AVS 63rd International Symposium & Exhibition | |
Plasma Science and Technology | Thursday Sessions |
Session PS2-ThM |
Session: | Plasma Processing of Challenging Materials |
Presenter: | Luke Minardi, University of California Los Angeles |
Authors: | L. Minardi, University of California Los Angeles N.D. Altieri, University of California Los Angeles E.L. Chen, University of California Los Angeles J.P. Chang, University of California Los Angeles |
Correspondent: | Click to Email |
Extreme ultraviolet (EUV) lithography is a promising candidate to replace optical lithography and extend Moore’s Law. EUV lithography requires reflective optics due to the strong absorption of EUV light by most materials. The exposed wafer area is defined by the absorbing and reflecting regions of the EUV mask. The absorber stack in the EUV mask consists of 2-10 nm TaON antireflective coating (ARC) and 50-60 nm TaN bulk absorber. The final EUV mask must have a ±1 nm absorber thickness uniformity and a mean-to-target critical dimension of 2.0 nm1,2. The etch process to pattern the EUV mask must be highly selective and anisotropic to meet the stringent requirements on mask dimension.
In this work, a generalized thermodynamic approach was used to screen viable etchants and predict selectivity. Gibbs energy minimization (GEM) was used to screen a library of halide containing etchants for efficacy on Ta-based compounds. Using GEM, Cl-based etchants indicated a high selectivity of Ta to Ta2O5. For example, at 350 K and 10-5 atm, 1 kmol of Ta(s) in the presence of 33 kmol of Cl2 forms 1 kmol of TaCl5(g) while 0.5 kmol Ta2O5(s) remains unreacted in 33 kmol of Cl2. Volatility diagrams were constructed for the Cl2-Ta and Cl2-Ta2O5 systems to compare etch product volatility at specified Cl2 pressure. For the Cl2-Ta system at a chlorine pressure of log(PCl2)=-5 atm, it is predicted that TaCl5(s) is in equilibrium with TaCl5(g) at a partial pressure of log(PTaCl5)=-6 atm. For the Cl2-Ta2O5 system at a chlorine pressure of log(PCl2)=-5 atm, it is predicted that Ta2O5(s) is in equilibrium with TaCl5(g) at a partial pressure of log(PTaCl5)=-40 atm. Using GEM and volatility diagrams it was predicted the etch rate of Ta(s) >> Ta2O5(s) in a Cl2 environment. Selectivity predictions have been tested and verified experimentally through etch rate experiments using an inductively coupled plasma etcher. Experiments conducted at 250 W power, 10 mTorr, and 20 sccm Cl2 determined selectivity to be 360 and 68 at bias powers of 0W and 10W, respectively. Although plasma processing is inherently non-equilibrium, thermodynamic prediction of product volatility is a powerful tool indicating trends in etch rate.
References:
1. B. Wu and A. Kumar, J. Vac. Sci. Technol. B 25 (6), 1743-1761 (2007)
2. J. Mathuni, J. Rau, F.-M. Kamm, G. Ruhl, Ch. Holfeld, F. Letzkus, C. Kopernik, and J. Butschke, Proc. SPIE Vol. 5504, 105-110 (2004)