Electronic analysis of sigma phase destabilization in a family of compositionally complex ferritic stainless steel substitutes.

Stainless steels are used extensively in industry due to a combination of desirable material properties, such as corrosion resistance and strength. However, ferritic stainless steels form a brittle sigma phase at moderately high processing temperatures, which limits their utility. Recent experimental results from our group found that small amounts of Al in the presence of Mn suppress the formation of the Fe-Cr sigma phase, leaving the desired ductile, body centered cubic phase. This first principles work uses the Crystal Orbital Hamilton Population method to explore the hypothesis that Al destabilizes sigma geometry by changing the electron distribution among the crystal's molecular orbitals. In order to create representative structures for this analysis, we combine a generalized cluster expansion with Monte Carlo simulations to determine the preferential placement of atomic species on each of the five symmetrically distinct lattice sites of the sigma crystal structure. Because the sigma phase is too complex for its energy to be cluster expanded using conventional automated fitting procedures, we implement an iterative method to efficiently select fitting structures that span the composition space of the alloy system. Beyond creating a model to assess the mechanisms for sigma phase destabilization in Fe-Cr-Mn-Al structures, this work pushes the limits on the size and complexity of metallic systems that have been modeled using cluster expansions.