1.4 W Passively Q-Switched Mode-Locked Tm:CALGO Laser with a MoS₂ Saturable Absorber

Abstract

The passively Q-switched mode-locked (QML) operation of a Tm:CaGdAlO₄ (Tm:CALGO) bulk laser pumped by a wavelength-tunable Ti:sapphire oscillator using molybdenum disulfide (MoS₂) as a saturable absorber was demonstrated. By using an output coupler with 3% transmittance, Q-switched mode-locked operation can be achieved with 3.56 W absorbed pump power. At a pump power of 6.77 W, laser pulses with the maximum average power of 1.44 W were obtained, corresponding to a slope efficiency of 21.3%. The laser delivered pulses centered at 1927 nm with a repetition frequency of 131.6 MHz. The experimental results confirm the promising application of the MoS₂ in high-power Q-switched mode-locked solid-state lasers at 2 µm.

1. Introduction

In recent years, the development of near-mid-infrared ultrafast laser has been rapid [1], especially the 2 μm mode-locked laser is widely used in communication, medical, and compact spectroscopy. Q-switched mode-locked (QML) lasers emitting at 2 µm are of great interest in many application fields, such as material processing, radar, medical treatment, and remote sensing due to the advantages of high output peak power and excellent beam quality [2,3,4,5]. The passively Q-switched technology with saturable absorbers was widely applied and many materials have been explored as the candidates to realize the QML operations for the near-infrared solid-state lasers. These materials include semiconductor saturable absorbers (SESAMs), carbon nanotubes (CNTs), transition metal sulfides (TMDC), HfS₃, and graphene, etc. In particular, SESAM has attracted a great deal of attention in recent years. Zhou et al. reported a 40 ps continuous mode-locked Tm:Y₂O₃ laser with a SESAM [6]. However, due to specific fabrication techniques, SESAMs can only operate at a particular spectral band [7]. Graphene was considered to be an excellent passively mode-locked candidate for mid-infrared lasers [8,9,10], but the manufacturing process is hard due to its abilities to easily clump. Single-walled carbon nanotubes (SWCNTs) and double-walled carbon nanotubes (DWCNTs) are reported to be able to realize the mode-locked operation for the lasers at a wavelength of 1–2 μm [11,12,13]. The low efficiency of these materials at 2 μm makes them quite difficult for the QML applications at a high power [14]. There is also a new two-dimensional material, HfS₃, with a structure similar to MOS₂. In 2024, Li et al. used two layers of HfS₃ material as SAs to generate 540 fs laser pulses in an erbium-doped fiber laser [15]. However, the preparation process of HfS₃ is relatively complex, requiring high-precision equipment and control, and it is difficult to ensure the high purity of the material.

Molybdenum disulfide (MoS₂) is one typical TMDC with a hexagonal structure similar to graphene, consisting of two layers of sulfur and one layer of molybdenum; this has attracted much attention and been explored in recent years as a promising candidate for the QML operation of infrared lasers due to their high damage threshold [16,17], quite good stability, and high modulation depth. In 2014, Zhang et al. demonstrated [18] that MoS₂ can be utilized for the laser mode-locking operation in a 1054 nm fiber laser with an output pulse duration of 800 ps. Using MoS₂ as the saturable absorber in an Er: Lu₂O₃ laser, Fan et al. achieved a watt-level output from a passive Q-switching operation at a central wavelength of 2.84 μm with a pulse duration of 335 ns and a repetition rate of 121 kHz [19]. The passive Q-switching operation at a wavelength of 964 nm was explored for a Nd:YAG crystal with a pulse duration of 280 ns [20]. In 2023, Lee et al. demonstrated a 1559.4 nm Erbium-doped fiber laser at a repetition rate of 1.88 MHz by inserting a MoS₂ saturable absorber into a laser resonator [21]. In 2024, H. L. Lian et al. reported an erbium-doped fiber laser with a MoS₂ film, achieving the Q-switched pulses with the shortest pulse duration of 4.75 μs at 89.9 kHz [22]. However, most of the relevant research reports mainly focus on the 1 μm laser, and the application of MoS₂ in a 2 μm high-average-power Q-switched mode-locked solid-state laser has not been reported so far.

Due to its disordered structure, CALGO host has quite good thermomechanical properties [23], high thermal damage threshold, and the broadband emission spectral bandwidth, which is considered as an excellent candidate for the high-average-power lasers with a short pulse output. Yb-doped CALGO lasers at 1µm have already been demonstrated [24], and high thermal damage threshold, which is an excellent substrate for high power laser. In 2015, Qin et al. realized the mode-locking operation of Tm:CALGO for the first time, with a maximum output power of 330 mW [25]. Then, Lan et al. realized the Q-switching operation of Tm:CALGO crystal at 13.9 kHz with a pulse width of 44 ns by Cr2+: ZnSe saturable absorbers [26]. In 2016, Wang et al. realized 650 fs continuous mode-locked operation of Tm: CALGO crystal with a maximum average power of 58 mW using SESAM [27]. In 2023, Dupont et al. reported a Tm: CALGO laser operating on the 3H4 → 3H5 transition and didn’t achieve mode-locking [28]. In conclusion, a stable high power mode-locked laser based on Tm:CALGO crystal remains to be further studied.

In this paper, we report for the first time a passively QML Tm:CALGO laser operating at a central wavelength of 1927 nm with a MoS₂ saturable absorber. An average power up to 1.44 W was achieved with a pulse repetition rate of 131.6 MHz when an output coupler with 3% transmittance was implemented. The output pulse spectral bandwidth is measured as 15 nm of the full-width at half maximum, which supports a transfer-limited pulse duration of 340 fs.

2. Experimental Setup

The schematic experimental setup of the QML laser based on the Tm:CALGO crystal is shown in Figure 1. A wavelength-tunable Ti:sapphire laser is used as the pump source, which provides a maximum output power of 7.6 W. The emission wavelength from the pump laser is fine tuned to 791.8 nm to match the peak absorption spectra of the Tm:CALGO crystal. The CALGO crystal doped with a 4 at.% of Tm³⁺ is 3 mm × 3 mm in aperture and 5 mm in thickness and is used as the gain medium. The Brewster-angle cut crystal was carefully wrapped with an indium film to remove the residual thermal load inside of the crystal; this was done with the help of water-cooled heat sinks, which are crucial for lasers operating at a high power. Then, the Tm:CALGO crystal was wrapped with an indium film and placed on water-cooled heat sinks and maintained at 12 °C. The pump beam was focused into the gain medium with a 150 mm focus-length lens L with a transmittance of 95%. The whole cavity consists of two dichroic concave mirrors, M1 and M2, with a 100 mm ROC, coated with a high-transmission film for the pump wavelength over 770–1050 nm and a highly reflective film for the lasing wavelength in the range of 1800–2075 nm, a flat concave mirror M3 with 75 mm ROC, a flat mirror M4, which is high-reflection coated (R > 99.9%) at lasering wavelength, a saturable absorber MoS₂, and a flat output coupler OC. Three different output couplers (OCs) with a transmittance of 1.5%, 3%, and 5% were used in the experiment. Through the calculation of the ABCD matrix, the beam size of the pump light is carefully controlled to match the cavity fundamental mode size, which is about 58 μm inside the Tm:CALGO crystal; the spot size on the MoS₂ was designed to be about 155 µm in diameter in order to reduce the damage risk of the high intracavity power.

3. Experimental Results and Discussion

The absorption efficiency of the Tm:CALGO crystal at different laser operation conditions is shown in Figure 2a. Around 89.2% and 68.9% of the pump absorption efficiency were measured for the laser at the lasing or non-lasering conditions, respectively, which is mainly due to the saturation effect of the upper level of the gain medium.

Figure 2b shows the CW laser performance with different transmissions of OCs. The maximum average powers of 1.22 W, 1.7 W, and 1.74 W are achieved at transmissions of 1.5%, 3%, and 5%, respectively, and the corresponding slope conversion efficiencies are 19.4%, 26.9%, and 30.1%, respectively. As the transmittance of the output mirror increases, the pumped laser threshold also increases due to an increase in cavity loss; the CW lasing thresholds are locked at 290 mW, 445 mW, and 526 mW, respectively. The mode-locking operation with a MoS₂ as the saturable absorber for this laser is demonstrated and the experimental results are shown in Figure 2c. For 1.5% OC, the lasing threshold is 390 mW. Stable QML operation is achieved when the pump-absorbed power reaches 2.42 W. As the pump-absorbed power rises, the maximum average output power can reach 1 W with a slope efficiency of 15.5%. When using 3% OC, the lasing threshold is 463 mW. With 3.561 W absorbed pump power, MoS₂ has a fluence of 266 µJ/cm². When the absorbed power of the pump is 6.77 W, the laser delivers pulses with the maximum average output power of 1.440 W, corresponding to a slope efficiency of 21.3%. However, with an output coupling transmission of 5%, it is difficult for the laser pulses to maintain a stable QML operation due to excessive loss in the cavity. Obviously, the average output power and slope efficiency in this case are higher than that of the 3% OC, and the mode-locking threshold is lower than that of the 1.5% OC. Therefore, we can choose the 3% transmittance couple according to the actual requirements.

The pulse spectrum from the mode-locked operation was measured with an optical spectrum analyzer (Avantes, Apeldoorn, The Netherlands) and is shown in Figure 3. The laser operates at a central wavelength of 1927 nm and the spectral bandwidth of the full-width at half maximum (FWHM) is about 14 nm, which corresponds to a transfer-limited pulse duration of 430 fs. The typical QML pulses train output from the laser was recorded with a fast photodiode (ET-5000, EOT, USA) and a digital oscilloscope (RIGOL, DS4024), which are shown in Figure 4 with different timescales. A 38.5 kHz QML repetition frequency was recorded. A repletion of the 131.6 MHz internal of the QML envelope is shown in Figure 4d, which is matched to the total length of the cavity of 1.15 m. Close to 100% of the modulation depth of the QML pulses was achieved.

Since the autocorrelator (APE, Berlin, Germany) is only suitable for measuring the shorter pulse duration of 12 ps, the accurate pulse duration cannot be measured. Based on the mode-locked pulse signal of the oscilloscope, the actual mode-locked pulse width can be estimated by the following Formula (1):

$${\mathrm{t}}_{\mathrm{m}}=\sqrt{{\mathrm{t}}_{\mathrm{r}}^2+{\mathrm{t}}_{\mathrm{p}}^2+{\mathrm{t}}_0^2},$$

(1)

t_m is the measured pulse rising edge time, t_r is the actual pulse rising edge time, t_p is the rising edge time of the diode detector, and t₀ is the rising edge time of the oscilloscope. The rising edge time of the oscilloscope can be calculated using the Formula (2):

$${\mathrm{t}}_0\times {\mathrm{W}}_{\mathrm{B}}=0.35~0.4$$

(2)

Among them, W_B is the bandwidth of the oscilloscope, and the bandwidth of the digital oscilloscope in this experiment is 200 MHz, so it can be estimated that t₀ = 1750 ps; Furthermore, it is known that the experimental measured pulse rising edge time is about 1900 ps, and the rising edge time of the 2 μm photodiode detector is 35 ps, so the actual mode-locked pulse rising edge time t_r can be estimated to be about 739.1 ps. The pulse width is about 1.25 times the rising edge time, so the actual mode-locked pulse width is about 923.8 ps.

Finally, we also refer to the recently published laser results for comparison with the current results in this work. In 2021, Zhang et al. obtained a maximum average output power of 208 mW and a pulse width of 412 ns using a Nd:YAG crystal as the laser gain medium and MoS₂ as the saturable absorber. In 2024, the Q-switched mode-locked Nd: GGG laser was reported, delivering a maximum power of 236.4 mW and a pulse duration of 447 ps using a MoS₂–SnSe₂ heterojunction as a saturable absorber [29]. However, we demonstrated a QML Tm:CALGO laser with a maximum average power of 1.4 W, which is the highest power output that can be achieved in a 2 μm ultrafast laser with a MoS₂ saturated absorber as a mode-locking element. In this work, MoS₂ shows good thermal and chemical stability compared with traditional saturated absorbing materials, especially in high-power laser experiments, and its durability is significantly excellent. Furthermore, MoS₂ has a high saturation absorption, which can effectively inhibit the continuous out-put of the laser and provides the possibility for mode-locked laser pulse generation. In addition, its two-dimensional structure allows MoS₂ to prepare extremely thin films, which not only reduces optical losses, but also effectively reduces the negative impact of thermal management and optical performance losses in high-power operations, thereby improving the overall efficiency of the laser.

4. Conclusions

In conclusion, we report for the first time a watt-level Tm:CALGO laser operating in QML mode at 1927 nm with a MoS₂ as the saturable absorber. Experimental results show that the lasing threshold is 390 mW with a 3% transmission output coupler and the stable QML operation can be realized at an absorbed pump power of 3.56 W. An average power up to 1.44 W was achieved at a pump power of 6.77 W, corresponding to a slope efficiency of 21.3%. The experimental results show that MoS₂ is a promising candidate as a saturable absorber for the high-power mode-locked solid-state lasers at 2 μm. The further power scaling of the Tm:CALGO lasers is currently in progress through the optimization of the cavity configuration, the transmission of the output coupler, and the fluence of the MoS₂ saturable absorber. The 2 μm laser is in urgent demand in the fields of national defense security, biomedicine, scientific research, and so on, but European and American countries that have developed rapidly in this field have imposed an embargo on 2 μm laser products and key devices in our country, which makes the research and development of high-power 2 μm lasers important.