In recent years, visible lasers have a wide range of applications in many fields^[1-4]. Among them, lasers with wavelengths of around 600 nm have been used in biomedicine, astronomy, lasers, metal processing, deep ultraviolet generation, and communications^[5-10]. Based on the characteristics of lasers with a wavelength of around 600 nm, such as the high absorption efficiency of blood proteins^[5], we believe that a high-performance 604 nm laser with narrow pulse duration and high energy will promote its applications in medical treatment.

The continuous-wave (CW) 604 nm laser has been mentioned many times in previous reports. In 2014, under the action of a linearly polarized 2ω-OPSL pump source with an output power of 5 W at 479 nm, Metz et al. obtained a CW 604 nm laser based on Pr:YLF crystal, whose maximum output power was 1.5 W^[11]. In 2016, Luo et al. obtained a CW 604 nm laser with a maximum output power of 0.6 W, and the pump source was a commercially available InGaN blue LD with a wavelength of 444 nm^[12]. In the same year, Fibrich et al. demonstrated a CW 604 nm laser with output power exceeding 1 W, and the pump source used was a fiber-coupled blue laser diode module from Necsel company^[13]. In 2022, the maximum output power of 3.28 W was obtained at wavelength of 604 nm based on a ∼0.12% (atomic fraction) doping Pr:YLF crystal^[14]. To the best of our knowledge, this is the maximum output power reported so far for a CW 604 nm laser.

There are two main methods to obtain lasers with narrow pulse duration and high pulse energy: mode-locking and Q-switching. Among them, Q-switching includes passive Q-switching and active Q-switching^[15,16]. Passively Q-switched praseodymium-doped (Pr3+) all-fiber pulsed lasers at 604 nm have been reported. In 2016, Li et al. proposed a passively Q-switched praseodymium (Pr3+)-doped all-fiber pulsed laser at 604 nm based on saturable absorbers of WS2 and MoS2; the narrowest pulse durations of the corresponding pulsed lasers were 435 ns and 602 ns, respectively. The maximum average output powers were 0.7 mW and 0.6 mW, respectively, and the maximum pulse energies were 6.4 nJ and 5.5 nJ, respectively^[17]. In 2017, Lin et al. demonstrated a passively Q-switched 604 nm praseodymium (Pr3+)-doped all-fiber pulsed laser, with Bi2Se3 as the saturable absorber. The narrowest pulse duration, the maximum average output power, and the maximum pulse energy of the pulsed laser were 494 ns, 0.5 mW, and 3.1 nJ, respectively^[18].

Passively Q-switched Pr:YLF pulsed lasers at 604 nm have been reported. In 2017, using a topological insulator (TI) Bi2Se3 nanosheet material as a saturable absorber, Cheng et al. proposed a passively Q-switched 604 nm pulsed laser based on Pr:YLF. The narrowest pulse duration, the maximum average output power, and the maximum pulse energy of the pulsed laser were 802 ns, 26 mW, and 0.2 µJ, respectively^[19]. One year later, Luo et al. proposed a passively Q-switched dual-wavelength (607 nm/604 nm) pulsed laser based on Pr:YLF, with a saturable absorber of few-layer Bi2Se3. The narrowest pulse duration, the maximum pulse energy, and the peak power were 263 ns, 0.19 µJ, and 0.71 W, respectively^[20]. In 2019, Tian et al. demonstrated a passively Q-switched dual-wavelength (607 nm/604 nm) pulsed laser and a dual-wavelength pulse vortex laser. The narrowest pulse duration and maximum average output power of the dual-wavelength pulsed laser were 409 ns and 63 mW, respectively. The narrowest pulse duration and maximum average output power of the dual-wavelength pulse vortex laser were 545 ns and 13 mW, respectively^[21]. Actively Q-switched Pr:YLF dual-wavelength pulsed lasers have been reported. In 2023, Jin et al. proposed a mechanism to obtain a dual-wavelength (604 nm/639 nm) pulsed laser based on Pr:YLF with a Fabry-Perot etalon and an acousto-optic modulator (AOM). The pulse duration, the energy of single pulse, and the maximum output power at the pulse repetition frequency (PRF) of 10 kHz were 100 ns, 4.9 µJ, and 98 mW, respectively^[22].

However, Q-switched 604 nm pulsed lasers with narrower pulse duration, higher power, and higher energy have not yet been reported.

In this work, we proposed an actively Q-switched Pr:YLF laser at a single wavelength of 604 nm with an AOM for the first time. A Lyot filter was used to select the oscillating laser in the resonant cavity. In CW laser operation, we achieved the maximum output power of 3.84 W at 604 nm. We obtained the maximum average output power, narrowest pulse duration, maximum single pulse energy, and peak power at repetition rates of 6 kHz (0.384 W, 44.5 ns, 64.1 µJ, 1.44 kW), 20 kHz (0.408 W, 56 ns, 20 µJ, 0.36 kW), and 50 kHz (0.454 W, 88 ns, 9.1 µJ, 0.1 kW) in Q-switched operation. The beam quality factors M2 were measured to be 2.87 (horizontal direction) and 2.40 (vertical direction).

As shown in Fig. 1, an LD array with a peak wavelength of 444 nm and a maximum output power of 36 W was used as the pump source. In our experiments, we measured an absorbed pump power efficiency of about 50%, with the maximum absorbed pump power of 18.06 W. The focus lens was an aspherical plano–convex lens with a focusing length of 75 mm. Under the action of the focus lens, the pump beam was focused into the crystal. The linear resonant cavity consisted of an input mirror (IM) and an output coupling mirror (OC). The IM was an end-coated plane mirror, and the OC was an end-coated plano–concave mirror with a curvature radius of 100 mm. The Pr:YLF crystal was encased in a copper block cooled with 18°C circulating water. The Pr:YLF was an a-cut crystal, with a 0.2% (atomic fraction) doping concentration of Pr3+ ions, and it had a length of 15 mm, with 3mm×3mm polished facets.

In CW laser operation, there was serious competition between the 604 nm laser in π-polarization direction and the 607 nm laser in σ-polarization direction, and it was difficult for the coupling mirrors (IM and OC) to suppress the output of the 607 nm laser efficiently. In order to solve this problem, we inserted a Lyot filter with a thickness of 2 mm at Brewster’s angle ∼56° in the resonant cavity for the selection of the 604 nm laser.

The AOM used in the experiment was the 3080-125 from Gooch & Housego, which has a modulation range of 400 nm to 850 nm. A filter was used to filter out the pump laser, which has a reflectivity of about 100% at 444 nm and a transmittance of about 90% at 604 nm.

As can be seen from Fig. 2, the transmittance of the IM at 444 nm was 91.5%, and the transmittance at 604 nm was almost 0%. The transmittance of the OC at 444 nm was about 96.7%, and the transmittance at 604 nm was about 0.08%.

In CW laser operation, we inserted no Lyot filter into the resonant cavity in the previous experiments. When the absorbed pump power was low, only the 604 nm laser was output. As the absorbed pump power increased, the 604 nm laser and 607 nm laser were output simultaneously, and finally only the 607 nm laser was output. The emission cross sections at wavelengths of 604 nm and 607 nm were ∼1.96×10−19cm2 and ∼1.57×10−19cm2, respectively^[23], which was one of the possible reasons for the experimental results. The CW output power versus absorbed pump power with no AOM in the experimental setup was presented in Fig. 3. No significant CW output power saturation was observed. The threshold absorbed pump power was 0.436 W, and the slope efficiency was 22.7%. The maximum CW output power at 604 nm was 3.84 W at the maximum absorbed pump power. The inset shows the spectrum of the CW laser at 604 nm; the central wavelength was 604.52 nm, and the full width at half-maximum (FWHM) was about 0.27 nm. We measured the power and spectrum with a Thorlabs S425C-L (detection sensitivity >2mW) and an Advantest Q8384 optical spectrum analyzer, respectively.

We recorded the experimental results in actively Q-switched operation, which showed that the range of the experimental tunable repetition rates was 6 kHz to 50 kHz. The average output power versus absorbed pump power at different repetition rates of 6 kHz, 20 kHz, and 50 kHz is presented in Fig. 4. With the same absorbed pump power, the average output power became gradually larger as the repetition rate within the range of the tunable repetition rate increased, and no saturation of the average output power was observed. Under the maximum absorbed pump power, the maximum average output powers at repetition rates of 6 kHz, 20 kHz, and 50 kHz were 0.384 W, 0.408 W, and 0.454 W, respectively. At different repetition rates of 6 kHz, 20 kHz, and 50 kHz, the corresponding slope efficiencies were 2.22%, 2.51%, and 2.77%, respectively. We believe that a better coupling system (mirrors) and a crystal with anti-reflection coating on the end faces in the experiment will improve the output power; as a result, a high slope efficiency will be obtained.

As shown in Fig. 5, we recorded the pulse durations versus absorbed pump power at repetition rates of 6 kHz, 20 kHz, and 50 kHz. Under the same repetition rate, as the absorbed pump power increased, the pulse duration gradually decreased. Under the same absorbed pump power, the smaller the repetition rate was within the range of tunable repetition rate, the narrower the pulse duration. At the maximum absorbed pump power, the pulse durations corresponding to 6 kHz, 20 kHz, and 50 kHz repetition rates were 44.5 ns, 56 ns, and 88 ns, respectively. The results were consistent with the law of dynamic evolution of population inversion density and photon density in the cavity at different repetition rates during the Q-switching operation. For example, when the repetition rate was the minimum value of 6 kHz, the population inversion density at the upper energy level was the largest, and when the photon density in the cavity evolved at the fastest speed, the pulse duration was narrowest.

For the equations E=P/f and Ppeak=E/τ^[24], E, P, f, Ppeak, and τ are the single pulse energy, average output power, repetition rate, peak power, and pulse duration, respectively. As shown in Fig. 6, under the same absorbed pump power, the smaller the repetition rate is within the range of tunable repetition rates, the greater the single pulse energy. At the maximum absorbed pump power, the maximum single pulse energy corresponding to repetition rates of 6 kHz, 20 kHz, and 50 kHz was 64.1 µJ, 20 µJ, and 9.1 µJ, respectively.

As shown in Fig. 7, under the same absorbed pump power, the smaller the repetition rate is within the range of tunable repetition rate, the greater the peak power. At the maximum absorbed pump power, the maximum peak power reached Watt level, corresponding to repetition rates of 6 kHz, 20 kHz, and 50 kHz and 1.44 kW, 0.36 kW, and 0.1 kW, respectively. The single pulse energy and peak power are the maximum values of the Q-switched 604 nm pulsed laser reported so far.

The traces of typical oscilloscope pulse trains and temporal pulse profiles corresponding to three repetition rates at maximum absorbed pump power are shown in Fig. 8. It can be seen that the 604 nm pulsed laser was relatively stable at actively Q-switched operation with an AOM at repetition rates of 6 kHz, 20 kHz, and 50 kHz. We recorded experiment data by an oscilloscope Agilent DSO-X 2012A.

Figure 9 presents the beam quality factors M2 and the beam profile under actively Q-switched operation with an AOM at repetition rate of 6 kHz. The beam quality factors M2 in the horizontal and vertical directions were 2.87 and 2.4, respectively. The inset is the beam profile captured by a CCD camera, and the results showed the formation of a Gaussian-like distribution. We believe that the thermal effects of the crystal have a great impact on the beam quality, and a better control of the crystal temperature will improve the quality factors of the beam.

We demonstrated an actively Q-switched Pr:YLF laser at a single wavelength of 604 nm with an AOM for the first time. In our experiments, we obtained the maximum CW output power of 3.84 W at 604 nm reported so far. The characteristics of the Q-switched pulsed laser at different repetition rates of 6 kHz, 20 kHz, and 50 kHz were compared. Under the maximum absorbed pump power, the pulse duration, average output power, peak power, and single pulse energy in Q-switched operation at the repetition rate of 6 kHz were 44.5 ns, 384 mW, 1.44 kW, and 64.1 µJ, respectively. As far as we know, there was no report on such a narrow pulse duration, high-power, and high-energy Q-switched pulsed laser at 604 nm. The pulse trains recorded by the oscilloscope were relatively stable. Apart from this, the beam quality factors Mx2 and My2 of the Q-switched 604 nm laser were 2.87 and 2.4, respectively. This work provided a method to obtain 604 nm pulsed lasers with narrow pulse duration, high power, and high energy, which promotes the application of orange pulsed lasers in various fields. In future work, we believe that a 604 nm pulsed laser with narrower pulse duration, higher power, and higher energy will be achieved by optimizing the experimental setup with a higher power pump source.