THEORETICAL-STUDY ON THE ALPHA-PROTON HFS CONSTANT IN METHYLSILYL RADICALS

Ab initio UHF/TZP calculations on the alpha-methyl substituted silyl radicals H(3-n)SiMe(n) (n = 0-2) show that methyl substitution does not appreciably change the geometry at silicon, which is computed to be nearly tetrahedral, as previously suggested on the basis of the significant variations observed experimentally in the alpha-proton hyperfine splitting constant, a(H-1(alpha)). UMP2/DZP//TZP calculations reproduce well these variations and show that they are mainly due to spin-polarization effects. The magnitude of a(H-1(alpha)) is, however, slightly underestimated. The agreement improves both by using higher correlated wavefunctions (UQCISD(T)) and by augmenting the basis set with tight and diffuse functions. The controversy on the sign of a(H-1(alpha)) has been definitely resolved, the sign is found to be positive for all the radicals.