Preparation and Properties of (YCa)(TiMn)O3−δ Ceramics Interconnect of Solid Oxide Fuel Cells

(YCa)(TiMn)O3–δ ceramics prepared using a reaction-sintering process were investigated. Without any calcination involved, the mixture of raw materials was pressed and sintered directly. Y2Ti2O7 instead of YTiO3 formed when a mixture of Y2O3 and TiO2 with Y/Ti ratio 1/1 were sintered in air. Y2Ti2O7, YTiO2.085 and some unknown phases were detected in Y0.6Ca0.4Ti0.6Mn0.4O3–δ. Monophasic Y0.6Ca0.4Ti0.4Mn0.6O3–δ ceramics were obtained after 1400–1500 °C sintering. Dense Y0.6Ca0.4Ti0.4Mn0.6O3–δ with a density 4.69 g/cm3 was observed after 1500 °C/4 h sintering. Log σ for Y0.6Ca0.4Ti0.6Mn0.4O3–δ increased from –3.73 Scm–1 at 350 °C to –2.14 Scm–1 at 700 °C. Log σ for Y0.6Ca0.4Ti0.4Mn0.6O3–δ increased from –2.1 Scm–1 at 350 °C to –1.36 Scm–1 at 700 °C. Increasing Mn content decreased activation energy Ea and increased electrical conductivity. Reaction-sintering process is proved to be a simple and effective method to obtain (YCa)(TiMn)O3–δ ceramics for interconnects in solid oxide fuel cells.


Introduction
Solid oxide fuel cells (SOFCs) transform chemical energy from fuels, such as natural gas, humidified hydrogen, into electrical energy with high conversion efficiency and low pollution. An SOFC includes three principal components: the electrolyte, the cathode and the anode. Each part of the SOFC needs to be compatible both physically and chemically with one another to minimize interfacial reactions. A dense electrolyte is needed to prevent gas mixing, whereas the cathode and the anode must be porous to allow gas transport to the reaction sites. SOFCs generate electricity through the oxidation of fuel at anode and the reduction of oxygen at cathode. To provide a high voltage and power output, interconnects are used to connect the cells in series.

Results and Discussion
The XRD profiles for the 2 h sintered YTiO3 (YT), Y0.6Ca0.4Ti0.6Mn0.4O3−  (YCTM4) and Y0.6Ca0.4Ti0.4Mn0.6O3−  (YCTM6) ceramics are illustrated in Figure 1. The reflections for YT in Figure 1a match well with those of Y2Ti2O7 (ICDD PDF # 00-042-0413) instead those of YTiO3 (ICDD PDF # 00-027-1481). This implies that Y2Ti2O7 formed more easily than YTiO3 as the raw materials with Y/Ti ratio 1/1 were heated in air. Weak peaks (+) around 30° and 50.4° 2θ for YTiO2.085 are seen in Figure 1a. Gill et al. prepared Y2Ti2O7 from Y2O3 and Ti2O3. After 800 °C/12 h calcining, Y2Ti2O7 phase along with weak YTiO2.085 phase was detected for pellets sintered at 1500 °C/12 h. Almost monophasic YTiO2.085 phase was detected for pellets sintered at 1550 °C/12 h [27]. The reflections for YCTM4 in Figure 1b show that a phase with similar crystal structure of YTiO3 formed as the major phase and Y2Ti2O7 phase still formed in YCTM4. Weak peaks for YTiO2.085 and some unknown phases are seen in Figure 1b. The reflections for major phase are also similar to those for Y0.33Ca0.67Ti0.67Mn0.33O3 reported by Kobayashi et al. [17]. The crystal structure for YCTM4 is different from YT as Ca and Mn were added. The reflections for YCTM6 in Figure 1c show that only a phase with similar crystal structure of YTiO3 and Y0.33Ca0.67Ti0.67Mn0.33O3 reported by Kobayashi et al. [17] formed. Y2Ti2O7 phase is not detected. More Mn addition inhibited the formation of Y2Ti2O7, YTiO2.085, and unknown phases. The reaction-sintering process is proved to be a simple and effective process to obtain YT, YCTM4 and YCTM6 ceramics. The calcination step of the conventional mixed oxide route was performed during the heating up period.
Relative density of YT, YCTM4 and YCTM6 ceramics sintered at various temperatures and soak time is shown in Figure 2. A low density, 69.1% of the theoretical density 4.98 g/cm 3 for Y2Ti2O7, was found for 1400 °C/2 h sintering YT pellets. Density of YT increased with sintering temperature and soak time. Dense YT with 91.4% of the theoretical density for Y2Ti2O7 was observed after being sintered at 1500 °C/6 h. Gill et al. prepared Y2Ti2O7 and obtained a low density 73% of the theoretical density after 800 °C/12 h calcining and 1500 °C/12 h sintering [27]. YCTM4 pellets show lower densities. YCTM4 with 77.9% of the theoretical density ~4.97 g/cm 3 for Y0.6Ca0.4Ti0.6Mn0.4O3−  was observed even after sintering at 1500 °C/6 h. Lower densities may be caused by the unknown phases, pores, and YTiO2.085 with a low density 4.28 g/cm 3 . A higher sintering temperature or a prolonged sintering period is suggested for obtaining dense YCTM4. YCTM6 pellets show higher densities. Dense YCTM6 with 93.6% of the theoretical density ~5.01 g/cm 3 for Y0.6Ca0.4Ti0.4Mn0.6O3−  was observed after being sintered at 1500 °C/4 h. The reaction-sintering process is proved to be effective at obtaining YT and YCTM6 ceramics with a relative density higher than 90%.  SEM photographs of as-fired YT ceramics sintered at various temperatures and soak time are presented in Figure 3. Porous pellets with grains smaller than 3 μm are seen for 1400 °C/2 h sintering YT pellets. Grain size increased with sintering temperature and soak time. Grains > 10 μm are seen for 1500 °C/6 h sintering YT pellets. Ding et al. prepared Y2Ti2O7 with 1 mol% La2O3 and found grains smaller than 4 μm after 1450 °C/2-4 h sintering [28]. SEM photographs of the YCTM4 ceramics sintered at various temperatures and soak time are presented in Figure 4. Porous pellets with grains smaller than 8 μm are seen for 1400 °C/2 h sintering YCTM4 pellets. Grain size increased with sintering temperature and soak time. Pores disappeared and grains >10 μm are seen for 1500 °C/6 h sintering YCTM4 pellets. Grain growth increased as Ca and Mn were added into YT. Grains <4 μm are seen for La0.4Ca0.6Ti0.6Mn0.4O3- via an EDTA-citrate method after 950 °C/5 h calcining and 1400 °C/10 h sintering [18]. Therefore, the reaction-sintering process is proved to be effective for grain growth in YT and YCTM4 ceramics. The calcinations stage and the following pulverization for the conventional solid-state reaction route could be bypassed. SEM photographs of the YCTM6 ceramics sintered at various temperatures and soak time are presented in Figure 5. Porous pellets with grains smaller than 6 μm are seen for 1400 °C/2 h sintering YCTM6 pellets. Grain size increased with sintering temperature and soak time. Pores disappeared and grains > 20 μm are seen for 1500 °C/6 h sintering YCTM6 pellets. More Mn addition increased the grain growth in YCTM6 pellets than in YCTM4 pellets. Grains <4 μm are seen for La0.4Ca0.6Ti0.4Mn0.6O3- via an EDTA-citrate method after 950 °C/5 h calcining and 1400 °C/10 h sintering [18]. It is noted some cracks are seen in YCTM6 pellets. These cracks propagated not only along the grain boundaries but also through the grains. Besides, the amount and the size of cracks increased with sintering temperature and soak time. Sintering at temperatures below 1450 °C for a prolonged period or adding sintering aids are suggested for obtaining dense YCTM6 pellets without cracks.   DC total conductivity of 1500 °C/6 h sintering YT, YCTM4, and YCTM6 ceramics are shown in Figure 6. Log σ is found from -8.19 Scm -1 at 350 °C to -4.94 Scm -1 at 700 °C for YT. Gill et al. prepared Y2Ti2O7 and obtained log σ about -7.8 Scm -1 at 700 °C and -6.636 Scm -1 at 900 °C after 800 °C/12 h calcining and 1500 °C/12 h sintering [27]. The reaction-sintering process is proved to be effective at obtaining YT ceramics with a higher conductivity even the calcination was bypassed. Log σ is found from -3.73 Scm -1 at 350 °C to -2.14 Scm -1 at 700 °C for YCTM4. Conductivity increased as Ca and Mn were added into YT. σ about 1.5 Scm -1 at 700 °C for La0.4Ca0.6Ti0.6Mn0.4O3- via an EDTA-citrate method after 950 °C/5 h calcining and 1400 °C/10 h sintering was obtained [18]. Log σ is found from -2.1 Scm -1 at 350 °C to -1.36 Scm -1 at 700 °C for YCTM6. σ about 7 Scm -1 at 700 °C for La0.4Ca0.6Ti0.4Mn0.6O3- via an EDTA-citrate method after 950 °C/5 h calcining and 1400 °C/10 h sintering was obtained [18]. Conductivity further increased as more Mn was added into YCTM4. A similar tendency was also observed in the study of La0.4Ca0.6Ti1-xMnxO3- by Hosseini et al. [18]. Conductivity increased as Mn content increased in La0.4Ca0.6Ti1-xMnxO3- . According to the Arrhenius relationship [29], the experimental activation energy Ea can be determined from the slope of the line when natural logarithm of conductivity (ln σ) is plotted against 1/T as shown in Figure 6. Ea for YT, YCTM4, and YCTM6 were derived as 1.18, 0.64, and 0.33 eV, respectively. Ea of 1.87 eV for La0.4Ca0.6TiO3- and Ea of 0.47 eV for La0.4Ca0.6Ti0.8Mn0.2O3- was obtained via an EDTA-citrate method after 950 °C/5 h calcining and 1400 °C/10 h sintering [18]. Hosseini et al. thought increasing Mn content decreased Ea and increased electrical conductivity in air. Increasing Mn content causes the number of oxygen vacancies to increase and this in turn causes an increase in electron hole concentration [18]. A similar tendency was also observed in this study.
The reflections of various phases for the sintered pellets were analyzed by XRD. Microstructures were analyzed by scanning electron microscopy (SEM). The density of the sintered pellets was measured using the Archimedes method. Ag electrodes were formed on both sides of the sintered pellets. Agilent 34970A Data Acquisition was used for electrical resistivity measurements at 350-700 °C.

Author Contributions
Yi-Cheng Liou formulated research ideas and wrote initial manuscript. Analysis of data was performed by Yi-Cheng Liou and Wen-Chou Tsai. Hao-Hsuan Yen and Yung-Chia Chang performed the experiments. All authors read and approved the manuscript.