1214 THE BELL SYSTEM TECHNICAL JOURNAL, SEPTEMBER 1956 



pie a sensitive galvanometer is connected. It is unimportant whether the 

 contact is made to the p-type or to the n-type side of the sample. The 

 best results are obtained on freshly lapped and clean surfaces. It is, 

 therefore, advisable to keep the sample on the steel cylinder while ap- 

 plying the thermoelectric probe. 



When applying the probe, the sample is moved in the x-direction to 

 the point of zero deflection on the galvanometer, whereby the point rests 

 on the junction. 



2. Point Rectification on the Surface 



This test is also usable on p-n junctions which are not rectifying. With 

 one fixed ohmic contact to the sample, the point rectification of the mov- 

 able needle can be displayed on an oscilloscope. By crossing the junction 

 with the needle, the characteristic changes from p-n to n-p. Thus, the 

 needle again can be placed on the junction. 



This test was applied on lightly doped Ge-layers. The oscilloscope pat- 

 tern is not very definite, since on a lapped surface the point rectification 

 is poor. However, with some experience the junction can be located. It 

 is advisable to repeat the measurements several times. Boiling the sam- 

 ple in water before applying the probe improves the surface. 



3. Photoelectric Probe 



This method requires that the junction exhibit rectifying properties. 

 It is most successfully applied to silicon. Between the needle and a con- 

 tact to either the p-type side or the n-type side of the sample, a high 

 impedance voltmeter is connected. While the sample is strongly illumi- 

 nated, it is moved in the x-direction. When the needle crosses the junc- 



n LAYER 

 P MATERIAL 



Fig. 4 — ^ Arrangement for Cu-plating the p-tj^pe side of a p-n junction. 



