|
Chemically amplified resists (CARs) are likely to continue to be the main resist materials platform for next generation lithography using shorter wavelength and higher energy radiation sources, such as extreme ultra-violet (EUV) and electron-beam (EB) lithography, to pattern features at and below the 32 nm technology node. As the cost of generating and manipulating high energy radiation in these techniques increases dramatically, photoacid generators (PAGs) with high sensitivities to these exposure sources are required to efficiently utilize such radiation and maintain high lithography tool throughputs. On the other hand, the high energy radiation used in current and next generation lithography tools can increasingly interact non-selectively with the PAG and polymer resin. Photoacid generation from PAG sensitization pathways involving the photoresist resin (e.g. polymer) becomes another potential route for boosting the photospeed of CARs if the PAG and matrix resin chemistry is selected properly. In this work, a fast, convenient, and material saving method which can measure the acid generation rate and yield under photolysis and radiolysis, as well as determine the efficiency of acid generation through direct PAG excitation and indirect PAG sensitization pathways has been developed. This method utilizes on-wafer ellipsometry to determine the absorption of protonated Courmarin 6 (C6) dye, which is incorporated into the polymer resin as a proton indicator. In this work, triphenylsulfonium triflate PAG is used in two different matrix resins, poly(methyl methacrylate) and poly(hydroxystyrene), to illustrate this point that matrix sensitization of the PAG can be important and can be studied using the method developed in this work. This study serves as the starting point for building the structure property relationships needed for intelligent PAG and matrix design to optimize exposure energy utilization in CARs. The validity of this new analytical method is verified by comparing our results in selected PAG-polymer combinations with results obtained from previous studies using different techniques. The potential measurement errors possible when using dye as the proton indicator in a solid-state medium, such as the probability of proton-dye contact and acid generation through dye radiation absorption and sensitization of the PAG, are also analyzed and discussed in this paper.
|
