{"id":"a1f786f8-a577-4a72-a0e2-29d6f39c6e41","arxiv_id":"2607.05854","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"First-principles calculations reveal that electrostriction and elastic modulus in wurtzites are strongly polarization-dependent, causing the piezoelectric response to be highly nonlinear and overturning the constant-Q approximation.","lead":"This paper uses first-principles DFT to show that the electrostriction coefficient in wurtzite materials like AlN is strongly polarization-dependent, overturning the standard assumption that it is constant. This matters because it reveals a new design principle for engineering piezoelectric and ferroelectric devices based on wurtzite crystals.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"The 1D reduction is appropriate for d333; the main quantitative risk is numerical derivative stability for the headline 354% figure, but the qualitative conclusion is visible in raw DFT data.","rationale":"The reader correctly identifies the 1D reduction as the paper's most visible limitation, and the paper itself acknowledges it. However, for the specific central claim about d333 in hexagonal wurtzites, the 1D reduction is not an approximation—it is the exact symmetry-allowed reduction. The Schur complement correctly accounts for transverse strain relaxation, and the polar mode is purely longitudinal by 6mm symmetry. The reader's concern would become load-bearing if the paper claimed to capture the full piezoelectric tensor or if off-axis polarization modes coupled significantly to the longitudinal response, but neither is the case here. The more pertinent risk is numerical: the headline 354% figure depends on local numerical second derivatives of a DFT free-energy surface, and Q(0) is sensitive to fitting range (Table S2). The paper would be strengthened by reporting error bars or sensitivity analysis on these derivatives. Nevertheless, the qualitative conclusion—that constant-Q severely overestimates the piezoelectric response in wurtzite AlN—is supported by the directly computed strain branch (Fig. 1a, 121% deviation visible without differentiation), reasonable experimental validation (d333 = 8 pC/N DFT vs ~6 pC/N experiment for AlN; 165 vs 137-156 pC/N for PbTiO3), and a physically transparent microscopic explanation (layered-to-tetrahedral bonding evolution). The thermodynamic framework (Eqs. 1-5, SM §I) is standard and correct. The PbTiO3 comparison provides a useful control showing the effect is wurtzite-specific. The circularity burden is low: results are derived from DFT free-energy surfaces with no fitted parameters. The verdict of ACCEPT is appropriate; the paper makes a substantive contribution with clearly stated limitations that do not undermine the core finding.","tokens_in":19342,"tokens_out":5828,"duration_ms":468757,"concrete_test":"Recompute b333(Ps) and Q(Ps) for AlN using (i) a finer grid (0.02 fractional step instead of 0.04) and (ii) three different local fitting windows (±1, ±2, ±3 grid points). If b333(Ps) shifts by more than 20% across these choices, the precise 354% figure needs revision with error bars. Additionally, independently verify b333(Ps) via direct constrained-DFT: apply a small polarization perturbation δP3 near Ps under σ=0 and measure δε33/δP3 directly, bypassing free-energy-surface differentiation entirely. If this direct measurement agrees with the surface-derivative value within 15%, the numerical differentiation is reliable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's concern about the 1D reduction is a real limitation in general but is not load-bearing for the specific central claim about d333. For a hexagonal 6mm system, the polar distortion is purely along P3 by symmetry, shear strains vanish, and the Schur-complement elimination of transverse strains under σ⊥=0 is an exact reduction—not an approximation—for the longitudinal response. The paper's own caveat (SM §I.D) about needing 'separate low-symmetry constrained-polarization perturbations' refers to computing other tensor components (d311, etc.), not to the validity of d333 itself. The more direct concern is numerical: the headline 354% overestimate of d333(Ps) depends on b333(Ps) = −f_Pε/f_εε, a ratio of second derivatives of the DFT free-energy surface evaluated via local least-squares fits (SM §II.B) on a grid with 0.04 fractional step size. No error bars or sensitivity analysis to fitting window or grid spacing are reported. The value of Q(0) used for the constant-Q baseline is itself sensitive to fitting range: Table S2 shows Q(0) varies from 0.137 (global quadratic fit) to 0.237 (local near P=0) to 0.269 (with higher-order terms) for AlN, which directly affects the 354% figure. However, the qualitative conclusion does not hinge on these numerical subtleties: the stress-free strain branch ε₀₃₃(P3) in Fig. 1(a)—a zeroth-order DFT quantity requiring no differentiation—already shows a 121% deviation from the Q(0)P² extrapolation at Ps. This visible, large deviation in the raw data makes the core finding (constant-Q breaks down in wurtzites) robust even if the precise 354% number shifts with derivative methodology.","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"This manuscript investigates the polarization dependence of electrostriction (Q) and the elastic modulus in strongly polar wurtzites, using AlN as the central example and PbTiO3 as a perovskite reference. The authors construct a two-dimensional DFT free-energy surface f_eff(P3, ε3) and compute local derivatives along the stress-free branch to extract the longitudinal piezoelectric coefficient d333, the strain-polarization factor b333, and the clamped-polarization elastic modulus C^P_3333. The central finding is that the constant-Q approximation (b ≈ 2Q(0)P) breaks down severely in AlN because the structural evolution from a layered nonpolar reference to a tetrahedrally coordinated wurtzite network strongly modifies the electromechanical coupling. This leads to a 354% overestimation of d333(Ps) by the constant-Q model, whereas the DFT-computed value (8 pC/N) agrees well with experiment (~6 pC/N). In contrast, PbTiO3 remains closer to the quadratic approximation, with only a 31% error. The thermodynamic reduction via Schur complement is mathematically clean, and the microscopic structural rationale (Al-N bond network evolution) is physically compelling.","tokens_in":19647,"tokens_out":1116,"duration_ms":182702,"significance":"The paper provides a clear, falsifiable first-principles demonstration that the widely used constant-Q approximation fails for strongly polar wurtzites. The DFT methodology is standard and appropriate (VASP, PBEsol, Berry-phase polarization). The core derivation is parameter-free: d333 is computed directly from derivatives of the DFT free-energy surface, not by fitting Q to target piezoelectric data. The qualitative conclusion is robustly supported by the raw, zeroth-order stress-free strain branch (Fig. 1a), which shows a 121% deviation from the Q(0)P^2 extrapolation without requiring any numerical differentiation. The finding that decreasing polarization can increase the electromechanical strain response is a non-trivial, anomalous prediction with direct design implications for wurtzite ferroelectrics.","major_comments":[{"comment":"SM §II.B and Table S2: The headline 354% overestimation of d333(Ps) depends on the choice of Q(0) for the constant-Q baseline. Table S2 shows that Q(0) for AlN varies significantly depending on the fitting procedure: 0.137 (global quadratic), 0.237 (local near P=0), and 0.269 (with higher-order terms). The manuscript does not clearly specify which Q(0) value is used for the 354% figure in Fig. 1(c). Since the magnitude of the overestimation is a central quantitative claim, the authors should explicitly state which Q(0) is used and briefly justify that choice. This is a presentation gap that directly affects the interpretation of the headline number.","section":null},{"comment":"SM §II.B: The local derivatives defining b333 and d333 are evaluated via local least-squares fits on a grid with a 0.04 fractional step size. No error bars or sensitivity analysis to the fitting window or grid spacing are reported. While the qualitative conclusion does not hinge on these numerical subtleties (as it is already visible in the raw strain data of Fig. 1a), the precision of the specific quantitative claims (e.g., d333(Ps) = 8 pC/N, 354% overestimate, 49% underestimate of C^P_3333) would be strengthened by a brief sensitivity check or an estimate of the numerical uncertainty associated with the local polynomial fits.","section":null}],"minor_comments":[{"comment":"Figures 1-4: The axis labels and legends contain formatting artifacts (e.g., '/s949', '/s916', '/s102') that appear to be rendering errors. These should be corrected for readability.","section":null},{"comment":"Fig. S4: 'Freqeuncy' is misspelled in the axis labels; should be 'Frequency'.","section":null},{"comment":"The abstract states the piezoelectric coefficient is a 'third-rank tensor' and electrostriction is a 'fourth-rank tensor'; while standard, these definitions are somewhat pedantic for the abstract and could be streamlined to focus more quickly on the central result.","section":null},{"comment":"Introduction: The analogy to Hartree-Fock self-consistent exchange fields (Refs. 16-18) is invoked but not elaborated upon. If this analogy is not used later in the paper, consider removing it to avoid confusion.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The reader's concern about the 1D reduction being a potential limitation is not load-bearing for the central claim about d333. For a hexagonal 6mm system, the polar distortion is purely along P3 by symmetry, and the Schur-complement elimination of transverse strains under σ⊥=0 is an exact reduction for the longitudinal response. The paper's own caveat in SM §I.D correctly refers to computing other tensor components, not to the validity of d333 itself. The more substantive concern is the numerical stability of the derivatives and the choice of Q(0) baseline, which I have raised as major comments requiring clarification, but these do not undermine the core qualitative finding. The paper is well-suited for the journal's scope."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. Both major comments identify legitimate presentation gaps that we will address in revision. The first concerns specifying which Q(0) value underlies the headline 354% overestimate; the second concerns adding sensitivity analysis for the local polynomial fits. Both are straightforward to remedy and do not affect the scientific conclusions.","responses":[{"response":"The referee is correct that this is a presentation gap. The 354% figure in Fig. 1(c) uses Q(0) = 0.237 m^4/C^2, the local curvature near P3 = 0 obtained from the stress-free strain branch. This is the natural choice because Q(0) is defined in the manuscript as Q_3333(0), the small-P3 curvature, and the constant-Q extrapolation is meant to represent what one obtains by measuring electrostriction near the reference state and extrapolating linearly. The global quadratic fit (0.137 m^4/C^2) is not a local Q(0) but a single-coefficient fit forced over the entire polarization range, which is a different approximation. We will revise the manuscript and SM to explicitly state which Q(0) is used in Fig. 1(c), add a sentence justifying the choice of the local curvature, and cross-reference Table S2 so the reader can see the range of fitted values and understand that the variation itself reflects the non-quadratic character of the strain-polarization relation.","revision_made":"yes","referee_comment":"SM §II.B and Table S2: The headline 354% overestimation of d333(Ps) depends on the choice of Q(0) for the constant-Q baseline. Table S2 shows that Q(0) for AlN varies significantly depending on the fitting procedure: 0.137 (global quadratic), 0.237 (local near P=0), and 0.269 (with higher-order terms). The manuscript does not clearly specify which Q(0) value is used for the 354% figure in Fig. 1(c). Since the magnitude of the overestimation is a central quantitative claim, the authors should explicitly state which Q(0) is used and briefly justify that choice."},{"response":"We agree that a sensitivity analysis would strengthen the quantitative claims and will add one. Specifically, we will vary the fitting window (number of neighboring grid points used in the local least-squares fits) and report the resulting spread in d333(Ps), b333(Ps), and C^P_3333(Ps). We expect the variations to be modest relative to the deviations from the constant-Q baseline, consistent with the referee's own observation that the qualitative conclusion is already visible in the raw strain data of Fig. 1(a). We will add a brief paragraph in SM §II.B summarizing the sensitivity check and quote representative uncertainty ranges for the headline numbers in the main text.","revision_made":"yes","referee_comment":"SM §II.B: The local derivatives defining b333 and d333 are evaluated via local least-squares fits on a grid with a 0.04 fractional step size. No error bars or sensitivity analysis to the fitting window or grid spacing are reported. While the qualitative conclusion does not hinge on these numerical subtleties (as it is already visible in the raw strain data of Fig. 1a), the precision of the specific quantitative claims (e.g., d333(Ps) = 8 pC/N, 354% overestimate, 49% underestimate of C^P_3333) would be strengthened by a brief sensitivity check or an estimate of the numerical uncertainty associated with the local polynomial fits."}],"tokens_in":19337,"tokens_out":772,"duration_ms":61091,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"The main result is clear and important: in strongly polar wurtzites like AlN, the electrostriction coefficient Q is strongly polarization-dependent, and the standard constant-Q approximation (d = 2QPε₀χ) overestimates the longitudinal piezoelectric response by a factor of 3.5. The PbTiO₃ comparison is a good control — same methodology, much smaller deviation (31%), consistent with perovskites staying closer to quadratic strain-polarization behavior. The computed d₃₃₃(Ps) = 8 pC/N for AlN matching experiment (~6 pC/N) and 165 pC/N for PbTiO₃ vs. 137–156 pC/N experimental gives confidence the DFT is sound. The microscopic structural explanation — layered-to-tetrahedral bond network evolution driving stiffening — is physically reasonable and supported by the bond-length data in Fig. 2(d). The thermodynamic reduction via Schur complement is mathematically clean, and for a hexagonal 6mm system the 1D longitudinal reduction is exact by symmetry, not an approximation. The stress-test concern about the 1D reduction being a limitation is overblown for d₃₃₃ specifically; the paper's own caveat in SM §I.D refers to other tensor components, not the longitudinal response. Where the paper is soft is numerical. The headline 354% figure depends on second derivatives of the DFT free-energy surface evaluated via local least-squares fits on a 0.04 fractional grid, with no error bars or sensitivity analysis to fitting window or grid spacing. Table S2 shows Q(0) for AlN ranges from 0.137 to 0.269 m⁴/C² depending on fitting protocol, which directly affects the constant-Q baseline and thus the 354% number. However — and this is the key point — the qualitative conclusion does not depend on these subtleties. The raw stress-free strain branch ε₀₃₃(P₃) in Fig. 1(a), a zeroth-order DFT quantity requiring no differentiation, already shows a 121% deviation from the Q(0)P² extrapolation at Ps. That visible deviation makes the core finding robust even if the precise 354% shifts with derivative methodology. No code or data release, which limits reproducibility of the numerical derivatives specifically. This paper is for anyone modeling wurtzite ferroelectrics or designing AlN-based electromechanical devices. The constant-Q assumption is embedded in a lot of phase-field and thermodynamic modeling, and this work shows it fails badly for an important material class. The result deserves a serious referee who can check the numerical derivative methodology and push for error bars on the headline figure.","headline":"Constant-Q electrostriction breaks down badly in wurtzites; the 354% overestimate headline is qualitatively robust but numerically soft.","tokens_in":20450,"tokens_out":645,"would_cite":true,"duration_ms":79661,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"Electrostriction in wurtzites is not constant","keywords":["piezoelectricity","electrostriction","wurtzite","AlN","ferroelectrics","first-principles","elastic modulus","polarization dependence"],"falsifier":"Measure the longitudinal piezoelectric coefficient d₃₃₃ of AlN or related wurtzites as a function of polarization (e.g., through alloying to tune P_s or under applied bias) and check whether it deviates from the linear-in-P scaling predicted by constant-Q. A measurement showing d₃₃₃ ∝ P would falsify the central claim.","tokens_in":19465,"feed_emoji":"⚡","tokens_out":890,"duration_ms":54484,"temperature":0.7,"pith_summary":"The paper argues that in strongly polar wurtzite materials like AlN, the electrostriction coefficient Q and the elastic modulus are strongly dependent on polarization, breaking the conventional assumption that Q is constant. Using first-principles free-energy surfaces, the authors show that the standard relation d = 2QPε₀χ overestimates the longitudinal piezoelectric coefficient d₃₃₃ in AlN by more than a factor of three (354%). The physical origin is structural: as polarization increases, the crystal evolves from a layered nonpolar reference toward a tetrahedrally coordinated wurtzite network, which stiffens the lattice and alters the strain-polarization coupling. This makes the piezoelectric response highly nonlinear in polarization and can even produce the anomalous effect that reducing polarization increases the electromechanical strain response. The perovskite PbTiO₃, by contrast, remains much closer to the conventional quadratic picture, with only a 31% error from the constant-Q approximation.","feed_headline":"Constant-Q model overestimates AlN piezoelectricity by 354%","feed_subtitle":"Electrostriction and elastic modulus in wurtzites vary strongly with polarization, making piezoelectric response highly nonlinear and upend","key_machinery":"A reduced one-dimensional free-energy surface f_eff(P₃, ε₃) constructed from DFT, from which local derivatives yield the polarization-dependent electrostriction Q₃₃₃₄(P₃), the strain-polarization factor b₃₃₃(P₃), the clamped-polarization modulus C̃ᴾ₃₃₃₃(P₃), and the piezoelectric coefficient d₃₃₃(P₃) = b₃₃₃ · ε₀χ^σ₃₃. Transverse strains are eliminated under zero transverse stress via Schur-complement reduction.","core_discovery":"The central finding is that the electrostriction coefficient Q₃₃₃₃ and the clamped-polarization elastic modulus C̃ᴾ₃₃₃₃ in wurtzite AlN are strongly polarization-dependent functions rather than constants. The constant-Q approximation, which treats the strain-polarization relation as purely quadratic (ε⁰ ≈ QP²), overestimates d₃₃₃ at the spontaneous polarization P_s by 354%. This breakdown arises because increasing polarization drives a structural transition from a layered hexagonal reference (long axial Al-N separation) to a tetrahedral wurtzite network (short, stiff bonds), causing the elastic modulus to nearly double and the mixed strain-polarization coupling to vary nonlinearly. The same非","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Constant-Q approximation overestimates AlN piezoelectricity by 354%","Wurtzite electrostriction and elastic modulus vary with polarization","Polarization-dependent electrostriction drives anomalous wurtzite piezoelectricity","Decreasing polarization can increase wurtzite electromechanical strain","Nonlinear electrostriction challenges constant-Q models in wurtzites"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The entire analysis is reduced to a one-dimensional polar-distortion path with P₁ = P₂ = 0, assuming that the longitudinal electromechanical response dominates and that transverse coupling is fully captured by eliminating in-plane strains under zero transverse stress. The paper itself notes that a full tensorial extension would require separate low-symmetry constrained-polarization perturbations, so if off-axis polarization responses contribute significantly, the 1D reduction","fun_headline_variants_meta":{"raw":{"variants":["Constant-Q approximation overestimates AlN piezoelectricity by 354%","Wurtzite electrostriction and elastic modulus vary with polarization","Polarization-dependent electrostriction drives anomalous wurtzite piezoelectricity","Decreasing polarization can increase wurtzite electromechanical strain","Nonlinear electrostriction challenges constant-Q models in wurtzites","Structural transition alters piezoelectric response in wurtzite AlN"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":1340,"prompt_tokens":613,"completion_tokens":727,"prompt_tokens_details":null},"tokens_in":613,"tokens_out":727,"duration_ms":34838,"temperature":1.0,"reasoning_tokens":628,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-08T22:19:51.732411+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"Measure the longitudinal piezoelectric coefficient d₃₃₃ of AlN or related wurtzites as a function of polarization (e.g., through alloying to tune P_s or under applied bias) and check whether it deviates from the linear-in-P scaling predicted by constant-Q. A measurement showing d₃₃₃ ∝ P would falsify the central claim.","supporting_citations":[],"review_version":1}