| pubmed-article:14629150 | pubmed:abstractText | Ab initio calculations using 6-311G**, cc-pVDZ, aug-cc-pVDZ, and a (valence) double-zeta pseudopotential (DZP) basis set, with (MP2, QCISD, CCSD(T)) and without (UHF) the inclusion of electron correlation, and density functional methods (B3LYP) predict that 1,n-homolytic transfers (n = 1-5) of silyl, germyl, and stannyl groups from group IV heteroatoms to carbon radicals can proceed via a frontside attack mechanism. At the B3LYP/DZP level of theory, energy barriers (DeltaE++) of 101.2, 98.8, 58.9, and 63.4 kJ/mol are calculated for the 1,2-, 1,3-, 1,4-, and 1,5-translocation reactions, respectively, of SiH3 between silicon atoms. Similar results are obtained for reactions involving germanium and tin with energy barriers (DeltaE++) of 85.9-113.1, 84.4-109.0, 41.7-73.3, and 48.5-78.2 kJ/mol for the 1,2-, 1,3-, 1,4-, and 1,5-translocation reactions, respectively. This study also predicts that four- and five-membered ring-closure reactions can be competitive with the 1,4- and 1,5-translocation reactions. These results suggest that while 1,2- and 1,3-translocation four-membered ring-formation reactions are unlikely to be synthetically viable, 1,4- and 1,5-transfers and five-membered ring-formation have synthetic possibilities. | lld:pubmed |
