Revealing the Spin Hydrodynamics of a Spin-Imbalanced Unitary Fermi Gas

Dechao Zhang; J. E. Thomas; Johannes Lang

Revealing the Spin Hydrodynamics of a Spin-Imbalanced Unitary Fermi Gas

T0 review · 3 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A periodic-potential quench in a uniform optical box yields the intrinsic spin diffusivity and spin Seebeck coefficient of a strongly interacting Fermi gas.

desk verdict Using a uniform box and a periodic-potential quench, this group reports the first homogeneous spin Seebeck coefficient and an intrinsic spin diffusivity that overrides the old trapped-cloud value; the result is plausible and well cross-checked internally, but it rests on a single-wavelength hydrodynamic model that the paper never tests. read the letter →

arxiv 2607.29647 v2 pith:2TCVUYUC submitted 2026-07-31 cond-mat.quant-gas

DeChao Zhang , Johannes Lang , J. E. Thomas This is my paper · ORCID

classification cond-mat.quant-gas

keywords unitaryFermigasspintransportSeebeckcoefficientdiffusionhydrodynamicscaloritronicsultracoldatomsopticalbox

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper reports a direct measurement of intrinsic spin-transport coefficients in a uniform strongly interacting Fermi gas. By quenching a periodic optical potential in an optical box and tracking the coupled relaxation of density and polarization ripples, the authors extract a spin diffusivity Ds = 2.20(10) ħ/m and an energy-spin Seebeck coefficient Sε = 2.20(06)P, the coefficient that lets a temperature gradient drive a spin current. The central claim is that these are homogeneous, parameter-free properties of a normal unitary Fermi gas, not averages spoiled by edge effects. This matters because spin caloritronics is hard to probe in strongly interacting quantum fluids, and microscopic theories now have direct benchmarks in this regime.

What carries the argument

The central object is a ~33 µm periodic optical potential superposed on a uniform-density box, creating matched density ripples in both spin states. After the quench, the model tracks Fourier components of density, polarization, and temperature perturbations. Currents are driven by gradients of temperature and spin chemical potential, and the reciprocal relation between off-diagonal coefficients leaves one Seebeck/Peltier parameter. The key identity is the third time derivative of the polarization being proportional to Sε times the second derivative of density, which isolates the Seebeck coefficient from the other fit parameters; a separate result shows the first-sound diffusivity is indepen

What would settle it

Repeat the quench measurement at a different perturbation wavelength (for example λ ≈ 20 µm) and test whether the extracted Ds and Sε stay constant within error; a systematic q-dependence, or a violation of the early-time identity δ'''P(0) = -(2/3)(ħq²/m)S̃εδ''ñ(0), would show the hydrodynamic description is incomplete.

Watch

Extended reading notes

Core claim

After the periodic perturbation is switched off, the gas evolves as a damped first-sound mode plus two diffusive heat-spin modes. The paper shows that the leading short-time curvature of the polarization is set by the energy-spin Seebeck coefficient alone through the identity δ'''P(0) = -(2/3)(ħq²/m)S̃ε δ''ñ(0), so Sε can be measured without knowing the diffusivity. Fitting the full three-variable evolution then gives Ds = 2.20(10) ħ/m and Sε = 2.20(06)P, nearly constant over T/TF ≈ 0.4–0.6 and polarization up to 0.4. These are claimed to be the intrinsic homogeneous transport coefficients; an earlier trapped-cloud value of about 6.3 ħ/m is argued to be inflated because density vanishes near

Load-bearing premise

The whole extraction rests on the assumption that the gas at the probed 33 µm wavelength is described by the linearized three-variable hydrodynamic model with reciprocal constitutive relations and no missing relaxation channels; if that model is incomplete, the fitted coefficients are effective fit parameters rather than intrinsic transport coefficients.

Editorial extensions

If this is right

If the extraction is correct, the intrinsic spin diffusivity of a normal unitary Fermi gas is about 2.2 ħ/m in this temperature window, roughly a third of the value inferred from trapped inhomogeneous clouds.
The energy-spin Seebeck coefficient is linear in polarization with slope 2.2(1) in units of ħ/m, so measurable spin currents can be generated by temperature gradients in spin-imbalanced gases.
The near-independence of Ds and Sε from temperature gives a specific target for microscopic transport calculations; two-body kinetic theory underestimates Sε by an order of magnitude, and medium-corrected calculations still deviate from the data.
No low-temperature growth characteristic of quantum degeneracy appears down to T/TF ≈ 0.4, so the approach of spin diffusion to the superfluid transition must be studied as a separate regime.
The first-sound diffusivity obtained from the same fits provides independent constraints on viscosity and energy conductivity, linking this spin measurement to existing sound-attenuation data.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The uniform-box protocol could be run at shorter wavelengths; if the fitted Ds and Sε change with wavelength, the hydrodynamic model breaks down and the numbers should be read as effective coefficients, giving a concrete way to chart the hydrodynamic limit.
The size of the correction relative to trapped-cloud results suggests that earlier inhomogeneous samples systematically overstate spin diffusivity; reanalyzing or repeating those experiments in nearly uniform traps could reconcile a decade of data.
A temperature gradient applied to a box-confined imbalanced gas should produce a measurable spin accumulation with the sign and magnitude predicted by Sε; measuring that would test the Seebeck coefficient independently of the relaxation fit.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: this is the clearest attempt yet to measure intrinsic spin transport in a uniform strongly interacting Fermi gas, and the numbers are probably right, but the paper overclaims what a single-wavelength model can establish. The new pieces are real: the spin-resolved periodic-potential quench in an optical box, the third-derivative constraint (Eq. 4) that isolates the energy-spin Seebeck coefficient from the initial curvature of the polarization, and the first direct measurement of D̃s in a homogeneous geometry, which puts the old trapped-cloud value (~6.3) to rest. I also give credit for the internal consistency: full-model fits and the short-time polynomial method agree to about 10% (2.20 vs 1.97 for S̃ε), and the measured csn = 0.58P matches the Luttinger-Ward EOS value 0.544P within 7%. The supplement’s derivation of the three-mode linear system is careful, and the Onsager structure is handled properly.

The soft spots are real but not fatal. The biggest is that all extracted values come from a single wavevector (λ≈33 μm). No q-dependence and no Knudsen number are given, so D̃s and S̃ε could be effective parameters of the assumed hydrodynamic closure rather than intrinsic coefficients. For a paper whose headline is 'intrinsic homogeneous transport coefficients,' that missing test is the main thing I’d want before believing the number is universal. Second, the quoted uncertainties are statistical only; there is no systematic budget, and model-form uncertainty is not even estimated. Third, the language about theory comparison is too kind: S̃ε = 2.20P is about 50% above the quantum Boltzmann prediction (1.44P), and D̃s = 2.20 is well above the Luttinger-Ward result (~1.4), yet the text says 'consistent.' That read is a stretch. Finally, no data or code are shipped, which matters for a claim to provide 'parameter-free benchmarks.'

None of this is a load-bearing flaw in the measurement itself. The central result is plausible, and the discrepancy with the trapped-cloud measurement is likely resolved. The paper deserves a serious referee. I’d send it to peer review and ask for a second wavelength or at least a Knudsen estimate, a systematic error budget, and a rewrite of the theory-agreement sentences.

Referee Report

3 major / 5 minor

Summary. The paper reports measurements of spin transport in a spin-imbalanced unitary Fermi gas confined in a uniform optical box. A spatially periodic optical potential is quenched off, and the subsequent relaxation of the Fourier components of the majority/minority density, total density, and polarization is tracked. The data are fitted to a three-variable linear-response hydrodynamic model (density, polarization, temperature; Eqs. S97) with Onsager-symmetric constitutive relations, yielding the energy-spin Seebeck coefficient, the spin diffusivity, the first-sound diffusivity, and related thermodynamic and transport parameters. The central claims are that the measured energy-spin Seebeck coefficient S̃ε = 2.20(06)P and spin diffusivity D̃s = 2.20(10)ℏ/m are the intrinsic homogeneous transport coefficients of the normal-phase unitary Fermi gas, and that the earlier trapped-cloud value D̃s ≈ 6.3 overestimated the spin diffusivity due to density inhomogeneity. The paper also presents a second, polynomial-based extraction of S̃ε (Fig. S2) and comparisons with two-body kinetic theory and quantum Boltzmann predictions.

Significance. If the extracted coefficients are truly intrinsic homogeneous transport coefficients, this is a substantial advance: it resolves a long-standing discrepancy between trapped-cloud spin-diffusion measurements and theory, and it provides the first direct measurement of the spin Seebeck/Peltier coefficients in a strongly interacting Fermi gas. The uniform box geometry, the use of a periodic-potential quench, and the separate extraction of S̃ε by two routes (full model and polynomial short-time analysis) are genuine strengths. The measured thermodynamic quantity csn agrees with the Luttinger-Ward EOS at the 7% level, providing a nontrivial check on the thermodynamic input. The paper is careful to distinguish the fundamental energy-spin Seebeck coefficient from the heat-spin Seebeck coefficient and to show that the first-sound diffusivity is independent of Ds and Sε (Eq. S111). If the hydrodynamic closure is valid at the probed wavelength, the results are important benchmarks for microscopic theories.

major comments (3)

[§II D, Eqs. S97; Fig. 2] The central claim that D̃s and S̃ε are intrinsic homogeneous transport coefficients rests on the validity of the three-variable linear-hydrodynamic model at a single wavevector λ ≈ 33 µm. No q-dependence test is reported, and no Knudsen number or estimate of gradient corrections is given. The 'long wavelength limit' comparison in §II D3 is an internal consistency check of the model against itself, not a test against data at another q. If higher-order gradient terms, viscoelastic memory, or box-edge coupling contribute at this wavelength, the fitted values are effective parameters of the assumed closure rather than intrinsic coefficients. Please provide either a measurement at a second wavevector, a quantitative bound on the Knudsen number, or an explicit estimate of the model-form error.
[§II D2, Eq. S102; Fig. S2] The agreement between the full-model fit (S̃ε = 2.20P) and the polynomial method (S̃ε = 1.97P) is presented as an independent cross-check, but the polynomial method uses Eq. S102, δ'''P(0) = -(2/3)S̃ε γ0 δ''n(0), which is derived from the same model Eqs. S97. Both routes therefore share the same hydrodynamic closure, and the 10% difference only quantifies the internal consistency of the fit, not the validity of the hydrodynamic assumption. The text should state this limitation explicitly and, if possible, validate the closure by an independent observable or a second wavelength.
[§I, Figs. 3–4; EOS input] The quoted uncertainties (e.g., D̃s = 2.20(10), S̃ε = 2.20(06)P) appear to be statistical only. The analysis uses the Luttinger-Ward T-matrix EOS of Ref. [28] for temperature calibration and all thermodynamic coefficients, and the measured csn deviates from the EOS prediction by 7% (Fig. S1). There is no systematic error budget that propagates the EOS uncertainty, the temperature-calibration uncertainty, or the uncertainty in the initial-condition determination into the transport coefficients. Please add a systematic-error analysis, or at least state the dominant systematic contributions and how they affect the quoted central values.

minor comments (5)

[Title/header] The title in the manuscript header reads 'Unit ary Fermi Gas'; should be 'Unitary Fermi Gas'.
[Discussion] 'spin-calorimetric properties' should likely be 'spin-caloritronic properties' for consistency with the rest of the text.
[Eq. (1)] The matrix of transport coefficients is written as (κǫ Pǫ; Sǫ 2σs) but without matrix brackets; this is ambiguous. Please display it as a proper 2×2 matrix or define the ordering explicitly.
[Abstract / §II C] The quantum Boltzmann prediction is described as 'parameter-free', but it uses the Luttinger-Ward EOS for thermodynamic consistency and the large-N expansion for the scattering kernel. The label is understandable but may overstate the absence of model dependence; a brief clarification would help.
[Fig. S3/S4] The density shift of 0.82 is introduced ad hoc for the first-sound diffusivity comparison; the text should state whether this shift is independently measured or fitted, and how its uncertainty is propagated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: transport coefficients are extracted from relaxation data by a hydrodynamic fit and benchmarked against independent microscopic calculations.

full rationale

The paper's central quantities, D̃s and S̃ε, are obtained as fit parameters in Eq. S97 from the measured time evolution of δn(q,t) and δP(q,t). They are not set equal to any theoretical prediction before the fit, and the comparison theories (quantum Boltzmann and Luttinger–Ward transport predictions) are computed independently from the measured relaxation. The short-time relation Eq. S102, δ'''P(0)=-(2/3)γ0 S̃ε δ¨n(0), is a derived consequence of the hydrodynamic model, not an input assumption equal to the measured data; using it for the polynomial estimate is a second estimator of the same model parameter, not a prediction-equals-fit reduction. The equation of state [28] is used for the temperature calibration and thermodynamic coefficients, but it does not itself contain the measured transport coefficients and is therefore not the target of the derivation; the co-authorship of J. Lang on [28] does not make the benchmark circular because the Luttinger–Ward EOS is independent of the values of Ds and Sε extracted from the relaxation data. The absence of a q-dependence test or Knudsen-number estimate is a model-validation concern about whether the fitted parameters are the intrinsic transport coefficients, not a circularity in which the output is equivalent to the input by construction.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

This is a measurement paper: the headline free parameters (D̃s, S̃ε) are the experimentally measured quantities, not ad hoc adjustments, so their presence is not a defect. The real burden is structural: everything rests on the validity of the three-variable hydrodynamic model and on the shared Luttinger-Ward EOS used both for temperature calibration and in the comparison theories. No new physical entities are postulated.

free parameters (7)

D̃s (dimensionless spin diffusivity) = 2.20(10)
Central measured result; joint χ² fit of the hydrodynamic model (Eqs. S97) to δP(q,t) and δñ(q,t). Reported nearly independent of T/TF and P.
S̃ε (dimensionless energy-spin Seebeck coefficient) = 2.20(06) x P (slope)
Central measured result; full-model fit gives slope 2.20P; independent polynomial method gives 1.97P (Fig. S2).
κ̃ε (dimensionless energy conductivity) = not reported in main text
Model fit parameter; enters the first-sound diffusivity through Eq. S111.
η̃ (dimensionless shear viscosity) = not reported in main text
Model fit parameter in the viscous damping term of Eq. S97.
ωT (isothermal sound frequency) = adjusted to match observed first-sound frequency
Determines reduced temperature T/TF via the theoretical EOS [28]; all universal-function inputs to the model depend on this calibration.
Polynomial-fit coefficients (A0, A2, A3, C0, C3, C4) = per-shot values
Auxiliary fits in the independent S̃ε extraction (Supplement II D 2).
Density shift 0.82 (D̃1 comparison) = 0.82
Imported from the authors' prior work [43], not fitted here; used to overlay the first-sound diffusivity prediction.

assumptions (5)

domain assumption The normal-phase unitary Fermi gas at T/TF ≈ 0.4–0.6 is described by the linearized three-variable hydrodynamic equations (S97) with the Onsager-symmetric constitutive relations of Eq. 1.
All extracted transport coefficients come from fitting this model; no q-dependence check or collision-rate estimate is provided. Main text Eq. 2; Supplement § II D.
domain assumption The Luttinger-Ward T-matrix EOS (Ref. [28], co-authored by J. Lang) supplies accurate universal functions fp, fh, fs1 for spin-imbalanced unitary gases.
Used to convert the fitted ωT into T/TF and to evaluate thermodynamic coefficients in Eqs. S97; comparison theories share this framework. Supplement § II A.
domain assumption Bulk viscosity ξB = 0 for a unitary Fermi gas.
Invoked to drop the bulk-viscous term in Eqs. S81–S85; based on scale invariance [26, 27] and a previous measurement [25].
domain assumption Box-confinement and perturbation-potential terms can be dropped from the evolution equations in the measurement region.
Measurement is of the central region before edge reflections, but the flatness of the box there is assumed. Supplement, text around Eq. S85.
domain assumption Initial state satisfies δh = 0, δT = 0, δṅ = 0, with δP(0) = -csn δñ(0), where csn is measured.
Sets the initial conditions for the model; csn is measured in situ (Fig. S1) and agrees with the EOS within 7%, so this is partly empirical.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Revealing the Spin Hydrodynamics of a Spin-Imbalanced Unitary Fermi Gas." pith.science (2026). https://pith.science/paper/2TCVUYUC

@misc{pith2026260729647,
  author       = {Pith},
  title        = {Pith review of: Revealing the Spin Hydrodynamics of a Spin-Imbalanced Unitary Fermi Gas},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2TCVUYUC}},
  note         = {Machine review of arXiv:2607.29647}
}

read the original abstract

Hydrodynamics governs diverse collective phenomena in nature, from the expansion of a quarkgluon plasma to viscous electron flow in quantum materials. In strongly interacting hydrodynamic fluids, the transport of spin remains poorly understood. Here, we investigate spin transport in the hydrodynamic regime of a spin-imbalanced unitary Fermi gas confined in a uniform optical box. By quenching a spatially periodic optical potential that modulates both the density and spin polarization, we observe the relaxation of the many-body system, which determines both the spin diffusivity and the spin Seebeck/Peltier coefficients in a homogeneous quantum gas. Our measurements provide parameter-free benchmarks for microscopic theories and establish an ultracold atom platform for studying spin caloritronics in strongly correlated matter.

Figures

Figures reproduced from arXiv: 2607.29647 by the authors.

**Figure 2.** FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** B shows the energy-spin Seebeck coefficient as a function of P. Since the temperature dependence is negligible over the range shown in Fig. 3A, we combine data for 0.38 < T /TF < 0.58, with an average hT /TF i = 0.47. S˜ ǫ appears to vary linearly with P, in qualitative agreement with predictions for a unitary Fermi gas in the twobody kinetic theory limit [13, 22]. The positive sign for a polarization P > 0, corr… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.