The Monte Carlo Method for the Orthonickelate Model

A.S. Moskvin; S.V. Nuzhin; V.S. Ryumshin; Yu.D. Panov

REVIEW 3 major objections 6 minor 13 references

The Monte Carlo Method for the Orthonickelate Model

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The triplet-boson model for orthonickelates, simulated with a classical Monte Carlo method that keeps boson density fixed, forms macroscopic phase-separated domains rather than homogeneous states with several nonzero order parameters.

desk verdict Plausible MFA/MC phase-separation story, but the MC algorithm has a non-symmetric proposal without Hastings correction that needs to be fixed. read the letter →

arxiv 2411.16957 v2 pith:GYCELV5Q submitted 2024-11-25 cond-mat.stat-mech cond-mat.str-el

Yu.D. Panov , S.V. Nuzhin , V.S. Ryumshin , A.S. Moskvin This is my paper · ORCID

classification cond-mat.stat-mechcond-mat.str-el MSC 82B2682B8082B20

keywords orthonickelatestripletbosonmodelphaseseparationmean-fieldapproximationclassicalMonteCarloMaxwellconstructionchargeorderingantiferromagnetism

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

This paper argues that the low-temperature states of the triplet-boson model for orthonickelates are not homogeneous phases with several simultaneous order parameters but spatial mixtures of pure phases. Using a classical Monte Carlo algorithm with kinematically fixed boson concentration, the authors find that for $z=4$, $V/J=4$, $t/J=1.5$ the system separates into boson-superfluid/charge-ordered domains for $0.15

What carries the argument

The central object is the quasi-classical on-site wave function $|\psi_i\rangle = c_{1,11}|1,11\rangle_i + c_{1,10}|1,10\rangle_i + c_{1,1-1}|1,1-1\rangle_i + c_{0,00}|0,00\rangle_i$, with four amplitudes parametrized by angles $\theta_i,\psi_i,\phi_i$ and uniformly sampled on the 8-dimensional sphere of coefficients. The Metropolis updates act on pairs of sites: a new density on one site is drawn from the inverse distribution function $F_1(n_1;\bar n)$ constructed from the density-weighted measure $f(n)=3n^2$, and the second site density is fixed by $\bar n$, which keeps the total boson concentration constant at every step. This machine lets the simulation reveal whether mixed phases persist or decompose, and it is compared against the MFA Maxwell construction built from chemical-potential equality $\mu_i(n,T)=\mu^*$ and free energy $f=m_1f_1+m_2f_2$.

What would settle it

Run an unbiased quantum calculation, such as exact diagonalization on a small cluster or a projective quantum Monte Carlo for Hamiltonian (1) at $z=4$, $V/J=4$, $t/J=1.5$ and $n=0.65$; if the ground state exhibits simultaneous nonzero CO and AFM order parameters on the same sites, or fails to show spatial separation into CO and AFM domains, the paper's central claim would be refuted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the mixed-order-parameter solutions of the mean-field approximation are not realized as homogeneous states when the full lattice is simulated. The classical Monte Carlo method with fixed boson density shows the same phase separation that the MFA Maxwell construction predicts: BS/CO coexistence for $0.15<n<0.50$ and CO/AFM coexistence for $0.5<n<1.0$ at $z=4$, $V/J=4$, $t/J=1.5$. The temperature scales of MFA and MC differ by a factor of about $6.3$, yet the ordering of critical temperatures of CO and AFM phases persists, while the BS phase occupies a smaller region in the MC diagram than in MFA.

Load-bearing premise

The classical Monte Carlo method replaces the quantum state by a product of single-site quasi-classical wave functions, so if quantum entanglement changes which phases are stable, the agreement between Monte Carlo and mean-field could reflect the shared classical approximation rather than the true quantum model.

Editorial extensions

If this is right

Homogeneous 'supersolid-like' mixed phases of the triplet-boson model are thermodynamically unstable at low temperature and decompose into macroscopic phase-separated regions.
The Maxwell construction in mean-field theory gives qualitatively correct binodals for the model, so MFA phase diagrams can be trusted for locating coexistence regions at least at the qualitative level.
The boson-superfluid phase is unstable at high concentrations and separates with the non-ordered phase, unlike the singlet local-boson model.
The antiferromagnetic phase is unstable at weak charge-charge interaction ($V/J<1$) and gives way to AFM/NO phase separation, while at $V/J>1$ homogeneous AFM is stable at all concentrations below its critical temperature.

Reading between the lines

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

If the product-state ansatz is what drives the qualitative agreement, an entangled quantum calculation could find that quantum fluctuations stabilize some homogeneous mixed phases; this is an inference, not a result of the paper.
The fixed-density pair-update algorithm could be adapted to other conserved-density hard-core boson models, since it avoids the flat-chemical-potential problem that makes grand-canonical simulations slow at low temperature.
The apparent shift of the CO critical-temperature maximum from $n=0.5$ to about $n=0.55$ in the MC data may be a threshold or finite-size artifact; a scaling-theory analysis, which the authors say is planned, would settle whether it is physical.

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.

Desk Editor's Note

read the letter

Short version: this is a modest but honest MFA plus classical Monte Carlo study of a triplet-boson model with fixed density. The new part is the MC implementation and the MC phase diagram, and the main qualitative conclusion—that the system phase separates instead of forming homogeneous mixed phases—is plausible. But there is a real algorithmic flaw in the MC proposal that the authors need to fix or explain: the pair update in Sec. 3 is not symmetric, and the paper gives no Hastings correction, so the sampled ensemble carries an extra n_i^2 weighting. That is a genuine problem, not a nitpick.

Credit where due: the MFA stability analysis is clean, and the Maxwell construction is clearly laid out. The comparison between MFA and MC is a real comparison, not a fit, since the MC is an independent numerical procedure. Fig. 7 shows nice snapshots of coexisting domains, and the authors are upfront that the 1% threshold is ad hoc and that finite-size scaling has not been done.

The soft spots, in order. First, the Metropolis-Hastings issue. The proposal density for new densities n1,n2 at fixed n̄ is proportional to n1^2 n2^2, which depends on the new values. Accepting with the Boltzmann factor alone targets a modified distribution. The bias is of order 2T ln n, comparable to the couplings at the simulated T. It could shift T_c and coexistence boundaries, including the reported MFA/MC temperature-scale ratio of about 6.3. The authors should add the ratio, or show numerically that it doesn't matter. Second, there are no error bars, no multiple seeds, and only one lattice size. For a phase-separation claim, you need at least some finite-size evidence. Third, the quasi-classical product-state ansatz is not benchmarked against exact quantum results on small clusters, so the quantitative agreement with MFA could partly come from the shared classical approximation. The authors acknowledge this indirectly by saying scaling analyses are planned.

Who is this for? People working on nickelate models and on constrained boson Monte Carlo. It is a valid piece of work, not a breakthrough, and the central message is probably right. I would send it to peer review: it deserves a serious referee. But I would ask the authors to fix the Hastings issue and add error bars or finite-size scaling before acceptance.

Referee Report

3 major / 6 minor

Summary. The paper studies a spin-triplet composite-boson model for orthonickelates (Hamiltonian (1)). In Sec. 2 the authors derive MFA stability conditions for homogeneous charge-ordered (CO), boson superfluid (BS), and antiferromagnetic (AFM) phases, and construct phase-separation binodals via the Maxwell construction. In Sec. 3 they introduce a classical Monte Carlo algorithm that samples product-state wave functions (Eq. (3)) with a kinematically fixed total boson density. In Sec. 4 they compare MC order-parameter onsets with the MFA phase diagram for z=4, V/J=4, t/J=1.5, after rescaling the temperature scales, and report phase separation into macroscopic BS/CO and CO/AFM domains. The central conclusion is that the classical MC simulation confirms the MFA Maxwell-construction picture.

Significance. The analytical stability conditions and the Maxwell-construction analysis (Sec. 2) are clear and internally consistent within the mean-field framework. The MC simulation is a genuinely independent numerical procedure that does not reduce by construction to the MFA result, so the qualitative agreement is a meaningful test of the Maxwell construction inside the product-state manifold. The snapshots in Fig. 7 directly show macroscopic domain coexistence. The main limitations are that the MC algorithm as described is not a correct Metropolis update for the stated Boltzmann weight, and the quasi-classical product-state ansatz is not validated against quantum benchmarks; the paper itself flags that scaling-theory estimates are 'planned to be made in the future' (Sec. 4).

major comments (3)

[Sec. 3, after Eqs. (4)-(12)] The pair-update proposal is not symmetric. The conditional density f1(n1; n̄) in Eq. (9) is proportional to n1^2 (2n̄ - n1)^2 = n1^2 n2^2, so for a move from an old state x to a new state x' on the same pair, q(x'|x) ∝ (n1' n2')^2 while q(x|x') ∝ (n1 n2)^2. The ratio q(x|x')/q(x'|x) is generally not unity. The statement in Sec. 3 that 'we use standard Metropolis algorithm' is therefore insufficient: standard Metropolis requires symmetric proposals. The acceptance probability must include the Metropolis-Hastings factor min(1, exp(-ΔE/T) (n1 n2 / n1' n2')^2). Without this correction, the stationary distribution is ∝ exp(-E/T) ∏_i n_i^2 instead of exp(-E/T), and since 2T ln n is of order unity over the simulated T/J range, the bias can shift phase boundaries and affect the order-parameter onsets. The authors should either implement the correction or demonstrate numerically that this bias does not alter the reported phase diagram.
[Sec. 3, Eqs. (3)-(7); Sec. 4] The quasi-classical product-state ansatz in Eq. (3) is an uncontrolled approximation to the quantum Hamiltonian (1). The MC simulation samples configurations in the manifold of product states with a uniform measure, and the MFA of Ref. [9] operates on the same manifold. Consequently, agreement between the MC and MFA results does not validate the ansatz itself; it only confirms the Maxwell-construction picture within the classical manifold. To support the claim that phase separation is a property of the quantum model, the paper should benchmark the classical MC against exact diagonalization or quantum Monte Carlo for small lattices, or explicitly restrict its conclusion to the quasi-classical version of the model.
[Sec. 4, Figs. 4-6] The critical temperatures are extracted from an arbitrary 1% threshold of the order parameter, with no error bars, finite-size scaling, or thermalization/blocking analysis. The paper itself states that 'more extensive analyses with the estimate of critical temperatures as per scaling theory are planned to be made in the future.' Since the temperature rescaling T_c,max(MFA)/T_c,max(MC) ≈ 6.3 is a central element of the comparison, the choice of threshold directly affects the reported scale ratio. Please provide sensitivity tests (e.g., different thresholds, system sizes, and numbers of MC steps) and statistical error estimates, or explicitly present the phase diagram as preliminary.

minor comments (6)

[Abstract] The phrase 'observed state of the system in numerical simulations' should be clarified as 'the state observed in the numerical simulations', since no experimental observation is made.
[Eq. (12)] The formulas for n1,min and n1,max could be derived explicitly; as written they are correct but not transparent. A short explanation that they enforce n1, n2 ∈ [0,1] would help the reader.
[Fig. 2 caption and Sec. 2] The formula for T_BS has ambiguous parentheses: T_BS = 4t(4n-3)[3 ln(n/(3(1-n)))]^{-1} would be clearer than the printed expression.
[Sec. 3] The text says 'uniformly distributed in section [0,1]' where 'interval' or 'segment' is meant; this is a language issue that should be corrected throughout.
[Sec. 3, step 7] The MC implementation does not specify how many initial steps are discarded for thermalization before averaging, nor whether the 4×10^6 steps include thermalization. Please specify the averaging protocol.
[Sec. 5, Conclusion] The statement that 'AFM phase is unstable at small inter-center charge-to-charge interaction, V/J < 1' is an MFA result from Sec. 2; the MC simulations were only performed at V/J = 4. The conclusion should distinguish between MFA-derived and MC-verified claims.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the classical MC simulation is an independent sampler for Hamiltonian (1) on an explicit product-state manifold, and it is not fitted to the MFA Maxwell-construction boundaries.

full rationale

The paper's central comparison is between MFA phase diagrams (imported from the same group's ref. [9]) and classical Monte Carlo simulations. The MC algorithm in Sec. 3 defines an explicit quasi-classical product-state ansatz, Eq. (3), and samples it with a density-conserving Metropolis update; it does not take the MFA binodals or Maxwell construction as an input. Both MFA and MC approximate the same quantum Hamiltonian (1) using classical product states, so their agreement is a consistency check within a shared ansatz rather than an independent validation of the quantum model against exact results. That shared-ansatz limitation, and the self-cited MFA baseline [9], reduce the strength of the conclusion but do not make the MC result equivalent to the MFA result by construction. The 1% order-parameter threshold used to estimate T_c is a standard finite-size estimator, and the paper explicitly notes that the topology is insensitive to the threshold. No fitted parameter is renamed as a prediction, and no uniqueness theorem is imported from the authors' prior work. The self-citations provide the model and the MFA comparison, but the MC phase-separation finding is an independent numerical result rather than a deterministic consequence of the MFA calculation. The Metropolis-Hastings proposal-ratio concern raised by a skeptical reader is a correctness issue, not a circularity, and does not affect the circularity score under the stated criteria.

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

The central claim depends on the triplet-boson Hamiltonian and the classical product-state MC approximation imported from the authors' earlier work, plus the arbitrary T_c threshold. No new entities are postulated in this paper. The free parameters listed are model and algorithm choices; the paper does not fit them to experimental data, which limits the results to qualitative model-level validity.

free parameters (4)

Hamiltonian coupling ratio V/J = 4
Chosen for the phase diagram in Fig. 4; the MFA/MC comparison is made only at this value, so conclusions are not shown to be parameter-independent.
Hopping ratio t/J = 1.5
Chosen together with V/J for Fig. 4; quantitative phase boundaries depend on it.
Lattice coordination in MC = z=4 (2D square lattice)
The MC is run on a 2D square lattice 'for clarity' rather than the 3D simple cubic lattice of the model; this changes critical temperatures and phase boundaries.
Order-parameter threshold for T_c = 1% of maximal possible value
Used to read off MC critical temperatures and the T_c,max=0.63J scale; the authors note the estimate is sensitive to this threshold.

assumptions (4)

domain assumption The anti-Jahn-Teller disproportionation picture maps RNiO3 to a triplet-boson model with Hamiltonian (1).
Invoked in Sections 1-2 on the basis of earlier papers [5-8]; the present paper provides no experimental or first-principles validation of this mapping.
ad hoc to paper Product-state quasi-classical wave functions, Eq. (3), represent the relevant thermal states of the quantum Hamiltonian.
Introduced in Section 3 for the MC method; no benchmark against exact diagonalization or quantum MC is given.
domain assumption The Maxwell construction, Eq. (2), is the correct way to describe phase coexistence in MFA.
Section 2 calls it a hypothesis that looks reasonable from the physical standpoint but needs independent confirmation. The MC is used as that confirmation, but the hypothesis itself is assumed for the MFA baseline.
domain assumption Metropolis sampling with 4e6 steps reaches equilibrium for the 96x96 lattice.
Section 4 gives the step count but no equilibration diagnostic, autocorrelation analysis, or finite-size scaling; ergodicity of the pair-update moves is assumed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The Monte Carlo Method for the Orthonickelate Model." pith.science (2026). https://pith.science/paper/GYCELV5Q

@misc{pith2026241116957,
  author       = {Pith},
  title        = {Pith review of: The Monte Carlo Method for the Orthonickelate Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GYCELV5Q}},
  note         = {Machine review of arXiv:2411.16957}
}

read the original abstract

The peculiarities of phase states of the triplet boson model for orthonickelates are investigated analytically and by means of numerical simulations. The conditions of thermodynamic stability of homogeneous phases are found. It is shown that the description of the phase inhomogeneous state in the mean-field approximation qualitatively agrees with the observed state of the system in numerical simulations by the classical Monte Carlo method

Figures

Figures reproduced from arXiv: 2411.16957 by the authors.

**Figure 1.** Chemical potential of CO phase 𝑧 = 4, 𝑉/𝐽 = 1 for various 𝑇 . Boundary concentrations are defined by the ratio 𝑇 = 4𝑉 𝑛(1 − 𝑛). in MFA for several temperature values. The boundary concentrations are defined by expression for critical temperature of the charge ordering, 𝑇𝐶𝑂 = 𝑧𝑉 𝑛 (1 − 𝑛), which coincides with the local bosons model case [10]. Condition of phase stability corresponds to positive value of derivative (… view at source ↗

**Figure 2.** a is the chemical potential of BS phase at 𝑧 = 4, 𝑡/𝐽 = 1, 𝑉 = 0 for different 𝑇 . b are the boundaries of BS phase stability at different values 𝑉 (𝑧 = 4, 𝑡/𝐽 = 1). Critical temperature 𝑇𝐵𝑆 = 4𝑡(4𝑛 − 3) [︁ 3 ln 𝑛 3(1−𝑛) ]︁−1 . to the phase separation into BS and NO phase macroscopic domains. The tricritical point 𝐴 divides the curve 𝑇𝐵𝑆 into transition lines of 2-d order on the left and 1-st order on the right of 𝐴… view at source ↗

**Figure 3.** a — chemical potential of AFM phase at 𝑧 = 4, 𝐽 = 1, 𝑉 = 0 for various 𝑇 . b are the boundaries of AFM phase stability at different values𝑉 (𝑧 = 4, 𝐽 = 1). Critical temperature 𝑇𝐴𝐹𝑀 = 8𝐽𝑛/3. 3 Boson concentration condition taken into account in classical Monte Carlo method The nature of chemical potential dependencies of various phase states of the model indicates the complexity of its numerical simulation by MC me… view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Comparison of phase diagrams, obtained in MFA and by MC method for [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

**Figure 5.** Figure 5: Temperature dependencies of order parameters for CO phase (a), BS [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]

**Figure 6.** Figure 6: Temperature dependencies of order parameters for CO phase (a), AFM [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]

**Figure 7.** Figure 7: Lattice state from 𝑁 = 96 × 96 sites for 𝑧 = 4, 𝑉/𝐽 = 4, 𝑡/𝐽 = 1.5 at a — 𝑛 = 0.25, 𝑇/𝑇𝑐,𝑚𝑎𝑥 = 0.02 (point 𝐴 in [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

[9]

Ryumshin, S.V

V.S. Ryumshin, S.V. Nuzhin, Yu.D. Panov, A.S. Moskvin. FTT 66, 7, 1047 (2024). (in Russian)

work page 2024
[1]

Medarde, J

M.L. Medarde, J. Phys.: Condens. Matter 9, 8, 1679 (1997)

work page 1997
[2]

Gawryluk, Y.M

D.J. Gawryluk, Y.M. Klein, T. Shang, D. Sheptyakov, L. Keller, N. Casati, P. Lacorre, M.T. Fern´ andez-D´ ıaz, J. Rodr´ ıguez-Carvajal, M. Medarde, Phys. Rev. B 100, 20, 205137 (2019)

work page 2019
[3]

Post, A.S

K.W. Post, A.S. McLeod, M. Hepting, M. Bluschke, Y. Wang, G. Cristiani, G. Logvenov, A. Charnukha, G.X. Ni, P. Radhakrishnan, M. Minola, A. Pasupa- thy, A.V. Boris, E. Benckiser, K.A. Dahmen, E.W. Carlson, B. Keimer, D.N. Basov, Nature Phys 14, 10, 1056 (2018)

work page 2018
[4]

Mundet, C

B. Mundet, C. Dom´ ınguez, J. Fowlie, M. Gibert, J.M. Triscone, D.T.L. Alexan- der, Nano Lett. 21, 6, 2436 (2021)

work page 2021
[5]

Mazin, D.I

I.I. Mazin, D.I. Khomskii, R. Lengsdorf, J.A. Alonso, W.G. Marshall, R.M. Ibberson, A. Podlesnyak, M.J. Mart´ ınez-Lope, M.M. Abd-Elmeguid, Phys. Rev. Lett. 98, 17, 176406 (2007)

work page 2007
[6]

Moskvin, J

A.S. Moskvin, J. Phys.: Condens. Matter 25, 8, 085601 (2013). 14

work page 2013
[7]

Moskvin, Yu.D

A.S. Moskvin, Yu.D. Panov. Phys. Solid State 65, 7, 1129 (2023)

work page 2023

Show all 13 references

[8]

Moskvin, Magnetochemistry 9, 11, 224 (2023)

A. Moskvin, Magnetochemistry 9, 11, 224 (2023)

2023
[10]

Micnas, J

R. Micnas, J. Ranninger, S. Robaszkiewicz, Rev. Mod. Phys. 62, 1, 113 (1990)

1990
[11]

Batrouni, R.T

G.G. Batrouni, R.T. Scalettar, Phys. Rev. Lett. 84, 7, 1599 (2000)

2000
[12]

Kapcia, S

K. Kapcia, S. Robaszkiewicz, R. Micnas, J. Phys.: Condens. Matter 24, 21, 215601 (2012)

2012
[13]

Kapcia, J Supercond Nov Magn 26, 8, 2647 (2013)

K. Kapcia, J Supercond Nov Magn 26, 8, 2647 (2013). 15

2013

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [9]

Ryumshin, S.V

V.S. Ryumshin, S.V. Nuzhin, Yu.D. Panov, A.S. Moskvin. FTT 66, 7, 1047 (2024). (in Russian)

work page 2024

[2] [1]

Medarde, J

M.L. Medarde, J. Phys.: Condens. Matter 9, 8, 1679 (1997)

work page 1997

[3] [2]

Gawryluk, Y.M

D.J. Gawryluk, Y.M. Klein, T. Shang, D. Sheptyakov, L. Keller, N. Casati, P. Lacorre, M.T. Fern´ andez-D´ ıaz, J. Rodr´ ıguez-Carvajal, M. Medarde, Phys. Rev. B 100, 20, 205137 (2019)

work page 2019

[4] [3]

Post, A.S

K.W. Post, A.S. McLeod, M. Hepting, M. Bluschke, Y. Wang, G. Cristiani, G. Logvenov, A. Charnukha, G.X. Ni, P. Radhakrishnan, M. Minola, A. Pasupa- thy, A.V. Boris, E. Benckiser, K.A. Dahmen, E.W. Carlson, B. Keimer, D.N. Basov, Nature Phys 14, 10, 1056 (2018)

work page 2018

[5] [4]

Mundet, C

B. Mundet, C. Dom´ ınguez, J. Fowlie, M. Gibert, J.M. Triscone, D.T.L. Alexan- der, Nano Lett. 21, 6, 2436 (2021)

work page 2021

[6] [5]

Mazin, D.I

I.I. Mazin, D.I. Khomskii, R. Lengsdorf, J.A. Alonso, W.G. Marshall, R.M. Ibberson, A. Podlesnyak, M.J. Mart´ ınez-Lope, M.M. Abd-Elmeguid, Phys. Rev. Lett. 98, 17, 176406 (2007)

work page 2007

[7] [6]

Moskvin, J

A.S. Moskvin, J. Phys.: Condens. Matter 25, 8, 085601 (2013). 14

work page 2013

[8] [7]

Moskvin, Yu.D

A.S. Moskvin, Yu.D. Panov. Phys. Solid State 65, 7, 1129 (2023)

work page 2023