Molecular adsorption at Pt(111). How accurate are DFT functionals?

Area of Science:

• Computational Chemistry
• Surface Science
• Materials Science

Background:

• Molecular chemisorption on metal surfaces is fundamental to catalysis, electrochemistry, and tribology.
• Density functional theory (DFT) is a primary tool for modeling these surface phenomena.
• Accurate prediction of adsorption energies is crucial for understanding and designing surface processes.

Purpose of the Study:

• To evaluate the accuracy of five exchange-correlation functionals in DFT for molecular chemisorption on Pt(111).
• To compare DFT predictions with experimental micro-calorimetric data for ten different molecules.
• To assess the influence of van der Waals interactions on chemisorption energy calculations.

Main Methods:

• Density functional theory (DFT) calculations on the Pt(111) surface.
• Comparison of five functionals: PBE, optPBE-vdW, optB86b-vdW, BEEF-vdW, and PBE-dDsC.
• Validation against experimental micro-calorimetric adsorption energy data for ten molecular systems.

Main Results:

• Functionals yield similar geometries and electronic structures but differ significantly in adsorption energies.
• optPBE-vdW and PBE-dDsC show the lowest mean absolute deviation (MAD) from experimental data (~0.2 eV).
• PBE and optB86b-vdW exhibit higher MAD (~0.45 eV), while BEEF-vdW is intermediate (~0.33 eV).
• π-bound unsaturated hydrocarbons are better described by optPBE-vdW and PBE-dDsC, while PBE and BEEF-vdW under-bind them.
• Vertically bound molecules (ethylidyne, CO) are over-bound by all tested functionals, with BEEF-vdW showing the best agreement.

Conclusions:

• The choice of exchange-correlation functional and the inclusion of van der Waals forces critically affect chemisorption energy accuracy.
• No single functional accurately predicts chemisorption energies across all tested molecular systems.
• Further development of DFT functionals is needed for universally accurate modeling of molecular chemisorption.