Pith. sign in

REVIEW 3 major objections 4 minor 60 references

Physically motivated iso-orbital indicator for meta-GGA exchange functionals

T0 review · 3 major / 4 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 This paper claims that replacing the denominator of the meta-GGA iso-orbital indicator with an orbital-free Pauli kinetic-energy enhancement factor removes the low-density divergence and systematically improves semiconductor band gaps.

desk verdict Novel parameter-free cure for the alpha divergence, but the band-gap gains likely come from a global rescaling of alpha in the interstitial region, not just the tail fix. read the letter →

arxiv 2607.17736 v1 pith:2DOVVB2Q submitted 2026-07-20 cond-mat.mtrl-sci

Jeet Sharma , Abhishek Bhattacharjee , Bikash Patra , Prasanjit Samal This is my paper · ORCID
classification cond-mat.mtrl-sci PACS 71.15.Mb71.20.-b
keywords meta-GGAiso-orbitalindicatorPaulikineticenergydensityorbital-freeDFTbandgapexchangefunctionalsemiconductorstheory
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

Meta-GGA density functionals use an iso-orbital indicator to distinguish single-orbital regions from overlapping-electron regions, but the standard indicator blows up in low-density atomic tails, causing unphysical exchange potentials and systematic band-gap errors. The authors replace the denominator of this indicator with a Pauli kinetic-energy enhancement factor taken from orbital-free DFT, which decays with density gradient and kills the divergence without any empirical parameter. Testing two such factors (LKT and PGS) inside the r2SCAN and MS2 exchange functionals, they find smoother exchange potentials, restored electron localization in atomic tails, and band-gap mean absolute errors reduced by 41.1% (r2SCAN@PGS) and 48.8% (MS2@PGS) on a ten-member semiconductor benchmark, while cohesive-energy accuracy is largely preserved. The paper argues this improvement is physical and transferable, not specific to one functional form, and motivates building a new meta-GGA around the corrected indicator.

What carries the argument

The key object is the modified iso-orbital indicator alpha_r2@OF = (tau - tau_vW)/(tau_UEG F(s) + tau_vW), where tau is the Kohn-Sham kinetic energy density, tau_vW the von Weizsäcker kinetic energy density, tau_UEG the uniform-electron-gas kinetic energy density, and F(s) a Pauli kinetic-energy enhancement factor from orbital-free DFT. Two forms are used: the LKT factor F(s)=1/cosh(1.3s) and the PGS factor F(s)=exp(-40/27 s^2), both satisfying F(0)=1 and decaying to zero as s grows. This decay makes the denominator nonvanishing in the low-density tail, eliminating the alpha divergence without an empirical regularization parameter, while keeping the indicator's physical limits intact.

What would settle it

Compute the modified indicator at the bond midpoint of a covalent solid such as diamond or silicon, where the reduced gradient s is near 1; PGS already gives F(s)~0.23 there, so if the indicator differs materially from the original alpha in that region, the claim that the correction is inert in bonding regions is falsified. Alternatively, apply the indicator to a meta-GGA not in the r2SCAN/MS2 family and check whether the band-gap improvement and potential smoothing persist.

Watch

Extended reading notes

Core claim

The central claim is that the low-density divergence of the iso-orbital indicator is a physical defect, not a numerical nuisance, and it can be cured by rescaling the uniform-electron-gas kinetic energy density in the denominator with a Pauli enhancement factor F(s) from orbital-free DFT. The proposed indicator is alpha_r2@OF = (tau - tau_vW)/(tau_UEG F(s) + tau_vW), where F(s) satisfies F(0)=1 and decays to zero as the reduced density gradient s grows, so the denominator survives the atomic tail. This preserves the single-orbital limit (alpha=0) and the slowly-varying limit (alpha=1) without reparametrization, unlike the alternative beta indicator. Within r2SCAN and MS2, the modified indica

Load-bearing premise

The paper assumes that the Pauli enhancement factors LKT and PGS are transferable physical inputs and that their decay leaves bonding and interstitial regions undisturbed, without reparametrizing the parent functionals to verify this.

Editorial extensions

If this is right

  • Meta-GGA functionals built on the modified indicator should be less prone to the grid-convergence problems, pseudopotential-generation instabilities, and oscillatory exchange potentials that plague the standard alpha.
  • Band-gap mean absolute error on a ten-member cubic semiconductor benchmark drops by 41.1% for r2SCAN@PGS and 48.8% for MS2@PGS, while cohesive-energy accuracy is largely preserved.
  • The exchange potential becomes smooth and monotonic in the atomic tail, suppressing oscillations that otherwise amplify into second functional derivatives such as TDDFT kernels.
  • Because the single-orbital and slowly-varying limits are unchanged, the indicator can be inserted into existing meta-GGAs without retuning interpolation functions, unlike the beta alternative.
  • The consistent improvement across two functionals with distinct constructions supports a physical, not functional-specific, origin and motivates development of a new meta-GGA with a constraint-satisfying iso-orbital indicator.
  • The crossover from cohesive-energy overestimation to underestimation between the LKT and PGS factors brackets the acceptable decay rate for future functional parametrizations.

Reading between the lines

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

  • The same construction could be ported to other meta-GGAs (e.g., SCAN or deorbitalized variants) and might cure not only atomic-tail divergence but also the spurious bump in the hydrogen exchange potential noted in deorbitalization work.
  • Because PGS's faster decay is what drives the band-gap improvement but also degrades r2SCAN's structural properties, a future meta-GGA that optimizes F(s)'s decay rate simultaneously with the enhancement-factor parameters could recover structural accuracy while keeping the band-gap gains.
  • The boundedness of the modified indicator in the tail suggests it could stabilize SCF iterations for large or low-density systems, and a natural test would be molecules with diffuse basis sets or surfaces and interfaces where tail regions matter.
  • The sign reversal of cohesive-energy mean error between LKT and PGS indicates that a factor with intermediate decay might offer the best compromise; this is testable by constructing a hybrid enhancement factor and benchmarking.
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.

Referee Report

3 major / 4 minor

Summary. The manuscript proposes replacing the denominator of the standard meta-GGA iso-orbital indicator α = (τ − τ_vW)/τ_UEG with τ_UEG F_θ(s) + τ_vW, using two Pauli kinetic-energy enhancement factors from the orbital-free DFT literature (LKT and PGS). The resulting indicator α_r2@OF is claimed to remove the low-density p-type-tail divergence of α without an empirical regularization parameter, to restore f_ELF → 1 in atomic tails, to suppress spurious oscillations in semilocal exchange potentials, and to reduce band-gap MAE on a ten-member semiconductor benchmark by 41.1% for r2SCAN@PGS and 48.8% for MS2@PGS while largely preserving cohesive energies. The asymptotic analysis in Appendix A supports the mathematical claim that α_r2@OF → 0 in p-type tails. However, the numerical evidence does not currently isolate the tail-region effect from a global rescaling of α in the interstitial region, and the claim that the correction is inert except in large-s tails is quantitatively inaccurate for PGS.

Significance. If established, the construction would be a simple, physically motivated cure for a documented numerical instability of meta-GGA functionals, with no new empirical regularization parameter. The paper provides concrete analytic asymptotics, explicit benchmark tables, and a comparison across two parent functionals. The Appendix A derivation is internally consistent and the central mathematical property — recovery of f_ELF → 1 in the p-type tail — is clearly demonstrated. The main limitation is interpretive: because the PGS/LKT parameters are imported without retuning, and because Eq. (7) substantially rescales α in the s≈1–3 interstitial region that the paper itself identifies as controlling semiconductor band gaps, the claimed physical, tail-specific origin of the band-gap improvement is not yet established.

major comments (3)
  1. [Sec. IV.B, Eq. (7), Fig. 1] The claim that F_θ(s) ≈ 1 for 'moderate' s and that the correction is inert in bonding/interstitial regions is quantitatively wrong for PGS. With τ_vW = (5/3) s^2 τ_UEG, the new denominator is τ_UEG[F(s) + (5/3)s^2]. At s=1, F_PGS = exp(−40/27) ≈ 0.23, so the denominator is ≈1.89 τ_UEG, whereas r2SCAN's denominator is ≈1.0017 τ_UEG (η=10^−3). At s=2, the PGS denominator is ≈6.67 τ_UEG. Section IV.C identifies s≈1–3 as the region controlling semiconductor band gaps, so α_r2@OF rescales α by a factor of roughly 2–7 in exactly that region. The band-gap improvements may therefore be caused by globally shifting the range of α sampled by the parent interpolation functions, rather than by removing the p-type tail divergence. A control calculation in which F_θ(s) is replaced by a step-like factor that is exactly 1 for s below, say, 1.5 would separate the tail effect from the global rescale.
  2. [Tables II and III; Sec. IV.C] The absence of retuning is not a sufficient test of transferability, and the structural/cohesive data show the collateral cost of an untuned global rescale. Table II reports r2SCAN@PGS degrading V0 MAE from 0.509 to 1.992 Å^3 and B0 MAE from 3.335 to 10.462 GPa. Table III shows the cohesive-energy mean error reversing sign from +0.046 to −0.198 eV/atom for r2SCAN@PGS. The paper itself states that 'the response can be restored by tuning the parameter μ in PGS (or a in LKT)'. This is a direct admission that the imported OFDFT parameters are not inert with respect to the parent functional's tuned interpolation. To support the claim that the improvement is 'physical rather than functional-specific', the authors need to show either that the improvement survives when α_r2@OF is designed to be nearly identical to α in the bonding/interstitial region, or that the parent interpolation functions r
  3. [Abstract and Sec. V] The phrase 'consistent improvement across two functionals' is overstated. Table II shows that r2SCAN@LKT degrades the band-gap MAE relative to r2SCAN (0.409 eV vs. 0.321 eV), while MS2@LKT improves it. Only PGS improves both parent functionals. The conclusion should be qualified to the PGS enhancement factor, and the claim that the approach is uniformly transferable across LKT and PGS is contradicted by the same table.
minor comments (4)
  1. [Sec. IV.C] The sentence 'whereas with LKT enhancement factor deteriorates for r2SCAN@LKT by 27.4% and improves MS2 by 21%' is grammatically awkward; consider 'the LKT factor degrades r2SCAN@LKT by 27.4% but improves MS2 by 21%'.
  2. [Table III caption] The abbreviation 'ME' is used in the caption but not defined; define it as the mean error (eV/atom) in the caption.
  3. [Sec. II, Table I footnote] The footnote for the noncovalent α_r2@OF row states that divergence occurs only at the strict τ_vW=0 bond center; it would be helpful to state explicitly that near, but not exactly at, the bond center the modified indicator still becomes large (though finite) because F_θ(s)≈1 for small s.
  4. [Sec. III] Minor typo: 'for consistency' beginning a sentence should be capitalized ('For consistency').

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the OFDFT enhancement factors are external inputs and the asymptotic analysis is parameter-free; band-gap results are empirical outputs.

full rationale

The central construction (Eq. 7) takes F_LKT = 1/cosh(1.3s) and F_PGS = exp(-40/27 s^2) directly from the orbital-free DFT literature (Refs. 31 and 32); the paper does not fit a=1.3 or mu=40/27 to its band-gap, volume, or cohesive-energy data. The claimed tail-divergence cure follows from F_theta(s)->0 as s->infinity, and the Appendix A asymptotics are derived in a parameter-free way from that decay and the exact shell decomposition of the Kohn-Sham KED; they are not outputs of the benchmark. The band-gap and cohesive-energy numbers are genuine computed results of modified r2SCAN and MS2 functionals, reported alongside the parent functionals' errors, so no fitted parameter is renamed as a prediction. Self-citations to Patra/Samal work (Refs. 13,14) appear only as background motivation for KED-sensitive exchange potentials; the load-bearing statement about the role of the iso-orbital indicator in band gaps is credited to external Tran/Blaha benchmarks (Refs. 33,34), so this is not a load-bearing self-citation chain. The skeptic's concern that PGS rescales alpha strongly in the s=1-3 interstitial region is a legitimate alternative explanation for the band-gap acceleration, not a circular reduction; the paper itself exposes the collateral structural degradation for r2SCAN in Table II and admits that r2SCAN's parameters were not retuned ('r2SCAN is a much complicated functional than MS2 with tightly fitted parameters adjusted for bar-alpha, which we did not retune for alpha_r2@OF in this work') and that it refrained from reparametrizing mu and a. These are scientific limitations and confounds, but not cases where the claimed result is equivalent by construction to its inputs. Therefore no circular step meeting the required evidentiary standard is present.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new entities; it repurposes two published Pauli KED enhancement factors as denominator screens. The main ledger items are the externally fitted LKT/PGS parameters and the domain assumption that asymptotic one-orbital scaling applies. A notable ad-hoc assumption is that the parent meta-GGA interpolation functions do not need retuning, which the paper itself contradicts for structural properties.

free parameters (2)
  • F_LKT parameter a = 1.3
    Taken from Luo-Karasiev-Trickey (2018). Not refitted here; controls the decay rate of the LKT enhancement factor and therefore which parent functional benefits.
  • F_PGS parameter mu = 40/27
    Taken from Constantin-Fabiano-Della Sala (2018). Not refitted here; controls the Gaussian decay rate of the PGS enhancement factor.
assumptions (4)
  • standard math Exact shell decomposition of the KS KED: tau_KS_nl = tau_vW[rho_nl] + l(l+1)/(2) rho_nl / r^2 (Eq. A1)
    Used throughout Appendix A to derive the p-type tail behavior. This is a known exact decomposition for spherically symmetric orbitals.
  • domain assumption In the asymptotic tail, only the HOMO contributes and rho ~ exp(-2 kappa r) with kappa = sqrt(-2 eps_HOMO)
    Used in Appendix A.3 to derive scaling of rho, tau_vW, and tau_UEG. Standard DFT asymptotic behavior.
  • domain assumption The Pauli enhancement factors satisfy F(0)=1 and decay monotonically to zero as s -> infinity, preserving the slowly varying and single-orbital limits
    Relies on the LKT/PGS papers. This is the design constraint that makes alpha_r2@OF reduce to alpha in the uniform limit and stay bounded in the tail.
  • ad hoc to paper The parent functionals' interpolation functions, fitted for alpha/bar-alpha, remain near-optimal under the new indicator without retuning
    Explicitly or implicitly assumed when reporting band-gap and cohesive-energy quality for unmodified r2SCAN/MS2 interpolation functions; the paper admits structural degradation for r2SCAN@PGS and a cohesive-energy sign reversal, indicating this assumption is not fully satisfied.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Physically motivated iso-orbital indicator for meta-GGA exchange functionals." pith.science (2026). https://pith.science/paper/2DOVVB2Q

@misc{pith2026260717736,
  author       = {Pith},
  title        = {Pith review of: Physically motivated iso-orbital indicator for meta-GGA exchange functionals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2DOVVB2Q}},
  note         = {Machine review of arXiv:2607.17736}
}
abstract

The iso-orbital indicator $\alpha = (\tau - \tau^\mathrm{vW})/\tau^\mathrm{UEG}$ is a key ingredient of meta-generalized gradient approximation (meta-GGA) functionals, but diverges in low-density tails , causing unphysical exchange potentials and systematic band gap errors as noted in [J. Chem. Phys. 150, 161101 (2019)]. We replace the denominator of $\alpha$ with a physically motivated Pauli KED drawn from the orbital-free DFT literature, eliminating the divergence in the low density atomic tail without any empirical regularization parameter. Testing two such enhancement factors: LKT and PGS, within the r$^2$SCAN and MS2 exchange functionals, we find that the modified indicators suppress spurious oscillations in the semilocal exchange potential and restore correct electron localization in atomic tails. For a ten-member cubic semiconductor benchmark, the band gap mean absolute error is reduced by 41.1 % for r$^2$SCAN@PGS and 48.8 % for MS2@PGS, while cohesive energy accuracy is largely preserved. The consistent improvement across two functionals with distinct constructions confirms a physical rather than functional specific origin, and motivates further development of meta-GGA functionals with constraint satisfying iso-orbital indicators.

Figures

Figures reproduced from arXiv: 2607.17736 by the authors.

Figure 1
Figure 1. FIG. 1. Pauli kinetic energy enhancement factors [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Comparison of electron densities and iso-orbital indicators for Li and C atoms. Spherically averaged electron densities [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Comparison of the electron localization function [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Comparison of semilocal exchange potential [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Comparison of semilocal exchange potential [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Boxplots of absolute errors in (left) band gap (eV), (center) bulk modulus(GPa), and (right) equilibrium volume [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7. Same as in Fig. 6 for MS2, MS2@LKT, and MS2@PGS on the same semiconductor test set. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Box plots of absolute errors in cohesive energy [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 1 linked inside Pith

  1. [1]

    The KS 9 KED equals the von Weizsäcker KED exactly:τ= 1 2 |∇ϕ|2 = |∇ρ|2 8ρ =τ vW

    Single orbital limit: T ail region ofs-type HOMO (l= 0) In the asymptotic tail, only the HOMO contributes, thus can be identified as single orbital region. The KS 9 KED equals the von Weizsäcker KED exactly:τ= 1 2 |∇ϕ|2 = |∇ρ|2 8ρ =τ vW. For ans-type HOMO,l= 0, and from Eq. (A1): τ KS n0 =τ vW[ρn0] + 0·1 2 ρn0 r2 =τ vW[ρn0].(A2) Thereforeτ=τ vW in eachs-s...

  2. [2]

    Sinces→0, both enhancement factors recoverF θ(0) = 1

    Slowly varying density limit Intheslowlyvaryinglimit, thereducedgradients→0, which implies: τ vW = |∇ρ|2 8ρ →0, τ≈τ UEG.(A5) The second condition is the gradient expansion result for the KS KED [8]. Sinces→0, both enhancement factors recoverF θ(0) = 1. Therefore, α≈¯α≈α r2OF ≈1, β≈ 1 2 .(A6) The slowly-varying limit≈1is preserved byα r2OF be- causeF θ(0) ...

  3. [3]

    For ap-type HOMO,l= 1, and from Eq

    T ail region:p-type HOMO (l= 1) This is the critical case. For ap-type HOMO,l= 1, and from Eq. (A1): τ KS n1 =τ W [ρn1] + 1·2 2 ρn1 r2 =τ W [ρn1] + ρn1 r2 .(A7) In the asymptotic region only the HOMO contributes, so the Pauli KED is: τ−τ vW = ρn1 r2 >0,(A8) which is strictly positive even asρ→0. This is the fundamental reason why the p-type tail is qualit...

  4. [4]

    Since ρ5/3 vanishes faster thanρ: α∼ ρ ρ5/3 =ρ −2/3 → ∞,(A21) ¯α∼ρ−2/3 → ∞,(A22) wherefor¯αtheηfloorvanishessinceτ vW = 0atthebond center, leaving the denominator≈τUEG ∝ρ 5/3 [10]

    Noncovalent density overlap region In the low-density interstitial region between two weakly interacting closed-shell systems, the important conditions are [30]: ρ→0, τ vW = 0at bond centers,(A20) withτ UEG ∝ρ 5/3 andτ−τ vW =τ∝ρ( [30]. Since ρ5/3 vanishes faster thanρ: α∼ ρ ρ5/3 =ρ −2/3 → ∞,(A21) ¯α∼ρ−2/3 → ∞,(A22) wherefor¯αtheηfloorvanishessinceτ vW = 0...

  5. [5]

    J. Sun, J. P. Perdew, and M. Seidl, Physical Review B 81, 085123 (2010)

    2010

  6. [6]

    Kohn and L

    W. Kohn and L. J. Sham, Phys. Rev.140, A1133 (1965)

    1965

  7. [7]

    a convenient means for preliminary investigation rather than a viable general functional

    offer a favorable accuracy-cost tradeoff, but the high- estsemilocalrung, themeta-GGA(MGGA),achievessig- nificantly improved accuracy by additionally incorporat- ing the KS kinetic energy density (KED)τ[8, 9]. The additional ingredientτencodes local orbital-overlap in- formation absent in LDA and GGAs, and enables MG- GAs to simultaneously describe molecu...

    arXiv 2026

  8. [8]

    R. G. Parr and W. Yang,Density-Functional Theory of Atoms and Molecules(Oxford University Press, Oxford, UK, and New York, USA, 1989), ISBN 0199878722

    1989

Show all 60 references
  1. [9]

    J. P. Perdew and S. Kurth, inA Primer in Density Func- tional Theory, edited by C. Fiolhais, F. Nogueira, and M. A. L. Marques (Springer-Verlag, Berlin, 2003), vol. 620 ofLecture Notes in Physics

    2003

  2. [10]

    J. P. Perdew and Y. Wang, Physical Review B45, 13244 (1992)

    1992

  3. [11]

    J. P. Perdew and A. Zunger, Phys. Rev. B23, 5048 (1981)

    1981

  4. [12]

    J. P. Perdew, K. Burke, and M. Ernzerhof, Physical Re- view Letters77, 3865 (1996)

    1996

  5. [13]

    J. Sun, A. Ruzsinszky, and J. P. Perdew, Phys. Rev. Lett. 115, 036402 (2015)

    2015

  6. [14]

    Tao and Y

    J. Tao and Y. Mo, Physical Review Letters117, 073001 (2016)

    2016

  7. [15]

    J. W. Furness, A. D. Kaplan, J. Ning, J. P. Perdew, and J. Sun, The Journal of Physical Chemistry Letters11, 8208 (2020)

    2020

  8. [16]

    A. D. Becke, International Journal of Quantum Chem- istry23, 1915 (1983)

    1915

  9. [17]

    A. D. Becke, Journal of Chemical Physics98, 5648 (1993)

    1993

  10. [18]

    Jana and P

    S. Jana and P. Samal, The Journal of Chemical Physics 148, 024111 (2018), ISSN 0021-9606

    2018

  11. [19]

    Patra, S

    B. Patra, S. Jana, L. A. Constantin, and P. Samal, Phys. Rev. B100, 155140 (2019)

    2019

  12. [20]

    J. Sun, B. Xiao, and A. Ruzsinszky, Journal of Chemical Physics137, 051101 (2012)

    2012

  13. [21]

    J. Sun, B. Xiao, Y. Fang, R. Haunschild, P. Hao, A. Ruzsinszky, G. I. Csonka, G. E. Scuseria, and J. P. Perdew, Physical Review Letters111, 106401 (2013)

    2013

  14. [22]

    Smiga, L

    S. Smiga, L. A. Constantin, F. Della Sala, and E. Fabi- ano, Computation7(2019), ISSN 2079-3197

    2019

  15. [23]

    Zhao and D

    Y. Zhao and D. G. Truhlar, The Journal of Chemical Physics125, 194101 (2006)

    2006

  16. [24]

    C. F. von Weizsäcker, Zeitschrift für Physik96, 431 (1935)

    1935

  17. [25]

    J. P. Perdew, A. Ruzsinszky, J. Sun, and K. Burke, The Journal of Chemical Physics140, 18A533 (2014)

    2014

  18. [26]

    Zhang, D

    Y. Zhang, D. Kitchaev, J. Yang, T. Chen, S. Dacek, R. Sarmiento-Perez, M. Marques, H. Peng, G. Ceder, J. Perdew, et al., npj Computational Materials4(2018)

    2018

  19. [27]

    J. Sun, R. C. Remsing, Y. Zhang, Z. Sun, A. Ruzsinszky, H. Peng, Z. Yang, A. Paul, U. Waghmare, X. Wu, et al., Nature Chemistry8, 831 (2016)

    2016

  20. [28]

    J. W. Furness, Y. Zhang, C. Lane, I. G. Buda, B. Barbi- ellini, R. S. Markiewicz, A. Bansil, and J. Sun, Commu- nications Physics1, 11 (2018)

    2018

  21. [29]

    Z.-h. Yang, H. Peng, J. Sun, and J. P. Perdew, Phys. Rev. B93, 205205 (2016)

    2016

  22. [30]

    E. R. Johnson, A. D. Becke, C. D. Sherrill, and G. A. DiLabio, The Journal of Chemical Physics131, 034111 (2009)

    2009

  23. [31]

    Yao and Y

    Y. Yao and Y. Kanai, The Journal of Chemical Physics 146, 224105 (2017)

    2017

  24. [32]

    Śmiga, S

    S. Śmiga, S. Siecińska, and E. Fabiano, Phys. Rev. B 101, 165144 (2020)

    2020

  25. [33]

    Della Sala, E

    F. Della Sala, E. Fabiano, and L. A. Constantin, Interna- tional Journal of Quantum Chemistry116, 1641 (2016)

    2016

  26. [34]

    A. P. Bartók and J. R. Yates, The Journal of Chemical Physics150, 161101 (2019)

    2019

  27. [35]

    J. W. Furness and J. Sun, Physical Review B99, 041119 (2019)

    2019

  28. [36]

    K. Luo, V. V. Karasiev, and S. B. Trickey, Physical Re- view B98, 041111(R) (2018)

    2018

  29. [37]

    L. A. Constantin, E. Fabiano, and F. Della Sala, The Journal of Physical Chemistry Letters9, 4385 (2018), 11 pMID: 30019904

    2018

  30. [38]

    Tran and P

    F. Tran and P. Blaha, Physical Review Letters102, 226401 (2009)

    2009

  31. [39]

    Tran and P

    F. Tran and P. Blaha, The Journal of Physical Chemistry A121, 3318 (2017)

    2017

  32. [40]

    Levy and H

    M. Levy and H. Ou-Yang, Physical Review A38, 625 (1988)

    1988

  33. [41]

    W. Mi, K. Luo, S. Trickey, and M. Pavanello, Chemical Reviews123(2023)

    2023

  34. [42]

    E. H. Lieb,Thomas-fermi and related theories of atoms and molecules(1980)

    1980

  35. [43]

    J. L. Gázquez and J. Robles, Journal of Chemical Physics 76, 1467 (1982)

    1982

  36. [44]

    Della Sala, E

    F. Della Sala, E. Fabiano, and L. A. Constantin, Physical Review B91, 035126 (2015)

    2015

  37. [45]

    V. V. Karasiev, T. Sjostrom, J. Dufty, and S. B. Trickey, Physical Review B89, 161108 (2014)

    2014

  38. [46]

    Kresse and J

    G. Kresse and J. Hafner, Physical Review B47, 558 (1993)

    1993

  39. [47]

    P. E. Blöchl, Physical Review B50, 17953 (1994)

    1994

  40. [48]

    Kresse and D

    G. Kresse and D. Joubert, Physical Review B59, 1758 (1999)

    1999

  41. [49]

    A. D. Kaplan and J. P. Perdew, Physical Review Mate- rials6, 083801 (2022)

    2022

  42. [50]

    V. Wang, N. Xu, J.-C. Liu, G. Tang, and W.-T. Geng, Computer Physics Communications267, 108033 (2021)

    2021

  43. [51]

    T. Koga, K. Kanayama, S. Watanabe, and . Thakkar, Ajit J., International Journal of Quantum Chemistry71, 491 (1999)

    1999

  44. [52]

    Q. Sun, X. Zhang, S. Banerjee, P. Bao, M. Bar- bry, R. Berger, S. N. Boyce, P.-F. Chang, C. Cheng, W. Derenzo, et al., The Journal of Chemical Physics153, 024109 (2020)

    2020

  45. [53]

    Weigend and R

    F. Weigend and R. Ahlrichs, Physical Chemistry Chem- ical Physics7, 3297 (2005)

    2005

  46. [54]

    A. D. Becke and K. E. Edgecombe, The Journal of Chem- ical Physics92, 5397 (1990)

    1990

  47. [55]

    Francisco, B

    H. Francisco, B. Thapa, S. B. Trickey, and A. C. Cancio, Phys. Rev. Mater.10, 043801 (2026)

    2026

  48. [56]

    P. Haas, F. Tran, and P. Blaha, Phys. Rev. B79, 085104 (2009)

    2009

  49. [57]

    L.Schimka, J.Harl, andG.Kresse, TheJournalofChem- ical Physics134, 024116 (2011), ISSN 0021-9606

    2011

  50. [58]

    Patra, S

    B. Patra, S. Jana, L. A. Constantin, and P. Samal, Phys. Rev. B100, 045147 (2019)

    2019

  51. [59]

    Sharma, B

    J. Sharma, B. Abhishek, P. Bikash, and P. Samal, sup- porting information (2026)

    2026

  52. [60]

    Lebeda, T

    T. Lebeda, T. Aschebrock, and S. Kümmel, Phys. Rev. Lett.133, 136402 (2024)

    2024

Pith tools

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