Martin Head-Gordon and David Chandler, University of California, Berkeley
and Lawrence Berkeley National Laboratory

Research Objectives
Our research centers on the development and application of methods that predict the electronic structure of interesting molecules. We seek to open new classes of chemical problems to study via electronic structure theory, including large molecules and extended systems, molecular excited states, and nonadiabatic interactions.

Computational Approach
Our approach includes electronic structure methods of the density functional theory type, Newtonian molecular dynamics, Car-Parinello ab initio molecular dynamics (CPMD), and transition path sampling methods.

Accomplishments
We have shown that a relatively simple theoretical approach, time-dependent density functional theory, can accurately describe electronic excited states of radicals including those with bielectronic character. Previously such states were considered very challenging to characterize.

We have produced calculations of unprecedented accuracy on large polycyclic aromatic hydrocarbon (PAH) cations of the form that arise as intermediates in combustion processes on the way to forming soot particles, and are also believed to play a significant role in interstellar carbon chemistry. Our calculations resulted in a new assignment of the visible spectrum of the perylene cation. Our results open the way to further studies of larger cation radicals.

Through the development of transition path sampling, we have created algorithms that permit the study of rare events without prior knowledge about mechanisms or transition states. The technique is an importance sampling of trajectory space that is based upon our discoveries of statistical mechanics and thermodynamics governing dynamical systems that can be far from equilibrium.

We have created a formulation of transition path sampling that can be interfaced with any trajectory algorithm, to harvest only those trajectories of interest or of importance. We have used this formulation to combine transition path sampling with CPMD.

We have devised a set of statistical methods for interpreting the behavior of transition paths in complex systems. These tools provide a type of pattern recognition for interpreting dynamics in a complex, high-dimensional system.

By harvesting reactive trajectories of autodissociation in liquid water, we have discovered molecular details of a fundamental chemical reaction. These results have established coordinates in liquid water that may be important for many bond-breaking and, in particular, proton transfer processes.

Configurations of liquid water within a few femtoseconds before (left) and after (right) liquid water has passed through the transition state surface of autoionization. The dashed lines show the hydrogen bond wire that must break in order for the system to cross the transition state surface. Yellow and blue identify the hydroxide and hydronium ions, respectively.

Significance
Electronic structure theory has emerged as a valuable counterpart to direct experiments for the study of reactive species that may not be characterized easily (if at all) in the laboratory. Our research on chemical and conformational transformations of biomolecules is beginning to yield a novel microscopic picture of biochemical dynamics. Our results may have significant implications for the general understanding of solvent roles in chemical and biochemical processes.

Publications
S. R. Gwaltney and M. Head-Gordon, "A second-order correction to singles and doubles coupled-cluster methods based on a perturbative expansion of a similarity-transformed Hamiltonian," Chem. Phys. Lett. 323, 21 (2000).

P.G. Bolhuis, C. Dellago, and D. Chandler, "Reaction coordinates of biomolecular isomerization," Proceedings of the National Academy of Sciences 97, 5877 (2000).

P. L. Geissler, C. Dellago, D. Chandler, J. Hutter, and M. Parrinello, "Ab initio analysis of proton tranfer dynamics in (H₂O) ₃H⁺," Chem. Phys. Lett. 321, 225 (2000).

http://www.cchem.berkeley.edu/~mhggrp