{"id":"565457fd-d474-483e-8947-9f303cfb74d0","arxiv_id":"2512.00512","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"In γ-TiAl, niobium atoms predominantly replace titanium but a rising minority replace aluminum, and those defects lower stacking-fault energies (aiding twinning) while raising Peierls stress (adding strength).","lead":"This paper uses computer simulations to show that niobium atoms in a titanium-aluminum alloy partly occupy aluminum sites, creating defects that simultaneously make the material stronger and more ductile. It explains a long-running disagreement between experiments and theory about how niobium improves this aerospace alloy.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"NNP transferability to dislocation cores and defect-SFE landscapes is the linchpin; without new DFT cross-checks, the strengthening ordering and SFE-reduction mechanism are unverified.","rationale":"The reader's weakest_assumption identifies exactly the most load-bearing concern: the study's conclusions depend on the transferability of a previously published NNP to properties—especially dislocation core energetics and composition-dependent stacking fault energies—that were not explicitly validated in the original potential development. This is more serious than a mere lack of code sharing because it directly underdetermines both halves of the 'dual role' claim: the SFE reduction attributed to NbAl/TiAl antisites and the strengthening ordering extracted from Peierls stress calculations. The manuscript provides no new DFT cross-checks, and the Peierls stress results are based on single random configurations for each defect type, making the quantitative ordering fragile. Nevertheless, the paper has genuine strengths: the MCMD method is appropriate, the site-occupancy results align with experimental ALCHEMI/FIM observations, and the qualitative trend that Ti-rich alloys (with more TiAl antisites) have lower SFE is consistent with experimental alloy comparisons. The concern is addressable via targeted DFT validation and configurational sampling, so a conditional verdict is appropriate rather than rejection. The reader's conditional verdict should therefore stand unchanged, with the listed requests (particularly DFT cross-checks and configurational averaging) as necessary conditions for full acceptance.","tokens_in":17267,"tokens_out":6959,"duration_ms":74177,"concrete_test":"Perform DFT (e.g., VASP/PBE) calculations of the GSFE curves for pure γ-TiAl and for representative 8 at.% NbTi, NbAl, TiAl, and AlTi defect cells, and compute the Peierls stress of a screw dislocation in a small supercell with the same defect arrangements using both DFT and the NNP. If the NNP's Peierls stress deviates from DFT by more than 50% or the defect-strengthening ordering changes, the central strengthening mechanism is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that NbAl and TiAl antisites simultaneously lower stacking fault energies and raise Peierls stresses enough to explain experimental SFE drops and Ti-rich strengthening rests entirely on the accuracy of the NNP of Ref. [17] for properties it was not explicitly validated on. The GSFEs for defect-containing cells (Fig. 4) and the Peierls stresses (Fig. 6) are computed with no new DFT cross-checks. Peierls stress is extremely sensitive to the dislocation core description; solute–dislocation interaction energies and barriers depend on the potential's fidelity in highly strained regions, which are rarely in the training set. The SFE trend is also nontrivial: the NNP gives a pure γ-TiAl SISF of ~152 mJ/m2, which is substantially above the cited experimental ~97 mJ/m2, so the absolute scale is already questionable. Moreover, the Peierls stress for each 8 at.% defect type is obtained from a single random defect arrangement in a single dislocation model, with no configurational averaging or sampling; thus the claimed ordering AlTi > TiAl > NbAl > NbTi may be an artifact of that one realization. If the NNP misrepresents either the defect–SFE coupling or the dislocation–defect interaction, both the mechanism and the ordering are unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses a previously developed neural network potential (NNP, ref. 17) to perform hybrid Monte Carlo/molecular dynamics (MCMD) simulations of Nb-doped γ-TiAl, together with generalized stacking fault energy (GSFE) and Peierls-stress calculations. It reports that Nb predominantly occupies Ti sites and forms short-range order with Al, but that a non-negligible fraction occupies Al sites (NbAl) and promotes TiAl antisite formation. The authors find that NbTi increases stacking fault energies while NbAl and TiAl decrease them, and that all these defects increase the Peierls stress of screw and edge dislocations. They conclude that the combined presence of NbAl and TiAl explains both the experimentally observed SFE reduction in Ti-rich high-Nb alloys and their higher strength relative to Al-rich alloys, thus resolving a long-standing DFT/experiment discrepancy.","tokens_in":17554,"tokens_out":4374,"duration_ms":47250,"significance":"If correct, the proposed mechanism is significant: it reconciles the experimental observation that Nb lowers SFE with DFT results showing NbTi raises SFE, by invoking minority NbAl and TiAl defects whose populations grow with Nb content. The study also provides a unified explanation for simultaneous strengthening and ductilization, with design implications for defect and composition engineering. The paper's strengths include large-scale NNP-based simulations, comparison of lattice parameters and site occupancy with experiments, and identification of a concrete, falsifiable microscopic mechanism. However, the central quantitative predictions rest entirely on NNP transferability to dislocation cores and defect-SFE landscapes that were not explicitly validated, and the paper lacks configurational averaging and a direct quantitative link between the MCMD defect populations and the reported experimental SFE drop.","major_comments":[{"comment":"The central claims rest on NNP predictions for properties not validated in ref. 17. No new DFT cross-checks are provided for defect-containing GSFE curves (Fig. 4) or Peierls stresses (Fig. 6, S10). The NNP pure γ-TiAl SISF (~152 mJ/m²) is ~50% above the experimental value (~97 mJ/m²) cited in the Introduction, yet this discrepancy is neither discussed nor benchmarked. Since Peierls stress is highly sensitive to the dislocation-core description and solute–dislocation interactions in highly strained regions, the quantitative ordering AlTi > TiAl > NbAl > NbTi and the absolute SFE trends are not established. The authors should provide DFT validation for at least representative defect-containing GSFE points and for the pure screw/edge core energetics, or explicitly state the associated uncertainty and treat conclusions as qualitative.","section":"§Models and Methods; §Defect effects on dislocation motions"},{"comment":"Each GSFE–concentration curve and each Peierls stress is computed from a single random defect arrangement. For 8 at.% defects and a finite dislocation cell, the local distribution of solutes around the dislocation line strongly affects the measured stress; the serrated flow in Fig. 6a further indicates configuration-dependent unlocking events. Without averaging over multiple independent random configurations and reporting standard deviations, the claimed strengthening ordering (AlTi > TiAl > NbAl > NbTi) may be an artifact of one realization. Similarly, the monotonic GSFE trends in Fig. 4 should be verified with at least a few independent configurations per concentration.","section":"§Peierls stress calculations; §Effect of defects on plastic deformation"},{"comment":"The paper proposes that NbAl and TiAl together cause the experimental SFE reduction, but it never computes the SFE for an MCMD-equilibrated alloy configuration at realistic defect populations. Figure 2d shows NbAl is only ~16% of Nb atoms at 10 at.% Nb, and the TiAl population in Fig. 2a-c is not quantified in the text; the isolated-defect trends in Fig. 4 are not combined to reproduce the measured drop from ~97 to ~34 mJ/m² for 10 at.% Nb. To make the central claim load-bearing, the authors should compute GSFEs for MCMD-derived configurations at the relevant temperatures and Nb contents, and compare quantitatively with experiment. This would also test whether the minority defect populations are sufficient to overcome the opposing effect of the dominant NbTi.","section":"§Several factors on site occupation; §Effect of defects on plastic deformation"}],"minor_comments":[{"comment":"The system size for the MCMD simulations is not stated in the main text. Please report the number of atoms and the simulation-cell dimensions, as the defect counts in Fig. 1d-f are otherwise hard to interpret.","section":"§Hybrid Monte Carlo and molecular dynamics simulations"},{"comment":"Equation (2) uses notation 'nWCP' and subscripts 'mn' in a confusing way; the Warren–Cowley parameter should be written as χ_mn with a clear definition of the summation over neighbor shells.","section":"§Characterization of short-range ordering"},{"comment":"Equation (3) is missing the explicit definition of the fault-plane area A and the displacement vector u in the text; also, E_fault(u) is used but not defined before the equation.","section":"§General stacking fault energies calculations"},{"comment":"In the 'Comparison with experimental results on site occupation' section, the text says 'as summarized by Hu et al.33' but reference 33 is Diao et al. Please correct the citation.","section":"§References"},{"comment":"The statement 'available from the corresponding authors upon request' would be strengthened by depositing the NNP parameters, input scripts, and analysis codes in a permanent repository, especially since the conclusions depend on a specific machine-learned potential.","section":"§Data and Code Availability"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents a plausible mechanism and a substantial simulation effort, but the central quantitative claims are currently under-validated. The lack of DFT cross-checks for the NNP on dislocation cores and defect-containing GSFE surfaces is a serious risk, as is the reliance on single random configurations. These are fixable with additional calculations, but they are not merely presentation issues. I recommend major revision rather than rejection because the overall direction is promising and the requested validations are within the scope of the study."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read. The paper gives a unified story: Nb prefers Ti sites, but the minority NbAl and TiAl antisites lower stacking fault energies (hence ductility) and raise Peierls stress (hence strength). That would explain why Ti-rich high-Nb TiAl alloys are both stronger and more ductile. It's an attractive resolution of a long-standing discrepancy, and the MCMD defect-population results are new and plausible: NbAl fraction rises from ~4.5% to ~16% as Nb goes from 2 to 10 at%, with TiAl antisites growing too. The SFE trends—NbTi raises, NbAl and TiAl lower—are consistent with earlier DFT and with the experimental trend for Ti-rich alloys. The lattice parameter comparisons with experiment are a nice validation.\n\nThe soft spots are real but addressable. The whole calculation sits on the NNP from the same group (ref 17), and the paper states it 'captures the key effects of Nb doping on stacking fault energies and formation energies'—meaning the SFE-reduction trends are largely reproducing the DFT data used in training, not independent predictions. The absolute SISF for pure gamma-TiAl comes out around 152 mJ/m2, versus the ~97 mJ/m2 experimental value cited in the introduction; that 50% mismatch is never mentioned. For the Peierls stress, each defect type is simulated in a single random arrangement in a single dislocation model, with no configurational averaging or error bars, so the ordering AlTi > TiAl > NbAl > NbTi could be an artifact of that one realization. There are also no new DFT cross-checks for dislocation cores or defect-SFE landscapes, and no data or code are shipped.\n\nNone of this kills the paper. The mechanism is plausible and the MCMD site-occupancy predictions are a real contribution. But the central claim—that NbAl and TiAl defects explain both SFE reduction and strengthening—would be much stronger if the authors (1) calculated SFEs directly on MCMD-equilibrated structures, (2) repeated Peierls-stress runs with several defect arrangements and reported error bars, (3) validated the NNP against DFT for at least a few dislocation and fault configurations, and (4) released the NNP and input files.\n\nThis paper deserves a serious referee—the topic is important and the issues are fixable. I'd send it to peer review with a request for major revision. Not a desk reject.","headline":"Plausible mechanism for Nb's dual role in gamma-TiAl, but the case hinges on the authors' own NNP, and the key Peierls-stress and SFE predictions are not cross-checked against DFT or experiments.","tokens_in":18100,"tokens_out":2953,"would_cite":false,"duration_ms":28584,"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":"Niobium's simultaneous strengthening and ductilizing of γ-TiAl stems from a minority population of Nb_Al substitutions and Ti_Al antisites, not from Nb on Ti sites alone.","keywords":["γ-TiAl alloys","niobium alloying","antisite defects","stacking fault energy","Peierls stress","solid-solution strengthening","deformation twinning","neural network potential"],"falsifier":"Compute the generalized stacking fault energy and Peierls stress from first principles for the same 8 at.% Nb_Al, Ti_Al, and Al_Ti random configurations; if ab initio results show these defects do not lower the relevant stacking fault energies and raise the Peierls stress in the order Al_Ti > Ti_Al > Nb_Al > Nb_Ti, the proposed dual-role mechanism fails.","tokens_in":17129,"feed_emoji":"⚙️","tokens_out":3925,"duration_ms":37688,"temperature":0.7,"pith_summary":"This paper argues that the long-standing puzzle of high-niobium γ-TiAl—why adding Nb raises both strength and ductility—can be traced to defect chemistry rather than to niobium's main substitution site. Atomistic simulations with a neural network potential show that most Nb atoms occupy Ti sites, but a fraction occupies Al sites and, at Ti-rich compositions, promotes Ti_Al antisite defects. Those minority defects sharply lower stacking fault energies and thus encourage deformation twinning, while all the defects raise the Peierls stress that dislocations must overcome. The result is a single mechanism that resolves the apparent contradiction between DFT calculations, which found Nb on Ti sites raises stacking fault energy, and experiments showing Nb lowers it.","feed_headline":"A defect minority explains niobium's dual strength-ductility boost","feed_subtitle":"Simulations trace both effects to Nb on Al sites and Ti_Al antisites lowering fault energy while raising Peierls stress.","key_machinery":"The central object is the defect population in the L1_0 γ-TiAl lattice: substitutional Nb on Ti sites (Nb_Ti), substitutional Nb on Al sites (Nb_Al), and antisite defects Ti_Al and Al_Ti. The paper uses a machine-learned neural network potential to drive hybrid Monte Carlo/molecular dynamics sampling of site occupancy, generalized stacking fault energy calculations on (111) planes, and Peierls stress calculations for 1/2[1-10] screw and edge dislocations at 8 at.% defect concentrations. The link between the two mechanical outcomes—ductility and strength—is that Nb_Al and Ti_Al lower both stable and unstable stacking fault energies, promoting twinning, while the same defects raise the stress","core_discovery":"On the paper's own terms, the central discovery is that the combined presence of Nb_Al antisite substitutions and Ti_Al antisite defects—present as a non-negligible minority in high-Nb, Ti-rich γ-TiAl—simultaneously lowers stacking fault energies and raises the Peierls stress of both screw and edge dislocations. This explains the experimentally observed drop in stacking fault energy with Nb content and the pronounced solid-solution strengthening, especially in Ti-rich alloys. The ordering of strengthening effectiveness is predicted to be Al_Ti > Ti_Al > Nb_Al > Nb_Ti > defect-free, and Ti-rich high-Nb alloys contain larger populations of the strongly strengthening defects, explaining why the","pith_inferences":["If the mechanism is correct, small populations of antisite defects may matter more than the majority solute site for mechanical behavior; composition design should optimize defect chemistry, not just solute type.","A direct experimental test would be atom-probe or channeling measurements quantifying Nb_Al occupancy and Ti_Al concentration in Ti-rich high-Nb alloys aged at 600–900 °C, to see whether the minority populations reach the levels the simulations predict.","The paper's reliance on a neural network potential leaves an open check: first-principles calculations of stacking fault curves and Peierls barriers for random 8 at.% defect configurations would confirm or overturn the predicted ordering Al_Ti > Ti_Al > Nb_Al > Nb_Ti."],"forward_implications":["The observed reduction of stacking fault energies in high-Nb γ-TiAl can be attributed to Nb_Al and Ti_Al antisites rather than to Nb on Ti sites, resolving the DFT–experiment discrepancy.","Ti-rich high-Nb alloys should show stronger solid-solution strengthening than Al-rich alloys because they harbor more Nb_Al and Ti_Al defects.","Controlling site occupancy through Nb content, stoichiometry, and heat treatment can tune short-range order and defect populations to balance twin-induced plasticity with dislocation hardening.","The mechanism generalizes beyond γ-TiAl: site-selective doping to create antisite defects could be used to tailor stacking fault energies and Peierls stresses in other ordered intermetallics."],"fun_headline_variants":["Niobium's rare antisites deliver both strength and ductility to TiAl","Two-for-one niobium defects explain TiAl's strength-ductility combo","Defect minority doubles TiAl's strength and ductility","Niobium antisites: the dual-action boost for TiAl alloys","How a few niobium antisites grant TiAl both strength and ductility"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole argument rests on the previously trained neural network potential being accurate for defect-rich dislocation cores and stacking fault landscapes, a regime the paper does not revalidate here with new first-principles calculations.","fun_headline_variants_meta":{"raw":{"variants":["Niobium's rare antisites deliver both strength and ductility to TiAl","Two-for-one niobium defects explain TiAl's strength-ductility combo","Defect minority doubles TiAl's strength and ductility","Niobium antisites: the dual-action boost for TiAl alloys","How a few niobium antisites grant TiAl both strength and ductility"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000856,"raw_usage":{"total_tokens":3539,"prompt_tokens":714,"completion_tokens":2825,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":2730}},"tokens_in":458,"tokens_out":2825,"duration_ms":21704,"temperature":1.0,"reasoning_tokens":2730,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T19:24:54.576411+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the generalized stacking fault energy and Peierls stress from first principles for the same 8 at.% Nb_Al, Ti_Al, and Al_Ti random configurations; if ab initio results show these defects do not lower the relevant stacking fault energies and raise the Peierls stress in the order Al_Ti > Ti_Al > Nb_Al > Nb_Ti, the proposed dual-role mechanism fails.","supporting_citations":[],"review_version":1}