Atomistic studies of grain boundaries in disordered and ordered binary alloys

This research is an atomistic study of structural and chemical features of grain boundaries in disordered and ordered binary alloys. The motivation for these studies is that grain boundaries are important in determining material properties, in particular in alloys. Computer simulations have been performed by molecular statics and Monte Carlo methods. The interatomic interactions have been described in the framework of Finnis-Sinclair many-body potentials. Construction of these potentials for certain alloy systems and testing of their applicability is an integral part of this research. First, general features of segregation in a model A-B alloy have been studied. The main finding is a strong dependence of the segregation propensity on solute sizes and boundary structure and formation of ordered boundary structures at saturation. The atomic structure of the $\Sigma 3\{111\}$ grain boundary facet, formed due to segregation in a Cu-Bi alloy, has been studied using potentials fitted to parameters obtained from ab initio LMTO full potential calculations. An excellent agreement between the calculated structure and the HREM observations has been attained which demonstrates the reliability of constructed empirical potentials and desirability of combined HREM and atomistic studies of interfaces. Finally, grain boundaries in $\rm Ni\sb3Al$ and $\rm Cu\sb3Au$ ordered alloys have been studied by Monte Carlo technique. The results show that in $\rm Cu\sb3Au$ grain boundaries become chemically disordered at temperatures well below the order-disordered transition temperature, while in $\rm Ni\sb3Al$ grain boundaries remain ordered practically up to melting. Such observations support the proposition that the difference in ordering strength between these two materials could be the reason for their different grain boundary mechanical properties. In the cases of off-stoichiometric compositions in $\rm Ni\sb3Al,$ it has been found that Ni segregation leads to extensive compositional disordering at grain boundaries, while Al segregation is strongly selective among atomic sites in the boundary region and an ordered grain boundary structure is formed as a result of Al segregation. This difference may be the principal reason why the ductility improvement of $\rm Ni\sb3Al$ alloys by boron doping can only be achieved in Ni rich $\rm Ni\sb3Al.