1. Introduction

Rare-earth-doped vanadate crystals are attractive gain mediums for producing small lasers. In particular, Nd-doped yttrium vanadate crystals show great potential as a laser medium. The thermal conductivity of Nd:YVO₄ is approximately the same as that of YAG, while its radiative lifetime is just under half that of YAG. The emission cross section of Nd:YVO₄ is two times larger that of YAG [1–3]. Consequently, Nd:YVO₄ can provide high gain at high repetition rates. Nd:YVO₄ is more effective than YAG for achieving pulse oscillation at high repetition rates of up to 1 MHz.

Operation of an active acousto-optic (AO) Q-switched transversely diode-pumped Nd-doped mixed gadolinium yttrium vanadate slab laser has been demonstrated. In addition, a high repetition rate of up to 700 kHz and a pulse width of less than 40 ns have also been achieved [4]. AO-Q-switching operation was obtained at repetition rates of up to 1 MHz in a grazing-incidence Nd:YVO₄ laser with a pulse-to-pulse instability of ±15% [5].

A coupled-cavity active electro-optic (EO)-Q-switched Nd:YVO₄ laser achieved a repetition rate of 2.25 MHz, a pulse width of 8.8 ns, and an average output power of 360 mW [6]. The coupled-cavity in this system consists of a laser medium and an EO-Q-switch crystal and does not have any polarizers. A short cavity length is effective for achieving high repetition rate pulse oscillation. However, since the laser cavity is formed by the facets of the laser medium coated with high-reflectance coatings and an EO-Q-switch crystal, this coupled-cavity requires parallel flat surfaces, which are difficult to fabricate.

We previously demonstrated a fiber-based oscillator-amplifier architecture that produces short laser pulses with a duration of a few nanoseconds at 100 kHz and an average power of 10 W for second harmonic generation [7, 8]. The oscillator is a compact EO-Q-switched Nd:YVO₄ laser that consists of a Nd:YVO₄ crystal, an output mirror and an EO deflector without polarizers.

In the present study, pulse oscillation at high repetition rates of up to 1.4 MHz is demonstrated by an oscillator with the same configuration.

2. Experiment

Three Nd:YVO₄ crystals were used for CW tests, which were conducted using diode-end-pumping prior to Q-switch operation. Details of three crystals are shown in table 1. The EO deflector was removed from the oscillator during the CW experiments. The pump wavelength was 808 nm and the cavity length was 40 mm. The temperature of the crystal holder was set to 21°C. The laser cavity was formed by a crystal facet with a high-reflectance coating and an output mirror. A concave mirror with a radius of curvature (ROC) of 0.2 m and a reflectance of 90% was used as the output mirror. The pump beam is delivered through a silica fiber with a core diameter of 200 µm. The image of the pump beam on the facet of the fiber is transferred to the Nd:YVO₄ crystal as a one-to-one image.

Table 1. Details of Three Crystals

View Table | View all tables in this article

Quadrupole EO deflectors [9, 10] were fabricated from 3 mm×3 mm×10 mm crystals. The EO deflector had a low optical loss and produced linear beam deflections. The EO deflector was used for Q-switching, and produced pulses with short widths and high peak powers at high repetition rates of over 1 MHz.

Figure 1 shows a schematic diagram of the pulse oscillation mode. Q-switch operation is performed with an EO deflector, which is operated by an external trigger and a driving voltage of 330 V. The period of constant high voltage was 100 ns or greater, but a duty cycle of less than 35% was used to prevent the EO deflector from the piezoelectric effect. The maximum driving rate of the EO deflector was 1.4 MHz, which is limited by the EO driver. A flat output mirror with a reflectance of 90% was used for pulse oscillation tests.

 

Fig. 1. Schematic diagram of the pulse oscillation mode.

Download Full Size | PDF

The laser beam is deflected when a high voltage is supplied to the EO deflector. If the output mirror is orientated so that it is normal to the axis of the deflected beam, as shown in Fig. 1(a), pulse oscillation is achieved only during the sustained period of high voltage. A low repetition rate pulse oscillation of less than 750 kHz is obtained in this “deflected beam oscillation” mode.

Conversely, if the output mirror is orientated normal to the axis of the undeflected beam, as shown in Fig. 1(b), pulse oscillation is achieved outside the sustained period of high voltage. This “undeflected beam oscillation” mode enables a long pulse build-up time and a low duty cycle of high voltage at a repetition rate higher than 750 kHz.

3. Results and discussion

Figure 2 shows the CW test results of the three crystals. The CW tests were conducted with an output mirror with a reflectance of 90% and a ROC of 0.2 m without the EO deflector. As Fig. 2 shows, crystal 3 had the best CW performance of the three crystals.

An output power of 3.3 W was obtained using crystal 1. The slope efficiency is 42% and the light-to-light conversion efficiency is 40%. The threshold pump power for lasing was 166 mW.

The slope efficiency of crystal 2 is 48%, and the light-to-light conversion efficiency is 45%, which is higher than that of crystal 1. The threshold pump power for lasing was 73 mW. A high Nd-doping of the vanadate crystal has a positive effect on the laser oscillation efficiency. However, the output power saturates at only 2.0 W. Because of the high Nd-doping of the vanadate crystal, serious thermal effects also simultaneously occur.

The slope efficiency of crystal 3 is 57%, and its light-to-light conversion efficiency is 55%. A maximum output power of 4.3 W was obtained. The threshold pump power for lasing was 352 mW. The total amount of Nd is equal to that of crystal 2. However, because the thickness of crystal 3 is twice that of crystal 2, the concentration of Nd is equal to that in crystal 1. Crystal 3 completely absorbs all the pump light energy, while not all of it was able to be absorbed by crystal 1, and deleterious thermal effects were not observed in crystal 3, unlike crystal 2.

 

Fig. 2. CW oscillation performance of the three Nd:YVO₄ crystals.

Download Full Size | PDF

Figure 3 shows the pulsed waveforms obtained using crystal 3 at repetition rates from 100 kHz to 1.4 MHz. For pulse operation tests, a flat mirror with a reflectance of 90% is used as the output mirror to avoid CW operation. To avoid optical damage of EO deflector, the pump power was restricted within 4.5 W in the initial stages. At the pump power of 4.5 W, the maximum achievable repetition rate of pulse oscillation was 1 MHz. The repetition rate of 1.4 MHz was achieved at the pump power of 6.5 W. Table 2 shows the average power of above conditions. The CW powers for the cases of both empty and deflector insertion were also indicated in the table. The degradation of average power was negligible except lower repetition rates. It is thought that the insertion loss of the EO deflector is vanishingly small.

 

Fig. 3. Results of the pulse waveform of crystal 3 at repetition rates from 100 kHz to 1.4 MHz.

Download Full Size | PDF

Figure 3(a) shows the pulse waveforms of “deflected beam oscillation” for crystal 3 at repetition rates from 100 kHz to 750 kHz. The pulse delay time (i.e., the pulse build-up time) and the pulse width increase with an increase in the repetition rate. Results of pulse oscillation tests are shown in table 2. The pulse width and the pulse delay time increase due to the reduction in the gain at higher repetition rates.

Figure 3(b) shows the results of the pulse oscillation tests with “undeflected beam oscillation” for crystal 3 at a high repetition rate of 1.0 MHz. Stable pulse operation is realized with a high repetition rate of 1.0 MHz. The pulse delay is 366 ns, and the pulse width is 44.6 ns.

The sustained period of high voltage of the EO deflector at a repetition rate of 100 kHz is 100 ns, which is equivalent to a duty cycle of 1%. At a high repetition rate of 750 kHz, the sustained period of high voltage for the EO deflector is 300 ns, which is equivalent to a duty cycle of over 20%. In order to reduce the piezoelectric effect, it is desirable to use a low duty cycle for the EO deflector. “Undeflected beam oscillation” is suitable for obtaining long pulse build-up times, because the pulse build-up occurs at zero-voltage. At a repetition rate of 1 MHz, a sustained period of high voltage of 100 ns was obtained, which is equivalent to a duty cycle of the EO deflector of only 10%.

Figure 3(c) shows the results of the pulse operation experiments with “undeflected beam oscillation” of crystal 3 at the highest repetition rate of 1.4 MHz. Pulse operation at a repetition rate of 1.4 MHz was obtained at a pump power of 6.5 W with a pulse delay of 357 ns, and a pulse width of 39.0 ns. This laser pulse cannot be effectively addressed by changing the pulse build-up time for repetition rates over 1.0 MHz. Consequently, it was decided to increase the pump power to achieve a higher gain.

Table 2. Results of Pulse Oscillation Tests for Crystal 3

View Table | View all tables in this article

Figure 4 shows the pulse-to-pulse instability of the crystal 3. Crystal 3 exhibits stable pulse oscillation at 500 kHz or less. The pulse-to-pulse instability is only 4%. The pulses become unstable with an increase in the repetition rate. The pulse-to-pulse instability for a repetition rate of 750 kHz is 9%. For pulse oscillation with “undeflected beam oscillation” of 1 MHz, the pulse-to-pulse instability was 11%. For a repetition rate of 1.4 MHz, a higher gain is achieved by increasing the pump power. The pulse-to-pulse instability was 20%.

In general, the pulse-to-pulse instability increases at higher repetition rates [5]. Table 3 shows the population inversion build-up time and pulse build-up time. The population inversion build-up time decreases at higher repetition rate. Therefore, pulse oscillation at higher repetition rate is affected by short term perturbation such as pump power fluctuation.

 

Fig. 4. Pulse-to-pulse instability of the crystal 3.

Download Full Size | PDF

Table 3. Population Inversion Build-up Time and Pulse Build-up Time

View Table | View all tables in this article

4. Conclusion

To achieve pulse oscillation at a high repetition rate over 1MHz, a compact EO-Q-switched Nd:YVO₄ laser cavity packed into a length of 40 mm was applied. The short cavity length and large net gain allow us to shorten the pulse build-up time. Then, Pulse operation at a high repetition rate of 1.4 MHz and a pulse width of 39 ns was successfully demonstrated using an EO deflector. An average output power of 2.7 W was obtained at a pump power of 6.5 W, and the light-to-light conversion efficiency was 42%.