You are here

Warning message

Attention! This event has already passed.

Advanced interpretation of electrochemical impedance spectra for the modelling of electrochemical processes

Tuesday, 18 April, 2006 - 17:00
Campus: Brussels Humanities, Sciences & Engineering campus
Laurence Pauwels
phd defence

Electrochemistry is at the heart of a multitude of industrial activities like electrowinning
and electrorefining of metals, plating, electrochemical forming and machining, etc.
Therefore, a thorough understanding of electrochemical reactions is indispensable.
Electrochemical impedance spectroscopy (EIS) is a powerful tool to elucidate the
reaction mechanism of an electrochemical process. However, the interpretation of
electrochemical impedance spectra is not straightforward and, therefore, modelling
techniques (curve-fitting techniques) are required to obtain insight in the observed
phenomena. With respect to this modelling, it was found that fundamental questions
are often left unresolved. For example, how to decide whether a fit is satisfactory? If it
is not, is this due to the poor quality of the experimental data or due to the inadequacy
of the postulated model? Therefore, the object in this work is to set up a methodology
for a sound statistical and accurate data modelling of electrochemical impedance
spectra. The strategy consists in addressing issues specific to the two key elements in
data modelling, namely the experimental data and the theoretical data.

With respect to the experimental data it is found that the intrinsic time dependency of
electrochemical processes (the electroreduction of silver thiosulphate complexes is
taken as a case study) has adverse influences on the acquisition of impedance
measurements as well as on the proper error analysis of the experimental data. By
combining a new electrode pretreatment procedure and a dedicated approach to error
analysis (the measurement model approach) it is found that unbiased impedance data
can be collected for an extended frequency region, and with a proper assessment of
the level of stochastic measurement noise. As a result, the procedure yields
experimental impedance spectra containing a higher amount of information with known
reliability. It is stressed that the knowledge of the reliability of the experimental data
enables to evaluate the statistical relevance of any subsequent regression.

In order to calculate theoretical data, it is proposed to build impedance models starting
from the basic physical-electrochemical laws. In view of treating these types of models,
a state-of-the-art numerical solver, Pirode, is evaluated by comparison with an
analytical approach. Focus is on electrode kinetics influenced by adsorption and/or
mass transfer since those phenomena are frequently encountered in industrial,
complex processes. The interest in the study of adsorption phenomena is furthermore
underlined due to the fact that the simulation of adsorption is new in Pirode. With the
current state of developments, it is found that analytical models (despite their general
lack of accuracy) are still needed to proceed to automate parameter estimation (i.e.
curve fitting). On the other hand it is demonstrated that the numerical model of Pirode
enables the simulation of theoretical data that describe more accurately the
electrochemical processes under investigation. Then again, fitting possibilities are
missing so far in Pirode. Further developments are therefore expected in the
integration of a statistically well-founded curve-fitting tool in powerful numerical solvers
like Pirode. This would enable both accurate and precise parameter estimation without
the need to elaborate analytical models.