Pseudo actinometry for the measurement of reactive species density.

Abstract

To compensate for the effect of plasma parameters in the signal of optical emission spectroscopy, we should normalize the emission intensity from the species against that of an inert gas (i.e., the actinometer). In many plasma processes in the semiconductor industry, plasma etching without using a neutral gas for the actinometer has become popular to achieve better critical dimension uniformity. We propose "pseudo actinometry" for normalization in the absence of an inert gas in the process plasma. Based on the theory of optical actinometry, we define a correction factor as the ratio of the emission intensity to the number density of the inert gas. As we reduced the density of the inert gas, we experimentally determined the correction factor by taking its convergence when the concentration of the inert gas was zero. As proof of concept, we applied pseudo actinometry to measure the density distribution of atomic chlorine in a photomask etching process without an inert gas. By comparing the distributions of chlorine radicals and the etch rate as measured by an ellipsometer, we calculated the correlation coefficient between the distributions. The correlation coefficient rapidly decreased to 0.60 when we used the correction factor determined at a flow rate for the inert gas of 20 standard cubic centimeters per minute at STP. By using pseudo actinometry, we successfully determined the distribution of chlorine radicals with a correlation coefficient of 0.98 in a plasma etching process without an inert gas.