🔢 mathematics

An isoperimetric inequality for Neumann eigenvalues with radial log-concave measures

This paper establishes a sharp isoperimetric inequality for the harmonic mean of the first $n$ nonzero Neumann eigenvalues of the Witten-Laplacian on origin-symmetric Lipschitz domains in space forms endowed with general radial log-concave measures, extending previous results by removing the requirement that the weight function be non-increasing.

Original authors: Kui Wang, Xiaomei Sun, Anqiang Zhu

Published 2026-08-11

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Kui Wang, Xiaomei Sun, Anqiang Zhu

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

Imagine you are a master architect trying to build the most efficient shape possible. In the world of mathematics, this isn't just about making a pretty house; it's about how things vibrate, heat up, or flow when they are trapped inside a container. This field is called spectral geometry, and it asks a simple but deep question: If you have a fixed amount of "space" (volume), which shape makes a drum vibrate at the lowest possible pitch? Or, conversely, which shape makes it hardest for heat to escape? For centuries, mathematicians have known that for a simple, empty room in flat space, the perfect sphere (or ball) is the champion. It is the most efficient shape for these physical properties.

But the real world isn't always flat, and space isn't always empty. Sometimes, the "floor" of our room is curved like the surface of a sphere or a saddle, and sometimes the air inside is thicker in some places than others. Mathematicians call these variations "weighted measures" or "log-concave measures." Think of it like a room where the air gets denser in a specific, smooth pattern as you move away from the center, or a room built on a curved hill. The big challenge has been figuring out if the ball is still the best shape when the rules of the room change. Previous attempts to solve this required the "air" to get thinner or stay the same as you moved outward, which ruled out many interesting scenarios, like the famous Gaussian distribution (the bell curve) where the density actually increases near the center before dropping off.

This paper by Sun, Wang, and Zhu steps into that gap. They prove a sharp, mathematical rule for a specific type of vibrating system (the Neumann eigenvalues of the Witten-Laplacian) inside these curved, weighted rooms. Their main finding is that even when the "air" density follows a complex pattern—specifically, when the underlying potential function is convex, allowing the density to increase and then decrease (like a bell curve)—the ball is still the unique champion that minimizes the sum of the reciprocals of the first few vibration frequencies. In simpler terms, if you want to minimize the "sluggishness" of the first $n$ vibrations in a space with a radial, log-concave weight, you must choose a ball. They didn't just guess this; they provided a rigorous, step-by-step proof that holds for a wide range of curved spaces (spheres, flat planes, and hyperbolic spaces) and weight functions, removing the old restriction that the weight had to be constantly decreasing. They showed that the ball remains the optimal shape even when the conditions are much more general than previously thought possible.

The Story of the Perfect Ball

Imagine you are holding a drum. When you hit it, it doesn't just make one sound; it makes a whole chord of notes, called eigenvalues. The first note is the deepest, the second is a bit higher, and so on. Mathematicians love to add up the "reciprocals" of these notes (which is like adding up how long each note lasts). The famous Szegő-Weinberger inequality is a rule that says: "If you have a fixed amount of rubber for your drum skin, the roundest drum (the ball) will give you the lowest possible sum of these notes."

For a long time, this rule only worked in perfectly flat, empty rooms. Then, scientists started asking: "What if the room is curved? What if the air inside is heavy in some spots and light in others?" This is where the paper comes in. The authors are dealing with a "Witten-Laplacian," which is just a fancy name for a vibration operator that includes a "drift" or a push caused by the weight of the air (the measure).

The authors tackle a specific type of weight called "radial log-concave." Imagine a fog that is centered around a point. "Log-concave" is a mathematical way of saying the fog is shaped like a smooth hill or a bowl—it doesn't have weird bumps or holes. "Radial" means the fog looks the same in every direction from the center. The tricky part is that previous rules said this fog had to get thinner and thinner as you moved away from the center. But in the real world (and in the Gaussian distribution used in statistics), the fog might get thicker first before it gets thinner. The old rules couldn't handle that.

Sun, Wang, and Zhu say, "Hold on, we can fix that." They prove that even if the density increases and then decreases, as long as the underlying potential function governing the weight is convex (a specific mathematical smoothness condition), the ball is still the winner.

How They Did It: The Detective Work

To prove this, the authors had to be like detectives solving a mystery about the shape of the drum's vibration.

The First Clue (The Eigenfunction): They started by looking at the very first vibration mode of a perfect ball in this weird, weighted space. They needed to know exactly what the vibration looked like. They proved that the vibration isn't just a simple bump; it has a specific structure where it grows from the center out to the edge, and it does so in a very predictable, monotonic way. It's like a wave that only goes up, never down, as it travels from the center to the rim.
The Second Clue (The Ratio): They then looked at the ratio of the vibration's height to the distance from the center. They discovered a hidden property: this ratio behaves in a very strict, decreasing manner. This was the key to unlocking the door. It meant that the vibration is "tightest" in the center and spreads out in a controlled way.
The Final Puzzle Piece (The Matrix Trick): Here is where it gets clever. They used a tool from linear algebra called a "matrix trace inequality." Imagine you have a bunch of numbers arranged in a grid (a matrix) representing the energy of the drum. They showed that no matter how you distort your drum (as long as it has the same volume and symmetry), the "energy grid" of your weird shape will always be "worse" (larger) than the energy grid of the perfect ball.

They used a "bathtub principle" (a fancy way of saying "fill the bottom first") to argue that if you take any weird shape and try to rearrange its mass to look more like a ball, you will always lower the sum of the reciprocals of the frequencies. Since the ball is the most efficient arrangement, any other shape must be less efficient.

The Big Reveal

The paper concludes with a definitive statement: If you have a domain (a shape) in a curved space with a radial log-concave weight, and you want to minimize the sum of the reciprocals of the first $n$ nonzero Neumann eigenvalues, you must choose a ball. There is no other shape that can do better. If you find a shape that does just as well, it must be a ball (up to a set of measure zero, which in plain English means it's the same shape, just maybe missing a few dust particles here and there).

This result is a big deal because it unifies many previous discoveries. It recovers the results for flat space, hyperbolic space, and the Gaussian space (the bell curve) as special cases. It also improves upon recent work that required the weight to be non-increasing. By removing that restriction, the authors have shown that the ball's dominance is much more robust than we thought. It's not just the champion in simple, boring rooms; it's the champion even in complex, curvy, and dynamically weighted environments.

So, the next time you see a bell curve or a sphere in a physics problem, remember: according to this paper, that sphere is the most efficient shape nature can possibly make, even when the rules of the game get a little complicated. The ball wins, every time.

Technical Summary: An Isoperimetric Inequality for Neumann Eigenvalues with Radial Log-Concave Measures

Problem Statement
This paper addresses the isoperimetric problem for the harmonic mean of the first $n$ nonzero Neumann eigenvalues of the Witten-Laplacian ( $\Delta_\phi = \Delta - \nabla \phi \cdot \nabla$ ) on origin-symmetric Lipschitz domains within space forms ( $M_\kappa$ , where $\kappa \in \{1, 0, -1\}$ ). The domains are endowed with a radial log-concave measure $d\gamma_\kappa = e^{-\phi(r)} d\text{Vol}_\kappa$ .

The central question is whether the geodesic ball $B_R$ of the same weighted volume uniquely minimizes the sum of the reciprocals of the first $n$ nonzero Neumann eigenvalues:
$\sum_{i=1}^n \frac{1}{\mu_i(\Omega)} \geq \frac{n}{\mu_1(B_R)}.$
This inequality implies the classical Szegő-Weinberger inequality ( $\mu_1(\Omega) \leq \mu_1(B_R)$ ) as a corollary.

Context and Limitations of Previous Work
The authors situate their work within a lineage of spectral geometry results:

Euclidean Space: The classical Szegő-Weinberger inequality establishes the ball as the maximizer of $\mu_1$ . Ashbaugh and Benguria conjectured the reciprocal sum inequality (1.1), which was proven for specific cases by Xia and Wang [21] and He, Li, and Tang [12].
Gaussian Space: Chiacchio and Di Blasio [8] extended the Szegő-Weinberger inequality to the Gaussian measure. Gao and Wang [11] and Chiacchio [9] subsequently established the sharp harmonic mean inequalities for the first $n-1$ and $n$ eigenvalues, respectively, in the Gaussian setting.
General Weighted Spaces: Chen and Mao [7] extended these results to space forms with log-concave measures. However, their approach required the weight potential $\phi(r)$ to be both convex and non-increasing. This restriction excludes the Gaussian measure (where $\phi(r) = r^2/2$ is increasing) and limits the applicability of the result.

Methodology
The proof relies on a refined analysis of the eigenfunctions on geodesic balls and a matrix trace inequality. The methodology proceeds in three main stages:

Spectral Analysis on Geodesic Balls:
The authors characterize the first nonzero eigenvalue $\mu_1(B_R)$ and its associated eigenfunctions. They prove that under the assumption that $\phi$ is convex, the first nonzero eigenvalue has multiplicity $n$ . The corresponding eigenfunctions take the form $u_i(x) = T_1(r)\psi_i(\theta)$ , where $\psi_i$ are linear coordinate functions restricted to the sphere, and $T_1(r)$ is the first eigenfunction of a specific radial Sturm-Liouville problem.
- Key Technical Lemma: They establish that $T_1(r)$ is strictly increasing and that the ratio $T_1(r)/S_\kappa(r)$ is monotonically decreasing, provided the condition $\left(\frac{C_\kappa(r)}{S_\kappa(r)}\phi'(r)\right)' \geq 0$ holds. This condition replaces the stricter "non-increasing $\phi$ " requirement of Chen-Mao [7].
Variational Characterization via Trial Functions:
To bound the sum of reciprocals, the authors employ Hersch's variational principle (Lemma 2.2). They construct $n$ trial functions $v_i(x) = G(r)\psi_i(\theta)$ on the domain $\Omega$ , where $G(r)$ is a truncated version of the radial eigenfunction $T_1(r)$ .
- They define Gram matrices $J$ (related to the $L^2$ inner product) and $K$ (related to the Dirichlet energy) using these trial functions.
- The inequality $\sum \mu_i^{-1} \geq \text{tr}(K^{-1}J)$ is derived.
Matrix Trace Inequality and Optimization:
The core of the proof involves estimating the matrices $J$ and $K$ using the "bathtub principle" (Lemma 2.3) applied to radial slices of the domain.
- Using the monotonicity properties of $T_1(r)$ and $T_1(r)/S_\kappa(r)$ , they show that the integrals defining $J$ and $K$ are bounded by expressions involving the volume difference between $\Omega$ and the ball $B_R$ .
- These bounds allow the application of a specific matrix trace inequality (Lemma 2.1 from He, Li, and Tang [12]): If $J \succeq aI + cZ$ and $0 \prec K \preceq \lambda aI - dZ$ with $\text{tr}(Z)=0$ , then $\text{tr}(K^{-1}J) \geq n/\lambda$ .
- By identifying the parameters $a, c, d, \lambda$ with spectral quantities of the ball, they derive the desired lower bound $n/\mu_1(B_R)$ .

Key Contributions and Results
The main result is Theorem 1.1, which states:
Let $\Omega \in E_\kappa$ be an origin-symmetric Lipschitz domain with weighted volume equal to that of a geodesic ball $B_R$ . If the radial weight $\phi(r)$ is convex and satisfies $\left(\frac{C_\kappa(r)}{S_\kappa(r)}\phi'(r)\right)' \geq 0$ , then:
$\sum_{i=1}^n \frac{1}{\mu_i(\Omega)} \geq \frac{n}{\mu_1(B_R)}.$
Equality holds if and only if $\Omega$ coincides with the ball $B_R$ .

Significance and Novelty
The paper claims its primary novelty lies in removing the non-increasing condition on the weight function $\phi(r)$ , which was a necessary constraint in the recent work of Chen and Mao [7].

Generalization: The new condition $\left(\frac{C_\kappa}{S_\kappa}\phi'\right)' \geq 0$ is satisfied by a broader class of convex potentials, including the Gaussian measure ( $\phi(r) = r^2/2$ ) and potentials of the form $\phi(r) = r^{2k}$ for $k \geq 1$ .
Unification: The result recovers the Gaussian space inequalities of Chiacchio and Di Blasio [8], Chiacchio [9], and Gao-Wang [11] as special cases, while simultaneously improving upon the Chen-Mao [7] results by covering cases they could not.
Methodological Shift: The authors note that their approach to characterizing the first eigenfunction on balls (Section 3) differs from Chiacchio and Di Blasio [8]. It is described as relatively straightforward and applicable to a broad class of convex radial weights without relying on the specific structure of the Gaussian density or the monotonicity of $\phi$ .

Conclusion
The paper provides a unified geometric framework for isoperimetric inequalities of Neumann eigenvalues on space forms with general radial log-concave measures. By establishing the sharp harmonic mean inequality under a relaxed convexity condition, it extends the scope of spectral isoperimetric theory beyond the restrictive settings of previous literature.

The Story of the Perfect Ball

How They Did It: The Detective Work

The Big Reveal

Technical Summary: An Isoperimetric Inequality for Neumann Eigenvalues with Radial Log-Concave Measures

More like this