I encountered a problem for measuring CPE exponent of a coated metal by impedance spectroscopy. The fitting results showed an equivalent circuit with two CPE. The logarithmic diagram of the imaginary part of impedance vs. frequency also has a slope of less than 1. How can I measure the CPE exponents by measuring the the slope of the curve? Which CPE exponent is related to this slope? Is this method valid for showing CPE behavior in circuits with more that single time constants?
why dont u use fitting programms like zview and zsimp instead of extrapolation of the EIS spectrum? they can measure the impedance parametera easily
Hi, thank you, I already used fitting and simulation, But I need to verify the parameters derived from EIS plots.
Dear Mehdi Tabatabaei Majd!
I agree with Mohammad Reza Derakhshandeh. Nowadays the best way for extraction different CPE elements is to use complex non-linear least squares fitting of the experimental data to theoretic circuit
Article Analysis of Impedance and Admittance Data for Solids and Liquids
There are a several freeware packages exist for this propose, e.g. LEVM by Ross Macdonald
http://jrossmacdonald.com/levmlevmw/
(In our laboratory we have developed NELM software for MatLab, which also free and could be obtained by request).
However, there are exists two another approaches for equivalent circuit identification: algebraic, see for example
http://jes.ecsdl.org/content/125/12/2062.short
and geometric, see for example
Article Nonlinear least-squares analyses of complex impedance and ad...
Unfortunately, the area of application of these approaches is very narrow. E.g. for algebraic method parameters of your equivalent circuit should have analytical representation via immittance features (zeros, poles, maximum and minimum positions etc). Moreover, both geometric and algeraic methods give better accurate result if "relaxation times" of CPEs are different significantly.
So, I advise to you to use CNLS for CPEs separation. Try to start it from different initial points in CNLS. If you anyway want to use geometric methods, please, attach the graphs with your data. It help to answer your question more correctly.
Sincerely,
Daniil D. Stupin
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Coating