Lowering the sintering temperature of a gadolinia-doped ceria functional layer using a layered Bi₂O₃ sintering aid for solid oxide fuel cells

Abstract

Solid oxide fuel cells are promising renewable energy devices due to their high efficiency and fuel flexibility. As they operate at a higher temperature than other fuel cells, ceramic materials, such as perovskite-based La_0.6Sr_0.4CoO₃ and La_0.6Sr_0.4Co_0.2Fe_0.8O₃, can be used as electrodes to replace expensive noble metals. However, when the corresponding electrode and yttria-stabilized zirconia electrolyte are sintered together, SrZrO₃ produced from a side reaction acts as an insulator and deteriorates the performance of the fuel cell. Thus, the dense functional layer of a ceria-based material should be introduced between the electrode and the electrolyte to suppress the formation of secondary phases. However, in the conventional cell manufacturing process, it is challenging to manufacture a dense functional layer under constrained sintering conditions. In this study, we develop a method for fabricating a dense gadolinia-doped ceria (GDC) functional layer, even under constrained sintering conditions, by using a sacrificial bismuth oxide, Bi₂O₃, sintering aid layer above the GDC layer. As thermal sintering progresses at 1000–1200 °C, the Bi₂O₃ sintering aid layer is sublimated, leaving only the pure GDC functional layer. The fabricated dense GDC functional layer characterized by various analysis methods shows improved solid oxide fuel cell performance.

Introduction

Fuel cells are eco-friendly hydrogen energy conversion devices that directly convert chemical energy into electrical energy. A typical fuel cell comprises an anode, a cathode, and an electrolyte, and it is classified according to the types and characteristics of the electrolyte [[1], [2], [3]]. In a fuel cell, electricity is generated by electrochemical reactions, namely the hydrogen oxidation reaction at the anode and the oxygen reduction reaction at the cathode. Electrolytes used in solid oxide fuel cells (SOFCs) are made of oxide materials, such as yttria-stabilized zirconia (YSZ), which is chemically stable and conducts oxide ions, that function as charge carriers. Unlike the electrolytes of water-containing fuel cells that conduct hydrogen or hydroxide ions as charge carriers, SOFCs conduct oxygen ions as charge carriers. Due to the lower ionic conductivity of oxide-based solid electrolytes compared to water-based hydrogen ion conductors, SOFCs require high operating temperatures (800–1000 °C) to minimize the ohmic losses caused by the high ion transfer resistance. High operating temperatures lead to problems, such as long start-up time, low durability, and limited material selection. Thus, lowering the operating temperature of SOFCs to intermediate temperature ranges (500–800 °C) has been suggested [[4], [5], [6]]. Studies have suggested that this problem can be solved by lowering the electrochemical reactivity and the operating temperatures, but it would degrade the performance of SOFCs; however, it should not be a concern as the operating temperature of SOFCs is still relatively higher than other fuel cells. This high operating temperature is advantageous because it allows the use of hydrocarbons as fuel through internal reforming [7]; moreover, it leads to high power generation efficiency because of the use of combined heat and power systems [8]. Furthermore, relatively inexpensive ceramic materials, such as perovskite-based La_0.6Sr_0.4CoO₃ (LSC) and La_0.6Sr_0.4Co_0.2Fe_0.8O₃ (LSCF), can be used as electrodes to replace expensive noble metal electrodes. Electrode materials comprising lanthanum and strontium, such as LSCF and LSC, possess both oxygen vacancies and electron holes to maintain charge neutrality due to the charge difference between La³⁺ and Sr²⁺. Thus, these cathode materials exhibit high electronic conductivity as well as high ionic conductivity similar to mixed ionic electronic conductors (MIECs). Since the electrochemical reaction of SOFCs occurs at a three-phase boundary, the use of MIEC electrodes will significantly increase the number of actual electrochemical reaction sites, as the entire surface will come into contact with the gaseous phase, forming an electrochemically active reaction site [2,[9], [10], [11]].

As MIEC cathode materials containing La and Sr, such as LSC and LSCF, are highly reactive to zirconia-based electrolytes during the high-temperature sintering process, side reactions at the electrolyte and cathode interface would form ionic insulating layers, such as La₂Zr₂O₇(LZO) or SrZrO₃(SZO), which greatly deteriorate fuel cell performance due to the increased ohmic resistance. To resolve this issue, studies have suggested replacing zirconia-based electrolytes with ceria-based electrolytes, which demonstrate higher ionic conductivity than YSZ at intermediate temperatures. However, at low oxygen partial pressure, ceria-based electrolytes become electrically conductive as Ce⁴⁺ is reduced to Ce³⁺, resulting in low open-circuit voltages and performance degradation. In particular, ceria-based functional layers, such as gadolinia-doped ceria (GDC), are widely used because of their low reactivity with electrodes and high oxygen ion conductivity [[9], [10], [11], [12], [13], [14]]. Thus, a zirconia-based electrolyte is applied at the anode interface and a functional layer is applied at the cathode interface. This bi-layered composite structure electrolyte is produced by fabricating a functional layer on the YSZ electrolyte.

Among the various methods of manufacturing the GDC functional layer, the screen printing method is commonly adopted since it is relatively inexpensive and is applicable to a large area. After depositing the functional layer via screen printing, a post-heat treatment process is essential to obtain the desired material properties. A high-temperature (∼1400 °C) heat treatment is required to prepare a dense GDC layer that suppresses the cation diffusion side reaction between the cathode and the electrolyte. However, at this sintering temperature, a (Ce,Zr)O₂-based solid solution would be formed at the GDC/YSZ interface, reducing the ionic conductivity and significantly degrading the performance of the fuel cell [[15], [16], [17]]. Therefore, it is necessary to devise a method for fabricating a GDC functional layer that prevents side reactions between the electrolyte interfaces during the high-temperature post-heat treatment process [[18], [19], [20]].

To obtain materials with high density at low sintering temperatures, many studies have suggested the use of sintering aids, such as Li, Co, Bi, Cu, and Fe, when sintering GDC [[21], [22], [23], [24], [25], [26]]. Various methods for applying sintering aids have been reported, such as doping a sintering aid material on a base material via powder synthesis. Another suggested method is using a mixture of a base powder and a sintering aid additive solution or a mixture of a base powder and a sintering aid powder. These studies have successfully produced highly dense oxide electrolytes with satisfactory ionic conductivity results, even at low sintering temperatures. However, most of the electrolytes fabricated from previous studies at lower sintering temperatures in the free sintering condition were thick pellets, and very few studies have confirmed the fabrication of thin electrolyte films under the constrained sintering conditions. In the free sintering condition of pelletized or tape-casted samples, shrinkages have been reported in all directions during the high-temperature heat treatment process; therefore, it is easy to lower the sintering temperature and control the density by adding sintering aids. However, under the constrained sintering conditions, there is some shrinkage of the layer on the rigid substrate in the lateral direction; therefore, it is very difficult to obtain a dense layer [[27], [28], [29]].

To fabricate high-performance SOFC structures for commercialization, extensive research has been conducted on anode-supported SOFCs, which have exhibited superior performance compared with electrolyte-supported SOFCs even at low operating temperatures [[30], [31], [32]]. In anode-supported SOFCs, the thick anode acts as a supporting structure that provides mechanical strength, allowing electrolytes and other components to be fabricated as thin films. Conventionally, the co-sintering process is used after screen printing the electrolyte and functional layers onto the supporting structure to shorten manufacturing time. However, this co-sintering process results in side reactions between electrolyte materials, as mentioned above. Therefore, the functional layer and the electrolyte must be sintered separately, where the functional layer has to be manufactured under the constrained sintering condition. To date, adopting an anode-supported structure that simultaneously uses a GDC functional layer on the cathode remains a challenge.

In this study, we successfully developed a method for fabricating a dense GDC functional layer, even under the constrained sintering condition, using a sacrificial bismuth oxide, Bi₂O₃, sintering aid layer on top of the GDC layer. Unlike the conventionally used doping or lattice penetration methods that are generally used when applying sintering aids, a layered structure was used to increase the liquid phase sintering effect and improve the sinterability. Depending on the type of material used as the sintering aid, there is crystallinity of the base material or penetration between the lattices, which degrade the performance of the sintering aid. Therefore, the sintering aid layer was sublimated by selecting an appropriate sintering aid material and optimizing the final sintering temperature, leaving only the functional layer. This method prevents the reduction of ionic conductivity due to the residual sintering aid material [23,33]. The Bi₂O₃ sintering aid layer melts at a relatively low temperature (817 °C) than other oxide materials, and sublimation occurs at approximately 1000 °C [[34], [35], [36]]. Consequently, a dense GDC functional layer could be fabricated by utilizing the liquid phase sintering process. As thermal sintering progressed, densification and grain growth of the GDC proceeded after solid Bi₂O₃ turned into the liquid phase. When the final sintering temperature was achieved, Bi₂O₃ sublimated, leaving only the pure GDC interlayer. The generation of a secondary phase was confirmed through the surface and cross-sectional analysis of the fabricated cell. Applying the Bi₂O₃ sintering aid layer could lower the sintering temperature of the GDC functional layer to 1200 °C, forming a dense film even in the constrained sintering condition. This dense GDC functional layer prevented the formation of a secondary phase, and the cell test results showed that its performance was 12 times better than that of the sample in which the secondary phase was generated. Performance of the fuel cell comprising the layered Bi₂O₃/GDC structure was compared with that of a conventionally manufactured fuel cell, it was proven that significant progress could be made toward the commercialization of fuel cells by reducing manufacturing costs through lower sintering temperature, improving performances through suppression of side reaction, and overcoming technological limitations.