Control of Defects and Impurities in Production of CdZnTe Crystals by the Bridgman Method
- PDF / 360,539 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 66 Downloads / 260 Views
The CdZnTe boules were grown by a modified vertical Bridgman [1,2] process with the material in a sealed, evacuated (fused) quartz ampoule. In most cases the melt and crystal were in a pyrolytic boron nitride (pBN) crucible which fit inside the ampoule. Some runs were performed without a crucible but with a pyrolytic carbon coating on the ampoule. Starting materials were Johnson Matthey XTAL Grade Cd, Zn and Te. GDMS (glow discharge mass spectroscopy) analysis typically shows almost no detectable impurities with the exception of carbon, nitrogen and oxygen, which are not reliably measured by this technique. Metallic impurities of particular interest, Cu, Ni, Na and Li are undetected. For these elements, detection limits are no higher than 2 or 3 ppba (parts per billion, atomic); except the detection limits for Cu in Zn may be greater than 50 ppba since Cu and Zn are adjacent elements in the periodic table. Cd and Zn were cut and chemically etched to remove surface contamination and to trim the weight. Te was broken into large pieces and then cast into a slug under flowing hydrogen. In most of the results reported here, the boules were 1.3 kg with a diameter of 55-57 mm. These are the smallest of the production boules, most of which are 3.5 kg with diameter 75-80 mm. Two different methods of charge preparation were used. In the in situ process, the Te slug, in the pBN crucible when used, was placed in the lower end of the ampoule. The Cd was placed in a quartz reservoir near the top of the ampoule. The Zn was loaded with either the Te or the 335
Mat. Res. Soc. Symp. Proc. Vol. 484 ©1998 Materials Research Society
Cd. During the initial part of the growth process, reaction of the elements took place as the Cd (and Zn if also in the reservoir) vapor was transported to the molten Te. The temperature was raised to keep the CdZnTe-Te solution molten as the reaction proceeded. After a soak above the CdZnTe melting point, the temperatures were equilibrated at their growth values and crystal growth was carried out at a translation rate of 1.1 mm/h. The second method of charge preparation, ex situ, was similar except the vapor transport reaction was performed in a separate, carbon coated ampoule using a horizontal furnace configuration. An appropriate weight of the resulting polycrystalline CdZnTe material was then loaded into a growth ampoule. Both compounding and growth were performed in the presence of excess Cd to provide a vapor'pressure (roughly one atmosphere) sufficient to avoid excessive deviation from stoichiometry. The correct amount of excess Cd had been determined empirically through measurements of infrared transmission and through observation of second phase particles
(precipitates) by infrared microscopy. These measurements were made on polished slices cut from the boules. When the Cd vapor pressure is correct, infrared transmission is at its theoretical value, >65%, across the 2.5-16 pm measured spectral range and precipitates, if detectable, are less than 10 pm in diameter and have a density no greater
Data Loading...
