Scintillation counters for the measurement of radon have been developed by Damon and Hyde (1), Malvicini (2), and Van Dilla and Taysum (3). Marked improvement in stability was obtained at this laboratory by a modification designed to ensure a reproducible distribution of radon and its daughters within the counter. In addition, lower backgrounds were obtained by the use of steel for the counter shell and quartz for the window. The usefulness was further augmented by adsorbing the radon from gases, liquids, and solids on charcoal. The radon was then transferred to the scintillation counter with a small volume of helium. The scintillation counter has a volume of 96 c.c. and is constructed from an inexpensive Kovar seal. It is coated on the inside with silver activated zinc sulfide to a thickness of 20 mg.∕sq. cm. for optimum pulse height. A clear quartz window having a transparent conductive tin coating on its inner surface is sealed on the open end of the counter. It was found that, in the absence of the conductive tin coating, the negative voltage on the photo-multiplier tube induced a charge on the quartz window which, in turn, concentrated thereon the decay products of radon. This caused a low counting efficiency, which varied considerably in the presence of gas impurities such as water vapor. The calculated efficiency of the counter for the case of radon in the gas phase and all the daughters on the window agreed well with the experimental value of 3.97 cpm∕μμLc radon obtained with non-coated windows and dry helium. The counting efficiency obtained with aliquots of radon from two N.B.S. calibrated radium solutions was 5.42 ± .06 cpm∕μμc radon (0.82 counts per alpha) for a counter with a coated window. This corresponds to the efficiency calculated for the case of radon in the gas and the daughters uniformly distributed on the counter walls and window. The efficiency and reproducibility of ten additional counters were obtained by comparing the counting rates of aliquots of radon from a large reservoir. The counting efficiency of any two counters was found to agree within 2 per cent, and the reproducibility for any one counter was ± 1 per cent. The counting efficiency has remained constant during the six-month period it has been observed. The background counting rate of newly constructed counters is 0.08 cpm. This background is stable over long periods of time except as increased by RaDEF residues from high counting rate samples. Instrumental spurious counts contribute approximately 0.003 cpm. Thus a stable, low-background, high-efficiency radon counting system has been developed. This system is suitable for the analysis of amounts of radon as small as 10−14 c or, when charcoal adsorption is used, concentrations of radon in air as low as 10−16 c radon∕liter.
Read full abstract