Description
Cosmological phase transitions, especially first order phase transitions, have attracted renewed interest as potential sources of gravitational waves. Accurately predicting the resulting gravitational wave signal requires a reliable estimate of the transition rate, which is governed by a saddle-point configuration known as the bounce solution. The seminal work of Coleman, Glaser, and Martin established that at zero temperature, any nontrivial bounce solution to the equations of motion that minimizes the Euclidean action is O(D)-symmetric in D-dimensional spacetime. At finite temperature, however, it has not been proven that an O(D-1)-symmetric bounce solution in the spatial directions indeed yields the minimal Euclidean action, despite this assumption being widely used in the literature. In this talk, we extend the Coleman–Glaser–Martin analysis to finite temperature. We rigorously prove that for a broad class of scalar potentials, any saddle-point configuration with the least action is necessarily O(D-1)-symmetric and monotonic in the spatial directions. This result provides a firm mathematical foundation for the symmetry properties widely assumed in studies of thermal vacuum decay and cosmological phase transitions.