{"id":"f7f28c39-4d17-4779-924c-ac6c2bbeb52b","arxiv_id":"2509.08955","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Resonant magnetic X-ray photon correlation spectroscopy on Cu0.88Mn0.12 reveals critical slowing down of spin-orientation fluctuations above Tg, with a relaxation time that fits both Vogel-Fulcher and power-law forms.","lead":"X-ray speckle measurements on a copper-manganese alloy show that slow spin fluctuations above the spin-glass transition follow the same Vogel-Fulcher temperature law used for structural glasses. This is presented as the first direct measurement of the time-dependent Edwards-Anderson order parameter fluctuations in a spin glass.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The mapping from g2 to chi_SG(tau) in Eqs. (8)-(10) is an uncontrolled factorization; without a test of the neglected four-spin cumulant, the extracted tau0 may not represent EA order-parameter fluctuations.","rationale":"The reader's weakest assumption identifies the decoupling approximation in Eqs. (8)-(10) as the point where the measured g2 becomes the EA overlap fluctuation correlation. I agree that this is the most load-bearing step: without it, the experiment measures some four-spin correlation whose connection to chi_SG(tau) is unproven. The concern is not that the Gaussian/Siegert factorization is obviously wrong; XPCS routinely uses such factorizations. The problem is that the entire claim of a direct measurement of the time-dependent spin-glass susceptibility rests on a factorization that is expected to be qualitatively modified near a critical point, and the paper supplies no quantitative check. The concrete Monte Carlo test is feasible and would settle whether the factorization error is negligible at the reported contrast level. I therefore do not change the reader's conditional verdict; the paper remains promising but requires this validation before the headline claim is accepted.","tokens_in":10167,"tokens_out":14154,"duration_ms":157171,"concrete_test":"Perform equilibrium Monte Carlo dynamics of a 3D Edwards-Anderson Ising spin glass (e.g., L=32, T/Tc in 1.1-1.5). Using the same 2-sec block averaging as in the experiment, construct the forward-scattering intensity I(q,t) from the spin configurations, compute the exact g2(q,tau) and the decoupled expression beta*/^2, and compare them. If the ratio differs from unity by more than ~10%, or if the tau0 obtained from fitting the exact g2 differs from that obtained from the factorized form by more than the reported error bars, the identification in Eqs. (8)-(10) is not quantitatively adequate.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central identity g2(q,tau)-1 = beta*chi_SG(tau)/^2 depends on Eq. (8), where the intensity autocorrelation is replaced by the same-pair term sum_ij _t and all other pairings are decoupled into ^2. This is a Gaussian/Siegert factorization. Near Tg, however, the spin-glass transition is driven by a non-Gaussian four-spin correlation: the connected four-point susceptibility diverges, and its contribution to g2(tau) need not be subleading compared with the factorized term. The paper provides no estimate of the neglected cumulant term and no independent check that factorization holds at the relevant temperatures and timescales. Furthermore, identifying the first term with requires the 2-sec block average of S_i(t) to act as the thermal average in Eq. (1). If those block-averaged spins are not faithful quasi-static configurations, q(t,tau) is not the Edwards-Anderson overlap. Every downstream claim -- chi_SG(tau), tau0(T), and the Vogel-Fulcher divergence -- inherits this identification. The power-law/VF degeneracy noted in Supp. F is a secondary issue; the factorization is the load-bearing step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents RM-XPCS measurements on Cu_0.88Mn_0.12 above the spin-glass transition and argues that the normalized intensity autocorrelation g2(q,tau) is directly proportional to the time-dependent spin-glass susceptibility chi_SG(tau), i.e., to the autocorrelation of the squared Edwards-Anderson overlap q(t,tau)^2. From the measured decay they extract a relaxation time tau0(T) that grows as T approaches Tg from above and claim it follows the Vogel-Fulcher law with T0=36.5 K, while acknowledging a power-law fit with Tg=44 K and B=2.7 is equally good. The paper concludes this is the first direct measurement of the temperature dependence of the time-dependent susceptibility related to SG(EA) order parameter fluctuations.","tokens_in":10585,"tokens_out":6239,"duration_ms":67673,"significance":"If the central mapping is valid, the manuscript is a significant experimental advance: it introduces a way to probe four-spin correlations associated with the EA order parameter on time scales inaccessible to neutron scattering, and the q-independence and off-resonance controls (Supp. C-E) are sensible checks. The potential extension to spin ices, spin liquids, and structural glasses is plausible. However, the paper provides no code or data availability statement and, more importantly, the mapping itself rests on an uncontrolled factorization for which the necessary validation is not supplied; until that is fixed, the quantitative claims (tau0(T), VF/power-law divergence) remain conditional.","major_comments":[{"comment":"The central identity g2-1 = beta chi_SG(tau)/^2 depends on Eq. (8), but Eq. (8) invokes 'By Eq. (7)' for ^2 while Eq. (7) is missing from the manuscript. More substantively, Eq. (8) factorizes the four-spin product into the same-pair term and a product of separated pair averages with only the assertion that different spin pairs are 'spatially uncorrelated.' This is a Gaussian/Siegert-type factorization, and no estimate is given for the neglected connected four-spin cumulant, which is precisely the quantity expected to grow near the spin-glass transition. Without a bound on this term, or an independent numerical/experimental test, the identification with chi_SG(tau) is not established.","section":"Eqs. (8)-(10)"},{"comment":"Eq. (1) defines q(t,tau) with a true thermal average _T, but the experiment replaces it with a block average over 2 sec (main text after Eq. (5)), while Supp. C says the time frame for collecting data was 5 sec. No demonstration is provided that this block is long enough to define a quasi-static spin configuration and short compared with the slow fluctuations of interest. If the 2-sec/5-sec block average is not a faithful ergodic average, q(t,tau) is not the Edwards-Anderson overlap and the subsequent chi_SG(tau) interpretation falls.","section":"Eqs. (1), (5)-(6); Supp. C"},{"comment":"The paper's abstract and conclusion emphasize the Vogel-Fulcher law, but Supp. f states the power-law Eq. (15) 'is as good as' the VF fit. Since the two forms are degenerate for this dataset, the claim 'consistent with the Vogel-Fulcher law' overstates what is demonstrated. The authors should provide quantitative model comparison (e.g., residuals, reduced chi^2, parameter uncertainties) and discuss what measurement range would distinguish VF from power law. The fitted power-law exponent B=2.7 also conflicts with earlier zv~7 estimates; this discrepancy is not resolved.","section":"Fig. (3) and Supp. f"},{"comment":"The relaxation time tau0(T) is extracted by fitting g2 to a single exponential plus constant. No justification is given for this functional form; if the true decay is stretched-exponential or has multiple steps, tau0(T) will be systematically biased, and this could affect the VF/power-law comparison. I request fits with alternative forms (e.g., exp[-(tau/tau0)^beta]) and a plot of residuals, or at least a statement that the conclusions are robust to the choice.","section":"Eq. (14)"}],"minor_comments":[{"comment":"'Vogel-Vulcher' should be 'Vogel-Fulcher' (twice).","section":"Abstract"},{"comment":"The power-law fit is said to be shown in 'Fig. (5)' in the main text but appears as Fig. (9) in the supplementary file; also Eq. (15) is referenced before it is displayed.","section":"Supp. f / main text"},{"comment":"The axis label uses omega (log10[omega/omega0]) whereas the text uses tau (log10(tau/tau0)); please make the notation consistent.","section":"Fig. 3"},{"comment":"Ref. [2] has a typo: 'Krikpatrick' should be 'Kirkpatrick'; check journal/volume formatting in Refs. [30] and [33].","section":"References"}],"recommendation":"major_revision","confidential_remarks":"This is a potentially important experiment, but the current manuscript overclaims. The strongest path to publication is to (i) provide the missing Eq. (7) and a careful derivation of the factorization, including an estimate of the connected four-spin contribution; (ii) resolve the 2-sec vs 5-sec time-frame inconsistency and justify the block average; and (iii) tone down the VF claim or provide a proper model comparison. I do not see the issues as fatal, but they are central and need to be addressed before the paper can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis paper is worth reading for the experiment. The authors have used resonant magnetic XPCS to follow spin fluctuations in CuMn over seconds to hours, and they see a clear, reproducible slowing down as Tg is approached from above. The controls are honest: off-resonance data are static, the magnetic signal is q-independent in the measured range, and the correlation functions do not depend on starting time. That part deserves credit.\n\nWhat is genuinely new is treating the speckle intensity autocorrelation as a four-spin correlation tied to the Edwards-Anderson order parameter. Standard XPCS measures density or charge fluctuations; this is a different observable. The claim that g2 - 1 is proportional to chi_SG(tau) goes through Eqs. (8)-(10), and here is the soft spot. The factorization that leaves only the same-pair term and decouples everyone else is uncontrolled. Near a spin-glass transition the connected four-point susceptibility is expected to diverge, so the term thrown away is not obviously subleading. The paper gives no estimate for the neglected cumulant and no independent check that the factorization holds at the temperatures and timescales measured. The 2-second block average that defines Si(t) also does a lot of work; if that is not a faithful quasi-static configuration, q(t,tau) is not the EA overlap.\n\nThe secondary issues are minor-ish: the Vogel-Fulcher form is not unique because a power law fits about as well (the paper admits this in the Supplement), and the claim that no correlation-time measurements exist in spin glasses overstates the neutron spin-echo literature. But the factorization is the load-bearing step.\n\nThe upshot: I would not use the extracted tau0(T) as a quantitative EA susceptibility relaxation time yet. The method, though, is promising enough that the paper deserves a serious referee, sent to someone who knows both XPCS and spin-glass theory. The authors should be asked to justify or test the factorization, report error bars, and present model comparison statistics for VF versus power law. If they can do that, the result could be important.","headline":"A genuinely new experimental probe of spin-glass dynamics, but the central mapping from speckle contrast to the Edwards-Anderson susceptibility rests on an uncontrolled factorization that needs much more support before the headline claim can be accepted.","tokens_in":11069,"tokens_out":2397,"would_cite":true,"duration_ms":28588,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["75.50.Lk","75.40.Gb"],"model":"deepseek-v4-flash","headline":"Resonant magnetic X-ray speckle autocorrelations directly measure fluctuations of the Edwards-Anderson spin-glass order parameter, and the measured relaxation time diverges as the Vogel-Fulcher law as Tg is approached.","keywords":["spin glasses","Edwards-Anderson order parameter","time-dependent susceptibility","x-ray photon correlation spectroscopy","critical slowing down","Vogel-Fulcher law","CuMn alloy","speckle correlations"],"falsifier":"A numerical simulation of a CuMn-like spin glass that computes both and the full four-spin X-ray intensity autocorrelation at small q would settle the identification: if the two disagree in shape or in tau0(T), the decoupling step in Eq. (8) is invalid; equivalently, measuring g2(q,tau) at two strongly different Mn concentrations and finding different reduced tau0(T-Tg) scaling would signal that the order-parameter identification fails.","tokens_in":10052,"feed_emoji":"🧲","tokens_out":6862,"duration_ms":74220,"temperature":0.7,"pith_summary":"This paper claims that the time-dependent correlations of the Edwards-Anderson order parameter above a spin-glass transition can be measured directly from the autocorrelation of magnetic X-ray speckle patterns. In resonant magnetic X-ray photon correlation spectroscopy, the normalized intensity autocorrelation g2(q,tau) minus one is shown to be proportional to a time-dependent spin-glass susceptibility chi_SG(tau), which is the squared, configuration-averaged Edwards-Anderson overlap. Applying this to Cu0.88Mn0.12, the authors observe exponential decay of the normalized autocorrelation with relaxation times from about 2 seconds to 2 x 10^4 seconds, slowing drastically as the temperature approaches Tg ≈ 45 K. The temperature dependence of tau0(T) is well described by the Vogel-Fulcher law with T0 = 36.5 K, and the paper notes it can also be fitted by a power law with Tg = 44 K and exponent 2.7. A sympathetic reader would care because this is, on the paper's own account, the first direct measurement of the temperature dependence of the time-dependent susceptibility tied to the Edwards-Anderson order-parameter fluctuations, and it links spin-glass dynamics to the Vogel-Fulcher phenomenology of structural glasses.","feed_headline":"Speckle autocorrelation reads the spin-glass order parameter directly","feed_subtitle":"In CuMn, the measured relaxation time follows the Vogel-Fulcher law as Tg approaches, linking spin-glass memory to structural glasses.","key_machinery":"The load-bearing object is the time-dependent Edwards-Anderson overlap q(t,tau) = (1/N) sum_i · averaged over the random spin orientations, whose infinite-time limit is the Edwards-Anderson order parameter. The argument works through the relation between the speckle intensity autocorrelation and this overlap: in the forward-scattering limit the magnetic scattering amplitude is a sum of thermally averaged spin components, and after decoupling distinct spin pairs the intensity autocorrelation becomes C^2 chi_SG(tau) + ^2, with chi_SG(tau) proportional to _t. This identity converts g2(q,tau)-1, normalized by the total intensity, into a direct measure of th","core_discovery":"The central claim is that the Edwards-Anderson order parameter has an observable time-dependent fluctuation spectrum above Tg, and that it appears as a four-spin correlation in the intensity of resonantly scattered coherent X-rays. Specifically, for forward scattering and with the usual decoupling of distinct spin pairs, the time-averaged magnetic intensity autocorrelation is C^2 chi_SG(tau) plus the square of the mean magnetic intensity, where chi_SG(tau) = (1/N^2) _t and q(t,tau) is the overlap of thermally averaged spin orientations at times t and t+tau. The experimentally normalized function g2(q,tau)-1 is therefore proportional to chi_SG(tau). The paper reports that in Cu0.8","pith_inferences":["If the identification of g2 - 1 with chi_SG(tau) survives closer scrutiny, the same normalized autocorrelation could serve as a model-free order-parameter-fluctuation thermometer for any spin glass, and its temperature dependence could be compared directly with the magnetization cusp to locate Tg uniquely.","A natural test of the decoupling assumption is to run a numerical simulation of the same CuMn-like spin Hamiltonian with X-ray scattering weights and check whether the intensity autocorrelation's shape matches ; a mismatch would mean the extracted tau0(T) is not the EA susceptibility time.","The near-q independence found here suggests that analogous speckle-correlation measurements in spin ices or quantum spin liquids would see a similar flat q dependence only if the fluctuations are genuine local-order-parameter fluctuations, offering a way to classify slow dynamics in those systems.","The paper's two-pulse XPCS remark implies a broader inference: the same four-spin-correlation quantity measured at nanosecond delays with X-ray free-electron lasers could bridge the gap between macroscopic memory and microscopic spin dynamics, connecting to neutron spin-echo results."],"forward_implications":["RM-XPCS provides a direct, real-time route to the time-dependent Edwards-Anderson susceptibility, covering time scales from seconds to tens of thousands of seconds that neutron scattering cannot reach.","In Cu0.88Mn0.12, the correlation time diverges as the temperature approaches Tg according to the Vogel-Fulcher law, with T0 ≈ 36.5 K, so the spin-glass memory time scale behaves like the viscous relaxation time of structural glasses.","The magnetic contribution to g2(q,tau) is nearly q-independent at small q, consistent with an order-parameter fluctuation that carries spatial randomness but no periodic spatial correlations.","The same four-spin intensity-correlation method is applicable to other systems with quenched or entangled spin fluctuations, including spin ices, quantum spin liquids, and possibly the structural glass transition.","If the power-law fit is used instead, the extracted dynamic exponent B ≈ 2.7 is appreciably smaller than the value ~7 from simulations of the cubic Ising spin glass, highlighting a discrepancy between the measured dynamics and standard spin-glass critical scaling."],"fun_headline_variants":["Speckle autocorrelation captures spin-glass memory above Tg","Spin-glass fluctuations above Tg measured via X-ray speckle","Vogel-Fulcher law emerges in spin-glass critical slowing","Four-spin correlations track spin-glass time order"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The central claim collapses if the decoupling approximation in Eqs. (8)-(10) fails—namely, if contributions from distinct spin pairs do not factor into ^2, or if the 2-second thermal average and forward-scattering limit cannot be taken, then the measured g2 is not a faithful reading of chi_SG(tau) and the extracted tau0(T) is not the Edwards-Anderson order-parameter relaxation time.","fun_headline_variants_meta":{"raw":{"variants":["Speckle autocorrelation captures spin-glass memory above Tg","Spin-glass fluctuations above Tg measured via X-ray speckle","Vogel-Fulcher law emerges in spin-glass critical slowing","Four-spin correlations track spin-glass time order"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000806,"raw_usage":{"total_tokens":3388,"prompt_tokens":771,"completion_tokens":2617,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":515,"completion_tokens_details":{"reasoning_tokens":2551}},"tokens_in":515,"tokens_out":2617,"duration_ms":19492,"temperature":1.0,"reasoning_tokens":2551,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T19:55:45.654733+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A numerical simulation of a CuMn-like spin glass that computes both and the full four-spin X-ray intensity autocorrelation at small q would settle the identification: if the two disagree in shape or in tau0(T), the decoupling step in Eq. (8) is invalid; equivalently, measuring g2(q,tau) at two strongly different Mn concentrations and finding different reduced tau0(T-Tg) scaling would signal that the order-parameter identification fails.","supporting_citations":[],"review_version":1}