REVIEW 3 major objections 5 minor 51 references
An ideal Josephson junction in an ultracold two-dimensional Fermi gas
T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read An ultracold two-dimensional Fermi gas with a weak link behaves like an ideal Josephson junction, demonstrating phase coherence and superfluidity in a strongly interacting 2D system.
desk verdict A clean experiment that makes the first 2D Fermi-gas Josephson junction with a convincing sinusoidal current-phase relation, but the claim of superfluidity across the entire BEC-BCS crossover is undercut by missing BCS-side thermometry. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The argument is carried by a lumped-element circuit model in which the junction is a nonlinear Josephson inductance $L_J(\varphi)=\hbar/(dI/d\varphi)$ in series with a bulk inductance $L_B$ and a capacitance $C$. For small phase excitations the oscillation frequency is $\omega=1/\sqrt{(L_B+L_{J,0})C}$, and the paper validates the model by showing that $L_{J,0}$ depends only on barrier height, not on system size. The nonlinearity is probed by measuring the downshift of the fundamental frequency at larger imprinted phases; inserting the ideal current-phase relation into the circuit equations gives the rescaled expression $I_0(\varphi_0)\approx 2I_C \sin(\varphi_0/2)$, which is the curve the data are compared against.
What would settle it
Measure the gas temperature directly, for example with radio-frequency spectroscopy that does not rely on the box being thermalized with its surroundings; if the true temperature were found to exceed roughly $T_c/T_F \approx 0.1$, the observed oscillations would no longer be evidence of superfluid phase coherence. Alternatively, test the predicted size dependence $I_C \propto L^{-\eta}$: if the critical current does not decrease with increasing box size, the phase-coherence explanation would be ruled out.
Extended reading notes
Core claim
The central claim is that a homogeneous 2D Fermi gas divided by a narrow tunnelling barrier forms an ideal Josephson junction: the measured dependence of the oscillation frequency on the imprinted phase matches the rescaled sinusoidal relation $I_0(\varphi_0)=2I_C \sin(\varphi_0/2)$, and the junction inductance extracted from frequency measurements collapses onto a single curve for several system sizes. From this, the paper concludes that the current across the junction is a supercurrent driven by the phase difference between two superfluids, directly demonstrating phase coherence and superfluidity in a strongly interacting 2D Fermi gas. In the molecular limit, the critical current is derived from the condensate density and a mean-field tunnelling amplitude, yielding a condensate fraction $n_c/n = 0.72$ that is consistent with a low-temperature 2D superfluid.
Load-bearing premise
The interpretation of the oscillations as Josephson oscillations assumes that the gas is a 2D superfluid at a temperature of about $T/T_F = 0.03$, well below the predicted critical temperature of about $T_c/T_F = 0.1$, but that temperature is estimated indirectly and thermalization between atoms inside and outside the box is not fully established.
Editorial extensions
If this is right
- If the central claim is right, a homogeneous 2D Fermi gas in the strongly interacting regime is phase-coherent, giving a clean model system for studying reduced-dimensionality effects on superfluidity.
- The demonstrated junction becomes a quantitative probe: the critical-current–condensate-density relation allows the condensate fraction to be extracted, and by varying the box size the algebraic scaling exponent of phase coherence can be measured.
- The near-constant critical current across the BEC-BCS crossover maps the interaction dependence of superfluid transport and provides a benchmark for future theories of 2D Josephson junctions in the crossover.
- The low damping observed at $T/T_F\approx 0.03$ points to phonons and vortex-antivortex nucleation as the main dissipation channels, which can be investigated in further experiments.
- The platform can be extended to periodically driven junctions and to spin-imbalanced gases, potentially enabling studies of driven coherent transport and exotic phases such as the Fulde-Ferrell-Larkin-Ovchinnikov state.
Reading between the lines
- The temperature estimate $T/T_F\approx 0.03$ comes from time-of-flight and equation-of-state comparisons, and the authors note that full thermalization between the box and its surroundings is not certain; a direct measurement that placed the temperature closer to the predicted $T_c/T_F\approx 0.1$ would weaken the superfluid interpretation.
- If the ideal current-phase relation holds across the entire crossover, the junction could serve as a phase-sensitive sensor for superfluid density, since the critical current is predicted to scale with the condensate fraction and the algebraic correlation exponent.
- A direct test of the phase-coherence mechanism would be to measure the critical current as a function of box size; the predicted algebraic decrease $I_C \propto L^{-\eta}$ would provide a measurement of the superfluid density and a direct check of the finite-size phase-coherence scenario.
- The damping channels identified in the simulations—phonons plus vortex-antivortex pairs in the low-density barrier region—could be compared with experiment by imaging vortex cores, turning a suspected limitation into a diagnostic of superfluid behavior.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports the realization of a Josephson junction in a homogeneous, strongly interacting two-dimensional Fermi gas of 6Li. A repulsive optical barrier splits a box-trapped 2D gas into two reservoirs; after imprinting a relative phase, the authors observe Josephson oscillations of the population imbalance and relative phase. They measure the oscillation frequency as a function of the imprinted phase and compare the extracted effective current with the sinusoidal current-phase relation of an ideal Josephson junction, using the critical current as a single fitted parameter. They also measure the critical current as a function of interaction strength across part of the BEC-BCS crossover and compare it with a bosonic c-field theory and an analytic expression for the tunnelling current. The central claim is that the observed oscillations demonstrate phase coherence and provide strong evidence for superfluidity in a strongly interacting 2D Fermi gas.
Significance. If fully supported, this would be a landmark experiment: it would demonstrate phase coherence across a weak link in a strongly interacting 2D Fermi gas and establish the sinusoidal current-phase relation in an ultracold atomic Josephson junction. The experiment exploits a uniform box potential, direct phase measurement via matter-wave interferometry, and a clean circuit-model analysis; the collapse of the Josephson inductance for different system sizes in Fig. 1F is a strong internal check. The authors also deposit data and simulation scripts, which is commendable. However, the breadth of the superfluid claim across the entire crossover rests on temperature information that is only established in the molecular regime, and one of the main theory comparisons contains a circular step. These issues do not invalidate the core observations but do require substantial qualification or additional measurements before the broad conclusions can be accepted.
major comments (3)
- [Fig. 3; Supplementary 'Equation of state' and 'Numerical simulations'] The claim that the observed oscillations demonstrate superfluidity across the entire BEC-BCS crossover presupposes that each reservoir is below its local BKT transition temperature. The paper establishes T/TF≈0.03 only in the molecular regime: the time-of-flight estimate is accompanied in the supplement by the caveat that thermalization between atoms inside and outside the box is 'a priori unclear', and the equation-of-state comparison in Fig. S4 is performed at ln(kF a2D)=−2.9. No independent temperature measurement is presented for the BCS-side points ln(kF a2D)=0.7 and 1.9, where the transition temperature is expected to be much smaller than the 0.1 TF quoted for the molecular side and, for weak binding, exponentially suppressed. The damping-based estimate in Fig. S3 comes from a bosonic c-field simulation and therefore does not by itself constrain the fermionic temperature in the weakly bound regime. A concrete test would be to measure the equation of state or compressibility at the BCS-side interaction strengths and compare with finite-temperature theory, or to restrict the superfluid claim to the interaction range in which T/Tc is established.
- [Fig. 3E; text following 'We use this theory to determine the condensate fraction'] The comparison between data and theory in Fig. 3E is not an independent test for the points used to fix the model. The condensate fraction nc/n=0.72(8) is obtained by inverting the measured critical current for ln(kF a2D)≤−0.9, and the blue theory curve is then calculated using this same nc/n. The line therefore matches the left-hand portion of the data by construction; the informative comparison is only the extrapolation to the BCS side, where the bosonic theory is acknowledged to be of uncertain quantitative accuracy. Please state explicitly which data points determine nc/n, report the residuals for the independent points separately, and avoid presenting the full-curve agreement as a parameter-free validation.
- [Fig. 2F; 'Current phase relation'] The sine-law comparison in Fig. 2F has one fitted parameter: the critical current IC is determined from the first three phase points, and since I0(φ0)≈IC φ0 for small φ0, those points fix the initial slope. The remaining data points do test the predicted sinusoidal reduction of the oscillation frequency, so the measurement is meaningful, but the phrase 'excellent agreement' overstates the strength of the test. The main text should state that the comparison has one fit parameter, specify the number of independent data points beyond the fitted ones, and report the scatter of the residuals.
minor comments (5)
- [Introduction] The phrase 'thed-wave symmetry' contains a missing space and should read 'the d-wave symmetry'.
- [Fig. 1F and Supplementary 'Numerical simulations'] The statement that the data show 'very good agreement with a full numerical simulation' should note that the barrier-height calibration is itself obtained by matching the data to that simulation. Only the collapse of the Josephson inductance across system sizes is an unconstrained check of the circuit model.
- [Supplementary 'Current phase relation'] The derivation of the rescaled expression I0(φ0)≈2IC sin(φ0/2) assumes the bulk inductance is negligible (LJ≫LB), whereas the experiment is performed with LJ/LB≈1.3. The numerical verification that the discrepancy is below 2% for φ0≲0.7π is reassuring, but this condition should be stated in the main text near Fig. 2F so readers can assess the validity of the comparison.
- [Supplementary Fig. S3] The temperature labels in the bosonic c-field simulations should be identified as bosonic T/TF values; as written, a reader could mistake them for direct fermionic thermometry on the BCS side.
- [Main text and Supplementary] The magnetic field is given as 731 G in the main text (footnote 29 for ln(kF a2D)=−2.4) but as 730 G in the supplementary equation-of-state section; please reconcile the convention.
Circularity Check
No significant circularity; the ideal current-phase relation and the crossover comparison retain independent content despite one fitted amplitude and self-citations that are not load-bearing.
full rationale
The paper's central claim is that the measured oscillation-frequency reduction with imprinted phase matches the rescaled ideal Josephson current-phase relation I0 = 2IC sin(φ0/2). Although IC is determined from the first three data points, this only fixes the overall amplitude; the sinusoidal shape at larger φ0 is an independent functional test. The critical-current crossover comparison in Fig. 3E uses a condensate fraction nc/n = 0.72 extracted from the measured critical current via the same theory that generates the blue curve, so the absolute scale at the molecular-side points is not a parameter-free prediction; however, the paper discloses this extraction and the interaction dependence of the curve is an independent comparison. The self-citations (Refs. 37 and 46) are used for simulation methodology and a future-proposal suggestion, not as load-bearing uniqueness or derivation steps. The supplementary limitations, including the statement that 'it is a priori unclear whether the atoms inside and outside the box potential are fully thermalised' and the note that quantitative accuracy beyond the bosonic case is left unverified, are correctness and support concerns rather than circularity. The derivation chain from the LC-circuit model to the effective current-phase relation is self-contained, and the observed phase-coherent oscillations and sinusoidal response are not equivalent by construction to the inputs. Overall, the paper is not significantly circular; the main honest caveats are the fitted amplitude in Fig. 2F and the partially calibrated theory curve in Fig. 3E, neither of which undermines the core inference.
Assumptions & free parameters
free parameters (3)
- Barrier height calibration factor =
Not stated numerically; DMD width W is mapped to V0/µ by matching data to the c-field simulation
- Critical current IC in Fig. 2F =
Determined from the first three data points of I0(phi0)
- Condensate fraction nc/n =
0.72(8)_stat (+0.1, -0.2)_sys
assumptions (5)
- domain assumption The junction and bulk can be modeled as a series LC circuit with a nonlinear Josephson inductance LJ(phi), a linear bulk inductance LB, and a capacitance C (Eq. 1 in the supplementary).
- domain assumption The Josephson-Anderson relation UJ = hbar dphi/dt and the relation dI/dphi = hbar/LJ hold for the atomic system.
- domain assumption The gas is in the two-dimensional superfluid phase, with T/TF approximately 0.03 below the predicted Tc/TF approximately 0.1.
- domain assumption The BKT scaling relations nc approximately n(L/r0)^(-eta) and eta = 2(n/ns)(T/TF) describe the finite-size two-dimensional condensate density.
- domain assumption The tunneling amplitude tk=0 can be computed from the Gross-Pitaevskii mean-field ansatz in Eqs. 13-20 of the supplementary.
Cite this review
Pith. "Pith review of An ideal Josephson junction in an ultracold two-dimensional Fermi gas." pith.science (2026). https://pith.science/paper/NI3NTFUC
@misc{pith2026190809776,
author = {Pith},
title = {Pith review of: An ideal Josephson junction in an ultracold two-dimensional Fermi gas},
year = {2026},
howpublished = {\url{https://pith.science/paper/NI3NTFUC}},
note = {Machine review of arXiv:1908.09776}
}
read the original abstract
The role of reduced dimensionality in high temperature superconductors is still under debate. Recently, ultracold atoms have emerged as an ideal model system to study such strongly correlated 2D systems. Here, we report on the realisation of a Josephson junction in an ultracold 2D Fermi gas. We measure the frequency of Josephson oscillations as a function of the phase difference across the junction and find excellent agreement with the sinusoidal current phase relation of an ideal Josephson junction. Furthermore, we determine the critical current of our junction in the crossover from tightly bound molecules to weakly bound Cooper pairs. Our measurements clearly demonstrate phase coherence and provide strong evidence for superfluidity in a strongly interacting 2D Fermi gas.
Figures
Reference graph
Works this paper leans on
-
[1]
B. D. Josephson, Phys. Lett. 1, 251 (1962)
1962
-
[2]
P. W. Anderson and J. M. Rowell, Phys. Rev. Lett. 10, 230 (1963)
1963
-
[3]
A. A. Golubov, M. Y. Kupriyanov, and E. Il’Ichev, Rev. Mod. Phys. 76, 411 (2004)
work page 2004
- [4]
- [5]
-
[6]
A. T. Sommer, L. W. Cheuk, M. J. H. Ku, W. S. Bakr, and M. W. Zwierlein, Phys. Rev. Lett. 108, 045302 (2012)
work page 2012
-
[7]
Makhalov, K
V. Makhalov, K. Martiyanov, and A. Turlapov, Phys. Rev. Lett. 112, 045301 (2014)
2014
-
[8]
W. Ong, C. Cheng, I. Arakelyan, and J. E. Thomas, Phys. Rev. Lett. 114, 110403 (2015)
work page 2015
Show all 51 references
-
[9]
Fenech et al., Phys
K. Fenech et al., Phys. Rev. Lett. 116, 045302 (2016)
2016
-
[10]
Mitra, P
D. Mitra, P. T. Brown, P. Schauß, S. S. Kondov, and W. S. Bakr, Phys. Rev. Lett. 117, 093601 (2016)
2016
-
[11]
Mazurenko et al., Nature 545, 462 (2017)
A. Mazurenko et al., Nature 545, 462 (2017)
2017
-
[12]
Levinsen and M
J. Levinsen and M. M. Parish, Strongly interacting two- dimensional fermi gases, in Annual Review of Cold Atoms and Molecules, pp. 1–75, World Scientific, 2015
2015
-
[13]
M. G. Ries et al., Phys. Rev. Lett. 114, 230401 (2015)
2015
-
[14]
Cataliotti et al., Science 293, 843 (2001)
F. Cataliotti et al., Science 293, 843 (2001)
2001
-
[15]
Albiez et al., Phys
M. Albiez et al., Phys. Rev. Lett. 95, 010402 (2005)
2005
-
[16]
S. Levy, E. Lahoud, I. Shomroni, and J. Steinhauer, Na- ture 449, 579 (2007)
2007
-
[17]
LeBlanc et al., Phys
L. LeBlanc et al., Phys. Rev. Lett. 106, 025302 (2011)
2011
-
[18]
Betz et al., Phys
T. Betz et al., Phys. Rev. Lett. 106, 020407 (2011)
2011
-
[19]
Spagnolli et al., Phys
G. Spagnolli et al., Phys. Rev. Lett. 118, 230403 (2017)
2017
-
[20]
Valtolina et al., Science 350, 1505 (2015)
G. Valtolina et al., Science 350, 1505 (2015)
2015
-
[21]
Burchianti et al., Phys
A. Burchianti et al., Phys. Rev. Lett. 120, 025302 (2018)
2018
-
[22]
C. Ryu, P. Blackburn, A. Blinova, and M. Boshier, Phys. Rev. Lett. 111, 205301 (2013)
2013
-
[23]
Watanabe, F
G. Watanabe, F. Dalfovo, F. Piazza, L. Pitaevskii, and S. Stringari, Phys. Rev. A 80, 053602 (2009)
2009
-
[24]
Spuntarelli, P
A. Spuntarelli, P. Pieri, and G. Strinati, Phys. Rev. Lett. 99, 040401 (2007)
2007
-
[25]
Ancilotto, L
F. Ancilotto, L. Salasnich, and F. Toigo, Phys. Rev. A 79, 033627 (2009)
2009
-
[26]
Eckel, F
S. Eckel, F. Jendrzejewski, A. Kumar, C. Lobb, and G. Campbell, Phys. Rev. X 4, 031052 (2014)
2014
-
[27]
Supplementary Materials
-
[28]
Hueck et al., Phys
K. Hueck et al., Phys. Rev. Lett. 120, 060402 (2018)
2018
-
[29]
This corresponds to an interaction strength of ln(kFa2D) =−2.4. 5
-
[30]
S 3), which is far below the predicted critical temperature Tc/TF ≈ 0.1 for 2D superfluidity at this interaction strength [13, 27, 41]
Remarkably, achieving this low damping requires a tem- perature of T/T F ≲ 0.03 (Fig. S 3), which is far below the predicted critical temperature Tc/TF ≈ 0.1 for 2D superfluidity at this interaction strength [13, 27, 41]
-
[31]
J. G. Lee, B. J. McIlvain, C. Lobb, and W. Hill III, Sci. Rep. 3, 1034 (2013)
2013
-
[32]
Eckel et al., Phys
S. Eckel et al., Phys. Rev. A 93, 063619 (2016)
2016
-
[33]
We note that achieving a sinusoidal current phase re- lation requires a barrier deep in the tunnelling regime which is fulfilled for our barrier depth of V0/µ≈ 1.5
-
[34]
Zaccanti and W
M. Zaccanti and W. Zwerger, Physical Review A 100, 063601 (2019)
2019
-
[35]
Hadzibabic and J
Z. Hadzibabic and J. Dalibard, RIV NUOVO CIMENTO 34 (2011)
2011
-
[36]
N. V. Prokof’ev and B. V. Svistunov, J. Exp. Theor. Phys. 127, 860 (2018)
2018
-
[37]
V. P. Singh, N. Luick, L. Sobirey, and L. Mathey, Phys. Rev. Research 2, 033298 (2020)
2020
-
[38]
J. J. Kinnunen, J. E. Baarsma, J.-P. Martikainen, and P. T¨ orm¨ a, Rep. Prog. Phys81, 046401 (2018)
2018
-
[39]
Hu et al., Nat
W. Hu et al., Nat. Mater. 13, 705 (2014)
2014
-
[40]
Okamoto, W
J.-i. Okamoto, W. Hu, A. Cavalleri, and L. Mathey, Phys. Rev. B 96, 144505 (2017)
2017
-
[41]
Petrov, M
D. Petrov, M. Baranov, G. Shlyapnikov, Phys. Rev. A 67, 031601 (2003)
2003
-
[42]
Luick et al., Zenodo, https://doi.org/10.5281/zenodo.3744797 (2020)
N. Luick et al., Zenodo, https://doi.org/10.5281/zenodo.3744797 (2020)
2020 doi
-
[43]
Luick et al., Zenodo, https://doi.org/10.5281/zenodo.3786298 (2020)
N. Luick et al., Zenodo, https://doi.org/10.5281/zenodo.3786298 (2020)
2020 doi
-
[44]
Petrov and G
D. Petrov and G. Shlyapnikov, Phys. Rev. A 64, 012706 (2001)
2001
-
[45]
Murthy et al., Phys
P. Murthy et al., Phys. Rev. A 90, 043611 (2014)
2014
-
[46]
V. P. Singh, C. Weitenberg, J. Dalibard, and L. Mathey, Phys. Rev. A 95, 043631 (2017)
2017
-
[47]
Turlapov and M
A. Turlapov and M. Y. Kagan, J. Phys. Condens. Matter 29, 383004 (2017)
2017
-
[48]
Boettcher et al., Phys
I. Boettcher et al., Phys. Rev. Lett. 116, 045303 (2016)
2016
-
[49]
Meier and W
F. Meier and W. Zwerger, Phys. Rev. A 64, 033610 (2001). 6 Supplementary materials Preparation of homogeneous 2D F ermi gases We prepare our homogeneous 2D Fermi gas follow- ing the scheme described in [28]. We start by evap- oratively cooling a spin mixture of 6Li atoms in th...
2001
-
[50]
for which φ′ 0 ≤ φ0. While this effective current is differ- ent from the instantaneous current, we can still relate the effective current phase relation I0(φ0) and the in- stantaneous current phase relation I(φ) by inserting the ideal current phase relation I(φ) = IC sin(φ) into Eq
-
[51]
1 numerically, but it is instructive to consider a simplified case which can be solved analytically
In principle, I0(φ0) can be found by solving Eq. 1 numerically, but it is instructive to consider a simplified case which can be solved analytically. If we assume that the dynamics of the system is dominated by the barrier (LJ≫ LB), we can neglect the bulk inductance LB in the ...
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.