Quang and John have, in a recent paper [Phys. Rev. A
56, 4273 (1997)], proposed a dressed-state Monte Carlo wave-function (MCWF) approach to a laser-driven two-level atom coupled to a structured radiation reservoir. In this Comment, we argue that this approach is at variance with the underlying physical idea of the MCWF technique, that it is at variance with basic formal requirements for the MCWF technique to apply, and that it produces spurious and unphysical quantitative results for wide ranges of parameters.
We introduce a dressed-state Monte Carlo wave-function technique to describe resonance fluorescence in a broad class of non-Markovian reservoirs with strong atom-reservoir interaction. The method recaptures photon localization effects which are beyond the Born and Markovian approximations, and describes the influence of the driving field on the atom-reservoir interaction. Using this approach, we predict a number of fundamentally new features in resonance fluorescence near the edge of a photonic band gap. In particular, the atomic population exhibits inversion for moderate applied field intensity. For a low external field intensity, the atomic system retains a long-time memory of its initial state.
We investigate an atomic $\ensuremath{\Lambda}$ system with one transition coupled to a laser field and a flat continuum of vacuum modes and the other transition coupled to field modes near the edge of a photonic band gap. The system requires simultaneous treatment of Markovian and non-Markovian dissipation processes, but the photonic-band-gap continuum cannot be eliminated within a density matrix treatment. Instead we propose a formalism based on Monte Carlo wave functions, and we present results relevant to the experimental characterization of a structured continuum.