We discuss the limitations for the applicability of the Lifshitz theory to describe the temperature dependence of the Casimir force between bulk lossy metal slabs of finite sizes. We pay attention to the important fact that Lifshitz’s theory is not applicable when the characteristic wavelength of the fluctuating field, responsible for the temperature-dependent terms in the Casimir force, are longer than the size of the sample. As a result, the widely discussed linearly decreasing temperature dependence of the Casimir force can be observed only for dirty and large metal samples at high enough temperatures. Moreover, for the correct description of the Casimir effect at low enough temperatures, a careful consideration of the concrete geometry of the interacting samples is essential.

• Received 3 December 2009

DOI:https://doi.org/10.1103/PhysRevA.82.032511