Experimental results and discussions

Figure 3.4 depicts the average output power versus the launched pump power in CW and passive Q-switching operation. The external cavity in the CW operation contained only a re-imagining lens and a reflective mirror without the saturable absorber. At a launched pump power of 16 W, the CW PCF laser was found to generate an output power of 8.7 W, corresponding to a slope efficiency of 78%. In the passive Q-switching operation, the average output powers at a launched pump power of 16 W were 7.1 W, 7.7 W, and 8.0 W for the lasers with the saturable absorbers of 3 × 50, 3 × 30, and 2 × 30 QWs, respectively. The signal intensity of the amplified spontaneous emission (ASE) is 40 dB below the lasing signal of 1030 nm measured by the optical spectrum analyzer, so the fraction of the ASE output power can be neglected. As a result, the Q-switching efficiency (the ratio of the average power of Q-switched operation to that of CW one) were approximately 82%, 89%, and 92% for the lasers with the saturable absorbers of 3 × 50, 3 × 30, and 2 × 30 QWs, respectively. The overall Q-switching efficiency was significantly superior to the results obtained with Cr⁴⁺:YAG crystals as saturable absorbers [11]. The lasing spectra for CW and passive Q-switching operations were quite similar with the peaks near 1030 nm and bandwidths

to be approximately 0.4 nm. The laser output was found to be linearly polarized with an extinction ratio of approximately 100:1, evidencing the function of the polarization maintaining in PCF. The M2 factor was found to be generally smaller than 1.3 over the entire output power range, owing to the low-NA feature of the PCF.

Figure 3.5 shows the pulse repetition rates in the passive Q-switching operation versus the launched pump power. Experimental results reveal that the pulse repetition rates for all cases increase monotonically with the pump power. At a launched pump power of 16 W, the pulse repetition rates were found to be 6.5 kHz, 16 kHz, and 23 kHz for the lasers with the saturable absorbers of 3 × 50, 3 × 30, and 2 × 30 QWs, respectively. With the experimental results of the average output power and the pulse repetition rate, we calculated the pulse energies versus the launched pump power. It was found that the pulse energies were nearly independent of the pump power and their average values were 1.1 mJ, 0.49 mJ, and 0.35 mJ for the lasers with the saturable absorbers of 3 × 50, 3 × 30, and 2 × 30 QWs, respectively. Fiber laser systems with energy of millijoule-class had been demonstrated with either actively Q-switched oscillator [15-17] or the master oscillator power fiber amplifier scheme [18-20]. To the best of our knowledge, this is the first time that the millijoule-level energy output was achieved with the passive Q-switching scheme in a PCF laser.

Fig 3. 4 Average output power with respect to launched pump power in CW and

Fig 3. 5 Pulse repetition rates in the passive Q-switching operation versus the launched pump power.

Figures 3.6(a)–3.6(c) depict typical oscilloscope traces for the single Q-switched pulses of the lasers with the saturable absorbers of 2 × 30, 3 × 30, and 3 × 50 QWs, respectively. It can be seen that the temporal shape of the single Q-switched pulse obtained with the absorber of 2 × 30 QWs is a simple pulse, whereas the temporal shape obtained with the absorber of 3 × 50 QWs reveal conspicuous modulation whose period is nearly equal to the round trip time. The self-modulation phenomenon inside the Q-switched envelope has been frequently observed in pulsed fiber lasers. This phenomenon is generally considered to arise from the stimulated Brillouin scattering (SBS) which can provide strong feedback to the cavity together with pulse compression.

The SBS-related pulses have been demonstrated in different fiber laser designs, such as self-Q switched [21-23], actively Q-switched [24,25], and passively Q-switched [26,27]

fiber lasers. Note that another self-modulation phenomenon was found in passively Q-switched Nd-doped crystal lasers with Cr⁴⁺:YAG crystals as saturable absorbers [28-31]; however, the origin is attributed to the excited-state absorption of the absorber and the fluctuation mechanism [32,33]. Our results reveal that the pulse energy obtained with the absorber of 3 × 30 QWs is just above the SBS threshold. As seen in Fig. 3.6(b), the rear end of the pulse exhibits a fast transient dynamics. On the other

hand, the intense SBS effect leads to the pulse to be strongly modulated, as seen in Fig.

3.6(c). With the numerical integration, we found the maximum peak powers were 7.4 kW, 12.8 kW, and 110 kW for the lasers with the saturable absorbers of 2 × 30, 3 × 30, and 3 × 50 QWs, respectively. The corresponding optical intensity on the 3 × 50 QWs was 350 MW/cm² which is quite close to the damage threshold of the saturable absorber, but no optical damage was observed. Figures 3.7(a)–3.7(c) show typical oscilloscope traces of a train of output pulses obtained with the saturable absorbers of 2

× 30, 3 × 30, and 3 × 50 QWs, respectively. It can be seen that for the laser with the absorber of 2 × 30 QWs the pulse-to-pulse amplitude fluctuation was generally less than 4% in rms. Even for the case of 3 × 30 QWs, just above the SBS threshold, the pulse-to-pulse amplitude fluctuation was also smaller than 4% in rms. Although the strong SBS effect might deteriorate the pulse stability to some extent, the pulse-to-pulse amplitude fluctuation could still be maintained to be 8.5% in rms for the laser with the saturable absorber of 3 × 50 QWs, as shown in Fig. 3.7(c). Compared with the previous results, the pulse stability was superior to that obtained in Ref. 9 and slightly diminished with respect to Ref. 11 as a result of the high pulse energy induced SBS effect. The overall pulse energy scaling was 2.4 times as high as the one in Ref. 9 and 1.8 times as that in Ref. 11.

Fig 3. 6 Typical oscilloscope traces for the single Q-switched pulses of the lasers with the saturable absorbers of (a) 2 × 30, (b) 3 × 30, and (c) 3 × 50 QWs,

respectively.

2x30 QWs 20 ns/div

3x30 QWs 20 ns/div

10 ns/div 3x50 QWs

(A)

(B)

(C)

Fig 3. 7 Typical oscilloscope traces for a train of output pulses of the lasers with the saturable absorbers of (a) 2 × 30, (b) 3 × 30, and (c) 3 × 50 QWs, respectively.

2x30 QWs

(A) 100 us/div

100 us/div (B) 3x30 QWs

3x50 QWs

3.1.4 Conclusion

In conclusions, we have, for the first time to my knowledge, demonstrated a millijoule-level passively Q-switched Yb-doped photonic crystal fiber laser with AlGaInAs QWs as a saturable absorber. At a launched pump power of 16 W, the average output powers were 7.1 W, 7.7 W, and 8.0 W for the lasers with the saturable absorbers of 3 × 50, 3 × 30, and 2 × 30 QWs, respectively. The pulse energies were found to be 1.1 mJ, 0.49 mJ, and 0.35 mJ for the lasers with the saturable absorbers of 3

× 50, 3 × 30, and 2 × 30 QWs, respectively. The maximum peak power could be up to 110 kW. The overall pulse-to-pulse amplitude fluctuation and the temporal jitter could be maintained to be well below 10% in rms. These high-pulse-energy high-peak-power passively Q-switched PCF lasers are potentially useful light sources for many technical applications.

3.2 A widely tunable eye-safe based on a passively