{"id":"431293d7-d78f-47d0-b05f-50cb93d87b86","arxiv_id":"2511.17403","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"First momentum-resolved photoemission of CrPS4 reveals Cr 3d/S 3p valence bands and orbital-selective Cr–S hybridization that links t2g states to magnetism and eg states to optical d–d transitions.","lead":"Researchers measured the electronic band structure of the layered antiferromagnet CrPS4 for the first time using angle-resolved photoemission, comparing data above and below its magnetic transition with DFT+U calculations. The results provide an experimental benchmark for a promising 2D spintronics material and show how Cr 3d and S 3p orbitals mix differently to control magnetism and optical properties.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 10-K para/antiferromagnetic comparison rests on an undocumented charging correction; if the shift is not rigid in time/momentum, the claimed magnetic-order insensitivity and the DFT+U benchmark for the ordered phase are not established.","rationale":"Reading the paper in good faith, the authors have produced the first ARPES study of CrPS4 with careful sample characterization: exfoliation onto a gold substrate to mitigate charging, AFM thickness/roughness checks, Raman verification of beam-damage absence, and raw data deposition. The 300 K ARPES results and their broad comparison to DFT+U are a plausible and useful benchmark. However, the strongest claim as stated includes the above/below Néel temperature comparison and the inference that magnetic order leaves the band structure essentially unchanged. That inference is the load-bearing bridge that allows the authors to compare 300 K paramagnetic ARPES data to antiferromagnetic DFT+U calculations. The only evidence for the bridge is the 10 K measurement, which the authors themselves describe as affected by positive surface charging accumulating to 2 eV over the measurement. The calibration is delegated to a supplementary section not available in the preprint text we were given, and the main-text comparison is qualitative, with no quantitative tolerance or error analysis. This is a genuine soft spot, not a manufactured one: if the charging shift is non-rigid, both the phase-comparison conclusion and the validation of AFM DFT+U are compromised. I do not think this warrants rejection, because the room-temperature benchmark can likely stand independently and the manuscript is transparent about the limitations. But the conditional verdict is appropriate, and the charging-correction sensitivity should be addressed before the stronger orbital-magnetism and orbital-optics language is taken at face value. I considered the additional concern that the t2g/eg hybridization mechanism is derived entirely from DFT+U and not directly from ARPES; that is real but less load-bearing, because the paper explicitly labels that part as a DFT+U-based analysis. The charging issue is the weakest link in the experimental argument that underpins the central claim.","tokens_in":12231,"tokens_out":5932,"duration_ms":64462,"concrete_test":"Obtain the raw 10 K ARPES data and the Supplementary Section S6 calibration procedure. Reproduce the energy calibration independently by tracking a fixed spectral feature or the secondary-electron cutoff in each time-stamped momentum frame, without assuming a global time-independent shift. Then quantify the residual shift as a function of time, momentum, and binding energy. Rerun the curvature analysis on the independently aligned data and compare 10 K versus 300 K band positions at several high-symmetry points (e.g., the VBM at Γ and the dispersive feature near Y). If the residual energy differences exceed the minor shifts claimed in similar materials (tens of meV), the conclusion that magnetic order leaves the band structure unchanged fails. If the raw data or alignment routine are not available, the claims about the antiferromagnetic-phase comparison should remain conditional.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central strategy is to compare 300 K ARPES with DFT+U calculations for the antiferromagnetic phase. The validity of that comparison rests on the claim that magnetic order leaves the band structure essentially unchanged, which is supported only by the 10 K data. The text states: 'Under photoemission at 10 K, the increased electrical resistivity induces positive surface charging, resulting in a photoelectron kinetic energy shift, which accumulates to 2 eV over 2 hours of measurement time. We carefully analyzed and calibrated the energy shift, as described in Supplementary Section S6.' A 2 eV cumulative shift over 2 hours is large relative to the ~6 eV valence-band width; if the shift is not strictly rigid across momentum, energy, and time, the corrected 10 K dispersions in Figure 4B can be systematically distorted. The main-text comparison is qualitative—curvature plots with traced lines and no quantitative residual or error bar. The conclusion explicitly notes that 'further progress in the preparation of the strongly charging semiconductor' is needed to identify nuanced changes, which is an admission that the 10 K data are at the edge of reliability. If the charging correction is non-uniform, the 'no major differences' finding collapses, and the 300 K ARPES comparison to AFM DFT+U loses its key justification. The 300 K band-structure benchmark may still stand, but the magnetic-order-related portion of the central claim would be unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the first angle-resolved photoemission spectroscopy (ARPES) study of the layered antiferromagnet CrPS4, measured above (300 K) and below (10 K) the Néel temperature, complemented by DFT+U calculations. The authors compare experimental momentum maps and band dispersions with DFT+U (PBE+D3+U_eff=2 eV) and report good agreement. They further analyze orbital-resolved DFT+U results and Wannier functions to argue that Cr t2g orbitals are weakly hybridized with S 3p and sustain local magnetic moments, while eg orbitals hybridize strongly, which relaxes dipole selection rules and explains sub-gap d-d optical transitions. The main claims are (i) an experimental benchmark for DFT+U in CrPS4, (ii) no major band-structure change across the magnetic transition, and (iii) an orbital-selective hybridization picture connecting magnetism and optical response.","tokens_in":12525,"tokens_out":3941,"duration_ms":39825,"significance":"If the central claims hold, this is a valuable experimental benchmark for CrPS4, a material of current interest for 2D antiferromagnetic spintronics and magneto-optics. The paper is careful in sample preparation and characterization (AFM thickness/roughness, Raman damage checks), uses C2/m symmetrization to reduce matrix-element effects, and makes raw data available in a repository. The Wannier-function analysis is a standard and appropriate tool for orbital character. However, the strength of the conclusions is currently limited: the 10 K comparison depends on a charging correction that is not documented in the main text, the experiment-theory agreement is presented only qualitatively, and the orbital-selective hybridization claims are theoretical projections rather than experimental observables. These issues are load-bearing for the stated conclusions, especially the magnetic-order insensitivity and the experimental validation of the AFM DFT+U band structure.","major_comments":[{"comment":"The conclusion that magnetic order leaves the band structure essentially unchanged rests entirely on the 10 K ARPES data. The text reports a cumulative charging shift of 2 eV over 2 h that is corrected using an analysis described only in Supplementary S6. A non-rigid shift (e.g., varying with momentum, energy, or time) would systematically distort the corrected dispersions and invalidate the para-/antiferromagnetic comparison. Please present the calibration evidence in the main text: time-dependent EDCs or curvature plots before and after correction, a test of rigid-shift behavior across the momentum field of view, and an estimate of the residual uncertainty. Until this is shown, the 10 K comparison and the justification for comparing 300 K ARPES with AFM DFT+U are not established.","section":"§4 / Figure 4B and Supplementary S6"},{"comment":"The 'excellent agreement' between ARPES and DFT+U is asserted from visual overlay of curvature plots and traced lines. The manuscript lacks a quantitative agreement metric, error bars on the experimental dispersion, or a residual analysis. This matters because U_eff is an adjustable parameter (2 eV, adopted from Ref. 23), and the Lorentzian broadening eta_arc is chosen to reproduce the experimental maps. Please add a quantitative comparison, e.g., the RMS deviation between traced experimental bands and DFT+U bands over a defined momentum range, or a spectral-function fitting with uncertainty propagation, so that the term 'benchmark' is supported rather than impressionistic.","section":"§3, Figures 3B–C and 4"},{"comment":"The orbital-resolved claims (weak t2g vs strong eg hybridization) are obtained from the same DFT+U calculation that is being benchmarked; the ARPES data are not orbital-resolved. The momentum-resolved spectral weight panels are theory projections, and the experimental comparison is only at the level of total intensity maps. Thus the statement that ARPES 'identifies' these hybridization regimes overstates the evidence. Please either (i) present experimental polarization- or photon-energy-dependent data that can distinguish orbital characters, or (ii) explicitly frame the t2g/eg assignment as a theoretical interpretation that is consistent with, but not uniquely determined by, the ARPES data. The same applies to the claim that eg–p hybridization 'relaxes dipole selection rules'; this is a plausible inference, not a demonstrated experimental result.","section":"§5, Figure 5"}],"minor_comments":[{"comment":"Typo: 'crystalllographic' should be 'crystallographic'.","section":"Figure 3 caption"},{"comment":"Several instances of 'The data was acquired' should be 'The data were acquired' (e.g., Figures 3 and 4 captions).","section":"General"},{"comment":"The phrase 'spin-resolved and atomic orbital-projected DOS' refers to theoretical pDOS; please label explicitly as 'calculated' to avoid implying experimental spin resolution.","section":"§2, Figure 2B"},{"comment":"There is a doubled period in '...with the Cr 3 d states..' and a possible missing space in '...states..' (just before 'The major contribution').","section":"§2, Figure 2B text"},{"comment":"The choice of U_eff = 2 eV is taken from Ref. 23. A brief discussion of sensitivity to U_eff (e.g., test with U_eff = 1 or 3 eV on the reported band features) would strengthen the benchmark claim.","section":"Methods / DFT+U"}],"recommendation":"major_revision","confidential_remarks":"The paper is promising and likely publishable after major revision. The essential missing piece is a transparent validation of the 10 K charging correction and a quantitative experiment-theory comparison. I did not see Supplementary S6; if it contains a rigorous calibration, the authors should move that analysis into the main text or at least summarize it in a way that the reader can evaluate without accessing supplementary material. The orbital-selective claims are interpretive and should be framed accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the measurement: this is the first ARPES on CrPS4, and the 300 K data with the DFT+U comparison makes a plausible benchmark for a material people care about in 2D magnetism. The sample preparation is careful — exfoliation onto gold, AFM thickness and roughness checks, Raman before and after to rule out beam damage — and the data are deposited on Zenodo. That is real, usable work, and the room-temperature band structure and momentum maps look like a solid experimental reference for future optical and transport studies.\n\nThe DFT+U comparison at 300 K is the strongest part. The Ueff = 2 eV choice is inherited from Susilo et al., and the calculated momentum maps and dispersions track the measured features well enough to support the claim that this parameterization is reasonable. The Wannier projection and orbital decomposition are standard but competently done, and the distinction between weakly hybridized t2g and strongly hybridized eg states is a useful organizing picture, even if it is not entirely new as a concept.\n\nWhere I get uneasy is the 10 K part. The authors themselves state that charging produces a photoelectron kinetic energy shift accumulating to 2 eV over two hours, and the calibration is described only in Supplementary Section S6. That is a large shift relative to the ~6 eV valence bandwidth, and if the shift is not rigid in momentum or time, the corrected dispersions in Figure 4B could be systematically distorted. The main-text comparison between 300 K and 10 K is also purely qualitative — traced lines overlaid on curvature plots, no residuals, no error bars. I do not think this invalidates the central benchmark, because the 300 K comparison to antiferromagnetic DFT+U does not strictly require the 10 K data if the magnetic transition is expected to be subtle. But the specific conclusion that 'magnetic order induces only minimal changes to the band structure' is only as strong as that charging correction, and the paper itself concedes that identifying nuanced changes needs better sample preparation.\n\nA smaller but real concern: the orbital-hybridization narrative is inferred from the same DFT+U calculation being benchmarked. That is not circular for the measured band structure, but statements like 't2g orbitals responsible for magnetic ordering' and 'eg hybridization enables optical transitions' are interpretive overlays from the calculation, not direct measurements. Also, the broadening parameter eta_arc = 0.2 eV for theoretical momentum maps is explicitly fitted to the experimental maps; that is fine for visualization, but it should not be counted as validation of lifetimes or matrix elements.\n\nOverall: this paper deserves a serious referee. The 300 K ARPES data are valuable and the DFT+U benchmark is mostly convincing. The 10 K discussion needs either a much more detailed main-text account of the charging calibration or a softened claim. I would send it to review and recommend major revisions with the charging stability and quantitative agreement metrics as the main asks.","headline":"First ARPES benchmark for CrPS4 with a credible 300 K band structure; the 10 K magnetic-order comparison rests on a charging correction that is deferred to the SI and needs scrutiny before the 'no change across TN' claim is taken as established.","tokens_in":13109,"tokens_out":983,"would_cite":true,"duration_ms":12824,"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":"This paper reports the first momentum-resolved photoemission study of the layered antiferromagnetic semiconductor CrPS4, showing that its valence band is made of Cr 3d and S 3p states, and that weakly hybridized t2g orbitals carry the magne","keywords":["CrPS4","angle-resolved photoemission spectroscopy","antiferromagnetic semiconductor","band structure","orbital hybridization","DFT+U","t2g/eg orbitals","charge-transfer gap"],"falsifier":"Measure the band structure at 10 K on a much thinner CrPS4 flake (a few monolayers, where charging is negligible) and compare the Y–Γ–Y dispersion with the 300 K data. If the dispersions differ by more than the experimental resolution (roughly 50 meV), the claim that magnetic ordering does not alter the band structure is wrong. A complementary test is to check whether the calibrated charging shift is identical at the zone center and at the Brillouin-zone boundary during the 10 K measurement; a non-rigid shift would directly invalidate the correction procedure.","tokens_in":12082,"feed_emoji":"🧲","tokens_out":4140,"duration_ms":39468,"temperature":0.7,"pith_summary":"This paper reports the first momentum-resolved photoemission (ARPES) study of the layered antiferromagnetic semiconductor CrPS4, both in its paramagnetic (300 K) and antiferromagnetic (10 K) phases. By combining the measurements with DFT+U calculations, the authors show that the valence band is built from Cr 3d and S 3p states and has a ligand-to-metal charge-transfer gap. The central result is an orbital-selective picture: weakly hybridized t2g orbitals stay localized and spin-polarized, sustaining the magnetic order, while strongly hybridized eg orbitals mix with S p states, relax dipole selection rules, and thereby explain the material's strong sub-gap optical absorption. This is the first experimental benchmark for band-structure calculations of CrPS4 and suggests that magnetic order barely changes the band dispersion.","feed_headline":"First ARPES band map of antiferromagnet CrPS4","feed_subtitle":"Photoemission above and below the Néel temperature matches theory and explains the magneto-optical fingerprint.","key_machinery":"The central object is the orbital-selective hybridization between Cr 3d t2g/eg states and S 3p states, analyzed through DFT+U band structure, spin- and orbital-projected densities of states, and Wannier functions. Wannier functions provide direct spatial evidence: the dxy t2g orbital stays localized on Cr, while the dz2 eg orbital has substantial weight on S ligands. This dichotomy is what carries the argument: t2g localization explains magnetism, eg–S p mixing explains optical activity.","core_discovery":"The paper establishes, for the first time, the experimental electronic band structure of CrPS4 by momentum-resolved photoemission spectroscopy above and below the Néel temperature (38 K), and shows it agrees with DFT+U calculations. In the valence band, Cr 3d and S 3p states dominate, and the band gap has ligand-to-metal charge-transfer character. Within the Cr 3d manifold, the t2g orbitals undergo only weak hybridization with sulfur p states, remain fully spin-polarized and localized, and are the carriers of the local moments that give the A-type antiferromagnetic order. The eg orbitals, by contrast, hybridize strongly with S p orbitals, forming bonding-antibonding pairs separated by about","pith_inferences":["If the orbital-selective picture is correct, tuning the Cr–S hybridization (for example by pressure, strain, or chemical substitution on the sulfur site) should shift the optical d-d oscillator strength and the magnetic anisotropy in a correlated way, because both are governed by the same t2g/eg splitting.","The near-identical band structures above and below the Néel temperature suggest the 38 K transition is driven primarily by interlayer exchange rather than by a change in the intralayer one-electron bands; this could be tested by measuring a monolayer, where interlayer coupling is absent.","The main technical vulnerability is the charging correction: the 2 eV energy shift that accumulates over two hours at 10 K must be strictly rigid in momentum for the claimed para/antiferromagnetic similarity to hold; a momentum-dependent or time-dependent shift would mimic band distortion.","The sequence crystal field → Hund exchange/Coulomb U → hybridization provides a transferable template for understanding other layered magnetic semiconductors where t2g and eg orbitals couple differently to ligands."],"forward_implications":["DFT+U with an effective Hubbard U of 2 eV is now experimentally benchmarked for the occupied valence band of CrPS4, making it a reliable starting point for future calculations of this material.","The ligand-to-metal charge-transfer character of the gap means that near-valence-band-maximum physics is as much about sulfur p states as about chromium d states, not a simple Mott-Hubbard picture.","The observed similarity of band dispersions at 300 K and 10 K indicates that the antiferromagnetic transition does not strongly reconstruct the one-electron bands, so paramagnetic-phase ARPES data can be compared with antiferromagnetic calculations.","The mechanism of eg–S p hybridization relaxing dipole selection rules provides a microscopic explanation for the strong sub-gap optical absorption near 1.6 and 1.8 eV in CrPS4.","The distinct hybridization characters of t2g and eg states open a route to separately probe magnetism and optical response via orbital-selective spectroscopies and doping experiments."],"fun_headline_variants":["First ARPES map reveals CrPS4's orbital dance","ARPES exposes magnetic orbitals in CrPS4","CrPS4 band structure finally mapped by ARPES","Photoemission maps CrPS4's magnetic orbitals"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The conclusion that magnetic order leaves the band structure essentially unchanged rests on the assumption that the 2 eV charging-induced photoelectron energy shift at 10 K is a rigid, homogeneous shift that can be fully calibrated; if the shift varies across momentum or drifts in time, the 10 K band dispersions and the claimed similarity between paramagnetic and antiferromagnetic phases would be compromised.","fun_headline_variants_meta":{"raw":{"variants":["First ARPES map reveals CrPS4's orbital dance","ARPES exposes magnetic orbitals in CrPS4","CrPS4 band structure finally mapped by ARPES","Photoemission maps CrPS4's magnetic orbitals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000538,"raw_usage":{"total_tokens":2399,"prompt_tokens":702,"completion_tokens":1697,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":446,"completion_tokens_details":{"reasoning_tokens":1632}},"tokens_in":446,"tokens_out":1697,"duration_ms":10325,"temperature":1.0,"reasoning_tokens":1632,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T20:54:59.014856+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the band structure at 10 K on a much thinner CrPS4 flake (a few monolayers, where charging is negligible) and compare the Y–Γ–Y dispersion with the 300 K data. If the dispersions differ by more than the experimental resolution (roughly 50 meV), the claim that magnetic ordering does not alter the band structure is wrong. A complementary test is to check whether the calibrated charging shift is identical at the zone center and at the Brillouin-zone boundary during the 10 K measurement; a non-rigid shift would directly invalidate the correction procedure.","supporting_citations":[],"review_version":1}