Understanding Artifacts in Impedance Spectroscopy

Four-terminal measurements of impedance spectra have long been troubled by the presence of high frequency artifacts that typically indicate unphysically large inductive behavior. We follow up on the observation of Fleig et al., that voltage and current are necessarily measured in different locations of the potentiostat circuit, and that, typically, the electrometer input is a virtual ground. In this case, the capacitance of coaxial cables that connect sample electrodes to the potentiostat provides a high frequency conduction path to ground, so that some of the current that passes through the sample bypasses the electrometer. In four-electrode measurements, this mechanism produces the observed inductive artifacts. We examine a variety of simulated samples, with calculations compared to measurements of relevant circuits, to quantitatively investigate the nature of the artifacts. Model results agree with measurements when the leakage capacitances are properly included in the circuit analyses. With understanding of the origin of the inductive artifacts, the four-electrode method can be effectively utilized, enabling a combination of two-, three- and four-electrode measurements to be used to best advantage. Using this combination of electrode configurations, temperature dependent measurements of SrTiO 3 , Y 2 O 3 -stabilized ZrO 2 , and In 2 O 3 films deposited on YSZ substrates are presented.