3:45–5:00 pm
GCIS W301/303 929 E 57th St
Speaker: Prof. Nancy Makri, University of Illinois at Urbana-Champaign
Title: Real-Time Path Integral Methodology for Large Systems in Condensed-Phase or Biological Environments
Since the early 1990s, the iterative, tensor-based quasi-adiabatic propagator path integral (QuAPI) methodology has enabled numerically exact, fully quantum mechanical simulations of system-bath dynamics. Recent work showed that one can further disentangle the path integral variables through the rigorous small matrix decomposition (SMatPI), which leads to iterative propagation with matrices of minimal size equal to that of the system's reduced density matrix (RDM). By eliminating tensor storage, the SMatPl algorithm enables the simulation of multistate systems and long-memory processes. Combination with modular path integral (MPI) decompositions enables iteration in space and time, leading to linear scaling with system size and propagation length. The methodology has been extended to systems coupled to some anharmonic environments. Recent work has shown that the quantum paths entering the path integral expression can be mapped on graphs, whose properties can be exploited to arrive at novel algorithms that allow the efficient treatment of static disorder and the simulation of processes that involve thousands of system states (regardless of the nature of the system Hamiltonian).
Faculty Host: Dipti Jasrasaria, Department of Chemistry