TOWARDS THE CONVERSION OF A SOLID OXIDE CELL INTO A HIGH TEMPERATURE BATTERY

C.M. Berger, O. Tokariev, P. Orzessek, A. Hospach, N.H. Menzler, M. Bram, W.J. Quadakkers and H.-P. Buchkremer

Institute of Energy and Climate Research (IEK), Forschungszentrum Jülich, Jülich, Germany

ABSTRACT

A Rechargeable Oxide Battery (ROB) is based on a high temperature Solid Oxide Cell (SOC) that works alternately in fuel cell and electrolyzer mode. Instead of externally storing the electrolytic hydrogen, a stagnant atmosphere composed of hydrogen and steam is used directly as an oxidizing and reducing agent for a metal-metal oxide material, which serves as integrated energy storage. As a consequence, all the expenses related to pumping losses, heat losses and further components can be avoided.

Storage components are manufactured by means of tape casting or extrusion using iron oxide (Fe2O3) based slurries or pastes with varying additions of CaO or CaCO3. Redox cycling as well as thermogravimetric studies and microstructural analyses of storage components using XRD and SEM have shown that the use of 28-66 mol.% CaO or CaCO3 provides promising redox cycle performance due to the reversible formation of mixed oxide phases.

INTRODUCTION

The need for energy storage technologies in the future electricity grid with an ever increasing share of renewable and thus fluctuating power sources is often stressed. In this context, also the use of hydrogen as a powerful energy carrier is emphasized. The major requirements for any storage technology are a high specific capacity, a high efficiency and a rapid response time. Additionally, a low impact on the environment, inherent safety and of course low cost is desirable. In order to meet these demands the development of a high temperature battery was reported by several authors [1, 2, 3]. Starting from a description of the general working principle of the battery the discussion in this paper focuses on the general development of the composition, the manufacturing process and the microstructure of the actual storage material.

Working principle

In a Rechargeable Oxygen Battery (ROB, Fig. 1) a solid oxide cell (SOC) operates alternately as fuel cell (SOFC) and as electrolyzer (SOEC). The SOC is augmented with a storage material which provides and absorbs reactants at the fuel side. Whereas a classical fuel cell or electrolyzer system needs continuous delivery and removal of fuel gases the integrated storage material is surrounded by a stagnant atmosphere. This enables the system to be operated as a battery in a semi-closed system, thus saving all cost related to gas delivery and conditioning. At the air side a simple passive air vent or a small fan is sufficient for transporting air to and from the electrode surface.

Fig. 1 Schematic of the working principle of an ROB in discharge and charge operation according to [4]

As shown in Fig. 1 in discharge operation a metal is oxidized by steam yielding hydrogen. In the same manner as in a fuel cell hydrogen is oxidized by oxygen ions that diffuse from the air electrode through the electrolyte. The electrons that are released in this reaction can be used to supply a load with electricity and the generated steam is used to further oxidize the storage material. The battery is completely discharged when the entire storage material is oxidized.

Conversely, in charging operation the metal oxide is reduced by hydrogen, leading to metal and steam at the fuel side which is further electrolyzed into hydrogen which acts as reducing agent for the metal oxide The battery is charged when all of the active storage material prevails in metallic state.

Operating Conditions

Since the ROB is based on a SOC the operating temperature lies around 800 °C. As mentioned above, the stagnant atmosphere surrounding the storage material consists of steam and hydrogen. During operation it is necessary to avoid local oxidation of nickel, which is part of the fuel electrode and also to avoid reduction of chromium oxide, which the main constituent of the protective layer on the surface of the ferritic steel interconnect.

Because of a common maximum fuel utilization of approximately 80% in fuel cell systems the ratio of steam to hydrogen is even further limited to 4:1 at the upper boundary and 1:4 at the lower boundary respectively. Within these ratios all the relevant reactions of the storage material take place (see Fig. 2). At 800 °C these gas compositions correspond to a range of oxygen partial pressure between 7-10-18 bar and 2.75-10-20 bar.

Fig. 2 Equilibrium oxygen partial pressure in various H2O/H2 mixtures as function of reciprocal temperature compared with data on the dissociation pressure of selected oxides, calculated with Fact Sage, [4]

THEORY

Generally, a number of metal/metal oxide systems is readily reduced and oxidized in the mentioned steam-hydrogen mixtures around 800 °C and could thus be applied in an ROB. Considering the thermodynamic stabilities, the metal-metal oxide systems based on germanium, tungsten, molybdenum and iron are potentially suitable candidates, depending on the exact range of oxygen partial pressure. However, due to the drawbacks of tungsten and molybdenum to form highly volatile or even liquid oxides and the quite high cost of germanium the present investigation focuses on iron based storage components.

Eq.1 describes the reversible reaction of the storage material whereas Eq.2 and Eq.3 represent the well-known reactions at the respective electrodes of the SOC.

Using an iron-based storage, a theoretical energy capacity of approximately 1300 Wh/kg with respect to the educts (iron and steam) can theoretically be achieved. However, pure iron rapidly degrades due to particle agglomeration and formation of oxide layers that lead to decreased reaction kinetics and eventually to irreversible capacity loss [5]. In order to mitigate degradation effects an additional oxide was introduced to the storage material.

For this purpose, several oxides were tested and benchmarked as described in [4]. The benchmarking process mainly dealt with experimental work consisting of repeated redox cycling tests and microstructural analysis of pressed samples as well as subsequent XRD measurements to estimate phase composition after various redox cycles. In this study especially the system Fe2O3-CaO was found to be a promising candidate because of apparent low degradation. Possibly, the reversible formation of mixed oxides leads to better degradation properties. However, the exact mechanisms were not fully clarified. In the following sections the system Fe2O3-CaO will be analyzed more closely.

EXPERIMENTAL

In all experiments reagent grade powders were used (Tab. 1). Before building a battery and testing it the main questions were, which composition and which starting powders (CaO or CaCO3) to use which manufacturing route to take and finally which sintering temperature to apply.

Table 1 In this investigation used reagent grade oxide powders

For analysis of the effect of calcia content on material behavior different compositions were prepared on the basis of experimental results described in [6, 7, 8] (Tab. 2). During variation of composition CaO was used, to avoid a corresponding change of porosity since CaCO3 would decompose into CaO and CO2 during sintering forming a different amount of pores, which could affect the later analysis.

Table 2 Label, composition and resulting theoretical stoichiometry of prepared samples.

Fig 3 shows the calculated phase diagram of the system Fe2O3-CaO in dependence of oxygen partial pressure at 800 °C, where in fact the mentioned phases are not considered although being described in the references [6, 7, 8] as well as in the XRD database.

Fig. 3 Phase Diagram CaO-FeO at 800°C calculated using FactSage software employing Databases FACTPS and FToxid

The indicated powder mixtures (Tab. 2) were pressed at 150 MPa to small pellets (0,5 g, ø10 mm) using a small scale laboratory press (P/O/WEBER GmbH, Germany). Pressing of specimens is useful when quick iterations and a general material benchmarking is necessary. For manufacturing larger...