Counterexamples to the $L^1$ and $L^{\infty}$ boundedness of the one-dimensional wave operators

This paper provides rigorous counterexamples demonstrating that the wave operators for one-dimensional Schrödinger operators with bounded, compactly supported potentials are unbounded on $L^1(\mathbb{R})$ and $L^{\infty}(\mathbb{R})$ in generic cases and specific exceptional cases, thereby completing the characterization of their $L^p$ boundedness.

The Gist

Imagine you have a vast, empty highway (this represents the "free" world where particles move without any obstacles). Now, imagine you place a few speed bumps or potholes on this road (this represents the potential, or the force field $V$). In the world of quantum mechanics, particles traveling on this road are described by a mathematical object called a wave function.

When a particle encounters these speed bumps, its wave changes shape. The Wave Operator is the mathematical machine that translates the particle's behavior on the empty highway into its behavior on the bumpy road.

For a long time, mathematicians knew exactly how this machine behaved for most types of traffic (mathematically, for most "sizes" of waves, known as $L^p$ spaces where $1 < p < \infty$). They knew the machine worked smoothly and didn't break the waves apart.

However, there were two tricky "endpoints" of traffic:

1. The "L1" case: Think of this as a single, very sharp, concentrated spike of traffic.
2. The "L-infinity" case: Think of this as a massive, flat wall of traffic stretching forever.

For decades, mathematicians suspected that the Wave Operator might break down (become "unbounded") when dealing with these two extreme types of traffic, especially when the road has a specific kind of "resonance" (a special vibration that gets stuck at zero energy). They suspected this because the math involved a tool called the Hilbert Transform, which is known to be very bad at handling these sharp spikes and flat walls. But, despite strong suspicions, no one had actually built a concrete example to prove it would break.

What this paper does:
Sisi Huang and Xiaohua Yao decided to build those concrete examples. They didn't just guess; they constructed specific, simple road maps (potentials) that are bounded and finite (like a short stretch of potholes) and showed exactly how the Wave Operator fails.

Here is the breakdown of their findings using simple analogies:

1. The "Generic" Road (The Normal Case)

Imagine a road where the speed bumps are just normal bumps. There is no special resonance.

• The Old Belief: People thought the Wave Operator might still work fine on the extreme traffic (L1 and L-infinity).
• The New Discovery: The authors proved that even on this "normal" road, if you send in a sharp spike of traffic (L1) or a flat wall of traffic (L-infinity), the Wave Operator breaks. It cannot handle the transformation. The output becomes infinitely large or messy.
• The Analogy: It's like trying to use a standard traffic camera to count a single, infinitely sharp needle or a wall of infinite width. The camera lens distorts the image so badly that the count becomes infinite.

2. The "Exceptional" Road (The Resonant Case)

Imagine a road where the speed bumps are arranged in a way that creates a "standing wave" or a trapped vibration at zero energy (a resonance).

• The Special Exception: There was one specific scenario where mathematicians knew the machine did work: if the road's vibration matched a very specific pattern (mathematically, if a certain limit equals 1).
• The New Discovery: The authors showed that if the road has a resonance but doesn't match that specific pattern (the limit is not 1), the machine breaks again.
• The Extra Twist: In this broken state, the machine doesn't just fail to keep the traffic finite; it fails to keep the "average chaos" of the traffic under control. In math terms, it maps a flat wall of traffic (L-infinity) into a space called BMO (Bounded Mean Oscillation), but it fails even there. It's like the machine not only blurs the image but makes the static noise so loud it becomes unmanageable.

The "Why" (The Hidden Culprit)

The paper explains that the reason for this breakdown is the Hilbert Transform.

• Think of the Wave Operator as a recipe. Most of the recipe is safe and smooth.
• However, the "low energy" part of the recipe (the part dealing with slow-moving particles) secretly includes the Hilbert Transform.
• The Hilbert Transform is like a blender that works great on smoothies but turns a single ice cube (L1) or a giant block of ice (L-infinity) into a chaotic, infinite mess.
• The authors proved that in the cases they studied, this "blender" is actually turned on and active, and there is no cancellation to save the day.

The Conclusion

This paper completes the puzzle.

• Before: We knew the Wave Operator worked for almost all traffic, and we knew it worked for extreme traffic only in one very specific resonant case.
• Now: We know that for every other case (the normal road, and the resonant road that doesn't match the specific pattern), the Wave Operator fails on the extreme traffic (L1 and L-infinity).

The authors didn't just say "it probably fails"; they built the specific road maps and showed the traffic jams happening, proving that the machine is indeed unbounded in these scenarios. This settles a long-standing question in the mathematical study of quantum waves.

Technical Summary

Technical Summary: Counterexamples to the $L^1$ and $L^\infty$ Boundedness of the One-Dimensional Wave Operators

Problem Statement
The paper addresses the $L^p$ boundedness of the wave operators $W_\pm(H, -\Delta)$ associated with the one-dimensional Schrödinger operator $H = -\Delta + V(x)$, where $V$ is a real-valued potential decaying as $|V(x)| \lesssim \langle x \rangle^{-\beta}$ for $\beta > 2$. While it is well-established that these operators are bounded on $L^p(\mathbb{R})$ for all $1 < p < \infty$ in both generic and exceptional cases, the behavior at the endpoints $p=1$ and $p=\infty$ has remained partially unresolved.

Previous work by Weder established that in the "exceptional case" (where zero is a resonance) specifically satisfying the condition $\lim_{x\to-\infty} f_+(0, x) = 1$ (where $f_+$ is the Jost function), the wave operators are bounded on $L^1$ and $L^\infty$. This boundedness was attributed to the vanishing of coefficients associated with the Hilbert transform in the low-energy decomposition of the wave operators. However, for the remaining cases—the generic case and the exceptional case where $\lim_{x\to-\infty} f_+(0, x) \neq 1$—it was long expected that the operators would be unbounded due to the presence of Hilbert transform terms. Despite this expectation, a rigorous proof demonstrating this unboundedness for bounded, compactly supported potentials was missing.

Methodology
The authors adopt a stationary representation approach to analyze the wave operators, decomposing them into high-energy ($W^H_-$) and low-energy ($W^L_-$) parts. Since the high-energy part is known to be bounded on all $L^p$ spaces, the analysis focuses exclusively on the low-energy part $W^L_-$.

1. Asymptotic Expansion: The core of the methodology involves deriving a precise asymptotic expansion of the perturbed resolvent $R^+_V(\lambda^2)$ near zero energy ($\lambda = 0$). The authors utilize the operator identity $R^+_V(\lambda^2)V = R^+_0(\lambda^2)v M^{-1}(\lambda)v$, where $M(\lambda) = U + vR^+_0(\lambda^2)v$. They compute the asymptotic expansion of the inverse operator $M^{-1}(\lambda)$ in powers of $\lambda$, distinguishing between the generic case ($W(0) \neq 0$) and the exceptional case ($W(0) = 0$).
2. Kernel Analysis: By substituting these expansions into the integral representation of $W^L_-$, the authors isolate singular operators. They analyze the integral kernels of these operators, classifying them by their order of singularity with respect to $\lambda$ as $\lambda \to 0^+$.
3. Boundedness Proofs: Using Schur tests and properties of specific integral kernels (related to the Hilbert transform and BMO spaces), the authors prove that most components of the low-energy operator are bounded on $L^p$ for $1 < p < \infty$ and map $L^\infty$ to BMO (Bounded Mean Oscillation) or $H^1$ (Hardy space) to $L^1$.
4. Counterexample Construction: To prove unboundedness at the endpoints, the authors construct specific test functions. For the generic case, they use characteristic functions of intervals. For the exceptional case, they utilize functions involving signs and characteristic functions of annuli. They demonstrate that the action of the singular operators on these test functions results in functions that fail to belong to $L^1$ or $L^\infty$ (specifically, their norms diverge as parameters tend to infinity).

Key Contributions and Results
The paper provides rigorous proofs for the unboundedness of wave operators in the previously open cases, completing the picture of $L^p$ boundedness for one-dimensional Schrödinger operators.

• Generic Case: For potentials of generic type with $\beta > 7$, the wave operators $W_\pm$ are unbounded on $L^1(\mathbb{R})$ and $L^\infty(\mathbb{R})$. However, they remain bounded from $L^\infty(\mathbb{R})$ to BMO$(\mathbb{R})$.
• Exceptional Case ($\lim_{x\to-\infty} f_+(0, x) \neq 1$): For potentials of exceptional type with compact support satisfying the condition that the Jost function limit is not 1, the wave operators are unbounded on $L^1(\mathbb{R})$ and $L^\infty(\mathbb{R})$. Furthermore, in this specific case, they are not bounded from $L^\infty(\mathbb{R})$ to BMO$(\mathbb{R})$.
• Distinction in Endpoint Behavior: The paper highlights a crucial distinction between the generic and exceptional cases regarding the $L^\infty \to \text{BMO}$ mapping. While the generic case preserves this boundedness, the exceptional case (with the non-unit Jost limit) destroys it. This is linked to the non-vanishing of a specific coefficient $c_2$ related to the resonance structure.

Significance
The paper claims to complete the theoretical understanding of the $L^p$ boundedness of one-dimensional wave operators. By providing explicit counterexamples and rigorous proofs for the unboundedness in the generic and specific exceptional cases, the authors resolve a long-standing gap in the literature. Their work clarifies that the presence of the Hilbert transform in the low-energy decomposition indeed leads to unboundedness on $L^1$ and $L^\infty$ unless specific cancellation conditions (as found in Weder's result) are met. Additionally, the paper refines the understanding of the mapping properties into BMO spaces, showing that the exceptional case with a non-unit Jost limit fails to map $L^\infty$ to BMO, unlike the generic case. These results rely on the construction of bounded, compactly supported potentials, demonstrating that the unboundedness is a fundamental feature of the operator structure rather than an artifact of long-range potential decay.