Dear Tingting,

EIS results are not so unexpected that you are mentioning because as you can see the GC/3L CNTs modified electrode behave as better electrocatalyst upon the ferri-/ferro- redox probe. That's why the semicircle size decreases. I advice you, first to use a +200 mV DC voltage along with the 5 mV amplitude that you are using and second (if it does not arise any problem to your experiments) use KNO3 or KCl as supporting electrolyte instead of Na2SO4, in order to retrieve better results and more well shaped Nyquist plots.

Regarding the data fitting process try the simplest and most common Randles equivalent circuit Rs(Qdl[RctW]), where Rs is the ohmic resistance of the electrolyte in the space between the working and the reference electrodes, Rct is the charge transfer resistance, which is due to electron transfer of the redox probe to the electrode, W the Warbung impedance resulting from the diffusion of the redox couple towards the electrode surface and Qdl, the double layer capacitance, which is represented by a constant phase element.

Good luck.

Dimitris

Dear Dimitris,

Thank you for your reply. I tried to fit the plot with the common equivalent circuit, but it doesn't fit well. Could you explain why should I use KNO3 and KCl instead of Na2SO4?Also use 200mV DC instead of OPC?

Thank you very much,

Tingting

Because KNO3 or KCl are stronger electrolytes.. have better  "ionic hydration number" when compared to Na2SO4 

Hello Tingting

I agree with Christian Punckt, you concentration of electrolyte and ferro/ferri are somewhat high. And also KCl is preferable to Na2SO4.

Maybe you could follow on Research Gate the discussion on the question:

"Why are electrochemcial impedance spectroscopy results not stable?" by Dan Wu.

Good luck.