Dislocation Density Reduction in Cadmium Telluride and Mercury Cadmium Telluride Grown on Silicon Using Thermal Cycle Annealing

Abstract

Mercury Cadmium Telluride (HgCdTe) is a material of great importance for infrared focal plane array applications. In order to produce large format detector arrays this material needs to be grown on a large area substrate, with silicon being the most mature substrate, it is the optimal choice for large format arrays. To help mitigate the effect of the lattice mismatch between the two materials, cadmium telluride (CdTe) is used as a buffer layer. The CdTe itself has nearly the same lattice mismatch (19.3%) to silicon, but due to the technological advantages it offers and compatibility with HgCdTe, it is the best buffer layer choice. The lattice mismatch between HgCdTe/CdTe and the silicon substrate leads to the formation of dislocations at densities in the mid 106 to low 107 cm−2 range in the epilayers. Such a high dislocation density greatly effects detector device performance quantities such as operability and sensitivity. Hence, the dislocation density should be brought down by at least an order of magnitude by adopting novel in situ and ex situ material processing techniques. In this work, in situ and ex situ thermal cycle annealing (TCA) methods have been used to decrease dislocation density in CdTe and HgCdTe. During the molecular beam epitaxial (MBE) growth of the CdTe buffer layer, the growth was interrupted and the layer was subjected to an annealing cycle within the growth chamber under tellurium overpressure. During the annealing cycle the temperature is raised to beyond the growth temperature (290 → 550 ◦C) and then allowed to cool before resuming growth again. This process was repeated several times during the growth. After growth, a portion of the material was subjected to a dislocation decoration etch in order to count the etch pit density (EPD) which has a direct correspondence with the dislocation density in the crystal. The crystalline quality was also characterized by x-ray diffraction rocking curves and photoluminescence. The in situ TCA resulted in almost a two order of magnitude reduction in the dislocation density, and factor of two reduction in the full width at half maximum of the x-ray rocking curves. Photoluminescence also suggested a decrease in the number of dislocations present in the material. This decrease is attributed to the movement of the dislocations during the annealing cycles and their subsequent interaction and annihilation. To decrease the dislocation density in HgCdTe layers grown on CdTe/Si composite substrates, ex situ TCA has been performed in a sealed quartz ampoule under a mercury overpressure in a conventional clam-shell furnace. The reduction in the dislocation density has been studied as a function of growth/annealing parameters such as the initial (as grown) dislocation density, buffer layer quality, Hg overpressure, annealing temperature, annealing duration, and the number of annealing cycles. It was found that the primary parameters that affect dislocation density reduction are the annealing temperature and the number of annealing cycles. Some secondary affects were observed by varying the duration spent at the maximum annealing temperature. Parameters such as the initial dislocation density and buffer layer quality did not play a significant role in dislocation reduction. Though no correlation between Hg overpressure and dislocation density was found, it did play a vital role in maintaining the quality of the surface. By using the ex situ TCA, a dislocation density of 1 × 106 cm−2 could be reliably and consistently achieved in HgCdTe layers that had a starting density ranging from 0.5 − 3 × 107 cm−2. Examination of the annealing parameters revealed an exponential decay in the dislocation density as a function of increasing number of annealing cycles. In addition, a similar exponential decay was observed between the dislocation density and the annealing temperature. The decrease in the dislocation density is once again attributed to moving dislocations that interact and annihilate. This behavior was modeled using a second order reaction equation. It was found that the results of the model closely agreed with the experimental values for a wide range of annealing temperatures and number of annealing cycles.