Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

The isostructural alpha-gamma phase transition in cerium from the perspective of meta-generalized gradient approximations

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A single DFT calculation reproduces both phases of cerium's volume collapse.

desk verdict A potentially important meta-GGA result for cerium, but the γ phase rests on an unvalidated spin-polarized solution that could be a metastable artifact. read the letter →

arxiv 2506.03578 v1 pith:X7HRXO6Q submitted 2025-06-04 cond-mat.mtrl-sci

Ashesh Giri , Chandra Shahi , Adrienn Ruzsinszky This is my paper
classification cond-mat.mtrl-sci
keywords ceriumalpha-gammaphasetransitionmeta-GGALAKfunctionalultranonlocalityf-electronlocalizationvolumecollapsedensitytheory
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 claims that the LAK meta-GGA — a density functional that uses the orbital kinetic energy density to go beyond standard semilocal approximations — reproduces both the α and γ phases of cerium and their correct energetic ordering in a single spin-polarized calculation. This matters because cerium's isostructural α–γ transition, driven by the localization and delocalization of f electrons, has resisted ordinary semilocal functionals, which see only the α phase. The same calculation yields lattice parameters, a magnetic-moment jump, a critical pressure, and a volume collapse close to experiment, suggesting that a carefully constructed meta-GGA can handle localized f electrons without exact exchange or Hubbard corrections.

What carries the argument

The load-bearing object is the LAK meta-GGA. It is a semilocal functional built from the density, its gradient, and the orbital kinetic energy density τ, entering through the variable α = (τ − τ_W)/τ_unif that separates one-electron regions from the uniform electron gas. LAK adjusts the relative weights of gradient and τ contributions so that its exchange enhancement factor satisfies ∂F_x(s, α)/∂α < 0, the ultranonlocality condition that mimics some effects of exact exchange. The double minimum appears only in the spin-polarized solution: allowing a magnetic moment at large volume localizes the f electrons and stabilizes γ-Ce, so the mechanism is the combination of ultranonlocality with spin polarization.

What would settle it

Repeat the LAK calculation with initial magnetic moments of 0, 1, 2, and 4 μB, and with antiferromagnetic starting configurations. If the γ-phase minimum disappears or the energy ordering reverses in any of these runs, the claim that LAK captures the phase transition from a single self-consistent calculation fails. A second check is to compare the predicted ferromagnetic moment of ~1.2 μB with magnetic-susceptibility or neutron-scattering data that show γ-Ce has no long-range magnetic order.

Watch

Extended reading notes

Core claim

The central discovery is that the LAK meta-GGA produces a double minimum in the total energy as a function of lattice constant for bulk cerium, but only when the calculation is spin-polarized and starts from a magnetic moment of 3.0 μB. The minimum at larger volume is γ-Ce, with a magnetic moment of about 1.2 μB per atom and f electrons localized at atomic sites; the minimum at smaller volume is α-Ce, with zero moment and delocalized f electrons. From that single self-consistent calculation the authors obtain lattice constants of 5.14 Å and 4.87 Å for the γ and α phases, a critical pressure of -0.4 GPa, and a volume collapse of about 16%, all in reasonable agreement with experiment. The projected electron-density difference shows more charge buildup at atomic sites in the γ phase and more interstitial charge in the α phase, consistent with the localization picture.

Load-bearing premise

The γ-phase minimum is obtained from a spin-polarized calculation that assumes an initial magnetic moment of 3.0 μB, and the paper treats this ferromagnetic-like solution as physical even though γ-Ce is experimentally paramagnetic with local moments.

Editorial extensions

If this is right

  • LAK captures both cerium phases at zero temperature without exact exchange, Hubbard U, or random-phase approximation, unlike LDA, PBE, SCAN, r2SCAN, and OFR2.
  • The zero-temperature lattice constants from LAK (4.87 Å for α, 5.14 Å for γ) are close to room-temperature experimental values.
  • The predicted volume collapse of about 16% is much closer to the experimental 14–15% than the roughly 30% collapse from the hybrid-based (EX+cRPA)@PBE0 calculation.
  • The magnetic moment jumps from 0 in α-Ce to about 1.2 μB in γ-Ce, marking the localization of f electrons at the transition.
  • If these results hold, ultranonlocal meta-GGAs could serve as a low-cost alternative to fourth- and fifth-rung methods for f-electron materials.

Reading between the lines

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

  • I infer that the same mechanism may apply to other f-electron metals with volume-collapse transitions, such as plutonium or americium, where a spin-polarized LAK calculation could be tested for a double minimum.
  • The sensitivity to the initial magnetic moment suggests a systematic scan over starting magnetizations is needed to determine whether the γ minimum is the ground state or a metastable ferromagnetic state.
  • A natural extension is to compute phonons and finite-temperature free energies with LAK to verify the experimental finding that phonon entropy does not drive the transition.
  • The ultranonlocality of LAK may also improve charge-transfer and band-gap descriptions in other metals or Mott insulators, not just cerium; this could be tested on simple oxides.
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

3 major / 4 minor

Summary. The manuscript applies the LAK meta-GGA to bulk cerium and reports, from a single self-consistent spin-polarized calculation, a double minimum in the energy-volume curve corresponding to the alpha and gamma phases, with lattice parameters 4.87 Å and 5.14 Å, a magnetic moment of about 1.2 μB, a critical pressure of -0.4 GPa, and a volume collapse of 15.8%. Two other meta-GGAs, OFR2 and r2SCAN, are shown to yield only the alpha phase. The paper argues that LAK's ultranonlocality is responsible for capturing localization/delocalization of f electrons and that this constitutes a parameter-free first-principles description of the isostructural alpha-gamma transition.

Significance. If the central claim holds, the result is significant: LAK would be the first semilocal-level functional without a Hubbard U or exact exchange to produce both cerium phases with the correct ordering and a realistic volume collapse, at a fraction of the cost of hybrid or RPA approaches. The study also provides a useful benchmark comparison of modern meta-GGAs on a correlated f-electron material. The paper's strengths are that it uses an external functional without fitting to cerium data, reports lattice constants close to experiment, and includes a computed volume collapse in good agreement with the measured value. However, the central claim depends entirely on a broken-symmetry spin-polarized solution whose stability and physical interpretation are not established, so the result is promising but not yet fully validated.

major comments (3)
  1. [Figure 1 and the paragraph describing the spin-polarized LAK calculation] The double minimum appears only in the spin-polarized calculation initialized with an initial magnetic moment of 3.0 μB; the spin-unpolarized LAK curve has only the alpha minimum. Because the gamma phase is carried entirely by this particular broken-symmetry magnetic solution, the manuscript must demonstrate that this minimum is a stable stationary point of the LAK functional rather than a metastable artifact. Concretely, the authors should scan initial magnetic moments (for example 0.5, 1.0, 2.0, 3.0, 4.0 μB), test an antiferromagnetic two-atom cell, and test a noncollinear ordering. Without this evidence, the central claim reduces to the weaker statement that LAK has only an alpha phase unless artificially initialized into a ferromagnetic-like state.
  2. [Figure 2 discussion and the magnetic-moment paragraph] The text states that the gamma phase is 'ferromagnetic in nature' and reports a magnetic moment of about 1.2 μB, but the Introduction correctly describes gamma-Ce as paramagnetic with Curie-Weiss susceptibility and localized moments. This contradiction is load-bearing because the physical identification of the calculated broken-symmetry state with the gamma phase depends on the local-moment picture. The paper should either provide evidence that the computed ferromagnetic-like solution represents the paramagnetic local-moment state (for example, by comparing with the expected Ce3+ effective moment or with the reduced moment expected from Kondo screening) or explicitly discuss the discrepancy.
  3. [Table II and the critical-pressure discussion] LAK's predicted critical pressure of -0.4 GPa is compared only with the value -0.8 GPa taken from Ref. [22], which is a theoretical study, while the experimental room-temperature values listed in Table II are +1.8 and +1.44 GPa. The sign of the predicted pressure is opposite to the room-temperature experimental sign, and the manuscript does not provide a temperature or pressure argument justifying the comparison to the extrapolated or theoretical value. This weakens the claim that LAK predicts the transition pressure 'reasonably close' to experiment; the authors should clarify what quantity is being compared and why a negative pressure is consistent with the measured phase transition.
minor comments (4)
  1. [Abstract and text near Figure 1] There are duplicated words in the text, for example 'the the total energy' and 'the the double minimum'; these should be corrected.
  2. [Figure 3 caption and the corresponding paragraph] The projected density difference is defined as n_alpha - n_gamma, and the text says there is more charge localization in the gamma phase at the atomic sites. With a conventional color scale, this would make the atomic sites negative (blue) rather than deep red. The caption and color scale need to be clarified so the sign convention is unambiguous.
  3. [Figure 4 discussion] The density of states figure is described as the sum of spin-up and spin-down states, but the figure does not show the spin-resolved components. Since the magnetic moment and the f-electron localization are central to the argument, showing the spin channels or at least reporting the f-occupancy in each phase would make the comparison more informative.
  4. [Table II and Ref. [22]] The same reference [22] is used for both the lattice parameters and the extrapolated pressure; the authors should explicitly state which entries in the tables are experimental and which are theoretical, since the text sometimes calls the -0.8 GPa value 'extrapolated to 0 K' without making clear that it is a theoretical estimate.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: LAK is an external, parameter-free functional and cerium results are computed outputs, not fitted inputs.

full rationale

This paper applies the LAK meta-GGA, an external functional defined in Lebeda, Aschebrock and Kümmel (PRL 133, 136402, 2024), to bulk cerium. No parameter of LAK is fitted to cerium or to the alpha-gamma transition; the authors explicitly aim for a 'parameter free' model. The central outputs (alpha/gamma lattice constants 4.87/5.14 Angstrom, volume collapse 15.8%, critical pressure -0.4 GPa) are computed from self-consistent E-V curves and compared against experiment; they are not inputs. The Birch-Murnaghan fit is a standard interpolation and does not inject the double minimum. Self-citations (SCAN [8], mTASK [11]) are contextual and not load-bearing: SCAN is invoked precisely because it fails to produce a gamma minimum, and mTASK appears only in the band-gap background discussion. The main weakness - the gamma minimum appears only in spin-polarized runs initialized at 3.0 micro-Bohr magnetons, and the paper calls gamma-Ce 'ferromagnetic in nature' while its own introduction cites Curie-Weiss paramagnetism with local moments - is a robustness and physical-validity concern (correctness risk), not a circularity: no derivation step equates an input with an output, and no fitted parameter is renamed as a prediction. On the scale defined here this is a 1: minor self-citation present but not load-bearing; the core claim stands as a benchmark application of an external functional with independent content.

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

The paper introduces no new functional or parameters. It relies on the published LAK functional, on the neglect of phonon entropy (justified by references to experiment), and on the VASP PAW potential. These are domain assumptions taken from the literature.

assumptions (3)
  • domain assumption The LAK meta-GGA accurately describes exchange-correlation effects in cerium, including f-electron localization.
    The entire paper is a test of LAK; its success is the central claim, and its validity is assumed for Ce.
  • domain assumption Phonon entropy contributions to the alpha-gamma phase transition are negligible.
    The paper cites experiments for this, and uses it to compare 0 K calculations with room-temperature data.
  • domain assumption The VASP PAW potential Ce_3, with 4f electrons as valence states, is accurate for cerium.
    The choice of PAW potential can affect the f-electron behavior; this is not tested.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The isostructural alpha-gamma phase transition in cerium from the perspective of meta-generalized gradient approximations." pith.science (2026). https://pith.science/paper/X7HRXO6Q

@misc{pith2026250603578,
  author       = {Pith},
  title        = {Pith review of: The isostructural alpha-gamma phase transition in cerium from the perspective of meta-generalized gradient approximations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X7HRXO6Q}},
  note         = {Machine review of arXiv:2506.03578}
}
read the original abstract

Meta-generalized gradient approximations (meta-GGAs) on the third rung of the functional hierarchy are gaining increasing relevance for the electronic structure. Meta-GGAs are constructed from numerous ingredients including the orbital kinetic energy density that make them more flexible than generalized gradient approximations (GGAs) including the heavily used PBE-GGA. Still, most meta-GGAs cope with the expected limitations of a semilocal density functional when band gaps or localization of electrons are needed. On the other hand, meta-GGAs are implicit functionals of the orbitals. This feature resembles hybrid density functionals with exact exchange. Efforts in recent years demonstrate that some meta-GGAs can rise beyond the accuracy of semilocal approximation when band gaps are computed. Cerium is an ideal testbed to challenge some recent meta-GGAs. Cerium shows an isostructural alpha - gamma phase transition with delocalized and localized f electrons in each phase, respectively. Since the phonon entropy term was found negligible in the alpha - gamma phase transition of cerium by accurate experiments, all changes in the transition are driven by electronic correlation. The correlation of f electron systems is hardly captured by semilocal approximations but the recent LAK meta-GGA with ultranonlocality steps out of the framework of conventional semilocal density functionals and delivers spectacular accuracy for the phase transition of cerium. LAK and further meta-GGAs inspired by the success of LAK can open a forefront of meta-GGAs for quantum materials with localized electrons.

Figures

Figures reproduced from arXiv: 2506.03578 by the authors.

Figure 1
Figure 1. The LAK total energy and magnetic moment [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Projected density difference nα − nγ projected onto the z = 0 plane. The density is averaged along z−axis at each (x, y) grid point. The color scale spans from -0.01 to +0.01 e/˚A 3 . The atomic positions are in the deep red regions. (a) α phase (b) γ phase [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 2
Figure 2. Variation of magnetic moment with lattice [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Density of States at equilibrium geometry of [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Artificial Symmetry Breaking by Self-Interaction Error

    cond-mat.mtrl-sci 2025-06 conditional novelty 6.0 of 10

    Self-interaction error alone can force semilocal density functionals to break symmetry in one-electron multicenter systems, and a purpose-built functional avoids the artifact.

Reference graph

Works this paper leans on

25 extracted references · 22 canonical work pages · cited by 1 Pith paper

  1. [22]

    Casadei, X

    M. Casadei, X. Ren, P. Rinke, A. Rubio, and M. Schef- fler, Physical Review B 93, 075153 (2016)

    work page 2016

  2. [1]

    D. C. Koskenmaki and K. A. Gschneidner Jr, Handbook on the physics and chemistry of rare earths1, 337 (1978)

    work page 1978

  3. [2]

    Eryigit, C

    S. Eryigit, C. Parlak, and R. Eryigit, Journal of Physics: Condensed Matter 34, 295402 (2022)

    work page 2022

  4. [3]

    Svane, Physical Review Letters 72, 1248 (1994)

    A. Svane, Physical Review Letters 72, 1248 (1994)

    work page 1994

  5. [4]

    Casadei, X

    M. Casadei, X. Ren, P. Rinke, A. Rubio, and M. Schef- fler, Physical Review Letters 109, 146402 (2012)

    work page 2012

  6. [5]

    Szotek, W

    Z. Szotek, W. Temmerman, and H. Winter, Physical Review Letters 72, 1244 (1994)

    work page 1994

  7. [6]

    Kohn and L

    W. Kohn and L. J. Sham, Physical Review 140, A1133 (1965)

    1965

  8. [7]

    J. P. Perdew, K. Burke, and M. Ernzerhof, Physical Review Letters 77, 3865 (1996)

    1996

Show all 25 references
  1. [8]

    J. Sun, A. Ruzsinszky, and J. P. Perdew, Physical Re- view Letters 115, 036402 (2015)

    2015

  2. [9]

    J. W. Furness, A. D. Kaplan, J. Ning, J. P. Perdew, and J. Sun, The Journal of Physical Chemistry Letters 11, 8208 (2020)

    2020

  3. [10]

    Aschebrock and S

    T. Aschebrock and S. K¨ ummel, Physical Review Re- search 1, 033082 (2019)

    2019

  4. [11]

    Neupane, H

    B. Neupane, H. Tang, N. K. Nepal, S. Adhikari, and A. Ruzsinszky, Physical Review Materials 5, 063803 (2021)

    2021

  5. [12]

    Lebeda, T

    T. Lebeda, T. Aschebrock, and S. K¨ ummel, Physical Review Letters 133, 136402 (2024)

    2024

  6. [13]

    A. D. Kaplan and J. P. Perdew, Physical Review Mate- rials 6, 083803 (2022)

    2022

  7. [14]

    Mejia-Rodriguez and S

    D. Mejia-Rodriguez and S. Trickey, Physical Review B 98, 115161 (2018)

    2018

  8. [15]

    Mej ´ ıa-Rodr ´ ıguez and S

    D. Mej ´ ıa-Rodr ´ ıguez and S. Trickey, Physical Review B 102, 121109 (2020)

    2020

  9. [16]

    Mott, Reviews of Modern Physics 40, 677 (1968)

    N. Mott, Reviews of Modern Physics 40, 677 (1968)

    1968

  10. [17]

    Johansson, Philosophical Magazine 30, 469 (1974)

    B. Johansson, Philosophical Magazine 30, 469 (1974)

    1974

  11. [18]

    Allen and R

    J. Allen and R. M. Martin, Physical Review Letters 49, 1106 (1982)

    1982

  12. [19]

    Chakrabarti, M

    B. Chakrabarti, M. Pezzoli, G. Sordi, K. Haule, and G. Kotliar, Physical Review B 89, 125113 (2014)

    2014

  13. [20]

    F. D. Murnaghan, Proceedings of the National Academy of Sciences 30, 244 (1944)

    1944

  14. [21]

    Kresse and J

    G. Kresse and J. Furthm¨ uller, Physical Review B 54, 11169 (1996)

    1996

  15. [23]

    Olsen, L

    J. Olsen, L. Gerward, U. Benedict, and J.-P. Iti´ e, Physica B+C 133, 129 (1985)

    1985

  16. [24]

    J. M. Lawrence, P. S. Riseborough, and R. D. Parks, Reports on Progress in Physics 44, 1 (1981)

    1981

  17. [25]

    Decremps, L

    F. Decremps, L. Belhadi, D. Farber, K. Moore, F. Oc- celli, M. Gauthier, A. Polian, D. Antonangeli, C. Aracne- Ruddle, and B. Amadon, Physical Review Letters 106, 065701 (2011)

    2011

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.