Answer to Question #9057 Submitted to "Ask the Experts"
The following question was answered by an expert in the appropriate field:
The relative amounts of potassium and thorium in beach sand can vary quite a lot. Typical potassium levels are 1 percent to 2 percent by weight (10,000–20,000 ppm), but values less than 1 percent and greater than 3 percent have frequently been observed. Thorium levels may also vary significantly. Values from 5 to 15 ppm by weight are common for ordinary beach sand. The well-known monazite sands contain much greater amounts of thorium, sometimes up to 10 percent by weight.
Regarding the interference that you cite, we should note that for a sand that contains a nominal 2 percent potassium, the expected potassium-40 activity concentration would be about 600 Bq kg-1 and, given the 10.67 percent yield of the 1,460.8 keV photon, the gamma-emission rate at this energy would be about 65 per second per kg of sand. If the same sand contained 10 ppm thorium-232, and the actinium-228 were in secular equilibrium with the thorium, the expected activity concentration of the actinium-228 would be about 41 Bq kg-1 and, given the 1.0 percent yield of the 1,459+ keV photon from actinium-228, the gamma-emission rate at this energy would be about 0.4 per second per kg of sand. Based on these numbers, it does not appear that the thorium influence on the potassium-40 peak would be significant, adding only about 0.6 percent to the expected counts in the peak region. This could change if you were looking at high thorium sands, such as the monazites where thorium activities could be thousands of times greater.
In cases where higher thorium content is suspected, you could do a fairly simple analysis to correct counts in the potassium region. This might require obtaining a sample of sufficient thorium content to allow good counting statistics (preferably a sample that reasonably simulates the physical characteristics of the sand you are measuring and with actinium-228 in equilibrium). The prepared sample should be sufficiently free of potassium so that the potassium-40 photopeak counts do not contribute appreciably to counts in the 1,459 keV thorium photopeak. When the sample is counted, the counts in photopeaks at one or more energies in addition to the 1,459 peak may be obtained. The gamma emissions of actinium-228 at 911 keV (27.7 percent yield) and 969 keV (16.6 percent yield) are good possibilities. You would determine the ratio of counts in the 1,459 keV peak to the counts in one (or each) of these other peaks. This ratio could then be applied when analyzing samples to correct for the thorium (actinium-228 1,459 keV) counts in the potassium-40 peak; that is, you could multiply counts in the sample 969 keV peak by the previously determined ratio of 1,459 keV photopeak counts to 969 keV photopeak counts to obtain counts to be subtracted from the sample 1,459 keV region to remove the thorium interference from the potassium region counts.
George Chabot, PhD, CHP