Experimental and theoretical studies of flame hydrolysis deposition process for making glasses for optical planar devices

Abstract

We have here presented experimental and theoretical studies of a flame hydrolysis deposition (FHD) process for making glasses for optical planar devices. FHD involves deposition of soot particles generated in the flame on a planar surface to form a porous layer. In order to function as a waveguide core, the porous soot deposit must be sintered at high temperatures to form a dense glass. However, these temperatures are high enough to cause the dopants to volatilize at the surfaces of the deposit. As a result, dopant concentration gradients and compositional inhomogeneities are created in the densified glass layer which result in inferior optical properties. If the layers could be deposited pre-sintered during laydown, these problems could be minimized or avoided. To understand the FHD process and its effect on the morphology of the resulting deposit, a number of models have been developed. We present models to predict the structure of the soot-laden flame along with the methods to estimate soot number density and mean size at different locations in the flame. A simple method is presented to predict morphology of the suspended soot and soot deposit formed during deposition. Motivated by the predictions of the model, process changes were made which resulted in the morphology of the core layers to change from sooty to pre-sintered during deposition. These process changes resulted in core glasses with significantly improved core roughness, index uniformity and thickness uniformity which helped reduce the straight waveguide losses from 0.2–0.3 to 0.02–0.05 dB/cm. Details of underclad and overclad glasses which resulted in low warpage and polarization sensitivity of 0.1 nm for phasar or phasar-arrayed waveguide devices are also presented.