Semiconductors

The advent of the semiconductor has revolutionised our lives, since it is the basis of all integrated circuits and microprocessors.

To distinguish between the electrical properties of materials we can group them into three sections:

(a) conductors,
(b) semiconductors and
(c) insulators.

You are probably aware of many conductors and insulators such as copper and rubber; semiconductors include materials such as silicon, germanium, carbon, selenium, gallium arsenide, lead sulphide.
The important difference between conductors, semiconductors and insulators lies in the number of free electrons present in the material. Perhaps the best way to consider the differences between them is to use the band theory of solids.

As you may know, electrons in an individual atom are restricted to well-defined energy levels and energy changes within the atom only take place between one level and another.

In a solid the atoms are linked together and the electrons can occupy a whole series of energy levels grouped into bands (see Figure 1). The difference in energy between levels within the band is very small compared with the energy gap between the bands. The electrical differences between one type of solid and another lie in the different arrangements of the bands.

The band structures of a conductor, semiconductor and insulator are shown in Figure 2.

Conductors

In a conductor the valence band is full of electrons, while the conduction band has some free electrons and many empty energy levels. The addition of a very small amount of energy will allow electrons to move within the conduction band, some rising to a higher level and others returning to lower levels. This movement of electrons is electrical conduction.

In some conductors the valence band and the con- duction band actually overlap. This effectively gives a partly filled top band.

Intrinsic semiconductors

We will deal first with the intrinsic semiconductor. This is a material that is a semiconductor 'in its own right' - nothing has been added to it.

In the intrinsic semiconductor the valence band is full once more, but the conduction band is empty at very low temperatures. However, the energy gap between the two bands is so very small that electrons can jump across it by the addition of thermal energy alone or even light energy of a suitable wavelength. In other words, heating the specimen or shining a light on it maybe sufficient to cause electrical conduction. The conductivity increases with temperature as more and more electrons are liberated. Semiconductors therefore have negative temperature coefficients of resistance.

For germanium the energy gap is 0.66 eV and for silicon it is 1.11 eV at 27 ^oC. When an electron jumps to the conduction band it leaves behind it a space or hole in the valence band. This hole is effectively positive and since an electron can jump into it from another part of the valence band it is as if the hole itself was moving! Conduction can take place either by negative electrons moving within the conduction band or by positive holes moving within the valence band.

A semiconductor may be thought of as similar to an almost full multi-storey car park, the cars representing the electrons and the spaces the holes (no cars are allowed to enter or leave the car park, however, only to drive round within it!). (Figure 3)


If this idea of holes seems odd to you, think of a pile of earth and the hole in the road from which it came. Both the pile (electron) and the hole (hole) have a physical effect on you if you run into them on a bike! Conduction by positive holes is rather like workmen digging up a road; in a way, they are only moving a hole from one place to another.

Insulators

In the insulator the valence band is full once again, but in these substances the energy gap between this and the empty conduction band is very large. It would take a great deal of energy to make an electron jump the gap and to cause the insulator to break down. At very high temperatures or under very large electric fields breakdown will occur, and like semiconductors the greater the temperature the greater the conduction. Insulators, like semiconductors, have negative temperature coefficients of resistance.

Extrinsic semiconductors

An extrinsic semiconductor is basically a semiconductor to which a very small amount of impurity has been added. About one atom per million is replaced by an impurity atom; this process is called doping.
Doping with an impurity can have quite marked effects on the electrical properties of the material. The addition of one impurity atom in one hundred million will increase the conductivity of germanium by twelve times at 300 K. Very precise doping may be achieved by neutron irradiation.

We will consider the effects of doping a piece of silicon. Silicon is made up of tetravalent atoms joined in a lattice, as shown in Figure 5. Two types of semiconductor can be made by doping with different impurities:

(a) n-type, by doping with pentavalent material such as phosphorus;

(b) p-type, by doping with trivalent material such as aluminium.

The effect of both types of doping is shown in the diagram. With the p-type each impurity atom has one fewer electron than the silicon atom, while with the n-type they have one extra electron.

Figures 6 and 7 show how the impurity atoms fit into the energy level diagram of the solid as a whole. In the p-type material the aluminium levels fall just above the full valence band of the silicon. These levels are very close to this band and so electrons can easily jump into them from the valence band. For this reason they are called acceptor levels. When an electron jumps up to these levels it leaves behind a hole in the valence band; it is the movement of holes within the valence band that causes the greatest conduction in a p- type material. In the n-type material the phosphorus energy levels fall just below the empty conduction band of the silicon, and very close to it. For this reason electrons can very easily jump from them into the conduction band, and they are therefore called donor levels. In n-type material conduction takes place mainly due to the movement of these electrons.