Date of Completion

5-15-2017

Embargo Period

11-11-2017

Keywords

X-ray dynamical diffraction, x-ray analysis, dislocation density, threading dislocations, dislocations

Major Advisor

Prof. John E. Ayers

Associate Advisor

Prof. Faquir Jain

Associate Advisor

Prof. Helena Silva

Field of Study

Electrical Engineering

Degree

Doctor of Philosophy

Open Access

Campus Access

Abstract

High-resolution x-ray diffraction characterization of pseudomorphic semiconductor device structures is well established, and enables the depth profiling of strain and composition. Recently, metamorphic growth has become increasingly important for the realization of high electron mobility transistors (HEMTs), light-emitting diodes (LEDs), laser diodes, solar cells, and other devices because of the freedom it affords in choosing the compositions and thicknesses of epitaxial layers. However, the traditional methods of x-ray diffraction (XRD) analysis involving dynamical simulations and curve-fitting to experimental diffraction profiles are based on the assumption of no dislocations and do not apply to metamorphic structures. The goal of the present work is therefore to extend the high-resolution x-ray analysis to metamorphic structures containing dislocations, using the mosaic crystal model recently developed by our research group. Toward this end, we have analyzed the high-resolution diffraction profiles for a number of metamorphic heterostructures, including step-graded, linearly-graded and S-graded buffers, superlattices, superlattice HEMTs, single quantum wells, and tandem solar cells. We have shown that, in many cases, the widths of primary Bragg peaks or Pendellösung fringes may be used to estimate the threading dislocation densities in these structures. We also show that the diffraction profiles are sensitive to the depth profile of the threading dislocation density, so that in principle it should be possible to extract the depth profile of threading dislocation density as well as the composition and strain by a curve-fitting procedure. Though such an approach is expected to be computational complex, our work establishes the foundation for its eventual realization.

COinS