 2 (X ϭ H, Cl, F) [1] [2] [3] [4] and We were not able to find a minimum for disilatetrahedrane (9). This compound represents a second order saddle C 2 H 2 Si [5] [6] species in earlier papers, we concentrate in this study on the generation and properties of the silyl-substi-point lying 64.9 kcal/mol above silacyclopropenylidene 1.
1,2-Disilacyclobutadiene also deserves a special comment. tuted silacyclopropenylidene 1. Irradiation of this compound was expected to open an entry to other C 2 H 4 Si 2 iso-Whereas nearly planar 1,3-disilacyclobutadiene ( 8) is a real minimum on the C 2 H 4 Si 2 energy hypersurface, there is, in mers, including not only open-chain silylenes but also such fascinating species like disilatetrahedrane and disilacyclobu-accordance with earlier calculations [7] , no minimum for a planar 1,2-disilacyclobutadiene (formula 7,C 2v ). Dropping tadienes.
of symmetry restrictions leads to the "real" minimum structure 7, which corresponds to a 1,2-disilabicyclobutane-2,4-Calculations diyl type structure. In addition to diradical 7, in which the C 2 H 4 Si 2 species are of considerable interest with respect hydrogens at the Si atoms are in a formal anti position there to the peculiar bonding properties of silicon. In this system exists a second minimum 5, in which these two hydrogens silicon can be di-or tetravalent, it can participate in double are syn oriented. bonds to silicon or carbon and it can be part of heterocy-Syntheses of Precursors 10, 11, and 12 clobutadienes or heterotetrahedrane.
In 1984 M. S. Gordon et al. calculated structures and Our standard method for the generation of silylenes conenergies of thirteen C 2 H 4 Si 2 isomers at the MP3/6-31G*//3sists in the pyrolytic α-elimination of trimethylsilane from 21G level of theory [7] . Unfortunately, the IR spectra of suitable precursors. The primarily formed noncyclic silylthese isomers were not calculated. Since we needed these enes undergo a thermal rearrangement to the corresponddata for the interpretation of our experimental IR spectra ing silacyclopropenylidenes [4] [5] . In the present case, three we repeated the calculations on the BLYP/6-31G* level of candidates can formally yield open-chain C 2 H 4 Si 2 silylenes theory [8] . Scheme 1 shows the structures and energies of upon extrusion of trimethylsilane, namely 5,5-dimethyleight C 2 H 4 Si 2 isomers, which are minima on the energy hy-1,4,5-trisilahex-2-yne (10), 1,1,1-trimethyl-2-ethynyltrisilane persurface and whose spectra were of interest to us.
(11), and 1,1,1-trimethyl-3-ethynyltrisilane (12). As expected, the global minimum is silylsilacycloprope-
The preparation of precursors 10, 11, and 12 turned out nylidene 1. Of nearly the same energy is disilacyclobutenylito be difficult, but we succeeded in preparing at least sufdene 2, which could be a suitable photochemical precursor ficient amounts of each for matrix-isolation studies. 5,5-Difor 1,2-disilacyclobutadiene. Similarly, photolysis of disilymethyl-1,4,5-trisilahex-2-yne (10) was synthesized in the follene 3 could result in the formation of 1,3-disilacyclobutadilowing way: A mixture of phenylsilane (13) and phenyldisilene 8. The weak SiϪSi bond in silylene 6 makes this comane 15 was treated with the stoichiometric amount of tripound much less stable than (silylethynyl)silylene 4. fluoromethanesulfonic acid yielding the corresponding triflates 14 and 16. Reaction of the resulting solution with a [᭛] Part 24: Ref. [6] .