The design of reflective filters based on Al(x)Ga(1-x)N multilayers
In this work we report a method fur designing reflective filters based on Al(x)Ga(1-x)N multilayers. Reflective filters have been designed using a genetic algorithm. The algorithm finds the optimal thickness and composition of each layer for a specified number of layers in order to achieve the desired reflectance characteristics over the specified wavelength range. The calculations take into account the dependence of both the real and imaginary parts of the index of refraction on wavelength and composition. The method is highly versatile, since it also enables the incorporation of constraints such as the limitation of the composition difference of the adjacent layers, so that lattice mismatch can be small. This feature is particularly important due to the large lattice mismatch between GaN and AIN, which causes surface roughness in AIN/GaN structures and consequently lower reflectance than that calculated.