Pith. sign in

REVIEW 3 major objections 4 minor 20 references

Beyond-Hubbard pairing in a cuprate ladder

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read In the cuprate ladder Sr14Cu24O41, the magnetic signal from doped holes is far weaker than the single-band Hubbard model predicts, and the paper traces the suppression to a strong nearest-neighbor attraction that enhances d-wave-like hole…

desk verdict A clear new RIXS observation of a suppressed quasiparticle spin-flip branch in a cuprate ladder, with a plausible but quantitatively conditional attribution to an attractive nearest-neighbor V. read the letter →

arxiv 2501.10287 v1 pith:RN6GZIFQ submitted 2025-01-17 cond-mat.str-el cond-mat.mtrl-scicond-mat.supr-con

classification cond-mat.str-elcond-mat.mtrl-scicond-mat.supr-con
keywords cuprateladderHubbardmodelnearest-neighborattractionresonantinelasticx-rayscatteringdynamicalspinstructurefactord-wavepairingdensitymatrixrenormalizationgroupSr14Cu24O41
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 uses high-resolution resonant inelastic x-ray scattering on the self-doped cuprate ladder Sr14Cu24O41 to isolate a low-energy branch of magnetic excitations coming from the doped holes. That branch is nearly absent at low temperature and only faintly present above the charge-order transition, in sharp contrast to the intense quasiparticle spin-flip branch predicted by density-matrix renormalization group calculations for a single-band Hubbard ladder. The authors argue the discrepancy is not disorder or charge order but an additional nearest-neighbor attractive interaction $V$ of order $-1.0t$ to $-1.25t$, which binds holes on the same rung and suppresses hole spin-flip scattering. In their model this attraction raises the hole-pair binding energy by almost an order of magnitude and preserves d-wave-like pairing correlations. If correct, the result identifies a missing ingredient beyond the Hubbard model that may be needed for robust d-wave superconductivity in the cuprates.

What carries the argument

The central object is the extended Hubbard model on a two-leg ladder, with leg hopping $t$, rung hopping $t_\perp=0.84t$, diagonal hopping $t'=-0.3t$, on-site repulsion $U=8t$, and a nearest-neighbor interaction $V$; the paper compares its dynamical spin structure factor $S(q,\omega)$, computed with the Krylov-space correction-vector DMRG method, against RIXS spectra normalized to a single-ion spin-flip cross section. The key mechanism is the response of $V$: a repulsive $V$ barely changes the spectra, while an attractive $V$ makes it energetically favorable for doped holes to occupy the same rung, restoring intact rung singlets that can be excited into the two-triplon continuum and suppressing the quasiparticle spin-flip branch. The pairing strength is quantified by the hole-pair binding energy $2E_{GS}(N-1)-E_{GS}(N)-E_{GS}(N-2)$, and the pairing symmetry by the relative sign of rung-rung and rung-leg singlet pair correlations.

What would settle it

Measure the quasiparticle spin-flip branch in a doped two-leg ladder where the nearest-neighbor interaction is known to be small, or compute the 260 K spectra with finite-temperature methods that include charge-order fluctuations; if the branch then appears with full Hubbard-model intensity, the inferred attractive $V$ is not required to explain the suppression.

Watch

Extended reading notes

Core claim

The central claim is that the magnetic response of the doped holes in Sr14Cu24O41, observed as a quasiparticle spin-flip branch in the dynamical spin structure factor $S(q,\omega)$, is strongly suppressed relative to the single-band Hubbard model, and that the suppression is a direct spectroscopic signature of enhanced hole pairing caused by a large nearest-neighbor attractive interaction. Using model parameters fixed by the undoped two-triplon dispersion, the paper shows the Hubbard model produces an intense spin-flip branch that is absent in the measured spectra. Introducing $V \approx -1.0t$ to $-1.25t$ in the extended Hubbard ladder reproduces the data: holes bind into rung pairs, the pair binding energy grows to about $0.074t$, and the quasiparticle spin-flip intensity decreases monotonically while the two-triplon continuum sharpens. The pair correlations retain a d-wave-like structure, with rung and leg pair correlations of opposite sign, and the measured orbital dynamics place the ladder in a crossover regime between one and two dimensions, supporting the extension of the conclusion to two-dimensional cuprates.

Load-bearing premise

The inference relies on the conversion of raw RIXS intensity into an absolute $S(q,\omega)$ using a single-ion spin-flip cross-section, and on the assumption that zero-temperature DMRG on an isolated 64x2 ladder represents the 260 K thermal state without additional broadening or charge-order precursors.

Editorial extensions

If this is right

  • The single-band Hubbard model, with only on-site repulsion, does not fully account for the magnetic excitations of doped cuprate ladders; an attractive nearest-neighbor interaction is required to match the measured spectra.
  • An attractive $V$ of about $-1.0t$ to $-1.25t$ increases the hole-pair binding energy by nearly an order of magnitude and leaves the pairing symmetry d-wave-like.
  • Because the suppression is observed above the charge-order transition, the enhanced pairing is intrinsic to the doped ladder and not a byproduct of the charge-order gap.
  • The combination of a gapped triplon continuum and a confined orbiton places this ladder in a dimensional crossover regime, supporting the relevance of the result to two-dimensional cuprates.
  • Similar values of $V$ inferred in one-dimensional chain cuprates suggest the attractive interaction is shared across cuprate families and may be a common ingredient for robust superconductivity.

Reading between the lines

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

  • If the claim holds, high-resolution RIXS on lightly doped two-dimensional cuprates should reveal a similarly suppressed quasiparticle spin-flip weight relative to Hubbard-model predictions, providing a direct test in the 2D case.
  • The inferred interaction is consistent with an electron-phonon origin; a sharper test would be an isotope-exchange experiment that shifts phonon energies and checks whether the quasiparticle spin-flip intensity recovers.
  • The paper's ladder geometry has no local $C_4$ symmetry, yet the pairing retains d-wave-like sign structure; an interesting extension is whether the same robustness persists in models with explicit symmetry-breaking disorder or anisotropic $V$.
  • Because the suppression sets in monotonically with $|V|$, a quantitative map of the quasiparticle spin-flip intensity across different ladder compounds could serve as a spectroscopic thermometer for the effective intersite attraction.
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 paper reports Cu L-edge RIXS measurements of the spin-ladder compound Sr14Cu24O41 and compares the extracted magnetic dynamical structure factor with DMRG calculations for a two-leg Hubbard ladder and an extended Hubbard ladder with nearest-neighbor interaction V. The central experimental observation is a weak, dispersive quasiparticle spin-flip branch above the charge-order temperature, whose intensity is strongly suppressed relative to the single-band Hubbard model. The authors show that introducing an attractive nearest-neighbor Coulomb interaction V in the range -1.0t to -1.25t, with U-V held fixed to preserve the exchange couplings, reproduces the suppression of this branch, and they further compute that this V enhances the hole-pair binding energy by almost an order of magnitude while retaining a d-wave-like pairing symmetry. The paper also presents orbiton dispersion data suggesting a crossover between one- and two-dimensional physics. The authors conclude that the Hubbard model is insufficient for cuprate ladders and that an attractive intersite interaction may be a universal ingredient for d-wave pairing in cuprates.

Significance. If the central inference is correct, the paper provides a rare experimental signature of beyond-Hubbard physics in a cuprate material and quantifies a large attractive nearest-neighbor interaction that could resolve the ongoing debate about whether the pure Hubbard model supports robust superconductivity. The experimental data are of high quality: the RIXS measurements have 35 meV resolution, the two-triplon dispersion is well characterized, the doping level is verified by XAS, and the charge-order transition is independently determined. The DMRG calculations are state-of-the-art, with clear documentation of the model Hamiltonian, parameter extraction via Bayesian optimization, and the dynamical spin structure factor computed on 64×2 clusters. The paper also explicitly discusses boundary effects and pair-pair correlations, which is commendable. However, as detailed below, the quantitative claims about V and the order-of-magnitude binding-energy enhancement rest on a single-model comparison and a normalization procedure whose validity for the quasiparticle spin-flip branch is not established.

major comments (3)
  1. [SM Section 2 (Extracting S(q,ω)) and Fig. 2] The conversion of raw RIXS intensity into S(q,ω) uses a single-ion spin-flip cross-section Rspin(ε,ε′,Ω_i) to normalize all magnetic features, including the quasiparticle spin-flip branch. The central inference of an attractive V depends on this branch being intrinsically weak after normalization. However, the RIXS matrix element for a quasiparticle spin flip in a doped ladder may differ substantially from that of an isolated Cu2+ ion: the local electronic configuration around the hole is strongly modified by doping, and the cited reference [12] primarily validates the single-ion mapping for the two-triplon continuum, not for the quasiparticle spin-flip channel. If the intrinsic cross-section for this branch is suppressed by the same many-body effects that the authors attribute to V, then dividing by the single-ion Rspin would artificially lower the extracted S(q,ω), mimicking the effect of V<0. This is a load-bearing assumption. The authors should either compute the RIXS cross-section for the doped ladder within their DMRG framework (e.g., along the lines of Ref. [30]) to verify that the matrix element for the quasiparticle spin-flip branch is unchanged from the single-ion value, or provide an independent calibration that rules out an energy- or momentum-dependent matrix-element artifact.
  2. [Fig. 3 and Appendix (DMRG calculations, Binding energy)] The quantitative conclusion that V is -1.0t to -1.25t is obtained by tuning V to reproduce the suppression of the quasiparticle spin-flip intensity in the same data set. The subsequent calculation that this V enhances the hole-pair binding energy by 'almost an order of magnitude' is then a consequence of the fitted V, not an independent prediction or test of enhanced pairing. This is a circularity in the presentation: the binding-energy enhancement is an output of the model whose key parameter was already fit to the magnetic response. The authors should state this explicitly, provide a goodness-of-fit metric or uncertainty estimate for the inferred V range, and avoid presenting the binding-energy enhancement as independent corroboration. If possible, they should identify a distinct, a priori prediction (for example, the momentum dependence of the suppression or the behavior under additional doping) that could be tested against new data.
  3. [Comparison of zero-temperature DMRG to 260 K data (Figs. 2 and 3)] The key experimental spectra were measured at 260 K, above TCO, while the DMRG calculations are zero-temperature. Thermal broadening, thermally excited quasiparticles, and fluctuation effects can independently reduce the intensity of the quasiparticle spin-flip branch relative to the T=0 Hubbard-model calculation, even for V=0. The paper does not assess finite-temperature effects, and the 40 K data are obscured by the charge-order gap, so there is no direct low-temperature check of the Hubbard-model prediction. Without a quantitative estimate of thermal effects on the quasiparticle spin-flip intensity, the comparison is not fully controlled. The authors should either estimate the temperature dependence (e.g., using finite-temperature DMRG or a simplified phenomenological model) or restrict the quantitative claim to a statement of qualitative suppression, deferring the V inference to a lower-temperature measurement where charge order does not interfere.
minor comments (4)
  1. [Appendix Eq. (5)] The definition of binding energy B.E. = 2E_GS(N-1) - E_GS(N) - E_GS(N-2) is standard, but for a 64×2 cluster with open boundary conditions the finite-size corrections to this quantity can be as large as the binding energy itself. Please provide an estimate of finite-size errors (e.g., by repeating the calculation for different cluster lengths) or a discussion of why they are negligible.
  2. [Appendix, model parameters] The choice to hold U-V fixed at 8t when varying V is an important modeling assumption. The text states that this keeps J constant, but the strong-coupling expression J = -4t^2/(U-V) is only exact in the strong-coupling limit and for a specific charge-transfer process. Please justify this choice more carefully and discuss whether alternative ways of introducing V (e.g., fixing U and varying V) would change the conclusions.
  3. [SM Section 2 and Fig. S7] In the description of the normalization procedure, the paper states that the orbital excitation integrated intensity is matched to the single-ion ED calculation. The figure shows the θ-dependent scaling factors, but the text does not describe how the normalization is propagated to the magnetic excitations or what uncertainty is introduced by the elastic and phonon subtraction. A short error-propagation analysis, or at least a statement of the typical scale of the normalization correction, would make the quantitative comparison more transparent.
  4. [Fig. 4 and main text, final paragraph] The orbiton dispersion and dimensional-crossover discussion in Fig. 4 is interesting but somewhat disconnected from the central pairing claim. The connection to the suppression of quasiparticle spin flips is not made explicit, and the statement that the results 'may be directly relevant' to two-dimensional cuprates is speculative. Consider either providing a more direct link (e.g., through the effect of dimensionality on V) or clearly labeling this as a separate observation.

Circularity Check

0 steps flagged · score 0.0 of 10

Derivation is self-contained: the V=0 Hubbard prediction fails independently, and the attractive V is explicitly fitted to the residual suppression rather than presented as a predicted observable.

full rationale

The paper's central chain is: (1) measure raw RIXS and subtract elastic and phonon backgrounds; (2) convert to S(q,omega) using a single-ion spin-flip cross-section, with doped-system validity supported by an external published RIXS calculation [SM Ref. 12]; (3) fit undoped Hubbard parameters (t, t_perp, t') to the 40 K two-triplon dispersion; (4) use those parameters to compute the doped two-leg Hubbard S(q,omega) at V=0, obtaining an intense quasiparticle spin-flip branch; (5) observe that the measured branch is faint, so the V=0 prediction fails; (6) introduce V and fit it to the suppression, finding V approx -1.0t to -1.25t; (7) compute the hole-pair binding energy from the same Hamiltonian (Appendix Eq. 5) and report the enhancement. No step equates an output to an input by definition. The V=0 comparison is a genuine prediction with independently fitted parameters; the failure of that prediction is the non-circular core of the beyond-Hubbard claim. The binding-energy enhancement is explicitly a DMRG consequence of the fitted V, not an independent confirmation, so it is a model interpretation rather than a predicted observable. The RIXS-to-S normalization is an assumption whose violation would be a systematic error, not a circularity. Self-citations (e.g., Refs. 12, 30, 39 in the SM) are published external calculations and are not used to define the central result; the central comparison uses the paper's own DMRG calculations.

Assumptions & free parameters 6 free parameters · 6 assumptions · 1 invented entities

The central claim rests on a small set of fitted hopping and interaction parameters. The hopping parameters are anchored to the measured two-triplon dispersion, which is good. The decisive parameter V is fitted to the same spectral suppression that motivates it, so the quantitative pairing-enhancement claim inherits a circularity burden. The RIXS-to-S(q,omega) normalization and the zero-temperature DMRG comparison are additional domain assumptions that the reader must accept.

free parameters (6)
  • leg hopping t = 0.38 eV
    Used as the global energy scale. The appendix says 'Setting t=0.38 eV and U=8t, we tune for t_perp and t_prime', so this value is adopted or fixed rather than independently measured here; it yields J consistent with neutron scattering.
  • rung hopping t_perp = 0.84 t
    Fitted to the experimental two-triplon dispersion at 40 K using Bayesian optimization.
  • diagonal hopping t_prime = -0.3 t
    Fitted together with t_perp to the two-triplon dispersion at 40 K.
  • on-site Hubbard repulsion U = 8 t
    Chosen so that superexchange J = 4 t^2 / U matches reported values from neutron scattering; effectively selected rather than derived from the RIXS data alone.
  • nearest-neighbor interaction V = -1.0 t to -1.25 t
    Fitted to reproduce the suppressed quasiparticle spin-flip intensity in the doped ladder spectra. This parameter drives the central claim of enhanced hole pairing.
  • single-ion crystal-field parameters Dq, Ds, Dt = Dq=0.164 eV, Ds=0.42 eV, Dt=0.19 eV
    Set to match experimental orbital excitation energies for the RIXS normalization procedure in SM Section 2; they affect the conversion of raw intensity into S(q,omega).
assumptions (6)
  • domain assumption The extended Hubbard Hamiltonian in Eq. (1) with on-site U and nearest-neighbor V captures the relevant magnetic excitations of Sr14Cu24O41.
    The entire comparison between RIXS data and DMRG assumes this two-leg ladder model is the correct minimal description of the material.
  • domain assumption RIXS intensity after subtracting backgrounds and normalizing by the single-ion spin-flip cross-section is proportional to the dynamical spin structure factor S(q,omega), including for the doped ladder.
    SM Section 2 invokes this mapping with a caveat from prior work; the weak quasiparticle branch amplitude depends on this normalization.
  • domain assumption One- and two-triplon sectors do not mix, and measurements at H=0 access only even-triplon-number excitations.
    This symmetry argument is used in the Appendix to isolate the two-triplon continuum and the quasiparticle spin-flip branch.
  • domain assumption DMRG on 64x2 ladders with up to 2000 kept states gives converged S(q,omega) for the parameters used.
    No explicit convergence checks against bond dimension or system size are shown for the dynamical structure factor.
  • ad hoc to paper When V is introduced, U - V is held fixed at 8t to keep the superexchange J constant.
    This is a modeling choice stated in the Appendix. It preserves the undoped magnetic energy scale but is not derived from data and affects how strongly V suppresses the quasiparticle branch.
  • domain assumption Zero-temperature DMRG ground-state spectra can be compared with RIXS data measured at 260 K, above the charge-order transition.
    The central comparison in Fig. 3 uses 260 K data against T=0 simulations; thermal population and charge-order precursor effects are not included in the model.
invented entities (1)
  • Attractive nearest-neighbor Coulomb interaction V < 0 in the extended Hubbard ladder independent evidence
    purpose: Suppresses the quasiparticle spin-flip spectral weight and binds doped holes into rung pairs, producing the enhanced d-wave-like pairing claimed in the paper.
    V is not directly measured in this work but is inferred from the RIXS suppression. Independent support comes from ARPES on 1D cuprate chains reporting V near -0.8t to -1.2t [40] and from prior theoretical studies [14-17,46], so the interaction is not invented from nothing.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Beyond-Hubbard pairing in a cuprate ladder." pith.science (2026). https://pith.science/paper/RN6GZIFQ

@misc{pith2026250110287,
  author       = {Pith},
  title        = {Pith review of: Beyond-Hubbard pairing in a cuprate ladder},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RN6GZIFQ}},
  note         = {Machine review of arXiv:2501.10287}
}
abstract

The Hubbard model is believed to capture the essential physics of cuprate superconductors. However, recent theoretical studies suggest that it fails to reproduce a robust and homogeneous superconducting ground state. Here, using resonant inelastic x-ray scattering and density matrix renormalization group calculations, we show that magnetic excitations in the prototypical cuprate ladder Sr$_{14}$Cu$_{24}$O$_{41}$ are inconsistent with those of a simple Hubbard model. The magnetic response of hole carriers, contributing to an emergent branch of spin excitations, is strongly suppressed. This effect is the consequence of d-wave-like pairing, enhanced by nearly an order of magnitude through a large nearest-neighbor attractive interaction. The similarity between cuprate ladders and the two-dimensional compounds suggests that such an enhanced hole pairing may be a universal feature of superconducting cuprates.

Figures

Figures reproduced from arXiv: 2501.10287 by the authors.

Figure 1
Figure 1. FIG. 1. Magnetic excitations of the cuprate ladder Sr [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Magnetic excitations from the doped holes. (a) RIXS spectra along the leg direction for momenta spanning [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Signatures of enhanced hole pairing due to an attractive nearest-neighbor interaction. (a-b), Theoretical [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Orbiton dispersion and dimensional crossover (a), RIXS spectrum (circles) and fit (solid line) of the [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages

  1. [12]

    Jia, C. et al. Persistent spin excitations in doped antiferromagnets revealed by resonant inelastic light scattering. Nature Communications 5, 3314 (2014)

  2. [1]

    Jang, H. et al. Time-resolved resonant elastic soft x-ray scattering at Pohang Accelerator Laboratory X-ray Free Electron Laser. Review of Scientific Instruments 91, 083904 (2020)

  3. [2]

    Larch: an analysis package for xafs and related spectroscopies, V ol

    Newville, M. Larch: an analysis package for xafs and related spectroscopies, V ol. 430, 012007 (IOP Publishing, 2013)

  4. [3]

    FLUO: Correcting XANES for self-absorption in fluorescence measurements (1999)

    Haskel, D. FLUO: Correcting XANES for self-absorption in fluorescence measurements (1999)

  5. [4]

    & Uchida, S

    Osafune, T., Motoyama, N., Eisaki, H. & Uchida, S. Optical Study of the Sr14−xCaxCu24O41 System: Evidence for Hole-Doped Cu2O3 Ladders. Physical Review Letters 78, 1980 (1997)

  6. [5]

    N ¨ucker, N. et al. Hole distribution in (Sr, Ca, Y , La)14Cu24O41 ladder compounds studied by x-ray absorption spectroscopy. Physical Review B 62, 14384 (2000)

  7. [6]

    Abbamonte, P. et al. Crystallization of charge holes in the spin ladder of Sr14Cu24O41. Nature 431, 1078–1081 (2004)

  8. [7]

    & Rice, T

    Troyer, M., Tsunetsugu, H. & Rice, T. Properties of lightly doped t-J two-leg ladders. Physical Review B 53, 251 (1996)

Show all 20 references
  1. [8]

    & Johnston, S

    Kumar, U., Nocera, A., Dagotto, E. & Johnston, S. Theoretical study of the spin and charge dynamics of two-leg ladders as probed by resonant inelastic x-ray scattering. Physical Review B 99, 205130 (2019)

  2. [9]

    Wang, Y ., Fabbris, G., Dean, M. P. M. & Kotliar, G. EDRIXS: An open source toolkit for simulating spectra of resonant inelastic x-ray scattering. Computer Physics Communications 243, 151–165 (2019)

  3. [10]

    J., Ghiringhelli, G., Sala, M

    Ament, L. J., Ghiringhelli, G., Sala, M. M., Braicovich, L. & van den Brink, J. Theoretical demonstration of how the dispersion of magnetic excitations in cuprate compounds can be determined using resonant inelastic X-ray scattering. Physical Review Letters 103, 117003 (2009)

  4. [11]

    Robarts, H. C. et al. Dynamical spin susceptibility in La2CuO4 studied by resonant inelastic x-ray scattering. Physical Review B 103, 224427 (2021)

  5. [13]

    Role of oxygen states in the low valence nickelate La 4 Ni 3 O 8

    Shen, Y .et al. Role of oxygen states in the low valence nickelate La 4 Ni 3 O 8. Physical Review X 12, 011055 (2022)

  6. [14]

    Schlappa, J. et al. Spin–orbital separation in the quasi-one-dimensional Mott insulator Sr 2CuO3. Nature 485, 82–85 (2012)

  7. [15]

    Wohlfeld, K., Nishimoto, S., Haverkort, M. W. & van den Brink, J. Microscopic origin of spin-orbital separation in Sr2CuO3. Physical Review B 88, 195138 (2013)

  8. [16]

    Propagation of an orbiton in the antiferromagnets: theory and experimental verification

    Wohlfeld, K. Propagation of an orbiton in the antiferromagnets: theory and experimental verification. arXiv preprint arXiv:1912.11363 (2019)

  9. [17]

    Kugel, K. I. & Khomski ˘ı, D. The Jahn-Teller effect and magnetism: transition metal compounds. Soviet Physics Uspekhi 25, 231 (1982). 17

  10. [18]

    & Chang, J

    Zhou, Z., Ye, W., Luo, H.-G., Zhao, J. & Chang, J. Robust superconducting correlation against inter-site interactions in the extended two-leg Hubbard ladder. arXiv preprint arXiv:2303.14723 (2023)

  11. [19]

    Scheie, A. et al. Cooper-pair localization in the magnetic dynamics of a cuprate ladder. To be published. (2024)

  12. [20]

    Nocera, A. et al. Doping evolution of charge and spin excitations in two-leg Hubbard ladders: Comparing DMRG and FLEX results. Physical Review B 97, 195156 (2018). 18

Pith tools

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