Few ns Pulse Duration of Gain-Switched Ho:YAG Laser Pumped by an Active Q-Switched Tm:YLF Laser

Abstract

This paper describes a gain-switched Ho:YAG laser that emitted at 2089 nm, driven by an actively Q-switched Tm:YLF laser as the pumping source. The laser attained few ns short pulse durations with high energies at controlled repetition rates due to the active Q-switch pump source. Using the gain-switch method, stable short pulse durations ranging from 3.5 to 7.2 ns, with an energy per pulse of 0.4 to 0.52 mJ, were achieved at repetition rates of up to 2.5 kHz. This design can have significant advantages in various fields, where accuracy in the repetition rate is essential and a passive Q-switch cannot be implemented due to its accuracy limitations, including sensing, LIDAR, medical procedures, and material processing.

1. Introduction

Pulsed lasers with a spectral range of 2 μm are used in a variety of applications. This spectral range is ideal for remote sensing applications, like wind and range measurements; medical surgery; fine material processing [1,2,3]; and as a pumping source for mid-infrared lasers [4]. The 2 μm range is “eye-safe” [5,6], which enables its use in populated areas, and has a high atmospheric transmittance of more than 80%/km within the 2–2.6 μm window [7]. This is critical for remote sensing applications, such as laser rangefinders and laser radar (LiDAR). Moreover, the 2 μm range is ideal for medical surgery and fine material processing due to its high absorption coefficient in water and plastics.

Some applications require the use of pulsed lasers with a wide range of repetition rates, short pulse durations, and high power. Accurate repetition rates, which enable synchronization of the laser pulse with a receiver or a camera, is necessary for autonomous safety devices, LiDAR, 3D imaging, laser tomography, time imaging, and spectroscopy. For instance, time-of-flight LiDAR operates by sending a short pulse over an object or scene, and a synchronized detector calculates the distance from the system to the object by measuring the time of the back-reflected laser signal from the object to the detector. The pulse duration is directly related to the measurement accuracy; shorter pulse durations result in a higher measurement resolution [8]. Short pulses are also advantageous in applications involving light–matter interactions [9]. The shorter the pulse duration, the less heat transfer from the irradiated area to the surrounding areas, and thus, less thermal damage.

Short-pulse-duration lasers are typically divided into the picosecond and the few nanosecond ranges. Lasers in the picosecond range utilize the mode-locking method and are generally characterized by higher costs and increased complexity. In contrast, lasers in the nanosecond range can be achieved using the comparatively more affordable and simpler Q-switching method. Several studies focused on Ho-doped and Tm-doped crystals to obtain short-pulse-duration and high-energy pulsed lasers operating in the 2 μm wavelength region. Currently, most studies on high-energy lasers with pulse duration in the tens of nanoseconds with Ho-doped or Tm-doped crystals have been performed using the Q-switching method by inserting a saturable absorber (SA) or an acousto-optical modulator (AOM) into the cavity [10].

Lan et al., 2006, reported an 11 ns pulse duration and a 25 μJ energy per pulse at 14.8 kHz PRF and a wavelength of 2089 nm using a ceramic Ho:YAG as the gain medium and ${\mathrm{Cr}}^{2+}$:ZnS as an SA [11]. Yuan et al., 2020, reported a 13.3 ns pulse duration and 1.02 mJ energy per pulse at a 10.1 kHz PRF and a wavelength of 2089 nm using the same gain medium and SA [10]. Dai et al., 2012, reported a 132 ns pulse duration and 7.6 mJ energy per pulse at a 100 Hz PRF and a wavelength of 2090.9 nm using a Ho:YAG crystal as a gain medium and an AOM as a switching component [12].

Another way of obtaining short pulses is through the gain-switching method, which is primarily used in lasers with a relatively short lifetime, where the use of the Q-switching method is impractical. In this method, the pumping source is pulsed, and due to the high peak powers of the pulses, significant population inversion is achieved compared with the threshold. This high gain leads to the generation of high-energy short pulses [13].

The advantage of this method is the ability to design a short cavity since it only consists of the gain medium. Shortening the cavity length is crucial for reducing the pulse duration, as it is linearly proportional to the cavity round trip according to the analytical approximation [14]. In contrast, the Q-switching method involves the use of relatively large components, such as the AOM, which extends the cavity length, or the SA, which also extends the cavity length due to the large beam diameter it requires because of its low damage threshold. Only two studies were conducted at 2100 nm using the gain-switching method. In a study conducted in 2010 [15], the shortest pulse duration achieved was 20 ns with a pulse energy of 4.1 mJ at a repetition rate of 2 kHz and a wavelength of 2130 nm. In a later study in 2016 [16], a pulse duration of 44 ns with a pulse energy of 90 μJ at a repetition rate of 20 kHz and a wavelength of 2089 nm was reported. These studies demonstrated short-pulse lasers, although their pulse durations were not shorter than other Q-switched lasers.

In a previous work [17], we demonstrated the capability of generating a short-pulse-duration gain-switched Ho:YAG laser using a passively Q-switched Tm:YLF laser. However, since in some applications, an accurate controlled pulse repetition rate is required, this study utilized an actively Q-switched Tm:YLF laser as the pump source that enables controlled and accurate repetition rates. Furthermore, this work explored the potential for obtaining shorter pulse durations using longer pumping pulse durations. We achieved a short pulse duration of 4.6 ns with an energy of 0.52 mJ per pulse at a repetition rate of 2.5 kHz. Additionally, a pulse duration of 3.5 ns with an energy of 0.4 mJ per pulse was attained at a repetition rate of 1 kHz. Moreover, this study demonstrated the ability to obtain a short pulse from a Ho:YAG laser using a relatively long pumping pulse. The experimental setup is described below.

2. Experimental Setup

The experimental setup for the gain-switched Ho:YAG laser is illustrated in Figure 1. The pumping laser employed was an actively Q-switched Tm:YLF laser that utilized the end-pumping structure with a cavity length of 100 mm. A 3 × 3 × 16 mm³ 3 at.% Tm:YLF crystal was pumped by a 793 nm fiber-coupled laser diode. To achieve the pulsed mode, a 34 mm long quartz crystal acousto-optic modulator (AOM) was utilized, operating at a radio frequency of 40.68 MHz. The laser wavelength was tuned to the absorption peak of the Ho:YAG crystal at 1879 nm by using a 100 μm uncoated YAG etalon, enabling precise operation in close proximity to the holmium crystal’s absorption peak with an accuracy exceeding 0.5 nm. The Tm:YLF laser had a pulse duration of 12.5 ns and an energy per pulse of 2.2 mJ (limited due to damage on the Tm:YLF crystal), with a beam quality parameter of ${\mathrm{M}}^2$ < 1.4. The laser operated at repetition rates of 0.5, 1, and 2.5 kHz. The beam was collimated and focused to a spot diameter of 220 μm using a pair of double convex lenses with anti-reflective (AR) coatings in the range of 1650–3000 nm and a focal length of 200 mm.

After the collimating lens, a $\lambda$/2 plate and a polarizing cube were placed to serve as an attenuation system for controlling the transmitted energy, allowing alteration of the transmitted energy without affecting other parameters of the Tm:YLF laser, such as the pulse duration. To prevent feedback that could damage the Tm:YLF laser, a $\lambda$/4 plate was positioned after the polarizing cube.

The Ho:YAG cavity consisted of a planar entrance mirror (HT at 1879 nm, HR at 2100 nm), an output coupler (R = 50 mm, HR at 1879 nm, PR = 60% at 2100 nm), and a 3 × 3 × 20 mm³ 1 at.% Ho:YAG crystal. The Ho:YAG facets were AR-coated for both the pump and laser wavelengths.

The stability conditions of the resonator and the laser beam diameter were calculated in various configurations using the ABCD method. The Ho:YAG laser cavity was designed with a total cavity length of approximately 25 mm to achieve the shortest possible pulse duration.

The output power was measured using an Ophir L50(150)A-35 power meter, while the pulse energy was measured with an Ophir PE50-C energy meter. The pulse temporal characterization was performed using an EOT ET-5000 extended InGaAs fast photodetector with a 28 ps rise time and a 1 GHz oscilloscope (Agilent MSO7104A). The laser spectrum was acquired using a BaySpec extended InGaAs 1D array spectrometer. The ${\mathrm{M}}^2$ measurement was conducted using a Thorlabs BP209-IR2 scanning slit optical beam profiler. Spatial profiling of the laser beam was performed using a pyroelectric camera (Pyrocam III-HR, Spiricon).

3. Results

The Ho:YAG laser emitted at a wavelength of 2089 nm in both the continuous wave (CW) and pulsed modes, as depicted in Figure 2. In the CW mode, the Ho:YAG laser emitted 2.15 W using 4.55 W of the Tm:YLF, resulting in a slope efficiency of 59.2% and an optical-to-optical conversion of 47.3%, as illustrated in Figure 3.

In the pulsed mode, the Ho:YAG laser was operated at three different repetition rates between 0.5 kHz and 2.5 kHz. These repetition rates were chosen from a general applications point of view (repetition rates in the range of few hundreds of Hz are common for medical applications and repetitions rates in the kHz range are generally used in material processing applications); above 2.5 kHz, the received pulse energies were lower than the threshold of the Ho:YAG laser. The Tm:YLF pumping pulse duration was 12.5 ns. The shortest pulse duration of 3.5 ns was achieved with a maximal pulse energy of 0.4 mJ for a repetition rate of 1 kHz and a pumping pulse energy of 2.2 mJ. This corresponded to a slope efficiency of 30.3%, optical-to-optical conversion of 18.2%, and peak power of 114 kW. The maximum pulse energy was 0.52 mJ, which was achieved with a shortest pulse duration of 7.2 ns for a repetition rate of 2.5 kHz and for a pumping pulse energy of 1.66 mJ. This corresponded to a slope efficiency of 48.8%, optical-to-optical conversion of 31.2%, and peak power of 72 kW. The temporal profile of the 3.5 ns pulse duration is shown in Figure 4. The pulse train at 1 kHz is shown in Figure 5, the enlarged variations can be explained by variations in the Tm:YLF, which are reinforced in the Ho:YAG since the slope efficiency is much higher than the optical-to-optical conversion.

The Ho:YAG energy vs. pumping energy is shown in Figure 6. For the 2.5 kHz measurement, the pumping energy did not exceed 1.6 mJ due to damage on the Ho:YAG cavity mirrors. Figure 6 may give the impression that the efficiency is higher at higher repetition rates. However, from the average output power vs. average input power point of view (which was achieved by multiplying each data point (both x and y values) by its repetition rate), the lines overlapped, as presented in Figure 7. This means that they had similar efficiencies.

The relationship between the pulse duration and pulse energy is evident in the graph shown in Figure 8. At lower energies, the pulse duration experienced a substantial decrease before asymptotically stabilizing at higher energies. This pattern was particularly observed at a repetition rate of 1 kHz, where increasing the pulse energy did not result in a significant reduction in pulse duration. The 0.5 kHz line in the graph had only the beginning of the saturation stage, suggesting that higher energies could potentially yield shorter pulse durations. However, interpreting the behavior of the 2.5 kHz line was challenging, as the discrepancy may be attributed to the increased thermal load that affected the cavity mode [18].

The measured beam radius in both axes (${\omega }_x$ and ${\omega }_y$) along the Z-axis and the beams spatial profile are shown in Figure 9. By fitting the measured data to hyperbolic curves, the beam quality factors (${\mathrm{M}}^2$) were calculated to be 1.32 on both axes.

4. Discussion

Table 1. Summary of pulsed Ho:YAG lasers mentioned in the introduction, including the present laser.

Wavelength	Repetition Rate	Energy per Pulse	Pulse Duration	Switching Method
2089 nm	14.8 kHz	25 μJ	11 ns	Passive Q-switch
2089 nm	10.1 kHz	1.02 mJ	13.3 ns	Passive Q-switch
2090.9 nm	100 Hz	7.6 mJ	132 ns	Active Q-switch
2130 nm	2 kHz	4.1 mJ	20 ns	Gain switch
2089 nm	20 kHz	90 μJ	44 ns	Gain switch
2089 nm	1 kHz	0.4 mJ	3.5 ns	Gain switch
2089 nm	2.5 kHz	0.52 mJ	4.6 ns	Gain switch

summarizes the performance of this laser and the lasers presented in the introduction. From the data presented, it is easy to distinguish the significant advantage of the gain switch method over the Q-switch method in the ability to receive short pulses of a few nanoseconds and also the improvement in energy per pulse achieved in this work compared with previous published works by applying the gain-switch method with Ho:YAG crystals.

In order to investigate the influence of the pump pulse duration on the pulse durations received from the Ho:Yag gain-switched laser, two longer pumping cavities were constructed. The first cavity had a length of 200 mm and utilized a pumping pulse duration of 70 ns, while the second cavity had a length of 500 mm with a pumping pulse duration of 260 ns. In the 70 ns pumping pulse duration configuration, the experiments yielded a minimal Ho:YAG pulse duration of 4.4 ns and a maximal pulse energy of 0.4 mJ at a repetition rate of 1 kHz. These results were achieved with a pumping pulse energy of 1.15 mJ, corresponding to a slope efficiency of 59.9%, an optical-to-optical conversion of 34.8%, and a peak power of 91 kW. Additionally, at a repetition rate of 2.5 kHz, the maximum pulse energy of 0.52 mJ was obtained with the shortest pulse duration of 4.6 ns. This was achieved using a pumping pulse energy of 1.07 mJ, resulting in a slope efficiency of 58.3%, an optical-to-optical conversion of 48.6%, and a peak power of 113 kW. Figure 10 and Figure 11 show the Ho:YAG energy and pulse duration performance with a pumping pulse duration of 70 ns.

When the 70 ns pumping pulse duration results were compared with the previous results of a 12.5 ns pumping pulse duration, it was demonstrated that short pulse durations could be achieved using a pumping pulse duration of 70 ns for all repetition rates. However, the comparison was not ideal because the cavity also exhibited a higher efficiency for CW performance. Additionally, for a pumping pulse duration of 70 ns at 2.5 kHz and similar energy per pulse values, a shorter Ho:YAG pulse duration was obtained at a lower pumping power. If it was possible to have pulse durations of 12.5 ns and 70 ns in the same setup, unknown variations could have been avoided, and a clear conclusion could have been drawn.

In the 500 mm long pumping cavity with a pumping pulse duration of 260 ns, multiple Ho:YAG pulses were obtained instead of a single pulse. These results indicate that 260 ns is an excessively long pumping pulse duration that does not result in a high-performance laser.

5. Conclusions

In conclusion, a few ns Ho:YAG (2089 nm) gain switch laser with a relatively high energy per pulse pumped by an active Q-switched Tm:YLF laser is presented. Through experimental efforts and the optimization of laser parameters, short pulses ranging from a few nanoseconds to less than 10 ns, with an energy per pulse up to 0.52 mJ at repetition rates of 1–2.5 kHz, were achieved using pumping pulse durations from 12.5 ns to 70 ns. The implementation of this design using an active Q-switch instead of a passive Q-switch in the pump source was demonstrated for the first time, and it is essential in applications that demand accuracy and stability of the repetition rate. In addition, the limits of this method to achieve less than 10 ns pulse durations, with similar output energies per pulse, with dependence on the pumping pulse duration, were tested. This exploration of longer pumping pulse duration and the use of an active Q-switch pumping source opens up new possibilities for generating short pulses with increased energies and controlled repetition rates, thereby offering promising opportunities in applications such as sensing, LiDAR, medical procedures, and material processing.