On the \(\mathcal F\)-hypercentre and the intersection of all \(\mathcal F\)-maximal subgroups of a finite group. (Q425281): Difference between revisions

In this report we will consider only finite groups. A class of groups \(\mathcal F\) is said to be a formation if it is closed under taking quotients and subdirect products. It is said to be hereditary when it is closed under taking subgroups. It is said to be saturated if \(G\in\mathcal F\) provided that \(G/\Phi(G)\in\mathcal F\). Given a group \(G\) and a class of groups \(\mathcal X\), the intersection of all \(\mathcal X\)-subgroups of \(G\) is denoted by \(\text{Int}_{\mathcal X}(G)\). Groups which do not belong to \(\mathcal X\) but have all proper subgroups in \(\mathcal X\) are called \(\mathcal X\)-critical groups. The \(\mathcal X\)-hypercentre \(Z_{\mathcal X}(G)\) of \(G\) is the product of all normal \(\mathcal X\)-hypercentral subgroups of \(G\). It is shown by \textit{R. Baer} [Trans.\ Am.\ Math.\ Soc.\ 75, 20-47 (1953; Zbl 0051.25702)] that when \(\mathcal X\) is the class \(\mathcal N\) of all nilpotent groups, then \(\text{Int}_{\mathcal N}(G)=Z_{\mathcal N}(G)\), but the equality does not hold for other classes of groups. The main object of this paper is to determine for which hereditary saturated formations \(\mathcal F\) the equality \(\text{Int}_{\mathcal F}(G)=Z_{\mathcal F}(G)\) holds for all groups \(G\) (respectively, for all soluble groups \(G\)). To do this, we say that a hereditary saturated formation \(\mathcal F\) with \(\pi(\mathcal F)\neq\emptyset\) satisfies the boundary condition (respectively, the boundary condition in the class of all soluble groups) if for any \(p\in\pi(\mathcal F)\), \(G\in\mathcal F\) whenever \(G\) is an \(F(p)\)-critical group (respectively, a soluble \(F(p)\)-critical group). Here \(F\) represents the canonical definition of the saturated formation \(\mathcal F\). According to Theorems~A and~B, the equality \(\text{Int}_{\mathcal F}(G)=Z_{\mathcal F}(G)\) holds for all groups (respectively, all soluble groups) when the formation \(\mathcal F\) satisfies the boundary condition (respectively, the boundary condition in the class of all soluble groups). Some properties of \(\text{Int}_{\mathcal X}(G)\) are presented in Theorem~C. Some classes of groups satisfying the boundary condition are shown. For instance, the formation \(\mathcal N^r\) of all soluble groups of nilpotent length at most \(r\) satisfies the boundary condition in the class of all soluble groups, and the formation \(\mathcal N\) of all nilpotent groups satisfies the boundary condition. Given a partition \(\{\pi_i\mid i\in I\}\) of the set \(\mathbb P\) of all prime natural numbers, the class of all (soluble) groups \(G\) of the form \(G=A_{i_1}\times\cdots\times A_{i_t}\), where \(A_{i_j}\) is a Hall \(\pi_{i_j}\)-subgroup of \(G\) and \(i_1,\dots,i_t\in I\), is a hereditary saturated formation satisfying the boundary condition (in the class of all soluble groups). Lattice formations (formations \(\mathcal F\) in which the set of all \(\mathcal F\)-subnormal subgroups of a group \(G\) forms a sublattice of the lattice of all subgroups of \(G\)) of soluble groups containing the class \(\mathcal N\) of all nilpotent groups are hereditary saturated formations satisfying the boundary condition in the class of all soluble groups. The class of all groups with nilpotent derived subgroup is also a hereditary saturated formation satisfying the boundary condition. However, other formations, like the classes of all \(p\)-soluble groups, all \(p\)-supersoluble groups, all \(p\)-nilpotent groups, and all soluble groups do not satisfy the boundary condition. The key to show it is Lemma~4.8, which is a restatement of Theorem~A and says that if \(\mathcal F\) is a non-empty saturated formation and for some prime \(p\) the equality \(F(p)=\mathcal F\) holds, then \(\mathcal F\) does not satisfy the boundary condition. Many known results about the influence of the elements of prime order and order~\(4\) or the maximal subgroups of the Sylow subgroups of a group on its structure can be restated with the help of the subgroup \(\text{Int}_{\mathcal F}(G)\). A \(p\)-decomposability criterion for groups with a triple factorisation \(G=A_1A_2=A_2A_3=A_1A_3\) where there exists a prime \(q\) such that \(\gcd(|G:A_i|,|G:A_j|)\) is a power of \(q\) for all \(i\neq j\) and \(A_i\cap A_j\leq\text{Int}_{\mathcal F}(A_i)\cap\text{Int}_{\mathcal F}(A_j)\) for \(i\neq j\), where \(\mathcal F\) is the class of all \(p\)-decomposable groups: these groups turn out to be \(p\)-decomposable. Finally, the author gives examples of groups \(G\) in which \(\text{Int}_{\mathcal U}(G)\), where \(\mathcal U\) is the class of all supersoluble groups, does not coincide with the generalised hypercentre of \(G\), the largest term of the chain of subgroups \(1=Q_0\leq Q_1\leq\cdots\leq Q_i\leq\cdots\) where \(Q_i/Q_{i-1}\) is the subgroup of \(G/Q_{i-1}\) generated by the set of all cyclic Sylow-permutable subgroups of \(G/Q_i\).