Plateau Bubbles and the Quintuple Bubble Theorem on $\mathbb{S}^n$
Abstract
Sullivan's multi-bubble isoperimetric conjectures in -dimensional Euclidean and spherical spaces assert that standard bubbles uniquely minimize total perimeter among all bubbles enclosing prescribed volume, for any . The double-bubble conjecture on was confirmed by Hutchings-Morgan-Ritor\'e-Ros (and later extended to ). The double-bubble conjecture on () and the triple- and quadruple- bubble conjectures on and (for and , respectively) were recently confirmed in our previous work, but the approach employed there does not seem to allow extending these results further. In this work, we confirm the quintuple-bubble conjecture on (), and as a consequence, by approximation, also the quintuple-bubble conjecture on () but without the uniqueness assertion. Moreover, we resolve the conjectures on and on (without uniqueness) for all , conditioned on the assumption that the singularities which appear at the meeting locus of several bubbles obey a higher-dimensional analogue of Plateau's laws. Another scenario we can deal with is when the bubbles are full-dimensional ("in general position"), or arrange in some good lower-dimensional configurations. To this end, we develop the spectral theory of the corresponding Jacobi operator (finding analogies with the quantum-graph formalism), and a new method for deforming the bubbles into a favorable configuration. As a by-product, we show that the Jacobi operator on a minimizing configuration always has index precisely and hence the corresponding isoperimetric profile is concave, answering a question of Heppes. Several compelling conjectures are proposed, which would allow extending our results to all unconditionally.
Cite
@article{arxiv.2307.08164,
title = {Plateau Bubbles and the Quintuple Bubble Theorem on $\mathbb{S}^n$},
author = {Emanuel Milman and Joe Neeman},
journal= {arXiv preprint arXiv:2307.08164},
year = {2024}
}
Comments
82 pages, addressed the slight difference in conventions from arXiv:2205.09102, added reference to Heppes' question on the concavity of the isoperimetric profile and additional recent references regarding isoperimetric problems on clusters