1. Introduction

High-repetition-rate high-power diode-pumped solid state lasers (DPSSL) are attractive devices in industrial processing, such as marking, drilling, cutting and so on, since its high efficiency scanning and excellent processing quality [1, 2]. The high-repetition-rate DPSSLs are also widely used in some others applications such as scientific research, medical treatment, military systems and so on [3, 4, 5, 6, 7]. The high repetition rate (> 100 kHz) operation can be achieved by Q-switching, especially actively Q-switching method for its stable pulse energy and low temporal jitter at high repetition rates. Acousto-optics (AO) Q-switching operation takes the advantages of its high repetition rate and high peak power of pulses; while, however, its pulse repetition frequencies (PRF) are usually less than 100 kHz. The lasers with grazing incidence configuration are often used to obtain high-repetition-rate output [8, 9, 10, 11, 12, 13, 14], but the output power are always low with relatively poor beam quality. End-pumped geometry is a common and useful configuration for high-power output with good beam quality; however, it is hard to obtain AO Q-switching operation with very high repetition rate in a common diode-end-pumped configuration, since the high gain should be obtained with high concentration crystal and high pump intensity which will result in the thermal fracture [15, 16].

In 2003, García-López et al. reported a 500 kHz AO Q-switched slab Nd:YVO₄ laser with 16 W average power output, and the grazing incidence cavity geometry was adopted [8]. In 2005, Minassian et al. demonstrated a 400 kHz AO Q-switching diode-side pumped bounce slab MOPA Nd:GdVO₄ laser with average power of 101 W [9, 10, 11]. In 2006, Omatsu et al. used an AO Q-switched Nd doped mixed gadolinium yttrium vanadate (Nd:Gd_xY_1–xVO₄) bounce laser, and obtained 17 W average power output with repetition rate of 650 kHz with the pulse duration less than 40 ns. They also predicted that the maximum pulse repetition frequency for stable Q-switching is expected to be around 1 MHz based on pure vanadate bounce laser [12]. The beam quality and the stability of pulse in Ref. [8]–[12] are not reported. In 2007, our group reported a 1 MHz AO Q-switched slab Nd:YVO₄ laser with a grazing incidence geometry and the stability of pulse peak value < 15% (RMS) [13]. In 2008, our group reported a 2.2 MHz AO Q-switching grazing laser with 3 at.% neodymium doped Nd:YVO₄ which is the highest pulse repetition rate ever reported based on AO Q-switching. The 8.6 W TEM₀₀ mode output at 2.1 MHz and 10 W multi-mode output at 2.2 MHz were obtained respectively [14]. In 2007, we also reported a 108 W average power AO Q-switched end-pumped MOPA Nd:YVO₄ at maximum pulse repetition frequency of 500 kHz, and the relatively poor beam quality was measured as about two-times diffraction-limit [17]. The composite Nd:YVO₄/YVO₄ was used, and the MOPA system included a oscillator stage and tow amplifier stages. The maximum pulse repetition frequency was limited by relatively weak gain at very high pulse repetition frequency which corresponds to the optimized design of the oscillator cavity, and the thermal distortion effect resulted in the poor beam quality of the MOPA laser.

In this work, we presented a high-average-power, high-repetition-rate and good-beam-quality AO Q-switched Nd:YVO₄ MOPA laser based on the dual-end-pumped configuration. The composite Nd:YVO₄ was used to prevent the thermal fracture under high power and high intensity pumping. The optimized master oscillator was amplified using the four-stage power amplifiers, and the pumping/gain units in the MOPA system were all the same while no coupler and isolator was used between the amplifiers for the compact and simple design. 183.5 W average power was obtained at the pulse repetition rate of 850 kHz with pulse instability less than 12%, and the beam quality was near diffraction-limit with M ² factor measured as M ² _x = 1.28 and M ² _y = 1.21 in the orthogonal directions respectively. 195 W output in CW operation was also achieved corresponding to the absorbed-output efficiency of 53.3%.

2. Experimental setup

The experimental setup of the MOPA system is shown schematically in Fig. 1.

 

Fig. 1. Experimental setup of the acoustic-optic Q-switched MOPA laser system.

Download Full Size | PDF

The Nd:YVO₄ MOPA laser system consisted of a high power oscillator stage and four single pass power amplifier stages. The Nd:YVO₄ crystal was used as the gain medium for its efficient 1064 nm four-level operation. Moreover, The Nd:YVO₄ crystal was with large stimulated emission cross section (σ ₂₁ ~ 25 × 10^-25cm²) and high effective absorption coefficient (α_eff ~ 28cm^-1 @1.1 at.%), which can provide high gain and are benefit to the high-repetition-rate operation. Meanwhile, its short upper-state lifetime (τ_f ~ 90μs) leads to the faster building-up of the pulses to achieve narrower pulse duration. Therefore, The Nd:YVO₄ crystal was an excellent laser medium for Q-switching operation at high repetition rate with high pulse peak power. The composite Nd:YVO₄ crystal was used in our experiment. The a-cut 0.3 at.% Nd-ion concentration bulk Nd:YVO₄ (3 mm × 3 mm × 16 mm) was thermally bonded with two YVO₄ end caps (3 mm × 3 mm × 2 mm). The low doping-concentration Nd:YVO₄ (peak absorption coefficient of α _p ~ 19 cm^-1 and effective absorption coefficient of α_eff ~ 10 cm^-1 @ 0.3 at. % Nd:YVO₄ respectively) leads to relatively weak thermal effects, especially the thermal fracture and the thermal lensing effect. In diode-pumped solid state lasers, the major sources of heat production are quantum defect [18], non-radiative transition [19], fluorimetric quenching induced by impurity and defect such as energy-transfer-upconversion (ETU), excited state absorption (ESA) and cross relaxation [20], and the Auger recombination [21]. Additional heat load produced by nonlinear processes mentioned above, rather than the quantum defect of the pumping, aroused more serious heat generation, especially for high doping concentration crystal and Q-switching operation. Therefore, the low doping-concentration Nd:YVO₄ was used to reduce the thermal loading density and achieve more uniform absorption under high intensity pumping. Using the thermal bonding end caps can achieve more uniform temperature distribution to prevent the thermal fracture and the thermally induced bulge of the pump face. Both end faces of the YVO₄ crystal were dichroic AR coated at 1064 nm and 808 nm. The crystal was wrapped with indium foil and placed in copper heat sinks using water-cooling.

The dual-end-pumped configuration was used to provide high pumping intensity to obtain high gain and high power output. The diode-laser-module of Jenoptik JOLD-45-CPXF-1L by fiber-coupled were applied as the pumping source in the experiment, and if without special description, the diode was with the full pump power of 44 W at 15°C. The fiber was with numerical aperture of 0.22 and diameter of 400μm, and the pump beam from the fiber was imaged onto the end face of the crystal with the adjustable battery of lens. The temperature of the laser diode was controlled by the thermoelectric cooling module(TECM), thus, the center wavelength of the laser diode could be tuned. In our experiment, the temperature of laser diodes were controlled at 15°C, corresponding to the center wavelength of 805 nm, which avoided the the peak value of absorption spectrum of Nd:YVO₄ at 808 nm and yielding to the uniform absorption and weak thermal effect. The dichroic mirrors were AR coated at 808 nm and HR coated at 1064 nm for lasers at the incidence angle of 45°. The NEOS 33041-20-1.5-I crystal quartz Q-switch was used for AO Q-switching. The external trigger signal for the Q-switch drive was provided by a high precision digital signal generator. The power of the radio-frequency driver was 20 W at 40.68 MHz. A planar-planar cavity optics was used in the oscillator, and the output coupler (OC) with a transmissivity of 48% was positioned a distance of l ₁ = 120 mm from the crystal. The HR mirror was positioned a distance of l ₂ = 95 mm from the crystal.

The pumping/crystal units in the amplifier stages were all the same as that of the oscillator. No isolator and coupler were used between the amplifier stages to obtain a compact and simple configuration. The thermal lensing effect of the laser medium was used to realize the mode matching between the pump volume and laser mode. One of the thermal lens in the crystal can image the laser beam into the next crystal with proper size, if the distance between these two amplifiers was optimized. In out experiment, the distance L_i was optimized as follows: L ₁ = 130mm, L ₂ = L ₄ = 170mm, and L ₃ = 230mm. The calculated mode size in the MOPA system with above optimal parameters is shown in Fig. 2., from which we can see that the mode sizes in the crystals are with approximately equal values. The more accurate mode-matching can be realized by finely adjusting the coupling system. It is worthwhile to mention that the size ratio ω_l/ω_p between signal size and pump size exerts a significant influence on the beam quality, where the mode size is defined as the radius of the mode, since the phase aberration always occurs in the wing of the pump region where temperature-gradient-induced thermal stress are higher. Therefore, the waist radius of the pump beam should be larger than the radius of laser mode in the crystal to prevent this aberration which would degrade the beam quality[22]. The waist radius of the pump beam was optimized by adjusting the coupling system and was set ~ 0.4 mm which corresponded to ω_l/ω_p ≈ 0.75, and the waist spot of the pump beam was focused into the crystal about 2.5 mm from the end face.

 

Fig. 2. The mode size in the MOPA system.

Download Full Size | PDF

3. Experimental results and discussion

 

Fig. 3. The spatial form and the pulse overlaping of the oscillator at 500 kHz (a). without the slit and (b). with and the slit.

Download Full Size | PDF

The laser power was measured using the OPHIR FL250A-LP1-DIF power meter. Without the insertion of the slit into the oscillator cavity, we can get 40 W multi-mode in CW operation. However, multi-mode operation can seriously affect the Q-switching operation at high repetition rate. It’s because that at high repetition rate the gain (i.e., population inversion) allocated to every order transverse mode is no longer high enough to maintain the stable Q-switching operation, and the gain competition between transverse modes will result in the uncertain building-up time of pulse, this in turn, the Q-switching operation under multi-mode at high repetition rate is with serous temporal jitter. The spatial form of the multi-mode output is shown in Fig. 3(a). Since the existence of the two side lobes in the vertical direction (c-axis of the crystal), the Q-switching operation at 500 kHz was observed unstably and the pulses were with serious temporal jitter. Then the slit with width of 1 mm was inserted into the oscillator cavity to provide addition loss for the high-order transverse mode in the vertical direction, and the side lobes were suppressed, leading to a near diffraction-limit output with output power of 36 W. The spatial form and the stable pulse overlapping are shown in Fig. 3(b).

 

Fig. 4. The output power and optical-optical efficiency of the MOPA system varying with the pump power in CW operation.

Download Full Size | PDF

The seed laser from the oscillator was amplified by the four-stage power amplifiers. In CW operation, 195 W output power was obtained from the MOPA system at the total pump power of 440 W, corresponding to the total optical-optical efficiency of 44.7% and the absorbed pump power to output power efficiency of 53.3%. Comparing with the values of 42.2% and 50.2% in Ref. 17, the improved efficient was resulted in the optimized design of the amplifiers’ configuration to achieve more optimized mode matching. The conversion efficiency slightly increased as the number of amplifier stages increased, which is due to that higher extraction efficiency can be obtained with higher signal laser intensity. Figure 4 shows the output power and optical-optical efficiency of the MOPA system varying with the pump power in CW operation.

In Q-switching operation, the MOPA system was stably Q-switched from 60 kHz to 850 kHz. The pulse characters of the output laser were detected using a high speed photoelectric detector and an Agilent Infiniium oscilloscope with the bandwidth of 1.5 GHz. The output power and pulse duration varying with the pulse repetition rate IS shown in Fig. 5. 183.5 W output was obtained at 850 kHz with the pulse duration of 72 ns, corresponding to the pulse peak power of 3 kW. The highest 219 kW pulse peak power was achieved at 60 kHz with pulse duration of 12.8 ns. The pulse repetition rate was kept no lower than 60 kHz to prevent the laser damage of the dichroic AR coating at the crystal exit of the 4-th amplifier, where the peak power intensity was as high as ≈90 MW/cm². The instability of pulse energy was described by the standard deviation jitters. At the typical pulse repetition rate of 500 kHz, the standard deviation jitters of the pulse energy was less than ±4% while degraded to less than ±12% at 850 kHz. The phenomenon of pulse missing was observed when the pulse repetition rate was more than 900 kHz, due to the lack of gain for Q-switching operated at very high repetition rate. The pulse series at repetition rate mentioned above are shown in Fig. 6.

 

Fig. 5. The pulse characters vary with the PRF.

Download Full Size | PDF

 

Fig. 6. The oscilloscope traces of the AO Q-switching pulse series at (a). 60 kHz, (b). 500 kHz, (c). 850 kHz and (d). 900 kHz (pulse missing).

Download Full Size | PDF

The beam quality factors were measured with a Spiricon M2–200s laser beam analyzer. The 90/10 Knife-Edge method was used in the measurement. The beam quality factors of 183.5 W output at 850 kHz were measured as M ² _x = 1.28 and M ² _y = 1.21 in the two orthogonal directions respectively (see Fig. 7(a)), i.e., a near diffraction-limit laser was obtained. The Gaussian-like spatial form is shown in Fig. 7(b). We also measured the beam quality factors while the signal laser was sequentially amplified with LD#3+…+LD#i(i = 4,…,10). The results are shown in Fig. 8. From which we can see that the beam quality was varied while amplified, which is contributed to the combined action of the thermal aberration effect and gain guiding effect in the process of amplifying [23].

 

Fig. 7. (a) The beam quality measurement and (b) the spatial form at 850 kHz.

Download Full Size | PDF

 

Fig. 8. The variation of beam quality while the signal laser sequentially amplified with LD#3+…+LD#i(i = 4,…,10).

Download Full Size | PDF

4. Conclusion

We have demonstrated a high-average-power, high-repetition-rate TEM₀₀ mode MOPA laser using acoustic-optic Q-switching. The composite Nd:YVO₄ crystal was used to prevent thermal fracture under high power and high intensity pumping. The seed laser from the AO Q-switched oscillator was scaled up to 183.5 W TEM₀₀ mode output at 850 kHz after amplifying by the four-stage power amplifiers, and the beam quality factors were measured as M ² _x = 1.28 and M ² _y = 1.21 in the orthogonal directions. In CW operation, 195 W with absorbed-output efficiency of 53.3% was also achieved.