On the infinite cluster of Bernoulli bond percolation in Scherk's graph (Q4537293)

Let \(G=(V,E)\) be a graph, with \(V\) the set of vertices and \(E\) the set of edges. The simple random walk on \(G\) is the Markov chain defined on the set \(V\) in which the probability of moving to neighboring vertices is given by \(1/\deg(v)\), where \(\deg (v)\) is the degree of the vertex \(v\). Then \(G\) is said to be recurrent or transient, depending on whether this simple random walk is recurrent or transient. For instance, the standard two-dimensional lattice \(L^2=(Z^2,E^2)\) is recurrent whereas the three-dimensional lattice \(L^3=(Z^3,E^3)\) is transient. Sherk's graph, denoted by \(S\), is a subgraph of \(L^3\) with the same set \(Z^3\) as set of vertices, and where two vertices \((x,y,z)\) and \((x',y',z')\) are adjacent if \(x=x'\) and \(|x-x'|+|z-z'|=1\), or \(z=z'=0\) and \(|x-x'|+|z-z'|=1\). It was shown by \textit{S. Markvorsen, S. McGuinness} and \textit{C. Thomassen} [Proc. Lond. Math. Soc., III. Ser. 64, No. 1, 1-20 (1992; Zbl 0772.05086)] that \(S\) is transient. In the Bernoulli bond percolation of \(G=(V,E)\), each edge \(e\in E\) is said to be open with probability \(p\) or closed with probability \(1-p\), independently of all other edges. In each realization, a connected component of the random subgraph \((V,\text{ all open edges})\) is called a cluster. Let \(p_c(G)=\inf\{p\): there is an infinite cluster in \(G\) with retention probability \(p\}\). The main result of this paper is the following Theorem: The infinite cluster of Bernoulli bond percolation in Scherk's graph is transient for \(p>{1\over 2}\) and is recurrent for \(p\in (p_c(S), {1\over 2})\).