{"id":"580f4bd5-cd53-439e-8f7c-df97b635ccd5","arxiv_id":"1908.02580","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"HSE06 hybrid DFT reproduces the experimentally observed charge-ordered structure of κ-D3(Cat-EDT-TTF/ST)2 where GGA fails, and predicts a close-lying noncentrosymmetric phase.","lead":"This paper compares two density functional methods on deuterated molecular conductors, and finds that a hybrid functional, HSE06, correctly reproduces the low-temperature charge-ordered structure that the simpler GGA method fails to stabilize. It also predicts a new noncentrosymmetric ordering pattern with an energy nearly equal to the observed phase.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"CO-phase stability is not actually established: the HSE06 relaxations start from the experimental LT geometry and hold lattice constants fixed at LT values, so the result may reflect initial-state and boundary-condition bias rather than a predicted instability of the high-T phase.","rationale":"I read the paper as doing two things: (1) benchmarking GGA vs HSE06 for internal-coordinate relaxation of an experimentally known CO structure; (2) claiming that HSE06 captures the structural stability of the CO phase. The first is well supported: starting from the experimental P-1 geometry with fixed experimental cell, HSE06 retains the CO distortion and reproduces the central C=C bond lengths, while GGA relaxes them to near equivalence; the footnote 46 HSE06-D3 check strengthens this. The second is not established by the reported calculations. Fixing the LT lattice constants and starting from the LT geometry means the CO phase is never tested as a global or even an unconstrained local minimum; the high-T C2/c phase is never relaxed, and the cell cannot respond to the D ordering. Because the O...O distance and D position are the order parameter, pinning the cell can bias the outcome. This is not an internal inconsistency, but it is a load-bearing gap in the 'stability' language. The 8 meV/f.u. P1/P-1 separation is within hybrid-functional error, so the prediction of a close-lying noncentrosymmetric phase should be read as tentative. The reader's classical-nuclei concern is real and related, but the fixed-cell/no-spontaneous-instability issue is the more direct threat to the stated stability claim. I therefore recommend no change to the CONDITIONAL verdict: the structural reproduction result is credible, but the stability claim needs a variable-cell and/or high-T-starting-point check before it is fully established.","tokens_in":12450,"tokens_out":10115,"duration_ms":109043,"concrete_test":"Run a variable-cell HSE06-D3 relaxation of κ-D3(Cat-EDT-TTF)2 starting from the experimental high-temperature C2/c structure (C2/c lattice parameters and internal coordinates), and separately relax the experimental P-1 structure with the cell free. If the C2/c starting point does not spontaneously develop the CO/D-ordering distortion, or if the free-cell P-1 relaxation moves the O...O distances by more than ~1% from the experimental values, then the claim that HSE06 predicts the CO phase to be structurally stable is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central stability claim (abstract: GGA fails to reproduce 'the structural stability of the CO phase' while HSE06 succeeds; Sec. IV B: HSE06 'highly accurate for reproducing the structures') rests on structural optimizations that (i) start from the experimental low-temperature P-1 structure and (ii) fix the lattice parameters to the experimental LT cell, as stated in Sec. III ('we use the experimental lattice parameters throughout this paper'). This protocol tests only whether that particular local minimum survives internal-coordinate relaxation. It does not test whether the CO phase is a true minimum on the HSE06 potential-energy surface, because the high-temperature C2/c phase is never used as a starting point and the cell is not allowed to relax. Since D ordering is coupled to the O...O hydrogen-bond distance, pinning the cell to the measured LT lattice can itself stabilize the CO internal distortion; the apparent GGA/HSE06 difference may therefore be partly a boundary-condition effect. The predicted noncentrosymmetric P1 phase is only 8 meV/f.u. above P-1, within the expected error of a hybrid functional, so the prediction of a close-lying phase is tentative. The classical-nucleus approximation (Sec. III) is a related limitation, but the fixed-lattice/initial-state gap is the more decisive threat to the paper's 'stability' language.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper investigates the low-temperature charge-ordered (CO) phase of the hydrogen-bonded molecular conductors κ-D3(Cat-EDT-TTF)2 and κ-D3(Cat-EDT-ST)2 using GGA-PBE and HSE06 calculations. The authors fix the lattice constants to the experimental values and relax the internal coordinates. They find that GGA-PBE relaxations starting from the experimental P-1 structure lose the CO distortion, while HSE06 retains it and reproduces the central C=C bond length asymmetry and hydrogen-bond geometry. They also construct a noncentrosymmetric P1 phase by reversing the D displacements and find it lies 8 meV/f.u. above the centrosymmetric phase. The paper concludes that HSE06 is highly accurate for the CO structure and that GGA fails to stabilize it.","tokens_in":12677,"tokens_out":6096,"duration_ms":68186,"significance":"If the central comparison is accepted, the paper provides a useful methodological benchmark: it shows that the standard HSE06 hybrid functional, used without parameter fitting, can preserve a charge-ordered distortion that GGA-PBE destroys, and it yields a concrete prediction of a close-lying ferroelectric-like CO phase that could be investigated experimentally. The authors also explicitly report a numerical check with HSE06-D3 and clearly state the limitations of the fixed-lattice protocol and of the classical treatment of D nuclei. However, because the relaxations begin from the experimental low-temperature geometry and never relax the cell, the paper overstates the thermodynamic stability of the CO phase; the significance of the results is correspondingly that of a conditional structural benchmark rather than a definitive stability calculation.","major_comments":[{"comment":"The central stability conclusion rests on structural relaxations that start from the experimental low-temperature P-1 structure and hold lattice constants fixed to the experimental LT cell, as stated in Sec. III ('we use the experimental lattice parameters throughout this paper') and Sec. IV B (the initial state is 'the experimental structures of the D localized phase'). This protocol tests whether the experimental internal distortion survives internal-coordinate relaxation, but it does not test whether the CO phase is a stable minimum on the HSE06 potential-energy surface, because the high-temperature C2/c phase is never used as a starting point and the cell is not relaxed. Since the D ordering is coupled to the O···O hydrogen-bond distance, pinning the LT cell can itself stabilize the distorted internal coordinates, and the apparent GGA/HSE06 difference may be partly a boundary-condition effect. The abstract and Sec. IV B use the language 'structural stability of the CO phase,' which is stronger than the evidence. I ask the authors to either rephrase the stability claim as 'the CO internal distortion is preserved under HSE06 relaxation at the experimental lattice' or add full cell-plus-internal relaxations from both the C2/c and P-1 starting points.","section":"Secs. III and IV B"},{"comment":"The predicted noncentrosymmetric P1 phase is reported to lie only 8 meV per formula unit above the centrosymmetric phase, with no convergence tests or error estimates for this energy difference, and with no analysis of the dependence on k-point sampling, plane-wave cutoff, or exact-exchange integration grid. Given that HSE06 total-energy errors for competing molecular packings can readily exceed this scale, the Summary's statement that a 'stable' noncentrosymmetric CO phase was found is not justified by the reported data. Please report convergence tests for the energy difference and either compute it with an independent method (e.g., different k-point mesh or HSE06-D3) or explicitly present it as a tentative near-degeneracy.","section":"Sec. IV B (energy difference between P-1 and P1)"},{"comment":"The calculation treats H/D as classical point charges and does not distinguish H from D, as stated in Sec. III: 'the present study does not consider the quantum effects of H or D atoms.' This is load-bearing for the paper's title claim about deuterium-coupled ordering, because the experimental phenomenology is isotope-specific (only the deuterated samples undergo the CO transition), and the authors cite multicomponent DFT studies showing that nuclear quantum motion can change the hydrogen-bond potential from double-well to single-well. The HSE06 classical-nucleus relaxation therefore cannot exclude the possibility that the CO structure is stabilized by neglecting zero-point motion, and it cannot address why H and D behave differently. The authors acknowledge this limitation, but it should be stated in the abstract and conclusions as a condition on the stability claim.","section":"Sec. III (classical treatment of D nuclei)"}],"minor_comments":[{"comment":"There is a typo in the phrase 'van dar Waals interactions'; it should read 'van der Waals interactions.'","section":"Sec. III"},{"comment":"The compound name appears as 'ethylenedithiote-tetrathiafulvalene'; the standard spelling is 'ethylenedithio-tetrathiafulvalene.'","section":"Sec. I"},{"comment":"The HSE06 result for D-Se is semimetallic while experiments show insulating behavior; this qualification is stated in the Discussion, but it should also appear in the Summary so that the accuracy claim for HSE06 is not overgeneralized.","section":"Sec. V"},{"comment":"In the Summary, 'We also proposed possible patterns' should be 'We also propose possible patterns'; in the Acknowledgements, 'One the authors' should be 'One of the authors.'","section":"Sec. VI and VII"},{"comment":"The optimized atomic coordinates for the HSE06 P-1 and P1 structures are not provided as supplemental data; making them available would significantly improve reproducibility and allow other groups to test the 8 meV energy difference.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"This is a competent and honest benchmark study, but the mismatch between the calculation protocol and the word 'stability' is the main obstacle. The stress-test concern about fixed lattice constants and initial-state bias is real: the paper explicitly says it relaxes internal coordinates at experimental lattice parameters, so the title claim of 'stability' is stronger than the evidence. Additional calculations that start from the high-temperature phase and relax the cell, plus convergence tests for the 8 meV energy difference, would move the paper from major revision to acceptance. I do not see evidence of misconduct or hidden errors."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the HSE06-vs-GGA contrast for this compound class: starting from the experimental low-temperature structure, GGA relaxes internal coordinates back toward the symmetric phase, while HSE06 keeps the charge-ordered geometry, with C=C bond lengths matching experiment within ±0.02 Å. That is a real, useful result for people who need a practical functional for these hydrogen-bonded molecular conductors. The prediction of a close-lying noncentrosymmetric P1 phase is also new at the first-principles level and worth taking seriously, even though the phase was anticipated by Naka and Ishihara's effective model.\n\nI want to be clear about the soft spots, because the abstract's word \"stability\" overshoots what the calculations actually show. The relaxations start from the experimental LT structure and the lattice parameters are fixed at the experimental LT values, as stated in Sec. III and IV B. That means the protocol tests whether that particular local minimum survives internal-coordinate relaxation; it does not test whether the CO phase is a true minimum reached from the high-temperature phase. The stress-test note is right: pinning the cell to the measured LT lattice can itself bias the internal distortion, and the GGA/HSE06 difference may be partly a boundary-condition effect. The paper should either relax the cell (with dispersion corrections, which they checked in a footnote do not change internal coordinates much) or at least run a relaxation from the C2/c starting point and report where it goes. Without that, the central claim should be phrased as \"HSE06 preserves the experimental CO geometry at the experimental cell,\" not \"HSE06 stabilizes the CO phase.\"\n\nThe other issues are real but less decisive. The 8 meV/f.u. energy difference between P-1 and P1 is tiny and well within hybrid-functional error, so the predicted noncentrosymmetric phase is tentative. The classical treatment of D nuclei is a stated limitation, and it matters because the phase transition is isotope-specific; but the paper is honest about that and the structural benchmark is still meaningful. The D-Se semi-metallic band structure versus insulating experiment is a further temper, and they acknowledge it.\n\nOverall, this is a careful computational study with clear reporting and no hidden fitting. The citation pattern is fair, including the prior GGA work and the effective-model paper. It deserves serious peer review, but the authors should be asked to soften the stability language and, if feasible, test relaxation from the high-temperature phase. I would send it out rather than desk-reject.","headline":"A useful DFT benchmark for a coupled D/CO phase: HSE06 reproduces the experimental LT geometry while GGA does not, but the 'stability' claim is weaker than the protocol supports because the cell is fixed and relaxations start from the LT structure.","tokens_in":13221,"tokens_out":1521,"would_cite":true,"duration_ms":18585,"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":"A screened hybrid functional stabilizes the charge-ordered crystal that GGA misses.","keywords":["charge ordering","hydrogen-bonded molecular conductors","HSE06 hybrid functional","deuterium isotope effect","density functional theory","Mott insulator","noncentrosymmetric phase"],"falsifier":"Take the optimized HSE06 potential-energy surface for the shared deuterium atom and recompute it with deuterium zero-point motion included, for example by path-integral or multicomponent DFT on the same crystal; if the double-well profile flattens into a single well, the classical HSE06 stabilization of the charge-ordered phase would not survive quantum treatment.","tokens_in":12237,"feed_emoji":"⚛️","tokens_out":5802,"duration_ms":61335,"temperature":0.7,"pith_summary":"This paper tries to show that a screened hybrid density functional, HSE06, can reproduce the experimentally observed charge-ordered structure of two deuterated hydrogen-bonded molecular conductors, while the standard GGA functional cannot. Getting this structure right matters because the ordering couples deuterium positions in hydrogen bonds to the distribution of π electrons, and reliable structures are needed to build effective models of the coupled order. The paper further claims that HSE06 predicts a second, noncentrosymmetric charge-ordered phase, close in energy to the measured one, with a different deuterium arrangement. If true, this turns a known failure of plain DFT into a quantitative tool for this class of molecular conductors.","feed_headline":"Hybrid functional stabilizes the charge-ordered crystal GGA misses","feed_subtitle":"A screened-exchange DFT calculation reproduces measured bond lengths and predicts a nearby noncentrosymmetric phase.","key_machinery":"The load-bearing object is the coupled order parameter formed by the off-center deuterium position in the O–D···O hydrogen bond and the charge disproportionation between the two types of dimers. The computational switch that makes the argument work is the fraction of exact exchange in HSE06, which localizes the π wavefunctions enough to stabilize the charge imbalance and the associated molecular distortions. The paper's explanatory device is a dimer molecular-orbital diagram: the deuterium-bearing monomer sits lower in HOMO energy, and the deuterium-free dimer has a shorter inter-monomer distance, so its antibonding level is pushed above the deuterium-bearing dimer's, emptying one level and filling the other.","core_discovery":"Starting from the measured low-temperature P-1 structure of κ-D3(Cat-EDT-TTF)2 and its selenium analog, structural relaxation with GGA-PBE returns to a symmetric, high-temperature-like geometry with essentially no charge disproportionation between the two molecular units. Relaxation with HSE06 instead settles near the measured geometry, reproducing central C=C bond lengths of 1.35 Å and 1.38 Å for the deuterium-bearing and deuterium-free units. HSE06 also opens a 0.04 eV indirect gap in the sulfur compound, while the selenium compound remains semi-metallic, in contrast to experiment. Using the HSE06 landscape, the paper constructs a noncentrosymmetric P1 structure by reversing the deuterium displacement pattern on half the hydrogen bonds; this phase optimizes to a stable structure only 8 meV per formula unit above the centrosymmetric ground state. The stabilization is attributed to two comparable effects: the lower HOMO energy of the monomer carrying the deuterium and the stronger dimerization of the deuterium-free units.","pith_inferences":["Beyond the paper: because the experimental transition occurs only in deuterated samples, the classical-HSE06 success on deuterated compounds suggests the functional may be capturing an enthalpy landscape that nuclear quantum motion then selects between; a quantum-nuclear calculation on the same surface would test this directly.","Beyond the paper: the 8 meV per formula unit separation between the two phases is close to the accuracy limits of the method, so a measured electric polarization or dielectric anomaly in the selenium compound could discriminate between the centrosymmetric and noncentrosymmetric arrangements.","Beyond the paper: if zero-point motion shifts the shared deuterium toward the center of the hydrogen bond, then the HSE06 double-well picture may over-stabilize charge order; path-integral or multicomponent DFT calculations could reveal whether the classical result is an artifact or the dominant physics."],"forward_implications":["HSE06 structural optimization can be used quantitatively for charge-ordered phases in this family of molecular conductors, including bond lengths that Raman measurements probe.","The predicted noncentrosymmetric P1 phase, only 8 meV per formula unit above the observed phase, is a concrete candidate to search for under pressure or with altered deuteration patterns.","The molecular-orbital energy differences provide first-principles parameters for effective models of coupled deuterium–π-electron order, filling a gap left by models that neglect the inter-monomer distance asymmetry.","The 0.04 eV gap in the sulfur compound matches its insulating behavior, while the semi-metallic result for the selenium compound marks a specific quantitative limit of HSE06 for this family."],"supporting_citations":[{"why":"Previous GGA relaxation of the hydrogen analog that showed too-small charge-order distortion; the baseline this paper improves on.","marker":"[16]"},{"why":"Reports the D-S low-temperature structure, phase transition, and Raman peaks used as the experimental target.","marker":"[19]"},{"why":"Provides the D-Se low-temperature structure and lattice parameters used in the calculations.","marker":"[20]"},{"why":"Effective model that predicted two coupled orders, including the noncentrosymmetric phase tested here.","marker":"[23]"},{"why":"Defines the GGA-PBE functional whose relaxations collapse back to the high-symmetry phase.","marker":"[24]"},{"why":"Defines the HSE06 hybrid functional whose structural relaxations reproduce the charge-ordered geometry.","marker":"[25]"},{"why":"Multicomponent DFT study of nuclear quantum effects on the H/D potential surface, which the authors position as complementary to their classical treatment.","marker":"[60]"}],"fun_headline_variants":["HSE hybrid functional stabilizes charge order that GGA misses","Screened-exchange DFT reproduces measured bond lengths and predicts new phase","GGA fails, HSE succeeds: charge order in hydrogen-bonded conductor","Hybrid functional predicts new noncentrosymmetric charge-ordered phase","Charge order in hydrogen-bonded conductor: only hybrid DFT gets it right"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculations treat the shared deuterium as a classical point charge, so any nuclear quantum motion that might explain why only deuterated samples order is left out.","fun_headline_variants_meta":{"raw":{"variants":["HSE hybrid functional stabilizes charge order that GGA misses","Screened-exchange DFT reproduces measured bond lengths and predicts new phase","GGA fails, HSE succeeds: charge order in hydrogen-bonded conductor","Hybrid functional predicts new noncentrosymmetric charge-ordered phase","Charge order in hydrogen-bonded conductor: only hybrid DFT gets it right"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001175,"raw_usage":{"total_tokens":4877,"prompt_tokens":982,"completion_tokens":3895,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":598,"completion_tokens_details":{"reasoning_tokens":3801}},"tokens_in":598,"tokens_out":3895,"duration_ms":26218,"temperature":1.0,"reasoning_tokens":3801,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:39:07.426240+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the optimized HSE06 potential-energy surface for the shared deuterium atom and recompute it with deuterium zero-point motion included, for example by path-integral or multicomponent DFT on the same crystal; if the double-well profile flattens into a single well, the classical HSE06 stabilization of the charge-ordered phase would not survive quantum treatment.","supporting_citations":[{"cited_title":"Tsumuraya , author H","cited_arxiv_id":null,"evidence_quote":"Previous GGA relaxation of the hydrogen analog that showed too-small charge-order distortion; the baseline this paper improves on."},{"cited_title":"Ueda , author S","cited_arxiv_id":null,"evidence_quote":"Reports the D-S low-temperature structure, phase transition, and Raman peaks used as the experimental target."},{"cited_title":"Ueda , author A","cited_arxiv_id":null,"evidence_quote":"Provides the D-Se low-temperature structure and lattice parameters used in the calculations."},{"cited_title":"Naka \\ and\\ author S","cited_arxiv_id":null,"evidence_quote":"Effective model that predicted two coupled orders, including the noncentrosymmetric phase tested here."},{"cited_title":"Yamamoto , author Y","cited_arxiv_id":null,"evidence_quote":"Multicomponent DFT study of nuclear quantum effects on the H/D potential surface, which the authors position as complementary to their classical treatment."}],"review_version":1}