{"id":"071de933-63dc-46c9-ac4e-8eb4ade35a35","arxiv_id":"2412.16103","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"In expanded-lattice FeSe superconductors, the amount of electron doping delivered by intercalated molecules tunes Tc more than the interlayer spacing alone.","lead":"This paper compares two versions of a pyridine-intercalated iron selenide superconductor to understand why the as-made one superconducts at 44 K while the annealed one drops to 39 K. X-ray and neutron measurements suggest that differences in electron doping from the intercalated molecules, not just the distance between the iron layers, set the critical temperature.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The electron-doping inference rests on an uncalibrated Se K-edge white-line difference and a model-dependent Debye softening; the doping–Tc link needs a quantitative, controlled test.","rationale":"The reader correctly identified the Debye/Einstein model dependence in the local-bond-softening argument. I go further: the more load-bearing weakness is the XANES-based doping inference, because the central claim is not merely about bond softening but about carrier doping as the key Tc parameter. The XANES comparison in Fig. 4 uses data from two separate studies, reports only a 'small' intensity difference, and lacks any uncertainty or normalization sensitivity analysis. Since the as-made and annealed samples differ in interlayer spacing, pyridine content, and Fe-site occupancy, the white-line difference cannot be uniquely attributed to electron doping without a controlled measurement or an independent electronic-structure probe. The Fe–Se bond-length difference, although statistically quoted, is also small and derived from different experiments. The paper otherwise contains useful, reproducible experimental work: careful sample characterization, temperature-dependent EXAFS, and INS PDOS with honest statements about the absence of a spin resonance and the limited resolution. The conditional verdict is appropriate: the manuscript presents a plausible but not yet fully supported central claim. The proposed single-session XANES comparison with full error propagation, together with an optional XES/HAXPES check, would either solidify the doping interpretation or reveal that the differences are within experimental uncertainty. No change to the reader's conditional verdict is needed.","tokens_in":17855,"tokens_out":7868,"duration_ms":82208,"concrete_test":"Measure as-made and annealed Li1.0(C5H5N)yFe2-zSe2 samples from the same synthesis batch in a single synchrotron session, with identical sample environment, E0 calibration, and normalization procedure. Quantify the white-line intensity difference of feature #A using Monte-Carlo propagation over repeated scans and over reasonable variations in the pre/post-edge normalization ranges. If the difference is within ~2σ, the electron-doping inference collapses; if it exceeds ~3σ, the inference is supported. As a complementary, model-independent check, record Se Kβ X-ray emission or core-level photoemission shifts on the same samples to directly test the valence change.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that carrier doping, rather than interlayer spacing alone, parametrizes Tc in the expanded-lattice regime rests on three proxies: reduced Se K-edge white-line intensity (#A) interpreted as increased Se 4p filling, a 0.005 Å Fe–Se bond lengthening, and a Debye-model softening of the local bond. The weakest link is the XANES proxy. In Fig. 4b–c the as-made and parent spectra are taken from a previous study (ref. 25) rather than measured side-by-side with the annealed sample; the #A difference is described as 'small' but is shown without error bars, repeated-scan statistics, or sensitivity tests to E0 determination and normalization range. Se 1s→4p white-line intensity in FeSe depends on Fe–Se hybridization and local coordination, not solely on the occupation of Se 4p states. The as-made and annealed phases differ in d (16.2 vs. 11.4 Å), pyridine content (y ~0.45 vs. 0.22), and Fe-site occupancy (N = 3.88 vs. 4.01 in Table 3), any of which can shift feature #A without a change in formal electron count. The Fe–Se bond difference is small (~0.005 Å, roughly 2.5σ from the quoted errors, and extracted from two different experiments). The Debye-based softening is also model-dependent: the correlated Einstein model (Section 2.2, Table 3) gives statistically identical Einstein temperatures across parent, annealed, and as-made phases, so the MSRD data alone do not establish a change in bond stiffness. Together, these unquantified or model-dependent proxies support the claim only weakly; the doping part of the central conclusion is plausible but not yet quantitatively established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports Se K-edge XANES and EXAFS measurements on the annealed Li1.0(C5H5N)yFe2-zSe2 superconductor, together with inelastic neutron scattering on the deuterated annealed phase, and compares them with the as-made intercalate and with parent β-FeSe. The central claim is that electron doping delivered by the intercalated alkali-molecule guests, rather than interlayer spacing alone, parametrizes Tc in the expanded-lattice regime: the as-made phase (Tc ≈ 44 K) is argued to be more electron-doped, to have a longer and softer Fe-Se bond, and to show a higher Se 4p filling than the annealed phase (Tc ≈ 39 K). Supporting evidence comes from XANES feature #A and #B intensities, a small Fe-Se bond elongation from EXAFS, a lower correlated Debye temperature, the absence of a neutron spin-resonance mode, and hardening of low-energy transverse-acoustic phonons attributed to chemical pressure from the intercalated guests.","tokens_in":18300,"tokens_out":6350,"duration_ms":61428,"significance":"If the central doping claim is established, the paper would make a useful contribution to the ongoing discussion of why Tc saturates in molecule-intercalated FeSe at large interlayer separations. The study combines element-selective local-structure probes with neutron spectroscopy on carefully prepared air-sensitive samples, and the EXAFS modelling follows standard practice with the supporting details given in the SI. The neutron PDOS, in particular the hardening of the lowest TA mode and the molecular vibrational features, is a valuable dataset in its own right. However, the paper's main conclusion is carried by three proxies for electron doping, and two of them are either small-scale or model-dependent: the XANES white-line difference is presented without error bars or normalization sensitivity tests, and the bond softening is inferred from a correlated Debye model even though the correlated Einstein model gives statistically identical local force constants for the three phases. No direct carrier-concentration measurement is provided. These weaknesses make the central claim plausible but not yet quantitatively supported.","major_comments":[{"comment":"The XANES-based inference of different Se 4p filling is not quantitatively supported as presented. The as-made and parent spectra in Fig. 4b,c are taken from a previous study (ref. 25) rather than measured side-by-side with the annealed sample, and the reported difference in feature #A is described only as 'small' with no error bars, scan-to-scan statistics, or sensitivity tests with respect to E0 determination and normalization range. Since the Se 1s→4p white-line amplitude also depends on Fe-Se hybridization and the local coordination environment, and since the as-made and annealed phases differ substantially in interlayer spacing (16.2 vs 11.4 Å), pyridine content (y ≈ 0.45 vs 0.22) and Fe-site occupancy (N = 3.88 vs 4.01, Table 3), the intensity difference alone does not establish a difference in Se 4p occupancy. This proxy is load-bearing for the abstract's claim that the annealed phase has reduced Se 4p filling and hence lower electron doping.","section":"III.C.1, Fig. 4b-d"},{"comment":"The conclusion that the higher-Tc as-made phase has a softer local Fe-Se bond is model-dependent. The correlated Einstein model gives statistically identical Einstein temperatures for parent, annealed, and as-made phases (θE = 318.2±4.8, 325.0±1.3, and 319.4±5.0 K), implying similar local force constants, whereas the softening appears only in the correlated Debye model (θD = 380.0±5.3, 368.4±3.2, and 312.1±4.4 K). The manuscript states that the Einstein model is a basic approximation but does not provide a goodness-of-fit comparison showing that the Debye model is statistically preferred, nor does it account for the likely correlation between θD and the strongly differing static disorder σs^2 in the fits. The MSRD data alone therefore do not demonstrate a change in local bond stiffness; the softening is an interpretation, not a direct measurement.","section":"III.C.2.2, Table 3"},{"comment":"The Fe-Se bond-length difference between the as-made and annealed phases that is used to support the doping scenario is very small and is extracted from two different experiments: the annealed data are measured in this work, while the as-made values come from ref. 25. The tabulated values (2.394(1) vs 2.3908(8) Å) differ by only about 0.003 Å, roughly 2.5σ, and systematic offsets in E0 alignment, edge-step normalization, or sample condition between separate beamtimes could easily produce a shift of this size. This metric should be treated as corroborative only, and the paper should state explicitly what statistical and systematic uncertainties are included.","section":"III.C.2.1, Table 2"}],"minor_comments":[{"comment":"In the sentence referring to the softening of the Fe-Se bond, 'Table 1' should be 'Table 3': the Debye temperatures are reported in Table 3, not Table 1.","section":"III.C.2.2"},{"comment":"The absence of a spin-resonance mode is used to speculate about deviations from optimal doping, but the text also acknowledges that phonon scattering could hide such a signal; without a background-subtracted Q-E difference map or an estimated upper limit on magnetic scattering, the absence should be presented only as an inconclusive null result, not as independent evidence for the doping scenario.","section":"III.B, Figs. 3 and S3"},{"comment":"For the pristine FeSe modes #3 and #4, the table lists a single range '15-18', while the intercalated values are two discrete energies (14.5 and 17.7 meV); please clarify whether the range covers two unresolved modes in the pristine compound.","section":"Table 1"},{"comment":"The figure caption states that the β-FeSe and as-made data are compiled from ref. 25, but the main text does not repeat this; a brief reminder in the text would help readers judge that the comparison is not from a single experimental run.","section":"III.C.1, Fig. 4"},{"comment":"The notation for interlayer spacings and Tc values is inconsistent (e.g., 'das-made', 'dan nealed', 'Tc as-made'); please use a uniform subscript/superscript convention throughout.","section":"I, III.C.2.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript fits the journal's scope and contains useful complementary XAS and neutron data, but I cannot support acceptance until the central doping inference is placed on firmer quantitative ground. The authors should either add a direct doping probe (e.g., core-level shifts, Hall measurements, or ARPES) or provide side-by-side XANES with proper error bars and a statistical comparison of the Debye and Einstein fits. The reliance on previously published spectra for the as-made and parent phases for the key comparison is a particular weakness. I see the issues as fixable within the manuscript's scope, hence major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a careful experimental paper with genuinely new data, but the central claim—that carrier doping, not just interlayer spacing, parametrizes Tc in these expanded-lattice FeSe phases—is built on small, partly model-dependent proxies. I would send it to review, but the referees should push for quantitative doping evidence.\n\nWhat's new is real. The annealed Li1.0(C5H5N)yFe2-zSe2 phase is characterized here for the first time with Se K-edge XAS and INS. The as-made vs annealed comparison is new, and so is the absence of a spin resonance in the annealed phase—a useful negative result for the field. The INS work is solid: the TA mode hardening relative to beta-FeSe is documented, the molecular modes are assigned, and the temperature dependence across Tc shows no anomaly. The EXAFS analysis follows standard practice, and the authors are honest enough to report both Einstein and Debye model fits.\n\nThe soft spots are in the interpretation. The XANES difference (feature #A) is small, shown without error bars, and the as-made and parent spectra come from a previous study, not from side-by-side measurements. Interpreting #A as a direct readout of Se 4p occupation is plausible but not quantitative; changes in hybridization, coordination number, or pyridine content could shift it. The Fe-Se bond length difference between as-made and annealed is only ~0.003 Å (roughly 2.5σ) and extracted from different experiments. The claimed bond softening rests entirely on the correlated Debye model: the correlated Einstein model gives statistically identical Einstein temperatures across parent, annealed, and as-made phases, as the authors themselves show. So the 'softer bond' is a model-dependent inference, not a direct measurement.\n\nThere is also a structural confound: the two intercalated phases differ in interlayer spacing (16.2 vs 11.4 Å), pyridine content (y ~0.45 vs 0.22), and Fe-site occupancy (N = 3.88 vs 4.01). Any of these can affect the XAS without a change in formal electron count. Attributing the observed differences specifically to electron doping needs a direct carrier measurement or a more controlled comparison.\n\nWho is this for? Researchers working on intercalated iron selenides and the Tc-versus-spacing puzzle. For them, the INS data and the no-resonance observation are worth having. For a broader audience, the conclusion overloads the evidence.\n\nRecommendation: send it to peer review. It deserves careful referee scrutiny. The data are new and the authors are transparent, but the manuscript should be revised to quantify the XANES contrast, address the Einstein/Debye discrepancy, and scale back the doping claim accordingly.","headline":"New XAS and INS data on the annealed pyridine-intercalated FeSe are worth having, but the doping-Tc claim rests on a small XANES contrast and a Debye-model softening that the Einstein fit does not back up.","tokens_in":18763,"tokens_out":3973,"would_cite":false,"duration_ms":33543,"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":"In far-apart FeSe layers, electron doping delivered by intercalated molecules, not interlayer spacing alone, sets the superconducting transition temperature.","keywords":["layered superconductors","iron chalcogenides","intercalation","electron doping","XANES","EXAFS","phonon density of states","inelastic neutron scattering"],"falsifier":"A direct measurement of the Fe-Se bond-stretching phonon energy (e.g., by momentum-resolved inelastic neutron or X-ray scattering on single crystals of the as-made and annealed phases) would confirm or refute the claimed local bond softening; if both phases have the same phonon energy, the Debye-model inference collapses.","tokens_in":1059,"feed_emoji":"🧲","tokens_out":4606,"duration_ms":78418,"temperature":0.7,"pith_summary":"This paper asks what controls the superconducting transition temperature in iron selenide superconductors once the FeSe layers are pried far apart by intercalated molecules. Comparing two Li-pyridine intercalates of FeSe—the as-made phase with a Tc near 44 K and an annealed derivative near 39 K—it argues that electron doping delivered by the intercalant, not just interlayer spacing, sets Tc. Se K-edge X-ray absorption shows the higher-Tc phase has more electrons in Se 4p states, longer Fe-Se bonds, and, under a correlated Debye model, a softer local Fe-Se bond. Neutron spectroscopy on the annealed phase finds no spin-resonance mode below Tc, suggesting it sits away from optimal doping. If correct, the result redirects attention to chemical control of carrier density rather than simply pushing layers apart.","feed_headline":"Intercalant doping, not layer spacing, sets FeSe Tc","feed_subtitle":"Se K-edge shows higher Tc phase carries more electrons and softer Fe-Se bonds","key_machinery":"The load-bearing probe is the Se K-edge X-ray absorption spectrum, which reads the local electronic and structural state of the FeSe layer. Its near-edge region (XANES) tracks the occupancy of Se 4p states near the Fermi level, while the extended region (EXAFS) isolates the nearest-neighbor Fe-Se bond length and its mean-square relative displacement (MSRD). The correlated Debye model converts those MSRDs into a Debye temperature θD, a one-parameter measure of local bond stiffness; lower θD means a softer Fe-Se bond. On the dynamics side, neutron time-of-flight spectroscopy supplies the phonon density of states, where the lowest transverse-acoustic mode and the internal pyridine vibrations act as sensors of chemical pressure exerted by the intercalated molecules.","core_discovery":"The central claim is that in the expanded-lattice regime (interlayer spacing d > 8.6 Å), the magnitude of Tc in Li1.0(C5H5N)yFe2-zSe2 is governed by the level of electron doping transferred from intercalated [Li-pyridine] guests to the FeSe layers, rather than by interlayer separation alone. Evidence comes from comparing the as-made polytype (d ~ 16.2 Å, Tc ~ 44 K) with the annealed polytype (d ~ 11.4 Å, Tc ~ 39 K). XANES at the Se K-edge shows greater filling of Se 4p orbitals in the as-made phase, while EXAFS gives a longer Fe-Se bond (2.394 Å vs 2.391 Å), consistent with higher electron doping. Fits of the Fe-Se mean-square relative displacements to the correlated Debye model yield a lower Debye temperature for the as-made phase (312 K vs 368 K annealed, 380 K parent), interpreted as softening of the local Fe-Se bond with doping. Neutron time-of-flight spectra on the annealed deuterated phase show that confined molecules act as chemical pressure on the Se-Fe-Se sheets, hardening the lowest transverse-acoustic mode relative to β-FeSe, while no collective spin resonance develops below Tc, suggesting the annealed compound is not optimally doped.","pith_inferences":["If doping rather than spacing controls Tc, then systematically varying the Li-to-pyridine ratio in the as-made phase should trace a Tc-vs-doping dome, a test the paper does not perform.","The Debye softening is model-dependent, so an independent probe of the Fe-Se bond's force constant—for example, measuring the zone-center optical phonon energy directly—would either corroborate or overturn the claimed correlation between Tc and bond softness.","A similar XAS and neutron study on ammonia-intercalated FeSe with controlled amide content could determine whether the same doping-vs-Tc relationship holds across different molecular guests."],"forward_implications":["In the expanded-lattice regime beyond 8.6 Å, interlayer spacing alone does not set Tc; electron doping does, so two phases with very different spacings can have different Tc if doping differs.","Annealing lowers the electron count in FeSe layers, seen as fewer filled Se 4p states and a shorter Fe-Se bond, lowering Tc from about 44 K to 39 K while healing Fe-site vacancies and sharpening the transition.","The local Fe-Se bond is softer in the higher-Tc, more heavily doped as-made phase, so local bond stiffness, not just bond length, correlates with Tc.","Intercalated pyridine molecules exert chemical pressure on the FeSe layers, evidenced by a roughly 20 percent hardening of the lowest transverse-acoustic phonon relative to parent β-FeSe.","The absence of a spin-resonance mode below Tc in the annealed phase implies the compound is not optimally doped, so its 39 K Tc is not the maximum for this intercalant family."],"supporting_citations":[{"why":"Provides the previous XAS study of the as-made Lix(C5H5N)yFe2-zSe2 series, including the parent and as-made data reproduced here for comparison.","marker":"[25]"},{"why":"Supplies the NH3-poor intercalate's interlayer spacing, Tc, and Fe-Se bond length used in the comparison table.","marker":"[15]"},{"why":"Supplies the NH3-rich intercalate's corresponding data, showing the opposite trend in Tc with bond length that motivates the doping interpretation.","marker":"[16]"},{"why":"Defines the 8.6 Å interlayer-spacing threshold beyond which Tc saturates, the regime this paper investigates.","marker":"[12]"},{"why":"Dominant electronic-structure interpretation that links reduced electron doping in alkali-ammonia intercalates to smaller Tc, which the paper extends to pyridine intercalates.","marker":"[13]"},{"why":"Provides the phonon density of states of β-FeSe and the assignment of the low-energy acoustic modes that the intercalated compound is compared against.","marker":"[39]"},{"why":"Supplies the vibrational energies of free pyridine used to identify the molecular modes observed in the intercalated phase.","marker":"[40]"},{"why":"Reports a spin resonance in molecular-intercalated FeSe, the phenomenon the paper looks for but does not find.","marker":"[23]"}],"fun_headline_variants":["Doping, not distance, sets FeSe superconducting Tc","Intercalant electrons, not spacing, control FeSe Tc","Tc in FeSe: doping wins over interlayer separation","Why FeSe Tc: electron doping beats lattice expansion"],"cache_read_input_tokens":20864,"weakest_assumption_plain":"The conclusion that the higher-Tc as-made phase has softer Fe-Se bonding rests on the correlated Debye model fit to EXAFS mean-square relative displacements; the correlated Einstein model gives statistically similar Einstein temperatures across all three compounds, so the softening is a model-dependent interpretation rather than a direct measurement.","fun_headline_variants_meta":{"raw":{"variants":["Doping, not distance, sets FeSe superconducting Tc","Intercalant electrons, not spacing, control FeSe Tc","Tc in FeSe: doping wins over interlayer separation","Why FeSe Tc: electron doping beats lattice expansion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000262,"raw_usage":{"total_tokens":1741,"prompt_tokens":1237,"completion_tokens":504,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":853,"completion_tokens_details":{"reasoning_tokens":437}},"tokens_in":853,"tokens_out":504,"duration_ms":5028,"temperature":1.0,"reasoning_tokens":437,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:46:41.384405+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of the Fe-Se bond-stretching phonon energy (e.g., by momentum-resolved inelastic neutron or X-ray scattering on single crystals of the as-made and annealed phases) would confirm or refute the claimed local bond softening; if both phases have the same phonon energy, the Debye-model inference collapses.","supporting_citations":[{"cited_title":"Deltsidis, L","cited_arxiv_id":null,"evidence_quote":"Provides the previous XAS study of the as-made Lix(C5H5N)yFe2-zSe2 series, including the parent and as-made data reproduced here for comparison."},{"cited_title":"Burrard-Lucas et al., Enhancement of the superconducting transition temperature of FeSe by intercalation of a molecular spacer layer, Nat","cited_arxiv_id":null,"evidence_quote":"Supplies the NH3-poor intercalate's interlayer spacing, Tc, and Fe-Se bond length used in the comparison table."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the NH3-rich intercalate's corresponding data, showing the opposite trend in Tc with bond length that motivates the doping interpretation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the 8.6 Å interlayer-spacing threshold beyond which Tc saturates, the regime this paper investigates."},{"cited_title":"Guterding, H","cited_arxiv_id":null,"evidence_quote":"Dominant electronic-structure interpretation that links reduced electron doping in alkali-ammonia intercalates to smaller Tc, which the paper extends to pyridine intercalates."},{"cited_title":"Ksenofontov, G","cited_arxiv_id":null,"evidence_quote":"Provides the phonon density of states of β-FeSe and the assignment of the low-energy acoustic modes that the intercalated compound is compared against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the vibrational energies of free pyridine used to identify the molecular modes observed in the intercalated phase."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports a spin resonance in molecular-intercalated FeSe, the phenomenon the paper looks for but does not find."}],"review_version":1}