Realization of a Superleak in a Quantum Gas

arXiv:2610.31185 · cond-mat.quant-gas, physics.atom-ph, quant-ph · Submitted 2026-10-22 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: We're looking at the paper: "Realization of a Superleak in a Quantum Gas".

Mira: Here's the gist of the paper. The gist The superfluid undergoes coherent Josephson transport, whereas the normal component exhibits slow, thermodynamically limited response, resulting in a flow-rate contrast of 500(49) Realization of Dynamical Selectivity The researchers realize a dynamical superleak by exploiting the intrinsically different dynamical responses of the superfluid and normal components Using an optical superlattice, the superfluid undergoes coherent transport, while the response of the low-dimensional normal component is strongly suppressed by slow thermalization This results in a pronounced separation of their characteristic timescales, with the superfluid flow rate exceeding that of the normal component by a factor of 500(49) A microscopic theory quantitatively captures the observed dynamics and reveals distinct scaling behaviors governed by coherent tunneling and thermalization, respectively Theoretical Framework The dynamical superleak is engineered using an optical superlattice where neighboring quasi-two-dimensional (2D) regions are coupled through optical weak links This bichromatic superlattice drives particle redistribution between the two sublattices with a timedependent energy offset ∆(t) The transport dynamics of each component are characterized by the flow rate Is,n = ∂t(No s,n − Ne s,n)/2 For the superfluid component, macroscopic phase coherence gives rise to phase rigidity with a superfluid velocity determined by the spatial phase gradient vs = ħ m∇ϕ (1) The corresponding particle-current density is given by js = nsvs Transport Dynamics Contrast The superfluid component exhibits phase rigidity, with the superfluid velocity determined by the spatial phase gradient In contrast, the normal component lacks phase coherence and its transport is governed by incoherent quantum tunneling The slow thermalization process sets the slow timescale of normal transport, leading to a delayed and dissipative response compared to the phase-coherent superfluid flow The normal transport relaxation rate scales as Γ ∝ Γth, where Γth is the cross-dimensional thermalization rate Dynamical Superleak and Purification The strong transport contrast enables a dynamical filtering process analogous to a conventional superleak This involves repeated filtering of the normal component, thereby purifying the superfluid The purification protocol combines dynamical separation with site-selective atom removal The long lattice provides additional confinement for the even sites in the y-z plane, allowing atoms on the odd sites to spill out when the transverse dipole trap is lowered Repeating this transport-and-removal cycle progressively filters out the normal component and purifies the superfluid Sustaining Coherent Dynamics The purification protocol continuously removes thermal excitations while retaining coherent oscillations during sinusoidal driving This ability to selectively remove entropy-carrying excitations provides a route to maintaining low-entropy quantum matter during driven dynamics and may offer a general tool for stabilizing coherent many-body transport The resulting superflow remains coherent with no detectable decay over 100 driving cycles, in contrast to the pronounced damping observed without purification Experimental Verification The experiment starts with ultracold gases of ≃ 2 × 10 5 87Rb atoms confined in a 3D optical dipole trap The condensate fraction can be finely tuned by controlling the final depth of the dipole potential By preparing samples with different condensate fractions, distinct transport timescales are revealed, with the superfluid component transferring within milliseconds whereas a thermal sample exhibits transport over hundreds of milliseconds The extracted superfluid transport rate exceeds the normal component by a factor of 500(49) Characterization of Timescales The characteristic timescale for the superfluid transport is set by the Josephson critical current, with Ic = 2J/ħ defining the upper bound for coherent flow through the weak link The measurements collapse when time is rescaled by Jτ /h, confirming that the coherent transport timescale is determined by the tunneling energy The normal component relaxation times are approximately two orders of magnitude longer than the superfluid transport timescale The two timescales remain clearly separated throughout the coexistence regime, showing only weak temperature dependence In summary, this work realizes a dynamical superleak in an ultracold atomic gas by exploiting the distinct transport mechanisms of the superfluid and normal components Combined with site-selective removal, this dynamical selectivity enables repeated filtering of the normal component, thereby purifying the superfluid and stabilizing coherent dynamics against excitations This approach provides a platform for engineering quantum fluids and exploring nonequilibrium many-body dynamics and dissipation The ability to maintain superfluid coherence through dynamical filtering may benefit atomtronic quantum devices and precision measurements Acknowledgments We thank P.

Kai: Hauke, S.

Mira: Jochim, Q.

Kai: Niu, A.

Title and authors: Kai: So, we're looking at this paper today, "Realization of a Superleak in a Quantum Gas," and the main thing here is they managed to get something they call dynamical selectivity. It’s about using an optical superlattice to separate how the superfluid and normal parts of the gas move.

Mira: Exactly, it's fascinating because it shows that you can make them respond totally differently, which leads to this flow-rate contrast of five hundred times between the two components. That's a big number for how distinct their responses are.

Lev: From an engineering standpoint, that contrast is what makes this useful for building better hardware; if we could control that separation, it would help us manage noise in these systems way more effectively.

Kai: Right, so the core idea they're presenting is exploiting that difference between coherent Josephson transport in the superfluid and the slow thermalization of the normal component. They use a bichromatic superlattice to drive particle redistribution with this time-dependent energy offset, (t).

Mira: And what makes it special is that while the superfluid has macroscopic phase coherence, which gives it phase rigidity and a velocity determined by its spatial phase gradient, the normal component lacks that coherence.

Lev: That means the normal transport is governed by incoherent quantum tunneling and this slow thermalization process sets its timescale, scaling with th, which is the cross-dimensional thermalization rate.

Kai: So they're essentially engineering a way to filter out the normal fluid because it takes much longer to respond compared to the superfluid flow. They use this contrast for a dynamical superleak, which is like filtering the normal component by repeatedly filtering it out.

Mira: That purification protocol seems like a clever way to selectively remove entropy-carrying excitations while keeping the coherent oscillations alive during sinusoidal driving, which is interesting for maintaining low-entropy quantum matter under continuous operation.

Lev: If we think about running this on actual hardware, the paper shows that this filtering process can sustain coherent oscillations for more than one hundred cycles without detectable decay, which is pretty good news for stability.

Kai: Experimentally, they start with ultracold gases of two times ten five eighty-seven rubidium atoms in a three-dimensional optical dipole trap and tune the condensate fraction to see the different timescales emerge.

Mira: And the numbers really back that up, showing that while the superfluid component transfers within milliseconds, a thermal sample takes over hundreds of milliseconds to move. The extracted superfluid transport rate exceeds the normal component by a factor of five hundred times, which is what they found in this work.

Lev: That five-hundred-fold difference is significant because it means the coherent flow is not just slightly faster; it's orders of magnitude faster under these specific conditions.

Kai: The characteristic timescale for the superfluid transport is set by the Josephson critical current, I c, which defines the upper bound for coherent flow through that weak link, and they confirm this when time is rescaled by J tau/h.

Mira: And it's clear that the normal component relaxation times are about two orders of magnitude longer than that superfluid transport timescale, and this separation stays pretty consistent across the coexistence regime.

Lev: That consistency across the temperature dependence is important; it suggests this dynamical selectivity isn't just a fluke happening at one specific point, but a feature of how these components fundamentally respond.

Kai: So, to wrap up on this "Realization of a Superleak in a Quantum Gas," the authors show that by exploiting the distinct transport mechanisms—coherent tunneling versus slow thermalization—they can achieve dynamical purification.

Mira: The implication is that we have a platform for engineering quantum fluids where we can actively purify them and stabilize coherence during driven operations, which is a route toward more robust coherent many-body dynamics in these systems.

Lev: For real hardware development, this suggests we could use such filtering protocols to continuously remove thermal excitations when running experiments on superconducting qubits or other strongly correlated systems.

Kai: Yeah, so while they show the physics here, the practical implication for us is a method for stabilizing coherence through dynamical filtering that could benefit atomtronic quantum devices and precision measurements down the line.

Mira: It’s a lot of work to build this optical superlattice and control all those parameters, but seeing that flow-rate contrast of five hundred times validates the approach they took in realizing this superleak.

Lev: We need to keep thinking about how much overhead that filtering process adds when you try to implement it on a real system where you have inherent dissipation from the environment.

Kai: Well, that's where we'll be next time, looking at how we can take these ideas and see what actual fidelity we can achieve in our superconducting circuits.

The paper's summary: Kai: So, to wrap up what we just heard about "Realization of a Superleak in a Quantum Gas," the core finding is that they actually managed to create this dynamical superleak using an optical superlattice that exploits how the superfluid and normal parts of the gas move at completely different speeds.

Mira: Exactly, it’s not just one component moving fast while the other moves slow; it’s about making sure their fundamental ways of transporting particles are totally distinct, which results in a flow-rate contrast that is five hundred times larger for the superfluid compared to the normal part.

Lev: From what I see on paper, that huge difference in transport time is what really matters because it lets them do this kind of dynamical filtering process, which essentially purifies the superfluid by repeatedly removing those slower, entropy-carrying excitations.

Kai: And that purification protocol is pretty neat because they managed to keep the coherent oscillations going for over one hundred driving cycles without any detectable decay, which is a big deal for stability in these driven quantum systems.

Mira: What makes this interesting from a theory standpoint is how they show this separation of timescales—the superfluid transport time being set by the Josephson critical current, I c, and the normal component relaxation time scaling with that cross-dimensional thermalization rate, th.

Lev: That scaling relationship between the thermalization rate and the slow response of the normal fluid is what tells us exactly how much overhead we’re looking at when trying to implement this kind of separation on a real quantum hardware platform.

Kai: So if we think about this for our own work, it means that if we can engineer systems with such strong dynamical selectivity, we might find ways to actively suppress noise or thermal excitations in our qubits during operations.

Mira: Right, and the implication is that this isn't just a curiosity for ultracold gases; it suggests a general tool for maintaining low-entropy quantum matter while you’re driving it around.

Lev: It changes how we think about error correction because instead of just trying to correct errors after they happen, you could potentially filter them out in real time using this kind of tailored transport mechanism.

Kai: So the takeaway here is that by building a system where coherent transport and thermal response are separated by a factor of five hundred, you can engineer a purification route for quantum fluids.

Mira: It's about showing that you can use site-selective atom removal as part of a dynamical filtering protocol to increase the condensate fraction from around fifty percent up to over ninety-four percent in these lattice systems.

Lev: That jump in purity is what makes it useful; we’re not just observing a phenomenon, we’re looking at how you can actively engineer high-purity quantum states for use in more complex quantum information processing.

Kai: So the next thing to look at is whether this purification method can be scaled up from a small optical superlattice setup to something more practical for large-scale many-body systems.

The paper's improvements: Tom: So we’re looking at what they suggest next for this "Realization of a Superleak in a Quantum Gas," and basically, they are proposing ways to make the system even cleaner and more stable by taking that purification idea further.

Kai: They talk about how you can use AI to model transport in systems with both coherent and thermal parts, which means we could get way more accurate predictions about quantum fluid dynamics than just a single-component model gives us.

Mira: I mean, the paper shows this dynamical filtering process, so the next step is using AI to develop methods that can stabilize coherent many-body states during operations in actual quantum devices.

Lev: If we take that idea of dynamically removing entropy-carrying excitations and use AI to guide that removal, it could help us maintain those coherent oscillations for more than one hundred cycles even when we’re doing continuous driving.

Kai: And they also suggest using this purification protocol to engineer super high-purity quantum states; they showed that by repeating the filtering four times, you can get the condensate fraction from about fifty percent up to over ninety-four percent in these lattice systems.

Mira: That improvement in purity is what’s huge because it means we could be preparing higher-quality superfluids for use in strongly correlated lattice systems, which is a big step for material science applications.

Lev: For hardware, that high purity translates directly into lower noise floors when you try to run those complex error correction schemes we’re building, so the overhead of maintaining coherence becomes much more manageable.

Kai: They also focus on optimizing the transport characteristics for quantum information processing by controlling flow rates through those weak links.

Mira: Specifically, they suggest that AI can figure out that shorter ramps requiring faster superfluid transport are needed to avoid strong excitations and keep the coherent flow below that critical value of Ic, which is two over J over h.

Lev: That’s a very practical point because if we can guide the driving parameters based on those theoretical constraints, we might be able to design pulse sequences that prevent the system from entering an excited state in the first place.

Kai: So they are moving beyond just observing this superleak and suggesting a way to actively control and tune the system through AI-driven feedback loops.

Mira: It’s about building an integrated control scheme where you use site-selective atom removal guided by that AI to keep things coherent while you’re doing the heavy lifting of driving the system.

Lev: What they don't explicitly cover is how this purification protocol scales up from a simple optical superlattice to something that works on a larger, more realistic quantum chip architecture.

Kai: That’s the gap for us, seeing if we can translate these dynamical selection principles into something that runs robustly in the real cryogenic environment we’re working in.

Conclusion: Kai: So to wrap up this discussion on "Realization of a Superleak in a Quantum Gas," we’re looking at how they actually managed to build this dynamical superleak by exploiting the difference between coherent Josephson transport and slow normal component response.

Mira: Right, it really shows that you can engineer these distinct dynamical scalings—coherent tunneling versus thermalization—which gives us a flow-rate contrast of five hundred times between the two components.

Lev: From where I’m looking at this for hardware, that massive factor of five hundred is what makes the separation so pronounced, and it’s not just theoretical; it dictates how much overhead we’re dealing with when you try to run these kinds of operations on real quantum hardware.

Kai: And they showed that by using a site-selective removal protocol, this dynamical filtering can purify the superfluid, keeping its coherent oscillations going for over one hundred cycles without any detectable decay.

Mira: That purification means we’re talking about preparing high-purity superfluids in strongly correlated lattice systems with condensate fractions jumping from about fifty percent to over ninety-four percent after just four cycles.

Lev: That level of purity is what makes it useful; it lowers the noise floor significantly when you try to run those complex error correction schemes we’re building for real hardware.

Kai: So, the implication here is that by mastering this dynamical selectivity, we get a platform for engineering quantum fluids and actively stabilizing coherence during driven operations.

Mira: It’s about showing that you can use this separation of timescales to maintain low-entropy quantum matter even when it’s being driven around sinusoidally.

Lev: The limitation they point out is that this whole purification protocol relies on having a long lattice for the confinement of those even sites in the y-z plane, which means it might not be immediately scalable to more complex geometries.

Kai: So, while they proved the physics here with ultracold gases, our next step needs to figure out how we translate this mechanism into something that actually runs reliably on superconducting qubits.

Mira: I think the next big thing is using AI to model this transport in systems with both coherent and thermal components so we can predict these dynamics much more accurately before we even start building the hardware.

Lev: And for error correction, it’s about figuring out how to integrate this filtering mechanism into a circuit design so that the overhead of keeping things pure doesn't kill our overall operation time.

Wan-Ling Yang, *Shao-Bo Liu, *Xiang Luo, *Chao Yang

State Key Laboratory of Quantum Functional Materials · Department of Physics and Guangdong Basic Research Center of Excellence for Quantum Science, Southern University of Science and Technology, Shenzhen 518055, China · Quantum Science Center of Guangdong-Hong Kong-Macao Greater Bay Area, Shenzhen 518045, China

cond-mat.quant-gas, physics.atom-ph, quant-ph

Submitted: 2026-10-22

Updated: 2026-10-22

Comments: 14 pages, 4 main figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The gist: The gist The superfluid undergoes coherent Josephson transport, whereas the normal component exhibits slow, thermodynamically limited response, resulting in a flow-rate contrast of 500(49)

Key concepts

Superfluid Component
This part of the gas exhibits macroscopic phase coherence, meaning all atoms move together in a synchronized manner. Its transport is governed by phase rigidity, allowing it to flow coherently through weak links without significant dissipation.
Normal Component
This component lacks long-range phase coherence and its transport is governed by incoherent quantum tunneling. It responds slowly because of slow thermalization, meaning it takes a long time to reach equilibrium compared to the superfluid.
Dynamical Superleak
This is a filtering process that uses the difference in transport speeds between the two components. By repeatedly filtering out the slower, normal component, researchers purify the superfluid and maintain its coherent flow even under external driving.

Summary

The gist The superfluid undergoes coherent Josephson transport, whereas the normal component exhibits slow, thermodynamically limited response, resulting in a flow-rate contrast of 500(49)

Realization of Dynamical Selectivity

The researchers realize a dynamical superleak by exploiting the intrinsically different dynamical responses of the superfluid and normal components Using an optical superlattice, the superfluid undergoes coherent transport, while the response of the low-dimensional normal component is strongly suppressed by slow thermalization This results in a pronounced separation of their characteristic timescales, with the superfluid flow rate exceeding that of the normal component by a factor of 500(49) A microscopic theory quantitatively captures the observed dynamics and reveals distinct scaling behaviors governed by coherent tunneling and thermalization, respectively

Theoretical Framework

The dynamical superleak is engineered using an optical superlattice where neighboring quasi-two-dimensional (2D) regions are coupled through optical weak links This bichromatic superlattice drives particle redistribution between the two sublattices with a timedependent energy offset ∆(t) The transport dynamics of each component are characterized by the flow rate Is,n = ∂t(No s,n − Ne s,n)/2 For the superfluid component, macroscopic phase coherence gives rise to phase rigidity with a superfluid velocity determined by the spatial phase gradient vs = ħ m∇ϕ (1) The corresponding particle-current density is given by js = nsvs

Transport Dynamics Contrast

The superfluid component exhibits phase rigidity, with the superfluid velocity determined by the spatial phase gradient In contrast, the normal component lacks phase coherence and its transport is governed by incoherent quantum tunneling The slow thermalization process sets the slow timescale of normal transport, leading to a delayed and dissipative response compared to the phase-coherent superfluid flow The normal transport relaxation rate scales as Γ ∝ Γth, where Γth is the cross-dimensional thermalization rate

Dynamical Superleak and Purification

The strong transport contrast enables a dynamical filtering process analogous to a conventional superleak This involves repeated filtering of the normal component, thereby purifying the superfluid The purification protocol combines dynamical separation with site-selective atom removal The long lattice provides additional confinement for the even sites in the y-z plane, allowing atoms on the odd sites to spill out when the transverse dipole trap is lowered Repeating this transport-and-removal cycle progressively filters out the normal component and purifies the superfluid

Sustaining Coherent Dynamics

The purification protocol continuously removes thermal excitations while retaining coherent oscillations during sinusoidal driving This ability to selectively remove entropy-carrying excitations provides a route to maintaining low-entropy quantum matter during driven dynamics and may offer a general tool for stabilizing coherent many-body transport The resulting superflow remains coherent with no detectable decay over 100 driving cycles, in contrast to the pronounced damping observed without purification

Experimental Verification

The experiment starts with ultracold gases of ≃ 2 × 10 5 87Rb atoms confined in a 3D optical dipole trap The condensate fraction can be finely tuned by controlling the final depth of the dipole potential By preparing samples with different condensate fractions, distinct transport timescales are revealed, with the superfluid component transferring within milliseconds whereas a thermal sample exhibits transport over hundreds of milliseconds The extracted superfluid transport rate exceeds the normal component by a factor of 500(49)

Characterization of Timescales

The characteristic timescale for the superfluid transport is set by the Josephson critical current, with Ic = 2J/ħ defining the upper bound for coherent flow through the weak link The measurements collapse when time is rescaled by Jτ /h, confirming that the coherent transport timescale is determined by the tunneling energy The normal component relaxation times are approximately two orders of magnitude longer than the superfluid transport timescale The two timescales remain clearly separated throughout the coexistence regime, showing only weak temperature dependence

Conclusion

In summary, this work realizes a dynamical superleak in an ultracold atomic gas by exploiting the distinct transport mechanisms of the superfluid and normal components Combined with site-selective removal, this dynamical selectivity enables repeated filtering of the normal component, thereby purifying the superfluid and stabilizing coherent dynamics against excitations This approach provides a platform for engineering quantum fluids and exploring nonequilibrium many-body dynamics and dissipation The ability to maintain superfluid coherence through dynamical filtering may benefit atomtronic quantum devices and precision measurements

Acknowledgments

We thank P. Hauke, S. Jochim, Q. Niu, A. Smerzi, S. Stringari, Y.-C. Wang, B. Wu, S.-Z Zhang, H.-Z Zhao for insightful discussions We thank Z.-B Yang and J.-Y Shang for their assistance in the initial studies This work is supported by the National Key R&D Program of China (Grant No. 2022YFA1405800), NSFC (Grant No. 12274199), Shenzhen Science and Technology Program (Grant No. KQTD20240729102026) and Guangdong Major Project of Basic and Applied Basic Research (Grant No. 2023B0303000011) X. L. acknowledges China postdoctoral science foundation (No. 2023M741547) The authors contributed equally to this work.

References

[1] P. Kapitza, Nature 141, 74 (1938)

[2] J. F. Allen and A. D. Misener, Nature 141, 75 (1938)

[3] L. Landau, Phys. Rev. 60, 356 (1941)

[4] A. Leggett, Quantum Liquids: Bose Condensation and Cooper Pairing in Condensed-matter Systems (Oxford University Press, Oxford, 2006)

[5] M. A Nielsen and I. L. Chuang, Quantum Computation and Quantum Information (Cambridge University Press, 2012)

[6] C. J. Pethick and H. Smith, Bose-Einstein Condensation in Dilute Gases (Cambridge University Press, 2008)

[7] I. Bloch, J. Dalibard, and W. Zwerger, Rev. Mod. Phys. 80, 885 (2008)

[8] L. Pitaevskii and S. Stringari, Bose-Einstein Condensation and Superfluidity (Oxford University Press, Oxford, 2016)

[9] C. Raman, M. K¨ohl, R. Onofrio, D. S. Durfee, C. E.

Improvements for AI systems

  1. Improved AI system can perform enhanced energy-efficient quantum simulation by modeling transport in systems with distinct coherent and thermal components. This allows for more accurate prediction of quantum fluid dynamics beyond single-component models by incorporating the distinct dynamical scalings observed between superfluid and normal transport.

  2. Improved AI system can develop novel methods for stabilizing coherent many-body states during driven operations in quantum devices. By utilizing the dynamical filtering process analogous to a conventional superleak, the AI can continuously remove entropy-carrying excitations, enabling maintaining coherent oscillations for more than 100 cycles even under continuous driving.

  3. Improved AI system can engineer high-purity quantum states using a purification protocol integrated into dynamic control schemes. The system can implement repeated filtering of the normal component, which increases the condensate fraction from approximately 50% to over 94% after four cycles, leading to preparing high-purity superfluids in strongly correlated lattice systems.

  4. Improved AI system can optimize transport characteristics for quantum information processing by controlling flow rates through weak links. The AI can determine that shorter ramps requiring faster superfluid transport are needed to avoid strong excitations, thereby ensuring the coherent flow stays below the critical value Ic = 2J/h where strong excitations emerge.

Related papers