PINK: physical-informed machine learning for lattice thermal conductivity

Abstract

Lattice thermal conductivity (${\kappa }_L$) is crucial for efficient thermal management in electronics and energy conversion technologies. Traditional methods for predicting \k{appa}L are often computationally expensive, limiting their scalability for large-scale material screening. Empirical models, such as the Slack model, offer faster alternatives but require time-consuming calculations for key parameters such as sound velocity and the Gruneisen parameter. This work presents a high-throughput framework, physical-informed kappa (PINK), which combines the predictive power of crystal graph convolutional neural networks (CGCNNs) with the physical interpretability of the Slack model to predict \k{appa}L directly from crystallographic information files (CIFs). Unlike previous approaches, PINK enables rapid, batch predictions by extracting material properties such as bulk and shear modulus from CIFs using a well-trained CGCNN model. These properties are then used to compute the necessary parameters for ${\kappa }_L$ calculation through a simplified physical formula. PINK was applied to a dataset of 377,221 stable materials, enabling the efficient identification of promising candidates with ultralow ${\kappa }_L$ values, such as Ag${}_3$Te${}_4$W and Ag${}_3$Te${}_4$Ta. The platform, accessible via a user-friendly interface, offers an unprecedented combination of speed, accuracy, and scalability, significantly accelerating material discovery for thermal management and energy conversion applications.

21 pages, 10 figures

References in corpus (23)

• BoltzTraP. A code for calculating band-structure dependent quantities
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• AFLOW: An automatic framework for high-throughput materials discovery
• Phononics: Manipulating heat flow with electronic analogs and beyond
• Atomistic Line Graph Neural Network for Improved Materials Property Predictions
• Universal Fragment Descriptors for Predicting Electronic Properties of Inorganic Crystals
• Self-consistent phonon calculations of lattice dynamical properties in cubic SrTiO${}_3$ with first-principles anharmonic force constants
• Lattice anharmonicity and thermal conductivity from compressive sensing of first-principles calculations
• Machine-learning interatomic potentials for materials science
• Benchmarking Materials Property Prediction Methods: The Matbench Test Set and Automatminer Reference Algorithm
• Force and heat current formulas for many-body potentials in molecular dynamics simulation with applications to thermal conductivity calculations
• Global Attention based Graph Convolutional Neural Networks for Improved Materials Property Prediction
• Orbital Graph Convolutional Neural Network for Material Property Prediction
• Ab initio Green-Kubo Approach for the Thermal Conductivity of Solids
• Anharmonicity Measure for Materials
• The role of high-order anharmonicity and off-diagonal terms in thermal conductivity: a case study of multi-phase CsPbBr3
• Exploring diamond-like lattice thermal conductivity crystals via feature-based transfer learning
• Combining the AFLOW GIBBS and Elastic Libraries for efficiently and robustly screening thermo-mechanical properties of solids
• Machine learning for predicting ultralow thermal conductivity and high ZT in complex thermoelectric materials
• An interpretable formula for lattice thermal conductivity of crystals
• Kirigami-Inspired Thermal Regulator
• Anomalous thermal conductivity in 2D silica nanocages of immobilizing noble gas atom

Cited by in corpus (2)

• Materials Informatics: Emergence To Autonomous Discovery In The Age Of AI
• LLM-driven discovery for carbon allotropes with bond-network entropy