Replacing hybrid density functional theory: motivation and recent advances

Area of Science:

• * Computational Chemistry
• * Materials Science
• * Quantum Physics

Background:

• * Density Functional Theory (DFT) is a cornerstone for electronic structure calculations in various scientific disciplines.
• * Hybrid DFT methods are widely used but suffer from inherent accuracy limitations, specifically over-delocalization and under-binding.
• * These limitations impact the reliability of simulations for chemical structures, mechanisms, and spectra.

Purpose of the Study:

• * To review the fundamental tradeoffs limiting hybrid DFT accuracy.
• * To introduce and discuss six modern approaches designed to surpass these limitations.
• * To provide practical guidance for DFT users in managing simulation accuracy.

Main Methods:

• * Review of existing literature on DFT approximations and their limitations.
• * Introduction of six advanced methods: DFT+U, self-interaction corrections, localized orbital scaling corrections, local hybrid functionals, real-space nondynamical correlation, and a novel rung-3.5 approach.
• * Analysis of how these methods address over-delocalization and under-binding issues.

Main Results:

• * Summary of the inherent accuracy limitations (tradeoffs) in standard hybrid DFT.
• * Detailed overview of six contemporary methods offering potential improvements.
• * Demonstration that these advanced methods aim to enhance accuracy without significantly increasing computational cost.
• * Identification of specific techniques to mitigate the impact of DFT tradeoffs.

Conclusions:

• * Hybrid DFT accuracy is constrained by delocalization and binding errors.
• * Several advanced methods show promise in overcoming these limitations.
• * Practical strategies can help researchers manage and mitigate these issues in their DFT simulations.