Although semiconductor conductivity (see Section 3.1) at temperatures above the absolute zero point is considerably higher than the conductivity of insulators, it is nonetheless very low. This conductivity can be substantially increased by adding suitable external atoms in a process known as semiconductor doping.
Table 3.1 Band gap energy and absorption mechanisms for selected semiconductor materials
Figure 3.4 Crystal lattice and band model of undoped silicon (intrinsic Si), EG = 1.12eV
If (as in Figure 3.5) a silicon atom is replaced by a phosphorus atom with five valence electrons in the outermost shell, one of these electrons cannot form a bond with one of the four adjacent atoms. As a result, the electron can readily liberate itself from its atomic nucleus, which can then regain its positive charge. Hence a phosphorus atom ‘donates’ an electron to the crystal lattice and is thus referred to as the donor,
Figure 3.5 Crystal lattice and band model of silicon (n-Si) that is doped with donors (P)
Crystal lattice and band model of silicon (p-Si) that is doped with acceptors (P)
while the counterpart electron is referred to as the donor electron. In the band model, the energy of the donor electron is only slightly below (by ED) the band threshold for the conduction band, which means that only a minute amount of energy needs to be obtained from the temperature-induced motion for the donor electron to migrate to the conduction band. Donor electrons engender conductivity by means of negative charge carriers; thus the semiconductor is n-conductive.
On the other hand if (as shown in Figure 3.6), a silicon atom on the outermost shell is replaced by a boron atom with only three valence electrons, only three of the four boron atoms bonding with the adjacent silicon atoms can be saturated with electrons, i. e. one of the four bonds is missing an electron, thus engendering a hole. An electron from an adjoining atom’s valence bond can fall into this hole, in which case the original hole disappears, a new one is created elsewhere and the boron atom becomes negatively charged.
In the band model, the energy available at the space for the electron not provided by the acceptor atom for the fourth adjacent atom is only slightly higher (by EA) than at the upper edge of the valence band. Hence an electron in the valence band needs only a minute amount of energy from the temperature-induced motion to fill the aforementioned space and in so doing leaves a hole in the valence band.
A hole in a valence band can move as freely as a free electron in the conduction band and is thus also referred to as a defective electron. A boron atom can ‘accept’ an electron and is thus called an acceptor. Hence acceptors contribute to conductivity through holes in the valence band, i. e. they are positive charge carriers, thus rendering semiconductors p-conductive.
Key to understanding the p-n junction process is that donor atoms that have donated an electron constitute permanently integrated positive charges (ions) in the crystal lattice; conversely, acceptor atoms that have absorbed an electron constitute permanently integrated negative charges (ions) in the crystal lattice.
Hence semiconductor doping is actually a process involving targeted semiconductor contamination that can only be carried out in a controlled manner if the semiconductor is extremely pure – which for silicon means one external atom per 1010 silicon atoms. Needless to say, attaining this level of purity is an extremely cost-intensive process, prompting manufacturers of inexpensive solar cells to reduce this cost by reducing the requisite purity gradient of the base material through the use of so-called solar silicon in lieu of purer electron silicon.
In addition to P, other five-valence elements such as As, Sb or Bi can serve as donors. As for acceptors, B, Al, Ga or In can also be used.