Density Functionals for Hydrogen Storage: Defining the H2Bind275 Test Set with Ab Initio Benchmarks and Assessment of...

Area of Science:

• Computational chemistry
• Materials science
• Hydrogen storage

Background:

• Efficient hydrogen storage is vital for a sustainable hydrogen economy.
• In silico methods, particularly Density Functional Theory (DFT), can accelerate the discovery of novel adsorbent materials.
• Accurate prediction of hydrogen adsorption enthalpies is key for material selection.

Purpose of the Study:

• To identify reliable Density Functional Theory (DFT) approximations for predicting hydrogen binding energies.
• To establish a comprehensive benchmark dataset (H2Bind275) for evaluating DFT methods.
• To guide high-throughput screening of materials for hydrogen storage applications.

Main Methods:

• Compilation of the H2Bind275 dataset, representing diverse hydrogen binding scenarios.
• Calculation of reference interaction energies using coupled-cluster theory.
• Assessment of 55 Density Functional Theory (DFT) approximations for H2 interaction energies.
• Evaluation of the def2-TZVPP basis set for accuracy and computational cost.

Main Results:

• Identification of five top-performing DFT functionals: ωB97X-V, ωB97M-V (hybrid), DSD-PBEPBE-D3(BJ), PBE0-DH (double hybrid), and B97M-V (semilocal).
• Recommendation to add empirical dispersion corrections to underbinding DFT functionals (e.g., revPBE, BLYP, B3LYP) for improved accuracy at low cost.
• Validation of the def2-TZVPP basis set for minimizing basis set errors (<1 kJ/mol).

Conclusions:

• Specific DFT functionals and basis sets are recommended for accurate and cost-effective screening of hydrogen storage materials.
• Computational chemistry, guided by robust benchmarks, significantly accelerates materials discovery for hydrogen energy.
• The findings provide a pathway for efficient in silico identification of high-capacity hydrogen adsorbents.