193-nm coatings resist excimer-laser damage

The best way to design and test coatings for use at 193 nm relies on two approaches: one for high pulse energy and low repetition rate, and the other for low pulse energy and high repetition rate.

Apr 1st, 2004
Th 146612

The best way to design and test coatings for use at 193 nm relies on two approaches: one for high pulse energy and low repetition rate, and the other for low pulse energy and high repetition rate.

Argon fluoride (ArF) lasers produce intense, short pulses of 193-nm radiation, making them useful light sources for a host of scientific, medical, and industrial applications. Eyesight correction (laser assisted in situ keratomileusis, or LASIK; and photorefractive keratotomy, or PRK), deep-UV and immersion microlithography, materials processing, photoablation and laser marking are just a few examples of applications that incorporate ArF-laser sources. As these laser-based applications continue to advance, they create new and challenging requirements for high-performance deep-UV optics and coatings. Systems operating at low to medium repetition rates are capable of average power levels of several hundred watts or greater, and recent generations of ArF lasers are now capable of operating at up to 4000 Hz at 20-W average power. In addition, lifetime requirements often exceed several billion pulses in industrial applications, creating significant demands on coated optics.

Many factors influence the damage resistance and lifetimes of 193-nm coatings. From a manufacturing standpoint, selection of coating materials, thin-film-coating design, the coating process, and accurate quality control are all crucial factors that determine the performance of a mirror. Equally important are application factors such as laser fluence, repetition rate, and environmental conditions.

Manufacturing considerations

Coating-material selection starts with an incoming inspection process designed to ensure that only the best materials are used in the production of 193-nm mirrors. The incoming quality-control procedure should include an initial monolayer coating prepared from any new lot of coating materials. The monolayer is measured to determine absorption at the laser wavelength; if required, optical constants can be derived from the measurement data.

Thin-film design enables the rapid development of coatings that address a broad array of requirements, including multiwavelength or multi-angle operation, electric-field-intensity (EFI) manipulation within coating layers, and selection of coating materials for specific environmental conditions.

The coating process is vitally important because this is where layers are deposited and mirrors are produced. Control of temperature, pressure, trace-gas levels, and layer thickness all contribute to the ultimate performance of a 193-nm laser mirror. Research on how temperature affects candidate 193-nm coating materials is important, as is incorporation of a residual-gas analyzer to monitor trace gases that can impact coating properties.

Accurate and reproducible monitoring of layer thickness is a key factor to successful production of optical thin films. To this end, an optical monitoring system was designed and manufactured with dual feedback for improved layer control. By incorporating both broad-spectrum and single-wavelength monitoring, layer thickness can be precisely controlled.

Quality control, the final step in the manufacturing process, ensures that the user receives the highest performance mirrors possible. In addition to a visual inspection, we recommend a quality-control process that includes a measurement of both reflection and transmission at the laser wavelength and angle of incidence, as this helps to determine absorption losses that may affect mirror performance.

Application factors

Laser fluence and repetition rate must be considered in the design stage if a mirror is to perform within the scope of demanding industrial applications. From an optical-design standpoint, we categorize mainstream industrial 193-nm laser applications into two general groups: relatively high laser fluence (up to and exceeding 500 mJ/cm2) with low repetition rates (up to about 200 Hz), and relatively low laser fluence (on the order of 5 to 30 mJ/cm2) with high repetition rates (up to 4 KHz). Of course, there will always be situations with different requirements and operating parameters that are addressed on an individual basis.

High fluence, low repetition rate

FIGURE 1. Nomarski-microscope images show the onset (left) and progression (right) of laser-induced damage within voids and pinholes in a conventional 193-nm mirror coating. These mirrors exhibit moderate and severe damage from laser exposure. In both instances, laser-induced damage started within microscopic voids in the coating. The severe damage at right occurred after 240,000 pulses at 225 mJ/cm2 at a 100-Hz repetition rate.
Click here to enlarge image

In applications with high fluence and low-repetition-rate operation, laser-induced damage typically begins within microscopic pinholes and voids in the coating (see Fig. 1). Depending on the coating design, materials selected, and laser characteristics, this type of laser-induced damage can either occur in the form of gradual enlarging of voids, leading to failure over time, or in some cases rapid and catastrophic coating failure. The impact of catastrophic failure is usually apparent. In the case of gradual deterioration, however, the user may observe reduced system throughput and possibly increased scatter over time. Surface contamination can also lead to coating failure, so it is important to ensure that the surfaces are clean before installation of coated optics. We recommend a high-purity methanol or acetone lens-tissue wipe to remove residual contaminants; however, the coating supplier should be contacted for instructions before attempting to clean a mirror surface.

Electric-field-intensity peaks within coating layers can play a role in laser-induced damage; this is the basis of the so-called "e-design" approach. Traditional 193-nm coatings manufactured on a commercial basis consist of quarter-wave stacks of high- and low-index materials such as lanthanum fluoride and magnesium fluoride. If materials are selected with low absorption at 193 nm, reflectance continues to increase as paired layers of materials are added to the stack. There are inherent stress properties that limit the number of layers that can be deposited, yet with the right material combinations, it is possible to produce a stack with 40 or more layers yielding 97% to 99% reflectance at 193 nm.

Some coating designs incorporate a half-wave low-index overcoat as a means of increasing resistance to laser damage. The purpose of this design is to position the highest EFI peak within the half-wave layer that offers greater resistance to laser damage. The e-design takes EFI correction to the next level by positioning and distributing EFI peaks within several layers, thus reducing the intensity within any single layer in the stack. This approach diminishes the effect of EFI-related damage, resulting in increased laser-damage resistance and longer coating lifetimes.

FIGURE 2. Results of laser-damage testing conducted at 225 mJ/cm2 and a 100-Hz repetition rate show improved resistance to 193-nm laser damage and increased lifetimes achieved with an EFI-corrected coating.
Click here to enlarge image

To compare the various design strategies, we prepared test coatings using identical coating materials and the same coating process. The first design incorporated a quarter-wave stack with half-wave overcoat and the second was manufactured using the e-design concept. The two test coatings, plus additional 193-nm mirrors from industry suppliers, were sent to a commercial laser lab for damage testing (Spica Technologies; Nashua, NH). When exposed to 225 mJ/cm2 at a 100-Hz repetition rate, the new e-design survived twice as long as the test coating with a half-wave overcoat, even though they incorporated the same materials and had similar reflectance characteristics (see Fig. 2). At this fluence level and repetition rate, most conventional coatings exhibited rapid failure. It is also important to note that when the laser fluence was increased to 300 mJ/cm2 and the repetition rate to 200 Hz, the e-design lifetime was a factor of ten greater than the closest test mirror. It appears that for high-fluence/low-repetition-rate applications, EFI correction—and specifically the e-design strategy—plays an important role in resistance to laser-induced damage.

High repetition rate, low fluence

For applications incorporating ArF lasers with kilohertz repetition rates and relatively low fluence levels, it appears that a major damage mechanism to thin-film coatings is a dehydration effect. Continuous exposure to 193-nm laser pulses can dehydrate the stack; when the coating is allowed to rehydrate, this can result in stress-related damage in the form of microscopic cracking or delaminating of the film.

While it is fairly easy to produce hard, dense coatings by means of heat or ion-assisted deposition, these coatings typically exhibit high absorption and substandard reflectance at 193 nm. So the objective is to establish the ideal combination of materials and processes that yield high reflectance at 193 nm with exceptional resistance to dehydration-induced damage. One approach was to develop a hybrid coating incorporating up to four different coating materials selected to balance stress and improve resistance to laser damage. In addition, the materials were chosen for low absorption at 193 nm to yield acceptable reflectance levels for the application.

To test the effectiveness of hybrid coating designs, a nitrogen-purged oven was set up to simulate the dehydration effects of high-repetition-rate operation. When a candidate coating design achieved acceptable reflectance at 193 nm, it was baked in the nitrogen environment for a 24-hour cycle. After a gradual cool-down period, coatings were removed and allowed to acclimate, or rehydrate. Damage usually became apparent within 30 minutes and was typically in the form of microscopic cracks in the coating. When a coating passed the initial bake test, it was then cycled through the same test several additional times to confirm resistance to dehydration-induced damage.

The hybrid approach has enabled production of coatings with 97% reflectance and exceptional resistance to laser-induced dehydration damage. Samples delivered for laser-damage testing have surpassed more than 8 billion laser pulses at 5 mJ/cm2 and a 4-kHz repetition rate with no deterioration or loss of reflectance.

Research continues in the effort to improve 193-nm high-performance optics and coatings. Promising new materials, improvements to optical-monitoring-system technology, and advances in thin-film metrology are just a few areas under development.

MICHAEL CASE is vice president of research and development at Acton Research Corporation, 530 Main St., Acton, MA 01720; email: mike.case@acton-research.com.

More in Optics
Lasers & Sources
LASER World of PHOTONICS 2019: Inspiring