Practical approximation of the non-adiabatic coupling terms for same-symmetry interstate crossings by using adiabatic potential energies only

A very simple equation, F-ij(App) = [(partial derivative(2)(V-i(a) - V-j(a))/partial derivative Q(2))/(V-i(a) - V-j(a))](1/2) /2, giving a reliable magnitude of non- adiabatic coupling terms (NACTs, F-ij's) based on adiabatic potential energies only (V-i(a) and V-j(a)) was discovered, and its reliability was tested for several prototypes of same- symmetry interstate crossings in LiF, C-2, NH3Cl, and C6H5SH molecules. Our theoretical derivation starts from the analysis of the relationship between the Lorentzian dependence of NACTs along a diabatization coordinate and the well- established linear vibronic coupling scheme. This analysis results in a very simple equation, alpha = 2 kappa/Delta(c), enabling the evaluation of the Lorentz function ff parameter in terms of the coupling constant and the energy gap Delta(c) (Delta(c) broken vertical bar V-i(a) - V-j(a)broken vertical bar(Qc)) between adiabatic states at the crossing point Q(C). Subsequently, it was shown that Q(C) corresponds to the point where F-ij(App) exhibit maximum values if we set the coupling parameter as kappa = [(V-i(a) - V-j(a)) center dot (partial derivative(2)(V-i(a) - V-j(a)) / partial derivative Q(2))](QC)(1/2) /2. Finally, we conjectured that this relation could give reasonable values of NACTs not only at the crossing point but also at other geometries near Q(C). In this final approximation, the pre- defined crossing point Q(C) is not required. The results of our test demonstrate that the approximation works much better than initially expected. The present new method does not depend on the selection of an ab initio method for adiabatic electronic states but is currently limited to local non-adiabatic regions where only two electronic states are dominantly involved within a nuclear degree of freedom. Published by AIP Publishing.