• 沒有找到結果。

High­Power 946 nm Nd:YAG Laser and Blue Laser at  473 nm in QCW operation

Diode­Pumped Nd:YAG Laser at 0.946 μm  and Intracavity Frequency Doubling

2.3    High­Power 946 nm Nd:YAG Laser and Blue Laser at  473 nm in QCW operation

With the rapid development of laser diode, high-power 2 D laser diode arrays (LDA) or diode stacks have been used in diode-pumped solid-state lasers to generate high output pulse energies in free-running or Q-switched operations for several years.

Compared with the conventional high-power pump source, flash lamps, the 2D diode stacks, consisting of several diode bars, possess many advantages of higher repetition rates, higher overall efficiency, much lower thermal loading, and higher reliability.

High-power diode stacks are operated in the pulse durations of several hundred micro-seconds, i.e. quasi-cw (QCW) operations, and emit high peak powers (on-time average powers) of several hundred watts. The QCW pumping technique provides not only the high-level power but also the average heat-loading reduction. In this part, we used a 2D diode stack in QCW operation as the pump source. A high-power, QCW diode-pumped Nd:YAG laser operating at 946 nm and its intracavity frequency doubling to 473 nm with a BiBO crystal were demonstrated.

2.3.1    High‐Power QCW pump source

 

Here, the pump source is a high-power QCW diode stack (Quantel laser diodes) that consists of three 10-mm-long diode bars generating 130 W per bar, for a total of 390 W at the central wavelength of 808 nm. The diode stack is designed with 0.4 mm spacing between the diode bars so the overall area of emission is approximately 10 mm (slow axis) × 0.8 mm (fast axis). The images of the diode stack are shown in Fig. 2.3.1. The full divergence angles in the fast and slow axes are approximately 35o and 10o, respectively. This divergence causes a great loss of pump power.

Therefore, the highly efficient and simple optical devices, coupling the output from a laser diode stack into a gain medium, are especially important for the design of high power lasers with large laser diode stack as the pump source.

A lens duct was reported to have good coupling efficiency and was used to

41 

end-pump solid-state lasers [19, 20]. The lens duct is a glass device that can consist of one spherical input surface and five planar surfaces, as shown in Fig. 2.3.2 (a).

The spherical surface is designed for efficient collection of the output radiation from a laser diode stack into the duct. Schematic of ray tracing inside a lens duct is depicted in Fig. 2.3.3. The light rays are totally reflected from the four side surfaces until they reach the planar output surface, and then injected into the gain medium. The output surface of lens duct has to adjoin the gain medium to reduce the coupling loss.

Fig. 2.3.2 (b) shows the lens duct assembly. For practical reasons the lens duct has to be made of two pieces glued together. The first piece is a slice of a commercial plano–cylindrical lens, which provides the curved face. The second piece is a lens duct all of whose faces are plane. The curved surface of the plano–cylindrical lens and the lens duct output face are not antireflection coated.

In comparison with other coupling methods such as optical fibers [21], gradient-index (GRIN) lenses [22], or aspheric lenses [23], the lens duct has the benefits of simple structure, high coupling efficiency, and impervious to slight misalignment. These benefits are practically important for the end-pumped solid-state lasers with laser-diode stacks in which there is a significant geometric mismatch between the effective diode emitter area and the available input aperture of the gain medium.

42 

Fig. 2.3.1. (a) Image of the Quantel laser diode and (b) the near-field image of laser diode emitters at 30 A.

(a)

(b)

43 

Fig. 2.3.2. (a) Schematic of a lens duct with five geometric parameters of r, L, H1, H2, and H3: r is the radius of the input surface, L is the length of the duct, H1 is the width of the input surface, H2 is the width of the output surface, and H3 is the thickness of the duct. (b) Lens duct assembly.

H

1

L

r H

2

H

3 (a)

(b)

44 

Fig. 2.3.3. Schematic of ray tracing inside a lens duct: (a) top view in the slow-axis plane [20], (b) side view in the fast-axis plane.

(a)

(b)

45 

2.3.2    Experimental Setup

 

In this experiment, the lens ducts were manufactured with the same parameters of r = 10 mm, L = 30 mm and H1 = 12 mm, but the various output cross-sections with H2

× H3 at mm2 of 2.1×2.1, 1.6×1.6 and 1.2×1.2, respectively. The coupling efficiencies of three lens ducts were experimentally found to be approximately 86, 78, and 72%, respectively. The active medium was 3×3×5 mm3 Nd:YAG crystal with 1.1 at.%.

Approximately 75% of the pump light was absorbed in the active medium. The entrance surface of the laser crystal was coated with high reflection at 946 nm (R>99.8%) and high transmission at 808 nm (T>90%) and 1064 nm (T>85%). The other surface of the laser crystal was coated for antireflection at 946 nm (R<0.2%).

Fig. 2.3.4 (a) shows the schematic diagram of QCW 946 nm Nd:YAG laser.

The output couplers were flat and coated for high transmission at 1064 nm (T>90%) and partial reflection at 946 nm (R=97%). The cavity length was approximately 10 mm. Fig. 2.3.4 (b) shows the experimental configuration of cw intracavity frequency-doubled Nd:YAG laser. The frequency doubler was a 5-mm-long BiBO crystal, cutting for type-I critical phase-matching (θ=161.7o, φ=90o) at room temperature. Both facets of the BiBO crystal were coated for anti-reflection at 946 and 473 nm (R<0.2%) to reduce the reflection loss in the cavity. The laser crystals were wrapped with indium foil and mounted in a water-cooled copper block. A plano-concave mirror of 50 mm radius was chosen to be the output coupler coated for high reflection at 946 nm (R>99.8%), high transmission at 1064 nm (T >70%), and high transmission at 473 nm (T>80%), respectively. The SHG cavity length of was approximately 30 mm.

The pulse temporal behavior was recorded by a LeCroy digital oscilloscope (Wavepro 7100; 10 G samples/sec; 1 GHz bandwidth) with a fast InGaAs photodiode.

The spectral information of the laser was monitored by an optical spectrum analyzer (Advantest Q8381A). The spectrum analyzer employing diffraction grating monochromator can be used for high-speed measurement of pulse light with the resolution of 0.1 nm.

46 

Fig. 2.3.4. Schematics of QCW diode-pumped Nd:YAG lasers at 946 nm in (a), and intracavity frequency-doubling at 473 nm in (b).

(a)

(b)

47 

2.3.3    Experimental Results and Discussion

 

First of all, the QCW free running operation of the Nd:YAG laser at 946 nm was measured without the BiBO crystal in the cavity to confirm the pump efficiency of the lens duct and the quality of the laser crystal. In this experiment, the laser-diode stack was set to emit pump pulses of 270 µs at a repetition rate of 35 Hz, as the duty cycle was approximately 1 %. Three kinds of output cross-sections of lens ducts were employed for comparisons in the quasi-continuous-wave free running operation. Fig.

2.3.5 (a) shows the experimental results of the output pulse energy at 946 nm with respect to the pump energy emitted from the laser diode in the free-running operation.

It can be found that the lower pump threshold and higher output energy were achieved by employing a lens duct with a smaller output dimension. With the lens duct with the smallest output surface of 1.2×1.2 mm2, the lower pump threshold of 36 mJ and the maximum output pulse energy of 6.4 mJ were obtained at a pump energy of 105 mJ. The experimental results confirm that a high-intensity pump light is necessary for such quasi-three-level laser. Note that the smallest dimension of a lens duct is limited by the emission area of laser diode stack. Fig. 2.3.5 (b) depicts the pulse train at repetition rate of 35 Hz, at the maximum pump energy. Fig. 2.3.6 (a)-(c) depict the temporal shapes of the single pulse exhibited relaxation-oscillation driven spikes for the employed lens ducts with output dimensions of 2.1×2.1, 1.6×1.6, and 1.2×1.2 mm2, respectively. It can be seen that the relaxation-oscillation driven spikes and the pulse shape of laser of a single pulse were dominated by the pump intensity. Moreover, due to the laser buildup time, the output laser pulse was shorter than the pump duration of 270 µs in the QCW operation. The on-time average output power can be estimated with the laser output pulse width and pulse energy.

The on-time average output power at 946 nm versus the on-time average pump power is plotted in Fig. 2.3.7. Using a lens duct with the output dimension of 1.2×1.2 mm2, the maximum on-time average output power was 34 W at the on-time average pump power of 392 W. The overall slope efficiency of the three curves were nearly similar to be 13 %.

The QCW diode-end-pumped Nd:YAG laser at 473 nm by intracavity frequency

48 

doubling was operated by inserting a nonlinear crystal BiBO and replacing of the above mentioned concave mirror. The experimental results of the output pulse energy at 473 nm versus the pump energy are plotted in Fig. 2.3.8. It reveals that the pump intensity is extremely critical for an efficient SHG of the 946-nm low-gain laser.

Using a lens duct with the output dimension of 1.2×1.2 mm2, the maximum output pulse energy of approximately 1.75 mJ was achieved for 105 mJ pump energy from the laser diode stack. Fig. 2.3.9 shows the estimated on-time average output power at 473 nm as a function of the on-time average pump power. The maximum on-time average output power at 473 nm of approximately 9 W was estimated at 392 W of the on-time average pump power. The conversion efficiency of blue light at 473 nm with respect to the average output power of the free running performance was greater than 26 %. To our best knowledge, it is the highest average power for intracavity frequency-doubling in the blue region at 473 nm of diode-end-pumped Nd:YAG laser.

Fig. 2.3.10 (a)-(b) depict the pulse train and temporal shape of the single pulse at the maximum pump energy of 105 mJ by using a lens duct with the output dimension of 1.2×1.2 mm2. Under the optimal alignment condition, the pulse-to-pulse amplitude fluctuation was estimated to be approximately ±10 %. The spatial distribution of the output blue-light beam was recorded with a CCD as displayed in Fig. 2.3.10(c). The beam quality factors were measured and found to be M < 12 and x2 M2y < 7, repetitively, where the x and y directions are parallel to the slow and fast axes of the laser-diode stack. The asymmetry of the M factors in the x and y directions was 2 due to the walk-off effect by nonlinear crystal.

49 

Fig. 8. Average output powers, with and without the ISA inserted into the cavity with the OC1, versus the incident pump power.

Fig. 2.3.5. Experimental results of (a) the free-running operation for the output pulse energy at 946 nm versus the pump energy for three kinds of output surfaces of lens ducts, and (b) pulse train at the maximum pump energy.

0 10 20 30 40 50 60 70 80 90 100 110

0 1 2 3 4 5 6 7

LensDuct 2.1x2.1 LensDuct 1.6x1.6 LensDuct 1.2x1.2

Output energy at 946 nm (mJ)

Pump energy at 808 nm (mJ)

(a)

(b)

50 

Fig. 2.3.6. Temporal shapes of the single pulse for the lens duct output surface of (a) 2.1×2.1 mm2; (b) 1.6×1.6 mm2; (c) 1.2×1.2 mm2.

51 

Fig. 2.3.7. Estimated on-time average output power at 946 nm versus the on-time average pump power in the free-running performance, for three kinds of output surfaces of lens ducts.

0 50 100 150 200 250 300 350 400 450

0 5 10 15 20 25 30 35 40

LensDuct 2.1 x 2.1 LensDuct 1.6 x 1.6 LensDuct 1.2 x 1.2

On-time output power at 946 nm (W)

On-time pump power at 808 nm (W)

52 

Fig. 2.3.8. Experimental results of the free-running operation for the output pulse energy at 473 nm versus the pump energy for three kinds of output surfaces of lens ducts.

0 10 20 30 40 50 60 70 80 90 100 110

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Output energy at 473 nm (mJ)

Incident pump energy at 808 nm (mJ) LensDuct 2.1x2.1

LensDuct 1.6x1.6 LensDuct 1.2x1.2

53 

Fig. 2.3.9. Estimated on-time average output power at 473 nm as a function of the on-time average pump power for the lens duct with an output surface of 1.2×1.2 mm2.

0 50 100 150 200 250 300 350 400

0 2 4 6 8 10

On-time output power at 473 nm (W)

On-time pump power at 808 nm (W) LensDuct 1.2x1.2

54 

(a)

(b)

(c)

Fig. 2.3.10. Experimental results for the lens duct with an output surface of 1.2×1.2 mm2: (a) pulse train at the maximum pump energy of 105 mJ; (b) temporal shape of the single pulse at the maximum pump energy of 105 mJ; (c) Spatial distribution of the output blue beam recorded with a CCD.

55 

2.4    High­Peak­Power Passively Q­Switched Nd:YAG Laser