Note on the Diophantine equation \(1+2p+(2p)^2+\dots+(2p)^n=y^p\) (Q950671)

In 1987, \textit{A Rotkiewicz} [Elem. Math. 42, No. 3, 76 (1987; Zbl 0703.11016)] showed that the repunits (numbers of the form \(111\dots 11=1+10+\dots+10^n\)) are never squares or cubes for \(n\geq 1\). For this he used a more general result on the Diophantine equations \(1+x+\dots+x^n=y^2\) and \(1+x+\dots+x^n=y^3\) proved by \textit{W. Ljunggren} in 1943 [Norsk Mat. Tidsskr. 25, 17--20 (1943; Zbl 0028.00901)]. In this paper the author gives an elementary proof (using Fermat's little theorem) of the following result implying that repunits are never fifth powers of integers: Theorem: Let \(p\) be an odd prime or a Carmichael number and let \(n\in \mathbb N\). Then the Diophantine equation \(1+2p+(2p)^2+\dots+(2p)^n=y^p\) has no solution \(y\).