PRODUCTION OF CONTINUOUS AND CHARACTERISTIC X- RADIATION 



4 lO 



-SO 



0-25 



05 



075 



«/«. 



lO 



0-2S 



06 



0-7S to 



Fig. 1. Calculated energy distribution of X-radiation from thin targets for an accelerating voltage 

 of 10 kV. The units are ergs/steradian/unit frequency interval/electron/atom-per-cm^. (After Kirk- 

 patrick and Wiedmann, 1945.) 



seen to be a useful approximation of 90°, al- 

 though is not applicable in the forward di- 

 rection. 



The intensity averaged over all angles, 

 (but at the high energy limit only), is illus- 

 trated in Fig. 2, as a function of ZV^', again 

 after Kirkpatrick and Wiedmann. The sim- 

 ple Kramer's theory predicts exact propor- 

 tionahty; and this is indeed very nearly so, 

 over a wide range of the parameters. 



Amrehn and Kuhlenkampff (1955) deter- 

 mined the spectral distribution at 90° from 

 thin targets of almninmn, nickel and tin. 

 The effective thicknesses of their targets 

 ranged from 4 X lO^^gm/cm^ (tin) to 1.2 X 

 10~^ gm/cm^ (aluminmn). By comparing this 



v/z'oo 



with the Aufhelhmgsdicke of 5-8 X 10-« 

 gm/cm^ (for the electron energies of 25 to 

 40 keV used by these authors) it can be 

 seen these represent a real approach to the 

 condition of ideal thinness. It is perhaps 

 more important to note that the energy loss 

 in these targets would not exceed about 0.15 

 keV, a small fraction of the incident energy. 

 The intensity from such targets is low, but 

 by using a proportional counter and pulse 

 height analyzer it was possible to obtain 

 data of good statistical accuracy in reasona- 

 ble times (30 seconds for each observation). 

 Fig. 3 shows the spectral distribution ob- 

 tained at right angles to the direction of the 



Fig. 2. Total intensity at the high energy limit 

 as a function of ZVV (After Kirkpatrick and 

 Wiedmann, 1945). The units are as in Fig. 1. 



electron beam, using an accelerating voltage 

 of 34 kV, and a target of almninum. The 

 corresponding theoretical curve is given 

 (normalized to give the best fit with the ex- 

 perunental data) and it is seen that the agree- 

 ment in shape is very good indeed. When 

 comparing aluminum and nickel, the total 

 intensity was found to be closely proportional 

 to Z^, (for equal atoms/cm^), although with 

 tin the agreement was somewhat less good. 



Angular Distribution. The angular dis- 

 tribution of the continuous spectrmn from 

 thin targets has been investigated by several 

 workers. The radiation from targets of alu- 

 minum 6000 A in thickness was examined by 

 Kulenkampff (1928) using electron energies 

 from 16 to 38 keV. It was observed that the 



655 



