{"id":"3eddcdea-6d03-4536-957d-c3ec72f8225e","arxiv_id":"2510.16710","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A 3D FRG scheme without a hard-core reference system describes the Lennard-Jones liquid near its critical point with thermodynamic consistency superior to HNC/PY/KH closures.","lead":"This paper extends a functional renormalization group (FRG) method to three-dimensional liquids and tests it on a Lennard-Jones fluid, reporting better thermodynamic consistency than the HNC, PY, and KH integral-equation closures. The method is a parameter-free analytical alternative to simulation for computing pressures and pair structures near the liquid–gas critical point.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"KSA truncation error is uncontrolled and flow-path dependence untested, so claimed thermodynamic consistency may be an artifact.","rationale":"The reader identified the KSA truncation accuracy as the weakest assumption. I agree that the uncontrolled truncation is the central vulnerability, but I sharpen it: the paper itself states in Sec. II C that the KSA introduces dependence on the λ-evolution of the interaction, and no test of this path dependence is provided. This makes the reported thermodynamic consistency potentially flow-specific rather than a robust property of the FRG scheme. The concrete test of using a different flow parameterization directly addresses the load-bearing issue. The reader's verdict of CONDITIONAL is appropriate; the paper should be revised to include such a test or explicit error bounds. My agreement is partial because the reader did not explicitly emphasize the untested path-dependence, which I believe is the most concrete manifestation of the truncation error.","tokens_in":14814,"tokens_out":5086,"duration_ms":45716,"concrete_test":"Recompute the FRG solution at (ρ*,T*)=(0.5,1.4) using a linear λ-switch vλ(r)=λv(r) instead of the range cutoff of Eq. (20), keeping the same grids and Lmax=10. Compare the resulting βP, F_ex/N, ψ_ex, and the virial/compressibility spread with the published values. If these quantities shift by more than ~5% or the route spread widens substantially, the KSA truncation is not robust and the central claim is flow-path dependent.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central claim (Sec. IV) is that the FRG with KSA truncation preserves thermodynamic consistency far better than HNC/PY/KH and tends to match MD. The entire argument rests on the validity of the second-order KSA, Eq. (14), over ρ*≤0.5, T*≈1.25–1.4. The authors provide only plausibility arguments: a small-λ equivalence to a density expansion, and 'partial inclusion' of higher-order correlations. No quantitative error estimate or convergence check is given. Moreover, Sec. II C explicitly admits that with KSA the final results depend on the chosen λ-evolution of the potential, yet only a single flow path (range cutoff, Eq. (20)) is used. This is a critical gap. If a different, equally plausible flow (e.g., linear switching vλ=λv) yields significantly different pressures or a larger spread among the three thermodynamic routes, then the observed consistency is a coincidence of the particular flow, not a robust property of the FRG framework. The abstract's additional claim of accuracy comparable to Rogers-Young is also unsubstantiated, as no RY calculation appears in the paper. These issues undermine the assurance that the method's success is not an artifact of the uncontrolled KSA truncation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper extends the authors' previous one-dimensional FRG scheme for classical liquids to three dimensions. Starting from an exact flow equation for the free-energy functional, the authors derive hierarchical flow equations for cavity distribution functions, truncate at second order using the Kirkwood superposition approximation, and reduce the relevant spatial integrals to a manageable form via Legendre expansion and a range-cutoff flow vλ(r)=v(r)θ(λr_cut−r). The method is applied to a Lennard-Jones fluid at T*=1.4 for ρ*≤0.5 and also near T*≈1.3. Pressure, excess free energy, excess chemical potential, and g(r) are compared with molecular dynamics and with HNC, PY, and KH integral-equation closures using virial and compressibility routes. The authors report that the FRG preserves thermodynamic consistency among flow, virial, and compressibility routes much better than the tested closures, reproduces MD results more closely, and captures pressure softening near the critical temperature. A numerical breakdown for ρ*>0.5 and in the spinodal region is also reported.","tokens_in":15119,"tokens_out":5306,"duration_ms":47812,"significance":"If the central consistency claim survives closer scrutiny, this would be a meaningful advance: a parameter-free, non-perturbative liquid-state method that avoids hard-core reference systems and competes with modern closures. The paper has clear strengths: no parameters are fitted to MD data; the comparison of flow, virial, and compressibility routes is a genuine internal consistency check; the Legendre-based reduction of the three-dimensional integrals is nontrivial and makes the calculation feasible; and the external benchmarks against MD are appropriate. The main caveats are the uncontrolled KSA truncation, the untested dependence of results on the chosen λ-flow, and the absence of the advertised Rogers-Young comparison. These caveats determine whether the observed consistency is a robust property of the FRG framework or a consequence of the specific numerical choices made here.","major_comments":[{"comment":"The abstract states that FRG can achieve accuracy comparable to the Rogers-Young closure, but no RY calculation appears anywhere in the manuscript. The benchmark set is limited to HNC, PY, and KH. This is a load-bearing comparative claim that is currently unsupported. Please either add RY results for the same state points and routes, or remove/qualify the claim.","section":"Abstract; §III A; §IV"},{"comment":"Section II C explicitly acknowledges that with the KSA truncation the final results depend on the chosen λ-evolution of the interaction, yet all numerical results are obtained with the single range-cutoff flow (20). This is load-bearing: if another plausible flow (e.g., linear switching vλ=λv, or a flow that mixes in attraction during the early stage) yields materially different pressures or a larger spread among the three thermodynamic routes, then the observed thermodynamic consistency could be an accident of the chosen flow rather than a property of the FRG scheme. Please provide a flow-path dependence test at least at one state point, e.g., T*=1.4, ρ*=0.5, and ideally an estimate of the KSA error, e.g., by comparing with the Abe correction cited as Ref. [53] or by retaining third/fourth-order correlations.","section":"§II B; §II C 2; Eqs. (14), (20)"},{"comment":"The conclusion states that the method 'successfully captur[es] the presence of both the critical point and the first-order phase transition,' but the evidence in Fig. 4 is a pressure curve with a decreasing branch in a region labeled metastable. No Maxwell construction, no coexistence calculation, no quantitative spinodal criterion, and no MD benchmark near T_c is provided. The pressure softening is suggestive, but the conclusion overstates what can be inferred from the present data. Please either add a quantitative criterion for the spinodal/metastable boundaries and a comparison with known coexistence data, or soften the claim to 'the pressure develops an inflection/softening consistent with the presence of a critical point.'","section":"§III B; Fig. 4; §IV"}],"minor_comments":[{"comment":"The sentence 'With rcut = 8σ, these quantities are seen to converge at λ=1' is unclear: Fig. 7 appears to show results for rcut=5σ. Please clarify whether this is a convergence check with respect to rcut and show the corresponding curves if not already shown.","section":"Appendix A"},{"comment":"The shaded regions labeled 'spinodal' and 'metastable' are described only as 'suggested by the results.' Please define the exact criterion used to assign these regions (e.g., where dP/dρ<0, or where the flow becomes numerically unstable).","section":"Fig. 4"},{"comment":"The sentence explaining how F_ex/N and ψ_ex are obtained from the virial and compressibility routes is grammatically tangled ('are obtained from the resulting P using Eq. (29) and the relation Eq. (30)'). Please rewrite for clarity.","section":"Eq. (29) and surrounding text"},{"comment":"There is a typo: 'confirme the convergence' should be 'confirm the convergence.'","section":"Appendix A"},{"comment":"References [17] and [49] appear to have incomplete bibliographic information (missing journal volume/year or article number). Please complete them.","section":"References"},{"comment":"The MD simulation description does not mention tail corrections or equilibration time; adding this would help assess the numerical MD benchmark.","section":"Appendix B 1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is potentially a solid contribution, but the advertised Rogers-Young comparison is missing and the flow-path dependence is untested despite being explicitly acknowledged as a consequence of the KSA truncation. I would condition acceptance on either adding the RY benchmark and a flow-path test, or substantially qualifying the claims in the abstract and conclusion. The paper fits the journal's scope and the numerical implementation is substantial."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper delivers a real 3D implementation of the FRG scheme from the authors' 1D work, with a clever Legendre-expansion reduction of the angular integrals and a range-cutoff flow that turns the potential derivative into a delta function. The numerical results near the LJ critical region are credible: the three thermodynamic routes (flow, virial, compressibility) stay close to each other and to MD at ρ*≤0.5, which is a clear improvement over HNC/PY/KH. That part, the central claim, I buy. The abstract then overshoots: it says accuracy comparable to Rogers-Young, but the paper never computes RY. That claim should either be deleted or substantiated. There isn't a single mention of RY in the body, so the abstract is simply not supported by the evidence.\n\nThe method itself looks sound. The derivation is competently laid out, the KSA truncation is explicit, and the authors are honest that the results depend on the chosen λ-flow once KSA is imposed. What they don't do is test that dependence. They argue the range-cutoff flow maps to a density evolution at small λ, which is plausible, but there is no second flow path (e.g., linear switching) to show the TC is robust rather than a property of this particular path. The same goes for the KSA: they give plausibility arguments but no error estimate or higher-order truncation check. That's the biggest soft spot, and it's the one a referee should press on. The MD benchmarks use 256 particles and no error bars—minor for a demonstration, but it limits the strength of the numerical accuracy claims.\n\nThe paper is also transparent about the method's current limits: it breaks down above ρ*~0.5 and near the spinodal. That confines the demonstration to a narrow window, which is fine as a proof of concept but means the 'new method' claim is still provisional. The water extension is speculation.\n\nNet: the central result holds as a demonstration; the RY claim is wrong as written; the flow-dependence/KSA-error gap is real but not fatal. I'd send this to peer review. The authors should add or retract the RY comparison, show error bars, and ideally run one alternative flow path (or justify why none is needed). Then it's a solid methods paper for the liquid-theory crowd.","headline":"A credible 3D FRG method with good thermodynamic consistency in a narrow LJ window, but the abstract overstates against Rogers-Young and the flow-path dependence is untested.","tokens_in":15534,"tokens_out":3377,"would_cite":true,"duration_ms":31127,"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":"A functional renormalization group flow for classical liquids, truncated at pair level, preserves thermodynamic consistency far better than integral-equation closures and reproduces molecular dynamics results for the three-dimensional Lenna","keywords":["functional renormalization group","classical liquids","Lennard-Jones","thermodynamic consistency","integral equation closures","Kirkwood superposition approximation","pair distribution function","critical point"],"falsifier":"At ρ* = 0.5, run the flow with a third-order truncation that solves for y^(3) from its own flow equation instead of imposing KSA on it; if the virial-route pressure shifts by more than the difference between FRG and MD, the KSA is the main source of error. Alternatively, compute the third virial coefficient from the flow's ρ→0 limit and compare with the exact virial coefficient for the Lennard-Jones potential.","tokens_in":14701,"feed_emoji":"💧","tokens_out":5129,"duration_ms":43799,"temperature":0.7,"pith_summary":"This paper extends a functional renormalization group (FRG) scheme for classical liquids from one to three dimensions, without ever needing a hard-sphere reference fluid. The authors show that a second-order truncation of the flow-equation hierarchy, closed with the Kirkwood superposition approximation, yields thermodynamic quantities (pressure, excess free energy, excess chemical potential) that agree far more closely among the three standard thermodynamic routes than do the HNC, PY, and KH integral-equation closures. Against molecular dynamics simulations of the Lennard-Jones liquid at T* = 1.4 and densities up to about 0.5, the FRG results are consistently more accurate than those closures, especially the pair distribution function's first peak. Near the critical temperature, the flow equations reproduce the softening of the pressure isotherm and break down in the spinodal region, which the authors read as evidence that the method captures the liquid-gas transition. The paper therefore claims FRG is a genuinely new, parameter-free route to liquid structure that is competitive with modern integral-equation theory.","feed_headline":"Renormalization-group flow preserves liquid thermodynamics","feed_subtitle":"Near the critical point, the flow equations beat standard closures against molecular dynamics.","key_machinery":"The engine is the exact functional flow equation for the free energy density functional with respect to a parameter λ that switches on the pair interaction. Rewriting the derived hierarchy in terms of cavity distribution functions y^(n) removes the apparent divergences of hard-core interactions, so the flow can start from the ideal-gas reference and switch on the full Lennard-Jones potential. At second order the hierarchy is closed by the Kirkwood superposition approximation (KSA), which expresses y^(3) and y^(4) as products of y^(2)'s while preserving cluster decomposability. The spatial integrals are made tractable by a Legendre-polynomial expansion that turns three-dimensional double inte","core_discovery":"The central claim is that the functional renormalization group—implemented as a set of λ-flow equations for cavity distribution functions, truncated at two-body level with the Kirkwood superposition approximation and with the three-dimensional integrals handled by Legendre expansion—preserves thermodynamic consistency much more effectively than integral equation methods and tends to reproduce molecular dynamics results with higher accuracy for the three-dimensional Lennard-Jones liquid near its critical point. The paper demonstrates this at T* = 1.4 for densities up to ρ* ≈ 0.5, where the virial, compressibility, and direct-flow routes to pressure stay close to one another, while HNC, PY, an","pith_inferences":["The observed high-route consistency might be partly accidental: KSA bias could shift all routes in the same direction at these moderate densities. A decisive check would be to compute third or fourth virial coefficients with and without KSA and compare with low-density expansions.","Since the λ-flow with a range cutoff is interpreted at early stages as a density evolution (packing fraction η = πρ(λ r_cut)^3/6), the method hints at a deeper mapping between 'switching on the interaction' and 'increasing density'—a link that could connect FRG liquid theory with density-functional formulations.","If the flow is extended to higher densities without encountering hard-core divergences, FRG could replace the closure step in reference-interaction-site-model solvation treatments, offering a thermodynamically consistent route to solvent effects in electronic-structure calculations.","The Legendre cutoff L_max = 10 and the numerical cost suggest the method is currently heavier than standard integral-equation solvers; a practical extension would need adaptive or accelerator schemes, but the convergence of the l-sum appears benign enough to warrant trying."],"forward_implications":["If the claim holds, FRG gives a parameter-free liquid-state method whose thermodynamic consistency rivals or surpasses closures, without hard-sphere reference data.","The method reproduces the critical-region pressure softening, so it can be used to locate critical and spinodal behavior without inputting coexistence data.","At the studied state points, FRG's pair distribution function, particularly the first peak near r/σ = 1, is closer to molecular dynamics than HNC, PY, or KH.","The present breakdown at ρ* > 0.5 is attributed to the interaction-range flow, so redesigning that flow (e.g., mixing in attraction during the repulsive stage) is expected to extend validity to higher densities and eventually to realistic solvents.","Inclusion of higher-order distribution functions via the flow equations (beyond KSA) is a concrete, in-principle path to still better accuracy."],"fun_headline_variants":["Renormalization flow preserves liquid thermodynamics better than closures","FRG beats HNC and PY in thermodynamic consistency for Lennard-Jones liquids","Thermodynamic consistency from renormalization flow in 3D liquids","Renormalization group provides accurate thermodynamics for Lennard-Jones liquids"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the Kirkwood superposition approximation at second order keeps the flow equations accurate over the density range studied; if higher-order correlations are actually significant there, the thermodynamic consistency the paper reports is a property of the truncation, not of the FRG framework.","fun_headline_variants_meta":{"raw":{"variants":["Renormalization flow preserves liquid thermodynamics better than closures","FRG beats HNC and PY in thermodynamic consistency for Lennard-Jones liquids","Thermodynamic consistency from renormalization flow in 3D liquids","Renormalization group provides accurate thermodynamics for Lennard-Jones liquids"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000991,"raw_usage":{"total_tokens":4031,"prompt_tokens":730,"completion_tokens":3301,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":474,"completion_tokens_details":{"reasoning_tokens":3227}},"tokens_in":474,"tokens_out":3301,"duration_ms":19929,"temperature":1.0,"reasoning_tokens":3227,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T09:10:07.202584+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"At ρ* = 0.5, run the flow with a third-order truncation that solves for y^(3) from its own flow equation instead of imposing KSA on it; if the virial-route pressure shifts by more than the difference between FRG and MD, the KSA is the main source of error. Alternatively, compute the third virial coefficient from the flow's ρ→0 limit and compare with the exact virial coefficient for the Lennard-Jones potential.","supporting_citations":[],"review_version":1}