Pith. sign in

REVIEW 3 major objections 3 minor 38 references

Approximate thermodynamics of the two-dimensional Ising model in an external magnetic field

T0 review · 3 major / 3 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read This paper claims that the two-dimensional Ising model in an external magnetic field, at high temperature and weak field, has its free energy given by a closed expression built from complete elliptic integrals and hyperbolic functions, with

desk verdict A genuine but uncontrolled approximation: the field-dependent closed forms are new and the H=0 limit is exactly Onsager, but the ψ→4ψ rescale is fit to numerics and the zero-field susceptibility is wrong at first order in βJ. read the letter →

arxiv 2607.16735 v1 pith:M46CDD6U submitted 2026-07-18 math-ph cond-mat.stat-mechmath.MP

Mathieu Gaudreault This is my paper · ORCID
classification math-phcond-mat.stat-mechmath.MP MSC 82B2082B2682B80
keywords two-dimensionalIsingmodelexternalmagneticfieldquaternionrepresentationsmall-angleapproximationtransfermatrixfreeenergymagnetizationsusceptibility
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 two-dimensional Ising model in a magnetic field has no exact solution, and the symmetry-broken transfer matrix resists the methods used for the zero-field case. This paper proposes that a small-angle rotation in the quaternion representation of the model produces a centrosymmetric transfer matrix whose largest eigenvalue can be written down explicitly, yielding a closed-form approximate free energy. From that free energy the paper derives analytic expressions for magnetization, susceptibility, internal energy, and specific heat that depend on the field strength. The paper reports that these expressions match BKL simulations in the regime T ≫ 2(H + J) and reduce exactly to the Onsager results when the field is zero. The importance lies in providing the first field-dependent closed-form thermodynamics for the model in any region, however restricted.

What carries the argument

The quaternion (Pauli-matrix) representation of the transfer matrix, in which the field-dependent single-spin operators are rotated by an angle φ relative to the coupling operators. The approximation takes φ → 0, dropping the tan(φ) coupling terms while rescaling ψ → 4ψ, so the transfer matrix becomes centrosymmetric and can be diagonalized with Kaufman's plane-rotation machinery. The resulting largest eigenvalue depends on the field through the effective parameters η and ωa, yielding the free energy and all derived thermodynamic quantities.

What would settle it

Compute the exact largest eigenvalue of the full 2D transfer matrix for a finite lattice (e.g., N = 20) at values of βJ and βH where φa is small but not tiny (say φa = 0.1), and compare it to the largest eigenvalue of T_a; if the relative difference does not vanish at a rate consistent with the claimed O(φ) truncation, the approximation's validity region is not controlled in the way the paper states.

Watch

Extended reading notes

Core claim

The central claim is that, for φa = arctan[sinh(2βH) exp(2βJ)] ≪ 1, the free energy of the 2D Ising model in a magnetic field is βFa = −log[2 cosh(2η) cosh(2ψ)] − (1/π)∫₀^{π/2} dθ log[½(1 + √(1 − κ² sin²θ))], where tanh η = [cosh(2ψ)e^{2βJ} − 1]/[cosh(2ψ)e^{2βJ} + 1], κ = 2 sinh(2η)/cosh²(2η), and ψ = βH. The paper derives the magnetization, susceptibility, internal energy, and specific heat as derivatives of this expression, correcting the magnetization by a factor of 1/4 to account for the small-angle rescaling ψ→4ψ. It claims these formulas agree with BKL simulations in the stated regime and reduce exactly to the known H = 0 results.

Load-bearing premise

The load-bearing premise is that the small-angle truncation (φ → 0, tan φ = 0) together with the rescaling ψ → 4ψ leaves the largest eigenvalue of the exact field-dependent transfer matrix essentially unchanged, an assumption supported only by numerical agreement and never by a derived error bound.

Editorial extensions

If this is right

  • If correct, the four thermodynamic functions (magnetization, susceptibility, internal energy, specific heat) are available in closed form, explicitly dependent on the external field, in the high-temperature weak-field regime.
  • The expressions reduce exactly to the Onsager free energy, internal energy, and specific heat at H = 0, providing a smooth interpolation between the known zero-field exact results and the field-dependent approximation.
  • The approximation defines a precise validity criterion, φa ≪ 1, equivalently T ≫ 2(H + J), which the paper checks against BKL simulations for J = H and J ≠ H.
  • The mechanism, applied to the one-dimensional model in the supplementary material, yields closed-form approximations that also reduce to the exact paramagnetic limit when J = 0.
  • The paper suggests the same construction can be extended to anisotropic couplings, other two-dimensional lattice types, and—if a quaternion representation exists—possibly the three-dimensional model.

Reading between the lines

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

  • The validity region is empirical: no error bound is derived for the φ → 0 truncation, so the claimed regime is defined by observed simulation agreement rather than by a controlled expansion, and the method is unlikely to be reliable near the critical temperature or at moderate fields.
  • The empirical factor of 4 in the ψ rescaling (and 2 in one dimension) suggests a coordination-number-dependent normalization; testing whether φa = arctan[sinh(2βH) exp(2βJ)] remains the correct angle for lattices with coordination numbers other than 4 would be a natural next step.
  • The closest existing benchmark for the susceptibility, the high-temperature perturbative series, is compared in the paper and found to be more accurate in some ranges; a head-to-head test of the two approximations against high-precision simulations at φa just below the stated validity threshold could separate truncation error from field-rescaling error.
  • The machinery might be extended to produce approximate spin-spin correlation functions, since the plane-rotation diagonalization used for the eigenvalue also yields eigenvectors, though the paper does not attempt this.
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 / 3 minor

Summary. The manuscript proposes approximate closed-form thermodynamics for the 2D Ising model in a magnetic field. Using the quaternion/Kaufman transfer-matrix representation, the author applies a 'small-angle' truncation (Sec. III.B) and a field rescale psi -> 4 psi; the resulting centrosymmetric transfer matrix is then diagonalized by Kaufman's method, giving the approximate free energy beta F_a in Eq. (51). Derivatives yield approximate expressions for M, chi, U, and C (Eqs. 52, 55, 57, 58), which reduce to the Onsager results at H=0 and are compared with BKL Monte Carlo data. A supplementary section applies the same construction to the 1D Ising model. The paper states the approximation is valid when phi_a << 1 and acknowledges failures near criticality.

Significance. If the derivation were controlled, this would be a useful contribution: explicit field-dependent formulas in a regime without an exact solution, with an exact H=0 limit, analytic derivatives, and direct comparison to simulations in a high-temperature weak-field window. The algebraic path from Eqs. (13)-(45) is coherent, and the Monte Carlo comparisons at J=0.01 show that the formulas capture the magnitudes of M, chi, U, and C in the tested regime. However, as detailed below, the small-angle truncation is uncontrolled, the field rescale is calibrated rather than derived, and the 1D companion section contains an algebraic error. The paper is therefore better viewed as a promising phenomenological approximation than as a derivation of the model's thermodynamics; in its current form it needs substantial revision.

major comments (3)
  1. [Sec. III.B, Eqs. (32), (36), (60)] The small-angle step is not a controlled truncation. Setting tan phi = 0 in Eq. (32) removes a term whose H-derivative at H=0 is nonzero; rescaling psi -> 4 psi in Eq. (36) cannot restore it. Consequently, the zero-field susceptibility Eq. (60) expands to beta chi_a = 1 + beta J + O((beta J)^3), while the exact square-lattice high-temperature series is beta chi = 1 + 4 v + ... = 1 + 4 beta J + ... . The leading interaction correction is missed by a factor of four. Moreover, phi_a = 0 identically at H=0 for any beta J, so the stated criterion phi_a << 1 does not imply beta J << 1; yet Eq. (60) is not exact at finite beta J. No error bound or next-order term is provided, so the field derivatives of Eq. (51) are not controlled in the claimed validity region.
  2. [Sec. III.B, Eq. (36); Sec. VI] The factor 4 in psi -> 4 psi (and 2 in 1D) is inserted after the truncation, and Sec. VI states it was 'necessary for the approximate expressions to agree with the numerical simulations.' This makes the factor a fitted parameter rather than a consequence of the approximation. The same numerics are also used to define the validity region, so agreement cannot independently validate the central claim. The paper should either derive this factor from a systematic expansion, or be reframed explicitly as a calibrated/phenomenological approximation.
  3. [Supplementary Material, Eqs. (C34)-(C37)] Eq. (C35) is not the derivative of Eq. (C34). Differentiation gives chi = 2 beta e^(2 beta J) [cosh(2 beta H)+e^(2 beta J)] / [1+e^(2 beta J) cosh(2 beta H)]^2, not the displayed expression. At J=0 the correct derivative is 2 beta / [1+cosh(2 beta H)], whereas Eq. (C35) gives 2 beta / [1+cosh(2 beta H)]^3; the exact value is beta sech^2(beta H). The sentence after Eq. (C37) claiming exact J=0 agreement is therefore false. This error in the companion validation should be corrected.
minor comments (3)
  1. [Secs. IV-V] The Monte Carlo comparisons give absolute-difference thresholds but no statistical errors or number of independent runs; finite-size effects between n=20 and n=50 are not quantified. Please add error bars or state uncertainties.
  2. [Sec. VI] The kappa' = 0 divergence is described as a 'sharp transition between ordered and disordered phase,' but for H != 0 the exact model has no phase transition; the text should call this an artifact of the approximation.
  3. [Sec. III.B, Eq. (36)] 'phi_a small implies 2 beta J << 1 and 2 beta H << 1' is not a valid implication when H=0 (phi_a=0 for any beta J); the condition T >> 2(H+J) should be justified separately.

Circularity Check

1 steps flagged · score 4.0 of 10

The ψ→4ψ rescale is a numerical-agreement-motivated constant, so the field-dependent formulas are partly calibrated to the BKL benchmark; the H=0 Onsager reduction and the non-fitted functional form provide independent content.

  1. fitted input called prediction [Sec. III.B (Eqs. 35-39); Sec. VI Discussion]
    "To ensure that the system responds to the full physical field H in the limit H → 0, we rescale ψ → 4ψ in Eq. (35). ... In this work, the external magnetic field strength was rescaled by a factor of 4 in two dimensions ... These factors were necessary for the approximate thermodynamic expressions to agree with the numerical simulations."

    The factor 4 is not derived from the quaternion rotation; it is a free calibration constant chosen, by the author's own account, so the approximate expressions match the BKL simulations (and the compensating 1/4 in Ma and Ua is then required by the chain rule). The same BKL data are used as the validation of Eqs. (52), (55), (57), (58), and the validity window φa<0.01–0.1 is read off from where agreement occurs. Hence the field-dependent predictions are partly fitted to the benchmark they are claimed to confirm. The H=0 Onsager limit and the non-fitted functional form are the independent parts.

full rationale

The only concrete circular element is the ad hoc ψ→4ψ rescale (and its 1/4 derivative correction), which the paper itself says was necessary for agreement with the BKL numerics; using those same numerics to validate the resulting formulas makes the agreement partly by construction. This prevents a 0-2 score. However, the derivation is not reducible to the fit: the free-fermion/Kaufman diagonalization gives a nontrivial H-dependence, the H=0 limit is exactly Onsager, and the formulas at small field reproduce the paramagnetic response. The more serious concerns raised by the skeptic—the discarded tan φ term has a nonzero H-derivative at H=0, Eq. (60) misses the 4v term of the exact high-temperature susceptibility expansion, and no error bound is given—are accuracy/validity defects, not circularity, and I do not score them as such. No self-citation chain is load-bearing. Overall: partial circularity from a calibrated input, with the central claim retaining independent content.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The derivation leans on one hand-set constant — the ψ→4ψ (2D) and ψ→2ψ (1D) rescaling — which the author states was necessary to make the approximate expressions agree with the simulations (Sec. VI). The validity thresholds (φa < 0.01–0.1 and the error budgets in Sec. V) are declared from the agreement itself. Standard mathematical machinery (Onsager's integral identity, Kaufman diagonalization) and the BKL simulation setup are imported from the literature; the small-angle truncation and the rescale are the paper's own assumptions, and neither carries a quantitative error estimate.

free parameters (2)
  • global field-rescale factor for ψ (ψ→4ψ in 2D; ψ→2ψ in 1D) = 4 (2D), 2 (1D)
    Chosen so that the approximate formulas reproduce the physical field response; the author states these factors 'were necessary for the approximate thermodynamic expressions to agree with the numerical simulations' (Sec. VI). Partially principled via the per-bond H/4 splitting of Eqs. 6–7, but the value is tuned to agreement.
  • validity thresholds and error budgets defining the agreement region = φa < 0.01 / 0.1 / 0.03; abs. differences < 1 / 0.22 / 0.03 (Sec. V)
    The validation region is declared post hoc from where the curves happen to agree; no a priori error estimate ties these thresholds to the truncation order.
assumptions (5)
  • standard math Onsager identity |z| = (1/π)∫₀^π dτ log[2cosh z − 2cos τ], used to reduce the double integral I
    Invoked at Eq. (48) and in Appendix B. Standard (Onsager 1944), but the domain of the acosh argument dropping below 1 (complex branch) is not addressed; the final real form is asserted by analytic continuation without proof.
  • domain assumption The reduced transfer matrix Ta (Eq. 43) is in the same form as the transfer matrix of the 2D model with H=0, so Kaufman's plane-rotation diagonalization applies and the largest eigenvalue is known
    Invoked in Sec. III.B to jump from Eq. (43) to the free energy Eq. (46); relies on prior literature [19] and on the centrosymmetric structure of Ta.
  • domain assumption BKL Monte Carlo on N=20–50 lattices with 100,800 configurations approximates the thermodynamic limit well enough to judge the analytic formulas
    Secs. IV–V; finite-size effects and Monte Carlo statistical error are not quantified (no error bars, no seed, no extrapolation).
  • ad hoc to paper The field rescale ψ→4ψ restores the full physical field coupling after the small-angle truncation
    Sec. III.B: 'To ensure that the system responds to the full physical field H in the limit H→0, we rescale ψ→4ψ'; the author later acknowledges the factor was necessary for numerical agreement (Sec. VI, Discussion).
  • ad hoc to paper The small-angle condition φa ≪ 1 (T ≫ 2(H+J)) is a valid small parameter for the exact transfer matrix; dropped terms are O(φ) and negligible
    Eq. (36) and surrounding text; this is the central truncation of the paper, and no error bound or next-order term is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Approximate thermodynamics of the two-dimensional Ising model in an external magnetic field." pith.science (2026). https://pith.science/paper/M46CDD6U

@misc{pith2026260716735,
  author       = {Pith},
  title        = {Pith review of: Approximate thermodynamics of the two-dimensional Ising model in an external magnetic field},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M46CDD6U}},
  note         = {Machine review of arXiv:2607.16735}
}
read the original abstract

We introduce approximate expressions for the thermodynamics of the two-dimensional Ising model in an external magnetic field. The external field breaks the system's symmetry, which complicates the exact calculation of the free energy. To address this, we write the model in its quaternion representation to apply a small angle approximation with respect to a reference frame. The approximation simplifies the transfer matrix to a centrosymmetric matrix for which the largest eigenvalue is known and depends on the external field strength. We therefore write an approximate expression for the system's free energy and derive the magnetization, susceptibility, internal energy, and specific heat. These results are compared with numerical simulations performed with the BKL algorithm. We find that the approximate expressions and numerical simulations agree in regions where the approximation is valid.

Figures

Figures reproduced from arXiv: 2607.16735 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p022_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p023_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p024_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p032_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5 [PITH_FULL_IMAGE:figures/full_fig_p033_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6 [PITH_FULL_IMAGE:figures/full_fig_p034_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7 [PITH_FULL_IMAGE:figures/full_fig_p035_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8 [PITH_FULL_IMAGE:figures/full_fig_p036_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references

  1. [1]

    Ising, Z

    E. Ising, Z. Phys. 31, 253 (1925)

    1925

  2. [2]

    Chandler, Introduction to Modern Statistical Mechanics (Oxford University Press, London,

    D. Chandler, Introduction to Modern Statistical Mechanics (Oxford University Press, London,

  3. [3]

    Onsager, Phys

    L. Onsager, Phys. Rev. 65, 117 (1944)

    1944

  4. [4]

    Odor, Rev

    G. Odor, Rev. mod. Phys. 76, 663 (2004)

    2004

  5. [5]

    Binder and K

    K. Binder and K. Schroder, Phys. Rev. B 14, 2142 (1976)

    1976

  6. [6]

    V. S. Pande, A. Y. Grosberg, and T. Tanaka, Biophys. J. 73, 3192 (1997)

    1997

  7. [7]

    Gaudreault and J

    M. Gaudreault and J. Vinals, Phys. Rev. E 80, 021916 (2009)

    2009

  8. [8]

    Mello, L

    I. Mello, L. Squillante, G. O. Gomes, A. C. Seridonio, and M. de Souza, Physica A 573, 125963 (2021)

    2021

Show all 38 references
  1. [9]

    Torquato, Phys

    S. Torquato, Phys. Biol. 8, 015017 (2011)

    2011

  2. [10]

    D. R. Chialvo, Nature Physics 6, 744 (2010)

    2010

  3. [11]

    Sornette, Rep

    D. Sornette, Rep. Prog. Phys. 77, 062001 (2014)

    2014

  4. [12]

    B. M. McCoy and T. T. Wu, Phys. Rev. D 18, 1259 (1978)

    1978

  5. [13]

    Zamolodchikov and I

    A. Zamolodchikov and I. Ziyatdinov, Nucl. Phys. B 849, 654 (2011)

    2011

  6. [14]

    K. A. Meissner, D. Ircha, W. Olszewski, J. Ruta, , and A. Slapek, J. Stat. Phys. 190, 1 (2023)

    2023

  7. [15]

    Belavin, A

    A. Belavin, A. Polyakov, and A. B. Zamolodchikov, Nucl. Phys. B 241, 333– (1984)

    1984

  8. [16]

    Xu and A

    H.-L. Xu and A. Zamolodchikov, J. High Energy Phys. (57), 1

  9. [17]

    X.-P. Kong, H. Au-Yang, and J. H. H. Perk, Phys. Rev. A 116, 54 (1986)

    1986

  10. [18]

    W. P. Orrick, B. Nickel, A. J. Guttmann, and J. H. H. Perk, Phys. Rev. Lett 86, 4120 (2001)

    2001

  11. [19]

    Kaufman, Phys

    B. Kaufman, Phys. Rev. 76, 1232 (1949)

    1949

  12. [20]

    B. M. McCoy and T. T. Wu, The Two-Dimensional Ising Model (Harvard university press, 1973)

    1973

  13. [21]

    R. J. Baxter, Exactly solved models in statistical mechanics (Academic Press, London, 1982) p. 502

    1982

  14. [22]

    Huang, Statistical mechanics (John Wiley and sons, New York, 1987)

    K. Huang, Statistical mechanics (John Wiley and sons, New York, 1987)

    1987

  15. [23]

    A. B. Bortz, M. H. Kalos, and J. L. Lebowitz, J. Comput. Phys. 17, 10 (1975)

    1975

  16. [24]

    W. P. Orrick, B. Nickel, A. J. Guttmann, and J. H. H. Perk, J. Stat. Phys. 102, 795 (2001)

    2001

  17. [25]

    T. T. Wu, B. M. McCoy, C. A. Tracy, and E. Barouch, Phys. Rev. B 13, 316 (1976)

    1976

  18. [26]

    Kaufman and L

    B. Kaufman and L. Onsager, Phys. Rev. 76, 1244 (1949)

    1949

  19. [27]

    R. M. F. Houtappel, Physica 16, 425 (1950)

    1950

  20. [28]

    Hucht, J

    A. Hucht, J. Phys. A: Math. Theor. 86, 065201 (2017)

    2017

  21. [29]

    T. D. Schultz, D. C. Mattis, and E. H. Lieb, Rev. Mod. Phys. 36, 856 (1964)

    1964

  22. [30]

    P. W. Kasteleyn, Physica 27, 1209 (1961). 20

    1961

  23. [31]

    C. N. Yang, Phys. Rev. 85, 808 (1952)

    1952

  24. [32]

    R. J. Baxter, J. Stat. Phys. 145, 518 (2011)

    2011

  25. [33]

    Abramowitz and I

    M. Abramowitz and I. A. Stegun (Eds.), Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables (Dover Publications Inc., New York, 1965) p. 1046. 21 Algorithm used in BKL sampling - Assume that the transition probabilities Pj are known. - Calculate ra...

    1965

  26. [34]

    1 0 0 1 # , X =

    One-dimensional Ising model with the quaternion algebra Consider a linear chain of n spins with state S = ±1 and periodic boundary conditions Sn+1 =S1. Let E be the interaction energy of the linear chain, defined by E = −J nX i=1 SiSi+1 − H 2 nX i=1 (Si +Si+1) . (C12) The part...

  27. [35]

    We rescale H → 2H to account for the two solutions of Eq

    Approximate thermodynamics of the one-dimensional Ising model In this section, we introduce the small angle approximation and apply it to the one- dimensional model. We rescale H → 2H to account for the two solutions of Eq. ( C23). In these terms, the small angle approximation...

  28. [36]

    Let S be the sum of all spin states of a configuration normalized by the 29 total number of spins computed as S = 1 n nX i=1 Si

    Numerical simulations The one-dimensional Ising model was simulated by using the BKL algorithm as described in the manuscript. Let S be the sum of all spin states of a configuration normalized by the 29 total number of spins computed as S = 1 n nX i=1 Si . (C38) The magnetizat...

  29. [38]

    ( C34) - ( C37)] are compared with numerical simulations as the angle φa is varied in Fig

    Comparison between analytical and numerical results The approximate thermodynamic expressions [Eqs. ( C34) - ( C37)] are compared with numerical simulations as the angle φa is varied in Fig. 4. The cases where J =H andJ >H were considered. The agreement between the approximate...

  30. [103]

    A total of 96,000 configurations were simulated

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.