{"id":"bc063976-841d-4ac0-9add-6c6f7a853110","arxiv_id":"2412.02213","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Nickel doping of the diamond-lattice antiferromagnet Co1.3Ir1.7S4 drives an insulator-to-metal crossover, suppresses antiferromagnetism near x=0.95, and yields non-Fermi-liquid behavior with a spin-glass tail.","lead":"This paper maps how adding nickel to a cobalt-iridium thiospinel changes it from an antiferromagnetic insulator to a metal with a frozen spin-glass state and unusual low-temperature electronic behavior. It provides a new material platform for studying how disorder interrupts a magnetic quantum phase transition.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The phase diagram's x-axis depends on the assumed A-site Ni occupancy; site inversion at x≥0.9 is inferred only from lattice-parameter deviation, not directly measured, so xc and the glassy-tail assignment are not fully secured.","rationale":"The reader's weakest assumption identifies exactly the issue that I consider most load-bearing: nominal x equals the active A-site Ni concentration. The paper's own admission of 4.6–6.7% site inversion at high doping (Table S2) shows this is not a hypothetical edge case but a documented deviation that is estimated only indirectly from the lattice-parameter behavior. Since the phase diagram, the critical concentration xc, and the quantum Griffiths interpretation are all calibrated against the x-axis, a larger or differently distributed inversion would directly alter the conclusions. I therefore agree with the reader's assessment that the central claim is probably sound but conditional on a direct structural/occupancy measurement. The paper does provide independent support: the XRD shows a single phase, the lattice contraction and DFT free-energy differences support A-site substitution, and the magnetic susceptibility and heat capacity data track each other across TN and TSG, which is a good internal consistency check. There is no evidence of internal inconsistency or overreach in the raw data; the concern is about the calibration of the independent variable. The reader's CONDITIONAL verdict is appropriate, and my stress test does not move that verdict. I would add that the QGP interpretation is explicitly provisional in the paper ('further investigations are still needed'), which further reduces the risk of the central claim being rejected outright. The concrete test proposed—direct site-occupancy refinement—would settle whether the inversion fraction is indeed as small as estimated, and if so, the phase diagram would stand on much firmer footing.","tokens_in":14117,"tokens_out":3474,"duration_ms":40455,"concrete_test":"Perform Rietveld refinement against neutron powder diffraction (or resonant X-ray diffraction at the Co and Ni K-edges) on samples with x = 0.6, 0.8, 0.95, and 1.0 to determine Co/Ni occupancies on the A and B sites. Compare the refined inversion fraction with the 4.6–6.7% estimate from Table S2. If the inversion is significantly larger (e.g., >15% at x=1) or strongly composition-dependent, re-extract the phase diagram using the refined A-site Co content as the true tuning parameter and re-evaluate xc and the glassy-tail interpretation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—the full phase diagram in Fig. 5 with an avoided AFM QCP at xc≈0.95 and a quantum-Griffiths-like glassy tail above it—is plotted against nominal Ni content x. The interpretation that Ni is a nonmagnetic electron donor occupying the A site is supported by the lattice contraction in Fig. 1(d) and DFT free-energy differences in Fig. 1(e), but the paper itself reports 4.6–6.7% site inversion for x≥0.9 (Table S2), estimated from the deviation of the lattice parameter from a linear Vegard-like fit using ionic radii (Shannon). This is an indirect calibration. If the inversion is larger or composition-dependent in the metallic/NFL region (x≥0.6), the effective A-site Co concentration becomes 1−x+y rather than 1−x, shifting the insulator-metal crossover, TN(x), and xc. Concretely, at x=1 the end member would no longer be Co0.3NiIr1.7S4 with 0.3 Co on A, but would have additional Co on A (and Ni on B), changing the magnetic dilution, the electron count, and the relevance of the disorder-driven QGP interpretation. Because the entire quantitative x-axis, the location of xc, and the identification of the glassy tail above xc rest on this occupancy assumption, this is the most load-bearing concern. The claim is plausible and the authors transparently flag the inversion, but it is not yet directly verified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a systematic Ni-doping study of the thiospinel Co1.3-xNixIr1.7S4 (0 ≤ x ≤ 1), combining powder XRD, dc magnetization, electrical resistivity, and specific heat. The authors track the evolution from an antiferromagnetic insulator (TN ≈ 292 K at x = 0) through an insulator-to-metal crossover near x ≈ 0.35, a spin-glass-like transition in the metallic state, and a non-Fermi-liquid regime for x ≥ 0.95 with resistivity exponent α ≈ 1.2–1.3 and divergent C/T. They construct a phase diagram (Fig. 5) in which the AFM quantum critical point is avoided and argue that the glassy tail above xc is consistent with a quantum Griffiths phase driven by quenched disorder. DFT calculations are used to support the A-site Ni occupancy and to estimate the bare electronic specific-heat coefficient.","tokens_in":14458,"tokens_out":3096,"duration_ms":34161,"significance":"If the phase diagram is correct, this paper provides a new doped diamond-lattice antiferromagnet in which disorder converts an avoided QCP into a glassy non-Fermi-liquid region. The key strengths are the use of three independent bulk probes (magnetization, resistivity, specific heat) that yield mutually consistent transition temperatures for the AFM order, and the transparent reporting of the site-inversion caveat. The DFT calculations are a useful complement, although they are not the main evidence. The claim of a quantum Griffiths phase is appropriately phrased as a possibility rather than a definitive identification, and the authors explicitly call for further investigation. The main limitation is the indirect determination of the A-site occupancy, which anchors the entire x-axis, and the lack of quantitative uncertainty estimates on the fitted exponents and phase boundaries.","major_comments":[{"comment":"The entire x-axis and the location of xc ≈ 0.95 in Fig. 5 are interpreted under the assumption that nominal Ni concentration equals the active A-site Ni content, with Ni acting as a nonmagnetic electron donor. The manuscript itself reports 4.6–6.7% site inversion for x ≥ 0.9, estimated only from the deviation of the lattice parameter from a linear fit (Table S2). No direct measurement of site occupancy (resonant XRD, XAS, neutron diffraction) is provided. If the inversion is larger or composition-dependent in the metallic/NFL region, the effective A-site Co concentration becomes 1 − x + y, which shifts the insulator-metal crossover, TN(x), and xc, and the assignment of the glassy tail to A-site Co dilution would need revision. Please add a direct occupancy determination or, failing that, a quantitative sensitivity analysis of the phase diagram to plausible inversion profiles.","section":"§III.A, Table S2"},{"comment":"The phase diagram in Fig. 5 is presented without uncertainty estimates on any of the boundaries, although TN is extracted from three different probes and the paper states only that the values are 'generally consistent.' Please define clearly how each transition temperature is determined (peak position, inflection point, onset, or fit criterion), provide error bars or at least representative uncertainties on the points in Fig. 5, and quantify the consistency among TχN, TρN, and TCN. This is load-bearing because the claimed suppression of TN to zero at xc ≈ 0.95 and the existence of the SG-like tail above xc rest on the precise evolution of these boundaries.","section":"§III.E, Fig. 5"},{"comment":"The specific-heat analysis uses C/T = γ + βT² for x ≤ 0.6 and C/T = γ + βT² − η ln T for 0.8 ≤ x ≤ 1, with no documented justification for this crossover or comparison of fit residuals. A logarithmic term can often mimic other low-temperature contributions, and the reported γ values at x ≥ 0.8 may be sensitive to the choice of fitting form and to the fitted temperature window. Please report the temperature ranges used, the resulting fit parameters with uncertainties, and a comparison against alternative fits (e.g., including a T³ term only, or a fixed ln T term at all x). This is important because the large enhancement of γ is used to support the effective-mass increase and the NFL interpretation.","section":"§III.D, Eq. (C/T fits)"}],"minor_comments":[{"comment":"The description of α ≈ 1.7 at x ≤ 0.9 as being 'close to the Fermi-liquid scenario (α = 2)' is somewhat generous; a 15% deviation is significant and deserves a comment on whether this represents a distinct crossover or simply a fit artifact. Please state the fitted temperature ranges and the statistical errors on α.","section":"§III.C, Fig. 3(d)"},{"comment":"The spin-glass-like transition is identified solely from ZFC/FC bifurcation and a broad anomaly in resistivity and specific heat. Additional evidence such as ac-susceptibility frequency dependence or aging/memory measurements would strengthen the assignment. If such measurements are unavailable, the text should more explicitly acknowledge that the 'SG-like' label is provisional.","section":"§III.B, Fig. 2(b)"},{"comment":"The linear fit in Fig. 1(d) is used both to demonstrate lattice contraction and to estimate the site-inversion degree, but the fit range and the uncertainty of the fitted slope are not stated. Please clarify whether the fit includes only x ≤ 0.8 and what the statistical error of the slope is, since the inversion estimate depends directly on this baseline.","section":"§III.A, Fig. 1(d)"},{"comment":"The DFT calculations are performed with nonmagnetic configurations, and it would be helpful to state explicitly how this choice affects the computed density of states and the comparison with the experimental γ, especially in the magnetically ordered region.","section":"§II, Methods (DFT)"},{"comment":"References [4] and [39] cite the same paper (J. Huang et al., Nat. Phys. 2024); please consolidate or distinguish the two citations. Also, the typo 'diamond-lattic e' appears in the abstract of the manuscript version provided; please correct it.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and is generally well organized. The main risk is not internal inconsistency but the unverified occupancy assumption that anchors the quantitative phase diagram. I would encourage the editor to request the authors address the major comments, especially the direct occupancy measurement or a sensitivity analysis, and to include numerical uncertainty estimates in the revised version. The quantum Griffiths interpretation is speculative but the paper is appropriately cautious; the main load-bearing issue is the x-axis composition."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper for the phase diagram, not for the quantum Griffiths claim. The authors have mapped out Co1.3−xNixIr1.7S4 from x=0 to 1, showing an insulator-to-metal crossover near x≈0.35, gradual suppression of antiferromagnetism, a spin-glass-like regime in the metal, and a non-Fermi-liquid region above x≈0.95 with T^1.2–1.3 resistivity and divergent C/T. That is a useful, new experimental map for a parent compound only reported last year, and the three bulk probes give mutually consistent transition temperatures. The DFT free-energy calculations supporting Ni's A-site preference are a nice touch, and the paper is honest that the QGP is tentative, explicitly calling it \"possible\" and saying more work is needed.\n\nThe soft spots are real but manageable. The fitted exponents (α, η, δ) come with no error bars, and there is no raw data deposit, so I cannot fully check the scaling collapse in the SM. The bigger issue, which the stress-test flagged, is site inversion: the x-axis is nominal Ni content, and the paper itself estimates 4.6–6.7% Co/Ni inversion for x≥0.9 from lattice-parameter deviation, not from a direct probe. That does not sink the phase diagram, because the transitions are still well-defined functions of nominal x, but it does muddy the electron-count interpretation and the magnetic dilution story near the putative critical concentration. A neutron or resonant-XRD study would settle it. The QGP identification is the weakest link: power-law fits to χ(T), M(H), and C/T with exponents around 0.4–0.8 are suggestive but not unique, and the scaling collapse is presented without residuals or error estimates. I would treat the avoided-QCP and QGP language as interpretive framing, not as established physics.\n\nAll that said, the measured phase diagram itself is probably right, and the paper deserves a serious referee. It is aimed at people working on doped spinels, disorder-driven quantum criticality, and NFL behavior in correlated metals. With revisions that add uncertainty estimates, raw data, and ideally a direct site-occupancy measurement, this would be a solid contribution. I would accept it for peer review and would cite the phase diagram in my own work.","headline":"A solid first doping study of a new diamond-lattice antiferromagnet, with a phase diagram that likely holds even if the quantum Griffiths interpretation is provisional.","tokens_in":15005,"tokens_out":1620,"would_cite":true,"duration_ms":21066,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["72.80.Ga","74.62.Dh","75.30.-m","74.70.Xa"],"model":"deepseek-v4-flash","headline":"Substituting nickel for cobalt in the diamond-lattice magnet Co1.3Ir1.7S4 produces a full phase diagram whose central feature is the replacement of the expected antiferromagnetic quantum critical point by a glassy non-Fermi-liquid metal…","keywords":["phase diagram","diamond-lattice antiferromagnet","thiospinel","insulator-metal crossover","non-Fermi liquid","quantum Griffiths phase","spin glass","Ni doping"],"falsifier":"A direct measurement of Ni and Co site occupancies—for example by neutron powder diffraction or extended X-ray absorption fine structure on samples near x=0.95–1—showing that the B-site Ni fraction is far above the estimated 6.7% would undermine the doping-axis interpretation; conversely, a clean single crystal of the same composition showing a conventional antiferromagnetic quantum critical point instead of the glassy non-Fermi-liquid tail would disprove the disorder-driven phase diagram.","tokens_in":13923,"feed_emoji":"🧲","tokens_out":8361,"duration_ms":79315,"temperature":0.7,"pith_summary":"Electron-doping the diamond-lattice antiferromagnet Co1.3Ir1.7S4 with nickel does more than suppress magnetic order: it yields a complete phase diagram in which the insulating magnet becomes metallic, the antiferromagnetic transition is pushed toward zero temperature, and the metal at the edge of magnetic order refuses to behave as a Fermi liquid. The paper constructs this diagram from resistivity, magnetization, and specific-heat measurements on a series of polycrystalline samples. The key claim is that the antiferromagnetic quantum critical point is avoided: instead of a clean continuous transition at the critical doping $x_c\\approx0.95$, the metallic state is disorder-dominated, with a spin-glass-like tail and non-Fermi-liquid behavior (resistivity exponent $\\alpha\\approx1.2$–$1.3$ and divergent $C/T$) persisting above $x_c$. If correct, the system becomes a concrete experimental platform for studying quantum Griffiths physics in a doped diamond-lattice magnet, with the disorder that suppresses Fermi-liquid behavior also likely inhibiting the superconductivity that electron doping of this family was expected to produce.","feed_headline":"Nickel turns a diamond-lattice antiferromagnet into a glassy metal","feed_subtitle":"Ni substitution maps a full phase diagram in which disorder replaces a quantum critical point.","key_machinery":"The load-bearing object is the phase diagram itself, built on the A-site diamond sublattice and the narrow charge-transfer gap of the parent Co[Co0.3Ir1.7]S4. Ni2+ substituted for Co2+ at the tetrahedral A site acts as a nominally nonmagnetic electron donor: the added electrons fill the antibonding Co-$t_2$/S-$p$ states, closing the gap and driving metallization, while diluting the magnetic Co sublattice and weakening the nearest-neighbor antiferromagnetic exchange. In the metallic state, the randomness of Co/Ni occupancy, including partial site inversion at high doping, frustrates the RKKY-coupled spins into a spin-glass-like phase. In the paper's interpretation this same disorder converts the would-be quantum critical endpoint into a finite region of quantum Griffiths behavior, replacing the conventional Fermi-liquid description.","core_discovery":"The authors establish that substituting Ni for Co on the A-site diamond sublattice of Co1.3Ir1.7S4 drives an insulator-to-metal crossover at $x\\approx0.35$, gradually suppresses the Néel order from 292 K at $x=0$ to 23 K at $x=0.7$, and completely extinguishes it near $x_c\\approx0.95$. In the metallic state at $x\\geq0.4$ a spin-glass-like transition emerges at low temperatures, and just above the magnetic phase boundary the resistivity follows a power law $T^\\alpha$ with $\\alpha\\approx1.2$–$1.3$ instead of the Fermi-liquid $T^2$, while $C/T$ grows logarithmically at low temperatures and the electronic specific-heat coefficient $\\gamma$ increases substantially. The paper concludes that an antiferromagnetic quantum critical point is avoided at $x_c$ and that the glassy tail above the critical concentration aligns with an extended quantum Griffiths phase produced by quenched disorder, supported by power-law fits and a scaling collapse of magnetization data.","pith_inferences":["A direct experimental determination of Ni versus Co site occupancy—for example by neutron or resonant X-ray diffraction—could test whether the glassy tail is intrinsic to A-site dilution or an artifact of larger-than-estimated Co–Ni site inversion; this is a measurement the paper does not perform.","Extending the doping beyond $x=1$ or applying pressure could tune the system closer to the suppressed quantum critical point, offering a test of whether the glassy NFL region expands or sharpens into a conventional QCP.","The scaling exponents reported for the Griffiths analysis (e.g., $\\eta$ values near 0.4–0.8 and the $H/T$ collapse) provide quantitative fingerprints that could be compared with infinite-randomness fixed-point predictions, a comparison the paper leaves to future work.","If the disorder-driven interpretation holds, similar non-Fermi-liquid and glassy behavior may appear in other doped diamond-lattice or charge-transfer-gap spinels, making this family a useful testing ground for theory."],"forward_implications":["At $x\\approx0.35$ the system crosses from an insulator to a metal, so electron doping of this diamond-lattice antiferromagnet provides a controlled route from a charge-transfer-gap insulator to a correlated metal.","The antiferromagnetic transition is suppressed smoothly, with $T_N$ decreasing from 292 K to 23 K by $x=0.7$ and vanishing near $x_c=0.95$, so the phase diagram gives a clear target concentration for magnetic quantum criticality.","In the metallic regime above $x_c$, the absence of Fermi-liquid behavior—$\\alpha\\approx1.2$–$1.3$ and divergent $C/T$ with enhanced effective mass—means the material joins the small family of spinel compounds showing non-Fermi-liquid physics.","The spin-glass-like tail persisting beyond $x_c$ implies that quenched disorder, not a clean quantum critical point, controls the low-temperature physics near the magnetic boundary.","The paper reports no superconductivity in the doped region, attributing its absence to the disorder that also produces the glassy non-Fermi-liquid state."],"supporting_citations":[{"why":"Characterizes the parent compound Co[Co0.3Ir1.7]S4: high Néel temperature of 292 K, insulating charge-transfer gap, and A-site diamond-lattice magnetism from which the doping study starts.","marker":"[13]"},{"why":"Predicts diamond-like Co-based chalcogenides as possible unconventional high-temperature superconductors, the motivation for choosing Ni electron doping in this family.","marker":"[19]"},{"why":"Provides the disorder-driven non-Fermi-liquid framework the paper uses to connect spin-glass freezing with NFL behavior.","marker":"[34]"},{"why":"Supplies the survey of non-Fermi-liquid behavior in d- and f-electron metals and the power-law/Griffiths signatures used for comparison.","marker":"[35]"},{"why":"The reference theory for Fermi-liquid instabilities at magnetic quantum phase transitions, used to interpret the enhanced effective mass and the avoided QCP.","marker":"[36]"},{"why":"Gives the quantum Griffiths and smeared-transition theory that the glassy tail above xc is compared with.","marker":"[41]"},{"why":"Provides the Ni1-xVx experimental analog of a quantum Griffiths phase in a disordered itinerant ferromagnet.","marker":"[43]"},{"why":"Supplemental material containing free-energy calculations and site-inversion estimates that justify assigning Ni to the A site and quantify disorder at high doping.","marker":"[27]"}],"fun_headline_variants":["Nickel turns a diamond antiferromagnet into a glassy metal","Nickel doping replaces a quantum critical point with a Griffiths phase","Diamond-lattice antiferromagnet goes glassy metal under nickel doping","Ni doping induces spin glass and avoids quantum critical point","Insulator-metal crossover leads to a disorder-driven Griffiths phase"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole x-axis story assumes that the nominal nickel content x is actually built into the A-site diamond lattice as nonmagnetic Ni2+ electron donors, but if Co–Ni site inversion at high doping is larger than the estimated 4.6–6.7% or varies differently with x, both the assignment of the glassy tail to A-site Co dilution and the precise value of xc would need revision.","fun_headline_variants_meta":{"raw":{"variants":["Nickel turns a diamond antiferromagnet into a glassy metal","Nickel doping replaces a quantum critical point with a Griffiths phase","Diamond-lattice antiferromagnet goes glassy metal under nickel doping","Ni doping induces spin glass and avoids quantum critical point","Insulator-metal crossover leads to a disorder-driven Griffiths phase"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002034,"raw_usage":{"total_tokens":7993,"prompt_tokens":1079,"completion_tokens":6914,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":695,"completion_tokens_details":{"reasoning_tokens":6826}},"tokens_in":695,"tokens_out":6914,"duration_ms":44521,"temperature":1.0,"reasoning_tokens":6826,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:43:44.907523+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of Ni and Co site occupancies—for example by neutron powder diffraction or extended X-ray absorption fine structure on samples near x=0.95–1—showing that the B-site Ni fraction is far above the estimated 6.7% would undermine the doping-axis interpretation; conversely, a clean single crystal of the same composition showing a conventional antiferromagnetic quantum critical point instead of the glassy non-Fermi-liquid tail would disprove the disorder-driven phase diagram.","supporting_citations":[{"cited_title":"Ji, S.-Q","cited_arxiv_id":null,"evidence_quote":"Characterizes the parent compound Co[Co0.3Ir1.7]S4: high Néel temperature of 292 K, insulating charge-transfer gap, and A-site diamond-lattice magnetism from which the doping study starts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Predicts diamond-like Co-based chalcogenides as possible unconventional high-temperature superconductors, the motivation for choosing Ni electron doping in this family."},{"cited_title":"Miranda and V","cited_arxiv_id":null,"evidence_quote":"Provides the disorder-driven non-Fermi-liquid framework the paper uses to connect spin-glass freezing with NFL behavior."},{"cited_title":"Stewart, Non-Fermi-liquid behavior in d-and f- electron metals, Rev","cited_arxiv_id":null,"evidence_quote":"Supplies the survey of non-Fermi-liquid behavior in d- and f-electron metals and the power-law/Griffiths signatures used for comparison."},{"cited_title":"Vojta, Quantum griffiths effects and smeared phase transitions in metals: Theory and experiment, J","cited_arxiv_id":null,"evidence_quote":"Gives the quantum Griffiths and smeared-transition theory that the glassy tail above xc is compared with."},{"cited_title":"Ubaid-Kassis, T","cited_arxiv_id":null,"evidence_quote":"Provides the Ni1-xVx experimental analog of a quantum Griffiths phase in a disordered itinerant ferromagnet."},{"cited_title":"3−xNixIr2S4 (0.95 ≤ x ≤ 1.15), the specific heat of NiIr 2S4, the density of states and band structure for Co 1−xNix[Co0","cited_arxiv_id":null,"evidence_quote":"Supplemental material containing free-energy calculations and site-inversion estimates that justify assigning Ni to the A site and quantify disorder at high doping."}],"review_version":1}