Planar Site Percolation, End Structure, and the Benjamini-Schramm Conjecture
Abstract
Let be an infinite, connected, locally finite planar graph and consider i.i.d.\ Bernoulli site percolation. Write and for the critical and uniqueness thresholds. Using a well--separated Freudenthal embedding , we introduce a cycle--separation equivalence on ends and associated ``directional'' thresholds . When the set of end--equivalence classes is countable, we show that and that for every there are almost surely infinitely many infinite open clusters. Combined with the theorem of Glazman--Harel--Zelesko for , this yields non--uniqueness throughout the full coexistence interval , and hence in this setting. This resolves the extension problem posed by Glazman--Harel--Zelesko for the upper half of the coexistence regime under a natural countability hypothesis. In contrast, for graphs with uncountably many end--equivalence classes we give criteria guaranteeing infinitely many infinite clusters above criticality, and we construct an explicit locally finite planar graph of minimum degree at least for which . Consequently, the Benjamini--Schramm conjecture (Conjecture 7 in \cite{bs96}) that planarity together with minimal vertex degree at least 7 forces infinitely many infinite clusters for all does not hold in full generality. Our proofs combine a cutset characterization of with a planar alternating--arm exploration organized by an end--adapted boundary decomposition.
Cite
@article{arxiv.2601.09958,
title = {Planar Site Percolation, End Structure, and the Benjamini-Schramm Conjecture},
author = {Zhongyang Li},
journal= {arXiv preprint arXiv:2601.09958},
year = {2026}
}