Recently, I made a python code to fit the impedance data of perovskite solar cell I have gathered with a certain model of equivalent circuit. I used Nelder-Mead optimization but I think I need some tweaking in order to reduce errors. Anybody has idea why some EIS data could fit into the chosen circuit or not?
I see something weird on the units when comparing the plots... is the log right?
On the circuit part, it seems that you used a RC-RC circuit. The data seems that starts with a resistance (R) then some capacitance behaviour (RC in parallel). Till here (logFreq from 8 to 4) we should fit a R-RC circuit. Then fix the capacitance obtained and add a secondary RC pair to continue the fit. The extra RC may contain a constant phase element (Y) instead of a capacitor (to cover the flattened second semicircle) that could be justified by certain porosity.
So, a better fit should be obtained with R-RC-RY or R-C(R-RY) models. Adding a diffusion element (Wb) may also help, specially to model the small 45º slope between the two events (if the Y alone do not fix that). Maybe something like this R-C(R-W)-RY or like this R-C(RW-RY).
But adding extra elements is dangerous, since the different parameters will counterbalance between them and the fitting algorithm will not take "physics meaning" decisions by himself (unless we tell to do it, of course).
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python