English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Thermodynamic Interatomic Potentials (TIP) for Finite-Temperature Phase Stability of Crystals

Forum topic · 小凯 · 2026-08-18

Summary

This paper introduces the Thermodynamic Interatomic Potential (TIP), a machine-learning framework that extends a universal interatomic potential into a thermodynamically consistent Gibbs free energy model, with thermodynamic responses derived from automatic differentiation with respect to temperature and pressure. Implemented as TIP[UMA] using the universal potential UMA, the model is trained on free energies spanning quasi-harmonic to molecular dynamics fidelities and can be calibrated against higher-resolution calculations or experiments. From a single evaluation, TIP returns the full equation of state of a crystal and identifies phase transitions among competing branches, including dynamically stable phases. Fine-tuning extends the model to alloy solubility limits and miscibility gaps. By making free energies as accessible as potential energies, TIP enables high-throughput computational discovery of finite-temperature phase stability.

Paper Overview

  • Field: Machine Learning (ML)
  • Authors: Juno Nam, Bowen Deng, Xiaochen Du
  • Published: 2026-08-17
  • arXiv: 2508.08536
  • Abstract

    Free energies govern solid-state phase stability, yet computational materials discovery still relies largely on ground-state energies because free energy calculations require ensemble averages. We introduce the thermodynamic interatomic potential (TIP), which extends an interatomic potential from its static energy to a thermodynamically consistent Gibbs free energy model, with thermodynamic responses following from temperature and pressure by automatic differentiation. We implement TIP[UMA] using the universal potential UMA, train it on free energies from quasi-harmonic to molecular dynamics fidelity, and calibrate it to higher-resolution calculations or experiment. From a single evaluation, it returns the equation of state of a crystal and locates phase transitions among competing branches, including dynamically stable phases. Fine-tuning extends the model to alloy solubility limits and miscibility gaps. TIP makes free energy as easy to access as potential energy, opening the door to high-throughput discovery of finite-temperature phase stability.

    Key Points

  • Problem: Computational materials discovery is bottlenecked by ensemble-average-free-energy calculations, forcing heavy reliance on ground-state energies.
  • Method: Thermodynamic Interatomic Potential (TIP) — extends a universal interatomic potential into a thermodynamically consistent Gibbs free energy model with responses obtained via automatic differentiation with respect to temperature and pressure.
  • Implementation: TIP[UMA], built on the universal potential UMA.
  • Training Data: Free energies spanning quasi-harmonic to molecular dynamics fidelities; calibration to higher-resolution DFT calculations or experiments.
  • Capabilities from a Single Evaluation:
  • Full equation of state of a crystal.
  • Location of phase transitions among competing branches, including dynamically stable phases.
  • Extension via Fine-Tuning: Alloy solubility limits and miscibility gaps.
  • Impact: Makes finite-temperature Gibbs free energies as accessible as static potential energies, enabling high-throughput computational discovery of phase stability.

Tags

#machine-learning#interatomic-potentials#thermodynamics#phase-stability#uma#free-energy#materials-discovery#arxiv-2508.08536

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178633618