Lake Michigan environmental survey : final report / [edited] by John C. Ayers.

Five planchets were prepared for each of the radioisotopes in Table A-15
with the exception of Cs-137; three Cs-137 planchets were prepared. The calibration planchets were prepared as follows:
1. One milliliter of a wetting agent was placed in the planchet to insure
uniform drying. Without the wetting agent the surface tension of any
liquid in the planchet would pull the last few drops to the side of
the planchet which would result in a ring source instead of a circular
source. The wetting agent used was Kodak Photo-Flo 200, diluted 1:100
with water.
2. The standard solutions were transferred by weight from the polyethylene
bottles to the planchets. The transfer was done using glass eye
droppers which had been treated with Siliclad to minimize adsorption
of the radioactive ions on the glass. The planchet was swirled carefully to insure that the radioactivity was dispersed uniformly across
the planchet.
3. The planchet was placed under a heat lamp in a hood, and the liquid
was evaporated to dryness.
4. One milliliter of a sealer was transferred to the planchet and evaporated to dryness. The sealer consisted of 0.5 grams of plexiglass
powder dissolved in 100 milliliters of 1, 2-dichloroethane.
Constant Counting Conditions. It is very important that the calibration
planchets be counted under the same conditions as the sample planchets. Parameters such as temperature can be kept constant by following good laboratory
practice, e.g. keeping the windows and door to the counting room closed at all
times. Contamination of the detector must be avoided. The gain and zero settings of the equipment must be constant while counting calibration planchets.
However, the spectrum analysis computer program can make the necessary corrections if the gain and zero settings of an environmental sample differ from the
calibration gain and zero settings.
In order to determine how large the activity of a calibration planchet
could be before induced gain shifts and live timer errors become significant,
two experiments were conducted. In both experiments the percent dead time of
the spectrometer was used as a measure of the activity of the sample being
counted.
The first experiment determined the effect of sample activity on induced
gain shifts. This was done by building a tower of 5 cm diameter, 1.5 cm tall

A-32