Search

Combined technique and setup have been created that combines capabilities of electrochemical methods, as well as high-temperature Raman spectroscopy (RS) at Osipyan Institute of Solid State Physics RAS. In order to study the processes in the electrochemically active zone, a special geometry of samples was developed on basis of optically transparent single crystal membranes of an anionic conductor with a toroidal shape counter-electrode. With the use of this combined technique and special geometry, studies of the kinetics of reduction of nickel in composite SOFC anodes were carried out. The influence of the composition of fuel on Raman spectra obtained from the internal interface in the current load mode was also investigated, and the correlations with the cell voltage were studied. RS is also used to study the structure of single-crystal samples of anionic conductors, including at SOFCs operating temperature. The new combined technique was used to study other components: sealing glasses for SOFCs and optical glasses.

Steam reforming is a highly endothermic process of hydrogen or synthesis gas production from methane or other hydrocarbon fuels. Metals from group VIII of the Periodic table are catalysts for steam reforming; nickel is the most commonly used among them. Platinoid catalysts: rhodium, ruthenium and palladium are more active, but less frequently used due to their high cost. Nickel catalysts have proven their effectiveness due to the simplicity of their production, stability and chemical activity [1]. It is known that the heat required for the endothermic process of steam reforming can be provided by an electrochemical reaction in a stack of high-temperature solid oxide fuel cells [2,3]. In the case of internal reforming, the process of steam reforming of methane proceeds directly at the anode of a solid oxide fuel cell (SOFC) due to the high operating temperature and the presence of nickel in the composition [4,5]. The advantage of internal reforming is not only thermal, but also chemical integration of reforming agent and generator – the water vapor produced during the SOFC anodic electrochemical reaction can be used for reforming without the need for anode recycling organization. Also, the endothermic effect of steam reforming can be used to control the temperature of the stack. As it was shown in work [6] feeding of the mixture containing 30% of methane into the electrolyte-supported SOFC stack to the input leads to 22% of methane observed at the output. Thus, based on the literature data, we cannot expect a high conversion rate in the process of internal methane reforming. Therefore, studies of the internal methane conversion process were carried out on an experimental short stack of two 100x100 mm electrolyte-supported SOFCs. A whole series of experiments were carried out on H2+CH4+H2O mixtures with different sets of conditions – an increase in the ratio and consumption of methane, and then a decrease in temperature. The ranges of methane consumption and other key parameters of the experiment were based on the results of the analysis of available literature data on the experimental study of the kinetics of methane steam conversion on cermet SOFC anodes. Prior to the experiments, discussions were mainly caused by the probable carbon deposition at the anodes and the associated rapid performance degradation, the capabilities of the anodes for internal conversion were upper estimated according to the fuel consumption of SOFCs, so it was assumed that, at worst case, the conversion provides only its internal consumption needs. The results of experiments at an operating temperature of 850°C showed that the kinetics of internal conversion was largely underestimated. The assembly of two SOFCs converts methane up to concentrations not exceeding the calculated equilibrium ones, up to the maximum available flow rate on the equipment used – 187 ml/min – even without current passing. This fuel flow rate is capable of providing a current of up to 53 A, while the rated operating current of these SOFCs is about 20 A, up to 30 A on methane-rich fuel. This means, that the conversion capabilities of SOFC even in the absence of current are at least twice higher than their own needs. The current further stimulates the conversion by generating additional steam. In order to discover the limits of conversion capabilities, the temperature was lowered to 750°C. The result of the conversion at maximum flow rate proceeding up to equilibrium values was repeated. Thus, in the course of the experiments, it was not possible to obtain methane concentrations in the exhaust gas that significantly exceed the equilibrium ones. This result turned out to be extremely unexpected and undoubtedly positive, since it demonstrates an extremely high potential of internal conversion and opens the way to a sharp decrease in the requirements for the degree of conversion of fuel supplied to SOFC, abandonment of a bulky fuel processor, reduction of costs for SOFC cooling and, thereby, an increase in efficiency and reducing the mass and dimensions of SOFC power plants. This work was carried out with financial support from the Russian Scientific Foundation, grant no. 17-79-30071. References 1. Tokyo Gas Co. Ltd., Japanese Patent No. JP 06-243881 (1994) 2. A.H. Fakeeha et al. J. King Saud Univ., Vol. 7, Eng. Sci. (Special Issue), pp. 171-189 3. S. H. Clarke et al. Catalysis Today 38 (1997) 41 l-423 4. A.L. Dicks Journal of Power Sources 71 (1998) 111–122 5. В.А. Собянин Ж. Рос. хим. об-ва им. Д.И. Менделеева (2003) XLVII №6 с. 62-70 6. Kupecki, K. Motylinski, J. Milewski Energy Procedia 105 (2017) 1700–1705

SOFC efficiency is directly determined by the electrodes, optimization of which is related to the study of the mechanisms of electrochemical current-generating reactions taking place. Such studies are tricky due to the high operating temperature of SOFCs, high current loads, and corrosive gas media. A promising research method is Raman spectroscopy (RS) under operating conditions of SOFC. At ISSP RAS, a combined experimental technique and setup have been created that combines the capabilities of traditional electrochemical methods, as well as high-temperature Raman spectroscopy [1]. In order to study the processes in the electrochemically active zone, a special geometry of samples was developed on basis of optically transparent single crystal membranes of an anionic conductor with a counter-electrode of a toroidal shape. With the use of this combined technique and special geometry of samples in previous works, studies of the kinetics of reduction and morphological changes of nickel in composite SOFC anodes were carried out. It was shown that the first reduction cycle differs significantly from the subsequent ones both in the initial delay and in the total time of the process. SEM studies have shown that these changes are associated with a morphological rearrangement that occurs during the first reduction cycle: the grain size of NiO is significantly reduced compared to the initial one. For model SOFCs investigated in the OCV mode [2] after the initial microstructural rearrangement in the redox cycle, the behavior of standard cermet anodes in a H2–N2 flowing atmosphere can be described using Avrami model. The influence of the composition of fuel on Raman spectra obtained from the internal interface in the current load mode was also investigated [3], and the correlations of obtained data with the cell voltage were studied. It was shown that the corresponding mechanisms that determine the reaction kinetics can be associated with the transfer of ions through the GDC|YSZ. Subsequent rate limiting steps such as electrochemical oxidation of hydrogen in the Ni-GDC layer are undetectable for the geometry under test due to the limited penetration depth of the laser beam. Comparative studies of model SOFCs with a supporting anode substrate (ASC) and a supporting solid electrolyte (ESC) were carried out using the in-situ RS technique [4]. It was shown that the addition of the GDC indicator layer makes it possible to carry out an in-situ study of the chemical potential of oxygen at the internal interface of the anode|electrolyte depending on current load or fuel mixture, but the GDC/YSZ two-layer thin-film electrolyte exhibits electronic or gas leakage due to poor stability in a reducing atmosphere. Despite the significantly lower internal resistance compared to the ESC, the quality of the ASC thin film membrane severely limits research depending on the operating temperature, current load and fuel composition. In this case, the use of thin-film membranes significantly reduces the effect of 8YSZ on the Raman spectra of the internal interface. After post-processing and normalization, the obtained Raman spectra show a similar effect of the current load or the hydrogen content in the fuel gas mixture on the GDC peak area (460 cm-1). The linear dependence of the OCV on the peak area of ~460 cm-1 makes it possible to assess the relationship between changes in the peak area and the evolution of the chemical potential of oxygen at the anode under a current load. Comparative analysis of the anodic impedance for different fuel mixtures suggests that spectroscopic measurements of the internal interfaces provide direct information on the contribution of the fuel oxidation reaction to the total losses in SOFC. In addition to studies of complete fuel cells, high-temperature Raman spectroscopy is used to study the structure of single-crystal samples of anionic conductors [5], including at the operating temperature of SOFCs. High-temperature RS allows one to obtain additional information on the microstructure of samples both at room temperature and at working temperature, where most other research methods do not allow research. New combined technique was used to study other components of high-temperature electrochemical devices: sealing glasses [6] for SOFCs and optical glasses [7]. This work was supported by Russian Scientific Foundation, grant no. 17-79-30071. [1] D.A.Agarkov et al. ECS Trans., vol.68, iss.1, pp.2093-2103 (2015). [2] D.A.Agarkov et al. Solid State Ionics, vol.302, pp.133-137 (2017). [3] D.A.Agarkov et al. Solid State Ionics, vol.319C, pp.125-129 (2018). [4] D.A.Agarkov et al. Solid State Ionics, vol.344, p.115091 (2020). [5] D.A.Agarkov et al. Solid State Ionics, vol.346, p.115218 (2020). [6] A.Allu et al. ACS Omega, vol.2, pp.6233-6243 (2017). [7] M.K.Kokila et al. Solid State Sciences, vol.107, p.106360 (2020). Figure 1