Gaussian Basis Sets for Crystalline Solids: All-Purpose Basis Set Libraries vs System-Specific Optimizations.

Area of Science:

• Computational Quantum Chemistry
• Materials Science
• Solid-State Physics

Background:

• Traditional basis set selection in quantum chemistry relies on size and method suitability, not system-specific chemistry.
• Solids exhibit diverse bonding (metallic, ionic, covalent, dispersive) requiring tailored basis sets, unlike more homogeneous molecules.
• A practical, routine algorithm for system-specific basis set optimization is needed for solid-state calculations.

Purpose of the Study:

• To develop a novel, system-specific basis set optimization algorithm for solid materials.
• To introduce the BDIIS (Basis set Direct Inversion in Iterative Subspace) algorithm for routine use.
• To demonstrate the advantages of optimized basis sets for solid-state calculations.

Main Methods:

• Developed the BDIIS algorithm, inspired by direct inversion in the iterative subspace.
• Minimized total energy alongside the overlap matrix condition number, following VandeVondele et al.
• Applied and tested the method on prototypical solids: diamond, graphene, NaCl, and LiH.

Main Results:

• Successfully optimized valence orbitals using the BDIIS method.
• Demonstrated advantages of basis set optimization for solids, even with large basis sets (e.g., quadruple-ζ).
• Showcased applicability at both Density Functional Theory (DFT) and Hartree-Fock levels.

Conclusions:

• The BDIIS algorithm provides a practical approach for system-specific basis set optimization in solids.
• Optimized basis sets offer significant advantages for accurate solid-state electronic structure calculations.
• This method enhances the reliability of computational chemistry for diverse solid materials.