Performance and testing of glass-ceramic sealant used to join anode-supported-electrolyte to Crofer22APU in planar solid oxide fuel cells

Abstract

This work describes the performance and testing of a glass-ceramic sealant used to join the ceramic electrolyte (anode-supported-electrolyte (ASE)) to the metallic interconnect (Crofer22APU) in planar SOFC stacks. The designed glass-ceramic sealant is a barium and boron free silica-based glass, which crystallizes by means of the heat-treatment after being deposited on substrates by the slurry technique.

Joined ASE/glass-ceramic seal/Crofer22APU samples were tested for 500 h in H₂–3H₂O atmosphere at the fuel cell operating temperature of 800 °C.

Moreover, the joined ASE/glass-ceramic seal/Crofer22APU samples were submitted to three thermal cycles each of 120 h duration, in order to evaluate the thermomechanical stability of the sealant.

The microstructures and elemental distribution at Crofer22APU/glass-ceramic and ASE/glass-ceramic interfaces were investigated.

SEM micrograph observations of joined samples that underwent cyclic thermal tests and exposure for 500 h in H₂–3H₂O atmosphere showed that the adhesion between the glass-ceramic and Crofer22APU at either interface was very good and no microstructural changes were detected at the interfacial boundaries.

The study showed that the use of the glass-ceramic was successful in preventing strong adverse corrosion effects at the Crofer22APU/glass-ceramic sealant interface.

Introduction

Solid oxide fuel cells (SOFCs) are highly efficient energy conversion devices which produce electricity by the electrochemical reaction between a fuel and an oxidant. Among the different SOFCs, the planar type, which is expected to be cost effective and mechanically robust, offers an attractive potential for increased power densities compared to other concepts. A fuel cell device consists of an anode electrode (exposed to fuel), an electrolyte, and a cathode electrode (exposed to oxidant) [1], [2].

The repeating unit of a planar solid oxide fuel cell (SOFC) is formed by anode–electrolyte–cathode and interconnects. The interconnect links the anode of one cell to the cathode of the neighbouring cell [3]. In most SOFC stack designs, the interconnect is sealed to the ceramic cell components [4]. A promising interconnect material is chromia-forming ferritic stainless steel. However, the seal between the stainless steel metal interconnect and the ceramic SOFC components presents a challenge [5], [6].

The sealants for planar SOFCs must meet some important requirements: they have to be hermetic in order to prevent mixing of the fuel and oxidant and should have a thermal expansion coefficient close to those of the interconnect and the electrolyte. Moreover, the sealant must be mechanically and thermochemically stable in both oxidizing and wet-reducing environments at 800 °C and must not undergo any reaction with the other cell components. The problem becomes even more challenging as there is also a requirement for thermal cycle stability for planar stacks in which dissimilar SOFC components are sealed together. The sealants have to survive for several hundreds of thermal cycles during SOFC operations. Any cracks that form in the sealants or at the interfacial regions can cause leakage that leads to lower cell performance and efficiency. A number of different approaches are currently being studied for sealing SOFCs including [7]: brazing [8], [9], [10], compressive seals, as well as glasses, glass-ceramic seals and glass-composite seals [11], [12], [13], [14], [15], [16], [17], [18], [19].

Glass-ceramics can be prepared by controlled sintering and crystallization of glasses and have superior mechanical properties and higher viscosity at the SOFC operating temperature than glasses. Furthermore, they can have thermal expansion coefficients very different from the parent glass, due to the different crystalline phases that form and their relative proportion. Glass-ceramics show better resistance to the severe service environment (both oxidizing and reducing) than brazing alloys, and by carefully choosing the glass composition, they can meet most of the requirements that need to be exhibited by the ideal sealant material.

In order to develop a suitable glass-ceramic sealant, it is therefore necessary to understand the crystallization kinetics, the sealing properties and the chemical interactions with other components of the cell. Barium aluminosilicate sealants have shown high reactivity with the metallic interconnect at 800–900 °C forming a porous and weak interface composed of barium chromate (BaCrO₄) and monocelsian (BaAl₂Si₂O₈), while borate glasses are not sufficiently stable in a humidified fuel gas environment [20], [21].

Moreover, glass-ceramic seals have good hermeticity and are thermally and environmentally stable. However, the inherent brittleness of glasses may cause cracks to develop in seals during thermal cycling or thermal shock. This can cause leakage that would lead to lower cell performance and efficiency.

Owing to these problems that have been observed for barium aluminosilicate sealants, the authors have developed an alternative sealant based on a sodium–calcium–aluminosilicate glass-ceramic [22] consisting of gehlenite (Ca₂Al₂SiO₇) and a sodium aluminosilicate (NaAlSiO₄₎ phase. In the study that is presented here, the testing and performance of the sodium–calcium–aluminosilicate glass-ceramic sealant that was used to join the Anode-supported-electrolyte to Crofer22APU in planar SOFCs are reported. The investigation involved static treatments in humidified hydrogen and thermal cycling in air for 500 h; this can be considered as a preliminary screening on the glass-ceramic sealant performance, taking in account that the target operational life of the cell should be thousand of h (e.g., 10 kh).