{"id":"e6ab65c1-8092-47ab-930b-5a3a389779f1","arxiv_id":"1908.07982","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"For non-centrosymmetric crystals with long-range Coulomb interactions, this paper derives Ewald-sum corrections to nonaffine response theory and shows that in alpha-quartz nonaffine relaxation lowers elastic constants by factors of 3 to 4, matching experiment.","lead":"This paper adds long-range electric forces into a lattice-dynamics theory that accounts for how atoms relax under strain, and applies it to quartz. The calculation matches measured elastic constants and shows that quartz is three to four times softer than the standard affine estimate.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Ewald nonaffine-force formulas in Eqs. 10 and 14 appear to omit the strain derivatives of the volume prefactor and reciprocal vectors, so the analytic nonaffine correction is not established.","rationale":"The paper's goal is to add Ewald long-range interactions to nonaffine response theory and to show that alpha-quartz has a large nonaffine correction. The VDOS comparison gives independent support for the underlying BKS model, and the numerical strain validation is a sensible check. However, the central new mathematics is the strain derivative of the Ewald force and energy. Direct differentiation of Eq. 8 produces explicit terms from dV/depsilon and dG/depsilon that do not appear in Eqs. 10 or 14; these terms include -F_i^Ewald and are not obviously zero. The reader's weakest assumption identifies exactly this gap, and I agree. I keep the reader's conditional verdict rather than escalating to reject because the proposed finite-strain check is simple and would settle the point; if the check confirms the missing terms are nonzero, the analytic formulas and the elastic constants in Table III must be recomputed before the central claim can be accepted.","tokens_in":9956,"tokens_out":16911,"duration_ms":179476,"concrete_test":"Finite-strain check on the 1350-atom relaxed alpha-quartz cell with the same BKS and Ewald parameters as Section II: compute only the reciprocal-space Ewald force F_i from Eq. 8 at epsilon_xx = +1e-5 and -1e-5, and compare the central-difference derivative [F_i(+epsilon)-F_i(-epsilon)]/2epsilon with the right-hand side of Eq. 10 for every atom. If the mean absolute difference is at rounding level, the omitted terms cancel and Eq. 10 is vindicated; if the difference is comparable to |F_i^Ewald|, the analytic nonaffine force is incomplete and Table III cannot be attributed to the printed derivation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Direct differentiation of Eq. 8 for a uniaxial x strain yields more than Eq. 10. Under epsilon_xx, V -> V(1+epsilon), G_x -> G_x/(1+epsilon), and G.R_ij is invariant. With I(u)=exp(-sigma^2 u/2)/u, dI/depsilon = I(sigma^2+2/G^2)G_x^2, d(1/V)/depsilon = -1/V, and dG/depsilon = (-G_x,0,0). Hence dF_i/depsilon contains the retained term I(sigma^2+2/G^2)G_x^2 G, but also the omitted terms -I G and -I G_x e_x (up to the sign convention of Eq. 10). The -I G term is exactly -F_i^Ewald, which is not zero at the relaxed positions; the other term has the same order as the retained one. No cancellation argument is given. The same pattern affects the shear derivative Eq. 14 and the second derivatives behind Eqs. 12-16. Since the nonaffine correction in Table III is 70-90% of the affine values, the printed analytic C_m vectors are not reliable, and the statement 'We checked by direct numerical calculations' is not backed by reproducible data or code.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper extends the nonaffine response theory of Lemaitre and Maloney to crystals with long-range Coulomb interactions treated by Ewald summation. The authors derive analytic expressions for the Ewald contribution to the affine elastic constants and to the nonaffine force field, apply them to alpha-quartz modeled with the BKS potential, and report static elastic constants that agree reasonably with experiment. The central physical result is that the nonaffine correction reduces the affine Born-Huang estimates by 70-90% for the elastic constants of alpha-quartz, which the authors attribute to the lack of centrosymmetry and connect to the boson peak.","tokens_in":10208,"tokens_out":13591,"duration_ms":124388,"significance":"If the Ewald extension is correct, the paper offers a useful formalism for computing elastic constants of non-centrosymmetric crystals from a single relaxed configuration, without explicit deformation simulations. The reported magnitude of nonaffine softening in a perfectly ordered crystal is a striking result with potential implications for piezoelectric materials and for the physics of the boson peak. The paper is also commendable for comparing directly with experimental data and for separating affine and nonaffine contributions. However, the analytic derivation is the load-bearing element of the method, and the manuscript as written does not make that derivation verifiable: the printed strain derivatives appear incomplete, and the claimed numerical checks are not documented.","major_comments":[{"comment":"The derivative of the Ewald force in Eq. (8) under strain is not complete. For the uniaxial case, using V -> V(1+epsilon) and reciprocal vectors G -> (G_x/(1+epsilon), G_y, G_z), with G.R_ij invariant, direct differentiation gives, up to the sign convention for Xi, an extra term -I(G^2)G - I(G^2)G_x e_x inside the sum over G, in addition to the retained I(G^2)(sigma^2+2/G^2)G_x^2 G term. Equation (10) contains only the last term. The omitted '-G' term is exactly -F_i^Ewald, which is not zero at the relaxed positions because only the total force (short-range plus Ewald) vanishes in alpha-quartz; the '-G_x e_x' term has the same order as the retained term. No cancellation argument is provided. The same omission affects the shear derivative in Eq. (14), where the derivative of the reciprocal vectors contributes a term of order I(G^2)G_x that is not shown, and it propagates into the second derivatives behind Eqs. (12), (13), and (16). Since Table III reports nonaffine corrections of 70-90% of the affine values, the printed analytic Xi vectors and the central methodological claim are not established by the manuscript as written.","section":"Section III B 2-3, Eqs. (8), (10), (14)"},{"comment":"The sentence 'We checked by direct numerical calculations that the analytical expressions described in previous Section predict faithfully the elastic constants' is the only numerical validation of the Ewald formulas, but no comparison is shown and no code, input files, or raw data are provided. Because Eqs. (10)-(16) are the paper's main new contribution, a quantitative comparison between the analytic and finite-strain values of Xi, C_Born, and the final elastic constants is necessary, or at minimum a reproducible data repository. As it stands, the agreement in Table III cannot be traced to the correctness of the Ewald formalism, especially given the incomplete derivative terms identified above.","section":"Section IV, second paragraph"}],"minor_comments":[{"comment":"The experimental row labeled Exp. [26] is missing the C66 entry; the row reads '86.6 106.4 58 6.74 12.4', which appears to omit or merge a value. Please reformat so all six elastic constants are listed.","section":"Table III"},{"comment":"The structure factor is written as 'S(G2)' in several places; this should be S(G).","section":"Eqs. (18)-(20)"},{"comment":"The numerical value of the Ewald Gaussian width sigma (or alpha) is never stated, although the real-space cutoff Rcut = 10 Angstrom is given. The Ewald calculation cannot be reproduced without this parameter.","section":"Section II B"},{"comment":"The sign convention for Xi is inconsistent between Eq. (6), which defines Xi as dF/depsilon, and Eq. (10), which appears to use -dF/depsilon for the Ewald part. Please state the convention explicitly.","section":"Eqs. (6) and (10)"},{"comment":"The paper states that the original BKS parameters were chosen because they provide the best agreement with experimental elastic constants of alpha-quartz [23]. If any of the experimental data in Table III were used in that selection, the comparison is not fully a posteriori and this should be stated clearly.","section":"Sections II B and IV"}],"recommendation":"major_revision","confidential_remarks":"The central issue is that the Ewald strain derivatives printed in Eqs. (10), (14), and the related second derivatives do not appear to be the complete derivatives of Eq. (8). I verified by direct differentiation that the omitted terms are nonzero and of the same order as the retained terms for a non-centrosymmetric crystal. The authors claim numerical checks but give no reproducible evidence. If the derivation can be corrected and the final numbers shown to be unchanged (or updated), the paper could be viable; as it stands, the methodological claim is not established. There is also a mild circularity in selecting the BKS parameter set and the 10 Angstrom cutoff for best agreement with the same elastic constants that are later reported as agreement with experiment; this should be clarified in revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Genuinely new piece: the Ewald extension of nonaffine lattice dynamics, and the alpha-quartz result that nonaffine relaxation softens elastic constants by factors of 3-4. The VDOS comparison and the numerical stress-strain checks are good instincts. But the central analytic derivation has a problem. I differentiated Eq. 8 by hand. Under uniaxial x strain, V -> V(1+eps) and G -> (G_x/(1+eps), G_y, G_z), with G·R invariant. The derivative of F_i picks up, besides the retained I(sigma^2+2/G^2)G_x^2 G term, two extra pieces: -I G and -I G_x e_x, multiplied by the same sin sum. The -I G piece is exactly -F_i^Ewald, which does not vanish at relaxed positions because only the total force vanishes. The -I G_x e_x piece is the same order as the retained term. No cancellation argument appears in Section III B 2. The shear case Eq. 14 has the same problem: a missing -I G_x e_y term (volume is preserved in shear, so no volume piece there). Since the nonaffine correction is assembled from these Xi vectors, Eqs. 10/14 control the 70-90% softening claim. That claim is therefore not established by the printed analytic route.\n\nCredit where due: the energy-derived affine formulas (Eqs. 11-13, 16, 18-20) appear to include the volume and reciprocal-vector derivatives; I spot-checked Eq. 16 and it works. And the numerical strain calculation, if backed by scripts or data, could still be a valid check — but no code or data are supplied, so \"we checked by direct numerical calculations\" is not independently verifiable. The potential choice (BKS plus 10 Angstrom cutoff) was selected for agreement with the same experimental elastic constants used for comparison; that is a moderate circularity, not fatal.\n\nBottom line: the paper asks the right question and the numerical quartz result may survive a corrected derivation, but the main methodological claim, as printed, is not correct. I would send it to review — a good referee could catch exactly this — but insist on correction and reproducible validation before acceptance. People working on nonaffine elasticity in crystals and piezoelectric oxides should read it for the quartz numbers and the noncentrosymmetry discussion, not for the Ewald force formulas in their current form.","headline":"Genuinely new Ewald extension of nonaffine lattice dynamics and a striking quartz result, but the printed strain-derivative formulas for the Ewald force miss terms from V and G, so the analytic nonaffine correction is not established.","tokens_in":10756,"tokens_out":8878,"would_cite":false,"duration_ms":90678,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["62.20.Dc","63.20.-e"],"model":"deepseek-v4-flash","headline":"Ewald-based nonaffine lattice dynamics predicts quartz elastic constants from static configurations; nonaffine relaxation cuts affine values by 70–90%, leaving true constants 3–4 times smaller (up to 15 times for C12/C13).","keywords":["nonaffine lattice dynamics","Ewald summation","elastic constants","α-quartz","BKS potential","non-centrosymmetric crystals","Born-Huang approximation","boson peak"],"falsifier":"Compute the derivative of the Ewald force in Eq. (8) with respect to a tensile or shear strain by keeping the full dependence of $V$ and $G$ on strain, and compare term by term with Eqs. (10), (14) and (16) at the α-quartz lattice constants used here. If the extra terms are nonzero, the analytic nonaffine correction changes; an independent check would be to recompute the six elastic constants by finite strain at fixed cell shape with full Ewald sums and compare with Table III.","tokens_in":9720,"feed_emoji":"💎","tokens_out":12174,"duration_ms":197755,"temperature":0.7,"pith_summary":"Nonaffine lattice dynamics, which writes elastic constants as an affine (Born) part minus a softening correction from atomic relaxations, is extended to crystals with long-range Coulomb forces. The extension is analytic: the Ewald reciprocal-space sum is differentiated with respect to strain and fed into the nonaffine sum rule, so no deformation simulation is needed. For α-quartz with the BKS potential, the resulting constants agree closely with experiment for all six independent elastic constants, while the affine-only estimate is 3–4 times too large for C11, C33, C44 and C66, and up to about 15 times too large for C12 and C13. The paper concludes that non-centrosymmetric crystals generically require the nonaffine correction and that the large softening is tied to the absence of inversion symmetry, the same structural feature it links to the boson peak.","feed_headline":"Atomic relaxation cuts quartz stiffness by 70-90 percent","feed_subtitle":"With Ewald Coulomb forces, static-lattice constants match experiment; affine-only estimates overshoot 15-fold.","key_machinery":"The carrying object is the nonaffine response sum rule, $C_{\\alpha\\beta\\kappa\\chi}=C^{\\rm Born}_{\\alpha\\beta\\kappa\\chi}-\\frac{1}{V}\\sum_{m=1}^{3N-3} C_{m,\\alpha\\beta}C_{m,\\kappa\\chi}/\\omega_m^2$, where the mode tensor $C_{m,\\kappa\\chi}$ is built from the nonaffine force field $\\Xi^{\\alpha}_{j,\\kappa\\chi}=\\partial F^{\\alpha}_i/\\partial \\epsilon_{\\kappa\\chi}$ projected onto eigenvectors of the dynamical matrix. The new machinery is the analytic strain differentiation of the Ewald long-range term $E_{LR}$ of Eq. (2), the Fourier-space part of the Coulomb energy. For an orthogonal cell, affine deformation changes the volume $V$, the reciprocal vectors $G$, and the interatomic separations $R_{ij}$, but leaves the dot products $G\\cdot R_{ij}$ and hence the structure factor $S(G)$ invariant; that invariance is what keeps derivatives tractable. Working from the Ewald force $F_i = \\frac{q_i}{V\\epsilon_0}\\sum_{G\\ne 0} I(G^2) G \\sum_j q_j \\sin(G\\cdot R_{ij})$, the paper derives closed-form expressions for the nonaffine force fields under tensile strain (Eq. 10) and shear strain (Eq. 14), and for the corresponding affine elastic constants (Eqs. 12, 13, 16). These expressions, together with the short-range pair part, are the inputs to the sum rule.","core_discovery":"The central claim, stated the way the authors would state it, is that the Ewald long-range Coulomb contribution can be treated analytically in nonaffine response theory, and that when this is done for α-quartz the elastic constants come out right only if the nonaffine term is included. Starting from one equilibrated static cell, the authors evaluate $C_{\\alpha\\beta\\kappa\\chi}=C^{\\rm Born}_{\\alpha\\beta\\kappa\\chi}-\\frac{1}{V}\\sum_m C_{m,\\alpha\\beta}C_{m,\\kappa\\chi}/\\omega_m^2$ with the reciprocal-space Ewald part of the dynamical matrix, the affine constants, and the nonaffine force field all expressed in closed form. The resulting constants $C_{11}=90.5$, $C_{33}=107.0$, $C_{44}=50.2$, $C_{66}=41.1$, $C_{12}=8.1$, $C_{13}=15.2$ GPa sit close to experimental values, whereas the affine-only values $375.6$, $329.6$, $189.2$, $125.4$, $125.2$, $189.1$ GPa are much too large. The nonaffine correction therefore removes about 70% of the affine stiffness for $C_{11}$, $C_{33}$, $C_{44}$, $C_{66}$, and about 90% for $C_{12}$ and $C_{13}$, and the paper reads this as evidence that non-centrosymmetry, not disorder, suffices to make nonaffine elasticity strong.","pith_inferences":["Not pursued in the paper: because the derivation of the reciprocal-space terms uses only the form $I(G^2)$ and the invariance of $G\\cdot R_{ij}$, the same analytic structure should extend to screened-Coulomb or dipole-dipole kernels; testing this on another ionic crystal would show whether the strong nonaffine softening is generic.","Not pursued in the paper: affine-only estimates for other non-centrosymmetric piezoelectric crystals with partial charges are likely inflated by a comparable factor, so existing Born-Huang calculations for wurtzite or perovskite structures may deserve revisiting with the Ewald nonaffine terms.","Not pursued in the paper: a finite-frequency extension of this formalism should show frequency-dependent elastic moduli that soften most near the boson-peak frequency; that prediction could be tested by comparing dynamic moduli from molecular dynamics with the static values reported here."],"forward_implications":["For α-quartz, the six independent elastic constants can be obtained from a single equilibrated static configuration, without simulating a deformation, and they match experimental values closely.","The affine-only Born-Huang estimate overestimates $C_{11}$, $C_{33}$, $C_{44}$, $C_{66}$ by a factor of 3–4 and $C_{12}$, $C_{13}$ by up to about 15, so omitting nonaffine relaxation is not an option for non-centrosymmetric crystals.","The nonaffine softening of about 70% (90% for $C_{12}$, $C_{13}$) arises because no atom sits at an inversion center; oxygen atoms, in the most asymmetric environments, undergo the largest nonaffine displacements.","The same model reproduces the experimental vibrational density of states and boson peak of α-quartz, supporting the proposed link between non-centrosymmetry, strong nonaffine elasticity, and the boson peak."],"supporting_citations":[{"why":"Derives the affine-minus-nonaffine sum rule for quasistatic elastic constants that Eq. (3) implements.","marker":"[2]"},{"why":"Defines the BKS potential, with Buckingham short-range terms and partial charges, used in all simulations.","marker":"[8]"},{"why":"Introduces Ewald's method for summing long-range Coulomb interactions, the starting point of the analytic derivation.","marker":"[12]"},{"why":"Provides the Ewald force and dynamical-matrix expressions that the paper differentiates with respect to strain.","marker":"[13]"},{"why":"Presents the Ewald summation formulation used for the long-range energy and structure factor.","marker":"[24]"},{"why":"Supplies the BKS parametrization and short-range cutoff that give the best agreement with experimental elastic constants.","marker":"[23]"},{"why":"Supplies experimental VDOS and boson-peak data that validate the vibrational spectrum of the model.","marker":"[15]"},{"why":"Provides experimental elastic constants of natural quartz used as reference for the computed values.","marker":"[25]"}],"fun_headline_variants":["Ewald method shows quartz stiffness falls by 70-90%","Nonaffine correction reduces quartz stiffness by up to 90%","Quartz affine-only elasticity overshoots experiment 15-fold","Ewald nonaffine theory matches quartz elastic constants","Non-centrosymmetry alone drives large nonaffine softening"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The formulas for how the long-range Coulomb force changes under strain assume that all contributions from the changing box volume and reciprocal-lattice vectors have been included; if one such contribution was missed, the predicted softening would be wrong.","fun_headline_variants_meta":{"raw":{"variants":["Ewald method shows quartz stiffness falls by 70-90%","Nonaffine correction reduces quartz stiffness by up to 90%","Quartz affine-only elasticity overshoots experiment 15-fold","Ewald nonaffine theory matches quartz elastic constants","Non-centrosymmetry alone drives large nonaffine softening"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001181,"raw_usage":{"total_tokens":4914,"prompt_tokens":1014,"completion_tokens":3900,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":630,"completion_tokens_details":{"reasoning_tokens":3814}},"tokens_in":630,"tokens_out":3900,"duration_ms":616497,"temperature":1.0,"reasoning_tokens":3814,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:54:53.878302+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the derivative of the Ewald force in Eq. (8) with respect to a tensile or shear strain by keeping the full dependence of $V$ and $G$ on strain, and compare term by term with Eqs. (10), (14) and (16) at the α-quartz lattice constants used here. If the extra terms are nonzero, the analytic nonaffine correction changes; an independent check would be to recompute the six elastic constants by finite strain at fixed cell shape with full Ewald sums and compare with Table III.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Derives the affine-minus-nonaffine sum rule for quasistatic elastic constants that Eq. (3) implements."},{"cited_title":"Lemaitre and C","cited_arxiv_id":null,"evidence_quote":"Defines the BKS potential, with Buckingham short-range terms and partial charges, used in all simulations."},{"cited_title":"Palyulin, C","cited_arxiv_id":null,"evidence_quote":"Introduces Ewald's method for summing long-range Coulomb interactions, the starting point of the analytic derivation."},{"cited_title":"Kholkin, N","cited_arxiv_id":null,"evidence_quote":"Provides the Ewald force and dynamical-matrix expressions that the paper differentiates with respect to strain."},{"cited_title":"Material properties of $\\alpha$-quartz that are relevant for its potential use in X-ray monochromators and analyzers","cited_arxiv_id":"1612.07049","evidence_quote":"Presents the Ewald summation formulation used for the long-range energy and structure factor."},{"cited_title":"Senguly and S","cited_arxiv_id":null,"evidence_quote":"Supplies the BKS parametrization and short-range cutoff that give the best agreement with experimental elastic constants."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies experimental VDOS and boson-peak data that validate the vibrational spectrum of the model."},{"cited_title":"Bragg and R","cited_arxiv_id":null,"evidence_quote":"Provides experimental elastic constants of natural quartz used as reference for the computed values."}],"review_version":1}