Jumping from one surface to another

Many interesting chemical processes involve non-adiabatic processes, i.e., they involve transitions between different potential energy surfaces and thus a break-down of the Born-Oppenheimer approximation. Calculating the rate of such processes (even in "toy" models, let alone full electronic structure theory calculations) is a great challenge.

One of the most stimulating talks I heard last week in Telluride was by Joe Subotnik. He gave a nice review of this problem and the "Tully problems" used as a benchmark.

Strategies for treating the problem with increasing sophistication (and computational cost) are something like:

• Ehrenfest mean-field dynamics
• Surface hopping
• Ehrenfest + decoherence (Prezhdo, Rossky, Schwartz...)
• Semi-classical dynamics (Miller–Meyer–Stock–Thoss formalism)
• Partial Wigner transforms (Kapral)
• Multiple spawning
• Full nuclear quantum dynamics

A recent J. Chem. Phys. paper by Joe also gives a good overview. This is all stuff I need to get a better handle on ....