First Principles Calculation of Defect Structure in Non-stoichiometric CoAl and CoTi

Abstract

First-principles electronic structure calculations have been performed for defect structure in non-stoichiometric CoAl and CoTi. In order to determine the type of constitutional defects, the compositional dependence curves both of formation energies and of lattice parameters are obtained by the calculations employing supercells in various sizes. The defect formation energies are calculated with taking into account the compositional dependence of the chemical potential. The calculated results suggest that the Co vacancy is the dominant thermal-excitation defect even in the Co-rich side near the stoiciometry in CoTi. Intermetallic compounds with the B2 structure have attracted considerable attention as promising candidates for high-temperature materials. One of the important features of intermetallic compounds with the B2 structure is having a wider range of compositions than other phases. In order to compensate the deviation from the stoichiometry, constitu- tional defects are introduced. Several properties such as mechanical and magnetic properties, therefore, show com- positional dependence originating from the constitutional defects. In addition, thermally excited defects play an important role in the diffusion of atoms and therefore the understanding of the defects are required for the alloy design and optimization of heat treatment. Electronic structure calculations using supercell models have been performed for non-stoichiometric B2 intermetallic compounds to understand the defect structure. The size of the supercell is one of the important factors for electronic structure calculations including defects. The defect formation energies in FeAl were calculated using the 8, 16 and 32 atoms supercells. 1) The calculations with the 16 and 54 atoms supercells were performed for convergence tests of the effective formation energies and the structural relaxation of the constitutional defects in NiAl. 2) In our previous work, we