(561e) Predicting the Adhesion Energies and Sinter Resistance of Metal Films on Carbide and Nitride Supports

Adhesion energy is an important property of supported metal catalysts that dictates the nanoparticle shape through Winterbottom construction (Winterbottom, Acta Metallurgica (1967)), determines the chemical potential of metal atoms (Campbell and Mao, ACS Catal. (2017)) and is a descriptor for sintering (Hu and Li, Science (2021)). We present a general model to estimate the adhesion energies of epitaxial transition metal films supported on 2D/3D carbides and nitrides. These materials are an emerging class of catalysts supports because of their exquisitely tuneable surface chemistry and bulk composition. Their surfaces can be terminated with unary or binary combinations of p-block elements like C*/F*/O*/N*. Density functional theory calculations confirm that the differential adhesion energies become independent of thickness after progressively adsorbing up to three metal layers. We construct a Born-Haber cycle to calculate the adhesion energy from the cohesive energy, strain energy, and binding energy of metal atoms on V₂C MXene. The latter is estimated using a linear model containing features like electronegativity of terminations, work function, and strain, which qualitatively describe the electronic structure and geometry of supported metal systems. Adhesion energies of metal on supports containing binary terminations are obtained using a linear combination of adhesion energies on the corresponding unary terminations. Our model is generalizable across a wide range of surface terminations with an accuracy of 0.05 eV/Ų and exhibits easier transferability to supports like Ti₃N₂, Mo₃N₂ and WC. Adhesion energies and adsorption energies of single atoms are linearly correlated with an error of 0.25 eV. Hence, our model also estimates the stability of single atom catalysts on supports. Changing the surface termination can alter adhesion energies such that the heterostructure lies at apex of the sintering volcano plot (Hu and Li, Science (2021)). The ability to predict adhesion energy from the elemental properties facilitates efficient identification of sintering-resistant catalysts.