Introduction Most advanced photoresists used in the manufacture of microelectronic devices rely on the principle of chemical-amplification. Rather than using light to directly cause a solubility switch, chemically amplified photoresists use light only to generate a catalytic species. The photo-generated catalyst, typically an acid, then promotes a solubility switching chemical reaction in the exposed regions of the photoresist. The resulting latent image can then be developed and transferred into the underlying substrate. While chemically amplified resists have many advantages, one inherent limitation is catalyst diffusion. That is, catalyst generated in exposed regions can potentially diffuse into unexposed regions, thus causing line-width spread or image blur. This blurring effect is an observed fact in microelectronic processing and is usually attributed to classical Fickian diffusion of the acid catalyst. Previous studies have inferred diffusion coefficients from lithographic performance (1). This method is valuable for providing workable engineering parameters, but does not provide much insight into the actual nature of the transport process. The goal of our work is to better understand the transport phenomena that results in image blur. As such, we have tried to design experiments that would allow direct measurement of diffusion coefficients without the many complicating factors inherent in a complete lithographic process. Several direct measurement techniques were tried unsuccessfully before we arrived at our current method. Our previous attempts were based on capacitance measurements, diffraction grating analysis, and atomic force microscope measurements. Our present technique for measuring acid diffusion is based on spectroscopy, either IR or fluorescence, and requires generation of trilayer polymer film stacks or “sandwiches”. The film stack consists of an acid feeder layer, an intermediate layer and a detector layer. The feeder layer contains a photoacid generator (PAG) which generates acid upon ultraviolet exposure. After exposure acid begins diffusing into the intermediate layer. Arrival of acid at the detector layer can be monitored spectroscopically and the diffusion time required to traverse through the intermediate layer can be recorded. The analyzed intermediate layer can be any acid inert polymer such as poly(ethylmethacrylate) (PEMA) or poly(4hydroxystrene) (PHS). The choice of detector layer depends upon which spectroscopic method is to be employed. For the fluorescence method, the detector layer is a polymer with an acridine based, covalently bonded fluorophore. For the IR method the detector layer is poly(tbutyloxycarbonyloxystyrene) (tBOC) which has an easily monitored acid labile protecting group. Results presented here are from the IR based method; results from the fluorescence method will be presented at a later date.

The deprotection kinetics of alicyclic polymer resist systems designed for 193 nm lithography was examined using IR and fluorescence spectroscopic techniques. A kinetic model was developed that simulates the deprotection of the resists fairly well. A new, simple, and reliable method for monitoring photoinduced acid generation in polymer films and in solutions of the kind used in 193 nm and deep-UV lithography was developed. This technique could find application in the study of diffusional processes in thin polymer films.