Geometric and arithmetic properties of Löbell polyhedra (Q6936508)

The Löbell polyhedra \( L_n \) are a family of polyhedra in hyperbolic \(3\)-space introduced by \textit{A. Yu. Vesnin} [Sib. Mat. Zh. 28, No. 5(165), 50--53 (1987; Zbl 0635.53020)], generalising a construction of \textit{F. Löbell} [Ber. Verh. Sächs. Akad. Leipzig 83, 167--174 (1931; Zbl 0002.40607)]. The groups \( \Gamma_n \) generated by the reflections in the faces of \( L_n \) are discrete cocompact subgroups of \( \mathrm{Isom}(\mathbb{H}^3)\).\N\NThe notion of quasi-arithmeticity of a lattice \( \Gamma \) over a number field \(k\) is a weakening of the notion of an arithmetic group; detailed definitions may be found in Section 2.2 of the paper, but roughly speaking one replaces the usual condition `\(\Gamma\) is commensurable with the \(\mathcal{O}_k\)-points of a suitable algebraic group' with `there exists a finite-index subgroup of \( \Gamma \) which is contained in the \(k\)-points of a suitable algebraic group'. In the paper under review, the authors show that the groups \( \Gamma_n \) are quasi-arithmetic if and only if \( n \in \{ 5,6,8,12 \} \), and that the group \( \Gamma_{12} \) is strongly quasi-arithmetic. This extends an earlier result of \textit{O. Antolín-Camarena} et al. [in: Complex dynamics. Families and friends. Wellesley, MA: A K Peters. 597--631 (2009; Zbl 1202.51020)], who showed that the groups \( \Gamma_n \) are arithmetic if and only if \( n \in \{5,6,8\} \). The proof uses the famous classification of arithmetic triangle groups due to \textit{K. Takeuchi} [J. Math. Soc. Japan 29, 91--106 (1977; Zbl 0344.20035)] to rule out all but finitely many cases which are then checked computationally using a theorem of \textit{È. B. Vinberg} [Mat. Sb., Nov. Ser. 72(114), 471--488 (1967; Zbl 0166.16303); translation from Mat. Sb., N. Ser. 72(114), No. 3, 471--488 (1967); correction 73(115), No. 2, 303 (1967)].\N\NAs well as the main result, the paper shows that the groups \( \Gamma_n \) come in infinitely many distinct commensurability classes and that the systoles of suitable manifold covers of the orbifolds \( \mathbb{H}^3/\Gamma_n \) approach \( 0 \) as \( n \to \infty \). They observe that the polyhedra \( L_n \) can be modified by contracting edges to ideal vertices to produce hyperbolic antiprisms, and that this produces an infinite family of incommensurable cusped hyperbolic \(3\)-manifolds with systoles approaching \(0\). They explain why such arguments do not work in higher dimension.