Page 260 - Chalcogenide Glasses for Infrared Optics

AMI Infrar ed Crystalline Materials 235

but they are randomly oriented. We experienced that same problem
in producing plates of CdTe. Since gallium is not volatile, only the
arsenic pressure must be in equilibrium with the stoichiometric
crystal at the melting point. For CdTe, both Cd vapor and Te vapor
must be in equilibrium with the stoichiometric crystal at the melting
point. Extremely small crystallites of As in GaAs may not appreciably
affect ir transmission measured with a spectrophotometer. However,
their presence may be verified by measuring transmission with a CO
2
laser. The inclusions heat up locally, causing an immediate drop in
infrared transmission. The bottom drawing in Fig. 9.16 illustrates
exactly how the real machine looks. The quartz chambers are placed
inside an alumina tube for support. The alumina tube rests inside a
very long Mullite tube supported at both ends. The furnace is much
larger in diameter. Operator viewing windows are cut in both the
furnace wall and the alumina tube. The operator watches the com-
pounding process, and when it is complete, he or she raises the
crystal end of the Mullite tube so the melt contacts the seed. Growth
begins then as the heater is moved slowly, so the growth moves down
the boat away from the seed. Crystal length is limited by the total
travel for the unit of 18 in.
Material grown in the AMI machine was fabricated into small
wafers for use as substrates in the production of second-generation
night vision goggles produced by the local company NYTEK in
Garland. AMI had a computer-controlled instrument designed and
built by Steve Boldish, formerly of NYTEK, to make Hall measure-
ments to verify the resistivity of single-crystal, polished substrate
slices. As part of the effort, GaAs slices undoped, lightly doped,
doped moderately, and heavily doped were measured to establish
resistivity as a function of doping level for AMI GaAs. Some of the
more than 50 samples used to form this plot were from outside
vendors, which allowed direct comparison of AMI results with those
of other suppliers. The agreement was good. The results plotted in
Fig. 9.17 for N-type Si doped GaAs cover 12 orders of magnitude in
carrier concentration and 7 orders of magnitude for resistivity. Note
the plot is almost a straight line until it reaches a carrier concentration
16
of 10 where the electron mobility begins to decrease. This plot was
used to guide AMI in producing Si doped GaAs of specific resistivity,
which was very important later in trying to produce windows to be
used with infrared systems in aircraft while also providing EMI
protection.
After the first Desert War, attention at AMI was turned again
toward using GaAs as an infrared optical material. First, the ZnS/
ZnSe avionics window did not survive well in the desert environ-
ment. Second, the infrared system in the U.S. tanks lost sensitivity in
the high temperatures of the desert due to thermal-generated carriers
in the germanium optics. The systems used in the tanks were complex
with many lens elements. The total optical path for germanium was