1. Introduction

Since the first demonstration of an efficient optically pumped alkali laser in 2003 [1], and after the first Diode Pumped Alkali Laser (DPAL) demonstration in 2005 [2], significant progress in DPAL development and power scaling was achieved. The most impressive results were achieved with Cs and Rb DPALs [3–5] including the demonstration of 1 kW output power with optical efficiency about 50% in CW regime for Cs DPAL [6]. Also, power scaling experiments with multiple diode laser pump sources [7–9] were performed, including experiments with transverse pumping [8] and an unstable cavity [9]. On the other hand, the potassium (K) DPAL, has not been extensively studied yet, in spite of its several advantages compared to Cs and Rb lasers. In particular, the K laser has higher quantum efficiency (99.6%) and can operate with low pressure He buffer gas [10], while Cs and Rb lasers can operate only with hydrocarbon buffer gases or with high pressure (several atmospheres) helium buffer gas. Both of these approaches have their disadvantages compared to the K laser. Hydrocarbon buffer gases can chemically react with alkali vapor and contaminate the gain medium with the reaction products (e.g. soot). High pressure Rb-He laser requires elevated temperatures and higher pump intensities that creates additional technical and fundamental problems especially when scaling to higher power levels.

There are only a few publications devoted to the study of potassium vapor lasers, many of which use so called “surrogate” (means other than diode laser) pump sources (see, for example [10, 11],). Only recently, a first demonstration of a CW diode pumped potassium laser buffered by atmospheric pressure helium was performed [12], and the slope efficiency achieved was not very high (about 25%). In this paper we present results of our experiments with a K DPAL operating in pulsed mode.

The K laser operates in a three level scheme (see Fig. 1). The optical pump source excites the D2 line of potassium atom (766 nm) and lasing occurs on the D1 line (770 nm), which is only 57.7 cm⁻¹ from the D2 line. To create a population inversion on the 4P_1/2 → 4S_1/2 transition, a fast (compared to the 4P_3/2 radiative lifetime of 27 ns) population transfer from 4P_3/2 to 4P_1/2 energy states is provided by collisional mixing of these states by a buffer gas. The K laser has a quantum efficiency of 99.6%, but the small separation of the pumped (4P_3/2) and lasing (4P_1/2) energy levels decreases the population inversion on the lasing transition and, hence, leads to lower gain coefficient of the active medium compared to Cs and Rb vapor lasers. Laser medium with a lower gain coefficient require a higher quality laser cavity, higher pump intensity and is more sensitive to intracavity losses. All parameters required for the small signal gain calculation for the K DPAL are provided in [12] and the calculated value is about 0.18 cm⁻¹, which is much smaller than the ones for Cs (4.5 cm⁻¹) and Rb (1.1 cm⁻¹). This means that the lasing threshold for the K DPAL has to be higher than the one for Cs and Rb DPALs.

 

Fig. 1 Potassium laser energy level diagram.

Download Full Size | PDF

2. Experimental apparatus and results

A diagram of the K DPAL is presented in Fig. 2. We used an L-shape laser cavity with longitudinal pumping of the gain medium, similar to the one described in our previous experiments [10]. The 1 cm long K vapor cell had AR coatings on both sides of the windows to minimize losses in the cell for both the laser and the pump wavelengths. The cell was filled with metallic potassium and 600 torr of helium at room temperature before being sealed.

 

Fig. 2 Diagram of the experimental setup (PBS – polarizing beam splitting cube).

Download Full Size | PDF

The sealed cell was assembled inside an oven that could control the cell temperature while keeping its windows approximately 5°C higher than the cell body. The cell optimal operating temperature of 180°C was determined experimentally by measuring laser slope efficiency at different temperatures. The small-signal absorption of the pump radiation in the K vapor cell at this temperature is close to 100% (see Fig. 3).

 

Fig. 3 Experimentally measured small signal transmission of the K vapor cell for the pump wavelength at different temperatures. Pump intensity is below several W/cm².

Download Full Size | PDF

The K vapor gain medium was pumped by a diode laser stack operating at 766 nm. The stack was operated in pulsed mode with a triangular pulse shape and duration of 30 μs and repetition rate 100 Hz (duty cycle is 0.3%). The maximum peak power of the pump delivered into the gain cell in these experiments was approximately 50W. The stack emission bandwidth was narrowed to less than 20 GHz (FWHM) and centered at 766 nm using technique similar to that described in [13]. This linewidth is reasonably well matched to the absorption line of potassium vapor broadened by the He buffer gas, which has an effective pressure at 180 C of 960 Torr and corresponding pressure broadened potassium absorption line is about 19 GHz.

The stack’s output beam had a rectangular cross sections with a vertical to horizontal sides ratio of about 4 to 1. To correct the beam and make it close to square before focusing into the gain medium, we used a system of cylindrical and spherical lenses with total focal length of about 20 cm. The beam was focused into the center of the K vapor cell and aligned collinearly with the laser cavity axis to provide longitudinal pumping. Such a combination of cylindrical and spherical focusing lenses provided a satisfactory pump beam size matching to the laser cavity mode size in the gain medium. The polarization of the pump beam was orthogonal to the laser beam polarization making it possible to separate the pump and lasing beams using polarization beam splitter (PBS). The stable 40 cm long laser resonator was constructed of a 50 cm radius concave mirror with 99.9% reflection at 770 nm and 766 nm and flat output coupler with an experimentally-optimized 60% reflectivity at 770 nm (see Fig. 4).

 

Fig. 4 Optimization of the output coupler reflectivity (Roc) at the cell temperature 180 C.

Download Full Size | PDF

The dependence of the K DPAL output peak power with respect to the pump peak power is presented in Fig. 5. The lasing threshold appeared to be about 22 W or approximately 4 kW/cm². The slope efficiency was 52% and the total optical-to-optical conversion efficiency (power out at 770 nm over power in at 766 nm)was about 31%. The maximum output power obtained was about 16 W. The demonstrated slope and optical-to-optical efficiencies in pulsed operation are significantly higher than those obtained in the same system operating in CW mode [12]. The differences show how effects such as thermal lensing and ionization can limit laser performance. The results of additional research aimed at studying the contribution of these limiting effects and possible ways to mitigate them will be published in separate paper.

 

Fig. 5 Potassium DPAL peak output power dependence on pump power demonstrating 52% slope efficiency (diamonds – experimental data, solid line – trendline).

Download Full Size | PDF

In spite of the relatively high Potassium DPAL slope efficiency achieved in this experiment (52%), it is still far from the maximum possible value determined by the quantum efficiency of the gain medium (99.6%). There are several limiting factors to consider. The dominant reason for the lower than theoretical slope efficiency is the mismatch between the pump beam size (about 0.5 mm x 0.8 mm FWHM) and the cavity mode size (about 0.5 mm FWHM). Additionally, there are some losses on all optical elements of the laser cavity, which reduce the efficiency of the system.

3. Conclusion

We have demonstrated a hydrocarbon free potassium DPAL operating in pulsed mode with a slope efficiency of 52% and optical conversion efficiency of 30% pumped by a narrowbanded diode laser stack. These numbers are significantly higher than demonstrated in [12] for the similar system but operating in CW mode (25% and 3% correspondingly) that show possible contribution of limiting effects, such as thermal lensing and ionization, when operating in CW.