{"id":"fa5740c9-149e-41f1-bb41-4763ec72851e","arxiv_id":"1909.02555","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":2,"one_line_summary":"In a homogeneous unitary Fermi gas, the sound diffusivity falls with temperature and saturates at about hbar/m, independent of temperature below the superfluid transition.","lead":"Physicists sent sound waves through a cloud of strongly interacting lithium atoms trapped in a uniform box and measured how quickly the sound faded. They found the damping reaches a universal quantum limit set by Planck's constant and the atomic mass, a result that connects to transport in neutron stars and the early universe.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The superfluid plateau D≈ℏ/m is extracted below Tc with a single-Lorentzian fit plus the Pethick–ter Haar f(ck/Γph) model, while the paper's own SI notes a second-sound peak in the two-fluid response; if that second peak biases the first-sound linewidth, the headline T-independent universal value…","rationale":"I read the central claim as two-part: a normal-state sound diffusivity that decreases monotonically with temperature, and a superfluid plateau near ℏ/m that is independent of temperature and condensate fraction. The normal-state part is well supported by direct quadratic Γ(k) scaling and by the parameter-free first-sound speed and sum-rule checks. The superfluid plateau, however, relies on the Pethick–ter Haar interpolation with a fitted Γ_ph, and the paper's own Supplementary Information acknowledges the two-peak structure of the full response without quantitatively demonstrating that the second-sound or beyond-hydrodynamic contribution does not bias the single-Lorentzian linewidth. The first-sound sum-rule check constrains the integrated weight, not the width, so it is not sufficient to rule out this bias. This makes the superfluid plateau the most load-bearing assumption in the paper. I do not regard the concern as a rejection: the f model is physically motivated, the crossover is visible, and the normal-state result is independent of it. But because the abstract's headline statement is precisely the superfluid plateau, I would make acceptance conditional on a numerical forward-model test that recovers D from synthetic two-fluid spectra using the same fitting procedure. This is a concrete, low-cost check that would restore the original ACCEPT verdict if it passes.","tokens_in":12460,"tokens_out":13692,"duration_ms":167810,"concrete_test":"Forward-model the density response with the two-fluid hydrodynamic theory of Hu–Zou–Liu (ref [64]), using the measured unitary Fermi gas equation of state (ref [18]) and the experimentally known superfluid density, at the same temperatures and box-mode wavenumbers as the data. Inject a known sound diffusivity D into the model, generate synthetic Im[χ(ω,k)] spectra, and run the paper's exact fitting pipeline (single Lorentzians plus f(ck/Γ_ph)) on them. If the recovered D differs from the injected value by more than the reported statistical errors, or if it shows a spurious T/n_C dependence, the superfluid plateau is an artifact of the fitting model. If it recovers D within errors, the concern is resolved.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Below Tc, the D values carrying the 'universal plateau' are not obtained by a parameter-free quadratic fit: Fig. 3B and the corresponding text use Γ = D k^2 f(ck/Γ_ph) with f(x) = tan^{-1}(x)/x and a fitted phonon damping rate Γ_ph = 0.27(8) k_BT/ℏ. This is a single-relaxation-time model. The Supplementary Information explicitly states that the full two-fluid response Im[χ(ω,k)]/ω contains a second, finite-frequency peak below Tc (second sound), yet the linewidth extraction fits each mode with one Lorentzian. The first-sound sum-rule check W1(k) in Fig. 6B verifies the integrated weight of the first-sound peak, not its width; a second-sound tail that is small in the integral could still broaden or shift the fitted Lorentzian. Because the reported flatness of D from n_C = 0 to n_C ≈ 0.8 is the central superfluid claim, such a bias is load-bearing: it could either create or mask a real temperature or condensate-fraction dependence of the sound diffusivity. The normal-state monotonic decrease is less exposed, since there Γ ∝ k^2 is direct and no f model is needed; the concern is specifically about the superfluid branch.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports measurements of sound propagation and attenuation in a homogeneous unitary Fermi gas confined in an optical box trap. By modulating an endcap wall and imaging the resulting density waves, the authors measure the dispersion relation ω(k), the speed of sound c, and the damping rate Γ of the density response. They find a linear dispersion with a scale-invariant speed of sound satisfying mc² = (10/9)E/N with no free parameters. From the damping rate they extract the sound diffusivity D = Γ/k². In the normal state, D decreases monotonically with decreasing temperature, approaching a value near ℏ/m, in contrast to the diverging 1/T² behavior of weakly interacting Fermi liquids. Upon crossing the superfluid transition, D is reported to be approximately temperature-independent and close to ℏ/m, independent of condensate fraction. The paper also derives constraints on the thermal conductivity and Prandtl number, finding Pr < 1 at all temperatures, and discusses implications for strongly interacting fermionic matter in other contexts.","tokens_in":12777,"tokens_out":5592,"duration_ms":59073,"significance":"If the results hold, they provide the first direct measurement of sound diffusivity in a homogeneous strongly interacting Fermi gas, establishing a quantum-limited value D ∼ ℏ/m that is universal across the superfluid transition. The clean homogeneous geometry, the parameter-free check of scale invariance through mc² = (10/9)E/N, and the high-temperature prediction D = 6.46(ℏ/m)(T/T_F)^{3/2} are notable strengths. The findings directly constrain shear viscosity and thermal conductivity and sharpen the comparison with other quantum fluids such as 4He, 3He, and strongly correlated systems in nuclear and particle physics. The experimental method is careful, and the paper is clearly written. However, the central superfluid claim rests on a linewidth extraction whose robustness against two-fluid (second-sound) contamination is not quantitatively demonstrated.","major_comments":[{"comment":"The central superfluid claim—D ≈ ℏ/m independent of temperature and condensate fraction—rests on the extraction of the first-sound linewidth Γ from single-Lorentzian fits below Tc. The SI itself states that below Tc the full two-fluid response Im[χ(ω,k)]/ω contains a second, finite-frequency peak (second sound) in addition to the first-sound peak. The first-sound sum-rule check W1(k) in Fig. 6B conserves the integrated weight of the first-sound peak; it does not certify that the fitted Lorentzian width is unbiased by a second-sound tail that is small in the integral but non-negligible near the resonance. Because the plateau from n_C = 0 to n_C ≈ 0.8 is the headline result, I ask the authors to demonstrate explicitly that the single-Lorentzian width is not biased: either fit the full two-fluid response of Refs. [47,48,64] or quantify the second-sound contribution to the linewidth at the wave numbers used. Without this, the temperature independence below Tc could be an artifact of the fitting model.","section":"Fig. 3B, Fig. 4, Supplementary Information 'The response function χ and its normalization'"},{"comment":"The below-Tc analysis uses Γ = D k² f(ck/Γ_ph) with f(x) = tan⁻¹(x)/x and a single fitted phonon damping rate Γ_ph = 0.27(8) k_BT/ℏ. The fitted Γ_ph absorbs the crossover from quadratic to linear scaling, and the extracted D may therefore inherit the model's assumption that one relaxation channel governs sound attenuation. The authors should state explicitly whether the D values in Fig. 4 are obtained from a direct low-k quadratic fit to Γ(k) or from the f-model, and should justify the applicability of this single-relaxation-time expression to the unitary Fermi gas by comparing it with the two-fluid hydrodynamic form. This distinction matters for the systematic uncertainty on D below Tc.","section":"Fig. 3B and the model Γ = D k² f(ck/Γ_ph)"}],"minor_comments":[{"comment":"The sentence 'Fermi's Golden Rule yields a rate Γ_ph ∝ k' appears to conflict with the fitted constant Γ_ph = 0.27(8) k_BT/ℏ; please clarify whether Γ_ph is momentum-dependent and how the argument of f(x) is defined in that case.","section":"Main text, paragraph after Fig. 3B"},{"comment":"The relation c_P = 5Pα/(2ρ) is used to reduce the sound diffusivity to D = 4η/(3ρ) + 4κT/(15P), but the derivation is not shown; a one-line derivation would help the reader verify the formula.","section":"Supplementary Information, 'Thermal conductivity and Prandtl number'"},{"comment":"The caption lists red circles, green squares, and blue triangles for T/T_F = 0.36(5), 0.21(3), and 0.13(2), respectively, but the text refers to 'red and green' above Tc and 'blue' below; consider adding a legend to the figure itself to avoid ambiguity.","section":"Fig. 3B caption"},{"comment":"Reference [34] appears to have an incomplete author list; please verify against the published version.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"This is a strong experimental paper with a clean geometry and parameter-free checks. The main concern is whether the below-Tc plateau in D is robust to second-sound contamination; this is addressable in revision by fitting the two-fluid response or by a quantitative estimate of the bias. If the authors provide such an analysis, I would be happy to accept."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Patel et al. report the first direct measurement of sound diffusivity D in a homogeneous unitary Fermi gas. The normal-state result—D decreasing monotonically as T drops, then tracking the T^{3/2} Boltzmann limit at high T—looks solid and is a nice contrast to the 1/T^2 Fermi-liquid divergence. The experimental work is careful: box trap, linear dispersion, parameter-free check of mc^2 = (10/9)E/N, Lorentzian fits to Im[chi]/omega, and a first-sound sum rule used to calibrate the drive. I believe the central qualitative claim: sound diffusivity in this system saturates around hbar/m and does not diverge at low T.\n\nThe soft spot is the superfluid branch. Below Tc, the D values that form the 'universal plateau' are not obtained from a clean quadratic Gamma vs k^2 fit. The data show a crossover to linear scaling, and the analysis uses Gamma = D k^2 f(ck/Gamma_ph) with f(x)=atan(x)/x and a fitted phonon damping rate Gamma_ph = 0.27(8) k_B T/hbar. That is a single-relaxation-time model. The SI explicitly states that the full two-fluid response contains a second-sound peak at finite frequency below Tc, yet each mode is fit with one Lorentzian. The first-sound sum rule checks the integrated weight of the peak, not its width. So a small second-sound tail could in principle bias the fitted linewidth, and that bias would matter precisely for the headline claim that D is temperature-independent from nC=0 to nC≈0.8. This is not a fatal objection—the data and the model are consistent, and the authors disclose the ~30% second-sound correction to the viscosity in the SI—but it is load-bearing enough that a referee should push for a quantitative two-fluid analysis or an estimate of the second-sound contribution to the linewidth before the plateau is taken as established.\n\nThe citation pattern is fine; the paper builds on the authors' own box-trap and equation-of-state work and on the relevant viscosity/thermal-conductivity literature. No invented entities or circular derivations.\n\nWho should read this: anyone working on transport in strongly interacting fermions, quantum hydrodynamics, or the unitary gas. It deserves a serious referee—not a desk reject. My recommendation: send to peer review, and in the review request ask specifically for an assessment of the second-sound contamination in the superfluid D extraction.","headline":"A careful, first-of-its-kind measurement of sound diffusivity in a homogeneous unitary Fermi gas, with a robust normal-state result and a believable quantum-limited plateau, though the superfluid extraction leans on a single-Lorentzian fit that deserves a hard look.","tokens_in":13298,"tokens_out":2279,"would_cite":true,"duration_ms":24610,"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 a strongly interacting Fermi gas, sound diffusivity stops at the quantum limit $\\hbar/m$ and remains flat across the superfluid transition.","keywords":["unitary Fermi gas","sound diffusivity","quantum limited transport","density response function","hydrodynamic damping","superfluid transition","scale invariance","optical box trap"],"falsifier":"A direct measurement of the shear viscosity $\\eta$ and thermal conductivity $\\kappa$ in the same homogeneous gas would settle the central claim: if $4\\eta/(3mn) + 4\\kappa T/(15P)$ differs from the reported $D$ by more than the combined uncertainties, the inferred sound diffusivity is not the hydrodynamic one. Alternatively, repeating the linewidth measurement at wavenumbers below $k \\simeq 0.05\\,mc/\\hbar$ would show whether $\\Gamma/k^2$ remains $k$-independent; a residual $k$-dependence would indicate contamination by second sound or collisionless damping.","tokens_in":12294,"feed_emoji":"🔊","tokens_out":4698,"duration_ms":48577,"temperature":0.7,"pith_summary":"This paper reports the first direct measurement of sound diffusivity in a homogeneous, strongly interacting Fermi gas of atoms tuned to a Feshbach resonance (the unitary Fermi gas), using sound waves generated and detected in an optical box trap. It aims to establish that, as temperature is lowered in the normal state, the sound diffusivity $D$ decreases monotonically instead of diverging as $1/T^2$ like in weakly interacting Fermi liquids. At the superfluid transition, the paper claims $D$ reaches a universal value close to $\\hbar/m$, the Heisenberg limit set by Planck's constant and particle mass, and stays essentially flat as the condensate fraction grows. A sympathetic reader would care because this gives a clean experimental benchmark for transport theories of strongly interacting fermions that also describe neutron matter, hydrodynamic electron flow, and the quark-gluon plasma.","feed_headline":"Sound diffusion in a Fermi gas hits the quantum floor","feed_subtitle":"As the gas cools through its superfluid transition, sound damping sticks near ħ/m instead of diverging.","key_machinery":"The central object is the density response function $\\chi(\\omega,k)$, whose poles locate sound modes and whose imaginary part, measured via the out-of-phase density modulation, gives the sound attenuation rate $\\Gamma$. The hydrodynamic relation $\\Gamma = D k^2 f(c k/\\Gamma_{\\rm ph})$ with $f(x)=\\tan^{-1}(x)/x$ connects the measured linewidth to the sound diffusivity $D$, with a phonon damping rate $\\Gamma_{\\rm ph}$ accounting for the crossover to collisionless damping at higher wavenumbers. Below the superfluid transition, two-fluid hydrodynamics is assumed to separate first-sound from second-sound peaks, and the first-sound sum rule $W_1 = n\\pi/(2mc^2)$ calibrates the drive amplitude.","core_discovery":"The central discovery is that the sound diffusivity of the unitary Fermi gas is quantum limited. Measuring the density response function $\\operatorname{Im}\\chi(\\omega,k)$ through steady-state response to a modulated optical wall, the authors resolve first-sound resonances at the normal-mode wavenumbers of the box, whose Lorentzian linewidth gives the damping rate $\\Gamma$. At low wavenumbers $\\Gamma = D k^2$, establishing diffusive sound attenuation, and the extracted $D$ falls smoothly from the high-temperature $T^{3/2}$ behaviour to a plateau near $\\hbar/m$ below the superfluid transition. The same data set confirms scale invariance through the relation $m c^2 = \\frac{10}{9} E/N$, and combined with a computed shear viscosity implies a Prandtl number strictly below unity at all temperatures, which excludes a relativistic conformal gravity dual for this system.","pith_inferences":["If the plateau at $\\hbar/m$ persists to arbitrarily low temperature, then the kinematic viscosity of the unitary Fermi gas cannot diverge as phonon damping would suggest, which would rule out one class of low-temperature transport scenarios.","The same box-trap response method could measure the spin diffusivity or a direct thermal conductivity in the same homogeneous geometry, providing an independent check of the decomposition $D = D_\\eta + D_\\kappa$.","A natural testable extension is to push to lower $T/T_F$: if a small Fermi-liquid-like rise in $D$ appears at the lowest temperatures, the universal plateau would be a crossover feature rather than an exact fixed point.","For neutron matter, where the same unitary limit approximately applies, a sound diffusivity near $\\hbar/m$ sets a floor on dissipative heating in neutron star mergers, with observable consequences for gravitational wave damping."],"forward_implications":["In the normal state, $D$ decreasing as $T$ falls rules out the $1/T^2$ Fermi-liquid divergence for a unitary Fermi gas in the explored range, sharpening the contrast with liquid $^3$He.","Below $T_c$, $D \\simeq \\hbar/m$ independent of temperature and condensate fraction means sound attenuation in strongly interacting fermionic superfluids is fixed by $h$ and $m$ alone, matching the behaviour of liquid $^4$He.","A Prandtl number below one for all temperatures implies that momentum diffusion dominates heat diffusion, a constraint on any kinetic-theory or holographic description of the unitary Fermi gas.","Because scale invariance makes $D$ a universal function of $T/T_F$, the measured values transfer directly to other unitary Fermi systems regardless of density or species.","The measured $D$ constrains the sum $4\\eta/(3mn) + 4\\kappa T/(15P)$, so any proposed shear viscosity and thermal conductivity must satisfy this combination."],"supporting_citations":[{"why":"Supplies the density response formalism and the damped-oscillator pole for sound that the measurement is built on.","marker":"[47]"},{"why":"Gives the interpolating damping formula $\\Gamma = D k^2 f(ck/\\Gamma_{\\rm ph})$ used to fit the data.","marker":"[53]"},{"why":"Provides the theoretical shear viscosity used to decompose the measured diffusivity and to compute the Prandtl number.","marker":"[21]"},{"why":"Provides the high-temperature kinetic-theory prediction for sound attenuation and thermal conductivity that anchors the $T^{3/2}$ curve.","marker":"[23]"},{"why":"Provides the equation of state, transition temperature $T_c = 0.17T_F$, and temperature calibration used throughout.","marker":"[18]"},{"why":"Establishes the prior quantum-limited spin diffusivity in the unitary Fermi gas that motivates the universal $\\hbar/m$ scale.","marker":"[5]"},{"why":"Describes the phonon damping crossover in a pair-condensed Fermi gas that explains the linear $\\Gamma(k)$ branch.","marker":"[32]"},{"why":"Gives the hydrodynamic definitions of sound diffusivity, viscosity, and thermal conductivity used to interpret the result.","marker":"[49]"}],"fun_headline_variants":["Universal sound diffusion limit confirmed in Fermi gas","Fermi gas sound damping reaches quantum limit","Sound diffusion in strongly interacting Fermi gas quantum-limited","ħ/m: Universal sound diffusion floor in Fermi gas"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis below the superfluid transition assumes that the measured first-sound linewidth is governed by two-fluid hydrodynamics with a clean separation from second sound, and that the linewidth at low wavenumbers obeys $\\Gamma = D k^2 f(ck/\\Gamma_{\\rm ph})$ with a fitted phonon damping rate; if second-sound contamination or beyond-hydrodynamic relaxation biases those linewidths, the inferred $D$ would not be the true sound diffusivity.","fun_headline_variants_meta":{"raw":{"variants":["Universal sound diffusion limit confirmed in Fermi gas","Fermi gas sound damping reaches quantum limit","Sound diffusion in strongly interacting Fermi gas quantum-limited","ħ/m: Universal sound diffusion floor in Fermi gas"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000203,"raw_usage":{"total_tokens":1346,"prompt_tokens":869,"completion_tokens":477,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":485,"completion_tokens_details":{"reasoning_tokens":418}},"tokens_in":485,"tokens_out":477,"duration_ms":5287,"temperature":1.0,"reasoning_tokens":418,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:46:57.048857+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of the shear viscosity $\\eta$ and thermal conductivity $\\kappa$ in the same homogeneous gas would settle the central claim: if $4\\eta/(3mn) + 4\\kappa T/(15P)$ differs from the reported $D$ by more than the combined uncertainties, the inferred sound diffusivity is not the hydrodynamic one. Alternatively, repeating the linewidth measurement at wavenumbers below $k \\simeq 0.05\\,mc/\\hbar$ would show whether $\\Gamma/k^2$ remains $k$-independent; a residual $k$-dependence would indicate contamination by second sound or collisionless damping.","supporting_citations":[{"cited_title":"Microscopic theory of su- perfluid helium,","cited_arxiv_id":null,"evidence_quote":"Supplies the density response formalism and the damped-oscillator pole for sound that the measurement is built on."},{"cited_title":"On the attenuation of sound in liquid helium,","cited_arxiv_id":null,"evidence_quote":"Gives the interpolating damping formula $\\Gamma = D k^2 f(ck/\\Gamma_{\\rm ph})$ used to fit the data."},{"cited_title":"Vis- cosity and scale invariance in the unitary Fermi gas,","cited_arxiv_id":null,"evidence_quote":"Provides the theoretical shear viscosity used to decompose the measured diffusivity and to compute the Prandtl number."},{"cited_title":"Thermal con- ductivity and sound attenuation in dilute atomic Fermi gases,","cited_arxiv_id":null,"evidence_quote":"Provides the high-temperature kinetic-theory prediction for sound attenuation and thermal conductivity that anchors the $T^{3/2}$ curve."},{"cited_title":"Revealing the superfluid lambda transition in the univer- sal thermodynamics of a unitary Fermi gas,","cited_arxiv_id":null,"evidence_quote":"Provides the equation of state, transition temperature $T_c = 0.17T_F$, and temperature calibration used throughout."},{"cited_title":"Universal spin transport in a strongly interacting Fermi gas,","cited_arxiv_id":null,"evidence_quote":"Establishes the prior quantum-limited spin diffusivity in the unitary Fermi gas that motivates the universal $\\hbar/m$ scale."},{"cited_title":"Three-phonon and four-phonon interaction processes in a pair-condensed Fermi gas,","cited_arxiv_id":null,"evidence_quote":"Describes the phonon damping crossover in a pair-condensed Fermi gas that explains the linear $\\Gamma(k)$ branch."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the hydrodynamic definitions of sound diffusivity, viscosity, and thermal conductivity used to interpret the result."}],"review_version":1}