A photomask is a high-precision plate that acts as a stencil for printing circuit patterns onto silicon wafers during chip manufacturing. It contains a carefully arranged pattern of transparent and opaque areas that light passes through (or gets blocked by) to create the microscopic circuitry inside every modern processor, memory chip, and sensor. If you think of semiconductor manufacturing as a kind of photography, the photomask is the negative.
How a Photomask Works
Chip manufacturing uses a process called photolithography, which literally means “writing with light.” A photomask sits between a light source and a silicon wafer that has been coated with a light-sensitive material. When light shines through the mask, the pattern of transparent and opaque areas casts a precise shadow onto the wafer below. Wherever light hits the coating, it triggers a chemical change that allows those areas to be etched away (or preserved, depending on the process), leaving behind the intended circuit pattern.
Modern lithography systems don’t project the mask pattern at its actual size. They use a 4x or 5x reduction ratio, meaning the features on the mask are four or five times larger than what ends up on the wafer. This is a critical advantage: it’s far easier to fabricate an accurate pattern at a larger scale, and the optical system shrinks it down to the nanometer dimensions needed on the final chip. A single chip may require dozens of separate masks, each defining a different layer of the circuit.
What Photomasks Are Made Of
The transparent substrate is typically made of fused quartz (also called fused silica), chosen because it transmits ultraviolet light with very little distortion and resists thermal expansion during the intense exposure process. On top of this quartz plate sits a thin layer of chromium, usually around 60 nanometers thick, which serves as the opaque material that blocks light. The chromium is etched away in precisely the right places to form the circuit pattern, while the remaining chromium defines the areas that stay dark.
The result is something that looks, at a glance, like a square glass plate with faint geometric patterns on it. But the precision involved is extraordinary. Even a tiny speck of dust or a scratch on the mask surface can ruin the pattern on every chip printed from it, which is why photomasks are manufactured and handled in ultra-clean environments.
Types of Photomasks
The simplest and most traditional type is the binary mask (sometimes called a binary intensity mask). It works on a straightforward principle: each spot on the mask either blocks light completely or lets it through. Chromium stays where you want darkness, quartz is exposed where you want light. This binary, on-or-off approach works well for larger features but starts to lose sharpness at extremely small dimensions because of the way light bends around edges.
Phase-shift masks solve this problem by manipulating not just whether light passes through, but how the light waves interact with each other. By adding a thin film (often a molybdenum silicide compound) that shifts the phase of light passing through certain areas, these masks cause neighboring light waves to cancel each other out at their boundaries. The result is sharper edges and better contrast in the final pattern. Phase-shift masks are more expensive and complex to manufacture, but they’re essential for printing the smallest features on advanced chips.
The two types also behave differently when fine-tuned. Phase-shift masks tend to produce higher image contrast when the pattern is made slightly smaller than the target, while binary masks perform better when the pattern is slightly larger. Engineers choose between them based on the specific feature sizes and performance requirements of each chip layer.
How Photomasks Are Made
Creating a photomask starts with a blank: a polished quartz plate coated with chromium and a layer of light-sensitive resist. The circuit pattern is then written onto this blank using one of two methods.
Electron-beam lithography is the primary technique for the most advanced masks. Instead of using light, a focused beam of electrons traces the pattern directly onto the resist, point by point. This is slow compared to optical methods (writing a single complex mask can take many hours), but it achieves the extreme precision required for cutting-edge chips. Think of it as the difference between printing a poster on a commercial press versus drawing it by hand with a fine-tipped pen. The electron beam approach sacrifices speed for resolution.
Laser writing is faster and used for masks where the feature sizes are somewhat larger. A focused laser beam scans across the resist to define the pattern. After the pattern is written by either method, the resist is developed (similar to developing a photograph), and the exposed chromium is chemically etched away. What remains is the finished mask pattern.
Defect Inspection and Repair
Because a single photomask may be used to print patterns on thousands of wafers, any defect on the mask gets replicated across every chip. This makes quality control extremely rigorous. Finished masks are scanned by automated inspection systems that compare the actual pattern against the intended design, flagging any spots where material is missing or where unwanted particles or residue have landed.
When defects are found, they can often be repaired rather than scrapping the entire mask. One modern approach uses femtosecond laser pulses (bursts of light lasting just quadrillionths of a second) to carefully remove unwanted material without damaging the underlying quartz substrate. Repairs are then verified using specialized imaging tools that simulate how the mask will actually perform during chip printing, along with cross-sectional analysis to confirm no hidden damage remains. For the most advanced masks used in extreme ultraviolet (EUV) lithography, both the inspection and repair processes have become significantly more complex and expensive.
Why Photomasks Matter to the Chip Industry
Photomasks represent a substantial and growing segment of the semiconductor supply chain. The global photomask market is valued at roughly $6.35 billion in 2026 and is projected to reach $7.92 billion by 2031, growing at about 4.5% per year. That growth tracks directly with the push toward smaller, more complex chip designs, which require more mask layers per chip and tighter manufacturing tolerances on each one.
A single set of photomasks for an advanced chip can cost millions of dollars, making them one of the most expensive consumables in semiconductor manufacturing. This cost is justified because the mask is the master template: get it right once, and you can stamp out millions of identical chips. Get it wrong, and every wafer printed from that mask produces defective parts. That combination of extreme precision, high stakes, and massive reuse is what makes the photomask one of the most critical components in modern electronics manufacturing.