Comparison of two direct LS methods for measuring 222Rn in drinking water using alpha/beta liquid scintillation spectrometry.

Direct liquid scintillation (LS) methods are widely used for surveying (222)Rn in drinking water. Two direct methods are used that differ in sample composition. In a two-phase sample, water lies below a water-immiscible cocktail, while in a homogeneous sample water is mixed with an emulsifying cocktail. Although these methods were developed in the late 1970s, their performances have not been simultaneously tested. Here, the methods were compared in two ways: by preparing both types of sample similarly from (222)Rn-bearing groundwater in one emulsifying and in three organic cocktails, and by calibrating the methods with a (226)Ra standard according to their respective procedures. The samples were measured using alpha/beta LS spectrometry. The standard deviations of parallel samples and the repeatability of the measurements were excellent for both methods, except two-phase (226)Ra samples, whose efficiencies decreased slightly over time. This instability was due to interference from (210)Pb, (210)Bi and (210)Po, which accumulated in the (226)Ra standard, and possibly also to the migration of (214)Pb and (214)Bi into the aqueous phase and deficient transfer of (222)Rn from the water to the cocktail.