p-n JUNCTIONS IN SEMICONDUCTORS 



457 



the transition region than in the n-region; the effective length of flow in the 

 //-region, being Lp , is greater than the width of the transition region. Con- 

 sequently, the variation of cpp shown in Fig. 6(c) is seen to be reasonable. 

 Similar considerations apply to <p„ . As is shown in Fig. 6(c), the application 

 of dip does not alter <pp — \f/ in the p-region nor <p„ — xp in the «-region. The 



p-REGION 



TRANSITION 

 REGION 



Xa ^Tp XTn 



DISTANCE THROUGH SAMPLE, X *- 



F'g 6 — Simplifiod model of a p-n junction. 



(a) Distribution of donors and acceplcrs. 



(b) Poienlials for thermal equililirium. 



(c) Eflccl of 8.p applied polcnlial in forward direction. 



reason, as discussed in connection with (2.31), is that in these regions elec- 

 trical neutrality requires an essentially constant value for tlie more abundant 

 carrier. Hence the relationships between the ip's and xp follow from (2.4). 



The nature of the potential distribution in the transition region has no 

 effect in the considerations just discussed. However, as shown in Section 2, 



