On numbers \(n\) for which the prime factors of \(\sigma(n)\) are among the prime factors on \(n\) (Q1885231)

Let \(\text{rad}(n)\) denote the product of all prime divisors of \(n\). At the West Coast Number Theory Conference in 2000, J.-M. DeKoninck asked for all positive integers \(n\) for which \(\sigma(n)=\text{rad}(n)^2\). Note that \(n=1\) and \(n=1782\) satisfy this equation. In this article, the author considers a more general question: he looks at integers \(n\) such that \(\sigma(n)\mid\text{rad}(n)^K\) for some \(K\). He proves that for any given \(T\), there are only finitely many such \(n\) with \(T\) prime factors. More precisely, let \(\omega(n)\) be the number of distinct prime factors of \(n\). Let \(K,L, T\) be fixed positive integers. Assume that \(\omega(n)=T\) and that \(\sigma(n)=am\) for some positive integer \(a\leq L\) and for some positive integer \(m\) such that \(m\mid\text{rad}(n)^ K\). The author proves that \(n<\exp((M\cdot T!)^{2^T})\), where \(M\) is any positive constant with \(M\geq K+\log L\). The proof is done by induction on \(T\); the arguments are elementary but intricate.