Inside every modern computer or other data-processing wonder is a microprocessor bearing millions of transistors sculpted from silicon y chemicals and light. Silicon, the second most abundant element on Earth, is used extensively in transistors, integrated circuits, photoelectric devices, and other electronic devices. A pure crystal of silicon does not conduct electricity unless traces of other substances are diffused or doped onto it. Therefore silicon must be manufactured in a specific way in order for it to yield electrons for a current.
All manufacturing of silicon takes place in a "clean room" which is an area where each one foot cube of air must contain fewer than 1000 tiny specks of dust and zero humidity. The temperature is maintained at a constant 68 degrees Fahrenheit and all workers have to wear coats, gloves, masks, and overshoes. This is necessary because even one dust particle or water droplet can ruin a batch of chip production.
The manufacturing of a silicon chip starts when silica, the main component of sand, is heated with carbon which makes 98 percent pure silicon. This is then dissolved in hydrochloric acid. The resulting liquid is fractionally distilled to separate almost all of the impurities.The remaining liquid is then heated in a hydrogen tmosphere, which produces the purest silicon possible. This silicon, however, is in the form of many crystals of different sizes and orientations. This silicon goes through the Czochralski pulled crystal process in which it is melted in a large crucible into which a probe, tipped with a small seed crystal is immersed. Silicon atoms attach themselves to the seed in perfect alignment with its structure while it is rotated and pulled slowly upward. The seed grows into a three foot long, cylindrical, single crystal.
Silicon this pure is hard, dark grey in color and lusterous. The giant crystal of silicon is next ground into a perfect cylinder, which is sliced by a diamond-tipped saw into wafers 1 mm thick. Using particles one-tenth of a micrometer wide, the faces of these wafers are polished to give a smooth base onto which up to two hundred dentical chips can simultaneously be photo-etched.
The base of the chip is next doped with small traces of boron. First the silicon base is coated with a layer of insulating silicon dioxide and photoresist, a light-sensitive material. This hardens only where ultraviolet light, projected through a mask, strikes it. The chip is immersed in solvent to wash away the soft resist shielded by the mask. Hot gases than etch away most of the dioxide, leaving a thin layer for insulation, and the rest of the resist is removed.
A layer of conducting polysilicon is now deposited, as well as a new layer of resist. A second exposure through a mask is made, and after washing the chip with solvent an L-shaped pattern remains on the chip. Etching removes the polysilicon not shielded by the resist as well as a thin layer of dioxide, exposing two wells of the silicon base. The rest of the resist is then removed, so the gate rises above the wells. Doping transforms the two wells of p-type into n-type silicon, which carries only negative charge.
Electrical connections are now added to the circuit. Layers of dioxide and resist are again deposited and a third masking is carried out. Etching creates shafts to the n-type silicon and to the gate. Aluminum is then spread over the surface. Masking and etching leaves three metal electrical conatacts.
Up to five hundred chips may be created on a single wafer. Each is tested, and working chips are separated with a diamond-tipped saw and mounted in a plastic casing. They are connected to metal legs by fine gold wire. Tiny chips of silicone are connected forming complex circuits that are used in computers, satellites and spacecraft.