{"id":"03d2f764-0fe7-4d5f-98b2-e658047a3a74","arxiv_id":"2506.15574","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Calculations with the PBE0 functional predict bulk NbOI2 has the largest piezoelectric stress coefficient (6.32 C/m2) among NbOX2 (X=Cl, Br, I), exceeding a cited PZT value.","lead":"This paper uses density functional theory to predict that the layered material NbOI2 has a very large piezoelectric response of 6.32 C/m2, higher than the other two niobium oxide dihalides and than a reference PZT value. A generalist reader might care because it points to a lead-free candidate for converting mechanical stress into electricity, though the result is a functional-dependent computational prediction.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Headline claim that NbOI2 'beats PZT' rests on a PZT benchmark (-4.7 C/m2) that appears to be the transverse e31 of a PZT film, not PZT's largest coefficient; standard PZT e33 values exceed 6.32 C/m2.","rationale":"The reader's CONDITIONAL verdict was based on structural and functional uncertainty. Those are legitimate concerns, but they are addressable with additional validation and do not by themselves show the headline claim is wrong. The more decisive issue is the PZT benchmark: the abstract's 'higher than PZT' clause depends on -4.7 C/m2 being PZT's largest piezoelectric coefficient, which is very likely a misreading of a transverse thin-film e31 value. This concern is external, concrete, and testable by reading ref. 35 and standard PZT data. I therefore keep the reader's CONDITIONAL verdict, but for a more specific reason: the manuscript's central comparison is unsupported as written, even though the underlying DFT data may be reproducible.","tokens_in":16648,"tokens_out":14305,"duration_ms":147020,"concrete_test":"Retrieve Cattan et al. (ref. 35) and identify which tensor component carries the -4.7 C/m2 value (section, setup, and symbol used; it is very likely e31,f of a thin film). Then obtain an accepted bulk PZT e33 value from a standard source (e.g., Berlincourt's tables or PZT-5A/PZT-5H datasheets). Recompute the abstract comparison with that e33. If e33(PZT) exceeds 6.32 C/m2, the authors should remove the 'even PZT' claim and replace it with a comparison against the same tensor component of a lead-free layered family; the data may otherwise stand.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is not just the DFT number; it is the statement that bulk NbOI2 is 'around 31% higher than ... even lead zirconate titanate.' The only PZT benchmark offered is -4.7 C/m2, attributed to ref. 35. In the piezoelectric literature, that value is the transverse e31 (or e31,f) coefficient of a PZT thin film, not PZT's largest piezoelectric coefficient. PZT's longitudinal stress coefficient e33 is typically 10-30 C/m2 (for example, PZT-5A has e33 around 15.8 C/m2 and PZT-5H around 23 C/m2). The computed NbOI2 coefficient, e22 = 6.32 C/m2, is also longitudinal (polar b-axis, yy strain). Comparing NbOI2 e22 with PZT e31,f mixes a longitudinal with a transverse coefficient. Against the correct PZT e33, 6.32 C/m2 is lower, not 31% higher. The 'even PZT' clause is therefore unsupported even if Table V is numerically correct. This is a benchmark-level error, not an attack on the DFT calculation itself.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a density functional theory study of bulk niobium oxide dihalides NbOX2 (X = Cl, Br, I) using the PBE and PBE0 functionals in CRYSTAL17. The authors compute optimized lattice parameters, band structures and gaps, elastic constants, and mechanical properties such as hardness, machinability, melting temperature, and Debye temperature, and they use the Berry phase method to obtain piezoelectric stress tensors. The headline result is a large piezoelectric coefficient e22 = 6.32 C/m2 for NbOI2 at the PBE0 level, which the abstract claims is about 31% higher than the other two compounds and even higher than lead zirconate titanate (PZT). The paper concludes that NbOI2 is a promising lead-free piezoelectric material for energy harvesting.","tokens_in":16860,"tokens_out":7537,"duration_ms":65441,"significance":"If the reported e22 value is correct and the comparison with PZT is properly framed, bulk NbOI2 would be a notable lead-free piezoelectric with a coefficient rivaling the best known materials. The work has the merit of using a hybrid functional and the Berry phase method, and it provides a consistent set of elastic and mechanical properties for three related layered compounds. However, the central claim is weakened by three issues: the PZT benchmark used is a thin-film transverse coefficient rather than PZT's largest coefficient; the computed piezoelectric coefficient varies by about 2.5 C/m2 across functionals (Table VI); and the optimized structures for NbOCl2 and NbOBr2 deviate significantly from earlier reported structures (Section III.A). These issues must be resolved before the headline claim is acceptable.","major_comments":[{"comment":"The comparison with PZT is based on the value -4.7 C/m2 attributed to ref. 35, which is the transverse e31,f coefficient of a PZT thin film, not the largest piezoelectric coefficient of PZT. PZT's longitudinal coefficient e33 is typically 10-30 C/m2 (for example, about 15.8 C/m2 for PZT-5A and about 23 C/m2 for PZT-5H). The computed NbOI2 coefficient e22 = 6.32 C/m2 is also longitudinal (yy strain). Comparing a longitudinal coefficient with a transverse one is inappropriate, and the claim that NbOI2 exceeds PZT is therefore unsupported as stated. The authors should compare e22 with the appropriate PZT e33 values or rephrase the claim to avoid the misleading 'even PZT' statement.","section":"Section I (also abstract and conclusion)"},{"comment":"The reported e22 for NbOI2 varies strongly with functional: 3.8 C/m2 (PBE), 5.3 C/m2 (PBE-D3), and 6.32 C/m2 (PBE0). The paper presents only the PBE0 value as the headline without discussing this spread or providing error estimates. In the absence of any experimental bulk data for these compounds, the authors should justify why PBE0 is quantitatively reliable for these van der Waals layered materials, or at least report the full functional range and discuss its effect on the conclusions regarding NbOI2's superiority.","section":"Section III.D and Table VI"},{"comment":"The optimized lattice parameters for NbOCl2 and NbOBr2 are nearly 1 Angstrom smaller along one axis than previously reported optimized structures (refs. 43 and 57). Since the elastic constants and piezoelectric coefficients are computed from these geometries, a structural discrepancy of this magnitude could substantially change the calculated properties. No experimental diffraction data or comparison with other functionals is provided to validate the chosen ground-state structure. The authors should demonstrate that the piezoelectric coefficients are robust to this structural difference, or use a validated structure and discuss the impact on the central claim.","section":"Section III.A"},{"comment":"The ferroelectric distortion is described as arising from Nb displacements along the a-axis, yet the largest piezoelectric coefficient is reported for the yy strain component (e22). Please clarify the relation between the polar axis and the Cartesian axes used in the tensor, and verify that e22 is indeed the appropriate longitudinal coefficient for the polar direction. If the polarization axis is along a, a large e11 might instead be expected; this point is important for interpreting the physical meaning of the headline value.","section":"Section III.A and III.D"}],"minor_comments":[{"comment":"Several entries are printed as '0.00e-01', which should simply be 0.00; also the value 4.995 C/m2 for NbOCl2 in Table V is reported as 4.99 C/m2 in Table VI and in the text. Unify the notation for the reported coefficients.","section":"Table V"},{"comment":"The text refers to the Born stability criteria as 'Eqs. ??', indicating a missing cross-reference. The equations in Eq. (2) should be properly referenced or numbered.","section":"Section III.C (elastic stability)"},{"comment":"Equations (8) and (9) for sound velocities are dimensionally inconsistent: the density should be inside the square root rather than dividing the square root of the modulus. The resulting velocity values in Table III should be recalculated. In addition, the mean velocity Vm for NbOCl2 (1250 m/s) is smaller than the transverse velocity Vt (2517 m/s), which is unphysical; this indicates an arithmetic error.","section":"Eqs. (8) and (9), Table III"},{"comment":"There are several typos and formatting issues, including 'Vicker's Hardness' (should be 'Vickers'), 'Kleinman' spelled incorrectly in two places, and 'F unctionals' in the header of Table VI. The reference list also contains numerous entries (refs. 86-156) that do not appear to be cited in the text; these should be removed or properly cited.","section":"Throughout the manuscript"},{"comment":"The statement that PBE-GGA underestimates the band gap 'because it fails to account for the vdW interactions' is not accurate; PBE's gap underestimation is primarily due to self-interaction error. This sentence should be corrected to avoid a misleading mechanistic explanation.","section":"Section III.B"}],"recommendation":"major_revision","confidential_remarks":"The use of a non-peer-reviewed preprint from the same group (ref. 11) as a comparison benchmark for 'Li-based perovskites' is questionable; it should be replaced by peer-reviewed data or omitted. The inflated reference list suggests the manuscript needs careful editorial trimming. If the structural and functional-dependence issues are addressed and the PZT comparison is corrected, the paper could be suitable for publication as a computational materials science study, but the current claim of superiority over PZT should not appear in the abstract or conclusion in its present form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper adds PBE0 hybrid-functional values and a full elastic/mechanical characterization for bulk NbOX2 (X = Cl, Br, I), but the headline claim that NbOI2 \"even beats PZT\" does not survive a look at the benchmark. The -4.7 C/m2 figure, from ref. 35, is a transverse e31 coefficient of a PZT thin film, not PZT's largest longitudinal coefficient. Standard PZT e33 values are 10-30 C/m2, so 6.32 C/m2 is lower than PZT, not 31% higher. That is a benchmark-level error in the abstract and conclusion, and it is the paper's main selling point.\n\nWhat is genuinely useful: the three halides are compared consistently under PBE, PBE-D3, and PBE0, and the qualitative ordering NbOI2 > NbOBr2 ~ NbOCl2 is stable across all three functionals. The piezoelectric tensor is computed with Berry phase, no fitting to the target result. The elastic constants, moduli, hardness, and Debye temperature are standard but complete for this family. If you need a quick PBE0 data point for bulk NbOX2 piezoelectrics, this is a reasonable source.\n\nThe soft spots are real. The optimized cells for NbOCl2 and NbOBr2 are nearly 1 Å smaller along the a-axis than previously reported optimizations, and no experimental diffraction data are used to validate the ground-state geometry. Piezoelectric coefficients are sensitive to polar displacement patterns, so this is not a cosmetic detail. The functional spread for NbOI2 is 3.8 to 6.32 C/m2, and no error bars are given. The comparison to prior bulk values in ref. 44 is already in the literature, so the new quantitative content is mostly the PBE0 column and the mechanics.\n\nAlso, the reference list contains a long run of unrelated entries (carbenes, quantum optics, supernova cosmology, etc.), which looks like a bibliography-template accident and undermines confidence in the manuscript's care level.\n\nFor a reader working on NbOX2 or lead-free piezoelectrics, this paper is a minor data point. It does not demonstrate a material that surpasses PZT, and the structure issue needs addressing before the numbers can be trusted. As a desk decision, I would send this back for major revision at best; the benchmark error alone is enough to preclude acceptance as written, and the incremental novelty makes it a weak candidate for a serious referee unless the authors fix the PZT comparison and validate the structures.","headline":"Routine DFT extension of known NbOX2 piezoelectrics whose headline 'beats PZT' claim is unsupported because it compares against a transverse PZT film coefficient, not PZT's longitudinal e33.","tokens_in":17460,"tokens_out":2259,"would_cite":false,"duration_ms":24823,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.15.Mb","77.65.-j","62.20.-x"],"model":"deepseek-v4-flash","headline":"Bulk NbOI2 is predicted to convert stress to charge with a piezoelectric coefficient of 6.32 C/m², beating PZT.","keywords":["NbOX2","niobium oxide dihalides","piezoelectricity","PBE0 hybrid functional","lead-free piezoelectrics","elastic constants","Berry phase","NbOI2"],"falsifier":"Grow bulk $\\mathrm{NbOI_2}$, apply a known uniaxial strain along the crystallographic $b$-axis, and measure the induced surface charge; if $e_{22}$ comes out near $6.32\\,\\mathrm{C/m^2}$, the central claim survives, while a value near the PBE-level $3.8\\,\\mathrm{C/m^2}$ would show the hybrid-functional enhancement is an artifact. A cheaper check is diffraction refinement of the crystal structure: if the niobium off-centering or the $\\mathrm{Nb}$–$\\mathrm{Nb}$ dimer alternation differs from the $C2$ model used here, the polarization calculation would have to be redone.","tokens_in":16449,"feed_emoji":"⚡","tokens_out":9115,"duration_ms":80406,"temperature":0.7,"pith_summary":"This paper uses density functional theory with a hybrid functional to predict the structural, electronic, elastic, and piezoelectric properties of three layered niobium oxide dihalides, $\\mathrm{NbOX_2}$ with $X=\\mathrm{Cl},\\mathrm{Br},\\mathrm{I}$. The central finding is that bulk $\\mathrm{NbOI_2}$ converts mechanical stress to electric charge more strongly than the benchmark piezoelectric ceramic PZT, with a computed piezoelectric stress coefficient $e_{22}=6.32\\,\\mathrm{C/m^2}$. The authors trace the large response to iodine's lower electronegativity raising the dynamical charges, on top of an off-center displacement of niobium inside distorted octahedra. If the prediction is right, $\\mathrm{NbOI_2}$ is a lead-free, layered, mechanically soft candidate for sensors, actuators, and energy harvesting.","feed_headline":"Niobium oxide iodide beats PZT at stress-to-charge conversion","feed_subtitle":"DFT predicts lead-free NbOI2 converts stress to charge at 6.32 C/m², about 31% above the PZT benchmark.","key_machinery":"The load-bearing setup is density functional theory with the PBE0 hybrid functional, in which polarization is obtained through the Berry-phase approximation and piezoelectric coefficients are evaluated as numerical derivatives of the Berry phase with respect to finite strains applied to the monoclinic $C2$ cell. The physical mechanism that creates the polar ground state is the pseudo-Jahn-Teller distortion: each niobium sits off-center in a distorted $\\mathrm{NbO_2X_4}$ octahedron, forming $\\mathrm{Nb}$–$\\mathrm{O}$ dipole pairs, while $\\mathrm{Nb}$–$\\mathrm{Nb}$ dimerization along the $b$-axis stabilizes the semiconducting state. The specific computed quantity that carries the headline claim is $e_{22}$, the change in $y$-direction polarization per unit $yy$ strain, whose value distinguishes $\\mathrm{NbOI_2}$ from its siblings and from PZT.","core_discovery":"The paper's central claim is that among bulk $\\mathrm{NbOX_2}$ ($X=\\mathrm{Cl},\\mathrm{Br},\\mathrm{I}$) in the monoclinic $C2$ structure, $\\mathrm{NbOI_2}$ has the largest direct piezoelectric stress coefficient, $e_{22}=6.32\\,\\mathrm{C/m^2}$ from the PBE0 hybrid functional. This is about 31% higher than the computed values for $\\mathrm{NbOCl_2}$ ($4.99\\,\\mathrm{C/m^2}$) and $\\mathrm{NbOBr_2}$ ($4.69\\,\\mathrm{C/m^2}$), and it exceeds the magnitude of the lead zirconate titanate value ($-4.7\\,\\mathrm{C/m^2}$) cited in the paper. The response is concentrated in the $y$-component of polarization under $yy$ strain; all other strain components give responses an order of magnitude smaller. The authors attribute the enhancement to higher dynamical charges in the iodide, which they link to iodine's lower electronegativity and the smaller off-center displacement in that compound.","pith_inferences":["Going beyond the paper: if the bulk $6.32\\,\\mathrm{C/m^2}$ value survives measurement, the contrast with the monolayer value $e_{22}=31.6\\times10^{-10}\\,\\mathrm{C/m}$ suggests that layer stacking amplifies rather than screens the polarization response; measuring flakes of increasing thickness would test this directly.","Going beyond the paper: the strong functional dependence of the coefficient ($6.32\\,\\mathrm{C/m^2}$ with PBE0 versus $3.8\\,\\mathrm{C/m^2}$ with PBE) makes exact-exchange mixing a testable ingredient; hybrid-functional calculations on isostructural iodide or bromide compounds would show whether this enhancement is systematic.","Going beyond the paper: since the response is concentrated in one tensor component, the natural energy-harvesting geometry for a macroscopic crystal is a uniaxial stretch or bend along the $b$-axis, a design pointer the paper does not spell out."],"forward_implications":["If the prediction holds, bulk $\\mathrm{NbOI_2}$ is a lead-free material with a piezoelectric stress response roughly 31% stronger than the PZT reference, and its layered structure leaves it open to exfoliation or composite integration.","With PBE0 band gaps between $2.13$ and $2.42$ eV, the three compounds are semiconductors, so a single crystal could in principle pair piezoelectric charge generation with light absorption or photovoltaic operation.","The negative Cauchy pressures and predicted brittleness mean any practical device would need encapsulation or protective coatings before repeated mechanical cycling.","Because the elastic constants satisfy the Born stability conditions, the three crystals are mechanically stable despite their weak van der Waals interlayer bonding.","The machinability index marks $\\mathrm{NbOCl_2}$ and $\\mathrm{NbOI_2}$, but not $\\mathrm{NbOBr_2}$, as suitable for device fabrication."],"supporting_citations":[{"why":"establishes the layered $C2$ crystal structure with coexisting ferroelectric and antiferroelectric phases and supplies the earlier optimized lattice parameters against which the paper's cells differ.","marker":"[43]"},{"why":"provides the monolayer $\\mathrm{NbOI_2}$ piezoelectric coefficient and the PBE and PBE-D3 bulk values that frame the functional dependence of the result.","marker":"[44]"},{"why":"supplies the cited PZT piezoelectric coefficient of $-4.7\\,\\mathrm{C/m^2}$ that the headline comparison beats.","marker":"[35]"},{"why":"reports synthesis of $\\mathrm{NbOI_2}$ crystals, grounding the material's experimental availability.","marker":"[57]"},{"why":"gives the $\\mathrm{Nb}$–$\\mathrm{Nb}$ and $\\mathrm{Nb}$–$\\mathrm{O}$ bond alternation distances that define the polar distortion used in the calculations.","marker":"[45]"},{"why":"supports the claim that dynamical charges increase as halide electronegativity decreases, explaining why the iodide is the strongest piezoelectric.","marker":"[85]"}],"fun_headline_variants":["Lead-free NbOI2 outdoes PZT piezoelectricity by 31%","Niobium iodide oxide sets piezoelectric record at 6.32 C/m²","DFT finds NbOI2 beats PZT in piezoelectric response","NbOI2: lead-free piezoelectric outperforms PZT","Iodide niobium oxide surpasses PZT's piezoelectric constant"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The prediction assumes that the optimized monoclinic $C2$ structures are the true ground states; the paper itself notes that its $\\mathrm{NbOCl_2}$ and $\\mathrm{NbOBr_2}$ cells are nearly 1 Å shorter along one axis than earlier optimized structures, and it validates the geometries only against other calculations, not against experimental diffraction data.","fun_headline_variants_meta":{"raw":{"variants":["Lead-free NbOI2 outdoes PZT piezoelectricity by 31%","Niobium iodide oxide sets piezoelectric record at 6.32 C/m²","DFT finds NbOI2 beats PZT in piezoelectric response","NbOI2: lead-free piezoelectric outperforms PZT","Iodide niobium oxide surpasses PZT's piezoelectric constant"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000549,"raw_usage":{"total_tokens":2609,"prompt_tokens":919,"completion_tokens":1690,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":535,"completion_tokens_details":{"reasoning_tokens":1598}},"tokens_in":535,"tokens_out":1690,"duration_ms":11716,"temperature":1.0,"reasoning_tokens":1598,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:32:22.052948+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Grow bulk $\\mathrm{NbOI_2}$, apply a known uniaxial strain along the crystallographic $b$-axis, and measure the induced surface charge; if $e_{22}$ comes out near $6.32\\,\\mathrm{C/m^2}$, the central claim survives, while a value near the PBE-level $3.8\\,\\mathrm{C/m^2}$ would show the hybrid-functional enhancement is an artifact. A cheaper check is diffraction refinement of the crystal structure: if the niobium off-centering or the $\\mathrm{Nb}$–$\\mathrm{Nb}$ dimer alternation differs from the $C2$ model used here, the polarization calculation would have to be redone.","supporting_citations":[{"cited_title":"Young , author P","cited_arxiv_id":null,"evidence_quote":"supports the claim that dynamical charges increase as halide electronegativity decreases, explaining why the iodide is the strongest piezoelectric."}],"review_version":2}