scispace - formally typeset
Search or ask a question

Showing papers by "Peter J. Rossky published in 1995"


Journal ArticleDOI
TL;DR: In this article, a reduction mapping of the quantum density matrix is proposed for mixed quantum-classical systems. But the quantum decoherence problem is not addressed in this paper, instead, the classical paths are restricted to a single path among all the quantum paths.
Abstract: We address the issue of quantum decoherence in mixed quantum‐classical simulations. We demonstrate that restricting the classical paths to a single path among all the quantum paths affects a coarse graining of the quantum paths. Such coarse graining causes the quantum paths to lose coherence as the various possible classical paths associated with each quantum state diverge. This defines a reduction mapping of the quantum density matrix, and we derive a quantum master equation suitable for mixed quantum‐classical systems. The equation includes two terms: first, the ordinary quantum Liouvillian which is parametrized by a single classical path, and second, a quantum decoherence term that includes both a coherence time and length scale which are determined by the dynamics of the classical paths. Model calculations for electronic coherence loss in nonadiabatic mixed quantum‐classical dynamics are presented as examples. For a model charge transfer chemical reaction with nonadiabatic transitions, application of ...

357 citations




Journal ArticleDOI
TL;DR: In this paper, the simulation results are directly compared to two available data sets obtained experimentally via ultrafast absorption spectroscopy, and it is pointed out that a number of available, and apparently different, hypotheses for the sequence of species present during electronic relaxation share key features with this mechanism.
Abstract: Previous analysis of the computer simulation of the relaxation of energetic excess electrons in liquid water (Keszei E.; et al. J . Chem. Phys. 1993, 99, 2004) has led to a detailed molecular level and kinetic picture of this process, including the presence of multiple pathways to the equilibrium ground state. In order to explore the validity of this view, simulation results are directly compared to two available data sets obtained experimentally via ultrafast absorption spectroscopy. The analysis is carried out, first, by convolution of the simulated instantaneous spectral response of the electron with an appropriate instrumental response function. The difference between the resulting data and the reported experimental observations is no larger than the difference between the two experimental data sets. It is further shown by separate analysis that the mechanism of relaxation apparent in the simulation is kinetically consistent with the available experimental data. It is pointed out that a number of available, and apparently different, hypotheses for the sequence of species present during electronic relaxation share key features with this mechanism. Taken together, these considerations support the validity of the microscopic processes evident in simulation and emphasize the limitations inherent in the analysis of the experimentally determined spectral dynamics.

39 citations



Journal ArticleDOI
TL;DR: In this article, the microscopic connections between solvation dynamics and the experimentally measured dynamic emission Stokes shift are explored by molecular dynamics simulation using the hydrated electron as a fully quantum mechanical probe.
Abstract: The microscopic connections between solvation dynamics and the experimentally measured dynamic emission Stokes shift are explored by molecular dynamics simulation using the hydrated electron as a fully quantum mechanical probe. Solvent response functions are computed using the method of spectral reconstruction at various time resolutions for both fluorescence and stimulated emission and compared directly to the dynamic evolution of the underlying electronic energy gap. The first spectral moment proves to be a better choice of characteristic frequency than the spectral maximum, and iterative deconvolution is found to significantly improve the spectrally determined solvent response function. Use of the appropriate characteristic frequency at zero time to produce solvent response functions which do not start at unity further improves the agreement between the spectrally determined and microscopic solvent response functions.

35 citations



Journal ArticleDOI
TL;DR: In this article, the authors explore the possible coupling between dielectric solvation, the response of the solvent to a change in charge distribution of the solute, and mechanical solvation.

17 citations


Journal ArticleDOI
TL;DR: In this paper, a new semiclassical formalism for nonadiabatic dynamics of a quantum subsystem interacting with an explicit bath is presented, based on a stationary phase approach to the bath and a variational principle for the quantum transition amplitudes, for quantum systems represented by approximate wave functions.
Abstract: We present a new semiclassical formalism for nonadiabatic dynamics of a quantum subsystem interacting with an explicit bath. The method is based on a stationary phase approach to the bath and a variational principle for the quantum transition amplitudes, for quantum systems represented by approximate wave functions. A new expression for the force exerted on a classical bath by a quantum subsystem is derived which, in the adiabatic limit, reduces to the gradient of the expectation value of the energy. Our new methods for adiabatic and nonadiabatic dynamics are applied to a test problem of vibrational relaxation. For adiabatic dynamics, we find that our new algorithm produces results which converge faster, with increasing basis set size, than calculations performed with the Hellmann–Feynman force; for a limited basis set, our new algorithm gives results that are in better agreement with exact results. For nonadiabatic dynamics, we also find that, in comparison to an earlier algorithm, our new algorithm produces results which converge more rapidly with increasing basis set size. In addition, we find that our new algorithm is more robust with respect to the size of the time step than the earlier algorithm, a result of the implementation of a nuclear coordinate dependent basis.

10 citations