Why does silicon not do organic chemistry?
Silicon has four valence electrons, forms four bonds, and sits directly below carbon in group 14. On paper it should be able to do everything carbon does.
It cannot, and the reason is a pair of bond energies.
The C–C bond is about 346 kJ/mol. The Si–Si bond is about 222 kJ/mol — markedly weaker. Silicon is a larger atom, so its bonding orbitals overlap less effectively and the bond it forms with itself is longer and looser. Chains of silicon atoms exist, but they are fragile: silanes beyond a handful of atoms decompose readily, while carbon chains run to millions of atoms in a polymer.
The second half of the problem is oxygen. The Si–O bond is about 452 kJ/mol, far stronger than Si–Si and stronger than C–O at roughly 358 kJ/mol. Any silicon chain in the presence of oxygen is thermodynamically driven to become silicon dioxide. Carbon has no equivalent vulnerability — C–C and C–O are comparable, so carbon chains persist in an oxygen atmosphere.
There is a third obstacle. Carbon forms strong double and triple bonds with itself, which is what makes alkenes, alkynes, benzene rings and the whole of aromatic chemistry possible. Silicon’s p orbitals are too diffuse to overlap side-on efficiently, so Si=Si bonds are weak and were not even synthesised in a stable compound until 1981.
The consequence at planetary scale is stark. Silicon is about 28% of Earth’s crust by mass and carbon about 0.02%, yet life is built from the scarce element. When silicon and oxygen meet they make rock, and rock does not do chemistry.
There is a final practical point often raised in discussions of silicon-based life. Carbon’s oxide is a gas at ordinary temperature, so an organism can exhale it. Silicon’s oxide is quartz. Respiration would fill the organism with sand.
Why is silicon the material of the semiconductor industry?
Silicon’s band gap is about 1.1 eV at room temperature, and that value is close to ideal for switching.
A conductor has no gap and cannot be turned off. An insulator has a gap too wide to cross at any reasonable voltage and cannot be turned on. A semiconductor sits between, and 1.1 eV is large enough that a silicon device is genuinely off when off — leakage stays low even at elevated temperature — while small enough that ordinary supply voltages turn it on.
Germanium, tried first, has a gap of only 0.67 eV. Germanium transistors leak badly as they warm and were displaced for that reason.
But band gap alone does not explain silicon’s dominance, because several materials have suitable gaps. The decisive advantage is the oxide.
Silicon dioxide grows directly on a silicon surface by simple thermal oxidation, and it is an outstanding electrical insulator, chemically stable, and forms an atomically clean interface with the silicon beneath. That interface is what the MOSFET gate is built on, and the MOSFET is the switch that essentially all modern computing runs on. Germanium’s oxide is water-soluble and useless. Gallium arsenide, faster than silicon in some respects, has no comparable native oxide.
Add to that abundance — silicon comes from sand — and the ability to grow single crystals of extraordinary purity, on the order of one impurity atom in a billion, and the choice was settled by the 1960s.
How do you read silicon’s electron configuration?
Silicon’s ground state is [Ne] 3s² 3p²: a neon core, two electrons in 3s, two in 3p. Electrons per shell: 2, 8, 4.
Two electrons across three 3p orbitals means two singly occupied orbitals with parallel spins, by Hund’s rule — the same arrangement as carbon one row above, and visible as two single arrows in the filling diagram below. The term symbol is ³P₀, identical to carbon’s.
Four valence electrons make silicon tetravalent, and in elemental silicon each atom bonds to four neighbours in the same diamond cubic lattice that carbon adopts as diamond. The difference is the bond strength: diamond’s rigid short bonds make it the hardest natural material with a band gap of 5.5 eV, an insulator. Silicon’s longer, weaker bonds give a much narrower gap, and a semiconductor.
This page classifies silicon as a metalloid. It sits on the diagonal band between metals and non-metals, and it behaves accordingly: it has a metallic lustre and conducts, but poorly and with a conductivity that rises with temperature rather than falling, which is the signature of a semiconductor rather than a metal.
What is silicon actually made into?
Semiconductors take a strikingly small share of world silicon production. The bulk goes into materials most people never think of as silicon at all.
Silicate minerals make up the majority of Earth’s crust, and construction runs on them. Cement, concrete, brick, glass and ceramics are silicate chemistry. Ordinary glass is silica melted with sodium carbonate and lime — the sodium lowers the melting point, the lime restores durability.
Ferrosilicon, an iron–silicon alloy, is used in steelmaking as a deoxidiser and consumes a large fraction of production. Silicones, the polymers with alternating Si–O backbones, exploit exactly the bond strength that prevents silicon biochemistry: the Si–O chain is so stable that silicone rubber tolerates temperatures and chemical attack that would destroy carbon-based polymers.
Photovoltaics are the fastest-growing use. A solar cell is a large-area silicon diode, and the same band gap that makes silicon a good switch makes it a reasonable absorber of sunlight — the solar spectrum peaks not far above 1.1 eV.
Solar-grade silicon needs less purity than electronics-grade, which is why the photovoltaic industry could scale on refining capacity that already existed.
How is ultrapure silicon made?
Semiconductor silicon requires purity of around one impurity atom per billion — sometimes described as nine nines. Sand is nowhere close, and the route from one to the other has three stages.
First, quartz is reduced with carbon in an electric arc furnace to give metallurgical grade silicon, about 98% pure. This grade is adequate for aluminium alloys and silicone manufacture and accounts for most production.
Second comes chemical purification, usually the Siemens process. The silicon is converted to trichlorosilane, a liquid at ordinary temperature, which is then distilled repeatedly. Distillation is the key move: separating impurities from solid silicon is very difficult, but separating one liquid from another by boiling point is routine chemical engineering. The purified trichlorosilane is then decomposed back to silicon on a heated rod, building up polysilicon.
Third, the polysilicon must be made into a single crystal, because grain boundaries in a polycrystalline material scatter charge carriers and ruin device performance. The Czochralski process melts the silicon and slowly withdraws a rotating seed crystal from the surface, pulling out a cylindrical boule in which the entire crystal lattice is continuous — often two metres long and 300 mm across.
That boule is sliced into wafers, and every integrated circuit begins as a slice of one.