Dear Mengqi Duan

I will recall my previous reply on the same question found in the RG link below. When the potential parameter is underestimate , the inclined line corresponding to the low frequency region become strongly potential dependent, and with increasing the open circuit potential to a higher value than you practiced practically, you can get an increasing tendency to move towards the real axis and finally becoming a semicircle in the medium frequency region and a straight line in the low frequency region at proper potential. Finally don't worry about the performance of the TiO2 coating onTi foil, before you get the right polarization potential value for the test and results analysis. Best regards ....

https://www.researchgate.net/post/What_does_absence_of_semicircle_in_a_Nyquist_plot_EIS_mean

Thank you, Mohammed Hussein j. H. Al'Atia !

The trend to directly interpret EIS spectra using physical models based on the perception of the user, puts the physical interpretation before the actual fitting of data. This results are complex commercial applications that researcher is required to choose an ECM and that promote automated fitting algorithms. The necessity of choosing a ECM confuses the novice user and creates diverse interpretations by experienced users. Automated fitting routines, although capable, require blind trust of the user.

In contrast, I am promoting to first fit the spectra with an expandable model where the number of responses (e.g a parallel combination of series R-CPE legs) are based number of parameters the data can support. Once this is determined, then physical models can be explored. This approach is focused on how the instrument responds to a perturbation, rather user perception.

I have published articles on LinkedIn and am in the process of updating a software program called PEDRA (Parallel Electrical Debye Response Analysis) running within IGOR Pro. This approach offers a simpler, focused application with incredible graphics capability and open source code. The advantages of the PEDRA approach are: 1) simplifies choice of ECM, 2) identifies the number of responses (i.e. parameters) that the data can support, 3) used a iterative, user involved, fitting process. I also uploaded a poster I presented at the Fall Material Research Symposium, December 2024.

Thank you, Michael Maguire!