Products of conjugacy classes in finite and algebraic simple groups. (Q1935197)

The paper under review contains many new and interesting results on products of conjugacy classes in simple groups, both finite and algebraic. \textit{Z. Arad} and \textit{M. Herzog} [Products of conjugacy classes in groups. Lect. Notes Math. 1112. Berlin: Springer-Verlag (1985; Zbl 0561.20004)] made the following conjecture: If \(S\) is a finite nonabelian simple group and \(A\), \(B\) are nontrivial conjugacy classes of \(S\), then \(AB\) is not a conjugacy class. This conjecture had been verified for alternating and sporadic groups, and for certain families of groups of Lie type. In this paper the authors prove the conjecture for all linear groups and for several other low-rank groups of Lie type. The proof is based on a simple idea: identify a small set of irreducible complex characters \(\Gamma\) of \(S\) such that for any pair of nontrivial classes \(A,B\), there exists a character \(\chi\in\Gamma\) which is non constant on \(AB\). A strong version of the Arad-Herzog conjecture is proved for all simple algebraic groups \(G\) (over algebraically closed fields) and this does not depend on the isogeny class of \(G\). Precisely, the authors prove that almost always the products of two conjugacy classes in a simple algebraic group is a union of infinitely many classes. As a corollary, the algebraic group analogue of Szep's conjecture immediately follows. In particular it is shown a complete classification of pairs of centralizers in a simple algebraic group which have dense product. Finally, these results lead to a generalization of the Baer-Suzuki theorem for \(p\)-elements with \(p\geq 5\).