Reply to "Comment on 'Surface-impedance approach solves problems with the thermal Casimir force between real metals' "

The preceding Comment discusses in detail the main idea of our paper [ Phys. Rev. A 67, 062102 (2003) ], namely that one cannot substitute the Drude dielectric function into the Lifshitz formula for the thermal Casimir force in the frequency region where a real current of conduction electrons leads to Joule heating in the metal. In that Comment, it is claimed that this idea would be in contradiction to the fluctuation-dissipation theorem. In this Reply we present an explicit explanation why there is no contradiction. In the second part of the Comment an alternative method is suggested, different from the one used in our paper, to calculate the thermal Casimir force in the framework of the impedance approach. This method is in support of a previous prediction by Svetovoy and Lokhanin, criticized by us, that there exists a relatively large thermal correction to the Casimir force between real metals at small separations. Here we present strong quantitative arguments in favor of the statement that the method of the Comment is in violation of the Nernst heat theorem. We also demonstrate that it is in contradiction with experiment. The approach of our paper is shown to be in agreement with both thermodynamics and experimental data.