In chip manufacturing workshops, there is a key material that is “invisible yet indispensable” — it acts like a high-precision “film” that accurately transfers chip design blueprints onto silicon wafers. Without it, even the most complex chip designs cannot be mass-produced. This material is the photomask, and its “core framework” — the blank mask — is monopolized by Japanese and South Korean companies, with EUV-level products even subject to embargoes against our country.
By 2025, as domestic chip manufacturing capacity expands, the photomask market is set to explode: the global scale will exceed $9.5 billion, with the blank mask market reaching $2.01 billion in 2024, and a projected CAGR of 9.07% over the next decade. This report on photomasks and blank masks, published by Huayuan Securities, dissects the technical logic, market landscape, and domestic opportunities in this “bottleneck” sector. Today, we will explain in simple terms why photomasks are important, the challenges of domestic substitution, and which companies have already made breakthroughs.
1. First, let’s understand: What is a photomask? Without it, chip design is just “talking on paper”.
In simple terms, a photomask is a “pattern transfer tool” for chip manufacturing — just like using a template to draw as a child, the photomask is engraved with the circuit patterns of the chip, and the lithography machine uses it to “project” the patterns onto silicon wafers coated with photoresist. After development and etching, the circuits appear on the silicon wafer.
1. The “dual identity” of photomasks
Structurally: It consists of a “transparent substrate + opaque layer” — the substrate is made of high-purity quartz (for high-precision scenarios) or soda glass (for mid-to-low-end scenarios), and the opaque layer is made of chromium or molybdenum-silicon alloy, with the patterns engraved on the opaque layer;
Functionally: It serves as a “bridge connecting IC design and manufacturing” — the chip layout produced by design companies must first be made into a photomask before it can be mass-replicated on silicon wafers. Without a photomask, chip design is just “talking on paper”.
2. Classified by precision, the most “valuable” is for chips
Different downstream scenarios have vastly different precision requirements for photomasks, with the technology used for semiconductor chips being the most challenging and valuable:
| Application Field | Minimum Line Width | CD Precision (Precision Index) | Typical Uses |
|---|---|---|---|
| Semiconductor Chips | 0.5μm | 0.02μm | Mobile SoC, AI Chips |
| Flat Panel Displays | 1.2μm | 0.10μm | Mobile Screens, TV Panels |
| Printed Circuit Boards | 10μm | 0.50μm | Mobile PCBs, Automotive Circuit Boards |
For example, a photomask used for 14nm chips can be worth up to $500,000 per piece, and a set (30 pieces) can cost $7.5 million, which is more than three times the cost during the 65nm era — the more advanced the process, the more expensive the photomask.
2. Blank masks: The “core framework” of photomasks, accounting for over 50% of value
If the photomask is the “film”, then the blank mask is the “unexposed film substrate” — it is the raw material for photomasks, consisting of “quartz substrate + opaque film + photoresist”; the patterns are later engraved onto it to create the photomask.
1. The “technical barriers” of blank masks: Requirements finer than a hair
The quality of blank masks directly determines the yield of chips, with three major technical challenges:
Ultra-high flatness: The surface roughness of the substrate must reach the nanometer level (1/10,000th the thickness of a hair), with a flatness requirement of ≤0.1μm for 14nm nodes; otherwise, pattern transfer will be distorted;
Extremely low defect rate: There must be no impurities (such as pinholes or grain-sized defects) on the substrate, or else the chips will have “bad spots”, leading to a sharp drop in yield;
Ultra-thin opaque layer: For nodes below 20nm, thin masks must be used, and the opaque layer must maintain its opacity even when thinned, a technical challenge comparable to “carving patterns on an eggshell”.
2. How important is the market? Expected to exceed $2 billion in 2024, with domestic share only at 1%
Blank masks are the most valuable raw material for photomasks — in 2023, blank masks (quartz substrates + soda substrates) accounted for 52% of the procurement costs for Longtu Photomask, making it the highest cost component.
However, this market is firmly monopolized by Japanese and South Korean companies:
Global landscape: Japan’s HOYA accounts for over 50%, followed closely by Shin-Etsu and South Korea’s Samsung SDI, with the three companies collectively holding over 80% of the global market share;
Domestic challenges: For mid-to-low-end photomasks (i-line/KrF), HOYA accounts for over 50%; for high-end ArF photomasks, almost all rely on HOYA and Shin-Etsu; the most advanced EUV blank masks are also embargoed, making them completely unavailable to domestic companies.
3. Technological iterations: From “chromium plates” to “OMOG”, domestic companies lag three generations
As chip processes have evolved from 90nm to 7nm, blank masks have undergone three major upgrades, while domestic companies are still stuck in the first generation:
1.0 Era (90nm and above): Chromium plate masks, single-layer chromium opacity, can be produced by domestic companies (such as Qingyi Optoelectronics);
2.0 Era (65-28nm): Phase-shift masks (PSM), using MoSi composite layers to enhance resolution, widely adopted abroad, with only a few domestic companies able to produce them in small batches;
3.0 Era (14nm and below): OMOG masks, made from molybdenum-silicon oxide materials, with three times the optical density, compatible with EUV lithography, still under development domestically.
3. Domestic breakthroughs: Two companies have made progress, with quartz masks as the breakthrough point
Despite the significant gap, by 2025, domestic photomask companies have found breakthroughs in the mid-to-high-end sector, with Longtu Photomask and Qingyi Optoelectronics making the most notable progress.
1. Longtu Photomask:
Quartz masks account for over 80%, with rapid growth in the semiconductor field
As one of the few domestic third-party manufacturers capable of producing advanced photomasks, Longtu Photomask’s core advantage lies in “semiconductor quartz masks”:
Product structure: In 2024, revenue from quartz masks is expected to exceed 200 million yuan, accounting for over 80% of total revenue, mainly used in power semiconductors and advanced packaging, with line widths reaching 0.5μm, close to international second-tier levels;
Performance growth: From 2020 to 2024, revenue CAGR is expected to reach 46.93%, with a net profit of 92 million yuan in 2024 (up 9.84% year-on-year), making it the most stable in profitability among domestic companies;
Customer breakthroughs: It has entered the supply chains of leading domestic power semiconductor manufacturers (such as Silan Microelectronics) and advanced packaging manufacturers (such as JCET), and is expected to penetrate the supply chain of SMIC’s mature processes by 2025.
2. Qingyi Optoelectronics: The largest in the country, driven by both display and semiconductor sectors
Qingyi Optoelectronics is the “big brother” of domestic photomasks, established earliest and the largest in scale, with advantages in both “display + semiconductor”:
Leading scale: In 2024, revenue is expected to reach 1.112 billion yuan (up 20.35% year-on-year), with a net profit of 172 million yuan (up 28.49% year-on-year), making it the only photomask company in the country with revenue exceeding 1 billion yuan;
Product upgrades: Revenue from quartz masks is expected to reach 1.02 billion yuan (up 20.43% year-on-year), with the proportion of semiconductor quartz masks increasing to 30%, capable of producing PSM masks for 45nm nodes;
Scene coverage: In addition to semiconductors, it holds over 20% market share in the display sector (AMOLED, LCD), being a core supplier for BOE and TCL, enhancing its risk resistance.
Besides these two, there are also SMIC’s own SMIC Mask (for self-use) and Wuxi Huaren Shanghua (focusing on mature processes), but among third-party manufacturers, Longtu and Qingyi are the most competitive in the market.
4. Industry summary: 2025 is a “critical window period” for domestic substitution, with both opportunities and challenges
1. Three major opportunities: Downstream demand + technological breakthroughs + policy support
Explosive downstream demand: The expansion of domestic 28nm and above mature process capacity (SMIC, Huahong Semiconductor) is expected to lead to an annual growth of 15% in photomask demand, providing “training opportunities” for domestic companies;
Partial technological breakthroughs: The purity of domestic quartz substrates (such as Quartz Co.) has reached 99.999%, meeting the requirements for ArF photomasks; coating equipment (such as North Huachuang) can also achieve partial substitution, reducing dependence on foreign equipment;
Policy support: Photomasks have been included in the key support areas of the “Big Fund”, and local governments (such as Shanghai and Guangdong) are providing subsidies for photomask projects, accelerating the expansion of domestic companies.
2. Two major challenges: Technical barriers + geopolitical risks
Difficulty in overcoming technological gaps: The multi-layer film technology for EUV blank masks and non-pattern detection technology have been monopolized by foreign countries for 20 years, and domestic companies will need at least 5-8 years to break through;
Geopolitical risks remain: If Japan and South Korea restrict exports of ArF blank masks, the supply of photomasks for advanced processes (28nm and below) in the country will be cut off, leaving only mature processes available.
3. Future trends: Focus on “mature processes + specialized techniques”, gradually penetrating into high-end markets
Domestic substitution will not happen “overnight”, but will proceed in three steps:
Short-term (2025-2027): Capture the mature process (28nm-90nm) and specialized technique (power semiconductors, MEMS) photomask markets, aiming to increase the domestic substitution rate from 10% to 30%;
Mid-term (2028-2030): Break through ArF photomask technology, achieve domestic production of blank masks for 14nm nodes, and enter the supply chains of SMIC and Yangtze Memory Technologies;
Long-term (after 2030): Develop EUV blank masks, break the overseas embargo, and support the development of domestic advanced processes of 7nm and below.
Conclusion: The “domestic battle” for photomasks must be steady and long-term
Photomasks may seem “niche”, but they are a “critical consumable” in chip manufacturing — without them, no matter how large the chip production capacity, operations cannot commence. By 2025, domestic companies have made small inroads into the mid-to-high-end sector, but to truly break the monopoly of Japan and South Korea, they need to solve three major challenges: “materials (high-purity quartz), equipment (coating/detection equipment), and processes (OMOG technology)”.
For investors, focusing on companies with “high quartz mask ratios and rapid breakthroughs in semiconductor clients” (such as Longtu Photomask and Qingyi Optoelectronics) can better capture the dividends of domestic substitution; for the industry, the domestic substitution of photomasks is not about “overtaking on a curve”, but rather about “steady progress” — only by first establishing a foothold in mature processes can they gradually break into high-end markets.
