Negative group velocity electromagnetic waves and the energy–momentum tensor

Abstract

V G Veselago's results (Usp. Fiz. Nauk 179 689 (2009) [Phys. Usp. 52 649 (2009)]) on the electromagnetic (EM) energy–momentum tensor in a medium are analyzed. It is shown that Veselago's statements on the Abraham tensor are wrong (this is not actually a tensor, and the Abraham force was introduced into the theory as an artificial auxiliary device). In discussing the EM energy–momentum tensor in a dispersive medium, it seems to have escaped the author's attention that the problem was resolved a long time ago: the electromagnetic energy–momentum tensor for a dispersive isotropic medium at rest is a symmetric 4-tensor which includes the Brillouin energy density, the energy flux density (Umov–Poynting vector), the momentum density (the Umov–Poynting vector divided by ), and the Pitaevskii tension tensor. For a mechanically and thermally equilibrium medium, it is shown that the spatial components of the Polevoi–Rytov tensor which is discussed in the analyzed paper cannot be interpreted as the field-dependent part of the Pitaevskii total tension tensor, unless for quasimonochromatic plane wave propagation. It is also shown that for arbitrary (not necessarily zero) reflection, the force an EM wave in an isotropic medium exerts on a solid can be expressed in terms of an appropriate component of the Polevoi–Rytov tension tensor.