Why only silicon in IC fabrication?

Answer 1

Transistors require semiconductor material to be able to function since a transistor must be able to change it's state of conductivity according to its working conditions. Although many elements these days are involved in manufacturing of transistors. Fundamentally two common semiconductors are described for educational purpose for BJT (bipolar junction transistors). They are Silicon (Si) and Germanium (Ge). Silicon is never intrinsic (pure) in transistors.

To form a p-n-p or n-p-n junction they are doped with pentavalent (5 valance electrons) and trivalent (3-valance electrons) impurities into their crystal lattice. Common impurities in silicon transistors may be trivalent Boron for p-type and pentavalent phosphorus for n-type. Germanium conducts better when in conductive state than silicon due to 32 electrons per atom, but due to high electron density the device can handle very little electrical current.

Germanium was used in the past for pre-amplifiers. Silicon does not have as good conductivity and also does not provide very high hfe values. The highest hfe value you will find in signal transistors would be approximately 300, whereas power transistors you would commonly have an hfe of about 25. Silicon only has 14 electrons per atom. The main advantage is with silicon is that it has a lower electron density when it is in conductive state; to allow larger currents and higher power dissipation.

In the past, difficulty was experienced with the practical use of silicon due to its lack of 'purity'. Once a purer form of silicon was produced, there was no stop to it. Silicon is more cost effective. In 1998 silicon sold for $10 p/kg compared to germanium which was almost at $1800 p/kg.

Germanium is showing some comeback again. Gallium arsenide (GaAs) in wireless communications devices are being replaced with Silicon-germanide (SiGe) and become more useful with modern high speed integrated circuits. Germanium is also commonly used in infrared night vision systems and fiber-optics.

Ultimately one cannot say that Silicon is the only element used in transistors, but what one can say is that it is probably the most commonly used and most fundamental for modern applications.

Answer 2

Actually ICs can be made with other materials than silicon (e.g. gallium arsenide), it is just more difficult and expensive (sometimes prohibitively so).

One example is germanium, although the earliest prototype ICs were made with germanium they were prohibitively expensive to mass produce because:

• germanium oxide is crumbly and water soluble so cannot be used to passivate the chip and support the wiring
• wiring had to be done by hand between the IC components, both limiting the number of components per IC and raising the cost

Answer 3

Silicon has several advantages over germanium:

• easier to process - silicon dioxide coating can be used as doping mask, germanium dioxide is not stable to use. etc.
• easier to package - silicon dioxide or silicon nitride passivation can protect junction edges from contamination allowing cheap molded plastic packaging, germanium cannot be passivated requiring expensive hermetically sealed metal or glass packaging to prevent junction edge contamination.
• wider temperature range - silicon will operate at junction temperatures up to 150C, germanium will only operate at junction temperatures up to 60C.
• etc.

Answer 4

Silicon is used because it has semi-conductor properties.

It can conduct or not, depending on the junction formed and the deliberate impurities added (doped).

Other substances can be used such as Germanium.

Silicon is an element and can be grown in a laboratory, to form a pure crystal structure, suitable for making semi-conductors.

It is easier and cheaper to grow silicon crystals than some other semiconductor materials.