Significant reduction of the lattice thermal conductivity in antifluorites via a split-anion approach

Searching for thermoelectrics with low lattice thermal conductivity ${\ensuremath{\kappa}}_{l}$ is crucial to address needs for waste heat recovery, but few light element dominated materials provide a low ${\ensuremath{\kappa}}_{l}$. Herein, we propose an effective strategy, i.e., a split-anion approach, to reduce the ${\ensuremath{\kappa}}_{l}$ of the antifluorite. After introducing the split-anion approach, the stable quasiantifluorite ${\mathrm{Li}}_{4}\mathrm{OSe}$ is obtained, and ${\ensuremath{\kappa}}_{l}$ is reduced by $\ensuremath{\sim}90%$ compared to the antifluorite ${\mathrm{Li}}_{2}\mathrm{Se}$. The ${\ensuremath{\kappa}}_{l}$ value of ${\mathrm{Li}}_{4}\mathrm{OSe}$ reaches as low as 0.49--1.75 W ${\mathrm{m}}^{\ensuremath{-}1}\phantom{\rule{4pt}{0ex}}{\mathrm{K}}^{\ensuremath{-}1}$, the range of thermoelectric applications, despite numerous light Li atoms and a large mass mismatch among the constituent atoms. This significant reduction of the thermal conductivity mainly results from the soft phonon branches and large optical bandwidth induced by lattice symmetry breaking after the split-anion approach. These findings offer ample scope for tuning the phonon band structure and thermal transport by reducing lattice symmetry like in the split-anion approach, leading to ultralow ${\ensuremath{\kappa}}_{l}$ of materials.