Publications de Changpeng Lin | EPFL Graph Search

Corrigendum to “Rapid prediction of phonon density of states by crystal attention graph neural network and high-throughput screening of candidate substrates for wide bandgap electronic cooling” [Mater. Today Phys. 50 (2025) 101632] (Materials Today Physics (2025) 50, (S2542529324003080), (10.1016/j.mtphys.2024.101632))

The authors regret that Figs. 1–3 in the published article were wrongly uploaded when the article was in the proof stage. The correct Figs. 1–3 are now presented in this Corrigendum. The figures shown here are the same ones that were reviewed by the review ...

2025

Synergistically optimizing electrical and thermal transport in layered BiCuSeO via biaxial strain modulation

Changpeng Lin

Decoupling electrical and thermal parameters in figure of merit (zT) is still a major challenge for thermoelectric materials. Layered oxyselenides have the advantages of superlattice structures, including the quantum confinement effect and independent domi ...

2025

Rapid prediction of phonon density of states by crystal attention graph neural network and high-throughput screening of candidate substrates for wide bandgap electronic cooling

Changpeng Lin

Machine learning has demonstrated superior performance in predicting vast materials properties. However, predicting a spectral-like continuous material property such as phonon density of states (DOS) is more challenging for machine learning. In this work, ...

2025

Unravelling ultralow thermal conductivity in perovskite Cs2AgBiBr6: dominant wave-like phonon tunnelling and strong anharmonicity

Changpeng Lin

Understanding the lattice dynamics and heat transport physics in the lead-free halide double perovskites remains an outstanding challenge due to their lattice dynamical instability and strong anharmonicity. In this work, we investigate the microscopic mech ...

Nature Portfolio2024

Carrier-phonon decoupling in perovskite thermoelectrics via entropy engineering

Changpeng Lin

Thermoelectrics converting heat and electricity directly attract broad attentions. To enhance the thermoelectric figure of merit, zT, one of the key points is to decouple the carrier-phonon transport. Here, we propose an entropy engineering strategy to rea ...

Nature Research2024

Unlocking phonon properties of a large and diverse set of cubic crystals by indirect bottom-up machine learning approach

Changpeng Lin

Although first principles based anharmonic lattice dynamics is one of the most common methods to obtain phonon properties, such method is impractical for high-throughput search of target thermal materials. We develop an elemental spatial density neural net ...

SPRINGERNATURE2023

Million-scale data integrated deep neural network for phonon properties of heuslers spanning the periodic table

Changpeng Lin, Hong Zhang, Chen Shen, Yong Zhao

Existing machine learning potentials for predicting phonon properties of crystals are typically limited on a material-to-material basis, primarily due to the exponential scaling of model complexity with the number of atomic species. We address this bottlen ...

NATURE PORTFOLIO2023

Anti-Jahn-Teller effect induced ultrafast insulator to metal transition in perovskite BaBiO3

Changpeng Lin, Ben Xu

The Jahn-Teller (JT) effect involves the ions M with a degenerate electronic state distorting the corner-sharing MO6 octahedra to lift the degeneracy, inducing strong coupling of electrons to lattice, and mediating the exotic properties in perovskite oxide ...

NATURE PORTFOLIO2022

General invariance and equilibrium conditions for lattice dynamics in 1D, 2D, and 3D materials

Nicola Marzari, Changpeng Lin, Samuel Poncé

The long-wavelength behavior of vibrational modes plays a central role in carrier transport, phonon-assisted optical properties, superconductivity, and thermomechanical and thermoelectric properties of materials. Here, we present general invariance and equ ...

NATURE PORTFOLIO2022