1. Introduction

Sum-frequency laser [1], copper vapor laser [2] and dye laser [3,4] are the traditional methods to obtain coherent light in the spectral range of 550-620 nm that are urgent for applications such as stimulated emission depletion microscopy [5], sodium guide star [6–8] and isotope separation [3,9], etc. However, due to drawbacks on complexity and high-cost for the traditional ways, new laser sources are desirable, and among them frequency doubled solid-state Raman laser have emerged as a promising alternative [10–15]. Generally, the performance of all-solid-state Raman lasers is closely related to the laser medium [16,17]. Due to the large stimulated emission cross section and short upper-state lifetime, neodymium-doped vanadate crystals such as Nd:YVO₄ and Nd:GdVO₄ are excellent laser media in high-repetition-rate Q-switched solid-state Raman lasers for scaling up the average power [16]. Particularly, the self-Raman lasers with Nd-doped vanadates remove the necessity of an extra Raman crystal, which improves the structural compactness and decreases the insertion loss of the whole system [13,18–20]. As a consequence, it further facilitate the scale-up of average output power in high-repetition-rate Q-switched operation. During the past decade, Nd-doped vanadates based self-Raman lasers have been investigated thoroughly. In 2009, Zhu et al. demonstrated an end-pumped acousto-optic (AO) Q-switched Nd:YVO₄ self-Raman laser intra-cavity frequency-doubled by LBO crystal. 7.9-W output at 588 nm with pulse repetition frequency (PRF) of 110 kHz was obtained under 26.5-W LD pumping [14]. In 2020, Chen et al. achieved a high-power AO Q-switched Nd:YVO₄ self-Raman yellow laser with high repetition rates (100-500 kHz) through optimizing the “gate-open” time. Under an incident pump power of 26 W, a maximum average output power of 8.8-W yellow lasers was obtained at a repetition rate of 200 kHz [18]. High-power and high-repetition-rate yellow pulsed lasers are the ideal source of material processing with high speed scanning [21] and sodium guide star (pulsed at Larmor frequency) [22]. Whereas, for applications on pumping source of dye lasers [23], isotope separation [3,9] and laser resistor trimming, yellow pulsed lasers with high-energy and low-repetition-rate are needed. The relatively short upper-laser-level lifetime of Nd:vanadates prevents self-Raman lasers from operating efficiently at PRFs of a couple of tens of kilohertz or lower. Therefore, the pulse energy is severely limited. Furthermore, extra thermal load from the SRS process resulted in serious thermal effects in self-Raman lasers, which decrease the Raman gain coefficient, cause cavity instability, or even result in thermal fracture. The phenomenon is more obvious for the Nd:vanadate crystals, whose thermal conductivity and mechanical properties are not comparable to those of Nd:YAG [24,25]. The Nd:YAG crystal, compared with Nd:vanadates crystal, has a much longer fluorescence lifetime. The long lifetime can raise the output pulse energy of a fundamental AO Q-switched laser, thereby improving the output pulse energy of the Stokes light [17]. However, Nd:YAG is not suitable for generating laser pulses at repetition rate higher than 100 kHz in Q-switched operation, because of the moderate gain and relatively long upper-state lifetime [26]. Overall, although a diode-end-pumped solid-state Raman yellow laser that could not only operate efficiently at low repetition rate but also at high repetition rate is necessary to fulfill the different requirements in applications [3,7,19–21], to our best knowledge, none of the kinds of lasers have been reported.

We hereby report on an efficient end-pumped intra-cavity frequency-doubled Raman laser with hybrid laser gain. Combining the advantages of Nd:YAG crystal and Nd:vanadates crystal, we realized an efficient solid-state Raman laser that could be capable of operating at low and high pulse repetition frequency. Since the ⁴F_3/2→⁴I_11/2 transition for Nd:YAG and the Nd:YVO₄ crystals corresponds to the relatively close emission wavelength [27], a Nd:YAG crystal and a Nd:YVO₄ crystal set in sequence would absorb the pump and provide the laser gain together. The majority of the pump is absorbed by the Nd:YAG, which can sustain much higher thermal load and allows higher pump power. The Nd:YVO₄ crystal, which provides much higher laser gain on π-polarization than on the σ direction, absorbs part of the pump to make sure that the 1064-nm fundamental laser is nearly linear polarized. Therefore, higher pump can be employed compared to the vanadate self-Raman lasers. In addition, the approximately linear polarized 1064-nm fundamental wave and 1176-nm Stokes wave enable the LBO type I phase matching, which is beneficial to the power scaling. The maximum average output power raises from 4.1 W to 7.6 W at an incident pump power of 42.0 W while the PRF raises from 10 kHz to 110 kHz.

2. Experimental setup

Experimental setup of the hybrid gain Raman laser is depicted in Fig. 1. The pump source is a 50-W fiber-coupled laser diode (LD) array emitting centered at ∼808 nm with bandwidth of ∼3 nm (FWHM). The core diameter and the NA of the fiber are 400 µm and 0.22, respectively. A 1:2 multi-lens coupler is used to refocus the pump into the laser crystals of a Nd:YAG and a Nd:YVO₄, which are both anti-reflectively (AR) coated at 800-1200 nm. Nd:YAG crystal with doping concentrations of 0.4-at.% and the dimensions of 3×3×6 mm³ is used in the experiment. The Nd:YVO₄ crystal is a-cut with the doping concentrations and dimensions of 0.18-at.% and 3×3×20 mm³, respectively. The Nd:YAG crystal is placed in the front to absorb the majority of the 808-nm pump, while the Nd:YVO₄ crystal is set close to the Nd:YAG afterwards. The two crystals are both wrapped in indium foil and clamped in an aluminium holder cooled by refrigerant water at the temperature of 10 ^°C. A flat mirror M1 coated for high transmissive (HT) at 808 nm and high-reflection (HR) at 1064 and 1176 nm makes the shared cavity of 1064-nm laser and 1176-nm Stokes with a concave mirror M2 which is coated for HR at 1064 and 1176 nm and HT at 588 nm. The radius of curvature (ROC) of mirror M2 is 500 mm. A 3×3×15-mm³ lithium triborate (LBO) crystal AR coated at 588, 1064 and 1176 nm on both facets is used for intra-cavity frequency doubling the 1176-nm Stokes wave. The crystal is cut at θ=90°, φ=0° for non-critical phase matching at 41.5 ^°C. M3 is a plano mirror coated for AR at 1064 and 1176 nm on both facets and HR at 588 nm on one facet, used to collect the backward propagating yellow light. The 20-mm-long, 80-MHz acousto-optic Q-switch driven by 10.0 W of RF power has AR coatings over fundamental laser and first Stokes wavelength range (R<0.2%) on both faces. The total cavity length is ∼92 mm. It is worth mentioning that the pump absorption of the Nd:YAG varies obviously as the LD wavelength drifts with the launched power because of the crystal’s relatively narrow absorption bandwidth. Therefore, the LD’s temperature is optimized for best absorption in the Nd:YAG crystal at the maximum pump power.

 

Fig. 1. Schematic of the Nd:YAG-Nd:YVO₄ hybrid Raman yellow laser.

Download Full Size | PDF

3. Results and discussions

Hybrid laser gain from the two laser crystals is expected to generate approximately polarized 1064-nm fundamental wave and hence approximately polarized 1176-nm Stokes wave. The polarization ratio of the 1064-nm fundamental laser is measured using a polarization beam splitter, the mirror M2 replaced by a 20% output coupler at 1064 nm with the same ROC, Q-switching PRF of 10 and 110 kHz, respectively, and the LBO crystal absent. In our experiments, ∼60% of incident pump is absorbed by Nd:YAG and most of the residual is absorbed by the Nd:YVO₄. The polarization ratio is ∼69:1 when the pump power is just above threshold and remains over 10:1 under the pump power of 42.0 W at the PRF of 110 kHz, while it decreased from 60.2:1 to 7.6:1 at the PRF of 10 kHz. The results are shown in Fig. 2. The fundamental field becomes more depolarized as the pump power increases. The severe thermal birefringence in the Nd:YAG crystal under high pump power should be responsible for this result. Note that, in our experiment, we found that the polarization ratio obviously varied with the stress of Nd:YAG crystal. Thus, the polarization ratio could be improved if the stress on Nd:YAG was controlled reasonably. However, we did not further investigate the problem because of lacking of equipment for controlling accurately the stress on Nd:YAG.

 

Fig. 2. The polarization ratio of fundamental laser as function of incident pump power with different PRFs.

Download Full Size | PDF

Thermal lens effects in both laser crystals are estimated following the approach presented by Innocenzi et al. [28], with thermal load from both lasing and SRS taken into consideration. The 42.0-W incident pumping with 400-µm pump spot radius (ω_p) in the 0.4-at.% doped, 6-mm-long Nd:YAG would result in 140-150 mm thermal focal length. For the Nd:YVO₄ crystal, the cumulative thermal focal length is estimated to be larger than 110 mm [29]. The cavity should be far from unstable. For the typical Nd:YVO₄ self-Raman laser, the thermal focal length of the crystal would be shorter than 65 mm under 42.0 W of pumping with the same ω_p and ω_s and the cavity. The result shows that the hybrid-gain structure has advantage in terms of alleviating thermal lens effects. Moreover, the thermal load of Raman medium employed in hybrid-gain structure is smaller than that in typical Nd:YVO₄ self-Raman laser, which ensures that the Raman gain coefficient does not decrease significantly with increasing of incident pump power. Therefore, higher pump power is allowed for hybrid gain Raman laser because of the two causes mentioned above.

Then the LBO crystal is inserted into the cavity and mirror M2 is used for yellow output. Figure 3 shows the average output power and optical efficiency of 588-nm emission as function of incident pump power with different PRFs. The maximum average output power of 7.6 W is obtained under the incident pump power of 42.0 W at the PRF of 110 kHz, corresponding to an optical efficiency of 16.9%. The optical efficiency at the high PRF is much lower than that reported in [14] and [18]. Furthermore, it is worth noting that, though we align the cavity at the maximum incident pump power, the optical efficiency begins to decrease at the relatively low incident pump power. This phenomenon is different with the traditional Nd:YVO₄ Raman laser. Imperfectly polarized Stokes wave instead of decline of Raman gain should be chiefly responsible for the results. We measured and estimated the 588-nm output beam quality factor M² through a knife-edge method. The 588-nm output beam quality factor M² in the x and y directions are measured to be 4.3 and 4.1, respectively, at the maximum output power of 7.6 W, while the root-mean-square (RMS) fluctuation of the output power within 10 min is ∼4.2%. The beam profile is shown as the inset of Fig. 3. Note that the hybrid Raman laser can operate efficiently at the low PRF. 4.1-W (0.41-mJ) yellow light is obtained under the incident pump power of 42.0 W at the PRF of 10 kHz. We also measured the 588-nm output beam quality factor M² at the 4.1-W output. The beam quality factor M² in the x and y directions are 5.1 and 4.7, respectively. The temporal behavior of the Raman lasers was recorded by using a photoelectric detector Thorlabs DET08C and an oscilloscope RIGOL DS1202Z-E. The overall peak-to-peak instability of pulse amplitude is ±6.1% at the PRF of 10 kHz, while it is ±4.6% at the PRF of 110 kHz as shown in Fig. 4. At the PRFs of 10 and 110 kHz, the pulse durations are 8 and 11 ns, respectively.

 

Fig. 3. Average output power at 588 nm and optical efficiency as function of incident pump power with different PRFs.

Download Full Size | PDF

 

Fig. 4. Oscilloscope traces of 588-nm light at repetition rates of (a) 10 and (b) 110 kHz. Left, pulse train; right, pulse duration.

Download Full Size | PDF

A typical self-Raman laser with only the Nd:YVO₄ crystal used as gain medium is carried out for comparison. Other cavity elements are kept the same except the cavity length is shortened to 84 mm after removing the Nd:YAG crystal. At the PRF of 110 kHz, 7.8-W yellow average output power is obtained under the incident pump power of 31.8 W, corresponding to an optical efficiency of 24.5%. However, the maximum average output power decreases apparently as the decline of the PRF and it is only 1.3 W at the PRF of 10 kHz. The hybrid Raman laser, since majority of the pump is absorbed by the Nd:YAG, could operate efficiently at relatively low PRFs, without serious decay in output power as what happened with typical Nd:YVO₄ self-Raman lasers, as shown in Fig. 5. As a result, both pulse energy and peak power can be enhanced. At the same time, a common Raman laser with Nd:YAG and YVO₄ used as fundamental and Stokes wave gain medium, respectively, is also carried out for comparison. A Nd:YAG crystal with typical parameters (3×3×10 mm³, 1.0-at.%) [11,17] instead of the Nd:YAG used in hybrid Raman laser is used for providing laser gain. This is because that the 0.4-at.% and 3×3×6-mm³ Nd:YAG cannot provide sufficient laser gain. A 3×3×20-mm³ YVO₄ crystal is used as Raman medium. A KTP crystal, cut at an angle of θ=68.7°, φ=0°, with dimensions of 3×3×15 mm³ is employed as the frequency-doubling medium. The total cavity length is ∼96 mm. At the PRF of 10 kHz, 2.0 W of yellow output is obtained under ∼25.2-W incident pumping, corresponding to an optical efficiency of 7.9%. It is significantly smaller than hybrid Raman laser. Unlike the other two regimes, the maximum average output power of yellow light decreases at high PRF. The intracavity power intensity of fundamental wave is limited by the allowed maximum pump power which is related to the thermal effect of Nd:YAG crystal. Moreover, the pulse energy decreases apparently with the increasing of the PRF. We deduce that these are the main causes of the phenomenon. The comparison results show that the hybrid Raman laser is a new, simple and reliable method to achieve efficient yellow source which is required for different fields [3,9,21–23].

 

Fig. 5. Maximum average output power of three kinds of yellow Raman laser at different PRFs.

Download Full Size | PDF

4. Conclusion

In this paper, we have demonstrated an efficient intracavity frequency-doubled Raman laser using both Nd:YAG and Nd:YVO₄ crystals to provide hybrid laser gain. The Nd:YAG crystal which could sustain much higher pump is used to absorb majority of pump for power scaling, while the Nd:YVO₄ absorbs the residual to provide stronger gain on π-polarization thus keeps the fundamental and Stokes wave nearly linear polarized. Therefore, the laser allows higher pumping and non-critical phase matching, and capable of operating efficiently with low PRF to achieve higher pulse energy and with high PRF to achieve higher average output power. The maximum 588-nm yellow output of 7.6 W at 110 kHz and 4.1 W at 10 kHz, respectively, are obtained under 42.0-W incident pumping.