{"id":"cbbee222-962e-401e-a9eb-8bf3f35dca56","arxiv_id":"2608.02243","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Hydrogen doping of NdNiO3 fills O 2p ligand holes and forms electron polarons that weaken Ni-O hybridization, driving Ni eg electrons into a localized d8 state and causing a Mott metal-insulator transition.","lead":"Using first-principles calculations, the authors propose that hydrogenating the perovskite NdNiO3 forces electrons into oxygen orbitals, creating 'polarons' that freeze nickel electrons and turn the material from a metal into an insulator. The work also compares proton movement in NdNiO3 and SmNiO3, which matters for designing fuel-cell electrolytes that conduct protons but block electrons.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central mechanism is extracted from a single DFT+U setting (PBEsol+U with Ueff=2.0 eV); no sensitivity check is reported for the ligand-hole vs Ni-eg occupation that the whole anti-doping/polaron claim hinges on.","rationale":"The reader identified Ueff=2.0 eV as the weakest assumption, and my reading agrees. The paper is otherwise internally consistent: structural results match experiment, the PDOS/COHP/Wannier/Bader analyses are mutually supportive, and the proton-transport discussion is explicitly framed as qualitative. But the central mechanistic claim is ultimately a statement about where added electrons go and why localization follows. That statement is determined by a single Hubbard parameter in a functional that is known to over-delocalize. The manuscript's only justification for U is consistency with previous work and the pristine magnetic ground state, neither of which directly constrains the hydrogenated charge-transfer balance. A U-sensitivity test is the natural, decisive check: DFT+U with varied U is cheap enough to run, and if the ligand-hole occupation is stable across U, the central claim is materially strengthened. If it is not, the paper would need to be repositioned as a prediction conditional on a particular electronic-structure model. The reader's verdict of CONDITIONAL already reflects this; I would not escalate to REJECT because the structural and experimental consistency provides independent support. I therefore recommend no change to the verdict, while emphasizing that the U-sensitivity test is the specific step that would most directly test the central claim.","tokens_in":16920,"tokens_out":11186,"duration_ms":108318,"concrete_test":"Repeat the key calculations for HNdNiO3 (PDOS, COHP, Wannier occupations, bond lengths, gap) with Ueff = 1, 2, 3, and 4 eV, and at least one independent functional such as HSE06 or a PBEsol+U+J variant. Monitor whether the added electron remains predominantly in O 2p ligand-hole states and whether the ~0.8 eV gap and Ni-O elongations persist. If the occupation shifts to Ni eg at U>2 eV or the gap closes/appears only over a narrow U window, the anti-doping/polaron mechanism is parameter-dependent and the central claim needs to be weakened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that electrons introduced by hydrogen occupy O 2p ligand-hole states of the Ni-O d8L hybrid rather than Ni eg states, and that this occupation drives polaron-mediated localization. That assignment is controlled by the relative Ni 3d – O 2p charge-transfer energetics, which in this calculation are set by Ueff=2.0 eV applied only to Ni 3d, with Nd/Sm 4f frozen in the core. The manuscript justifies this U value by citing earlier PBEsol+U work and by the statement that larger U yields a monoclinic ferromagnetic ground state inconsistent with experiment, but it reports no sensitivity study for the hydrogenated phase itself. The relevant quantity is not the pristine magnetic ground state but the energy of adding an electron to the d8L hybrid; self-interaction errors and the small U could easily bias the occupation toward O 2p. If the true charge-transfer parameter is larger, the added electrons would instead occupy Ni eg, and the polaron-mediated weakening of Ni-O hybridization would not be the cause of localization. The distinction from the conventional 'electrons go into Ni' Mott picture would collapse. Since the entire narrative—anti-doping, electron polarons, weakened hybridization, d8 Mott gap—is read out of this one electronic-structure setting, the absence of a U/functional check is a load-bearing gap. This is not a disagreement with consensus; it is an untested sensitivity in the central causal chain.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents first-principles DFT+U calculations on hydrogenated NdNiO3 and SmNiO3, aiming to explain the hydrogen-induced metal-insulator transition (MIT) and proton conduction. The central claim is that electrons introduced by hydrogen doping occupy O 2p ligand-hole states of the Ni-O d8L hybrid, forming electron polarons that, together with proton polarons, weaken Ni-O hybridization and drive the itinerant Ni eg electrons toward a localized d8 configuration, opening a Mott gap. The authors support this with PDOS, COHP, MLWF, Bader charge, and structural distortion analyses, and also report hydrogen insertion energies and proton migration barriers/diffusion coefficients for NdNiO3 and SmNiO3.","tokens_in":17255,"tokens_out":4197,"duration_ms":41697,"significance":"If the proposed mechanism is correct, it provides a coherent atomistic picture of the hydrogen-induced MIT in rare-earth nickelates, distinguishing it from the simpler electron-filling Mott picture and connecting it to polaron formation. The paper is strengthened by the use of multiple complementary electronic-structure diagnostics, the agreement of the calculated ground-state structural parameters with experiment, the absence of parameter fitting beyond a literature-inherited Ueff, and the open data statement. The proton-transport results, while more qualitative, offer comparative insight for electrolyte design. However, the central mechanistic conclusion rests on a single DFT+U parameter setting and on an under-specified electron-doped reference calculation, so the current evidence is not yet conclusive.","major_comments":[{"comment":"The entire anti-doping/polaron mechanism depends on the relative energy of Ni eg vs O 2p ligand-hole states, which in this calculation is controlled by Ueff = 2.0 eV applied only to Ni 3d (with Nd 4f frozen in the core). The authors justify this value by consistency with prior PBEsol+U work and by the statement that larger U yields an incorrect magnetic ground state for pristine nickelates, but no sensitivity check is reported for the hydrogenated phase or for the electron-doped reference. The pristine magnetic ground state does not directly constrain the charge-transfer energy relevant to adding an electron to the d8L hybrid. A larger U could move the added electron from O 2p into Ni eg, collapsing the proposed mechanism. I request calculations at additional Ueff values (e.g., 1.0, 3.0, 4.0 eV) or a hybrid-functional test on HNdNiO3, reporting the PDOS/COHP/MLWF occupations, band gap, a","section":"Computational Method, Ueff"},{"comment":"The electron-doped NdNiO3 calculation is central to isolating the electron-polaron effect from the proton-polaron effect, but the computational protocol is not specified. Adding 'eight extra electrons' to a periodic 40-atom supercell requires charge neutralization; no compensating background or other procedure is mentioned. If a uniform background was used, its effect on total energies, structural relaxations, and the resulting band gap should be discussed. It is also unclear whether the electron-doped structures were fully relaxed with the same criteria as the hydrogenated ones, and what magnetic ordering was assumed. Without this information, the 0.55 eV gap and the bond-length elongations reported for this reference cannot be properly evaluated.","section":"Results and Discussion B"}],"minor_comments":[{"comment":"The MLWF evidence for weakened hybridization and eg localization is presented only qualitatively ('the weight ... is reduced', 'shapes become closer to ideal'). Providing numeric MLWF occupation matrices or oxygen weights would make the central electron-localization claim more quantitative and easier to assess.","section":"Results and Discussion B"},{"comment":"The migration barriers and diffusion coefficients are obtained with the CHGNet machine-learning potential, not with the DFT+U approach used for the electronic-structure analysis. The manuscript does state that CHGNet is used for qualitative trends, but for a journal publication it would be helpful to benchmark at least one barrier (e.g., the intraoctahedral barrier in NdNiO3) against direct DFT+U NEB, or to state the expected MLIP error for these correlated oxides.","section":"Proton conduction C"},{"comment":"There are several presentation issues: 'electron-doped symstem' should be 'system'; the isosurface unit in Fig. 3(e) reads '0.01 e−3' and should presumably be 'e Å⁻³' or similar; the d8L notation should be defined at first use; and in the Wannier-function description it should be stated explicitly why only the spin-up channel is considered, given that spin-down states also contribute to the insulating gap.","section":"General / Fig. 3"},{"comment":"The column layout of Table I is hard to parse: for pristine NdNiO3, the experimental values are given alongside PBEsol+U and LDA+U+J, but for HNdNiO3 no experimental structural parameters are listed even though the text refers to agreement with experiment for the volume expansion. Please clarify which entries are experimental and cite the corresponding sources in the table caption.","section":"Table I"}],"recommendation":"major_revision","confidential_remarks":"The manuscript makes a strong and interesting mechanistic claim, but the current evidence is too dependent on one U value and on an under-specified electron-doped calculation. The requested sensitivity tests are standard and feasible; I would be willing to look at a revised version. The proton-conduction part is more exploratory and should be framed as such if no DFT validation of the MLIP barriers is provided."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The useful core of this paper is the mechanistic bridge it builds between ligand-hole filling and Ni eg localization. The anti-doping observation itself was already reported by Gao et al., but the authors make a stronger claim: the added electrons occupy O 2p ligand-hole states, form electron polarons, and together with proton polarons weaken Ni-O hybridization enough to drive the eg electrons into a localized d8 configuration. That is a new synthesis, and the evidence they marshal for it is more consistent than most DFT+U papers I see. The PDOS, COHP, MLWF shapes, Bader charges, structural distortions, and the electron-doped reference calculation all point the same way, and the lattice parameters match experiment well.\n\nThe proton-transport section is also a real addition. The comparison between Nd and Sm — smaller A-site cation gives easier hydrogen uptake but slower diffusion — is clear and physically plausible. The finding that hydrogenation creates a fast 1D intraoctahedral pathway while blocking 3D percolation is concrete and testable.\n\nWhere I hesitate is the Ueff = 2.0 eV sensitivity. The whole narrative depends on the added electrons going into O 2p ligand holes rather than Ni eg states. That occupation is controlled by the charge-transfer energy, and U is precisely the parameter that shifts that balance. The authors justify 2.0 eV by citing prior work and by noting that larger U gives a magnetic ground state inconsistent with experiment in the pristine phase. But the relevant quantity is not the pristine magnetic ordering — it is the energy of adding an electron to the d8L hybrid in the hydrogenated structure. They never test that. I don't think this sinks the paper, because the ligand-hole assignment is independently supported by prior XAS work and by their own electron-doped calculation, but it does mean the central causal chain has a parameter-dependence that is asserted, not demonstrated.\n\nTwo smaller issues. The electron-doped reference calculation adds eight electrons to a neutral supercell without mentioning a compensating background or finite-size corrections; for a qualitative occupation analysis this is probably minor, but it should be acknowledged. And the proton barriers and diffusion coefficients rely entirely on the CHGNet machine-learning potential with no DFT cross-check. The authors explicitly call that part qualitative, which is honest, but then the numbers in the abstract and figures (0.19 eV barrier, D values) read as more quantitative than they are.\n\nIf I were refereeing this, the main request would be: run the hydrogenated phase at a couple of other U values (and maybe one functional) and show that the occupation and the polaron picture survive. That is a few calculations, not a rewrite. The paper deserves peer review — the mechanism is plausible, the work is careful, and the proton-transport trends are useful. Send it out, with a referee asked to dig into the U sensitivity.","headline":"A coherent polaron-mediated MIT mechanism in hydrogenated nickelates, with genuinely new proton-transport trends, but the central charge-transfer assignment rests on a single Hubbard U and needs a sensitivity check.","tokens_in":17754,"tokens_out":2082,"would_cite":true,"duration_ms":19953,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.30.+h"],"model":"deepseek-v4-flash","headline":"The hydrogen-driven metal-insulator transition in nickelate perovskites is caused by electron and proton polarons that localize nickel's eg electrons into a Mott-insulating d8 state, not by simple electron filling of the nickel orbitals.","keywords":["hydrogen-doped nickelates","metal-insulator transition","polaron","Mott insulator","ligand hole","proton conduction","rare-earth nickelates","NdNiO3"],"falsifier":"Resonant inelastic X-ray scattering or X-ray absorption at the Ni L-edge and O K-edge on hydrogenated NdNiO3 could directly show whether the added electrons sit in oxygen ligand holes or in Ni eg states. Alternatively, repeating the first-principles calculation with the on-site Coulomb parameter varied from 1 to 4 eV would reveal whether the d8-ligand-hole occupation and polaron localization survive; if they disappear at moderate interaction strengths, the mechanism is an artifact of the parameter choice.","tokens_in":16799,"feed_emoji":"⚡","tokens_out":6259,"duration_ms":55835,"temperature":0.7,"pith_summary":"This paper uses first-principles electronic-structure calculations to explain how hydrogenation turns NdNiO3 from a metal into an insulator. It argues that electrons added by hydrogen do not directly reduce nickel but instead occupy oxygen ligand-hole states in the nickel–oxygen hybridized d8L configuration. That occupation stretches the Ni–O bonds and forms electron polarons; the inserted protons act as additional polarizing centers. Together these polarons weaken nickel–oxygen hybridization, drive the nickel eg electrons into a localized d8 (t2g^6 eg^2) configuration, and open a Mott gap of about 0.81 eV. The paper also examines proton conduction, showing that hydrogenation creates fast one-dimensional proton pathways along the c-axis while suppressing three-dimensional diffusion, and that smaller rare-earth ions (Sm vs Nd) promote hydrogen uptake but impede proton mobility.","feed_headline":"Hydrogen turns nickelates insulating by polarons, not electron filling","feed_subtitle":"Hydrogen's electrons land on oxygen, not nickel; the lattice distortion locks the material into a Mott insulator.","key_machinery":"The central object is the Ni–O hybridized d8L electronic configuration—a nickel d8 state carrying a hole on an oxygen ligand. The paper's mechanism is that hydrogen doping fills that ligand hole, and the resulting lattice distortion forms an electron polaron; the proton adds a second polaronic distortion. Polaron formation weakens Ni–O hybridization, which is the step that localizes the eg electrons into a d8 Mott state. For proton transport, the key distinction is between intraoctahedral transfer (proton hopping between oxygen sites within the same NiO6 octahedron) and interoctahedral transfer (proton moving through the A-site plane), whose computed energy barriers and percolation pathways","core_discovery":"In hydrogenated NdNiO3, the electrons introduced by hydrogen fill the O 2p ligand hole of the Ni–O antibonding d8L states rather than directly converting Ni3+ to Ni2+. Filling these antibonding states elongates the Ni–O bonds, forming electron polarons, and the protons themselves act as polarizing centers. Together these polarons weaken the Ni–O σ hybridization, converting the itinerant Ni eg electron into a localized t2g^6 eg^2 configuration. The half-filled eg shell then experiences strong on-site Coulomb repulsion, opening a Mott gap of about 0.81 eV. The same calculations map proton transport: in pristine NdNiO3 the intraoctahedral proton-transfer barrier is 0.35 eV and the interoctahedr","pith_inferences":["A similar polaron-mediated mechanism may operate in other strongly correlated oxides that spontaneously absorb hydrogen, such as cobaltites and vanadates, offering a screening criterion: materials with O 2p ligand-hole states in their metallic phase are prime candidates.","Because electron-polaron formation is driven by filling antibonding states, strain or electrostatic doping could be used instead of hydrogen to tune the transition, with hydrogen adding proton-polaron effects on top.","The suppression of three-dimensional proton diffusion under full hydrogenation suggests an intermediate hydrogen concentration may best balance polaron formation against proton mobility; measuring conductivity across H concentrations could reveal a peak.","The computed role of A-site radius implies that mixing Nd and Sm on the A-site could tune both hydrogen uptake and proton mobility, an experimentally accessible chemical substitution."],"forward_implications":["The hydrogen-induced insulating phase of NdNiO3 should be classified as a polaron-driven Mott insulator, not a simple band-filling or charge-transfer insulator.","Proton conductivity in rare-earth nickelates is set by two competing effects: smaller A-site cations increase hydrogen uptake but raise migration barriers, implying an optimal cation size for electrolyte performance.","Hydrogenation creates fast one-dimensional proton channels along the c-axis while blocking three-dimensional percolation, so the material may still work well at electrolyte–electrode interfaces that exploit that direction.","The calculated 0.81 eV gap in hydrogenated NdNiO3 provides a concrete target for optical and transport measurements to confirm the insulating state.","The finding that purely electron-doped NdNiO3 has a smaller gap (0.55 eV) than hydrogenated material supports a measurable role for the proton polaron beyond electron filling."],"fun_headline_variants":["Polarons, not electron filling, drive nickelate insulator switch","Hydrogen's electrons land on oxygen, making nickelate Mott insulator","Polarons lock nickelate into Mott insulator after hydrogen doping","Electron polarons, not nickel reduction, cause nickelate MIT","Hydrogen-induced polarons, not electron count, flip nickelates to insulators"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire polaron mechanism rests on the choice of the on-site Coulomb correction parameter (Ueff = 2.0 eV) and on freezing the rare-earth f electrons in the core; if the real nickel–oxygen energy balance differs, the added electrons might occupy nickel states instead of oxygen ligand holes, and the polaron-driven localization would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Polarons, not electron filling, drive nickelate insulator switch","Hydrogen's electrons land on oxygen, making nickelate Mott insulator","Polarons lock nickelate into Mott insulator after hydrogen doping","Electron polarons, not nickel reduction, cause nickelate MIT","Hydrogen-induced polarons, not electron count, flip nickelates to insulators"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000604,"raw_usage":{"total_tokens":2696,"prompt_tokens":828,"completion_tokens":1868,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":572,"completion_tokens_details":{"reasoning_tokens":1789}},"tokens_in":572,"tokens_out":1868,"duration_ms":12028,"temperature":1.0,"reasoning_tokens":1789,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T10:48:32.394749+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Resonant inelastic X-ray scattering or X-ray absorption at the Ni L-edge and O K-edge on hydrogenated NdNiO3 could directly show whether the added electrons sit in oxygen ligand holes or in Ni eg states. Alternatively, repeating the first-principles calculation with the on-site Coulomb parameter varied from 1 to 4 eV would reveal whether the d8-ligand-hole occupation and polaron localization survive; if they disappear at moderate interaction strengths, the mechanism is an artifact of the parameter choice.","supporting_citations":[],"review_version":1}