{"id":"bf84832d-25f5-4b21-953c-b3844815a87a","arxiv_id":"2607.28838","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"At electronic temperatures above about 2 eV, α-quartz loses elastic and dynamical stability, its polar optical electron-phonon coupling drops by orders of magnitude, and Si and O atoms initially equilibrate at different temperatures.","lead":"This paper trains deep neural network potentials on quantum simulations of α-quartz to study what happens to the crystal in the first moments after its electrons are heated by an intense laser or ion beam. It finds the lattice softens and expands, polar electron-phonon coupling collapses, and silicon and oxygen atoms briefly reach different temperatures.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fröhlich suppression is computed from harmonic LO phonons and dielectric tensors at Te=2.2–2.6 on a trigonal saddle point with imaginary branches and an extrapolated EOS volume; the quantitative 'Te>2 eV' claim is not yet tied to a well-defined structure.","rationale":"The stress-test pass confirms the reader's conditional verdict. The qualitative narrative—lattice destabilization, volume expansion, bulk modulus collapse, bond weakening, and the strong reduction of the Fröhlich coupling—is well supported by the DFT/DNNP data at Te ≤ 2.1 eV, where the structure is stable and the EOS minimum is sampled. The problematic step is specifically the quantitative use of harmonic Fröhlich parameters at Te = 2.2 and 2.6 eV, where the reference structure is already unstable or fluid-like. At these temperatures an LO phonon frequency and a static dielectric constant are not well-defined within the same harmonic framework used to derive Eq. (3), and the extrapolated EOS volume adds further uncertainty. Because this is the basis for the headline statement 'strongly suppressed at Te > 2 eV,' the claim should remain conditional and be accompanied by the caveat already partly present in the text, or by calculations on the relaxed low-symmetry minimum. The DNNP transferability concern is real but secondary: the qualitative conclusions are corroborated by AIMD, and DNNP validation errors are small, so the Fröhlich ill-posedness is the more load-bearing issue. No fatal inconsistency or misconduct is present; the paper's core contribution is still valuable, but a specific quantitative claim is not yet fully established.","tokens_in":21701,"tokens_out":4211,"duration_ms":54459,"concrete_test":"At Te = 2.2 eV, redo the Fröhlich calculation on the dynamically stable distorted local minimum obtained by the simulated-annealing relaxation described in Sec. III D, using DFPT dielectric tensors and NAC-corrected LO frequencies at a volume sampled near the fitted EOS minimum (e.g., V0 ≈ 231.5 Å^3/cell) and at ±5% around it. If α remains ≤ 0.02 and ε_s/ε_∞ ≤ 1.1, the suppression threshold is robust; if α rises significantly or the LO mode remains imaginary after relaxing the saddle-point constraint, the quantitative Te > 2 eV claim should be weakened or removed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim—polar optical scattering collapses for Te > 2 eV—rests on Eq. (3), whose inputs (m*, ω_LO, ε_s, ε_∞) are extracted at fixed optimized volumes from harmonic/DNNP phonons and DFT dielectric calculations. At Te = 2.2 eV the trigonal structure is already elastically and dynamically unstable (Secs. III A and III D; Table S1), and the paper itself states that this symmetry-constrained structure is a saddle point; at Te = 2.6 eV several phonon branches are imaginary. Yet Table I reports positive LO frequencies and Fröhlich α values for these conditions. The EOS fit for Te = 2.2 eV has its minimum outside the sampled volume range (Fig. 1; SM Table S2), so V0 and B0 are extrapolations. In an unstable or saddle-point structure, the harmonic LO frequency and the macroscopic dielectric response—especially ε_s, which enters α through (1/ε_∞ − 1/ε_s)—are not well-defined. The two-order-of-magnitude reduction in α comes largely from ε_s/ε_∞ → 1 (e.g., 103.16/95.35 at 2.2 eV), a near-cancellation that may be an artifact of evaluating linear-response quantities at the edge of mechanical instability. Consequently, the magnitude and threshold of the claimed nonpolar crossover are not quantitatively established, even though the qualitative softening trend is independently supported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines finite-electronic-temperature DFT, AIMD, and electronic-temperature-dependent deep neural network potentials (DNNPs) to study the response of α-quartz to sudden electronic excitation. It reports that raising the electronic temperature T_e weakens Si–O bonding, expands the lattice, lowers the bulk modulus, and eventually makes the trigonal crystal elastically and dynamically unstable, with an instability threshold between T_e=2.1 and 2.2 eV. It further estimates the Fröhlich coupling constant from band-structure and phonon inputs, concluding that polar optical scattering is strongly suppressed for T_e>2 eV, and uses large-cell DNNP-MD to show that the lattice does not reach a Maxwell–Boltzmann equilibrium in the first few hundred femtoseconds, with Si and O atoms initially equilibrating at different kinetic temperatures.","tokens_in":22125,"tokens_out":8114,"duration_ms":89626,"significance":"If the central claims hold, the paper offers a valuable multiscale framework for nonthermal lattice destabilization and provides one of the more complete pictures of electronically excited SiO2, connecting elastic, dynamical, bonding, and transport-related indicators. The main strengths are the triangulation of destabilization from multiple independent probes (elastic constants, EOS, pCOHP, Bader charges, DFT/DNNP phonons, and AIMD/DNNP MD), the explicit reporting of DNNP validation errors, and the honest acknowledgment of several limitations (NAC omission, extrapolated EOS at T_e=2.2 eV, and the saddle-point nature of the high-T_e structure). The qualitative destabilization story is well supported; however, the quantitative Fröhlich-suppression claim is partly based on harmonic quantities evaluated at structures the paper itself identifies as unstable, and this overreach needs to be corrected before the quantitative conclusions can be accepted.","major_comments":[{"comment":"Table I reports α=0.0070 (2.2 eV) and positive f_LO values for structures that the paper itself identifies as unstable. Sec. III A/Table S1 show that at 2.2 eV elastic conditions (iii) and (iv) fail (λ_min=-10.10 GPa); at 2.6 eV conditions (i) and (iv) fail. Sec. III D states that at 2.2 eV the symmetry-constrained trigonal structure is a saddle point and that at 2.6 eV several branches are imaginary. In such structures the harmonic ω_LO and macroscopic ε_s are not well-defined, and the near-cancellation ε_s/ε_∞=1.0819 that produces the small α may be an artifact. The EOS volume at 2.2 eV is also an extrapolation beyond the sampled V (Fig. 1, Table S2). Because T_e=2.0 and 2.1 eV are stable and already give α=0.0115 and 0.0098, roughly two orders below the ground state, the qualitative suppression conclusion can be retained; but the quantitative 'T_e>2 eV' threshold and the 2.2/2.6 eV en","section":"§III E, Eq. (3), Table I"},{"comment":"The volume-independent HSE correction ΔF_HSE(V_GS0,T_e) is applied to PBE free energies at all volumes. The fitted equilibrium volumes expand from 120.7 Å^3/cell (ground state) to 151.3 Å^3/cell (2.0 eV) and 164.0 Å^3/cell (2.1 eV), and to an extrapolated 231.5 Å^3/cell (2.2 eV). A correction fixed at the ground-state volume is not obviously representative over this range. This directly affects the EOS-derived B_0 (68.1→2.0 GPa), which is a central destabilization metric, and the volumes used for phonon and dielectric inputs. Please quantify the volume dependence of the HSE correction, for example by computing HSE at the expanded volume for T_e=2.0 eV, or soften the quantitative EOS claims.","section":"§II B, Eq. (2)"},{"comment":"The DNNP validation reports force MAEs of 0.014–0.028 eV/Å on validation frames drawn from the same 72-atom NVE trajectories used for training; no independent test set from a separate trajectory or from a different thermodynamic state is described. The 3087-atom DNNP-MD at T_e=2.2 and 2.6 eV explores strongly expanded and disordered/fluid-like states (Secs. III F and III G) that lie far from the original training distribution. The non-equilibration and separate T_Si/T_O conclusions are nonetheless corroborated by direct 72-atom AIMD (Fig. 5a), so this is not fatal, but the paper should report an ensemble or uncertainty estimate for the large-cell runs, or explicitly discuss the extrapolation risk.","section":"§III F, SM Table S3"}],"minor_comments":[{"comment":"Typo: 'over a range of electronic temperature relevant relevant for experiments' contains a duplicated word; later 'election-ion relaxation' should read 'electron-ion relaxation'.","section":"Section I"},{"comment":"The text states that at T_e=2.6 eV the −IpCOHP drops to 0.6 eV, but Fig. 2 only shows pCOHP up to 2.2 eV. Add the 2.6 eV panel or refer explicitly to a supplementary figure.","section":"§III C"},{"comment":"No details are given for how m* is extracted from the band structure, nor are uncertainties provided. State the fitting procedure and estimate errors, especially at elevated T_e where the band structure is strongly renormalized.","section":"Table I"},{"comment":"The sentence 'we find that T_O > T_Si, as expected since m_Si > m_O' is ambiguous: if both species had the same kinetic temperature, equipartition would give T_O=T_Si. Clarify that the observation reflects different kinetic energies per atom in the transient regime.","section":"§III F"}],"recommendation":"major_revision","confidential_remarks":"The strongest and most convincing part of the manuscript is the multiscale demonstration of nonthermal lattice destabilization. The quantitative Fröhlich suppression is overextended because the headline values at T_e=2.2 and 2.6 eV are computed for unstable or saddle-point structures. This is fixable by reframing the quantitative claim to T_e≤2.1 eV or by recomputing at the annealed stable minimum, so I recommend major revision rather than rejection. No concerns about novelty or citation patterns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing you should know: this paper is worth reading for the DNNP resource and the qualitative destabilization picture, but the headline Fröhlich claim is overreach. The authors train electronic-temperature-dependent deep neural network potentials from AIMD, run 3087-atom MD, and sweep Te finely. That is genuinely new relative to prior 72-atom AIMD studies, and the force errors (0.014–0.028 eV/Å) and ground-state phonon comparison (RMSE 0.44 THz) are respectable. The elastic-instability threshold between 2.1 and 2.2 eV, the volume expansion, the bulk modulus collapse, the pCOHP antibonding occupation, and the Bader charge reduction all point the same way. I believe the lattice destabilization is real.\n\nThe soft spot is the quantitative Fröhlich coupling at Te = 2.2 and 2.6 eV. Those structures are already elastically and dynamically unstable — the paper itself calls the trigonal structure a saddle point at 2.2 eV. Computing a harmonic LO frequency and static dielectric constants on that surface and plugging them into Eq. (3) is not well-defined. The two-order-of-magnitude drop in α comes largely from εs/ε∞ → 1, which is exactly the near-cancellation you would expect when linear response is evaluated at the edge of mechanical instability. Also, the 2.2 eV EOS minimum lies outside the sampled volume range, so that volume is an extrapolation. I would not cite those Table I numbers as quantitative predictions. The qualitative trend — α decreasing, the system becoming less polar — is plausible and consistent with Bader charges, but the precise threshold \"Te > 2 eV\" for scattering suppression is not established.\n\nThe non-Maxwellian velocity analysis is suggestive but has no error bars or statistical tests; fine as a qualitative observation. The authors should also deposit the DNNP models and training data, since that is the most reusable part.\n\nWho is this for: people modeling radiation damage, laser-matter interaction, or nonthermal melting in SiO2. They will get a useful proof-of-concept and a coherent qualitative picture. The paper deserves peer review — the central destabilization claim is well supported, and the Fröhlich overreach is fixable with explicit caveats or by restricting the claim to the stable regime. Send it to referees, but ask for revision: recast the Te ≥ 2.2 eV Fröhlich values as not well-defined, or recompute them on a properly relaxed lower-symmetry structure; specify the EOS extrapolation; and deposit the models.","headline":"The qualitative lattice-destabilization story is solid and the Te-dependent DNNPs are a useful resource, but the quantitative Fröhlich-suppression numbers at Te>2 eV rest on harmonic inputs in an unstable structure and should not be taken at face value.","tokens_in":22657,"tokens_out":1935,"would_cite":true,"duration_ms":24400,"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":"The paper claims that in electronically excited α-quartz, raising the electronic temperature to ~2 eV suppresses long-range polar optical-phonon scattering by more than two orders of magnitude, even though the lattice itself is not yet ther","keywords":["α-quartz","silicon dioxide","electronic temperature","deep neural network potential","Fröhlich coupling","polar optical phonon scattering","nonthermal melting","two-temperature model"],"falsifier":"Compute the dynamic dielectric and LO phonon spectral function at Te = 2.2 and 2.6 eV directly from a 3087-atom DNNP-MD trajectory (e.g., from the dipole time-correlation function) instead of from harmonic force constants. If the LO peak in the dielectric loss spectrum remains strong and well defined with a large εs − ε∞ splitting, then the claim that polar optical scattering collapses by more than two orders of magnitude is wrong; alternatively, a THz or infrared pump-probe measurement of the LO phonon reflectivity in laser-excited quartz that shows no loss of oscillator strength near 34 THz","tokens_in":21585,"feed_emoji":"⚡","tokens_out":7913,"duration_ms":79843,"temperature":0.7,"pith_summary":"At high electronic temperature, α-quartz (SiO2) stops behaving like a polar insulator. Training deep neural network potentials on finite-electronic-temperature DFT molecular dynamics lets the authors follow a 3087-atom quartz cell through the first picoseconds after a sudden electronic excitation. They find that the crystal loses mechanical rigidity between 2.1 and 2.2 eV electronic temperature, the bulk modulus drops from 68 to about 2 GPa, and the Fröhlich coupling constant collapses from 1.36 to 0.007 by 2.2 eV. A similar drop in Bader charges indicates a crossover to a nearly nonpolar phase, so polar optical phonon scattering should be strongly suppressed above Te ≈ 2 eV. At the same time, distinct Si and O kinetic temperatures show the lattice does not reach a Maxwell-Boltzmann equilibrium within the first few hundred femtoseconds.","feed_headline":"2 eV of electronic heat strips quartz of polar phonon scattering","feed_subtitle":"The Fröhlich constant drops more than 100-fold, and silicon and oxygen ions fail to equilibrate within the first 100 fs.","key_machinery":"The central device is a set of deep neural network potentials trained separately at several fixed electronic temperatures (Mermin-DFT finite-temperature AIMD on 72-atom cells) that reproduce energies, forces, and virials well enough to yield converged phonon band structures and MD on a 3087-atom supercell. The physical identity that carries the argument is the Fröhlich formula α = e²/(4πε0ħ) √(m*/(2ħωLO)) (1/ε∞ − 1/εs), evaluated at each Te: although ωLO softens, the inverse-dielectric difference collapses as εs→ε∞, so α drops. Elastic stability criteria for trigonal crystals provide the instability threshold, while pCOHP and Bader analysis identify antibonding occupation and ionicity loss a","core_discovery":"Under an instantaneous rise of the electronic temperature Te, α-quartz undergoes a nonthermal destabilization: at Te between 2.1 and 2.2 eV the trigonal structure violates elastic stability criteria, the equilibrium volume expands substantially, and the bulk modulus falls from 68.1 to roughly 2.0 GPa. The mechanism is electronic: conduction-band antibonding states become occupied, weakening Si–O bonds (integrated pCOHP falls from 4.0 to 0.6 eV/bond at 2.6 eV) and reducing Bader charges. Using the electronic- and phonon-band-structure data, the authors estimate the Fröhlich constant α, which drops by more than two orders of magnitude by Te ≈ 2 eV because the dielectric contrast εs/ε∞ approach","pith_inferences":["If the Fröhlich collapse is real, it suggests the electron-phonon coupling parameter used in thermal-spike models is not constant: it should drop sharply once Te passes ~2 eV, softening the predicted track radii—an extension the paper does not quantify.","The species-separated thermalization could be probed directly by time-resolved, element-specific electron or x-ray diffraction; a Debye-Waller measurement distinguishing O and Si sublattice disorder within the first 100 fs would test the DNNP prediction.","Because the α values at Te ≥ 2.2 eV are computed for structures with imaginary harmonic phonons, the 'nonpolar crossover' may actually be a loss of well-defined polar phonons; if so, the suppression is robust, but the threshold value of α may be an artifact of treating a saddle point as a crystal.","A non-adiabatic extension, allowing Te to decrease as the lattice heats, could show whether the nonthermal suppression survives electron cooling; without that, the long-time outcome remains open."],"forward_implications":["At Te ≳ 2 eV, polar optical phonon scattering in quartz is so strongly suppressed that electron-lattice energy exchange and carrier mobility are dominated by nonpolar, deformation-potential channels.","The failure of lattice Maxwell-Boltzmann equilibration within the first few hundred fs means that single-lattice-temperature two-temperature models can misestimate transient energy transfer after ultrafast excitation.","The elastic/shear instability threshold between Te = 2.1 and 2.2 eV gives a quantitative nonthermal melting or amorphization criterion for ion-track and laser-damage modeling.","The predicted crossover to a nearly nonpolar, low-ionic-charge state implies measurable changes in infrared and LO-TO properties that could be seen in transient optical or x-ray experiments.","DNNP-MD at this scale shows that incipient-melting regimes of a 3087-atom cell are accessible with near-DFT accuracy, so similar electronic-temperature-dependent potentials can be built for other radiation-tolerant dielectrics."],"fun_headline_variants":["Nonthermal electronic heat strips quartz of polar phonon coupling","2 eV electronic jolt turns quartz from polar to nonpolar","Femtosecond heat surge collapses quartz's bulk modulus","Electronic heat suppresses polar phonon scattering in quartz","Electronic heating strips quartz of polar scattering"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premises are that harmonic LO-phonon and dielectric parameters remain well defined in structures that are already mechanically unstable at Te ≥ 2.2 eV, and that DNNPs trained on 72-atom molecular dynamics faithfully transfer to a 3087-atom cell across the incipient melting regime; if either fails, the quantitative magnitude and threshold of the Fröhlich suppression are not established.","fun_headline_variants_meta":{"raw":{"variants":["Nonthermal electronic heat strips quartz of polar phonon coupling","2 eV electronic jolt turns quartz from polar to nonpolar","Femtosecond heat surge collapses quartz's bulk modulus","Electronic heat suppresses polar phonon scattering in quartz","Electronic heating strips quartz of polar scattering"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00102,"raw_usage":{"total_tokens":4239,"prompt_tokens":942,"completion_tokens":3297,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":686,"completion_tokens_details":{"reasoning_tokens":3222}},"tokens_in":686,"tokens_out":3297,"duration_ms":23474,"temperature":1.0,"reasoning_tokens":3222,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T01:31:55.329693+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the dynamic dielectric and LO phonon spectral function at Te = 2.2 and 2.6 eV directly from a 3087-atom DNNP-MD trajectory (e.g., from the dipole time-correlation function) instead of from harmonic force constants. If the LO peak in the dielectric loss spectrum remains strong and well defined with a large εs − ε∞ splitting, then the claim that polar optical scattering collapses by more than two orders of magnitude is wrong; alternatively, a THz or infrared pump-probe measurement of the LO phonon reflectivity in laser-excited quartz that shows no loss of oscillator strength near 34 THz","supporting_citations":[],"review_version":1}