Temporal coherence, laser spectroscopy

In the case of coherent laser light, the pulses are characterized by well-defined phase relationships and slowly varying amplitudes (Haken, 1970). Such quasi-classical light pulses have spectral and temporal distributions that are also strictly related by a Fourier transformation, and are hence usually refered to as Fourier-transform-limited. They are required in the typical experiments of coherent optical spectroscopy, such as optical nutation, free induction decay, or photon echoes (Brewer, 1977). Here, the theoretical treatments generally adopt a semiclassical procedure, using a density matrix or Bloch formalism to describe the molecular system subject to a pulsed or continuous classical optical field, which generates a macroscopic sample polarization. In principle, a fully quantal description is possible if one represents the state of the field by the coherent or quasi-classical state vectors (Glauber, 1965 Freed and Villaeys, 1978). For our purpose, however. 

Many experiments in laser spectroscopy depend on the coherence properties of the radiation and on the coherent excitation of atomic or molecular levels. Some basic ideas about temporal and spatial coherence of optical fields and the density-matrix formalism for the description of coherence in atoms are therefore discussed at the end of this chapter. 

Two techniques, which appear well suited to the diagnostic probing of practical flames with good spatial and temporal resolution, are coherent anti-Stokes Raman spectroscopy (CARS) and saturated laser fluorescence. The two techniques are complementary in regard to their measurement capabilities. CARS appears most appropriate for thermometry and major species concentration measurements, saturated laser fluorescence to trace radical concentrations. With electronic resonant enhancement (6), CARS may be potentially useful for the latter as well. Fluorescence thermometry is also possible (7, 8) but generally, is more tedious to use than CARS. In this paper, recent research investi-... 

The earliest pulsed laser quantum beat experiments were performed with nanosecond pulses (Haroche, et al., 1973 Wallenstein, et al., 1974 see review by Hack and Huber, 1991). Since the coherence width of a temporally smooth Gaussian 5 ns pulse is only 0.003 cm-1, (121/s <-> 121 cm"1 for a Gaussian pulse) nanosecond quantum beat experiments could only be used to measure very small level splittings [e.g. Stark (Vaccaro, et al., 1989) and Zeeman effects (Dupre, et al., 1991), hyperfine, and extremely weak perturbations between accidentally near degenerate levels (Abramson, et al., 1982 Wallenstein, et al., 1974)]. The advent of sub-picosecond lasers has expanded profoundly the scope of quantum beat spectroscopy. In fact, most pump/probe wavepacket dynamics experiments are actually quantum beat experiments cloaked in a different, more pictorial, interpretive framework,... 

Over-the past decade, not only have pulse durations decreased from 10 to 10" s but there has been a dramatic increase in the tunability of lasers, such that tunable coherent radiation can now span the VUV to the very long wavelength laser radar. Femtosecond spectroscopy, like most advances, has begun in the visible region and considerable research and development is necessary to expand this present spectral range around 600 nm (4). However, it is also the case that for many problems in photo dynamics, for which the state selectivity or the nature of the optically prepared initial state is of paramount importance, the spectral line-width (Av) of the pulse must remain narrow. Thus the transform-limited bandwidth relationships (AvA K) govern the temporal properties of the laser pulse and, for example, a 5 ns pulse of 0.01 cm" linewidth prepares a different ensemble than a 300 fs pulse of 26 cm linewidth at the same wavelength. 

A number of laser based temporal domain experiments have evolved in the past two decades which have no equivalents in conventional spectroscopic technique as they exploit the coherence inherent to stimulated sources. The origin of these methods traces directly to techniques which are widely utilized in NMR and ESR spectroscopy to determine various relaxation and dephasing times. In fact, much of the terminology from magnetic resonance processes has been carried over to laser driven coherent transient studies. See Shoemaker (1978), Levenson (1982) and Brewer and DeVoe (1984). 

---