🔢 mathematics

Counterexample to the second eigenfunction having one zero for a non-local Schrodinger operator

This paper demonstrates that the second eigenfunction of a perturbed fractional Laplace operator on a bounded interval can exhibit two sign changes, thereby providing a rigorous counterexample to the classical expectation that such eigenfunctions possess exactly one zero.

Original authors: Ben Andrews, Sophie Chen

Published 2026-08-18

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Ben Andrews, Sophie Chen

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 world of physics, many systems are described by equations that predict how energy moves and settles. For centuries, scientists have relied on a set of rules derived from classical physics to understand these systems. One of the most trusted rules concerns the "second eigenfunction," a mathematical description of the second simplest way a system can vibrate or hold energy. In the familiar world of classical physics, if you look at a simple, bounded space like a drumhead or a straight line, this second vibration always has exactly one point where it crosses from positive to negative, like a wave cresting and then dipping below the water line. This single crossing point, or "zero," is a fundamental expectation, a rule of thumb that has held true for standard equations for a very long time. However, the universe is not always limited to these standard rules. In recent decades, physicists and mathematicians have turned their attention to "non-local" systems, where the behavior at one point depends on conditions far away, not just its immediate neighbors. These systems are described by fractional operators, which act like a blurred version of the standard equations, allowing for long-range interactions. The question that has lingered is whether the old, reliable rule about the single crossing point still holds when these long-range effects are introduced.

A team of researchers has now provided a definitive answer: the old rule does not always hold. In a new study, they demonstrated that for a specific type of non-local system, the second vibration can have two crossing points instead of one. This discovery challenges the classical intuition that has guided mathematical physics for generations. The researchers focused on a simplified model: a line segment with a potential energy landscape that varies along its length. In their setup, they created a scenario where the energy landscape had three distinct regions, with the middle region having a significantly higher energy value than the two regions on either side. By carefully analyzing this configuration using advanced perturbation theory—a method that studies how a system changes when a small disturbance is applied—they found that the second vibration mode did not behave as expected. Instead of a single smooth transition from positive to negative, the vibration pattern flipped signs twice, creating a shape with two distinct crossings.

To reach this conclusion, the team first constructed a theoretical model using an "infinite potential well," a concept where the energy outside specific small regions is so high that the system is forced to exist only within those regions. They treated this as a matrix problem, a simplified mathematical representation where the interactions between the three regions could be calculated directly. Their analysis showed that when the middle region was much more energetic than the sides, the mathematical solution for the second vibration required two sign changes. They then used a rigorous mathematical argument to show that this result was not just an artifact of the simplified infinite model. By gradually lowering the energy barrier from infinity to a very high but finite value, they proved that the behavior persisted. As the barrier became finite, the solutions to the equations smoothly transitioned, carrying the two-crossing pattern with them. This meant that even in a physically realizable system with finite energy barriers, the second eigenfunction could exhibit this unexpected double-crossing behavior.

The researchers were careful to specify the conditions under which this occurs. Their detailed proof focused on a specific case where the non-local interaction parameter was one-half, a value associated with a process known as the Cauchy process. While they strongly suspect that similar phenomena occur for other rational values of this parameter, their rigorous proof is currently limited to this specific case. They also noted that the counterexample relies on the potential energy being non-convex, meaning the middle region is a peak rather than a valley. It remains an open question whether a smooth, convex potential would preserve the traditional single-crossing rule. Nevertheless, the finding is significant because it provides one of the first rigorous insights into the qualitative behavior of eigenfunctions for perturbed non-local Schrödinger operators. It shows that the intuitive geometric properties we expect from classical systems can break down when long-range interactions are introduced, revealing a richer and more complex landscape of possibilities in the mathematics of quantum and non-local physics.

Technical Summary: Counterexample to the Second Eigenfunction Having One Zero for a Non-local Schrödinger Operator

Problem Statement
The paper investigates the qualitative behavior of eigenfunctions for perturbed non-local Schrödinger operators, specifically focusing on the fractional Laplacian $(-\Delta)^s$ with Dirichlet boundary conditions on a bounded interval $\Omega = (-1, 1)$ . The central problem addresses a long-standing expectation in spectral theory: that the second eigenfunction of such operators should possess exactly one zero (a single sign change), analogous to the classical Laplace operator ( $s \to 1^-$ ) and the unperturbed fractional Laplacian for $s=1/2$ (the Cauchy process). The authors aim to construct a rigorous counterexample demonstrating that, under specific bounded potential perturbations, the second eigenfunction can exhibit two sign changes, thereby violating the classical nodal property.

Methodology
The authors employ a three-step construction strategy, combining finite-dimensional matrix analysis with infinite-dimensional functional analytic techniques:

Matrix Approximation (Step 1): The authors first analyze an extreme case involving an "infinite potential well" where the potential is finite only on $k$ mutually disjoint, infinitesimally small sub-intervals $I_i$ and infinite elsewhere. By approximating the solution space with weighted sums of indicator functions on these sub-intervals, the infinite-dimensional eigenvalue problem is reduced to a $k \times k$ symmetric matrix eigenvalue problem. Using direct algebraic analysis and the Perron-Frobenius theorem, they analyze the eigenvectors of this matrix. They demonstrate that if the potential values $V_i$ on the sub-intervals are non-convex (specifically, $V_2 \gg V_1, V_3$ for $k=3$ ), the second eigenvector of the matrix exhibits two sign changes.
Infinite Potential Well Counterexample (Step 2): To rigorously connect the matrix result to the operator problem, the authors treat the infinite potential well problem as a regular perturbation of a limiting operator. They utilize a rescaling technique to map the varying Hilbert spaces $L^2(\cup I_i)$ to a fixed Hilbert space $H = \prod L^2(I)$ . By introducing a parameter $\beta = \epsilon^{1/q}$ (where $s=p/q$ is rational), they establish that the rescaled operator family is an analytic family of type (A) in the sense of Kato. This allows the application of the Kato–Rellich theorem for degenerate eigenvalues, proving that the spectral properties (including the sign patterns of eigenvectors) of the finite-dimensional matrix persist in the infinite-dimensional operator limit for sufficiently small perturbations.
Finite Potential Well Counterexample (Step 3): Finally, the authors extend the result from the idealized infinite potential well to a physically realizable finite potential well. They define a sequence of finite potential wells that converge pointwise to the infinite potential as the well depth increases (parameterized by $\delta \to 0$ ). Using a compactness argument and energy minimization principles within the fractional Sobolev space $H^s_0(\Omega)$ , they prove that the eigenfunctions of the finite potential well converge in $L^2$ to the eigenfunctions of the infinite potential well. Since the limit eigenfunction has two sign changes, the eigenfunctions of the sufficiently deep finite potential well must also exhibit two sign changes for small enough $\delta$ .

Key Contributions and Results

Existence of Counterexamples: The paper provides the first rigorous proof that the second eigenfunction of a perturbed fractional Laplacian on a bounded interval can have two sign changes. This contradicts the classical intuition derived from the local Laplacian and the unperturbed fractional case.
Role of Potential Geometry: The counterexample relies on a bounded, non-convex potential (specifically a triple-well configuration where the middle well is significantly deeper than the outer wells). The authors show that the non-convexity of the potential values drives the splitting of the degenerate ground state in a manner that forces the second eigenfunction to adopt a $(-, +, -)$ or $(+, -, +)$ sign pattern across the wells.
Analytic Perturbation Framework: The authors successfully adapt Kato–Rellich perturbation theory to the non-local setting. A significant technical contribution is the handling of the singular spectral scale arising from shrinking intervals and the construction of an analytic family of operators for rational $s \in (0, 1)$ .
Convergence Results: The paper establishes the $L^2$ convergence of eigenfunctions from finite to infinite potential wells, ensuring that the qualitative nodal properties of the idealized model are preserved in the finite case.

Significance and Claims
The paper claims to provide one of the first rigorous insights into the qualitative behavior of eigenfunctions for perturbed non-local Schrödinger operators. The authors emphasize that their result challenges the assumption that nodal properties of unperturbed fractional eigenfunctions are robust under bounded perturbations. While the detailed analysis focuses on the case $s=1/2$ (the Cauchy process), the methodology suggests that similar phenomena occur for other rational values of $s$ in $(0, 1)$ . The authors modestly note that extending these results to irrational $s$ remains an open question due to the analyticity requirements of the Kato–Rellich theory, and whether similar counterexamples exist for convex potentials is also left as an open problem. The work fundamentally alters the understanding of spectral nodal properties in non-local settings, showing they are not universally preserved under perturbation.

More like this