{"id":"06801d8f-d2f5-479a-b462-876fb3bb38d3","arxiv_id":"2505.20169","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A benchmark of ten classical force fields identifies REAXFF R4/R3 and MEAM M1/M2 as the most reliable for simulating titanium carbide and nitride MXene structures and elastic constants, and reports the first computed linear compressibilities for these MXenes.","lead":"The authors tested ten classical molecular dynamics force fields to see which ones correctly reproduce the structure and stiffness of six titanium MXene sheets, comparing against density functional theory results from the literature. Only four force fields passed their tests, and the study reports new values for the linear compressibility of these 2D materials.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Force-field ranking is threshold-dominated: the 33.3% thickness criterion admits borderline structures, so the 'adequate' list and derived compressibilities are not robust.","rationale":"The paper is a useful and largely reproducible benchmarking study: the MD protocols (minimization cycles with NVE kinetic-energy convergence checks, controlled stretching) are described in unusual detail, and the authors are transparent about force fields that fail or produce distorted structures. The qualitative conclusion that the force fields behave differently and that no single potential works for all MXenes is well supported. The concern I find most load-bearing is not the use of DFT references (their spread is acknowledged and ranges are given), but the arbitrariness of the pass/fail thresholds, especially the 33.3% thickness criterion. Several force fields that are later called 'adequate' pass only because of this very permissive threshold, and a stricter but still reasonable threshold would remove some of them from the final list. This directly affects the central claim that R4, R3, M1, and M2 are the force fields researchers should use. The reader identified threshold sensitivity and the derived C66 as weaknesses; I agree with threshold sensitivity and additionally note the unexplained exclusion of R2 for Ti3C2, which reinforces that the final selection is not transparently derivable from the stated criteria. The derived-C66 issue is real but less central because the paper explicitly states Eq. (3) and uses it; it mainly affects the reporting of G values, not the force-field ranking. Overall, the qualitative ranking is likely correct, but the specific adequacy list and the numerical elastic/compressibility values should be stated as conditional on the chosen thresholds, with a sensitivity check. Therefore the verdict should remain CONDITIONAL rather than unchanged or full acceptance.","tokens_in":27429,"tokens_out":6190,"duration_ms":62442,"concrete_test":"Re-run the Section 2.4.2 acceptance tests using stricter, literature-standard thresholds (e.g., thickness deviation ≤10% instead of 33.3%, lattice deviation ≤2% instead of 5%) on the same Tables 5 and 6 data, and recompute the Section 4 list. If the set of adequate force fields changes, for example if M1 is dropped for Ti3C2/Ti4C3 or R5 for Ti2C, then the central ranking is threshold-controlled and must be reported with a robustness statement.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central deliverable is the list of force fields declared adequate (Section 4: R4 for Ti2C/Ti3C2/Ti4C3, R3 for Ti2N, M1/M2 for Ti3C2/Ti4C3/Ti3N2/Ti4N3). That list depends directly on the acceptance thresholds defined in Section 2.4.2: lattice parameter within 5% of the DFT interval extremes, and thickness deviating by no more than 33.3%. The thickness threshold is extremely loose. R5 passes Ti2C with Δt/t = 33.3% (exactly the limit); M1 passes Ti2C with 31.2%, Ti3C2 with 24.6%, and Ti4C3 with 18.2%. If a more standard 10% thickness tolerance were used, M1 would be removed from Ti3C2 and Ti4C3, and R5 from Ti2C, altering the headline recommendation. No sensitivity analysis is provided, so the adequacy claims are not robust to a plausible change in criterion. A related internal inconsistency: for Ti3C2, Table 8 shows R2 and R4 have identical C11 (391 N/m) and R2's C12 (10 N/m) is closer to the DFT value (40 N/m) than R4's (105 N/m), yet Section 4 lists R4 and M1 while excluding R2 with no stated justification. This shows the final selection is not fully determined by the published tests. Since the linear compressibility values are computed from elastic constants of these selected force fields, the 'first time' novelty claim inherits the same threshold sensitivity.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript benchmarks ten classical molecular dynamics force fields (COMB3 C1/C2, REAXFF R1–R6, MEAM M1/M2) against literature DFT data for six titanium MXenes (Ti2C, Ti3C2, Ti4C3, Ti2N, Ti3N2, Ti4N3). It proposes three sequential tests: qualitative structural regularity, quantitative agreement of lattice parameter and thickness with DFT intervals, and comparison of elastic constants. On this basis it identifies R4 (Ti2C, Ti3C2, Ti4C3), R3 (Ti2N), M1 (Ti3C2, Ti4C3), and M2 (Ti3N2, Ti4N3) as the adequate force fields, and it uses these to compute Young's modulus, Poisson's ratio, shear modulus, and linear compressibility, claiming the first reported linear compressibilities for titanium MXenes.","tokens_in":27680,"tokens_out":10273,"duration_ms":95031,"significance":"If the force-field ranking is robust, the paper provides a practical selection guide for classical MD simulations of titanium MXenes, an area where users currently choose potentials without systematic comparison. The study is also transparent in important respects: the MD protocols for C11, C22, and C12 are described step by step; the evaluation is against external DFT data; no parameters are fitted. However, the claimed new β values are algebraic derivatives of previously published Cij, the selection of “adequate” force fields depends on hand-chosen tolerances, and the C66 values are not obtained by a shear simulation. These issues affect the reliability of the central deliverable and need to be addressed before the benchmark can be used with confidence.","major_comments":[{"comment":"No MD protocol for shear strain is described; Section 2.3 contains only uniaxial (2.3.2) and biaxial (2.3.3) stretching protocols. The reported C66 and G values are therefore evidently obtained from Eq. (3), i.e., derived from C11, C12, and C22, not simulated. This should be stated explicitly in Sections 2.3 and 4, and Eq. (3) is not validated by the symmetry and stability tests. For M1 Ti2C, where C11 ≠ C22 by about 20%, Eq. (3) gives C66 = 34 N/m while the usual hexagonal in-plane relation (C11 − C12)/2 gives about 42 N/m; the paper does not justify which formula applies. The shear modulus G in Table 10 is thus not an independent MD prediction.","section":"Section 2.3 and Section 4 (Tables 8–10)"},{"comment":"The thickness acceptance criterion is stated inconsistently. Section 2.4.2 says that MD values of both a0 and t must fall within the DFT interval or deviate by no more than 5%, with an additional thickness criterion of 33.3%. Section 3.1.3 and Tables 4–6 apply 5% for a0 and 33.3% for t. Under the literal text of Section 2.4.2, M1 would fail for Ti2C, Ti3C2, and Ti4C3 because Δt/t = 31.2%, 24.6%, and 18.2% all exceed 5%, and R3 would fail for Ti2N with Δt/t = 6.6%. The 33.3% threshold is also very loose; R5 passes Ti2C at exactly 33.3%. No sensitivity analysis is provided, so the final list of adequate force fields is not robust to a plausible stricter thickness criterion such as 10%.","section":"Section 2.4.2, Section 3.1.3, Tables 4–6"},{"comment":"The third test does not, as described, determine the final selection. No quantitative scoring rule is given for the elastic-constant comparison. For Ti3C2, Table 8 shows R2 and R4 have the same C11 (391 N/m), yet R2's C12 (10 N/m) is much closer to the DFT value (40 N/m) than R4's (105 N/m), so the exclusion of R2 and selection of R4 is unexplained. For Ti4C3, R4's C11 (521 N/m) is far outside the DFT range [312–366] and is much worse than M1's 365 N/m, yet both are listed as adequate in Section 4. A reproducible selection requires either an explicit scoring metric or a clear statement that the final list is a subjective judgement.","section":"Section 3.2.2 and Section 4"},{"comment":"The DFT reference data are incomplete, yet Table 10 reports DFT-derived E, β, ν, and G for Ti3N2 and Ti4N3. Table 2 lists only C11 for these structures (263 and 369 N/m, respectively), but the Table 10 footnote indicates the DFT elastic quantities were computed from Cij values in the references. The full set of DFT Cij used should be tabulated with provenance, including any conversion from 3D GPa values to 2D N/m. In addition, the literature C11 values have substantial spread (e.g., 130–151 N/m for Ti2C, 312–366 N/m for Ti4C3, acknowledged in Section 2.4.3), and the ranking should be shown to be stable with respect to the choice of DFT reference or to the spread itself.","section":"Tables 2 and 10"}],"minor_comments":[{"comment":"The text refers to \"Tin+1Cn\" when the test also applies to nitrides; it should read \"Tin+1Xn\" or explicitly include both carbide and nitride MXenes.","section":"Section 2.4.2"},{"comment":"The Poisson ratios νxy and νyx are described as \"ratios of the areas under the stress-strain curves,\" which is not the standard definition; Poisson's ratio is the negative ratio of transverse to axial strain.","section":"Section 2.2"},{"comment":"The stability condition lists \"C33 > 0,\" which is not defined for the 2D elastic constants used in this paper; the relevant conditions are C11 > 0, C22 > 0, C11C22 − C12^2 > 0, and C66 > 0.","section":"Section 2.4.3"},{"comment":"The column header reads \"a0, and thickness, a0, and thickness, t\"; the duplicated phrase should be removed.","section":"Table 5"},{"comment":"The word \"Figura\" appears in the caption; it should be \"Figure.\"","section":"Figure 3 caption"},{"comment":"The footnote for the symbol \"–\" says it indicates that a0 and/or t are within the DFT range, but the symbol appears in both the Δa0/a0 and Δt/t columns; the footnote should apply to both columns.","section":"Section 3.1.4 and Table 5"},{"comment":"The claim that the linear compressibility values \"were presented for the first time\" should be qualified: since β is an algebraic combination of Cij via Eq. (2), and the Cij have been published in the cited DFT works, the contribution is the compilation and MD-based evaluation of these quantities, not the discovery of new data.","section":"Abstract and Section 4"},{"comment":"The phrase \"square symmetry\" is inaccurate for MXenes, which are hexagonal; the expected equality C11 = C22 follows from the in-plane symmetry conditions, and the terminology should be corrected.","section":"Section 2.4.3"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about threshold sensitivity is verified: applying a 10% thickness tolerance would remove M1 from the Ti3C2 and Ti4C3 lists and R5 from Ti2C, so the headline recommendations are not robust. The absent C66 protocol and the unexplained R2/R4 selection in Table 8 are the most concrete technical gaps. The 'first time' novelty claim is weak because β is a derived quantity from already published Cij. The paper fits the journal's scope and the benchmark is worth publishing after the selection procedure is made fully reproducible."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Honest take: this is a useful, clearly-written paper that practitioners will want on file. It systematically benchmarks ten force fields (COMB3, REAXFF, MEAM) against literature DFT for six titanium MXenes, and the authors are transparent about failures. The protocols for structure optimization and in-plane elastic constants are reproducible in detail, and the qualitative ranking—R4 for Ti2C/Ti3C2/Ti4C3, R3 for Ti2N, M1/M2 for the thicker carbides/nitrides—looks plausible. The paper deserves a serious referee.\n\nThat said, the central deliverable, the list of declared 'adequate' force fields, is softer than it appears. The acceptance thresholds in Section 2.4.2 do a lot of work: the 33.3% thickness tolerance is extremely loose. R5 passes Ti2C exactly at the limit; M1 passes Ti3C2 and Ti4C3 with thickness errors of 24.6% and 18.2%. If a more standard 10% tolerance were used, M1 would drop out of Ti3C2/Ti4C3 and R5 out of Ti2C. No sensitivity analysis is provided, so the headline recommendation is not robust to a plausible change in criterion. This matters because the linear compressibility values, presented as 'first time,' inherit the same threshold sensitivity.\n\nThere is also an internal inconsistency in the selection logic. For Ti3C2, Table 8 shows R2 and R4 give identical C11 (391 N/m), and R2's C12 (10 N/m) is much closer to the DFT value (40 N/m) than R4's (105 N/m) — yet R4 is selected and R2 is excluded without any stated justification. Either the selection criteria need to be shown to exclude R2, or R2 should be listed. Also, no shear protocol for C66 is described anywhere; the reported C66 values match Eq. (3) derived from the fitted C11, C12, C22. That is not a fatal flaw—it is a legitimate isotropic approximation—but it should be labeled as derived, not simulated.\n\nMinor but real: the DFT 'ground truth' has a wide spread (e.g., Ti4C3 C11 from 312 to 366 N/m) and no error bars, so the force-field ranking is only as good as the chosen reference interval. No input files, force field parameters, or data tables are provided, which limits independent checking.\n\nWho is this for? Anyone doing classical MD on Ti MXenes who needs a starting point for force-field choice. The qualitative guidance is useful even if the exact thresholds are arbitrary. I would accept this for peer review, with a request for sensitivity analysis, a justification of the R2/R4 choice, and explicit labeling of the derived C66. The core benchmark is valuable and the manuscript is honestly written, but the claimed precision needs qualification.","headline":"A genuinely useful force-field benchmark for Ti MXenes, but the 'adequate' list is threshold-sensitive and the shear moduli are derived, not simulated.","tokens_in":28285,"tokens_out":1366,"would_cite":true,"duration_ms":15691,"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":"Four of ten MD force fields reproduce titanium MXene structure and elasticity.","keywords":["MXenes","titanium carbide","titanium nitride","molecular dynamics","COMB3","REAXFF","MEAM","linear compressibility"],"falsifier":"Rerunning the MD protocol for Ti$_3$C$_2$ with R4 and M1 but extracting $C_{66}$ from a direct shear deformation instead of the symmetry relation would settle whether the reported shear moduli are trustworthy; equivalently, reapplying the structure tests with a 2% lattice-parameter threshold would immediately show whether R4, R3, M1, and M2 survive as the only adequate force fields.","tokens_in":1740,"feed_emoji":"🧪","tokens_out":2010,"duration_ms":100633,"temperature":0.7,"pith_summary":"This paper asks whether three classical molecular-dynamics potentials, COMB3, REAXFF, and MEAM, with their published parameter sets, ten force fields in total, can reproduce the lattice parameter, thickness, and elastic constants of six bare titanium MXenes: Ti$_2$C, Ti$_3$C$_2$, Ti$_4$C$_3$, Ti$_2$N, Ti$_3$N$_2$, and Ti$_4$N$_3$. Benchmarking against DFT values from the literature, it finds that none of the ten force fields works for all six, but four are dependable: REAXFF R4 for the three carbide MXenes, REAXFF R3 for Ti$_2$N, MEAM M1 for Ti$_3$C$_2$ and Ti$_4$C$_3$, and MEAM M2 for Ti$_3$N$_2$ and Ti$_4$N$_3$. Using those four, the paper computes Young's modulus, Poisson's ratio, shear modulus, and linear compressibility for all six MXenes, reporting the linear compressibility values as new reference data. A researcher planning MD simulations of titanium MXenes can use this screening to choose a parameter set instead of testing potentials from scratch.","feed_headline":"Four of ten MD force fields pass titanium MXene tests","feed_subtitle":"Match DFT structure and stiffness, so MD users can pick the right parameter set; first linear compressibility values included.","key_machinery":"The screening is carried by an MD protocol adapted from graphyne elastic-constant calculations. A minimization cycle, energy minimization, box relaxation, then a short NVE run, is repeated until the accumulated kinetic energy stays below $10^{-6}$, giving the optimized structure and $a_0$, thickness, and $U_0$. A stretching protocol then applies ten 0.1% tensile steps along x, along y, or biaxially, keeping the transverse box fixed, fits the energy-strain curves to parabolas to obtain $C_{11}$, $C_{22}$, and $C_{12}$, and obtains $C_{66}$ through the square-symmetry relation $C_{66} = \\frac{1}{4}(C_{11} - 2C_{12} + C_{22})$; these constants feed the formulas for Young's modulus, linear compressibility, Poisson's ratio, and shear modulus.","core_discovery":"The central claim is that a three-stage screening identifies the only MD parameter sets worth using for bare titanium MXenes. In the first stage many force fields fail outright: they flatten the sheet to one atom thick, amorphize it, give multiple lattice parameters, or crash with NaN errors. In the second stage, lattice parameter and thickness are compared with DFT intervals, requiring agreement within 5% for $a_0$ and within 33.3% for thickness. In the third stage, elastic constants from strain-energy fits are checked against DFT, together with $C_{11}=C_{22}$ symmetry and positive-definite strain energy. The surviving fields are R4, R3, M1, and M2, while COMB3 either fails structurally or severely overestimates stiffness. The paper also claims first-reported linear compressibility values for all six titanium MXenes.","pith_inferences":["The same screening could be run on terminated MXenes with O, OH, F, or Cl surface groups, since termination changes both relaxed geometry and stiffness; the paper treats only bare sheets.","Because $C_{66}$ is derived from a symmetry identity rather than a direct shear deformation, the reported shear moduli inherit any systematic error in $C_{11}$, $C_{12}$, and $C_{22}$; a direct shear-strain simulation would settle that.","Future experimental measurements, such as nanoindentation of single MXene flakes, would provide an independent check on which recommended force field is truly predictive rather than merely consistent with DFT."],"forward_implications":["Choose R4 for bare Ti$_2$C, Ti$_3$C$_2$, and Ti$_4$C$_3$; choose R3 for Ti$_2$N; choose M1 for Ti$_3$C$_2$ and Ti$_4$C$_3$; choose M2 for Ti$_3$N$_2$ and Ti$_4$N$_3$.","No single published force field describes all six titanium MXenes, so simulation studies must switch parameter sets by material.","The reported $\\beta_x$ and $\\beta_y$ linear compressibilities for all six MXenes can be cited as new reference values for 2D titanium carbide and nitride.","COMB3's surviving structures overestimate stiffness so strongly that COMB3 should not be used for elastic predictions, even where it passes structural tests.","Earlier MD studies using other REAXFF fields or COMB3/C2 may still be reasonable for structure, but their elastic conclusions are not supported by this screening."],"supporting_citations":[{"why":"Supplies the MEAM parameter sets M1 (Ti-C) and M2 (Ti-N) that pass the screening for the thicker MXenes.","marker":"[84]"},{"why":"Supplies the REAXFF R4 parameter set found adequate for Ti2C, Ti3C2, and Ti4C3.","marker":"[79]"},{"why":"Supplies the REAXFF R3 parameter set found adequate for the titanium nitride MXenes, especially Ti2N.","marker":"[78]"},{"why":"Supplies the REAXFF R1 parameter set that is tested and rejected for amorphizing Ti2C and crashing on nitrides.","marker":"[77]"},{"why":"Supplies the REAXFF R2 parameter set that reproduces carbide structures but fails the elastic-constant comparison.","marker":"[58]"},{"why":"Provides DFT structural and elastic reference values for titanium nitride and carbide MXenes used in the second and third tests.","marker":"[37]"},{"why":"Provides DFT lattice-parameter and elastic-constant values for early transition metal carbides, including Ti2C, Ti3C2, and Ti4C3.","marker":"[36]"},{"why":"Provides DFT elastic constants for titanium carbide MXenes used to judge the elastic agreement.","marker":"[47]"},{"why":"Supplies the minimization-and-stretching MD protocol that the paper adapts to compute optimized structures and elastic constants.","marker":"[88]"},{"why":"Identifies the LAMMPS package in which all simulations are run.","marker":"[89]"}],"fun_headline_variants":["Only 4 of 10 MD force fields pass Ti MXene tests","Ti MXene force fields: 4 pass, COMB3 fails","Best MD force fields for Ti MXenes identified","First linear compressibility values for Ti MXenes","MD force fields for Ti MXenes: 4 winners"],"cache_read_input_tokens":30336,"weakest_assumption_plain":"The ranking rests on trusting published DFT numbers as the exact reference and on hand-chosen pass thresholds; if those numbers or thresholds are not the right benchmark, the list of recommended force fields changes.","fun_headline_variants_meta":{"raw":{"variants":["Only 4 of 10 MD force fields pass Ti MXene tests","Ti MXene force fields: 4 pass, COMB3 fails","Best MD force fields for Ti MXenes identified","First linear compressibility values for Ti MXenes","MD force fields for Ti MXenes: 4 winners"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000545,"raw_usage":{"total_tokens":2653,"prompt_tokens":1038,"completion_tokens":1615,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":654,"completion_tokens_details":{"reasoning_tokens":1532}},"tokens_in":654,"tokens_out":1615,"duration_ms":12928,"temperature":1.0,"reasoning_tokens":1532,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:57:33.047330+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Rerunning the MD protocol for Ti$_3$C$_2$ with R4 and M1 but extracting $C_{66}$ from a direct shear deformation instead of the symmetry relation would settle whether the reported shear moduli are trustworthy; equivalently, reapplying the structure tests with a 2% lattice-parameter threshold would immediately show whether R4, R3, M1, and M2 survive as the only adequate force fields.","supporting_citations":[{"cited_title":"Kim, B.-J","cited_arxiv_id":null,"evidence_quote":"Supplies the MEAM parameter sets M1 (Ti-C) and M2 (Ti-N) that pass the screening for the thicker MXenes."},{"cited_title":"Huygh, A","cited_arxiv_id":null,"evidence_quote":"Supplies the REAXFF R4 parameter set found adequate for Ti2C, Ti3C2, and Ti4C3."},{"cited_title":"Monti, V","cited_arxiv_id":null,"evidence_quote":"Supplies the REAXFF R3 parameter set found adequate for the titanium nitride MXenes, especially Ti2N."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the REAXFF R1 parameter set that is tested and rejected for amorphizing Ti2C and crashing on nitrides."},{"cited_title":"Kurtoglu, M","cited_arxiv_id":null,"evidence_quote":"Provides DFT lattice-parameter and elastic-constant values for early transition metal carbides, including Ti2C, Ti3C2, and Ti4C3."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides DFT elastic constants for titanium carbide MXenes used to judge the elastic agreement."}],"review_version":1}