# Machine Learning Discovery of Record‐Low Lattice Thermal Conductivity in Double Perovskites

## TL;DR

A machine learning model helps discover a double perovskite material with extremely low thermal conductivity, useful for thermoelectric and insulation applications.

## Contribution

A deep learning interatomic potential enables efficient high-throughput screening of phonon properties in double perovskites.

## Key findings

- Cs2HgPtCl6 has a record-low lattice thermal conductivity of 0.071 Wm−1K−1 at room temperature.
- 1,597 dynamically stable double perovskite structures were identified from 9,709 screened compounds.
- Phonon transport calculations including four-phonon scattering confirmed ultralow thermal conductivity.

## Abstract

Double perovskites (ABC2D6) are versatile materials with applications in photovoltaics, optoelectronics, and thermoelectrics, where phonon‐mediated thermal transport is critical. However, high‐throughput phonon calculations by density functional theory (DFT) are computationally prohibitive due to the large supercells required. We develop a deep learning interatomic potential, Elemental‐SDNNFF, trained directly on DFT‐calculated forces within an active learning framework, enabling efficient prediction of phonon properties across thousands of double perovskites. Using this model, we screened 9709 cubic double perovskite structures, identifying 1597 dynamically stable candidates. Their lattice thermal conductivities (LTCs) were predicted by coupling Elemental‐SDNNFF with the Boltzmann Transport Equation, including off‐diagonal contributions. For the most promising compounds, DFT validation and four‐phonon scattering calculations revealed ultralow LTCs (<0.1 Wm−1K−1). Remarkably, Cs2HgPtCl6 was found to possess a bandgap of 0.35 eV and an LTC of 0.071 Wm−1K−1 at room temperature—the lowest ever reported for isotropic bulk materials, comparable to air. The result was independently confirmed by molecular dynamics simulations with a DeePMD potential and phonon lifetime extraction using DynaPhoPy. This work establishes an efficient machine learning‐assisted framework for fast screening of dynamic stability and accurate prediction of phonon transport in complex materials, highlighting double perovskites as Double Perovskites, High‐Throughput DFT, Machine Learning, Phonon Boltzmann Transport, Record‐Low Lattice Thermal Conductivity promising candidates for thermoelectric and thermal insulation applications.

A deep learning interatomic potential is introduced to predict forces for computing phonon properties and thermal transport behavior in double perovskites. Screening 9,709 compounds identifies 1,597 stable materials, and Boltzmann transport calculations including both three and four‐phonon scattering suggests a record‐low lattice thermal conductivity of 0.071Wm‐1K‐1 for Cs2HgPtCl6, which is comparable to air.