Semiconductor Photoresists Market

A few litres of resist decide whether a wafer worth hundreds of thousands of dollars yields or scraps, which is why qualification takes years and switching almost never happens. Douglas Insights values the semiconductor photoresists market at USD 3.33 billion in 2025 and expects USD 7.87 billion by 2035, compounding at 8.99%. The estimate is built from the bottom up: roughly 6.4 million litres of photoresists and associated lithography materials consumed in 2025 across ArF immersion and ArF dry photoresists, KrF photoresists, extreme ultraviolet photoresists, and g-line, i-line and ancillary materials, at an average realised price of USD 520 per litre, triangulated against wafer starts by node, lithography layer counts and supplier disclosures. Volume grows 5.2% a year as wafer output and lithography layers increase, while realised price rises 3.6% a year as extreme ultraviolet and advanced ArF resists take share. This study sits within our semiconductor materials coverage and follows the published Douglas Insights methodology.

The event this edition is anchored on is JSR’s announcement on 10 April 2026 that it had set up JSR Micro Advanced Manufacturing Taiwan Co., Ltd. with LCY Chemical and Wah Lee Industrial, a joint venture that will serve as a Taiwan manufacturing base for advanced electronic materials and, in JSR’s words, “establish a stable supply system for the growing advanced semiconductor market” (JSR news release, 10 April 2026). A Japanese resist leader building in Taiwan, home of most leading-edge foundry output, captures the three forces in this report: leading-edge wafer growth, qualification lock-in and the push to localise supply. The drivers, supplier and scenario sections below return to it.

Why does a handful of Japanese companies control the semiconductor photoresists market?

Japanese chemical makers spent decades building the formulation, purity control and chipmaker qualifications that photoresists demand, and rivals did not. Because a qualified resist is rarely replaced once it sits in a production flow, that head start compounds: JSR, Tokyo Ohka Kogyo, Shin-Etsu Chemical, Sumitomo Chemical and Fujifilm hold the large majority of 2025 value, most firmly at the leading edge.

A photoresist is a light-sensitive polymer coated onto a wafer; when exposed through a mask by a lithography tool, its chemistry changes so that patterns can be developed and etched into the chip. At advanced nodes, resists must resolve features only a few nanometres across, with extreme uniformity, minimal defects and precisely controlled sensitivity, and tiny variations in formulation or purity can ruin yields across thousands of wafers. Chipmakers therefore qualify resists through long, expensive processes and are extremely reluctant to switch suppliers once a resist is proven in a production process. The same qualified-supplier discipline now reaches back into research materials, as the Strontium Titanate Single Crystal Substrate Market report shows, where programme buyers have borrowed semiconductor incoming-inspection practice for crystal substrates. That concentration became a geopolitical issue when Japan updated its licensing policies for exports of controlled items to South Korea, including certain photoresists, on 1 July 2019 (METI, update of licensing policies in relation to the Republic of Korea), prompting Korean and later Chinese efforts to build domestic supply. The exclusive chapter of this report matrices resist demand by lithography node, because node transitions decide which resists and suppliers grow.

Which lithography materials count as semiconductor photoresists in this study?

Semiconductor photoresists here means the resists and closely tied lithography materials used in wafer fabrication, worth USD 3.33 billion in 2025: ArF immersion and dry, KrF, extreme ultraviolet, and g-line, i-line and ancillary products. Resists for circuit boards, displays and packaging outside the fab, plus lithography tools and masks, are excluded.

ArF immersion and ArF dry photoresists cover resists for 193 nanometre lithography, the workhorse for many critical layers at advanced and mainstream nodes, including immersion lithography. KrF photoresists cover resists for 248 nanometre lithography used for less critical layers, thick-film applications and mature nodes. Extreme ultraviolet photoresists cover resists for 13.5 nanometre lithography used at the most advanced nodes, including chemically amplified and metal oxide resists. g-line, i-line and ancillaries cover resists for older wavelengths plus the developers, anti-reflective coatings and rinse materials sold alongside resists. Other wet chemicals also sit outside the boundary. The model is cut four ways: by resist category, by node (leading edge below 7nm, advanced 7 to 28nm, mature above 28nm), by chip type (logic and foundry, DRAM, NAND flash, analogue and power) and by six regions. Value is measured at the price chipmakers pay for resist delivered to the fab.

What makes extreme ultraviolet resists so hard?

Extreme ultraviolet photoresists must catch far fewer, far more energetic photons than older resists while still printing smooth edges a few nanometres wide. Balancing sensitivity against roughness and defects is why the segment is the costliest per litre and the fastest growing, at 14.59% a year from USD 599 million in 2025.

Extreme ultraviolet light has a very short wavelength that allows much smaller features to be printed, but each photon carries far more energy, and far fewer photons reach the wafer than in older lithography because generating extreme ultraviolet light is difficult and inefficient. With fewer photons, random statistical variation in how many land in a given spot, known as shot noise, causes rough and inconsistent feature edges and defects. Resists must therefore be highly sensitive to make use of scarce photons, yet also produce precise features and survive the etch steps that follow, competing requirements that chemists struggle to balance. Conventional chemically amplified resists have been pushed hard, and new approaches, including metal oxide resists that absorb extreme ultraviolet more efficiently, are being developed and adopted. The move to high numerical aperture extreme ultraviolet tools, which print even finer features, intensifies these challenges. Success at the leading edge increasingly depends on resist innovation as much as on lithography tools.

What lifts semiconductor photoresist consumption from 6.4 to about 10.6 million litres?

Four forces lift semiconductor photoresist consumption: more wafer starts, more lithography layers per wafer, wider extreme ultraviolet use and new fabs outside Asia. Together they add 5.2% a year to litres consumed, from about 6.4 million in 2025 to about 10.6 million in 2035, while resist mix adds 3.6% a year to price.

Wafer output is the base layer. Rising demand for chips for artificial intelligence accelerators, data centres, vehicles and phones increases wafer starts, and in the model wafer starts multiplied by lithography layers and resist per layer produce the 5.2% volume leg. Each 1-point change in that volume growth moves the 2035 market by roughly USD 700 million. The same advanced-node wafer starts also drive demand for the fab automation covered in our Semiconductor Wafer Handling Robots Market report and the tools in our Semiconductor Test Equipment Market report, while the chips patterned with these resists end up on the substrates covered in our Semiconductor Advanced Packaging Substrates Market report.

Layer count and node advancement come second. Advanced logic and memory need more lithography layers and more multiple-patterning steps, which raises resist use per wafer. ArF immersion and dry resists, the workhorse at 7 to 28nm, carry this effect: the category grows 8.28% a year, from USD 1.40 billion in 2025 to USD 3.10 billion in 2035, almost exactly the market rate, because added layers offset the layers that migrate to extreme ultraviolet.

Extreme ultraviolet adoption is the third and most valuable driver. As extreme ultraviolet spreads to more layers in logic and foundry production below 7nm and into leading DRAM, and as high numerical aperture tools arrive, extreme ultraviolet resists rise from USD 599 million to USD 2.34 billion, taking their share from 18% to about 30%. That segment alone supplies about 38% of the USD 4.54 billion the market adds between 2025 and 2035, and it is the main reason the realised price leg reaches 3.6% a year.

New fab construction is the fourth. Fabs being built in the United States, Europe and Japan under supply chain policies add new resist demand close to the customer, which is why North America is the fastest-growing region at 10.4% a year, from USD 399.4 million in 2025 to USD 1.07 billion in 2035. JSR’s April 2026 Taiwan joint venture shows the same logic in Asia: resist makers are moving manufacturing next to the fabs whose output drives the volume leg.

Which risks could cap semiconductor photoresist consumption?

Three risks are modelled for semiconductor photoresists: chip-cycle downturns, European limits on fluorinated chemistry and supply concentration under export controls. Combined in the downturn scenario, they cut volume growth to 3.0% and price growth to 2.0%, leaving the 2035 market near USD 5.43 billion, about USD 2.44 billion below the base case.

Semiconductor cyclicality is the first risk. Resist demand follows wafer output, so a memory or broader chip downturn cuts litres directly; in the model, every point of volume growth lost to weaker wafer starts removes roughly USD 700 million from the 2035 figure, and the 2.2-point volume cut in the downturn case accounts for most of the USD 2.44 billion gap.

Fluorinated chemistry is the second. Many photoresists rely on photoacid generators and other components that fall under per- and polyfluoroalkyl substances. On 20 August 2025 the European Chemicals Agency published an updated EU-wide restriction proposal after the five national authorities behind it reviewed more than 5,600 consultation comments, and electronics and semiconductors were among the sectors for which alternative, conditional restriction options were assessed (ECHA, 20 August 2025). Reformulating a qualified resist would restart qualification, which is why the downturn scenario also trims the price leg by 1.6 points.

Supply concentration and geopolitics are the third. Dependence on a few suppliers in one country creates supply risk, and the July 2019 Japanese licensing change for exports to South Korea showed how quickly trade policy can interrupt deliveries. Efforts to localise production face multi-year qualification times, so the model phases share gains in slowly rather than as a step change.

Which resist categories carry the value?

ArF immersion and dry photoresists carry the most semiconductor photoresist value, 42% or USD 1.40 billion in 2025, while extreme ultraviolet resists grow fastest at 14.59% a year. KrF holds 22%, and older wavelengths with ancillaries hold 18%, matching extreme ultraviolet’s 2025 value of USD 599 million.

KrF photoresists, at USD 732.2 million, are used widely for less critical layers and mature nodes, including large volumes for NAND flash and for analogue and power devices above 28nm, and grow 6.28% a year. g-line, i-line and ancillaries, at USD 599.0 million, cover older wavelengths and the developers and coatings every litho cell needs, and grow 6.08%. Extreme ultraviolet photoresists overtake KrF before 2035 and finish as the second-largest category behind ArF. Each category is modelled through 2035 by node and region.

Where are photoresists consumed?

Asia Pacific consumes 78% of semiconductor photoresist value, USD 2.60 billion in 2025, because most wafer fabrication and almost all leading-edge output sit in Taiwan, South Korea, China and Japan. North America is the fastest-growing region at 10.4% a year as new fabs ramp.

Asia Pacific grows 8.8% a year to USD 6.04 billion in 2035. North America, at 12% of 2025 value, reaches USD 1.07 billion on fab construction under supply chain policies. Europe holds 8%, USD 266.2 million, growing 8.6% with fabs in Germany, Ireland and elsewhere. The Middle East, Latin America and Africa together account for about 2% of resist value. Six regional models sum to the global figure, with country tables in the Excel model, and the split follows where wafers are exposed, not where resists are formulated.

Who supplies photoresists?

Five Japanese companies lead semiconductor photoresist supply: JSR, Tokyo Ohka Kogyo, Shin-Etsu Chemical, Sumitomo Chemical and Fujifilm. DuPont and Merck KGaA supply resists and ancillaries, Korea’s Dongjin Semichem grew after 2019, and Chinese suppliers are gaining at mature nodes under localisation policy.

JSR, taken private under a Japanese government-backed investment fund, holds leading positions in ArF and extreme ultraviolet resists, and its April 2026 Taiwan venture with LCY Chemical and Wah Lee Industrial puts manufacturing in Taiwan, close to leading-edge foundry customers. Tokyo Ohka Kogyo is a leading supplier across resist categories, while Shin-Etsu, Sumitomo and Fujifilm hold strong qualified positions. Chinese suppliers are developing domestic resists, particularly for mature nodes. The competitive chapter profiles each supplier’s resist portfolio by wavelength, extreme ultraviolet capability, qualification positions with major chipmakers and regional production.

How are photoresists priced?

Semiconductor photoresists average USD 520 per litre in 2025, rising to about USD 539 in 2026 in the model, with a very wide spread from cheap g-line resists to premium extreme ultraviolet grades. The price leg adds 3.6% a year, driven by mix rather than list-price increases.

Older g-line and i-line resists cost relatively little per litre, KrF somewhat more, advanced ArF immersion resists considerably more, and extreme ultraviolet resists the most, reflecting their difficulty and the value of the chips they enable. Because resist is a small share of total chip cost but critical to yield, chipmakers rank performance and reliability above price, and qualification creates switching costs that hold pricing. Long-term supply agreements with major chipmakers are common. As extreme ultraviolet climbs from 18% to about 30% of value, the average litre sold becomes more expensive, which is the reason for the positive price leg. The pricing chapter publishes price bands by resist category.

Which 2035 path, chip downturn or leading-edge boom, would move photoresists off USD 7.87 billion?

The base case for semiconductor photoresists combines 5.2% volume growth and 3.6% price growth for an 8.99% CAGR and USD 7.87 billion in 2035. A chip downturn with costly PFAS reformulation lands near USD 5.43 billion; an AI-led leading-edge boom passes USD 10.6 billion.

In the chip-downturn scenario, semiconductor demand weakens and fluorinated chemical restrictions force reformulation, setting the legs at 3.0% and 2.0%. In the leading-edge scenario, artificial intelligence sustains advanced-node expansion and extreme ultraviolet and high numerical aperture adoption accelerate, setting them at 6.8% and 5.0%, about USD 2.69 billion above the base. Moves such as JSR’s Taiwan joint venture point toward the upper path, because they add supply capacity aimed at leading-edge customers. The published outside growth estimates we reconciled against span 6% to 14%, bracketing all three cases.

How could EU PFAS regulation and export controls reshape photoresist supply?

Three regulatory layers shape semiconductor photoresists: EU limits on PFAS chemistry, export controls on advanced materials, and subsidy programmes that reward local supply. The August 2025 ECHA update is the nearest test, because its outcome for semiconductor uses decides whether qualified resists must be reformulated.

Chemical regulation comes first: the restriction proposal first submitted in January 2023 and updated in August 2025 could affect photoacid generators and other fluorinated components in many photoresists, and the final decision rests with the European Commission and member states after ECHA’s scientific committees give their opinions. Export control and trade policy are second: photoresists for advanced nodes are subject to export controls in some jurisdictions, and the 2019 Japanese measures toward South Korea drove a wave of localisation. Industrial policy is third: semiconductor subsidy programmes in the United States, Europe, Japan and elsewhere encourage local supply of critical materials, including photoresists, influencing where capacity is built. The regulatory chapter maps these requirements.

Can the supply concentration be broken?

Only slowly. Korean and Chinese photoresist makers are taking share in mature-node and less critical resists, but at the leading edge, where extreme ultraviolet resists grow 14.59% a year, Japanese suppliers keep their lead because new suppliers must survive years of chipmaker qualification first.

South Korea responded to the 2019 export restrictions by accelerating domestic development, and Korean suppliers have made progress, particularly in less critical resists, while Korean chipmakers diversified sourcing. China, facing export controls on advanced semiconductor technology, has made localisation of photoresists a priority, with domestic suppliers gaining share in mature-node resists but lagging significantly at the leading edge. The barriers are considerable: advanced resist formulation requires deep chemistry expertise and access to high-purity raw materials, and above all a new supplier must be qualified by a chipmaker, a lengthy process that chipmakers are reluctant to undertake for critical layers because a resist failure can ruin enormous volumes of wafers. Formulation know-how is also heavily patented, and the portfolios and renewal decisions behind it are tracked in the systems sized in the Enterprises Intellectual Property Management Software Market report. Japanese suppliers are also responding by moving closer to customers, as JSR’s Taiwan venture shows. The model reflects gradual share gains by Korean and Chinese suppliers concentrated in mature-node resists, with Japanese suppliers retaining leadership in extreme ultraviolet and advanced ArF.

Douglas Exclusive: the lithography node and resist demand matrix

The lithography node and resist demand matrix is a Douglas Insights model, not an official register: its values are modelled estimates. It is built across 288 cells, four resist categories by three node bands by four chip types by six regions, and converts wafer output into semiconductor photoresist litres and value.

For each cell the matrix carries wafer starts, lithography layers by wavelength, resist consumption per layer, extreme ultraviolet adoption by layer and supplier qualification positions, so a change in a node roadmap flows through to resist demand by category, supplier and region. Licence holders receive it as a maintained tab in the Excel model.

How was the 6.4 million litre photoresist baseline built and checked?

The semiconductor photoresists baseline multiplies about 6.4 million litres by an average realised USD 520 per litre to reach USD 3.33 billion in 2025. Volume comes from wafer starts, layers and resist per layer; the growth path was reconciled against eight published estimates ranging from 6% to 14% a year.

The model is built bottom-up from volume: wafer starts by node, chip type and region, lithography layer counts by wavelength, resist consumption per layer and wafer, extreme ultraviolet adoption, and realised prices from supplier disclosures, with circuit board, display and packaging resists, lithography equipment and masks, and other wet chemicals excluded. The revenue CAGR of 8.99% decomposes into 5.2 points of volume and 3.6 points of price and mix, and the six regional models sum exactly to the global figure. Every figure carries a numbered source and a confidence grade in the fact sheet above, and the working model ships with every licence. The next scheduled review of this study is September 2027.

How this report was made: automated tools help assemble drafts, check that every figure matches the model and flag missing sources, while the Douglas Insights research desk chooses the sources, builds and owns the model, and approves the published report, as set out in our editorial standards.