Bond order potentials for bcc transition metals and molybdenum silicides

One of the most challenging areas in materials science is modeling of the mechanical behavior of materials. Macroscopic properties, such as plastic flow and brittleness, are determined by processes occurring at the atomic level and are mainly influenced by the properties of extended defects. Understanding the energetics and structural properties of dislocations, grain boundaries and other extended defects enables to gain a valuable insight into the deformation mechanisms and allows to design specific ways to overcome intrinsic limitations of materials. In this thesis, we study the properties of extended defects in four bcc transition metals—Nb, Ta, Mo and W, using the bond order potentials (BOP). These potentials are based on the real-space parameterized tight-binding method and are eminently suitable for modeling of extended defects in materials with predominantly metallic and covalent type of bonding. Our results confirm that the unusual plastic behavior of bcc transition metals is governed by the properties of the a/2[111] screw dislocations and reveal quantitatively the overall invalidity of the well-known Schmid law in these materials. The most recent improvement of BOP has been a derivation of analytic environmental dependence of bond integrals. This development extends the transferability of the BOP model and its applicability to multi-component systems. The accuracy of this, so-called, screened bond order potential (SBOP) formalism was studied for the first time for the case of pure Mo as well as the intermetallic compound MOSi₂. Molybdenum silicides are prospective high-temperature structural materials and full understanding of the deformation behavior of these compounds is a necessary prerequisite for their extensive use in industrial applications.