First-Principles Predictions of Ordered Structures in Fe-Ni Alloys

The observed ordered fcc Fe-Ni compounds are L1$_2$ FeNi$_3$ and (possibly metastable) L1$_0$ FeNi. Little is known about ordering tendencies in Fe-rich fcc alloys (stable at high $T$ at all compositions), despite the immense technological importance of such materials. We studied the $T=0$ and finite-$T$ ordering in fcc Fe-Ni alloys, employing the mixed-basis cluster expansion approach built on {\em ab initio} input formation enthalpies. We address explicitly the fact that Fe-Ni alloys may occur in either of three distinct ferromagnetic phases: high-spin (HS) fcc, low-spin (LS) fcc, and bcc phases. We distinguish between fcc- and bcc- based structures by analyzing the coordination of the nearest neighbors, and between HS and LS fcc phases by analyzing the total magnetic moment and performing series of fixed-moment calculations. We cluster expand the properties of HS ferromagnetic fcc alloys, providing a special attention to the unstable HS fcc Fe end- point. We further compare the stability of our studied HS fcc compound structures with alternative LS fcc and with bcc compound structures. For Fe-rich alloys, we find that due to instability of HS fcc Fe with respect to tetragonal deformations, the alloys have an exceptionally strong tendency to form (100) superlattices. For Ni-rich alloys, we predict new unsuspected ordered Fe-Ni compounds.