Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

The anisotropic Heisenberg model close to the Ising limit: triangular lattice vs. effective models

T0 review · 3 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read DMRG and exact-diagonalization results on up to 72 sites indicate a gapped zero-field ground state for the easy-axis triangular-lattice Heisenberg model with α≲0.3–0.5, and a crossover/transition to gapless behavior at larger α.

desk verdict A careful but not decisive numerical case for a gapped ground state in the easy-axis triangular lattice; the effective-model comparison is the cleanest part, the extrapolations are the weak link. read the letter →

arxiv 2510.12667 v2 pith:IMXQ3XOV submitted 2025-10-14 cond-mat.str-el

Martin Ulaga , Jure Kokalj , Takami Tohyama , Peter Prelovšek This is my paper · ORCID
classification cond-mat.str-el
keywords alphalatticemodeltriangularmodelsanisotropiceffectiveheisenberg
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

Imagine tiny magnetic arrows on a triangular grid. When the material strongly prefers arrows to point up or down (easy-axis anisotropy), the zero-field ground state is debated: either a supersolid—a frozen pattern that also has a weak rotating component—or a plain gapped solid with no rotating component. This paper attacks the question with two numerical techniques and with simpler lattice models.

The authors study the spin-1/2 Heisenberg model with anisotropy α = J⊥/Jz < 1. Using density-matrix renormalization group (DMRG) on clusters up to N=72 sites and exact diagonalization on smaller clusters, they extract the magnetization m(h) near zero field. A nonzero intercept h* from the N→∞ extrapolation would mean a magnon gap—zero magnetization until the field exceeds h*. They also compute the spin stiffness ρs, which should vanish for a gapped state. Both probes point to a gap for α below roughly 0.3–0.5, and the inferred h* scales approximately like ζ α J. They emphasize that linear spin-wave theory, which predicts gapless modes and a finite transverse component, fails in this regime because magnons repel each other strongly, similar to electrons forming a Mott insulator.

In parallel, they freeze one-third of the triangular-lattice spins and arrive at effective honeycomb- and square-lattice models. These reproduce the triangular-lattice magnetization and transverse component for partially polarized states, but at the point meant to represent the triangular zero-field case they remain gapless with finite m⊥. So the simpler lattices are useful analogues, yet they miss the very gap the triangular lattice develops. The paper leaves open whether the gapped-to-gapless change at α* is a true transition or a smooth crossover.

Extended reading notes

Core claim

At zero field and α≪1, the triangular-lattice AHM has a gapped ground state with m⊥=0: 'the extrapolated results would be consistent with quite different marginal fields (effective magnon gaps) h*=Δ1=ζαJ' and 'extrapolated ρs/(αJ)→0 in the regime α<α*∼0.3'. The phase diagram (Fig. 7) therefore contains a 'gapped spin solid' for h<h*(α), with a crossover/transition to gapless supersolid at α*≲0.5.

Load-bearing premise

The deduced gap rests on N→∞ extrapolations of small-cluster quantities. h* is read off quadratic fits to ED/DMRG m(h) data for N≤72 (Fig. 4, Appendix A), and ρs is extrapolated by 1/N from N≤36 ED plus N=48 DMRG (Fig. 6). If the true thermodynamic limit has h*→0 or ρs>0 with a different scaling (the data are strongly size-dependent and the two probes give different α*≈0.5 vs 0.3), the gapped-solid claim collapses. The authors concede 'it seems beyond present numerical capabilities to clarify whether we are dealing with a transition or a crossover'.

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 / 6 minor

Summary. The paper studies the spin-1/2 easy-axis (XXZ) Heisenberg model on the triangular lattice (TL), with exchange anisotropy α = J⊥/Jz < 1 and a longitudinal field h, motivated by experiments on K2Co(SeO3)2 (α ≈ 0.07). It first compares the full TL model with effective models on honeycomb and square lattices obtained by freezing one third of the spins; at the correspondence point (m = 1/2 on the bipartite lattices, which corresponds to h = 0 on the TL), these effective models remain gapless with finite transverse magnetization m⊥. The central and more conjectural claim is that the TL model itself has a gapped ground state at h = 0 for α ≪ 1, with h* = Δ1 = ζαJ, and a transition/crossover to a gapless supersolid at α* ≲ 0.5. Evidence is drawn from (i) the zero-crossing of polynomial fits to magnetization curves m(h) from ED/DMRG on N = 30–72 (Sec. III A, Appendix A), and (ii) the 1/N-extrapolated spin stiffness ρs/(αJ), which vanishes for α ≲ 0.3 (Sec. III B). LSWT is shown to fail at h ∼ 0, with the failure attributed to effective magnon repulsion. The resulting phase diagram (Fig. 7) contains a gapped spin solid for h < h*(α). The authors explicitly concede that a transition vs a crossover at α* cannot be distinguished with present numerics.

Significance. Should the gapped-solid scenario survive, the paper would help settle an active controversy: the easy-axis TL model at α ≪ 1 would not be a supersolid at h = 0, directly affecting the interpretation of KCSO neutron-scattering and thermodynamic experiments, and it identifies a qualitative failure mechanism for LSWT (magnon repulsion / Mott-like gap). The numerical work is substantial: systematic ED/DMRG magnetization curves up to N = 72, a spin-stiffness analysis including a DMRG point at N = 48, and finite-size scaling of m⊥ on the effective honeycomb model. The multi-probe design (m(h), ρs, m⊥) and the explicit α > 0 vs α < 0 comparison are strengths, as is the falsifiable prediction h* = ζαJ with ζ extracted from data. The weaknesses are that both gap diagnostics are extrapolations with untested asymptotic forms and no error bars, and the two probes give inconsistent α* values (≈0.5 vs ≈0.3). The central claim is therefore plausible but not established by the present data; the stress-test concern about the unvalidated finite-size extrapolation lands.

major comments (3)
  1. [Sec. III A / Figs. 4,5; Appendix A] The gapped-solid claim rests on the marginal field h* extracted as the zero-crossing of a polynomial fit to m(h) for N = 30–72 (Fig. 4; Appendix A), not on a direct measurement of the one-magnon gap Δ1 = E(S_z = 1) − E0. For a gapless system with downward curvature in m(h), a polynomial fit can produce a spurious positive zero-crossing; the good data collapse does not fix the intercept. At α = 0.1 the extrapolated h* ≈ ζαJ ≈ 0.01–0.02 J (Fig. 5) is comparable to or smaller than the finite-size gap scale expected for a gapless system on the N = 72 cluster (∝ αJ/L, of order 0.07 J), so h* is not yet separated from finite-size effects. Fig. 5 has no error bars and the fit form/window are not specified. I recommend adding a direct one-magnon-gap analysis: extrapolate Δ1(N) = E(S_z = 1) − E0, already available in the DMRG runs restricted to S_z^tot ≤ 4, and test the sensitivity of h* to the f
  2. [Sec. III B / Fig. 6] The spin stiffness is extrapolated to N → ∞ with a bare 1/N ansatz using only four points (N = 18, 30, 36 by ED; N = 48 by DMRG), a fixed twist θ = 0.1 with no convergence check in θ, and no error bars. The conclusion that ρs/(αJ) → 0 for α ≲ 0.3 is fragile; alternative scalings (e.g., 1/N^2, exponential convergence, or an added curvature term) should be tested and shown not to change the zero crossing. In addition, the two probes are quantitatively inconsistent — α* ≈ 0.5 from h*(α) in Sec. III A vs α* ≈ 0.3 from ρs in Sec. III B — which under-determines the boundary drawn in Fig. 7. This discrepancy should be reconciled or presented as an explicit uncertainty range for α*.
  3. [Sec. V / Fig. 7] The authors state that it is 'beyond present numerical capabilities to clarify whether we are dealing with a transition or a crossover at a particular α*.' This ambiguity is load-bearing: the abstract and the phase diagram assert a gapped GS phase for h < h*(α), but a genuine thermodynamic phase requires h*(∞) > 0. Given the strong size dependence in Figs. 4 and 6 and the mismatch between the two probes, the abstract's wording ('confirm the existence of the gap', 'indicate a transition/crossover') overstates what the data establish. Unless the direct-gap extrapolation requested above is supplied, the central claim should be presented as 'consistent with a gapped solid'.
minor comments (6)
  1. [Fig. 5 (Sec. III A)] The procedure producing h* is not fully described: specify the polynomial degree, the h/(αJ) fitting window, and whether the fit is applied to the pooled finite-N data or to an extrapolated m(h). Add error estimates (e.g., bootstrap over N or over the fit range).
  2. [Eq. (3), Sec. III B] The stiffness formula appears as 'ρs = (1/N)∂²E0/∂²θ', which should read ∂²E0/∂θ² evaluated at θ = 0. Also, the fixed value θ = 0.1 is used without a convergence check; a brief test of θ-dependence would strengthen the DMRG points in Fig. 6.
  3. [Eq. (8), Sec. IV] In HJ,BC the two hopping terms are printed identically (a_i a†_j appears twice); presumably the second should be a†_i a_j. Please correct and verify the subsequent algebra.
  4. [Fig. 2 caption] The caption says 'DSSP' in two places; the text defines DSSF (dynamical spin structure factor). Unify the acronym.
  5. [Fig. 6 legend] The legend 'N 18, 30, 36, ∞, 48' is confusing: ∞ sits between cluster sizes, and the DMRG N = 48 points are listed last although the text introduces them as crosses. Reorder and explain the symbols in the caption.
  6. [Abstract / Sec. I] The abstract says 'several additional numerical studies ... confirm the existence of the gap at α ≪ 1.' The cited literature is divided (Refs. 13–15 and 27 report gapless/supersolid behavior in parts of this regime). 'Confirm' overstates the current state of evidence; consider 'support' or 'are consistent with'.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the gapped-solid claim rests on fresh ED/DMRG extrapolations; minor self-citations are contextual, not load-bearing.

full rationale

The central gapped-solid conclusion is not an input to the calculation. h* is read from polynomial extrapolations of m(h) computed with ED/DMRG for N=30...72 (Sec. III A, Fig. 4, Appendix A), and rho_s is obtained from a twist second derivative (Eq. 3) with 1/N extrapolation (Sec. III B, Fig. 6). Neither observable is defined to produce the claimed phase; the m(h) threshold and vanishing rho_s are the empirical evidence. The equality h* = Delta_1 is an operational identity for a gapped phase, not a construction that forces the result. Self-citations [25,26] supply context, methods, and previous smaller-N evidence, but the present N=72 data and stiffness extrapolations stand independently; no central claim reduces to a self-citation. The acknowledged unresolved transition-vs-crossover ambiguity and the alpha* discrepancy are limitations on certainty, not circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

No new particles, forces, or entities are postulated. The 'gapped spin solid' is a label for a known model state, not an invented entity.

free parameters (3)
  • marginal field h* (scaled gap ζ = h*/(αJ)) = α-dependent; e.g., for α=0.1 a small ζ≲0.2 implied by Fig. 5; ζ decreases toward 0 at α*≈0.5
    Obtained by quadratic extrapolation of finite-size m(h) data; this fitted gap is the quantitative basis for the gapped-solid phase and α*.
  • spin stiffness extrapolation ρs(N→∞) = ≈0 for α<~0.3; ≈0.05J at α=1 (from 1/N fits)
    Used as the second independent indicator of the gap; the N→∞ value is fitted from N=18–48 data and carries no uncertainty.
  • α* (gap-closing anisotropy) = ≲0.5 from h* data; ~0.3 from ρs data
    Extracted visually/semi-quantitatively as the point where ζ and ρs/(αJ) drop to zero; the discrepancy between the two estimates is part of the soundness caveat.
assumptions (5)
  • domain assumption Finite-size ED/DMRG spectra with PBC (N≤72, S_z^tot≤4 near h=0) represent the thermodynamic-limit low-energy manifold.
    The h* and ρs extrapolations assume the smallest magnetization sectors control m(h→0); not proven.
  • domain assumption The 1/N (or quadratic in N) scaling ansatz for h* and ρs is valid.
    No microscopic justification for the scaling form; different forms could change whether h*→0.
  • domain assumption Freezing one third of the triangular-lattice spins and mapping to HcL/SqL at m=1/2 corresponds to the TL at h=0.
    This mapping is exact only at α→0 on the m=1/3 plateau; at finite α and at m=1/2 it is an approximation that may miss the gap physics.
  • domain assumption ρs=0 is a reliable signature of a gapped ground state.
    Standard relation from Kohn/Scalapino analogies, but for a spontaneously symmetry-broken spin state the finite-size extrapolation is delicate.
  • domain assumption LSWT with quadratic bosonization is valid near the saturation/plateau; discarded quartic magnon repulsion terms dominate at low m.
    The paper's explanation for LSWT failure rests on this assumed hierarchy of terms, not on a controlled calculation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The anisotropic Heisenberg model close to the Ising limit: triangular lattice vs. effective models." pith.science (2026). https://pith.science/paper/IMXQ3XOV

@misc{pith2026251012667,
  author       = {Pith},
  title        = {Pith review of: The anisotropic Heisenberg model close to the Ising limit: triangular lattice vs. effective models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IMXQ3XOV}},
  note         = {Machine review of arXiv:2510.12667}
}
abstract

Stimulated by recent experiments on materials representing the realization of the anisotropic Heisenberg spin-$1/2$ model on the triangular lattice, we explore further properties of such a model in the easy-axis regime $\alpha = J_\perp/J_z < 1$ and the plausibility of finding effective models that capture similar physics. We show that, at finite fields, the magnetization curve as well as the transverse magnetization (superfluid) order parameter $m_\perp$ of the triangular lattice model are indeed qualitatively reproduced by anisotropic Heisenberg models on the honeycomb or the square lattice. At the point of correspondence to the zero-field triangular lattice model, however, the bipartite models are qualitatively different as they remain gapless even at $\alpha \ll 1$ with a small but finite $m_\perp >0 $. Conversely, we present several additional numerical studies of the full model on the triangular lattice which support the appearance of a gap at zero field and $\alpha \ll 1$. In particular, the magnetization curve $m(h)$ as well as the spin stiffness $\rho_s$ indicate a transition/crossover from gappless to gapped regimes at $\alpha \sim \alpha^*$ with $\alpha^* \lesssim 0.5$. We also show that deviations from the linear spin-wave theory and the emergence of the gap can be traced back to the strong effective repulsion between magnon excitations, showcasing similarity to strongly correlated systems.

Figures

Figures reproduced from arXiv: 2510.12667 by the authors.

Figure 1
Figure 1. Magnetization curves m(h) vs. renormalized fields h˜ = h/(zSJ) for (a) the honeycomb lattice (HcL) and ( b) the square lattice (SqL), for different anisotropies α = 0.1, 0.2, 0.5, obtained with DMRG calculation on lattices with N = 72 sites and N = 64 sites, respectively. The gray dashed lines indicate m = 1/2 and h˜ = 1, corresponding to m = 0 and h = 0, respectively, within the TL model. The coloured dashed lines … view at source ↗
Figure 2
Figure 2. Transverse order parameter m2 ⊥ vs. magnetization m, ex￾tracted from ED results for DSSP on systems with N = 36 and N = 40 sites on (a) HcL, and (b) SqL, for different α = 0.1 − 1.0. Ver￾tical m = 1/2 line indicates the correspondence to the h = m = 0 model on TL. The gray dashed curve represents the expected LSWT dependence m2 ⊥ = ζ(1 − m2 ) with ζ = 1/8. 0.00 0.02 0.04 0.06 1/N 0.00 0.02 0.04 0.06 0.08 0.10 m2⊥ (a… view at source ↗
Figure 3
Figure 3. Transverse order parameter m2 ⊥ at m = 1/2 for HcL : (a) finite-size 1/N scaling of results for different α, (b) m2 ⊥, obtained on N = 40 sites, and extrapolated N → ∞ values vs. α. There, one can give an explanation with the simplest classical result of the LSWT, where m = cos(θ) and m⊥ = S sin(θ) (note our different definitions of m ≤ 1 and m⊥ ≤ 1/2) and consequently m2 ⊥ ∝ 1 − m2 . Still, deviations become very p… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Magnetization curve m vs. normalized field h/(αJ), for α = 0.1, obtained with ED for N = 30, 36 lattices and via DMRG for N = 48−72 lattices, with the quadratic extrapolation (red dashed line). The gray dashed line represents the result of LSWT, explained and discussed…
Figure 5
Figure 5. Figure 5: Scaled gap h ∗ /(αJ) vs. α, as obtained by extrapolation of m(h) results presented in [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]
Figure 6
Figure 6. Figure 6: Normalized spin stiffness ρs/(αJ) vs. anisotropy α, as calculated on TL via ED on systems with different sizes N = 18, 30, 36, together with the extrapolated value N → ∞, obtained via the 1/N scaling. Crosses denote DMRG results on N = 48 sites. TL [39]. For the presen…
Figure 7
Figure 7. Figure 7: The ground-state phase diagram h/J vs. α obtained in this work, displaying three phases: the gapped spin solid (GS) phase for h < h∗ , the supersolid Y (SSY) phase for h∗ < h < hc and the polarized UUD phase on the magnetization plateau h > h∗ . ing α < α∗ ∼ 0.5 , as w…
Figure 8
Figure 8. Figure 8: Magnetization curves m vs. normalized fields h/(αJ) for various anisotropies −0.2 ≤ α ≤ 1. The thin gray lines represent LSWT results, while the red lines are simple polynomial fits to the data for all cluster sizes N for small m. LSWT result for α = 1 is obtained for …

Discussion (0). Continue with ORCID 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. Emergent Spin Supersolids in Frustrated Quantum Materials

    cond-mat.str-el 2026-01 unverdicted novelty 2.0 of 10

    Spin supersolids featuring coexisting longitudinal spin order breaking lattice symmetry and transverse order breaking spin U(1) symmetry have been established in frustrated quantum magnets through consistent experimen...

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.