1. Introduction

Stimulated Raman Scattering (SRS) can widely spread the wavelength of laser emission and has been proved to be an efficient method to obtain some significant all-solid-state coherent radiation sources, such as 1.17-μm [1–9] and 1.5x-μm [10,11] laser sources, etc. By exploiting a Raman medium with a certain Raman frequency shifting, the fundamental laser emitting from a common laser gain medium could be shifted to the required Stokes light [1–3], which has been the main and sophisticated approach over the last few decades. Generally, they can perform excellently at relatively low pulse repetition frequencies (PRFs) (a couple of tens of kilohertz or lower) [1,2]. With the development of the solid-state Raman lasers, researchers proposed a concept of self-Raman laser based on some crystals, such as Nd:YVO₄ [4–7], and Nd:GdVO₄ [8–10], which possesses excellent properties of laser and Raman gain simultaneously. Removing an extra Raman crystal, the self-Raman lasers could be more compact, lower insertion loss and more cost-effective [5–7]. Furthermore, the a-cut Nd:vanadates can generate π-polarized laser beam so that walk-off-free non-critical phase-matching with longer interaction length is allowed for the second harmonic generation (SHG) process, thus enhancing the overall conversion efficiency as well as the beam quality. Despite the advantages above, self-Raman lasers suffer some problems. For instance, serious thermal effect caused by extra thermal load from the SRS process cannot be ignored. Besides, particularly, the relatively short upper-laser-level lifetime of Nd:vanadates blocks self-Raman lasers from operating efficiently at low PRFs, and consequently the pulse energy is limited.

In this article, we report an efficient actively Q-switched 1176-nm hybrid gain intracavity Raman laser with laser gain from sequentially placed Nd:YAG and Nd:YVO₄ and Raman gain from the Nd:YVO₄. The combination of the two laser gain media enables the device to sustain higher pump and operate efficiently at lower PRFs compared with common vanadate self-Raman lasers. Meanwhile it generates the first-Stokes wave with high polarization ratio, which is helpful to further frequency conversion based on efficient type I phase matching but is difficult for traditional Nd:YAG-vanadate Raman lasers. With an incident pump power of 42.0 W, the highest pulse energy of 0.54 mJ was obtained at the PRF of 10 kHz. At the PRF of 100 kHz, the maximum average output power of 9.80 W was efficiently generated for the 1176-nm first Stokes wave, corresponding to the optical efficiencies of 23.3%. To our best knowledge, it is the highest pulse energy and average output power at 1176 nm ever reported for end-pumped linear intracavity Raman lasers.

2. Experimental arrangement

The experimental setup of the 1176-nm Nd:YAG/Nd:YVO₄ hybrid gain Raman laser is depicted in Fig. 1(a). A fiber-coupled laser diode array (core diameter of 400 μm, NA = 0.22) emitting at 808 nm was employed as the pump source. The pump beam was reimaged into a Nd:YAG and an a-cut composite Nd:YVO₄ crystal by a homemade 1:1.25 multi-lens coupler. The waist diameter of the pump beam was ~500 μm. The Nd:YAG crystal had a doping concentration of 0.4-at.% and dimensions of 3 × 3 × 6 mm³. The composite Nd:YVO₄-YVO₄ crystal which consisted of a 3 × 3 × 15 mm³ 0.2-at.% Nd-doped segment and a 3 × 3 × 5 mm³ un-doped segment was employed as both laser and Raman gain medium. The Nd:YAG crystal was placed in the front to absorb the majority of the 808-nm pump, while the doped segment of the Nd:YVO₄-YVO₄ composite crystal was set close to the Nd:YAG. According to [12,13] and our previous works, a lower temperature can improve the thermo-optic properties of the laser gain medium (Nd:YAG), consequently improve the laser performance especially for high pump power. Therefore, the Nd:YAG crystal was wrapped in indium foil and mounted in an aluminum holder which was cooled by refrigerant water at a low temperature of 5°C. The Nd:YVO₄-YVO₄ crystal was wrapped in indium foil and mounted in another aluminum holder. The water temperature was maintained at 20°C. The 20-mm-long acousto-optic Q-switch used here was driven by 10-W rf power with ultrasonic frequency of 80 MHz and was coated for AR (R<0.2%) at fundamental and Stokes wavelengths. A flat mirror M1 was coated for AR (R<5%) at the pumping wavelength of 808 nm and was high-reflection (HR) coated for at 1064 and 1176 nm. The output mirror M2 coated for HR at 1064 nm and partially transmissive (T = 23.8%) at 1176 nm was a concave mirror with a 200-mm radius of curvature (ROC). The cavity was set as short as possible to 69 mm.

 

Fig. 1 Schematics of (a) the 1176 nm Nd:YAG/Nd:YVO₄ hybrid gain Raman laser (b) the 1176 nm Nd:YVO₄ self-Raman laser (c) the 1176 nm Nd:YAG/YVO₄ Raman laser.

Download Full Size | PDF

For comparison, the performance of a Nd:YVO₄ self-Raman laser and a common Nd:YAG/YVO₄ Raman laser were also investigated with the same pump source. The arrangement of the self-Raman laser is shown in Fig. 1(b). We set the un-doped segment of the Nd:YVO₄ crystal which was identical to the one in Fig. 1(a) close to the pump source for relieving thermal effect. Figure 1(c) depicts the schematic of Nd:YAG/YVO₄ Raman laser. A 1.0-at.%-doped 10-mm Nd:YAG crystal and an un-doped YVO₄ crystal with dimensions of 3 × 3 × 20 mm³ were adopted as the laser medium and Raman medium, respectively. They were all anti-reflectively (AR) coated at 1000-1200 nm on both facets. The cavity mirrors were identical to them in Fig. 1(a). The cavities lengths of the self-Raman and Nd:YAG/YVO₄ Raman laser were 62 and 72 mm, respectively.

3. Experimental results and discussion

We first studied the performance of Q-switched Nd:YAG/Nd:YVO₄, Nd:YVO₄, and Nd:YAG/YVO₄ lasers at 1064 nm, respectively. As we know, the gain spectra of Nd:YAG and π-polarization of Nd:YVO₄ are different and it usually varies with temperature, stress, etc [14–16]. Therefore the detailed spectra of output laser at different PRFs should be learned for further investigating the Nd:YAG/Nd:YVO₄ hybrid laser gain Raman laser. Figures 2(a) and 2(b) are the normalized spectra of the Nd:YAG, Nd:YVO₄, and Nd:YAG/Nd:YVO₄ lasers at the PRFs of 10 and 100 kHz, respectively. The spectra of the Nd:YAG, Nd:YVO₄ and Nd:YAG/Nd:YVO₄ lasers were recorded under the incident pump power of 42.0, 17.0 and 42.0 W, respectively, with corresponding absorbed pump power of ~25.0, ~16.7 and ~41.0 W(without considering the loss of facets between Nd:YAG and Nd:YVO₄), respectively. Comparing the two graphics in Figs. 2(a) and 2(b), respectively, the Nd:YAG contributes more for the output of the Nd:YAG/Nd:YVO₄ hybrid gain laser at the PRF of 10 kHz and is as important as Nd:YVO₄ at the PRF of 100 kHz, which is consistent with our expectation.

 

Fig. 2 Spectra of the Nd:YAG, Nd:YVO₄, and Nd:YAG/Nd:YVO₄ lasers at (a) 10 kHz and (b) 100 kHz.

Download Full Size | PDF

Figure 3(a) depicts the average output power at 1064 nm versus the incident pump power at the PRFs of 10 and 100 kHz, respectively. Put behind the Nd:YAG (0.4-at.%), the Nd:YVO₄ served as a laser medium which can fully absorb the residual pump light, while shared the heat produced in the laser emission process, consequently guaranteed the Nd:YAG/Nd:YVO₄ hybrid gain laser to maintain working efficiency. With the incident pump power of 42.0 W, 19.1-W average output power at 1064 nm was obtained at the PRF of 100 kHz, which contained a ~17.0-W π-polarization emission and a ~2.1-W σ-polarization emission, respectively. The polarization ratio reduced with the decreasing of the PRF and decreased to ~5:1 at the PRF of 10 kHz. All of them are less than 15:1 which was mentioned in [17]. It is worth mentioning that the polarization ratio can vary with the alignment state of cavity. In addition, as we can see that the average output power of the Nd:YAG/YVO₄ laser reached saturated at high pump power. The serious thermal effects should be responsible for the inferior performance. Though, compared with the other two lasers, the average output power of the Nd:YVO₄ laser at 100 kHz were always the maximum, its performances at 10 kHz were the worst, which should be imputed to the short upper-laser-level lifetime of Nd:vanadates media.

 

Fig. 3 Average output power of the (a) fundamental laser and (b) the first-Stokes light versus the incident pump power at different PRFs.

Download Full Size | PDF

We further investigated the operation of the first Stokes at 1176 nm with a Stokes output coupler (OC). We chose three mirrors with transmission of 9.8%, 15.6% and 23.8%, respectively, as the first-Stokes light OC. Maximum Stokes light outputs of the three lasers were achieved with different OCs, respectively (9.8% for Nd:YVO₄ self-Raman laser and 23.8% for the other two Raman lasers). The results given bellow are all with the suitable OC. Figure 3(b) shows the average output power of the three Raman lasers at 1176 nm versus the incident pump power. Under low pump power (less than ~23 W), the output of the Nd:YVO₄ self-Raman laser at 100 kHz is significantly higher than the other two Raman lasers. The lower loss caused by the using of low-transmission output coupler may be the main cause. The output power of Nd:YVO₄ self-Raman laser and Nd:YAG/YVO₄ Raman laser became saturated and even rolled over with the increasing of the pump power. Such behavior should be attributed to the serious thermal effects under high pump power. It is noteworthy that, though the saturation point can be improved with a more proper gain medium and cavity design, only the low-PRF performance of Nd:YAG/YVO₄ Raman laser and only high-PRF performance of Nd:YVO₄ self-Raman laser would be as good as hybrid gain Raman laser. In addition, we can observe that the output power of Nd:YVO₄ self-Raman laser decreased sharply under incident pump power of ~31 W at the PRF of 100 kHz. We recorded the beam profile by an Ophir Pyrocam III at the time, as shown in the bottom-right inset of Fig. 3(b). Obviously, higher transverse mode rather than the fundamental mode oscillated, which may be the major cause of the sharp decline. Though we did not observe similar behavior for the hybrid gain Raman laser, the optical efficiency declined obviously where the pump power exceeded 30.1 W at the PRF of 100 kHz. It declined from 27.4% to 23.3% while the pump power increased from 30.1 W to 42.0 W. Usually, the performance of Stokes light has much to do with the fundamental light. However, we can indicate from Fig. 3(a) that the output of 1064 nm still performed quite well when the pump power exceed 30.1 W. Therefore it is reasonable to question whether the degraded beam quality deteriorates the Stokes light. We measured the M² of 1176-nm Stokes light at the pump power of 30.1 W, as shown in the inset of the Fig. 3(b). The beam quality (M² = ~2.9) at 30.1-W pump power was apparently much better compared with that at the maximum pump power (M² = 5.05). Therefore it is reasonable to believe that beam quality degradation is the major cause of the issue. Under incident pump power of 42.0 W, 9.80 W 1176-nm light was generated for hybrid gain Raman lasers at the PRF of 100 kHz, with corresponding optical efficiencies to be 23.3%. The SRS threshold was 5.30 W. When the PRF decreased to 10 kHz, the SRS threshold decreased to 3.15 W because of the increasing of pulse energy and peak power of fundamental light. Maximum average output power was 5.42 W under the incident pump power of 42.0 W, with corresponding optical efficiencies of 12.9%. We measured the pulse energy of hybrid gain Raman laser at the maximum average output power and it was larger than 0.54 mJ. Compared with the other two Raman lasers, the hybrid gain Raman laser can keep operating efficiently at the PRFs from hundred to ten kilohertz. We deduce that it benefits from two reasons: firstly, the 1064-nm oscillation was relatively insensitive to the PRFs changes because of the long upper-laser-level lifetime of Nd:YAG crystal; secondly, the construction of the double crystals contributed to alleviate the thermal effect, and both of them could provide laser gain (hybrid laser gain), which guaranteed the quality of fundamental light. It is worth mentioning that the output may be higher if more pump power can be provided.

Using a knife-edge method, the size of the laser spot at different beam locations were measured at the maximum output power as shown in Fig. 4. The 1176-nm output beam quality factor M² was calculated to be ~5.89 in the horizontal direction at the PRF of 10 kHz, and ~5.05 at the PRF of 100 kHz, respectively. The insets of Fig. 4 are the beam profiles achieved at the maximum output power.

 

Fig. 4 M² measurement and beam profile at the maximum output power at the PRF of (a) 10 kHz and (b) 100 kHz.

Download Full Size | PDF

The temporal behavior of the Raman lasers was recorded by using a photoelectric detector Thorlabs DET08C and an oscilloscope Tektronix DPO2024B. Figure 5 shows the output pulse duration as functions of incident pump power. The pulse durations shortened along with increasing pump power. The shortest pulses duration of the hybrid gain Raman laser is 10.6 ns at the maximum pump power with the PRF of 10 kHz. We integrated the area of the pulses and found that it was around 61.2% for the first peak. Thus, the peak power was estimated to be ~31.0 kW. The inset is the typical oscilloscope trace of the hybrid gain Raman laser at maximum power with 10 kHz PRF. A satellite pulse being analogous to that mentioned in [18] was also observed in the oscilloscope trace. The Raman laser works above the optimum value of Raman gain may cause this phenomenon [18,19].

 

Fig. 5 The output pulse durations as functions of incident pump power at different conditions.

Download Full Size | PDF

It is noteworthy that we observed blue luminescence emitting from the Raman-active crystal, which was also reported by some researchers with different explanations [20–22]. In order to verify the origin of this phenomenon, we focused the 1064- and 1176-nm laser, respectively, into a segment of Tm³⁺-doped double cladding silica fiber. As we can see from the results shown in Figs. 6 (a) and 6(b), both first-Stokes and fundamental light can stimulate the blue emission and the intensity of the former was obvious stronger. In addition, faint red light was also observed in the spectra. The center wavelengths of the blue and red emission are ~475 and ~650 nm, respectively, which is consistent with the result shown in Fig. 6(c). Thus, we consider that the unwanted Tm³⁺ impurity involved in Raman-active medium should be responsible for the blue emission, and this explanation coincides with the idea in [20].

 

Fig. 6 Spectra detected from the side of the Tm³⁺-doped fiber and corresponding photos of the blue luminescence were pumped by (a) 1064-nm fundamental light, (b) 1176-nm first-Stokes light. (c) The image of the blue-luminescence spectrum and photo of Nd:YVO₄ crystal with the blue luminescence when Nd:YVO₄ self-Raman laser was operating.

Download Full Size | PDF

4. Conclusion

In summary, an efficient actively Q-switched 1176-nm Nd:YAG/Nd:YVO₄ hybrid gain intracavity Raman laser with the ability that can operate efficiently at the PRFs from ten to hundred kilohertz was demonstrated for the first time. The maximum pulse energy of 0.54 mJ is achieved at the PRF of 10 kHz, and the maximum average output power of 9.80 W is generated at the PRF of 100 kHz. Thanks to the construction of coaxial double crystals, we achieve the high output of the first-Stokes wave with relatively high polarization ratio. We believe that, through further optimization, it can be a promising method for generation of efficient yellow emission by frequency-doubling of the 1176-nm Stokes light.