SUPEREXCHANGE PATHWAY CALCULATION OF LONG-DISTANCE ELECTRONIC COUPLING IN H2C(CH2)M-2CH2 CHAINS

Citation
La. Curtiss et al., SUPEREXCHANGE PATHWAY CALCULATION OF LONG-DISTANCE ELECTRONIC COUPLING IN H2C(CH2)M-2CH2 CHAINS, Chemical physics, 176(2-3), 1993, pp. 387-405
Citations number
72
Categorie Soggetti
Physics, Atomic, Molecular & Chemical
Journal title
ISSN journal
03010104
Volume
176
Issue
2-3
Year of publication
1993
Pages
387 - 405
Database
ISI
SICI code
0301-0104(1993)176:2-3<387:SPCOLE>2.0.ZU;2-P
Abstract
In a recent communication (C.A. Naleway, L.A. Curtiss and J.R. Miller, J. Phys. Chem. 95 (1991) 8434) we outlined a method for calculation a nd analysis of long-distance couplings in electron transfer reactions making use of McConnell's superexchange formalism with matrix elements from ab initio molecular orbital wave functions transformed to locali zed bond orbitals using Weinhold's natural bond orbital method. In thi s paper we present a detailed investigation into several aspects of th e superexchange (SE) pathways method for calculation of anion and cati on pi couplings in H2C(CH2)m-2CH2, m=4-8, trans alkyls. The convergenc e of the couplings calculated from this method is examined as a functi on of energy threshold. Good convergence is found for wave functions ( neutral triplet beta, dication) for which paths through unoccupied orb itals dominate paths through occupied orbitals; convergence is not as good for wave functions (neutral triplet a, anion) for which occupied paths are dominant. Reasons for these results are explored. The coupli ngs calculated from the SE method are in reasonable agreement with Koo pman's theorem when convergence of the couplings with threshold is att ained and the 3-2 1 G basis set did well compared to larger basis sets . In agreement with previous studies, pathways with hops over bonds ma ke up the bulk of the interaction and most of these involve CC bonds o r CC antibonds. The SE pathways calculations, some of which included more than 10 million different paths, indicate that contributions from long paths tend to cancel so that the shorter paths (2-5 jumps) are l argely responsible for determining the magnitude of the coupling.