Analyzing the Grain Boundary Resistance in Oxygen Ion Conductors By Static and Dynamic Impedance Measurements

Abstract

Electrochemical Impedance Spectroscopy (EIS) is a very powerful method for analyzing electrochemical systems with respect to the factors impacting conductivity, charge transfer and diffusion. A great advantage of EIS is that the system under test (SUT) can be measured in operando so that the SUT can be characterized under conditions as they prevail in a fully functional technical system – as compared to model cells where only certain properties are tested under well-defined conditions. The small-signal excitation ensures that the SUT remains in its predefined operating point. Those aspects are of great importance when trying to optimize devices for commercial application.However, some characteristics can only be assessed through dynamic measurements with macroscopic external triggers. For example, grain boundary potential barriers in oxygen ion conductors can be well characterized in steady-state by EIS. With an external bias voltage, the resistance of the grain boundaries can be decreased, and the magnitude of this decrease can also be determined very accurately by EIS. However, while the resistance shrinks, further information about how the bias impacts the grain boundaries remains elusive. Is it a direct voltage effect or does the charge carrier and dopant concentration change within and adjacent to the grain boundary? The dynamic evolution of relevant parameters during the transition when changing the bias voltage can help to provide the additional information required for a more incisive analysis.Classic EIS is not applicable for such measurements because the steady-state condition is one of the fundamental requirements for EIS. In the field of semiconductors, deep-level transient spectroscopy (DLTS) is an established technique to determine the energy levels of certain defects. Hereby, the capacity at high frequencies is monitored continuously during voltage or light pulses at different temperatures. The temperature at which the pulses impact the capacity is an indicator for the thermal energy required to activate those defects. With this information, their energy level within the band structure can be calculated.This is comparable to the concept of single-frequency EIS (SFEIS), where the impedance at one specific frequency is measured continuously. For example, high-frequency SFEIS measurements have been used to determine the temperature evolution in a fuel cell or battery by relating the Ohmic resistance to the electrolyte temperature.It might not be possible to measure full spectra during an externally triggered system transition, but it is possible to dynamically monitor the impedance at a specific frequency and analyze its dynamics in the time domain. Resistance or capacitance can both be used as indicative parameters for a specific process or behavior.In this study, we analyze the grain boundary resistance of a 3% Gd-doped CeO2 (GDC) thin film. GDC is one of the most prominent oxygen ion conductors but suffers, at lower doping levels, from a large grain boundary resistance that makes polycrystalline samples two to three orders of magnitude more resistive than single crystals. The additional resistance is caused, in large part, by the space-charge zone and associated potential barrier due to the positive charge of the grain boundary core. It is known that the grain boundary resistance of GDC decreases upon biasing with an external voltage. Recently, our group has demonstrated that illuminating the sample with light above the band-gap has a similar effect on the resistance. Basically, the light creates electrons and holes, that are able to compensate the space-charge and flatten the bands as generally assumed for photoelectrochemical systems.We measure the dynamic evolution of resistance and capacitance of a GDC thin film upon bias with voltage and light pulses. Then, we analyze the obtained time domain evolutions, to clarify the exact mechanisms that leads to the resistance decrease in each case.The method presented here nicely compliments a classic EIS analysis, where spectra recorded in two different static operating points represent the characteristics of the system in both states in equilibrium. The evolution of relevant parameters, such as capacitance and resistance at a certain frequency, can provide additional information about the causes and dynamics of the changes and thereby provide more information on the exact mechanism and the driving force.This work was supported by JSPS Core-to-Core Program, A. Advanced Research Networks: “Solid Oxide Interfaces for Faster Ion Transport”, Department of Energy, Basic Energy Sciences and the Swiss National Science Foundation. Figure 1