{"id":"c2021c7d-0ece-4ace-a214-10da671242d4","arxiv_id":"2412.16421","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Unbiased Hartree-Fock and exact diagonalization find altermagnetic Mott insulating and metallic phases in the Lieb lattice Hubbard model with interactions only on the transition-metal sublattices.","lead":"Using computer simulations of a modified Lieb lattice Hubbard model, the authors find that repulsive electron interactions alone can produce altermagnetic Mott insulators at two different electron fillings, with spin-split bands and magnons. The result connects microscopic theory to the anti-CuO2 layered oxychalcogenides where d-wave altermagnetism was recently reported.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The n=2 altermagnetic Mott insulator is established only for an unjustified parameter regime (U_A=0, epsilon_A=10|t1|); without a microscopic estimate for real oxychalcogenides, the central claim overreaches.","rationale":"I agree with the reader's identification of the weakest assumption. The n=2 altermagnetic insulator is the most novel part of the claim (n=4 at epsilon_A=0 is less surprising given the bipartite geometry), and it is achieved only for a specific parameter regime (U_A=0, epsilon_A=10|t1|) that is neither derived from ab initio estimates nor tested for robustness. The reader's internal-consistency note about Fig. 3(e) is a separate, correctable issue: the representative point (U=5, t2=0.4, epsilon_A=10) appears to lie in the non-coplanar region of Fig. 3(b), so that figure should be recomputed or relabeled. But even if that figure is fixed, the parameterization concern remains. The paper's ED and spin-model results are supportive but, on clusters of 12-32 sites, cannot settle the finite-U_A question. Thus a CONDITIONAL verdict is appropriate: the model Hamiltonian as defined almost certainly supports altermagnetic Mott states, but the claim that this 'establishes' altermagnetism for oxychalcogenides requires a microscopic justification of U_A approximately 0 and epsilon_A=10|t1|.","tokens_in":14075,"tokens_out":9578,"duration_ms":75899,"concrete_test":"Perform unrestricted Hartree-Fock at n=2 for a representative altermagnet point (e.g., t2=-0.5, U=20, epsilon_A=10, following the methods of Sec. I in the SM) while varying U_A on the A sublattice through {0, U/4, U/2, U} and epsilon_A through {2, 4, 10}|t1|. If the lowest-energy state at any finite U_A or epsilon_A<10 ceases to be the collinear altermagnet and instead becomes the non-coplanar or ferromagnetic state found at epsilon_A=0, the n=2 central claim is not robust to physically plausible model parameters.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the modified Lieb lattice Hubbard model 'adapted to the anti-CuO2 structure' establishes altermagnetic Mott insulators at n=2 and n=4 rests on two undefended model choices: the Coulomb repulsion on the oxygen-like A sublattice is set to zero (U_A=0) and the n=2 altermagnet requires a large onsite energy epsilon_A=10|t1| to empty the A sublattice. The paper provides no microscopic estimate for U_A or the charge-transfer energy in a specific oxychalcogenide, and its own Fig. 3(a) shows that the n=2 altermagnet disappears at epsilon_A=0. Since U on oxygen 2p orbitals is generically finite (typically several eV), the n=2 phase may be an artifact of the fine-tuned parameter regime rather than a robust property of the material class. This is a gap between the model Hamiltonian and the material-relevance claim, and it directly undermines the 'establishes ... n=2' part of the central claim. The n=4 altermagnet, stable already at epsilon_A=0, is less exposed. A separate internal issue is that the representative point in Fig. 3(e) (U=5, t2=0.4, epsilon_A=10) appears to lie in the non-coplanar region of Fig. 3(b), so that figure needs correction; however, even after correcting that figure, the parameterization concern remains.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the Hubbard model (Eq. 1) on the Lieb lattice with Coulomb repulsion U restricted to the B and C sublattices, as a minimal model of the anti-CuO2 layers of quasi-2D oxychalcogenides. Using unrestricted Hartree-Fock on 12x12 systems with multiple magnetic unit cells and random initial conditions, together with exact diagonalization on 12- and 18-site clusters, the authors map t2-U phase diagrams at fillings n=2 and 4, identify collinear altermagnetic insulating regions with d-wave spin splitting, construct an effective checkerboard Heisenberg model whose magnon spectra show sublattice-resolved splitting, and report altermagnetic metallic phases upon doping away from n=2 and n=4.","tokens_in":14373,"tokens_out":7173,"duration_ms":64198,"significance":"If correct, the paper would provide one of the first unbiased interacting-model demonstrations of d-wave altermagnetism in a realistic lattice geometry, with concrete and falsifiable predictions: spin-split electron bands with d_{x^2-y^2} symmetry, sublattice-resolved magnon splitting observable by neutron scattering, and quasi-1D altermagnetic Fermi surfaces upon doping. The numerical cross-checks are a genuine strength: the Hartree-Fock calculations are unrestricted rather than ansatz-based, and the ED results provide correlated-system input. The main qualification is that the n=2 altermagnet is established only in a special parameter regime, namely epsilon_A=10|t1| and U_A=0, so the material-relevance claim in the conclusions is broader than the evidence currently supports.","major_comments":[{"comment":"The n=2 altermagnetic Mott insulator is demonstrated only for epsilon_A=10|t1| and with zero Coulomb repulsion on the A sublattice. The paper's own Fig. 3(a) shows that at epsilon_A=0 the n=2 ground state is a vortex or block state with no altermagnet, and no microscopic estimate is provided for epsilon_A or for the oxygen-like A-site Coulomb repulsion. Since real oxychalcogenides have finite Coulomb repulsion on oxygen 2p orbitals and moderate charge-transfer energies, the conclusion that the model 'establishes the presence of altermagnetic Mott insulators' at n=2 in this material class overreaches. Please provide a quantitative estimate of epsilon_A and U_A, or a calculation with finite U_A, and if that is not possible, restrict the material-level claim to n=4 and present n=2 as a conditional model prediction.","section":"Eq. (1), Fig. 3(a), Conclusions"},{"comment":"The representative band structure in Fig. 3(e) is computed at U=5, t2=0.4, epsilon_A=10, which lies in the t2>0 part of the phase diagram. According to Fig. 3(b) and the accompanying text, for t2>0 the ground state is either a non-coplanar insulator at weak and intermediate U or a saturated ferromagnet at large U, not the altermagnet. Either the phase diagram labels are incorrect or the band structure in Fig. 3(e) is not representative of the altermagnetic phase; this inconsistency must be corrected before the d-wave spin splitting shown in Fig. 3(e) can be used as evidence for the central claim.","section":"Fig. 3(b) and Fig. 3(e)"},{"comment":"The exact-diagonalization support for the n=2 altermagnet is obtained on a 12-site (2x2) cluster with periodic boundary conditions, and the momentum-resolved spin splitting delta_updown(k) is computed with a symmetry-breaking Zeeman field of 0.05 added on one corner site. This is a standard finite-size probe, but it is not an independent confirmation of spontaneous altermagnetic order: the same fine-tuned parameter set (epsilon_A=10, U_A=0) is used, and the authors explicitly acknowledge that the competing valence-bond-solid state cannot be ruled out on larger systems. Please clarify what the ED results add beyond showing collinear AF correlations, and consider a larger-scale calculation (for example DMRG or variational Monte Carlo) to substantiate the n=2 claim.","section":"Fig. 4(d) and Fig. 4(f)"}],"minor_comments":[{"comment":"The supplemental material is referenced only as '[40] Supplementary link', a placeholder; a permanent link or an appendix is needed for the reader to verify the Hartree-Fock, ED, and spin-wave details.","section":"Supplementary material"},{"comment":"There are several typographical and formatting errors: 'arrises' in the section on altermagnetic metal by doping, 'Physl Rev. Lett.' in Ref. [29], and 'subalttice' in SM Section III.","section":"Main text and references"},{"comment":"The color scale label 'm_s tilde' in Fig. 4(d) is not defined in the caption; please specify that it is the ED staggered magnetization defined in the text.","section":"Fig. 4(d) caption"},{"comment":"The estimates of J2/J1 are order-of-magnitude only, and the main text should state explicitly that the spin-model results are qualitative because J2/J1 is not derived quantitatively from the Hubbard parameters for the specific altermagnetic phases.","section":"SM Section IV"},{"comment":"The non-coplanar magnetic state is discussed only in the SM; a one-sentence definition in the main text would help readers interpret the 'Non-Coplanar (I)' regions in the phase diagrams.","section":"Fig. 2(a) and Fig. 3(b)"}],"recommendation":"major_revision","confidential_remarks":"The paper is potentially suitable for a rapid-communication journal if the parameterization and figure issues are resolved. The concurrent work on the standard Lieb lattice Hubbard model is acknowledged, but the paper's distinguishing choice, U_A=0, is precisely the choice that needs physical justification; the authors should engage with whether the oxygen-like A sublattice can realistically be treated as non-interacting. The n=4 altermagnet at epsilon_A=0 is less exposed and could support a revised claim, but the current abstract and conclusions claim more than the n=2 evidence supports."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on Kaushal and Franz (2412.16421). The genuinely new piece is that they treat the Lieb lattice Hubbard model with interacting B/C sites and a non-interacting A site, and show that altermagnetic order emerges spontaneously from the Coulomb interaction, rather than being put in by hand as in earlier static\\-moment studies. Both unrestricted Hartree\\-Fock and small\\-cluster ED point the same way, which is a real plus. The n=4 altermagnetic Mott insulator looks robust, and the doping study with altermagnetic metals is a useful addition. The effective spin model with altermagnetic magnon splitting is a nice extra.\n\nThat said, the central 'establishes n=2' claim is too strong. The n=2 altermagnet requires a large epsilon_A=10 and zero U on the A sublattice, with no microscopic estimate for real oxychalcogenides. The standard Lieb model with U on all sublattices is already covered by a concurrent paper, and their own Fig. 3(a) shows no n=2 altermagnet at epsilon_A=0. So the n=2 claim is conditional on a parameter regime that may or may not be physical.\n\nThere's also a concrete figure error. The representative band structure in Fig. 3(e) is at U=5, t2=0.4, epsilon_A=10, but their phase diagram Fig. 3(b) places that point in the non-coplanar region, not the altermagnet. The text says the altermagnet stabilises for t2<0, so either the figure or the phase diagram is mislabeled. A referee should catch this.\n\nA subtlety not in the reader's report: the effective spin model assumes J2/J1 > 1, and their own strong\\-coupling estimates suggest for n=2 and t2<0 the sign of J2/J1 could go the other way. So the magnon\\-splitting calculation may not be describing the same parameter region as the Hubbard\\-model altermagnet. Worth addressing.\n\nNo code or data, which would help but is not fatal. The ED clusters are very small, though they are used as a consistency check.\n\nOverall: a competent, useful paper that deserves refereeing, but the n=2 claim needs an honest reset\\-\\-either material\\-specific parameter justification or a softer conclusion. I'd send it to review with the expectation of revision.","headline":"Solid numerical study of interaction-driven altermagnetism on the Lieb lattice, but the n=2 claim rests on fine-tuned parameters and a figure error; still worth refereeing.","tokens_in":14941,"tokens_out":5660,"would_cite":true,"duration_ms":44052,"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":"This paper establishes that a modified Lieb lattice Hubbard model, with repulsion only on the transition-metal sublattices, hosts altermagnetic Mott insulators at electron densities n=2 and n=4, with d-wave spin splitting, and that doping…","keywords":["altermagnetism","Lieb lattice","Hubbard model","Mott insulator","Hartree-Fock","exact diagonalization","spin splitting","oxychalcogenides"],"falsifier":"Solving the same Hubbard model with a finite Coulomb repulsion U_A on the A sublattice (for example U_A = U) and checking whether the staggered magnetization and the d-wave spin splitting at n=2 persist; if they vanish, the central n=2 claim is falsified. The paper's own Fig. 3(a) indicates that at εA=0 the n=2 altermagnet is absent, so varying U_A or εA is the direct test.","tokens_in":13828,"feed_emoji":"🧲","tokens_out":3808,"duration_ms":33387,"temperature":0.7,"pith_summary":"The paper claims that repulsive electron interactions alone, modeled by a Hubbard term placed only on the two transition-metal sublattices of the Lieb lattice, stabilize altermagnetic Mott insulators at two filling values. Altermagnets are collinear antiferromagnets with zero net magnetization but spin-split electron bands, a property that is promising for spintronics. The authors support this with unrestricted Hartree-Fock phase diagrams and exact diagonalization on small clusters, and they also find altermagnetic metals upon doping the Mott states. If correct, this gives a minimal interacting model that produces altermagnetism without assuming magnetic order by hand, on a lattice geometry that mirrors real quasi-2D oxychalcogenides.","feed_headline":"Hubbard model yields altermagnetic Mott insulators at two fillings","feed_subtitle":"A minimal Lieb-lattice model with repulsion only on magnetic sites produces d-wave spin-split magnets and doped metals.","key_machinery":"The central object is the modified Lieb lattice Hubbard Hamiltonian H = t1 Σ_{⟨i,j⟩,σ} c†_{i,σ} c_{j,σ} + t2 Σ_{⟨⟨i,j⟩⟩,σ} c†_{i,σ} c_{j,σ} + U Σ_{i∈{B,C}} n_{i,↑} n_{i,↓} + εA Σ_{i∈A} n_i, with U only on the B and C sublattices that host magnetic atoms, while the A sublattice is non-magnetic. The argument is carried by unrestricted Hartree-Fock calculations that test multiple magnetic unit cells and initial conditions, corroborated by exact diagonalization on 12- and 18-site clusters, and by an effective spin-1/2 checkerboard Heisenberg model whose spin-wave spectra reproduce the altermagnetic splitting.","core_discovery":"For the Lieb lattice with Hubbard U on sublattices B and C only, and with an onsite energy εA on sublattice A, the paper finds collinear antiferromagnetic states with zero net magnetization and d_{$x^{2}$-$y^{2}$}-wave spin-split bands at average electron densities n=2 and n=4, over broad ranges of U and next-nearest-neighbor hopping t2. These are identified as altermagnetic Mott insulators. Doping them produces metallic states with d-wave spin splitting, including quasi-one-dimensional electron pockets when electron-doping the n=2 insulator with εA large. The paper also constructs an effective spin-1/2 Heisenberg model on a checkerboard lattice and shows that its magnon spectra exhibit the same altermagnetic splitting.","pith_inferences":["The n=2 altermagnetic phase depends on the A sites being effectively non-interacting (U_A ≈ 0) and on a large εA; if real oxygen p orbitals carry a comparable Hubbard U, the n=2 phase is likely suppressed, so the model is most securely applied to systems where the A sublattice is genuinely itinerant and weakly correlated.","The altermagnetic metals found here might be a platform for correlation-driven superconductivity, since doped Mott insulators on bipartite lattices are known to host pairing instabilities; the paper mentions this as a future direction but does not compute it.","The quasi-one-dimensional electron pockets in the electron-doped n=2 altermagnet suggest that transport and spin-splitting will be strongly anisotropic, which could be tested by measuring resistivity or ARPES along the x and y directions.","The Hartree-Fock tendency to overestimate ferromagnetic order, noted in the text, implies that the altermagnetic metal phase may actually occupy a larger portion of the phase diagram than shown."],"forward_implications":["The n=2 and n=4 altermagnetic Mott insulators are stable against quantum fluctuations, as the exact diagonalization results show staggered magnetization in regions where Hartree-Fock predicts the altermagnet.","The altermagnetic band splitting is d_{x^2-y^2}-wave with nodal lines along k_x = ±k_y, protected by the C4 rotational symmetry, which makes the splitting observable with angle-resolved photoemission.","Doping the n=2 insulator with electrons and U > εA produces altermagnetic metals with quasi-one-dimensional Fermi-surface pockets, which can be detected by quantum oscillations or ARPES.","The effective spin model predicts that the sublattice-resolved dynamical spin structure factor shows altermagnetic magnon splitting, measurable by inelastic neutron scattering.","The results apply to quasi-2D oxychalcogenides with the anti-CuO2 structure and d1 electronic configuration, providing a minimal model for their altermagnetic behavior."],"supporting_citations":[{"why":"Define the altermagnetic state as a collinear magnet with zero net magnetization and non-relativistic spin splitting, establishing the phenomenon the paper aims to realize.","marker":"[1–4]"},{"why":"Specify the symmetry condition for altermagnetism (rotation or mirror relation between opposite-spin sites) that underlies the d-wave splitting in the Lieb lattice.","marker":"[5, 6]"},{"why":"Propose La2O3Mn2Se2 as a quasi-2D d-wave altermagnetic insulator, providing the material motivation for studying the anti-CuO2 structure.","marker":"[19, 20]"},{"why":"Report experimental evidence for room-temperature d-wave altermagnetism in oxychalcogenides KV2Se2O and Rb1-δV2Te2O, supporting the relevance of the model.","marker":"[24, 25]"},{"why":"Show that a spin-fermion model on the Lieb lattice with static moments gives d_{x^2-y^2}-wave band splitting, the starting point that this paper extends by making magnetism emerge from the Hubbard interaction.","marker":"[26, 27]"},{"why":"Establishes the particle-hole symmetry of the conventional Lieb lattice, which the modified model breaks, justifying the separate treatment of n=2 and n=4.","marker":"[35]"},{"why":"Concurrent study of the standard Lieb lattice with U on all sublattices, whose difference from this work (no Coulomb repulsion on sublattice A) frames the key modeling choice.","marker":"[51]"}],"fun_headline_variants":["Lieb lattice Hubbard model yields altermagnetic Mott insulators","Two fillings of Lieb lattice Hubbard model give altermagnetic Mott states","Doping Lieb lattice altermagnet produces d-wave spin-split metals","Altermagnetic Mott insulators and metals from modified Lieb lattice"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model assumes that the A sublattice (oxygen sites) has no Hubbard repulsion and, for the n=2 altermagnet, that its onsite energy is very large, so the A sites act as a passive reservoir; if real oxygen orbitals have a comparable U_A or a smaller εA, the n=2 altermagnetic insulator does not survive.","fun_headline_variants_meta":{"raw":{"variants":["Lieb lattice Hubbard model yields altermagnetic Mott insulators","Two fillings of Lieb lattice Hubbard model give altermagnetic Mott states","Doping Lieb lattice altermagnet produces d-wave spin-split metals","Altermagnetic Mott insulators and metals from modified Lieb lattice"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000537,"raw_usage":{"total_tokens":2541,"prompt_tokens":871,"completion_tokens":1670,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":487,"completion_tokens_details":{"reasoning_tokens":1592}},"tokens_in":487,"tokens_out":1670,"duration_ms":10209,"temperature":1.0,"reasoning_tokens":1592,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:38:33.448908+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solving the same Hubbard model with a finite Coulomb repulsion U_A on the A sublattice (for example U_A = U) and checking whether the staggered magnetization and the d-wave spin splitting at n=2 persist; if they vanish, the central n=2 claim is falsified. The paper's own Fig. 3(a) indicates that at εA=0 the n=2 altermagnet is absent, so varying U_A or εA is the direct test.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the particle-hole symmetry of the conventional Lieb lattice, which the modified model breaks, justifying the separate treatment of n=2 and n=4."}],"review_version":1}