Pith. sign in

REVIEW 5 major objections 5 minor 6 references

Optical Anisotropy and Phase Matching in Non-Centrosymmetric Perovskite Oxides from DFT+U and DFT+U+V Functionals

T0 review · 5 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A low-cost DFT fix reproduces four ferroelectrics' birefringence.

desk verdict Useful benchmark of DFT+U+V for optical anisotropy, but the PbTiO3 geometry error and a cherry-picked experimental comparison undercut the central claim. read the letter →

arxiv 2608.06759 v1 pith:RDC66EP4 submitted 2026-08-07 cond-mat.mtrl-sci

Mohamed S.M.M. Ali , Ismaila Dabo (Department of Materials Science , Engineering , Wilton E. Scott Institute for Energy Innovation , Carnegie Mellon University) This is my paper · ORCID
classification cond-mat.mtrl-sci
keywords DFT+U+VopticalanisotropybirefringenceperovskiteoxidesHubbardparametersdensity-functionalperturbationtheoryphasematchingferroelectric
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

The paper tries to establish that the extended Hubbard functional DFT+U+V, with Hubbard parameters computed by density-functional perturbation theory and applied only to transition-metal onsite plus transition-metal–oxygen intersite terms, predicts the anisotropic optical response of BaTiO3, LiNbO3, KNbO3, and PbTiO3 in agreement with measured refractive indices and birefringence. The result matters because it offers a computationally cheap route to predictive optical anisotropy and phase-matching calculations without hybrid functionals or many-body perturbation theory. The paper also shows that plain DFT overestimates refractive indices while DFT+U alone suppresses anisotropy, and that the choice of orbital manifold is decisive. As a demonstration, it predicts that Bi–Mn co-substituted BaTiO3 develops a fourfold larger birefringence and a large visible dichroic ratio, with implications for polarization-sensitive photodetectors and integrated photonics.

What carries the argument

The load-bearing object is the extended Hubbard functional DFT+U+V, whose correction energy adds an onsite U term and an intersite V term between atom-centered orbital occupations. The parameters are obtained non-empirically as second derivatives of the total energy with respect to orbital occupancies through density-functional perturbation theory, which includes electronic screening. The decisive choice is the projection manifold: the paper finds that set II, with onsite Hubbard terms on transition metals plus intersite transition-metal–oxygen terms but no onsite oxygen U, reproduces the measured anisotropy. The optical response is then evaluated from the independent-particle momentum-space dielectric function, and phase-matching behavior is summarized by two new dimensionless metrics, the phase-matching fraction and the renormalized Dirichlet energy of the refractive-index-mismatch function.

What would settle it

Fix the relaxed geometry to the experimental lattice parameters for PbTiO3 (c/a = 1.063 instead of the calculated 1.150) and recompute birefringence with the same DFT+U+V set-II parameters; if the large birefringence that matches Ref. [43] collapses, the central agreement is an artifact of an over-distorted geometry rather than of the functional.

Watch

Extended reading notes

Core claim

The central claim is that correcting self-interaction errors with DFT+U alone is insufficient for non-centrosymmetric perovskite oxides because it localizes electrons too strongly and suppresses optical anisotropy, whereas adding intersite Hubbard V terms between transition-metal and oxygen orbitals restores the covalent hybridization that controls birefringence. Using DFPT-derived Hubbard parameters within the manifold {UTM, V_TM-O}, the paper reproduces the measured refractive-index dispersion and birefringence of BaTiO3, LiNbO3, KNbO3, and PbTiO3, identifies LiNbO3 as having the most favorable SHG phase-matching landscape among these, and extends the same parameter-free framework to substituted BaTiO3 solid solutions. The paper further claims that ACBN0, an alternative parameterization, systematically overestimates ferroelectric distortions and birefringence for these oxides, and that including onsite Hubbard U on oxygen within DFT+U+V destroys the anisotropy because oxygen orbitals respond weakly to the applied perturbation, making their U parameters spuriously large.

Load-bearing premise

The whole predictive chain assumes the Hubbard-corrected relaxed ferroelectric geometry is close enough to experiment and that omitting excitonic and local-field corrections does not change the anisotropy; PbTiO3's overestimated c/a of 1.150 versus the experimental 1.063 shows how tight that assumption is.

Editorial extensions

If this is right

  • If the central claim is right, DFT+U+V with DFPT parameters and the set-II manifold can be used to screen ferroelectric oxides for birefringence and phase matching at a fraction of the cost of hybrid or GW calculations.
  • Plain DFT overestimates refractive indices and DFT+U underestimates anisotropy in these materials, so uncorrected or U-only calculations should not be trusted for optical-device design in this class.
  • The phase-matching fraction and renormalized Dirichlet energy give dimensionless scalar descriptors for ranking materials by phase-matching bandwidth in screening workflows.
  • Bi–Mn co-substituted BaTiO3 is predicted to combine a fourfold birefringence increase with Type-I SHG phase matching over the 0.75–1.0 eV range, making it a concrete candidate for polarization-sensitive photodetection and integrated photonics.

Reading between the lines

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

  • The paper's PbTiO3 agreement leans on the birefringence data of Ref. [43] rather than the Sellmeier fit of Ref. [37]; if experimental consensus shifts toward the fit, the claimed agreement for PbTiO3 would need revisiting.
  • Because birefringence in these crystals is extremely sensitive to the ferroelectric distortion, the same functional should be tested on other phases and on compounds with smaller distortions to see whether the set-II manifold remains optimal.
  • The independent-particle dielectric function omits excitonic and local-field effects; comparing these DFT+U+V spectra against BSE-level spectra would show whether the predicted anisotropy survives those corrections.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The manuscript benchmarks DFT+U and DFT+U+V against the anisotropic optical response of BaTiO3, LiNbO3, KNbO3, and PbTiO3, with Hubbard parameters computed non-empirically using DFPT and, for comparison, ACBN0, for three orbital manifolds: {U_TM,U_O}, {U_TM,V_TM-O}, and {U_TM,U_O,V_TM-O}. It reports that the DFPT set-II functional (TM onsite plus TM-O intersite) gives the best refractive indices and birefringence, proposes two scalar phase-matching metrics, and applies the method to Zn- and Bi/Mn-substituted BaTiO3, predicting enhanced birefringence and dichroism from Mn-induced mid-gap states. The paper's central claim is that DFPT set II reproduces both the experimental crystal geometries and optical anisotropies of the four benchmark oxides at mean-field cost.

Significance. The work is significant in its methodological comparison and in providing concrete, falsifiable predictions, notably the blue-shifted Type-I SHG phase-matching frequency and enhanced visible-range dichroic ratio in Ba7BiTi7MnO24. Its main strength is that U and V are not fitted to optical data: the DFPT and ACBN0 parameter sets follow well-defined linear-response prescriptions, and the manifold comparison is systematic. The LiNbO3 birefringence agreement and the phase-matching landscape analysis are convincing. However, the benchmark evidence is not as uniform as the conclusion claims: the PbTiO3 geometry errors are large, the BaTiO3 internal displacements are also overestimated by roughly 50%, and the optical response omits excitonic and local-field effects. Because these issues bear directly on the central predictive claim, the paper needs major revision rather than acceptance.

major comments (5)
  1. [Table 2, PbTiO3 row; Assessment of optical anisotropy calculations] Table 2 shows that DFPT set II relaxes PbTiO3 to c/a = 1.150 versus the experimental 1.063, an error of about 8%, and internal displacements are overestimated by roughly 44% (e.g., Delta z_O1 = -0.671 Å versus -0.456 Å, and Delta z_O3 = -0.656 Å versus -0.456 Å). Birefringence in a uniaxial ferroelectric is highly sensitive to the tetragonal distortion, so agreement with the birefringence data of Ref. [43] does not independently validate the electronic-structure treatment: it may reflect compensation between an over-polarized structure and other errors. The claim in the Assessment section that DFPT set II reproduces the experimental crystal geometries and optical anisotropies with good overall accuracy across the four benchmark oxides is directly contradicted by this row. Please add a sensitivity test, for example computing birefringence at the experimental geometry, or substantially weaken the geometry part of the claim for PbTiO3.
  2. [Table 2, BaTiO3 row] The geometry problem is not confined to PbTiO3. For BaTiO3, DFPT set II gives internal oxygen displacements Delta z_O1 = -0.145 Å and Delta z_O3 = -0.090 Å versus experimental values of -0.093 Å and -0.056 Å, i.e., roughly 50-60% too large, even though c/a happens to be close. This pattern suggests that the good BaTiO3 birefringence agreement may also involve compensation between an overestimated polar distortion and other errors. Please quantify the sensitivity of the computed birefringence to the ferroelectric distortion, for instance by repeating the optical calculation at the experimental lattice parameters and internal coordinates.
  3. [Computational Methods, dielectric response] The dielectric tensor is evaluated in the independent-particle approximation with a damping parameter, and Eqs. (12)-(14) derive the refractive index and extinction coefficient from it; no excitonic or local-field corrections are included. The comparisons in Fig. 3 extend into spectral regions near and above the optical gap (for example PbTiO3 below 500 nm), where excitonic and local-field effects can change both the magnitude and the anisotropy of the response. To support the predictive claim, either estimate these corrections for at least one benchmark compound or explicitly restrict the claim to the transparency region where independent-particle response is expected to be reliable.
  4. [Influence of the selected projection manifold] The recommendation of set II {U_TM,V_TM-O} is made after comparing sets I-III against the same experimental birefringence curves shown in Fig. 3. The Hubbard U and V values themselves are non-empirical, but the choice of the orbital manifold is benchmark-informed. This is a form of model selection on the target data and should be stated as a limitation. The predictive claim would be strengthened by demonstrating transferability on a held-out compound or property, or by a physical criterion that selects set II without reference to the optical benchmarks.
  5. [Assessment of optical anisotropy calculations; Conclusions] The paper acknowledges that KNbO3 is the only exception, with ACBN0 agreeing better with experiment because of compensating errors, yet the conclusion still states that DFPT set II reproduces the measured anisotropic optical properties across the benchmark compounds. A claim of consistency across four materials should be weakened to three of four, or the KNbO3 failure should be analyzed more quantitatively, for example by decomposing the birefringence error into contributions from the gap error and the distortion error.
minor comments (5)
  1. [Computational Methods] The code name is misspelled as 'Qunatum ESPRESSO'; it should read 'Quantum ESPRESSO'.
  2. [Computational Methods, dielectric response] The dielectric-response formula in the Computational Methods is not numbered, but Eqs. (12)-(14) refer to its real and imaginary parts and to the refractive index; please number it consistently.
  3. [Birefringent phase-matching landscapes] The phase-matching metrics in Eqs. (10) and (11) use a Gaussian broadening parameter sigma that is not defined in the table caption for Table 4; please state the value and the integration domain explicitly at the point where the metrics are introduced.
  4. [Assessment of optical anisotropy calculations; Figure 3] The text discusses two conflicting experimental birefringence datasets for PbTiO3 (Ref. [37] and Ref. [43]) but does not clearly state which one is plotted in Fig. 3. Please specify the dataset in the figure caption and justify the selection.
  5. [Data availability] The data availability statement only offers data 'on a reasonable request'; for a benchmarking study of this type, depositing structures, Hubbard parameters, and computed dielectric tensors in a public repository would significantly improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Hubbard parameters are computed from DFPT linear response, not fitted to optical data; the optical benchmark is an external test.

full rationale

The central derivation chain is non-circular. The Hubbard U and V parameters are obtained from density-functional perturbation theory as second derivatives of the DFT energy with respect to orbital occupations (Eqs. (1)-(8) and the Computational Methods), with no fitting to refractive-index data. The frequency-dependent dielectric function is then computed from the independent-particle expression (Eq. (11)) using these non-empirical parameters. Thus the benchmark comparison between DFPT set II and measured birefringence is an external test rather than a reduction of the output to the input. The paper's choice of orbital manifold (set II) is informed by comparing the three manifolds to the same experimental data, so the four-compound comparison is a model-selection benchmark rather than a blind prediction for that discrete choice; but this is a methodological caveat, not an equation-level circularity, and the doped solid-solution predictions are out-of-sample. Self-citations (e.g., Refs. [23] and [32]) support standard methodological points and are not load-bearing. The PbTiO3 geometry error (c/a=1.150 vs 1.063) and the choice to credit Ref. [43] over Ref. [37] in the birefringence comparison are correctness/benchmark concerns, not circularity; the core Hubbard-parameter derivation is self-contained and parameter-free with respect to the optical targets.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard DFT/Hubbard machinery plus an independent-particle optical calculation. The Hubbard U and V are not fitted to optical data, which keeps circularity low. However, the choice of manifold is a post-hoc modeling decision, and the optical response relies on the adequacy of Kohn-Sham eigenvalues and relaxed geometries.

free parameters (3)
  • Broadening (damping) parameter in dielectric response = not stated
    Appears in Eq. (11); controls spectral linewidth of the imaginary dielectric function and can affect near-gap dispersion. No numerical value is provided in the manuscript.
  • Orbital manifold for Hubbard corrections = set II = {U_TM, V_TM-O}
    The paper recommends excluding onsite oxygen U after comparing multiple manifolds to experimental birefringence and geometries. This is a categorical modeling choice, selected using the same validation data.
  • Gaussian broadening sigma in phase-matching metrics = 5e-3
    Introduced in Eqs. (10)-(11) to weight the phase-matching landscape; chosen by the authors with no derivation from experiment or from a physical broadening mechanism.
assumptions (4)
  • domain assumption The independent-particle approximation, without excitonic or local-field corrections, is adequate for birefringence and refractive indices of these oxides.
    Eq. (11) computes the dielectric tensor from Kohn-Sham states; no BSE or local-field effects are included, yet the central comparisons are to experimental birefringence.
  • domain assumption DFPT linear-response Hubbard parameters computed from the DFT electronic ground state are transferable to the optical response of the same and substituted compounds.
    Relies on the hpcode and DFPT formalism from Refs. [21,25,32]; there is no independent verification of the optical response beyond the four benchmark oxides.
  • domain assumption PBEsol with PseudoDojo norm-conserving pseudopotentials provides accurate lattice geometries and wavefunctions for these ferroelectrics.
    This is the starting DFT approximation in the Computational Methods section; all Hubbard corrections and optical calculations build on it.
  • domain assumption Kohn-Sham eigenvalue differences in Eq. (11) can be interpreted as optical transition energies without a scissor correction.
    The paper compares computed refractive indices to experiment without adding a self-energy shift, while Table 3 shows that DFT+U+V bandgaps still deviate from GW and experimental values.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optical Anisotropy and Phase Matching in Non-Centrosymmetric Perovskite Oxides from DFT+U and DFT+U+V Functionals." pith.science (2026). https://pith.science/paper/RDC66EP4

@misc{pith2026260806759,
  author       = {Pith},
  title        = {Pith review of: Optical Anisotropy and Phase Matching in Non-Centrosymmetric Perovskite Oxides from DFT+U and DFT+U+V Functionals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RDC66EP4}},
  note         = {Machine review of arXiv:2608.06759}
}
abstract

Optical anisotropy underpins the operation and performance of a broad range of photonic and quantum technologies. In this work, we critically examine the accuracy of density functional theory approximations with onsite and intersite Hubbard corrections (the DFT+$U$ and DFT+$U$+$V$ functionals) in predicting the anisotropic optical response of the non-centrosymmetric perovskite oxides, such as BaTiO$_3$, LiNbO$_3$, KNbO$_3$, and PbTiO$_3$. It is found that correcting self-interaction errors using DFT+$U$ alone does not capture the optoelectronic response of these materials, often leading to a suppression of their optical anisotropy. While intersite Hubbard interactions restore this anisotropy, the choice of the (inter)atomic orbital manifold that defines the Hubbard correction remains critical to its accuracy. The predictive performance of the resulting, systematically validated DFT+$U$+$V$ functional is achieved at a fraction of the computational cost of hybrid functionals and many-body perturbation theory calculations. As benchmarks, we investigate Zn- and (Bi,Mn)-substituted BaTiO$_3$ solid solutions; the latter exhibit polarization-dependent bandgap narrowing from mid-gap states, substantially enhancing the dichroic ratio and birefringence with promising implications for polarization-sensitive photodetectors and integrated photonics.

Figures

Figures reproduced from arXiv: 2608.06759 by the authors.

Figure 1
Figure 1. Optical anisotropy enables a broad range of applications. (a) Dispersion of a uniaxial crystal where [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Refractive index dispersion in non-centrosymmetric oxides ( [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. To enable a direct comparison between the two [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 3
Figure 3. Figure 3: Birefringence dispersion of non-centrosymmetric oxides ( [PITH_FULL_IMAGE:figures/full_fig_p006_3.png]
Figure 4
Figure 4. Figure 4: Birefringent phase-matching landscapes of BaTiO [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Density of states of cation-substituted BaTiO3. The energy scale is referenced to the Fermi level, EF, such that E ≠ EF =0 eV corresponds to the Fermi energy; negative values denote valence states and positive values denote conduction states. (a) Projected density of s…
Figure 6
Figure 6. Figure 6: (a) Birefringence dispersion and (b) SHG cross section of Ba [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Absorption spectra in the extraordinary (a) and ordinary (b) directions, and dichroic ratio (c) of [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

6 extracted references · 6 canonical work pages

  1. [1]

    [1]Eltes, F.et al.An integrated optical modulator operating at cryogenic temperatures.Nat. Mater. 19, 1164–1168 (2020). [2]Warner, H. K.et al.Coherent control of a super- conducting qubit using light.Nature Physics21, 831–838 (2025). [3]Abel, S.et al.Large pockels eect in micro- and nanostructured barium titanate integrated on sili- con.Nat. Mater.18, 42–...

    work page 2020

  2. [2]

    Theoretical Investigation on Electronic Properties of ZnO Crystals Using DFT-Based Cal- culation Method.Adv

    Rusydi, A. Theoretical Investigation on Electronic Properties of ZnO Crystals Using DFT-Based Cal- culation Method.Adv. Mater. Res.1112, 41–44 (2015). 11 [14]Jain, M., Chelikowsky, J. R. & Louie, S. G. Reli- ability of Hybrid Functionals in Predicting Band Gaps.Phys. Rev. Lett.107, 216806 (2011). [15]Garza, A. J. & Scuseria, G. E. Predicting Band Gaps wit...

    work page 2015

  3. [3]

    Schmidt, W. G. Do we know the band gap of Lithium Niobate?Phys. Status Solidi C7, 362– 365 (2010). [18]Schmidt, F.et al.Quasiparticle and excitonic eects in the optical response of KNbO3.Phys. Rev. Mater. 3, 054401 (2019). [19]Pacchioni, P., Varrassi, L. & Franchini, C. Coupling between ferroelectric distortions and excitonic prop- erties in PbTiO3.Appl. ...

    work page 2010

  4. [4]

    & Timrov, I

    Marzari, N. & Timrov, I. Understanding the role of hubbard corrections in the rhombohedral phase of BaTiO3.Phys. Rev. B108, 235171 (2023). [27]Agapito, L. A., Curtarolo, S. & Buon- giorno Nardelli, M. Reformulation of DFT+Uas a Pseudohybrid Hubbard Density Functional for Accelerated Materials Discovery.Phys. Rev. X5, 011006 (2015). [28]Lee, S.-H. & Son, Y...

    work page 2023

  5. [5]

    Birefringence of lead titanate (PbTiO3)

    Miga, S. Birefringence of lead titanate (PbTiO3). J. Appl. Crystallogr.37, 551–554 (2004). [44]Wemple, S. Polarization Fluctuations and the Optical-Absorption Edge in BaTiO3.Phys. Rev. B 2, 2679–2687 (1970). [45]Bhatt, R.et al.Urbach Tail and Bandgap Analysis in Near Stoichiometric LiNbO3Crystals.Phys. Status Solidi A209, 176–180 (2012). [46]Wiesendanger,...

    work page 2004

  6. [6]

    Rep.13, 10487 (2023)

    Electronic, Optical and Mechanical Properties of Ba1≠xCaxTiO3Perovskite from First-Principles In- vestigation.Sci. Rep.13, 10487 (2023). [53]Yabuta, H.et al. Enhancement of Tetragonal Anisotropy and Stabilisation of the Tetragonal Phase by Bi/Mn-Double-Doping in BaTiO3Fer- roelectric Ceramics.Sci. Rep.7, 45842 (2017). [54]Giannozzi, P.et al.Quantum ESPRES...

    work page 2023

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.