{"id":"f2a2ba40-56bb-4feb-a0b7-08ce85ff11de","arxiv_id":"2506.00778","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In Au-Al-Gd 1/1 approximants, the largest magnetocaloric entropy change appears in the antiferromagnetic region near the FM/AFM boundary, not in the ferromagnetic phase.","lead":"Researchers measured the magnetic entropy change in Au-Al-Gd quasicrystal approximants across a wide composition range and found the largest response in an antiferromagnetic region near a phase boundary, reaching 7.2 J/K per mole of Gd under a 5 T change. The result suggests tuning the electron-per-atom ratio across a magnetic phase boundary could improve low-temperature magnetic refrigeration materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (2)'s Taylor expansion is invalid at the measured H/TC values, so the claimed MFT breakdown and the flat ΔSM_max(TC) are not established without an exact mean-field calculation.","rationale":"The reader's weakest assumption identified exactly the load-bearing issue: Eq. (2) is a small-argument Taylor expansion of the Brillouin function, and the paper does not demonstrate its validity at the measured H/TC ratios. This concern is central because the headline scientific claim includes a specific deviation from the mean-field TC^{-2/3} law; if the comparison baseline is incorrect, the 'clear breakdown of MFT' is not established. The AFM-region enhancement is a separate empirical result that may still hold, but the paper itself interprets it as 'presumably associated with the breakdown of the MFT,' so the theoretical interpretation is tied to the same questionable comparison. A concrete numerical check with the exact MFT solution would settle whether the observed flatness and exponent deviations are genuine beyond-mean-field effects or merely artifacts of the low-field expansion. The existing CONDITIONAL verdict already captures the need for this check, so no verdict change is required; the condition should be made explicit and the test performed before the MFT-breakdown claim is accepted. The paper has real strengths: the comprehensive ΔSM map across e/a is new, the phase diagram and magnetization data are internally consistent, and the experimental values are reported directly. Those strengths do not, however, rescue the unvalidated MFT comparison, which is the most fragile link in the central argument.","tokens_in":12647,"tokens_out":9962,"duration_ms":102015,"concrete_test":"Numerically solve the self-consistent Weiss molecular-field equation m = N gJ μB B_J(gJ μB(H + λ m)/(kB T)) with J = 7/2, λ fixed by each reported TC, and the same Gd concentration as the samples. Compute M(T,H) over dense T and H grids from 0 to 5 T, then obtain ΔSM(T,H) via the Maxwell relation μ0 ∫ (∂M/∂T)_H dH. Extract ΔSM_max and the exponent n as done in the paper, and overplot the results on Fig. 5. If the exact MFT reproduces the flat ΔSM_max(TC) trend and the n values, the claim of MFT breakdown is refuted; if it still predicts a clear TC^{-2/3} decay and n ≈ 2/3, the claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central argument that mean-field theory (MFT) breaks down rests on comparing the measured ΔSM_max with Eq. (2), which is derived by Taylor-expanding the Brillouin function in the limit of small argument. At the experimental fields, the Brillouin-function argument for Gd3+ (gJ = 2) is x = gJ μB H / (kB T) ≈ 1.34 H/T. With μ0H = 5 T and TC between 10 and 30 K, x ranges from ≈0.67 down to ≈0.22. The Taylor expansion used in Eq. (2) is not justified at the upper end of this range. The paper states: \"This discrepancy cannot be attributed to the approximation used in deriving Eq. (2), but is simply due to the breakdown of the validity of the BJ function.\" That assertion is unsupported, because a failure of the expansion is not distinguishable from a true breakdown of MFT at these x values. The correct MFT benchmark is the numerical solution of the Weiss molecular-field equations for J = 7/2 at each measured TC, from which ΔSM can be computed directly. Without this benchmark, the observed flat ΔSM_max(TC) and the field exponents n = 0.65–0.88 could be the exact MFT behavior at non-infinitesimal H/TC rather than evidence against MFT. The absence of error bars further weakens the claim that the TC dependence is flat within uncertainty.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports magnetization measurements on polycrystalline AuxAl86-xGd14 (x = 51–73) Tsai-type 1/1 approximant crystals, constructs a magnetic phase diagram versus electron-per-atom ratio e/a, and uses the Maxwell relation to compute the magnetic entropy change ΔSM across the spin-glass (SG), ferromagnetic (FM), and antiferromagnetic (AFM) regions. The central experimental results are a comprehensive ΔSM map over e/a = 1.54–1.98, a nearly T_C-independent maximum entropy change ΔSMmax within the FM region (T_C ≈ 10–30 K), and an enhanced ΔSMmax of about 7.2 J/K mol-Gd under a 5 T field change in the AFM region near the FM/AFM boundary. The authors interpret these observations as evidence for a breakdown of mean-field theory (MFT) near the FM phase boundaries and propose tuning the magnetic ground state across a phase boundary as a route to improved magnetocaloric performance.","tokens_in":12912,"tokens_out":5586,"duration_ms":58729,"significance":"If the conclusions hold, the paper would provide a valuable design principle for magnetocaloric materials in Tsai-type approximants and potentially in rare-earth intermetallics more broadly. The comprehensive ΔSM map across an isostructural series with continuously varying ground state is original and experimentally demanding, and the reported high ΔSMmax in an AFM region under high fields is an interesting falsifiable claim. Credit is also due for using the Maxwell relation on directly measured M-T data and for presenting the field and temperature dependences systematically. However, the theoretical interpretation rests on a comparison with a low-field Taylor expansion of the Brillouin function, and the manuscript does not yet establish that the observed deviations are genuine MFT breakdowns rather than finite-field effects. With an exact mean-field benchmark, error estimates, and a more transparent comparison, the central claim could become convincing.","major_comments":[{"comment":"The central MFT-breakdown claim rests on Eq. (2), which is a small-argument Taylor expansion of the Brillouin function. For Gd3+ (J = 7/2, g = 2), the Brillouin-function argument is x = gJ μB H / (kB T) with gJ = 7, giving x ≈ 2.35 for μ0H = 5 T and T_C ≈ 10 K, and x ≈ 0.78 for T_C ≈ 30 K. The upper end is far outside the small-x regime where Eq. (2) is valid. The assertion in the text that the discrepancy 'cannot be attributed to the approximation used in deriving Eq. (2)' is therefore unsupported: at these x values, a failure of the Taylor expansion is not distinguishable from a true breakdown of MFT. The authors should replace Eq. (2) with the numerical solution of the Weiss molecular-field equations for J = 7/2 at each measured T_C and compute ΔSmax from the full Brillouin function; only then can the observed flat T_C dependence be attributed to a breakdown of MFT.","section":"§III, Eq. (2), Fig. 5(a)"},{"comment":"No error bars or uncertainty estimates are provided for ΔSmax, T_C, or the exponent n. The claim that ΔSmax is 'almost independent' of T_C over 10–30 K and that n deviates systematically from the mean-field value n = 2/3 is not quantitatively supported without such uncertainties. The reported n values are 0.62–0.71 in the central FM region, with only the two border compositions giving n = 0.85 and 0.88; within experimental uncertainty this could be consistent with a constant n ≈ 2/3. The authors should propagate uncertainties from magnetization noise, field and temperature calibration, and the numerical integration in Eq. (1), and show that the flatness and exponent deviations are statistically significant.","section":"Fig. 5(a), Fig. 6, Fig. S5"},{"comment":"The field exponent n is obtained by fitting |ΔSmax| = A (μ0H)^n over the entire 0–5 T range, but Eq. (2) is a low-field asymptotic law. Fitting a power law across the full field range mixes the low-field and high-field regimes, so the extracted n is not a direct test of the MFT prediction away from the low-field limit. An exact MFT calculation of ΔSmax(H,T_C) would provide the proper benchmark for both the T_C dependence and the field dependence, and would clarify whether the observed exponent variation near the phase boundaries is anomalous or simply a finite-field effect.","section":"§III, Fig. S5"}],"minor_comments":[{"comment":"In the abstract, 'a n effective strategy' contains a typo and should read 'an effective strategy'.","section":"Abstract"},{"comment":"The Maxwell relation in Eq. (1) is written symbolically; the manuscript should state that the integration was performed numerically and give the magnetic-field step size or the number of field points used, since the accuracy of ΔSM depends on the discretization.","section":"Eq. (1) and Fig. 4"},{"comment":"The solid line representing the MFT prediction in Fig. 5(a) is not described with explicit parameters; the definitions of C and K in Eq. (2) and the prefactor used to draw the line should be reported so the reader can verify the normalization.","section":"Fig. 5(a)"},{"comment":"The e/a values used for the phase diagram and ΔSM map should state whether they are computed from nominal or analyzed compositions; Table S1 shows analyzed compositions differing by up to about 4 at.% from nominal, so an error estimate on e/a should be included.","section":"Table S1 and Fig. 3"},{"comment":"The statement that 7.2 J/K mol-Gd is the highest ΔSM ever reported in Tsai-type compounds should be qualified as 'to the authors' knowledge' and should specify the comparison conditions (field variation, temperature range, and normalization per mole of Gd), since literature values are often quoted under different field amplitudes and units.","section":"§III, Fig. 6"},{"comment":"The inset of Fig. 2(d) shows a metamagnetic anomaly near 0.5 T at 2 K for x = 73. The manuscript should clarify whether the same metamagnetic transition is present in the AFM samples (e/a = 1.54–1.57) that contribute to the high-field ΔSmax enhancement, and whether the transition temperature or field changes with e/a.","section":"Fig. 2(d) and Fig. 6"}],"recommendation":"major_revision","confidential_remarks":"The experimental data set is original and the e/a-driven ΔSM map is a useful contribution. The main theoretical claim, however, is not yet supported because the MFT benchmark is a low-field expansion applied outside its validity range. The requested exact mean-field calculation and uncertainty estimates are well within the scope of a revision, so I do not see grounds for rejection. The manuscript relies heavily on prior work from the same group for the phase diagram and earlier conclusions, but the ΔSM measurements themselves are independent evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick verdict: the experimental core of this paper is solid and worth having. The authors synthesized a single-phase series of Au–Al–Gd 1/1 approximants spanning e/a from 1.54 to 1.98, measured M–T and M–H, and produced a systematic ΔSM map via the Maxwell relation. That scan across the SG–FM–AFM phase diagram is new, and the high-field enhancement in the AFM region near the FM/AFM boundary (≈7.2 J/K mol-Gd at Δ(μ0H) = 5 T) is a genuine observation. It will interest the magnetocaloric and quasicrystal communities, and it is a defensible result on its own.\n\nThe trouble is the paper's interpretive frame. The claim that mean-field theory breaks down near the FM boundaries rests on comparing the measured ΔSM,max with Eq. (2), a small-argument Taylor expansion of the Brillouin function. At the measured fields and TC values, the Brillouin-function argument x = gJ μB H/(kB T) is as large as ~0.67 (5 T, TC ≈ 10 K). Eq. (2) is not valid there. The paper states, without support, that the discrepancy cannot be attributed to this approximation. But it can—at least, nothing in the paper rules it out. The correct benchmark is a numerical solution of the Weiss molecular-field equations for J = 7/2 at each TC, computing ΔSM from the exact M(H,T). Without that check, the flat ΔSM,max(TC) and the exponents n = 0.65–0.88 are consistent with exact mean-field behavior at non-infinitesimal H/TC. The MFT-breakdown conclusion is therefore conditional, not proven.\n\nA second, more minor issue is the absence of error bars on ΔSM,max, TC, and the exponent n. The flatness of ΔSM,max versus TC may look different once uncertainties are included. Also, the concluding design principle for 'general rare-earth intermetallic compounds' goes beyond the single alloy family studied; it is stated as a suggestion, which is fine, but it should not be read as a demonstrated rule.\n\nWho is this for? Experimentalists working on Tsai-type quasicrystals and approximants, and anyone hunting for low-temperature magnetocaloric materials. Both groups will get useful data and a clear map. For a referee, the paper is worth engaging: the measurements appear careful, the Maxwell integration is standard, and the AFM enhancement is an interesting finding independent of the MFT framing. My recommendation is to send it to peer review, with the request that the authors replace the truncated expansion with an exact MFT calculation and add uncertainty estimates. With those fixes, the paper would be solid; as is, the headline claim overreaches.","headline":"The experimental ΔSM map and the AFM-region enhancement are real and useful; the MFT-breakdown claim needs an exact mean-field benchmark before it can stand.","tokens_in":13480,"tokens_out":3220,"would_cite":true,"duration_ms":31391,"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":"In Au-Al-Gd quasicrystal approximants, the largest magnetic entropy change appears in the antiferromagnetic region near the FM/AFM boundary, reaching about 7.2 J/K mol-Gd at 5 T.","keywords":["quasicrystal approximant","Tsai-type","Au-Al-Gd","magnetocaloric effect","magnetic entropy change","mean-field breakdown","e/a tuning","antiferromagnetism"],"falsifier":"Plot the field-cooled magnetization at $T_C$ for all ferromagnetic samples against $\\mu_0 H/T_C$: if the data collapse onto one universal Brillouin-type curve, the claimed breakdown of mean-field theory is an artifact of the expansion's range, and the near-independence of $\\Delta S_M^{\\max}$ on $T_C$ would be saturation, not anomaly.","tokens_in":12456,"feed_emoji":"🧲","tokens_out":10785,"duration_ms":98287,"temperature":0.7,"pith_summary":"This paper tries to establish that the magnetic entropy change in Au–Al–Gd 1/1 quasicrystal approximants is governed more by where the material sits in the magnetic phase diagram than by the mean-field scaling usually assumed for ferromagnets. By substituting Au for Al across a wide electron-per-atom range, the authors move the same crystal structure through spin-glass, ferromagnetic, and antiferromagnetic states and map the magnetocaloric response across all of them. They find the largest response, about $\\Delta S_M^{\\max}=7.2$ J/K mol-Gd for a 5 T field change, inside the antiferromagnetic region near the FM/AFM boundary, contrary to the usual expectation that ferromagnets give larger entropy changes. Within the ferromagnetic region the maximum entropy change is nearly independent of Curie temperature between 10 and 30 K, in conflict with the mean-field $T_C^{-2/3}$ law. If these results hold, tuning the magnetic ground state across a phase boundary becomes a practical design route for magnetic refrigeration materials.","feed_headline":"Magnetocaloric peak of 7.2 J/K mol-Gd lands at magnetic phase border","feed_subtitle":"Composition tuning across the FM/AFM boundary, not mean-field scaling, decides the cooling response near 10 K.","key_machinery":"The central object is the electron-per-atom ratio $e/a$ as a continuous composition knob in the isostructural Tsai-type 1/1 approximant, combined with the thermodynamic Maxwell relation that converts field-cooled magnetization curves into the magnetic entropy change $\\Delta S_M$. The comparison that exposes the anomaly is the mean-field expression $\\Delta S_M^{\\max}(T_C,\\mu_0 H) = \\tfrac{1}{2} C (C^2/K)^{2/3} (\\mu_0 H/T_C)^{2/3}$, obtained from a Taylor expansion of the Brillouin function, which forces $T_C^{-2/3}$ scaling and a field exponent $n = 2/3$. The phase diagram of AFM, FM, and SG states as a function of $e/a$ is what lets $\\Delta S_M^{\\max}$ be tested in the same crystal structure across all magnetic ground states, and the Brillouin-function scaling relation is the identity whose failure near the phase boundaries carries the paper's claim.","core_discovery":"On its own terms, the paper claims that in the Tsai-type 1/1 Au–Al–Gd approximant the maximum magnetic entropy change $\\Delta S_M^{\\max}$ departs from mean-field behavior and is largest exactly where the mean-field picture fails. Across the ferromagnetic window $e/a = 1.60$–$1.86$, where $T_C$ runs from 10 to 30 K, $\\Delta S_M^{\\max}$ for a 5 T field change stays nearly flat instead of following the $T_C^{-2/3}$ law from Eq. (2), and the field exponent $n$ (0.65–0.88) moves away from the mean-field value $2/3$ as $e/a$ approaches either FM phase boundary. At the FM/AFM border the response increases toward the antiferromagnetic side under high fields, reaching about 7.2 J/K mol-Gd for a 5 T field change, about 1.4–1.9 times the values previously reported for Tsai-type approximants and comparable to candidate materials for low-temperature magnetic refrigeration. The paper interprets the effect as an anomalous high-field magnetic response, presumably tied to enhanced spin fluctuations near phase boundaries, and proposes that tuning the magnetic ground state across a phase boundary by electron concentration is a route to larger magnetocaloric responses in rare-earth intermetallics.","pith_inferences":["Beyond the paper: a direct test of the proposed spin-fluctuation origin would be to measure the NMR relaxation rate or specific heat across the same $e/a$ range and check that fluctuations intensify near the FM/AFM and FM/SG borders.","Beyond the paper: the high-field enhancement in the AFM region is likely connected to the metamagnetic transition near 0.5 T observed at 2 K; measuring $\\Delta S_M$ through that transition at several temperatures could separate the forced-ferromagnetic contribution from the anomalous boundary effect.","Beyond the paper: applying the same analysis to Au–Al–Tb or Au–Al–Dy approximants, or to 2/1 approximants, would test whether the enhancement is generic to Tsai-type compounds or specific to the Gd system.","Beyond the paper: the nearly flat $\\Delta S_M^{\\max}$ versus $T_C$ trend suggests that at fixed field the working temperature of a refrigerant could be shifted by composition without sacrificing entropy change, which is useful for matching a specific cooling load."],"forward_implications":["If the claim is right, replacing Au by Al (or vice versa) to move $e/a$ toward the FM/AFM boundary can raise the magnetocaloric response of Tsai-type approximants to levels competitive with established low-temperature refrigerants near 10 K.","Mean-field scaling laws such as $\\Delta S_M^{\\max} \\propto T_C^{-2/3}$ are not reliable design rules near magnetic phase boundaries, so composition-dependent measurements are needed instead of extrapolating from one Curie temperature.","Antiferromagnetic compounds near a FM/AFM boundary should not be excluded as magnetocaloric candidates at high fields, because their entropy change can exceed that of ferromagnets of the same rare earth at the same field.","If $e/a$ tuning works through RKKY interactions generally, other rare-earth intermetallic systems with adjustable electron count can be pushed toward a magnetic phase boundary to enhance their $\\Delta S_M$.","The comprehensive $\\Delta S_M$ map identifies the $e/a$ windows worth optimizing for cooling applications in related approximants and quasicrystals."],"supporting_citations":[{"why":"Establishes the $e/a$-dependent magnetic phase diagram (AFM/FM/SG) that the $\\Delta S_M$ map is built on.","marker":"[8]"},{"why":"Provides the earlier $\\Delta S_M$ values in FM Au–Al–RE 1/1 approximants that this study compares with and exceeds.","marker":"[25]"},{"why":"Supplies Eq. (2), the mean-field $T_C^{-2/3}$ prediction whose breakdown is the paper's central anomaly.","marker":"[34]"},{"why":"Documents the conventional trend of $\\Delta S_M^{\\max}$ decreasing with $T_C$ and gives Laves-phase comparison values.","marker":"[33]"},{"why":"Provides the survey showing few rare-earth compounds exceed 6 J/K mol-RE for 5 T in the 2–40 K range, supporting the significance of the 7.2 J/K mol-Gd value.","marker":"[41]"},{"why":"States the theoretical expectation that ferromagnets give larger $\\Delta S_M$ than antiferromagnets for the same rare earth, which the observed AFM enhancement contradicts.","marker":"[42]"},{"why":"Earlier critical-behavior and magnetocaloric study of a FM Gd–Au–Si approximant that concluded mean-field behavior, in contrast to the present result.","marker":"[35]"},{"why":"Earlier comparative study of ferromagnetism in non-Heisenberg approximants that reported mean-field behavior, the limited $e/a$ window this paper widens.","marker":"[36]"}],"fun_headline_variants":["Anomalous magnetocaloric boost at FM/AFM boundary in Au-Al-Gd","Au-Al-Gd magnetocaloric anomaly peaks at FM/AFM boundary","Mean-field fails: magnetocaloric peak shifts to AFM side","7.2 J/K mol-Gd peak lands where mean-field breaks"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the mean-field formula used for comparison is accurate at the measured field and temperature values; if the Taylor expansion of the Brillouin function fails when 5 T is large relative to a 10 K Curie temperature, the flat $T_C$ dependence could be a saturation effect rather than a true breakdown of mean-field theory.","fun_headline_variants_meta":{"raw":{"variants":["Anomalous magnetocaloric boost at FM/AFM boundary in Au-Al-Gd","Au-Al-Gd magnetocaloric anomaly peaks at FM/AFM boundary","Mean-field fails: magnetocaloric peak shifts to AFM side","7.2 J/K mol-Gd peak lands where mean-field breaks"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00059,"raw_usage":{"total_tokens":2857,"prompt_tokens":1125,"completion_tokens":1732,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":741,"completion_tokens_details":{"reasoning_tokens":1651}},"tokens_in":741,"tokens_out":1732,"duration_ms":12685,"temperature":1.0,"reasoning_tokens":1651,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:57:55.879675+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Plot the field-cooled magnetization at $T_C$ for all ferromagnetic samples against $\\mu_0 H/T_C$: if the data collapse onto one universal Brillouin-type curve, the claimed breakdown of mean-field theory is an artifact of the expansion's range, and the near-independence of $\\Delta S_M^{\\max}$ on $T_C$ would be saturation, not anomaly.","supporting_citations":[{"cited_title":"Ishikawa, T","cited_arxiv_id":null,"evidence_quote":"Establishes the $e/a$-dependent magnetic phase diagram (AFM/FM/SG) that the $\\Delta S_M$ map is built on."},{"cited_title":"Kikugawa, T","cited_arxiv_id":null,"evidence_quote":"Provides the earlier $\\Delta S_M$ values in FM Au–Al–RE 1/1 approximants that this study compares with and exceeds."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies Eq. (2), the mean-field $T_C^{-2/3}$ prediction whose breakdown is the paper's central anomaly."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents the conventional trend of $\\Delta S_M^{\\max}$ decreasing with $T_C$ and gives Laves-phase comparison values."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the survey showing few rare-earth compounds exceed 6 J/K mol-RE for 5 T in the 2–40 K range, supporting the significance of the 7.2 J/K mol-Gd value."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the theoretical expectation that ferromagnets give larger $\\Delta S_M$ than antiferromagnets for the same rare earth, which the observed AFM enhancement contradicts."},{"cited_title":"Shiino, G","cited_arxiv_id":null,"evidence_quote":"Earlier critical-behavior and magnetocaloric study of a FM Gd–Au–Si approximant that concluded mean-field behavior, in contrast to the present result."},{"cited_title":"Labib, T","cited_arxiv_id":null,"evidence_quote":"Earlier comparative study of ferromagnetism in non-Heisenberg approximants that reported mean-field behavior, the limited $e/a$ window this paper widens."}],"review_version":1}