DFT+U+V is equivalent to DFT+U with density-dependent hybridized projectors

Alberto Carta; Edward Linscott; Nicola Marzari

REVIEW 3 major objections 2 minor 69 references

DFT+U+V is equivalent to DFT+U with density-dependent hybridized projectors

T0 review · 3 major / 2 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read To first order in V/U, the inter-site +V correction of DFT+U+V is exactly an on-site DFT+U correction evaluated on a density-dependent set of Hubbard projectors hybridized with the orbitals of neighboring sites.

desk verdict Good physical idea and mostly clean proof, but the central algebra has a conjugation error for complex density matrices that needs fixing before the equivalence claim is safe. read the letter →

arxiv 2607.18071 v1 pith:PH3I2SHG submitted 2026-07-20 cond-mat.mtrl-sci cond-mat.str-elphysics.chem-phphysics.comp-ph

Edward Linscott , Alberto Carta , Nicola Marzari This is my paper · ORCID

classification cond-mat.mtrl-scicond-mat.str-elphysics.chem-phphysics.comp-ph

keywords DFT+U+Vinter-siteHubbardcorrectionsprojectorsself-interactionerrorcovalencystronglycorrelatedmaterialsortho-atomicscheme

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that the inter-site Hubbard correction +V, used in the DFT+U+V method, does not introduce a new kind of interaction: to first order in the ratio V/U, a DFT+U+V energy is exactly the standard DFT+U energy evaluated on a rotated set of Hubbard projectors that are hybridized with orbitals of neighboring sites. The rotation is generated by an explicit anti-Hermitian operator built from the inter-site density matrix and the Hubbard parameters, and the projectors are necessarily density-dependent because the generator is a functional of the density. The reason a reader should care is that this explains why +V works: it moves the on-site occupation-curvature penalty (the self-interaction correction) onto orbitals that already contain the covalent character of the bond, counteracting the over-localization and broken bonds that plain +U causes. For practitioners it also reshapes the theory: the choice of projectors, Hubbard parameters, and whether to include +V are not independent decisions.

What carries the argument

The load-bearing object is the anti-Hermitian generator S^ (Eq. 5) that rotates the 'block-diagonalizing' projectors |φ^{Iσ}_i⟩—the natural orbitals that diagonalize each site's occupation matrix—into hybridized projectors e^{S^}|φ^{Iσ}_i⟩ that mix in the orbitals of V-coupled neighbors. The proof reduces the equivalence to the commutator identity [S^, V^U] = (1/2)V^V, which is solvable precisely because both the inter-site energy and potential are quadratic in the density, so 2E_V = Tr(ρ V^V). This single identity forces the form of S^, makes the projectors density-dependent, and yields the energy and potential equalities order by order.

What would settle it

In a plane-wave PAW code (which uses non-orthogonal projectors), construct S^ from Eq. (5) and compare the DFT+U+V energy and potential against DFT+U with the rotated projectors over a range of V/U. If the difference scales as V/U rather than (V/U)^2, the ortho-atomic assumption is load-bearing. Alternatively, in an ortho-atomic code, drive an orbital occupation toward 1/2 and check whether the residual ceases to be O(V^2/U).

Watch

Extended reading notes

Core claim

Central result: to first order in V/U, E^U[S^] = E^U + E^V + O(V^2/U), and the same for the potential. E^U[S^] is the DFT+U energy with projectors e^{S^}|φ^{Iσ}_i⟩, where S^ is the anti-Hermitian generator whose matrix elements mix site I and J through V^{IJ}, the inter-site occupancy Λ^{JIσ}_{ji}, and the denominator U^I(1-2λ^{Iσ}_i)-U^J(1-2λ^{Jσ}_j). Because S^ is built from the density matrix, the projectors are density-dependent. The corollary is that +V is the same self-interaction penalty as +U, but measured on projectors that carry the bond's covalent character—which is why +V restores covalency. With frozen projectors, the equality holds for the energy with e^{S^} and for the potenti

Load-bearing premise

The proof assumes the Hubbard projectors are orthonormal across sites (the ortho-atomic scheme); with non-orthogonal projectors, inter-site overlap matrices enter the operator algebra and the simple generator form cannot be assumed to hold.

Editorial extensions

If this is right

Any DFT+U+V calculation can be replaced, to first order in V/U, by a DFT+U calculation on hybridized density-dependent projectors, bringing the full DFT+U implementation ecosystem (response properties, extensions) to bear on +V problems.
The +V correction targets the same local occupation-curvature self-interaction as +U; it should be parametrized by linearizing the energy in the occupation of the hybridized projectors, a condition the usual off-diagonal response-matrix recipe does not satisfy.
With frozen projectors, no single rotation reproduces the DFT+U+V potential: the spectrum requires the e^{2S^} rotation while the energy requires e^{S^}, so frozen-projector DFT+U approximates DFT+U+V only up to the residual W^ term.
Applying +U to the hybridized projectors enhances covalency for less-than-half-filled antibonding orbitals (the minority-spin e_g states in NiO), directly counteracting the over-localization caused by +U on bare projectors.
Projector choice, U values, and the decision to include +V are coupled degrees of freedom, not independent choices; V is, at heart, a choice of projectors.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the all-orders diagonalization of the corrective potential remains of +U type (the paper shows the first-order term is, while the second-order term is not), then extended Hubbard functionals beyond Dudarev-type +V might also reduce to some projector redefinition, unifying the extended-Hubbard family.
The divergence of the generator as an occupation approaches 1/2 delimits the practical range of the equivalence; comparing energies in systems with a nearly half-filled strongly hybridized orbital (e.g., mixed-valence oxides) would reveal where the first-order statement breaks.
The density-dependent projectors suggest a self-consistent scheme in which the rotation is updated during the SCF cycle; one testable prediction is that such a scheme exactly matches DFT+U+V energies and potentials while sidestepping the frozen-projector mismatch.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Here's the short version: the paper's interpretation of DFT+U+V as DFT+U with density-dependent hybridized projectors is probably right, and the explicit generator is a nice piece of work. But the central derivation has a conjugation error for complex off-diagonal density-matrix elements. It's fixable, but it has to be fixed before the equivalence claim is safe.

What's genuinely new: the explicit first-order generator S in Eq. (5), the proof that DFT+U+V reproduces DFT+U on rotated projectors to order V/U, and the argument that the +V correction is really an on-site curvature penalty applied to projectors that already contain covalency. The frozen-projector obstruction in S7/S8 is also new and useful, and the paper is honest about the limits: first-order validity, failure near lambda=1/2, ortho-atomic projectors only. The NiO illustration is nice and the data are archived.

The soft spot is not the disclosed limitations — it's an actual error. Eq. (11) defines the inter-site potential with Lambda^{JI}_{ji}, but Eq. (18), the functional derivative of E^V, gives Lambda^{IJ}_{ij}. For real density matrices these are the same; for complex ones they differ by conjugation. This propagates into the generator Eq. (5), and the identity Eq. (19) that connects Tr[rho V_V] to 2E^V only holds when Lambda is real. So the first-order equivalence as written fails for a complex inter-site density matrix. The stress-test note shows the fix is straightforward — conjugate the numerator in S — and once that's done the proof goes through. But the current text is not correct in the general case, and the numerical test at Gamma in NiO probably doesn't expose it because those matrix elements are real or effectively so.

That's a load-bearing flaw in the formal proof, though not in the physical message. If the authors repair the conjugation, the paper would be solid. I'd send it to review — the topic matters, the idea is important, and the rest of the algebra is careful — but the reviewers should insist on a corrected derivation and a check that the potential equivalence holds for complex Lambda. For my own work, I wouldn't cite this version; I'd wait for the revision.

Referee Report

3 major / 2 minor

Summary. The paper claims that, to first order in V/U, the inter-site Hubbard correction DFT+U+V is exactly equivalent to a standard on-site DFT+U functional evaluated on a density-dependent, redefined set of Hubbard projectors hybridized with neighboring-site orbitals. The central result is an explicit rotation-like generator (Eq. (5)) and an energy/potential identity (Eqs. (22)-(23)). The authors also analyze the frozen-projector case, show that only a partial equivalence holds there, and illustrate the formalism on bulk NiO.

Significance. If the equivalence is correct, it gives the first formal interpretation of the widely used +V correction: +V does not introduce a new interaction beyond the +U occupation-curvature penalty, but rather changes the subspace on which that penalty is imposed, thereby making the on-site correction aware of covalency. This has direct implications for how V is computed and for the broader question of what the Hubbard projectors represent. The paper is clearly written, includes open data and reproducible scripts, and provides a numerical demonstration. However, the central algebraic derivation contains a sign/conjugation error that must be corrected before the claim is reliable.

major comments (3)

[Appendix, Eq. (11)] The definition of the inter-site potential is incorrect. From Eq. (18), δE^V/δρ = -∑ V^{IJ} P^I ρ P^J, so the matrix element in the φ basis is -V^{IJ} Λ^{IJ}_{ij} = -V^{IJ} ⟨φ_i^I|ρ|φ_j^J⟩. Eq. (11) instead writes -V^{IJ} Λ^{JI}_{ji}, which is the complex conjugate. This is not a harmless reordering: for a complex density matrix the two operators differ. This error propagates into the generator: solving Eq. (20) with the correct V_V gives S_{ij}^{IJ} = V^{IJ} Λ^{IJ}_{ij}/D for the (I,J) block, not V^{IJ} Λ^{JI}_{ji}/D as in Eq. (5)/(21).
[Appendix, Eqs. (5), (20)] The generator as printed is not anti-Hermitian. For real Λ, Eq. (5) gives S_{ij}^{IJ}=S_{ji}^{JI}, i.e. a symmetric matrix, whereas the derivation assumes an anti-Hermitian T so that e^T preserves orthonormality. With this symmetric S, [S,V_U] is antisymmetric (e.g. in a two-site/one-orbital model, [S,V_U]_{12}=-aD/2, [S,V_U]_{21}=+aD/2), so Eq. (20) cannot be satisfied with the symmetric V_V of Eq. (11). Consequently the first-order energy change Tr[V_U[ρ,S]] vanishes for real Λ, and Eq. (22) does not hold as stated. The proof can be repaired by choosing an antisymmetric S, e.g. S_{ij}^{IJ}=V^{IJ}Λ^{IJ}_{ij}/D for an ordered pair and the negative for the swapped pair, but this is a substantive correction, not a typo.
[§4, Fig. 2 and SI S4] The numerical validation uses NiO at the Γ point, where the inter-site density-matrix elements are real; it therefore cannot distinguish Λ^{IJ} from Λ^{JI}. Given that the paper's claim is general, the authors should either state explicitly that the equivalence holds only for real density matrices in the current derivation, or provide a test with complex inter-site elements (e.g. spin-orbit coupling or a non-Γ k-point) after correcting the generator. The conjugation issue is otherwise hidden by the test system.

minor comments (2)

[Abstract and §2] The abstract says 'to first order in V/U'; the expansion in fact also involves the occupations through the denominator D, and the authors themselves note the failure as λ_i→1/2. This caveat appears only later in the text; a sentence in the abstract or introduction would help set expectations.
[Notation, Eq. (5)] The symbols Λ^{JI}_{ji} are easy to misread. After correcting the generator, define the ordered-pair convention explicitly (e.g. for I<J and I>J separately) so that the reader can verify anti-Hermiticity.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the generator is solved from a commutator condition and the proof is self-contained.

full rationale

The central claim is established by an explicit construction, not by fitting or by importing the result from a prior work. The generator S is obtained by solving the commutator condition [S,V_U]=1/2 V_V (Appendix, Eq. 20), which is itself derived from the requirement that the first-order energy change match E^V. This is a representation theorem: the existence of such an S is nontrivial, and the paper shows that no frozen-projector rotation can reproduce the full potential (SI S8), so the equivalence has real content and is not vacuous. No parameters are fitted to data and then renamed as predictions; U and V enter only as given interaction strengths. The self-citations that appear are confined to motivation, background, and illustrative linear-response practice, and none of them carries the load of the derivation. The stated ortho-atomic orthonormality assumption is an explicit scope condition, not a circular appeal. Possible algebraic issues (e.g., the Hermitian conjugation in Eq. 11 for complex density-matrix elements) would be correctness concerns, not circularity. For these reasons the derivation is self-contained and no circular step is identified.

Assumptions & free parameters 0 free parameters · 5 assumptions · 1 invented entities

The derivation introduces no fitted free parameters: U and V are external inputs to the parent DFT+U+V functional, and the numerical values in the NiO test are taken from linear response, not fitted to the equivalence. The proof depends on standard linear algebra plus the domain assumptions listed above, most notably orthonormal projectors and the validity of the V/U expansion away from half-filling.

assumptions (5)

domain assumption Hubbard projectors are orthonormal across sites (ortho-atomic scheme, ⟨φ_i^{Iσ}|φ_j^{Jσ}⟩=δ_IJ δ_ij).
The Appendix states this explicitly and notes non-orthogonal projectors would introduce inter-site overlap matrices, changing the operator algebra.
domain assumption The +V functional has the Dudarev/FLL quadratic form, Eq. (3), with V_II=0.
The proof targets this specific functional form; other double-counting schemes or spin-dependent V would alter the algebra.
domain assumption First-order perturbation theory in V/U is valid, requiring the denominators U_I(1−2λ_i)−U_J(1−2λ_j) to stay away from zero.
The paper warns the expansion fails as λ→1/2 even for V≪U; the equivalence is therefore not uniform over occupations.
standard math The density operator can be represented in a basis where on-site occupation matrices are diagonalizable with eigenvalues λ_i^{Iσ}.
Used to block-diagonalize the on-site potential; standard spectral decomposition, no new physical assumption.
standard math Rotations e^T with anti-Hermitian T preserve orthonormality, and standard commutator/BCH expansions are used.
Required to express E_U[T] to first order in T and to derive the generator S.

invented entities (1)

Density-dependent hybridized Hubbard projectors e^S|φ_i^{Iσ}⟩ (and doubly hybridized e^{2S})
purpose: A rotated set of projectors that makes the DFT+U functional reproduce the DFT+U+V energy/potential to first order in V/U; used to explain how +V restores covalency.
They are a mathematical reparametrization defined through Eq. (5); they are constrained by equivalence to the parent functional and illustrated on NiO, but no independent experimental handle is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of DFT+U+V is equivalent to DFT+U with density-dependent hybridized projectors." pith.science (2026). https://pith.science/paper/PH3I2SHG

@misc{pith2026260718071,
  author       = {Pith},
  title        = {Pith review of: DFT+U+V is equivalent to DFT+U with density-dependent hybridized projectors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PH3I2SHG}},
  note         = {Machine review of arXiv:2607.18071}
}

abstract

Hubbard-corrected density-functional theory (DFT+$U$) is a popular tool for first-principles modeling of materials with localized $d$ or $f$ electrons, but its on-site corrections tend to over-localize charge and break covalent bonds. Inter-site $+V$ corrections were introduced to counter this and are now widely used, but a formal justification has been lacking. Here we show that -- to first order in $V/U$ -- inter-site corrections are exactly equivalent to on-site DFT+$U$ evaluated on a density-dependent redefinition of the Hubbard projectors, hybridized with those of neighboring sites, providing insight into the explicit mechanism by which $V$ affects covalency. If the projectors are held frozen, as is common practice, the equivalence partially breaks down. Beyond reinterpreting the formalism, these results sharpen the questions of how $V$ should be computed and what the Hubbard subspaces fundamentally are.

Figures

Figures reproduced from arXiv: 2607.18071 by the authors.

**Figure 2.** FIG. 2. The frozen-projector approximation reproduces the exact DFT+ [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 2 canonical work pages

[1]

Hohenberg and W

P. Hohenberg and W. Kohn, Inhomogeneous electron gas, Phys. Rev.136, B864 (1964)

1964
[2]

Kohn and L

W. Kohn and L. J. Sham, Self-consistent equations in- cluding exchange and correlation effects, Phys. Rev.140, A1133 (1965)

1965
[3]

R. O. Jones, Density functional theory: Its origins, rise to prominence, and future, Rev. Mod. Phys.87, 897 (2015)

2015
[4]

Marzari, A

N. Marzari, A. Ferretti, and C. Wolverton, Electronic- structure methods for materials design, Nat. Mater.20, 736 (2021)

2021
[5]

Van Noorden, These are the most-cited research pa- pers of all time, Nature640, 591 (2025)

R. Van Noorden, These are the most-cited research pa- pers of all time, Nature640, 591 (2025)

2025
[6]

Terakura, T

K. Terakura, T. Oguchi, A. R. Williams, and J. K¨ ubler, Band theory of insulating transition-metal monoxides: Band-structure calculations, Phys. Rev. B30, 4734 (1984)

1984
[7]

V. I. Anisimov, J. Zaanen, and O. K. Andersen, Band theory and Mott insulators: HubbardUinstead of Stoner I, Phys. Rev. B44, 943 (1991)

1991
[8]

Hubbard, Electron correlations in narrow energy bands, Proc

J. Hubbard, Electron correlations in narrow energy bands, Proc. R. Soc. A276, 238 (1963)

1963

Show all 69 references

[9]

A. I. Liechtenstein, V. I. Anisimov, and J. Zaanen, Density-functional theory and strong interactions: Or- bital ordering in Mott-Hubbard insulators, Phys. Rev. B 52, R5467 (1995)

1995
[10]

S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys, and A. P. Sutton, Electron-energy-loss spec- tra and the structural stability of nickel oxide: An LSDA+Ustudy, Phys. Rev. B57, 1505 (1998)

1998
[11]

M. T. Czy˙ zyk and G. A. Sawatzky, Local-density func- tional and on-site correlations: The electronic structure of La2CuO4 and LaCuO3, Phys. Rev. B49, 14211 (1994)

1994
[12]

V. I. Anisimov, I. V. Solovyev, M. A. Korotin, M. T. Czy˙ zyk, and G. A. Sawatzky, Density-functional theory and NiO photoemission spectra, Phys. Rev. B48, 16929 (1993)

1993
[13]

V. I. Anisimov, F. Aryasetiawan, and A. I. Lichtenstein, First-principles calculations of the electronic structure and spectra of strongly correlated systems: The LDA+U method, J. Phys. Condens. Matter9, 767 (1997)

1997
[14]

W. E. Pickett, S. C. Erwin, and E. C. Ethridge, Refor- mulation of the LDA+Umethod for a local-orbital basis, Phys. Rev. B58, 1201 (1998)

1998
[15]

They are also usually spin-independent, but we have as- signed them a spin indexσfor the sake of generality
[16]

D. A. Scherlis, M. Cococcioni, P. Sit, and N. Marzari, Simulation of heme using DFT+U: A step toward accu- rate spin-state energetics, J. Phys. Chem. B111, 7384 (2007)

2007
[17]

H. Hsu, P. Blaha, M. Cococcioni, and R. M. Wentzcov- itch, Spin-state crossover and hyperfine interactions of ferric iron in MgSiO 3 perovskite, Phys. Rev. Lett.106, 118501 (2011)

2011
[18]

Bajdich, M

M. Bajdich, M. Garc ´ ıa-Mota, A. Vojvodic, J. K. Nørskov, and A. T. Bell, Theoretical investigation of the activity of cobalt oxides for the electrochemical oxidation of water, J. Am. Chem. Soc.135, 13521 (2013)

2013
[19]

Cococcioni and S

M. Cococcioni and S. de Gironcoli, Linear response ap- proach to the calculation of the effective interaction pa- rameters in the LDA+Umethod, Phys. Rev. B71, 7 035105 (2005)

2005
[20]

H. J. Kulik, M. Cococcioni, D. A. Scherlis, and N. Marzari, Density functional theory in transition-metal chemistry: A self-consistent HubbardUapproach, Phys. Rev. Lett.97, 103001 (2006)

2006
[21]

The Supplemental Material includes Refs

See Supplemental Material, appended to this preprint, for a discussion of global versus local curvature; the linear-response calculation of the Hubbard parameters; computational details; the form of the rotation gener- ator in periodic systems; a decomposition of the first- or...
[22]

H. J. Kulik and N. Marzari, Transition-metal dioxides: A case for the intersite term in Hubbard-model functionals, J. Chem. Phys.134, 094103 (2011)

2011
[23]

H. J. Kulik and N. Marzari, Accurate potential energy surfaces with a DFT+U(R) approach, J. Chem. Phys. 135, 194105 (2011)

2011
[25]

G. C. Moore, M. K. Horton, E. Linscott, A. M. Ganose, M. Siron, D. D. O’Regan, and K. A. Persson, High- throughput determination of HubbardUand HundJ values for transition metal oxides via the linear response formalism, Phys. Rev. Mater.8, 014409 (2024)

2024
[26]

Leiria Campo Jr and M

V. Leiria Campo Jr and M. Cococcioni, Extended DFT+U+Vmethod with on-site and inter-site elec- tronic interactions, J. Phys. Condens. Matter22, 055602 (2010)

2010
[27]

Cococcioni and N

M. Cococcioni and N. Marzari, Energetics and cathode voltages of LiMPO 4 olivines (M= Fe, Mn) from ex- tended Hubbard functionals, Phys. Rev. Mater.3, 033801 (2019)

2019
[28]

Ricca, I

C. Ricca, I. Timrov, M. Cococcioni, N. Marzari, and U. Aschauer, Self-consistent DFT+U+Vstudy of oxy- gen vacancies in SrTiO 3, Phys. Rev. Res.2, 023313 (2020)

2020
[29]

Tancogne-Dejean, Parameter-free hybridlike functional based on an extended Hubbard model: DFT+U+V, Phys

N. Tancogne-Dejean, Parameter-free hybridlike functional based on an extended Hubbard model: DFT+U+V, Phys. Rev. B102, 155117 (2020)

2020
[30]

Mahajan, A

R. Mahajan, A. Kashyap, and I. Timrov, Pivotal role of intersite Hubbard interactions in Fe-dopedα-MnO 2, J. Phys. Chem. C126, 14353 (2022)

2022
[31]

Timrov, F

I. Timrov, F. Aquilante, M. Cococcioni, and N. Marzari, Accurate electronic properties and intercalation voltages of olivine-type Li-Ion cathode materials from extended Hubbard functionals, PRX Energy1, 033003 (2022)

2022
[32]

J. Yang, T. Zhu, and S. Liu, Onsite and intersite elec- tronic correlations in the Hubbard model for halide per- ovskites, Phys. Rev. B106, 195159 (2022)

2022
[33]

Binci, M

L. Binci, M. Kotiuga, I. Timrov, and N. Marzari, Hy- bridization driving distortions and multiferroicity in rare- earth nickelates, Phys. Rev. Res.5, 033146 (2023)

2023
[34]

R. L. Kam, L. Binci, A. D. Kaplan, K. A. Persson, N. Marzari, and G. Ceder, Interplay between electron localization, magnetic order, and Jahn-Teller distortion dictates LiMnO 2 phase stability, Phys. Rev. B111, 245132 (2025)

2025
[35]

W. Yang, I. Timrov, F. Aquilante, and Y.-W. Son, Com- paring Hubbard parameters from linear-response theory and a Hartree-Fock-based approach, npj Comput Mater 10.1038/s41524-026-02123-9 (2026)

2026 doi
[36]

Haddadi, E

F. Haddadi, E. Linscott, I. Timrov, N. Marzari, and M. Gibertini, On-site and intersite Hubbard corrections in magnetic monolayers: The case of FePS 3 and CrI 3, Phys. Rev. Mater.8, 014007 (2024)

2024
[37]

Lee and Y.-W

S.-H. Lee and Y.-W. Son, First-principles approach with a pseudohybrid density functional for extended Hubbard interactions, Phys. Rev. Res.2, 043410 (2020)

2020
[38]

D. D. O’Regan, N. D. M. Hine, M. C. Payne, and A. A. Mostofi, Projector self-consistent DFT+Uusing nonorthogonal generalized Wannier functions, Phys. Rev. B82, 081102 (2010)

2010
[39]

I. Dabo, A. Ferretti, N. Poilvert, Y. Li, N. Marzari, and M. Cococcioni, Koopmans’ condition for density- functional theory, Phys. Rev. B82, 115121 (2010)

2010
[40]

N. L. Nguyen, N. Colonna, A. Ferretti, and N. Marzari, Koopmans-compliant spectral functionals for extended systems, Phys. Rev. X8, 021051 (2018)

2018
[41]

E. B. Linscott, N. Colonna, R. De Gennaro, N. L. Nguyen, G. Borghi, A. Ferretti, I. Dabo, and N. Marzari, koopmans: An open-source package for accurately and efficiently predicting spectral properties with Koopmans functionals, J. Chem. Theory Comput.19, 7097 (2023)

2023
[42]

C. Li, X. Zheng, N. Q. Su, and W. Yang, Localized orbital scaling correction for systematic elimination of delocal- ization error in density functional approximations, Natl. Sci. Rev.5, 203 (2018)

2018
[43]

N. Q. Su, A. Mahler, and W. Yang, Preserving symmetry and degeneracy in the localized orbital scaling correction approach, J. Phys. Chem. Lett.11, 1528 (2020)

2020
[44]

D. Wing, G. Ohad, J. B. Haber, M. R. Filip, S. E. Gant, J. B. Neaton, and L. Kronik, Band gaps of crystalline solids from Wannier-localization–based optimal tuning of a screened range-separated hybrid functional, Proc. Natl. Acad. Sci.118, e2104556118 (2021)

2021
[45]

Ma and L.-W

J. Ma and L.-W. Wang, Using Wannier functions to im- prove solid band gap predictions in density functional theory, Sci. Rep.6, 24924 (2016)

2016
[46]

Marzari, A

N. Marzari, A. A. Mostofi, J. R. Yates, I. Souza, and D. Vanderbilt, Maximally localized Wannier functions: Theory and applications, Rev. Mod. Phys.84, 1419 (2012)

2012
[47]

Mahler, J

A. Mahler, J. Z. Williams, N. Q. Su, and W. Yang, Wan- nier functions dually localized in space and energy (2022), arXiv:2201.07751

2022 arXiv
[48]

G. Ohad, M. Camarasa-G´ omez, J. B. Neaton, A. Rama- subramaniam, T. Gould, and L. Kronik, Foundations of the Ionization Potential Condition for Localized Electron Removal in Density Functional Theory, Phys. Rev. Lett. 136, 026401 (2026)

2026
[50]

Bastonero, C

L. Bastonero, C. Malica, E. Macke, M. Bercx, S. Hu- ber, I. Timrov, and N. Marzari, First-principles Hubbard parameters with automated and reproducible workflows, npj Comput Mater11, 183 (2025)

2025
[51]

Uhrin, A

M. Uhrin, A. Zadoks, L. Binci, N. Marzari, and I. Tim- rov, Machine learning Hubbard parameters with equiv- ariant neural networks, npj Comput Mater11, 19 (2025). 8

2025
[52]

Beida, G

W. Beida, G. Bihlmayer, C. Friedrich, G. Michalicek, D. Wortmann, and S. Bl¨ ugel, Implementation and appli- cation of a DFT+U+Vapproach within the all-electron FLAPW method (2026), arXiv:2511.08002 [cond-mat]

2026
[53]

Zhang, D

Y. Zhang, D. Ke, R. Maniar, T. Lebeda, P. Zhang, J. Sun, and J. P. Perdew, Electron localization in noncompact covalent bonds captured by the r2SCAN+V approach, Proc. Natl. Acad. Sci.123, e2529764123 (2026)

2026
[56]

Giannozzi, S

P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococ- cioni, I. Dabo, A. Dal Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F. ...

2009
[65]

DFT+U+Vis equivalent to DFT+Uwith density-dependent hybridized projectors

I. Timrov, N. Marzari, and M. Cococcioni, Self-consistent Hubbard parameters from density-functional perturba- tion theory in the ultrasoft and projector-augmented wave formulations, Phys. Rev. B103, 045141 (2021). Supplemental Material for “DFT+U+Vis equivalent to DFT+Uwith d...

2021
[66]

Q. Zhao, E. I. Ioannidis, and H. J. Kulik, Global and local curvature in density functional theory, J. Chem. Phys.145, 054109 (2016)

2016
[67]

Cococcioni and S

M. Cococcioni and S. de Gironcoli, Linear response ap- proach to the calculation of the effective interaction pa- rameters in the LDA+Umethod, Phys. Rev. B71, 035105 (2005)

2005
[68]

Timrov, N

I. Timrov, N. Marzari, and M. Cococcioni, Hubbard parameters from density-functional perturbation theory, Phys. Rev. B98, 085127 (2018)

2018
[69]

(S2) to obtain spin-indexed response matricesχ σσ′ IJ

A spin-resolved extension of this linear-response method also exists, which excludes the sums over spin channels in Eq. (S2) to obtain spin-indexed response matricesχ σσ′ IJ . See Ref. [16] for more details
[70]

Giannozzi, S

P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, 6 Ni (maj)Ni (min)OO -3.45 0.01 -0.01 -3.45 0.01 -0.01 -3.47 -3.47 -3.47 1.98 -0.22 0.22 1.98 -0.22 0.22 -3.47 -3.47 -3.47 0.01 -0.22 0.01 -0.22 -0.01 0.22 -0.01 0.22 (a) VU + V -3.45 -0.01 -0.01 -3.45 -0.01 -0.01 -3.47 ...

2009
[71]

Giannozzi, O

P. Giannozzi, O. Andreussi, T. Brumme, O. Bunau, M. Buongiorno Nardelli, M. Calandra, R. Car, C. Cavaz- zoni, D. Ceresoli, M. Cococcioni, N. Colonna, I. Carn- imeo, A. Dal Corso, S. de Gironcoli, P. Delugas, R. A. DiStasio, A. Ferretti, A. Floris, G. Fratesi, G. Fugallo, R. Ge...

2017
[72]

J. P. Perdew, A. Ruzsinszky, G. I. Csonka, O. A. Vydrov, G. E. Scuseria, L. A. Constantin, X. Zhou, and K. Burke, Restoring the density-gradient expansion for exchange in solids and surfaces, Phys. Rev. Lett.100, 136406 (2008)

2008
[73]

M. J. van Setten, M. Giantomassi, E. Bousquet, M. J. Verstraete, D. R. Hamann, X. Gonze, and G. M. Rig- nanese, The PseudoDojo: Training and grading a 85 ele- ment optimized norm-conserving pseudopotential table, Comput. Phys. Commun.226, 39 (2018)

2018
[74]

D. R. Hamann, Optimized norm-conserving Vanderbilt pseudopotentials, Phys. Rev. B88, 085117 (2013)

2013
[75]

Marzari, D

N. Marzari, D. Vanderbilt, and M. C. Payne, Ensem- ble Density-Functional Theory for Ab Initio Molecular Dynamics of Metals and Finite-Temperature Insulators, Phys. Rev. Lett.79, 1337 (1997)

1997
[76]

Timrov, N

I. Timrov, N. Marzari, and M. Cococcioni, Self-consistent Hubbard parameters from density-functional perturba- tion theory in the ultrasoft and projector-augmented wave formulations, Phys. Rev. B103, 045141 (2021)

2021
[77]

Timrov, N

I. Timrov, N. Marzari, and M. Cococcioni, HP – A code for the calculation of Hubbard parameters using density- functional perturbation theory, Comput. Phys. Commun. 279, 108455 (2022)

2022
[78]

Carta, I

A. Carta, I. Timrov, P. Mlkvik, A. Hampel, and C. Ed- erer, Explicit demonstration of the equivalence between DFT+Uand the Hartree-Fock limit of DFT+DMFT, Phys. Rev. Res.7, 013289 (2025)

2025
[79]

Linscott, A

E. Linscott, A. Carta, and N. Marzari, DFT+U+V is equivalent to DFT+Uwith density-dependent hy- bridized projectors, Materials Cloud Archive2026.X, 10.24435/materialscloud:zb-sz (2026)

2026 doi
[80]

X. Wang, T. Tian, and U. Aschauer, Beautiful Atoms: A Python library for the manipulation and visualization of atomistic structure using Blender,https://github. com/beautiful-atoms/beautiful-atoms(2022)

2022
[81]

E. B. Linscott, D. J. Cole, M. C. Payne, and D. D. O’Regan, Role of spin in the calculation of HubbardU and Hund’sJparameters from first principles, Phys. Rev. B98, 235157 (2018)

2018

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [1]

Hohenberg and W

P. Hohenberg and W. Kohn, Inhomogeneous electron gas, Phys. Rev.136, B864 (1964)

1964

[2] [2]

Kohn and L

W. Kohn and L. J. Sham, Self-consistent equations in- cluding exchange and correlation effects, Phys. Rev.140, A1133 (1965)

1965

[3] [3]

R. O. Jones, Density functional theory: Its origins, rise to prominence, and future, Rev. Mod. Phys.87, 897 (2015)

2015

[4] [4]

Marzari, A

N. Marzari, A. Ferretti, and C. Wolverton, Electronic- structure methods for materials design, Nat. Mater.20, 736 (2021)

2021

[5] [5]

Van Noorden, These are the most-cited research pa- pers of all time, Nature640, 591 (2025)

R. Van Noorden, These are the most-cited research pa- pers of all time, Nature640, 591 (2025)

2025

[6] [6]

Terakura, T

K. Terakura, T. Oguchi, A. R. Williams, and J. K¨ ubler, Band theory of insulating transition-metal monoxides: Band-structure calculations, Phys. Rev. B30, 4734 (1984)

1984

[7] [7]

V. I. Anisimov, J. Zaanen, and O. K. Andersen, Band theory and Mott insulators: HubbardUinstead of Stoner I, Phys. Rev. B44, 943 (1991)

1991

[8] [8]

Hubbard, Electron correlations in narrow energy bands, Proc

J. Hubbard, Electron correlations in narrow energy bands, Proc. R. Soc. A276, 238 (1963)

1963