Smite: A quasiclassical trajectory (QCT) program for bimolecular collisions and unimolecular dynamics on ab-initio, and machine-learned potential energy surfaces
The paper introduces Smite, a versatile Python toolkit for quasiclassical trajectory simulations of bimolecular and unimolecular dynamics that supports flexible initial state preparation, on-the-fly or machine-learned potential energy surfaces, non-adiabatic surface hopping, and comprehensive analysis tools.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
For decades, chemists have sought to understand the invisible choreography of atoms as they collide, break apart, and reassemble into new substances. At the heart of this inquiry lies a simple but profound idea: if you know the forces that push and pull on every atom, you can predict how a chemical reaction will unfold. This requires two things. First, a map of the energy landscape, known as a potential energy surface, which shows the hills and valleys that atoms must traverse. Second, a way to simulate the motion of those atoms as they roll across this terrain. While quantum mechanics offers the most precise description of these movements, it is often too computationally heavy to handle anything more complex than a handful of atoms. For larger molecules, scientists have long relied on a practical middle ground called quasiclassical trajectory simulations. This approach treats the heavy atomic nuclei as classical particles moving along definite paths, while still respecting the quantum rules that govern their starting positions and energies. It has become a standard tool for watching chemical reactions play out in slow motion, revealing details that experiments alone cannot capture.
Into this field steps a new software toolkit called Smite, developed by researchers at the Belgian Institute for Space Aeronomy and KU Leuven. The program is designed to be a versatile engine for simulating how molecules collide, react, or simply bounce off one another. Unlike previous tools that often required researchers to build custom code for each new problem, Smite brings together the preparation of starting conditions, the actual motion of atoms, and the analysis of the results into a single, unified framework. The researchers built this system to handle a wide variety of scenarios, from simple collisions between two molecules to the complex breakup of a single unstable molecule, and even interactions between a gas and a solid surface. Its core strength lies in its ability to prepare reactants with extreme precision, allowing scientists to set the exact vibration and rotation of molecules before they collide, or to start from a snapshot of a previous computer simulation. This flexibility lets researchers test specific hypotheses about how energy moves through a system, such as whether a molecule reacts faster when it is vibrating in a particular way.
The program distinguishes itself by offering multiple ways to set the initial state of a molecule. A user can choose to start with a molecule vibrating at a specific quantum level, or with a distribution of energies that mimics a hot gas. It can also generate starting points based on the ground state of a molecule using a statistical method that accounts for the inherent uncertainty of quantum mechanics. For diatomic molecules, like oxygen or nitrogen, the software can treat the vibration and rotation as a coupled system, acknowledging that spinning a molecule faster can stretch its bond. Furthermore, Smite can take a saved snapshot from a previous, long-running simulation and use it as a starting point, preserving complex, anharmonic motions that simple models might miss. Once the molecules are set in motion, the software can calculate their path using either pre-made mathematical maps of the energy landscape or by calling up powerful electronic structure calculations on the fly. This means it can handle reactions where the energy surface is too complex to be written down in a simple formula, allowing for the study of systems with ten or more atoms with high accuracy.
Beyond just watching atoms move, Smite includes tools to analyze what happens when the simulation ends. It can identify which atoms have bonded together to form new products and calculate how the energy is shared among them. It tracks the direction of the scattered molecules and the rotation they acquire, providing a detailed picture of the collision's outcome. The software also includes specialized modules for studying how molecules interact with light, such as when a neutral molecule is hit by a photon and becomes an ion, or when a molecule hops between different electronic states during a reaction. These features allow researchers to simulate conditions relevant to combustion, atmospheric chemistry, and even the chemistry of interstellar space. For instance, the program can model how organic radicals capture oxygen in the atmosphere to form aerosols, or how molecules in a star's atmosphere absorb light during collisions. By providing a clear, modular way to set up and analyze these events, the software helps scientists move beyond statistical averages to understand the specific, individual pathways that lead to chemical change.
The developers tested Smite on a variety of known systems to ensure its reliability, comparing its results against established data and other simulation methods. They verified that the program correctly reproduces the expected distributions of energy and momentum for simple collisions and that it can handle the complex dynamics of reactions involving multiple pathways. The software is designed to be open and accessible, with its code available for other scientists to use and improve. By making these advanced simulation techniques easier to apply and more flexible, Smite offers a powerful new way to explore the microscopic world of chemical reactions. It does not claim to solve every problem in reaction dynamics, but it provides a robust and transparent platform for investigating the detailed mechanisms that govern how matter transforms, from the flames of a fire to the chemistry of distant planets.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.