Minimal degree of an element of a number field with respect to its quadratic extension (Q6113097)

Let \(\beta\) be an algebraic number and let \(L\) be a number field containing \(\beta\). In [\textit{C.-M. Park} and \textit{S. W. Park}, Int. J. Number Theory 18, No. 3, 485--500 (2022; Zbl 1494.11085)] the minimal degree \(\deg_L(\beta)\) of \(\beta\) with respect to \(L\) is defined as the smallest natural number \(d\) such that there exists a primitive element \(\alpha\) of \(L\) over \(\mathbb Q\) and a polynomial \(f\in\mathbb{Q}[x]\) of degree \(d\) for which \(\beta=f(\alpha)\). It is easy to see that \(\deg_L(\beta)\geq [L:\mathbb Q(\beta)]\), but in general there is no reason to expect an equality. In the paper under review equality is proved in the special situation that \(\beta\) is quadratic and \([L:\mathbb Q(\beta)]=2\). On the other hand if \([\mathbb Q(\beta):\mathbb Q]\geq 3\) several cases are studied in which there exist infinitely many quadratic extensions \(L/\mathbb Q(\sqrt{\beta})\) with \(\deg_L(\beta)>2\). In particular this is proved for all totally real algebraic numbers \(\beta\) of degree at least \(3\) over \(\mathbb Q\). For the convenience of the reader I would like to remark that in the proof of Lemma 7 the formula \(f(x)=g(x)^2\) is a recurring typo and should be always substituted by \(f(x)=g(x^2)\).