Eigenvalue bounds for non-self-adjoint Schrödinger operators and pseudodifferential generalizations

This survey paper compiles existing results on spectral bounds for deterministic and random non-self-adjoint Schrödinger operators with complex potentials on Euclidean spaces and compact manifolds, while also presenting a new theorem extending these bounds to fractional Laplacians using $L^p$-norm estimates of the potentials.

The Gist

Imagine you are a physicist trying to predict how a tiny particle, like an electron, will move. Usually, we use a mathematical tool called the Schrödinger operator to do this. Think of this operator as a giant, complex machine that takes an input (the particle's current state) and spits out an output (how it will behave).

In the "old days" of physics, this machine was built to be perfectly balanced, or self-adjoint. This meant the machine was stable: if you put energy in, you got a predictable, real number out. It was like a well-tuned piano; every key produced a clear, real note.

The Problem: The Machine Gets "Unbalanced"

However, in the real world, things aren't always so neat. Sometimes, the environment around the particle is messy or "leaky" (like a radioactive atom decaying). To model this, physicists started using complex potentials. In math terms, this means the "settings" on our machine are no longer just real numbers; they include imaginary numbers.

When you add these complex settings, the machine loses its balance. It becomes non-self-adjoint.

• The Consequence: Instead of producing clear, real notes, the machine starts producing "ghost notes" (complex eigenvalues).
• The Danger: These ghost notes are unstable. A tiny change in the machine's settings can make the notes jump wildly to completely different places. It's like trying to balance a pencil on its tip; it's possible, but it's incredibly sensitive and hard to predict.

The Goal: Drawing a Safety Net

The main job of this paper is to act as a safety net. The author, Eduard Stefanescu, wants to answer a simple question: "If we know how messy the environment is (the potential), can we draw a circle around where these unstable 'ghost notes' might appear?"

He doesn't just want to say "it's unpredictable." He wants to say, "If the messiness is measured by $X$, then the ghost notes will definitely stay inside this specific circle."

The Journey of the Paper

1. The History Lesson (Sections 3 & 4)
The paper starts by looking back. In the past, mathematicians figured out how to draw these safety nets for the "balanced" machines (real potentials). They used clever tricks involving:

• The Birman-Schwinger Principle: A way of translating the problem of finding a ghost note into a different, easier problem (like translating a riddle into a math equation).
• Lieb-Thirring Inequalities: Rules that limit how many ghost notes can exist based on how "heavy" the messy environment is.

2. The New Challenge: The "Fractional" Machine (Section 6)
Most of these safety nets were built for standard machines (the classical Laplacian). But in modern physics, we sometimes need to model "fractional" behavior—where particles move in weird, non-standard ways (like jumping instead of walking smoothly). This is modeled by a Fractional Laplacian.

The paper's big new result is extending the safety net to these fractional machines, but specifically on compact manifolds.

• Analogy: Imagine the standard machine works on an infinite flat floor ($\mathbb{R}^d$). The new result works on a closed, finite surface, like the surface of a sphere or a donut (a compact manifold).
• The Result: Stefanescu proves that even on these curved, closed surfaces, if you know the "size" (the $L^p$ norm) of the messy environment, you can still draw a precise circle around where the unstable eigenvalues will hide.

3. Randomness vs. Determinism (Section 5)
The paper also discusses two types of messiness:

• Deterministic: The mess is fixed and known. The safety nets here are strict but sometimes leave large gaps.
• Random: The mess is generated by rolling dice (random variables). Surprisingly, the paper notes that if the mess is random, the safety nets can be much tighter! It's like if you shake a box of marbles, they tend to settle in a predictable pile, whereas if you arrange them by hand, they might be scattered everywhere.

The "How" (Section 7)

How did he do it? He didn't reinvent the wheel. He took the methods used by other mathematicians (Cuenin and Sogge) for the standard machines and tweaked them to work for the fractional ones.

• He used a special curve (a contour in the complex plane) to separate the "safe" zone from the "danger" zone.
• He proved that the "ghost notes" cannot escape a specific region defined by the size of the potential.

Summary

In simple terms, this paper is a survey and an extension.

1. Survey: It collects all the known rules for predicting where unstable quantum particles will go when the environment is messy.
2. Extension: It takes those rules, which previously only worked for standard machines on flat or curved surfaces, and proves they also work for fractional machines (weird, jumping particles) on closed surfaces (like spheres).

The paper provides a mathematical "fence" that guarantees these unstable particles won't wander off into the infinite unknown, as long as we know how "rough" the terrain they are walking on is.

Technical Summary

Technical Summary: Eigenvalue Bounds for Non-Self-Adjoint Schrödinger Operators and Pseudodifferential Generalizations

Problem Statement
The paper addresses the spectral analysis of Schrödinger operators $H := -\Delta + V$ (and their fractional generalizations) where the potential $V$ is complex-valued. Unlike the self-adjoint case, where the spectrum is real and variational principles apply, complex potentials lead to non-self-adjoint operators with complex spectra, spectral instability, and the breakdown of standard variational methods. The central problem is to establish quantitative bounds on the location of eigenvalues (spectral inclusion) in terms of the $L^p$-norms of the potential $V$. These bounds are sought for operators defined on both Euclidean space $\mathbb{R}^d$ and compact Riemannian manifolds.

Methodology and Analytic Framework
The paper employs a synthesis of operator-theoretic tools and harmonic analysis techniques:

1. Birman–Schwinger Principle: The core mechanism for reducing spectral information of $H$ to estimates on the associated Birman–Schwinger operator $K(\lambda) = |V|^{1/2}(H_0 - \lambda)^{-1}\text{sgn}(V)|V|^{1/2}$.
2. Schatten–von Neumann Trace Ideals: The analysis relies on proving that $K(\lambda)$ belongs to specific Schatten classes $S_p$, allowing for the control of discrete spectra and eigenvalue sums via singular value estimates.
3. Resolvent Estimates: A critical component involves deriving $L^p \to L^{p'}$ bounds for the resolvent of the unperturbed operator (the Laplacian or fractional Laplacian). The paper utilizes Sogge's spectral cluster bounds and Sobolev embeddings to control these resolvents, particularly near the spectrum of the free operator.
4. Complex Scaling and Contour Deformation: For the fractional case, the proof strategy involves defining a specific contour $\Gamma$ in the complex plane (related to the principal logarithm) and a region $\Xi$. Phragmén–Lindelöf arguments are used to extend resolvent estimates from the boundary of this region to its interior.
5. Randomization: For operators on $\mathbb{R}^d$, the paper reviews methods where randomizing the potential (Anderson-type models) allows for improved eigenvalue bounds, effectively doubling the short-range exponent compared to the deterministic case, at the cost of an exceptional set of measure zero.

Key Contributions and Results

• Survey of Deterministic and Random Bounds: The paper systematically reviews existing results for non-self-adjoint Schrödinger operators.

  • On $\mathbb{R}^d$, it details the sharpness of Frank's estimates for short-range potentials ($d/2 < q \leq (d+1)/2$) and discusses counterexamples (by Bögli and Cuenin) showing the failure of these bounds for $q > (d+1)/2$.
  • It summarizes randomization results (Cuenin–Merz) which improve deterministic bounds for random potentials.
  • On compact manifolds, it reviews Cuenin's [11] results for the classical Laplacian $-\Delta_g + V$, which provide spectral inclusion bounds in terms of $L^q(M)$ norms of $V$.

• New Theorem: Fractional Laplacians on Compact Manifolds:
  The primary novel contribution is Theorem 6.1, which extends spectral inclusion bounds to fractional Schrödinger operators of the form $(-\Delta_g)^{\alpha/2} + V$ on a closed Riemannian manifold $(M, g)$ of dimension $d \geq 2$.

  • Parameters: The result holds for $\frac{2d}{d+1} \leq \alpha \leq d$ and specific ranges of $q$ depending on $\alpha$ and $d$.
  • Result: The spectrum is contained within a union of discs centered at the eigenvalues of the fractional Laplacian, $\lambda_k^\alpha$, with radii proportional to $\|V\|_{L^q(M)}(1+\lambda_k)^{2\sigma(q)}$, and a region defined by a power-law decay in $|z|$.
  • Exponents: The function $\sigma(q)$ governing the decay rates is explicitly defined, distinguishing between regimes $d/\alpha < q \leq (d+1)/2$ and $(d+1)/2 \leq q \leq \infty$.
  • Proof Sketch: The proof adapts Cuenin's [11] methodology, utilizing the Birman–Schwinger principle and establishing resolvent estimates $\|((-\Delta_g)^{\alpha/2} - z)^{-1}\|_{L^p \to L^{p'}}$ that depend on the distance to the spectrum and the complex parameter $z$.

Significance and Claims
The paper positions itself primarily as a survey that consolidates the current state of the art regarding eigenvalue bounds for non-self-adjoint operators. Its specific contribution is the extension of Cuenin's compact manifold results to the fractional setting.

• Generalization: The author claims that the structural framework developed in prior works (Cuenin, Frank, Schimmer, etc.) allows for the extension of these spectral bounds from the classical Laplacian to fractional Laplacians and, potentially, to positive self-adjoint elliptic pseudodifferential operators of arbitrary positive order.
• Modesty: The paper explicitly states that the full proof of Theorem 6.1, including broader extensions and additional results, will be published in a forthcoming publication. The current text provides only a proof idea and the statement of the theorem.
• Context: The work is motivated by the need to understand spectral instability in open quantum systems (modeled by complex potentials) and the mathematical challenge of quantifying eigenvalue locations without the aid of self-adjointness.

The paper does not propose new experimental applications or future physical theories but serves to bridge gaps in the mathematical theory of spectral bounds, specifically adapting techniques from restriction theory and harmonic analysis to fractional operators on manifolds.