{"id":"ceb54a27-100c-4855-9497-c1b3f9b1fe95","arxiv_id":"1908.00464","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In a simulated binary metallic glass, annealing closer to the glass transition temperature lowers the potential energy and raises the shear modulus and yield stress, while higher annealing temperatures produce ductile, homogeneous deformation.","lead":"Molecular dynamics simulations of a model metallic glass show that changing the annealing temperature before a rapid quench changes the glass's internal energy, structure, and how it deforms under shear. The results connect a processing variable to a switch between brittle shear-band failure and more homogeneous, ductile flow.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Ta–fictive temperature identification is not established at low Ta: incomplete relaxation at Ta=0.32–0.34 means the final glass state depends on annealing time and cooling history, so the reported low-Ta trends may not isolate Ta.","rationale":"The reader's weakest assumption identifies the same issue, and I agree. The paper is a careful MD study of a standard KA model with a plausible protocol, and the high-Ta data (Ta ≥ 0.38) are likely equilibrated and support the qualitative trend of lower energy and more ordered structure with decreasing Ta. The concern is limited to the low-Ta end, where the paper itself admits incomplete relaxation and a non-monotonic U(Ta) below Ta≈0.30. Because the paper includes Ta=0.32 and 0.34 in its key figures and in the 'approaches Tg from above' narrative, the equating of Ta with fictive temperature is not established there. The proposed test—varying the annealing time—would directly settle whether the low-Ta results are a unique function of Ta or an artifact of the fixed 2e5 tau window. This does not overturn the paper's qualitative conclusions for well-annealed states, so the appropriate verdict remains CONDITIONAL, unchanged from the reader's.","tokens_in":9035,"tokens_out":8051,"duration_ms":78759,"concrete_test":"Repeat the preparation protocol at Ta=0.32 and Ta=0.34 with annealing times of 2e5, 1e6, and 5e6 tau (continuing until the drift in U during the annealing interval is below the 15-sample run-to-run noise), then quench to TLJ=0.01 and measure U, the shear modulus G, and the yielding peak σY from the stress-strain curve at one strain rate (e.g., 1e-5). If G or σY changes by more than the 15-sample standard error between annealing times, the final glass state is not determined by Ta alone and the Ta=fictive temperature assumption fails at low Ta; if properties are insensitive, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise is that the instantaneous quench from Ta to TLJ=0.01 preserves the structure equilibrated at Ta, so that Ta can be equated with the fictive temperature. Section III states this explicitly: 'In the limiting case of an infinitely fast quenching rate, used in the present study, the fictive temperature essentially coincides with the annealing temperature.' This requires the system to be at equilibrium at Ta before the quench. The paper's own Fig. 1 shows that the potential energy continues to decrease during the 2x10^5 tau annealing interval at Ta <= 0.36, and the reported MSD of 1.64 sigma^2 at Ta=0.32 over the same interval indicates substantial structural relaxation. The text further reports (without showing data) that for Ta=0.28-0.30, U measured at 0.01 increases when Ta is reduced, which is attributed to an increased effective cooling rate from the initial melt. This non-monotonicity demonstrates that the final glass energy is not a single-valued function of Ta; it depends on the entire cooling and aging history. Since Figs. 3 and 4 include Ta=0.32 and 0.34 (below Tg≈0.35), the mechanical trends at low Ta could reflect partial aging rather than the equilibrium structure at Ta. The central claim that thermal history alone, specifically Ta, sets the energy, short-range order, and mechanical properties is therefore not fully supported without showing that longer annealing or an independent fictive-temperature measurement leaves the trends unchanged.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports molecular-dynamics simulations of the Kob-Andersen 80:20 binary Lennard-Jones mixture (N = 60,000) aimed at isolating the effect of the annealing temperature Ta on the structure and mechanical response of the resulting glass. The preparation protocol is: cool from T = 1.0 to Ta in 10^4 tau, hold at Ta for 2 x 10^5 tau, quench instantaneously to T = 0.01, relax for 10^4 tau, then shear at constant volume at strain rates from 10^-5 to 10^-3 tau^-1. The paper reports that the potential energy after quenching decreases as Ta approaches Tg ≈ 0.35 from above; that the nearest-neighbor peak of g_BB becomes less pronounced at lower Ta; that the shear modulus and stress-overshoot peak increase strongly as Ta decreases toward Tg; that the shear modulus becomes strain-rate dependent only for Ta > Tg; and that nonaffine-displacement snapshots show a transition from shear-band localization at Ta = 0.32 to homogeneous deformation at Ta = 0.50. The authors frame these results as evidence that thermal history, specifically the fictive temperature set by Ta, controls the energy, short-range order, and mechanical properties of the model metallic glass.","tokens_in":9368,"tokens_out":7161,"duration_ms":70477,"significance":"If the central premise is accepted, the paper provides a clean computational demonstration of fictive-temperature control over the mechanical properties of a model metallic glass, complementing recent experimental work on toughening transitions. The qualitative trends—higher energy and more disordered structure for higher Ta, larger modulus and yield stress near Tg from above, and a brittle-to-ductile crossover—are physically reasonable and are supported by the stress-strain curves, energy curves, and nonaffine snapshots. The paper also has strengths in explicit multi-sample averaging over 15 independent samples, a well-defined preparation protocol, and a clear attempt to connect the results to the fictive-temperature framework. However, the load-bearing identification of Ta with the fictive temperature is not established at low Ta, where structural relaxation is incomplete; because the Ta = 0.32 and 0.34 data in Figs. 3-5 are in this regime, the claimed annealing-temperature trends are not fully isolated from aging and cooling-history effects. The omitted low-Ta data and the absence of error bars in Fig. 4 further limit the quantitative robustness of the transition claim.","major_comments":[{"comment":"The claim that Ta equals the fictive temperature is load-bearing and is not supported for Ta ≤ 0.36. Figure 1 shows that the potential energy continues to decrease during the entire 2 x 10^5 tau annealing interval at these temperatures, and the reported MSD of about 1.64 sigma^2 at Ta = 0.32 indicates ongoing structural relaxation. Therefore the instantaneous quench does not freeze an equilibrated Ta structure, and the trends for Ta = 0.32 and 0.34 in Figs. 3-5 may reflect partial aging during the finite annealing window rather than the equilibrium state at Ta. The authors should demonstrate that longer annealing (or an independent measurement of the fictive temperature, e.g., from the inherent-structure energy) leaves the reported trends unchanged, or restrict the central claims to the equilibrated range.","section":"III, after Fig. 1; fictive-temperature identification"},{"comment":"The text states that for 0.28 ≤ Ta ≤ 0.30 the potential energy U measured at TLJ = 0.01 increases as Ta is decreased, but these data are not shown. This nonmonotonicity means that the potential energy is not a single-valued function of Ta, and the abstract's statement that 'glasses prepared at higher annealing temperatures are relocated to higher energy states' is therefore valid only over the restricted range 0.32-0.50. The omitted data should be included, and the abstract and conclusions should be qualified accordingly.","section":"III, inset to Fig. 1; abstract"},{"comment":"No error bars or other measures of statistical uncertainty are given for sigma_Y and G despite the statement that the data are averaged over 15 independent samples. The central quantitative claims—that the mechanical properties change sharply when Ta crosses Tg and that the shear modulus becomes strain-rate dependent only for Ta > Tg—cannot be assessed without an indication of the sample-to-sample scatter. Error bars (or a table of mean plus or minus standard error) should be added to Fig. 4 and its inset.","section":"III, Fig. 4"},{"comment":"The value Tg ≈ 0.35 is taken from the author's prior work [13] and was obtained with a cooling-rate protocol (10^-5 epsilon/kB tau); this value is then used to locate the mechanical transition in Fig. 4. Since the glass transition temperature depends on the cooling rate and the thermal protocol, the location of the observed transition relative to Tg should either be confirmed within the present preparation protocol or discussed as potentially shifted relative to the cited value.","section":"III, definition of Tg"}],"minor_comments":[{"comment":"Exponents are frequently missing in the text (e.g., '10 4tau' and '2 x 10 5tau'); these should be typeset as 10^4 tau and 2 x 10^5 tau.","section":"Throughout"},{"comment":"The sentence containing 'the nonaddine measure' should read 'the nonaffine measure'.","section":"III, nonaffine-displacement definition"},{"comment":"Reference [39] on laser welding of glasses appears unrelated to the discussion of fictive temperature; please verify whether it was intended to be cited or replace it with a more relevant reference.","section":"References"},{"comment":"The brittle-to-ductile interpretation is based on snapshots from individual samples; the authors should state whether the deformation-mode change was observed consistently in all 15 samples.","section":"III, Figs. 5-7"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies heavily on the author's own prior publications for the glass transition temperature and for the interpretation of nonaffine displacements. This is not disqualifying in a focused line of work, but the editor may wish to request that the Tg value be independently derived or compared with literature values for the Kob-Andersen model. The paper fits the journal's scope, and the main concern is the incomplete equilibration at low Ta rather than any fundamental inconsistency in the deformation data."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nYou should know this is a straightforward MD study of how the annealing temperature Ta before a rapid quench changes the energy, structure, and mechanics of the Kob-Andersen glass. It is not a headline result, but it is a useful systematic map: lower Ta gives deeper energy, more ordering in g_BB, higher shear modulus and yield stress, and a transition from shear-band localization to homogeneous flow as Ta rises above Tg. The trends line up with the experimental fictive-temperature picture, which is nice but not new; parts overlap with the author's earlier heat-treatment MD papers. The cleanest contribution is the Ta scan at fixed quench protocol, with 15 independent samples for the key quantities.\n\nThe paper does several things well. The stress-strain curves, energy traces, and nonaffine displacement snapshots are mutually consistent. The strain-rate dependence of G only for Ta>Tg is an interesting detail that matches experiment. The writing is clear and the protocol is reproducible in principle.\n\nNow the soft spots, mostly minor-to-moderate. Fig. 4 has no error bars despite averaging 15 samples; that should be easy to add. The low-Ta data (Ta=0.28–0.30) are mentioned but not shown, and the nonmonotonicity in U at 0.01 deserves a figure because it reveals that the effective cooling rate changes, not just Ta. The bigger conceptual issue is equating Ta with fictive temperature. The instantaneous quench freezes structure, but at Ta=0.32–0.34 the system is clearly still aging: MSD of 1.64 σ² over 2×10⁵τ means atoms are making cage jumps. So the lowest-Ta points may partly reflect aging history rather than the equilibrium state at Ta. That does not break the qualitative trend, but it does mean the paper overstates “thermal history alone” unless it can show the trends are robust to longer annealing or an independent fictive-temperature measurement.\n\nTg is taken from a self-cited paper; acceptable given the same model and parameters, but a direct determination in this work would be cleaner. No code or data are deposited, which limits reuse.\n\nBottom line: the central claim holds up in its main lines. This paper deserves a serious referee, not a desk reject. I would send it out with a request for error bars, the missing low-Ta data, and a careful discussion of the equilibration issue.\n\nBest,\n[You]","headline":"A competent, systematic MD scan of annealing temperature in the Kob-Andersen glass; the qualitative trends are believable, but the low-Ta data and the Ta=fictive-temperature assumption need more support before the strong claims are published as-is.","tokens_in":9864,"tokens_out":1917,"would_cite":true,"duration_ms":21026,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":null,"created_at":"2026-08-14T15:54:13.449277+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":null,"supporting_citations":[],"review_version":1}