Programmable Quantum Junctions Turn Entropy Flow Into a Circuit Variable

A new July 2026 theory paper argues that driven qubit circuits can do something more subtle than move heat: they can program the physical flow of entropy itself.

Most energy stories are told in watts, joules and efficiency. Quantum thermodynamics adds a stranger accounting problem. In a quantum device, useful work, waste heat, information, coherence and measurement back-action are not cleanly separated. The difference matters because future quantum processors, sensors and refrigerators will not just need to dump heat. They will need to manage entropy without destroying the fragile states that make the hardware useful.

That is why “Entropy Transport in Programmable Quantum Junctions,” submitted to arXiv on July 14, 2026 by Radhika Joshi, Yuli V. Nazarov and Mohammad H. Ansari, is an important addition to the Floquet energy map. The paper studies driven qubit junctions coupled to thermal reservoirs and asks whether entropy conductance can be governed by the microscopic quantum dynamics of a circuit, not only by reservoir temperatures. Its answer is yes: with periodic driving and qubit architecture as control knobs, entropy current becomes a programmable transport quantity.

The paper’s most provocative message is not that a quantum circuit can carry heat. It is that a driven quantum circuit can reshape entropy flow through coherence, resonance and junction design.

Why entropy transport is not ordinary heat transport

Heat is energy exchanged because of temperature differences. Entropy is broader. It measures missing information, disorder and the number of microscopic configurations compatible with what we observe. In thermodynamics, heat flow and entropy production are tightly linked. In quantum devices, however, entropy can also be carried, suppressed or reshaped by coherence and correlations.

That makes entropy harder to treat like a normal current. Charge current is associated with the expectation value of a charge-flow operator. Energy current has a Hamiltonian bookkeeping. The Joshi-Nazarov-Ansari paper emphasizes that information-theoretic entropy flow, including Rényi and von Neumann entropy conductance, is nonlinear in the reservoir density matrix. In plain English: entropy current is not simply “another meter reading” on the same footing as voltage or heat. It needs a deeper transport framework.

The new work builds that framework for a general N-qubit junction coupled to M thermal environments, then examines two simple cases: a driven single-qubit junction and a driven two-qubit junction. Each qubit can have a transition frequency, pairwise interaction with another qubit and an external periodic drive. The reservoirs are treated through a dissipative master-equation description, while the entropy flow into a weakly coupled probe reservoir is evaluated using full-counting-statistics and Keldysh-style tools.

The Floquet connection: periodic driving as a thermodynamic dial

Floquet engineering enters because the junction is periodically driven. A time-periodic field can make a qubit absorb or emit packets of drive energy, opening resonant pathways that are absent in a static circuit. For energy applications, the point is not just to shake the system. The point is to tune which transitions are available and how strongly they contribute to transport.

In the paper’s model, the drive amplitude and frequency influence the quantum state of the junction, and that state determines the entropy conductance into the probe reservoir. The result is a thermodynamic control layer: reservoirs still matter, but the circuit’s quantum logic and drive protocol matter too. This is exactly the kind of design language Floquet energy research needs if it is to move from exotic phases toward devices.

The two-qubit result is especially interesting. According to the abstract and paper body, the two-qubit junction enhances entropy transfer while requiring substantially lower driving power than its single-qubit counterpart. That is a qualitative design lesson: adding a controlled quantum degree of freedom can make entropy movement more efficient, not merely more complicated.

Coherent entropy current: the nonclassical part of the flow

One of the paper’s key findings is that entropy flow separates into incoherent and coherent contributions. The incoherent term is closer to the thermal intuition: transitions exchange excitations with a reservoir. The coherent term is more quantum. It depends on off-diagonal components of the system state and appears only under resonant driving.

For non-physicists, resonance is the familiar swing-pushing condition: energy transfer becomes effective when the drive is timed to the system’s natural motion. In a qubit junction, resonant driving can sustain coherent superpositions that alter how entropy enters the probe reservoir. The authors report a sizable coherent contribution to entropy current, showing that coherence is not just a computational resource; it can be a transport resource.

Coherence is often treated as something thermodynamics destroys. Here it becomes part of the entropy-current mechanism, provided the drive is tuned into the right resonant regime.

This matters for quantum refrigeration and reservoir protection. If a circuit can suppress entropy flow into one reservoir while allowing it into another, it could help shield sensitive quantum components. If it can move entropy with lower drive power, it could support cooling strategies for superconducting qubits or nanoscale sensors. The paper is theoretical, so those device claims remain future-facing, but the control principle is clear.

Negative differential entropy conductance

The second non-intuitive effect is negative differential entropy conductance. In ordinary language, increasing the thermal bias does not always increase entropy flow into the probe reservoir. Under the modeled conditions, increasing the bias can suppress the entropy flow. This resembles negative differential conductance in electronic devices, where more voltage can produce less current, but the transported quantity here is entropy.

That is not a violation of thermodynamics. The external drive and quantum state preparation remain part of the balance sheet. But it is a reminder that driven quantum systems do not behave like passive thermal pipes. The route from a hot reservoir to a probe can be blocked, redirected or reshaped by coherence and architecture. For engineering, negative differential entropy conductance could become a switching effect: a way to protect a reservoir or trigger feedback when a thermal gradient becomes too large.

How this fits with July’s quantum heat-engine work

The timing is useful because another July 2026 arXiv paper, “Extracting Work from Discrete Quantum Polytropic Processes” by Vishal Anand, Swarup Kumar Giri, Avijit Misra, Subhadip Mitra and Samyadeb Bhattacharya, approaches quantum thermodynamics from the heat-engine side. That work derives an upper bound on extractable work for time-dependent, non-Markovian quantum heat engines operating with finite baths. It isolates penalties from system-bath correlations, bath non-equilibrium and residual interaction energy.

Together, the two papers sharpen the same message. Quantum thermodynamic advantage is not free. The polytropic-process paper argues that maximal efficiency requires quasi-static operation to harvest coherent, non-Markovian resonances, while maximum power pushes the engine into a finite-time regime where discretization noise can wipe out the delicate memory effects. The entropy-junction paper shows a more constructive side: if the goal is not simply “maximum work” but controlled entropy routing, driven circuit architecture may offer useful knobs.

This distinction is important for floquet.ca’s mission. The most credible path to quantum-energy usefulness may not be a single engine that “beats Carnot” in a simplistic sense. It may be a collection of components that manage entropy more intelligently: sideband heat routers, driven quantum junctions, protected reservoirs, finite-time protocols and feedback loops that treat information as part of the thermodynamic design.

Why experimentalists should care

The authors explicitly point toward superconducting-qubit platforms, where driven qubits, microwave resonators and engineered reservoirs are already standard tools. That does not mean the proposed entropy junction is a finished chip. Measuring entropy currents is harder than measuring energy currents. But the building blocks are familiar: qubits with tunable drives, controlled coupling strengths, cryogenic reservoirs and increasingly sophisticated full-counting-statistics measurements.

If experiments can isolate the coherent contribution and observe negative differential entropy conductance, the result would be a milestone for quantum thermodynamics. It would show that entropy management can be engineered at the circuit level, rather than inferred after the fact from heat budgets. It would also provide a more practical bridge between quantum information hardware and energy science. The same platforms used to protect qubits from decoherence might be used to study how entropy can be directed, delayed or removed.

What to watch next

The immediate questions are concrete. Can the two-qubit advantage survive realistic noise, parameter disorder and measurement back-action? How much external drive power is required once the full microwave-control chain is included? Can a probe reservoir be monitored without disturbing the entropy flow being measured? And can feedback use the predicted conductance effects to protect a quantum memory or improve a refrigerator?

The broader lesson is already valuable. Entropy is becoming a design variable. In programmable quantum junctions, periodic driving does not merely inject energy; it selects thermodynamic pathways. That is a serious step toward quantum energy hardware whose purpose is not magic efficiency, but disciplined control over where disorder goes.

Sources and further reading

• Radhika Joshi, Yuli V. Nazarov and Mohammad H. Ansari, “Entropy Transport in Programmable Quantum Junctions,” arXiv:2607.12581, submitted July 14, 2026. Read the arXiv abstract.
• Vishal Anand, Swarup Kumar Giri, Avijit Misra, Subhadip Mitra and Samyadeb Bhattacharya, “Extracting Work from Discrete Quantum Polytropic Processes,” arXiv:2607.01811, submitted July 2, 2026. Read the arXiv abstract.
• Yingying Hong, Longxing Xu, Weiwei Zhang, Jie Ren and Jianhui Wang, “Geometric Mode Steering of the Quantum Mpemba Effect,” arXiv:2607.14501, submitted July 16, 2026. Read the arXiv abstract.