RADIOACTIVITY METHODS 



1013 



more energetic a-rays, the self-absorption of as in the sample is usually great enough 

 to make their contribution also negligible. 



For uranium, thorium, and potassium work, the choice of a proper counter "is 

 usually determined by the available amount of sample. One- or two-gram samples are 

 best examined by mica-window counters, but larger samples can be tested with alum- 

 inum or glass thin-walled counters, which are generally much cheaper than the mica- 

 window type. Normally only 2 - 3 g. of sample are required for a /3 assay. To improve 

 the geometrical relation of counter and sample for work with low-grade materials, 

 several thin-wall counters can be connected in parallel and placed over the sample 

 tray, or the sample may be placed to coaxially surround the counter.f The latter 

 method is awkward for routine operation, but has the most efificient geometry, and is 

 generally preferred for rocks and minerals of very low )3 activity (the order of 0.001 

 per cent uranium or equivalent) when large counting intervals are used and clean-up 

 time becomes relatively unimportant. Even under most favorable conditions, however, 

 ultimate sensitivity of the jS counter will be less than that of the more sensitive a 

 detectors, largely because of the cosmic-ray$ and contamination background to which 

 all jS counters are subject. 



?• «/cm' 



?• e/cm» 



?,«A 



Fig. 630. — Beta counting rate plotted against apparent density of samples with constant radioactive 

 content: A, 0.43 per cent UgOg-, B, 0.4 per cent ThO^; C, 12 per cent Kj,0. 



Because of the uncertainties in the energy distribution of ^3 spectra, back-scattering, 

 and absorption, theoretical calibration of beta counters can be only very rough. An 

 empirical calibration is therefore essential. Beta counting of samples of known radio- 

 active content and a statistical comparison of the beta activities of these "standard" 

 samples to the activities of unknown samples, in general, constitute sufficient calibra- 

 tion, if all samples are essentially identical in their apparent density,§ and in the pro- 

 portion of radioactive elements contributing to the total activity, which also implies 

 constancy of the factors of radioactive equilibrium and emanating power. This condi- 

 tion may be met, to a greater or lesser extent, in various types of rocks, and the simple 

 standard-sample calibration will give reasonably accurate results when judiciously 

 applied. 



For more accurate work with heterogeneous samples it is usually desirable to make 

 a density correction.ff The effect of variable apparent density can be established very 

 simply by diluting, in a given proportion, a known amount of radioactive material with 



t A. M. Gaudin and J. H. Pannell, "Radioactive Determination of Potassium in Solids," Anal. 

 Chem. 20, 1154-6 (1948). 



t J. M. Jauch, "Cosmic Rays," Nucleonics 4, No. 4, 39-51, and No. 5, 44-58 (April, May, 1949). 



§ Apparent density is here defined as the weight of the actual sample divided by its volume 

 when placed under the counter. 



tt H. Faul and G. R. Sullivan, "Density Correction in Beta-Ray Assaying of Rock and Mineral 

 Samples," Nucleonics 4, No. 1, 53-6 (Jan., 1949). 



