Simple zeros of primitive Dirichlet \(L\)-functions and the asymptotic large sieve (Q2922409)

The authors give their results in the present paper, under the generalized Riemann hypothesis (GRH), namely all the non-trivial zeros of Dirichlet \(L\)-functions (here \(\chi\) is a Dirichlet character modulo \(q\)) NEWLINE\[NEWLINE L(s,\chi):=\sum_{n=1}^{\infty}\chi(n)n^{-s} NEWLINE\]NEWLINE lie on the critical line \(\text{Re}(s):=\sigma=1/2\). For a positive (\(>\frac{11}{12}\)) proportion of these zeros, over primitive characters \(\chi\), the authors a prove simplicity result, see Theorem 1.1 for the details.NEWLINENEWLINEThey note in passing that the previously known lower bound (\(>0.8688...\)) of \textit{A. E. Özlük} [J. Number Theory 59, No. 2, 319--351 (1996; Zbl 0863.11056)] was obtained for \textit{all} Dirichlet characters, rather than for all primitive ones; this, in turn, lowered this constant: in section \(6\) the authors explain how.NEWLINENEWLINEThe second result of the paper (from which Theorem 1.1 follows like in Özlük paper) is Theorem 1.2, giving some kind of Montgomery pair correlation conjecture (we are always under GRH) for the \(q-\)adic case (i.e., for \(L(s,\chi)\) with modulus \(q\) instead of \(\zeta(s)\), the Riemann zeta function), following Özlük approach. Through a version of the explicit formula for primes in arithmetic progressions (namely, for partial sums of \(-L'/L\), the logarithmic derivative of \(L\) above), they express the asymptotic formula for these Montgomery pair-correlation functions, in terms of an asymptotic formula for the mean-square, over characters and moduli, of the von Mangoldt function (i.e., \(\Lambda(n):=\log p\) for \(n=p^k\), \(p\) prime, \(k\) natural, \(\Lambda(n):=0\) otherwise), twisted by \(\chi(n)\), resembling much the large sieve quantities: see Proposition 1.1.NEWLINENEWLINEThe last result of the paper is a rather straightforward consequence of the Corollary 1.1. This gives an asymptotic formula which is a kind of Barban-Davenport-Halberstam asymptotic formula for mean-squares of primes in arithmetic progressions: they highlight this connection explicitly in section 7.NEWLINENEWLINEThe main parts of the paper are devoted to the proof of Proposition 1.1, whose core consists of the asymptotic large sieve.NEWLINENEWLINEThe main novelty of this paper, with respect to Özlük's previous result, is a much stronger technical tool, namely the asymptotic large sieve developed by \textit{B. Conrey} et al. [``Asymptotic large sieve'', Preprint, \url{arXiv:1105.1176}].