Evolutionary algorithms applied to minimize the thermal transport in atomic nanoribbons

In the pursuit of sustainable energy solutions, the development of materials with tailored transport properties is essential for enhancing device performance. In thermoelectric applications, materials with low thermal conductivity are particularly desirable to improve the dimensionless figure of merit and the efficiency of the Seebeck effect. In this study, we present an atomistic investigation of lattice thermal transport in a binary nanoribbon composed of alternating atomic blocks with a rock-salt structure, coupled to semi-infinite reservoirs at both ends. Using the Born interatomic potential within the harmonic approximation and the lattice thermal Kubo–Greenwood formula based on Green’s function techniques, we simulate phonon transport at the atomic scale. The accuracy of this approach depends critically on the selection of interatomic potential parameters. To address this, we apply evolutionary algorithms, specifically Differential Evolution, to optimize atomic masses and force constants in order to minimize the lattice thermal conductivity. These algorithms efficiently explore high-dimensional parameter spaces and identify optimal configurations corresponding to global or near-global minima. Our results, supported by ab initio calculations for real materials with a rock-salt structure, demonstrate an effective strategy for engineering low-thermal-conductivity materials, which are desirable for the design of high-performance thermoelectric devices. • Minimize thermal conductivity by adjusting interatomic potential parameters. • Evolutionary algorithms enable discovery of low thermal conductivity materials. • Ab initio calculations confirm our results for rock-salt structured materials.