GENERATION, CONTBOL, AND MEASUREMENT 



235 



1.0 



0.0 1 



o 



< 00 1 



fatigue, and the sensitivity decreases with time; the decrease increases 

 rapidly with intensity, and if exposed to very high intensities, the cell 

 may be permanently damaged owing to excessive current on the later 

 dynode stages. Therefore the photomultiplier is seldom used for precise 

 measurements at high flux intensities. It is especially useful at very low 

 intensities. 



The photomultiplier has a dark current consisting of three components: 

 ohmic leakage, amplified thermionic 

 emission, and regenerative ioniza- 

 tion (Fig. 3-33). Ohmic leakage 

 is due to traces of conducting ma- 

 terials on the stem and insulating 

 elements inside the tube which are 

 unavoidably deposited during man- 

 ufacture. It is an inherent prop- 

 erty of any tube and varies con- 

 siderably among tubes. Ohmic 

 leakage predominates at low values 

 of applied voltage. At normal 

 voltages per stage of 80-100 v for 

 the 931-A, thermionic emission is 

 predominant and sets the limit to 

 the lowest flux intensity that can 

 be measured. Regenerative ioniza- 

 tion occurs only when potentials of 

 110 V or more are applied to the 

 electrodes. It is caused by elec- 

 trons ionizing the small trace of residual gas present in the tube. Unless 

 limited, regenerative ionization will cause the complete destruction of the 

 multiplier. 



The photomultiplier is extremely linear in its response to intensity 

 (Fig. 3-34). Commercial tubes (Engstrom, 1947) have a maximum devi- 

 ation of 3 per cent from linearity over a millionfold range of 10^'" to 10~^ 

 lumen, and this linear behavior appears to extend down to 10~'^ lumen, 

 where single-electron counting must be used. 



Only those photosurfaces which are insensitive to the longer wave 

 lengths in the red, as the S-4, S-5, and S-8 responses, are used in photo- 

 multipliers in order to limit the dark current. The red-sensitive silver- 

 cesium oxide-cesium (S-1) surface is not used in commercial tubes, partly 

 because of its high sensitivity in the near infrared and partly because of 

 technical difficulties of manufacture. The dark current due to thermionic 

 emission from an S-1 surface would be excessive at room temperatures. 



The application of photomultiplier tubes for the measurement of low 

 flux intensities has been discussed by Engstrom (1947) and for scintil- 



0001 



140 



20 40 60 80 100 120 

 POTENTIAL PER STAGE. V 



Fig. 3-33. Effect of dynode voltage on 

 the dark current in a 931-A-type photo- 

 multiplier tube. {From Engstrom, 1947.) 



