Carbon formation or "coking" on solid oxide fuel cell (SOFC) anodes adversely affects performance by blocking catalytic sites and reducing electrochemical activity. Quantifying these effects, however, often requires correlating changes in SOFC electrochemical efficiency measured during operation with results from ex situ measurements performed after the SOFC has been cooled and disassembled. Experiments presented in this work couple vibrational Raman spectroscopy with chronopotentiometry to observe directly the relationship between graphite deposited on nickel cermet anodes and the electrochemical performance of SOFCs operating at 725 °C. Raman spectra from Ni cermet anodes at open circuit voltage exposed to methane show a strong vibrational band at 1556 cm(-1) assigned to the "G" mode of highly ordered graphite. When polarized in the absence of a gas-phase fuel, these carbon-loaded anodes operate stably, oxidizing graphite to form CO and CO(2). Disappearance of graphite intensity measured in the Raman spectra is accompanied by a steep ∼0.8 V rise in the cell potential needed to keep the SOFC operating under constant current conditions. Continued operation leads to spectroscopically observable Ni oxidation and another steep rise in cell potential. Time-dependent spectroscopic and electrochemical measurements pass through correlated equivalence points providing unequivocal, in situ evidence that identifies how SOFC performance depends on the chemical condition of its anode. Chronopotentiometric data are used to quantify the oxide flux necessary to eliminate the carbon initially present on the SOFC anode, and data show that the oxidation mechanisms responsible for graphite removal correlate directly with the electrochemical condition of the anode as evidenced by voltammetry and impedance measurements. Electrochemically oxidizing the Ni anode damages the SOFC significantly and irreversibly. Anodes that have been reconstituted following electrochemical oxidation of carbon and Ni show qualitatively different kinetics of carbon removal, and the electrochemical performance of these systems is characterized by low maximum currents and large polarization resistances.

Novel integration of in situ near infrared (NIR) thermal imaging, vibrational Raman spectroscopy, and Fourier-transform infrared emission spectroscopy (FTIRES) coupled with traditional electrochemical measurements has been used to probe chemical and thermal properties of Ni-based, solid oxide fuel cell (SOFC) anodes operating with methane and simulated biogas fuel mixtures at 800 °C. Together, these three non-invasive optical techniques provide direct insight into the surface chemistry associated with device performance as a function of cell polarization. Specifically, data from these complementary methods measure with high spatial and temporal resolution thermal gradients and changes in material and gas phase composition in operando. NIR thermal images show that SOFC anodes operating with biogas undergo significant cooling (ΔT = -13 °C) relative to the same anodes operating with methane fuel (ΔT = -3 °C). This result is general regardless of cell polarization. Simultaneous Raman spectroscopic measurements are unable to detect carbon formation on anodes operating with biogas. Carbon deposition is observable during operation with methane as evidenced by a weak vibrational band at 1556 cm(-1). This feature is assigned to highly ordered graphite. In situ FTIRES corroborates these results by identifying relative amounts of CO2 and CO produced during electrochemical removal of anodic carbon previously formed from an incident fuel feed. Taken together, these three optical techniques illustrate the promise that complementary, in situ methods have for identifying electrochemical oxidation mechanisms and carbon-forming pathways in high temperature electrochemical devices.

In situ vibrational Raman scattering has been employed to examine rates of carbon formation and removal from Ni/YSZ cermet anodes in functioning, electrolyte-supported solid oxide fuel cells (SOFCs). Specifically, Raman scattering characterized the ability of different gas phase species commonly used as reforming agents to remove carbon that had accumulated on Ni/ YSZ cermet anodes at 730 °C. Anodes held at open circuit voltage (OCV) were exposed first to a dry methane feed and then to an inert carrier gas containing either H 2 O (g) , CO 2 , or O 2 . Carbon deposits began to form within 5 s of methane exposure. Vibrational Raman spectra showed that the carbon deposits consisted of highly ordered graphite as evidenced by a single pronounced feature in the spectra at 1556 cm −1 . Changing the incident gas phase environment over the anode to Ar containing either H 2 O (2%), CO 2 (6%), or O 2 (6%) led to quantitative removal of the carbon and partial or complete oxidation of the Ni as evidenced by the growth of a NiO vibrational band (at 1080 cm −1 ) in the Raman spectra. Carbon removal rates from the Ni/YSZ anode were fastest with vapor phase H 2 O, then O 2 , and finally slowest with CO 2 . The extent of Ni oxidation was much more pronounced with O 2 than with either H 2 O or CO 2 . These chemical processes observed directly in the Raman spectra were reflected in the device's open circuit voltage (OCV). Correlating findings from these two methodsin situ Raman spectroscopy and voltage measurementsprovided a direct connection between the chemical composition of SOFC anodes and the electrochemical condition of the device. These results inspire confidence that any of the reforming agents usedH 2 O, O 2 , and CO 2 will remove carbon from Ni anodes quantitatively on a time scale of ∼10 to ∼125 s. However, H 2 O and CO 2 appear less likely to damage the cell following carbon removal, as H 2 O and CO 2 do not quantitatively oxidize the Ni in the cermet anode. In contrast, exposure to O 2 leads to much more extensive Ni oxidation and an OCV that approaches 0.0 V, implying that the Ni/NiO equilibrium sustained by H 2 O and CO 2 is driven completely to NiO by O 2 .

Supporting InformationElectrochemically-induced Cooling NIR thermal images recorded temperature changes across the anode surface, in particular near the fuel inlet, during different anode processes. The surface temperature cooled near the end of a ten minute exposure to methane or biogas relative to the initial surface temperature under hydrogen. Figure S2 below indicates that this amount of cooling was even more pronounced when the cell was also polarized during fuel exposure. This was true in all cases, as seen from the negative d(∆T). The additional, electrochemically-induced cooling was strongest under methane fuel at 800 °C.

Experiments performed in this work explored how Ni-YSZ cermet anodes infiltrated with 1% Sn or 1% BaO mitigate carbon formation compared to undoped Ni-YSZ anodes in functioning solid oxide fuel cells (SOFCs). In situ vibrational Raman spectroscopy was used to study the early stages of carbon accumulation on the SOFC anodes at 730 °C with methane and under open circuit voltage (OCV) conditions. Additionally, carbon removal with different gas phase reforming agents was evaluated. The effects of these phenomenacarbon accumulation from methane and carbon removal by reforming agentson the electrochemical capabilities of a device were monitored with electrochemical impedance and voltammetry measurements. Vibrational spectra showed that the undoped and 1% Sn infiltrated anodes were very susceptible to carbon formation from methane while considerably less carbon accumulated on the 1% BaO anodes. Electrochemical data, however, implied that carbon accumulated in different regions of the anode and that both Sn and BaO effectively reduced carbon accumulation but also inhibited electrochemical oxidation. For each anode, H 2 O was the most effective reforming agent for removing carbon followed by O 2 and then CO 2 . H 2 O and CO 2 , however, left the anode only partially oxidized, while prolonged exposure to O 2 completely oxidized Ni to nickel oxide. The spectroscopic and electrochemical data showed strong correlations that provide mechanistic insight into the consequences of adding secondary materials to SOFC anodes with the intent of reducing carbon accumulation.