A. T. Paxton, C. Elsässer

Physical Review B, 87 (2013), 224110

Recent density functional theory (DFT) calculations by Först et al. [Phys. Rev. Lett. 96, 175501 (2006)] have predicted that vacancies in both low and high carbon steels have a carbon dimer bound to them. This is likely to change the thinking of metallurgists in the kinetics of the development of microstructures. While the notion of a C2 molecule bound to a vacancy in Fe will potentially assume a central importance in the atomistic modeling of steels, neither a recent tight binding (TB) model nor existing classical interatomic potentials can account for it. Here we present a TB model for C in Fe, based on our earlier work for H in Fe, which correctly predicts the structure and energetics of the C2 dimer at a vacancy in Fe. Moreover the model is capable of dealing with both concentrated and dilute limits of carbon in both ?-Fe and ?-Fe as comparisons with DFT show. We use both DFT and TB to make a detailed analysis of the dimer and to come to an understanding as to what governs the choice of its curious orientation within the vacancy.