Date of Graduation


Document Type


Degree Type



Eberly College of Arts and Sciences



Committee Chair

Harry O. Finklea

Committee Co-Chair

Fabien Goulay

Committee Member

Justin Legleiter

Committee Member

Alan Stolzenberg

Committee Member

John Zondlo


Solid oxide fuel cells (SOFCs) are promising electrochemical energy conversion devices due to their advantages of high theoretical efficiency, fuel flexibility (including hydrogen and carbon monoxide), scalability, and low emission. An important problem for commercially using the SOFCs is to improve the long term stability for the SOFCs. To improve the lifetime of SOFCs and develop innovative electrode microstructures; we need to understand the individual process including the degradation processes to the cathode or the anode. To identify the independent anode and cathode contributions to the total impedance, one conventional electrochemical tool for evaluation is the reference electrode. Although reference electrode simulations predict that overpotential/current and impedance data assigned to one electrode will inevitably contain contributions from the other electrode, many current experimental reports persistently use reference electrodes. We have developed several configurations of reference electrodes and cell designs, and have experimentally examined these systems using various gas flows and two temperatures, and have also compared experimental results to simulated predictions. Cyclic Voltammetry (CV) and Electrochemical Impedance Spectroscopy (EIS) were used to collect data. The results for all of our reference electrode designs show agreement with simulation predictions. For the ring reference electrode on the anode-supported cell, the data support that the cathode impedance includes impedance from the anode in the 3-electrode measurement. On the electrolyte-supported cell, using both the ring reference electrode and a micro-reference electrode close to the cathode, the apparent inductance loops in the 3-electrode measurement show up for both reference electrode configurations. The inductive loops are artifacts which indicate the inability to accurately separate impedances for the cathode and anode.;We developed a better protocol to identify the minimum number of processes contributing to the total impedance of a SOFC associate with cathode and anode without the use of reference electrodes. The protocol is based on deconvolution combined with equivalent circuit fitting of impedance data. This protocol uses a spreadsheet program. The impedance / deconvolution technique supplies information of characteristic relaxation time distributions. This information, together with systematic changes in gas composition to the anode or cathode, is used to build the equivalent circuits. We are able to identify the impedance components associated with each electrode. The deconvolution and equivalent circuit analyses are applied successfully to three types of commercially button SOFCs including two anode-supported SOFCs and an electrolyte-supported SOFC. The results show the ability to assign the physical processes associated to cathode or anode in real electrode systems.;In the Strontium-doped lanthanum manganite (LSM) cathode of SOFCs, one special phenomenon of degradation was noticed as a changing polarization resistance over tens of hours in a cell at open circuit voltage (Aging effect). This aging effect was reversible during temperature cycles between 800ºC and 700ºC. We explored this phenomenon via electrochemical techniques (EIS and CV), with the purpose of discovering the true physical source. From the initial investigation, we proposed to relate two processes to the phenomenon we observed: (1) the wetting behavior of the LSM on the Yttria stabilized zirconia (YSZ) surface (2) the segregation of cations. Another approach for separating the processes at different frequencies is deconvolution of impedance spectra combined with complex non-linear regression fitting of the impedance spectra. We performed the extended tests with different cathode and electrolyte compositions under open circuit. This protocol provides the trends in parameters for the equivalent circuit during thermal aging for the impedance data from the extended tests, especially, the frequency ranges associated with aging. Based on results from deconvolution, a hypothesis is proposed to address the aging effect.