Effect of laser cavity parameters on saturation of light – current characteristics of high-power pulsed lasers

Theory of direct and indirect effect of two-photon absorption on nonlinear optical losses in high power semiconductor lasers

“…We have verified that for the values of the laser cavity geometry parameters L and w and antireflection and highreflection coated facet reflectance coefficients R AR , R HR (often used in practice, see e.g. [5,24]) used in the calculations, the lumped model provides good accuracy, not just qualitatively, but quantitatively as well. As an illustration, figure 5 includes, in addition to the lines, also results calculated for the three representative active layer positions (l a =1.17 μm near the OCL centre and l a =1.8 μm near the p-cladding as in figure 1(a), and in addition l a =1.45 μm near the optimum) using a distributed model similar to those of [22,23,29]) with power P, and hence ( ) a TPA FC and ( ) a , TPA mod calculated locally, as functions of z.…”

“…It has to be noted that figures 6-7 were calculated assuming that the laser in all cases operates in a single, fundamental transverse mode at all currents. For the case of l a =1.17 μm, this can indeed be confidently expected, because this position is optimised for best modal discrimination [24]. For the other two positions, reliable single-mode operation cannot be guaranteed and operation in the second order mode, at least for lower currents, as discussed in [31], cannot be ruled out.…”

“…and (14) In this equation, the constant built-in absorption (built in) in   included the effects of cladding doping, structure imperfections, and free carrier absorption by the thermally escaping OCL carriers (the N bT term in (10)); the value for it was taken as 0.4 cm -1 [24]. We treat the case of pulsed (quasi-CW) operation so the thermal roll-over is not considered in the simulations, and the value of N bT can be approximately taken as constant.…”

“…The first is a broadened, symmetric waveguide structure, intended for operation at λ=1.06 μm. The prototype version of the structure, shown in figure 1(a) (solid line) alongside the corresponding modal power distribution, is a broad symmetric waveguide with h=2 μm and the AL positioned near the centre (slightly shifted towards the p-cladding to aid higher order mode discrimination [5,24]); however in the simulations, both the OCL thickness and the AL position (the extreme case of the AL located near the OCL/p-cladding interface being shown in a dashed line as case 2 in figure 1(a)) could be varied.…”

“…Within this procedure, the power P as function of current i (or current density j=i/wL, recalculated using the cavity length L=3 mm, width w=100 μm) is determined from a transcendental equation value i th did not affect the results much, given the range of currents i?i th . The internal loss was calculated as a sum of all contributions included the effects of cladding doping, structure imperfections, and free carrier absorption by the thermally escaping OCL carriers (the N bT term in (10)); the value for it was taken as 0.4 cm −1 [24]. We treat the case of pulsed (quasi-CW) operation so the thermal roll-over is not considered in the simulations, and the value of N bT can be approximately taken as constant.…”