Matsumoto and Yoshimura have recently argued that the number density of heavy particles in a thermal bath is not necessarily Boltzmann-suppressed for T ≪ M , as power law corrections may emerge at higher orders in perturbation theory. This fact might have important implications on the determination of WIMP relic densities. On the other hand, the definition of number densities in a interacting theory is not a straightforward procedure. It usually requires renormalization of composite operators and operator mixing, which obscure the physical interpretation of the computed thermal average. We propose a new definition for the thermal average of a composite operator, which does not require any new renormalization counterterm and is thus free from such ambiguities. Applying this definition to the annihilation model of Matsumoto and Yoshimura we find that it gives number densities which are Boltzmann-suppressed at any order in perturbation theory. We discuss also heavy particles which are unstable already at T = 0, showing that power law corrections do in general emerge in this case.DFPD 00/TH/41

No abstract

The one-loop effective potential calculated for a generic model that originates from 5-dimensional theory reduced down to 4 dimensions is considered. The cut-off and dimensional regularization schemes are discussed and compared. It is demonstrated that the prescriptions are consistent with each other and lead to the same physical consequences. Stability of the ground state is discussed for a U(1) model that is supposed to mimic the Standard Model extended to 5 dimensions. It has been shown that fermionic Kaluza-Klein modes can dramatically influence the shape of the effective potential shifting the instability scale even by several orders of magnitude.