Guohua Tao

Rotational coherence and a sudden breakdown in linear response seen in room-temperature liquids

Highly energized molecules normally are rapidly equilibrated by a solvent; this finding is central to the conventional (linear-response) view of how chemical reactions occur in solution. However, when a reaction initiated by 33-femtosecond deep ultraviolet laser pulses is used to eject highly rotationally excited diatomic molecules into alcohols and water, rotational coherence persists for many rotational periods despite the solvent. Molecular dynamics simulations trace this slow development of molecular-scale friction to a clearly identifiable molecular event: an abrupt liquid-structure change triggered by the rapid rotation. This example shows that molecular relaxation can sometimes switch from linear to nonlinear response.

The molecular origins of nonlinear response in solute energy relaxation: The example of high-energy rotational relaxation

A key step in solution-phase chemical reactions is often the removal of excess internal energy from the product. Yet, the way one typically studies this process is to follow the relaxation of a solute that has been excited into some distribution of excited states quite different from that produced by any reaction of interest. That the effects of these different excitations can frequently be ignored is a consequence of the near universality of linear-response behavior, the idea that relaxation dynamics is determined by the solvent fluctuations (which may not be all that different for different kinds of solute excitation). Nonetheless, there are some clear examples of linear-response breakdowns seen in solute relaxation, including a recent theoretical and experimental study of rapidly rotating diatomics in liquids. In this paper we use this rotational relaxation example to carry out a theoretical exploration of the conditions that lead to linear-response failure. Some features common to all of the linear-response breakdowns studied to date, including our example, are that the initial solute preparation is far from equilibrium, that the subsequent relaxation promotes a significant rearrangement of the liquid structure, and that the nonequilibrium response is nonstationary. However, we show that none of these phenomena is enough to guarantee a nonlinear response. One also needs a sufficient separation between the solute time scale and that of the solvent geometry evolution. We illustrate these points by demonstrating precisely how our relaxation rate is tied to our liquid-structural evolution, how we can quantitatively account for the initial nonstationarity of our effective rotational friction, and how one can tune our rotational relaxation into and out of linear response.

Understanding electronically non-adiabatic relaxation dynamics in singlet fission.

Nonadiabatic relaxation of one singlet state into two triplet states is the key step in singlet fission dynamics, the understandings of which may help design next generation solar cells. In this work we perform the symmetrical quasi-classical (SQC) nonadiabatic molecular dynamics (MD) simulation [Cotton and Miller, J. Phys. Chem. A, 2013, 117, 7190; Meyer and Miller, J. Chem. Phys. 1979, 70, 3214] for a model system to study the real-time fission dynamics. The dependence of the nonadiabatic relaxation dynamics on energy levels, electronic couplings, and electronic-phonon couplings has been examined, in comparison with other analytical approximations, such as Förster theory and Marcus theory. Unlike many other methods, the SQC nondiabatic MD simulation approach is able to describe fission dynamics efficiently and accurately enough to provide microscopic insights into singlet fission.

Time-dependent importance sampling in semiclassical initial value representation calculations for time correlation functions

An efficient time-dependent importance sampling method is developed for the Monte Carlo calculation of time correlation functions via the initial value representation (IVR) of semiclassical (SC) theory. A prefactor-free time-dependent sampling function weights the importance of a trajectory based on the magnitude of its contribution to the time correlation function, and global trial moves are used to facilitate the efficient sampling the phase space of initial conditions. The method can be generally applied to sampling rare events efficiently while avoiding being trapped in a local region of the phase space. Results presented in the paper for two system-bath models demonstrate the efficiency of this new importance sampling method for full SC-IVR calculations.

Time-dependent importance sampling in semi-classical initial value representation calculations for time correlation functions. III. A state-resolved implementation to electronically non-adiabatic dynamics

The recently developed time-dependent (TD) Monte Carlo (MC) importance sampling method [Tao and Miller; JCP 135, 024104 (2011)] provides an efficient implementation of the semi-classical (SC) initial value representation (IVR) methodology for the evaluation of time correlation functions. The key idea in this TD-SC-IVR method is to perform importance sampling of trajectories for the MC averages in the SC calculations with a sampling function that includes information about the final (time-evolved) values of the coordinates and momenta of trajectories in addition to (the usual) information of their initial values. This paper shows how this approach deals with electronically non-adiabatic dynamics, i.e. dynamics that involves coupled multiple electronic states. We suggest that a state-resolved sampling function may facilitate the efficient implementation of TD-SC-IVR for such processes. This is illustrated by application to the calculation of the nuclear momentum distribution function (i.e. the nuclear energy-loss spectrum) for a benchmark non-adiabatic scattering problem.

A multi-state trajectory method for non-adiabatic dynamics simulations.

A multi-state trajectory approach is proposed to describe nuclear-electron coupled dynamics in nonadiabatic simulations. In this approach, each electronic state is associated with an individual trajectory, among which electronic transition occurs. The set of these individual trajectories constitutes a multi-state trajectory, and nuclear dynamics is described by one of these individual trajectories as the system is on the corresponding state. The total nuclear-electron coupled dynamics is obtained from the ensemble average of the multi-state trajectories. A variety of benchmark systems such as the spin-boson system have been tested and the results generated using the quasi-classical version of the method show reasonably good agreement with the exact quantum calculations. Featured in a clear multi-state picture, high efficiency, and excellent numerical stability, the proposed method may have advantages in being implemented to realistic complex molecular systems, and it could be straightforwardly applied to general nonadiabatic dynamics involving multiple states.

Efficient importance sampling in semiclassical initial value representation calculations for time correlation functions

The semiclassical (SC) theory based on an initial value representation (IVR) methodology provides a practical way to describe quantum effects in complex molecular systems. The efficiency of SC–IVR calculations for time correlation functions depends heavily on how to perform the Monte Carlo sampling of initial conditions. Here, we compare a variety of possibilities of sampling initial conditions in the SC calculations by choosing the sampling function to be either time-dependent or time-independent (TI). The implementation of these importance sampling protocols to two benchmark system-bath models demonstrates its advantages over the standard sampling method. In particular, the recently developed TI importance sampling which incorporates path correlation in the bath degrees of freedom shows a great potential in describing many-body quantum dynamics efficiently and accurately.