I want to calculate ionic conductivity of my lithium containing amorphous solid electrolytes but I am kind of confused in the output values of ZFit tool. I used CPE because it gave me better fit compared to Warburg. Even if I used Warburg my question still applies: How to calculate ionic conductivity by using CPE or Warburg coefficients? I don't want to use extrapolation of the linear part or determining the diameter of the semi circle approaches because they give quite different results for my samples.
Dear sir,
In my opinion, if you are lost with the ZFit results, it is because you left the physics aside and just do mathematical fitting.
A Warburg impedance only describes mass transfer limitations in the lowest frequencies range, not a phenomenon occuring in the high frequency range, like a CPE does.
Looking at your spectra and the very high number of points that are present in the high frequency range, several thoughts and questions come to my mind...
What is the highest frequency at which your capacitive loop appear? Because, in electrochemical devices, the lower the upper limit of frequency of the capacitive loop is, the slower the corresponding phenomenon is. Moreover, the overall ohmic resistance of a battery should be very low and the ionic resistance should be much higher than the electronic one. In this case, the latter is negligible vs the former and the cloud of points that I see in the high frequency range of the spectrum should correspond to the ionic resistance. By the way, is it a cloud of points or is there a particular shape? A very small semi-circle with an angle of departure significantly lower than 90° for instance.
I am also wondering about the physical phenomenon that would justify a change in the ionic conductivity with the frequency. This change might be expected in a porous medium because the depth of the signal penetration increases as the frequency decreases but this would mean that the ionic and the electronic resistances would have the same order of magnitude. I am skeptical about this....
Another question that comes to my mind is whether the spectrum is represented in an orthonormal scale. I am wondering about this because the straight line at low frequencies does not have a 45° slope, which should be the case if this was a Warburg impedance. I know that there some specific behavior for porous media, (like the battery electrodes) but, in this case, the Warburg model is not appropriate. Unless the lower limit of measured frequency (how much is it?) is too high to see the complete formation of the 45° straight line,... So, to be able to "extrapolate the linear part" properly, you have to decrease the lower limit of frequency until which you record your spectra.
I hope this helps...
Dear Orhan Kıbrıslı ,
just a few additional comments:
1. What is the rationale behind the choice of such an electrical equivalent circuit ?
2. As far as I understand, the ionic conductivity (or resistivity) can be simply obtained by measuring the ohmic resistance, that is to say the series resistance R1 at high frequencies
3. Your plot is not orthonormal, here is a small topic that explains why it should be orthonormal: https://www.biologic.net/topics/nyquist-plots-why-use-the-orthonormal-scale/
4. To have orthonormal scale, you need to go to Graph Properties->Axis->Options and check XY1 proportional.
I also hope this helps...
Best regards,
Nicolas
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs