HOLE /.yjKCTlOX /.V CERSf A \ I IM' 3,A<-> 



injecting carriers of the same type; hoarier, such i)icr eases can he produced bv 

 injecting the opposite type since the space charge of the latter can he neutralized 

 by an increased co)icenlration of the type normally present. 



Thus \vc conclude that the existence of tu'o processes of electronic conduction 

 in semiconductors, corresponding respectively to positive and negative mobile 

 charges, is a major feature in several forms of transistor action. 



In terms of the description given al)ove, the experiment of Figure 1 is 

 readily interpreted. The instantaneous rise at ti is simply the ohmic contribu- 

 tion due to the changing total currents in the right branch when the emitter 

 current starts to flow. After this, there is a time lag until the holes injected 

 into the germanium drift down the specimen and arrive at the collector. 

 When the current is turned off at /s , a similar sequence of events occurs. 



The measured values of the time lag of ti — /•.. , the field E and the distance 

 L can be used to determine the mobility of the holes. The fact that holes, 

 rather than electrons, are involved is at once evident from the polarity of 

 the effect; the disturbance produced by the emitter point flows in the di- 

 rection of E, as if it were due to positive charges; if the electric field is re- 

 versed, the signal produced at /-j is entirely lacking. The values obtained by 

 this means are found to be in good agreement with those predicted from the 

 Hall effect and conductivity data. The Hall mobility values obtained on 

 single crystal filaments of ;/- and p-type germanium'* are 



IJL„ = 1700 cm sec per volt/cm 



Hn — 2600 cm/ sec per volt cm 



The agreement between Hall effect mobility and drift mobilitv, as was 

 pointed out at the beginning of this section, is a very gratifying confirmation 

 of the general theoretical picture of holes drifting in the direction of the 

 electric field. 



We shall next consider a more quantitative embodiment of the experi- 

 ment just considered. In Fig. 3, we show the experimental arrangement. 

 In this case it is essential in order to obtain large eft'ects that the cross-section 

 of the germanium filament be small. A thin piece of germanium is cemented 

 to a glass backing plate and is then ground to the desired thickness. After 

 this the undesired portions are removed by sandblasting while the desired 

 portions are protected by suitable jigs consisting of wires, scotch tape, metal 

 plates, etc. After the sandblasting, the surface of the germanium is etched. 

 In this way specimens smaller than 0.01 X 0.01 cm in cross-section have 

 been produced. The ends of the filament are usually made very wide so as 

 to simplify the problem of making contacts. 



Under experimental conditions, a battery like Bi, of Figure 1 applies a 

 ''sweeping" field in the filament so that any holes injected by the emitter 



