On asymptotic properties of a number theoretic function arising out of a spin chain model in statistical mechanics (Q5953667)

Let \(A:= \left(\begin{smallmatrix} 1 &0\\ 1 &1 \end{smallmatrix} \right)\), \(B:= \left(\begin{smallmatrix} 1 &1\\ 0 &1 \end{smallmatrix} \right)\), and for \(n\in \mathbb{N}\), let \(\Phi(n)\) be the number of matrices \(C\) which are products of \(A\)'s and \(B\)'s where both \(A\) and \(B\) must occur, such that the trace \(\text{Tr} (C)=n\). It had been conjectured that \(\Phi(n)\sim n\log n\), \(n\to\infty\). In this paper it is shown that the summatory function \(\Psi(n)= \sum_{j\leq n}\Phi(j)\) behaves as follows \[ \Psi(n)= \tfrac{6}{\pi^2} n^2\log n+O(n^2\log\log n), \qquad n\to \infty. \] This disproves the conjecture, as can be easily shown by using partial summation to relate the behaviours of \(\Phi(n)\) and \(\Psi(n)\), even if the conjecture would be modified in the way \(\Phi(n)\sim \frac{12}{\pi^2}n\log n\), \(n\to\infty\). In fact, \textit{M. Peter} has shown that the limit \(\lim_{n\to\infty} \frac{\Phi(n)}{n\log n}\) does not exist [J. Number Theory 90, 265-280 (2001; Zbl 1002.11072)]. On the other hand, via partial summation the main result of the paper shows that for the partition function one has \[ Z(\beta)= \sum_{n\geq 1} \frac{\Phi(n)}{n^\beta}= \frac{12}{\pi^2} (\beta-2)^{-2}+ O((\beta-2)^{-1} \log(\beta-2)^{-1}), \qquad \beta\to 2+0. \] Thus the singularity of the partition function at \(\beta=2\) is of order 2 from the right. The authors do not get any information on the analytic continuation of \(Z(\beta)\) beyond its abscissa of convergence. For the proof of the main result, mean value formulae for certain arithmetical functions are needed which are defined in terms of reduced quadratic irrationalities. Since it is difficult to work with them directly, they are approximated in the paper by multiplicative functions.