Silicon

• 1. Introduction

  ›

  Element 14 makes up more than a quarter of Earth's crust by weight, second only to oxygen, yet for most of human history nobody had ever seen it in pure form. Silicon, a hard, brittle, blue-grey crystalline solid, clings to oxygen so tightly that chemists spent decades trying and failing to break the two apart. How did chemists finally isolate a substance so stubborn, and how did an element mined mostly as sand and clay come to define the modern digital economy? The answer runs from an 18th-century chemist's failed guess to the transistor labs of Bell Labs and Fairchild Semiconductor, and ends with an element now central to plant biology, marine ecosystems, and nearly every computer chip on Earth.

• 2. Ancient Use, Naming, and Isolation

  ›

  Predynastic Egyptians shaped silicon-bearing rock crystal into beads and small vases thousands of years before anyone identified silicon as an element, and by around 1500 BC Egyptian and Phoenician glassmakers were manufacturing glass from silica. Natural silicate compounds also went into the mortar of some of the earliest human dwellings. In 1787, the chemist Antoine Lavoisier suspected that silica was the oxide of an undiscovered element, but the material's chemical affinity for oxygen was too strong for the tools of his day to break it free. Sir Humphry Davy tried and failed to isolate the element in 1808, though he proposed the name silicium, combining the Latin silex, meaning flint, with an -ium ending because he assumed it was a metal. Joseph Louis Gay-Lussac and Louis Jacques Thénard are thought to have produced impure amorphous silicon in 1811 by heating potassium metal with silicon tetrafluoride, but they never purified or formally identified their product as a new element. The Scottish chemist Thomas Thomson renamed the element silicon in 1817, keeping part of Davy's coinage but swapping the ending because he believed the element behaved more like a nonmetal, similar to boron and carbon. Jöns Jacob Berzelius finally purified the element in 1824 by reducing potassium fluorosilicate with molten potassium metal and repeatedly washing the resulting brown powder, and for that work he is generally credited as the discoverer. The same year he prepared silicon tetrachloride, decades after Carl Wilhelm Scheele had first made silicon tetrafluoride in 1771 by dissolving silica in hydrofluoric acid. It took another 31 years for anyone to produce silicon in its more familiar crystalline form: the chemist Henri Deville achieved it by passing silicon chloride vapor over pure aluminum, yielding hard octahedral crystals. Friedrich Wöhler discovered the first volatile silicon hydrides, synthesizing trichlorosilane in 1857 and silane itself in 1858, though a detailed study of these silanes waited until the early 20th century and the work of Alfred Stock. Charles Friedel and James Crafts made the first organosilicon compound, tetraethylsilane, in 1863, but thorough characterization of organosilicon chemistry again had to wait, this time for Frederic Kipping in the early 1900s. Starting in the 1920s, William Lawrence Bragg's work in X-ray crystallography finally explained the atomic structure of the silicates, building alongside Linus Pauling's crystal chemistry and Victor Goldschmidt's geochemistry.

• 3. Building the First Transistor

  ›

  The earliest semiconductor devices did not use silicon at all. German physicist Ferdinand Braun built a crystal detector from galena in 1874, and Indian physicist Jagadish Chandra Bose built a radio crystal detector from the same mineral in 1901. The first silicon semiconductor device, a silicon radio crystal detector, came later still, developed by American engineer Greenleaf Whittier Pickard in 1906. Decades passed before silicon's electronic potential became clear. In 1940, Russell Ohl discovered the p-n junction and the photovoltaic effect in silicon, and the following year, wartime demand for radar microwave detectors drove the development of techniques for producing high-purity germanium and silicon crystals. In 1947, physicist William Shockley theorized a field-effect amplifier built from germanium and silicon but could not get a working device to function, and turned to germanium instead; that same year, John Bardeen and Walter Brattain, working under Shockley, built the first working transistor, a point-contact design. Silicon's breakthrough came in 1954, when physical chemist Morris Tanenbaum fabricated the first silicon junction transistor at Bell Labs. The following year, Carl Frosch and Lincoln Derick, also at Bell Labs, stumbled onto the discovery that silicon dioxide could be grown directly on silicon, and by 1957 they had published the first manufactured semiconductor oxide transistor, a planar design with its drain and source sitting side by side on the same surface. In 1959, Robert Noyce built the first silicon-based integrated circuit at Fairchild Semiconductor, improving on earlier work by Jack Kilby that had relied on germanium instead.

• 4. Atomic Structure and Electrical Behavior

  ›

  A silicon atom carries fourteen electrons arranged in the ground-state configuration Ne3s2 3p2, four of them valence electrons that let the atom form four bonds and complete a stable octet, much like its lighter relative carbon. Its first four ionisation energies, 786.3, 1576.5, 3228.3, and 4354.4 kilojoules per mole, are high enough to rule out any simple positively charged silicon ion. Its single-bond covalent radius of 117.6 picometers sits between carbon's 77.2 picometers and germanium's 122.3 picometers, following the expected trend down the periodic table. At room temperature, pure silicon is technically an insulator, because a small energy gap separates its highest filled electron states from the empty ones above. Adding a trace of an element like phosphorus, arsenic, or antimony introduces extra electrons that jump into the empty conduction band, creating what is called an n-type semiconductor, while adding boron, aluminum, or gallium instead creates electron-accepting gaps that produce a p-type semiconductor. Joining n-type and p-type silicon together forms a p-n junction that lets current flow more easily in one direction than the other, acting as a diode, and stacking a thin layer of p-type silicon between two n-type regions creates a transistor capable of amplifying a signal.

• 5. Crystal Form and Isotopes

  ›

  Silicon crystallizes at standard conditions into a diamond cubic lattice, the same arrangement that gives it a high melting point of 1414 degrees Celsius, since breaking that dense network of covalent bonds takes a lot of energy. It boils at 3265 degrees Celsius, lower than the temperature at which carbon sublimes, at 3642 degrees Celsius, reflecting the fact that a silicon-silicon bond is weaker than a carbon-carbon bond. Under higher pressure, silicon can adopt other crystal structures, including a hexagonal close-packed form known as Si-VII that appears around 40 gigapascals, and a body-centered cubic form called BC8 that remains stable even after the pressure is released. Researchers have also managed to build silicene, a single-layer form of the element analogous to graphene. Naturally occurring silicon consists of three stable isotopes: 28Si, making up 92.24 percent, 29Si at 4.67 percent, and 30Si at 3.07 percent, all forged inside Type Ia supernovae during the oxygen-burning process. The fusion of 28Si with alpha particles inside stars, known as the silicon-burning process, is the final stage of stellar nucleosynthesis before a massive star collapses and explodes as a type II supernova. Of twenty-two characterized radioisotopes, the longest-lived is 32Si, with a half-life of about 157 years, followed by 31Si at 2.62 hours; every other known radioactive isotope decays in under seven seconds. Because 31Si can be produced by neutron activation and detected through its characteristic beta decay, it is useful for quantitative chemical analysis. Silicon also reaches the oceans through groundwater and river transport, and scientists have measured a gradient of more than 0.3 parts per thousand in the isotopic composition of deep water between the Atlantic and Pacific oceans.

• 6. Silicates, Glass, and Alloys

  ›

  More than 90 percent of Earth's crust consists of silicate minerals, and most silicon reaches commercial use this way, barely processed from its natural mineral form as clay, silica sand, or building stone. Silicates go into Portland cement for mortar and stucco, and when mixed with silica sand and gravel they form the concrete underlying most large construction projects. Silicate clays also fire into whiteware ceramics such as porcelain, while silica itself is melted into traditional soda-lime glass for windows and containers, drawn into optical fiber, and spun into fiberglass and glass wool for structural support and insulation. Silicon carbide, another silicon compound, serves as an industrial abrasive and a component of high-strength ceramics. Silicone polymers, built on a silicon backbone, show up in waterproofing treatments, molding compounds, mechanical seals, high-temperature greases, caulking, breast implants, and contact lenses; Silly Putty began as silicone oil mixed with boric acid. In metallurgy, elemental silicon added to molten cast iron as ferrosilicon or silicocalcium alloys improves the casting of thin sections and stops cementite from forming when the metal is exposed to air; it also absorbs oxygen from the melt, helping control a steel's carbon content precisely, and it modifies the resistivity and magnetic properties of transformer steel. About 55 percent of the world's metallurgical-grade silicon goes into aluminum-silicon casting alloys for the automotive industry, where a silicon content near 12 percent forms a eutectic mixture that solidifies with very little shrinkage, reducing cracking and improving the hardness and wear resistance of cast aluminum parts.

• 7. Sand to the Silicon Age

  ›

  Producing 96 to 99 percent pure silicon starts by carbothermically reducing quartzite or sand with pure coke inside an electric arc furnace, usually alongside scrap iron to yield ferrosilicon, an alloy that accounts for about 80 percent of the world's elemental silicon output. China alone supplies roughly two-thirds of that output, about 4.6 million tonnes, mostly as ferrosilicon, followed by Russia at 610,000 tonnes, Norway at 330,000 tonnes, Brazil at 240,000 tonnes, and the United States at 170,000 tonnes. Only about 20 percent of that silicon is refined further to metallurgical grade, and of that, an estimated 15 percent is refined again to semiconductor purity, typically expressed as nine nines, or 99.9999999 percent, through the Czochralski process, which grows the monocrystalline silicon wafers used in semiconductors and high-efficiency solar cells. Reaching that purity means distilling and reducing volatile compounds such as tetrachlorosilane or trichlorosilane using very pure zinc metal, then growing the resulting spongy silicon into cylindrical single crystals and refining them further by zone refining; transistor manufacturing demands impurity levels below one part in ten billion. Because pure silicon barely conducts electricity, manufacturers dope it with small amounts of boron and phosphorus to control its conductivity for use in transistors and solar cells, and in an integrated circuit a silicon wafer supports transistors separated by a thin insulating layer of silicon dioxide grown directly on its surface, an oxide that, unlike germanium's equivalent, resists dissolving in water. Cheaper polycrystalline and amorphous silicon, produced in volumes comparable to monocrystalline material, go into liquid-crystal displays and thin-film solar cells; by 2013, polycrystalline silicon output for solar cells alone was projected to reach 200,000 metric tons a year. Because of silicon's dominance in electronics, commentators call the late 20th and early 21st centuries the Silicon Age, comparing it to how the Stone, Bronze, and Iron Ages were defined by their materials. Less than 15 percent of highly purified silicon goes toward the semiconductor chips essential to smartphones and modern computers, yet in 2019 networks and communications devices alone made up 32.4 percent of the semiconductor market, one projected to reach 726.73 billion dollars by 2027. California's Santa Clara Valley lent its nickname, Silicon Valley, to a long list of imitators worldwide, including Silicon Wadi in Israel, Silicon Forest in Oregon, Silicon Hills in Austin, Silicon Slopes in Salt Lake City, Silicon Saxony in Germany, Silicon Fen in Cambridge, Silicon Roundabout in London, Silicon Glen in Scotland, Silicon Alley in New York, and Silicon Beach in Los Angeles. Meanwhile, silicon quantum dots, made by thermally processing hydrogen silsesquioxane into nanocrystals just a few nanometers wide, glow at different colors depending on their size and now appear in displays, solar concentrators, and sensors that detect hazardous chemicals through the quenching of their light.

• 8. Silicon in Living Things

  ›

  Few organisms use silicon directly despite its abundance, but some sea sponges, diatoms, and radiolaria secrete skeletal structures from silica, and many plants deposit silica in their tissues, taking it up as orthosilicic acid and transporting it through the xylem into the cell wall. That silica strengthens cell walls, reduces insect damage and disease, and in some plants triggers the production of defensive compounds called phytoalexins; several horticultural crops rely on it so heavily that fungicides can fail without adequate silicon nutrition. Because it can form complex, stable molecules with four covalent bonds, astrobiologists sometimes name silicon as a hypothetical alternative to carbon for extraterrestrial biochemistry. In the ocean, diatoms take up dissolved silicic acid through a dedicated transport protein and build it into cell-wall structures called frustules, and when diatom cells die and sink in a process called marine snow, they carry silicon and other nutrients from the sunlit upper ocean down to greater depths. The Southern Ocean alone produces roughly a third of the world's marine biogenic silica, so little of which escapes the region that oceanographers call it a biogeochemical divide, while diatoms in the North Atlantic and North Pacific subtropical gyres contribute only about 5 to 7 percent of global marine silica production. In humans, studies have linked higher dietary silicon intake to greater bone density in premenopausal women, and silicon supplementation has been shown to increase bone volume in patients with osteoporosis; the element is also needed to synthesize elastin and collagen, of which the aorta holds the greatest concentration in the body. The International Plant Nutrition Institute has classified silicon as a nutritive element, while the Association of American Plant Food Control Officials calls it a beneficial substance. Workplace exposure carries its own risks: the Occupational Safety and Health Administration caps total silicon exposure at 15 milligrams per cubic meter over an eight-hour workday, and inhaling crystalline silica dust can cause silicosis, a lung disease marked by inflammation and nodular scarring in the lungs' upper lobes.