Passive Q-switching Nd: YAG Lasers Using Bulk Semiconductors

Abstract

Passive Q-switching has advantages over the other Q-switch schemes to generate short duration pulses with high peak power: light weight, low cost, design simplicity and absence of extra power sources. In addition to dyes, passive Q-switching of Nd:YAG lasers using saturable absorbers has been reported using crystals with color centers, and chromium-doped garnetse2 We report in this letter a passive Q-switching device using a bulk semiconductor epitaxial layer. Semiconductor-based Q-switches have the advantage that the absorption can be tuned so that the length of the device can be of several microns or even thinner. With the existing technology in epitaxial growth, band gaps of compound semiconductors can be engineered to cover almost the entire range of the spectrum from visible to infrared. Solid state lasers can be Q-switched by employing suitable compound semiconductors. For example, Vodopyanov et a1.3 have reported Q-switching an Er, Cr:Y SGG laser using InAs. The Nd:YAG 1.064 pm laser used had a rod p16.3X80 mm in a resonator formed by two flat dielectric mirrors of 100% and 55% reflectivity, separated by 30 cm. The rod was pumped by two Xe flashlamps with typical pump energy between 12 to 30 J in single-shot mode. The Qswitch used was a film of InGaAsP of band gap wavelength 1.09 pm grown by liquid phase epitaxy on an Fe-doped semi-insulating InP substrate. The back side of the substrate was mechanically polished after growth. Fig. 1 shows the measured transmission spectrum at room temperature. The small signal transmittance of the Q-switch was 17%. The saturable absorption measurement was performed with the same Nd:YAG laser but Qswitched by a plastic film coated with dye. The pulse width was 20 nsec. The InGaAsP sample was uncoated. Fig. 2 shows the transmittance as a function of the pump pulse energy at 1.064 pm. The saturated transmittance was 47%. Taken into account the Fresnel reflection loss resulting from both uncoated surfaces, the internal transmission was 90% at maximum. The saturation intensity was about 6x105 W/cm2. The sample was then placed in between the rear mirror and the laser rod, aligned at normal incidence. The Q-switched pulses, shown in Fig. 3, were recorded using a fast InGaAs detector. They had energy of lmJ/pulse and the FWHM of the pulses was 30 nsec. Increasing the pump energy resulted in multiple pulses, with the energy per pulse fluctuation less than 5%. The total energy in the Q-switched pulses was well below the corresponding energy under the free running condition with the same pump energy. The unbleachable loss introduced to the cavity by the Qswitch sample is a primay cause of the low output. No substructure in the pulse was measured, due to the resolution limit of our detection system. Damage to the semiconductor sample was not observed in the saturable absorption measurement, where the highest pump intensity was 8x lo6 W/cm2. However, some damage was observed while experimenting Q-switching effec’i for several days. The damage is suspected to be related to the thermal effects. Q-switching operation did not cease and the energy per pulse did not deteriorate.