Enhancement of the Lithographic Process using Supercritical Carbon Dioxide in the Development Step

Abstract

As microelectronic features reach the 45 nm-node and beyond, extreme ultraviolet (EUV) lithography and other techniques are being explored as the next generation lithographic process. The current and continuing challenges faced by these practices are the reduction of image collapse and line width roughness (LWR). This dissertation covers two techniques utilizing CO₂ in the development step with the means to reduce these challenges: a CO₂ drying method to remove the development rinse solution and a carbon dioxide compatible salt ⁄ supercritical carbon dioxide (CCS ⁄ scCO₂) direct development.

The CO₂ drying method uses scCO₂ to reduce the surface tension of the water rinse solution after development. This method has potential to reduce image collapse but not in a timely manner due to the water solubility in CO₂ being too low for chemical removal and yet too high for mechanical removal.

On the other hand, the CCS ⁄ scCO₂ direct development of standard EUV photoresists achieves reduction of both line width roughness and image collapse in high aspect ratio features. The CCS ⁄ scCO₂ one step development takes advantage of the scCO₂ low surface tension to help prevent image collapse and the plasticizing properties of CO₂ in polymers to assist in reduction of line width roughness. The CCS, a fluorinated ammonium salt, associates with the photoresist Brönsted acid groups in the unexposed regions promoting the photoresist dissolution into the scCO₂ rich phase, which results in a reverse development.

A simplified rate model and quartz crystal microbalance (QCM) rate experiments were employed to understand the kinetics and overall mechanism of photoresist dissolution into the high pressure CCS ⁄ scCO₂ solution. At 5mM CCS, the zero order photoresist removal confirmed that the photoresist phase transfer, photoresist mass transfer, or both were the rate limiting steps which was the premise used for the rate equation. Increasing temperature (35°C-50°C) at a density of 0.896 g⁄ml was found to increase the removal rate due to phase transfer limitations and followed an Arrhenius behavior (Ea = 79.0 kJ⁄mol). Increasing pressure (4000-5000 psig) at 40°C also increased the removal rate due to an increasing CO₂ solubility parameter and phase transfer coefficient, but at 50°C pressure had little effect on the removal rate where phase transfer limitations were no longer present. When the CCS concentration was in global excess of Brönsted acid groups 2400:1 at 5mM, the CCS ⁄ scCO₂ developer removed the photoresist linearly with time. At lower CCS concentrations but still in global excess of Brönsted acid groups, the photoresist removal slowed (0.5mM CCS, ˜240:1) or was prevented (0.03mM CCS, ˜15:1) due to partitioning of the CCS between the CO₂ rich phase and the film.

The fundamentals of CO₂ and CCS adsorption onto the SiO₂ substrate and CO₂ absorption into the photoresist film were also investigated using the high pressure QCM at 35°C and 50°C. Adsorption studies of scCO₂ were comparable to other flat plate geometry studies. The adsorption of CCS was found to begin at 8.0 MPa with a temperature of 35°C and at 9.4 MPa at a temperature of 50°C, where the adsorption of the CCS was driven by entropy. The absorption of CO₂ into the glassy photoresist resin was also measured with QCM and found comparable to CO₂ absorption in glassy polystyrene for 35°C and 50°C. The diffusion behavior during CO₂ absorption was found to be comparable to fluorescent molecule diffusion in CO₂ swollen glassy polystyrene.