 |
|
|
|
|
 |
 |
|
RESOURCES > EIS > FITTING CIRCUITS Once you have taken the EIS data, the next step is to figure out what it means. Often, that means searching for an equivalent circuit model that fits the data.
All curve-fitting software should report some measure of the "goodness of fit." Often this is the chi-squared parameter ( X2 ) or a value related to it. Boukamp makes the recommendation that the value of X2 should decrease by tenfold if a new circuit element is introduced into the circuit model. The tenfold decrease provides the justification for including the new circuit element. If the inclusion of an additional circuit element does not substantially improve the goodness-of-fit (as evidenced by the decrease in the X2 value), then based on Occam's Razor, you should keep the simpler model, or continue your search for an improved one. The old joke about the ability to "fit an elephant" if you use enough parameters is all too true with impedance data. Each component added to the model should have a physical explanation. Adding components only because they make the fit look better (smaller X2) without a physical interpretation is the equivalent to "fitting an elephant."
|
|
|
|
Electrochemistry The Bookstore Tell Us ! |
One guideline that I have heard
recommended (although I can not give a reference for it) is that data over a
decade range of frequency is required to support each circuit component.
|
|
|
|
|
|
