THEORY OF THE SWEPT INTRINSIC STRUCTURE 1241 



[concentration (holes plus electrons) would be symmetrical with a maxi- 

 [mum in the center. As the applied bias is increased the hole and electron 

 distributions are further displaced relative to one another and the space 

 charge increases. Finally, at high enough biases, so many of the carriers 

 are swept out immediately after being generated that few carriers are 

 left in the intrinsic region. Now the space charge decreases with increas- 

 ing bias until there is negligible space charge, and a relatively large and 

 constant electric field extends through the intrinsic region from junction 

 to junction. This may happen at biases that are still much too low to 

 appreciably affect the high fields right at the junction or in the extrinsic 

 layers, which remain approximately as they were for zero bias. 



The current will increase with voltage until the total number of 

 carriers in the intrinsic region becomes small compared to its normal 

 value. After that, there is negligible further increase of current with 

 voltage. All the carriers generated in the intrinsic region are being sw^ept 

 out before recombining. In general, the current will saturate while the 

 minimum field in the intrinsic region is still small compared to the 

 average field. 



Comparison with the NP Structure 



The analysis is more difficult than in a simple reverse-biased NP 



structure. In the NP case there is a well defined space charge region in 



I which carrier concentration is negligible compared to the fixed charge of 



i the chemical impurities; so the field and potential distributions are easily 



found from the known distribution of fixed charge. Outside of the space 



i charge region are the diffusion regions in which the minority carrier con- 



j centration rises from a low value at the edge of the space charge region 



; to its normal value deep in the extrinsic region. However, there is no 



, space charge in this region because the majority carrier concentration, 



I by a very small percentage variation, can compensate for the large per- 



I centage variation in minority carrier density. The minority carriers flow 



by diffusion. Since the disturbance in carrier density is small compared to 



the majority density, the recombination follows a simple linear law 



(being directly proportional to the excess of minority carriers). Thus 



the minority carrier distribution is found by solving the simple diffusion 



equation with linear recombination. 



None of these simplifications extend to the NIP or NI or IP structure. 

 There is, in the intrinsic region, no fixed charge; hence the space charge 

 is that of the carriers. There is no majority carrier concentration to 

 maintain electrical neutrality outside of a limited space charge region. 



