1 Introduction

A development of mid-IR laser sources is of particular interest in spectroscopy, photochemistry, isotope separation, and other fields. Carbon monoxide laser (CO laser) has an extremely broad spectrum numbering about one thousand spectral lines in fundamental (wavelength λ ~ 4.7–8.7 µm) [1, 2] and first-overtone (λ ~ 2.5–4.2 µm) [3] spectral bands. It makes CO laser very attractive for different applications, such as, e.g., gas analysis and atmosphere sensing [1]. Therefore, nowadays different teams [4,5,6] develop compact and reliable RF discharge CO lasers including commercial ones [6].

Often in practice, short radiation pulses with high peak power are needed. In particular, they are required for a frequency conversion of CO laser radiation in nonlinear crystals, and allowed us to densely cover wide domain of mid-IR range from 2.5 to 17 µm [7]. Previously, we reported [8] on the development of a compact cryogenic Q-switched CO laser pumped by a slab RF discharge. The laser operated on fundamental band vibrational–rotational transitions of CO molecule and simultaneously emitted about one hundred spectral lines within wavelength range of 5–7 µm. The laser pulses had microsecond duration and kW-range peak power at 100 Hz pulse repetition rate. In our paper [9], a broadband sum frequency generation in ZnGeP2 crystal was demonstrated for the first time for a compact cryogenic Q-switched slab RF discharge CO laser with frequency conversion efficiency up to ~ 8%. To increase the conversion efficiency, an improvement of CO laser radiation parameters is required: pulse duration shortening and peak power increasing.

It should be noted that our studies [8, 9] were carried out only using high-oxygen content CO:O2:He gas mixture as the laser active medium, which was really optimal for repetitively pulsed free-running laser operation but likely not for Q-switching mode. CO laser peak power was about 2 kW [8, 9], but the paper [8] reported that it was increased up to 3 kW for some other conditions. Therefore, the objective of this paper was peak power enhancement of the cryogenic slab RF discharge Q-switched CO laser by means of parametric study of its output characteristics for different gas mixture compositions and RF pumping parameters.

2 Experimental setup

Our experiments (an optical scheme is presented in Fig. 1) were carried out with the repetitively pulsed Q-switched cryogenic slab RF discharge CO laser described in detail in [4]. An electrode system of the laser consisting of two hollow brass electrodes of 400 mm length and 16 mm height was placed inside the laser chamber of ~ 15 l internal volume. The RF discharge gap in the experiments was 5 mm. The electrode system was cooled by liquid nitrogen, so the gas mixture inside the active RF discharge volume was at the temperature of ~ 100 K, while the laser chamber walls were at room temperature. The laser was operated in quasi-sealed-off mode (without any gas mixture forced exchange).

Fig. 1
figure 1

Optical scheme of the experiment (see the text for details)

Full size image

The laser chamber 1 was covered by CaF2-window 2 installed at the Brewster’s angle to the cavity axis. The double-pass V-type laser cavity was formed by flat rear mirror 3 (Au-coated fused silica substrate), rotating flat Al-mirror 4 (6000 rpm), retro-reflector 5 (180 cm radius of curvature Au-coated fused silica substrate) and output coupler 6 (plane-parallel silicon plate).

A diaphragm 7 of 5 mm in diameter was placed in front of the output coupler. An alignment of the optical scheme was performed by He–Ne laser beam 8 injected through flat mirror 9 with a small hole. Parts of the CO laser output radiation splitted by flat BaF2 plates 10, 11 were directed to power meter 12 (12A-SH, Ophir Optronics Solutions Ltd) and to photo-detector 13 (PEM-L-3, VIGO system S.A.), respectively. The main part of the laser radiation was focused by CaF2 lens 14 onto the entrance slit of spectrometer 15 based on monochromator (MDR-3, LOMO JSC). RF power supply (CESAR 6010, Advanced Energy Industry, Inc.) was synchronized with rotating mirror 4 by additional electro-optical system (not presented in Fig. 1) and provided repetitively pulsed RF discharge excitation of the laser active mixture in desired time moments with pulse repetition rate of 100 Hz. RF discharge power pulse had a rectangular shape of sub-millisecond duration and amplitude of 0.75 kW. Thus, a single microsecond laser pulse corresponded to a single sub-millisecond RF discharge pulse.

3 Experimental results

First experiments were performed with three gas mixtures: CO:O2:He = 1:0.3:10; CO:O2:N2:He = 1:0.3:3:7; CO:O2:Ar:He = 1:0.3:3:7 at the same total pressure of 22 Torr, that corresponded to gas density of ~ 0.09 Amagat. It should be noted that to have a long laser operating cycle (until lasing failure due to the active medium degradation), all gas mixtures have to contain anomalously large amount of oxygen (up to 50% with respect to the CO concentration) [10]. We studied gas mixtures with oxygen concentration up to 30% because higher percentage of oxygen decreased CO laser power [10]. The CO laser output characteristics for different gas mixtures were measured at the same RF discharge pump conditions: pulse duration τ = 0.5 ms and power amplitude of 0.75 kW. Waveforms of the CO laser output are presented in Fig. 2a, that demonstrates the laser pulse duration to be about one microsecond (FWHM). The active medium composition variation affected the laser peak power rather than low-power pulse tail.

Fig. 2
figure 2

Waveforms (at optimal Δt = 0.7 ms) (a) and peak power versus time delay Δt (b) of CO laser pulses for active medium compositions: 1—CO:O2:N2:He = 1:0.3:3:7; 2—CO:O2:He = 1:0.3:10; 3—CO:O2:Ar:He = 1:0.3:3:7

Full size image

The dependence of Q-switched CO laser radiation peak power P on time delay Δt between RF power pulse beginning and Q-switching moment is presented in Fig. 2b. In general, the moment of maximum population inversion in the CO laser active medium in conditions of pulsed excitation depends on both pumping parameters and active medium composition, because it is determined by the balance between rates of vibrational–vibrational exchange and vibrational–translational relaxation in the active medium (see, for example [1]). In our experiments the optimal time delay Δt was about the same for all studied gas mixtures, namely 0.7 ms (Fig. 2b). This can be explained by the fact that the time required for CO molecules to go up the ladder of vibrational levels and to form a “plateau” in the vibrational distribution function (about 0.1 ms [1]) is far less than the RF discharge pump pulse duration. In other words, the active medium excitation in our experiments was quasi-CW. When the discharge pulse ended, the active medium started to be cooled diffusively down by cold electrode system and the active medium gain slightly increased. At the same time, the population inversion decreased due to vibrational–translational relaxation. Thus, the optimal time delay Δt (~ 0.7 ms for all studied gas mixtures) was determined by a competition between these two opposite processes in conditions of identical energy of RF pump pulse and at the same concentrations of CO molecules in all the three active gas mixtures.

Several papers (see, e.g., [11, 12]) considered gas mixtures containing argon as appropriate for a CO laser operation. Maximal ever-known CO laser efficiency (63 ± 15%) was obtained in the gas mixture CO:Ar = 1:10 [12]. In our experiments, a partial replacement of He by Ar in the initial three-component gas mixture CO:O2:He = 1:0.3:10 resulted in laser peak power decrease probably due to the higher heating of the active medium by repetitively pulsed RF discharge pumping (because the thermo-conductivity of Ar is about 6 times lower than that of He). It should be also noted that RF discharge was not at all ignited in gas mixture CO:O2:Ar = 1:0.3:10 for our experimental conditions.

The partial replacement of He by N2 in the initial gas mixture resulted in noticeable enhancement of the laser peak power, which was observed in the next experimental series, where we studied an influence of nitrogen on the laser characteristics. The dependence of Q-switched CO laser peak power on nitrogen content X is presented in Fig. 3. In this case, the active medium composition was CO:O2:N2:He = 1:0.3:X:(10-X). Maximal laser pulse peak power for gas mixture CO:O2:N2:He = 1:0.3:X:(10-X) was obtained at X = 1.2. In this case, the ratio between oxygen and nitrogen concentrations was 1:4, which was approximately similar to that of the atmospheric air. Therefore, the latter could be considered as a simple replacement of nitrogen and oxygen for cryogenic CO laser active medium. The atmospheric air is a complicated gas mixture, but CO2 and H2O are its only components (certainly, except for nitrogen and oxygen) with noticeable concentrations which are frozen at cryogenic internal elements of the laser setup. Indeed, the CO laser peak power measured for the air gas mixtures was close to the result obtained with pure gases mixture. Results for gas mixtures CO:Air:He = 1:1:9.3 (corresponding to X = 0.78) and CO:Air:He = 1:1.5:8.8, (corresponding to X = 1.17) are presented in Fig. 3.

Fig. 3
figure 3

CO laser peak power for CO:O2:N2:He = 1:0.3:X:(10-X), CO:Air:He = 1:1:9.3 (X = 0.78) and CO:Air:He = 1:1.5:8.8 (X = 1.17) gas mixtures versus nitrogen content X. τ = 0.5 ms, Δt = 0.7 ms

Full size image

To increase the Q-switched CO laser peak power we slightly increased specific energy input to the active medium by increasing RF pump pulse duration τ and maintaining RF pulse power at the constant level (0.75 kW). The CO laser peak power (for the best gas mixture CO:O2:N2:He = 1:0.3:1.2:8.8) versus time delay Δt for different RF pump pulse duration τ is presented in Fig. 4.

Fig. 4
figure 4

CO laser peak power versus time delay Δt for different RF pump pulse duration τ. Gas mixture CO:O2:N2:He-1:0.3:1.2:8.8, τ = 0.5 ms (1); 0.6 ms (2); 0.7 ms (3); 0.8 ms (4)

Full size image

The enhancement of RF pump pulse duration (and, respectively, RF pump pulse energy) lead to moderate growth of maximal laser peak power with its subsequent saturation at τ longer than 0.6 ms. Maximal laser peak power 2.6 kW was obtained for τ = 0.7 ÷ 0.8 ms at Δt = 0.9 ÷ 1.0 ms. It should be noted that temporal position of the laser peak power maximum was shifted about synchronously with changing τ in such a way that the time interval between it and the end of RF pump pulse remained constant (~ 0.2 ms). This fact confirmed that optimal time delay Δt was determined by a competition between active medium cooling by cold electrode system and population inversion decrease due to vibrational–translational relaxation.

In the next series of the experiments, we increased gas pressure of the active medium and a speed of the rotating mirror (see Fig. 5). Active medium composition was CO:Air:He = 1:1.5:8.8. Maximal laser pulse peak power was obtained at the active medium gas pressure being in the interval of 30–37 Torr (gas density ~ 0.12–0.15 Amagat). For higher pressure, optimal discharge duration τ and time delay Δt were 0.6 and 0.7 ms, respectively (versus τ = 0.7 ÷ 0.8 ms at Δt = 0.9 ÷ 1.0 ms for gas pressure of 22 Torr), due to, probably, faster V–T relaxation. Further gas pressure increase caused an irregularity of the discharge in used gas mixtures. A faster speed of the rotating mirror also allowed us to slightly decrease pulse duration (Fig. 5b) and increase the laser pulse peak power. The maximal peak power was obtained at the rotating mirror speed of 7800 rpm (pulse repetition rate of 130 Hz), which was limited by an ability of the device. Peak power reached 4 kW which was 1.3 times higher than that of Ref. [8].

Fig. 5
figure 5

CO laser peak power versus the active medium pressure (a) and laser pulse waveforms (b) for pulse repetition rate of 130 Hz (1) and 100 Hz (2)

Full size image

We also measured the spectra of Q-switched slab CO laser for different active medium compositions. As an example, the spectra measured at τ = 0.5 ms, Δt = 0.7 ms for the laser operating on the gas mixture CO:Air:He-1:1:9.3 are presented in Fig. 6.

Fig. 6
figure 6

Output spectra of Q-switched CO laser with CO:Air:He active medium composition. a Uncoated Si output coupler of ~ 50% reflectivity; b dielectric coated CaF2 plate of ~ 90% reflectivity

Full size image

The first one (Fig. 6a) consisted of about 52 rotational–vibrational components in the wavelength range of 5.05–6.0 µm and was obtained using the output coupler of ~ 50% reflectivity (a plane-parallel uncoated Si plate). The spectra obtained with other gas mixtures and RF pulse durations were pretty much the same; therefore, special spectral features of the laser using different gas mixtures were not found. The spectrum of Q-switched slab CO laser may be varied by the time delay Δt [8] or, additionally, by changing the reflectivity of laser cavity output coupler. The spectrum measured for the same conditions with ~ 90% output coupler reflectivity (a dielectric coated CaF2 plate) (Fig. 6b) consisted of 64 components in the wavelength range of 4.95–6.0 µm. Thus, higher Q-factor of the laser cavity allowed us to expand CO laser spectrum in its short-wave part. Moreover, the changing of output coupler resulted in lasing on different rotational components in the long-wave part of the spectrum. It may be a result of complicated Fabry–Perot resonator, when we have a comparable reflection from front and back surfaces of the Si output coupler (Fig. 6a) (compare with Fig. 6b, where the spectrum has more regular vibrational–rotational structure).

At the end of the paper we would discuss about scalability of such a laser system. Investigated CO laser was initially created for CW or repetitively pulsed operation with stable or hybrid (unstable-waveguide) laser resonator in which mirrors are placed near the electrodes inside the laser chamber. To investigate Q-switched CO laser mode, Brewster window and some external elements of the laser resonator were installed. Such a scheme has significant additional optical losses. The placement of the optical scheme (including rotating mirror) inside the laser chamber is not a very simple technical procedure, but it is possible and will increase the power and performance of the system. In this case we estimate the laser pulse peak power to be 10–12 kW and higher. Another way to increase radiation power of Q-switched repetitively pulsed cryogenic slab RF discharge CO laser is an active medium enlargement or other kind of a Q-switching technic (electro-optic or acousto-optic modulation), i.e., the development of the new installation of special design suitable for Q-switched operation.

4 Conclusions

Multi-parametric optimization of the repetitively pulsed cryogenic slab RF discharge CO laser operating in Q-switched mode allowed us to increase its output peak power up to 4 kW. The partial replacement of He by N2 in the initial gas mixture resulted in noticeable enhancement of the laser peak power. Maximal peak power was obtained with gas mixture CO:O2:N2:He = 1:0.3:1.2:8.8, where O2:N2 ratio was 1:4 as approximately in the atmospheric air. It was shown that the atmospheric air can be applied instead of pure nitrogen and oxygen without any laser power losses. It was found also that optimal time delay for Q-switching moment (maximal active medium gain) was after the end of RF discharge pulse and not dependent on the active gas mixture composition.

An influence of Ar, N2 and the air additives to CO:O2:He active gas mixture on spectral and temporal parameters of the cryogenic repetitively pulsed slab RF discharge Q-switched CO laser appeared to be insignificant for long (≥ 0.5 ms) RF discharge pump pulse, when characteristic times of formation of CO molecule vibrational distribution function are significantly shorter.

In general, the effect of variations in the active gas mixture composition on the energy (power) characteristics of the laser is not so obvious and requires, along with detailed experimental studies, model calculations under conditions that are as close as possible to experimental ones including timing of RF discharge pumping and the cavity Q-factor, diffusive heat exchange between gas and cooled walls of the discharge gap, etc.