A new generation of effective core potentials: Selected lanthanides and heavy elements.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Effective core potentials (ECPs) are crucial for simplifying electronic structure calculations of heavy atoms.
• Correlation-consistent ECPs (ccECPs) offer improved accuracy by systematically including electron correlation.
• Heavy and f-elements present unique challenges in ECP construction due to large cores and complex electronic structures.

Purpose of the Study:

• To construct novel correlation-consistent effective core potentials (ccECPs) for key heavy and f-elements.
• To ensure these ccECPs balance accuracy, valence space size, and transferability.
• To develop ccECPs suitable for use in plane wave computational codes.

Main Methods:

• Construction of spin-orbit (SO) averaged relativistic effective potentials (AREP) within a relativistic coupled-cluster framework.
• Optimization using objective function one-particle characteristics for enhanced convergence.
• Adjustment of transferability via binding curves of hydride and oxide molecules.
• Specialized core-valence partitioning for f-elements, including the 4f subshell in the valence space.

Main Results:

• Development of accurate and transferable ccECPs for Y, Zr, Nb, Rh, Ta, Re, Pt, Gd, and Tb.
• Successful handling of challenges associated with f-elements, such as large cores and near-degeneracies.
• ccECPs demonstrate a good balance between accuracy and computational efficiency.
• The developed potentials are compatible with plane wave codes at reasonable energy cutoffs.

Conclusions:

• The newly developed ccECPs provide reliable and efficient tools for computational studies of heavy and f-elements.
• These potentials facilitate accurate investigations in materials science and chemistry.
• The approach offers a robust method for constructing ECPs for challenging elements.