{"id":"4176fb18-87e5-4612-a1fc-45d6a38d2f03","arxiv_id":"2507.12944","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"DFT+DMFT predicts 64 symmetry-related Weyl nodes in CeRu4Sn6, with the nearest node 0.5 meV below the Fermi level.","lead":"This paper uses a combined density-functional and dynamical mean-field calculation to predict that the heavy-fermion compound CeRu4Sn6 hosts 64 Weyl nodes near its Fermi energy, the closest only 0.5 meV below it. If correct, this makes CeRu4Sn6 a concrete platform for studying how strong electron correlations produce topological semimetal behavior.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's own MaxEnt glitch admission (Sec. III A) undermines the ReΣ(0) used in Hqp; since the closest Weyl node sits only 0.5 meV below the Fermi level, an uncompensated meV-level error can shift or destroy the central claim.","rationale":"The central claim is that DFT+DMFT predicts Weyl nodes in CeRu4Sn6, with the most experimentally relevant node only 0.5 meV below the Fermi level. For this claim to hold, the quasiparticle Hamiltonian Hqp(k) built from Eq. (2) must correctly capture the low-energy band structure. Hqp relies on Z and ReΣ(0), which are extracted from the maximum-entropy continuation of the DMFT self-energy. The paper itself states in Sec. III A that the mj=±1/2 component has numerical glitches near ω=0 arising from analytic continuation. Since node III is a Ce 4f5/2-dominated node at -0.5 meV, any meV-scale error in ReΣ(0) for that component directly changes the node energy; the paper gives no uncertainty estimate. This is not a disagreement with consensus but an internal correctness risk: the authors' own text flags the weak link. The quantum criticality question raised in the conclusion is a real limitation of the DMFT quasiparticle picture, but the authors explicitly acknowledge it, and it affects the low-temperature physical interpretation rather than the internal consistency of the calculation. The proposed check, a Matsubara-based low-frequency fit of ReΣ(0) that does not rely on MaxEnt, would settle whether the node at -0.5 meV survives. This is exactly the same weakest assumption identified by the reader, so the verdict should remain conditional until such a test is performed.","tokens_in":13761,"tokens_out":6314,"duration_ms":65539,"concrete_test":"Recompute ReΣ(0) and Z for mj=±1/2 from the stored Matsubara self-energy by fitting the lowest 20 frequencies to ReΣ(iω_n) = a + b ω_n^2, which avoids MaxEnt near ω=0, and rebuild Hqp(k) with the new values. Then rerun the Weyl-node search, or at least the band crossing near node III. If node III moves by more than about 1 meV, changes its sign relative to the Fermi level, or disappears, the headline prediction is not robust to the analytic-continuation artifact the paper itself flags.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III A reports for the mj=±1/2 component \"minor glitches near ω = 0, which likely arise from numerical artifacts in the analytic continuation.\" This is the same component whose self-energy determines the quasiparticle parameters used in Eq. (2). Table I gives ReΣ(0)=1.013 eV for mj=±1/2 at 23 K, and the direct Kondo gap is only about 5 meV. The closest Weyl node (node III in Table II) is just 0.5 meV below the Fermi level and carries 77% Ce 4f5/2 spectral weight, so it is precisely the mj=±1/2 quasiparticle band crossing a conduction band. Any error in ReΣ(0) extracted from the glitchy continuation directly shifts the f-level in Hqp(k), moving the node in energy. An error of a few meV would move node III across the Fermi level or, if the crossing is marginal, could annihilate it; even a sub-meV error makes the \"0.5 meV\" locating statement unquantified. The authors provide no error bar on ReΣ(0) and no sensitivity test of node positions to this quantity.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a DFT+DMFT study of the heavy-fermion compound CeRu4Sn6. The authors construct a 90-orbital tight-binding model from DFT, solve the correlated Ce 4f5/2 problem with DMFT at several temperatures, extract quasiparticle parameters Z and ReΣ(0) from the DMFT self-energy, and use the resulting renormalized quasiparticle Hamiltonian to search for Weyl nodes via Berry-curvature monopoles. They report five inequivalent Weyl nodes (64 in total after accounting for D2d symmetry and time reversal), with node energies between -3.8 meV and +22.1 meV relative to the Fermi level, including one node only 0.5 meV below the Fermi level. The nodes are interpreted as bridging the direct Kondo-insulating gap, and the reduced Weyl velocities are argued to be consistent with the experimentally observed nonlinear Hall effect.","tokens_in":14056,"tokens_out":6675,"duration_ms":79092,"significance":"If the central result holds, this would be a valuable step beyond previous DFT+Gutzwiller treatments: it would provide a DFT+DMFT-based identification of Weyl nodes in CeRu4Sn6 and connect the Kondo hybridization directly to the low-energy topology. The calculation has notable strengths: the Weyl-node positions are an emergent output rather than a fit to experiment, the chemical potential is set self-consistently, the symmetry analysis of the 64 nodes is explicit, and the use of publicly available DMFT and analytic-continuation codes supports reproducibility. The predicted very small node energy (-0.5 meV) and low Weyl velocities are falsifiable statements that can be confronted with transport and photoemission experiments. However, the quantitative precision of the main claim is not yet established, because the quasiparticle Hamiltonian that determines the node positions is built from a self-energy whose analytic continuation is explicitly reported to have glitches near ω=0 in the very mj=±1/2 channel used for the Weyl search.","major_comments":[{"comment":"The manuscript states in Section III A that the mj=±1/2 component of the self-energy exhibits \"minor glitches near ω = 0, which likely arise from numerical artifacts in the analytic continuation.\" This is the same component whose ReΣ(0)=1.013 eV and Z are inserted into Eq. (2) to build Hqp(k), and the closest Weyl node (node III in Table II) lies only 0.5 meV below the Fermi level. Because ReΣ(0) enters Eq. (2) linearly, an unresolved uncertainty of even 1 meV in the analytic continuation shifts the f-level and hence the node energies by a comparable amount, which is enough to move node III across the Fermi level or qualitatively change the central claim. The paper provides no error bar on ReΣ(0) or Z and no sensitivity test. Please add such a test, for example by varying ReΣ(0) and Z within the plausible MaxEnt uncertainty and reporting the resulting changes in Table II, or by cross-checking ReΣ(ω) with an independent analytic-continuation method such as Padé.","section":"Section III A/B, Eq. (2), Table I/II"},{"comment":"The Weyl-node search is performed entirely within the linearized quasiparticle Hamiltonian Hqp(k), whose parameters Z and ReΣ(0) are evaluated at ω=0 only. The self-energy has strong frequency dependence with poles on the scale of an eV (Fig. 5), while the Weyl nodes span a window of roughly ±22 meV. Figure 9 validates the quasiparticle bands against the DMFT spectral function for only two of the five inequivalent nodes. The authors should quantify the error of the quasiparticle approximation at the momenta of all five nodes, or include the next-order frequency dependence of ReΣ in Eq. (2), to ensure that the nodes and their type-I/type-II classification are not artifacts of the linearization.","section":"Section III B, Eq. (2), Fig. 9"},{"comment":"Node III has an energy of only -0.5 meV and a minimum distance of |Δkmin|=0.0021 Å^-1 between symmetry-related nodes of opposite Chern number. These are extremely small scales compared with the direct gap of about 5 meV and the input uncertainties of the calculation. The manuscript should demonstrate that node III survives under small variations of the Hubbard U, the double-counting correction, and the chemical potential, and should report how the node energy changes in such a robustness check. Without this, the \"0.5 meV\" statement is not yet supported at the stated precision.","section":"Table II, Section III B"}],"minor_comments":[{"comment":"The notation \"mj = (−)1/2\" is ambiguous: it should be clarified whether the Weyl search uses both mj=+1/2 and mj=-1/2 (which are degenerate in a paramagnetic calculation) or only one projection, and if only one, why this is sufficient.","section":"Section III B"},{"comment":"The text calls nodes I and IV \"high symmetry points,\" but their fractional coordinates are not obviously high-symmetry points of the Brillouin zone; this phrase should be rephrased, e.g., as points whose symmetry orbit is reduced because they are invariant under a combined symmetry operation.","section":"Section III B / Table II"},{"comment":"The paper uses the phrase \"ab initio results\" while the Hubbard U=5.5 eV, Hund's coupling J=0, and the double-counting correction VDC are taken from previous literature/Anisimov formula; the abstract could be more precise by saying \"DFT+DMFT with a literature value of U\" to avoid overclaiming a fully parameter-free calculation.","section":"Abstract / Section II A"},{"comment":"The color scale in Fig. 8 is normalized as An(k)=A(k,ω)/max(A(k,ω)), but the color bar label should state that these are normalized spectral weights; currently the units and normalization are clear only from the caption.","section":"Fig. 8"},{"comment":"In Eq. (2), the role of the term −µI−VDC should be stated explicitly in the main text: it should be clear whether H(k) already includes the chemical potential or whether µI is subtracted to define the band structure relative to the DMFT chemical potential.","section":"Eq. (2)"},{"comment":"No data or code availability statement is provided; for reproducibility, it would be helpful to make available the tight-binding parameterization, the DMFT input files, and the Weyl-search settings.","section":"Data and code availability"}],"recommendation":"major_revision","confidential_remarks":"The main reason for the major revision is the unresolved quantitative uncertainty in ReΣ(0) for the mj=±1/2 channel, which enters directly into the quasiparticle Hamiltonian that produces the closest Weyl node. This is not a problem of circularity or fitting; it is a precision problem that the authors can address with sensitivity tests and an independent analytic-continuation check. I would not require the authors to prove that the nodes are stable to arbitrarily large parameter variations, but they should provide enough quantitative information to support the 0.5 meV claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plainly: this is a well-executed DFT+DMFT study that produces a concrete, falsifiable prediction of 64 Weyl nodes in CeRu4Sn6, and it is the first full DMFT inventory for this material. The main thing you should know is that the closest node sits only 0.5 meV below the Fermi level, and the paper itself flags numeric glitches in the analytic continuation of the mj=±1/2 self-energy — the very object used to build the quasiparticle Hamiltonian. There are no error bars on ReΣ(0) and no sensitivity test of node positions. That is a legitimate soft spot, not a nitpick.\n\nWhat is genuinely new: the complete set of five inequivalent nodes with coordinates, types, velocities, and Ce-4f spectral weights; the demonstration that the direct Kondo gap is bridged by these nodes; and the stark velocity reduction compared to weakly correlated Weyl semimetals. The method is appropriate, the open-source codes are standard, and the calculation is not circular — nodes are computed, not fitted. This is a clear step beyond the earlier DFT+Gutzwiller prediction.\n\nThe weakness is real but not fatal. The admitted glitches near ω=0 in the mj=±1/2 channel matter because that channel carries 77% of the spectral weight at the closest node. An uncompensated error of just a few meV in ReΣ(0) could shift or annihilate that node. The authors need to quantify this sensitivity. The temperature mismatch is a second, softer issue: the Weyl search is done at 23 K in a Fermi-liquid description, while the experiment sees quantum criticality below 10 K. The authors note this and do not overclaim, but it leaves open whether the zero-temperature topology is the one computed.\n\nMinor: the tight-binding basis and the static treatment of the f7/2 manifold are standard but not tested for their effect on the node energies.\n\nWho this is for: researchers working on correlated topology, heavy-fermion semimetals, or DMFT-based topological prediction. It deserves peer review. If I were refereeing, I would request an error analysis on the analytic continuation, a sensitivity check of node energies to the glitch, and a brief discussion of extrapolation to T→0. The central claim is plausible and the paper is an honest, reproducible advance.","headline":"A credible DFT+DMFT prediction of 64 Weyl nodes in CeRu4Sn6, but the meV-scale proximity of the closest node demands an error analysis the paper does not provide.","tokens_in":14617,"tokens_out":3365,"would_cite":true,"duration_ms":37720,"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":"A DFT+DMFT calculation finds 64 Weyl nodes in CeRu4Sn6, the closest only 0.5 meV below the Fermi level, establishing the compound as a correlated Weyl semimetal.","keywords":["CeRu4Sn6","Weyl semimetal","Kondo insulator","heavy fermion","DFT+DMFT","Berry curvature","nonlinear Hall effect","topological band crossings"],"falsifier":"Recompute $H_{\\rm qp}(\\mathbf{k})$ with $\\mathrm{Re}\\,\\Sigma(0)$ for $m_j = \\pm 1/2$ shifted by the uncertainty range of the maximum-entropy continuation and re-run the Berry-curvature search: if the node at fractional coordinates [0.244, -0.127, 0.060], 0.5 meV below the Fermi level, disappears, the central claim fails.","tokens_in":13571,"feed_emoji":"🌀","tokens_out":9067,"duration_ms":95671,"temperature":0.7,"pith_summary":"This paper asks whether the heavy-fermion compound CeRu4Sn6 is a topological Weyl semimetal, and answers yes on the basis of a correlated electronic-structure calculation. The authors combine density functional theory with dynamical mean-field theory (DFT+DMFT), the standard tool when electron correlations are strong, and they track the Berry curvature through the renormalized quasiparticle bands. They find five inequivalent Weyl nodes (64 symmetry-related copies in total), both type I and type II, located inside the Kondo-hybridization gap, with the closest node only 0.5 meV below the Fermi level. If this is right, the compound becomes a concrete material platform where the Kondo effect itself generates topological band crossings, which is what experiments on the nonlinear Hall effect and specific heat had suggested.","feed_headline":"CeRu4Sn6 is a correlated Weyl semimetal with 64 nodes","feed_subtitle":"Closest Weyl point lies just 0.5 meV below the Fermi level; Kondo correlations, not bare bands, create the topology.","key_machinery":"The load-bearing construction is the renormalized quasiparticle Hamiltonian $H_{\\rm qp}(\\mathbf{k}) = \\sqrt{Z}\\,(H(\\mathbf{k}) + \\mathrm{Re}\\,\\Sigma(0) - \\mu I - V_{\\mathrm{DC}})\\sqrt{Z}$, built from the DFT tight-binding Hamiltonian $H(\\mathbf{k})$ and the DMFT self-energy $\\Sigma(\\omega)$ for the Ce $4f_{5/2}$ orbitals. The quasiparticle weight $Z = [1 - \\partial \\mathrm{Re}\\,\\Sigma(\\omega)/\\partial\\omega|_{\\omega=0}]^{-1}$ and the zero-frequency real part of the self-energy carry all the correlation renormalization; the search algorithm then follows the Berry curvature field lines of this Hamiltonian to points where the integrated Chern number is $\\pm 1$, identifying each Weyl node and its topological charge. The direct but not indirect Kondo gap of about 5 meV is what the nodes bridge.","core_discovery":"The central claim is that CeRu4Sn6 is a strongly correlated Weyl semimetal in which the Weyl nodes are a product of correlation physics, not of the bare band structure. Treating the Ce 4f electrons with DMFT at 23 K produces a direct Kondo gap of about 5 meV across the Brillouin zone, but the gap is bridged by 64 Weyl nodes that arise from the hybridization of the correlated f states with conduction bands. Five of these nodes are inequivalent; two sit at high-symmetry positions and have eight replicas each, three have sixteen each, and the Chern numbers balance at +32 and -32. The closest node lies only 0.5 meV (about 6 K) below the Fermi level, and the geometric-mean Weyl velocities are one to two orders of magnitude smaller than in weakly correlated Weyl semimetals. The paper argues these features explain the low-temperature spontaneous nonlinear Hall effect and the $T^{3}$ specific-heat contribution observed in experiments, and they position CeRu4Sn6 as a model for studying interaction-driven band topology.","pith_inferences":["Because the two closest opposite-charge nodes (node III) sit only $|\\Delta k_{\\min}| = 0.0021$ Å$^{-1}$ apart, small strain or disorder could pair-annihilate them; a strain-dependent calculation would give a testable prediction for how the low-temperature Hall signal degrades with sample quality.","Sweeping the Coulomb interaction $U$ around the chosen 5.5 eV would show how robust the 0.5 meV node is; if its energy crosses zero under a modest change, CeRu4Sn6 would sit near a doping- or pressure-driven topological transition.","The same approach could be applied to other 'failed' Kondo insulators, with $\\mathrm{Re}\\,\\Sigma(0)$ near zero frequency as the controlling quantity; wherever the analytic continuation is trustworthy, similar correlation-generated Weyl nodes should appear.","A calculation beyond DMFT that includes non-local correlations could test whether the predicted Fermi-liquid Weyl nodes survive the Kondo-destruction quantum criticality hinted at by neutron-scattering $\\omega/T$ scaling below 10 K."],"forward_implications":["CeRu4Sn6 becomes a concrete, ab initio example of a correlated Weyl semimetal, giving a materials-specific origin for the observed spontaneous nonlinear Hall effect.","The smallest energy scale, 0.5 meV between the closest Weyl node and the Fermi level, sets a low-temperature scale of about 6 K below which topological transport should be most visible.","The calculated Weyl velocities, roughly 7 to 51 km/s, are orders of magnitude below those of noninteracting Weyl semimetals, which would show up as enhanced low-temperature specific heat and amplified nonlinear responses.","The DMFT picture refines the 'failed Kondo insulator' description: the material has a direct hybridization gap but no indirect gap, because both trivial bands and Weyl nodes cross the Fermi level.","Because the Weyl nodes are not pinned to the Fermi energy by filling constraints, their energy offsets (from -3.8 to +22.1 meV relative to the Fermi level) predict a temperature- or doping-dependent evolution of the transport signatures."],"supporting_citations":[{"why":"Supplies the crystal structure and space group ($I\\bar{4}2m$) on which all band-structure and symmetry calculations rest.","marker":"[14]"},{"why":"Experimental NMR evidence that CeRu4Sn6 forms a Kondo insulating gap, the state that the Weyl nodes bridge.","marker":"[17]"},{"why":"Previous DFT+Gutzwiller study that first proposed CeRu4Sn6 as a correlated Weyl semimetal and provides the comparison baseline.","marker":"[18]"},{"why":"Experimental report of the spontaneous nonlinear Hall effect and $T^3$ specific-heat contribution that motivated the search and supports the interpretation.","marker":"[20]"},{"why":"Earlier DFT+DMFT study of the same compound using a different quantum Monte Carlo solver; this work extends it to lower temperatures.","marker":"[29]"},{"why":"The Weyl-node search algorithm that locates nodes by following Berry curvature field lines, the method used to find the 64 nodes.","marker":"[38]"},{"why":"Demonstration of the same DMFT-based Weyl-node search in Ce3Bi4Pd3, validating the approach for heavy-fermion systems.","marker":"[39]"},{"why":"The Weyl-Kondo semimetal model that defines the expected phase and connects Kondo hybridization to Weyl nodes near the Fermi level.","marker":"[11, 12]"}],"fun_headline_variants":["CeRu4Sn6: 64 Weyl nodes from strong correlations","Correlated Weyl semimetal CeRu4Sn6 has 64 nodes","Weyl nodes bridge Kondo gap in CeRu4Sn6","Heavy-fermion Weyl semimetal with 64 nodes","CeRu4Sn6: Weyl nodes just 0.5 meV below Fermi"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the analytic continuation of the DMFT self-energy being reliable near zero frequency: the paper states that the $m_j = \\pm 1/2$ component has numerical glitches near $\\omega = 0$, and the closest Weyl node lies only 0.5 meV below the Fermi level, so a small error in $\\mathrm{Re}\\,\\Sigma(0)$ could shift that node or remove it.","fun_headline_variants_meta":{"raw":{"variants":["CeRu4Sn6: 64 Weyl nodes from strong correlations","Correlated Weyl semimetal CeRu4Sn6 has 64 nodes","Weyl nodes bridge Kondo gap in CeRu4Sn6","Heavy-fermion Weyl semimetal with 64 nodes","CeRu4Sn6: Weyl nodes just 0.5 meV below Fermi"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000215,"raw_usage":{"total_tokens":1446,"prompt_tokens":981,"completion_tokens":465,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":597,"completion_tokens_details":{"reasoning_tokens":363}},"tokens_in":597,"tokens_out":465,"duration_ms":4535,"temperature":1.0,"reasoning_tokens":363,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:34:30.529280+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute $H_{\\rm qp}(\\mathbf{k})$ with $\\mathrm{Re}\\,\\Sigma(0)$ for $m_j = \\pm 1/2$ shifted by the uncertainty range of the maximum-entropy continuation and re-run the Berry-curvature search: if the node at fractional coordinates [0.244, -0.127, 0.060], 0.5 meV below the Fermi level, disappears, the central claim fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the crystal structure and space group ($I\\bar{4}2m$) on which all band-structure and symmetry calculations rest."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Experimental NMR evidence that CeRu4Sn6 forms a Kondo insulating gap, the state that the Weyl nodes bridge."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Previous DFT+Gutzwiller study that first proposed CeRu4Sn6 as a correlated Weyl semimetal and provides the comparison baseline."},{"cited_title":"Emergent Topological Semimetal","cited_arxiv_id":"2404.15924","evidence_quote":"Experimental report of the spontaneous nonlinear Hall effect and $T^3$ specific-heat contribution that motivated the search and supports the interpretation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Weyl-node search algorithm that locates nodes by following Berry curvature field lines, the method used to find the 64 nodes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstration of the same DMFT-based Weyl-node search in Ce3Bi4Pd3, validating the approach for heavy-fermion systems."}],"review_version":1}