{"id":"15faaeec-ca92-4ff0-bd40-6f8d776126b5","arxiv_id":"2607.15224","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A muon-derived spin relaxation model is applied to metal-halide perovskites, yielding localization radii and correlation times and predicting longer spin lifetimes for lighter halogens and cations.","lead":"The authors adapt a muon-spin-relaxation model to describe how long electron and hole spins survive in metal-halide perovskites. The model works for any correlation time, fits two perovskite samples, and predicts that lighter halogens and metal cations give longer spin lifetimes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'necessity' of the dynamic Kubo-Toyabe model is not demonstrated: no fit comparison to mono-exponential/short-τc models is shown, and the PFR curves may be underdetermined by the parameters extracted.","rationale":"The reader identified the uncertain hyperfine constants as the weakest assumption and gave a CONDITIONAL verdict. I agree that this is a real concern, especially for quantitative localization radii and composition trends. However, the more load-bearing issue for the central claim is the lack of a direct statistical comparison proving the exact dynamical model is necessary: the paper's abstract and Section IV claim 'demonstrating the necessity' but no model comparison is shown. This is not a contradiction of the reader's verdict; it strengthens the case for keeping the verdict CONDITIONAL, so I leave the verdict unchanged. The proposed test (model comparison with identifiability analysis) would directly settle whether the experimental data actually require the dynamical Kubo-Toyabe treatment, independent of the hyperfine-constant uncertainty.","tokens_in":20607,"tokens_out":4582,"duration_ms":40456,"concrete_test":"Re-fit the PFR longitudinal-field data of Figure 6(c,d) using (i) the dynamic Kubo-Toyabe model and (ii) a mono-exponential / Smirnov effective-T1 model, keeping the same number of free parameters and including the electron/hole amplitude ratio. Compare fits via AIC/BIC and residual analysis; if the simpler model yields ΔAIC < 2 or residuals within noise, the necessity claim is not supported. Also run a profile-likelihood or bootstrap analysis to check whether ΔB_N,e, τc,e, ΔB_N,h, τc,h are uniquely identifiable from these curves.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the exact dynamical Kubo-Toyabe strong-collision solution is necessary to describe spin relaxation in MHPs and that the extracted ΔB_N, τc, and a0 (Table 6) are reliable. The experimental support rests entirely on fitting the PFR longitudinal-field curves in Figure 6(c,d). However, no quantitative comparison is provided between the dynamical model and a mono-exponential or Smirnov effective-T1 fit to the same data. The text asserts that the short-τc approximation gives correlation times 'approximately half as long' by referencing prior work [24], but does not show residuals, confidence intervals, or an information criterion for the present samples. The observable is a single curve (PFR vs. longitudinal field at a fixed 13 ns delay) per sample, while the fit involves at least four free parameters (ΔB_N,e, τc,e, ΔB_N,h, τc,h), and the relative amplitudes of electron and hole contributions are not specified as fixed or free. Without an identifiability analysis or error bars, the extracted parameters may be non-unique. If a simpler mono-exponential model fits the curves within noise, the 'necessity' claim collapses, and the localization radii would inherit the uncertainty. This is a load-bearing gap in the experimental demonstration, distinct from the acknowledged uncertainty in hyperfine constants.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper adapts the dynamical Kubo–Toyabe strong-collision model, originally developed for muon spin relaxation, to describe longitudinal hyperfine-driven spin relaxation of localized charge carriers in metal-halide perovskites for arbitrary hyperfine correlation time τc. After presenting the static MER relaxation function and the short-τc (mono-exponential) approximation, the authors derive a general relaxation function G(t) as a Poisson-weighted convolution of static relaxation functions, and compute electron and hole nuclear-field variances ΔBN for perovskite Bloch states. They predict that electron spin relaxation is strongly halogen-dependent while hole relaxation is governed by the metal cation, with lighter halogens and Sn substitution leading to longer spin lifetimes. The model is then applied to photo-induced Faraday rotation data on FAPbI3 and MAPbI3 at 13 ns delay, extracting ΔBN,e, ΔBN,h, τc,e, τc,h and, via Eqs. (19)–(20), localization radii a0. The paper concludes that bulk MHPs lie in the intermediate regime τc/T1 ∼ 1, requiring the exact dynamical treatment.","tokens_in":21042,"tokens_out":2870,"duration_ms":27813,"significance":"If the central claims hold, the paper provides a useful theoretical framework: the strong-collision Kubo–Toyabe formalism is a standard and internally consistent way to interpolate between the static MER limit and motional narrowing, and the composition trends (halogen and cation substitution) are concrete, falsifiable predictions. The use of literature hyperfine constants rather than a fit to the model is a strength: the predictions are not circular. The paper also explicitly identifies the uncertainty in the choice of atomic hyperfine constants (Ref. [47] vs. [50]) and the lack of bulk experimental hyperfine values. However, the experimental demonstration—which is the basis for the 'necessity of the exact dynamical solution' and for the extracted parameters in Table 6—is not supported by the analysis shown: no error bars, no comparison with simpler models, and no identifiability check are presented. Since the quantitative reliability of ΔBN, τc, and a0 is load-bearing for the paper's main applied claims, the manuscript needs substantial revision before acceptance.","major_comments":[{"comment":"The extracted parameters in Table 6 are the central experimental result, but the fitting procedure is not established as unique or even necessary. Each fit is to a single PFR-vs-longitudinal-field curve at a fixed 13 ns delay, with four dynamical parameters (ΔBN,e, τc,e, ΔBN,h, τc,h) plus the relative electron/hole amplitudes, which are not explicitly fixed or reported. No error bars, confidence intervals, residual plots, or parameter-correlation/identifiability analysis are given. The statement that a short-τc approximation gives 'approximately half as long' correlation times is supported only by earlier work [24], not by fits to the present data. Since the paper claims that the exact dynamical model is necessary, the authors should quantitatively compare the dynamical model with mono-exponential, Smirnov, and static-MER fits to the same data (e.g., via residuals or an information crite","section":"Section IV, Fig. 6(c,d), Eqs. (19)–(20)"},{"comment":"The hyperfine coupling constants are taken from Roothaan–Hartree–Fock calculations (Ref. [47]) because no bulk perovskite experimental values exist; the authors explicitly acknowledge that Morton–Preston values (Ref. [50]) differ. Since ΔBN^2 is proportional to A^2 (Eqs. 19–20), and a0 ∝ A^(−2/3), a factor-of-2 uncertainty in A would shift a0 by about 40% and could soften or alter the predicted halogen and cation trends in Figs. 4 and 5. The acknowledgment is honest, but the manuscript provides no quantitative sensitivity analysis. Please propagate the [47] vs. [50] difference through the predicted relaxation functions and the extracted localization radii, or give a clear argument for why the chosen set is the more reliable one in this context.","section":"Section III C, Tables 1–2"},{"comment":"The fits for FAPI and MAPI use the stable low-temperature phases (tetragonal P4/mbm and orthorhombic Pnma, Table 5) but apply Bloch states and hyperfine coupling matrices derived for a cubic structure. The text says the change will 'slightly modify' the Bloch states, but no estimate of the resulting uncertainty in ΔBN or a0 is provided. Given that the extracted localization radii are close to 5 nm and that a0 depends on ΔBN^−2/3, even a modest phase-induced change in the hyperfine matrix elements may be non-negligible. A quantitative or referenced justification for treating the cubic form as sufficient would strengthen the reliability of Table 6.","section":"Section IV and Appendix A, Eqs. (11)–(14)"}],"minor_comments":[{"comment":"Typographical errors: 'particulaly' (first sentence of the Abstract), 'te dynamical model' (Section II B), 'Ti:Saphirre' (Section IV), 'MaSnI3' in the caption of Fig. 5, and inconsistent use of 'a0, e' vs 'a0,e' in Table 6. These should be corrected before publication.","section":"Abstract and Introduction"},{"comment":"The list of three regimes in the text should be checked for numerical consistency: the text describes ωN τc = 0.01 as the short-correlation regime and Figure 1 shows curves for 0.01, 0.1, 1, 10, 100, but the discussion of Fig. 2(a) states that at ωN τc = 100 the dynamical and MER functions are 'very similar', which is consistent; however, the intermediate-regime comparison at ωN τc = 2.0 in Fig. 2(b) would benefit from specifying the parameter values and definition of reduced time more explicitly in the caption.","section":"Section II B, Eq. (5) and Fig. 2"},{"comment":"Reference [36] duplicates Reference [29] (same Hayano, Uemura, Imazato, Nishida, Yamazaki, Kubo paper) and should be consolidated. Also, the g-factor signs in Table 6 (ge and gh) are given with signs, but the ωN values are all positive; the text should clarify whether |g| is used in the Larmor-frequency conversion.","section":"References"},{"comment":"The caption and text do not specify whether the electron and hole contributions in the PFR fits were weighted by their respective spin polarizations or by independent amplitude parameters. Please clarify the amplitude treatment in the fitting procedure and state the number of data points per curve.","section":"Section IV, Table 6"}],"recommendation":"major_revision","confidential_remarks":"The theoretical part of the manuscript is solid and could be a useful contribution to the field, but the experimental section is the main weakness. The 'necessity' claim and the extracted parameters in Table 6 rest on fits to single PFR curves with no statistical comparison to simpler models and no error bars. I would ask the authors to add a quantitative model-comparison and an identifiability/uncertainty analysis; without that, the central applied claim is not demonstrated. The hyperfine-constant sensitivity is also worth addressing explicitly, although the authors have already flagged the choice honestly. This is a major-revision rather than a rejection because the concerns are fixable within the scope of the manuscript."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should read this for the model and the composition predictions; the experimental demonstration is not yet convincing.\n\nWhat's new: applying the dynamical Kubo–Toyabe/strong-collision formalism (Hayano 1979; Allodi/Renzi 2014) to localized carriers in metal-halide perovskites, with explicit formulas for the relaxation function that interpolate between the MER and short-correlation limits. The paper shows correctly that in the intermediate regime ωN τc ~ 1 the full solution differs markedly from both approximate limits. The application to MHPs gives two clean predictions: electron spin relaxation is controlled by the halogen species, so lighter halides should give longer electron lifetimes (hundreds of ns for Cl at a0=5 nm, τc=4 ns), and hole relaxation is dominated by lead, so Sn substitution should roughly double hole lifetimes. These are new and testable. The authors are honest about inputs: they state the atomic hyperfine constants are theoretical (Koh–Miller) because no bulk perovskite experimental values exist, and that predictions assume a0 and τc.\n\nThe soft spots are real. The experimental section fits two samples with two electron parameters (ΔBN,e, τc,e) and two hole parameters (ΔBN,h, τc,h) from a single PFR curve per sample (signal vs longitudinal field at 13 ns). No error bars are given on the data or the extracted parameters, no residuals are shown, and no comparison is made to mono-exponential or Smirnov effective-T1 fits of the same data. The paper says other models can fit the data, and the necessity claim rests on that being false—but that's exactly what isn't demonstrated. The relative electron/hole weighting isn't specified as fixed or free, which makes identifiability a real question. If a simpler model fits within noise, the extracted a0 and τc inherit the ambiguity. The uncertainty in the hyperfine constants is acknowledged; a factor of 2 in A changes a0 by about a factor 1.6, so it's the largest systematic.\n\nThe theoretical part is coherent, the limits are checked, and the predictions are falsifiable. The paper deserves a serious referee. It will be useful to people working on perovskite spin dynamics and hyperfine interactions. My recommendation: send to review, but the authors need to add error bars, an explicit fit comparison with residuals or an information criterion, and state the amplitude treatment. Data and code would help a lot.","headline":"A useful adaptation of the muon Kubo–Toyabe model to hyperfine spin relaxation in perovskites, with testable composition predictions, but the experimental section does not yet show the dynamical model is necessary.","tokens_in":21531,"tokens_out":4997,"would_cite":true,"duration_ms":42191,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims an exact Kubo–Toyabe relaxation function for localized carrier spins in metal-halide perovskites, and uses it to extract localization radii and correlation times from Faraday-rotation data.","keywords":["metal-halide perovskites","spin relaxation","hyperfine interaction","Kubo-Toyabe model","carrier localization","Faraday rotation","nuclear spin bath","spin dynamics"],"falsifier":"Measure the longitudinal-field Faraday-rotation recovery on MAPbI3, MAPbBr3, and MAPbCl3 films with matched localization volumes at 2 K: the model predicts electron spin lifetimes spanning from a few nanoseconds (iodine) to hundreds of nanoseconds (chlorine), so a flat halogen dependence would falsify the electron-halogen channel; alternatively, a zero-field time-resolved trace at omega_N tau_c ~ 2 should show the non-exponential shape of G(t), which no single-exponential fit can reproduce.","tokens_in":20517,"feed_emoji":"🧲","tokens_out":8691,"duration_ms":67790,"temperature":0.7,"pith_summary":"Localized electrons and holes in metal-halide perovskites relax their spins through the hyperfine field of lattice nuclei, and this paper claims that the decay can be described exactly, at any correlation time, by a dynamical version of the Kubo–Toyabe model borrowed from muon-spin relaxation. The exact relaxation function G(t) interpolates between the static Merkulov-Rosen-Efros limit and the short-correlation mono-exponential limit, and it is needed because bulk perovskites sit in the intermediate regime where those approximations fail. The model reveals distinct channels: electron spins are governed by halogen nuclei, while hole spins are governed by metal (lead) nuclei, so lighter halogens or lighter metals should give longer spin lifetimes. Fitting time-resolved Faraday rotation on MAPbI3 and FAPbI3 films yields nuclear-field widths, nanosecond hyperfine correlation times, and localization radii around 5 nm. If correct, the model turns spin-relaxation measurements into a quantitative probe of carrier localization in perovskites.","feed_headline":"Lighter halogens stretch perovskite spin lifetimes","feed_subtitle":"Electrons relax via halogens, holes via lead; the model predicts hundred-nanosecond spins in chlorine perovskites.","key_machinery":"The central object is the dynamical Kubo–Toyabe strong-collision relaxation function G(t), built by summing over random hops that reset the local hyperfine field: g^(0)(t) = exp(-nu t) g(t) and g^(n)(t) as repeated convolutions (Eqs. 8–10), with hopping rate nu = 1/tau_c. This machinery extends the static Merkulov-Rosen-Efros function g(t) to finite correlation times, and the discrete-time strong-collision numerical scheme makes it computable. The companion identities (Eqs. 19–20) connect the variance DeltaB_N^2 of the Gaussian nuclear-field distribution to the localization radius a0 and the hyperfine coupling constants.","core_discovery":"The paper's central claim is that the longitudinal spin relaxation of a localized carrier in a fluctuating nuclear field is governed by the dynamical Kubo–Toyabe relaxation function G(t), defined as the sum over all numbers of strong-collision hops n of the functions g^(n)(t) in Eqs. (8)–(10). Each hop randomly redraws the local hyperfine field from the Gaussian distribution of Eq. (3), so the function interpolates exactly between the static Merkulov-Rosen-Efros limit (long correlation time) and the short-correlation mono-exponential limit, and it differs from both when omega_N tau_c ~ 1. For metal-halide perovskites the model shows that electron spins feel mainly halogen nuclei while hole s","pith_inferences":["The predicted halogen trend is directly testable: in MAPbI3, MAPbBr3, and MAPbCl3 films with comparable localization volumes, the electron spin lifetime should scale roughly with the halogen hyperfine constants; observing a weaker trend would point to a different relaxation channel.","Because the extracted localization radii inherit the uncertainty of the atomic hyperfine constants, an independent measurement of a0 (for example from temperature dependence of the localization or from the magnetic-field dependence of the spin signal) would calibrate the constants and sharpen the composition predictions.","The strong-collision assumption—complete randomization of the nuclear field after each hop—could be relaxed in a future model; comparing the extracted tau_c values with independently estimated hopping rates would show how much of the correlation time is truly hopping versus field memory.","The framework is not limited to perovskites; any localized carrier system with a finite hyperfine correlation time (defects, nanocrystals, organic semiconductors) could be analyzed with the same G(t), turning spin-relaxation data into localization radii there too."],"forward_implications":["In bulk lead-halide perovskites, carrier spins sit in the intermediate regime tau_c/T1 ~ 1, so neither the static frozen-bath model nor the short-correlation mono-exponential formula fits the data; the full G(t) is required and recovers both as limits.","Electron spin relaxation is controlled by halogen nuclei; replacing iodine with bromine or chlorine lengthens electron spin lifetimes from nanoseconds toward hundreds of nanoseconds in MAPbCl3.","Hole spin relaxation is controlled by metal nuclei; replacing lead with tin approximately doubles the hole spin lifetime for the same localization volume.","Fitting the field-dependent Faraday-rotation signal at a fixed delay extracts the nuclear-field width DeltaB_N, the hyperfine correlation time tau_c (about 1–5 ns), and localization radii of 4.9–6.2 nm for electrons and holes in MAPbI3 and FAPbI3.","The same exact function unifies the two limiting regimes, so parameters extracted from experiments in any correlation-time regime can be compared on equal footing."],"fun_headline_variants":["Halogen weight dictates perovskite spin lifetime","Lighter halogens extend perovskite spin lifetimes","Perovskite spin duration scales with halogen size","Halogen choice controls perovskite spin relaxation","Electron spins in perovskites follow halogen mass"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The extracted localization radii and the predicted halide/cation trends rest on hyperfine coupling constants computed from atomic wavefunctions, which the authors chose because no bulk perovskite experimental values exist; if those constants are wrong by a factor of two, the radii shift by roughly 40 percent and the trends could soften.","fun_headline_variants_meta":{"raw":{"variants":["Halogen weight dictates perovskite spin lifetime","Lighter halogens extend perovskite spin lifetimes","Perovskite spin duration scales with halogen size","Halogen choice controls perovskite spin relaxation","Electron spins in perovskites follow halogen mass"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000401,"raw_usage":{"total_tokens":1950,"prompt_tokens":784,"completion_tokens":1166,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":528,"completion_tokens_details":{"reasoning_tokens":1100}},"tokens_in":528,"tokens_out":1166,"duration_ms":9463,"temperature":1.0,"reasoning_tokens":1100,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T23:47:19.446309+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the longitudinal-field Faraday-rotation recovery on MAPbI3, MAPbBr3, and MAPbCl3 films with matched localization volumes at 2 K: the model predicts electron spin lifetimes spanning from a few nanoseconds (iodine) to hundreds of nanoseconds (chlorine), so a flat halogen dependence would falsify the electron-halogen channel; alternatively, a zero-field time-resolved trace at omega_N tau_c ~ 2 should show the non-exponential shape of G(t), which no single-exponential fit can reproduce.","supporting_citations":[],"review_version":1}