🔬 mesoscale physics

Magnetic Double-Wells: Lower Bounds on Tunneling

This paper establishes lower bounds on tunneling rates for generic double-well systems under strong magnetic fields and deep potentials, complementing previous findings that demonstrated vanishing tunneling in specially constructed cases.

Original authors: Charles L. Fefferman, Jacob Shapiro, Michael I. Weinstein

Published 2026-06-19

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Charles L. Fefferman, Jacob Shapiro, Michael I. Weinstein

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

The Big Picture: A Quantum Game of Hide and Seek

Imagine a tiny particle (like an electron) trapped in a landscape with two deep valleys, or "wells." In the world of quantum mechanics, this particle doesn't just sit still; it has a magical ability called tunneling. It can spontaneously disappear from one valley and reappear in the other, even if there is a high mountain separating them.

The speed at which this happens is called the tunneling rate. In a normal world (without magnets), this rate is always positive. The particle will always eventually hop over, though it might take a very, very long time if the mountains are high.

The Twist: The authors of this paper studied what happens when you turn on a very strong magnetic field.

The Discovery: When the Magic Stops (and When It Doesn't)

In a previous study, these same authors found a very strange, specific case: if you build the valleys in a very particular, "non-radial" (lopsided) shape and turn on a strong magnetic field, the tunneling can stop completely. The particle gets stuck in one valley forever. It's as if the magnetic field creates a perfect "lock" that prevents the particle from ever crossing over.

However, the authors realized this "perfect lock" is a fluke. It only happens for very specific, carefully engineered shapes of the valleys.

This new paper proves the opposite: For almost any other shape of the valleys (what they call "generic" cases), the tunneling never stops completely. Even with a strong magnetic field, there is always a non-zero chance the particle will hop over. The paper provides a mathematical guarantee (a "lower bound") that the tunneling rate will never be zero, except for a tiny, negligible set of special cases.

The Analogy: The Spinning Coin

To understand the math, imagine the particle is a spinning coin trying to jump from one side of a table to the other.

No Magnet: The coin spins and jumps randomly. It will eventually cross.
The "Perfect Lock" Case (Previous Work): If you arrange the table and the coin in a very specific, weird way, the magnetic field makes the coin spin in a pattern where the "heads" and "tails" cancel each other out perfectly. The coin vibrates in place but never crosses.
The "Generic" Case (This Paper): The authors say, "If you change the shape of the table even a tiny bit, or if you pick a random spot on the table, that perfect cancellation breaks." The coin will wobble, maybe spin weirdly, but it will still eventually cross.

The paper proves that while you can build a table where the coin never crosses, you cannot build a table where it almost never crosses for a long list of different shapes. For almost all shapes, the crossing is guaranteed to happen, even if it's incredibly slow.

How They Proved It: The "Time Travel" Trick

The math behind this is complex, but the strategy is clever. The authors used a technique called Analytic Continuation.

Think of the tunneling rate as a function that changes as you tweak the strength of the magnetic field or the size of the valleys.

The Problem: Directly calculating the tunneling rate for a strong magnetic field is like trying to walk through a foggy swamp; you can't see the path, and the math gets messy and breaks down.
The Solution: The authors imagined a "Time Travel" path. They started in a world where the math is easy and clear (a world with no magnetic field). They knew the particle definitely jumps there.
Then, they slowly "rotated" the problem into the complex mathematical world (the foggy swamp) where the magnetic field exists. They proved that the path from the "easy world" to the "magnetic world" is smooth and continuous.
Because the path is smooth, if the particle jumps in the "easy world," it must also jump in the "magnetic world," unless it hits a specific "wall" (a zero point).
They then proved that these "walls" are so rare (mathematically, they have "zero density") that for any random setup you pick, you will almost certainly not hit a wall. The particle will jump.

The "Mesoscopic Annuli" (The Onion Layers)

To make this "Time Travel" work, they had to deal with the fact that the magnetic field makes the math explode (go to infinity) if you look at the whole universe at once.

They solved this by peeling the problem like an onion. They broke the space around the valleys into many thin rings (annuli).

Inner Ring: Close to the valley, the math looks like a simple spring (a harmonic oscillator).
Outer Rings: Far away, the math looks like a free particle.
Middle Rings: They built a bridge between these two worlds using advanced tools called "pseudodifferential operators" (think of these as specialized lenses that let them focus on one ring at a time without the math breaking).

By stitching these rings together, they could prove that the "Time Travel" path works all the way from the easy world to the complex magnetic world.

Summary of the Main Result

The Phenomenon: Quantum tunneling in a double-well system with a strong magnetic field.
The Exception: There are specially crafted, rare shapes where tunneling stops completely (zero rate).
The Rule: For almost all other shapes (generic cases), the tunneling rate is strictly positive. It might be incredibly small (exponentially small), but it is never zero.
The Takeaway: You cannot rely on a strong magnetic field to permanently trap a particle in one well, unless you are extremely careful to build a very specific, unnatural trap. In the real world, with random or generic traps, the particle will always find a way to escape eventually.

The paper does not discuss medical applications, future technologies, or how to build better batteries. It is purely a mathematical proof about the fundamental behavior of particles in magnetic fields.

Technical Summary: Magnetic Double-Wells: Lower Bounds on Tunneling

Problem Statement
The paper investigates the phenomenon of quantum tunneling in double-well systems subjected to strong, constant magnetic fields. In the absence of a magnetic field, a particle localized in one of two deep potential wells has a non-zero probability of tunneling to the other, characterized by an exponentially small eigenvalue splitting ( $\Delta_0$ ) between the two lowest energy levels.

Previous work established that while radial single-well potentials in strong magnetic fields still exhibit a strictly positive lower bound on tunneling, specially constructed non-radial potentials can lead to the exact vanishing of tunneling (zero eigenvalue splitting). The central question addressed here is whether this vanishing phenomenon is generic or exceptional. Specifically, the authors aim to prove that for generic coupling constants $\lambda$ , tunneling does not vanish, providing rigorous lower bounds on the eigenvalue splitting and the associated hopping coefficient $\rho_0(\lambda)$ .

Methodology
The authors employ a complex analytic approach to establish lower bounds for the tunneling rate. The core strategy involves treating the coupling constant $\lambda$ as a complex variable and analyzing the analytic properties of the hopping coefficient $\rho_0(\lambda)$ and the eigenvalue splitting $\Delta_0(\lambda)$ .

Analytic Continuation via Cutoffs: A primary obstruction to analytic continuation is that the resolvent of the Landau Hamiltonian (the unperturbed part of the system) does not extend analytically off the real axis for complex magnetic fields. To circumvent this, the authors introduce a spatial cutoff function $\chi$ supported on a compact set containing the potential wells. They construct an analytic extension of the resolvent $(h_\lambda - z_\lambda)^{-1}$ acting only on functions supported within this compact set. This is achieved by:
- Defining a "cutoff Hamiltonian" $h^\theta_\lambda$ with a cut-off vector potential.
- Constructing a parametrix (approximate inverse) using a dyadic decomposition of the space into "mesoscopic annuli."
- Utilizing the theory of pseudodifferential operators to invert the operator locally in each annulus, leveraging the ellipticity of the symbol in these regions.
- Using a Neumann series to correct the approximate inverse to an exact inverse for the cutoff operator.
Riesz Projection and Ground State Approximation: The ground state $\phi_\lambda$ of the single-well Hamiltonian is represented via a Riesz projection of an approximate ground state (derived from the magnetic harmonic oscillator) onto the actual ground state. By ensuring the approximate state has compact support (via a cutoff), the authors ensure the analyticity of the projected state $\tilde{\phi}_\lambda$ in a wedge-shaped region of the complex plane $\Omega_{\Lambda, \epsilon}$ .
Complex Analytic Lemma: The authors apply a specialized complex analysis lemma (Lemma 1.2, derived from Blaschke factorization and Poisson integrals). This lemma states that if an analytic function $F(z)$ on a wedge satisfies an exponential upper bound and is non-zero at a single point (specifically at $b=0$ , where non-magnetic lower bounds are known), then $|F(t)|$ satisfies a lower bound of the form $\exp(-t^{1+\epsilon})$ for all real $t$ outside a set of density zero.
Extension to Double-Well: The analysis is extended from the single-well hopping coefficient to the double-well eigenvalue splitting $\Delta_0(\lambda)$ . This involves constructing a $2 \times 2$ matrix representation of the Hamiltonian restricted to the low-energy subspace, using analytically continued basis vectors derived from magnetic translations of the single-well ground states.

Key Contributions and Results
The main result is Theorem 1.1, which establishes that the vanishing of tunneling is a non-generic phenomenon. Specifically:

Lower Bounds: For a generic set of coupling constants $\lambda$ (specifically, outside a set of density zero), the hopping coefficient $\rho_0(\lambda)$ and the eigenvalue splitting $\Delta_0(\lambda)$ satisfy lower bounds of the form:
$|\rho_0(\lambda)|, |\Delta_0(\lambda)| \geq \exp(-\lambda^{1+\epsilon})$
for any arbitrarily small $\epsilon > 0$ , provided the magnetic field parameter $b$ is sufficiently small.
Sparsity of Vanishing Points: The set of $\lambda$ values where tunneling might vanish (or be exponentially smaller than the generic bound) is sparse. The authors prove that the sum of the negative powers of these exceptional points converges, implying they are isolated and do not accumulate in a way that would allow for a continuous interval of vanishing tunneling.
Averaged Bounds: The paper provides averaged lower bounds for $-\log|\rho_0|$ and $-\log|\Delta_0|$ over intervals $[\lambda, 2\lambda]$ , confirming that the tunneling rate does not vanish on average.

Significance and Claims
The paper claims to resolve the question of whether exact tunneling suppression in magnetic double-wells is a robust feature or a fine-tuned anomaly.

Contrast with Non-Magnetic Case: While non-magnetic tunneling is generally robust, the magnetic case allows for exact vanishing for specific, specially constructed potentials. This work demonstrates that such vanishing is not generic; for almost all coupling strengths, tunneling persists, albeit potentially at a rate that is exponentially small.
Complement to Previous Work: This result complements the authors' previous construction [FSW25b] of potentials where tunneling vanishes exactly. It clarifies that such potentials form a "measure zero" (in a density sense) subset of the parameter space.
Methodological Advance: The paper provides a rigorous framework for handling analytic continuation in the presence of strong magnetic fields, where the standard Landau resolvent fails to extend analytically. The use of spatial cutoffs and mesoscopic annuli to construct analytic resolvents is a key technical innovation.

The authors remain modest regarding the sharpness of the exponent $1+\epsilon$ , noting that it is an open question whether the bound can be improved to $\exp(-c\lambda)$ (i.e., taking $\epsilon \to 0$ ). They also note that extending these results to large magnetic fields $b$ or higher Landau levels remains an open direction.