Search

Absorption spectra and extinction coefficient values of various optical films are evaluated by the recently developed TRACK method, which is particularly appropriate for non-ideal coatings. Results are compared to those obtained by the LID technique.

A new concept enhances the capability of photo-thermal absorption measurements with transversal probe beam guiding by overcoming drawbacks like sensitivity lack for materials with low photo-thermal response, non-transparent and/or round substrates.

Photolithography is a key technolgoy for the production of semiconductor devices. It supports the continuing trend towards higher integration density of microelectronic devices. The material used in the optics of lithography tools has to be of extremely high quality to ensure the high demand of the imaging. Due to its properties CaF2 is a material of choice for the application in lithography systems. Because of the compexity of the lithography tools single lenses or lens system modules cannot be replaced. Therefore the lens material has to last the full lifetime of the tool without major degradation. According to the roadmap for next generation of optical lithography tools, like immersion lithography, the requirements of CaF2 for radiation hardness are increasing considerably. We will present a detailed analysis of the key factors influencing the laser hardness covering the complete production chain. Some aspects of the evaluation methods for testing CaF2 laser durability will be presented.

The combination of in situ transmission and laser induced fluorescence (LIF) measurements of CaF 2 at 193 nm and 157 nm laser irradiation reveals a correlation between selected fluorescence bands and the laser transmission. At 193 nm irradiation, the fluence dependent transmission |dT/dH| of calcium fluoride shows a significant dependence on the temporal pulse shape of the laser source. Furthermore, a quantitative correlation between transmission properties and fluorescence intensity of calcium fluoride is reported for the first time in case of a LIF band at around 740 nm. Newly defined LIF detection conditions yield a remarkable increase of the sensitivity for bands with short lifetimes. Furthermore, different excitation mechanisms for the investigated fluorescence bands are found from both, fluence and pulse number dependent LIF measurements. At 157 nm irradiation, a fluence dependence of the transmission (|dT/dH|, T 0 ) is obtained which is comparable to that of 193 nm excitation. LIF investigations at 157 nm excitation reveal an increase in complexity of the spectra compared to those of 193 nm excitation. It is found that the LIF spectra at 157 nm excitation mostly consist of the same bands as for 193 nm irradiation. Some samples, however, show a LIF band vanishing relative to its intensity at 193 nm excitation or the appearance of new bands characteristic only for 157 nm excitation. From comparing two calcium fluoride samples at 193 nm and 157 nm irradiation it is assumed that the presence of a characteristic 157 nm excited LIF band at around 225 nm is responsible for a drop in transmission at 157 nm. Keywords: Laser induced fluorescence, transmission, CaF

Under 193 nm excimer laser irradiation the laser induced deflection technique (LID) is applied to investigate directly the bulk absorption α of high quality fused silica and calcium fluoride. Fused silica samples are characterized by their fluence H dependent absorption α(H). Their small signal absorption coefficients α 0 are extrapolated by an appropriate fitting model. All investigated standard samples with high H2 content fulfill the requirement for optical lithography which is determined by an α0 of less than (formula available in paper). Prolonged direct absorption measurements at relatively high fluences of 10 and 20 mJ/cm2 by the LID technique are compared to state of the art marathon durability tests for H2 poor fused silica at a H = 1.3 mJ/cm2. The very good agreement of the results demonstrates that the measurement time for durability tests of fused silica can be reduced considerably by increasing the applied fluencs H. Calcium fluoride is investigated by both, direct bulk absorption (LID) and conventional transmission measurements. A very good agreement is found by comparing the results of both experiments. For investigations at 157 nm laser irradiation a new compact LID measurement device is introduced. Calibration measurements show that the sensitivity is significantly increased compared to the previous setup. The detection limit of the new setup is estimated to α values of (formula available in paper) for calcium fluoride and fused silica, respectively.

An experimental setup was established for in situ transmission and laser induced fluorescence (LIF) measurements of CaF2 at 193 nm laser irradiation. The known rapid damage process in CaF2 upon ArF laser irradiation is shown to terminate for all tested samples within 3×104 laser pulses for the applied fluences. Furthermore, it is demonstrated that for typical application values the fluence dependent transmission (FDT) at the end of the rapid damage process is independent of the irradiation history and determined by the specific crystal quality. From the lifetimes and signal strengths of different present fluorescence bands the excitation and recording conditions for LIF investigations are derived. The results of laser induced fluorescence measurements at 193 nm excitation make evident that certain impurities or defects are responsible for the different transmission properties even of high purity CaF2 crystals. Comparing transmission and LIF data a quantitative correlation was found between selected emission bands and ArF laser stability of CaF2 material.

Luminescence measurements have been set up in order to study the interaction of UV-laser radiation with dielectric thin films. The pulsed laser excitation was carried out at 193-nm (6.4eV), the coating materials comprised wide-band-gap oxides and fluorides. Experiments show the significant optical response of single- and multilayer coatings on the low fluence excitation at sub-band-gap energy. Time- and spectrally-resolved measurements indicate characteristic emission bands of color centers in the deep-UV and vacuum-UV coating materials. An assignment of these optical transitions can be derived from the comparison with known bulk-material studies.

A compact experimental setup, based on the laser induced deflection technique (LID), measures small absorption coefficients in fused silica upon 193-nm irradiation with high sensitivity and accuracy. For that, two probe laser beams are passed numerous times through a sample and are deflected by a refractive index gradient which is generated by the power absorbed within the material. The absorption coefficient of the sample is determined by applying a comfortable and precise electrical calibration procedure. The investigation of two equivalent fused silica samples of different thickness confirmed that the setup allows to exclusively measure the bulk absorption of the material without contributions from the irradiated surfaces. Furthermore, influences of irradiation parameters like repetition rate and pulse width on the absorption coefficient of fused silica at a fixed applied energy density have been investigated. The results confirm the complexity of the absorption mechanism present in fused silica upon laser irradiation. In order to separate linear and nonlinear absorption two fused silica samples have been irradiated with different energy densities keeping the repetition rate and the pulse width constant. The results show a nonlinear dependence of the absorption coefficient on the energy density which can qualitatively be explained by the two-step absorption mechanism in fused silica.

The laser induced absorption of CaF2 caused by ArF excimer laser light has been observed at energy densities of F equals 2-30 mJ/cm2 per pulse and a repetition rate of R equals 50 Hz. The experiments show that the transmission of CaF2 samples depends on the pulse energy density. The change of the absorption coefficient with the time of irradiation can be described by an exponential model. Different experiments were performed where the energy density was increased and decreased stepwise. They prove that color centers not only are formed but also are annihilated by irradiation. Laser induced decrease of absorption was observed in all samples as soon as the energy density was decreased. Coloring and bleaching of the samples are completely reversible processes. The level of transmission depends on the energy density of the laser light and the quality of the material but not on the history of irradiation. The damage resistance of the material can be adjusted by the appropriate choice of the raw material and the process parameters. The reversibility of the laser induced absorption can be explained by a reaction equilibrium. This leads to a model where the concentration of absorbing defects depends on the current irradiation conditions. Using these equations the reversibility and the observed exponential dependence of the change of transmission with time can be explained. Assuming different dependencies of the reaction constants of coloring and bleaching on the energy density, the change of the absorption coefficient with pulse energy density can be calculated.

Large area photomask substrates of fused silica and a size of up to 9 inches by 9 inches are locally evaluated by laser induced fluorescence and H2 measurements. Using a narrow-bandwidth KrF excimer laser, pulsed UV Raman spectroscopy is applied to measure the content of molecular hydrogen as an indicator of laser durability. Due to expeditious H2 measurements, this technique allows to determine the local H2 concentrations across an entire 6 inch by 6 inch photomask substrate within one hour. ArF excimer laser induced fluorescence (LIF) measurements are applied to correlate the H2 concentration and the laser induced defect formation in fused silica. In most cases the NBOH defect density, determine by the intensity of the peak at (lambda) equals 650 nm in the LIF spectrum, shows a symmetric curve across the substrate diagonals with its lowest values in the center. The associated H2 concentrations are found to show a maximum in the center. Furthermore, pulsed UV Raman spectroscopy is used to monitor the H2 concentration during irradiation with several 105 pulses of high KrF laser fluence. At first a rise in the H2 concentration is observed. After a maximum value further irradiation leads to a drop of molecular hydrogen within the illuminated volume.