We report a study of the application of CF4 and CHF3 electron cyclotron resonance (ECR) discharges to selective etching of SiO2 over Si. Due to significant fluorocarbon film deposition for plasma operation without rf sample bias in the pressure range below 10 mTorr, rf biasing is required for etching of SiO2 and Si. The rf threshold voltage for etching is 55 V for CHF3 and 35 V for CF4 at a pressure of 1 mTorr. At 100 V rf bias, silicon dioxide etch rates were greater than ≂600 nm/min in CF4 and 450 nm/min for 1000 W plasmas at 1 mTorr pressure. A plot of the oxide etch rate vs rf bias exhibits a fluorocarbon film suppression regime at low rf voltages and an oxide sputtering regime at higher rf voltages. In the fluorocarbon suppression regime, the etch rate is primarily determined by fluorocarbon deposition which results in a thin fluorocarbon film being present on the SiO2 surface during steady-state etching. In the oxide sputtering regime, the oxide etch rate increases linearly with the ion current to the wafer and the square root of the ion energy. The etch yields decrease with increasing microwave power and decreasing pressure and are in the range 0.5–2 atoms per incoming ion. The silicon etch rate is much lower in CHF3 than in CF4, which translates into better SiO2-to-Si etch selectivity in CHF3 (≂15) than in CF4 (≂5). The lower Si etch rate in CHF3 is due to a greater thickness of the fluorocarbon film present on the silicon surface during steady-state etching. The fluorocarbon film thickness is ≂5.5 nm in CHF3 as compared to ≂2.5 nm in a CF4 discharge (at a rf bias of 100 V). The oxide surface is free of fluorocarbon film for the same conditions. The etch depth of ≂2.5 μm deep contact holes etched using 1 mTorr CHF3 plasmas into photoresist patterned SiO2 was measured by scanning electron microscopy as a function of the feature width. The etch depth decreased by ≂10% as the feature size was reduced from 1.3 to 0.6 μm.

Resonance absorption at 130 nm was used to measure absolute oxygen atom concentrations in O2-containing distributed electron cyclotron resonance plasmas. The dissociation fraction [O]/[O2] in pure O2 plasmas (1–6 mTorr) was in the range 0.01–0.06, but was significantly increased by the addition of SF6, N2 or Kr. At 2 mTorr total pressure a maximum [O]/[O2] of 0.3 was observed for 10% SF6 added. The results were compared to those obtained by optical emission actinometry measurements. The quantity I0 (844 nm)/IAr (750 nm) (with 0.1 mTorr Ar added) was poorly correlated with [O] but well correlated with [O2]. This suggests that, for dissociation fractions lower than 0.1, dissociative excitation, O2+e→O*(3p 3P)+O, is the most important mechanism for the production of 844 nm emission.

This work focuses on the etching of different porous methylsilsesquioxane materials (spin on SiOCH, k=2.2) with different porosity (30%, 40% and 50%) in fluorocarbon-based plasmas (CF4∕Ar). The etching of these materials is performed on blanket wafers in a magnetically enhanced reactive ion etcher. The surface and bulk modification after partial etching are studied using different surface analysis techniques such as quasi-in-situ x-ray photoelectron spectroscopy (XPS), infrared spectroscopy (FTIR), and attenuated total reflection spectroscopy (FTIR-ATR). Similar to nonporous SiOCH materials, a decrease in etch rate of porous SiOCH films is observed with either increasing Ar dilution or polymerizing gas addition (CH2F2), which can lead in this last case to an etch stop phenomenon. The etch rate increases with higher porosity in the SiOCH film, since less material per unit thickness needs to be removed as the porosity increases. The XPS results show that a fluorocarbon layer is formed at the surface of the porous material and complementary angle resolved XPS analyses reveal that fluorocarbon species diffuse through the pores into the material. After partial etching, FTIR and ATR analyses reveal a carbon depletion in the remaining film, which indicates that the porous material is altered during plasma exposure. The film degradation is more important as the porosity increases. The etch rate evolution and film degradation are discussed and interpreted in terms of etching mechanisms and plasma surface interaction.