Silicon is made from sand through a high-temperature chemical reaction that strips oxygen atoms away from silicon dioxide, the compound that makes up most sand and quartz. The core process is straightforward in concept: heat sand with carbon to extreme temperatures, and the carbon pulls the oxygen off, leaving behind elemental silicon. In practice, this requires industrial-scale furnaces running above 1,600°C and careful control of raw materials, energy, and impurities.
Why Sand Is the Starting Point
Sand is mostly silicon dioxide (SiO₂), a molecule where one silicon atom is locked to two oxygen atoms. Silicon is the second most abundant element in Earth’s crust, but it almost never exists in pure form. It’s always bonded to oxygen in rocks, quartz, and sand. The goal of the entire production process is breaking that bond.
Not just any beach sand works. Industrial silicon production uses high-purity quartz with at least 99% silicon dioxide content. For semiconductor applications, producers need high-purity quartz containing 99.99% SiO₂ or higher. The raw quartz is mined, crushed, and sorted to remove visible contaminants before it enters the furnace.
The Carbothermic Reduction Reaction
The core chemistry is called carbothermic reduction. Carbon acts as the reducing agent, meaning it reacts with the oxygen in silicon dioxide and carries it away as carbon monoxide gas. The simplified equation looks like this:
SiO₂ + 2C → Si + 2CO
In plain terms: silicon dioxide plus carbon yields pure silicon plus carbon monoxide gas. The carbon sources used in industry are a mix of charcoal, coal, petroleum coke, and wood chips. Each carbon source has different properties that affect how evenly the reaction proceeds inside the furnace, so producers blend them in specific ratios.
The real chemistry inside the furnace is far more complex than that single equation suggests. Intermediate compounds form along the way, including silicon monoxide gas and silicon carbide. These intermediates cycle through their own reactions at different temperature zones within the furnace before ultimately producing liquid silicon that collects at the bottom.
Inside the Electric Arc Furnace
The reaction takes place in a submerged arc furnace, a massive vessel lined with heat-resistant materials. Three large graphite electrodes plunge into the mix of quartz and carbon from above, and an enormous electrical current arcs between them. This arc generates the heat needed to drive the reaction.
The furnace operates in distinct temperature zones. The upper region, where raw materials are loaded, stays cooler as incoming charge material absorbs heat. The lower zone near the electrodes reaches 1,600 to 2,000°C, and the arc zone itself can exceed 2,000°C. Liquid silicon, which melts at 1,414°C, pools at the bottom of the furnace and is periodically tapped off, meaning it flows out through a hole into ladles for casting.
Most plants consume 11 to 13 kilowatt-hours of electrical energy per kilogram of silicon produced. That’s a significant energy demand. A single furnace can run continuously for months or even years between rebuilds, with workers feeding fresh quartz and carbon into the top while tapping molten silicon from the bottom.
What Comes Out: Metallurgical-Grade Silicon
The silicon that flows out of the arc furnace is called metallurgical-grade silicon, and it’s roughly 99% pure. That 1% of impurities includes iron, aluminum, calcium, and other elements carried in by the raw materials. For making aluminum alloys or silicone chemicals, this purity level is perfectly adequate.
For electronics or solar cells, 99% is nowhere near pure enough. Semiconductor-grade silicon requires what the industry calls “six nines” purity: 99.9999% or higher. The most advanced chips need nine-nines purity (99.9999999%), where impurity concentrations drop to roughly one atom per billion. Getting from 99% to that level requires an entirely separate set of refining steps.
Purifying Silicon for Electronics
The dominant purification method is the Siemens process, which converts crude silicon into a gas, purifies the gas, then converts it back into solid silicon. It works in three stages.
First, metallurgical-grade silicon reacts with hydrochloric acid to form a compound called trichlorosilane, a liquid that boils at just 32°C. This conversion step leaves most impurities behind because they form different compounds that separate out. Second, the trichlorosilane is distilled repeatedly, the same basic principle behind distilling alcohol but with tighter control. Each distillation pass removes more trace contaminants.
Third, the purified trichlorosilane gas is fed into a sealed chamber containing thin silicon rods heated to around 1,000 to 1,175°C. In the presence of hydrogen gas, the trichlorosilane breaks down and deposits pure silicon atoms onto those heated rods, building them up into thick cylinders over time. This chemical vapor deposition step produces polycrystalline silicon of extraordinary purity.
The Siemens process is energy-intensive. A newer alternative, the fluidized bed reactor method, uses 80 to 90% less electrical energy and runs continuously rather than in batches. It produces small spherical granules of silicon instead of large rods, which are actually easier to handle in downstream processing. Both methods achieve the purity needed for solar cells and electronics.
Growing a Single Crystal
Purified polycrystalline silicon still isn’t ready for computer chips. The silicon atoms are arranged in small, randomly oriented crystal grains. Chip manufacturing requires monocrystalline silicon, a single continuous crystal lattice with no grain boundaries.
The Czochralski method is how most of the world’s single-crystal silicon is made. The purified polysilicon is loaded into a cylindrical crucible and melted. A technician dips a tiny seed crystal, just a few millimeters across, into the surface of the molten silicon. A small portion of the seed melts to form a clean contact with the liquid. Then the seed is slowly pulled upward while rotating.
As the seed rises, molten silicon crystallizes onto its bottom surface, extending the single-crystal structure atom by atom. The pulling speed, rotation rate, and heating power are carefully adjusted to control the diameter of the growing crystal. Over the course of many hours, what started as a tiny seed becomes a cylindrical ingot that can weigh over 100 kilograms. This ingot is then sliced into the thin wafers that become solar cells or semiconductor chips.
Environmental Cost of Production
Because the process fundamentally relies on carbon to strip oxygen from silicon dioxide, CO₂ emissions are baked into the chemistry. Producing one ton of metallurgical-grade silicon releases roughly 5.9 tons of CO₂ from the chemical reaction alone, plus another 0.9 tons of biogenic CO₂ from charcoal and wood-based carbon sources. Additional byproducts include nitrogen oxides, sulfur oxides, and fine particulate matter.
One useful byproduct does emerge: microsilica, the ultra-fine dust captured from furnace exhaust. About 270 kilograms of microsilica are produced per ton of silicon. This material is widely used as an additive in high-performance concrete, where it fills gaps between cement particles and dramatically increases strength and durability.
Because carbon is used for its chemical properties rather than its energy content, there’s no way to reduce these process emissions below stoichiometric levels without replacing carbon entirely as the reducing agent. Some researchers are exploring aluminum-based reduction and other alternatives, but carbon-based smelting remains the industry standard.
From Beach Sand to Microchip: The Full Chain
The complete journey from raw quartz to a finished silicon wafer passes through at least four major stages, each one increasing purity by orders of magnitude. Quartz mining and preparation yields material that is 99%+ SiO₂. Carbothermic reduction in an arc furnace produces 99% pure metallurgical-grade silicon. Chemical purification through the Siemens process or fluidized bed reactors reaches 99.9999% or higher. Crystal growth via the Czochralski method transforms that purified silicon into a single perfect crystal ready for wafer slicing.
Each stage requires specialized equipment, enormous energy inputs, and precise process control. The simplicity of the underlying chemistry, carbon pulls oxygen off silicon, belies the engineering complexity of doing it at scale and at the purity levels modern technology demands.