Walk into any semiconductor fabrication facility on Earth — from TSMC's gigafabs in Tainan to Intel's operations in Chandler, Arizona, to Samsung's complex in Pyeongtaek, South Korea — and you will find one material present at virtually every critical process step. Not silicon, though that is obviously there too. Not the exotic photoresists or the ultra-pure gases. Quartz. Specifically, high purity quartz that has been processed to 99.99% SiO2 or higher.
It is one of the great hidden dependencies in modern technology. The semiconductor industry hit $527 billion in global revenue in 2023 and is widely projected to surpass $1 trillion by 2030. Every dollar of that revenue depends, at some point in the manufacturing chain, on components made from high purity quartz. Yet outside of the materials science community and the procurement departments of major chipmakers, almost nobody talks about it.
That is starting to change, and for good reason.
Where Quartz Shows Up in the Fab
To understand why high purity quartz is so indispensable, you need to understand where it appears in the semiconductor manufacturing process. It is not a single application — it is woven through the entire production flow.
Czochralski (CZ) crystal growing crucibles. This is the big one. The Czochralski process is how the vast majority of single-crystal silicon ingots are produced. A seed crystal is dipped into molten polysilicon held at approximately 1,425 degrees Celsius and slowly pulled upward, forming a cylindrical ingot that will eventually be sliced into wafers. The container holding that molten silicon is a quartz crucible, typically 800mm to 1,000mm in diameter for 300mm wafer production. These crucibles are consumable — each one is used for a single pull and then discarded, because the extreme temperatures and chemical interactions with molten silicon degrade the quartz structure. A single large fab can go through 300 to 500 crucibles per year. Each crucible requires 50 to 100 kilograms of high purity quartz feedstock. Multiply those numbers across the dozens of crystal growing operations worldwide and you start to see the scale of demand.
300–500 crucibles per year, per fab.
Each CZ crucible requires 50–100 kg of HPQ and is discarded after a single crystal pull. This is the largest single driver of quartz consumption in semiconductors.
Diffusion furnace tubes. After wafers are cut and polished, many of the key processing steps — oxidation, doping, annealing — take place inside horizontal or vertical tube furnaces. The tubes themselves are made from fused quartz, because the material can withstand sustained temperatures above 1,100 degrees Celsius while introducing essentially zero contamination into the process environment. A single furnace tube can weigh 30 to 80 kilograms and typically needs to be replaced every few hundred process cycles as thermal cycling causes devitrification and micro-cracking.
Wafer carriers and boats. Inside those furnace tubes, the wafers need to be held in place. Quartz wafer carriers (sometimes called boats or cassettes) hold 25 to 50 wafers in precise slots, maintaining uniform spacing so that gases flow evenly across every wafer surface. These carriers are machined from solid fused quartz and, like everything else in a fab, are eventually consumed through use and replaced.
Photomask substrates. At the most advanced process nodes, the photomasks that define circuit patterns are fabricated on ultra-flat fused quartz substrates. ASML's EUV lithography systems use reflective masks on specialized substrates, but for DUV lithography — which still handles the majority of layers in a modern chip — transparent fused quartz remains the standard substrate material. The purity requirements here are extraordinary, because any impurity in the substrate can distort the optical properties and ruin the pattern transfer.
Etch chamber components. Plasma etch processes use quartz components — windows, liners, shower heads, and focus rings — inside the process chambers. These parts are exposed to aggressive fluorine-based chemistries and high-energy plasmas, which gradually erode them. They are consumable items that need regular replacement, and they must be made from high purity quartz to avoid introducing contaminants that would kill device yields.
The Purity Problem: Why Ordinary Quartz Will Not Cut It
There is plenty of quartz on Earth. It is one of the most abundant minerals in the crust. So why is the semiconductor industry willing to pay premium prices for material from a handful of sources? Because semiconductor manufacturing is, at its core, a contamination control exercise, and quartz purity matters enormously.
When a CZ crucible contains trace amounts of metals — iron, aluminum, sodium, potassium, lithium, calcium — those metals can migrate into the molten silicon during the crystal pull. Even parts-per-billion levels of certain contaminants can affect the electrical properties of the resulting wafers, leading to yield loss or reliability problems in finished chips. The tighter the design rules, the less contamination can be tolerated.
This is where the grading system becomes critical. Quartz graded at 4N (99.99% SiO2) has total metallic impurities below 100 parts per million. This grade is widely used for the outer layers of CZ crucibles, which provide structural integrity. It also serves as the feedstock that specialized processors further purify to 5N (99.999%) or higher for the inner crucible layers that actually contact the molten silicon. Think of 4N quartz as the critical entry point — without a reliable supply of high-quality 4N material, the entire downstream purification chain breaks down.
4N quartz is the foundation.
It feeds the outer crucible layers directly and serves as feedstock for further purification to 5N+ grades. No 4N supply means no crucibles, period.
The CHIPS Act Effect: A Tidal Wave of New Demand
If the existing demand picture was not enough, the policy environment is about to make it significantly more intense. The United States CHIPS and Science Act committed $52 billion to domestic semiconductor manufacturing, and the results are visible on the ground. TSMC is building three fabs in Phoenix, Arizona. Intel is expanding in Ohio, Arizona, and New Mexico. Samsung is investing $17 billion in Taylor, Texas. Micron is building a $100 billion complex in Clay, New York over the next two decades.
The European Chips Act has mobilized an estimated 43 billion euros in public and private investment to boost the continent's semiconductor production. Intel is building a massive fab complex in Magdeburg, Germany. TSMC has a facility planned in Dresden. STMicroelectronics and GlobalFoundries are expanding in France.
Japan has lured TSMC to Kumamoto for a fab that is already operational, with a second one under construction. South Korea's government has backed a $230 billion semiconductor investment plan. India has committed over $10 billion and has fab projects from Tata and others in the pipeline.
Add it all up and you are looking at roughly 40 to 60 new semiconductor fabs expected to begin operations between 2024 and 2030. Each of those fabs will need a steady supply of quartz crucibles, furnace tubes, wafer carriers, etch components, and more. The incremental demand for high purity quartz from this construction wave alone is measured in tens of thousands of metric tons per year.
AI and Advanced Nodes: More Steps, More Quartz
There is another dynamic at play that compounds the demand story. The artificial intelligence revolution is not just driving more chip volume — it is driving more complex chips that consume more quartz per unit.
An advanced AI accelerator like NVIDIA's Blackwell or AMD's Instinct MI300 series contains billions of transistors manufactured at 3nm, 4nm, or 5nm process nodes. These advanced nodes require more lithography layers, more etch steps, more deposition cycles, and more thermal processing steps than the mature nodes used for commodity chips. More process steps mean more time in furnace tubes, more wafer carrier cycles, more etch chamber component wear. The quartz consumption per wafer pass increases as the technology advances.
Looking ahead, TSMC's 2nm process (N2) and Intel's 18A node will push this even further. Gate-all-around (GAA) transistor architectures require additional processing steps compared to FinFET, and each additional step touches quartz components somewhere in the process flow. The industry is moving toward a future where the quartz intensity per chip is going up, not down, even as engineers find efficiencies elsewhere.
Advanced nodes = more quartz per wafer.
3nm and 2nm processes require more lithography layers, more etch steps, and more thermal cycles. Each additional step wears through quartz components faster.
The Companies That Make It Happen
The quartz component supply chain for semiconductors is dominated by a small number of specialized manufacturers, each of which depends on a reliable flow of high purity quartz feedstock.
Heraeus (Germany) is one of the world's leading producers of fused quartz and quartz glass products for the semiconductor industry. Their Conamic division manufactures crucibles, tubes, and components used in fabs globally. Shin-Etsu Quartz Products (Japan) is a major supplier of CZ crucibles and quartz components, closely tied to Shin-Etsu Chemical's dominant position in silicon wafer manufacturing. Tosoh SGM (Japan) specializes in high-purity quartz glass for semiconductor applications. Momentive Performance Materials (USA) produces fused quartz products under legacy brands that have served the industry for decades.
All of these companies share a common challenge: securing enough high purity quartz feedstock to meet rapidly growing demand. They source from the same limited pool of high-grade natural quartz deposits, and they are all feeling the squeeze as downstream demand accelerates.
The Supply Chain Vulnerability That Keeps Chipmakers Up at Night
The semiconductor industry has spent the past five years obsessing over supply chain resilience. The COVID-era chip shortage, the geopolitical tensions around Taiwan, the concentration of HPQ supply in Spruce Pine, North Carolina — all of these have forced a reckoning with how fragile the technology supply chain really is.
The Spruce Pine disruption in late 2024 was particularly instructive. When Hurricane Helene cut road access to the mines that produce much of the world's highest-grade quartz, the shock rippled through the entire supply chain within days. Crucible manufacturers began rationing allocations. Spot prices spiked. Crystal growing operations started auditing their inventory buffers with newfound urgency. The disruption was ultimately short-lived, but the lesson was clear: a supply chain that depends on a single geographic point of failure is a supply chain waiting to break.
This is why semiconductor companies, crucible manufacturers, and their procurement teams are actively seeking diversified sources of high purity quartz. The search is not just about finding more material — it is about finding material from different geologies, different countries, and different risk profiles.
Why Diversified Sourcing Is No Longer Optional
The math is straightforward. If you are building a new fab that will cost $15 to $20 billion and take four years to construct, you cannot afford to have your crystal growing operations held hostage by a supply disruption affecting a single quartz source. The cost of the quartz itself is a rounding error compared to the cost of an idle fab. But without that quartz, the fab does not produce wafers, and without wafers, the entire investment sits idle.
This reality is driving a structural shift in how semiconductor supply chains approach quartz procurement. Multi-sourcing strategies that were considered unnecessary five years ago are now becoming standard. Companies are qualifying alternative quartz sources, running extensive purity testing and process validation trials, and building longer-term supply agreements with producers outside the traditional Spruce Pine and Norway axis.
Origins like Sri Lanka, with its naturally high-purity vein quartz deposits, are attracting attention precisely because they offer geological diversity. Sri Lankan vein quartz forms through hydrothermal processes that produce material with inherently low metallic impurities, making it well-suited as feedstock for the crucible supply chain. The deposits are substantial — operations like Quartz.lk are developing reserves exceeding one million metric tons — and the geographic position offers efficient shipping access to growth markets across Asia.
Looking Ahead: The Quartz Bottleneck Is Real
The semiconductor industry is heading into a period of unprecedented expansion. Forty to sixty new fabs. A trillion-dollar revenue target. AI chips demanding more process steps per wafer. Advanced nodes requiring tighter contamination control. Government subsidies removing the capital barriers to new construction.
All of this points in one direction for high purity quartz demand: sharply upward. And the supply side has not yet demonstrated that it can keep pace. New quartz processing capacity takes time to build and qualify. Mining permits take years. Purity validation with end customers is a rigorous, multi-stage process that cannot be rushed.
The companies that will navigate this successfully — whether they are chipmakers, crucible manufacturers, or quartz producers — are the ones that are making supply chain decisions today based on where the market will be in 2028 and 2030, not where it is right now. Waiting to see how the demand curve develops before locking in diversified quartz supply is a strategy that has a name. It is called being too late.
The semiconductor industry was built on silicon. But silicon starts its journey in a quartz crucible, processed in quartz furnace tubes, carried on quartz boats, and patterned through quartz photomask substrates. Take away the quartz and nothing else in the fab matters. That is the fundamental truth that is finally getting the recognition it deserves.
