A technique is described that uses a gas-phase species to mitigate the oxidation of a Mo/Si multilayer optic caused by either extreme UV (EUV) or electron-induced dissociation of adsorbed water vapor. It is found that introduction of ethanol (EtOH) into a water-rich gas-phase environment inhibits oxidation of the outermost Si layer of the Mo/Si EUV reflective coating. Auger electron spectroscopy, sputter Auger depth profiling, EUV reflectivity, and photocurrent measurements are presented that reveal the EUV/water- and electron/water-derived optic oxidation can be suppressed at the water partial pressures used in the tests (∼2×10−7–2×10−5 Torr). The ethanol appears to function differently in two time regimes. At early times, ethanol decomposes on the optic surface, providing reactive carbon atoms that scavenge reactive oxygen atoms before they can oxidize the outermost Si layer. At later times, the reactive carbon atoms form a thin (∼5 Å), possibly self-limited, graphitic layer that inhibits water adsorption on the optic surface.

Multilayer lifetime has emerged as one of the major issues for the commercialization of extreme-ultraviolet lithography (EUVL). We describe the performance of an oxidation-resistant capping layer of Ru atop multilayers that results in a reflectivity above 69% at 13.2 nm, which is suitable for EUVL projection optics and has been tested with accelerated electron-beam and extreme-ultraviolet (EUV) light in a water-vapor environment. Based on accelerated exposure results, we calculated multilayer lifetimes for all reflective mirrors in a typical commercial EUVL tool and concluded that Ru-capped multilayers have approximately 40x longer lifetimes than Si-capped multilayers, which translates to 3 months to many years, depending on the mirror dose.

A technique is described that uses radiation and a gas-phase species to produce a protective carbon coating on extreme ultraviolet (EUV) optics. A specific example is given in which a ∼5 Å carbon coating is deposited on EUV Mo/Si optics via coexposure to radiation (EUV photons, electrons) and ethanol vapor. Auger electron spectroscopy, sputter Auger depth profiling, and EUV reflectivity measurements are presented that suggest a carbon coating that is substantially void free and protects the optic from water-induced oxidation at the water partial pressures used in the tests (∼2×10−7 Torr). The coating is also resistant to atmospheric degradation, and to gasification by the combination of electrons and molecular oxygen. The protective coating reduces the relative reflectivity (ΔR/R0) of an optic by ∼0.5%.

Carbon contamination removal was investigated using remote RF-O-2, RF-H-2, and atomic hydrogen experiments. Samples consisted of silicon wafers coated with 100 A sputtered carbon, as well as bare Si-capped Mo/Si optics. Samples were exposed to atomic hydrogen or RF plasma discharges at 100 W, 200 W, and 300 W. Carbon removal rate, optic oxidation rate, at-wavelength (13.4 nm) peak reflectance, and optic surface roughness were characterized. Data show that RF- O-2 removes carbon at a rate approximately 6 times faster RF- H-2 for a given discharge power. However, both cleaning techniques induce Mo/Si optic degradation through the loss of reflectivity associated with surface oxide growth for RF-O-2 and an unknown mechanism with hydrogen cleaning. Atomic hydrogen cleaning shows carbon removal rates sufficient for use as an in-situ cleaning strategy for EUVoptics with less risk of optic degradation from overexposures than RF-discharge cleaning. While hydrogen cleaning (RF and atomic) of EUV optics has proven effective in carbon removal, attempts to dissociate hydrogen in co-exposures with EUV radiation have resulted in no detectable removal of carbon contamination.

The performance of Mo/Si multilayer mirrors (MLMs) used to reflect UV (EUV) radiation in an EUV + hydrocarbon (NC) vapor environment can be improved by optimizing the silicon capping layer thickness on the MLM in order to minimize the initial buildup of carbon on MLMs. Carbon buildup is undesirable since it can absorb EUV radiation and reduce MLM reflectivity. A set of Mo/Si MLMs deposited on Si wafers was fabricated such that each MLM had a different Si capping layer thickness ranging form 2 nm to 7 nm. Samples from each MLM wafer were exposed to a combination of EUV light + (HC) vapors at the Advanced Light Source (ALS) synchrotron in order to determine if the Si capping layer thickness affected the carbon buildup on the MLMs. It was found that the capping layer thickness had a major influence on this 'carbonizing' tendency, with the 3 nm layer thickness providing the best initial resistance to carbonizing and accompanying EUV reflectivity loss in the MLM. The Si capping layer thickness deposited on a typical EUV optic is 4.3 nm. Measurements of the absolute reflectivities performed on the Calibration and Standards beamline at the ALS indicated the EUV reflectivity of the 3 nm-capped MLM was actually slightly higher than that of the normal, 4 nm Si-capped sample. These results show that he use of a 3 nm capping layer represents an improvement over the 4 nm layer since the 3 nm has both a higher absolute reflectivity and better initial resistance to carbon buildup. The results also support the general concept of minimizing the electric field intensity at the MLM surface to minimize photoelectron production and, correspondingly, carbon buildup in a EUV + HC vapor environment.

Carbon deposition and removal experiments on Mo/Si multilayer mirror (MLM) samples were performed using extreme ultraviolet (EUV) light on Beamline 12.0.1.2 of the Advanced Light Source, Lawrence Berkeley National Laboratory (LBNL). Carbon (C) was deposited onto Mo/Si multilayer mirror (MLM) samples when hydrocarbon vapors were intentionally introduced into the MLM test chamber in the presence of EUV at 13.44 nm (92.3eV). The carbon deposits so formed were removed by molecular oxygen + EUV. The MLM reflectivities and photoemission were measured in-situ during these carbon deposition and cleaning procedures. Auger Electron Spectroscopy (AES) sputter-through profiling of the samples was performed after experimental runs to help determine C layer thickness and the near-surface compositional-depth profiles of all samples studied. EUV powers were varied from similar to0.2 mW/mm(2) to 3 MW/mm(2) (at 13.44 nm) during both deposition and cleaning experiments and the oxygen pressure ranged from similar to5 x 10(-5) to 5 x 10(-4) Torr during the cleaning experiments. C deposition rates as high as similar to8 nm/hr were observed, while cleaning rates as high as similar to5 nm/hr could be achieved when the highest oxygen pressures were used. A limited set of experiments involving intentional oxygen-only exposure of the MLM samples showed that slow oxidation of the MLM surface could occur.

The first environmental data from the Engineering Test Stand (ETS) has been collected. Excellent control of high-mass hydrocarbons has been observed. This control is a result of extensive outgas testing of components and materials, vacuum compatible design of the ETS, careful cleaning of parts and pre-baking of cables and sub assemblies where possible, and clean assembly procedures. As a result of the hydrocarbon control, the residual ETS vacuum environment is rich in water vapor. Analysis of witness plate data indicates that the ETS environment does not pose a contamination risk to the optics in the absence of EUV irradiation. However, with EUV exposure, the water rich environment can lead to EUV-induced water oxidation of the Si-terminated Mo/Si optics. Added ethanol can prevent optic oxidation, allowing carbon growth via EUV "cracking" of low-level residual hydrocarbons to occur. The EUV environmental issues are understood, mitigation approaches have been validated, and EUV optic contamination appears to be manageable.

Extreme Ultraviolet Lithography (EUVL) is a candidate for future application by the semiconductor industry in the production of sub-100 nm feature sizes in integrated circuits. Using multilayer reflective coatings optimized at wavelengths ranging from Ii to 14 nm, EUVL represents a potential successor to currently existing optical lithography techniques. In order to assess lifetimes of the multilayer coatings under realistic conditions, a series of radiation stability tests has been performed. In each run a dose of EW radiation equivalent to several months of lithographic operation was applied to Mo/Si and Mo/Be multilayer coatings within a few days. Depending on the residual gas concentration in the vacuum environment, surface deposition of carbon during the exposure lead to losses in the multilayer reflectivity. However, in none of the experimental runs was structural damage within the bulk of the multilayers observed. Mo/Si multilayer coatings recovered their full original reflectivity after removal of the carbon layer by an ozone cleaning method. Auger depth profiling on Mo/Be multilayers indicate that carbon penetrated into the Be top layer during illumination with high doses of EW radiation. Subsequent ozone cleaning fully removed the carbon, but revealed enhanced oxidation of the area illuminated, which led to an irreversible loss in reflectance on the order of 1%.

The spectral response and outgassing characteristics of new, low-voltage phosphors for application in field-emission flat-panel displays, are presented. A variety of tested phosphor materials include combustion synthesized powders and RF diode or magnetron sputtered thin films. These cathodoluminescent materials are tested with e-beam excitation at currents up to 50 /spl mu/A within the 200-2000V (eg. "low-voltage") and 3-8 kV (eg. "medium voltage") ranges. The spectral coordinates are reasonable compared to industrial low-voltage P22 phosphors. Phosphor outgassing is measured with a residual gas analyzer. We find that levels of outgassing stabilize to low values after the first few hours of excitation. The desorption rates measured for powder phosphor layers with different. thicknesses are compared to desorption from thin films.

The development of a laboratory EUV lithography tool based on a laser plasma source, a 10x Schwarzschild camera, and a magnetically levitated wafer stage is presented. Interferometric measurements of the camera aberrations are incorporated into physical-optics simulations to estimate the EUV imaging performance of the camera. Experimental results demonstrate the successful matching of five multilayer reflecting surfaces, coated to specification for a wide range of figure and incidence angle requirements. High-resolution, 10x-reduction images of a reflection mask are shown.

Laser plasma sources convert 1 - 2% of the incident laser energy into soft x rays that can be used in multilayer-based reflective systems. These sources are useful in the laboratory for development of soft-x-ray projection lithography (SXPL). In the commercialization of SXPL technology, the laser plasma source offers the advantages of modularity and lower cost, when compared to the alternative synchrotron source. The characteristics of the source define requirements for other system components. The condensing system, which collects radiation from the plasma source and directs it onto the mask, must be designed to match the source size and the aperture of the imaging objective. The first surface of the condenser is subject to damage by unwanted debris from the plasma source. This paper discusses several of the major issues involved in using laser plasma sources in SXPL experiments and provides examples of experimental solutions. Simulated and actual soft-x-ray images are shown.

The performance of a new 10×-reduction Schwarzschild system for projection imaging at 13.4 nm wavelength is reported. The optical design is optimized to achieve 0.1 μm resolution over a 0.4 mm image field of view, an increase in area of a factor of 100 over previous designs. An offset aperture, located on the convex primary, defines an unobscured 0.08 numerical aperture. The system is illuminated using extreme ultraviolet (EUV) radiation emitted from a laser plasma source and collected by an ellipsoidal condenser. A 45° turning mirror is used to relay the collected EUV radiation onto a near-normal reflecting mask. Multiple sets of primary and secondary elements were fabricated, matched, and clocked to minimize the effects of small figure errors on imaging performance. Optical metrology indicates that the wave-front error within the subaperture used is within a factor of 2 of the design value. Images recorded in poly(methyl methacrylate) and ZEP 520 (Nippon Zeon) resists reveal good imaging fidelity over much of the 0.4 mm field with equal line/space gratings being resolved to 0.1 μm.

A 10× reduction Schwarzschild system is being developed for projection imaging at 13 nm wavelength. The optical design is optimized over a 0.4 mm field of view in the image field using 0.1 µm design rules. An off-set aperture, located on the convex primary, defines an unobscured 0.08 numerical aperture. Multiple primary and secondary elements were fabricated, matched and clocked to minimized the effects of small figure errors on imaging performance. The system is illuminated using soft x-rays emitted from a laser plasma source and collected by an ellipsoidal condenser. A 45° turning mirror is required to direct the collected x-rays onto a near-normal reflecting mask. Precise matching of the Mo/Si multilayer coatings is required on all five reflective surfaces in the system. Incidence angles vary on the Schwarzschild primary as a function of radius from the optical axis, requiring a tapered multilayer coating to maintain peak reflection at the operating wavelength.

Soft-x-ray projection imaging is demonstrated by the use of 14-nm radiation from a laser plasma source and a single-surface multilayer-coated ellipsoidal condenser. Aberrations in the condenser and the Schwarzschild imaging objective are characterized and correlated with imaging performance. A new Schwarzschild housing, designed for improved alignment stability, is described.

High resolution soft-x-ray projection images are recorded by using a laser plasma source (LPS) and multilayer coated optics. Soft x-rays are generated by focusing a KrF excimer laser beam on a moving gold target. The resulting 15-eV plasma emits an intense burst of soft x-rays from a spot 150 μm in diameter. The soft x-rays are collected by an off-axis ellipsoidal condenser to illuminate a transmission mask composed of a germanium absorber supported by a silicon membrane. A magnified image of the LPS is projected by the condenser into the entrance pupil of the imaging objective to provide Kohler illumination of the mask. The imaging objective is a 20× reduction Schwarzschild having a numerical aperture of 0.08 and operating at a wavelength of 14 nm. Features as small as 0.1 μm have been recorded in polymethylmethacrylate (PMMA). A second-generation system is being fabricated based on a 10× Schwarzschild objective and a near-on-axis ellipsoidal condenser. This system is designed to image a reflecting mask onto a resist-coated wafer.

Projection imaging experiments are described. A Schwarzschild objective is illuminated with 14-nm radiation using an ellipsoidal condenser and a laser plasma source.

The Advanced Limiter Test-II (ALT-II) is a large area toroidal pump limiter in the TEXTOR tokamak. Discrete target plates located in channels at eight toroidal positions behind the main limiter surface neutralize a portion of the plasma efflux from the core. The resulting gas is exhausted by eight external pumps. The primary experimental goals of ALT-II are aimed at power loading studies and plasma density control during long pulse (4 s), high power (6 MW) tokamak discharges. It is found that both the power and the plasma flow to the limiter are asymmetric and depend on line density. Peak neutral pressures of 0.8 m torr and removal rates of up to 0.15 torr⋅L/s per pump station are achieved in the Ohmic phase. The projected exhaust efficiency of ALT-II with full pumping is 5-10%. During ion cyclotron resonance frequency heating, the particle removal rate exceeds 0.4 torr⋅L/s per blade, and the exhaust efficiency is 4-5% for power levels up to 2.6 MW.