A second limit formula for higher rank twisted Epstein zeta functions and some applications (Q763972)

Let \(Q\) be a positive definite quadratic form on \(\mathbb{R}^n\). Given \(u,v \in \mathbb{R}^n\), the twisted Epstein zeta function is defined as \[ \zeta_n(s, u, v, Q) = \sum_{m \in \mathbb{Z}^n \setminus \{0\}} e^{2\pi i m \cdot n} Q[m]^{-s}. \] This series is absolutely convergent for \(s \in \mathbb{C}\) with real part \(\geq n/2\). It is also known that \(\zeta_n(s,u,v,Q)\) admits a meromorphic continuation to \(s \in \mathbb{C}\), with a simple pole at \(s=n/2\) if \(u \in \mathbb{Z}^n\), otherwise it is an entire function. It also satisfies a functional equation with respect to \(s \mapsto n/2 - s\). The twisted Epstein zeta function has applications in arithmetic. For example, the case \(n=2\) is related to the study of quadratic number fields. Assume hereafter that \(u \notin \mathbb{Z}^n\). A second limit formula for \(\zeta_n\) is a formula for \(\zeta_n(n/2, u, 0, Q)\). The case \(n=2\) is just the celebrated formula of Kronecker. When \(n \geq 3\), \textit{I. Efrat} [Duke Math. J. 55, 349--360 (1987; Zbl 0607.10019)] derived a second limit formula, and used it to compute the determinant of the Dirac operators and Laplacians for certain spaces of quasi-periodic functions on a torus. The aim of this article is to generalize Efrat's approach to general \(n\). The author obtains an explicit second limit formula in terms of \(\zeta_{n-1}(n/2, \cdots)\), as well as a \(K\)-Bessel expansion which generalizes the Chowla-Selberg formula. His arguments are based on elaborate computations. Using this and the functional equation, he is able to compute the determinant of the Laplacian on a function space \(S^u(T)\).