Accurate and Efficient Description of Acidic Zeolites with Plane-Wave Density Functional Theory Using Range-Separated...

Area of Science:

• Computational chemistry
• Materials science
• Catalysis

Background:

• Brønsted acidic zeolites are crucial catalysts, but their reactivity is challenging to model accurately.
• Generalized gradient approximation (GGA) density functional theory (DFT) often yields significant errors in reaction barriers and carbocation stability.
• High-accuracy ab initio methods are typically limited to smaller, nonperiodic cluster models.

Purpose of the Study:

• To evaluate the performance of various density functionals and the random phase approximation (RPA) against high-level ab initio benchmarks.
• To identify accurate and computationally efficient methods for studying zeolite reactivity.
• To enable reliable prediction of reaction energies and barriers in periodic zeolite systems.

Main Methods:

• Comparison of density functionals (including range-separated hybrids) and RPA with DLPNO-CCSD(T) and CBS-extrapolated MP2 calculations.
• Utilized large cluster models for reference calculations.
• Applied selected accurate functionals within periodic plane-wave DFT.

Main Results:

• Range-separated hybrid functionals, specifically ωB97M-D4, ωB97M-V, ωB97X-D4, ωB97X-V, and ωB97-D, demonstrated high accuracy.
• These functionals achieved mean absolute errors (MAE) below 8 kJ/mol compared to reference calculations.
• ωB97M-D4 exhibited the best performance with an MAE of 5.1 kJ/mol, outperforming even CBS-extrapolated MP2.
• Range-separated functionals were found to perform well in periodic DFT calculations.

Conclusions:

• Range-separated hybrid functionals provide accurate reaction energies and barriers for Brønsted acidic zeolites.
• These methods enable efficient and accurate direct calculations on periodic zeolite models.
• This advancement facilitates more reliable computational studies of zeolite-catalyzed reactions.