Hyperbolic spaces at large primes and a conjecture of Moore. (Q1430396)

A space \(X\) is said to have a homotopy exponent at a prime \(p\) if there is \(n\) such that \(p^n\) annihilates the \(p\)-torsion in \(\pi_*(X)\). A finite 1-connected complex X is called elliptic if \(\dim \pi_*(X)\otimes \mathbb{Q}\) is finite and hyperbolic otherwise. A conjecture of Moore states: A \(1\)-connected finite CW-complex \(X\) has a homotopy exponent at \(p\) if and only if \(X\) is elliptic (according to this, the existence of an exponent is independent of the prime \(p\)). By a result of \textit{C. A. McGibbon} and \textit{C. W. Wilkerson} [Proc. Am. Math. Soc. 96, 698--702 (1986; Zbl 0594.55006)], an elliptic space has an exponent for almost all primes. For hyperbolic spaces Selick has verified the conjecture for suspensions with torsion-free homology [\textit{P. Selick}, Math. Proc. Camb. Philos. Soc. 94, 53--60 (1983; Zbl 0522.55010)], and Anick proved it for hyperbolic spaces of Lusternik-Schnirelmann category two at large primes [\textit{D. Anick}, Lect. Notes Math. 1370, 24--52 (1989; Zbl 0671.55009)]. The main theorem proved in this paper states that ``A hyperbolic space \(X\) which is \(r\)-connected of dimension \(m\) and with \(H_*(\Omega(X);\mathbb{Z}_{(p)})\) torsion free has no homotopy exponent at \(p\) if \(p> (m-1)/r+(2p-2)/(2p-3)\).'' Finally, a version of this theorem for a class of formal spaces is given in the last section of the paper.