{"id":"ec2e740e-2be3-46bf-b7c2-dd03b215082a","arxiv_id":"2504.16372","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In compressively strained superconducting La2PrNi2O7 films, the flat 3dz2 band is observed about 70 meV below the Fermi level, contradicting expectations that it crosses the Fermi level, and the cuprate-like band shows unexpected three-dimensional kz dispersion.","lead":"Researchers measured the electronic bands of a strained, superconducting nickelate film with photoemission and found that the key flat band sits about 70 meV below the Fermi energy, not at it. This challenges the idea that this band directly drives the superconductivity and points to other mechanisms such as phonons or spin fluctuations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed factor-of-5-to-10 discrepancy rests on comparing strained La2PrNi2O7 films with bulk La3Ni2O7; without a Pr-matched unstrained baseline, Pr substitution effects could account for part of the apparent strain shift.","rationale":"The empirical ARPES finding that the g band lies roughly 70 meV below the Fermi level is well supported within the paper: it appears across MBE and PLD films, in as-grown and ozone-treated states, at multiple photon energies, and in multiple thicknesses. The histograms and direct side-by-side comparison give a credible internal cross-check. The soft spot is not the measurement but the attribution of the measured difference to strain alone. The paper's strongest quantitative statement, that the g band shifts a factor of 5 to 10 less than DFT predicts, depends on comparing La2PrNi2O7 films to bulk La3Ni2O7 data and to DFT calculations for La3Ni2O7. Pr substitution is a controlled variable that is left uncontrolled in both the experiment and the theory. This does not invalidate the observation that the g band stays below EF in superconducting films, but it does weaken the paper's conclusion that a large class of DFT-based flat-band-crossing scenarios is ruled out by this measurement. The reader's CONDITIONAL verdict already captures this type of concern; my emphasis is on the missing Pr-matched baseline rather than on the Hubbard U choice or oxygen stoichiometry, which are secondary. A single Pr-substituted DFT calculation or an unstrained LPNO reference measurement would settle the issue, so no verdict change is warranted relative to the reader's conditional assessment.","tokens_in":14698,"tokens_out":4773,"duration_ms":52682,"concrete_test":"Compute DFT+U band structures for La2PrNi2O7 and La3Ni2O7 under identical unstrained and 2%-compressed/1%-c-axis-expanded structures, extracting the g-band top energy in each case. If the strain-induced shift in La2PrNi2O7 remains close to the 210 meV found for La3Ni2O7, the Pr-substitution concern is refuted. If the shift drops below roughly 50 meV or changes sign, the paper's factor-of-5-to-10 discrepancy must be re-evaluated. An even more direct check is ARPES on an unstrained La2PrNi2O7 reference (bulk crystal or a film on a lattice-matched substrate) using the same end station; if its g-band top is already near -70 meV, then most of the apparent shift is not strain-driven.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is the strain-driven shift of the g-band top: bulk La3Ni2O7 sits near -50 meV (ref. 7), while the strained La2PrNi2O7 films sit at -54/-91/-71 meV, and DFT+U predicts a 210-350 meV downward shift under compressive strain. This comparison varies two parameters at once: strain and chemical composition. The experimental baseline is La3Ni2O7 bulk, not La2PrNi2O7, and the DFT calculations are also for La3Ni2O7 with rare-earth f-electrons treated as core. Pr substitution changes the ionic radius and can modify the local structure, octahedral distortions, and the dz2-derived g-band energy independently of epitaxial strain. If Pr substitution alone lowers the g band by several tens of meV, the true strain-induced shift could be much smaller, and the factor-of-5-to-10 discrepancy would be largely an artifact of comparing unlike compounds. The paper itself notes that its Pr content differs from a previous film report, but it does not provide a strain-free La2PrNi2O7 reference, nor a Pr-substituted DFT calculation, to separate the two effects. This concern is specific, load-bearing, and testable.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports an in-situ synchrotron ARPES study of compressively strained La2PrNi2O7 thin films grown by oxide MBE, with supplementary PLD-grown films, both as-grown and after ozone treatment (superconducting, onset Tc ~40 K). The central finding is that the strongly renormalized flat 3dz2-derived g band remains approximately 70 meV below the Fermi level in the strained films, rather than shifting downward by the 210–350 meV predicted by DFT+U calculations for the same nominal strain. The authors also report a photon-energy-dependent kz dispersion of the 3dx2-y2-derived b band, which they attribute to a low-symmetry Ni sublattice and a finite inter-bilayer dx2-y2 hopping, and they note that the ~70 meV g-band energy coincides with oxygen phonon and spin-excitation energies. On this basis they argue that superconductivity in these films is not driven by a Fermi-level-crossing flat band.","tokens_in":15037,"tokens_out":5779,"duration_ms":56707,"significance":"If the central observation holds, the paper provides a strong experimental constraint on theories of superconductivity in strained bilayer nickelates: it rules out the common scenario in which the flat 3dz2 band crosses the Fermi level and supplies the large density of states that drives pairing. The robustness of the observation is a clear strength—the g band below EF is reproduced across MBE and PLD samples, as-grown and ozone-treated states, and multiple photon energies, and the paper carefully controls matrix-element effects that had obscured the band in prior work. The kz dispersion of the dx2-y2 band, if confirmed, is also an important input for structural and electronic models. The main weakness is that the quantitative claim of a factor-of-5-to-10 discrepancy relies on comparing strained La2PrNi2O7 films to bulk La3Ni2O7 data and to DFT calculations for La3Ni2O7, without separating strain effects from Pr-substitution effects. The paper is well positioned to influence the field, but this compositional confound must be addressed.","major_comments":[{"comment":"The quantitative claim that the g band shifts by only ~20 meV instead of the predicted 210–350 meV rests on comparing strained La2PrNi2O7 thin films with bulk La3Ni2O7 ARPES data (ref. 7) and with DFT calculations performed for La3Ni2O7. This varies strain and chemical composition simultaneously. Pr substitution alone can alter the Ni-O bond lengths, octahedral tilts, and the dz2-derived g-band energy independently of epitaxial strain. The paper itself notes in §3 that its Pr content differs from a previous film report, yet it provides no unstrained or weakly strained La2PrNi2O7 reference, nor a Pr-substituted DFT calculation. Without separating the composition effect from the strain effect, the reported factor-of-5-to-10 discrepancy could be substantially overestimated. I recommend adding an ARPES measurement of an unstrained or weakly strained LPNO film (or of a La2PrNi2O7 bulk crystal) and DFT calculations that include Pr, so that the strain-driven shift is isolated from the substitution effect.","section":"§2, Fig. 2h, Table I"},{"comment":"The DFT benchmark uses U = 3.5 eV and treats rare-earth f-electrons as core states. The size of the strain-induced g-band shift is likely sensitive to these choices. Fig. S1b shows a calculation with U = 1 eV, but the text does not report the g-band position for this U, nor does it provide a systematic U sweep to demonstrate that the 210 meV shift is robust. If the predicted shift decreases at lower U or with explicit Pr 4f treatment, the claim of 'fundamental missing pieces in DFT' would be less well supported. A simple U sweep (e.g., U = 0–5 eV) for the strained and unstrained structures, and ideally a Pr-substituted supercell, would make the disagreement quantitative and less model-dependent.","section":"§2, Table I, Supplementary Information (DFT+U section)"}],"minor_comments":[{"comment":"The Gaussian fits for the as-grown and ozone-treated g-band distributions are based on a small number of samples, and the as-grown distribution appears broad. The paper should state the number of samples in each histogram bin and provide error bars on the fitted centers, since the inferred ~20 meV shift is comparable to the spread between samples.","section":"§2, Fig. 2h"},{"comment":"The statement that 'the only reason that ARPES data shows kz-dispersion must be from the fact that the single Ni layer takes a lower symmetry case' is an inference by elimination, not a direct structural determination. The SI later acknowledges that the structure is inferred, but the main text should more carefully phrase this as a suggestion consistent with the data and DFT, rather than a definite conclusion.","section":"Supplementary Information, 'Lattice structure indicated from kz-dispersion'"},{"comment":"The tight-binding model is fitted to the measured dispersions, so the resulting kz dispersion is not an independent prediction. The SI notes this caveat, but the main text would benefit from stating explicitly that the model is a parameterization of the observed bands, not a verification of the low-symmetry sublattice scenario.","section":"Supplementary Information, 'Construction of a simple tight-binding model'"},{"comment":"The column headers of Table I are difficult to parse: the row for the c-lattice constant lists three values, and it is unclear which columns correspond to 'Relaxed bulk', 'LPNO thin film', and the two DFT columns. Please reformat the table so that each column is unambiguously labeled.","section":"Table I"}],"recommendation":"major_revision","confidential_remarks":"The experimental work is of high quality and the central observation is likely to be influential. The main risk is the compositional confound between strain and Pr substitution in the quantitative comparison; this is a load-bearing issue that the authors can address by providing an appropriate baseline or by softening the quantitative claims. I would also encourage the editor to have a theorist with expertise in DFT+U for nickelates assess the sensitivity of the strain shift to U, as the current supplement does not adequately explore this."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nMy read: this is the best ARPES so far on compressively strained La2PrNi2O7 films, and the central empirical finding is solid—the flat 3dz2 (g) band sits roughly 70 meV below the Fermi level in both as-grown and ozone-treated superconducting films. That contradicts the earlier report from Li et al. (2501.09255) claiming a Fermi-level crossing. The cross-checks are genuinely good: MBE and PLD films, as-grown and ozone-treated states, several photon energies, and two thicknesses. The matrix-element explanation for the earlier report's apparent crossing is plausible and documented.\n\nWhat's new: the g-band position in a superconducting film, and a photon-energy-dependent kz dispersion of the dx2-y2 (b) band. The kz observation is interesting but the inference that it requires a low-symmetry doubled Ni sublattice is indirect—it rests on a tight-binding model that is fitted, not predictive.\n\nSoft spots. The headline quantitative claim—that the g band shifts only 1/5 to 1/10 as much as DFT predicts—compares strained La2PrNi2O7 films to unsubstituted bulk La3Ni2O7. That varies strain and Pr content at once. The paper offers no Pr-matched unstrained baseline and no Pr-substituted DFT. The stress-test concern lands: part of the apparent strain shift could be a Pr-substitution effect. Looking at Table I, the as-grown g band is at -54 meV, essentially unchanged from bulk La3Ni2O7's -50 meV; the ozone-treated films sit at -71 and -91 meV. So the experimental \"strain shift\" is small and entangled with oxygen content. The factor-of-5-to-10 discrepancy is therefore not as clean as stated. The paper flags that its Pr content differs from the earlier film report and notes oxygen-content spread, but it does not disentangle these.\n\nThe central conclusion—the g band does not cross the Fermi level in these films—holds up, because it does not depend on the bulk comparison. That is load-bearing and rules out the simple flat-band-crossing scenario for superconductivity here. The kz and effective-pressure claims are more speculative but labeled as such.\n\nMinor: band positions have no error bars, and the histograms are small-n.\n\nBottom line: this deserves a serious referee. The g-band result is likely correct and important for the nickelate subfield. The comparison to theory needs tightening—ideally a Pr-substituted baseline and error bars—but that's revision, not rejection. I'd bring it to reading group and would cite the g-band result.","headline":"Solid ARPES benchmark showing the g band sits ~70 meV below EF in strained superconducting LPNO films, but the claimed factor-of-5-to-10 discrepancy with DFT rests on comparing to unsubstituted bulk La3Ni2O7.","tokens_in":15637,"tokens_out":3530,"would_cite":true,"duration_ms":31446,"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":"In compressively strained superconducting La2PrNi2O7 films, the flat 3dz2 band stays about 70 meV below the Fermi level, not crossing it as density functional theory predicts, so the high density of states of that band cannot be what…","keywords":["bilayer nickelate superconductors","compressive epitaxial strain","ARPES band structure","flat 3dz2 band","Fermi level","kz dispersion","thin films","collective mode coupling"],"falsifier":"Measure the gamma-band top in the same strained film system across a continuous strain series, or in Pr-free La3Ni2O7 films at identical strain, using photon energies and matrix elements that clearly resolve the band: if the band top reaches or crosses the Fermi level in a superconducting film, or if the measured strain shift approaches the 210 meV DFT value, the paper's central claim would be refuted.","tokens_in":2020,"feed_emoji":"⚛️","tokens_out":2071,"duration_ms":75782,"temperature":0.7,"pith_summary":"This paper uses angle-resolved photoemission spectroscopy on compressively strained La2PrNi2O7 films, both as-grown and ozone-treated with onset Tc around 40 K, to test how strain reshapes the electronic structure of bilayer nickelates. It finds that the flat 3dz2 (gamma) band does shift downward under strain, as density functional theory predicts, but by only about one-fifth to one-tenth of the predicted amount: roughly 20 meV instead of 210 meV, leaving the band about 70 meV below the Fermi level in superconducting samples. The authors argue this contradicts the widely assumed mechanism in which superconductivity is driven by the high density of states of this flat band at the Fermi level. They also resolve a clear kz dispersion of the 3dx2-y2-derived band, indicating nontrivial inter-bilayer hopping that two-dimensional models miss. A reader should care because the result narrows which electronic ingredients can explain ambient-pressure superconductivity in strained bilayer nickelates.","feed_headline":"Flat band stays ~70 meV below Fermi level in strained nickelates","feed_subtitle":"ARPES shows the 3dz2 band barely shifts under strain, undercutting a leading explanation for 40 K superconductivity.","key_machinery":"The load-bearing object is the flat, strongly renormalized 3dz2 bonding band, called the gamma band, whose band top is measured by in situ synchrotron ARPES at several photon energies in both MBE- and PLD-grown films. Its position relative to the Fermi level is what distinguishes this paper's picture from earlier ones: the measured band top stays about 70 meV below EF, whereas an earlier report extrapolated it to a Fermi-level crossing. The quantitative argument is carried by comparing the measured strain-induced band shifts with DFT+U predictions computed at fixed experimental lattice constants, and the kz dispersion is captured by a 16-orbital tight-binding model that adds a small inter-bilayer 3dx2-y2 hopping term.","core_discovery":"The central discovery is that in compressively strained La2PrNi2O7 thin films, including ozone-treated films with onset Tc near 40 K, the strongly renormalized flat 3dz2 (gamma) band sits about 70 meV below the Fermi level rather than crossing it. DFT+U calculations using the measured 2% in-plane compression and 1% c-axis expansion predict this band should move down by roughly 210 meV, so the measured downward shift is a factor of 5-10 smaller. The paper therefore concludes that superconductivity in these films is not explained by the flat band's high density of states at the Fermi level, and instead notes that the 70 meV position coincides with the energy of collective modes, namely oxygen-related phonons and the peak of the spin excitation spectrum. As a separate finding, the 3dx2-y2-derived band shows a clear kz dispersion, which the paper attributes to inter-bilayer hopping enabled by buckled Ni-O planes rather than to strain relaxation or simple two-dimensional band structure.","pith_inferences":["Editorial inference beyond the paper: if the gamma band never crosses the Fermi level under any accessible strain, then the gamma band may contribute to superconductivity indirectly, for example through interlayer scattering or by mediating coupling to collective modes, rather than by supplying a large density of states; a testable consequence is that Tc would not track the gamma-band density of s","Editorial inference beyond the paper: the kz dispersion of the 3dx2-y2 band suggests that the c-axis coherence of the electronic structure is stronger than previously assumed, so photon-energy-dependent ARPES at more kz points could map the full three-dimensional Fermi surface and test whether inter-bilayer hybridization affects the superconducting gap anisotropy.","Editorial inference beyond the paper: the coincidence between the gamma-band energy and the 70 meV mode could mean the flat band acts as a sensitive probe of collective-mode coupling; a future experiment tracking the strength of the 70 meV renormalization kink across Tc could discriminate between phonon-mediated and spin-fluctuation-mediated pairing."],"forward_implications":["A Fermi-level crossing of the flat 3dz2 band is not required for 40 K superconductivity in strained bilayer nickelates, so theories built on the gamma band's density of states at EF need to be revised.","The gamma band's position near 70 meV, matching the energy of oxygen phonons and the spin-excitation peak, suggests collective-mode coupling may be more relevant to pairing than the flat band's density of states.","The effective pressure in these films is about 5 GPa, considerably larger than the naive expectation from the DFT-relaxed structure, which changes how strain experiments are compared with hydrostatic-pressure experiments on bulk crystals.","The observed kz dispersion of the 3dx2-y2 band implies that three-dimensional inter-bilayer hopping and buckled Ni-O planes must be included in realistic models of the superconducting state.","The measured band positions and renormalization factors provide a quantitative benchmark for future correlated-electron calculations of strained bilayer nickelates."],"supporting_citations":[{"why":"Provides the bulk La3Ni2O7 ARPES band positions and renormalization factors used as the reference for comparing thin-film band shifts.","marker":"[7]"},{"why":"Reports ambient-pressure superconductivity in thin film La3Ni2O7, establishing the sample regime this paper studies.","marker":"[18]"},{"why":"Reports ambient-pressure superconductivity onset above 40 K in (La,Pr)3Ni2O7 films, setting the Tc context for the ozone-treated samples.","marker":"[19]"},{"why":"Describes the PLD-grown compressively strained La2PrNi2O7 films and their transport, supplying the PLD sample synthesis used here.","marker":"[20]"},{"why":"Supplies the DFT+U calculations predicting the strain-driven gamma-band shift and effective pressure that the paper tests quantitatively.","marker":"[21]"},{"why":"Provides an independent theoretical prediction of strain-induced electronic structure changes in Ruddlesden-Popper nickelates.","marker":"[22]"},{"why":"Offers another DFT-based prediction of Fermi surface reconstruction under compressive strain in La3Ni2O7.","marker":"[23]"},{"why":"Is the earlier ARPES report claiming a gamma-band Fermi-level crossing in similar films, which this paper directly re-examines and contradicts.","marker":"[24]"},{"why":"Reports an anomalous energy gap and collective-mode coupling in superconducting La2.85Pr0.15Ni2O7 films, providing the comparison for the 70 meV mode feature.","marker":"[25]"},{"why":"Identifies buckled Ni-O planes in strain-engineered La3Ni2O7 films, the structural mechanism the paper invokes to explain the observed kz dispersion.","marker":"[32]"}],"fun_headline_variants":["Flat band hovers 70 meV below Fermi in strained nickelates","Nickelate flat band misses Fermi level under strain","Strained nickelates: flat band 70 meV off, DFT off by 10x","Nickelate superconductivity not linked to flat band Fermi crossing","ARPES: strained nickelate flat band stays 70 meV below Fermi"],"cache_read_input_tokens":17664,"weakest_assumption_plain":"The load-bearing premise is that DFT+U with U = 3.5 eV, using the measured 2% in-plane compression and 1% c-axis expansion, is a reliable baseline for how much the gamma band should shift; if the effective Hubbard U, oxygen stoichiometry, or actual strain state differs, the claimed factor-of-5-to-10 discrepancy could shrink, and the comparison to bulk La3Ni2O7 could also be affected by Pr substitution rather than strain alone.","fun_headline_variants_meta":{"raw":{"variants":["Flat band hovers 70 meV below Fermi in strained nickelates","Nickelate flat band misses Fermi level under strain","Strained nickelates: flat band 70 meV off, DFT off by 10x","Nickelate superconductivity not linked to flat band Fermi crossing","ARPES: strained nickelate flat band stays 70 meV below Fermi"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000314,"raw_usage":{"total_tokens":1848,"prompt_tokens":1077,"completion_tokens":771,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":693,"completion_tokens_details":{"reasoning_tokens":675}},"tokens_in":693,"tokens_out":771,"duration_ms":7362,"temperature":1.0,"reasoning_tokens":675,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:04:59.047406+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the gamma-band top in the same strained film system across a continuous strain series, or in Pr-free La3Ni2O7 films at identical strain, using photon energies and matrix elements that clearly resolve the band: if the band top reaches or crosses the Fermi level in a superconducting film, or if the measured strain shift approaches the 210 meV DFT value, the paper's central claim would be refuted.","supporting_citations":[],"review_version":1}