{"id":"aa3ccc53-360e-401a-a45d-e620648af690","arxiv_id":"2607.19123","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In the refractory high-entropy alloy TiZrHfNbTa, high-pressure torsion lowers both electrical and thermal conductivity, but thermal conductivity partially recovers at high strain due to a rise in specific heat, while electrical conductivity does not.","lead":"This paper measures how extreme mechanical twisting changes how a five-metal alloy conducts heat and electricity. It finds electrical conductivity only keeps dropping, while heat conduction dips then partially recovers because of how dislocations and grain boundaries affect vibrations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The ~4.5% thermal-conductivity recovery between N=1 and N=10 is within the reported ±5% DSC/measurement uncertainty; without error bars the central non-monotonic claim is unsupported.","rationale":"The reader's weakest assumption is exactly the one I identify: the small thermal-conductivity recovery is treated as a real effect despite being comparable to the stated measurement uncertainty. This is indeed the most load-bearing concern because the entire narrative—partial thermal recovery without electrical recovery, attributed to the dislocation-to-grain-boundary transition and anharmonic grain-boundary vibrations—hinges on this small difference. If the recovery is scatter, the paper's main conclusion collapses to a more mundane story of monotonic degradation followed by saturation, and the Wiedemann-Franz decomposition becomes meaningless. I also considered the use of the ideal Lorenz number L0; while this is a real assumption, it would only shift the absolute electron/phonon split, not the presence or absence of the recovery itself. Similarly, the ω-phase volume fraction is unquantified, but the authors acknowledge it would reduce conductivity and thus does not drive the recovery. Therefore the measurement-uncertainty issue is the first-order concern. The paper is otherwise careful and internally consistent, and the conclusion is appropriately hedged with 'slight' and 'modest,' so the correct verdict remains CONDITIONAL. I see no reason to change the reader's verdict, so verdict_should_be is UNCHANGED.","tokens_in":13131,"tokens_out":4100,"duration_ms":47118,"concrete_test":"Remount N=1 and N=10 samples and measure CP (DSC) and α (LFA) on at least five independent specimens per condition, then compute κ = ραCp with propagated uncertainties. Perform a two-sample t-test on the N=1 vs N=10 κ values; if the 95% confidence intervals overlap or p>0.05, the claimed recovery is not statistically supported and the central conclusion should be downgraded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central novelty is the non-monotonic thermal conductivity: a 'slight recovery' from ~8.9 to ~9.3 W/m/K between N=1 and N=10 (Fig. 5b), driven by an increase in specific heat capacity (210–211 to 228–230 J/kg/K) that overcomes a slight decrease in thermal diffusivity (4.31–4.53 to 4.10–4.39 mm²/s). Because κ = ραCp, the 4.5% recovery in κ depends on the product of two measured quantities, each with substantial uncertainty. The DSC is certified at ±5% relative error, and the authors report a relative standard deviation below ±5% for repeated CP measurements; LFA uncertainty is not quantified. Propagating even conservative uncertainties (e.g., ±5% in Cp, ±3% in α) yields a relative uncertainty in κ of roughly ±6%, larger than the 4.5% recovery. No error bars or raw repeat measurements are shown in Figs. 4–5. If the N=1 vs N=10 difference is within experimental scatter, then the claimed recovery, the subsequent electron/phonon decomposition, and the dislocation-to-grain-boundary explanation are not supported. This is the load-bearing assumption of the paper, and it is not secured by the evidence presented.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports thermal and electrical transport measurements on equiatomic TiZrHfNbTa after homogenization and HPT processing for N=1 and N=10 turns. It claims that HPT reduces specific heat capacity and both thermal and electrical conductivity at low strain, then produces a slight recovery in thermal conductivity at high strain due to a transition from dislocation-dominated to grain-boundary-dominated phonon scattering, while electrical conductivity remains monotonic. Using the Wiedemann-Franz law and kinetic theory, the paper decomposes thermal conductivity into electron and phonon contributions, concluding that electrons dominate (~89% after HPT) and that the phonon mean free path decreases from 1.0 nm to 0.5 nm after one turn and partially recovers to 0.6 nm after ten turns.","tokens_in":13461,"tokens_out":4484,"duration_ms":42866,"significance":"If the reported effects are real, this is a useful contribution to a sparse literature: it provides transport data for a refractory high-entropy alloy after severe plastic deformation and offers a mechanistic explanation based on different scattering efficiencies of dislocations and grain boundaries. The electron/phonon decomposition is carried out without free parameters, using independently measured total thermal conductivity, electrical conductivity, specific heat capacity, and a literature sound velocity; the microstructural characterization by XRD and TEM is appropriate. The central claim is falsifiable and interesting. However, because the non-monotonic thermal-conductivity recovery is comparable to the stated measurement precision, the current evidence does not yet secure the paper's main conclusion.","major_comments":[{"comment":"The recovery from ~8.9 to ~9.3 W m^-1 K^-1 between N=1 and N=10 is ~4.5%, and this is the paper's central non-monotonic claim. The DSC is certified at ±5% relative error and repeated CP measurements have relative standard deviation below ±5%; no LFA uncertainty is given. Since κ = ραC_P, even a conservative ±5% uncertainty in C_P alone propagates to a ~±5% relative uncertainty in κ, which is larger than the recovery. No error bars or raw repeat data are shown in Figs. 4–5. The authors should report the per-condition repeat measurements, add error bars, and propagate uncertainties through κ = ραC_P, or temper the claim of recovery.","section":"Section 3, Fig. 5b"},{"comment":"The non-monotonic specific heat capacity is load-bearing because it is invoked as the cause of the thermal-conductivity recovery. The increase from 210–211 to 228–230 J kg^-1 K^-1 (~9%) is larger than the κ recovery, but it is still within the reported ±5% RSD envelope if that uncertainty applies per measurement. The statement that at least five repeated measurements were made is not accompanied by the measured values, error bars, or a statistical test. Without this, the partial recovery of C_P is not established, and the explanation for the κ recovery is undermined.","section":"Section 3, Fig. 4"},{"comment":"The electron-phonon decomposition uses κ_e = L0 T σ, but the temperature T is not stated explicitly. Electrical conductivity is measured at 'ambient temperature', while thermal conductivity is reported at 318 K in Fig. 5. If the two temperatures differ, the absolute values of κ_e and hence the 11–23% phonon fraction change. The paper also states 'approximately 77 of the total' without a percent sign. Please clarify T, give exact values, and report the uncertainty on the electron/phonon split.","section":"Section 4, Fig. 7"}],"minor_comments":[{"comment":"The caption says '(a-f) 1 turn and (f, g) 10 turns', but 'f' is used twice. It should probably be '(a–e) 1 turn and (f, g) 10 turns'.","section":"Fig. 3 caption"},{"comment":"'approximately 77 of the total thermal conductivity' is missing a percent sign; should read '~77%'.","section":"Section 4, paragraph 1"},{"comment":"'properties of the alloy is significantly affected' should be 'are significantly affected'.","section":"Abstract/Introduction"},{"comment":"'relative standard deviation was calculated to be smaller than ±5%' should be 'smaller than 5%' since a standard deviation cannot be negative; also specify whether the ±5% applies to absolute accuracy or repeatability.","section":"Section 2, DSC paragraph"},{"comment":"The sentence 'thermal conductivity increases with temperature, a behavior typical of disordered alloys where phonon scattering is dominant' is imprecise because the paper later concludes electron scattering dominates. Please rephrase to avoid apparent contradiction.","section":"Section 3, Fig. 5a text"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope and the methodology is largely sound. The main risk is that the central non-monotonic claim rests on a 4.5% difference that may be within the stated measurement uncertainty. I recommend requesting the raw repeated measurements, error bars, and uncertainty propagation before acceptance. There are no citation or authorship concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take on the Hidalgo-Jimenez et al. paper. It reports the first electrical and thermal transport measurements on TiZrHfNbTa after HPT, and that dataset alone is worth having. The microstructural work is solid: XRD peak broadening, TEM showing grain refinement to ~40 nm, dislocation substructures, and the omega phase after N=1. The main qualitative trends — hardness saturating, electrical conductivity dropping monotonically, and thermal conductivity decreasing after one turn — are all plausible and consistent with prior SPD work on other metals. The Wiedemann-Franz split is a standard back-of-the-envelope step, and the conclusion that electrons carry most of the heat is not surprising for a metallic alloy.\n\nThe soft spot is exactly where the stress test points. The headline claim is the partial recovery between N=1 and N=10: thermal conductivity goes from ~8.9 to ~9.3 W/m/K, a 4.5% bump that rides on a ~9% increase in CP and a ~4% decrease in thermal diffusivity. The DSC is certified at ±5% relative error and the authors state repeat measurements have RSD below ±5%, yet no error bars or individual runs appear anywhere. Propagating those uncertainties conservatively gives a relative error in κ that is comparable to or larger than the 4.5% recovery. The CP increase from 210 to 229 J/kg/K is larger than 5% and is suggestive, but the κ recovery is the load-bearing result, and it just barely clears the noise floor. The authors need to show the repeat data and propagate errors through κ = ραCp.\n\nThe second caveat is the ideal Lorenz number used for the electron/phonon decomposition. In a severely distorted BCC HEA, L0 is not guaranteed; the phonon contribution of 11–23% could shift by several points if the true Lorenz number differs. This does not change the dominance of the electronic channel, but it weakens the quantitative split.\n\nThe non-monotonic CP story is interesting and plausible, but the proposed mechanism — dislocation suppression of low-frequency modes versus grain-boundary anharmonicity — is asserted rather than tested. That's fine for a first report, but it should be framed as a hypothesis.\n\nBottom line: the paper deserves a serious referee and could be accepted with major revision. The experiment is reproducible and the data are portable. I'd ask for error bars, uncertainty propagation, a discussion of the Lorenz number assumption, and maybe one more intermediate strain state if it exists. For my own work I'd cite the baseline transport numbers, but not the recovery claim until it is secured statistically.","headline":"Useful first transport dataset for HPT-processed TiZrHfNbTa, but the headline non-monotonic recovery of thermal conductivity sits close to the stated measurement uncertainty and needs error bars before it carries the interpretation.","tokens_in":13920,"tokens_out":3334,"would_cite":true,"duration_ms":32401,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Severe torsional deformation of the refractory high-entropy alloy TiZrHfNbTa first lowers and then partially restores thermal conductivity while electrical conductivity keeps falling, showing that dislocations and grain boundaries scatter p","keywords":["high-entropy alloys","severe plastic deformation","high-pressure torsion","thermal conductivity","electrical conductivity","specific heat capacity","phonon scattering","Wiedemann-Franz law"],"falsifier":"Measure specific heat capacity and thermal diffusivity on at least five separately processed N = 1 and N = 10 discs, propagate the standard deviations through κ = αρCP, and check whether the N = 1 to N = 10 difference in κ exceeds the combined uncertainty. If the two conditions overlap within error, the claimed recovery and the dislocation-to-grain-boundary explanation lose support. Quantifying the ω-phase fraction would provide an additional check on the attribution of the recovery to grain boundaries.","tokens_in":13082,"feed_emoji":"🔥","tokens_out":7970,"duration_ms":62384,"temperature":0.7,"pith_summary":"The paper shows that high-pressure torsion (twisting a disc under several gigapascals of pressure) can decouple heat and charge transport in the refractory high-entropy alloy TiZrHfNbTa. After one turn, both specific heat capacity and thermal conductivity drop; after ten turns, heat capacity and thermal conductivity partially recover, but electrical conductivity continues to decline without recovery. The authors attribute this to a microstructural transition from dislocation-dominated to grain-boundary-dominated defect states: grain boundaries scatter electrons strongly but let more phonons through than dislocation strain fields do. Using the Wiedemann-Franz law, they find electrons carry most of the heat, with the phononic contribution between 11 and 23 percent, so the partial recovery is essentially a phonon effect. If correct, this offers a way to engineer thermal and electrical transport separately in refractory high-entropy alloys.","feed_headline":"Deforming TiZrHfNbTa alloy raises heat flow but not electron flow","feed_subtitle":"Grain boundaries still block electrons, but let phonons through, so heat flow partly returns.","key_machinery":"The decomposition relies on the Wiedemann-Franz law plus a kinetic-theory formula for the phonon mean free path. Total thermal conductivity is measured as κ = αρCP from laser-flash diffusivity, Archimedes density, and DSC heat capacity; electrical conductivity is measured by four-point probe. The electronic share is taken as κe = L0Tσ with the standard Lorenz number L0 = 2.44 × 10−8 W·Ω·K−2, and the phonon share is the remainder κp = κ − κe. The phonon mean free path then follows from κp = CPrρvsΛp/3 using a literature sound velocity. This machinery is what turns the bulk measurements into the conclusion that the high-strain recovery is vibrational, not electronic.","core_discovery":"The central claim is that severe plastic deformation does not simply degrade transport in TiZrHfNbTa. It first lowers specific heat capacity and thermal conductivity because dislocations suppress low-frequency vibrational modes and scatter both phonons and electrons. After ten turns, a nanocrystalline structure with high-angle grain boundaries forms, and heat capacity partially recovers via anharmonic vibrations at these interfaces; thermal conductivity follows a slight recovery (from about 8.9 to 9.3 W m−1·K−1) as the phonon mean free path grows from 0.5 to 0.6 nm, while electrical conductivity stays at its reduced steady state because grain boundaries remain strong electron scatterers. The","pith_inferences":["If the non-monotonic heat capacity is real, measurements at intermediate turn numbers (N = 2–5) should show a smooth dip and rise; locating the minimum would directly test the dislocation-to-grain-boundary transition.","The stated ±5% DSC error margin is comparable to the ~4.5% recovery in thermal conductivity between N = 1 and N = 10; direct measurement of CP and α on multiple specimens with propagated uncertainties would settle whether the recovery is a material effect or noise.","The paper leaves the ω-phase volume fraction unquantified; if the ω phase grows with turns, its interphase boundaries would add phonon scattering and oppose the recovery, so measuring that fraction could reconcile the magnitude of the effect.","A natural extension would be to anneal the N = 10 material at temperatures just below ω-phase dissolution: if the recovery persists after annealing, the grain-boundary explanation would be strongly supported; if it disappears, the ω phase may be more important than assumed."],"forward_implications":["The reported 77–89% electronic share of thermal conductivity means that in this refractory HEA, efforts to lower heat transport must primarily reduce electronic conduction, not phonon conduction.","Because the slight thermal recovery is driven by specific heat capacity rather than diffusivity, engineering high-angle grain boundaries is a plausible route to tune heat capacity and thermal conductivity without sacrificing electrical conductivity.","The drop in phonon mean free path from 1.0 to 0.5 nm after one HPT turn quantifies how strongly dislocations scatter phonons; recovery to 0.6 nm suggests phonons can transmit across high-angle grain boundaries.","The contrasting evolution of specific heat capacity (non-monotonic) and electrical conductivity (monotonic) implies that vibrational and electronic responses to nanostructuring are not locked together in this material."],"fun_headline_variants":["Heat recovers in deformed alloy, electrons stay stuck","Dislocations cut heat, grain boundaries restore it","Phonons recover, electrons don't in HPT alloy","Deformed alloy: heat flows back, electron flow stays low","Grain boundaries let heat through but block electrons"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the small rise in thermal conductivity between one and ten turns (from about 8.9 to 9.3 W m−1·K−1) is a genuine material response rather than experimental scatter, since a ~4.5% change sits at the edge of the paper's own ±5% DSC heat-capacity error budget and no error bars are reported.","fun_headline_variants_meta":{"raw":{"variants":["Heat recovers in deformed alloy, electrons stay stuck","Dislocations cut heat, grain boundaries restore it","Phonons recover, electrons don't in HPT alloy","Deformed alloy: heat flows back, electron flow stays low","Grain boundaries let heat through but block electrons"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000234,"raw_usage":{"total_tokens":1346,"prompt_tokens":772,"completion_tokens":574,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":496}},"tokens_in":516,"tokens_out":574,"duration_ms":5021,"temperature":1.0,"reasoning_tokens":496,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T13:21:15.339367+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure specific heat capacity and thermal diffusivity on at least five separately processed N = 1 and N = 10 discs, propagate the standard deviations through κ = αρCP, and check whether the N = 1 to N = 10 difference in κ exceeds the combined uncertainty. If the two conditions overlap within error, the claimed recovery and the dislocation-to-grain-boundary explanation lose support. Quantifying the ω-phase fraction would provide an additional check on the attribution of the recovery to grain boundaries.","supporting_citations":[],"review_version":1}