{"id":"6f419d74-2750-4135-8fec-14b5bfd61a49","arxiv_id":"2412.09955","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Cr1+δTe2 crystals host an orthogonal ferromagnetic state with alternating in-plane and out-of-plane magnetized layers, accompanied by abrupt spin-flop-like transitions.","lead":"Magnetic measurements and theory show that a layered chromium telluride crystal holds a previously unrecognized magnetic state, with adjacent atomic layers magnetized in perpendicular directions. The finding may settle a long-running dispute about the material's magnetic order and could point toward new spintronic switching devices.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The unique claim of alternating orthogonal magnetic layers is not directly established: bulk magnetization and single-k spin-ARPES cannot exclude a uniform canted ferromagnet, so 'definitive evidence' overreaches.","rationale":"I read the paper as claiming a new bulk magnetic phase, orthogonal ferromagnetism, with alternating in-plane and out-of-plane ferromagnetic layers, and abrupt spin-flop transitions. The most load-bearing assumption is that the observed magnetization and photoemission reflect this specific two-sublattice arrangement rather than a more conventional uniform canted ferromagnet. The paper's own caveat about magnetic domains in the spin-ARPES measurement (Fig. 1j) is an explicit admitted limitation, and no layer-resolved magnetic probe is reported. I considered the ordered 2x2 intercalant assumption identified by the reader; it matters for the DFT energetics, but the defining 'alternating layers' feature would survive some in-plane intercalant disorder, so the underdetermination of the magnetic structure is the sharper concern. The proposed neutron diffraction experiment would settle it directly. I therefore keep the reader's CONDITIONAL verdict rather than escalating: the claim is plausible and the DFT plus magnetization are real evidence, but 'definitive' is too strong until a layer-resolved magnetic structure determination is available.","tokens_in":13227,"tokens_out":15457,"duration_ms":194821,"concrete_test":"Perform single-crystal neutron diffraction with polarization analysis (or resonant magnetic X-ray scattering at the Cr L3 edge if crystals are too small) on Cr1.25Te2 below the magnetic ordering temperature. Refine the magnetic structure with two crystallographically independent Cr sites (CrTe2-layer Cr and intercalated Cr) and test whether the moment directions on the two sites are mutually orthogonal. If the data can be fit equally well by a single global canted-moment model, the orthogonal-ferromagnetism claim is not established; if the alternating-layer orthogonal model is uniquely required, it is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing element of the central claim is the atomically alternating in-plane/out-of-plane moment arrangement shown in Fig. 2a-b. This arrangement comes solely from LDA non-collinear DFT on a periodic 2x2 cell with one intercalated Cr; no measurement resolves the magnetic structure on a layer-by-layer basis. The SQUID data (Fig. 2d-e, 3a; supplementary Figs. S7-S9) are bulk integrals over all Cr sites and would look similar for a uniform canted ferromagnet with one global easy cone: square out-of-plane loops, non-saturating in-plane M(H), and kinks from spin reorientation are generic to competing anisotropy and exchange, not a unique signature of alternating orthogonal blocks. The spin-ARPES result (Fig. 1j) is taken at a single k-point, and the authors explicitly caveat that the measured vector could arise from a single domain within the 10 um spot ('likely magnetic domains comparable to the spot size...') or a higher proportion of domains with consistent orientation. A single-domain uniform canted ferromagnet would also break time-reversal symmetry and show nonzero Sx, Sy, and Sz. The DFT 2x2 ordered intercalant supercell is a computational choice; no XRD, Raman, STEM, or diffraction data verify the assumed in-plane intercalant order, and even perfect layer-wise intercalation would not by itself prove the orthogonal moment orientation without a layer-resolved magnetic probe. Therefore the distinctive part of the claim, alternating atomically sharp orthogonal blocks, is underdetermined by the evidence as presented.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes that Cr1+δTe2 (δ = 0.25–0.50) hosts a previously unrecognized magnetic ground state, termed 'orthogonal ferromagnetism,' in which atomically sharp single layers of in-plane and out-of-plane ferromagnetically ordered Cr moments alternate and are coupled by antiferromagnetic exchange. The authors support this with DFT total-energy minimization over multiple initial non-collinear configurations (Supplementary Fig. S2), spin-ARPES at the Γ point showing finite Sx, Sy, and Sz (Fig. 1j), and SQUID magnetometry showing square out-of-plane loops and non-saturating in-plane M(H) with kinks interpreted as spin-flop-like transitions (Figs. 2d–e, 3a). They argue this resolves earlier discrepancies between canted-ferromagnet descriptions and magnetization downturns, and they extend the claim to Cr1.5Te2.","tokens_in":13448,"tokens_out":5010,"duration_ms":50551,"significance":"If correct, the discovery of an intrinsic single-phase orthogonal ferromagnet would be of considerable interest: it would constitute a new magnetic state in a quasi-2D van der Waals system, with potential implications for spintronics and for reconciling conflicting reports on Cr1+δTe2. The paper's methodology is multimodal and internally consistent: the DFT calculations are not fitted to the magnetization data, the ARPES band structure is compared to DFT, and the magnetization measurements show clear anisotropy. However, the load-bearing evidence for the layer-resolved alternating moment arrangement is indirect; the most distinctive aspect of the claim, the atomically alternating orthogonal order, rests on a DFT supercell model whose intercalant ordering is not experimentally verified, and the bulk and single-k measurements cannot exclude a uniform canted ferromagnet. The paper would be strengthened by a quantitative model of the M(H) response or a layer-resolved magnetic probe.","major_comments":[{"comment":"The alternating in-plane/out-of-plane moment arrangement is obtained only from LDA non-collinear DFT in a 2×2 supercell with a single intercalated Cr per cell. The assumed ordered intercalant arrangement is not verified by any experiment (e.g., electron diffraction, STEM, or XRD superstructure). Without such verification, or at least a discussion of the sensitivity of the ground state to intercalant disorder or different ordering patterns, the computed orthogonal state may not represent the actual bulk magnetic structure. The authors should either provide structural evidence for the ordered 2×2 intercalant lattice or explicitly present the orthogonal phase as a candidate ground state whose experimental confirmation requires further structural and magnetic characterization.","section":"Fig. 2a-b and Methods (DFT calculations)"},{"comment":"The spin-ARPES data at the Γ point show nonzero Sx, Sy, and Sz, which demonstrates broken time-reversal symmetry, but this is not a unique fingerprint of the alternating orthogonal state. A uniform canted ferromagnet with a single domain would also show all three spin components. The authors themselves note that the observed polarization could arise from a single magnetic domain within the 10 µm spot ('likely magnetic domains comparable to the spot size'). The claim of 'definitive evidence' in the Abstract and Conclusion is therefore too strong; at minimum, the paper should acknowledge that spin-ARPES at a single k-point is consistent with, but does not uniquely establish, the orthogonal arrangement.","section":"Fig. 1j and associated text"},{"comment":"The in-plane magnetization curves show a kink and non-saturating behavior, which the authors interpret as a two-stage spin reorientation with a spin-flop-like transition. However, no quantitative model is presented to show that the proposed orthogonal state reproduces the observed M(H) curves, including the field positions of the kinks and the magnetization values. Since non-saturating M(H) and kinks are generic to systems with competing magnetic anisotropies and exchange, these data alone do not discriminate between the orthogonal state and a uniform canted ferromagnet. The authors should simulate M(H) for their spin Hamiltonian and compare the predicted kink fields and magnetization steps to experiment, also comparing to the canted-ferromagnet scenario.","section":"Fig. 3a-b and the spin-flop interpretation"}],"minor_comments":[{"comment":"The description of the DFT initial configurations is internally inconsistent: the text states that both the φi = 0 and φi = π starting configurations 'maintained ... antiferromagnetic alignment,' but then refers to '98 meV and 40 meV for the ferromagnetic and antiferromagnetic alignment, respectively.' Presumably one of these is ferromagnetic; please correct this discrepancy.","section":"Supplementary Fig. S2 and main text"},{"comment":"The parameter V0 = 8 eV is introduced without definition; presumably it is the inner potential used for the kz mapping, but this should be stated explicitly.","section":"Fig. 1g"},{"comment":"The word 'ploarized' appears instead of 'polarized'; please proofread the Methods section.","section":"Methods (Photoelectron Spectroscopy)"},{"comment":"The phrase 'single-ride magnetization curves' should be 'single-cycle' or 'single-ramp' magnetization curves.","section":"Supplementary Fig. S8 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper presents an intriguing candidate magnetic ground state and a substantial multimodal dataset, but the headline claim of 'orthogonal ferromagnetism' currently outruns the evidence. The most distinctive assertion—atomically alternating orthogonal layers—is supported only by DFT on an assumed ordered supercell and by measurements that are equally compatible with a uniform canted ferromagnet. I would encourage the editor to request either additional experimental verification (e.g., layer-resolved XMCD, neutron diffraction, or magnetic imaging) or a significant softening of the claims and a quantitative comparison with the canted-ferromagnet alternative. The paper may be better suited to a specialized magnetism venue if the evidence remains indirect."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe thing to know about this paper: it claims a new magnetic phase, orthogonal ferromagnetism, in Cr1+δTe2 (δ=0.25–0.50), with alternating in-plane and out-of-plane ferromagnetic layers coupled antiferromagnetically. The claim is interesting and the observed abrupt spin-flop-like transitions are a genuine departure from the gradual reorientation reported earlier. But the distinctive part of the claim—the atomically sharp alternating blocks—rests almost entirely on LDA non-collinear DFT in an ordered 2×2 supercell. No measurement resolves the magnetic structure layer by layer. The phrase “definitive evidence” is too strong.\n\nWhat the paper does well: the DFT is careful. They tested a wide range of initial spin configurations and found this state to be most stable, with a clear energy separation (40–98 meV) from collinear states. The magnetization data are consistent with the model: square out-of-plane loops, non-saturating in-plane curves with kinks, and sensible anisotropy values. Extending the results to Cr1.5Te2 strengthens the case that this phenomenology is robust. The authors also openly note that the spin-ARPES signal might come from a single domain within the 10 µm spot, which is honest.\n\nThe soft spots are real. The spin-ARPES is at one k-point and cannot distinguish alternating orthogonal blocks from a uniform canted ferromagnet—a single-domain canted ferromagnet would also give nonzero Sx, Sy, Sz at Γ. Bulk SQUID integrates over all Cr sites, so it is blind to the layer-resolved structure. The 2×2 ordered intercalant supercell is an assumption; no XRD, STEM, or diffuse scattering evidence shows that the excess Cr actually orders that way. If the intercalants are disordered or arranged differently, the computed ground state may not be the true bulk state. These caveats do not kill the idea, but they mean the central claim is underdetermined by the evidence as presented.\n\nWho is this for: specialists in 2D magnetism and chromium tellurides. It deserves a serious referee, but the referee should push for toning down the definitive language, addressing the domain ambiguity quantitatively, and ideally providing structural or layer-resolved magnetic data—or at least a clear argument why the ordered DFT supercell is representative.\n\nIn short: plausible, possibly important, but not proven. Send it to peer review, but expect significant revision.","headline":"A plausible new magnetic phase in Cr1+δTe2, but the layer-resolved orthogonal arrangement is not directly proven; the paper overstates its definitive evidence.","tokens_in":14165,"tokens_out":2708,"would_cite":true,"duration_ms":30068,"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":"The paper claims that Cr1+δTe2 hosts an “orthogonal ferromagnet” phase — alternating out-of-plane and in-plane ferromagnetic monolayers — that resolves the disputed canted-ferromagnet picture and explains abrupt spin-flop-like transitions.","keywords":["orthogonal ferromagnetism","Cr1+δTe2","self-intercalated van der Waals magnet","spin-flop transition","non-collinear magnetism","spin-ARPES","SQUID magnetometry","density functional theory"],"falsifier":"Perform bulk-sensitive magnetic diffraction (neutron or resonant X-ray magnetic scattering) on the same Bridgman-grown Cr1.25Te2 crystals below the ordering temperature and look for reflections from a unit cell in which intercalant moments are transverse to layer moments; if only a single canted moment direction is found, or the intercalant moments are disordered, the orthogonal phase is not the ground state. On the same crystals, raster the micro-focused spin-ARPES beam: spin-polarization maps that flip sign between regions would confirm domains comparable to the spot size, while a uniform polarization would favor the single-domain reading.","tokens_in":12980,"feed_emoji":"🧲","tokens_out":8924,"duration_ms":90631,"temperature":0.7,"pith_summary":"Cr1+δTe2 has been studied for years, yet its magnetic ground state is disputed: some see an out-of-plane ferromagnet, others a canted ferromagnet with gradual spin reorientation. This paper argues that both descriptions miss the actual low-temperature state, which it calls orthogonal ferromagnetism: within each CrTe2 monolayer the Cr moments align ferromagnetically out of plane, while the Cr atoms intercalated in the van der Waals gap align nearly in plane, so the crystal is a natural stack of alternating in-plane and out-of-plane ferromagnetic monolayers coupled by antiferromagnetic exchange. The paper presents SQUID magnetometry, polarization-dependent ARPES, non-collinear DFT, and micro-focused spin-ARPES to argue that this single-phase alternating order explains the disputed magnetization data, including a downturn that gradual-reorientation models could not accommodate. The relevance is practical as well as fundamental: if this phase is real, Cr1+δTe2 is a bulk material that performs the orthogonal spin architecture usually achieved only by stacking different magnetic layers, with abrupt spin-flop-like switching.","feed_headline":"CrTe2 hosts hidden orthogonal ferromagnet phase","feed_subtitle":"The phase flips spins abruptly under in-plane fields and explains a long-disputed magnetization signal.","key_machinery":"The argument is carried by a 2×2×1 supercell of CrTe2 containing one intercalated Cr atom, relaxed with non-collinear LDA including self-consistent spin-orbit coupling from a wide set of initial canting angles; all non-collinear starting points converge to the same orthogonal arrangement, while collinear ferromagnetic and antiferromagnetic starting points end at energies 98 meV and 40 meV higher. The DFT energies are mapped onto a classical Heisenberg model $E = -\\frac{1}{2}\\sum_{ij} J_{ij}\\,\\mathbf{u}_i\\cdot\\mathbf{u}_j$, whose fitted exchanges and the 8 meV anisotropy fix the coexistence of in-plane and out-of-plane ferromagnetic blocks. Experimental confirmation relies on micro-focused spin-ARPES at the Γ point measuring all three spin components, finding both in-plane and out-of-plane polarization consistent with the calculated ground state, plus polarization-dependent ARPES that benchmarks the DFT band structure.","core_discovery":"The central claim is that the magnetic ground state of Cr1+δTe2 (δ = 0.25–0.50) is a previously overlooked phase termed orthogonal ferromagnetism. In the DFT-relaxed structure, Cr moments in the CrTe2 layers are ferromagnetically aligned out of plane, while intercalated Cr moments are nearly in plane (an 83° cant relative to the c-axis), and the nearest layer Cr atoms tilt by 17°, creating local frustration. The paper reports an energy difference of 8 meV per unit cell between the out-of-plane layer spins and the in-plane intercalant spins, traced to spin-orbit coupling of Te. Exchange parameters extracted from a Heisenberg fit are J1 = −106 meV between intercalant and layer Cr, J2 = +8 meV for next-nearest neighbors, and Jint = 4 meV between neighboring intercalants, giving an effective antiferromagnetic interlayer coupling. The authors interpret the in-plane magnetization curves, which do not saturate and show a kink, as evidence of a double spin reorientation: domain alignment, then a spin-flop-like transition, then continuous rotation; they contrast this with earlier reports of a smooth reorientation. The same phenomenology is reported for Cr1.5Te2, indicating persistence up to 50% intercalation.","pith_inferences":["Whether the orthogonal order survives depends on how excess Cr atoms are arranged; disordered intercalants would likely restore canted or glassy behavior, which may be why earlier samples differed — this is my inference, not the paper's claim.","A bulk-sensitive probe such as neutron or resonant X-ray magnetic scattering on the same crystals could directly image the alternating layer moments; the paper does not report such a measurement.","The three equivalent in-plane easy directions suggest a multi-level switching landscape that could be exploited for multi-state magnetic memory, a direction the paper leaves implicit.","Because the spin-ARPES polarization is also consistent with a single-domain sample, a rastered spatial map of the spin components would discriminate the bulk alternating phase from single-domain saturation."],"forward_implications":["The long-standing disagreement between an out-of-plane ferromagnet and a canted ferromagnet description of Cr1.25Te2 is resolved in favor of alternating orthogonal layers.","The magnetization downturn that gradual-reorientation models could not explain is reinterpreted as the signature of an abrupt spin-flop-like transition.","Orthogonal ferromagnetism is achieved in a single chemical phase, without the interfaces required by crossed-magnet heterostructures.","The effect extends to at least 50% excess Cr, making Cr1+δTe2 a doping-tunable platform for spin-flop generation and orbitronic devices."],"supporting_citations":[{"why":"Earlier characterization of Cr1.25Te2 as a ferromagnet with out-of-plane polarization, the interpretation the paper contrasts with its orthogonal phase.","marker":"[17]"},{"why":"The canted-ferromagnet model and gradual spin-reorientation data that the new magnetization curves and DFT state are claimed to supersede.","marker":"[18]"},{"why":"Structural study linking the c-axis lattice parameter to Cr intercalation, used to establish the 1.25:2 stoichiometry of the crystals.","marker":"[24]"},{"why":"Raman characterization of trigonal Cr5Te8 used to corroborate stoichiometry and high crystallinity.","marker":"[25]"},{"why":"PEEM observations of magnetic domains in related CrTe2 flakes, cited to support the 'domains comparable to spot size' interpretation of spin-ARPES.","marker":"[26]"},{"why":"Electronic-structure code and projector-augmented wave method used for the DFT relaxations and non-collinear calculations.","marker":"[31]"},{"why":"Updated description of the same electronic-structure code used in the DFT workflow.","marker":"[32]"},{"why":"Band-unfolding procedure used to compare supercell DFT bands with the measured ARPES dispersions.","marker":"[33]"}],"fun_headline_variants":["Hidden orthogonal ferromagnet phase found in CrTe2","CrTe2 reveals abrupt spin-flop not gradual reorientation","Spin-flop-like transitions in CrTe2 hint at hidden phase","Abrupt spin-flop in CrTe2: hidden orthogonal phase"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The identification of the orthogonal phase assumes that the ordered 2×2 supercell with one intercalated Cr per cell used in DFT matches the real excess-Cr arrangement, and that the measured spin-ARPES polarization reflects the bulk alternating state rather than a single-domain sample; the authors themselves note 'likely magnetic domains comparable to the spot size'.","fun_headline_variants_meta":{"raw":{"variants":["Hidden orthogonal ferromagnet phase found in CrTe2","CrTe2 reveals abrupt spin-flop not gradual reorientation","Spin-flop-like transitions in CrTe2 hint at hidden phase","Abrupt spin-flop in CrTe2: hidden orthogonal phase"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000797,"raw_usage":{"total_tokens":3585,"prompt_tokens":1100,"completion_tokens":2485,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":716,"completion_tokens_details":{"reasoning_tokens":2422}},"tokens_in":716,"tokens_out":2485,"duration_ms":21371,"temperature":1.0,"reasoning_tokens":2422,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T16:31:41.396445+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform bulk-sensitive magnetic diffraction (neutron or resonant X-ray magnetic scattering) on the same Bridgman-grown Cr1.25Te2 crystals below the ordering temperature and look for reflections from a unit cell in which intercalant moments are transverse to layer moments; if only a single canted moment direction is found, or the intercalant moments are disordered, the orthogonal phase is not the ground state. On the same crystals, raster the micro-focused spin-ARPES beam: spin-polarization maps that flip sign between regions would confirm domains comparable to the spot size, while a uniform polarization would favor the single-domain reading.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier characterization of Cr1.25Te2 as a ferromagnet with out-of-plane polarization, the interpretation the paper contrasts with its orthogonal phase."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The canted-ferromagnet model and gradual spin-reorientation data that the new magnetization curves and DFT state are claimed to supersede."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Structural study linking the c-axis lattice parameter to Cr intercalation, used to establish the 1.25:2 stoichiometry of the crystals."},{"cited_title":"Purbawati, S","cited_arxiv_id":null,"evidence_quote":"Raman characterization of trigonal Cr5Te8 used to corroborate stoichiometry and high crystallinity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"PEEM observations of magnetic domains in related CrTe2 flakes, cited to support the 'domains comparable to spot size' interpretation of spin-ARPES."},{"cited_title":"Ikeda, K","cited_arxiv_id":null,"evidence_quote":"Electronic-structure code and projector-augmented wave method used for the DFT relaxations and non-collinear calculations."},{"cited_title":"Enkovaara, C","cited_arxiv_id":null,"evidence_quote":"Updated description of the same electronic-structure code used in the DFT workflow."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Band-unfolding procedure used to compare supercell DFT bands with the measured ARPES dispersions."}],"review_version":1}