Groups whose Chermak-Delgado lattice is a quasi-antichain (Q2414569)

All groups in this report will be finite. Let \(G\) be a group and let \(H\le G\). The Chermak-Delgado measure of \(H\) in~\(G\) is defined as \(m_G(H)=\lvert H\rvert \lvert {\operatorname{C}_G(H)}\rvert\). The maximal Chermak-Delgado measure of all subgroups of~\(G\) is denoted by \(m^*(G)\). The set \(\mathcal{CD}(G)\) of all subgroups \(H\le G\) such that \(m_G(H)=m^*(G)\) is a sublattice of the lattice of all subgroups of \(G\), called the Chermak-Delgado lattice of~\(G\). In [\textit{B. Brewster} et al., Arch. Math. 103, No. 4, 301--311 (2014; Zbl 1308.20019)], it was proved that the Chermak-Delgado lattice of a finite group can be an antichain of width \(w\) (that is, composed by a maximum, a minimum, and \(w\) atoms) if, and only if, \(w=1+p^a\) for some positive integer \(a\) and some prime \(p\). It was also proved that the number of abelian atoms must be of the form \(t=p^b+1\) for some integer \(b\) (where \(t\ge 2\)). That paper proposed as open questions: (Q1) Which values of \(t\) are possible in a quasi-antichain Chermak-Delgado lattice of width \(w=p^a+1\) when \(a>1\)? (Q2) Are there examples of groups \(G\) with \(G\in\mathcal{CD}(G)\) and \(\mathcal{CD}(G)\) a quasi-antichain where \(t=0\) and \(p\equiv 1\pmod{4}\)? The aim of this paper is to solve these questions. The main theorem of the paper under review gives the answer the first qustion. It shows that under these hypothesis, \(a=b\) or \(a=2b\) and that, if \(\lvert G/{\operatorname{Z}(G)}\rvert=p^{2n}\), then \(a\mid n\). Theorems~3.3 and~3.4 show constructions of groups with both possible relations. Theorem~3.5 shows a construction of groups with Chermak-Delgado lattice a quasi-antichain and with exactly one (for \(p=2\)) or two (for \(p\ne 2\)) abelian atoms. Theorem~3.6 gives the construction corresponding to no abelian atoms, related to the second open question.