Extremal properties of the first eigenvalue and the fundamental gap of a sub-elliptic operator

Hongli Sun, Weijia Wu, Donghui YangarXiv 2023 · DOI · arXiv

This paper, recently published by researchers at Central South University, delves into the specialized world of sub-elliptic operators. While standard elliptic operators (like the Laplacian) are the bread and butter of physics and engineering, sub-elliptic operators are "degenerate"—they lose their strength at certain points in space.

The authors investigate how to "shape" a potential field $V$ to achieve two goals: pushing the ground-state energy as high as possible and making the energy gap between the first two states as small as possible.

TL;DR

The authors prove the existence and uniqueness of optimal potential functions that maximize the first eigenvalue and minimize the "fundamental gap" of a degenerate sub-elliptic operator under $L^p$ constraints. This work extends classical spectral optimization results from standard Schrödinger equations to more complex, degenerate systems.

What Problem Does This Paper Address?

In quantum mechanics, the eigenvalues of an operator correspond to energy levels. The first eigenvalue ($\lambda_1$) is the ground state, and the second ($\lambda_2$) is the first excited state. The difference between them, $\Gamma = \lambda_2 - \lambda_1$, is known as the fundamental gap.

Understanding this gap is vital because it dictates physical phenomena like the tunneling effect and the stability of quantum systems. Mathematically, the gap is linked to the "sharpness" of Poincaré inequalities and the speed of convergence in various numerical simulations.

While these properties are well-studied for the standard Laplacian operator ($-\Delta$), they are much harder to pin down for sub-elliptic operators. In this paper, the authors look at a specific degenerate equation: $$-\left(y^{2\alpha_1} \partial_x^2 u + x^{2\alpha_2} \partial_y^2 u\right) + Vu = \lambda u$$ Here, the "strength" of the derivatives depends on the coordinates $x$ and $y$. If $y=0$, the first term vanishes, making the operator degenerate. The challenge is to find a potential $V$ (within a fixed "budget" defined by an $L^p$ norm) that extremizes these spectral properties.

Methods

The authors employ a rigorous functional analysis approach to bridge the gap between standard elliptic theory and sub-elliptic degeneracy:

  1. Custom Sobolev Spaces: They define a specialized Hilbert space $H(\Omega)$ that incorporates the weights $y^{\alpha_1}$ and $x^{\alpha_2}$. They prove a compact embedding theorem, showing that even with degeneracy, the operator still behaves "nicely" enough to have a discrete set of eigenvalues.
  2. Variational Calculus: To find the optimal potential, they use the Lax-Milgram theorem to ensure weak solutions exist and then apply Gateaux derivatives to find the "shape" of the potential that maximizes the first eigenvalue.
  3. Perturbation Theory: To minimize the fundamental gap, the authors use first-order perturbation techniques. They analyze how a tiny "nudge" to the potential $V$ shifts the eigenvalues $\lambda_1$ and $\lambda_2$, allowing them to find the point where the gap is at its minimum.

Datasets

Not discussed in the paper. This is a theoretical mathematical study; no empirical datasets or external benchmarks were used.

Key Insights and Ideas

  • The "Shape" of the Best Potential: The authors found that the potential $V_0$ that maximizes the ground state $\lambda_1$ is not arbitrary. It is directly proportional to the first eigenfunction $u_0$ raised to a specific power determined by the $L^p$ constraint. Essentially, the potential "mirrors" the density of the ground state.
  • Non-Degeneracy of the Second State: A major hurdle in gap optimization is the possibility of "degenerate" eigenvalues (where multiple states share the same energy). The authors successfully prove that for the gap-minimizing potential, the second eigenvalue is non-degenerate (simple), which is a crucial requirement for the stability of the solution.
  • Support of the Potential: For the fundamental gap problem, the optimal potential $V^*$ doesn't necessarily cover the whole domain. The authors show that $V^*$ has specific support properties, and outside that support, the difference between the squares of the first two eigenfunctions ($|u_2|^2 - |u_1|^2$) must be non-negative.

Key Results

  • Existence and Uniqueness: The paper proves that there exists a unique potential $V_0 \in L^p$ that maximizes $\lambda_1(V)$.
  • Characterization of $V_0$: The maximizing potential is explicitly characterized as: $$V_0 = M \|u_0\|_{L^q}^{1-q} |u_0|^{2(q-1)}$$ where $q$ is the conjugate exponent of $p$.

  • Gap Minimization: For $p > 2$, the authors prove that the fundamental gap $\Gamma(V)$ attains a minimum within the allowed class of potentials.

  • The "Gap Condition": They establish that at the minimum, the potential satisfies $|u_2^*|^2 - |u_1^*|^2 = c(V^*)^{p-1}$ on its support (where $c$ is a negative constant).

Limitations and Future Work

  • Domain Specificity: The results are derived for a specific rectangular domain $\Omega = (0,1) \times (0,1)$. While the methods are general, the specific constants and embedding limits might change for more complex geometries.
  • Constraint Range: The minimization of the fundamental gap is specifically characterized for $p > 2$. The behavior for $1 < p \le 2$ remains an open question or requires different techniques.
  • Higher Eigenvalues: This study focuses only on the first two eigenvalues. Future research could explore the "higher gaps" ($\lambda_n - \lambda_{n-1}$) or the distribution of the entire spectrum in sub-elliptic settings.

Summary updated

Abstract

We consider the problems of extreming the first eigenvalue and the fundamental gap of a sub-elliptic operator with Dirichlet boundary condition, when the potential $V$ is subjected to a $p$-norm constraint. The existence results for weak solutions, compact embedding theorem and spectral theory for sub-elliptic equation are given. Moreover, we provide the specific characteristics of the corresponding optimal potential function.
BibTeX
@misc{sun2023extremal,
  author = {Hongli Sun and Weijia Wu and Donghui Yang},
  title = {{Extremal properties of the first eigenvalue and the fundamental gap of a sub-elliptic operator}},
  year = {2023},
  eprint = {2306.05825},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2306.05825},
}
Figure