Abstract:We present an orbital-resolved extension of the Hubbard $U$ correction to density-functional theory (DFT). Compared to the conventional shell-averaged approach, the prediction of energetic, electronic and structural properties is strongly improved, particularly for compounds characterized by both localized and hybridized states in the Hubbard manifold. The numerical values of all Hubbard parameters are readily obtained from linear-response calculations. The relevance of this more refined approach is showcased by its application to bulk solids pyrite (FeS$_2$) and pyrolusite ($\beta$-MnO$_2$), as well as to six Fe(II) molecular complexes. Our findings indicate that a careful definition of Hubbard manifolds is indispensable for extending the applicability of DFT+$U$ beyond its current boundaries. The present orbital-resolved scheme aims to provide a computationally undemanding yet accurate tool for electronic structure calculations of charge-transfer insulators, transition-metal (TM) complexes and other compounds displaying significant orbital hybridization.

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve The paper aims to improve the predictive capability of Density Functional Theory (DFT) for energy, electronic, and structural properties by introducing orbital-resolved Hubbard U correction (orbital-resolved DFT+U). Specifically, the paper focuses on enhancing the accuracy and applicability of the DFT+U method when dealing with compounds that include localized and hybridized states. ### Main Issues and Challenges 1. **Self-Interaction Error (SIE)**: - Standard semi-local exchange-correlation functionals (such as LDA and GGA) suffer from severe self-interaction errors when dealing with strongly localized electrons, leading to inaccurate predictions of properties like energy and band gaps. - Conventional shell-averaged DFT+U methods can partially alleviate this issue but may still lack precision in certain cases. 2. **Selection of Hubbard U Parameter**: - In traditional DFT+U methods, the Hubbard U parameter is often treated as an adjustable parameter, determined by comparing with experimental results. This approach lacks universality and relies on experimental data. - Ab initio methods (such as constrained DFT, cDFT) can calculate the U parameter, but these methods are computationally expensive. 3. **Definition of the Hubbard Manifold**: - Choosing which electronic states require Hubbard correction is a critical issue. The traditional approach is to correct the entire d or f shell as a whole, which may not be optimal. - This shell-averaged approach can lead to inaccurate results, especially when dealing with compounds with significant orbital hybridization. ### Solutions 1. **Orbital-Resolved DFT+U Method**: - Introduce orbital-resolved Hubbard U correction, assigning different U parameters to each orbital instead of using the same U value for the entire shell. - This method can obtain all Hubbard parameters ab initio through linear response calculations (LR-cDFT), reducing the need for manual adjustments. 2. **Application Examples**: - The paper applies the orbital-resolved DFT+U method to bulk solids such as pyrite (FeS₂) and manganite (β-MnO₂), as well as six Fe(II) molecular complexes. - Results show that the orbital-resolved method significantly improves the prediction of structural and energy properties, aligning more closely with experimental observations. 3. **Computational Efficiency**: - The orbital-resolved DFT+U method also has advantages in computational efficiency, maintaining high accuracy while reducing computational costs. ### Conclusion By introducing the orbital-resolved DFT+U method, the paper addresses the self-interaction error and parameter selection issues present in traditional DFT+U methods when dealing with compounds with localized and hybridized states. This method enhances the accuracy and applicability of calculations, providing a low-cost yet accurate tool for electronic structure calculations, particularly suitable for charge-transfer insulators, transition metal complexes, and other compounds exhibiting significant orbital hybridization.

Orbital-resolved DFT+U for molecules and solids

Orbital-Resolved DFT +U for Molecules and Solids

Noncollinear DFT+$U$ and Hubbard parameters with fully-relativistic ultrasoft pseudopotentials

High-throughput determination of Hubbard and Hund values for transition metal oxides via the linear response formalism

Self-consistent on-site and inter-site Hubbard parameters within DFT+U+V for UO$_2$ using density-functional perturbation theory

Explicit demonstration of the equivalence between DFT+U and the Hartree-Fock limit of DFT+DMFT

Extensive Benchmarking of DFT+U Calculations for Predicting Band Gaps

DFT+U+J with linear response parameters predicts non-magnetic oxide band gaps with hybrid-functional accuracy

Structure and energetics of solvated ferrous and ferric ions: Car-Parrinello molecular dynamics in the DFT+U formalism

Hubbard $U$ through polaronic defect states

Communication: Comparing Ab Initio Methods of Obtaining Effective U Parameters for Closed-Shell Materials

DFT+ within the framework of linear combination of numerical atomic orbitals

A Self-consistent DFT+DMFT scheme in the Projector Augmented Wave : Applications to Cerium, Ce2O3 and Pu2O3 with the Hubbard I solver and comparison to DFT+U

Modeling the self-energy and wavefunction relaxation in the orbitals

Machine learning Hubbard parameters with equivariant neural networks

Machine learning the Hubbard U parameter in DFT+U using Bayesian optimization

Exploring Metastable States in UO$_2$ using Hybrid Functionals and Dynamical Mean Field Theory

Molecular orbital projectors in non-empirical jmDFT recover exact conditions in transition metal chemistry

Orbital-Free DFT Correctly Models Quantum Dots When Asymptotics, Nonlocality and Nonhomogeneity Are Accounted For

Accurate Electron-phonon Interactions from Advanced Density Functional Theory

Orbital breathing effects in the computation of x-ray d-ion spectra in solids by ab initio wave-function-based methods