Stability of 41 metal–boron systems at 0 GPa and 30 GPa from first principles

• Largest ab initio metal boride database reveals existing discrepancies and new trends.
• Stability agreement for observed and calculated metal borides is over 82%.
• New stable phases are proposed at ambient and high pressures.
• Magnetic moment is shown to vary linearly with metal content in Fe, Co, and Ni borides.

A multitude of observed boron-based materials have outstanding superconducting, mechanical, and refractory properties. Yet, the structure, the composition, and the very existence of some reported metal boride (M–B) compounds have been a subject of extensive debate. This density functional theory work seeks to set a baseline for current understanding of known metal boride phases as well as to identify new synthesizable candidates. We have generated a database of over 12,000 binary M–B entries for pressures of 0 and 30 GPa producing the largest scan of compositions and systems in this materials׳ class. The 175 selected crystal structures include both observed prototypes and the new ones found with our evolutionary ground state search. The metals considered are Al, Ag, Au, Ba, Be, Ca, Cd, Co, Cr, Cs, Cu, Fe, Hf, Hg, Ir, K, La, Li, Mg, Mn, Mo, Na, Nb, Ni, Os, Pd, Pt, Rb, Re, Rh, Ru, Sc, Sr, Ta, Tc, Ti, V, W, Y, Zn, and Zr. Based on the formation energy calculated at zero pressure and temperature 4 new M–B phases or structures have been predicted, while a number of previously reported compounds have been shown to be unstable. At 30 GPa, changes in the convex hulls are expected to occur in 18 out of 41 M–B systems, which is used to indicate regions of the periodic table (for metal borides) that require further investigation from the community. Analysis of the collected information has revealed a nearly linear relationship between the magnetic moment per atom and the metal content for all the Fe–B, Co–B, and Ni–B structures within 0.15 eV/atom of the stability tie line. Both GGA-PBE and LDA-PW functionals were used to provide an understanding of the systematic error introduced by the choice of the exchange-correlation functional.