No abstract

In this paper, we want to optimize photo-resist coating recipe by using orthogonal experiment and statistical regression analysis. We combined two major factors influenced coating performance -"wafer coating time" & "wafer spinning speed" in TRACK unit with IDI pump dispense recipe. Eventually, IDI pump dispense recipe contained of four steps dispense time and dispense pressure. By using orthogonal experiment and statistical regression method we obtained a set of process parameters which have good predicted performance values. Based on those data, we could apply it on different viscosity type resist to get good coating performance as decreasing resist dispense volume from 3 cc/pcs to 1.6 cc/pcs. Results showed the performance of CD (Critical Dimension) and CP (Chip Probe) yield was comparable to original 3 cc/pcs coating recipe condition. Actually, we could optimize coating recipe to reach resist usage decreasing purpose.