Abstract

The first insights into index-guided diode laser array behavior were provided by using the effective index method to solve for the modes of the individual waveguides and coupled-mode theory to describe how these mix to form the array modes. Although this procedure has been fruitful for the study of conventional arrays, new devices now being fabricated employ large index steps for defining the channels, and thus are not amenable to the effective index approximation. Even worse, the modes of these devices, when calculated numerically, are very complex. This fact, together with the enormous parameter space resulting from the many possible shapes and spacings of the buried ridges, lead to serious difficulties in mode categorization and device design. Sidestepping the effective index approximation, we use coupledmode theory directly by considering the active region and the buried ridges as separate 2-D waveguides. The device modes are then expressed as linear combinations of these waveguide modes. This picture serves to unify our understanding of index-guided arrays in general, and offers a means of understanding leaky-mode array operation in particular by dramatically reducing the parameter space to consist only of the propagation constants of the lowest-order waveguide modes and their coupling coefficient.

© 1989 Optical Society of America