{"id":"85484e1d-901b-4d53-8872-1a0caa17224a","arxiv_id":"2607.29014","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Across rare-earth-substituted cobalt perovskites, oxygen capacity falls from La to Lu despite easier vacancy formation, due to O−-like species and rare-earth redox; ICOHP bond strength is proposed as the descriptor in this regime.","lead":"This paper measures and simulates how much oxygen a family of cobalt-oxide crystals can store, and finds the usual predictor—how easy it is to create an oxygen vacancy—fails across the rare-earth series. The authors identify a second charge-compensation mechanism involving oxidized lattice oxygen and reduced rare-earth cations, and propose metal–oxygen bond strength as a better descriptor when that mechanism dominates.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central mechanism hinges on Bader-charge-derived reduction of heavy rare-earth cations and O− formation; Bader charges are not oxidation states and are unreliable for Ln 4f/O holes. Without direct spectroscopic or alternative charge-partition evidence, the competing-pathway explanation, and henc","rationale":"The reader's weakest assumption is precisely that the mechanistic attribution depends on indirect Bader-charge, pDOS, and O1s evidence rather than a direct quantitative probe. I agree: the single most load-bearing step in the paper is the assignment of O− formation and Ln reduction based on Bader charge differences. If this assignment is wrong, the central claim that oxygen capacity is governed by competing charge-compensation mechanisms loses its mechanistic foundation, and the ICOHP descriptor, while possibly correlating with the experimental trend, would not be validated as a physically meaningful descriptor for anionic-redox-dominated compositions. The reader's verdict of CONDITIONAL is appropriate: the paper presents a substantial data set and a plausible interpretation, but the load-bearing mechanistic inference needs corroboration. I do not think the concern warrants rejection, because the experimental TGA trend is solid and the computational trend in EV,O is reported; however, the mechanism as proposed should be treated as provisional until direct spectroscopic evidence or a robust alternative charge-partitioning analysis confirms the Ln-reduction/O− picture. The composition mismatch and lack of out-of-sample validation are additional reasons for the conditional status, but the Bader-charge evidence is the most critical because it underpins the proposed mechanism itself.","tokens_in":11423,"tokens_out":8652,"duration_ms":99971,"concrete_test":"Perform Ln L3-edge HERFD-XANES and O K-edge XAS on pristine and TGA-reduced La0.2Sr0.8CoO3 and Lu0.2Sr0.8CoO3 (plus one intermediate, e.g., Tb). If the Lu L3 edge shows no valence/covalency change after reduction and the O K-edge pre-peak (O 2p holes) does not grow from La to Lu, the proposed Ln-reduction/O−-stabilization mechanism is falsified. As a computational cross-check, recompute atomic charges with DDEC6 (or with Ln 4f in valence and a Hubbard U) on the same relaxed structures; if the Ln charge changes vanish or reverse, the Bader-based assignment is an artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's core mechanistic claim—that in heavy-rare-earth compositions charge compensation shifts from Co reduction to Ln reduction plus formation of stabilized O−-like species, which suppresses O2 release—rests on comparing Bader charges of pristine and single-vacancy supercells (Fig. 3d–f). This is the load-bearing step: if the Bader-based partitioning is not a faithful measure of oxidation-state change, the 'competing charge-compensation pathways' are not established, and ICOHP becomes an empirically fitted correlate rather than a mechanistic descriptor. Bader charges are known to be basis-set/pseudopotential-sensitive, are not oxidation states, and are especially problematic for localized Ln 4f states and for oxygen holes (where the charge density change is delocalized). The claim that 'heavy rare-earth cations are reduced' is chemically surprising for Lu3+ and most Tb3+–Lu3+ ions, which typically have stable trivalent configurations; the observed Bader shifts could reflect electrostatic relaxation or the choice of reference compounds rather than true redox. The composition mismatch (DFT Ln0.25Sr0.75 vs experimental Ln0.2Sr0.8) also leaves open the possibility that the calculated mechanism is not representative of the measured series.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines TGA, XPS, and first-principles DFT to establish composition–oxygen capacity relationships in Ln_xSr_{1−x}CoO_3 (Ln = La–Lu). It shows that increasing Sr content expands the Co^{4+} cationic redox reservoir and that in this regime oxygen-vacancy formation energy (EV,O) correlates with measured oxygen capacity. For the rare-earth series Ln_0.2Sr_0.8CoO_3, the authors report that oxygen capacity decreases from La to Lu even though calculated EV,O decreases, and they attribute this to a competing anionic charge-compensation pathway involving O−-like lattice-oxygen species and reduction of heavy rare-earth cations. They propose average metal–oxygen bond strength, quantified by ICOHP, as the descriptor for oxygen capacity when anionic redox dominates.","tokens_in":11726,"tokens_out":4043,"duration_ms":45616,"significance":"If the mechanistic attribution is correct, the paper makes a useful contribution by showing that no single thermodynamic descriptor governs oxygen capacity in perovskites and by proposing ICOHP as a physically motivated descriptor for anionic-redox regimes. The experimental dataset is broad (fourteen rare-earth compositions) and the observation that EV,O fails to describe the rare-earth trend is clearly documented. The paper also gives explicit credit to the limitations of EV,O as a universal descriptor, which is valuable. However, the central mechanistic claim—heavy-rare-earth compositions stabilize O−-like species and undergo Ln reduction—currently rests on Bader charge analysis and projected-density-of-states features rather than a direct quantitative probe of oxygen or lanthanide oxidation states. This weakens the link between mechanism and descriptor and is the main barrier to acceptance.","major_comments":[{"comment":"DFT calculations are performed on Ln_0.25Sr_0.75CoO_3 (stated in the Fig. 3 caption), while the experiments and the central claim concern Ln_0.2Sr_0.8CoO_3 (Fig. 2). Fig. 3a plots these two sets on the same axes with no acknowledgment of the composition mismatch. If the vacancy formation energies or Bader charge responses are sensitive to the Ln/Sr ratio, the calculated inverse trend is not directly established for the measured compositions. Please either compute at the experimental composition, or demonstrate quantitatively that the 0.25/0.75 and 0.2/0.8 A-site configurations give the same mechanism and energetics.","section":"§Competing cationic and anionic charge-compensation pathways; Fig. 3a"},{"comment":"The load-bearing mechanistic assignment—that heavy rare-earth compositions form O−-like species and reduce Ln cations—is inferred from comparing Bader charges with reference compounds, but no reference compounds, thresholds, or uncertainties are given. Bader charges are not oxidation states and are especially unreliable for localized Ln 4f states and delocalized O 2p holes. The pDOS hole features and O 1s binding-energy shifts are consistent with O− formation but are not quantitative or species-specific. Since the entire 'competing charge-compensation pathway' explanation depends on this assignment, please provide more direct evidence (e.g., XAS/RIXS, DFT+U with occupation-matrix control, or a systematic benchmark of the Bader-based oxidation-state assignment against reference systems).","section":"§Competing cationic and anionic charge-compensation pathways; Fig. 3d–f"},{"comment":"ICOHP is selected as the descriptor after the oxygen-capacity trend has been observed on the same dataset, so its predictive power is not independently tested. The text calls ICOHP 'physically meaningful' and the framework 'predictive,' but the paper contains no out-of-sample validation or quantitative correlation (e.g., leave-one-Ln-out or prediction of an unmeasured composition). Without such a test, ICOHP is a post-hoc correlate, not a demonstrated descriptor. Please add a genuine predictive test or temper the predictive claim accordingly.","section":"§Metal-oxygen bonding provides a descriptor; Fig. 4"}],"minor_comments":[{"comment":"The notation for the oxygen-hole species is inconsistent: 'O−', 'O1−', 'O⁻-like', and 'O−-like' are used interchangeably. Please standardize to a single notation, e.g., 'O^{−}-like (oxygen hole)'.","section":"Throughout"},{"comment":"The 'fractions of partially oxidized O-anions' are mentioned but the definition and threshold for 'partially oxidized' are not given in the main text. Specify the Bader-charge criterion and how the fraction is computed, or move this definition to a clearly referenced Supplementary Note.","section":"Fig. 3f caption and §Competing pathways"},{"comment":"The sentence after 'systematic decrease in the Co^{4+} fraction' appears to be truncated or misformatted: it reads ' (O2-) binding energy (Fig. 1f).' This needs repair. Also, the Co 2p spectra are said to be 'nearly unchanged' across the rare-earth series, but Fig. 3d shows Co Bader-charge changes; please reconcile these two statements in the text.","section":"§Rare-earth substitutions; Fig. 2d"}],"recommendation":"major_revision","confidential_remarks":"The experimental dataset and the clear demonstration that EV,O fails for the rare-earth series are valuable and likely publishable. The main concern is the Bader-charge-based mechanistic attribution and the composition mismatch between DFT and experiment. Both are fixable with additional calculations and/or a more careful framing. I would not reject, but the central claim as currently stated is not fully supported."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper: the experimental dataset is genuinely useful, and the central mechanistic claim is shakier than the authors admit.\n\nThe new piece is the fourteen-composition Ln0.2Sr0.8CoO3 series. The TGA data showing oxygen capacity dropping from La to Lu while vacancy formation energies decrease is a clean, important counterexample to the field's default descriptor. I believe the experiments: the trend is monotonic, the structural characterization backs the phase purity, and the XPS O 1s shift toward higher binding energy is consistent with more electron-deficient lattice oxygen in heavy lanthanides. The paper also does a good job showing that EV,O alone cannot explain oxygen capacity — that part is solid.\n\nWhat gives me pause is the mechanism they attach to the trend. The whole 'competing charge-compensation' story rests on Bader-charge differences between pristine and reduced supercells. As the stress-test notes, Bader charges are not oxidation states, and they are especially unreliable for Ln 4f and O 2p holes. The claimed reduction of heavy rare-earth cations is chemically surprising for Tb–Lu, which usually stay trivalent, and the observed Bader shifts could just as easily be electrostatic relaxation or reference-charge artifacts. The O− species are inferred from pDOS holes and Bader counts, not from a direct probe like XAS or EELS at the oxygen edge. So the mechanism is plausible, but it is not established on the current evidence.\n\nThere are two additional soft spots. The DFT supercells are Ln0.25Sr0.75CoO3 while the experiments are Ln0.2Sr0.8CoO3; Figure 3a overlays both without acknowledging the mismatch. If the mechanism is composition-sensitive, this matters. And the ICOHP descriptor is identified retrospectively on the same data that it is supposed to predict. No holdout, no candidate screening. That makes the 'predictive design' claim overstated for now.\n\nI still think this deserves a serious referee. The dataset is real, the counterintuitive trend is important, and the authors are honest about the EV,O failure. The referee should push for direct evidence of oxygen holes (RIXS, O K-edge XAS) and validation of ICOHP on at least one composition outside the fitted series. If they can supply that, the paper becomes a strong contribution. If not, the experimental half remains a solid achievement but the explanation remains speculative.\n\nRecommendation: send it to peer review with a request for major revisions on the mechanistic evidence and descriptor validation.","headline":"Solid experimental dataset showing a true counterexample to EV,O, but the O-/Ln-reduction mechanism rests on Bader charges and the ICOHP descriptor is post-hoc — deserves review, not acceptance as is.","tokens_in":12262,"tokens_out":2912,"would_cite":true,"duration_ms":28649,"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":"Oxygen capacity in perovskite oxides is governed by competing charge-compensation mechanisms, so the predictive descriptor depends on whether cationic or anionic redox dominates.","keywords":["oxygen capacity","perovskite oxides","charge compensation","anionic redox","oxygen vacancy formation energy","ICOHP","rare-earth substitution","chemical looping"],"falsifier":"Measure O K-edge X-ray absorption and Ln M-edge or Ln 3d XPS on reduced heavy-rare-earth samples such as Lu0.2Sr0.8CoO3−δ: if no signature of oxidized oxygen (hole states) or reduced rare-earth ions appears, the proposed anionic-redox mechanism is wrong. Alternatively, compute the equilibrium oxygen capacity for a heavy-rare-earth composition using grand-canonical DFT at the operating temperature and compare directly with TGA; if the measured capacity is reproduced without invoking O− stabilization, the mechanism is unnecessary.","tokens_in":11273,"feed_emoji":"🧪","tokens_out":3686,"duration_ms":36639,"temperature":0.7,"pith_summary":"The paper tries to establish that oxygen capacity in perovskite oxides is not described by a single universal descriptor such as oxygen-vacancy formation energy. Instead, it depends on which charge-compensation pathway is active: when cobalt cations absorb the electrons left by oxygen removal, vacancy formation energy predicts capacity; when lattice oxygen is oxidized to O−-like species and rare-earth cations are reduced, the average metal–oxygen bond strength (ICOHP) is the controlling descriptor. Using the Ln0.2Sr0.8CoO3 series, the paper shows that oxygen capacity decreases from La to Lu even though vacancy formation becomes easier, because heavy rare earths stabilize oxidized lattice oxygen and suppress net O2 release. This matters because it offers a mechanism-dependent framework for designing oxygen carriers for chemical looping and oxygen storage.","feed_headline":"Smaller lanthanides trap oxygen despite easier vacancy formation","feed_subtitle":"Heavy rare-earth perovskites stabilize oxidized oxygen, so metal–oxygen bond strength—not vacancy energy—predicts capacity.","key_machinery":"The central objects are the two competing charge-compensation pathways—cationic redox (Co4+/Co3+/Co2+) and anionic redox (O2− → O−, i.e., oxygen-hole formation)—and the two descriptors tied to them: oxygen-vacancy formation energy (EV,O) and integrated crystal orbital Hamilton population (ICOHP), a bond-strength metric that sums the energy-weighted bonding interactions between each oxygen and its neighboring metal cations. ICOHP does the explanatory work in the anionic-redox regime by quantifying how strongly residual lattice oxygen is stabilized after vacancy formation; stronger bonding (more negative ICOHP) traps the oxygen as O−-like species and prevents its release as O2. Bader charge an","core_discovery":"The central claim is that oxygen capacity is governed by competing cationic and anionic charge-compensation pathways, and the correct predictive descriptor depends on which pathway dominates. In LaxSr1−xCoO3, increasing Sr content raises the Co4+ concentration and expands the cationic redox reservoir; here oxygen-vacancy formation energy (EV,O) captures the measured capacity trend. Across the isovalent rare-earth series Ln0.2Sr0.8CoO3, however, EV,O decreases from La to Lu while measured oxygen capacity decreases—the opposite of what vacancy thermodynamics predicts. The paper attributes this to an alternative pathway in which lattice oxygen is partially oxidized to O−-like species during oxy","pith_inferences":["The same competition between cationic and anionic redox may explain oxygen-capacity trends in other oxide families (e.g., Fe- or Mn-based perovskites) whenever heavy lanthanides or highly covalent B-site metals promote oxygen-hole formation; this is a testable generalization the paper hints at but does not demonstrate.","If O− stabilization is the suppressing factor, then substituting a redox-active B-site cation that competes for electrons (or disrupting the Co2Ln2Sr2 and Co2Sr4 coordination motifs) could recover oxygen capacity in heavy-rare-earth perovskites—an experiment the paper does not run.","ICOHP could be paired with high-throughput screening to rank oxygen carriers without computing full vacancy thermodynamics, but the paper does not validate this scaling.","The DFT supercells used for the rare-earth series are Ln0.25Sr0.75CoO3, while the experiments use Ln0.2Sr0.8CoO3; if the mechanism is sensitive to the exact Sr content, the predicted descriptor trend might shift quantitatively, though not necessarily qualitatively."],"forward_implications":["For compositions where cationic redox dominates, oxygen-vacancy formation energy remains a valid and simple descriptor for oxygen capacity.","For compositions where anionic redox dominates, ICOHP—the average metal–oxygen bond strength—can serve as a descriptor, replacing the ineffective vacancy-formation-energy metric.","High oxygen capacity requires balancing three factors: a large cationic redox reservoir, metal–oxygen bonds labile enough to permit O2 evolution yet strong enough to keep the structure intact, and avoidance of excessive stabilization of O−-like species.","In the Ln0.2Sr0.8CoO3 family, La0.2Sr0.8CoO3 delivers the highest measured oxygen capacity (0.179 mol O2 per mole of oxide, about 72% of the theoretical yield for reduction to the O2.5 stoichiometry).","The mechanism-dependent descriptor framework is proposed as a general strategy for designing oxygen-exchange oxides beyond the specific perovskite series studied."],"fun_headline_variants":["Mechanism dictates descriptor for perovskite oxygen capacity","Heavy rare-earths store less oxygen despite easier vacancies","Bond strength, not vacancy energy, drives oxygen capacity in heavy lanthanides","Anionic redox flips the oxygen-storage rulebook in cobaltites","Vacancy energy fails for heavy rare-earth perovskites"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The claim that heavy rare-earth compositions suppress oxygen release by stabilizing O−-like oxygen and reducing rare-earth cations rests on Bader-charge assignments and binding-energy shifts rather than direct measurements of oxygen or rare-earth oxidation states, and on DFT supercells at Ln0.25Sr0.75CoO3 standing in for the experimentally measured Ln0.2Sr0.8CoO3.","fun_headline_variants_meta":{"raw":{"variants":["Mechanism dictates descriptor for perovskite oxygen capacity","Heavy rare-earths store less oxygen despite easier vacancies","Bond strength, not vacancy energy, drives oxygen capacity in heavy lanthanides","Anionic redox flips the oxygen-storage rulebook in cobaltites","Vacancy energy fails for heavy rare-earth perovskites"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000794,"raw_usage":{"total_tokens":3362,"prompt_tokens":800,"completion_tokens":2562,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":544,"completion_tokens_details":{"reasoning_tokens":2489}},"tokens_in":544,"tokens_out":2562,"duration_ms":19257,"temperature":1.0,"reasoning_tokens":2489,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T15:18:16.418901+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure O K-edge X-ray absorption and Ln M-edge or Ln 3d XPS on reduced heavy-rare-earth samples such as Lu0.2Sr0.8CoO3−δ: if no signature of oxidized oxygen (hole states) or reduced rare-earth ions appears, the proposed anionic-redox mechanism is wrong. Alternatively, compute the equilibrium oxygen capacity for a heavy-rare-earth composition using grand-canonical DFT at the operating temperature and compare directly with TGA; if the measured capacity is reproduced without invoking O− stabilization, the mechanism is unnecessary.","supporting_citations":[],"review_version":1}