Results And Discussion

1) In-situ X-ray diffraction analysis of the degradation process of alloy catalysts. In order to monitor the whole corrosion process of alloy catalysts and also to get the catalysts at any corrosion levels, a specially designed cell was prepared, which is similar to a conventional single cell consisting of graphite plates with gas flow channels but added a window covered with a polyimide film on one of the plates for the X-ray diffraction. The paste (Pt/Co/Ni alloy supported on carbon black + 100% H3P04) sandwichcd between thin carbon papers was mounted to the window side of the cell as a working electrode, which was paired with the conventional gas-difiusion electrode supplied H2 gas, across a SiC matrix electrolyte layer filled with H3P04. The corrosion of the alloy catalysts was performed under controlled potential, temperature and humidity conditions. The results were discusscd in combination with the post-analysis, such as the chemical analysis of their compositions or TEM and EPMA analysis of the catalyst particles. The following results have been found;

a) The whole corrosion process under any condition can be monitored by in-situ X-ray diffraction and samples at any corrosion levels can be obtained.

b) Under the condition near the practical PAFC operation, the dominant factor affecting the catalyst degradation is the temperature; e.g., the effect by the elevation of temperature from 200°C to 240°C is equivalent to that of the potential elevation from 0.7 V to 0.9V.

c) The degradation appeared both in the non-noble metal components loss at the gross alloy composition and the surface area loss by the growth of the particle size.

d) Accompanied with the corrosion, a Pt single phase increases with the decrease of the original Pt/(Co+Ni)=l/l phase of fee, but intermediate phases with some distributions, which are simply abbreviated as Pt/(Co+Ni)=3/l phase of fee, are kept constant at the integrated intensity of the diffraction peaks and also the mean particle size. The corrosion mechanism is under clarification in combination with other analytical methods [1].

2) Effect of the electrolyte flll-lcvel on electrode performance.

A conventional gas-diffusion electrode at PAFCs is composed of a mixture of fine particles of catalyst supported carbon black, abbreviated as Pt/CB, and polytetrafluoroethylene (PTFE), which forms the electrolyte- and gas-network, respectively. Both the gas permeability and utilization of the catalyst are controlled by a fill-level of electrolyte in the electrode. It has been recognized that optimization of the catalyst utilization is essential for high cell performance and over-filling brings about a shorter life-time. However, due to the lack of an appropriate and convenient method to evaluate the fill-level experimentally, there have been few reports [2,3] dealing with the relationship. We have proposed a new evaluation method of the fill-level, or the degree of Pt/CB wetted with the electrolyte [4], based on a double layer capacitance at the surface of CB itself by masking Pt surface with CO poisoning, because the former is considered to be much more stable than the latter during the long term operation of PAFCs. A cell, which can introduce the electrolytes of different concentrations and different reactant gases, was specially designed. The cell is not only use for the measurements of conventional polarization curves or 02 gain at high temperature in 100% H3P04 but also for the measurements of the double layer capacitance as well as the Pt surface area by the conventional hydrogen adsorption method at room temperature in 40% H3POj. By repeating the process, the electrolyte fill-level was substantially accelerated. All of the measurements are performed automatically by a computer controlled system. The major results found so far are:

a) The double layer capacity for CB becomes a good measure for the electrolyte fill-level evaluation.

b) Each contribution of the gas-diffusion loss and the catalytic activity loss to the cell performance losses during the cell operation could be separated clearly by the proposed experimental method.

c) The critical fill-level leading to the life-time has been found at the standard electrode, where the gas-diffusion loss becomes predominant; the primary pores (<ca. 0.09 JUL) formed between the primary CB particles were filled with electrolyte by 45% but the secondary pores

0.09 JUL) formed between the CB agglomerates and acting as gas-networks were filled up to 70%, although those at the optimum condition were 44% and 16%, respectively.

3) Corrosion of cell constructing materials and the effect on the electrode performances.

In this project, we focused our attention on the corrosion of CB substrate in Pt/CB under various conditions in comparison with that of Pt alloy particles supported. Electrochemical and physical methods have been used to separate the changes in the surface areas of CB and Pt, respectively.

4) The rate of electrolyte loss under various operating conditions of the model cell.

At the model cell, the electrolyte (PA) loss into exhaust gases of the anode and cathode has been measured by condensing it in cold traps and analyzing the amounts as a function of the operating temperature, utilization of reactant gases or current density. We are also going to identify species of phosphoric acids in the exhausting gases with a mass-spectrometer of ultra-high sensitivity. The following have been found:

a) The amount of electrolyte loss increases with an increase of the operating temperature at a constant utilization of each anode and cathode reactant gases and can be characterized with a mean linear velocity of the gas flow in ribs of the cell plates, e.g., at the cell operation of 220°C, U¿,=60% and Um=80%, the concentration of PA in the cathode exhausting gas was nearly constant (ca. 27mg/Nm3) in the region of 7-60 cm/s but that of the anode reaches nearly to the equilibrium concentration (ca. 57mg/Nm3) at 1 cm/s and decreases with an increase of the velocity and coincides with that of the cathode in the region larger than 13 cm/s. These precise data must be very useful for the simulation of the amount of PA loss during the practical operation of large size PAFCs and for the cell operation or the cell design taken into account better electrolyte management.

0 0

Post a comment