EUV mask blank supply is structurally concentrated in two qualified substrate-and-coating vendors (AGC Inc. and Hoya), with two additional names (S&S Tech, Applied Materials) cited as active challengers in published market rosters [S3][S5]. The result is a strategic chokepoint sitting one layer above the wafer-fab capex that drives the rest of the front-end equipment cycle.
EUV mask blanks are ultra-flat substrates, dominated by low thermal expansion material grades (72.40% of 2025 market share by substrate material), coated with alternating molybdenum and silicon layers (typically 40-50 pairs) tuned for 13.5 nm reflectivity, and capped with a thin protective layer for high-NA scanner compatibility [S4][S5]. Any imperfection on the surface can distort reflected light and collapse wafer yield, which is why blank qualification cycles are measured in years rather than quarters.
Who actually makes qualified EUV mask blanks
AGC and Hoya are the only two suppliers with merchant-shipped, customer-qualified EUV mask blanks at the multilayer level for high-volume logic production, a position that has been publicly visible since the 2018-2022 capacity ramp [S5][S9]. AGC confirmed in January 2022 that it would double EUVL mask blank production capacity to keep pace with circuit-shrinking demand [S8]; in January 2024, AGC's subsidiary AGC Electronics began a phased expansion targeting approximately 30% additional capacity by 2025 [S4]. S&S Tech (Korea) and Applied Materials are listed as additional market participants [S3], with Applied Materials positioned as a deposition-equipment supplier that is also vertically integrating into blank production capability [S4][S5].
For fab-side specifiers, the practical implication is a single-digit supplier list. Procurement teams that need to qualify a second source typically have one, not three, alternatives to fall back on, and that constraint is now codified in published market research as a "supply chain concentration risk with few qualified vendors" that exposes manufacturers to potential bottlenecks [S3]. The same report also flags "ultra-high purity material requirements" as a cost-side restraint, which compounds the supply-side bottleneck: even if a new entrant raises capital, the qualification cycle (defect density, multilayer uniformity, flatness, capping-layer chemistry) stretches the practical time-to-revenue well beyond a normal fab-tool startup.
Market sizing and where the growth pressure lands
Published market estimates diverge, which is itself a data point about how opaque the supply base is. Reportprime projects the EUV mask blanks market at USD 220.60 million in 2025, growing at 10.30% CAGR to USD 438.16 million by 2032 [S3]. SNS Insider values the same market at USD 232.00 million in 2025 with a 14.05% CAGR to USD 864.00 million by 2035 [S4]. Strategic Market Research uses a wider scope and pegs the global figure at roughly USD 1.9 billion in 2024 expanding to nearly USD 3.6 billion by 2030 at a 10.8% CAGR [S5], while 360iResearch forecasts growth by USD 1,939.25 million at a 12.78% CAGR through 2032 [S6]. The wide spread is driven by whether the report counts only merchant multilayer blanks or the broader blank-plus-pellicle-plus-substrate ecosystem, but the directional read is the same: double-digit growth, and most of it absorbed by a two-vendor supply base.
Demand-side drivers are concentrated too. Sub-5nm logic nodes (where EUV is now the default exposure technology), AI accelerator capacity, and high-bandwidth memory expansion are all pulling on the same blank inventory pool [S3][S5]. Reportprime's segmentation puts "7nm and 5nm semiconductor processes" as the primary application bucket [S3], while SNS Insider notes that multilayer mask blanks held 68.50% market share by type in 2025, with phase-shifting mask blanks growing fastest at 15.80% CAGR over 2026-2035 [S4]. For specifiers working in adjacent industrial domains, the same supply discipline shows up in pressure transmitter sourcing for safety instrumented functions, where a qualified-vendor list of two or three is the norm rather than the exception for harsh-service or safety-rated positions.
Capacity, substrates, and the high-NA inflection
The substrate is the physical chokepoint. The substrate must hold its figure to sub-nanometer flatness under scanner-stage acceleration and EUV-induced heating, which is why only a small number of glass melts have been qualified for the application. Once the substrate is made, the multilayer stack (Mo/Si, roughly 40-50 period pairs) is deposited by ion-beam or magnetron sputtering, with a ruthenium or similar capping layer added for high-NA EUV compatibility [S5][S7].
High-NA EUV, the next scanner generation from ASML, is tightening the blank specification further. MLC (multilayer plus capping layer) blanks are estimated at roughly 58% of 2024 global market revenue, driven by better defect protection and high-NA compatibility [S5]. SNS Insider's trend list explicitly includes "rising qualification activity for high-NA EUV compatible blanks supporting next-generation scanner roadmaps" and "expanding focus on defectivity reduction as sub-3-nanometer nodes tighten acceptable defect density thresholds" [S4]. Each of these tightening axes (flatness, defect density, multilayer thickness control, capping-layer uniformity) is a place where a single qualified supplier can throttle the entire fab ramp.
Geopolitical layer: dual-use classification and export controls
The supply concentration is now being reflected in trade policy. SNS Insider's 2025 trend list includes "increasing classification of EUV mask blanks as strategic dual-use items under export control frameworks" [S4], and Strategic Market Research frames the same dynamic as a "geopolitical and technological lever" sitting at the intersection of advanced optics, semiconductor scaling, and supply-chain risk [S5]. The U.S. domestic market is small in absolute terms (USD 28 million in 2025 per SNS Insider) but is being actively subsidized because the material is treated as a strategic dependency rather than a commodity input [S4].
For process engineers and procurement leads outside the front-end fab, the relevant lesson is structural: a category with a two-vendor supply base, multi-year qualification cycles, and active export-control classification behaves more like a defense material than a commercial chemical. Similar supply-discipline patterns surface in epoxy resin markets, where BPA-feedstock regulation and substitution pathways are reshaping the qualified-supplier list on a similar multi-year clock.
Comparison: the four named EUV mask blank participants
Across the publicly cited market rosters, four names recur, and the spec-relevant differences are positional rather than purely commercial: [S3]
AGC Inc. (Japan): incumbent merchant supplier, the only one with a publicly announced capacity-doubling program (January 2022) and a phased 30% expansion at AGC Electronics through 2025 [S4][S8]. Photronics Inc. (US) is positioned as a photomask maker rather than a blank substrate manufacturer, with 2025 investment in advanced photomask capacity rather than blank substrate [S4]. Hoya Corporation (Japan) is the second incumbent with merchant EUV multilayer blanks, cited in strategic-market-research stakeholder mapping alongside AGC and JSR [S5]. S&S Tech (Korea) and Applied Materials appear as additional market participants in published rosters [S3], with Applied Materials bringing deposition-equipment vertical integration into the blank production chain [S4][S5].
For a fab evaluating second-source qualification, the realistic choice set is AGC and Hoya today, with Applied Materials as the most credible challenger because it controls the deposition tool set used in blank production.
What this means for downstream specifiers
Blank supply concentration does not directly affect most industrial specifiers, but it does affect any factory whose production roadmap touches sub-7nm silicon: AI accelerator demand, advanced driver-assistance silicon, and high-bandwidth memory all sit downstream of the same two-vendor bottleneck. Reportprime's restraint list names the exact failure mode: "supply chain concentration risk, with few qualified vendors, exposes manufacturers to potential bottlenecks" [S3]. SNS Insider's 2025 trend list reads in the same direction, with "growing capacity investment among a concentrated group of vertically integrated coated-blank suppliers" listed as a defining market feature [S4].
For industrial process engineers reading the same kind of concentration risk into adjacent categories, the playbook is similar: qualify a second source early, track capex announcements at the supplier level, and treat multi-year qualification cycles as a hard constraint rather than a soft preference. The dust mask and industrial UPS categories show the same structural shape, a small qualified-vendor list, long qualification cycles, and a strategic-materials framing, and the same sourcing discipline applies. A single new entrant breaking into merchant EUV mask blank supply, or an AGC or Hoya capacity expansion slipping, would be the most actionable signal to watch over the next 12-24 months.