Benchmarking the GW Approximation against Coupled-Cluster Theory for 3d Transition Metals

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Transition-metal systems pose significant computational challenges for accurate electronic structure calculations.
• Benchmarking quantum chemical methods is crucial for reliable predictions of ionization potentials (IP) and electron attachment (EA) energies.
• Open-shell 3d transition metals require specialized methods due to their complex electronic configurations.

Purpose of the Study:

• To assess the performance of the GW approximation and equation-of-motion coupled-cluster singles and doubles (EOM-CCSD) for IPs and EAs.
• To compare these methods against a high-level reference, ΔCCSD(T), for a benchmark set of transition-metal atoms and molecules.
• To evaluate the computational efficiency and accuracy of G0W0, evGW, and qpGW methods for these systems.

Main Methods:

• Calculated IPs and EAs using the G0W0 approximation with the PBE0 functional as a starting point.
• Employed equation-of-motion coupled-cluster singles and doubles (EOM-CCSD) theory.
• Utilized ΔCCSD(T) as the reference method for benchmarking.

Main Results:

• G0W0 demonstrated accuracy comparable to higher-level wave function methods, with mean absolute errors of 0.30–0.47 eV.
• EOM-CCSD showed slightly higher accuracy (0.19–0.33 eV MAE) but was computationally more expensive.
• Self-consistent GW calculations (evGW, qpGW) did not significantly improve accuracy over G0W0 and increased computational cost.

Conclusions:

• G0W0 provides a computationally efficient and reliable alternative for calculating IPs and EAs in open-shell 3d transition-metal systems.
• Both G0W0 and EOM-CCSD methods achieve mean absolute errors below 0.6 eV compared to experimental data.
• The G0W0 approximation is particularly compelling for large-scale studies of transition-metal materials due to its favorable cost-performance ratio.