Pith. sign in

REVIEW 3 major objections 4 minor 2 cited by

Nonlinear Spectroscopy as a Magnon Breakdown Diagnosis and its Efficient Simulation

T0 review · 3 major / 4 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read This paper establishes a geometric rule for reading two-dimensional coherent spectroscopy data: in partially polarized magnets, conventional magnons confine the second-order response to the diagonal and vertical lines of the frequency…

desk verdict A genuinely useful method for computing 2DCS response functions, wrapped around a diagnostic for magnon breakdown that is solid but rests on a conjecture the authors are honest about. read the letter →

arxiv 2502.01746 v1 pith:5B6B2VIO submitted 2025-02-03 cond-mat.str-el

David A. S. Kaib , Marius Möller , Roser Valenti This is my paper · ORCID
classification cond-mat.str-el
keywords two-dimensionalcoherentspectroscopysecond-ordersusceptibilitymagnonbreakdownquantumspinliquidLanczosalgorithmKitaevmodelalpha-RuCl3linearspin-wavetheory
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 gives experimentalists a simple rule for reading two-dimensional coherent spectroscopy (2DCS) data: in a partially polarized magnet, the second-order response $\chi^2_\parallel(\omega_t,\omega_\tau)$ of a conventional magnon system must be concentrated on the diagonal line $\omega_t=\omega_\tau$ and the vertical line $\omega_t=0$ of the two-frequency plane, so strong intensity away from those lines is a direct signature that the excitations are not conventional magnons. To make this rule usable, it derives the universal form expected from linear spin-wave theory, and it introduces a Lanczos-based algorithm that computes second-order susceptibilities directly in the frequency domain at far lower cost than explicit time-evolution simulations. Applying the algorithm to an extended Kitaev model for $\alpha$-RuCl$_3$ under in-plane magnetic field, the paper finds a spread-out response just above the ordering field $B_c$ and a clean two-line response at high field, matching the expectation that magnons break down at intermediate fields and are restored at high fields. The authors conclude that terahertz 2DCS measurements of $\alpha$-RuCl$_3$ can directly decide whether the observed excitation continuum is conventional two-magnon type or of different nature.

What carries the argument

The argument rests on two pieces. The first is the universal form of $\chi^2_\parallel$ for conventional magnons: under linear spin-wave theory the parallel magnetization is quadratic in magnon operators, and the $n=m$ terms in the three-state product produce poles only on $\omega_t=\omega_\tau$ (the non-rephasing signal) and $\omega_t=0$ (the rectification signal); the $n\neq m$ terms, which would place intensity elsewhere, are the inter-band processes that the paper proves to be absent on Bravais lattices and conjectures to be negligible on non-Bravais lattices. The second is the computational method: for each distinct operator among $A,B,C$ in the susceptibility, one builds a Lanczos Krylov space from the normalized operator action on the ground state, uses exact moment identities inside that space to keep all contributions through order $L$ in the time variables, and evaluates $\chi^2_{ABC}(\omega_t,\omega_\tau)$ directly in the frequency domain as a sum over Lanczos eigenenergies; the calculation converges with $L\approx 50$–$150$ on the systems tested and costs about the same as a ground-state calculation.

What would settle it

Take a partially polarized non-Bravais magnet whose low-energy physics is accurately described by linear spin-wave theory and measure its $\chi^2_\parallel$ with 2DCS: strong intensity clearly away from the diagonal and vertical lines would refute the universal form. Even without new data, a zero-temperature LSWT calculation on a non-Bravais lattice with strong inter-band mixing showing summed $n\neq m$ intensity comparable to the $n=m$ intensity would falsify the conjecture.

Watch

Extended reading notes

Core claim

The central claim is that the second-order susceptibility $\chi^2_\parallel(\omega_t,\omega_\tau)$, the leading nonlinear magnetization response measured in 2DCS with the light field parallel to the ordered moment, acts as a magnon-breakdown diagnosis. Within linear spin-wave theory, the magnetization operator in this channel is quadratic in magnon operators, so $\chi^2_\parallel$ is built from three-state products $\langle 0|M|n\rangle\langle n|M|m\rangle\langle m|M|0\rangle$. Contributions with $n=m$ put spectral weight exclusively on the diagonal line $\omega_t=\omega_\tau$ and the vertical line $\omega_t=0$, while off-diagonal $n\neq m$ contributions require a magnon to switch bands and are shown to vanish for Bravais lattices and to be negligible for non-Bravais lattices on the basis of symmetry and a detailed honeycomb-lattice calculation. Exact-diagonalization results for the $\alpha$-RuCl$_3$ model show a strong off-diagonal, off-vertical response at $B\approx 1.17 B_c$ and a clean diagonal-plus-vertical response at $B\approx 3.33 B_c$. The paper concludes that deviations from the universal form in a measured $\chi^2_\parallel$ are direct evidence that the continuum is not of conventional two-magnon type.

Load-bearing premise

The load-bearing premise is that on non-Bravais lattices the inter-band $n\neq m$ matrix elements contribute negligibly to $\chi^2_\parallel$ compared with the $n=m$ terms; the paper proves this exactly only for Bravais lattices and supports the general case with one numerical example.

Editorial extensions

If this is right

  • A terahertz 2DCS measurement on $\alpha$-RuCl$_3$ at fields just above $B_c\approx 7$ T should see strong $\chi^2_\parallel$ intensity away from the diagonal and vertical lines, directly indicating that the observed continuum is not conventional two-magnon type.
  • At high fields ($B\gg B_c$) the same channel should return to the universal two-line shape, providing an internal cross-check of the diagnosis within one material.
  • For any partially polarized frustrated magnet, a measured $\chi^2_\parallel$ that departs from the two-line form signals a breakdown of the linear-spin-wave description of the excitations.
  • The new Lanczos algorithm makes full two-dimensional frequency-plane simulations of $\chi^2$ practical on clusters that already permit ground-state exact diagonalization, at roughly the same computational cost.
  • The method applies to operators beyond magnetization, so nonlinear responses through electric-field couplings can be simulated with the same machinery.

Reading between the lines

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

  • The off-diagonal intensity pattern itself may be a fingerprint: different breakdown mechanisms (decaying magnons, bound states, fractionalized partons) should leave different frequency-plane textures, though the paper does not attempt that classification.
  • A cheap auxiliary test of the conjecture would be to compute the full LSWT $\chi^2_\parallel$ for another non-Bravais magnet with strong inter-band coupling; agreement with the two-line form would strengthen the diagnosis, disagreement would require a threshold for what counts as negligible.
  • Finite temperature and magnon interactions in an otherwise conventional magnet will generate some off-diagonal weight, so operational use of the criterion will need a quantitative estimate of those backgrounds to set the detection threshold.
  • The same pole-location logic could extend to third-order susceptibilities, giving geometric criteria for three-magnon or higher-order processes in future 2DCS experiments.
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 / 4 minor

Summary. The paper introduces a Lanczos-based method for computing zero-temperature second-order susceptibilities χ2(ω_t, ω_τ) directly in the frequency domain, avoiding costly two-dimensional time-grid evolution. The method is derived in Eqs. (5)–(8), uses separate Krylov subspaces for the operators appearing in the nested commutator, and is benchmarked against time-evolution exact-diagonalization results for the transverse-field Ising model. The authors then apply the method to an extended Kitaev model for α-RuCl3 at two in-plane field values, reporting a qualitatively different χ2∥ response at B = 1.17Bc (intensity spread away from the diagonal and vertical axes) versus B = 3.33Bc (dominant poles on Fdiag and Fvert). To interpret these results, the authors derive the linear spin-wave theory (LSWT) form of χ2∥ for partially polarized magnets and argue that conventional magnons produce a universal response concentrated on Fdiag and Fvert. Deviations from this form are proposed as a direct experimental diagnostic of magnon breakdown or unconventional excitations.

Significance. If the central claim holds, the paper offers both a practical numerical tool and a simple experimental criterion: measured intensity away from ωt = ωτ and ωt = 0 in χ2∥ would identify a continuum as non-conventional. The Lanczos method is efficient and convincingly benchmarked against the TFIM, with convergence in the Krylov dimension L demonstrated. The LSWT universal form is derived without fitting parameters, and the α-RuCl3 prediction at B ≳ 7 T is falsifiable. The main weakness is that the universality statement for arbitrary partially polarized magnets rests on a conjecture about non-Bravais lattices, and the material-specific ED prediction relies on a single finite-size cluster with ad hoc broadening. These gaps are load-bearing but appear addressable with additional numerical tests and analytical work.

major comments (3)
  1. [Section "χ2∥ for Conventional Magnons" and Appendix C, Eq. (A7)] The claim that the LSWT response of any partially polarized magnet is dominated by n = m contributions and hence restricted to Fdiag and Fvert is not established for non-Bravais lattices. Appendix C proves that n ≠ m contributions vanish identically for Bravais lattices (Z = 1), but for Z > 1 the matrix elements in Eq. (A7) are nonzero. The paper then relies on the statement that these inter-band contributions "generally have tiny intensity," which is explicitly labeled a conjecture in the main text. Only one non-Bravais example (honeycomb α-RuCl3 at B = 1.17Bc) is checked. Because the diagnostic would misclassify a conventional two-magnon continuum as unconventional if this conjecture fails in some material, this is a load-bearing gap. The authors should either provide an analytical suppression argument (e.g., a bound in terms of band separations and Bogoliubov coefficients) or systematically test multiple non-Bravais lattices, parameter regimes, and field strengths.
  2. [Figures 3(c,d), ED results] The central α-RuCl3 prediction—qualitatively different off-diagonal intensity at B = 1.17Bc versus B = 3.33Bc—is based on a single 24-site C3-symmetric cluster with a single broadening η = 0.2 meV. No finite-size analysis, no variation of η, and no explicit convergence check for the α-RuCl3 model are shown (the convergence discussion in Fig. 2 is for the TFIM only). Since excitation continua are represented by discrete states in finite-size ED, the appearance of off-diagonal poles could in principle be affected by cluster geometry or by the chosen broadening. The authors should demonstrate robustness of the qualitative contrast with respect to cluster size/shape, η, and L before the material-specific claim is fully supported.
  3. [Appendix C, LSWT numerical check] The numerical confirmation that n ≠ m contributions are at least two orders of magnitude smaller than n = m contributions is reported for one field value (B = 1.17Bc) and one model. The text states that "LSWT results at other field strengths retain this form," but no such results are shown. Given that the field strength is a key parameter in the proposed crossover from magnon breakdown to conventional magnons, the field dependence of the inter-band matrix elements should be documented or at least explicitly argued. This would also strengthen the universality claim for the partially polarized phase beyond a single point in parameter space.
minor comments (4)
  1. [Abstract] The abstract contains a typesetting error: "χ^{2}\omega_t,\omega_τ)" is missing the opening parenthesis, making it read as "χ2ω_t,ω_τ)" rather than "χ2(ω_t,ω_τ)."
  2. [Figure 4 and main text] The main text refers to "Fig. 4(c)" and "Fig. 4(d)" when discussing the two-magnon continuum bottom and the LSWT χ2∥ results, but the displayed Fig. 4 only shows panels (a) and (b). The figure or the references should be updated to match.
  3. [Appendix C, Eq. (A9)] The function g(Em, En, ωt, ωτ) in Eq. (A9) is written with arguments Em and En, but the assignment of these energies to the intermediate states in Eq. (A8) is not explicitly stated. Please clarify the correspondence between g's arguments and the states |n⟩, |m⟩ to avoid ambiguity.
  4. [Table I] In Table I, the row for "fully polarized (no quantum fluctuations)" lists "zero" for both linear and nonlinear response. It may be worth adding a brief sentence in the text explaining that this follows from the absence of fluctuations in the strict fully polarized limit, since the main text focuses on the partially polarized cases.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the LSWT universal form is derived parameter-free, the Lanczos method is benchmarked against independent TFIM data, and the α-RuCl3 prediction is not fitted to the target.

full rationale

The paper's load-bearing steps do not reduce to their own inputs. The universal form of χ2∥(ωt,ωτ) for conventional magnons is derived analytically in Appendix C from the Holstein-Primakoff expansion and a Bogoliubov transformation, with no fitted parameter encoding the target prediction; the restriction to Fdiag and Fvert follows from Eq. (A6) for n=m contributions. The claim that n≠m inter-band matrix elements are negligible is explicitly flagged as a conjecture and checked numerically for the honeycomb model, so it is a stated assumption rather than a hidden circular step. The Lanczos algorithm is benchmarked against an independent time-evolution calculation for the transverse-field Ising model (Ref. [33]), establishing that the numerical machinery is not calibrated on the predicted α-RuCl3 signal. Model parameters for α-RuCl3 are taken from Refs. [37,38], whose overlap with the present authors is real but whose results are prior published and reproduce linear-response data independent of the nonlinear prediction; no constant in the χ2 calculation is fitted to the 2DCS result. The diagnostic claim is therefore a genuine derivation plus a numerical prediction, with the non-Bravais conjecture being a correctness risk rather than a circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The main free parameters are the α-RuCl3 model couplings taken from earlier fits; the central prediction depends on them. The key added assumption is the conjectured suppression of inter-band matrix elements in non-Bravais lattices, which is not proven. No new physical entities are postulated.

free parameters (4)
  • Kitaev model parameters (K, Γ, J, J3, g∥) = K=-5, Γ=2.5, J=-0.5, J3=0.5 meV, g∥=2.3
    Taken from fits to α-RuCl3 in Refs. [37,38]; the ED prediction of deviation at B≈1.17Bc depends on these values.
  • Field strengths in ED (B/Bc) = 1.17 and 3.33
    Chosen to represent the intermediate and high-field regimes of the model; Bc=6 T within ED.
  • Broadening η = 0.2 meV
    Lorentzian broadening in frequency-domain evaluation; chosen ad hoc to display raw results.
  • Krylov dimension L = 150
    Numerical truncation; convergence is demonstrated for TFIM but only asserted for Kitaev model.
assumptions (5)
  • domain assumption Standard linear spin-wave theory (Holstein-Primakoff) applies to the partially-polarized phase
    Used in the universal-form derivation, assumes small quantum fluctuations around field-polarized state for B>Bc.
  • ad hoc to paper The n≠m matrix elements are negligible in non-Bravais lattices
    Conjectured in main text; numerically supported only for the honeycomb α-RuCl3 model.
  • domain assumption The 24-site C3-symmetric cluster captures the relevant two-magnon continuum
    Finite-size ED used for Fig. 3; continuum appears as discrete poles.
  • domain assumption The extended Kitaev model with parameters of Ref. [37,38] is representative of α-RuCl3
    Model parameters are fitted to reproduce linear response anomalies.
  • standard math Krylov subspace matrix elements ⟨ψn|H^a|ψm⟩ = δnm(ϵn)^a for a < L
    Used in Eq. (6) to derive the frequency-domain expression.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Nonlinear Spectroscopy as a Magnon Breakdown Diagnosis and its Efficient Simulation." pith.science (2026). https://pith.science/paper/5B6B2VIO

@misc{pith2026250201746,
  author       = {Pith},
  title        = {Pith review of: Nonlinear Spectroscopy as a Magnon Breakdown Diagnosis and its Efficient Simulation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5B6B2VIO}},
  note         = {Machine review of arXiv:2502.01746}
}
abstract

Identifying quantum spin liquids, magnon breakdown, or fractionalized excitations in quantum magnets is an ongoing challenge due to the ambiguity of possible origins of excitation continua occurring in linear response probes. Recently, it was proposed that techniques measuring higher-order response, such as two-dimensional coherent spectroscopy (2DCS), could resolve such ambiguities. Numerically simulating nonlinear response functions can, however, be computationally very demanding. We present an efficient Lanczos-based method to compute second-order susceptibilities $\chi^{2}\omega_t,\omega_\tau)$ directly in the frequency domain. Applying this to extended Kitaev models describing $\alpha$-RuCl$_3$, we find qualitatively different nonlinear responses between intermediate magnetic field strengths and the high-field regime. To put these results into context, we derive the general 2DCS response of partially-polarized magnets within the linear spin-wave approximation, establishing that $\chi^2(\omega_t,\omega_\tau)$ is restricted to a distinct universal form if the excitations are conventional magnons. Deviations from this form, as predicted in our (Lanczos-based) simulations for $\alpha$-RuCl$_3$, can hence serve in 2DCS experiments as direct criteria to determine whether an observed excitation continuum is of conventional two-magnon type or of different nature.

Figures

Figures reproduced from arXiv: 2502.01746 by the authors.

Figure 1
Figure 1. FIG. 1. (a) Sketch of a 2DCS measurement protocol. Two [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Comparison of computed [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. FIG. 4. (a) Linear terahertz response Im [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Two-dimensional nonlinear dynamical response of the magnetoelectrically driven dimerized spin-$1{/}2$ chain

    cond-mat.str-el 2025-07 conditional novelty 6.0 of 10

    The 2D nonlinear response of a magnetoelectrically coupled dimerized spin-1/2 chain shows a galvanoelectric line whose width measures spinon scattering rates, plus RPA vertex corrections that create bound states and t...

  2. Probing quantum geometry with two-dimensional nonlinear optical spectroscopy

    cond-mat.mes-hall 2025-06 conditional novelty 6.0 of 10

    Two-dimensional coherent spectroscopy can isolate and measure the imaginary part of the multi-band quantum connection through the diagonal second-order conductivity under time-reversal symmetry.

Reference graph

Works this paper leans on

57 extracted references · 49 canonical work pages · cited by 2 Pith papers

  1. [1]

    Mukamel, Annu

    S. Mukamel, Annu. Rev. Phys. Chem. 51, 691 (2000)

    work page 2000

  2. [2]

    Khalil, N

    M. Khalil, N. Demird¨ oven, and A. Tokmakoff, J. Phys. Chem. A 107, 5258 (2003)

    work page 2003

  3. [3]

    P. K. Johansson, L. Schm¨ user, and D. G. Castner, Top. Catal. 61, 1101 (2018)

    work page 2018

  4. [4]

    Sankar and R

    P. Sankar and R. Philip, in Characterization of Nanoma- terials, Micro and Nano Technologies (Woodhead Pub- lishing, 2018) pp. 301–334

    work page 2018

  5. [5]

    Garmire, Phys

    E. Garmire, Phys. Today 47, 42 (1994)

    work page 1994

  6. [6]

    Kuehn, K

    W. Kuehn, K. Reimann, M. Woerner, T. Elsaesser, and R. Hey, J. Phys. Chem. B 115, 5448 (2011)

    work page 2011

  7. [7]

    J. Lu, X. Li, H. Y. Hwang, B. K. Ofori-Okai, T. Kurihara, T. Suemoto, and K. A. Nelson, Phys. Rev. Lett. 118, 207204 (2017)

    work page 2017

  8. [8]

    W. Choi, K. Lee, and Y. Kim, Phys. Rev. Lett. 124, 117205 (2020)

    work page 2020

Show all 57 references
  1. [9]

    M. K. Negahdari and A. Langari, Phys. Rev. B 107, 134404 (2023)

    2023

  2. [10]

    Wan and N

    Y. Wan and N. Armitage, Phys. Rev. Lett. 122, 257401 (2019)

    2019

  3. [11]

    Savary and L

    L. Savary and L. Balents, Rep. Prog. Phys. 80, 016502 (2016)

    2016

  4. [12]

    Knolle and R

    J. Knolle and R. Moessner, Annu. Rev. Condens. Matter Phys. 10, 451 (2019)

    2019

  5. [13]

    Broholm, R

    C. Broholm, R. Cava, S. Kivelson, D. Nocera, M. Nor- 6 man, and T. Senthil, Science 367, eaay0668 (2020)

    2020

  6. [14]

    Kasahara, T

    Y. Kasahara, T. Ohnishi, Y. Mizukami, O. Tanaka, S. Ma, K. Sugii, N. Kurita, H. Tanaka, J. Nasu, Y. Mo- tome, et al., Nature 559, 227 (2018)

    2018

  7. [15]

    Yokoi, S

    T. Yokoi, S. Ma, Y. Kasahara, S. Kasahara, T. Shibauchi, N. Kurita, H. Tanaka, J. Nasu, Y. Motome, C. Hickey, S. Trebst, and Y. Matsuda, Science 373, 568 (2021)

    2021

  8. [16]

    Bruin, R

    J. Bruin, R. Claus, Y. Matsumoto, N. Kurita, H. Tanaka, and H. Takagi, Nat. Phys. 18, 401 (2022)

    2022

  9. [17]

    Czajka, T

    P. Czajka, T. Gao, M. Hirschberger, P. Lampen-Kelley, A. Banerjee, N. Quirk, D. G. Mandrus, S. E. Nagler, and N. P. Ong, Nat. Mater. 22, 36 (2023)

    2023

  10. [18]

    Lefran¸ cois, J

    ´E. Lefran¸ cois, J. Baglo, Q. Barth´ elemy, S. Kim, Y.-J. Kim, and L. Taillefer, Phys. Rev. B 107, 064408 (2023)

    2023

  11. [19]

    Dhakal, D

    R. Dhakal, D. A. Kaib, S. Biswas, R. Valenti, and S. M. Winter, arXiv preprint arXiv:2407.00660 (2024)

    2024 arXiv

  12. [20]

    T.-H. Han, J. S. Helton, S. Chu, D. G. Nocera, J. A. Rodriguez-Rivera, C. Broholm, and Y. S. Lee, Nature 492, 406 (2012)

    2012

  13. [21]

    Banerjee, J

    A. Banerjee, J. Yan, J. Knolle, C. A. Bridges, M. B. Stone, M. D. Lumsden, D. G. Mandrus, D. A. Tennant, R. Moessner, and S. E. Nagler, Science 356, 1055 (2017)

    2017

  14. [22]

    Z. Wang, S. Reschke, D. H¨ uvonen, S.-H. Do, K.-Y. Choi, M. Gensch, U. Nagel, T. R˜ o˜ om, and A. Loidl, Phys. Rev. Lett. 119, 227202 (2017)

    2017

  15. [23]

    Kermarrec, A

    E. Kermarrec, A. Zorko, F. Bert, R. Colman, B. Koteswararao, F. Bouquet, P. Bonville, A. Hillier, A. Amato, J. Van Tol, et al., Phys. Rev. B 90, 205103 (2014)

    2014

  16. [24]

    Z.-L. Li, M. Oshikawa, and Y. Wan, Phys. Rev. X 11, 031035 (2021)

    2021

  17. [25]

    R¨ uckriegel, D

    A. R¨ uckriegel, D. Tarasevych, J. Krieg, and P. Kopietz, Phys. Rev. B 110, 144416 (2024)

    2024

  18. [26]

    Krupnitska and W

    O. Krupnitska and W. Brenig, Phys. Rev. B 108, 075120 (2023)

    2023

  19. [27]

    Brenig and O

    W. Brenig and O. Krupnitska, J. Phys. Condens. Matter 36, 505806 (2024)

    2024

  20. [28]

    Kanega, T

    M. Kanega, T. N. Ikeda, and M. Sato, Phys. Rev. Res. 3, L032024 (2021)

    2021

  21. [29]

    Qiang, V

    Y. Qiang, V. L. Quito, T. V. Trevisan, and P. P. Orth, Phys. Rev. Lett. 133, 126505 (2024)

    2024

  22. [30]

    G. Sim, J. Knolle, and F. Pollmann, Phys. Rev. B 107, L100404 (2023)

    2023

  23. [31]

    G. Sim, F. Pollmann, and J. Knolle, Phys. Rev. B 108, 134423 (2023)

    2023

  24. [32]

    Q. Gao, Y. Liu, H. Liao, and Y. Wan, Phys. Rev. B 107, 165121 (2023)

    2023

  25. [33]

    Watanabe, S

    Y. Watanabe, S. Trebst, and C. Hickey, Phys. Rev. B 110, 134443 (2024)

    2024

  26. [34]

    Woerner, W

    M. Woerner, W. Kuehn, P. Bowlan, K. Reimann, and T. Elsaesser, New J. Phys. 15, 025039 (2013)

    2013

  27. [35]

    non-rephasing signal

    will generate a basis for the L-dimensional Krylov subspace span O |0⟩ , HO |0⟩ , H2O |0⟩ , . . . ,HL−1O |0⟩ . (4) Fixing notation, we name the d-dimensional ( d = dim(H)) orthonormal basis vectors generated during the Lanczos routine as |ϕO m⟩. Diagonalization of the tridi- a...

  28. [36]

    Lanczos, J

    C. Lanczos, J. Res. Natl. Bur. Stand. 45, 255 (1950)

    1950

  29. [37]

    Lehoucq, D

    R. Lehoucq, D. Sorensen, and C. Yang, ARPACK Users’ Guide: Solution of Large-scale Eigenvalue Problems with Implicitly Restarted Arnoldi Methods, Software, Environ- ments, and Tools (Society for Industrial and Applied Mathematics, 1998)

    1998

  30. [38]

    S. M. Winter, K. Riedl, P. A. Maksimov, A. L. Cherny- shev, A. Honecker, and R. Valent ´ ı, Nat. Commun. 8, 1152 (2017)

    2017

  31. [39]

    S. M. Winter, K. Riedl, D. Kaib, R. Coldea, and R. Va- lent ´ ı, Phys. Rev. Lett.120, 077203 (2018)

    2018

  32. [40]

    Sahasrabudhe, D

    A. Sahasrabudhe, D. A. S. Kaib, S. Reschke, R. Ger- man, T. C. Koethe, J. Buhot, D. Kamenskyi, C. Hickey, P. Becker, V. Tsurkan, et al. , Phys. Rev. B 101, 140410(R) (2020)

    2020

  33. [41]

    Dagotto, Rev

    E. Dagotto, Rev. Mod. Phys. 66, 763 (1994)

    1994

  34. [42]

    Hart and R

    O. Hart and R. Nandkishore, Phys. Rev. B 107, 205143 (2023)

    2023

  35. [43]

    3(k) of Ref

    Note that the corresponding LSWT plot in Fig. 3(k) of Ref. [38] shows no intensity for B > Bc, as only one- magnon states were considered there

  36. [44]

    Jakliˇ c and P

    J. Jakliˇ c and P. Prelovˇ sek, Phys. Rev. B49, 5065 (1994)

    1994

  37. [45]

    Jakliˇ c and P

    J. Jakliˇ c and P. Prelovˇ sek, Adv. Phys.49, 1 (2000)

    2000

  38. [46]

    Holstein and H

    T. Holstein and H. Primakoff, Phys. Rev. 58, 1098 (1940)

    1940

  39. [47]

    We restrict k and k′ to one half of the Brillouin zone, in order to not double-count identical two-magnon states |kl, −kl⟩ = |−kl, kl⟩

  40. [48]

    Vladimirov, D

    A. Vladimirov, D. Ihle, and N. M. Plakida, J. Exp. Theor. Phys. 122, 1060 (2016)

    2016

  41. [49]

    P. A. Maksimov and A. L. Chernyshev, Phys. Rev. Res. 2, 033011 (2020)

    2020

  42. [50]

    R. Smit, S. Keupert, O. Tsyplyatyev, P. Maksimov, A. L. Chernyshev, and P. Kopietz, Phys. Rev. B 101, 054424 (2020). 7 Appendix Appendix A: Numerical implementation We explain the algorithm for the case of diagonal sus- ceptibilities χ2 AAA, i.e. A = B = C in Eq. (2). Then the...

    2020

  43. [51]

    Compute the ground state |0⟩ of H and its energy E0, for example via a standard Lanczos routine

  44. [52]

    using a random start vector, or related meth- ods [36]

  45. [53]

    Generate |ϕA 0 ⟩ and N A 0 according to Eq. (3)

  46. [54]

    ,|ϕA L−1⟩} as well as the eigenvalues ϵA m and eigenvectors vA m of the tridiagonal matrix

    Using the Lanczos algorithm with |ϕA 0 ⟩ as a start vector, generate and store the basis vectors {|ϕA 0 ⟩ , |ϕA 1 ⟩ , . . . ,|ϕA L−1⟩} as well as the eigenvalues ϵA m and eigenvectors vA m of the tridiagonal matrix

  47. [55]

    Compute all matrix elements ⟨ϕA l |A|ϕA p ⟩ for l, p∈ {0, 1, . . . , L− 1}. For this, it might be efficient to iterate over p, generating |ϕA p ′ ⟩ = A |ϕA p ⟩ and com- puting the overlaps ⟨ϕA l |ϕA p ′ ⟩ = ⟨ϕA l |A|ϕA p ⟩ for all l ≤ p. For A† = A, the elements with l > pfoll...

  48. [56]

    Obtain all Xn,m via Eq. (9). Note that the sum in Eq. (9) can be efficiently computed by expressing it as a matrix multiplication

  49. [57]

    (8) with a chosen broadening η >0 for all de- sired frequencies ωt, ωτ

    Evaluate χ(2) AAA(ωt, ωτ ) (here, Yn,m = Xn,m) via Eq. (8) with a chosen broadening η >0 for all de- sired frequencies ωt, ωτ . The d-dimensional eigenvectors in the Krylov sub- space ( |ψm⟩) do not need to be assembled explicitly at any point. The computationally expensive st...

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.