Substituent effects on scalar 2J(19F,19F) and 3J(19F,19F) NMR couplings:: A comparison of SOPPA and DFT methods

Substituent effects for (2)J(F,F) couplings in aliphatic and olefinic CF2 moieties and (3)J(F,F) couplings in fluorinated derivatives of ethylene were studied using both high level ab initio and DFT/B3LYP calculations. Where possible, J variations have been compared with experimental values. In general, the SOPPA (second-order polarization propagator approximation) methodology matches absolute experimental values reasonably well, whereas the DFT/B3LYP approach performs poorly in describing (2)J(F,F) couplings. Fortunately, substituent effects for DFT J couplings are notably better reproduced. For a vinyl CF2 moiety, the accurate prediction of (2)J(F,F) couplings is a challenging task even for high level ab initio methods such as SOPPA and SOPPA(CCSD) (second-order polarization propagator approximation with coupled cluster singles and doubles amplitudes). Aliphatic (2)J(F,F) couplings are very sensitive to the electronegativity of substituents placed alpha to the CF2 group. The latter J perturbations are dominated largely by the noncontact PSO and SD Ramsey contributions, whereas the influence of the FC term is rather small. Substituent effects on (2)J(F,F) and (3)J(F,F) couplings in fluorinated derivatives of ethylene are also dominated by non-Fermi contributions. Because DFT/B3LYP strongly underestimates the FC contribution, but generally assesses the non-Fermi terms similar to SOPPA, the latter accounts for DFT's ability to predict substituent effects reasonably well.