Moments of random matrices and hypergeometric orthogonal polynomials (Q2311921)

In random matrix theory, the moments \(\mathbb{E}(\mathrm{Tr}X_n^k)\) of Gaussian, Laguerre and Jacobi ensembles of \(n\!\times\!n\) random matrices \(X_n\), regarded as a function of \(n\), have been intensively studied in the case \(k \in \mathbb{N}\). In this paper, the authors investigate \(\mathbb{E}\mathrm{Tr}(X_n^k)\) as a function of a complex variable \(k\in\mathbb{C}\), and discover a new connection between moments of random matrices (properly normalized) and hypergeometric orthogonal polynomials (described in terms of the Askey scheme), reflection symmetry, zeros on a critical line in \(\mathbb{C}\), and orthogonality relations. The considered complex moments \(\mathbb{C}\ni k\mapsto \mathbb{E}\mathrm{Tr}(X_n^k)\) can be regarded as averages of spectral Riemann zeta functions, defined as maximal analytic continuation of \(s\mapsto \mathbb{E} \zeta_{X_n}(s)\!=\!\mathbb{E}(\mathrm{Tr}\, |X_n|^{-s})\). The relation of the obtained results with Mellin transform of products and Wronskians of pairs of classical orthogonal polynomials is investigated.