Characterization of an electron cyclotron-resonance-heated mirror plasma as a soft x-ray source for x-ray lithography.

Abstract

There exists an international race to reduce the linewidths of integrated circuits to the sub-micron level. The results of this race are likely to have a strong impact on both the economic and military independence of this country as our society relies more and more heavily on highly sophisticated electronic equipment in our daily lives and in the defense of the country. One of the leading technologies in attaining sub-micron linewidths is x-ray lithography. Results from the ECRIAXS computer code have suggested that a laboratory-sized electron-cyclotron-resonance-heated mirror plasma may be a viable source for x-ray lithography. This work attempts to investigate this suggestion. A foil-filtered x-ray PIN diode and an x-ray pinhole camera have been used to measure x-ray intensities from krypton and neon plasma. The spatially resolved spectral density of the negatives from the x-ray pinhole camera has been measured. A film modeling code (FILM) has been written which predicts the film response. The code includes the integrated spectral response of the film to plasma x-ray emission that has been filtered by a beryllium foil filter. The PIN diode also yields electron temperatures for the plasmas (T$\sb{\rm e} \sim$ 1.3-4.0 keV). The intensity measurements are compared to computations from the ECRIAXS code. The measurements of the spectrally integrated x-ray intensity are supplemented by the measurement of plasma parameters that can be directly compared with some of the ECRIAXS code outputs. It has been found that the plasma densities that are predicted by the code (e.g. 1.75 $\times$ 10$\sp{12}$ cm$\sp{-3}$) are significantly higher than the experimental measurements (e.g. 4.5 $\times$ 10$\sp{11}$ cm$\sp{-3}$). Modifications to the code can explain some of the discrepancies; however, the results are reliable only to within an order of magnitude.