The p-n Junction

The n-p Junction. Before beginning a discussion of electron transfer at interfaces between H-type semiconductor/solution interlaces, it is helpful to describe something of the theory of the famous n-p junction. This is not a part of electrode-process chemistry (which deals with electron-transfer reactions between electronically and ionically conducting phases), but it is the basis of so much modem technology (e.g., the transistor in computers) that an elementary version of events at the junction should be understood. Further, knowing about the n-p junction makes it easier to understand electrochemical interfaces involving semiconductors. 

The n-p junction was discussed in Section 7.4.1.2. In the original concept, this junction resulted from the transfer of electrons from one semiconductor to another. In the figures in Sec. 7.4, potential-distance relations for the junction of n and p semiconductors are shown. It is clear that here the transfer of one charge carrier from one semiconductor to the next in an uphill direction can be thought of as being opposed by the electrical potential hill shown. Such potential hills are termed Schottky barriers. ... 

The n-p junction is formed by letting lithium diffuse into a p-type silicon. The diffusion can be accomplished by several methods. Probably the simplest method consists of painting a lithium-in-oil suspension onto the surface... 

A more effective carrier confinement is offered by a double heterostmcture in which a thin layer of a low band gap material (the active layer) is sandwiched between larger band gap layers. The physical junction between two materials of different band gaps, and chemical compositions, is called a heterointerface. A schematic representation of the band diagram of such a stmcture is shown in Figure 4. Electrons injected under forward bias across the p—N junction into the lower band gap material encounter a potential barrier, AE at thep—P junction which inhibits their motion away from the junction. The holes see a potential barrier of AE at the N—p heterointerface which prevents their injection into the N region. The result is that the injected minority... 

Fig. 2. (a) A schematic diagram of a n—p junction, including the charge distribution around the junction, where 0 represents the donor ion 0, acceptor ion , electron °, hole, (b) A simplified electron energy band diagram for a n—p junction cell in the dark and in thermal equilibrium under short-circuit... 

As Figure 10 shows, the n—p—n bipolar junction transistor (BJT) may be regarded as two back-to-back p—n junctions separated by a thin base region (26,32,33). If external voltages are applied so that the base-emitter (BE) junction is forward biased and the base-coUector (BC) junction is reverse biased, electrons injected into the base from the emitter can travel to the base-coUector junction within their lifetime. If the time for minority carrier electrons to... 

Eig. 10. The n—p—n transistor biased ia its active region, where 7 = current, (------) indicate depletion regions at the p—n junctions, and S is the electric field ... 

Much of the discussion of this subsection has been based on the behavior of hydrogenated diodes annealed under reverse bias. Annealing under forward bias has also been studied, though less extensively, and some of the observations have suggested the possibility of a new type of thermal breakup of BH complexes, namely BH + e— B + H° (Tavendale et al., 1985, 1986a). These authors reported breakup of BH in a few hours at 300 K under forward bias, both in Schottky diodes and in n+-p junctions. However, in a similar experiment with an n+-p junction, Johnson (1986) found a slight buildup of BH under forward-bias anneal. Available details of the various experiments are too sketchy to allow useful speculation on the reasons for the different outcomes or possible mechanisms for accelerated breakup. 

If we place n- and p-type semiconducting crystals in contact (a p-n junction), we create a device that conducts electricity preferentially in one direction this is the basis of action of the semiconductor diodes used in the electronics industry, although specially refined silicon (Section 17.8.2) is usually employed rather than Ge. Transistors and electronic chips are designed using similar basic principles—typically with n-p-n or p-n-p junctions. We consider chemical aspects of electronic devices in more detail in Chapter 19. 

Now think of the n —>p hole current. When the holes from the n-type of medium reach the junction [see Fig. 7.21(c)], they do not see any barrier due to an electrical potential difference, so they simply tumble over the potential drop. Hence, the n —> p hole current density at equilibrium is controlled only by diffusion and is simply proportional19 to the number of holes, nhn, in the rz-type of material ... 

One is now in a position to compare the current-potential law for an electrode/elec-tiolyte interface20 (which has been referred to as an e-ijunction) with that for any n-p junction ... 

It is seen, therefore, that there are basic similarities in the i vs. r laws for both types of interfaces, but there is an important difference. There is no symmetry factor P in the exponential / vs. rj law for semiconductor n-p junctions. Why ... 

The situation in the case of the transfer of holes (or electrons) across n-p junctions is different. First, the only difficulty the electron has to overcome is that due to the electric field. When there is no field, there is no barrier. This is because the barrier is not an expression of the energies involved in atomic movements there are no atomic movements as prerequisites to the movements of holes or electrons. Second, whereas potential-energy barriers for atom movements and reactions are like hills, the barrier for hole and electron movements is like a cliff with its attendant implication that falling over the cliff does not involve an activation energy [Fig. 7.21(d)]. Finally, since the holes and electrons reach the barrier top only after traversing the whole distance over which the field extends, the entire e0rj—notafraction (1 - P)e0T]—affects the hole and electron movements. 

There is therefore one essential conclusion from the comparison of electrodic e-i junctions and semiconductor n-p junctions The symmetry factor P originates in the atomic movements that are a necessary condition for the charge-transfer reactions at electrode/electrolyte interfaces. Interfacial charge-transfer processes that do not involve such movements do not involve this factor. By understanding this, ideas on P become a tad less underinformed. Chapter 9 contains more on this subject. 

It is easy to see how the concentration of electrons at the surface will depend on the overpotential. As shown above, in dealing with the rate of electron transfer at an n-p junction, the form of the expression [outside the reversible region, i.e., for T > (RTIF)] is given by... 

---