Experimental Results and Discussions

We firstly use a simple laser setup for CW operation at 1342 nm to investigate the improvement of the thermal lensing effect in a double-end diffusion-bonded

Nd:YVO4 crystal [35]. For this investigation an output coupler with partial reflection at 1342 nm is used instead of the above-mentioned Raman cavity output coupler. The optimum reflectivity of the output coupler is found to be approximately 92–94%. The effective focal lengths of the thermal lens are estimated based on the fact that the laser system will start unstable for a cavity length longer than the critical length related to the thermal lensing. Even though the absolute accuracy is not easily achieved, this method is confirmed to provide the high relative accuracy for the effective focal lengths of the thermal lens [35]. Figure 4-2 shows the experimental data and fitted lines of thermal lensing power in a conventional crystal and a double-end diffusion-bonded crystal with the same dopant concentration. It can be seen that the effective focal length in a double-end diffusion-bonded crystal is nearly 1.6 times that in a conventional Nd:YVO4 crystal. As a result, the thermal effects can be substantiated to be significantly reduced in a double-end diffusion-bonded crystal.

When the Raman cavity output coupler is used in the laser cavity, the pumping threshold for the Raman laser output is found to be 2–3 W for the pulse repetition rates within 20-40 kHz. The beam quality factor is found to be better than 1.5 over the entire operating region. The spectrum of laser output is measured by an optical spectrum analyzer (Advantest Q8381A) employing a diffraction lattice monochromator with a resolution of 0.1 nm. As shown in Fig. 4-3, the optical spectrum for the actively Q-switched self-Raman output displayed that the fundamental laser emission is at 1342 nm and the Stokes component is at 1525 nm.

The frequency shift between Stokes and laser lines is in good agreement with the optical vibration modes of tetrahedral VO4-3 ionic groups (890 cm^-1) [36].

6 9 12 15 18 0

2 4 6 8

1/ f ( m-1 )

Input pump power (W) double-end

conventional

Fig. 4-2. Dependences of thermal lensing power on input pump power for conventional and double-end diffusion-bonded Nd:YVO4 CW laser at 1342 nm.

1300 1350 1400 1450 1500 1550 1600

1E-7 1E-6 1E-5 1E-4

Intensity (arb. units)

Wavelength (nm) 1342 nm

1525 nm

ω_R=890cm-1

Fig. 4-3. Optical spectrum of the diode-pumped actively Q-switch Nd:YVO4 self-Raman laser.

Figure 4-4 shows the experimental results of the average output power at 1525 nm with respect to the input pump power for the present self-Raman laser at pulse repetition rates of 20 and 40 kHz. For comparison, the previous results obtained by

Chen [14] with a conventional 0.2%-doped Nd:YVO4 crystal at a repetition rate of 20 kHz is also depicted in the same figure. Note that there were no experimental data for a conventional 0.2%-doped Nd:YVO4 crystal at a pulse repetition rate of 40 kHz because of the high lasing threshold. It can be seen that the Raman lasing threshold for a double-end diffusion-bonded Nd:YVO4 crystal is approximately 2.0 W that is substantially lower than the lasing threshold of 8.5 W for a conventional Nd:YVO4

crystal at the repetition rate of 20 kHz. Moreover, the lasing threshold at a pulse repetition rate of 40 kHz for present self-Raman laser is below 3.0 W. A rather low lasing threshold for high pulse repetition rates comes from the fact that the undoped part of the composite crystal increases the interaction length and then enhances the Raman gain.

Input pump power at 808 nm (W)

0 2 4 6 8 10 12 14 16 18 20

Average output power at 1525 nm (W)

0.0 0.5 1.0 1.5 2.0 2.5

40 kHz (double-end) 20 kHz (double-end) 20 kHz (conventional)

Fig. 4-4. The average output power at 1525 nm with respect to the input pump power at pulse repetition rates of 20 and 40 kHz shown as the down-triangle and circle symbols respectively for the double-end diffusion-bonded Nd:YVO4 crystal and that at 20 kHz shown as the square symbol for a conventional Nd:YVO4 crystal reported by Chen [14].

It has been experimentally evidenced that the maximum output power for a conventional self-Raman laser is limited by the critical pump power that induces a large temperature gradient in the gain medium to lead to the Raman gain lower than the cavity losses [1]. Consequently, the output power begins to saturate when the pump power exceeds the critical pump power. As shown in Fig. 4-4, the critical pump power for the self-Raman laser with a double-end diffusion-bonded Nd:YVO4 crystal can exceed 17.2 W that is limited by the available pump power and is considerably greater than the critical pump power of 13.5 W with a conventional Nd:YVO4 crystal.

As a result, at the pulse repetition rate of 20 kHz, the self-Raman laser with a double-end diffusion-bonded Nd:YVO4 crystal can generate the maximum average output power up to 1.72 W that is approximately 43% higher than the result with a conventional 0.2 %-doped Nd:YVO4 crystal [14]. At a repetition rate of 40 kHz, the maximum power at 1525 nm is even up to 2.23 W with an input pump power of 17.2 W, corresponding to a conversion efficiency of 13%. To the best of our knowledge, this is the highest average power for diode-pumped eye-safe self-Raman laser.

3.2ns

11ns Fundamental (1342 nm)

Raman (1525 nm)

ns/div 50

3.2ns

11ns Fundamental (1342 nm)

Raman (1525 nm)

ns/div 50

Fig. 4-5. Temporal characteristics of the fundamental and Raman pulses at a pulse repetition rate of 40 kHz with a pump power of 17.2 W.

The temporal traces for the fundamental and Raman pulses are recorded by a LeCroy digital oscilloscope (Wavepro 7100, 10 Gsamples/s, 1-GHz bandwidth) with two fast p-i-n photodiodes. At a repetition rate of 40 kHz the pulse energy is up to 56 μJ with an input pump power of 17.2 W and the pulse width is measured to be approximately 3.2 ns, as shown in Fig. 4-5. The corresponding peak power is higher than 17 kW. At the pulse repetition rate of 20 kHz, the maximum pulse energy is up to 86 μJ. Figure 4-6 shows the pulse width at a pulse repetition rate of 20 kHz with a pump power of 17.2 W. It can be seen that although a second tiny Raman pulse usually follows the main first peak, its contribution is rather limited. Consequently the peak power can be generally higher than 22 kW. Since the fundamental energy is remained after first Raman pulse, the sub-pulse of fundamental wave is formed shown as Fig. 4-5 and Fig. 4-6. At a pulse repetition rate of 20 kHz, the remaining energy is sufficient to reach Raman gain and a second tiny Raman pulse is produced shown as Fig. 4-6. The sub-pulse would not be generated if the reflectivity of output coupler was lowered.

3.8ns 7.6ns

Fundamental (1342 nm)

Raman (1525 nm)

ns/div 50

Fig. 4-6. Temporal characteristics of the fundamental and Raman pulses at a pulse repetition rate of 20 kHz with a pump power of 17.2 W.

4.3 High-Efficiency Q-Switched Dual-Wavelength Emission