Evaluación final de requerimiento total del caudal de aire para la Zona Gisela

Waveguide channels were fabricated in z-cut Nd:MgO:LiNbOg doped with 0.1% Nd and 5% mol MgO. The waveguide fabrication was described in section 6.3.3.1. The resonator was fornied by coating 99 and 96% reflecting miirors on the substrate end­ faces. With the dye laser set at 592 nm and with the pump light TM polarised, the laser characteristic was obtained for the z-cut sample. In this case the electric field is still parallel to the c axis and so the spectroscopic properties should be identical to those for proton exchange x-cut waveguides. This was confirmed by measuring the absorption as a function of wavelength and comparing it to the x-cut sample. A correction was made to the results, since the waveguide loss was -40%, and due to a larger waveguide mode size, the coupling efficiency improved to 60%. The following graph plots the channel waveguide laser output as a function of the absorbed pump power

120 100 80 a . 6 0 4 0 20 20 3 0 4 0 5 0

Absorbed pump power (mW)

Figure 6.33. Graph of the Z-cut laser characteristic for a nominal 96% output reflectance (592 nm pump).

The pump power threshold was 28±2 mW, and a maximum output power of 0.11 mW was obtained. The high threshold is presumably caused by high waveguide losses, due to increased scattering, and poor miiTor quality (confirmed in section 6.3.4). The waveguides were wider than those for the x-cut sample (5.5 fim compared to 4.5 pm) and so alignment losses would also be expected to be higher, (according to the analysis in Chapter 3.7).

6.4.2 Diode laser set-up

The Nd:MgO:LiNbOg channel waveguide laser was also pumped with a collimated gain-guided, single stripe laser diode, emitting 50 mW at 811.5 nm figure 6.34. As discussed in Chapter 5.4 a disadvantage of the gain-guided diode laser is that the beam is highly diverging and elliptical, as well as being astigmatic. To circumvent this problem, which prohibits efficient coupling to the waveguide, a system of beam shaping optics was used to modify the beam. A cylindrical lens compensated for the 40 pm of astigmatism inherent in the device, and an anamoi*phic prism pair was adjusted to provide an expansion of two, in order to shift the beam into a nearly circular profile. An 8 mm focal length objective lens focussed the beam onto the waveguide, and a 15 mm focal length lens collected the waveguide output. No attempt was made to isolate the sample from the pump, and this might have led to some

feedback to the diode laser. The waveguide laser output was passed through two RG 1000 filters, in order to suppress the pump light, and through a pinhole placed before the entrance slit of the monochromator.

N d:M gO :L iN bO g sam p le C ylindrical lens C ollim atin g lens G ain-guided laser diode RG 1000 b lock filter To m o n o clu o m a to r A n am oip hic prism pair F ocu sing lens C ollim ator

Figure 6.34. Experimental set-up for diode laser pumping of a channel waveguide.

The diode laser had an emission wavelength of 811.5 nm at 20 ®C. In order to achieve better absorption in the waveguide, the laser diode case temperature was increased to 25 ^C. This shifted the laser diode wavelength to 813 nm, corresponding to -75% absorption in the waveguide.

6.4.2.1 X-cut laser characteristic (96% output reflectance)

The minor characteristic was the same as that for the initial laser experiment, namely input and output reflectances of -98% and 96% respectively. The diode laser beam was launched into the 4.5 pm-wide waveguide. The waveguide laser began to oscillate when the pump power input reached 42±2 mW. If the absorption of the pump light, over the length of the waveguide, is 75%, the transfer efficiency is 16%, and the waveguide propagation loss (at 1.084 pm) is 1.3 dB/cm (32% loss), the pump power absorbed at threshold is given by

Pj, =42 mWxO.75 x 0.16 x 0.68 = 3.4±0.6 mW

This value agrees reasonably closely with that obtained from theory (Pth

mW). The uncertainties in the theoretical value include the waveguide loss, and the value of a, for which the largest published value was assumed

With the diode laser source we could only pump at or just above threshold, so a laser characteristic could not be obtained. However emission specüa showing the laser just below threshold, and above threshold are displayed in figures 6.35 and 6.36 respectively

3 5

I

i

Figure 6.35.Diode pumped emission spectrum just below threshold.

The spectrum exhibits several peaks below threshold, however the peak at 1084 nm is beginning to dominate the spectrum, indicating that we are approaching the threshold for laser oscillation.

bandpass =1.2 nm

1075 1080 1085 1090 1095

Wavelength (nm)

Above the threshold for laser oscillation the spectrum narrows, and there is a fifty-fold increase in the waveguide laser output. The waveguide laser output power was estimated by comparing the output on the chart recorder with that from a similar set­ up, and using the dye laser as the pump. The same sample was used in both cases, as well as identical collecting optics, filters, and entrance slit width. In this way the maximum waveguide laser power output (displayed in figure 6.36) was estimated to be 0.15 ±.05 mW.