SnTe alloys, which have the same crystal structure as PbTe, have attracted increasing attention. Here, we demonstrate that the synergistic effect of band structure modification and chemical bond softening can be realized simultaneously in In & Mn doped SnTe bulk alloys. The Seebeck coefficient and power factor are synergistically improved by co-doping of In and Mn. In doping is known to introduce a resonance level. Mn doping reduces the separation of light- and heavy-valence bands. The combination of these effects significantly enhances the Seebeck coefficient at room temperature owing to around a factor of five times increase in the band effective mass. The reduction of thermal conductivity is from the decrease of both the electronic and phononic parts. The electronic thermal conductivity is decreased by the increase in defect scattering, as can be confirmed by the carrier mobility. The force constant of the bonds around the Te site is decreased due to the co-doping of In & Mn, which indicates that the chemical bonds are softened, which leads to low sound velocity and lower lattice thermal conductivity. As a result, the peak thermoelectric figure of merit, zT = 1.03 has been achieved for Sn_0.89In_0.01Mn_0.1Te at 923 K. This strategy of using the synergistic effect of band structure modification and chemical bond softening could be applicable to other thermoelectric materials.

Interatomic potentials based on neural-network machine learning (ML) approach to address the long-standing challenge of accuracy versus efficiency in molecular-dynamics simulations have recently attracted a great deal of interest. Here, utilizing Pd-Si system as a prototype, we extend the development of neural-network ML potentials to compounds exhibiting various types of bonding characteristics. The ML potential is trained by fitting to the energies and forces of both liquid and crystal structures first-principles calculations based on density-functional theory (DFT). We show that the generated ML potential captures the structural features and motifs in $\mathrm{P}{\mathrm{d}}_{82}\mathrm{S}{\mathrm{i}}_{18}$ and $\mathrm{P}{\mathrm{d}}_{75}\mathrm{S}{\mathrm{i}}_{25}$ liquids more accurately than the existing interatomic potential based on embedded-atom method (EAM). The ML potential also describes the solid-liquid interface of these systems very well. Moreover, while the existing EAM potential fails to describe the relative energies of various crystalline structures and predict wrong ground-state structures at $\mathrm{P}{\mathrm{d}}_3\mathrm{Si}$ and $\mathrm{P}{\mathrm{d}}_9\mathrm{S}{\mathrm{i}}_2$ composition, the developed ML potential predicts correctly the ground-state structures from genetic algorithm search. The efficient ML potential with DFT accuracy from our study will provide a promising scheme for accurate atomistic simulations of structures and dynamics of complex Pd-Si system.

Magnetic materials are essential for energy generation and information devices, and they play an important role in advanced technologies and green energy economies. Currently, the most widely used magnets contain rare earth (RE) elements. An outstanding challenge of notable scientific interest is the discovery and synthesis of novel magnetic materials without RE elements that meet the performance and cost goals for advanced electromagnetic devices. Here, we report our discovery and synthesis of an RE-free magnetic compound, Fe ₃ CoB ₂ , through an efficient feedback framework by integrating machine learning (ML), an adaptive genetic algorithm, first-principles calculations, and experimental synthesis. Magnetic measurements show that Fe ₃ CoB ₂ exhibits a high magnetic anisotropy ( K ₁ = 1.2 MJ/m ³ ) and saturation magnetic polarization ( J _s = 1.39 T), which is suitable for RE-free permanent-magnet applications. Our ML-guided approach presents a promising paradigm for efficient materials design and discovery and can also be applied to the search for other functional materials.