Structure, bonding, and adhesion at the TiC(100)/Fe(110) interface from first principles

Abstract

Metal carbide ceramics offer potential as protective coatings for steels. Here we report a pseudopotential-based density functional (DFT) investigation of one such coating, wherein we predict the atomic structure, bonding, and the ideal work of adhesion ${\mathrm{(W}}_{\mathrm{ad}}^{\mathrm{ideal}})$ of the interface between a TiC(100) coating and a bcc Fe(110) substrate. Calibration of the DFT approximations used yields TiC and Fe bulk properties in reasonable agreement with experiment. Subsequent characterization of the low-index TiC and Fe surfaces reveals that all surfaces retain near bulk termination, in agreement with experiment. Stabilities of both TiC and Fe surfaces increase with their packing densities, i.e., $\text{(110)<(111)<(100)}$ for TiC and $\text{(111)<(100)<(110)}$ for bcc Fe. We estimate that the minimum critical stress required for crack propagation in bcc Fe is 27% larger than that in TiC. The TiC(100)/Fe(110) interface exhibits a lattice mismatch of $\text{∼2.1\%,}$ leading to a smooth interface with only a small structural relaxation, except for the ultrathin 1 monolayer (ML) coating. A mixture of metallic and covalent bonding dominates across the interface, due to significant C p-Fe d interaction and somewhat less pronounced Ti d-Fe d mixing; the latter is found to decrease with increasing coating thickness, but reaches a saturation value for 3-ML-thick coating. The asymptotic value of $W_{\mathrm{ad}}^{\mathrm{ideal}}$ for the TiC(100)/Fe(110) interface is predicted to be $\text{∼2.56\hspace{0.17em}}{\mathrm{J/m}}^2$ and is reached for a 3-ML-thick coating of TiC on Fe. This interface strength is considerably smaller than the energy required for cracking TiC or Fe, but may still be strong enough to survive as a coating for steel in extreme environments.