Electronic-structure-based material descriptors: (in)dependence on self-interaction and Hartree-Fock exchange.

Area of Science:

• Materials Science
• Computational Chemistry
• Solid State Physics

Background:

• Rational design of transition metal materials relies on electronic structure descriptors.
• Current methods like density functional theory (DFT) are prone to self-interaction and static correlation errors.
• These errors limit the accuracy of predicting material properties.

Purpose of the Study:

• To identify electronic structure descriptors for transition metals that are robust against computational errors.
• To evaluate the general validity of these descriptors across different computational methods.
• To provide reliable descriptors for the rational design of improved transition metal-based materials.

Main Methods:

• Analysis of d-band centers (original and width-corrected) for all 30 transition metals.
• Application of Hilbert transform highest peak descriptors.
• Comparison with results from hybrid functionals to assess static correlation treatment.

Main Results:

• Original and width-corrected d-band centers are unaffected by self-interaction errors.
• Hilbert transform highest peak descriptors also show robustness against self-interaction.
• Hybrid functionals demonstrate an unbalanced description due to poor static correlation treatment.

Conclusions:

• The proposed descriptors possess general validity, independent of specific computational methods.
• These robust descriptors can significantly improve the rational design of transition metal materials.
• This work offers a more reliable foundation for computational materials discovery.