{"id":"d26fcc21-c871-4ab7-b2be-7938d6c6227d","arxiv_id":"2505.01511","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"YAsS and SmAsS are predicted to be topological crystalline insulators with glide-protected hourglass fermions, and SmAsS is a candidate heavy-fermion material whose surface states should survive Kondo correlations.","lead":"This paper re-examines the crystal structures of the RAsS compounds, identifies YAsS and SmAsS as topological insulators with hourglass surface states, and argues that SmAsS's surface states survive strong electron correlations, making it a heavy-fermion topological material. The corrected structures and the heavy-fermion prediction give experimentalists a new target for testing how topology behaves in strongly correlated materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Heavy-fermion claim rests on an undocumented Sommerfeld coefficient: SmAsS orders antiferromagnetically at 7.5 K, and no C/T extraction is shown, so γ=160 mJ mol⁻¹ K⁻² may not be genuine electronic specific heat.","rationale":"The reader's weakest assumption was the Pnma structural assignment; that concern is real but partially mitigated by the presented high-resolution synchrotron data and detailed refinement tables. The more fragile link is the experimental mass-enhancement evidence. The abstract's novelty is heavy-fermion physics, and the only quantitative support is an undocumented γ value in a magnetically ordered compound. The Anderson model is not fit to data and cannot independently establish heavy-fermion behavior; it only shows that a chosen set of parameters preserves the topological invariants. The topology itself—YAsS and SmAsS as Z4=2 TCIs—is standard and internally consistent, so rejection is not warranted; however, the SmAsS-specific heavy-fermion claim should be explicitly conditioned on a documented electronic specific heat. This is consistent with the reader's CONDITIONAL verdict, so no change in verdict is needed.","tokens_in":18995,"tokens_out":4317,"duration_ms":47849,"concrete_test":"Re-analyze the raw specific heat of SmAsS: plot C/T versus T² for H=0 and H=9 T, subtract the YAsS lattice contribution and a nuclear-Schottky term C_N = A/T², and report the T→0 intercept with an uncertainty estimate. Also state whether TN is suppressed by 9 T and whether γ is extracted above or below TN. If the intercept is not stable under these subtractions, γ=160 mJ mol⁻¹ K⁻² cannot support the heavy-fermion classification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central experimental claim—SmAsS shows 'emergent heavy-fermion physics' and belongs with SmB6 and YbB12—rests on a single quoted Sommerfeld coefficient, γ=160 mJ mol⁻¹ Sm K⁻², stated in Sec. II and repeated in the Conclusions. The Methods (Sec. VII B) report only that specific heat was measured from 0.4 K to 100 K at H=0 and 9 T; no C/T versus T² plot, no lattice subtraction using the nonmagnetic YAsS analog, no nuclear-Schottky correction, no error bar, and no statement of which field or temperature interval was used. SmAsS orders antiferromagnetically at TN≈7.5 K (Fig. 1d). Below TN, C/T can receive non-electronic contributions from AFM spin waves and nuclear Schottky tails, so the quoted γ may not be a genuine linear electronic term. The phrase 'in the normal state' is also ambiguous because the measurements extend well below TN. In parallel, the entire Z4=2/hourglass analysis assumes a paramagnetic TRS-symmetric state; the paper never states that the surface states refer to the paramagnetic regime or what happens below TN. If γ is not robust, the SmAsS-specific heavy-fermion conclusion loses its experimental anchor and the paper reduces to a DFT/TQC prediction plus a parameterized Anderson-lattice model.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a combined experimental and computational study of the RAsS family (R = Y, La, Sm). Using synchrotron powder and single-crystal diffraction, the authors re-determine the crystal structures: YAsS and SmAsS crystallize in the orthorhombic space group Pnma (SmAsS type), while LaAsS adopts a √2×√2 in-plane modulation (GdPS type). From DFT and topological quantum chemistry, they show that YAsS and SmAsS have Z4=2 symmetry indicators and predict glide-protected hourglass fermions on the (0-11) surface, whereas LaAsS is near a topological transition. A minimal two-orbital tight-binding model reproduces the topological band structure, and an Anderson-lattice model with strongly correlated Sm 4f electrons indicates that the hourglass surface states persist in the correlated regime, shifted in energy. Resistivity and specific-heat measurements are reported; SmAsS is claimed to show a large Sommerfeld coefficient γ=160 mJ mol−1 K−2, which the authors interpret as evidence of heavy-fermion physics.","tokens_in":19263,"tokens_out":5333,"duration_ms":55131,"significance":"If the structural assignment and the specific-heat analysis hold, this family would constitute a new platform for correlated topological crystalline insulators, and the demonstration that hourglass surface states survive in a saddle-point Anderson-lattice treatment would be a valuable step beyond weakly correlated TCI classification. The paper makes good use of standard symmetry-indicator tools (Vasp2trace, Bilbao server, WannierTools) and the minimal model is constructed transparently from Wannier functions, making the symmetry-indicator part reproducible. The prediction that LaAsS is close to a topological transition is a concrete, falsifiable statement. The central limitations are that the heavy-fermion claim rests on a single undocumented specific-heat value and that the topological surface-state calculation assumes a paramagnetic, time-reversal-symmetric state for a compound that orders antiferromagnetically at 7.5 K.","major_comments":[{"comment":"The central experimental evidence for heavy-fermion behavior is the quoted value γ=160 mJ mol−1 Sm K−2, but no specific-heat analysis is shown: there is no C(T) curve, no C/T versus T^2 plot, no lattice subtraction using the nonmagnetic YAsS analog, no nuclear-Schottky correction, no error bar, and no statement of whether the value is obtained at H=0 or H=9 T or in which temperature interval. Because SmAsS orders antiferromagnetically at T_N≈7.5 K and the data extend down to 0.4 K, the C/T ratio below T_N can contain substantial contributions from AFM spin waves and nuclear Schottky tails. The authors must present the extraction of γ and demonstrate that it is a genuine linear electronic term, or else the heavy-fermion conclusion should be correspondingly qualified.","section":"Sec. II and Sec. VII B"},{"comment":"The DFT, symmetry-indicator, and Anderson-lattice calculations all assume a paramagnetic, time-reversal-symmetric state, yet SmAsS orders antiferromagnetically at about 7.5 K (Fig. 1d). The paper does not discuss whether the AFM order breaks the glide+TRS protection that produces the hourglass fermions, nor whether the surface states survive below T_N. Without this discussion, the claim that SmAsS is a heavy-fermion topological crystalline insulator in the experimentally relevant low-temperature state is not established; at minimum the surface-state prediction should be explicitly stated to apply to the paramagnetic phase, and the fate of the hourglass states below T_N should be addressed.","section":"Sec. IV and Sec. VI"},{"comment":"The assignment to space group Pnma is load-bearing for the Z4=2 indicator and the hourglass-fermion prediction. The authors state that laboratory diffraction was insufficient and that only synchrotron data allowed unambiguous identification, but the main text shows only one selected diffraction region. Given the near-equal twin ratio reported for YAsS (0.518:0.482) and the previously reported monoclinic angle very close to 90° (γ=90.26°–90.37°), a direct comparison of the Pnma refinement against the monoclinic P1121/n model (or an equivalent pseudomerohedral twin model) is needed to exclude the possibility that the orthorhombic model is an average over twinned monoclinic domains. The authors should report the residual factors, goodness of fit, and reflection-splitting criteria for both structural models.","section":"Sec. II and Appendix C, Table S.IV"},{"comment":"The Anderson-lattice calculation is presented as showing that 'the surface states persist despite f-electron interactions and shift downward in energy'. Because the itinerant Hamiltonian h(k) is the Wannier model fitted to the same DFT bands whose topology is being tested, and the parameters ε_f, V, and Q are chosen to pin the f-level at the Fermi level, the calculation is a consistency check for this parameter set in the saddle-point approximation rather than a parameter-free robustness proof. This should be stated explicitly, and the sensitivity of the result to the chosen parameters (particularly V and ε_f) should be described so that the reader can gauge how general the predicted robustness is.","section":"Sec. IV A and Sec. VII A"}],"minor_comments":[{"comment":"The sentence 'Note that the calculations haven been done using the standard Pnma setting' contains a typo: 'haven' should be 'have'.","section":"Sec. II"},{"comment":"The Methods section contains an incomplete sentence: 'For YAsS and SmAsS, single crystal diffraction data were collected using with Mo...' This should be completed or rephrased.","section":"Sec. VII B"},{"comment":"In the Conclusions, the formula 'YB12' should be written as 'YbB12' to match the standard notation used in the Introduction and References.","section":"Sec. VI"},{"comment":"The symbol λ is used both for the spin-orbit coupling parameter (Sec. III B) and for the Lagrange multiplier enforcing the f-electron constraint (Sec. IV A, Eq. (2)). These should be renamed to avoid confusion, for example λ_SOC and λ_constraint.","section":"Sec. III A and Appendix B"},{"comment":"The Wilson-loop integration directions are described separately for the monolayer and bilayer models; for clarity, please state explicitly in one place which momentum direction is integrated and which is the loop parameter in each case, and match this to the axes of Fig. 4c and 4d.","section":"Sec. III B and III C"},{"comment":"The notation for irreps (e.g., '4Γ+1⊕1Γ−1⊕...') is compressed; a one-sentence explanation of how the inversion-eigenvalue imbalance is converted to the Z4 indicator would help non-specialist readers follow the symmetry analysis.","section":"Table I and Table II"}],"recommendation":"major_revision","confidential_remarks":"The paper's experimental anchor for the heavy-fermion claim is a single γ value with no supporting extraction; this should be checked by the experimental coauthors before acceptance. The structural assignment appears plausible but the near-equal twin ratio for YAsS and the close-to-90° monoclinic angle make an explicit comparison with the monoclinic model necessary. The paramagnetic assumption for a magnetically ordered compound is a substantive gap that should be addressed in the revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know about arXiv:2505.01511. The topological classification part is solid and likely correct: YAsS and SmAsS are Z4=2 hourglass-fermion TCIs in the paramagnetic state, and the corrected crystal structures are a genuine contribution. The heavy-fermion claim for SmAsS, however, rests on a single Sommerfeld coefficient that isn't documented, and the paper never reconciles the 7.5 K antiferromagnetic order with the TRS-protected surface states it predicts.\n\nWhat's actually new: the structural revision of the RAsS family. The old monoclinic assignment was wrong; Y/SmAsS are orthorhombic Pnma with zigzag As chains, and LaAsS is a sqrt(2) x sqrt(2) GdPS-type variant. The synchrotron powder and single-crystal twin refinements look careful, with residual factors in the 2-3% range. That alone is worth publishing. The DFT/TQC analysis is standard: Z4=2, Wilson-loop winding, surface hourglass dispersions. The minimal model built from As p_y/p_z orbitals reproduces the band inversion and the hourglass states, and it's a nice layer-construction realization. Extending it to an Anderson lattice model shows the f-electrons don't destroy the gaps at the saddle-point level; that's a consistency check, not a derivation, and the paper is reasonably honest about that.\n\nSoft spots: the gamma = 160 mJ/mol K^2 is the entire experimental basis for the 'heavy-fermion' label. The Methods say specific heat was measured from 0.4 to 100 K at 0 and 9 T, but no C/T vs T^2 plot, no lattice subtraction using YAsS, no nuclear-Schottky correction, no error bar, and no stated fit interval. Given that SmAsS orders antiferromagnetically at 7.5 K, the linear term could easily be contaminated by magnetic or Schottky contributions. The AFM order is also not reconciled with the topology: the Z4=2 and hourglass states rely on TRS, which is broken below TN. The paper doesn't state that the surface states are properties of the paramagnetic phase or say what happens to them in the ordered state. That's a real omission, not a nitpick.\n\nWho this is for: readers interested in new TCI candidates and structure-property relations in square-net materials. The topology part is solid enough that a serious referee should engage with it. But the heavy-fermion framing needs major revision: either document the gamma extraction properly (with a lattice reference and error bar) or drop the SmB6/YbB12 comparison to a measured claim.","headline":"A solid TCI classification for Y/SmAsS with corrected structures, but the heavy-fermion claim for SmAsS is under-evidenced and the AFM order is never squared with the TRS-protected surface states.","tokens_in":19917,"tokens_out":5034,"would_cite":true,"duration_ms":45182,"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":"SmAsS and YAsS are topological crystalline insulators whose glide-protected hourglass surface states survive f-electron correlations.","keywords":["heavy-fermion topological crystalline insulator","hourglass fermions","glide symmetry","Z4 symmetry indicator","SmAsS","YAsS","Anderson lattice model","Kondo physics"],"falsifier":"If high-resolution ARPES on a glide-preserving (011) surface of YAsS or SmAsS shows no hourglass-shaped crossing in the surface Brillouin zone along $\\bar{\\Gamma}{-}\\bar{Y}$ or $\\bar{X}{-}\\bar{S}$, the central claim fails. Equivalently, a synchrotron refinement that fits the data equally well in the previously reported monoclinic space group P1121/n would remove the glide symmetry that the $Z_4 = 2$ indicator requires, and the hourglass prediction would not follow.","tokens_in":18713,"feed_emoji":"⌛","tokens_out":7669,"duration_ms":71857,"temperature":0.7,"pith_summary":"This paper claims that the rare-earth arsenic sulfides YAsS and SmAsS are topological crystalline insulators whose nontrivial topology is enforced not by spin-orbit coupling but by a glide symmetry of the orthorhombic Pnma structure. The central result is that SmAsS combines this crystalline topology with heavy-fermion behavior, as signaled by a measured Sommerfeld coefficient of $\\gamma = 160$ mJ mol$^{-1}$ Sm K$^{-2}$, placing it in the small class of materials, alongside SmB$_6$ and YbB$_{12}$, where topological surface states coexist with Kondo-like correlations. Using high-resolution synchrotron diffraction, the paper corrects a decades-old monoclinic structural assignment, derives a $Z_4 = 2$ symmetry indicator, and predicts hourglass fermions on the (011) surface. A minimal Anderson-lattice model reproduces the topology and predicts that the hourglass surface states survive the f-electron correlations, shifted downward in energy. If correct, this establishes a new family in which crystalline symmetry and strong correlations work together, and makes SmAsS a concrete target for photoemission checks.","feed_headline":"SmAsS and YAsS host hourglass fermions as topological insulators","feed_subtitle":"Glide symmetry protects the surface states, and SmAsS's heavy electrons shift them into reach of ARPES.","key_machinery":"The central object is the hourglass fermion protected by the nonsymmorphic glide $\\{m_x|\\tfrac12\\tfrac12\\tfrac12\\}$ in space group Pnma. Along a glide-invariant line, Kramers pairs at the zone center have the same glide eigenvalue while at the zone boundary they have opposite eigenvalues ($\\pm i$), so the bands are forced to cross in an hourglass shape that cannot be removed without breaking the glide. This protection is quantified by the $Z_4$ symmetry indicator, which takes the value 2 in YAsS and SmAsS. The argumentative engine is a layer construction: two glide-related As monolayers, each a $Z_2 = 1$ two-dimensional topological insulator, coupled by interlayer hoppings, reproduce the drumhead and hourglass surface spectra; for SmAsS the same model is extended by an Anderson-lattice term with an infinite-$U$ constraint solved at saddle point.","core_discovery":"On the paper's own terms, the discovery is that SmAsS and YAsS are glide-symmetry-protected topological crystalline insulators, with SmAsS additionally a heavy-fermion candidate. The load-bearing invariant is $Z_4 = 2$, obtained from inversion eigenvalues at $\\Gamma$ and $U$; in the nonmagnetic Pnma setting this indicator forces hourglass surface states along the $k_x = 0$ (and $\\pi$) glide-invariant planes. Without spin-orbit coupling, the same $Z_4 = 2$ gives $z'_2 = 1$ nodal lines, which appear as drumhead surface states. The paper shows that the topological bands come from As $p_z$ and $p_y$ orbitals arranged in two glide-related layers, each a $Z_2 = 1$ two-dimensional topological insulator; coupling them with the glide produces the hourglass. Including the Sm 4f electrons as an Anderson lattice at saddle-point level leaves the $Z_4 = 2$ invariant unchanged, so the f-states do not destroy the surface states but shift the topological gap downward by roughly 0.4 eV.","pith_inferences":["A direct test is to measure the (011) surface with photoemission at low temperature: if the hourglass dispersion is absent while the bulk remains semiconducting, the symmetry-indicator prediction would be ruled out.","Because SmAsS orders antiferromagnetically near 7.5 K, the paper's paramagnetic model leaves open whether the hourglass states survive in the magnetic phase; measuring the surface spectrum below $T_N$ would test whether the glide remains a good symmetry there.","The structural correction suggests that other members of the RAsS family reported forty years ago as monoclinic may deserve re-examination; compounds with heavier or magnetic rare earths could realize the same hourglass phase with different correlation strengths."],"forward_implications":["Angle-resolved photoemission on the (011) surface of SmAsS should reveal hourglass surface bands, shifted roughly 0.4 eV closer to the Fermi level by the f-electron states, making the predicted states spectroscopically accessible.","SmAsS becomes a third platform, after SmB$_6$ and YbB$_{12}$, where topological surface states and Kondo coherence coexist, but with topology enforced by crystalline glide symmetry rather than by parity alone.","LaAsS sits close to a topological transition: its experimental structure gives $Z_4 = 2$ while the optimized structure gives $Z_4 = 0$, so modest strain or pressure could switch it into or out of the hourglass phase.","The minimal two-layer model transfers to other ZrSiS-type square-net materials with the same As $p_z$/$p_y$ orbital content, predicting where hourglass fermions should appear in that broader family."],"supporting_citations":[{"why":"Prior monoclinic (P1121/n) assignment for the RAsS series; the paper's synchrotron data overturn it, establishing the glide-bearing Pnma structure.","marker":"26"},{"why":"Earlier single-crystal work that gave SmAsS an orthorhombic Pcmn cell, the structural basis this paper refines.","marker":"33"},{"why":"Introduced hourglass fermions and their glide-plus-time-reversal protection; the surface states predicted here are of this type.","marker":"44"},{"why":"Supplies the layer-construction scheme and the $Z_4$-to-$Z_2$-layers mapping used to build the minimal model.","marker":"38"},{"why":"Relates $Z_4 = 2$ to the $z'_2$ nodal-line invariant that yields the drumhead states in the spinless limit.","marker":"37"},{"why":"Shows the same layer construction in the related square-net material LaSbTe, whose topology the present model generalizes.","marker":"28"},{"why":"ErAsS is the closest prior candidate compound with hourglass fermions; the paper positions SmAsS beside it.","marker":"29"},{"why":"Auxiliary-boson treatment of the infinite-$U$ Anderson lattice is the framework for the saddle-point correlated model.","marker":"48"}],"fun_headline_variants":["SmAsS and YAsS reveal hourglass fermions under glide symmetry","Heavy-fermion topological insulators: hourglass states in SmAsS","Glide symmetry spawns hourglass fermions in new topological insulators","SmAsS joins heavy-fermion topological insulators with hourglass surfaces","Topological hourglass states persist in correlated SmAsS"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole hourglass prediction rests on the newly assigned orthorhombic Pnma structure for Y/SmAsS (and Pmnb for LaAsS); if the true symmetry were the previously reported monoclinic P1121/n, the glide that protects the surface states would not exist and the $Z_4 = 2$ indicator would not apply.","fun_headline_variants_meta":{"raw":{"variants":["SmAsS and YAsS reveal hourglass fermions under glide symmetry","Heavy-fermion topological insulators: hourglass states in SmAsS","Glide symmetry spawns hourglass fermions in new topological insulators","SmAsS joins heavy-fermion topological insulators with hourglass surfaces","Topological hourglass states persist in correlated SmAsS"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1353,"prompt_tokens":1017,"completion_tokens":336,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":633,"completion_tokens_details":{"reasoning_tokens":239}},"tokens_in":633,"tokens_out":336,"duration_ms":3458,"temperature":1.0,"reasoning_tokens":239,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:18:40.541104+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If high-resolution ARPES on a glide-preserving (011) surface of YAsS or SmAsS shows no hourglass-shaped crossing in the surface Brillouin zone along $\\bar{\\Gamma}{-}\\bar{Y}$ or $\\bar{X}{-}\\bar{S}$, the central claim fails. Equivalently, a synchrotron refinement that fits the data equally well in the previously reported monoclinic space group P1121/n would remove the glide symmetry that the $Z_4 = 2$ indicator requires, and the hourglass prediction would not follow.","supporting_citations":[{"cited_title":"Ceolin, N","cited_arxiv_id":null,"evidence_quote":"Prior monoclinic (P1121/n) assignment for the RAsS series; the paper's synchrotron data overturn it, establishing the glide-bearing Pnma structure."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier single-crystal work that gave SmAsS an orthorhombic Pcmn cell, the structural basis this paper refines."},{"cited_title":"Alexandradinata, R","cited_arxiv_id":null,"evidence_quote":"Introduced hourglass fermions and their glide-plus-time-reversal protection; the surface states predicted here are of this type."},{"cited_title":"Quantitative mappings between symmetry and topology in solids","cited_arxiv_id":null,"evidence_quote":"Supplies the layer-construction scheme and the $Z_4$-to-$Z_2$-layers mapping used to build the minimal model."},{"cited_title":"Diagnosis for nonmagnetic topological semimetals in the absence of spin-orbital coupling","cited_arxiv_id":null,"evidence_quote":"Relates $Z_4 = 2$ to the $z'_2$ nodal-line invariant that yields the drumhead states in the spinless limit."},{"cited_title":"Layer construction of topological crystalline insulator lasbte","cited_arxiv_id":null,"evidence_quote":"Shows the same layer construction in the related square-net material LaSbTe, whose topology the present model generalizes."},{"cited_title":"Topological crystalline insulator candidate erass with hourglass fermion and magnetic-tuned topological phase transition","cited_arxiv_id":null,"evidence_quote":"ErAsS is the closest prior candidate compound with hourglass fermions; the paper positions SmAsS beside it."},{"cited_title":"The Kondo problem to heavy fermions","cited_arxiv_id":null,"evidence_quote":"Auxiliary-boson treatment of the infinite-$U$ Anderson lattice is the framework for the saddle-point correlated model."}],"review_version":1}