{"id":"bee8240b-f99e-47b8-86eb-47ffa48c9531","arxiv_id":"2501.16187","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A liquid heat capacity model using a linear-in-frequency density of states for low-frequency shear modes, plus intramolecular vibrations, outperforms the earlier phonon model on 23 liquids.","lead":"Scientists compared four ways to compute how much heat a liquid stores, focusing on how slow 'sloshing' motions are counted. Treating those slow motions as overdamped, liquid-like excitations rather than solid-like vibrations matched measured heat capacities most accurately across liquids.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Liquid-like model's edge rests on an incomplete quasi-harmonic derivative: Eq. (17)'s T-dependent DOS is inserted into Eq. (9), which was derived assuming a T-independent g, so the reported cv omits ∂g/∂T terms.","rationale":"I read the paper as an attempt to replace the Debye DOS for low-frequency shear modes with a linear-density-of-states description while keeping the phonon-free-energy framework. The authors honestly flag the original model's questionable treatment of overdamped modes, but their own liquid-like model repeats a version of the same move: Eq. (17) is put into Eq. (9) without re-deriving the free energy for a T-dependent, damped spectrum. The most concrete internal symptom is the quasi-harmonic derivative. Since τ(T) enters g_s explicitly, the (1 + αT/2) factor in Eq. (9) cannot capture ∂g_s/∂T; the missing term is not necessarily small because A(T) and the crossover frequency shift substantially over the temperature ranges in Figs. 2-4. This concern is independent of whether one accepts that overdamped modes are describable by harmonic oscillators, and it is more decisive than the ωF 2π inconsistency alone. The proposed check—computing cv both with and without the ∂g_s/∂T terms—would settle the issue. If the correction is negligible, the liquid-like model's empirical advantage survives and the paper's conclusion is reasonable, though the ωF convention should be fixed. If the correction is significant, the reported ranking is not a reliable test of the physical nature of low-frequency shear modes. I therefore retain the reader's CONDITIONAL verdict. Agreement is partial: the reader located the weakness in the free-energy functional, and I agree, but I identify a specific T-dependent-DOS derivative that is not mentioned in the reader's rationale.","tokens_in":12971,"tokens_out":19132,"duration_ms":189487,"concrete_test":"Recompute the liquid-like model for two representative systems, e.g., Ar and Ga, in two ways. (A) Follow the paper: insert Eq. (17) with τ(T) into Eq. (9), evaluate E(T), and take dE/dT numerically. (B) Start from the free energy in Eq. (7) with the same T-dependent g_s(ω; τ(T)), and evaluate F(T) numerically while including the explicit ∂g_s/∂T contribution (equivalently, add the missing term -k_B T^2 ∫ log(1 - e^{-ℏω/kBT}) (∂g_s/∂T) dω to method A). Compare the resulting cv(T) curves and the errors in Fig. 3 and Table I. If methods A and B differ by more than the reported 2-3x gap between the liquid-like and phonon models, the central claim is not established until the correct thermodynamic formula is used.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The most load-bearing assumption is that the liquid-like model can be evaluated by substituting the shear DOS of Eq. (17) into the harmonic phonon energy of Eq. (9) and then differentiating with respect to T. Equation (9) is derived from Eqs. (7)-(8) under the quasi-harmonic assumption dω/dT = -αω/2 with a fixed density of states g(ω); it contains no term for an explicit T-dependence of g. In the liquid-like model, however, g_s(ω) = A(T) ω^2 sqrt(1 + 1/(4τ(T)^2ω^2)) depends on T through the Maxwell time τ(T), and A is renormalized at each T to keep the integral over all shear modes equal to 2N. Starting from Eq. (7) with this T-dependent g_s adds a term -k_B T^2 ∫ log(1 - e^{-ℏω/kBT}) (∂g_s/∂T) dω, plus boundary terms at ωF, to the energy in Eq. (9). The paper does not compute or mention this term. Therefore the liquid-like model's cv(T) is not the thermodynamic derivative of the model it defines; its apparent 2-3x error reduction in Fig. 3 and Table I could be an artifact of an incomplete derivative rather than evidence about the nature of overdamped modes. The unresolved factor-of-2π versus factor-of-1 definition of ωF (Theory section vs Overdamped section) compounds this ambiguity, because the split in Eq. (6) changes with that convention.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript revisits the phonon theory of liquid heat capacity (Bolmatov, Brazhkin, and Trachenko) by questioning the standard treatment of low-frequency shear modes. It proposes a 'liquid-like' model in which modes below the Frenkel frequency are described by a linear-in-frequency density of states, Eq. (17), instead of a Debye DOS, and it adds intramolecular vibrational contributions, Eq. (19), for molecular liquids. The authors compare four models (original phonon, gas-like kinetic, zero, and liquid-like) against NIST REFPROP data for a set of liquids and against a neutron-scattering DOS benchmark for Ga. They report that the liquid-like model gives the best agreement, reducing the percentage error by a factor of 2-3 relative to the original phonon model, and that the gas-like and zero models fail at high temperature. The paper includes an extended comparison for 23 liquids and a discussion of water and heavy water.","tokens_in":13260,"tokens_out":19344,"duration_ms":172913,"significance":"If the central claim holds, the work supports the view that low-frequency shear modes in liquids are overdamped excitations with a linear DOS, and that a phonon-type free energy built on that DOS is thermodynamically viable. The model is parameter-free in the sense that no heat-capacity data are used as input: tau is obtained from NIST viscosity and G_infinity, alpha from prior literature, and vibrational frequencies from standard tables. The inclusion of intramolecular vibrations is a useful extension that lets the framework reproduce the non-monotonic and increasing cv(T) behavior of CO2 and C5H12. The Ga comparison against the experimentally measured DOS is a particularly clean test. However, the strength of the conclusion is currently limited by the incomplete thermodynamic treatment of the T-dependent DOS and by the small number of systems for which the liquid-like model is quantitatively compared with the other models.","major_comments":[{"comment":"The liquid-like model evaluates Eq. (9), which is derived from Eqs. (7)-(8) under the quasi-harmonic assumption domega/dT = -alpha*omega/2 with a temperature-independent density of states g(omega), using the T-dependent shear DOS g_s(omega) = A(T) omega^2 sqrt(1 + 1/(4 tau(T)^2 omega^2)) from Eq. (17). Because tau(T) and the normalization A(T), fixed by the condition integral g_s domega = 2N, depend on T, the correct energy obtained from Eq. (7) contains the additional term -k_B T^2 integral log(1 - exp(-hbar*omega/k_B T)) (partial g_s/partial T) domega, plus boundary terms from the T-dependent split at omega_F in Eq. (6). This term is not computed or estimated. Since the reported 2-3x error reduction in Fig. 3 and Table I is the central claim, the authors must either include this contribution in cv or demonstrate that it is negligible; otherwise the liquid-like model's cv is not the thermodynamic derivative of the model it defines.","section":"Overdamped liquid-like modes, Eqs. (9) and (17)"},{"comment":"The Frenkel frequency is defined inconsistently as omega_F = 2*pi/tau in the Theory section and as omega_F = 1/tau in the Overdamped section. The position of the split in Eq. (6) and the numerical value of the integrals over the shear DOS depend on omega_F, so the two conventions differ by a factor 2*pi and will produce different predictions. Please adopt a single convention, state the relation to the Maxwell relaxation time tau_M, and check the sensitivity of the reported errors to this choice.","section":"Theory section (after Eq. (1)) and Overdamped section (after Eq. (16))"},{"comment":"The liquid-like model continues to compute the energy of the modes below omega_F from the canonical phonon free energy, Eq. (7), and the quasi-harmonic result, Eq. (9), which describe harmonic oscillators with real frequencies. This is the same class of assumption that the paper criticizes in the original model (Section 'Theory', paragraph after Eq. (6)), since modes below omega_F are described in the text as overdamped and non-oscillatory. The paper should justify why the harmonic-oscillator free energy is applicable to overdamped modes (or state the approximation explicitly) before concluding that the model is physically better motivated.","section":"Overdamped liquid-like modes, Eqs. (7)-(9)"},{"comment":"The quantitative claim that the liquid-like model is the most accurate over the whole temperature range is established only for four liquids in Fig. 3 and for Ga in Table I. The extended analysis in Appendix A reports theoretical values for the liquid-like model only and does not compare its errors against the phonon, gas-like, and zero models for the 23 liquids. Please extend the model-by-model error comparison to the full data set, or qualify the universality claim accordingly.","section":"Appendix A and Fig. 3"},{"comment":"Eq. (17), the central ingredient of the liquid-like model, is imported from Refs. [28,35,36] without derivation. Since the model's improvement over the original phonon model depends entirely on this DOS, the paper should either reproduce the derivation or state clearly the assumptions leading to Eq. (17), including the meaning of tau in the overdamped regime and the explicit form of the normalization A(T).","section":"Overdamped liquid-like modes, Eq. (17)"}],"minor_comments":[{"comment":"The word 'straigthfroward' should be 'straightforward'.","section":"Introduction"},{"comment":"The name 'Brazkhin' should be 'Brazhkin' where Ref. [28] is mentioned.","section":"Overdamped liquid-like modes"},{"comment":"The text says the last term in Eq. (6) coincides with Eq. (14), but Eq. (14) is labeled E_s(omega < omega_F), while the last term in Eq. (6) is E_s(omega < omega_F)/2; please clarify which quantity is being set to N(omega < omega_F) k_B T/2.","section":"Eq. (14) and surrounding text"},{"comment":"The comparison section states that data were collected for 21 liquids, while the text before Appendix A and the appendix itself refer to 23 liquids; please reconcile the counts.","section":"Comparison to experimental data and Appendix A"},{"comment":"For the molecular liquids N2 and CH4, the caption says cv denotes the intermolecular part, but the procedure for subtracting rotational and intramolecular contributions from the NIST data is not described; please state the subtraction method or clarify what is plotted.","section":"Fig. 2 caption"}],"recommendation":"major_revision","confidential_remarks":"The missing partial g/partial T term in the liquid-like model is the most serious technical issue; if it proves numerically large, the central claim would be compromised. The inconsistent definition of omega_F and the limited quantitative comparison should be fixable in revision. The paper's use of Refs. [28,35,36], some coauthored by the present authors, is not itself problematic given the independent neutron-scattering support for the linear DOS."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper's liquid-like model is a genuine step forward. Plugging the linear-in-frequency shear DOS into the phonon free energy instead of the Debye form, and adding intramolecular vibrations, is a natural extension that clearly helps for molecular liquids like CO2 and C5H12. The Ga comparison using the experimentally measured DOS is a nice touch. But the central numerical claim is undercut by a thermodynamic gap that the authors don't address.\n\nThe problem is that Eq. (9) was derived assuming a T-independent g(ω). The paper then inserts the T-dependent g_s from Eq. (17) into that formula and differentiates. But the true energy obtained from F = E0 + kBT∫log(1−e^{−ℏω/kT})g(ω,T)dω contains an extra term, −kBT²∫log(1−e^{−ℏω/kT})(∂g/∂T)dω, plus boundary terms at ωF when g changes shape with T. That term is not computed or mentioned. So the cv values plotted and compared are not the thermodynamic derivative of the model the paper defines. The reported 2–3x error reduction could be an artifact of this incomplete derivative, not evidence about the nature of overdamped modes. This is the load-bearing flaw; the qualitative physics may well survive, but the quantitative claim isn't established.\n\nOther soft spots are minor by comparison. The Frenkel frequency is defined as 2π/τ in the Theory section and 1/τ in the Overdamped section, which changes the split in Eq. (6). The 23-liquid comparison in Appendix A is presented as tables without uncertainties or aggregate error statistics, so it's hard to judge significance. Eq. (17) is cited rather than derived, though it has independent experimental support; that's acceptable.\n\nWho this is for: people working on liquid thermodynamics, phonon models, and collective excitations. The paper deserves a serious referee, but the authors need to re-derive the energy from the free energy with T-dependent g, include the missing terms, and redo the comparison. As written, the main claim is not yet supported.","headline":"The liquid-like model is a good idea, but the main result rests on an incomplete thermodynamic derivative; the comparison needs to be redone before the claim is credible.","tokens_in":13792,"tokens_out":4960,"would_cite":false,"duration_ms":42985,"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":"The most accurate model of liquid heat capacity treats low-frequency shear modes as overdamped, not as propagating Debye waves, and adds intramolecular vibrations.","keywords":["liquid heat capacity","phonon theory of liquids","overdamped shear modes","density of states","Frenkel frequency","intramolecular vibrations","k-gap dispersion","Debye model"],"falsifier":"A direct test would be to compute the heat capacity of a liquid from molecular dynamics by explicitly separating the energy of shear modes below the Frenkel frequency and checking whether their contribution follows the harmonic-oscillator expression with a linear DOS, or whether it obeys a different statistical mechanics; if the overdamped modes do not contribute to the free energy in the assumed harmonic form, the predicted $c_v$ would deviate systematically at temperatures where those modes dominate the count.","tokens_in":12738,"feed_emoji":"🔬","tokens_out":2819,"duration_ms":26116,"temperature":0.7,"pith_summary":"This paper argues that the standard phonon theory of liquid heat capacity rests on a wrong assumption: it treats low-frequency shear modes as propagating waves with a Debye density of states. The authors propose instead that these modes are overdamped and non-propagating, with a density of states linear in frequency, and they add the missing contribution from intramolecular vibrations. Comparing four theoretical models against experimental data for 21 to 23 liquids, they find that this liquid-like model agrees with data across the whole temperature range, reducing the typical error by a factor of two to three relative to the original phonon model. If right, this settles a basic question about what kinds of excitations carry heat in liquids and provides a parameter-free way to predict heat capacity for simple and molecular liquids alike.","feed_headline":"Overdamped shear modes beat Debye phonons in liquid heat capacity","feed_subtitle":"Revising phonon theory with a linear density of states and intramolecular vibrations cuts error two- to threefold.","key_machinery":"The central object is the liquid-like density of states for shear modes, $g_s(\\omega) = A \\omega^2 \\sqrt{1 + 1/(4\\tau^2\\omega^2)}$, which interpolates between a Debye $\\omega^2$ form at high frequency and a linear $\\omega$ form below the Frenkel frequency $\\omega_F = 1/\\tau$. This DOS is inserted into the canonical phonon free energy $F_{ph} = E_0 + k_B T \\sum_i \\log(1 - e^{-\\hbar\\omega_i/k_B T})$ with a quasi-harmonic temperature shift, replacing the Debye DOS for the shear-mode terms in the total energy decomposition of Eq. (6). The second ingredient is the intramolecular vibrational heat capacity, $c^{vib}_v = R \\sum_i (\\Theta_{v,i}/T)^2 e^{\\Theta_{v,i}/T}/(e^{\\Theta_{v,i}/T}-1)^2$, which is added to the intermolecular (translational plus rotational) contribution to capture the rising heat capacity of molecular liquids.","core_discovery":"The central claim is that the heat capacity of liquids is best described by treating shear modes with frequency below the Frenkel frequency as overdamped liquid-like excitations with a density of states $g_s(\\omega) \\propto \\omega$ at low frequency, rather than as Debye-like propagating phonons. The paper derives this density of states from the gapped (k-gap) dispersion relation of collective shear waves, Eq. (16), yielding Eq. (17), which reduces to a linear DOS below $\\omega_F$. Using this DOS in the canonical phonon free energy, together with an explicit intramolecular vibrational contribution Eq. (19), the model reproduces experimental heat capacities for noble liquids, liquid metals, and molecular liquids, including non-monotonic temperature dependence in CO$_2$ and monotonically increasing $c_v$ in C$_5$H$_{12}$ that the original phonon model cannot capture. The paper shows that alternative treatments of the low-frequency modes as gas-like kinetic excitations or as completely absent produce significantly worse agreement, especially at high temperatures, and that the liquid-like model also outperforms existing approaches for liquid Ga when compared with the experimentally measured density of states.","pith_inferences":["If the linear low-frequency DOS is the correct thermodynamic counting for overdamped shear modes, the same density of states should also enter other liquid thermodynamic properties, such as entropy and free energy differences across the melting line, and could be tested in simulations that resolve mode contributions directly.","The model's success suggests that the distinction between propagating and overdamped excitations matters for thermodynamics, not just dynamics; one could extend the framework to supercritical fluids to see whether the liquid-like DOS also captures the heat capacity crossover near the Widom line.","The failure of the gas-like model at high temperature hints that any two-phase description of liquids that assigns a fixed gas-like fraction to low-frequency shear modes will systematically underestimate $c_v$; a temperature-dependent assignment based on the overdamped DOS might reconcile such approaches.","The water deviations may be reduced by adding a configurational term derived from the temperature derivative of the pair distribution function, which would be a direct test of whether the phonon framework can be extended to hydrogen-bonded networks."],"forward_implications":["The liquid-like model provides a parameter-free prediction of liquid heat capacity that is more accurate than the original phonon model across the full temperature range, cutting typical percent error by roughly a factor of two to three.","The success of the model implies that low-frequency shear modes in liquids are not propagating Debye waves; they are overdamped, non-propagating excitations whose thermodynamic contribution should be counted with a linear-in-frequency density of states.","Adding intramolecular vibrations makes the theory capable of describing molecular liquids whose heat capacity increases with temperature, such as CO$_2$ and n-pentane, a behavior the original phonon model cannot reproduce.","Treating low-frequency shear modes as gas-like kinetic excitations leads to heat capacities below $2R$ at high temperatures, a value often used as a marker of the liquid-to-gas crossover, so the gas-like picture is physically inconsistent with liquid thermodynamics.","The model correctly captures the isotope effect between H$_2$O and D$_2$O at moderate temperatures, though it does not yet handle water's strong configurational contributions near its anomalous regime."],"supporting_citations":[{"why":"Supplies the original phonon theory of liquid thermodynamics that this work extends and corrects.","marker":"[19]"},{"why":"Proposes the gas-like kinetic treatment of low-frequency shear modes and criticizes the Debye assumption.","marker":"[27]"},{"why":"Provides the argument that shear modes below the Frenkel frequency are overdamped and yields the linear DOS form used in Eq. (17).","marker":"[28]"},{"why":"Derives the linear low-frequency density of states from instantaneous normal mode theory, supporting the liquid-like DOS.","marker":"[34]"},{"why":"Derives the same linear behavior from the gapped dispersion relation of collective shear waves.","marker":"[35]"},{"why":"Provides the experimentally measured density of states and the Ga heat capacity comparison used in Table I.","marker":"[38]"},{"why":"Supplies the NIST REFPROP experimental heat capacity data for all liquids compared in the paper.","marker":"[4]"},{"why":"Gives the framework for computing the infinite-frequency shear modulus used to obtain the Maxwell relaxation time.","marker":"[41]"},{"why":"Supplies the intramolecular vibrational frequencies used in Eq. (19) for molecular liquids.","marker":"[42]"}],"fun_headline_variants":["Overdamped shear modes improve liquid heat capacity","Linear shear DOS best for liquid heat capacity","Phonon theory revised: overdamped shear modes for liquids","Shear modes as overdamped linear DOS improve liquid cv","Overdamped shear modes beat Debye phonons in liquids"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that overdamped, non-propagating shear modes below the Frenkel frequency can still be described by the canonical phonon free energy, that is, as harmonic oscillators with real frequencies and a quasi-harmonic temperature shift; if the statistical mechanics of overdamped excitations differs fundamentally from harmonic phonons, the liquid-like model's advantage would be an artifact of using a better-fitting density of states in the wrong free-energy functional.","fun_headline_variants_meta":{"raw":{"variants":["Overdamped shear modes improve liquid heat capacity","Linear shear DOS best for liquid heat capacity","Phonon theory revised: overdamped shear modes for liquids","Shear modes as overdamped linear DOS improve liquid cv","Overdamped shear modes beat Debye phonons in liquids"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000936,"raw_usage":{"total_tokens":4062,"prompt_tokens":1060,"completion_tokens":3002,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":676,"completion_tokens_details":{"reasoning_tokens":2924}},"tokens_in":676,"tokens_out":3002,"duration_ms":19455,"temperature":1.0,"reasoning_tokens":2924,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T13:36:37.413851+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct test would be to compute the heat capacity of a liquid from molecular dynamics by explicitly separating the energy of shear modes below the Frenkel frequency and checking whether their contribution follows the harmonic-oscillator expression with a linear DOS, or whether it obeys a different statistical mechanics; if the overdamped modes do not contribute to the free energy in the assumed harmonic form, the predicted $c_v$ would deviate systematically at temperatures where those modes dominate the count.","supporting_citations":[{"cited_title":"(2) is significantly smaller than the other terms, rendering it negligible","cited_arxiv_id":null,"evidence_quote":"Supplies the original phonon theory of liquid thermodynamics that this work extends and corrects."},{"cited_title":"Trachenko and V","cited_arxiv_id":null,"evidence_quote":"Provides the argument that shear modes below the Frenkel frequency are overdamped and yields the linear DOS form used in Eq. (17)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Derives the linear low-frequency density of states from instantaneous normal mode theory, supporting the liquid-like DOS."},{"cited_title":"Dawidowski, F","cited_arxiv_id":null,"evidence_quote":"Derives the same linear behavior from the gapped dispersion relation of collective shear waves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the experimentally measured density of states and the Ga heat capacity comparison used in Table I."},{"cited_title":"The Debye frequency is obtained from the Debye tem- perature of each system, Θ D = ℏωD/kBT","cited_arxiv_id":null,"evidence_quote":"Supplies the NIST REFPROP experimental heat capacity data for all liquids compared in the paper."},{"cited_title":"Emergence of Debye scaling in the density of states of liquids under nanoconfinement","cited_arxiv_id":"2307.11429","evidence_quote":"Gives the framework for computing the infinite-frequency shear modulus used to obtain the Maxwell relaxation time."},{"cited_title":"Frenkel, Kinetic theory of liquids, International series of monographs on physics (Clarendon Press Oxford, Ox- ford, 1946)","cited_arxiv_id":null,"evidence_quote":"Supplies the intramolecular vibrational frequencies used in Eq. (19) for molecular liquids."}],"review_version":1}