Mixing different chemical species and decreasing dimensions to the nanoscale are two powerful approaches for improving materials. In both cases new properties emerge, and structure, composition, and chemical ordering can be tuned to tailor materials for specific purposes. To exploit the potential of these materials, it is crucial that they are fundamentally understood, and to this end, computational methods have emerged as an important complement to experiment. This thesis presents the development and application of methods for modeling alloys, nanoparticles, and nanoalloys on the atomic scale, with the purpose of guiding the search for new materials, in particular those related to plasmonic sensing of hydrogen.
A software for creating and sampling alloy cluster expansion has been developed partially in connection to this thesis, and is applied to hydrogenation of Pd and Pd–Au. For Pd–Au, the impact of chemical order on hydrogen uptake is studied, and two kinds of phase diagrams are calculated; one in which the Pd/Au atoms are fixed, and one in which they rearrange in response to hydrogen. These phase diagrams are constructed under the assumption that phase separation occurs with incoherent interfaces. This is not always the case, in particular not during hydrogenation of small Pd nanoparticles. Coherent interfaces lead to strain, and a methodology for studying this significantly more complex case is developed and applied to Pd–H, showing that there are three temperature intervals with qualitatively distinct hydrogenation behaviors.
Moreover, a software for creating Wulff constructions for the prediction of equilibrium nanoparticle shapes has been developed as part of this thesis and is used to study the impact of halides on the shapes of Au and Pd nanoparticles. Furthermore, an algorithm for finding equilibrium shapes of nanoparticles on the atomic scale is detailed, and the results indicate that an ensemble of nanoparticles in thermodynamic equilibrium in general should be expected to contain multiple different shapes. Moreover, nanoalloys of Ag–Cu and Pd–Au are studied on the atomic scale with the aim to understand how chemical ordering is impacted on the nanoscale, which reveals an interplay between chemistry and strain that can give rise to a rather complex distribution of the components throughout a nanoalloy. Finally, the dielectric functions of ten metallic alloys are calculated with first-principles methods and benchmarked with experiment, providing a library of reference data to aid modeling of nanoplasmonic systems. The latter results have also been made available in the form of a web application.