{"id":"7dd62770-9f97-4692-80f3-cf0c9f912563","arxiv_id":"2412.21150","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"DFT phonon calculations show the P4/mmm phase of single-layer-trilayer La3Ni2O7 is unstable at all pressures up to 30 GPa, and the lowest-energy distortions combine two instabilities, contrary to experimental refinements.","lead":"A computational study finds that the parent crystal structure of the nickelate superconductor La3Ni2O7 in its single-layer-trilayer form remains mechanically unstable under pressure up to 30 GPa, with several nearly degenerate distorted structures competing. The calculations predict distortions that differ from experimental refinements, which may explain why the experimental structure is hard to pin down.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Nonmagnetic PBE without Hubbard U leaves sub-meV structural energy differences and the 20-30 GPa phonon instability unvalidated; this is the load-bearing assumption.","rationale":"Agreeing with the reader's weakest assumption, I identify the neglect of magnetism and Hubbard U as the most load-bearing approximation. The paper's quantitative claims hinge on sub-meV per atom energy differences and on an instability that persists at 20–30 GPa, where experiments see an undistorted P4/mmm phase. For nickelates, both spin polarization and correlation strength (U) are known to alter structural and electronic properties; without testing them, the experimentally conflicting central claim is not firmly established. The reader's conditional verdict is appropriate: the concern is addressable by additional calculations, but the paper as written does not fully support the claim. Hence no verdict change.","tokens_in":13303,"tokens_out":4989,"duration_ms":67597,"concrete_test":"Recompute the phonon dispersions at 20 and 30 GPa using spin-polarized PBE and PBE+U (U_eff ≈ 4–6 eV on Ni d) with the same 4×4×2 q-grid and ultrasoft pseudopotentials, checking that the nondegenerate branch at M−A remains imaginary; additionally cross-check the 30 GPa result with VASP PAW frozen-phonons to rule out pseudopotential artifacts.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II states that all calculations use PBE without Hubbard U, spin polarization, or spin-orbit coupling. The paper's central claim—that the P4/mmm parent is dynamically unstable up to 30 GPa, and that the lowest-energy distortions at 0/10 GPa combine nondegenerate and doubly-degenerate instabilities—rests on energy differences of 0.2–0.4 meV/atom (Tables I–III). These are comparable to the expected accuracy of semilocal DFT in a strongly correlated oxide, where static correlations and spin fluctuations are known to affect structural energetics (e.g., Refs. 24, 33, 38, 40–41). No spin-polarized or DFT+U test is reported, so the conflicting experimental observation of a P4/mmm phase above ~12.8 GPa could simply reflect a functional artifact. If the nondegenerate M−A branch becomes stable under PBE+U or with spin polarization, the headline claim collapses.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports first-principles phonon and total-energy calculations for the single-layer-trilayer (SL-TL) polymorph of La3Ni2O7 in the parent P4/mmm structure at 0, 10, 20, and (for one branch) 30 GPa. Using density functional perturbation theory in Quantum ESPRESSO, the author finds a nondegenerate unstable branch along the M–A Brillouin-zone edge at all investigated pressures, with additional doubly degenerate instabilities at 0 and 10 GPa. Isotropy-subgroup enumeration based on the M2+, M5+, A4-, and A5- irreps yields 26 candidate order parameters; structures are generated from the computed phonon eigenvectors and relaxed in VASP. The main results are that the lowest-energy structures at 0 and 10 GPa condense both the nondegenerate and doubly degenerate instabilities, that the experimental Fmmm and Imma refinements are higher in energy, that Cmmm cannot be stabilized, and that at 20 GPa the parent phase remains unstable with a small energy gain, in contrast to the experimental observation of a P4/mmm phase above about 12.8 GPa.","tokens_in":13451,"tokens_out":7802,"duration_ms":79154,"significance":"If the results are quantitatively correct, they offer a coherent explanation for the difficulty of refining the ambient-pressure structure of SL-TL La3Ni2O7: many distorted structures are nearly degenerate because of weakly dispersive unstable branches and weak interlayer coupling. The systematic group-theoretical enumeration and the construction of candidate structures from computed phonon eigenvectors are methodologically clean and involve no fitted parameters. The prediction that the lowest-energy distortions require simultaneous condensation of modes at M and A, rather than only the A5--type modes used in experimental refinements, is concrete and testable by future diffraction or transmission-electron-microscopy work. However, the quantitative conclusions rest entirely on nonmagnetic PBE calculations without Hubbard U, and the energy differences involved are at the 0.1-0.4 meV/atom level, which is near the expected accuracy of the functional.","major_comments":[{"comment":"Section II: The manuscript reports no convergence tests for the plane-wave cutoff, k-point grid, or phonon q-grid. The phonon dispersions are computed in Quantum ESPRESSO with an 8x8x2 k-grid and a 4x4x2 q-grid at 60/600 Ry cutoffs, while the structural relaxations are done in VASP with an 8x8x3 k-grid and a 500 eV cutoff, and the two steps use different pseudopotential families (ultrasoft vs PAW). Because Tables I-III compare energy differences as small as 0.2 meV/atom, the lack of convergence data and the code/pseudopotential mismatch leave open the possibility that the reported energy rankings are numerical artifacts. Please provide convergence tests and, for the key structures, repeat the relaxations with the same code/pseudopotentials as the phonon calculations.","section":"II"},{"comment":"All calculations are nonmagnetic PBE without Hubbard U or spin-orbit coupling, and no test of spin-polarized or correlated treatments is reported. This matters because the central claims--that the M2+/A4- branch remains unstable at all pressures up to 30 GPa and that the lowest-energy structures at 0 and 10 GPa combine the nondegenerate and doubly-degenerate instabilities--rest on PBE imaginary frequencies and energy gains as small as -0.4 meV/atom (Table III). The existing literature on La3Ni2O7 (e.g., Refs. 24, 33, 38, 40-41) shows that spin and orbital correlations affect the structural and electronic energetics of this material. A concrete test would be to recompute the phonon instabilities and relaxed energies at 20 GPa with spin-polarized PBE and PBE+U; if the M2+ branch stabilizes under correlation, the pressure-dependence claim collapses. This is a load-bearing validation, not an optional refinement.","section":"II and Tables I-III"},{"comment":"The abstract and conclusion state that the P4/mmm phase is unstable at all pressures 'up to 30 GPa,' but the only supporting statement in the body is 'I did calculations also at 30 GPa and found that this branch remains unstable.' No 30 GPa frequencies, dispersion, or relaxation data are shown. Since the experimental study reports a transition to P4/mmm above about 12.8 GPa, the 30 GPa result is one of the most striking claims of the paper. Please report the 30 GPa data (at minimum the M and A frequencies and a brief convergence check) or revise the abstract and conclusion to state the pressure range actually documented.","section":"III, fourth paragraph, and Section IV"},{"comment":"At 20 GPa the three distorted structures lie within 0.2 meV/atom of one another and only 0.4 meV/atom below the parent, yet the text concludes that structural distortions are energetically favorable at 20 GPa, contrary to experiment. The manuscript itself acknowledges that the 10 GPa near-degeneracies are 'within numerical accuracy'; the same caveat applies a fortiori to the 20 GPa energy gain. Without a quantitative uncertainty estimate, this statement is not supported. Please provide an error bar from convergence tests or soften the claim.","section":"III, Table III"}],"minor_comments":[{"comment":"Introduction, first paragraph: 'supercondutivity' is a typo for 'superconductivity.'","section":"I"},{"comment":"The caption contains 'Structural distortiosns,' which should be 'Structural distortions.'","section":"Fig. 2 caption"},{"comment":"The text says the lowest energy gain at 10 GPa is '-1.9 eV/atom,' but Table II and the surrounding discussion indicate that this should be '-1.9 meV/atom.'","section":"III, paragraph after Table II"},{"comment":"'Methfesse-Paxton smearing' should be 'Methfessel-Paxton smearing.'","section":"II"},{"comment":"'less then 50 cm-1' should be 'less than 50 cm-1.'","section":"III, paragraph on 20 GPa"}],"recommendation":"major_revision","confidential_remarks":"I see no evidence of circularity or invented parameters; the calculations are self-contained first-principles simulations, and the group-theoretical enumeration is not circular. The main risk is methodological validation: the paper's quantitative conclusions depend on nonmagnetic PBE energy differences that are at or below the expected functional accuracy, and the 30 GPa claim is not actually documented in the text. The author should be asked to provide convergence data, a spin-polarized or DFT+U check at key pressures, and the 30 GPa results. The manuscript is within the scope of a condensed-matter theory journal, and a data availability statement or input-file repository would substantially improve reproducibility given the near-degeneracies reported."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this is the first phonon calculation for the single-layer-trilayer La3Ni2O7 phase, and the central result—that P4/mmm is dynamically unstable up to 30 GPa and the lowest-energy distortions at 0/10 GPa combine the nondegenerate and doubly-degenerate branches—is clearly presented and a useful challenge to the experimental refinements. The group-theoretical enumeration is thorough, and the paper is honest about the disagreement with experiment at high pressure.\n\nWhere it earns credit: the phonon dispersions are systematic, the isotropy-based mapping of 26 order parameters is careful, and the tables give a complete picture of the near-degenerate energy landscape. The observation that Cmmm relaxes back to P4/mmm is a concrete negative result that helps constrain the experimental models.\n\nThe soft spots are real, though. The energy differences that separate the candidate ground states are 0.2–0.4 meV/atom, right at the edge of what PBE can be trusted to resolve in a correlated oxide, and the paper does not test spin polarization or Hubbard U at all. The stress-test concern lands: if the nondegenerate M−A branch stabilizes under PBE+U, the headline about the combined condensation collapses. The stability of the nondegenerate branch up to 30 GPa is the load-bearing claim, and it is only checked at the plain PBE level. There are also no convergence tests reported for k-points, cutoffs, or q-grids, and the phonons and relaxations use two different codes and pseudopotentials, which is a minor consistency issue but should be addressed. The lack of input data (structures, pseudopotentials) also makes independent verification harder.\n\nI would not call any of this fatal. The phonon instability pattern itself is probably robust qualitatively, and the paper's own near-degeneracy results make the fragility of the ground-state ordering clear. The author's suggestion that the distortions have short coherence length is a plausible reading of the experimental difficulty. But as written, the structural ground-state prediction is a PBE statement, not a settled one.\n\nThis paper deserves a serious referee. The right revision adds spin-polarized and DFT+U checks on the two key branches, convergence tests, and ideally makes the input structures available. I would send it out.","headline":"First phonon study of the SL-TL La3Ni2O7 polymorph; the instability pattern is solid PBE physics, but the sub-meV ground-state claims need Hubbard U and magnetic checks.","tokens_in":13964,"tokens_out":2447,"would_cite":true,"duration_ms":24665,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The undistorted P4/mmm model of single-layer-trilayer La3Ni2O7 is dynamically unstable at all pressures up to 30 GPa, and the lowest-energy structures condense both a nondegenerate and a doubly-degenerate phonon instability, contrary to…","keywords":["layered nickelates","La3Ni2O7","phonon instabilities","structural distortion","density functional theory","high pressure","superconductivity"],"falsifier":"A high-pressure X-ray diffraction measurement at 20 GPa that can detect a roughly 16.5 degree rotation of the middle-layer NiO6 octahedra in the trilayer would settle whether the predicted small distortion exists, since experiments currently refine this pressure to P4/mmm.","tokens_in":13063,"feed_emoji":"⚛️","tokens_out":7467,"duration_ms":64966,"temperature":0.7,"pith_summary":"The paper uses first-principles phonon calculations to show that the tetragonal P4/mmm model of single-layer-trilayer La3Ni2O7, a member of the layered nickelate family with superconducting signatures, is dynamically unstable at every pressure studied, from 0 to 30 GPa. A nearly dispersionless, nondegenerate phonon branch is unstable along the Brillouin-zone edge M–A at all pressures, and at 0 and 10 GPa additional doubly-degenerate branches are also unstable. By generating and relaxing the distortions allowed by symmetry, the author finds that the lowest-energy structures at 0 and 10 GPa involve condensation of both the nondegenerate and doubly-degenerate instabilities, in contrast to the experimental Fmmm and Imma refinements that involve only the doubly-degenerate branch. At 20 GPa, distortions are still energetically favorable, contradicting experiments that see only P4/mmm. The work implies that the experimentally observed structure may be a superposition of many nearly degenerate distorted phases with short coherence length.","feed_headline":"Predicted: nickelate distorts at every pressure up to 30 GPa","feed_subtitle":"The true low-energy structure mixes two lattice instabilities, unlike the experimental refinements.","key_machinery":"The central objects are the unstable phonon branches of the P4/mmm phase along the Brillouin-zone edge M→A. The nondegenerate branch (irreps M+2 and A-4) is a rotation of the middle-layer NiO6 octahedra within the trilayer; the doubly-degenerate branches (M+5 and A-5) involve rotations of all octahedra in planes parallel to c. The paper uses density-functional perturbation theory for phonons and group-theoretical enumeration (isotropy subgroups and order parameters) to generate candidate distorted structures from the unstable-mode eigenvectors, then relaxes them with DFT total-energy calculations. The flatness of the unstable branch and the weak interlayer coupling are what create the large manifold of nearly degenerate structures.","core_discovery":"The central discovery is that the undistorted P4/mmm phase of single-layer-trilayer La3Ni2O7 is not the true ground state at any pressure up to 30 GPa: a nondegenerate phonon mode with irreps M+2 at M and A-4 at A is unstable along the entire Brillouin-zone edge M–A at all pressures, and at lower pressures two doubly-degenerate branches (M+5/A-5) are also unstable. The lowest-energy relaxed structures at 0 and 10 GPa condense both the nondegenerate and doubly-degenerate instabilities, with five structures within 0.4 meV/atom at 0 GPa and ten within numerical accuracy at 10 GPa. This near degeneracy, caused by the layered stacking and the flatness of the unstable branch, means the octahedral rotations are essentially uncorrelated along the c axis, so any actual crystal would show short coherence length. The experimentally proposed Cmmm structure relaxes back to P4/mmm and cannot be stabilized.","pith_inferences":["If the true structure at high pressure is a disordered stack of nearly degenerate distortions, then electronic-structure models built on the ideal P4/mmm phase may miss short-range structural effects on the Ni d-orbitals and on pairing.","The pressure evolution of the instabilities suggests a crossover between 10 and 20 GPa from two coexisting instabilities to a single one; careful diffraction or Raman experiments in that range could look for changes in the distortion pattern.","Testing the role of electron correlations, for example with DFT+U or hybrid functionals, could either stabilize or suppress the nondegenerate mode; such a calculation would provide a sharp test of whether the predicted instability is an artifact of the PBE functional.","The prediction that Cmmm cannot be stabilized at 0 GPa could be checked by re-examining the published refinements: if the Cmmm model is truly the best fit, it may be stabilized by factors not captured here, such as oxygen vacancies or strain."],"forward_implications":["The superconducting phase of single-layer-trilayer La3Ni2O7 above roughly 13 GPa is likely not the undistorted P4/mmm structure; the predicted small distortions with short coherence length may be why diffraction refinements see P4/mmm.","The experimentally proposed Fmmm and Imma structures are not the lowest-energy distortions; lower-energy structures involve condensation of the nondegenerate branch, with middle-layer octahedral rotations.","The large degeneracy of distorted structures at 0 and 10 GPa implies that the material will display stacking disorder and short out-of-plane coherence of the structural distortions, complicating structure determination.","At 20 GPa the energy gain is only about 0.4 meV/atom, yet the middle-layer octahedra rotate by about 16.5 degrees; such a distortion may be detectable by local probes even if diffraction sees an average P4/mmm."],"supporting_citations":[{"why":"Reports the SL-TL phase, its P4/mmm, Fmmm, Cmmm, and Imma refinements, and the finding that P4/mmm describes the high-pressure phase above about 12.8 GPa; this is the experimental baseline the calculations challenge.","marker":"[45]"},{"why":"Reports the Cmmm structure as the best fit to ambient-pressure X-ray data for the SL-TL phase; the paper shows this Cmmm structure relaxes back to P4/mmm.","marker":"[46]"},{"why":"Also reports Cmmm as the best fit and provides an independent experimental refinement of the same phase; used as a comparison for the instability analysis.","marker":"[47]"},{"why":"Earlier DFT study that found Fmmm and P4/mmm to be the lowest-energy structures at 0 and 16 GPa; this paper's relaxation results are compared with it.","marker":"[48]"}],"fun_headline_variants":["La3Ni2O7 never undistorted up to 30 GPa","Two instabilities condense in nickelate's true low-energy shape","Calculations find nickelate distortions at every pressure","Nickelate's ground state combines two lattice instabilities"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculations assume that the PBE exchange-correlation functional, without Hubbard U, spin polarization, or spin-orbit coupling, correctly captures sub-meV per atom energy differences between competing distorted structures and the phonon instabilities in this correlated oxide.","fun_headline_variants_meta":{"raw":{"variants":["La3Ni2O7 never undistorted up to 30 GPa","Two instabilities condense in nickelate's true low-energy shape","Calculations find nickelate distortions at every pressure","Nickelate's ground state combines two lattice instabilities"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000505,"raw_usage":{"total_tokens":2547,"prompt_tokens":1111,"completion_tokens":1436,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":727,"completion_tokens_details":{"reasoning_tokens":1365}},"tokens_in":727,"tokens_out":1436,"duration_ms":14283,"temperature":1.0,"reasoning_tokens":1365,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:00:23.146595+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A high-pressure X-ray diffraction measurement at 20 GPa that can detect a roughly 16.5 degree rotation of the middle-layer NiO6 octahedra in the trilayer would settle whether the predicted small distortion exists, since experiments currently refine this pressure to P4/mmm.","supporting_citations":[{"cited_title":"Puphal, P","cited_arxiv_id":null,"evidence_quote":"Reports the SL-TL phase, its P4/mmm, Fmmm, Cmmm, and Imma refinements, and the finding that P4/mmm describes the high-pressure phase above about 12.8 GPa; this is the experimental baseline the calculations challenge."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the Cmmm structure as the best fit to ambient-pressure X-ray data for the SL-TL phase; the paper shows this Cmmm structure relaxes back to P4/mmm."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Also reports Cmmm as the best fit and provides an independent experimental refinement of the same phase; used as a comparison for the instability analysis."},{"cited_title":"Zhang, L.-F","cited_arxiv_id":null,"evidence_quote":"Earlier DFT study that found Fmmm and P4/mmm to be the lowest-energy structures at 0 and 16 GPa; this paper's relaxation results are compared with it."}],"review_version":1}