Ab initio Green-Kubo (aiGK) simulations of heat transport in solids allow for assessing lattice thermalconductivity in anharmonic or complex materials from first principles. In this work, we present a detailed accountof their practical application and evaluation with an emphasis on noise reduction and finite-size corrections insemiconductors and insulators. To account for such corrections, we propose strategies in which all necessarynumerical parameters are chosen based on the dynamical properties displayed during molecular dynamicssimulations in order to minimize manual intervention. This paves the way for applying the aiGK method insemiautomated and high-throughput frameworks. The proposed strategies are presented and demonstrated forcomputing the lattice thermal conductivity at room temperature in the mildly anharmonic periclase MgO, andfor the strongly anharmonic marshite CuI.

The anharmonicity of atomic motion limits the thermal conductivity in crystalline solids. However, amicroscopic understanding of the mechanisms active in strong thermal insulators is lacking. In this Letter,we classify 465 experimentally known materials with respect to their anharmonicity and perform fullyanharmonic ab initio Green-Kubo calculations for 58 of them, finding 28 thermal insulators withκ < 10 W=mK including 6 with ultralow κ ≲ 1 W=mK. Our analysis reveals that the underlying stronganharmonic dynamics is driven by the exploration of metastable intrinsic defect geometries. This is atvariance with the frequently applied perturbative approach, in which the dynamics is assumed to evolvearound a single stable geometry.

Machine-learning potentials provide computationally efficient and accurate approximations of the Born–Oppenheimer potential energy surface. This potential determines many materials properties and simulation techniques usually require its gradients, in particular forces and stress for molecular dynamics, and heat flux for thermal transport properties. Recently developed potentials feature high body order and can include equivariant semi-local interactions through message-passing mechanisms. Due to their complex functional forms, they rely on automatic differentiation (AD), overcoming the need for manual implementations or finite-difference schemes to evaluate gradients. This study discusses how to use AD to efficiently obtain forces, stress, and heat flux for such potentials, and provides a model-independent implementation. The method is tested on the Lennard-Jones potential, and then applied to predict cohesive properties and thermal conductivity of tin selenide using an equivariant message-passing neural network potential.

The lanthanum-hydrogen system has attracted significant attention following the report of superconductivity in LaH10 at near-ambient temperatures and high pressures. Phases other than LaH10 are suspected to be synthesized based on both powder X-ray diffraction and resistivity data, although they have not yet been identified. Here, we present the results of our single-crystal X-ray diffraction studies on this system, supported by density functional theory calculations, which reveal an unexpected chemical and structural diversity of lanthanum hydrides synthesized in the range of 50 to 180 GPa. Seven lanthanum hydrides were produced, LaH3, LaH~4, LaH4+δ, La4H23, LaH6+δ, LaH9+δ, and LaH10+δ, and the atomic coordinates of lanthanum in their structures determined. The regularities in rare-earth element hydrides unveiled here provide clues to guide the search for other synthesizable hydrides and candidate high-temperature superconductors. The hydrogen content variability in lanthanum hydrides and the samples’ phase heterogeneity underline the challenges related to assessing potentially superconducting phases and the nature of electronic transitions in high-pressure hydrides.