In recent years there have been concerted efforts to develop high energy diode-pumped alkali vapor lasers (DPAL) [1–4]. These hybrid gas phase/solid state laser systems offer possibilities for constructing high-powered lasers that have high beam quality. Considerable progress has been made, but there are technical challenges resulting from the chemically aggressive nature of the alkali metal vapors. In the presence of an intense light field, these reactants destroy the inner surfaces of the gain cell windows.

Reactions with the collision partner that is used to induce spin-orbit relaxation between the ${}^{2}P$ levels also cause problems. In a typical DPAL, He buffer gas is used to pressure broaden the absorption lines to improve the spectral overlap with the pump laser radiation. Unfortunately, He has a low cross sections for alkali metal spin-orbit relaxation, and the heavier rare gases are not effective [5]. As the use of very high gas pressures will degrade the beam quality of the laser, the most commonly used approach has been to add a second buffer gas component that induces rapid spin-orbit relaxation, without significantly quenching the ${}^{2}P$ levels. Small hydrocarbons such as methane and ethane have proven to be effective. The difficulty with these reagents is that they react with the alkali metals to produce hydrides (“laser snow”) [6] and carbonaceous deposits (particularly on the windows, where the reactions are facilitated by photoexcitation). Solutions to these problems are currently being investigated, such as the use of gas curtains to protect the windows.

There is interest in using elements other than the alkali metals for this type of laser, in order to circumvent the chemical problems and expand the range of pump and lasing wavelength that can be employed. The ${np}^5(n+1)s$ configurations of rare gas atoms (Rg) support metastable $(n+1)s{[3/2]}_2$ (${}^3{P}_2$ in LS notation) and $(n+1)s^{\prime }{[1/2]}_0$ (${{}^{3}P}_0$) states that are readily populated under mild electric discharge conditions [7]. Strongly allowed optical transitions connect these levels to states of the $n{p}^5(n+1)p$ manifold [8]. A laser scheme that mimics the alkali vapor system involves excitation of the $(n+1)p{[5/2]}_3\leftarrow (n+1)s{[3/2]}_2$ transition and lasing on the $(n+1)p{[1/2]}_1\to (n+1)s{[3/2]}_2$ line, as shown in Fig. 1.

 

Fig. 1. Energy level scheme for an optically pumped ${\mathrm{Rg}}^{*}$ laser.

Download Full Size | PDF

The $p\leftarrow s$ transitions of the metastable rare gas atoms (${\mathrm{Rg}}^{*}$) are entirely analogous to those of the alkali metals. They have similar transition probabilities, upper state radiative lifetimes, lasing wavelengths, pressure broadening coefficients, and ionization potentials [8]. Advantages of using ${\mathrm{Rg}}^{*}$ for the lasing medium are that heating of the system is not required and rare gas collision partners (He in particular) can be used for the spin-orbit relaxation step [9]. Hence, the system is chemically inert. In the present work we have investigated the possibilities of using optically pumped ${\mathrm{Ar}}^{*}$, ${\mathrm{Kr}}^{*}$, and ${\mathrm{Xe}}^{*}$ as atomic gas lasers. Measurements that demonstrate both optical gain and lasing are described.

In order to achieve rapid collisional energy transfer in these experiments, it was necessary to use a gas discharge that could operate at relatively high pressures of He. For proof-of-principle studies we used the discharge cavity of a commercial excimer laser (Lambda Physik EMG 102) to produce discharges in $\mathrm{Rg}/\mathrm{He}$ mixtures and pure Ar. Observations of gain and lasing from optically pumped ${\mathrm{Rg}}^{*}$ were carried out using the configuration shown in Fig. 2. The pulsed discharge was operated at its preset minimum voltage of 18 kV. The electrodes in this cavity are separated by approximately 2.5 cm, and have a length of 80 cm. The output coupler, M1, was an uncoated ${\mathrm{MgF}}_2$ window. For gain measurements, M2 was also an uncoated window. To observe lasing, M2 was replaced by a total reflector and the probe laser was not used.

 

Fig. 2. Apparatus used to study gain and lasing of optically pumped ${\mathrm{Rg}}^{*}$ atoms.

Download Full Size | PDF

Measurements were conducted with static fills of the gas mixtures. The cavity was first pumped down to 0.5 mbar and then flushed using pure He. To produce gas mixtures, a small quantity of the working gas (typically around 30 mbar of Ar, Kr, or Xe) was added to the chamber. The system was then filled with He to the desired total pressure (200–2000 mbar). In one series of experiments the discharge cavity was filled with pure Ar.

Measurements were conducted using a repetition frequency of 10 Hz. Following the discharge pulse (nominally 10 ns duration), the ${\mathrm{Rg}}^{*}$ metastables were excited by the 10 ns pulses from a tunable optical parametric oscillator (OPO) (LaserVision). For the wavelengths of interest, the OPO yielded pulse energies of around 0.25 mJ with a linewidth of 30 GHz. The pump beam diameter inside the discharge cavity was 0.4 cm. An electronic delay generator was used to control the relative timing of the discharge, OPO pump, and dye laser probe pulses. The delay was required to permit optimization of the metastable number density. Some care was needed for this adjustment, as high number densities resulted in strong absorption of the pump radiation. When the optical thickness was too great, only a limited segment of the optical path could be pumped to inversion.

A pulsed tunable dye laser (Lambda Physik CompexPro 201/FL3002) was used to test for optical gain. This instrument had a linewidth of 10 GHz and a pulse duration of approximately 10 ns. The dye laser output was attenuated to avoid saturation. As shown in Fig. 2, the horizontally polarized pump beam and the vertically polarized probe beam were counter-propagated through the discharge cavity. On exiting the cavity, the probe beam was directed to a photodiode by a Glan—Taylor polarizer.

To observe lasing, the rear cavity window (M2 in Fig. 2) was replaced by a total reflector. The light exiting the cavity was directed by the polarizer to a monochromator/photodetector combination which was used to record the output spectrum. A 0.20 m monochromator (Bausch & Lomb) was used for these measurements. The output power was measured directly after the polarizer, using a commercial power meter (Ophir model 3A-P-V1-ROHS). The lasing spot was also imaged using an IR viewing card.

The first experiments examined optical pumping of ${\mathrm{Kr}}^{*}$ in $\mathrm{Kr}/\mathrm{He}$ mixtures. To test the apparatus, the $5p{[3/2]}_1\leftarrow 5s{[3/2]}_2$ transition was excited (769.67 nm, vacuum), and gain was measured on the $5p{[3/2]}_1\to 5s{[3/2]}_1$ line (830.04 nm). As these transitions have a common upper level, measurements were made for the lowest possible total pressure (200 mbar, set by the interlock on the power supply) to minimize collisional relaxation. Single pass gains in excess of 100 were readily observed.

The objective of the second series of experiments was to see if gain could be observed using the DPAL analog pumping scheme (Fig. 1, pumping at 811.51 with lasing at 893.12 nm). The initial investigation was carried out using 27 mbar of Kr in 973 mbar of He, and the results are shown in Fig. 3. Baseline signal levels were established, in the absence of the discharge, for the probe laser alone and the combination of the pump and probe lasers (left hand segment of the trace). The discharge was then turned on, resulting in a further slight increase of the baseline. Finally, gain was observed when the probe laser was scanned over the $5p{[1/2]}_1-5s{[3/2]}_2$ line. When on-resonance, the probe laser was entirely absorbed when the pump laser was blocked or tuned off-resonance.

 

Fig. 3. Demonstration of gain on the $5p{[1/2]}_1-5s{[3/2]}_2$ transition of Kr following optical pumping of the $5p{[5/2]}_3-5s{[3/2]}_2$ line. The pump and probe lasers were fired 7 μs after the discharge pulse.

Download Full Size | PDF

The population inversion achieved in the second experiment relied on rapid collisional energy transfer between the $5p{[5/2]}_3$ and $5p{[1/2]}_1$ levels. To explore the effect of the pressure on the inversion kinetics, gain measurements were performed for Kr (27 mbar) in He at total pressures ranging from 200–2000 mbar. Results from these measurements are presented in Fig. 4, where the pump and probe lasers were tuned to the $5p{[5/2]}_3-5s{[3/2]}_2$ and $5p{[1/2]}_1-5s{[3/2]}_2$ lines, respectively. The traces are time-resolved signals, with preprobe pulse data that indicate undetectable light levels (on this sensitivity scale) from the discharge. Gain is apparent for pressures above 600 mbar. The gain increased with increasing pressure up to approximately 1400 mbar, reaching a plateau at higher pressures. This behavior is readily understood in terms of the competition between radiative and collisional relaxation of the $5p{[5/2]}_3$ level.

 

Fig. 4. Dependence of the ${\mathrm{Kr}}^{*}$ $5p{[1/2]}_1-5s{[3/2]}_2$ gain on He pressure. These time-resolved traces were recorded for 27 mbar of Kr in He, with optical pumping of the $5p{[5/2]}_3-5s{[3/2]}_2$ transition. Going from the lowest to the highest trace, the total pressures for these measurements were 200, 400, 600, 700, 800, 900, 1000, 1200, 1400, 1600, 1800, and 2000 mbar. The delay between the discharge and laser pulses was optimized for each pressure, within the range 5–25 μs.

Download Full Size | PDF

Lasing of the ${\mathrm{Kr}}^{*}/\mathrm{He}$ mixtures was observed when the rear window of the discharge cavity was replaced by a total reflector. The output, viewed using an IR phosphor card approximately 4 m away from the polarizer, was a well-defined beam. For 27 mbar of Kr in 973 mbar of He, the output was completely dominated by the $5p{[1/2]}_1\to 5s{[3/2]}_2$ laser line. Pumping of higher energy levels of the Kr $4p^{5}5p$ manifold was also examined. Gain was observed on the $5p{[1/2]}_1\to 5s{[3/2]}_2$ line for excitation of the $5p{[5/2]}_2$, $5p{[3/2]}_1$ and $5p{[3/2]}_2$ levels.

Following these observations, preliminary experiments were conducted to test for optically pumped lasing of ${\mathrm{Ar}}^{*}$ and ${\mathrm{Xe}}^{*}$. For both $\mathrm{Ar}/\mathrm{He}$ and $\mathrm{Xe}/\mathrm{He}$ mixtures (27 mbar Rg in 973 mbar He), excitation of $(n+1)p{[5/2]}_3$ yielded lasing on the $(n+1)p{[1/2]}_1\to (n+1)s{[3/2]}_2$ transitions. The pump and lasing wavelengths were 811.75 and 912.55 nm for ${\mathrm{Ar}}^{*}$ and 882.18 and 980.24 nm for ${\mathrm{Xe}}^{*}$. As Ar is an inexpensive reagent, we were able to investigate the effect of using high mole fractions in the gas mixture. It was found that lasing could be achieved using pure Ar at a pressure of 1000 mbar.

The laser performance characteristics of the $\mathrm{Rg}/\mathrm{He}$ mixtures were all quite similar for $\mathrm{Rg}=\mathrm{Ar}$, Kr, and Xe. Power measurements were made for Ar (27 mbar) in He (973 mbar). With a pump pulse energy of 0.26 mJ, the lasing output energy measured after the polarizer was 0.017 mJ. As the laser cavity did not include any polarizing elements, the output power would be twice this value (at least, as this estimate neglects the reflection and scattering losses of the polarizer). An output energy of 0.034 mJ corresponds to a photon conversion efficiency (pump to lasing) of 13% and a peak output intensity of approximately $27\mathrm{kW}{\mathrm{cm}}^{-2}$. This is equivalent to $1.6\times {10}^{13}\mathrm{photons}{\mathrm{cm}}^{-3}$, which implies that the total ${\mathrm{Ar}}^{*}$ number density was greater than $4\times {10}^{13}{\mathrm{cm}}^{-3}$. Note that the long cavity used in these measurements was far from optimal, as the optical round trip time was comparable to the pump pulse duration.

A matter of practical interest is the discharge power that would be needed to sustain the required ${\mathrm{Rg}}^{*}$ $p{[3/2]}_2$ number density for a CW laser. We have estimated this for a volume of $10{\mathrm{cm}}^3$ and a ${\mathrm{Kr}}^{*}$ number density of ${10}^{13}{\mathrm{cm}}^{-3}$. With the conservative estimates of an effective ${\mathrm{Kr}}^{*}$ lifetime of 10 μs, and a discharge excitation efficiency of 1%, the required CW discharge power would be 1.6 kW.