SEMICLASSICAL FORMULATION OF OPTIMAL CONTROL THEORY

A. Kondorskiy 1,2), G. Mil'nikov 1,3) and H. Nakamura 1,4)

1) Institute for Molecular Science, Myodaiji, Okazaki 444-8585, Japan
2) Lebedev Physical Institute, Leninsky pr., 53, Moscow, 119991, Russia
3) Institute for Structural Macrokinetics, Chernogolovka, Moscow Reg., 142432, Russia
4) The Graduate University for Advanced Studies, Myodaiji, Okazaki 444-8585, Japan

Controlling molecular processes by laser pulses is a subject of active research in chemical physics. One of the most natural and flexible approaches in this area is an optimal control theory (OCT) [1] based on the idea that the controlling laser pulse should maximize a certain functional so that the variational principle can be used to design the pulse. The procedure leads to a set of equations for optimal laser field, which include two Schroedinger equations to describe dynamics starting from the initial and target state wave packets. The optimal laser field is given by the imaginary part of the correlation function of these two wave packets. The system of equations must be solved iteratively in general; hence its numerical cost becomes huge for multi-dimensional systems.
It is thus strongly desired to incorporate semiclassical approaches like the Herman-Kluk wave packet propagation method [2] into the OCT. Such kinds of direct implementation face significant numerical difficulties, however, because calculation of correlation function requires a double summation with respect to a big number of trajectories to be performed at each time step.
In the present work semiclassical formulation of optimal control theory is made by combining conjugate gradient search method with new approximate semiclassical expressions for correlation function [3]. Two expressions for correlation function, which contains only one summation with respect to trajectories, are obtained. The simplest one requires calculation of coordinates and momentums of classical trajectories only. The second one requires extra calculation of common semiclassical values; as a result additional quantum effects are treated.
The efficiency of the method is demonstrated by controlling nuclear wave packet motion in one- and two-dimensional model systems.

References

[1] S. A. Rice, M. Zhao, Optical Control of Molecular Dynamics, USA(2000).
[2] E. Kluk, M. Herman, H. Davis, J. Chem. Phys. 84, 326 (1986).
[3] A. Kondorskiy, H. Nakamura, J. Theor. Comp. Chem. (submitted).


This work was supported by a Grant-in-Aid for Specially Promoted Research on 'Studies of Nonadiabatic Chemical Dynamics based on the Zhu-Nakamura theory' from MEXT, Japan