Effects of Resist Thickness and Substrate Reflectance on Critical Dimension Bias of Isolated-Dense Pattern

Abstract

The critical dimension (CD) uniformity of an isolated-dense pattern has been the most critical issue in optical lithography below the wavelength of the illumination source and much effort has been exerted to improve it. The CD difference between isolated and dense patterns (ID bias), which is caused by proximity effects, is known to be one of the key factors leading to the deterioration of CD uniformity. Because ID bias is such a complicated phenomenon and affected by many factors, it would be very difficult to control it even though we understand all the factors that lead to it. Much research on ID bias has been conducted; however, the effect of resist thickness on ID bias has not yet been clearly studied. From our investigations, we have investigated the effects of resist thickness on ID bias and found that ID bias is strongly affected by resist thickness. ID bias changes periodically with resist thickness and the minimum ID bias is obtained by controlling the resist thickness for a specified pattern. We also found that the antireflective coating (ARC) process is an effective method for controlling ID bias.

Swing effects in alternating phase shift mask lithography: Implications of low σ illumination

“…Increased depth-of-focus ͑DOF͒ is the biggest advantage of using alt PSM; however, it very much depends on the partial coherence factor of the exposing radiation-a measure of the range of angles used for illuminating the reticle ͑also known as degree of partial coherence, the pupil filling function, or just the partial coherence͒. [9][10][11][12][13][14][15][16] However, none of them uses alt PSM. Therefore to make good use of alt PSM, one has to use the lowest possible , which is ϳ0.3 in the state-ofthe-art lithography projection tools.…”