The following paper describes a design of experiments investigation of the deep reactive of pillar structures on a silicon wafer. The etched wafers would subsequently be used as masters for the fabrication of nickel mould inserts for microinjection moulding. Undercuts occur when the pillar base has a smaller cross-section than the apex of the pillar. They therefore affect tolerances of the subsequent nickel mould, its strength and its de-mouldability from the silicon form. The response measured in these experiments was the degree of undercut of micro-scale (10 μm x 10 μm x 40 μm, 5 μm x 5 μm x 40 μm and 2 μm x 2 μm x 40 μm) The literature suggests that gas pressure, platen power, gas flow rate, phase switching times and mask size can all affect the degree of undercut. After examination of this literature, and of manufacturers guidelines, three parameters were selected for experimental testing: platen power, C 4F 8 gas flow rate during the passivation phase and switching times. Switching times was found to be the only statistically significant parameter for both 10x10 μm and 5x5 μm pillars. The 2x2 μm pillars were not successfully replicated and could therefore not undergo statistical evaluation.
展开▼
