 |
03 Jun 2005
Ultraviolet (UV) light is used for everything from semiconductor lithography and eye surgery to water sterilization and micromachining. Volker Schmidt from Berliner Glas talked to Optics.org about the current status of optics developed for this important spectral region.
Q. What type of materials are used to make transmission optics for the UV?
The ultraviolet wavelength range is separated into five ranges: near UV (450–350 nm); mid UV (350-300 nm); deep UV (DUV) (300–190 nm); vacuum UV (190-50 nm); and extreme UV (50-1 nm).
In the deep-UV range, the choice of materials includes fused silica, sapphire and several fluorides like calcium fluoride or magnesium fluoride. Each has its own advantages and disadvantages. Sapphire shows superior mechanical strength but is strongly birefringent and dispersive. In contrast, fused silica is mechanically strong, thermally very stable and has the advantages of no birefringence and low dispersion. However, even its purest forms strongly absorb light with a wavelength below 190 nm.
For wavelengths shorter than 190 nm and for chromatic correction of lens systems calcium fluoride is the most popular choice. Of all the fluorides on offer, it best satisfies the ideal characteristics of a material with low water absorption (hygroscopic susceptibility), low birefringence and low dispersion. Compared to fused silica it is soft and thermally instable. Finally, magnesium fluoride tends to be used at vacuum uv wavelengths and in applications where good heat conductivity is required.
Q. What is the typical performance of the best optics to date?
Deep UV applications involving KrF (248 nm), ArF (193 nm) and F2 (157 nm) lasers often require high damage-threshold optical materials made from either fused silica, calcium fluoride or magnesium fluoride.
Today, very high optical transmission can be achieved with these materials. For example, KrF-quality fused silica transmits more than 99.8%/cm at average power levels of up to 8 J/cm2.
In comparison, ArF-quality fused silica is available with a transmission of better than 99.5%/cm at average power levels of up to 3 J/cm2. Calcium fluoride provides a transmission of 99.5%/cm at the ArF-wavelength and 99.0% at the F2-wavelength.
It is important to note that the above transmission values are for low power levels. At higher powers a nonlinear process called two-photon-absorption starts and can dramatically degrade the transmission. Another well known process that affects transmission of DUV-optics is colour centre formation or fluorescence which is due to impurities in the glass.
Both these troublesome phenomena can be reduced by using extremely high purity materials. ArF and KrF-quality glass contain about 1000 times less aluminium and iron then typical glass for visible applications.
Q. Why are high damage-threshold UV optics so difficult to manufacture?
The manufacturing of high-damage threshold UV-optics is very demanding because the optics must be very pure and very smooth to avoid unwanted losses in transmission.
It is well known that surface scatter is a significant contributor to energy loss in DUV-optics. To keep scatter low at short wavelengths “super polishing” is often applied to keep the surface roughness below 0.8 nm rms. In addition, errors in the shape of the optic (topographic surface deviations) must be minimised as they have a much stronger effect on the wavefront error than in the visible region.
Any “cosmetic” surface error like scratches and chips can become centres for component degradation that ultimately lead to its failure. For this reason high damage threshold DUV optics typically require a scratch and dig specification of 10-5.
This becomes especially difficult for calcium fluoride because it is a soft and crystalline material which is easily chipped and scratched during polishing.
To achieve the highest damage-thresholds special processes involving magnetorheological finishing (MRF), acid etching and laser conditioning have been developed. As described in [1] damage threshold levels of 15 J/cm2 were achieved with these methods to generate optics for the National Ignition Facility (NIF) at Lawrence Livermore in the US.
Conventional polishing process using, for instance, zirconia-based slurry unavoidably leaves hidden pits and scratches in the surfaces. These sub-surface-damages become visible after etching and are a well known source of surface degradation in high power DUV-applications. In contrast, the MRF-process leaves nearly no residual pits and scratches.
Q. What kinds of coatings are available?
Most DUV-optical components require coatings. For anti-reflection coatings oxides and fluorides are typically used such as SiO2 and Al2O3 for the ArF and KrF wavelengths. However, the wavelength of 157 nm requires fluorides like MgF2 and LaF3.
Fluorides are more resistant to high power laser radiation than oxides, but oxides are tougher and easier to deposit using modern coating technologies like sputtering or Advanced Plasma Source. Aluminium is often used to generate high reflection mirrors in the DUV. It offers high reflectivity (90%) over a broad wavelength range but is not as effective as specialised fluoride interference coatings.
Q. What are the most important applications for high damage-threshold UV optics?
The most dominant application of high-damage threshold UV optics is in projection lithography. Currently, most systems use 248 nm sources, but there is a move towards 193 nm systems.
The expected move to F2-laser sources with a wavelength of 157 nm was a strong driver for the development of DUV optics, especially calcium fluoride material and fluoride coatings. With the decision of the semiconductor industry to drop this development and to utilize 193 nm immersion optics instead, much has changed.
The focus is now on higher power 193 nm optics, especially low absorption fused silica and better coatings. Besides UV-lithography there are several DUV-applications in micromachining, astronomy, synchrotron radiation optics, materials processing, TFT annealing and medicine.
Examples for micromachining are hole drilling, microstructuring, marking for electronic packaging and the direct writing of fiber bragg gratings. The most important medical application of DUV-optics is the Lasik (Laser In-Situ Keratomileusis) eye surgery. This vision correction technique uses a DUV laser to remove a precise amount of tissue from the cornea to alter its curvature and correct vision.
In TFT annealing excimer laser radiation is used for the polycrystallization of amorphous silicone thin film on substrates. It is applied now for the mass production of polysilicon TFT liquid crystal display panels. Another very promising application for high-power DUV-Lasers is the surface treatment of cylinders in vehicle engines. The idea is to create a non-metallic surface structure on the internal cylinder surface that results in reduced oil consumption and a longer service life.
Q. What do scientists need to consider when purchasing UV optics?
When purchasing DUV-optics the following criteria need to be taken into account:
• Is the material appropriate for wavelength and power levels of the application in mind? For demanding applications so called KrF or ArF-quality material is available from various suppliers.
• Is birefringence a problem in the application? Some materials like sapphire cannot be used if this is the case.
• Is the wave front error defined for the wavelength in question? Surface irregularities are often defined in fractions of wavelengths and often specified using visible sources such as a standard He-Ne-wavelength (633 nm).
• Surface cosmetics and surface roughness should always be well defined.
• If possible a coating with verified damage threshold should be used.
• Special care should be taken to choose a supplier with full knowledge of all the production steps including material quality and coating processes.
Volker Schmidt is development manager at Berliner Glas, one of Europe’s leading manufacturer of speciality glass products and precision optical components.
References:
[1] “MRF Improves Laser Damage Resistance of UV Optics”, Stephen D. Jacobs, Newsletter of the Center for Optics Manufacturing, Volume 10, Number 1 , January/February 2002.
 |  |  |  |  |  |  |
| © 2024 SPIE Europe |
|
