Rigidity and flexibility under spectral Ricci lower bounds and mean-convex boundary

Gioacchino Antonelli, Yangyang Li, Paul Sweeney Jr · 2026

Read Rigidity and flexibility under spectral Ricci lower bounds and mean-convex boundary by Gioacchino Antonelli, Yangyang Li, Paul Sweeney Jr (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-11384.txt.

Introduction

Gioacchino Antonelli, Yangyang Li and Paul Sweeney study manifolds with mean-convex boundary and spectrally nonnegative Ricci curvature. Their first result extends Kasue's rigidity theorem in sharp form. Under a spectral inequality and a nonnegative edge mean bend, a complete manifold with a split edge and a compact part must split as a product of an interval with a surface. Their second result concerns the relative fundamental group. In dimensions other than four, suitable spaces admit a metric of positive sectional curvature and a strictly mean-convex edge.

Catalog record

CREATORGioacchino Antonelli, Yangyang Li, Paul Sweeney Jr
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE101147
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.11384

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Rigidity and flexibility under spectral Ricci lower bounds and mean-convex boundary Gioacchino Antonelli, Yangyang Li, Paul Sweeney Jr Abstract . We study Riemannian manifolds ( M n , g ) (M^{n},g) with mean-convex boundary whose Ricci curvature is nonnegative in a spectral sense. Our first main result is a sharp spectral extension of a rigidity theorem by Kasue: we prove that under the conditions λ 1 ​ ( − γ ​ Δ + Ric ) ⩾ 0 , H ∂ M ⩾ 0 , \lambda_{1}(-\gamma\Delta+\mathrm{Ric})\geqslant 0,\qquad H_{\partial M}\geqslant 0, and in the sharp range 0 ⩽ γ < 4 0\leqslant\gamma<4 if n = 2 n=2 , and 0 ⩽ γ < n − 1 n − 2 0\leqslant\gamma<\frac{n-1}{n-2} if n ⩾ 3 n\geqslant 3 , a (possibly noncompact) complete manifold with disconnected boundary, with at least one compact boundary component, must split isometrically as a product [ 0 , L ] × Σ [0,L]\times\Sigma . Our second main contribution is a topological rigidity result for the relative fundamental group π 1 ​ ( M , ∂ M ) \pi_{1}(M,\partial M) , combined with a deep theorem of Lawson–Michelsohn. We prove that, in dimensions n ≠ 4 n\neq 4 , any compact manifold with boundary satisfying the two inequalities above, with at least one of them strict, admits a metric with positive sectional curvature and strictly mean-convex boundary, provided γ ⩾ 0 \gamma\geqslant 0 if n = 2 n=2 , and 0 ⩽ γ ⩽ n − 1 n − 2 0\leqslant\gamma\leqslant\frac{n-1}{n-2} if n ⩾ 3 n\geqslant 3 . This range of γ \gamma is sharp for the latter result to hold. 1. Introduction Let n\geqslant 2 , and let (M^{n},g) be a smooth complete manifold, possibly with boundary. In this paper, all the manifolds are assumed to be connected. We define the function \mathrm{Ric}:M\to\mathbb{R} by \mathrm{Ric}(x):=\inf_{\begin{subarray}{c}v\in T_{x}M\\ g(v,v)=1\end{subarray}}\mathrm{Ric}_{x}(v,v). Note that \Ric\in\mathrm{Lip}_{\text{loc}}(M) . For a constant \gamma\geqslant 0 , we say that M satisfies \lambda_{1}(-\gamma\Delta+\Ric)\geqslant 0 if either of the two following equivalent conditions holds: (i) for all \varphi\in C^{1}_{c}(M) it holds (1) \int_{M}\gamma|\nabla\varphi|^{2}+\Ric\cdot\varphi^{2}\geqslant 0; (ii) there exist \alpha\in(0,1) and u\in C^{2,\alpha}(M) such that u>0 , -\gamma\Delta u+\Ric\cdot u\geqslant 0 on M , and \langle\nabla u,\nu\rangle=0 on \partial M . Here \nu denotes the unit outward normal to the boundary. The equivalence of the two conditions is readily seen on compact manifolds, and follows from the argument of [ 12 ] on noncompact manifolds. The study of manifolds with Ricci lower bounds in the spectral sense has recently attracted substantial attention: see, e.g., [ 1 , 2 , 6 , 17 , 7 ] , and the references therein. This study has been connected to the study of stable minimal hypersurfaces, and in particular to the solution of the stable Bernstein problem in \mathbb{R}^{n} ; we refer to [ 9 , 8 , 24 , 5 ] . In [ 2 ] , the authors prove that, when n\geqslant 3 , if (M^{n},g) is compact then (2) \lambda_{1}\left(-\frac{n-1}{n-2}\Delta+\mathrm{Ric}\right)>0\Rightarrow\pi_{1}(M)\,\text{is finite}, and the constant \frac{n-1}{n-2} is sharp for ( 2 ) to hold. Moreover, in [ 4 ] the authors prove that, when n\geqslant 3 , if (M^{n},g) is compact then (3) \lambda_{1}\left(-\frac{n-1}{n-2}\Delta+\mathrm{Ric}\right)\geqslant 0\Rightarrow b_{1}(M)\leqslant n. In [ 3 ] , the authors prove that when n\geqslant 3 and \gamma>\frac{n-1}{n-2} , then the condition \lambda_{1}(-\gamma\Delta+\mathrm{Ric})>0 is stable under connected sums. Thus, one can observe that also in ( 3 ) the constant \frac{n-1}{n-2} is sharp for the property to hold. These results hint at the following natural question. Question 1 . Let n\geqslant 2 . Define \displaystyle C(n):=\Bigl\{\gamma\geqslant 0: \displaystyle\text{Any compact $M^{n}$ admitting a $C^{\infty}$ metric $g$ with $\lambda_{1}(-\gamma\Delta_{g}+\mathrm{Ric}_{g})\geqslant 0$,} \displaystyle\text{also admits a $C^{\infty}$ metric $\widetilde{g}$ such that $\mathrm{Ric}_{\widetilde{g}}\geqslant 0$}\Bigr\}. Find the value of \gamma(n):=\sup C(n). By using Gauss–Bonnet theorem, \gamma(2)=+\infty . In view of the results in [ 2 , 3 ] explained above, we also have \gamma(n)\leqslant\frac{n-1}{n-2} for every n\geqslant 3 . Problems related to 1 with \mathrm{Scal} instead of \mathrm{Ric} have been considered in [ 15 , 6.1.2] , and, e.g., in [ 22 , 16 ] . One of the aims of this paper is to settle a sharp boundary analogue of 1 for compact manifolds with mean-convex boundary in dimensions n\neq 4 ; see Theorem 3 . To the best of the authors’ knowledge, 1 remains widely open in dimensions n\geqslant 5 . 1.1. Main results The understanding of 1 in the setting of compact manifolds with mean-convex boundary rests on two key results. The first is a sharp spectral generalization of a rigidity theorem first proved by Kasue [ 19 , Theorem B(1)] (see also [ 18 ] ): a complete manifold with \Ric\geqslant 0 and disconnected mean-convex boundary, at least one component of which is compact, must split isometrically. See also [ 10 ] for a generalization, and the recent related work [ 11 ] . We stress that in this paper H_{\partial M} denotes the mean curvature with respect to the outward unit normal \nu to the boundary \partial M . Theorem 1 . Let n\geqslant 2 . Let (M^{n},g) be a complete (possibly noncompact) manifold with nonempty boundary \partial M . Let 0\leqslant\gamma<4 if n=2 , or 0\leqslant\gamma<\frac{n-1}{n-2} if n\geqslant 3 . Let us assume that: (1) There is 0<u\in C^{2,\alpha}(M) such that (4) -\gamma\Delta u+u\mathrm{Ric}\geqslant 0\,\,\text{on $M$},\qquad\text{and}\qquad H_{\partial M}+\gamma u^{-1}\langle\nabla u,\nu\rangle\geqslant 0\,\,\text{on $\partial M$}. (2) \partial M has at least two connected components, at least one of which is compact. Then \mathrm{Ric}\geqslant 0 , H_{\partial M}\geqslant 0 , and (M^{n},g) splits isometrically as ([0,L]\times\Sigma^{n-1},\mathrm{d}t^{2}+g_{\Sigma}) , where (\Sigma,g_{\Sigma}) is a compact Riemannian manifold with \mathrm{Ric}_{g_{\Sigma}}\geqslant 0 , and 0<L<\infty . Remark 1 . The following remarks are in order. (1) The fact that at least one connected component of \partial M is compact is necessary in Theorem 1 when n\geqslant 4 . Indeed, in \mathbb{R}^{4} the minimal catenoid lies within two parallel hyperplanes. Thus, there is an open set A in \mathbb{R}^{4} bounded by a hyperplane and a minimal catenoid. Notice that A satisfies the assumptions in Theorem 1 (with \gamma=0 and u\equiv 1 ) but it does not split isometrically. We remark that when n\in\{2,3\} , a manifold M^{n} with at least two mean-convex boundary components (regardless of the compactness assumption) and \mathrm{Ric}\geqslant 0 always splits isometrically. When n=2 this is readily seen: one can show that there exists a geodesic line in M , and conclude using the splitting theorem. When n=3 , it was recently proved in [ 11 , Corollary 3] . It would be interesting to understand if the compactness assumption in Item (2) of Theorem 1 can be dropped when n\in\{2,3\} . Since this question is not in the scope of this paper, we do not address it here. (2) The thresholds 0\leqslant\gamma<4 if n=2 , and 0\leqslant\gamma<\frac{n-1}{n-2} if n\geqslant 3 are sharp for Theorem 1 to hold. See, respectively, the interesting hyperbolic example in Proposition 3 , and the example in Proposition 2 (inspired by [ 2 ] ). (3) When n=2 , if the manifold (M^{n},g) in Theorem 1 is compact, we obtain the same result with the weaker assumption \gamma\geqslant 0 . In fact, the compactness of M rules out the possible escaping at infinity

Read the complete text