First-principles insight into the degeneracy of ground-state LiBH4structures

Abstract

Recently, a number of ground-state structures of LiBH${}_{4}$ have been proposed, both from experimental and computational works. The results show controversy between computational and experimental ground-state crystal structures of LiBH${}_{4}$. In order to determine which is truly the lowest in energy, we study LiBH${}_{4}$ in a variety of crystal structures using density functional theory (DFT) calculations of the free energy ($T=0$ K total energy plus vibrational thermodynamics), employing a variety of DFT methods and exchange-correlation functionals. Our calculations show that the experimentally observed structures are lowest in energy in DFT. However, multiple LiBH${}_{4}$ structures are degenerate with the experimental ground-state crystal structure and there exists a relatively flat potential energy landscape between them. These degenerate structures include the recently theoretically predicted LiBH${}_{4}$ structure [Tekin, Caputo, and Z\"uettel, Phys. Rev. Lett. 104, 215501 (2010)], which the authors claimed to be 9.66 kJ/(mol LiBH${}_{4}$) (or $\ensuremath{\sim}$100 meV/fu) lower in energy than the experimentally XRD determined LiBH${}_{4}$ structure [Souli\'e, Renaudin, \ifmmode \check{C}\else \v{C}\fi{}ern\'y, and Yvon, J. Alloys Compd. 346, 200 (2002)]. Our calculations do not support these previous claims, and hence resolve this discrepancy between DFT and experiment.