Kohn-Sham (KS) and post-KS density functional approximations (DFAs) often lack the chemical accuracy and may fail to capture qualitative trends for molecular interactions with ionic materials. This lack of reliability has hampered their use toward modeling surface chemistry and physics. Our studies on MgO(001) surface, show that the root cause for such discrepancies is density-driven errors rather than deficiencies in the exchange-correlation functional or errors from empirical dispersion corrections. For the post-KS random phase approximation (RPA) method, which incorporates dispersion interactions in a parameter-free manner, the density corrections based on the RPA singles (RPAS) method reduces the errors in RPA significantly, leading to remarkable agreement with high-level wavefunction methods while correctly predicting the order of small-molecule interactions with MgO(001) surface. Similar improvements are observed for the ${\mathrm{r}}^2\mathrm{SCAN}$-D4 functional, an empirically dispersion-corrected meta-GGA functional, by using the Hartree-Fock density. These consistent improvements across different classes of DFAs show that density-corrected methods can be used as a reliable and efficient approach to model molecule-surface interactions for ionic materials.

Close

PrevNext

View figure in article

• Adsorption
• Physisorption
• Reactions at surfaces & chemisorption
• Density functional calculations