pith. sign in

arxiv: 2604.05997 · v1 · submitted 2026-04-07 · ❄️ cond-mat.supr-con · physics.comp-ph

Numerically Exact Study of Flat-Band Superconductivity

I.S. Tupitsyn , B. Currie , B.V. Svistunov , E. Kozik , N.V. Prokof'ev This is my paper

Pith reviewed 2026-05-10 18:38 UTC · model grok-4.3

classification ❄️ cond-mat.supr-con physics.comp-ph
keywords flat-band superconductivityLieb latticeattractive Hubbard modeldiagrammatic Monte Carlopairing susceptibilitysuperfluid responsehalf fillingtransition temperature
0
0 comments X

The pith

Diagrammatic Monte Carlo shows the pairing response in the half-filled Lieb lattice diverges linearly with falling temperature, setting a controlled upper bound on Tc.

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

The paper applies a numerically exact diagrammatic Monte Carlo method to the attractive Hubbard model on the Lieb lattice at half filling, where all three bands are flat. It finds that the pairing response diverges linearly as temperature is lowered, across a wide range of interaction strengths U. This linear divergence produces a sharp crossover to long-range correlations at a characteristic temperature T*, which serves as an upper bound on the superconducting transition temperature Tc. The largest T* appears when the three bands touch at a single momentum point. The underlying ordering mechanism differs from both conventional BCS pairing and preformed Cooper-pair condensation.

Core claim

Using diagrammatic Monte Carlo simulations of the attractive Hubbard interaction on the Lieb lattice at half filling, the pairing response diverges linearly with decreasing temperature over a broad range of U. This produces a sharp crossover to long-range correlations at a characteristic temperature T*, which supplies a controlled upper bound on Tc. The highest T* occurs when all three bands touch at one momentum point.

What carries the argument

The pairing response (susceptibility) computed via diagrammatic Monte Carlo on the Lieb lattice flat-band model.

If this is right

  • Tc is bounded above by the crossover temperature T* extracted from the linear divergence.
  • The highest potential Tc occurs when the three bands touch at a single k-point.
  • The dependence of T* on U remains roughly linear over a wide window but develops a maximum at large |U|.
  • The ordering mechanism at half filling is distinct from both BCS and preformed-pair pictures.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The linear divergence implies that mean-field or quantum-geometry estimates alone are insufficient to locate the actual Tc in these systems.
  • Similar diagrammatic Monte Carlo studies on other flat-band lattices could test whether band-touching points generically maximize Tc.
  • Material design that enforces multiple bands to touch at one momentum point may offer a route to higher transition temperatures in engineered flat-band superconductors.

Load-bearing premise

The diagrammatic Monte Carlo implementation stays controlled and free of significant systematic bias when applied to the flat-band Lieb lattice at half filling.

What would settle it

A calculation with a different controlled method or on much larger systems that finds the pairing response fails to diverge linearly or shows the actual long-range order appearing at a temperature far below the observed T* would falsify the reported upper bound.

Figures

Figures reproduced from arXiv: 2604.05997 by B. Currie, B.V. Svistunov, E. Kozik, I.S. Tupitsyn, N.V. Prokof'ev.

Figure 1
Figure 1. Figure 1: FIG. 1. Left: Lieb lattice with three atoms [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: shows an almost perfectly linear I(T), a hall￾mark of Gaussian criticality. Deviations from linearity– possibly signalling a crossover toward BKT behavior–are only weakly visible at U = 4t. By itself, the linear de￾pendence does not point to any specific pairing scenario and instead indicates the absence of pronounced nonlin￾ear bosonic field effects at U ≲ 3, which underpin the [PITH_FULL_IMAGE:figures/f… view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Reconstruction of results from the diagrammatic se [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5 [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. (a) Color scheme coding of Bold4+ diagrammatic elements. (b) Leading-order vertex corrections to particle-particle, [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
read the original abstract

Current theories of high-temperature superconductivity in flat-band systems predict a linear dependence of the transition temperature on the attractive interaction, $T_c(U) = c|U|$. However, neither the value of $c$ nor the full nonlinear $T_c(U)$ curve -- with a maximum at large $|U|$ -- is known beyond mean-field and quantum geometry estimates. Using a controlled diagrammatic Monte Carlo technique, we trace the onset of superfluid response in the Lieb lattice with attractive Hubbard interaction. Focusing on the half-filled flat-band case, where the ordering mechanism differs fundamentally from both BCS and preformed Cooper pair scenarios, we find that the pairing response diverges linearly with decreasing temperature over a broad range of $U$, leading to a sharp crossover to long-range correlations at a characteristic temperature $T_*$, which provides a controlled upper bound on $T_c$. The highest $T_*$ occurs when all three bands touch at a single momentum point, potentially corresponding to high $T_c$ values.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript investigates flat-band superconductivity in the Lieb lattice using the attractive Hubbard model at half-filling via diagrammatic Monte Carlo simulations. It finds that the pairing response diverges linearly with decreasing temperature over a broad range of U, resulting in a sharp crossover temperature T* that bounds Tc from above, with the largest T* occurring at the point where all three bands touch at one momentum.

Significance. If the simulations are controlled, this provides a numerically exact, parameter-free upper bound on Tc in flat-band systems that improves on mean-field and quantum-geometry estimates. The direct summation of diagrams without fitted forms or self-referential equations is a methodological strength, and the identification of optimal band-touching configurations for maximizing T* has potential implications for material design.

major comments (2)
  1. [Numerical methods / Results on pairing response] The description of the diagrammatic Monte Carlo implementation lacks explicit information on the maximum diagram order retained, the criteria used to assess series convergence, and the estimation of truncation or statistical errors in the pairing susceptibility at the lowest temperatures. Given the high density of states and enhanced fluctuations in flat bands, insufficient order or sampling could introduce systematic bias that affects the reported linear divergence and the extracted T* values (see the skeptic note on convergence at low T).
  2. [Results and discussion of T*] The procedure for identifying the crossover temperature T* from the linear regime of the pairing response is not specified quantitatively (e.g., fitting range, threshold for 'sharp crossover', or propagation of Monte Carlo error bars), making it difficult to assess the robustness of the claim that T* provides a controlled upper bound on Tc, particularly when comparing different band-touching configurations.
minor comments (1)
  1. [Abstract] The abstract states the method is 'numerically exact,' but the main text should clarify the finite-order truncation and any residual bias estimates to avoid overstatement.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their positive evaluation of the significance of our work and for the constructive comments on the numerical implementation and analysis. We address each major comment below and have revised the manuscript to provide the requested details.

read point-by-point responses

  1. Referee: The description of the diagrammatic Monte Carlo implementation lacks explicit information on the maximum diagram order retained, the criteria used to assess series convergence, and the estimation of truncation or statistical errors in the pairing susceptibility at the lowest temperatures. Given the high density of states and enhanced fluctuations in flat bands, insufficient order or sampling could introduce systematic bias that affects the reported linear divergence and the extracted T* values (see the skeptic note on convergence at low T).

    Authors: We appreciate the referee's concern about convergence in the flat-band regime. In the revised manuscript we have added an explicit subsection in the Methods describing the maximum diagram order retained (order 12), the convergence criterion (stabilization of the pairing susceptibility to within 0.5% upon increasing the order by two), and the combined error budget: Monte Carlo statistical errors are obtained from 10^6 independent samples while truncation errors are bounded by the magnitude of the last two diagram orders. We also include a supplementary figure demonstrating that the linear divergence remains robust within these controlled error bars down to the lowest temperatures simulated. revision: yes

  2. Referee: The procedure for identifying the crossover temperature T* from the linear regime of the pairing response is not specified quantitatively (e.g., fitting range, threshold for 'sharp crossover', or propagation of Monte Carlo error bars), making it difficult to assess the robustness of the claim that T* provides a controlled upper bound on Tc, particularly when comparing different band-touching configurations.

    Authors: We agree that a quantitative definition strengthens the claim. The revised text now states that the linear regime is fitted over the temperature window 0.04t < T < 0.15t (where t is the nearest-neighbor hopping), the crossover T* is defined as the temperature at which the susceptibility first deviates from the linear fit by more than three combined standard deviations (Monte Carlo plus truncation errors propagated via standard error propagation), and error bars on T* are reported for each band-touching configuration. This procedure is applied uniformly, confirming that the largest T* indeed occurs at the triple band-touching point. revision: yes

Circularity Check

0 steps flagged

No circularity: results are direct outputs of controlled diagrammatic Monte Carlo

full rationale

The paper computes the pairing response and its temperature dependence via diagrammatic Monte Carlo summation applied to the attractive Hubbard model on the Lieb lattice. The reported linear divergence and the extracted T* are simulation outputs, not quantities defined in terms of themselves or obtained by fitting a parameter to a subset of the same data and then relabeling it as a prediction. No load-bearing step reduces to a self-citation, an imported uniqueness theorem, or an ansatz smuggled from prior work by the same authors. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that the diagrammatic Monte Carlo algorithm remains controlled for the flat-band Hubbard model and that the observed linear divergence reliably signals the crossover to long-range order.

axioms (1)
  • domain assumption Diagrammatic Monte Carlo provides numerically exact results for the attractive Hubbard model on the Lieb lattice when sufficiently high diagram orders are included.
    Invoked when the abstract states that the technique is controlled and traces the onset of superfluid response.

pith-pipeline@v0.9.0 · 5494 in / 1371 out tokens · 52496 ms · 2026-05-10T18:38:54.658525+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

30 extracted references · 30 canonical work pages

  1. [1]

    (7) with c(DMFT) ∗ = 0.062 for the standard Lieb lattice

    predicted that the mean-field transition tempera- tureT (DMFT) ∗ –an analog of ourT ∗–follows Eq. (7) with c(DMFT) ∗ = 0.062 for the standard Lieb lattice. However, DMFT allows one to compute the superfluid stiffness and this was used to obtain the transition temperature via the universal Nelson-Kosterlitz relation [27] with the re- sultT (DMFT) c ≈0.047U...

    work page

  2. [2]

    Khodel and V.R

    V.A. Khodel and V.R. Shaginyan, Superfluidity in sys- tems with condensed Fermi surface, JETP Lett.,51(9), 553-555, (1990)

    work page 1990

  3. [3]

    Kopnin, T.T

    N.B. Kopnin, T.T. Heikkil¨ a, and G.E. Volovik, High- temperature surface superconductivity in topological flat-band systems, Physical Review B83(22), 220503 (2011)

    work page 2011

  4. [4]

    Heikkil¨ a, N.B

    T.T. Heikkil¨ a, N.B. Kopnin, and G.E. Volovik, Flat bands in topological media, JETP Lett.94, 233 (2011)

    work page 2011

  5. [5]

    Peotta and P

    S. Peotta and P. T¨ orm¨ a, Superfluidity in topologically nontrivial flat bands, Nature Communications,6(1), 8944 (2015)

    work page 2015

  6. [6]

    Julku, S

    A. Julku, S. Peotta, T.I. Vanhala, D.-H. Kim, and P. T¨ orm¨ a, Geometric origin of superfluidity in the Lieb- lattice flat band, Phys. Rev. Lett.117, 045303 (2016)

    work page 2016

  7. [7]

    Liang, T.I

    L. Liang, T.I. Vanhala, S. Peotta, T. Siro, A. Harju, and P. T¨ orm¨ a, Band geometry, Berry curvature, and super- fluid weight, Phys. Rev. B95, 024515 (2017)

    work page 2017

  8. [8]

    Liu, X.-B

    T. Liu, X.-B. Qiang, H.-Zh. Lu, and X.C. Xie, Quantum geometry in condensed matter, National Science Review 12(3), nwae334 (2025)

    work page 2025

  9. [9]

    J. Yu, B.A. Bernevig, R. Queiroz, E. Rossi, P. T¨ orm¨ a, B.-J. Yang, Quantum geometry in quantum materials, npj Quantum Materials10, 101 (2025)

    work page 2025

  10. [10]

    Huhtinen, J

    K.-E. Huhtinen, J. Herzog-Arbeitman, A. Chew, B.A. Bernevig, and P. T¨ orm¨ a, Revisiting flat band supercon- ductivity: dependence on minimal quantum metric and band touchings. Phys. Rev. B106, 014518 (2022)

    work page 2022

  11. [11]

    Herzog-Arbeitman, A

    J. Herzog-Arbeitman, A. Chew, K.-E. Huhtinen, P. T¨ orm¨ a, and B.A. Bernevig, Many-body super- conductivity in topological flat bands. Preprint at https://arxiv.org/abs/2209.00007 (2022)

    work page arXiv 2022

  12. [12]

    T¨ orm¨ a, S

    P. T¨ orm¨ a, S. Peotta, B.A. Bernevig, Superconductivity, superfluidity and quantum geometry in twisted multi- layer systems, Nature Review Physics4, 528 (2022)

    work page 2022

  13. [13]

    Penttil¨ a , K.-E

    R.P.S. Penttil¨ a , K.-E. Huhtinen, and P. T¨ orm¨ a, Flat- band ratio and quantum metric in the superconductivity of modified Lieb lattices, Commun. Phys.8, 50 (2025)

    work page 2025

  14. [14]

    Prokof’ev and B.V

    N.V. Prokof’ev and B.V. Svistunov, Polaron problem by diagrammatic quantum Monte Carlo, Phys. Rev. Lett. 81, 2514 (1998)

    work page 1998

  15. [15]

    Prokof’ev and B

    N. Prokof’ev and B. Svistunov, Bold Diagrammatic Monte Carlo Technique: When the Sign Problem Is Wel- come, Phys. Rev. Lett.99, 250201 (2007)

    work page 2007

  16. [16]

    Van Houcke, E

    K. Van Houcke, E. Kozik, N. Prokof’ev, B. Svistunov, Di- agrammatic Monte Carlo, Physics Procedia6, 95 (2010)

    work page 2010

  17. [17]

    Kozik, K

    E. Kozik, K. Van Houcke, E. Gull, L. Pollet, N. Prokof’ev, B. Svistunov, and M. Troyer, Diagrammatic Monte Carlo for Correlated Fermions, Euro Phys. Lett. 90, 10004 (2010)

    work page 2010

  18. [18]

    Currie and E

    B. Currie and E. Kozik, Fractional quantum Hall states by Feynman’s diagrammatic expansion, https://arxiv.org/abs/2412.21064 (2026)

    work page arXiv 2026

  19. [19]

    Lieb, Two Theorems on the Hubbard Model, Phys

    E.H. Lieb, Two Theorems on the Hubbard Model, Phys. Rev. Lett.62, 1201 (1989)

    work page 1989

  20. [20]

    Connor Lenihan, Aaram J. Kim, Fedor ˇSimkovic IV, and Evgeny Kozik, Evaluating Second-Order Phase Transi- tions with Diagrammatic Monte Carlo: N´ eel Transition in the Doped Three-Dimensional Hubbard Model, Phys. Rev. Lett.129, 107202 (2022)

    work page 2022

  21. [21]

    ˇSimkovic IV, Y

    F. ˇSimkovic IV, Y. Deng, N.V. Prokof’ev, B.V. Svis- tunov, I.S. Tupitsyn, and E. Kozik, Magnetic correlations in the two-dimensional repulsive Fermi-Hubbard model, Phys. Rev. B96, 081117(R) (2017)

    work page 2017

  22. [22]

    Kozik, Combinatorial summation of Feynman dia- grams, Nature Communications,15, 7916 (2024)

    E. Kozik, Combinatorial summation of Feynman dia- grams, Nature Communications,15, 7916 (2024)

    work page 2024

  23. [23]

    Baker Jr, Application of the Pad´ e approximant method to the investigation of some magnetic properties of the Ising model, Phys

    G.A. Baker Jr, Application of the Pad´ e approximant method to the investigation of some magnetic properties of the Ising model, Phys. Rev.124, 768 (1961)

    work page 1961

  24. [24]

    D. L. Hunter and G. A. Baker, Jr., Phys. Rev.B19, 3808 (1979)

    work page 1979

  25. [25]

    ˇSimkovic IV and E

    F. ˇSimkovic IV and E. Kozik, Determinant Monte Carlo for irreducible Feynman diagrams in the strongly corre- lated regime. Phys. Rev. B100, 121102(R) (2019)

    work page 2019

  26. [26]

    Rossi, F

    R. Rossi, F. Werner, N. Prokof’ev, and B. Svistunov, Shifted-action expansion and applicability of dressed dia- grammatic schemes, Phys. Rev. B93, 161102(R) (2016)

    work page 2016

  27. [27]

    Kozik, M

    E. Kozik, M. Ferrero, and A. Georges, Nonexistence of the Luttinger-Ward Functional and Misleading Conver- 6 gence of Skeleton Diagrammatic Series for Hubbard-Like Models, Phys. Rev. Lett.114, 156402 (2015)

    work page 2015

  28. [28]

    Nelson and J.M

    D.R. Nelson and J.M. Kosterlitz, Universal Jump in the Superfluid Density of Two Dimensional Superfluids, Phys. Rev. Lett.39, 1201 (1977)

    work page 1977

  29. [29]

    self-energy

    Beckett, G., Beech-Brandt, J., Leach, K., Payne, Z., Simpson, A., Smith, L., Turner, A., Whiting, A. ARCHER2 Service Description. (2024) End-Matter Bold4+ Scheme The Bold4+ scheme starts with solving a set of four self-consistent equations accounting for renormalization of single- and two-particle channels for contact interaction using the lowest-order di...

    work page 2024

  30. [30]

    Note that this estimate has nothing to do with the lattice type or band structure. This estimate for the required strength of the e-ph coupling corresponds to moderate values of the dimensionless parameterλin the Migdal-Eliashberg theory forconventional superconductors with dispersive conduction bands. Indeed, for (8) we have λ=N F g2 ω2 0 ,(10) whereN F ...

    work page