Maximizing the Accuracy of QCT Cross Sections via Binning Methods

Faculty Sponsor: Brian Stewart

Jacob Ruben

Jacob Ruben is a rising junior (class of ’27) with a major in physics. He is conducting research in simulations of molecular collisions with Professor Stewart in the Physics Department. He is from New York City, and outside of school he likes swimming, playing basketball, and hiking.

Abstract: In molecular collision studies, there are some phenomena that can be captured only by quantum mechanics. Quantum simulations, however, are much more computationally expensive than those using classical mechanics. Quasiclassical Trajectory (QCT) simulations require the binning of the continuous classical actions into discrete quantum mechanically-allowed rovibrational states. Traditionally, this takes the form of standard histogram binning, or a Gaussian weighting function with a full width at half maximum (FWHM) of 0.1. We use microreversibility, a necessary consequence of time reversal symmetry, as an internal consistency check on calculated cross sections. For initially nonrotating molecules, trajectories do not fully populate the bin centered around the initial state, causing standard histogram binning to sample too large an area, resulting in poor microreversibility ratios. Microreversibility is greatly improved using a Gaussian weighting function, but the traditional narrow Gaussian causes a drastic overestimation of state-to-state cross sections. We propose different binning strategies for different regions of rovibrational action space, emphasizing microreversibility ratios close to one while retaining fidelity to computed quantum mechanical cross sections.

summer2025poster_JR