A particle on a ring or: how I learned to stop worrying and love $θ$-vacua

Abstract

Recently, Ai, Cruz, Garbrecht, and Tamarit (arXiv:2001.07152, arXiv:2404.16026, arXiv:2511.04216) claimed that the strong CP problem can be avoided by adopting a particular order of limits in the Euclidean path integral, in which the spacetime volume is taken to infinity before summing over all topological sectors. We critically examine this proposal using exactly solvable examples of one-dimensional quantum mechanics on a ring, namely the quantum rotor and the quantum pendulum. These systems provide fully controlled settings with known $θ$-dependent spectra. We find that the ACGT procedure fails to reproduce the correct energy spectrum. Since the spectrum is a direct physical observable, this result demonstrates that the proposed order of limits cannot be justified and conclusions about CP conservation in QCD cannot be based on this prescription alone.