ABSTRACT
Imagine that we have a perfect single crystal of a semiconductor, cooled, as described in the last chapter, so that leakage currents are negligible. If an X-ray photon of moderate energy (a few keV) enters the crystal, it will, in all probability, excite a photoelectron, which will have the energy of the photon, less the ionization energy. The photoelectron will travel through the crystal (typically, over less than 1 µm), exciting further electronhole pairs as it goes. It is easy to see that this process will continue until all the electrons have a kinetic energy less than the band gap. Hence the average energy to produce an electron-hole pair will be between one and two times the band gap. In fact there are other ways that an electron can lose energy, without exciting further ionizations; the overall average energy per electron-hole pair in reality is therefore rather higher than our simple argument would predict. In Si, for example, the band gap is 1.1 eV, while it takes on average about 3.8 eV to excite each electron-hole pair. In Ge, the values are, respectively, 0.67 eV and 2.98 eV Nevertheless, the end result is a cloud of electrons and holes, whose number is proportional to the energy of the incoming X-ray. If an electric field is maintained across the crystal, the electrons and holes will be attracted to opposite electrodes; the result will be the appearance of a charge at the electrode, proportional to the energy of the incoming X-ray. Although laborious to describe, this entire process takes place very fast, so the charge appears to come as a single pulse.
