Reaction barriers at metal surfaces computed using the random phase approximation: Can we beat DFT in the generalized...

Area of Science:

• Computational Chemistry
• Surface Science
• Heterogeneous Catalysis

Background:

• Density functional theory (DFT) methods, particularly generalized gradient approximation (GGA) functionals, often inaccurately predict reaction barriers for molecule dissociation on metal surfaces.
• A known issue is the collective underestimation of barriers by GGA-type functionals for specific dissociative chemisorption reactions, with inconsistencies in performance across different systems.
• The reasons behind this inconsistent behavior of GGA-DFT and whether other theoretical methods share these limitations are not well understood.

Purpose of the Study:

• To investigate the accuracy of advanced theoretical methods, specifically hybrid functionals and the random phase approximation in the adiabatic-connection fluctuation-dissipation theorem (ACFDT-RPA), in predicting reaction barriers for dissociative chemisorption.
• To compare the performance of these methods against GGA-DFT for two distinct systems: H2 on Al(110) (where GGA underestimates barriers) and H2 on Cu(111) (where GGA scatters around true barriers).
• To identify potential origins for the inconsistent performance of GGA-based functionals by analyzing electronic structure and comparing results across various theoretical approaches.

Main Methods:

• Calculation of barrier heights for dissociative chemisorption reactions using hybrid functionals with varying amounts of exact exchange.
• Application of the adiabatic-connection fluctuation-dissipation theorem in the random phase approximation (ACFDT-RPA) to determine reaction barriers.
• Comparative analysis of results from GGA, meta-GGA, GGA exchange + van der Waals correlation functionals, hybrid functionals, and ACFDT-RPA for H2 dissociation on Al(110) and Cu(111).

Main Results:

• Hybrid functionals demonstrate improved relative accuracy for the two systems compared to GGA-DFT but do not achieve chemical accuracy.
• ACFDT-RPA significantly improves barrier height predictions, yielding chemically accurate results for both H2 dissociation on Al(110) and Cu(111).
• Analysis of density of states and comparison with various DFT functionals provide insights into the origins of GGA's inconsistent performance.

Conclusions:

• ACFDT-RPA offers a reliable and accurate approach for calculating reaction barriers in dissociative chemisorption, overcoming limitations of GGA-DFT.
• Hybrid functionals offer an improvement over GGA but are insufficient for achieving chemical accuracy in these systems.
• Understanding the electronic structure and the specific interactions is crucial for explaining the varying performance of DFT functionals in catalysis.