Merlin a versatile optimization environment applied to the design of metallic alloys and intermetallic compounds

An important step in the design of alloys and intermetallic compounds using semi-empirical potentials is to determine the appropriate parameters, which best describe experimental and/or quantum mechanical ab initio results. This task is quite difficult as the data are not always consistent and complete and furthermore, they contain errors. To facilitate the modelling we use the optimization environment of Merlin http://merlin.cs.uoi.gr/. This was applied to study a particular class of intermetallic compounds and alloys, which are very interesting, the so-called super-alloys, such as Ni–Al. We have fitted the properties of such intermetallic alloys and compounds utilizing a semi-empirical tight-binding potential in the second moment approximation. The potentials, which were produced in this way, were tested for properties at various temperatures, including segregation to surfaces and interfaces, and also for dynamical properties like the phonon density of states and mean-square displacements. We find a very good agreement to known experimental results and also a wealth of interesting information has revealed. Therefore the produced interatomic potentials present a realistic way to test scenarios which appear in the materials design.