DUV Window Material Selection: UV Fused Silica vs. CaF₂ – A Practical Look at Transmittance Degradation
2026-9-8
When you're designing deep ultraviolet (DUV) optical systems—those operating at wavelengths below 200 nm—the choice of window material is far from trivial. It's one of those decisions that directly dictates how long your system will survive and how reliably it performs over its operational lifetime.
The real headache kicks in when you're dealing with high-energy photon sources like 193 nm ArF or 248 nm KrF excimer lasers. Under that kind of relentless irradiation, transmittance degradation isn't just a theoretical concern—it's a practical problem that creeps in, pulse by pulse, eating away at your system's throughput.
As it stands, only two mainstream materials are truly up to the job for DUV window applications: UV-grade fused silica and single-crystal calcium fluoride (CaF₂). Both are proven performers, but here's the kicker—their degradation behaviors are fundamentally different.
Fused silica tends to suffer from compaction and color-center formation—a classic case of glass fatigue under photon bombardment.
CaF₂, by contrast, degrades primarily through impurity-induced defects, and it doesn't compact at all.
These differences ripple through everything: degradation rates, dose tolerance, and long-term reliability in real-world systems. If you're an optical engineer—or anyone serious about building DUV systems that actually last—understanding these distinctions isn't optional. It's engineering essentials.
Because when it comes to DUV windows, the details matter. And what you don't know about degradation will come back to bite you.
1. Intrinsic Transmittance Performance: The Baseline
1.1 UV-Grade Fused Silica
UV-grade fused silica—think of industry-standard grades like JGS1—is made from high-purity synthetic quartz feedstock, with hydroxyl groups and metal impurities held down to the parts-per-million (ppm) level. Its transmission range stretches from 185 nm all the way out to 2100 nm, and at 193 nm, a 10 mm thick sample delivers internal transmittance north of 90%. The intrinsic absorption coefficient at 193 nm? Less than 0.002 cm⁻¹ for high-quality material—pretty impressive.
On top of that, fused silica has a very low coefficient of thermal expansion (about 0.55×10⁻⁶/°C), excellent mechanical strength, and it polishes beautifully down to angstrom-level surface roughness. All of this makes it an economical and practical choice for large-aperture optical components.
1.2 Calcium Fluoride (CaF₂)
Calcium fluoride crystals take things a step further—their transmission range can push all the way down to 125 nm or even shorter, giving them an edge over fused silica in the deep UV. For lithography-grade CaF₂, the internal transmittance at 193 nm for 10 mm thickness exceeds 99.7%, with an absorption coefficient below 0.0005 cm⁻¹—just one-quarter that of fused silica.
Optically, CaF₂ is a dream: low refractive index (1.4338@587.6 nm), high Abbe number (~95), and extremely low dispersion. But there's a trade-off. CaF₂ is a single-crystal structure with a Mohs hardness of only 158.3, and its thermal expansion coefficient is a whopping 18.85×10⁻⁶/°C—about 34 times higher than fused silica. It's also slightly hygroscopic, meaning moisture can be a concern in humid environments.
2. Physical Mechanisms of Transmittance Degradation
2.1 Fused Silica: Two-Photon Absorption Driving Compaction and Color Centers
When fused silica gets hit by 193 nm DUV radiation, the degradation is essentially a two-photon absorption-driven bulk effect. High-energy photons get absorbed, triggering microscopic densification of the SiO₂ network—that's the "compaction" we keep hearing about. At the same time, defect centers like E′ centers and NBOH (non-bridging oxygen hole) centers form, introducing new UV absorption bands. The net result? Transmittance drops and refractive index shifts.
Studies have shown that at 193 nm, color centers form 4 to 5 times faster than at 213 nm—consistent with the wavelength dependence of the two-photon absorption coefficient. And here's the kicker: different manufacturers and different grades of fused silica can show wildly different degradation behavior. Compaction follows a power-law relationship, but both the exponent and the coefficient depend heavily on the material's manufacturing history. There's no clear damage threshold—transmittance loss simply climbs as fluence increases.
2.2 CaF₂: Impurity-Induced Color Centers Take Center Stage
CaF₂ is crystalline, so it doesn't suffer from compaction—that's a major advantage right out of the gate. Its degradation is primarily driven by impurity elements, especially transition metal ions, that form color centers under irradiation. Since CaF₂'s intrinsic absorption at 193 nm is already extremely low, material purity—particularly the levels of potassium (K), magnesium (Mg), and other trace elements—becomes the single most important factor governing transmittance stability. Even impurities at the ppm level can have a noticeable impact on post-irradiation performance.
3. Quantitative Comparison of Degradation Behavior
3.1 Low-Fluence Performance: Both Materials Shine
Under low-fluence conditions (<15 mJ/cm², 30 Hz), UV-grade fused silica lenses can push past 2 billion pulses without significant transmittance loss. For low-power DUV systems, that's more than enough to get the job done.
CaF₂ performs even better under comparable conditions. One experiment showed a CaF₂ lens surviving beyond 2 billion pulses at 90 mJ/cm², 150–170 Hz. Another data point: high-quality CaF₂ crystals exposed to a peak fluence of 240 mJ/cm² at 193 nm suffered a transmittance loss of only about 0.15% after 1 billion pulses—a truly remarkable level of radiation hardness.
3.2 High-Fluence Conditions: The Divergence
Once the fluence exceeds roughly 0.1 J/cm², fused silica's transmittance loss starts climbing steeply—and again, there's no abrupt threshold. CaF₂, thanks to its immunity to compaction, maintains a much flatter degradation curve under high fluence. This is exactly why CaF₂ is the go-to material for 193 nm lithography applications—like high-end objective lenses and polarizing prisms—where lifetime and stability are absolutely critical.
4. Engineering Trade-offs: Beyond Just Transmittance
4.1 Transmittance Isn't the Whole Story
Here's a reality check: the actual service life of a DUV window is often not limited by bulk material degradation. More often than not, the bottleneck is surface coating damage or environmental contamination buildup.
Uncoated fused silica surfaces lose about 4% per surface to Fresnel reflection, so anti-reflection coatings are a must for DUV windows. And in many cases, the coating's damage threshold and its thermal expansion match with the substrate become the true lifetime limiters.
CaF₂ is significantly harder to polish than fused silica, and coating it is even more demanding—coatings must be applied at elevated substrate temperatures, and electron-beam evaporation processes can even turn optical elements pink if not carefully controlled. Plus, CaF₂'s high thermal expansion makes coating cracking a real risk under temperature cycling—much more so than with fused silica.
4.2 Decision Framework: Which One Should You Choose?
Here's a quick cheat sheet to help you decide:
Your decision boils down to this:
High-power, long-lifetime, mission-critical DUV applications (e.g., 193 nm lithography inspection, high-energy excimer lasers) → Go with high-purity lithography-grade CaF₂. Just be prepared to seal it properly and control temperature/humidity.
Moderate power, cost-sensitive, outdoor, or wide-temperature-range systems → UV-grade fused silica is your workhorse. It's robust, environmentally stable, and much easier on the budget.
One golden rule: Never substitute regular fused silica for UV-grade JGS1. Standard material has absorption bands in the 220–280 nm region—it simply won't cut it for DUV.
Final Thoughts
Transmittance degradation curves give you a great starting point, but they're not the final answer. CaF₂ wins on sheer performance; fused silica wins on engineering practicality and balance. If your system design allows for periodic window replacement, silica is a perfectly viable and cost-effective choice. If you need decade-long, hands-off stability, then investing in CaF₂ is money well spent.
Optical design is, at its heart, the art of compromise. Choosing a DUV window material? No different. Understand the physics, respect the engineering trade-offs, and you'll make the right call every time.
The real headache kicks in when you're dealing with high-energy photon sources like 193 nm ArF or 248 nm KrF excimer lasers. Under that kind of relentless irradiation, transmittance degradation isn't just a theoretical concern—it's a practical problem that creeps in, pulse by pulse, eating away at your system's throughput.
As it stands, only two mainstream materials are truly up to the job for DUV window applications: UV-grade fused silica and single-crystal calcium fluoride (CaF₂). Both are proven performers, but here's the kicker—their degradation behaviors are fundamentally different.
Fused silica tends to suffer from compaction and color-center formation—a classic case of glass fatigue under photon bombardment.
CaF₂, by contrast, degrades primarily through impurity-induced defects, and it doesn't compact at all.
These differences ripple through everything: degradation rates, dose tolerance, and long-term reliability in real-world systems. If you're an optical engineer—or anyone serious about building DUV systems that actually last—understanding these distinctions isn't optional. It's engineering essentials.
Because when it comes to DUV windows, the details matter. And what you don't know about degradation will come back to bite you.
1. Intrinsic Transmittance Performance: The Baseline
1.1 UV-Grade Fused Silica
UV-grade fused silica—think of industry-standard grades like JGS1—is made from high-purity synthetic quartz feedstock, with hydroxyl groups and metal impurities held down to the parts-per-million (ppm) level. Its transmission range stretches from 185 nm all the way out to 2100 nm, and at 193 nm, a 10 mm thick sample delivers internal transmittance north of 90%. The intrinsic absorption coefficient at 193 nm? Less than 0.002 cm⁻¹ for high-quality material—pretty impressive.
On top of that, fused silica has a very low coefficient of thermal expansion (about 0.55×10⁻⁶/°C), excellent mechanical strength, and it polishes beautifully down to angstrom-level surface roughness. All of this makes it an economical and practical choice for large-aperture optical components.
1.2 Calcium Fluoride (CaF₂)
Calcium fluoride crystals take things a step further—their transmission range can push all the way down to 125 nm or even shorter, giving them an edge over fused silica in the deep UV. For lithography-grade CaF₂, the internal transmittance at 193 nm for 10 mm thickness exceeds 99.7%, with an absorption coefficient below 0.0005 cm⁻¹—just one-quarter that of fused silica.
Optically, CaF₂ is a dream: low refractive index (1.4338@587.6 nm), high Abbe number (~95), and extremely low dispersion. But there's a trade-off. CaF₂ is a single-crystal structure with a Mohs hardness of only 158.3, and its thermal expansion coefficient is a whopping 18.85×10⁻⁶/°C—about 34 times higher than fused silica. It's also slightly hygroscopic, meaning moisture can be a concern in humid environments.
2. Physical Mechanisms of Transmittance Degradation
2.1 Fused Silica: Two-Photon Absorption Driving Compaction and Color Centers
When fused silica gets hit by 193 nm DUV radiation, the degradation is essentially a two-photon absorption-driven bulk effect. High-energy photons get absorbed, triggering microscopic densification of the SiO₂ network—that's the "compaction" we keep hearing about. At the same time, defect centers like E′ centers and NBOH (non-bridging oxygen hole) centers form, introducing new UV absorption bands. The net result? Transmittance drops and refractive index shifts.
Studies have shown that at 193 nm, color centers form 4 to 5 times faster than at 213 nm—consistent with the wavelength dependence of the two-photon absorption coefficient. And here's the kicker: different manufacturers and different grades of fused silica can show wildly different degradation behavior. Compaction follows a power-law relationship, but both the exponent and the coefficient depend heavily on the material's manufacturing history. There's no clear damage threshold—transmittance loss simply climbs as fluence increases.
2.2 CaF₂: Impurity-Induced Color Centers Take Center Stage
CaF₂ is crystalline, so it doesn't suffer from compaction—that's a major advantage right out of the gate. Its degradation is primarily driven by impurity elements, especially transition metal ions, that form color centers under irradiation. Since CaF₂'s intrinsic absorption at 193 nm is already extremely low, material purity—particularly the levels of potassium (K), magnesium (Mg), and other trace elements—becomes the single most important factor governing transmittance stability. Even impurities at the ppm level can have a noticeable impact on post-irradiation performance.
3. Quantitative Comparison of Degradation Behavior
3.1 Low-Fluence Performance: Both Materials Shine
Under low-fluence conditions (<15 mJ/cm², 30 Hz), UV-grade fused silica lenses can push past 2 billion pulses without significant transmittance loss. For low-power DUV systems, that's more than enough to get the job done.
CaF₂ performs even better under comparable conditions. One experiment showed a CaF₂ lens surviving beyond 2 billion pulses at 90 mJ/cm², 150–170 Hz. Another data point: high-quality CaF₂ crystals exposed to a peak fluence of 240 mJ/cm² at 193 nm suffered a transmittance loss of only about 0.15% after 1 billion pulses—a truly remarkable level of radiation hardness.
3.2 High-Fluence Conditions: The Divergence
Once the fluence exceeds roughly 0.1 J/cm², fused silica's transmittance loss starts climbing steeply—and again, there's no abrupt threshold. CaF₂, thanks to its immunity to compaction, maintains a much flatter degradation curve under high fluence. This is exactly why CaF₂ is the go-to material for 193 nm lithography applications—like high-end objective lenses and polarizing prisms—where lifetime and stability are absolutely critical.
4. Engineering Trade-offs: Beyond Just Transmittance
4.1 Transmittance Isn't the Whole Story
Here's a reality check: the actual service life of a DUV window is often not limited by bulk material degradation. More often than not, the bottleneck is surface coating damage or environmental contamination buildup.
Uncoated fused silica surfaces lose about 4% per surface to Fresnel reflection, so anti-reflection coatings are a must for DUV windows. And in many cases, the coating's damage threshold and its thermal expansion match with the substrate become the true lifetime limiters.
CaF₂ is significantly harder to polish than fused silica, and coating it is even more demanding—coatings must be applied at elevated substrate temperatures, and electron-beam evaporation processes can even turn optical elements pink if not carefully controlled. Plus, CaF₂'s high thermal expansion makes coating cracking a real risk under temperature cycling—much more so than with fused silica.
4.2 Decision Framework: Which One Should You Choose?
Here's a quick cheat sheet to help you decide:
|
Consideration |
UV-Grade Fused Silica |
CaF₂ |
|
193 nm absorption |
~0.002 cm⁻¹ |
~0.0005 cm⁻¹ |
|
Degradation mechanism |
Compaction + color centers |
Impurity-driven color centers |
|
High-fluence tolerance |
Moderate |
Excellent |
|
Mechanical strength |
High |
Low (soft, easy to scratch) |
|
Thermal expansion |
0.55×10⁻⁶/°C |
18.85×10⁻⁶/°C (34× higher) |
|
Polish/coat difficulty |
Low |
High |
|
Cost |
Lower |
Significantly higher |
High-power, long-lifetime, mission-critical DUV applications (e.g., 193 nm lithography inspection, high-energy excimer lasers) → Go with high-purity lithography-grade CaF₂. Just be prepared to seal it properly and control temperature/humidity.
Moderate power, cost-sensitive, outdoor, or wide-temperature-range systems → UV-grade fused silica is your workhorse. It's robust, environmentally stable, and much easier on the budget.
One golden rule: Never substitute regular fused silica for UV-grade JGS1. Standard material has absorption bands in the 220–280 nm region—it simply won't cut it for DUV.
Final Thoughts
Transmittance degradation curves give you a great starting point, but they're not the final answer. CaF₂ wins on sheer performance; fused silica wins on engineering practicality and balance. If your system design allows for periodic window replacement, silica is a perfectly viable and cost-effective choice. If you need decade-long, hands-off stability, then investing in CaF₂ is money well spent.
Optical design is, at its heart, the art of compromise. Choosing a DUV window material? No different. Understand the physics, respect the engineering trade-offs, and you'll make the right call every time.
