search

jdhuang-csm / hybrid-drt

Probabilistic electrochemical analysis with the distribution of relaxation times (DRT)
BSD 3-Clause "New" or "Revised" License
9 stars 2 forks source link

Possibility to extract frequency range information for the indentified peaks from DRT? #7

Open jianzuo opened 3 weeks ago

jianzuo commented 3 weeks ago

Hi, Thank you for checking my question!

When applying hybrid-drt to analyze impedance spectra for a practical electrochemical system, the identified peaks are associated with different physical processes. From the DRT inversion results, we can get information such as the time constants of those peaks.

We often see from the literature, talking about the frequency range of different processes such as ``charge transfer in fuel electrode between 1–10 kHz'' from https://doi.org/10.1016/j.jpowsour.2022.232352.

Could we also obtain the frequency range of the peak, rather than just the time constant?

Thank you!

jdhuang-csm commented 3 weeks ago

Hello,

Yes - luckily this is very easy! You can convert time constants to frequency with the simple formula: f=1/(2pitau).

jianzuo commented 3 weeks ago

Hi, Thanks for the quick response. I understand what you mean. But what I mean from the question is to obtain the frequency range of a peak, namely (f_min, f_max) not a characteristic frequency f which somehow lies in the middle of this range, as you have answered. Not sure it is clear, if not, I will think about it and try to put a clearer description.

Thank you!

jdhuang-csm commented 3 weeks ago

I see, thanks for the clarification.

As far as I understand, when authors discuss the "frequency range" of a particular process, they mean the range in which the characteristic frequency most often appears. Thus, I still think that the characteristic frequency of the peak is the most informative and relevant.

However, if you really need to identify a range of frequencies for a single process in a single spectrum, I can think of a couple ways to approach this. First, you could simply look at the DRT, but convert the timescale axis to a frequency axis (or add a second x-axis with matplotlib) and observe the extent of each peak. The second option would be to determine the frequency range of the impedance response of the peak. To do this, you would first need to separate the peak of interest (either using dual_fit_eis or predict_peak_distributions), then calculate its impedance, and finally plot the imaginary impedance vs. frequency.

Let me know if this answers your question, or if you are looking for something different.

jianzuo commented 3 weeks ago

Thank you Jake! Yes, that is what I want. Thanks for your insightful comment. I am going to try the possible ways you suggested.