arXiv:2609.40135
Smite: A quasiclassical trajectory (QCT) program for bimolecular collisions and unimolecular dynamics on ab-initio, and machine-learned potential energy surfaces
Péter Szabó, Jenne van Veerdeghem, Jérôme Loreau, Jean-François Müller, Jeremy N. Harvey
Abstract
We present Smite, a Python toolkit for quasiclassical trajectory (QCT) simulations of reactive and inelastic bimolecular collisions, unimolecular dynamics, and molecular collisions with finite surface models. A central feature is the explicit control of reactant preparation: harmonic normal modes admit fixed quantum numbers, prescribed energies, thermal quantum populations, or ground-state Wigner sampling; rotation is prepared at fixed angular-momentum magnitude or from thermal distributions appropriate to diatomic and polyatomic rotors. Rotating-Morse sampling provides a coupled anharmonic treatment of diatomic vibration and rotation, while saved molecular-dynamics phase points provide an alternative source of initial conditions. The same nuclear propagators operate with on-the-fly electronic-structure calculations or user-supplied analytical and machine-learned potential energy surfaces. Additional modules provide constrained rigid-fragment dynamics, thermostat-based preparation, photoionization initial conditions with energy and recoil constraints, and fewest-switches surface hopping (FSSH) on multiple PESs with approximate curvature-derived couplings. Geometry optimization, transition-state searches, crossing-point optimization, intrinsic-reaction-coordinate following, and thermochemistry use the same energy and derivative interfaces. Analysis tools connect trajectories to product energy partitioning, stereodynamical correlations, time-resolved vibrational signatures, and independent-atom x-ray scattering.
AI Summary
Smite is a Python toolkit for quasiclassical trajectory simulations of reactive and inelastic molecular collisions, unimolecular dynamics, and collisions with finite surface models. It addresses how to combine explicitly specified reactant ensembles with interchangeable potential-energy surfaces (PESs), propagation methods, and trajectory analysis. The paper describes mode-resolved harmonic preparation using fixed quantum numbers, prescribed energies, thermal populations, or ground-state Wigner sampling; rotational preparation at fixed angular momentum or from thermal distributions; coupled rotating-Morse diatomic sampling; and reuse of molecular-dynamics phase points. A shared energy/derivative interface supports analytical and machine-learned PESs and several electronic-structure backends. Additional modules cover rigid constraints, photoionization launches, stationary-point searches, and fewest-switches surface hopping with curvature-derived approximate couplings. Analysis includes product energy partitioning, stereodynamical correlations, velocity spectra, and independent-atom x-ray scattering. Explicit sampling weights and cross-section/rate estimators connect trajectory records to ensemble observables. The contribution presented is primarily the software framework and its methodological specification: the validation section describes automated checks and example systems, but reports no quantitative benchmark outcomes or performance comparisons. Stated limitations include absent tunneling and interference, zero-point-energy leakage, and the approximate nature of harmonic preparation on anharmonic PESs. Wigner sampling is ground-state only; angular-momentum correction can change prepared mode energies. Surface hopping currently uses PES calculators rather than general electronic-structure coupling interfaces and requires independent validation beyond isolated two-state crossings. General product-action assignment, Gaussian binning, and several ensemble observables require external post-processing; scattering omits bonding-density and electronic-coherence effects.
Error found: The manuscript describes implemented simulation capabilities and repository tests but reports no quantitative validation results demonstrating sampling accuracy, conservation, timestep convergence, or agreement with benchmark dynamics.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.