arXiv subjects

Galen T. Craven

Publications and source records attributed to Galen T. Craven.

At least 19 recordsLinked to original sources

arXiv2026

Stoichiometrically-informed symbolic regression for extracting chemical reaction mechanisms from data

A data-driven computational method is introduced to extract chemical reaction mechanisms from time series chemical concentration data. It is realized through the use of dynamic symbolic regression in which a sparse analytical form for a dynamical system is discoverable from the underlying data. We specifically develop the stoichiometrically-informed symbolic regression (SISR) method to address a standing challenge in complex chemical reaction networks: Given a time-series dataset of concentrations of several components, what is the mechanism and the associated rate constants? SISR finds the optimal mechanism, kinetic equations and rate constants by combining differential optimization with a genetic optimization approach that searches a symbolic space of possible reaction mechanisms. Use of SISR in several paradigmatic examples spanning linear and nonlinear reaction schemes results in excellent agreement between true and predicted mechanisms, including when the method is applied to noisy data. The advantages of a stoichiometrically-informed approach such as SISR to address reaction discovery is illustrated through comparison with the use of generic state-of-the-art data-driven approaches.

physics.chem-ph↗

arXiv2025

Machine learning method to determine concentrations of structural defects in irradiated materials

The formation and subsequent growth of structural defects in an irradiated material can strongly influence the material's performance in technological and industrial applications. Predicting how the growth of defects affects material performance is therefore a pressing problem in materials science. One common computational approach that is used to examine defect growth is cluster dynamics, a method which employs a system of mean-field rate equations to track the time evolution of concentrations of individual defect types. However, the computational complexity of performing cluster dynamics can limit its practical implementation, specifically in the context of exploring a broad set of physical conditions corresponding to, for example, different temperatures and pressures. Here, we present a machine learning approach to circumvent the computational challenges of performing cluster dynamics while maintaining high accuracy in the prediction of defect concentrations. The method is illustrated on the nuclear material uranium nitride but is broadly applicable to other materials. The developed data-driven method is shown to accurately capture complex correlations between material properties, temperature, irradiation conditions, and the concentration of defects.

cond-mat.mtrl-sci↗

arXiv2025

Neuromorphic heat transport effects in a molecular junction

Understanding energy transport at the nanoscale is an open and fundamental challenge in the molecular sciences with direct implications for the design of new electronics, computing devices, and materials. While nanoscale energy transport under steady-state conditions has been studied extensively, there is much less known about energy transport under time-dependent driving forces, particularly in the far-from-equilibrium regime. In this work, we use nonequilibrium molecular dynamics simulations and stochastic thermodynamics to investigate energy transport in a well-studied nanoscale system, a molecular junction, subjected to a time-periodic temperature gradient. The primary observation is that molecular junctions can exhibit heat transport hysteresis, a phenomenon in which the heat flux through a system depends not only on the instantaneous value of a time-dependent temperature bias but also on the temporal history of that bias. The presented findings illustrate that molecular junctions can exhibit the specific memory effect, heat transport hysteresis, that is essential for the design of thermal neuromorphic computers. This work elucidates a potential pathway toward the realization of such devices.

cond-mat.mes-hall↗

arXiv2025

Teacher-student training improves accuracy and efficiency of machine learning interatomic potentials

Machine learning interatomic potentials (MLIPs) are revolutionizing the field of molecular dynamics (MD) simulations. Recent MLIPs have tended towards more complex architectures trained on larger datasets. The resulting increase in computational and memory costs may prohibit the application of these MLIPs to perform large-scale MD simulations. Here, we present a teacher-student training framework in which the latent knowledge from the teacher (atomic energies) is used to augment the students' training. We show that the light-weight student MLIPs have faster MD speeds at a fraction of the memory footprint compared to the teacher models. Remarkably, the student models can even surpass the accuracy of the teachers, even though both are trained on the same quantum chemistry dataset. Our work highlights a practical method for MLIPs to reduce the resources required for large-scale MD simulations.

physics.chem-ph↗

arXiv2025

Ensemble Knowledge Distillation for Machine Learning Interatomic Potentials

The quality of machine learning interatomic potentials (MLIPs) strongly depends on the quantity of training data as well as the quantum chemistry (QC) level of theory used. Datasets generated with high-fidelity QC methods are typically restricted to small molecules and may be missing energy gradients, which make it difficult to train accurate MLIPs. We present an ensemble knowledge distillation (EKD) method to improve MLIP accuracy when trained to energy-only datasets. First, multiple teacher models are trained to QC energies and then generate atomic forces for all configurations in the dataset. Next, the student MLIP is trained to both QC energies and to ensemble-averaged forces generated by the teacher models. We apply this workflow on the ANI-1ccx dataset where the configuration energies computed at the coupled cluster level of theory. The resulting student MLIPs achieve new state-of-the-art accuracy on the COMP6 benchmark and show improved stability for molecular dynamics simulations.

physics.chem-ph↗

arXiv2025

Heat Transport Hysteresis Generated through Frequency Switching of a Time-Dependent Temperature Gradient

A stochastic energetics framework is applied to examine how periodically shifting the frequency of a time-dependent oscillating temperature gradient affects heat transport in a nanoscale molecular model. We specifically examine the effects that frequency switching, i.e., instantaneously changing the oscillation frequency of the temperature gradient, has on the shape of the heat transport hysteresis curves generated by a particle connected to two thermal baths, each with a temperature that is oscillating in time. Analytical expressions are derived for the energy fluxes in/out of the system and the baths, with excellent agreement observed between the analytical expressions and the results from nonequilibrium molecular dynamics simulations. We find that the shape of the heat transport hysteresis curves can be significantly altered by shifting the frequency between fast and slow oscillation regimes. We also observe the emergence of features in the hysteresis curves such as pinched loops and complex multi-loop patterns due to the frequency shifting. The presented results have implications in the design of thermal neuromorphic devices such as thermal memristors and thermal memcapacitors.

cond-mat.mes-hall↗

arXiv2024

Thermodynamic Transferability in Coarse-Grained Force Fields using Graph Neural Networks

Coarse-graining is a molecular modeling technique in which an atomistic system is represented in a simplified fashion that retains the most significant system features that contribute to a target output, while removing the degrees of freedom that are less relevant. This reduction in model complexity allows coarse-grained molecular simulations to reach increased spatial and temporal scales compared to corresponding all-atom models. A core challenge in coarse-graining is to construct a force field that represents the interactions in the new representation in a way that preserves the atomistic-level properties. Many approaches to building coarse-grained force fields have limited transferability between different thermodynamic conditions as a result of averaging over internal fluctuations at a specific thermodynamic state point. Here, we use a graph-convolutional neural network architecture, the Hierarchically Interacting Particle Neural Network with Tensor Sensitivity (HIP-NN-TS), to develop a highly automated training pipeline for coarse grained force fields which allows for studying the transferability of coarse-grained models based on the force-matching approach. We show that this approach not only yields highly accurate force fields, but also that these force fields are more transferable through a variety of thermodynamic conditions. These results illustrate the potential of machine learning techniques such as graph neural networks to improve the construction of transferable coarse-grained force fields.

physics.chem-ph↗

arXiv2024

The effect of temperature oscillations on energy storage rectification in harmonic systems

Rectification, the preferential transport of a current in one direction through a system, has garnered significant attention in molecules because of its importance for controlling thermal and electronic currents at the nanoscale. Here, we report the presence of energy storage rectification effects in a molecular chain. This phenomenon is generated by subjecting a harmonic molecular chain to an oscillating temperature gradient and showing that the energy absorption rate of the system depends on the direction of the gradient. We examine how the energy storage rectification ratios in the chain are affected by the oscillating gradient, asymmetry in the chain, and the system parameters. We find that energy storage rectification can be observed in harmonic lattice structures with time-dependent temperatures and that, correspondingly, anharmonicity is not required to generate this rectification mechanism in such systems.

cond-mat.mes-hall↗

arXiv2024

Molecular heat transport across a time-periodic temperature gradient

The time-periodic modulation of a temperature gradient can alter the heat transport properties of a physical system. Oscillating thermal gradients give rise to behaviors such as modified thermal conductivity and controllable time-delayed energy storage that are not present in a system with static temperatures. Here, we examine how the heat transport properties of a molecular lattice model are affected by an oscillating temperature gradient. We use analytical analysis and molecular dynamics simulations to investigate the vibrational heat flow in a molecular lattice system consisting of a chain of particles connected to two heat baths at different temperatures, where the temperature difference between baths is oscillating in time. We derive expressions for heat currents in this system using a stochastic energetics framework and a nonequilibrium Green's function approach that is modified to treat the nonstationary average energy fluxes. We find that emergent energy storage, energy release, and thermal conductance mechanisms induced by the temperature oscillations can be controlled by varying the frequency, waveform, and amplitude of the oscillating gradient. The developed theoretical approach provides a general framework to describe how vibrational heat transmission through a molecular lattice is affected by temperature gradient oscillations.

cond-mat.mes-hall↗

arXiv2024

Optical and Transport Properties of Plasma Mixtures from Ab Initio Molecular Dynamics

Predicting the charged particle transport properties of warm dense matter / hot dense plasma mixtures is a challenge for analytical models. High accuracy ab initio methods are more computationally expensive, but can provide critical insight by explicitly simulating mixtures. In this work, we investigate the transport properties and optical response of warm dense carbon-hydrogen mixtures at varying concentrations under either conserved electronic pressure or mass density at a constant temperature. We compare options for mixing the calculated pure species properties to estimate the results of the mixtures. We find that a combination of the Drude model with the Matthiessen's rule works well for DC electron transport and low frequency optical response. This breaks down at higher frequencies, where a volumetric mix of pure-species AC conductivities works better.

physics.plasm-ph↗

arXiv2023

Data-driven methods for diffusivity prediction in nuclear fuels

The growth rate of structural defects in nuclear fuels under irradiation is intrinsically related to the diffusion rates of the defects in the fuel lattice. The generation and growth of atomistic structural defects can significantly alter the performance characteristics of the fuel. This alteration of functionality must be accurately captured to qualify a nuclear fuel for use in reactors. Predicting the diffusion coefficients of defects and how they impact macroscale properties such as swelling, gas release, and creep is therefore of significant importance in both the design of new nuclear fuels and the assessment of current fuel types. In this article, we apply data-driven methods focusing on machine learning (ML) to determine various diffusion properties of two nuclear fuels, uranium oxide and uranium nitride. We show that using ML can increase, often significantly, the accuracy of predicting diffusivity in nuclear fuels in comparison to current analytical models. We also illustrate how ML can be used to quickly develop fuel models with parameter dependencies that are more complex and robust than what is currently available in the literature. These results suggest there is potential for ML to accelerate the design, qualification, and implementation of nuclear fuels.

cond-mat.mtrl-sci↗

arXiv2023

Energy transport between heat baths with oscillating temperatures

Energy transport is a fundamental physical process that plays a prominent role in the function and performance of myriad systems and technologies. Recent experimental measurements have shown that subjecting a macroscale system to a time-periodic temperature gradient can increase thermal conductivity in comparison to a static temperature gradient. Here, we theoretically examine this mechanism in a nanoscale model by applying a stochastic Langevin framework to describe the energy transport properties of a particle connecting two heat baths with different temperatures, where the temperature difference between baths is oscillating in time. Analytical expressions for the energy flux of each heat bath and for the system itself are derived for the case of a free particle and a particle in a harmonic potential. We find that dynamical effects in the energy flux induced by temperature oscillations give rise to complex energy transport hysteresis effects. The presented results suggest that applying time-periodic temperature modulations is a potential route to control energy storage and release in molecular devices and nanosystems.

cond-mat.stat-mech↗

arXiv2023

Analytical solutions of the Arrhenius-Semenov problem for constant volume burn

Analytical solutions to the Semenov thermal ignition problem for constant volume burn governed by Arrhenius reaction kinetics are derived. Specifically, an approximate analytical solution technique for the Arrhenius-Semenov differential equation is derived for reaction orders $n \in \mathbb{R}_{>0}$ and exact solutions are also constructed for reaction orders $n \in \mathbb{N}: n \leq 3$. The approximation technique relies on expansion of the respective nondominant terms in the differential equation at the lower and upper bounds of the reaction progress variable in order to create a pair of integrable series. The two integrated series are then connected to create a single continuous analytical solution. Excellent agreement is observed between the analytical approximation and solutions obtained numerically. The presented approximation constitutes a simple and robust strategy for solving the Arrhenius-Semenov problem analytically.

physics.chem-ph↗

arXiv2023

Electron hopping heat transport in molecules

The realization of single-molecule thermal conductance measurements has driven the need for theoretical tools to describe conduction processes that occur over atomistic length scales. In macroscale systems, the principle that is typically used to understand thermal conductivity is Fourier's law. At molecular length scales, however, deviations from Fourier's law are common in part because microscale thermal transport properties typically depend on the complex interplay between multiple heat conduction mechanisms. Here, the thermal transport properties that arise from electron transfer across a thermal gradient in a molecular conduction junction are examined theoretically. We illustrate how transport in a model junction is affected by varying the electronic structure and length of the molecular bridge in the junction as well as the strength of the coupling between the bridge and its surrounding environment. Three findings are of note: First, the transport properties can vary significantly depending on the characteristics of the molecular bridge and its environment; second, the system's thermal conductance commonly deviates from Fourier's law; and third, in properly engineered systems, the magnitude of electron hopping thermal conductance is similar to what has been measured in single-molecule devices.

cond-mat.mes-hall↗

arXiv2019

Wiedemann-Franz Law for Molecular Hopping Transport

The Wiedemann-Franz (WF) law is a fundamental result in solid-state physics that relates the thermal and electrical conductivity of a metal. It is derived from the predominant origin of energy conversion in metals: the motion of quasi-free charge-carrying particles. Here, an equivalent WF relationship is developed for molecular systems in which charge carriers are moving not as free particles but instead hop between redox sites. We derive a concise analytical relationship between the electrical and thermal conductivity generated by electron hopping in molecular systems and find that the linear temperature dependence of their ratio as expressed in the standard WF law is replaced by a linear dependence on the nuclear reorganization energy associated with the electron hopping process. The robustness of the molecular WF relation is confirmed by examining the conductance properties of a paradigmatic molecular junction. This result opens a new way to analyze conductivity in molecular systems, with possible applications advancing the design of molecular technologies that derive their function from electrical and/or thermal conductance.

cond-mat.mes-hall↗

arXiv2019

Electron transfer at thermally heterogeneous molecule-metal interfaces

The rate of electron transfer between a molecular species and a metal, each at a different local temperature, is examined theoretically through implementation of a bithermal (characterized by two temperatures) Marcus formalism. Expressions for the rate constant and the electronic contribution to a heat transfer mechanism which is induced by the temperature gradient between molecule and metal are constructed. The system of coupled dynamical equations describing the electronic and thermal currents are derived and examined over diverse ranges of reaction geometries and temperature gradients. It is shown that electron transfer across the molecule-metal interface is associated with heat transfer and that the electron exchange between metal and molecule makes a distinct contribution to the interfacial heat conduction even when the net electronic current vanishes.

cond-mat.mes-hall↗

arXiv2019

Upside/Downside statistical mechanics of nonequilibrium Brownian motion. I. Distributions, moments, and correlation functions of a free particle

Statistical properties of Brownian motion that arise by analyzing, separately, trajectories over which the system energy increases (upside) or decreases (downside) with respect to a threshold energy level, are derived. This selective analysis is applied to examine transport properties of a nonequilibrium Brownian process that is coupled to multiple thermal sources characterized by different temperatures. Distributions, moments, and correlation functions of a free particle that occur during upside and downside events are investigated for energy activation and energy relaxation processes, and also for positive and negative energy fluctuations from the average energy. The presented results are sufficiently general and can be applied without modification to standard Brownian motion. This article focuses on the mathematical basis of this selective analysis. In subsequent articles in this series we apply this general formalism to processes in which heat transfer between thermal reservoirs is mediated by activated rate processes that take place in a system bridging them.

cond-mat.stat-mech↗

arXiv2019

Upside/Downside statistical mechanics of nonequilibrium Brownian motion. II. Heat transfer and energy partitioning of a free particle

The energy partitioning during activation and relaxation events under steady-state conditions for a Brownian particle driven by multiple thermal reservoirs of different local temperatures is investigated. Specifically, we apply the formalism derived in a previous article [G. T. Craven and A. Nitzan, J. Chem. Phys. 148, 044101 (2018)] to examine the thermal transport properties of two sub-ensembles of Brownian processes, distinguished at any given time by the specification that all the trajectories in each group have, at that time, energy either above (upside) or below (downside) a preselected energy threshold. Dynamical properties describing energy accumulation and release during activation/relaxation events and relations for upside/downside energy partitioning between thermal reservoirs are derived. The implications for heat transport induced by upside and downside events are discussed.

cond-mat.stat-mech↗