Nanosecond pulsed lasers based on thulium-doped solid-state lasers and emission in the 1.9–2 µm band (${{}^3\mathrm{F}}_4\to {{}^3\mathrm{H}}_6$) have wide application values in laser ranging, spectroscopy, optical medicine, and metrology^[1–4]. In the past two decades, passively $Q$-switched (PQS) lasers using saturable absorbers (SAs) have strongly promoted the development of high-quality pulsed lasers, largely due to the rapid development of SAs. Therefore, it is very important to explore high-performance SAs to improve the performance of the PQS laser. In the past, Cr:ZnS^[5], Cr:ZnSe^[6], and semiconductor SA mirrors (SESAMs)^[7] have been widely used as SAs in the wavelength region of 1.9–2 µm. However, the disadvantages of SESAMs, such as complex manufacturing, high cost, and narrow absorption band width, greatly limit the application. Cr:ZnS and Cr:ZnSe have large absorption cross sections and are very suitable for generating high-energy pulsed lasers. The PQS thulium-doped lasers based on Cr:ZnSe and Cr:ZnS have been proved to be able to obtain millijoule pulse lasers; however, the corresponding repetition rates are rather low^[5,6].

In the recent years, SAs based on two-dimensional (2D) materials, including graphene^[8], black phosphorus (BP)^[9], transition metal dichalcogenide (TMD)^[10], topological insulators, and MXene^[11,12], have been widely investigated due to their excellent saturable absorption properties. Moreover, MXene also exhibits large optical modulation depth, high damage threshold, excellent conductivity, tunable bandgap, high electric capacity, and so on^[13–17].

Recently, a few MXenes (${\mathrm{Ti}}_3{\mathrm{C}}_2{\mathrm{T}}_{\mathit{x}}$, ${\mathrm{Ti}}_2{\mathrm{CT}}_{\mathit{x}}$, ${\mathrm{Ti}}_4{\mathrm{N}}_3{\mathrm{T}}_x$, ${\mathrm{Nb}}_2{\mathrm{CT}}_{\mathit{x}}$) have been applied to pulse lasers with different working mechanisms^[18–22]. In 2019, a ${\mathrm{Ti}}_3{\mathrm{C}}_2{\mathrm{T}}_{\mathit{x}}$$Q$-switched $\mathrm{Tm},\mathrm{Gd}\text{:}{\mathrm{CaF}}_2$ laser with 2.4 µs pulse width was reported by Zu et al., whose working wavelength was 1974.5 or 1929.7 nm, which mainly depended on the transmittance of the output coupler (OC)^[19]. In 2021, Huang et al. demonstrated that ${\mathrm{Ti}}_2{\mathrm{CT}}_{\mathit{x}}$ could be a promising SA in a 1.06 µm $\mathrm{Nd}\text{:}{\mathrm{Y}}_3{\mathrm{Al}}_5{\mathrm{O}}_{12}$ (Nd:YAG) solid-state laser^[20]. In 2020 and 2021, solid-state PQS $\mathrm{Er}\text{:}{\mathrm{Lu}}_2{\mathrm{O}}_3$ lasers at 2.85 µm with ${\mathrm{Ti}}_4{\mathrm{N}}_3{\mathrm{T}}_x$ and ${\mathrm{Nb}}_2{\mathrm{CT}}_{\mathit{x}}$ SAs were achieved, respectively^[21,22]. However, common MXenes such as ${\mathrm{Ti}}_2{\mathrm{CT}}_{\mathit{x}}$ and ${\mathrm{Ti}}_3{\mathrm{C}}_2{\mathrm{T}}_{\mathit{x}}$ have the disadvantage of easy degradation under ambient conditions, so their practical application is greatly limited^[3,23]. Fortunately, ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$ nanosheets have better stability than ${\mathrm{Ti}}_3{\mathrm{C}}_2{\mathrm{T}}_{\mathit{x}}$ and ${\mathrm{Ti}}_2{\mathrm{CT}}_{\mathit{x}}$ nanosheets^[23].

In 2020, MXene ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$ was first used as an SA in a 1.56 µm mode-locked fiber laser, showing excellent saturable absorption performance^[23]. However, the saturable absorption behavior of ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ in a $Q$-switched laser has not been studied. Compared with a fiber laser, a solid-state laser has smaller nonlinear pulse splitting^[24], so the research of the ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$$Q$-switched solid-state laser is of great significance to obtain high-energy short pulses.

In this paper, the linear and nonlinear absorption properties of a home-made ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ were studied. With the ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$, a typical PQS operation at 1.94 µm was realized in a $\mathrm{Tm}\text{:}{\mathrm{YAlO}}_3$ (Tm:YAP) laser. To the best of our knowledge, the first demonstration of a ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$$Q$-switched laser is reported in this paper.

The multilayer ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$ powder (10 mg) and anhydrous ethanol (10 mg) were mixed together in a glass bottle, and ultrasound was used in an ultrasound machine for 1 h. We then used a rotary coater (KW-4A, Chinese Academy of Sciences) to coat the ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$ solution on one side of the quartz plate. Based on the above operation, a home-made ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ was ready. The surface structure of ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$ nanosheets was captured at 1000× magnification, as shown in Fig. 1(a). The component of ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}$ nanosheets is shown in Fig. 1(b), which was analyzed by energy dispersive spectroscopy (EDS). The oxygen and fluorine peaks indicated the generation of numerous surface termination groups, and the weak aluminum peak represented some residue of ${\mathrm{V}}_2\mathrm{AlC}$. The linear absorption of the ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ at 1000–2200 nm is shown in Fig. 1(c). The presence of surface functional groups is the main cause of the broadband absorption spectrum^[18,25]. Figure 1(d) shows the relationship between nonlinear absorption and optical intensity. The experimental data were measured by a home-made ${\mathrm{Tm}}^{3+}\text{-}{\mathrm{Ho}}^{3+}$ co-doped laser (2119.3 nm, 55 ns, 10 kHz). The fitting curve was obtained by the following formula: $$T(I)=1-\mathrm{\Delta }T·\exp \left(\frac{-I}{I_{\mathrm{sat}}}\right)-T_{\mathrm{ns}}.$$

Herein, $I$ and $T$ are the optical intensity and transmittance, and the calculated non-saturated absorption loss ($T_{\mathrm{ns}}$), saturation intensity ($I_{\mathrm{sat}}$), and modulation depth ($\mathrm{\Delta }T$) were 7.9%, $208.95\mathrm{kW}/{\mathrm{cm}}^2$, and 11.6%, respectively.

The measurement results show that ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ has the advantage of large modulation depth, so it is very suitable for obtaining high energy and short pulse width.

Figure 2 shows the experiment setup. The pump laser was a 793 nm laser diode (LD). After passing through the lenses f1 (25 mm) and f2 (50 mm), the pump spot in the Tm:YAP crystal was 210 µm. The size of the crystal ($b$-axis cut, 3% doping concentration) was $3\mathrm{mm}\times 3\mathrm{mm}\times 8\mathrm{mm}$. The heat generated in the crystal can be taken away in time by a water cooler to ensure the operating temperature is 286 K. In this experiment, a dual mirror linear resonator was used to carry out the study, and the physical cavity length was 38 mm. The input coupler (IC) was a flat concave mirror, and its curve radius was 100 mm. Three OCs (plane mirror) were used here, and the transmittances ($T$) at 1900–2100 nm were 1.5%, 2.5%, and 5.0%, respectively. The generated laser was separated by the dichroic mirror (DM). The prepared ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ was inserted between the OC and Tm:YAP crystal, which served as the $Q$-switch. The laser radii on the Tm:YAP crystal and ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ were 187 and 174 µm, respectively.

Figure 3(a) shows the average output power obtained in the PQS mode. When the incident power was increased to 4.26, 5.01, and 4.25 W, respectively, the PQS laser with the OC of $T=1.5\%$, 2.5%, and 5.0% started the $Q$-switching operation. The maximum average output power available was 190, 220, and 350 mW, corresponding to the slope efficiencies of 3.5%, 4.4%, and 6.3%, respectively.

For comparison with the PQS mode, the continuous-wave (CW) output power obtained with the 5.0% output mirror was also studied. Up to 1.48 W output power with a slope efficiency of 20.5% was achieved, and the laser threshold was approximately 1.6 W, as shown in Fig. 3(b). Compared with the CW mode, the laser threshold of the PQS laser increased, and the output power decreased, which can be attributed to the insertion loss of the SA. Figure 4 shows the pulse characteristics achieved in PQS mode. By using the OC of $T=5.0\%$, a shortest pulse width of 528 ns with a 65.9 kHz repetition rate was obtained. By using the OCs of $T=1.5\%$ and 2.5%, the minimum pulse widths were 636 and 919 ns, and the maximum repetition rates were 69.3 and 74.6 kHz. The pulse energy tends to be constant, as shown in Fig. 5(a), while the peak power increases roughly linearly, as shown in Fig. 5(b). The maximum pulse energy and peak power were 6.33 µJ and 10.06 W, which were obtained at the pump power of 7.21 and 8.69 W, respectively. By using the OCs of $T=1.5\%$ and 2.5%, the maximum pulse energies of 2.74 and 2.95 µJ were obtained, and the corresponding peak powers were 4.31 and 3.21 W, respectively.

If we continue to increase the incident pump power, the SA will be destroyed with the 1.5% output mirror. Therefore, we can infer the damage threshold of $\mathrm{MXene}\text{-}{\mathrm{V}}_2{\mathrm{CT}}_x$ based on the intracavity energy, and its value is estimated to be $0.42\mathrm{J}/{\mathrm{cm}}^2$. By comparing the laser performance of the three output mirrors, it can be concluded that the optimum performance was achieved with the 5.0% output mirror, that is, smaller pulse width and repetition rate as well as higher energy and peak power can be obtained.

Figure 6 shows the temporal profiles and pulse trains obtained under different conditions. With the 5.0% output mirror, the measured output spectra are shown in Fig. 7. When the incident pump powers were 5.73 and 8.69 W, the peak emission wavelengths in the CW operation were 1992.1 and 1991.7 nm, corresponding to the linewidths of 4.7 and 5.6 nm, respectively. The peak emission wavelengths were decreased to 1937.8 and 1937.4 nm in the PQS operation, and the corresponding linewidths were 1.1 and 2.2 nm, respectively.

The shift of laser wavelengths from CW operation to PQS operation is due to the fact that the excited-state population fraction in PQS operation is much higher than that in CW operation, which results in the peak of the gain cross section moving to the shorter wavelength. Also, the linewidths of PQS operation are narrower than those of CW operation, which can be mainly attributed to the etalon effect caused by the SA mirror. The beam quality of the PQS laser at 350 mW output power was evaluated by a BP109-IR2M beam profiler. The measured $M^2$ factor in the X direction was 1.15, and the value in the Y direction was 1.18, which are shown in Fig. 8(a). The spatial distributions of the beam are shown in Figs. 8(b) and 8(c).

In summary, a multilayer ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ was prepared. The linear and nonlinear absorption properties of the SA were characterized. By using the ${\mathrm{V}}_2{\mathrm{CT}}_{\mathit{x}}\text{-}\mathrm{SA}$ in a Tm:YAP laser, a typical PQS operation at 1.94 µm was realized. The optimum performance was achieved with the 5.0% output mirror. At a repetition rate of 65.9 kHz, the maximum peak power and average output power were 10.06 W and 350 mW, respectively, corresponding to the minimum pulse width of 528 ns. Maximum pulse energy of 6.33 µJ was achieved at the 38.7 kHz repetition rate. Furthermore, the obtained $M^2$ factor was less than 1.2.