Laser saturated absorption experiments

We report on saturated absorption experiments in Na2> realized with a tunable and stabilized argon laser. These experiments provide both spectroscopic and physical results, which help in understanding the behavior of optically pumped alkali dimer lasers. We briefly describe a new double resonance experiment which enables us to study the gain line-shapes of the dimer laser and to demonstrate the backward-forward gain competition. 

Experimental Techniques. A block diagram of the experimental set-up used for saturated absorption experiments is shown in Figure 1. The argon laser is a commercial 4W tube in a home made cavity. This cavity is made of three Invar rods, decoupled from the tube in order to avoid vibrations. Line selection is made with a prism, and single frequency operation is obtained with a Michel son interferometer. The laser can be frequency locked to a stable Fabry-Perot resonator with a double servo-loop acting on a fast PZT for line narrowing and on a galvo-plate for wide tuna-bility. This results in a linewidth of less than 10 KHz and a continuous tunability of 6 GHz. 

Classical saturated absorption experiments on fast beams require different laser wavelengths to saturate and probe the same transition. However, a three-level system in V or A configuration can be realized to use the same direct and retroreflected laser beam interacting with the velocity class p under the condition... 

With special techniques it is possible to stabilize the laser frequency down to some 10 sec" 38,39) and promising experiments with the infrared line X = 3.39p of the He-Ne laser indicatethat a stability of 10 cycles/sec or better may be obtained when using the saturated absorption of molecules inside the laser resonator as the stabilizing element (see Section IV.3). 

The electronic structure of fluorenes and the development of their linear and nonlinear optical structure-property relationships have been the subject of intense investigation [20-22,25,30,31]. Important parameters that determine optical properties of the molecules are the magnitude and alignment of the electronic transition dipole moments [30,31]. These parameters can be obtained from ESA and absorption anisotropy spectra [32,33] using the same pump-probe laser techniques described above (see Fig. 9). A comprehensive theoretical analysis of a two beam (piunp and probe) laser experiment was performed [34], where a general case of induced saturated absorption anisotropy was considered. From this work, measurement of the absorption anisotropy of molecules in an isotropic ensemble facilitates the determination of the angle between the So Si (pump) and Si S (probe) transitions. The excited state absorption anisotropy, rabs> is expressed as [13] ... 

Saturated Laser Induced Fluorescence Spectroscopy. The development of saturated laser induced fluorescence spectroscopy is more recent than CARS and is less published. Even though this is the case, this introductory review will not be comprehensive. I will likely miss some work and I apologize in advance to those authors. I will not attempt to discuss laser absorption experiments or laser induced fluorescence experiments in the low laser power, i.e., non-saturated, limit. There is much work in the latter area of merit and several important papers on LIF in this conference. 

Taking into account all available data of frequency differences obtained during the course of the matrix measurements, and correcting for different iodine pressures, different iodine cells and different HFS-separations, we derive a combined frequency reproducibility of the two laser systems in the experiment of better than 1.5 0.7 kHz. This is a notable result, given the fundamental differences between the two iodine spectrometers as far as saturated absorption signal detection, laser frequency stabilization and laser set-ups are concerned. 

The second method is based on a time-of-flight measurement. The laser beam is again split, but now both partial beams cross the molecular beam perpendicularly at different positions zi and Z2 (Fig- 4.12). When the laser is tuned to a molecular transition /) k) the lower level /> = (Vi, Ji) is partly depleted due to optical pumping. The second laser beam therefore experiences a smaller absorption and produces a smaller fluorescence signal. In the case of molecules even small intensities are sufficient to saturate a transition and to completely deplete the lower level (Sect. 2.1). 

The analysis of molecular spectra is complicated because of the very large number of lines that is obtained simultaneously in normal excitation or absorption experiments. With narrow-band laser excitation an individual excited rotational-vibrational level can be populated selectively and only the decays originating in the excited state are observed. A similar simphfica-tion in absorption measurements is very desirable. Through the possibility of saturating optical transitions, a certain lower level can be labelled by depleting the population with a laser pump laser). If this laser is switched on and off repetitively, all absorption lines originating in the labelled level will be modulated when induced with a second (probe) laser [9.69, 9.70]. A number of schemes for modulation detection are indicated in Fig. 9.6. Several schemes can be used to ascertain that absorption has ocemred, as discussed... 

However, when the frequency is coincident with the center frequency of the Doppler profile, the weak probe wave interacts with molecules whose absorption has already been reduced by the strong counterrunning wave. Consequently, the absorption of the probe wave has a resonant minimum equal in width to the homogeneous width and centered exactly on the Doppler-broadened absorption line. This method has been demonstrated in experiments using a CO2 laser operating at 10/um and SFe molecules (Basov et al. 1969), and now it is universally accepted in laser saturation spectroscopy. 

J") is partly depleted due to optical pumping. The second laser beam therefore experiences a smaller absorption and produces a smaller fluorescence signal. In case of molecules already small intensities are sufficient to saturate a transition and to completely deplete the lower level (Sect. 7.1). 

The resolution obtainable in laser spectroscopy experiments is limited ultimately by the bandwidth of the multimode laser output. This is not usually important in laser-induced fluorescence or Raman scattering experiments. However, for Brillouin scattering, for saturated absorption, and for two-photon absorption experiments, single-frequency lasers are essential. These are also necessary in interferometry and holography whenever the optical path difference exceeds 30 cm. Many different schemes for obtaining singlefrequency output from lasers have been reported. We shall consider only a few of the most commonly used techniques. We assume throughout that the laser has been constrained to... 

This work was extended by Hansch et at. (1971) using a krypton ion laser, and the apparatus shown in Fig.13.13. In this experiment the iodine absorption cell was placed outside the laser cavity and a standing wave was obtained by arranging two laser beams, each of 10 mW power, to intersect at an angle of less than 2 mrad inside the 20 cm long absorption cell. From equation (13.40) the saturated absorption coefficient when both beams are present in the cell is given approximately by... 

These results are in good agreement with theory and are rather more accurate than previous measurements made by conventional vacuum ultraviolet spectroscopic techniques. Lee et al. (1975), however, have already substantially improved the precision of these experiments by resolving the fine-structure components of the Balmer B line using the saturated absorption technique. It appears, therefore, that the development of atomic and molecular spectroscopy using tunable dye lasers will continue to be very rapid and the reader will of necessity have to consult the current literature to learn the present state of the art. 

The fluorescence signal is linearly proportional to the fraction/of molecules excited. The absorption rate and the stimulated emission rate 1 2 are proportional to the laser power. In the limit of low laser power,/is proportional to the laser power, while this is no longer true at high powers 1 2 <42 j). Care must thus be taken in a laser fluorescence experiment to be sure that one is operating in the linear regime, or that proper account of saturation effects is taken, since transitions with different strengdis reach saturation at different laser powers. 

Plotted curves illustrating this relation, Fig. 5, resemble very much the curves of Fig. 3. Consequently, one cannot infer from a measured intensity or energy saturation curve reliable values of molecular data without additional information, as for instance an independent measurement of ksr Another possibility is a measurement of the temporal characteristics of the bleaching as demonstrated in an experiment by Hercher et al. 14>. These authors bleached a thoroughly degassed solution of metal-free phthalocyanine in 1-chloronaphthalene by a ruby laser pulse (694.3 nm) of about 59 nsec pulse width. At the same time they measured the absorption at 632.8 nm using a He-Ne-laser, and the result of this measurement is shown in Fig. 6. It clearly demonstrates that the sample was almost completely bleached even before the laser pulse reached its maximum intensity, and that almost all of the molecules were stored in the triplet state because the transmission did not decrease with the fall of the laser intensity for at least 100 nsec. A small residual absorption indicates triplet-triplet absorption. 

---