Finite temperature fermion Monte Carlo simulations of frustrated spin-Peierls systems

Fakher F. Assaad; Jeroen van den Brink; Jo\~ao C. In\'acio; Toshihiro Sato

REVIEW 4 major objections 6 minor 43 references

Finite temperature fermion Monte Carlo simulations of frustrated spin-Peierls systems

T0 review · 4 major / 6 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read Fermion-based Monte Carlo handles spin-Peierls systems without worsening the sign problem.

desk verdict Sound extension of Abrikosov-fermion AFQMC to spin-Peierls systems, with a useful sign-problem observation but thin numerical validation in the adiabatic regime it claims to enable. read the letter →

arxiv 2502.06565 v1 pith:B7JRRXUL submitted 2025-02-10 cond-mat.str-el

João C. Inácio , Jeroen van den Brink , Fakher F. Assaad , Toshihiro Sato This is my paper · ORCID

classification cond-mat.str-el

keywords auxiliary-fieldquantumMonteCarloAbrikosovfermionsspin-PeierlssystemsnegativesignproblemKitaev-HeisenbergmodelEinsteinphononsspinliquidHeisenberg

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 extends a fermion-based auxiliary-field quantum Monte Carlo method to spin-Peierls models where every magnetic bond is coupled to an Einstein phonon. Its central claim is that the addition of phonons, in the physically relevant adiabatic limit, does not make the negative sign problem more severe. If correct, this enables exact finite-temperature simulations of frustrated spin-phonon systems down to temperatures about twice below the magnetic energy scale, a regime where world-line methods fail because of the sign problem. The method is benchmarked against the SSE wormhole algorithm on a Heisenberg chain and then applied to the Kitaev-Heisenberg model on the honeycomb lattice.

What carries the argument

The central object is the Abrikosov fermion representation, where each spin-1/2 is written as a two-component fermion with a single-occupancy constraint imposed by a large Hubbard-U term. Squared fermion bilinears from spin-spin interactions are decoupled with discrete Hubbard-Stratonovich fields via Gauss-Hermite quadrature, while the bond phonons are kept as continuous fields $\phi_{b,\tau}$ on each imaginary-time slice. This machinery carries the argument because it turns the spin-phonon Hamiltonian into a fermion determinant problem whose sign can be optimized over a manifold of equivalent actions, and because the cost per Monte Carlo sweep does not scale with the phonon Hilbert space. The key mechanism controlling the sign is the realness of $\sqrt{1+\phi_{b,\tau}}$: as long as phonon fluctuations keep this factor real, the average sign stays close to its phonon-free value.

What would settle it

Run the Kitaev-Heisenberg AFQMC at $\lambda=0.2$, $\omega_0=0.75$, $T/A=1/1.8$ on a 32-site honeycomb cluster from many independent seeds and compare the estimated uniform susceptibility and average sign; if the estimates do not converge or the phonon autocorrelation time exceeds the simulated $10^4$ sweeps, the premise that local updates produce unbiased results fails. Alternatively, a global phonon update that changes all time slices at once should reproduce the same susceptibility, and a discrepancy would show the local-update sampling is biased.

Watch

Extended reading notes

Core claim

Using Abrikosov fermions with a Hubbard-U constraint and decoupling the spin interactions with Hubbard-Stratonovich fields, the paper writes the partition function of a generic spin-Peierls Hamiltonian with bond phonons as a sum over discrete auxiliary fields and a path integral over continuous phonon displacements. The phonon fields are sampled with local updates on the imaginary-time lattice. The central result is that for the frustrated Kitaev-Heisenberg model at temperature $T/A=1/1.8$ on a 32-site honeycomb cluster, the average sign with bond phonons at coupling $\lambda<0.2$ and frequency $\omega_0=0.75$--$1.0$ stays essentially as large as without phonons. The sign problem only develops when $\sqrt{1+\phi_{b,\tau}}$ becomes imaginary, which occurs for large phonon frequency fluctuations. The authors interpret this as evidence that including optical phonons in the adiabatic limit does not significantly worsen the sign problem, so the same low-temperature regime remains accessible.

Load-bearing premise

The method's quantitative results assume that the local single-spin-flip update for the continuous phonon fields is ergodic and that autocorrelation times of $10^3$--$10^4$ sweeps are short enough to make measured susceptibilities unbiased; if phonon dynamics freeze on longer timescales at lower temperatures, stronger coupling, or larger systems, the reported numbers could be biased even though the sign stays large.

Editorial extensions

If this is right

Exact finite-temperature QMC for frustrated spin-Peierls systems becomes possible down to $T/A \sim 1/1.8$ at low phonon coupling in the adiabatic regime.
The per-sweep computational cost does not grow when phonons are added; only the autocorrelation time increases, so the method can be pushed to larger lattices.
For $\lambda<0.2$ and $\omega_0=0.75$--$1.0$, the average sign remains nearly as large as in the phonon-free Kitaev-Heisenberg model, keeping the same temperature scales accessible.
On a non-frustrated Heisenberg chain, the AFQMC structure factor matches the SSE wormhole benchmark, validating the phonon-sampling procedure.
Antiferromagnetic Heisenberg chains and square lattices show precursors of a spin-Peierls gap (reduced uniform susceptibility) that ferromagnetic systems do not show, in line with bosonization and parton arguments.

Reading between the lines

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

If the adiabatic-limit claim holds, phonon spectral anomalies such as Landau damping would be most observable where Heisenberg exchange dominates, since the Kitaev point and ferromagnetic sectors remain effectively decoupled from phonons at the simulated temperatures.
The sign only degrades when $\sqrt{1+\phi_{b,\tau}}$ becomes imaginary, so alternative phonon parametrizations or shifted integration contours might extend the method to higher $\omega_0$ and stronger $\lambda$ while preserving the sign.
The strong autocorrelation growth with phonon coupling suggests that global phonon updates, such as hybrid Monte Carlo, could reduce the computational cost of generating independent configurations and should be tested as a direct extension.

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

You should know two things up front. First, the central algorithmic step is a straightforward extension of the authors' own Abrikosov-fermion AFQMC (Refs. 11,12) to bond Einstein phonons, and they do it cleanly: the derivation is standard, the benchmark against an independent SSE wormhole code (Fig. 1) is a real check, and the observation that phonons do not worsen the sign problem at moderate coupling and frequency is a genuinely new numerical result. Second, the paper is less convincing where it needs to be most: the adiabatic regime that motivates the whole effort is exactly where the local phonon update is slowest, and the manuscript gives you no error bars, no binning analysis, no independent-seed comparison, and no autocorrelation data at the small omega0 values used in the main susceptibility figures. That is the soft spot, and it is a real one, not a manufactured quibble.

What is actually new: the generalization of the auxiliary-field method to include dynamical bond phonons, and the specific finding that, for the Kitaev-Heisenberg model at T/A=1/1.8, adding phonons with lambda<0.2 and omega0=0.75-1.0 leaves the average sign essentially unchanged. The SSE benchmark gives the core implementation credibility, and the physics discussion—spin-Peierls precursors in 1D and 2D Heisenberg, the decoupling of phonons from ferromagnetic and Kitaev-point susceptibilities—is plausible and appropriately hedged. The citations are honest; the only heavy self-reference is to the base AFQMC method, which is prior work and not used circularly.

Where it wobbles: the stress-test concern about ergodicity of the local phonon update holds up on reading. Eq. (9) has the phonon kinetic term with coefficient m proportional to 1/omega0^2, so as omega0 drops the imaginary-time coupling stiffens and a single-site Metropolis move becomes rare. The paper reports autocorrelation times of 10^3-10^4 sweeps at omega0=0.5 (Fig. 2) and then shows susceptibility data at omega0=0.5, 0.75, and 1.0 without any convergence diagnostics. The sign-problem claim itself is supported for the tested parameters, but the quantitative susceptibility curves—especially the gap-precursor interpretation in Figs. 5(a), 6(a), and 8—rest on the unexamined assumption that the phonon field is decorrelated. That is correctable with binning analysis and a few autocorrelation curves, but it is not a minor cosmetic gap; it is the difference between a method paper and a method paper that actually demonstrates its reach.

Who this is for: people working on frustrated spin-phonon models, particularly Kitaev materials, who want a finite-temperature QMC option beyond SSE. It deserves a serious referee: the algorithmic idea is sound, the benchmark is genuine, and the sign-problem observation is worth publishing even if the quantitative physics sections need more work. My recommendation is to send it to review, with the clear expectation that the authors add error bars, convergence checks, and autocorrelation data in the adiabatic regime, or soften the claims accordingly.

Referee Report

4 major / 6 minor

Summary. The manuscript extends the Abrikosov-fermion auxiliary-field quantum Monte Carlo (AFQMC) method of Refs. [11,12] to spin-Peierls systems with bond Einstein phonons. After a Trotter decomposition and Hubbard-Stratonovich transformation, the continuous bond-phonon fields are sampled with local single-spin-flip updates alongside the discrete auxiliary fields. The method is benchmarked against a wormhole SSE calculation for the one-dimensional Heisenberg chain, with and without spin-phonon coupling (Fig. 1). The authors then study the average sign for the Kitaev-Heisenberg model on the honeycomb lattice, reporting that weak phonon coupling (λ<0.2 at ω0=0.75–1.0) does not significantly worsen the sign problem at T/A=1/1.8 for N=32 (Fig. 3). The remainder of the paper presents temperature-dependent uniform susceptibilities for the Heisenberg chain, square-lattice Heisenberg model, and Kitaev-Heisenberg model, interpreting the spin-phonon effects in terms of bosonisation, parton mean-field theory, and Kitaev physics. The central claim is that the AFQMC approach can address frustrated spin-Peierls systems down to temperatures around twice the magnetic scale without the sign problem becoming more severe in the adiabatic limit.

Significance. If the central claim holds, this is a useful methodological advance: it would bring finite-temperature sign-problem-optimized AFQMC to spin-phonon Hamiltonians, a regime where Hilbert-space methods are severely limited and where existing world-line methods suffer from the sign problem in frustrated models. The manuscript reports an independent benchmark against SSE (Fig. 1), which is a genuine strength and rules out circularity in the implementation check. The paper also states its central limitation honestly: local phonon updates increase autocorrelation times. However, the evidence for the central claim is presently thin in several load-bearing places: no error bars on the main susceptibility figures, no Trotter or finite-U extrapolation, and no demonstration that the local phonon update is ergodic and unbiased at the parameters used for the physics results. These gaps prevent the paper from being accepted as it stands, but they are addressable and do not appear to require a fundamentally different method.

major comments (4)

[§III.A, Fig. 3] The central claim that phonons do not worsen the sign problem for the Kitaev-Heisenberg model is supported only by one system size (N=32), one temperature (T/A=1/1.8), and a narrow range of couplings (λ<0.2, ω0=0.75–1.0). The sign problem in fermion QMC generally scales with system size and inverse temperature, so a single parameter set cannot establish that 'the same temperature regime is accessible' for larger systems or lower temperatures. Please provide sign data as a function of system size and, if possible, at lower temperatures or a scaling analysis.
[§III.A, Fig. 2 and Eq. (9)] The local single-spin-flip update for the continuous bond-phonon fields is the main sampling bottleneck, but the manuscript provides no convergence or ergodicity analysis at the parameters used in the physics sections. In Eq. (9), the phonon kinetic term couples adjacent imaginary-time slices with coefficient m∝1/ω0², so as ω0→0 the proposed moves become increasingly costly. Fig. 2 reports autocorrelation times of 10^3–10^4 sweeps already at ω0=0.5, and Figs. 5–8 use ω0=0.5 and 0.75 without binning analysis, multiple independent seeds, or error bars. Without such diagnostics, the susceptibility curves and even the averaged sign could be biased by slow phonon dynamics.
[§II, Eq. (8) and §III.B] The paper invokes the limit U→∞ to enforce the single-occupancy constraint, but no information is given about the finite value of U used in the simulations, nor is any finite-U extrapolation shown. Similarly, no Trotter-step (Δτ) convergence check is presented for the susceptibility data. Since the quantitative results in Figs. 5, 6, and 8 depend on the projection onto the physical subspace and on the Trotter error, the absence of these extrapolations leaves the reported curves without a demonstrated systematic-error control.
[§III.B, Figs. 5, 6, 8] The main physical figures show no error bars, making it impossible to assess whether the differences between the phonon-coupled and phonon-free susceptibility curves are statistically significant. For example, in Fig. 5(a) the reduction of χ at ω0=0.5 appears monotonic, but without error estimates or a comparison of independent runs, it could be within statistical noise. Please provide error bars or an equivalent statistical analysis for all susceptibility data points.

minor comments (6)

[§III.A, text after Fig. 4] There is a typo in 'Fo a fixed spin-phonon coupling'—it should read 'For a fixed spin-phonon coupling.'
[§III.B.1, text after Eq. (15)] The phrase 'If the spin-phonon coupling λ is ufficiently large' contains a typo ('ufficiently' should be 'sufficiently').
[§III.B.3, text near Fig. 8(c)] The word 'magnetropic susceptibility' appears to be a typo; it should likely be 'magnetic susceptibility.'
[Reference [22]] The reference to 'axXiv:2412.11382' contains a typo: 'axXiv' should be 'arXiv.'
[Figs. 3 and 4] The average-sign plots would benefit from error bars or at least a statement of the statistical uncertainty, since the central conclusion is about the absence of a significant sign-problem increase.
[Fig. 2 caption] The caption says the autocorrelation time is 'as a function of the spin-phonon coupling constant λ', but the horizontal axis is t_{MC}; the figure shows autocorrelation curves for different λ values. Please clarify the wording.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method is benchmarked against an independent SSE wormhole algorithm and the central sign-problem claim is a direct measurement, not a fitted input.

full rationale

The paper's derivation chain is self-contained and does not reduce to its own inputs. The central algorithmic claim—that AFQMC can simulate spin-Peierls systems with Abrikosov fermions and bond Einstein phonons—is validated in Fig. 1 against an independent SSE wormhole algorithm provided by M. Weber, as acknowledged. The partition function in Eq. (8) and phonon action in Eq. (9) follow from the Hamiltonian in Eq. (1) via a standard Trotter decomposition and Hubbard-Stratonovich transformation; no parameter in these expressions is fitted to the quantities later reported. The second central claim, that including phonons at low coupling does not significantly worsen the negative sign problem, is a direct measurement of the average sign defined in Eq. (12) at different λ and ω0 values, not a prediction derived from a fitted constant. The comparison with the no-phonon case is an empirical comparison of two AFQMC runs, not a construction that forces the result. Self-citations to Refs. [11,12] supply the base AFQMC methodology and prior Kitaev-Heisenberg results, but these are not load-bearing in the sense of defining the present results or forbidding alternatives; no uniqueness theorem is imported. The reader's concern about ergodicity of the local phonon update and absence of binning/error-bar analysis is a correctness or methodological risk, not circularity, and therefore does not affect the circularity score.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central result is a numerical method built on standard AFQMC machinery. The ledger lists the numerical parameters and domain assumptions the simulations depend on. No new physical entities are introduced. The spin-phonon coupling and phonon frequency are physical model inputs, not fitted parameters.

free parameters (2)

Trotter time step Delta_tau = LTrotter=100 for Heisenberg, LTrotter=20-200 for Kitaev-Heisenberg
Imaginary-time discretization parameter; no extrapolation to Delta_tau to 0 is shown, so systematic Trotter error is an unquantified numerical parameter.
Hubbard U constraint projection = treated as effectively infinite
The Abrikosov single-occupancy constraint is imposed by a Hubbard U in the limit U to infinity; practical ALF runs use finite U, and finite-U systematic error is not discussed.

assumptions (4)

domain assumption The U to infinity projection onto the odd-fermion-parity sector exactly implements the single-occupancy constraint and commutes with the phonon-coupled Hamiltonian.
Invoked in Sec. II after Eq. (6); if inexact at finite U or finite Trotter step, the method simulates a different model.
standard math The discrete Hubbard-Stratonovich transformation with four Gauss-Hermite quadrature points exactly decouples the squared fermion bilinears for each imaginary-time slice.
Standard AFQMC ingredient from Ref. 7; not re-derived here.
domain assumption The average sign computed by sampling Re[e^{-S}] divided by |Re[e^{-S}]| is an unbiased estimator of the severity of the sign problem.
Standard reweighting in AFQMC; relies on the partition function being real.
domain assumption The continuum phonon field, sampled in position basis with local updates, is ergodic on the simulated lattice sizes and temperatures.
Assumed implicitly; autocorrelation times are reported but no ergodicity or convergence proof is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Finite temperature fermion Monte Carlo simulations of frustrated spin-Peierls systems." pith.science (2026). https://pith.science/paper/B7JRRXUL

@misc{pith2026250206565,
  author       = {Pith},
  title        = {Pith review of: Finite temperature fermion Monte Carlo simulations of frustrated spin-Peierls systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B7JRRXUL}},
  note         = {Machine review of arXiv:2502.06565}
}

read the original abstract

The Abrikosov fermion representation of the spin-1/2 degree of freedom allows for auxiliary-field quantum Monte Carlo simulations of frustrated spin systems. This approach provides a manifold of equivalent actions over which the negative sign problem can be optimised. As a result, we can reach temperature scales well below the magnetic scale. Here, we show how to generalise this algorithm to spin-Peierls systems. In contrast to exact diagonalisation approaches, Monte Carlo methods are not Hilbert space bound such that the computational effort per sweep remains invariant when adding phonons. However, the computational effort required to generate independent configurations increases in the presence of phonons. We also show that, for the specific case of the Kitaev-Heisenberg model, the inclusion of phonons does not render the negative sign problem more severe. This new algorithm hence allows us to investigate the interplay between phonon degrees of freedom and magnetic frustration. We present results for frustrated and non-frustrated spin systems.

Figures

Figures reproduced from arXiv: 2502.06565 by the authors.

**Figure 2.** FIG. 2. Autocorrelation time for the equal time correlation [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** FIG. 3. Proposed ground-state phase diagram for the original [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: FIG. 4. Average sign as a function of the phonon frequency [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

**Figure 6.** Figure 6: FIG. 6. Temperature dependence of the uniform magnetic [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

**Figure 7.** Figure 7: FIG. 7. Momentum resolved static dimer structure factor for [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]

**Figure 8.** Figure 8: FIG. 8. Temperature dependence of the uniform magnetic susceptibility for the Kitaev-Heisenberg model with phonons at [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 35 canonical work pages

[1]

The results are presented for antiferromagnetic (a) and ferromagnetic (b) interactions

Heisenberg Model in One-Dimension Figure 5 shows the magnetic susceptibility as a func- tion of temperature for the Heisenberg chain without (black line) and with phonons at spin-phonon coupling λ = 0.2 for different phonon frequencies. The results are presented for antiferromagnetic (a) and ferromagnetic (b) interactions. Immediately we see a noticeable ...

work page
[2]

Heisenberg Model in Two-Dimensions We now consider the same Heisenberg model but on a two-dimensional square lattice. Figure 6 presents the temperature dependence of the magnetic susceptibil- ity for the antiferromagnetic (a) and ferromagnetic (b) Heisenberg models on the square lattice with λ = 0 .2 and various ω0 values. We again observe that, in the fe...

work page
[3]

Kitaev-Heisenberg model We now turn to the results and discussion for the frus- trated Kitaev-Heisenberg model on the honeycomb lat- tice. In Fig. 8 we present the temperature dependence of the magnetic susceptibility for different values of ϕ/π. Unlike the previous section, here we use a smaller value of the electron-phonon coupling λ = 0.1. We again ado...

work page
[4]

Balents, Nature 464, 199 (2010)

L. Balents, Nature 464, 199 (2010)

2010
[5]

G. D. Mahan, Many-Particle Physics, 2nd ed. (Plenum Press, New York, 1990)

work page 1990
[6]

L. J. Sandilands, Y. Tian, K. W. Plumb, Y.-J. Kim, and K. S. Burch, Phys. Rev. Lett. 114, 147201 (2015)

2015
[7]

Blankenbecler, D

R. Blankenbecler, D. J. Scalapino, and R. L. Sugar, Phys. Rev. D 24, 2278 (1981)

1981
[8]

White, D

S. White, D. Scalapino, R. Sugar, E. Loh, J. Gubernatis, and R. Scalettar, Phys. Rev. B 40, 506 (1989)

work page 1989

Show all 43 references

[9]

Assaad and H

F. Assaad and H. Evertz, in Computational Many- Particle Physics , Lecture Notes in Physics, Vol. 739, edited by H. Fehske, R. Schneider, and A. Weiße (Springer, Berlin Heidelberg, 2008) pp. 277–356

2008
[10]

F. F. Assaad, M. Bercx, F. Goth, A. G¨ otz, J. S. Hofmann, E. Huffman, Z. Liu, F. P. Toldin, J. S. E. Portela, and J. Schwab, SciPost Phys. Codebases , 1 (2022)

2022
[11]

Sylju ˚ asen and A

O. Sylju ˚ asen and A. Sandvik, Phys. Rev. E66, 046701 (2002)

2002
[12]

Weber, F

M. Weber, F. F. Assaad, and M. Hohenadler, Phys. Rev. Lett. 119, 097401 (2017)

2017
[13]

Weber, Phys

M. Weber, Phys. Rev. B 105, 165129 (2022)

2022
[14]

Sato and F

T. Sato and F. F. Assaad, Phys. Rev. B 104, L081106 (2021)

2021
[15]

T. Sato, B. J. Ramshaw, K. A. Modic, and F. F. Assaad, Phys. Rev. B 110, L201114 (2024)

2024
[16]

Stolpp, T

J. Stolpp, T. K¨ ohler, S. R. Manmana, E. Jeckelmann, F. Heidrich-Meisner, and S. Paeckel, Computer Physics Communications 269, 108106 (2021)

2021
[17]

Shi and S

H. Shi and S. Zhang, Phys. Rev. E 93, 033303 (2016)

2016
[18]

Ulybyshev and F

M. Ulybyshev and F. Assaad, Phys. Rev. E 106, 025318 (2022)

2022
[19]

G¨ otz, S

A. G¨ otz, S. Beyl, M. Hohenadler, and F. F. Assaad, Phys. Rev. B 105, 085151 (2022)

2022
[20]

F. F. Assaad and T. C. Lang, Phys. Rev. B 76, 035116 (2007)

2007
[21]

Karakuzu, K

S. Karakuzu, K. Seki, and S. Sorella, Phys. Rev. B 98, 201108 (2018)

2018
[22]

G¨ otz, M

A. G¨ otz, M. Hohenadler, and F. F. Assaad, Phys. Rev. B 109, 195154 (2024)

2024
[23]

Chaloupka, G

J. Chaloupka, G. Jackeli, and G. Khaliullin, Physical Review Letters 110, 097204 (2013), publisher: American Physical Society

2013
[24]

S. Beyl, F. Goth, and F. F. Assaad, Phys. Rev. B 97, 085144 (2018)

2018
[25]

Huang, N

C. Huang, N. Parthenios, M. Ulybyshev, X. Zhang, F. F. Assaad, L. Classen, and Z. Y. Meng, axXiv:2412.11382 (2024), arXiv:2412.11382 [cond-mat.str-el]

2024 arXiv
[26]

Troyer and U.-J

M. Troyer and U.-J. Wiese, Physical Review Letters 94, 170201 (2005), publisher: American Physical Society

2005
[27]

Weber, A

L. Weber, A. Honecker, B. Normand, P. Corboz, F. Mila, and S. Wessel, SciPost Physics 12, 054 (2022)

2022
[28]

Hangleiter, I

D. Hangleiter, I. Roth, D. Nagaj, and J. Eisert, Sci- ence Advances 6, eabb8341 (2020), publisher: American Association for the Advancement of Science

2020
[29]

F. Alet, K. Damle, and S. Pujari, Physical Review Let- ters 117, 197203 (2016), publisher: American Physical Society

2016
[30]

Reingruber, N

A. Reingruber, N. Caci, S. Wessel, and J. Richter, Phys- ical Review B 109, 125120 (2024), publisher: American Physical Society

2024
[31]

Kitaev, Annals of Physics 321, 2 (2006)

A. Kitaev, Annals of Physics 321, 2 (2006)

2006
[32]

Giamarchi, Quantum Physics in One Dimension(Ox- ford University Press, 2003)

T. Giamarchi, Quantum Physics in One Dimension(Ox- ford University Press, 2003)

2003
[33]

Hofmeier, J

D. Hofmeier, J. Willsher, U. F. P. Seifert, and J. Knolle, Physical Review B 110, 125130 (2024), publisher: Amer- ican Physical Society. 10

2024
[34]

Saremi and P

S. Saremi and P. A. Lee, Physical Review B 75, 165110 (2007), publisher: American Physical Society

2007
[35]

Dalla Piazza, M

B. Dalla Piazza, M. Mourigal, N. B. Christensen, G. J. Nilsen, P. Tregenna-Piggott, T. G. Perring, M. Enderle, D. F. McMorrow, D. A. Ivanov, and H. M. Rønnow, Na- ture Physics 11, 62 (2015), publisher: Nature Publishing Group

2015
[36]

H. Shao, Y. Q. Qin, S. Capponi, S. Chesi, Z. Y. Meng, and A. W. Sandvik, Physical Review X 7, 041072 (2017), publisher: American Physical Society

2017
[37]

Powalski, K

M. Powalski, K. P. Schmidt, and G. S. Uhrig, SciPost Physics 4, 001 (2018)

2018
[38]

A. W. Sandvik, Physical Review Letters 98, 227202 (2007), publisher: American Physical Society

2007
[39]

E. H. Lieb, Physical Review Letters 73, 2158 (1994), pub- lisher: American Physical Society

1994
[40]

Affleck and J

I. Affleck and J. B. Marston, Phys. Rev. B 37, 3774 (1988)

1988
[41]

T. C. Lang, Z. Y. Meng, A. Muramatsu, S. Wessel, and F. F. Assaad, Phys. Rev. Lett. 111, 066401 (2013)

2013
[42]

J. Nasu, M. Udagawa, and Y. Motome, Physical Re- view B 92, 115122 (2015), publisher: American Physical Society

2015
[43]

Alexandru, G

A. Alexandru, G. m. c. Ba¸ sar, P. F. Bedaque, and N. C. Warrington, Rev. Mod. Phys. 94, 015006 (2022)

2022

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.

[1] [1]

The results are presented for antiferromagnetic (a) and ferromagnetic (b) interactions

Heisenberg Model in One-Dimension Figure 5 shows the magnetic susceptibility as a func- tion of temperature for the Heisenberg chain without (black line) and with phonons at spin-phonon coupling λ = 0.2 for different phonon frequencies. The results are presented for antiferromagnetic (a) and ferromagnetic (b) interactions. Immediately we see a noticeable ...

work page

[2] [2]

Heisenberg Model in Two-Dimensions We now consider the same Heisenberg model but on a two-dimensional square lattice. Figure 6 presents the temperature dependence of the magnetic susceptibil- ity for the antiferromagnetic (a) and ferromagnetic (b) Heisenberg models on the square lattice with λ = 0 .2 and various ω0 values. We again observe that, in the fe...

work page

[3] [3]

Kitaev-Heisenberg model We now turn to the results and discussion for the frus- trated Kitaev-Heisenberg model on the honeycomb lat- tice. In Fig. 8 we present the temperature dependence of the magnetic susceptibility for different values of ϕ/π. Unlike the previous section, here we use a smaller value of the electron-phonon coupling λ = 0.1. We again ado...

work page

[4] [4]

Balents, Nature 464, 199 (2010)

L. Balents, Nature 464, 199 (2010)

2010

[5] [5]

G. D. Mahan, Many-Particle Physics, 2nd ed. (Plenum Press, New York, 1990)

work page 1990

[6] [6]

L. J. Sandilands, Y. Tian, K. W. Plumb, Y.-J. Kim, and K. S. Burch, Phys. Rev. Lett. 114, 147201 (2015)

2015

[7] [7]

Blankenbecler, D

R. Blankenbecler, D. J. Scalapino, and R. L. Sugar, Phys. Rev. D 24, 2278 (1981)

1981

[8] [8]

White, D

S. White, D. Scalapino, R. Sugar, E. Loh, J. Gubernatis, and R. Scalettar, Phys. Rev. B 40, 506 (1989)

work page 1989