Machine Learning for Improved Density Functional Theory Thermodynamics

2025-03-07

AI Summary

Density functional theory (DFT) predictions of alloy formation enthalpies suffer from insufficient energy resolution, leading to substantial errors in ternary phase stability calculations. To address this, we propose a physics-informed neural network correction framework: for the first time, it integrates structured elemental features—such as atomic number, electronegativity, and atomic radius—with concentration-dependent chemical interaction terms, yielding an interpretable and compositionally generalizable enthalpy-difference prediction model. A multilayer perceptron regressor is employed, trained on rigorously curated datasets and validated via leave-one-out and k-fold cross-validation. Applied to the Al–Ni–Pd and Al–Ni–Ti systems, the method significantly improves phase diagram accuracy, reducing the mean absolute error between predicted and experimental formation enthalpies by over 60%. This work establishes the first ab initio thermodynamic correction paradigm that simultaneously ensures physical consistency and high predictive reliability.

Technology Category

• Machine Learning: Feature Construction/Reformulation
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The predictive accuracy of density functional theory (DFT) for alloy formation enthalpies is often limited by intrinsic energy resolution errors, particularly in ternary phase stability calculations. In this work, we present a machine learning (ML) approach to systematically correct these errors, improving the reliability of first-principles predictions. A neural network model has been trained to predict the discrepancy between DFT-calculated and experimentally measured enthalpies for binary and ternary alloys and compounds. The model utilizes a structured feature set comprising elemental concentrations, atomic numbers, and interaction terms to capture key chemical and structural effects. By applying supervised learning and rigorous data curation we ensure a robust and physically meaningful correction. The model is implemented as a multi-layer perceptron (MLP) regressor with three hidden layers, optimized through leave-one-out cross-validation (LOOCV) and k-fold cross-validation to prevent overfitting. We illustrate the effectiveness of this method by applying it to the Al-Ni-Pd and Al-Ni-Ti systems, which are of interest for high-temperature applications in aerospace and protective coatings.

Problem

• Corrects DFT errors in alloy formation enthalpy predictions.
• Uses ML to improve ternary phase stability calculations.
• Enhances reliability of first-principles thermodynamic predictions.

Innovation

• Machine learning corrects DFT enthalpy errors
• Neural network predicts DFT-experiment discrepancy
• MLP regressor optimized with cross-validation