Thermal Cycling Stability of NiO/YSZ Anode-Supported SOFC Button Cells: An Experimental Study

Abstract

Solid oxide fuel cell (SOFC) technology is an electrochemical power generation apparatus that enables the direct conversion of chemical fuel energy into electrical energy. To address the issue of thermal cycling stability, which is critical for commercialization, a thermal cycling stability test was performed on a NiO/YSZ anode-supported SOFC button cell. This study investigates the influence of key thermal cycling parameters (heating/cooling rate and number of thermal cycles) on the cell’s electrochemical performance and microstructure evolution. The main findings are as follows: thermal cycling adversely affects the electrochemical performance of the SOFC, with the degree of degradation directly correlated to both the number of cycles and the heating/cooling rate. After 20 thermal cycles at a rate of 5 °C/min, the peak power density decreased by 20.57%. Furthermore, thermal cycling leads to an increase in both ohmic and activation polarization, with the performance degradation predominantly governed by the rise in ohmic polarization. It was demonstrated that the number of thermal cycles has a more significant impact on ohmic losses than the heating/cooling rate. This work offers valuable insight into the degradation mechanisms induced by thermal cycling in SOFC button cells.

1. Introduction

Solid oxide fuel cell (SOFC) is an electrochemical power generation apparatus that has been developed to enable the direct conversion of chemical fuel energy into electrical energy [1,2,3]. The typical operating temperature of SOFCs is currently in the range of 700 °C and 800 °C [2]. It is imperative to note that, in order to function correctly, the SOFC stack must undergo repeated thermal cycles between the ambient temperature and the operating temperature during the operation of an SOFC. Insufficient thermal cycling stability remains a key challenge that hinders the commercialization of SOFCs [4].

The core components of an SOFC comprise a porous anode layer, a porous cathode layer, and a dense electrolyte layer. It is important to note that the materials constituting each layer are distinct and that these materials possess different thermal expansion coefficients. Consequently, this may result in the induction of significant thermal stress within the component, and/or between the various layers, during the process of thermal cycling. It has been demonstrated that exposure to elevated temperatures can result in the delamination or fracture of the cell interface, thereby exerting a detrimental effect on SOFC performance [4]. In the study conducted by Lee et al. [5], the open circuit voltage (OCV) of the SOFC stack was examined in relation to thermal cycling. The findings of the research indicate that varying thermal cycle temperature ranges exert divergent effects on cell performance. It has been demonstrated that the use of gradient electrodes and composite sealants with sufficient elasticity at elevated temperatures can markedly enhance the thermal cycling stability of the SOFC stack. Yang et al. [6] conducted a study to assess the durability of the SOFC stack at 750 °C and its thermal cycling stability within the temperature range of 450 °C to 750 °C. The research findings indicate that the degradation of SOFC stack performance is primarily attributable to two factors: firstly, partial decomposition of the contact material LaCo_0.6Ni_0.4O_3-d in the cathode and, secondly, oxidation of the metal interconnector. The selection of appropriate sealing materials in conjunction with the enhancement of the metal inter-connector has the potential to elevate the stack degradation rate of the research group to 5%/1000 h. Pan et al. [7] conducted a study to evaluate the thermal cycling stability of a SOFC stack, subjecting it to temperatures ranging from room temperature to 750 °C. The findings of the research indicate that the mean voltage degradation rate per thermal cycle is 0.8%. Shin et al. [8] studied the effect of thermal cycle on the cell performance using NiO/YSZ anode-supported button cell, varying the heating/cooling rates from 1 °C/min to 5 °C/min at temperatures ranging from 400 °C to 700 °C. The findings of the research indicate that following the thermal cycling test at heating/cooling rates of 1 °C/min and 5 °C/min, respectively, the degradation rate of cell voltage was determined to be 8.8% (82 μV/h) and 19.1% (187 μV/h). Further resistance spectrum analysis results show that, during thermal cycling, the ohmic polarization of the cell remains consistent with the initial value, and electrode degradation (especially severe degradation of the anode and Sr segregation of the cathode) is the primary cause of cell performance deterioration. Zheng et al. [9] examined the coupled electrochemical and mechanical degradation of SOFC stacks during thermal cycling between 35 °C and 750 °C. Their findings demonstrated that, after 10 cycles, the SOFC stack voltage at 40 A decreased by 12.75%, while the PEN flexural strength fell by 54.22%. The post-mortem SEM/EDS analysis revealed a number of significant findings: severe coarsening, migration and depletion of nickel (Ni) in the anode, accompanied by strontium (Sr) segregation in the cathode. Further ECM fitting showed that 57% of the voltage loss originated from ohmic polarization increases (interconnect oxidation, sealing aging) and 37% from anode charge-transfer degradation. Wu et al. [10] evaluated the thermal cycling stability of a three-cell flat-tube SOFC short stack over 105 cycles (750–250 °C) and found that the SOFC stack retained good sealing (pressure drop < 2 kPa) and exhibited an OCV degradation of merely 0.027% per cycle and an operational-voltage degradation of 0.063% per cycle at a current density of 567 mA/cm². The experimental evidence, as presented in the EIS and DRT, indicated that the performance loss of the SOFC stack was primarily caused by an increase in ohmic polarization due to oxidation of the uncoated SUS441 interconnect. In addition, Kim et al. [11] conducted a series of SOFC thermal cycling tests, ranging from 200 °C to 750 °C, utilizing NiO/YSZ anode-supported button cells under various conditions. The experimental conditions comprised three distinct scenarios: condition 1, involving the absence of hydrogen supply; condition 2, involving the provision of hydrogen supply; and condition 3, involving the application of a negative current. The findings of the research indicate that, under conditions 2 and 3, there is no discernible change in cell performance and anode microstructure following the thermal cycling test. However, under condition 1 (slow), a significant degradation in cell performance is observed due to the re-oxidation of nickel in the anode. Han et al. [12] investigated the degradation behavior of NiO-YSZ anode-supported SOFCs under repeated thermal cycling (400–700 °C, 10 °C/min) at different current densities and found that the operating current density significantly influences both the structural integrity and electrochemical performance of SOFCs during thermal cycling.

The stability of SOFC can be categorized into three distinct types: long-term operation stability under a fixed operating condition, thermal cycling stability, and redox stability [13]. However, it should be noted that the aforementioned thermal cycling experiments were all conducted with a protracted operational time. For instance, in the experiment conducted by Pan et al. [7], the cell was discharged for a period of two hours prior to the performance testing in each thermal cycle. Similarly, in the experiment undertaken by Shin et al. [8], the cell was discharged for a duration of six to twelve hours prior to the performance testing in each thermal cycle. This indicates that the cell has undergone a significant number of hours of operation while being subjected to thermal cycling, amounting to dozens or even hundreds of hours. It has previously been established that the performance of the cell in the early stages of operation is characterized by significant instability [14]. It is therefore logical to hypothesize that the degradation in cell performance evidenced in the aforementioned study may be attributable to the concomitant effects of thermal cycling and long-term operation. Furthermore, investigations dedicated to SOFC thermal-cycle stability are both time-intensive and expensive. The existing literature on this subject is limited, and there is currently no consensus among researchers regarding the dominant degradation mechanisms [9,15]. For example, Kim et al. [11] observed negligible changes in both cell performance and microstructure under “condition 1 (fast)”, a finding that conflicts with the conclusions drawn in ref. [8].

A commercial SOFC stack comprises multiple parts, and degradation can be influenced by the compatibility among these parts as well as by each part in isolation. Previous investigations into the thermal cycling stability of SOFCs have predominantly focused on stack or system levels, where performance degradation is influenced by a combination of factors, including interconnect corrosion, sealant failure, and manifold issues. While critical for application, these studies make it challenging to decouple the intrinsic degradation of the cell components from extrinsic system-level failures. The present study distinguishes itself by employing a NiO/YSZ anode-supported SOFC button cell configuration. This approach intentionally isolates the electrode and electrolyte materials, allowing for a fundamental investigation solely into the impact of thermal cycling parameters (i.e., heating/cooling rate, number of thermal cycles) on the electrochemical performance and microstructure of the cell itself. This work provides valuable insights into the thermal cycling-induced degradation mechanisms in SOFC button cells.

2. Experimental Setup and Methods

The SOFC button cell used in this study is produced by Zhongfu (Wuxi) New Energy Co., Ltd., Wuxi, China. The basic structure and the materials used in the cell are as follows: The anode comprises NiO/YSZ, with a thickness of 400 μm and a porosity of approximately 0.4. The electrolyte consists of YSZ, with a thickness of 4 μm. The barrier layer is GDC, with a thickness of 5 μm. The cathode is LSC, with a thickness of 20 μm. The anode and electrolyte layers were sintered at 1400 °C, while the barrier layer and cathode were sintered at 1250 °C and 1050 °C, respectively. The diameters of the anode and the electrolyte were measured at 17 mm, while the cathode has a diameter of 10 mm. The photograph of the button cell used for the test is shown in Figure 1.

During testing, the cathode side of the button cell was directly exposed to ambient air, while the fuel side was sealed with ceramic paste. Silver wires and silver paste were used for current collection to minimize contact resistance and avoid the influence of metallic interconnects on cell performance.

The experimental steps are outlined below:

First, the SOFC button cell is placed in a muffle furnace, and different thermal cycling processes are achieved by setting different temperature heating/cooling rates. The temperature range of the thermal cycling process is set between 200 °C and 800 °C.

Secondly, the SOFC button cell was thermally cycled and fixed to the button cell testing bench. The electrochemical performance of the cell was then tested at 750 °C with a hydrogen gas flow rate of 100 SCCM. In order to provide a comprehensive comparison, the electrochemical performance of the button cell was also tested in the absence of thermal cycling. Figure 2 presents the button cell testing bench used in this study. This comprises a hydrogen generator, a nitrogen generator, two gas flow meters, a gas scrubber bottle, an electric furnace, an electrochemical workstation, a computer, gas pipes and wires. The red lines signify the gas path, while the green lines denote the electric circuit.

Finally, the SOFC button cell was retrieved from the testing bench following the completion of the electrochemical performance test. A specimen of the button cell’s central reaction part was prepared for further analysis, after which the cross-section of the button cell was examined using a scanning electron microscope (SEM). In order to establish a basis for comparison, the cross-section of the button cell was also tested in the absence of thermal cycling.

It should be noted that the selection of the specific heating/cooling rates (1 °C/min and 5 °C/min) was based on their relevance to practical SOFC operation, as this range (approximately 1–5 °C/min) represents common thermal cycling conditions encountered in engineering applications. Furthermore, to ensure the reliability and reproducibility of the results, three independent button cells were tested for each experimental condition. The observed degradation trends and conclusions were highly consistent across all replicates. Therefore, for the sake of clarity and conciseness in data presentation, this paper presents the results from a single, representative cell, which accurately reflects the consistent behavior observed across all samples.

3. Results and Analysis

In the course of the thermal cycling experiment, the same batch of SOFC button cells was exposed to 10 and 20 thermal cycles, respectively. With regard to the heating/cooling rate, the thermal cycles were conducted at a rate of 1 °C/min and 5 °C/min, respectively. The present study investigates the impact of heating/cooling rate and the number of thermal cycles on the electrochemical performance and microstructure change in the SOFC button cell.

3.1. Electrochemical Performance

As demonstrated in Figure 3, the electrochemical performance of the SOFC button cell that has experienced thermal cycling is worse than that of the button cell that has not experienced thermal cycling. When the heating/cooling rate is 1 °C/min, the peak power density of the SOFC button cell after 0, 10 and 20 thermal cycles is 441.2 mW/cm², 429.8 mW/cm² and 406.3 mW/cm², respectively. When the heating/cooling rate is 5 °C/min, the peak power density of the SOFC button cell after 0, 10 and 20 thermal cycles is 443.9 mW/cm², 388.1 mW/cm² and 352.6 mW/cm², respectively. It is evident that an increase in the number of thermal cycles results in a decline in the electrochemical performance of the SOFC button cell after thermal cycling. It has been demonstrated that the electrochemical performance of the SOFC button cell is adversely affected by a faster heating/cooling rate during thermal cycling. As the heating and cooling rates increased from 1 °C/min to 5 °C/min, the peak power density of the SOFC button cell after 20 thermal cycles decreased by 7.9% and 20.6%, respectively, in comparison with the peak power density of the SOFC button cell that had not undergone thermal cycling.

3.2. EIS Performance

Following the electrochemical performance test, electrochemical resistance spectroscopy (EIS) analysis was also conducted on the SOFC button cell at an output voltage of 0.3 V. To deconvolute the contributions from different resistive processes, the impedance spectra were fitted using an LR(QR)(QR) equivalent circuit model, a standard approach for SOFC analysis. In this model, the series inductance (L) and resistance (R) represent lead inductance and the total ohmic loss, respectively, while the two parallel (QR) elements correspond to the electrode polarization processes at different frequency ranges. The fitting results showed good agreement with the measured data, confirming the validity of the model. As shown in Figure 4, when the heating/cooling rate is 1 °C/min, and the ohmic polarization of the SOFC button cell after 0, 10 and 20 thermal cycles is 0.509 Ω*cm², 0.596 Ω*cm² and 0.610 Ω*cm², respectively, while the activation polarization of the SOFC button cell after 0, 10 and 20 thermal cycles is 0.036 Ω*cm², 0.055 Ω*cm² and 0.055 Ω*cm², respectively. In contrast, at a heating/cooling rate of 5 °C/min, the ohmic polarization of the SOFC button cell after 0, 10 and 20 thermal cycles is 0.504 Ω*cm², 0.616 Ω*cm² and 0.600 Ω*cm², respectively. Meanwhile, the activation polarization of the SOFC button cell after thermal cycling is 0.039 Ω*cm², 0.107 Ω*cm² and 0.149 Ω*cm², respectively.

Therefore, as a consequence of the results obtained above, it can be concluded that thermal cycling exerts a significant effect on the ohmic polarization of the SOFC button cell. It has been demonstrated that, when the heating/cooling rates are set at 1 °C/min and 5 °C/min, respectively, there is an increase in the ohmic polarization of the SOFC button cell of 19.8% and 19.0% after 20 thermal cycles when compared with that of the SOFC button cell that has not undergone thermal cycling. This increase is primarily attributed to the degradation of interfacial integrity. The thermally induced stresses, resulting from the mismatch in coefficients of thermal expansion between cell components, can cause micro-cracking and delamination, which consequently impair the continuous paths for electron transport and increase the ohmic resistance. Evidently, the magnitude of the increase in the ohmic polarization of the SOFC button cell after thermal cycling is directly proportional to the number of thermal cycles. However, it is also to be noted that there is no significant difference in the impact of different heating/cooling rates on the ohmic polarization increase.

Furthermore, it has been demonstrated that, in the instance of the heating/cooling rates being set at 1 °C/min and 5 °C/min, respectively, the activation polarization of the SOFC button cell undergoes a 52.8% and 282% increase after 20 thermal cycles when compared with that of the SOFC button cell that has not undergone thermal cycling. This significant rise, particularly at the higher heating/cooling rate, is directly linked to the damage at the electrode–electrolyte interface. The same microstructural degradation that increases ohmic resistance also reduces the number of active triple-phase boundaries and disrupts ion transport pathways, thereby leading to a more substantial increase in activation polarization. It is evident that rapid fluctuations in temperature result in a concomitant increase in the activation polarization of the SOFC button cell. However, it should be noted that, although thermal cycling causes a significant change in the activation polarization of the SOFC button cell, given that the absolute value of the activation polarization is substantially smaller than that of the ohmic polarization, the increase in ohmic polarization is widely considered to be the primary factor contributing to the cell performance degradation during thermal cycling.

It should be noted that the finding in this work highlights the dominant role of ohmic polarization, providing a complementary perspective to some stack-level studies where activation polarization or other factors were reported as more significant. The disparity can be reasonably attributed to the fundamental difference in research focus: stack-level studies [8,9,10] inherently include complexities such as interconnect oxidation and sealant failure, which may introduce additional resistance sources and mask the intrinsic degradation behavior of the cell components themselves. In contrast, the present button cell study effectively isolates the effect of thermal stress on the cell, thereby revealing the core mechanism that might be convoluted in more complex systems. This fundamental understanding is crucial for guiding targeted material and interface improvements in both single cells and stacks.

3.3. Microstructural Analysis

In order to clarify the morphological changes in each functional layer, the button cell after the electrochemical test was removed from the test bench. A sample of the reaction part in the center of the button cell was prepared, and the sample’s cross-section was characterized by means of scanning electron microscopy (SEM).

As demonstrated in Figure 5, the cross-section of a newly fabricated SOFC button cell is presented, prior to undergoing any testing procedures. The microscopic morphology of each functional layer of the SOFC button cell is clearly visible (from top to bottom in the image: they are the cathode layer, the barrier layer, the electrolyte layer and the anode layer, respectively). In the repeated characterization of the new SOFC button cell, the cell microstructure showed the same structural integrity, uniformity and consistency, indicating that the manufacturing process of the new SOFC button cell is relatively mature and thus the consistency of the initial conditions of the experiment is ensured.

As illustrated in Figure 6, the cross-sections of the SOFC button cell following thermal cycling exhibit distinct features depending on the heating/cooling rates employed, namely, 1 °C/min and 5 °C/min, respectively. It is evident that the electrolyte exhibits uniformity and high density after thermal cycling, suggesting that it possesses adequate thermal cycling resistance.

However, as shown in Figure 6, the microstructure of the barrier layer and the cathode layer undergoes changes after thermal cycling. It has been established that, when the heating/cooling rate is 1 °C/min, there are preliminary signs of delamination between the barrier layer and the cathode layer. This is indicated by an increase in the gap between the two layers and a decrease in tightness. Moreover, a comparison of Figure 6a,b demonstrates a positive correlation between the observed gap and the number of thermal cycles experienced. It is evident that the discrepancy diminishes after 10 cycles and becomes more pronounced after 20 cycles. This phenomenon has also been observed in cell samples following thermal cycling at a rate of 5 °C/min. In such cases, the delamination between the barrier layer and the cathode layer is more readily apparent, as illustrated in Figure 6c,d.

Furthermore, a positive correlation between the delamination phenomenon and the heating/cooling rate during thermal cycling is observed in Figure 6. It is evident that an increased heating/cooling rate will result in a more pronounced delamination phenomenon. This phenomenon has been documented in previous studies [16,17]. Delamination has been identified as a factor that can lead to an increase in contact resistance between the layers. This, in turn, has the potential to significantly increase the ohmic polarization of the SOFC button cell. The validity of this assertion is reinforced by the findings derived from the Electrochemical Impedance Spectroscopy (EIS) analysis, as outlined in Section 3.2.

Finally, as demonstrated in Figure 6d, the cathode of the SOFC button cell exhibited cracking after 20 thermal cycles at a heating/cooling rate of 5 °C/min. This phenomenon may be attributed to the utilization of an LSC cathode by the cell. The thermal expansion coefficient of LSC is greater than that of YSZ, and, therefore, structural fatigue will occur due to the stress accumulation after repeated rapid thermal cycling, resulting in a decrease in the SOFC electrochemical performance. This finding is in accordance with the conclusions of earlier research [9,18], which indicated that the instability of the cathode leads to severe degradation in SOFCs during rapid thermal cycling. It is evident that the destruction of the cathode structure exerts a significant effect on ion/electron transport in the cathode. Consequently, this results in an increase in both activation and ohmic polarization. These outcomes are consistent with the EIS results obtained in Section 3.2, thereby validating the experimental method and providing a robust foundation for further research.

It should be noted that some researchers have attributed SOFC performance degradation during operation primarily to Sr segregation in the cathode [19]. To examine this, we conducted EDS analysis on the button cell before and after thermal cycling. As shown in Figure 7, a direct comparison of the EDS results before thermal cycling and after 20 thermal cycles at 5 °C/min demonstrates uniform elemental distribution in both cases. Quantitative analysis reveals the Sr content changed only marginally from 7.64 Wt% to 9.99 Wt%, which does not represent a significant variation. Furthermore, the elemental mapping of the cell after thermal cycling (Figure 8) also shows homogeneous distribution of Sr, La, and Co in the cathode side. These results confirm that no noticeable changes in elemental distribution were detected after thermal cycling, thereby supporting our conclusion that performance degradation is mainly caused by crack formation induced by thermomechanical stress rather than by Sr diffusion.

4. Conclusions

In the present study, the thermal cycling stability test was conducted on a NiO/YSZ anode-supported SOFC button cell at an industrial scale. The present study investigates the impact of various thermal cycling parameters (i.e., heating/cooling rate, number of thermal cycles) on the electrochemical performance and microstructure change in the SOFC button cell. The following key conclusions can be drawn from this study:

Firstly, it is evident that thermal cycling has a detrimental effect on the electrochemical performance of SOFCs. It has been demonstrated that the extent of performance degradation observed after thermal cycling is directly proportional to the number of thermal cycles the cell undergoes and the heating/cooling rate during these cycles. Following 20 thermal cycles at a heating/cooling rate of 5 °C/min, a decline of 20.57% in the peak power density of the SOFC button cell was observed.

Secondly, it has been demonstrated that thermal cycling concurrently elevates ohmic and activation polarization. It is of crucial significance to understand that the performance degradation of SOFC button cells is predominantly governed by elevated ohmic polarization. The findings of the present study demonstrate that the impact of thermal cycles on ohmic losses is more significant than that of the heating/cooling rate.

Thirdly, it is important to note that thermal cycling has the potential to induce delamination between the cathode layer and the barrier layer. Furthermore, this process has the potential to result in the formation of cracks within the cathode. Consequently, this has the potential to result in a compromise to the structural integrity and performance of the SOFC button cell.

Finally, the observed performance degradation and microstructural damage collectively highlight the critical impact of thermo-mechanical stress on SOFC durability. These findings provide practical insights for the engineering of SOFC single cells and stacks. From an engineering perspective, two feasible pathways are suggested to mitigate this issue. At the cell level, transitioning from a single-phase LSC cathode to an LSC-GDC composite and/or a functionally graded electrode structure represents a promising strategy for improving thermo-mechanical compatibility. Beyond these approaches, developing cathode materials with negative thermal expansion (NTE) properties has emerged as another research focus to actively compensate for thermal mismatch and enhance cycling stability [20]. At the stack level, optimizing the structural design to enhance thermal management and minimize internal temperature gradients is equally critical [21]. The implementation of these strategies is expected to significantly improve the thermal cycling performance and operational longevity of SOFC systems.