Hi,
i carry out a galvanostatic EIS measurament on a SOFC cell of 16.48 cm^2 active area with a Kikusui in parallel with a Reference 3000 Gamry device.
I use DC current of 3 Amps and AC amplitude of 0.1 Amps
The results of Nyquist plot and K-K test are in attach.
How can i improve the quality of data mainly in high frequency range??
suggestions?
Thank you
Best regards
Bruno Conti
I would highly recommend averaging over more cycles and using a delay of at least a quarter cycle (several cycles at higher frequency) to get out any cyclic transients. I typically reduce the cycles that I average over as I go down in frequency to reduce the segment time for the EIS experiment. The form in which these adjustments are available vary by potentiostat. I have utilized both of them with a Bio-Logic potentiostat, but I am not too sure how averaging and delay adjustments are implemented on a Gamry, but I am guessing options are available.
I actually doubt you need more cycles (though if you would like to try the easiest way to do that is to switch from fast to norm or norm to low noise...and I can help you to customize that further if necessary). From what I can see the primary issue seems to be that the Kikusui is too slow to get data past 10kHz and maybe not even past 1 kHz. If you're using the Reference 3000, and you are only at 3 A DC you can adjust your signal down a little bit (e.g. 2.85 A DC, and 0.1 A AC) to do this without the electronic load.
Considering the impdedance you are getting back isn't actually that low (~15 mOhm, or 250 mOhm*cm^2...the Ref3000 can take you down below 0.1 mOhm) you also probably do not need that large of an AC current to get an accurate measurement. So I would recommend taking it down an order of magnitude which will let you get very close to that 3A level with the Reference 3000 alone...and you may be able to go down 2 orders of magnitude, in which case you can use the full 3A as there is a little head space built in.
For a better, more detailed response, please post the raw data file (*.dta) or email it to Gamry support.
If I understand correctly, you put parallel a DC load and an impedance analyzer? If so, you are also measuring the (finite) impedance of the load.
In "Impedance Spectroscopy Theory, Experiment, and Applications" by Barsoukov/Macdonald, it is described how to connect an external load with an impedance analyzer (Figure 4.1.20, original by Jensen/Poulsen).
Can't find the original at the moment, sorry about that.
Best,
Dino
Dino:
That is true, but at lower frequencies (exactly what that means varies with the load) the impedance of the electronic load is >> fuel cell, so the measured impedance in this setup tends to be dominated by the fuel cell. At higher frequencies, this falls apart. For most electronic loads this happens in the kHz range (though some can be below 1 kHz) and the best loads still tend to limit you to under 20 kHz. The good news is that most of the relevant impedance data for [most] fuel cells is below 1 kHz.
Jacob,
Yes, I agree with your comment on the loads.
However, especially SOFCs show relevant processes above 1 kHz.
Well, apart from that, large cells require large absolute currents (amplitudes and DC) and the inductive effects are much worse so the restrictions come from both influence of the load and the wiring, making it very difficult to measure beyond 1...10 kHz for the cell size, Bruno is describing.
I did not mean to be picky here, but I think for the cell type he is describing, 1 kHz as maximum frequency is not enough.
Best,
Dino
For the test i carry out the range of frequencies has to be from 0.01 Hz to 10000 Hz, if i use Gamry device with a booster for current of 3 Amps of DC current all it is ok. The problem is that we want to use the electronic load instead of booster
- Badges
- Science topic
- Similar topics
- Chemistry