🔬 mesoscale physics

Transmission representation of the Josephson free energy beyond the short-junction limit

This paper generalizes Beenakker's transmission representation of Josephson free energy beyond the short-junction limit to finite temperatures, proving a specific bound on the quartic-to-quadratic coefficient ratio and disproving a previous conjecture through a counterexample, with all mathematical derivations formally verified using the Lean proof assistant.

Original authors: Anton R. Akhmerov

Published 2026-10-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Anton R. Akhmerov

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the quiet world of superconductors, electricity flows without any resistance, a state that allows currents to persist indefinitely. When two such superconductors are separated by a tiny barrier, they can still talk to each other, allowing a special current to tunnel across the gap. This phenomenon, known as the Josephson effect, is the heartbeat of many modern quantum devices, including the qubits that power the most advanced quantum computers. The strength of this current depends on a hidden variable called the phase difference, a kind of internal clock setting between the two superconductors. For decades, scientists have relied on a simple rule to describe how the energy of this system changes as that clock setting shifts. This rule works perfectly when the barrier is extremely thin, but it breaks down when the barrier is thicker or more complex, leaving researchers without a reliable map for many real-world materials.

A researcher has now drawn a new map that works for junctions of any length, revealing that the energy landscape is far more flexible than previously thought. By developing a new mathematical framework that treats the flow of electrons as a collection of independent channels, each with its own frequency-dependent weight, they were able to calculate the energy of these systems with unprecedented precision. This approach allowed them to test a long-standing belief about the shape of the energy curve. For years, it was assumed that the energy curve could only bend in one direction, a constraint that would have simplified the design of quantum devices. However, the new calculations prove this assumption wrong. The researcher found specific configurations where the energy curve bends the other way, opening up new possibilities for engineering quantum circuits that were previously thought impossible.

The key to this discovery was a method that translates the complex quantum behavior of electrons into a sum of simple, independent contributions. Imagine the flow of electrons not as a chaotic flood, but as a series of distinct lanes, like cars on a multi-lane highway. In the simplest cases, the number of cars in each lane is fixed. But in the more complex, longer junctions studied here, the number of cars in each lane changes depending on the speed of the traffic. The researcher showed that even with this changing traffic, the total energy can still be calculated by adding up the contributions from each lane. This insight allowed them to derive strict mathematical limits on how much the energy curve can bend. They proved that while the curve can bend in either direction, there are hard limits on how extreme that bending can be.

One of the most significant outcomes of this work is the overturning of a specific conjecture that had guided the field for some time. Previous numerical studies had suggested that the energy curve could never bend in a way that would make the system behave in a certain "positive" manner. The new research explicitly disproves this, showing that such positive bending is not only possible but can be achieved in a simple, four-site junction. The researcher constructed a specific example of such a junction, a tiny square arrangement of atoms, and calculated its properties to confirm that it indeed exhibits this positive bending. This finding is not just a theoretical curiosity; it suggests that the materials used to build quantum computers might have more tunable properties than scientists realized, potentially allowing for more stable and versatile quantum bits.

To ensure that these complex results were beyond doubt, the researcher employed a rigorous verification process. They used a computer program designed to check mathematical proofs with absolute certainty, verifying every single equation in their paper. This level of scrutiny was necessary because the derivations were generated with the help of artificial intelligence, which, while powerful, can sometimes make subtle errors. By having the computer check every step against the rules of logic, the team guarantees that their conclusions are mathematically sound. The result is a robust new understanding of how superconducting junctions behave, replacing an old, incomplete rule with a comprehensive theory that works for any length of barrier.

The implications of this work extend beyond just fixing a formula. By proving that the energy curve can bend in unexpected ways, the researcher has opened a door to new types of quantum devices. The ability to engineer a junction with a specific, positive bending of its energy curve could lead to qubits that are less sensitive to noise and more reliable. While the paper focuses on the fundamental physics, it hints at a future where the design of quantum materials is guided by these precise, verified constraints. The work stands as a testament to the power of combining human insight with computational tools, using the latter to explore vast mathematical landscapes and the former to interpret the findings with care and precision. The journey from a simple, thin barrier to a complex, multi-site junction has revealed a richer, more nuanced world of quantum behavior, one where the rules are not as rigid as once believed.

Technical Summary: Transmission Representation of the Josephson Free Energy Beyond the Short-Junction Limit

Problem Statement
The paper addresses the challenge of describing the phase-dependent free energy of a Josephson junction beyond the "short-junction" limit. While C. W. J. Beenakker established a universal formula for short junctions where the energy-phase relation is a sum over independent transport channels with fixed transmission eigenvalues ( $\tau_n$ ), this representation fails for junctions of arbitrary length. In the general case, the spectrum depends on the full phase-dependent scattering matrix, and the coefficients of the free energy expansion ( $F(\phi) - F(0) = c_2\phi^2 + c_4\phi^4 + \dots$ ) are not easily constrained. Specifically, previous numerical studies on disordered, time-reversal-symmetric junctions led to a conjecture that the normalized quartic coefficient $\alpha_r = 12c_4/c_2$ is bounded by $-1 \le \alpha_r \le 0$ . This paper seeks to generalize the transmission representation to arbitrary junction lengths and rigorously test the validity of the upper bound $\alpha_r \le 0$ .

Methodology
The author derives a generalized transmission representation for the free energy of a time-reversal-symmetric $s$ -wave Josephson junction. The derivation proceeds through three main steps:

Matsubara Summation and Factorization: The phase-dependent part of the free energy is converted into a Matsubara sum. The Bogoliubov–de Gennes (BdG) Hamiltonian is factored using Takagi (for $\eta=-1$ ) or skew-Takagi/Youla (for $\eta=+1$ ) decompositions to isolate the superconducting phases onto a "channel space."
Scattering Matrix Reduction: Using a virtual-leads construction, the determinant of the BdG Hamiltonian is reduced to a scattering problem on the channel space. By imposing time-reversal symmetry and $s$ -wave pairing, the scattering matrix $S_\omega$ is shown to satisfy specific transpose symmetries.
Generalized Eigenvalue Pairing: The determinant ratio is expressed in terms of generalized eigenvalues of a pair of matrices derived from the scattering matrix. The author proves that non-endpoint eigenvalues occur in opposite-sign pairs ( $\pm \mu_n$ ), allowing the determinant to be factored into scalar terms dependent on frequency-dependent transmission eigenvalues $T_n(\omega)$ .

The resulting formula expresses the free energy difference as a sum over channels:
$F_\beta(H(\phi)) - F_\beta(H(0)) = -\frac{2}{\beta} \sum_{m=0}^\infty \sum_n \log\left(1 - T_n(\omega_m) \sin^2 \frac{\phi}{2}\right)$
where $T_n(\omega)$ are the Matsubara transmission eigenvalues, which reduce to the normal-state eigenvalues $\tau_n$ in the short-junction limit ( $\omega \to 0$ ).

Key Contributions and Results

Generalized Transmission Representation: The paper provides the first channel-based representation of the Josephson free energy for junctions of arbitrary length. Unlike the short-junction limit, the channel weights are now functions of Matsubara frequency, $T_n(\omega_m)$ , rather than fixed constants.
Bounds on Nonlinearities:
- Positivity: The representation proves that $\phi=0$ is a global free-energy minimum ( $c_2 \ge 0$ ).
- Quartic Coefficient Bounds: By analyzing the constraints on the frequency dependence of $T_n(\omega)$ $T_{n} (ω)$ (specifically using the Schwarz–Pick lemma on the analytic scattering matrix), the author derives tighter bounds on the normalized quartic coefficient $\alpha_r$ $α_{r}$ .
  - At finite temperature: $-1 \le \alpha_r \le 7/80$ .
  - At zero temperature: $-1 \le \alpha_r \le (16-5\pi)/[4(4-\pi)] \approx 0.085$ .
- These bounds allow for positive quartic coefficients ( $c_4 > 0$ ), contradicting the previous conjecture that $\alpha_r \le 0$ .
Disproof of the $\alpha_r \le 0$ Conjecture: The author constructs a specific four-site SNS junction (a square geometry with rank-one pairing blocks) that yields $\alpha_r \approx 0.0063 > 0$ . A larger realization of this construction yields $\alpha_r \approx 0.021$ . This explicitly disproves the conjectured upper bound of zero.
Finite-Length Corrections: The representation is used to derive the next-to-leading order correction to the short-junction energy. The author shows that the logarithmic correction term, previously thought to universally suppress the response, can have either sign depending on the lead dispersions.

Significance and Claims
The paper claims to resolve the question of whether the channel description of Josephson physics survives beyond the short-junction limit, affirming that it does, albeit with frequency-dependent weights. The primary theoretical contribution is the rigorous derivation of bounds on the anharmonicity of the Josephson potential, demonstrating that positive quartic nonlinearity is possible in conventional SNS junctions.

The author emphasizes the reliability of their results through formal verification. Every numbered equation in the manuscript is generated from a statement checked in the Lean proof assistant. This ensures that the derivations are free of algebraic errors, allowing readers to trust the results without repeating the complex derivations, provided the physical hypotheses (time-reversal symmetry, $s$ -wave pairing) match their system of interest.

The paper also serves as a case study in the use of AI models for theoretical physics research. The author discloses that they formulated the problem and directed the research, but the derivations, code, and manuscript text were produced by various AI models (GPT-5.2 through GPT-6.1, Claude Fable 5.1, and Claude Opus 5.5). The author notes that while the models became reliable once the tasks were correctly formulated, their self-assessment of task completion was often untrustworthy, necessitating rigorous human validation and formal verification.

Limitations and Open Questions
The paper does not claim that positive quartic coefficients are common in all real junctions, but rather that they are theoretically possible within the standard model of non-interacting time-reversal-symmetric $s$ -wave junctions. The author suggests that flat-band Josephson junctions might be a promising system to observe this effect experimentally. They also note that extending this transmission representation to bulk superconductors (where the phase varies continuously) remains an open problem. Additionally, while computer-assisted extensions suggest even tighter bounds (e.g., $\alpha_r \le 0.028$ ), these specific extensions have not yet been formally verified to the standard of the main manuscript.

Technical Summary: Transmission Representation of the Josephson Free Energy Beyond the Short-Junction Limit

More like this