Cubic boron arsenide (BAs) was predicted to have an exceptionally high thermal conductivity $\left(k\right)\sim 2000\mathrm{W}{\mathrm{m}}^{-1}{\mathrm{K}}^{-1}$ at room temperature, comparable to that of diamond, based on first-principles calculations. Subsequent experimental measurements, however, only obtained a $k$ of $\sim 200\mathrm{W}{\mathrm{m}}^{-1}{\mathrm{K}}^{-1}$. To gain insight into this discrepancy, we measured phonon dispersion of single-crystal BAs along high symmetry directions using inelastic x-ray scattering and compared these with first-principles calculations. Based on the measured phonon dispersion, we have validated the theoretical prediction of a large frequency gap between acoustic and optical modes and bunching of acoustic branches, which were considered the main reasons for the predicted ultrahigh $k$. This supports its potential to be a super thermal conductor if very-high-quality single-crystal samples can be synthesized.

• Received 16 August 2016

DOI:https://doi.org/10.1103/PhysRevB.94.220303