Increase in computation power has enabled usage of ever more accurate quantum mechanics based ab initio methods in thermodynamic modeling. As these methods do not need empirical parameters they can be used to supplement thermodynamic databases and to explain processes at atomistic level. They can also be used to predict properties of materials and thus guide experimental work to most promising areas. However, determination of statistical properties and screening of materials by the ab initio methods is computationally too heavy. This problem of limited computation power can be overcome by using coarse-grained models. In this Ph.D. thesis, the density functional theory and coarse-grained cluster expansion method was used to explain thermodynamics of the Cu-Ni-Pd and Ni-Rh alloys. Although these alloys have various applications, for example in catalysis, some their properties are still unknown. Coding of a new type of machine learning cluster expansion program was a significant part of this work. Finite temperature phase diagrams predicted by the coarse-grained model are in good agreement with experimental results. This shows that both the cluster expansion model and used density functional theory are sufficiently accurate to give realistic predictions. By studying enthalpies of formation at 0 K, it was observed that the Cu-Ni-Pd alloys do not form ordered ternary structures. Instead, the free-energy is minimized by mixing L12 and B2 type of CuPd compounds and fcc elements. In dilute Cu1-xNi0.5xPd0.5x, environments of Ni and Pd remains unchanged compared to corresponding binary alloy. In the Ni-Rh system, no L12 structure forms. Instead, Rh-rich clusters formed in the system. The magnetic enhancement effect in Ni-Rh alloys was shown to be due to Rh atoms inducing larger magnetic moment on surrounding Ni atoms at low X(Rh). Although Rh is non-magnetic, MoPt2 type of Ni2Rh1 compound was shown to be the most magnetic potential Ni-Rh structure. Enthalpies of formation modeled at finite temperatures are in excellent agreement with available experimental results. At concentration Cu57Ni1Pd42, the phase transition from ordered bcc structure to disordered fcc structure occurs at ca. 865 K. Critical point of the miscibility gap in the Ni-Rh system was predicted to locate at point T = 765 K, X(Rh) = 0.73.
|Publication status||Published - 2016|
|MoE publication type||G5 Doctoral dissertation (article)|
- density functional theory