Population trapping, quantum interference

The effect of quantum interference on spontaneous emission in atomic and molecular systems is the generation of superposition states that can be manipulated, to reduce the interaction with the environment, by adjusting the polarizations of the transition dipole moments, or the amplitudes and phases of the external driving fields. With a suitable choice of parameters, the superposition states can decay with controlled and significantly reduced rates. This modification can lead to subnatural linewidths in the fluorescence and absorption spectra [5,10]. Furthermore, as will be shown in this review, the superposition states can even be decoupled from the environment and the population can be trapped in these states without decaying to the lower levels. These states, known as dark or trapped states, were predicted in many configurations of multilevel systems [11], as well as in multiatom systems [12],... 

In this section we present a brief historical account of LICS without going into too many details, as our main concern is aspects of quantum interferences. LICS was initially suggested by Heller and Popov [4] and by Armstrong et al. [27], who termed the effect "pseudo-autoionization." Early theoretical investigations on LICS focused on coherent population trapping [28, 29], multichannel effects [30-33], and Raman-type transitions [2, 34 0]. 

The three-level saturation spectroscopy of A- and V-level configurations flourished in the subsequent years and led to the discovery of munerous new effects, such as coherent population trapping (Arimondo 1996), electromagnetically induced transparency (Harris 1997), and lasing without population inversion (Kocharovskaya 1992). These effects are beyond the scope of the present book. The control of quantum coherence and interference in laser-driven three-level systems has been treated in detail in the excellent reviews cited above. 

Figure 5 shows the ratio of trapped to ionized population for several peak intensities and three pulse widths. This ratio is close to zero for peak intensities ( peai) less than I s, as expected. For > Ires the ratio grows, and can be as much as one half. The trapped population is found mostly in states with / > 4 and quantum numbers > 7. Longer pulse widths result in more excitation and ionization, but a lower overall trapped/ionized ratio. Eventually, as the peak intensity is raised and the channel closing occurs earlier in the pulse, the ratio decreases since most of the ionization then occurs non-resonantly. The structure in the curves may be indicative of interference between resonant excitation on the rising and falling edges of the pulse, a subject we will return to in the next section. 

---