Abstract

The microelectronics industry has demonstrated the need for device structure characterization during photomask processing since the line quality on a lithographic mask is critical to the eventual device quality and reliability. Applying laser scatterometry as a characterization technique for measuring line edge roughness on photomasks partially fulfills this need. A novel, two dimensional scatterometer consisting of a frosted dome intercepts the scattered light from the illuminated sample. A CCD camera images the dome's surface, and the images are stored on a computer. A unique image combining algorithm was developed and implemented, increasing the dynamic range of the recorded scattered light image. In addition, a computer model was developed to create two dimensional diffraction grating images with line edge roughness. Taking the FFT of the image produced a simulation of the far field scatter pattern. A neural network was trained on a set of simulated grating scatter data. The neural network performed very well when verified against further simulated data with line edge roughness. Measurement of surface roughness on etched phase shifting masks (PSM) was performed by using angle resolved scatterometry. The power spectral density (PSD) function was obtained, and a rms roughness was calculated. Reactive ion etching (RIE) of quartz photomasks resulted in increased surface scatter rendering them unsuitable for PSM use. A comparison of surface roughness values is presented for reactive ion etched and neutral ion beam milled photomask surfaces.

© 1992 Optical Society of America