Casimir effect in dielectrics: Bulk energy contribution

In a recent series of papers, Schwinger discussed a process that he called the dynamical Casimir effect. The key essence of this effect is the change in zero-point energy associated with any change in a dielectric medium. (In particular, if the change in the dielectric medium is taken to be the growth or collapse of a bubble, this effect may have relevance to sonoluminescence.) The kernel of Schwinger’s result is that the change in Casimir energy is proportional to the change in the volume of the dielectric, plus finite-volume corrections. Other papers have called into question this result, claiming that the volume term should actually be discarded, and that the dominant term remaining is proportional to the surface area of the dielectric. In this paper, which is an expansion of an earlier Letter on the same topic, we present a careful and critical review of the relevant analyses. We find that the Casimir energy, defined as the change in zero-point energy due to a change in the medium, has at leading order a bulk volume dependence. This is in full agreement with Schwinger’s result, once the correct physical question is asked. We have nothing new to say about sonoluminescence itself.