Richard A. Ferguson

EECS Department
University of California, Berkeley
Technical Report No. UCB/ERL M91/78
September 1991

http://www2.eecs.berkeley.edu/Pubs/TechRpts/1991/ERL-91-78.pdf

A general and comprehensive methodology has been developed for the characterization, modeling, and simulation of advanced technologies and complex resist materials for optical lithography. The foundation of this methodology is the new lithography simulation program, SAMPLE-ARK, which simulates reaction kinetics and diffusion, and their effect upon chemical species concentrations within the resist, during post-exposure processing. The implementation of fundamental mechanisms such as multiple chemical reactions, simultaneous reaction-diffusion, concentration-dependent diffusion, diffusion from outside sources, and multiple species dissolution rate expressions has resulted in a general purpose line-edge profile simulator that has demonstrated the capability of simulating a wide range of complex resist technologies including image reversal, chemical amplification, and silylation processes.

In order to develop mechanistic resist models for simulation in SAMPLE-ARK, material characterization techniques have been evaluated and refined for monitoring the resist behavior during the exposure, post-exposure bake, and development steps. These techniques include measurements of optical transmission, FTIR spectroscopy, and interferometry. New modeling software has been written to facilitate the conversion of experimental data to mechanistic models. These programs range from quantitative FTIR analysis tools to parameter extraction routines for fitting kinetic models to exposure and bake data.

The validity of this methodology has been confirmed through the systematic application to two state-of-the-art deep-UV chemical amplification resists. Predictive models as well as an increased understanding of the important factors that affect the performance of these resists have resulted. For Shipley SNR 248, an acid hardening resist, monitoring of the crosslinking reaction with FTIR spectroscopy lead to the development of a kinetic post-exposure bake model that required an acid loss mechanism to account for quenching of the crosslinking reaction. The derivation of a dissolution rate expression based upon crosslinking-induced molecular weight variations accurately described the development data. For an AT&T t-BOC resist, the resist behavior depended strongly upon chemical composition. Use of an onium salt acid generator minimized sensitivity to the bake conditions as a result of acid loss during the bake. Increasing the loading of either the onium salt or the tosylate acid generator provided improved resist contrast.

Advisor: Andrew R. Neureuther

BibTeX citation:

@phdthesis{Ferguson:M91/78,
    Author = {Ferguson, Richard A.},
    Title = {Modeling and Simulation of Reaction Kinetics in Advanced Resist Processes for Optical Lithograpy},
    School = {EECS Department, University of California, Berkeley},
    Year = {1991},
    Month = {Sep},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/1991/1808.html},
    Number = {UCB/ERL M91/78},
    Abstract = {A general and comprehensive methodology has been developed for the
characterization, modeling, and simulation of advanced technologies
and complex resist materials for optical lithography. The foundation
of this methodology is the new lithography simulation program,
SAMPLE-ARK, which simulates reaction kinetics and diffusion,
and their effect upon chemical species concentrations within
the resist, during post-exposure processing. The implementation
of fundamental mechanisms such as multiple chemical reactions, 
simultaneous reaction-diffusion, concentration-dependent diffusion,
diffusion from outside sources, and multiple species dissolution
rate expressions has resulted in a general purpose line-edge profile
simulator that has demonstrated the capability of simulating a
wide range of complex resist technologies including image reversal,
chemical amplification, and silylation processes. 

In order to develop mechanistic resist models for simulation 
in SAMPLE-ARK, material characterization techniques have been
evaluated and refined for monitoring the resist behavior during
the exposure, post-exposure bake, and development steps. These 
techniques include measurements of optical transmission, FTIR
spectroscopy, and interferometry. New modeling software has been
written to facilitate the conversion of experimental data to
mechanistic models. These programs range from quantitative FTIR
analysis tools to parameter extraction routines for fitting kinetic
models to exposure and bake data.

The validity of this methodology has been confirmed through the
systematic application to two state-of-the-art deep-UV chemical
amplification resists. Predictive models as well as an increased
understanding of the important factors that affect the performance of
these resists have resulted. For Shipley SNR 248, an acid hardening
resist, monitoring of the crosslinking reaction with FTIR spectroscopy
lead to the development of a kinetic post-exposure bake model
that required an acid loss mechanism to account for quenching of
the crosslinking reaction. The derivation of a dissolution rate
expression based upon crosslinking-induced molecular weight
variations accurately described the development data. For an
AT&T t-BOC resist, the resist behavior depended strongly upon
chemical composition. Use of an onium salt acid generator minimized
sensitivity to the bake conditions as a result of acid loss during
the bake. Increasing the loading of either the onium salt or the
tosylate acid generator provided improved resist contrast.}
}

EndNote citation:

%0 Thesis
%A Ferguson, Richard A.
%T Modeling and Simulation of Reaction Kinetics in Advanced Resist Processes for Optical Lithograpy
%I EECS Department, University of California, Berkeley
%D 1991
%@ UCB/ERL M91/78
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/1991/1808.html
%F Ferguson:M91/78