{"id":"317859a8-78b0-46c7-b7cf-8467d79a7279","arxiv_id":"1909.00432","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For driven lattice gases in slow contact, chemical potentials can be defined only if the contact dynamics factorizes and obeys macroscopic detailed balance, and they fail to obey an equation of state.","lead":"This paper develops a theory of chemical potentials for two driven particle systems joined by a slow contact. It shows that such potentials exist only under special contact dynamics, and that they depend on the contact itself, not just on each system.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Finite-epsilon corrections to the slow-exchange factorization (Eq. 11) are not quantified; without an epsilon-to-zero extrapolation the KLS validation of mu_cont equalization remains incomplete.","rationale":"The reader's weakest assumption matches my main concern. The factorization Eq. (11) is not an internal inconsistency; it is a quasi-static assumption that becomes exact as tau_c/tau_b tends to infinity. The authors are transparent about this and about the generic breakdown for finite epsilon. What is missing is numerical control of the limit in the KLS application: epsilon = 0.01 is used without an extrapolation, and since the KLS stationary distribution is not known analytically, agreement at a single epsilon is weaker evidence than the exact-model check. I do not see a flaw in the formal derivation of additivity from macroscopic detailed balance plus factorization; the Hamilton-Jacobi reduction (Eqs. 13, 19, 21) is standard and the algebraic construction of mu_cont (Eqs. 28-31) follows from the stated assumptions. The unproved Eq. (44) is a separate limitation, already acknowledged by the authors; it affects only the relation mu_cont = mu_iso + eta, not the main construction of contact chemical potentials or the equalization prediction. Hence the reader's CONDITIONAL verdict is appropriate and I recommend no change.","tokens_in":32834,"tokens_out":20860,"duration_ms":192528,"concrete_test":"Repeat the KLS contact simulations of Fig. 3 (JA = JB = 1, fA = 6, fB = 0, 20x20 lattices, exponential contact rule) for contact rates epsilon = 0.1, 0.01, 0.001, and 0.0001. Plot the measured stationary densities rho_A, rho_B and the predicted equalization curve from Eq. (69) as functions of epsilon. If the measured densities converge to the equalization prediction in the epsilon-to-zero limit, the finite-epsilon concern is resolved; if they do not converge, the KLS validation is not in the slow-exchange regime and the empirical support weakens.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central additivity criterion and the resulting contact chemical potentials rest on the slow-exchange factorization of conditioned microstate distributions, Eq. (11), justified by the timescale separation tau_c >> tau_b in Sec. III A. This is the premise that makes the coarse-grained transition rates Eq. (10) correct: if it fails, I'(rho_A|rho_bar) is no longer given by Eq. (21) and mu_cont equalization is not valid. The authors explicitly state in Sec. VII that additivity is generically broken beyond the slow-exchange limit. Yet the numerical tests of the KLS model (Sec. V E) and the exactly solvable model (Sec. V D) use only epsilon = 0.01 and do not show convergence as epsilon decreases. The exact model has analytic large-deviations predictions that match at that epsilon, but the KLS check is purely numerical, and no data or code are provided. Thus finite-epsilon corrections, which the theory does not control, are an unquantified threat to the empirical part of the central claim. The logical derivation in the stated limit is not affected; the concern is whether the simulations actually realize that limit.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a large-deviations framework for two driven lattice gases exchanging particles through a slow contact. In the limit where contact jumps are rare compared to bulk relaxation, the coarse-grained density dynamics is governed by a Hamilton-Jacobi equation. The authors show that if the macroscopic transition rates at contact satisfy a macroscopic detailed-balance relation and factorize into system-dependent factors, then the stationary large-deviations function is additive and one can attach contact chemical potentials μ_cont_A, μ_cont_B that equalize in the steady state. They argue that these potentials generally depend on the contact dynamics and therefore do not obey an equation of state, and they relate them to equilibrium or isolated-system chemical potentials through excess terms. The framework is applied to an exactly solvable driven lattice gas and to the KLS model using numerical simulations.","tokens_in":32979,"tokens_out":12100,"duration_ms":114908,"significance":"If the central claims hold, the paper clarifies the microscopic conditions under which a nonequilibrium chemical potential can be defined via additivity of the density large-deviations function, and it sharpens earlier phenomenological discussions by Pradhan, Seifert, Sasa, Tasaki, and Chatterjee et al. The Hamilton-Jacobi/additivity derivation in Secs. III–IV is internally consistent, and the central equalization prediction is parameter-free. The exactly solvable model provides a nontrivial analytic test, and the KLS comparison is a genuine numerical check rather than a fit. The negative corollary that the nonequilibrium chemical potential is contact-dependent and lacks an equation of state is clearly formulated. The paper is also careful to label Eq. (44) as a postulate, which limits the generality of the μ_cont–μ_iso relation but does not affect the core additivity criterion.","major_comments":[{"comment":"The slow-exchange factorization P(CA,CB|ρA,ρB)=PA(CA|ρA)PB(CB|ρB) is the load-bearing premise for the coarse-grained rates (10) and all subsequent results, and the authors correctly note in Sec. VII that additivity is generically broken beyond this limit. The numerical validations in Secs. V D and V E, however, use only ε=0.01 and provide no ε-dependence or extrapolation toward ε=0; the KLS comparison is purely numerical and no data or code are supplied. Since the theory does not quantify corrections in ε, agreement at a single finite value does not by itself establish that the simulations realize the ε→0 limit. Please add an ε-dependence study or at least an estimate of the leading correction, or explicitly present the comparisons as illustrative rather than as direct validation of the limit.","section":"Secs. V D and V E, Eq. (11)"},{"comment":"The authors state that they 'postulate, without proof' the representation PVk,k(Ck|ρk)=Zneq^{-1}e^{-βHk+Υneq}. The subsequent relation μcont_k=μiso_k+ln(φ_{k,ΔΥneq}/φ_k), Eq. (46), is therefore conditional on an unproved ansatz. The central additivity result does not depend on Eq. (44), but the paper should either present Eq. (46) explicitly as a conjecture or provide evidence for the assumed form of the stationary distribution.","section":"Sec. IV C 3, Eq. (44)"},{"comment":"The text says that for the Kawasaki rule the coarse-grained rates do not factorize, 'so that the large-deviations function is not additive, implying that a chemical potential cannot be defined.' This inference is not valid as stated, because Sec. IV A 1 explicitly notes that factorization is a sufficient but not necessary condition. One needs to show that the ratio φ(ρA,−1)/φ(ρA,+1) in the Kawasaki example cannot be written as a difference of single-density functions, or the conclusion should be qualified as restricted to the class of factorized contacts.","section":"Sec. V C 2"},{"comment":"The printed formulas are inconsistent with the definitions. Eq. (30) defines μcont_k as the logarithm of a ratio, but Eqs. (69) and (72) present μcont_k as a bare ratio, omitting the logarithm. In Eq. (63), the numerator should follow from Eq. (B8) with a factor μ(n+1) in the exponent rather than μn. These inconsistencies make the theoretical curves in Figs. 2, 3, and 5 irreproducible from the text; please correct the formulas and state explicitly which quantity was actually computed in the simulations.","section":"Secs. V C–V E, Eqs. (63), (69), (72)"}],"minor_comments":[{"comment":"The caption says 'Densities ρA (red) and ρB (blue) versus time,' but the horizontal axis is labeled f_A; please correct the caption or the axis label.","section":"Fig. 2 caption"},{"comment":"The KLS simulation results are reported without error bars or statistical uncertainty estimates; please add standard errors or state that fluctuations are smaller than the symbol size.","section":"Figs. 3–6"},{"comment":"No data or code availability statement is included; given that the KLS validation is an important part of the paper, a reproducibility statement would be helpful.","section":"Reproducibility"},{"comment":"The statement that macroscopic detailed balance is 'always verified' for single-particle exchange would be clearer if the two-term cancellation in Eq. (20) were shown explicitly for ΔNA=±1.","section":"Sec. III C"},{"comment":"Ref. [52] is cited as 'to be published'; if the perturbative solution discussed in Sec. III C is needed, please either include the relevant calculation or remove the forward reference.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"To the editor: the manuscript is honest about the unproved Eq. (44), and I do not consider the use of the authors' previous exact solution for the solvable model problematic. The main reasons for major revision are the missing finite-ε extrapolation in the numerical validation and the equation-level inconsistencies in Sec. V. I would also encourage requesting a data/code availability statement for the KLS simulations."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a solid theory paper, not a breakthrough. Its real contribution is a clean set of sufficient conditions for the large-deviations function of densities to be additive when two driven lattice gases are weakly coupled: macroscopic detailed balance at the contact plus factorization of the coarse-grained exchange rates. When those hold, each system gets a contact chemical potential that equalizes at the steady state; when they fail (Kawasaki rates), additivity breaks and no such potential exists. That gives a general criterion for when Sasa-Tasaki's chemical potentials actually work, and it goes beyond earlier short-range-correlation results. It also confirms that these potentials depend on the contact, so no equation of state—a useful negative result.\n\nThe paper does several things well. The derivation in Secs. III–IV is internally consistent, and the slow-exchange limit is clearly stated. The exactly solvable model gives parameter-free formulas that match the numerics without any fitting. The KLS simulations are a genuine test: the theoretical μ_cont functions predict the density shift, and the two contact rules (exponential vs Sasa-Tasaki) give different shifts, which is exactly what the framework predicts. The authors also flag their own weak spots: Eq. (44) is admitted to be a postulate, and the conclusion explicitly notes that additivity is generically broken beyond the slow-exchange limit.\n\nWhere are the soft spots? The load-bearing factorization Eq. (11) is asymptotic in ϵ→0, and all simulations are done at ϵ=0.01. There is no epsilon-dependence check, so we do not see whether the results extrapolate to the limit or whether 0.01 is simply small enough. This is a real gap in the empirical confirmation, though not in the logical derivation itself. The KLS numerics come with no code or data, which makes it hard to verify that the measurement procedure was implemented as described. The μ_cont=μ_iso+η relation rests on the unproved Eq. (44) ansatz; this is the least secure part of the paper, but the main additivity criterion does not depend on it.\n\nWho should read this? People working on nonequilibrium thermodynamics of driven lattice gases, especially contact phenomena and zeroth-law violations. It is the kind of paper I would send to a careful referee rather than desk-reject: the central argument holds up, the gaps are identifiable and probably fixable. I would ask for an epsilon-sweep or at least a clear statement about finite-ϵ corrections, and for the KLS data and code. With those, this should be published.","headline":"A clean large-deviations criterion for when contact chemical potentials exist in driven lattice gases, with honest but unquantified finite-epsilon caveats in the numerics.","tokens_in":33548,"tokens_out":3079,"would_cite":true,"duration_ms":29978,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82C22","60F10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper establishes that two driven systems in slow particle exchange acquire well-defined contact chemical potentials, and that those potentials—not the bulk alone—determine the steady-state densities.","keywords":["nonequilibrium chemical potential","large deviations","lattice gas","slow exchange limit","additivity","contact dynamics","zeroth law","driven systems"],"falsifier":"Simulate one pair of bulk driven systems with the same parameters but two different factorized contact rules and compare the measured steady-state densities over a range of overall densities; the paper predicts the two curves differ, with the difference set by the contact chemical potentials. If the densities coincide, the claimed contact-dependence fails. Conversely, if a non-factorized heat-bath contact rule still yields an additive joint density large-deviations function $I(\\rho_A,\\rho_B)=\\gamma_A I_A(\\rho_A)+\\gamma_B I_B(\\rho_B)$, the sufficiency claim for factorization would be falsified.","tokens_in":32524,"feed_emoji":"⚖️","tokens_out":7481,"duration_ms":67703,"temperature":0.7,"pith_summary":"This paper tries to establish when two driven, out-of-equilibrium systems that exchange particles through a weak contact can be assigned chemical potentials at all, and what those potentials predict. The answer is conditional and partly negative: if the contact is slow enough and the exchange dynamics factorizes between the two systems, the large-deviations function of the densities becomes additive, so a contact chemical potential can be attached to each system and equalizes at the steady state. That potential, however, is not a function of the bulk state alone; it carries an imprint of the contact dynamics, so no equation of state exists generically. If the contact rule does not factorize, no chemical potential can be defined, even though the most probable densities still exist. A sympathetic reader would care because this pins down when the familiar equilibrium notion of a chemical potential survives in driven steady states, and why earlier numerical violations of the zeroth law appeared.","feed_headline":"Driven contacts get chemical potentials, but no equation of state","feed_subtitle":"Large-deviations theory shows the potential equalizes at contact, yet depends on the contact's own dynamics.","key_machinery":"The engine of the argument is the large-deviations (Hamilton–Jacobi) equation for the joint density distribution in the slow-exchange limit, together with two structural conditions imposed on it. Macroscopic detailed balance, $I'(\\rho_A|\\bar\\rho)=\\ln[\\phi(\\rho_A,-1)/\\phi(\\rho_A,1)]$, lets the ratio of the forward and backward coarse-grained exchange rates determine the derivative of the large-deviations function. Factorization, $\\phi(\\rho_A,\\Delta N_A)=\\nu_0\\phi_A(\\rho_A,\\Delta N_A)\\phi_B(\\rho_B,\\Delta N_B)$, splits that derivative into a difference of two single-system terms, yielding the contact chemical potential $\\mu^{\\rm cont}_k(\\rho_k)=\\ln[\\phi_k(\\rho_k,-1)/\\phi_k(\\rho_k,1)]$. The factorization is inherited from a microscopic contact rule of the form $T_c(C'_A,C'_B|C_A,C_B)=\\nu_0\\theta_A(C_A,C'_A)\\theta_B(C_B,C'_B)$, which holds for exponential and high-barrier Arrhenius rules but fails for heat-bath and Metropolis rules.","core_discovery":"The central claim is that additivity of the density large-deviations function, $I(\\rho_A,\\rho_B)=\\gamma_A I_A(\\rho_A)+\\gamma_B I_B(\\rho_B)$, is the precise condition under which a nonequilibrium chemical potential can be defined for two driven lattice gases in slow contact. The paper shows that two sufficient conditions make additivity hold: macroscopic detailed balance at contact, $I'(\\rho_A)=\\ln[\\phi(\\rho_A,-1)/\\phi(\\rho_A,1)]$, and factorization of the coarse-grained exchange rate, $\\phi(\\rho_A,\\Delta N_A)=\\nu_0\\,\\phi_A(\\rho_A,\\Delta N_A)\\phi_B(\\rho_B,\\Delta N_B)$. When both hold, each system carries a contact chemical potential $\\mu^{\\rm cont}_k(\\rho_k)=\\ln[\\phi_k(\\rho_k,-1)/\\phi_k(\\rho_k,1)]$, and the steady-state densities satisfy $\\mu^{\\rm cont}_A(\\rho_A^*)=\\mu^{\\rm cont}_B(\\rho_B^*)$. Because $\\mu^{\\rm cont}_k$ depends on the contact factors $\\phi_k$, it is not a bulk equation of state: different factorized contact rules move the predicted densities, as confirmed in the exactly solvable model and in KLS simulations, while a non-factorized rule destroys additivity and with it the chemical-potential description.","pith_inferences":["A natural extension would be to use a driven system with a known contact chemical potential as a probe of another driven system, with the caveat that the measured value will depend on the probe's own contact rule, so comparisons require identical contact dynamics.","One could test the slow-exchange boundary directly by measuring the joint density large-deviations function at increasing exchange rates; the paper's framework predicts that additivity breaks generically away from the slow limit, so observing additivity persist at finite rates would mark the regime where a chemical-potential description still works.","The same additivity criterion could be applied to other conserved quantities, such as energy or volume, yielding analogous contact temperature or contact pressure whose contact dependence would mirror the chemical-potential result.","Since macroscopic detailed balance alone is insufficient, a multi-particle exchange rule that factorizes in the bulk but not at the contact would provide a sharp numerical test of the factorization condition's role."],"forward_implications":["When macroscopic detailed balance and factorization both hold, the steady-state densities of two systems in contact are fixed by equality of contact chemical potentials, $\\mu^{\\rm cont}_A(\\rho_A^*)=\\mu^{\\rm cont}_B(\\rho_B^*)$, and this equality predicts the densities measured in simulations.","The contact chemical potential is not a bulk property: it depends on which factorized microscopic rule realizes the contact, so no equation of state exists generically; only a contact fine-tuned with the drive restores one.","The zeroth law of thermodynamics holds only within classes of systems that share a factorized contact rule, effectively including half of the contact in each system; transitivity fails otherwise.","$\\mu^{\\rm cont}_k$ can be expressed as the equilibrium chemical potential $\\mu^{\\rm eq}_k$ or the isolated-system potential $\\mu^{\\rm iso}_k$ plus an excess term measuring the nonequilibrium modification of the stationary distribution and/or extra work at the contact.","For non-factorized contact rules, the large-deviations function is non-additive and no chemical potential can be defined, even though macroscopic detailed balance holds for single-particle exchange and the most probable densities remain well defined."],"supporting_citations":[{"why":"Supplies the phenomenological slow-exchange, high-barrier contact dynamics and the chemical-potential definition that the paper formalizes and generalizes.","marker":"[2]"},{"why":"Introduces the additivity condition of the large-deviations function as a route to a nonequilibrium chemical potential for a single system, extended here to systems in contact.","marker":"[10, 11]"},{"why":"Reports earlier contact simulations showing zeroth-law deviations that the paper interprets as contact-dependent, non-additive behavior.","marker":"[16, 17]"},{"why":"Proposed conditions for additive large-deviations functions under short-range correlations and microscopic detailed balance; the paper contrasts its own factorized, drive-independent contact class with this approach.","marker":"[21]"},{"why":"Defines the KLS driven lattice gas used for the numerical validation of the predicted contact chemical potentials.","marker":"[28]"},{"why":"Supplies the large-deviations/Hamilton-Jacobi formulation used to connect the derivative of the rate function to the contact current and the stationary state.","marker":"[42]"},{"why":"Provides the exactly solvable driven lattice-gas model whose stationary distribution makes the analytic test of the chemical potentials possible.","marker":"[66]"}],"fun_headline_variants":["Chemical potentials for driven contacts: no equation of state","Large deviations define chemical potentials at driven contacts","Driven contacts: chemical potentials exist, but no bulk law","When additivity holds, contact chemical potentials appear","Contact dynamics shape chemical potentials, breaking equations of state"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole construction rests on a strict separation of time scales: particle exchanges across the contact are so rare that, conditioned on the current densities, the two systems' microstates are statistically independent, each in its own isolated steady state.","fun_headline_variants_meta":{"raw":{"variants":["Chemical potentials for driven contacts: no equation of state","Large deviations define chemical potentials at driven contacts","Driven contacts: chemical potentials exist, but no bulk law","When additivity holds, contact chemical potentials appear","Contact dynamics shape chemical potentials, breaking equations of state"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000283,"raw_usage":{"total_tokens":1705,"prompt_tokens":1009,"completion_tokens":696,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":625,"completion_tokens_details":{"reasoning_tokens":621}},"tokens_in":625,"tokens_out":696,"duration_ms":65621,"temperature":1.0,"reasoning_tokens":621,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:52:56.818417+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate one pair of bulk driven systems with the same parameters but two different factorized contact rules and compare the measured steady-state densities over a range of overall densities; the paper predicts the two curves differ, with the difference set by the contact chemical potentials. If the densities coincide, the claimed contact-dependence fails. Conversely, if a non-factorized heat-bath contact rule still yields an additive joint density large-deviations function $I(\\rho_A,\\rho_B)=\\gamma_A I_A(\\rho_A)+\\gamma_B I_B(\\rho_B)$, the sufficiency claim for factorization would be falsified.","supporting_citations":[{"cited_title":"Bertin, O","cited_arxiv_id":null,"evidence_quote":"Defines the KLS driven lattice gas used for the numerical validation of the predicted contact chemical potentials."},{"cited_title":"Chatterjee, P","cited_arxiv_id":null,"evidence_quote":"Supplies the large-deviations/Hamilton-Jacobi formulation used to connect the derivative of the rate function to the contact current and the stationary state."},{"cited_title":"Spohn, J","cited_arxiv_id":null,"evidence_quote":"Provides the exactly solvable driven lattice-gas model whose stationary distribution makes the analytic test of the chemical potentials possible."}],"review_version":1}