Quality-Indicator-Based Preprocessing for the Distribution of Relaxation Times Method

Abstract

The deconvolution of impedance data to their respective distribution of relaxation times (DRT) is accompanied by an increase in the resolution of the frequency domain and is, thus, advantageous. Yet, most of the techniques used for computing DRT spectra from experimental data are valid only for impedance spectra that solely show polarization processes. Therefore, most experimental electrochemical impedance spectra must be corrected from perturbing inductive, capacitive, or diffusive contributions. This preprocessing is time consuming, and its quality has an enormous impact on the obtained DRT spectra. It also prohibits using the DRT method within an automated impedance evaluation. In this work, we propose a software-aided approach for deriving optimal preprocessing. For this purpose, we first introduce a quality indicator that reflects the quality of a certain preprocessing. Based on this quality indicator, we furthermore introduce a new batch fitting approach for deriving the optimal equivalent electrical circuit within the preprocessing of experimental impedance spectra. Besides, the batch fitting displays the optimal frequency range for modeling the unwanted impedance fractions. Based on this information, the considered impedance data can finally be optimally corrected from any non-polarization processes. The new methods and their application are tested for simulated data and experimental impedance data of a lithium-ion cell.