Primitive roots in quadratic fields. II. (Q880068)

Given a non-trivial unit in a real quadratic number field, for any rational prime \(p\) which is inert, the maximal order of the unit modulo \((p)\) is \(p+1\). An extension of Artin's primitive root conjecture is that there are infinitely many such primes \(p\). Under the Generalized Riemann Hypothesis this conjecture is known to be true. Unconditionally the author showed in Part~I [Int. J. Number Theory 2, No. 1, 7--23 (2006; Zbl 1099.11061)] that for any choice of 7 units in different real quadratic number fields satisfying a simple restriction, at least one of the units satisfies the above version of Artin's conjecture. Here the author deals with the case of non-units in any (not necessarily real) fixed quadratic field. Given a non-unit \(\alpha\) from such a quadratic field \(K\), for a rational prime \(p\) which is inert in the field, the maximal possible order of \(\alpha\) modulo \((p)\) is \(p^2-1\). It is shown that for any choice of 113 algebraic integers of \(K\) satisfying a simple restriction, at least one of the algebraic numbers has order at least \((p^2-1)/48\bmod (p)\) for infinitely many inert primes \(p\). This result supplements a result of \textit{W. Narkiewicz} [J. Reine Angew. Math. 381, 110--115 (1987; Zbl 0619.10044)], who in the much easier case of split primes \(p\) obtained a much stronger result (with 113 replaced by 3). A crucial ingredient in the author's proof is Selberg's lower bound sieve.