Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Electrical and Computer Engineering

First Advisor

Pratul K. Ajmera


The use of infrared piezobirefringence for characterization of defects in cubic semiconductor materials is investigated in this work. Under stress, these normally isotropic materials become birefringent. The stress can result from an applied external load or from defects present in the material. Defects can be generated in the material during its growth and/or processing. As a first step towards defect characterization, the case of diametrically loaded discs of semiconductor materials was simulated. This was done to obtain a better understanding of the simulation algorithm prior to its subsequent use in dislocation characterization. A dark-field plane polariscope was constructed using a He-Ne laser tuned to 1.15 $\mu$m wavelength as the light source. The computer simulated images matched well with the experimentally observed ones on diametrically loaded discs of silicon and gallium arsenide. The behavior of the stress-optic coefficient C was also investigated, which has been treated as a constant by other investigators in earlier works. In this work, it was found that for (100) oriented Si and GaAs discs under diametrical compression, the stress-optic coefficient C is a strong function of position for a given load, and also changes with the direction of the applied load with respect to the principal crystal axes. However, no such dependence was found for (111) oriented Si and GaAs discs under diametrical compression, as expected from the crystal symmetry. The values of C for (100) oriented Si disc under diametrical compression ranged from 2.0 $\times$ 10$\sp{-12}$ cm$\sp2$/dyne to 3.0 $\times$ 10$\sp{-12}$ cm$\sp2$/dyne and for (100) oriented GaAs disc under diametrical compression ranged from 0.8 $\times$ 10$\sp{-12}$ cm$\sp2$/dyne to 2.6 $\times$ 10$\sp{-12}$ cm$\sp2$/dyne for the cases investigated. The corresponding figures for (111) oriented Si and GaAs discs under diametrical compression are 2.33 $\times$ 10$\sp{-12}$ cm$\sp2$/dyne and 1.94 $\times$ 10$\sp{-12}$ cm$\sp2$/dyne respectively. Next, an accurate algorithm was developed for the simulation of the images of dislocations as a function of the sample orientation, the orientation of the dislocation line with respect to the principal crystal axes, the orientation of the Burger vector with respect to the dislocation line, and the polarization angle of the incident light. An image data bank has been created for different dislocations. This data bank can be utilized to obtain an accurate characterization of the image of a dislocation by comparing the experimentally obtained images with the simulated ones. In this work, images of the dislocations were observed experimentally and were matched with the simulated images. This approach provides a fast, non-destructive alternative technique for defect characterization in semiconductor materials.