Abstract

Plasma etch rates for SiO2 and AZ5214 photoresist have been measured from various combinations of input parameters, such as power level, gas flow rate, and etch gas using a Samco RIE-1C table- top reactive ion etcher. Etches of silicon dioxide were performed with both CHF3 and CF4 as the etchant. Gas flow rates of 5, 10, and 20 mL/min and power levels of 50 and 100 w RF were used. The data were obtained by etching each wafer for a fixed time and then measuring the remaining film thickness using a Filmetrics F20 thin film measuring system. The photoresist etch rates in O2 were obtained were obtained in a similar manner. It was observed that higher power settings produce faster etches for both oxide and photoresist. In etching silicon dioxide, lower etchant gas flow rates result in faster etches, whereas, in etching photoresist higher gas flow rates speed up etching. An empirical model that predicts etch rate with any combination of input parameters has been developed. Scanning electron microscopy was used to compare the etch profiles of wafers produced with dry etch to those made with wet etches. The dry etched wafers had much more anisotropic etch profiles than the wet etched wafers which was as expected.