What Are Glass Windows Made Of and How Are They Made?

Most glass windows are made from silica sand, the same basic material that makes up ordinary beach sand, melted at extremely high temperatures and combined with a handful of other minerals. The standard recipe for flat window glass, known as soda-lime-silica glass, has remained remarkably stable for over a century, though the manufacturing process was revolutionized in the 1950s by the invention of the float glass method. What has changed dramatically is everything that happens after that basic sheet of glass is formed: tempering, laminating, coating, and assembling glass into increasingly sophisticated window systems that manage heat, light, sound, and even their own cleanliness.

What Goes Into the Mix

About 70 to 75 percent of a typical window pane is silicon dioxide, or silica. This comes from high-purity silica sand, which is essentially quartz ground down by geological time. Pure silica would make excellent glass, but it melts at roughly 1,700 °C, which is impractically hot for large-scale manufacturing. To bring the melting point down to a more manageable range of around 1,500 °C, manufacturers add soda ash (sodium carbonate), which typically makes up about 12 to 15 percent of the batch. The trade-off is that soda makes the glass water-soluble, so limestone (calcium carbonate) goes in at around 10 percent to stabilize the structure and make it durable enough to sit in a wall for decades without dissolving in the rain.

Small amounts of other ingredients round out the formula. Magnesium oxide and aluminum oxide improve chemical durability and scratch resistance. A dash of iron oxide is almost always present as a natural impurity in the sand, and it is what gives ordinary window glass its faint green tint when you look at it edge-on. For specialty glass, manufacturers might add boron oxide for heat resistance (borosilicate glass, the kind used in lab equipment) or lead oxide for brilliance (crystal glass), but the vast majority of flat glass in buildings is plain soda-lime-silica.

At the atomic level, what makes glass different from crystalline materials like quartz is that its atoms never settle into a neat, repeating lattice. When molten glass cools, the silicon and oxygen atoms form a disordered network, with each silicon atom bonded to four oxygen atoms in a tetrahedral shape but without any long-range order. Neutron scattering studies have shown that the local atomic arrangement in amorphous silica closely resembles certain high-temperature crystalline phases of silica, but with none of the periodic repetition that defines a crystal.

The Float Glass Process

Nearly all flat glass made today is produced by the float process, invented by Alastair Pilkington in the late 1950s. Before that, flat glass was made by drawing molten glass upward through rollers or grinding and polishing cast sheets, methods that were expensive and produced glass with visible distortions. The float process changed everything by using gravity and surface tension to create glass that is naturally flat and smooth on both sides.

The process begins in a furnace where the raw materials are melted together at about 1,500 °C into a homogeneous molten mass. This melt then flows continuously onto a bath of molten tin, which is kept at around 1,000 °C in a sealed chamber filled with nitrogen and hydrogen to prevent the tin from oxidizing. Because molten glass is less dense than molten tin, the glass floats on top and spreads out into a thin, perfectly flat ribbon. The thickness of the ribbon is controlled by the speed at which it is pulled across the tin bath and by mechanical devices called top rollers that stretch or compress the glass. Standard window glass runs about 4 to 6 millimeters thick, but the process can produce sheets as thin as 0.5 millimeters or as thick as 25 millimeters.

The tin bath is not just a passive surface. The flow patterns and heat transfer within the molten tin directly affect the optical quality of the finished glass. Research into the fluid dynamics of the tin bath has shown that convective currents in the tin can introduce subtle thickness variations in the glass ribbon, which is why manufacturers carefully control the temperature profile across the entire bath.¹ One side of every sheet of float glass has trace amounts of tin embedded in its surface from resting on the bath. This “tin side” is usually invisible to the naked eye but matters when applying certain coatings later on.

Annealing and the Problem of Internal Stress

Glass that cools too quickly develops internal stresses that make it fragile and prone to cracking unpredictably. After leaving the tin bath at around 600 °C, the glass ribbon enters a long, tunnel-like oven called a lehr, where it is cooled in a carefully controlled sequence over the course of several hundred meters. This process is called annealing, and its entire purpose is to let the glass relax into a stress-free state.

The physics of why this matters involves two different kinds of stress. As glass cools through its transition range (roughly 550 to 500 °C for soda-lime glass), the material shifts from a liquid-like state where atoms can rearrange themselves to a rigid state where they are locked in place. If the outside of the glass cools faster than the inside, the outer layers contract first and end up in tension once the interior catches up. Laboratory simulations of industrial annealing have found that about 40 percent of the residual stress in annealed glass comes from volume relaxation effects during cooling, while the remaining 60 percent comes from the thermal gradients that develop as the glass cools further.² Modern annealing lehrs use precisely calibrated heating zones and cooling rates to minimize both types, producing glass that can be cut, drilled, and handled without spontaneous fracture.³

Once annealed, the continuous glass ribbon is inspected for defects using automated optical scanners, then cut into large stock sheets, typically around 3 by 6 meters. These sheets are the starting point for all the processing that turns plain glass into a finished window product.

Tempered Glass and Why It Shatters Into Cubes

Tempered glass, sometimes called toughened glass, is ordinary annealed glass that has been reheated and then rapidly cooled to create a specific stress profile. The glass is heated to above its transition temperature (about 620 °C for soda-lime glass) and then blasted with jets of cold air. The surface solidifies quickly while the interior remains soft and continues to cool and contract afterward. This sets up a permanent state where the surface is in compression and the core is in tension, following a roughly parabolic stress distribution through the thickness of the glass.⁴

That compressive skin is what makes tempered glass so much stronger than regular glass. Surface flaws, microscopic scratches and nicks that every piece of glass acquires during handling, are the usual starting points for cracks. In annealed glass, any tensile stress can pry these flaws open. In tempered glass, the surface compression clamps them shut, so the glass can absorb much more force before cracking. Tempered glass is typically four to five times stronger than annealed glass of the same thickness.

The other well-known property of tempered glass is how it breaks. When a crack finally does manage to penetrate through the compressed surface into the tensioned core, the stored energy releases all at once, and the entire pane disintegrates into small, roughly cubic fragments rather than sharp daggers. This is a safety feature: those blunt granules are far less likely to cause serious cuts than the long shards that annealed glass produces. Building codes require tempered glass in locations where people might walk into it or fall against it, like shower doors, sliding patio doors, and glass near floor level.

One important limitation is that tempered glass cannot be cut or drilled after tempering. Any attempt to modify the surface disrupts the stress balance and causes the whole pane to shatter. All cutting, hole-drilling, and edge-finishing must happen before the tempering step.

Laminated Glass and How It Holds Together

Laminated glass takes a completely different approach to safety. Instead of strengthening a single sheet, it bonds two or more sheets of glass together with a transparent polymer interlayer, usually polyvinyl butyral (PVB) or ethylene-vinyl acetate (EVA). The glass layers provide the structural rigidity while the polymer interlayer absorbs impact energy. When laminated glass breaks, the shards stick to the interlayer rather than scattering, which is why it has been the standard for car windshields since the 1930s.⁵

In building windows, laminated glass serves multiple roles beyond impact resistance. It is the standard choice for hurricane-rated windows and storefronts in high-crime areas because even after the glass layers crack, the interlayer keeps the opening sealed against wind, water, and intrusion. The polymer layer also blocks most ultraviolet radiation, which helps protect furniture and artwork from fading. And because the flexible interlayer dampens vibration transmission, laminated glass provides better sound insulation than a single sheet of the same total thickness.

You can combine laminating with tempering. A common configuration for high-security windows uses two sheets of tempered glass bonded to a thick interlayer. This gives you both the strength of tempered glass and the post-breakage holding power of lamination.

Low-Emissivity Coatings

The biggest revolution in window performance over the past few decades has not been in the glass itself but in what gets deposited on its surface. Low-emissivity (low-E) coatings are microscopically thin metallic layers, usually silver, sandwiched between layers of metal oxide, that allow visible light through while reflecting infrared radiation. The practical effect is that heat energy bounces off the coated surface instead of passing through the glass.

In cold climates, the low-E coating is placed on the interior surface of the outer glass pane, facing the air gap. It reflects radiant heat back into the room, reducing heat loss. In hot climates, the coating goes on the exterior surface of the inner pane, reflecting solar heat back outside. The choice of coating position and the specific formulation let manufacturers tune windows for different climates.

Modern low-E coatings achieve remarkable optical performance. Research into optimized five-layer coating stacks (alternating layers of dielectric material and silver) has demonstrated visible light transmittance of about 72 percent with a neutral color appearance, meaning the glass does not look obviously tinted.⁶ That is a slight reduction from the roughly 90 percent transmittance of uncoated clear glass, but the energy savings more than compensate. Pairing low-E coatings with anti-reflective treatments can recover much of that lost light. One study found that applying anti-reflective coatings to both panes of a low-E double-glazed window boosted light transmittance by about 15 percentage points, from 74 to 89 percent, while leaving the low-E properties intact. The higher solar transmittance also reduced annual heating demand by roughly 4 percent in the simulated conditions.⁷

Insulated Glass Units

A single pane of glass, even with a low-E coating, is a poor insulator. The standard solution is the insulated glass unit (IGU), which sandwiches a sealed air space between two or three panes of glass. The air gap interrupts conductive and convective heat transfer, and filling that gap with a gas that conducts heat less readily than air improves performance further.

Argon is the most common fill gas because it is inexpensive, nontoxic, and about 34 percent less thermally conductive than air. Krypton conducts even less heat and is used when a thinner gap is needed, as in triple-glazed units where space is tight. Xenon is the best insulator of the three but is expensive enough that it rarely appears outside specialized applications. Research has also investigated vacuum-insulated glass, where the gap between panes is evacuated entirely, eliminating both conduction and convection through the gas layer.⁸ Vacuum glazing allows much thinner profiles than gas-filled units, which matters for retrofitting older buildings with frames designed for single panes.

The edge seal of an IGU is a critical weak point. A typical unit uses a metal spacer bar around its perimeter, filled with desiccant to absorb any moisture, and sealed with layers of butyl and polysulfide or silicone. If the seal fails, moisture enters the gap and the window fogs permanently between the panes, a common complaint with aging double-glazed windows. Seal failures typically begin appearing after 15 to 20 years, though quality varies enormously by manufacturer.

Tinted Glass and How Color Gets In

There are two ways to color window glass. Body tinting adds metal oxide colorants directly to the glass melt so the color runs all the way through the thickness of the sheet. Iron oxide produces green or blue-green tints, cobalt oxide gives a deep blue, and selenium compounds create bronze or gray tones. The tint absorbs a portion of incoming solar energy, reducing glare and heat gain. The thicker the glass, the darker it appears, which is why tinted glass is often specified at 6 millimeters or more.

The alternative is applying a thin film to the surface after the glass is formed. Metallic or ceramic coatings sprayed or sputtered onto glass can achieve more precise control over which wavelengths are transmitted, reflected, or absorbed. Early solar-control coatings used thin films of metals like silver, copper, and gold to create reflective surfaces that bounced back solar energy while still transmitting some visible light.⁹ Modern versions have moved toward multi-layer ceramic coatings that are less mirror-like and more color-neutral.

Body-tinted glass has one disadvantage worth knowing about: because it absorbs solar energy rather than reflecting it, the glass itself heats up significantly in direct sunlight. This thermal stress can cause cracking if the glass is partially shaded, because the hot zone expands while the shaded zone does not. Tempered tinted glass handles this better, but the risk is one reason that reflective or low-E coatings are generally preferred over heavy body tints for large commercial facades.

Electrochromic Smart Glass

Electrochromic glass can switch between transparent and tinted states with the application of a small voltage. The technology typically uses a thin stack of coatings, including an electrochromic layer (often tungsten oxide), an ion-conducting electrolyte, and a counter electrode. When voltage is applied, ions migrate into the electrochromic layer, causing it to absorb visible light and darken. Reversing the voltage drives the ions back out and the glass clears.¹⁰

Early electrochromic windows were slow to switch, often taking several minutes to go from clear to dark. That sluggishness limited their appeal, but recent advances have dramatically compressed the timeline. One laboratory demonstration achieved coloring in 1.6 seconds and bleaching in 0.8 seconds, with a transmittance contrast of over 64 percent between the clear and dark states.¹¹ That kind of speed would make the glass responsive enough to track passing clouds or adjust to changing sun angles throughout the day.

Electrochromic windows are already installed in some commercial buildings and high-end residential projects, where they replace motorized blinds or shades. They offer the advantage of controlling heat and glare without blocking the view, since even in their darkest state most designs still transmit some light. The main barriers to wider adoption are cost, which remains several times higher than conventional low-E insulated glass, and durability over decades of daily cycling.

Self-Cleaning Glass

Self-cleaning glass relies on a thin coating of titanium dioxide (TiO₂) on the exterior surface. Titanium dioxide is a photocatalyst: when exposed to ultraviolet light from the sun, it generates reactive molecules on its surface that break down organic dirt and grime. The coating also becomes highly hydrophilic under UV exposure, meaning water spreads into a flat sheet across the surface rather than forming droplets.¹² When it rains, this sheeting action washes away the loosened dirt without leaving streaks or water spots.

Self-cleaning glass works best in locations that get regular rainfall and reasonable sun exposure. In very dry climates, you still need to hose it down occasionally, and the photocatalytic effect cannot remove inorganic deposits like mineral scale from hard water or heavy construction dust. The coating is permanent and built into the manufacturing process, so it does not wear off with normal cleaning. It does, however, add a modest cost premium, and the glass has a faint blue or gray hue that is visible when held up against uncoated glass.

Sound Insulation Through Glass

Windows are typically the weakest link in a building’s sound barrier. A single pane of glass transmits noise readily because it is thin and stiff, vibrating easily at frequencies that match common environmental sounds like traffic. Thicker glass helps, but mass alone has limits. Laminated glass outperforms monolithic glass of the same weight because the polymer interlayer converts vibration energy into heat, damping the transmission. Using panes of different thicknesses in an IGU also helps, because the two panes resonate at different frequencies, avoiding the “coincidence dip” where a single thickness becomes acoustically transparent.

Research into metamaterial-based window designs has explored adding engineered structures to the glass that target specific frequency ranges. These designs use localized resonance units to dissipate acoustic and vibrational energy, broadening the range of frequencies the window can block while maintaining light transmission.¹³ This technology is still largely in the research stage, but it points toward a future where windows can be tuned for specific acoustic environments, like airport flight paths or busy urban intersections, without sacrificing natural light.

Decarbonizing the Furnace

Melting glass is energy-intensive. Conventional float glass furnaces burn natural gas around the clock, running continuously for 15 to 20 years before they are shut down for rebuilding. The glass industry has been under growing pressure to reduce its carbon footprint, and several pathways are being explored simultaneously.

A comprehensive techno-economic assessment of low-emission furnace designs found that a hybrid furnace combining oxy-combustion with partial electrification could reduce emissions by about 33 percent compared to a conventional gas-fired furnace. Fully electric furnaces cut emissions by roughly 41 percent, and hydrogen-fired furnaces achieved the largest reduction at about 50 percent by eliminating combustion emissions altogether.¹⁴ The catch with electric and hydrogen furnaces is that their climate benefit depends entirely on the energy source. An electric furnace powered by coal-generated electricity may produce more total emissions than a well-run gas furnace. Hydrogen furnaces require green hydrogen, which is not yet widely available at industrial scale.

Recycled glass, or cullet, is another lever. Every 10 percent of cullet added to the batch reduces energy consumption by roughly 2.5 to 3 percent, because cullet melts at a lower temperature than raw materials. Container glass (bottles and jars) already uses high percentages of cullet, but flat glass recycling is more challenging because the tight optical specifications mean contaminants that would be harmless in a bottle can produce visible defects in a window pane. Sorting and cleaning technologies are improving, and some manufacturers now routinely incorporate 20 to 30 percent cullet into their float lines.