Long term operation is an essential subject in the commercialization of the Molten Carbonate Fuel Cell (MCFC). Material stability is important for the development of the MCFC, particularly for long term operatioa In this paper, the specification and the stabilization of MCFC matrix arc investigated, with the aim of producing 40000 hours of operation.

It is common knowledge that matrix thickness has a large influence on shorting time, as shorting is caused by the dissolution of the nickel oxide cathodes (1,2). Therefore, the optimum thickness of a matrix designed for 40000 hours operation without the nickel shorting was sought The influences of different electrolytes and matrix specifications on the shorting time were measured with accelerated cell tests. The internal resistance of the matrix was also estimated.

Gamma( y )-lithium aluminate (LiAlOi) powder with a sub-micron particle diameter is commonly used for a raw material of matrix to retain molten carbonate electrolytes. This is because most researchers found that T -LiAIQ> was the most stable material in the MCFC environment among the three allotropic forms alpha ( a ), beta ( j3 ), and y (3,4). However, two problems with the stability of y -LiAlOi are being vigorously discussed, especially in Japan: particle growth causes decreasing electrolyte retention (5,6), and the transformation of y to a (6,7). This transformation contradicts the accepted opinion that y is the most stable form. In this paper, the particle growth and the phase transformation of LiAlOi are examined with post-test analyses. The influence of matrix degradation on cell performance is also considered.

Accelerated single cell tests have been performed to evaluate the nickel shorting time for three cells with the different electrolytes and matrices. The specifications arc shown in Table 1. To accelerate the shorting, thinner matrices (0.5 ± 0.01 mm) were used, and the average partial pressure of the catbon dioxide in the cathode gas (Pco:_avg) was 1.72 atm. The anode was made of Ni-Al alloy. Lithiated nickel oxide was used for the cathode. The matrix was made of y -LiAIOa by tape casting method. A comparison of shorting times of Li/K and Li/Na carbonate was carried out by testing cells A and B (Table 1). A new matrix which had a structure delaying the shorting time was used with the Li/Na carbonate in cell C Each cell was usually operated at 150mA/cm 2 . The operation was continued until the open circuit voltage obviously dropped due to the nickel shotting.

Post-test analyses were performed to estimate the in-cell stability of the LiAlOi for several

0 0

Post a comment