Composition and formation enthalpy of random alloy?

Alloy Theoretic Automated Toolkit (ATAT)
1

Hello Dr Axel and atat users,

I’m a newbie with alloy and ATAT. I just spent a couple of days to understand how to use ATAT as well as theory of cluster expansion and sqs method because my current project concerns these topics. Our concern is a new binary alloy. From XRD, we knew that is random alloy with fcc lattice, however, composition is not actually clear. As far I know, MAPS can make a convex hull for ordered alloys with determined type of crystal, but is there anyway to have a convex hull including random alloys?

For formation enthalpy of radom alloy, it looks like one of the best approach is using SQS method. For some cases of well-known alloy, the differences between theory and experiment are about 1kJ/mol (~10meV/atom). However, for new random alloy, is it safe to use this method? or there is other method implemented in ATAT can improve final results.

Thank you very much in advance.

2

When looking at a disordered solution, you should not look at the convex hull of the energies. Since disordered solution occur at high temperature, you have to include a configurational entropy term at least: the simplest way to do this is to add sum_i x_i ln(x_i) (where x_i are compositions) to your SQS energies.

Even better would be to fit a polynomial through your sqs energies. The sqs2tdb automates a lot of this and produce a working .tdb file for use with thermocalc, for instance.