Sec. 13.10] STANDARDIZATION OF RADIOACTIVE SAMPLES 



377 



sample supports consisting of films of materials, such as nylon, polyethylene, 

 or aluminum, which have low atomic weights. Scattering by the intervening 

 air and parts of the apparatus can, of course, be greatly reduced by proper 

 design and by placing the entire unit in an evacuated box, as shown in Fig. 

 103. 



Routine counting of samples usually necessitates deposition of several 

 milligrams of active material on some kind of standardized sample holder, 

 as described in Chap. 17. For a thin sample the computed and experimental 



0.25 



0.5 



1.75 



2.00 



0.75 1.00 1.25 1.50 



UNITS OF COUNTER DIAMETER 



Fig. 108. Relative counting efficiency (for P 32 ) as a function of the lateral displacement of 

 the source from the axis of the counter. <i> denotes the geometric efficiency when the sample 

 is on the counter axis. [Unpublished data of H. Anger.] 



efficiency decreases as a function of the distance the sample is displaced 

 laterally from the axis of the counter in a manner indicated in Fig. 108. 

 Samples with finite thickness introduce an additional effect, a combination 

 of absorption and scattering in the sample material, often called self-absorp- 

 tion. Along with this there is also distortion of the isotropic angular dis- 

 tribution of the beta particles. A counter, however, may be calibrated for 

 measuring samples of any thickness by obtaining first a standard self-absorp- 

 tion curve for a given substance. This may be done in two different ways. 

 A dilution curve is obtained if a sample of known radioactivity and negligible 

 mass is progressively diluted by addition of stable carrier. With samples of 

 identical intensity and sample area the counting rate decreases progressively 

 as the mass of carrier is increased. Another method of measuring self- 

 absorption is by the use of radioactive samples with the same specific activity 

 but different total mass or thickness. Instead of a linear relationship between 

 sample thickness and counting rate, one obtains a typical saturation curve 



