100
1000 Angstrom
correction
at.%
78.4
9.4
12.2
1500 Angstrom
correction
wt.%
78.7
15.4
5.9
1500 Angstrom
correction
at.%
78.0
9.4
12.7
1800 Angstrom
correction
wt.%
78.6
15.4
6.0
1800 Angstrom
correction
at.%
77.7
9.3
13.0
The value corrected for a thickness of 1800 Angstroms
is quite close to the actual composition of the alloy.
The actual thickness of the sample is unknown. The data
show, however, a 20% difference in total aluminum as a
result of absorption. In all subsequent spectra, a
specimen thickness of 1000 Angstroms was presumed.
In this research, every effort was made to reduce
specimen related artifacts by orienting the specimen such
that the absorption path length was minimized between the
flourescing area of the sample and the detector, and by
orienting the crystal so that only the desired phase was
analyzed. Because the detector was a horizontal detector,
this was a significant limitation. It was necessary to
orient the crystal as close as possible to a B =C.100^
direction in order to minimize gamma/gamma prime phase
overlap.
A further complication was contamination. The
effects on spatial resolution are well documented (Hren,
1979). In order to generate sufficient signal for
adequate statistics, it was necessary to count for very
long times, usually in excess of 100 seconds. No data