Answers are grounded in this paper and cite their sources.
Generally, the adiabatic approximation is defined via a projection onto an adiabatically moving electronic base. The Hamiltonian then is separated into three constituents (‘‘adiabatic,’’ ‘‘diagonal-nonadiabatic,’’ ‘‘nondiagonal-nonadiabatic’’). A model system is considered for which the correct zero-order eigenfunctions are known. It is shown that with these any single of the three Hamiltonian constituents yields diverging expectation values; only their sum yields convergence. From this the conclusion seems stringent that any result pertaining to electronic transitions, which is based on the adiabatic approximation, is an artifact.
No takes yet. Share an insight, caveat, or question.
Max Wagner (1985) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: