The Theory of p-n Junctions in Semiconductors and p-n Junction Transistors

S IS well known, silicon and germanium may be either «-type or ^>-type semiconductors, depending on which of the concentrations N,I of donors or NA of acceptors, is the larger. If, in a single sample, there is a transition from one type to the other, a rectifying photosensitive p-n junction is formed. 1 The theory of such junctions is in contrast to those 1 For a review of work on silicon and germanium during the war see H. C. Torrey and C. A. Whitmer, Crystal Rectifiers, McGraw-Hill Book Company, Inc., New York (1948). P-n junctions were investigated before the war at Bell Telephone Laboratories by R. S. Ohl. Work on p-n junctions in germanium has been published by the group at Purdue 435 436 BELL SYSTEM TECHNICAL JOURNAL of ordinary rectifying junctions because, on both sides of the junction, both electron flow and hole flow must be considered. In fact, a major portion of the hole current may persist into the w-type region and viceversa. In later sections we show how this feature has a number of interesting consequences, which we shall describe briefly in this introduction. A p-n junction may act as an emitter in the transistor sense, since it can inject hole current into w-type material. The a-c. impedance of a p-n junction may exhibit a frequency dependence characterized by this diffusion of holes and of electrons. For high frequencies the admittance varies approximately as (/co)1'2 and has comparable real and imaginary parts. When a p-n junction makes contact to a piece of /z-type material containing a high concentration of injected holes, it acts like a semipermeable membrane and tends to come to a potential which corresponds to the hole concentration.