## 2

Assuming quasi-steady state for the deposition process, material balance of reactant i in the cylindrical pore gives:

where J-, is the flux of i. Transport of Ni2+ through the electrolyte in the matrix pore can be described by diffusion, migration and convection. Since the migration of Ni2+due to the gradient of electric field in the matrix is minor [1], we neglect migration term in Jv Then flux of species i becomes:

where D, denotes an effective diffusion coefficient of i, and v denotes the convective flow rate. Effective diffusion coefficient is correlated with bulk diffusivity (Dib) by:

where x is tortuosity factor and e is porosity of the matrix. Since the poroity is a function of pore radius, Dj changes with axial position and time during the deposition. The convective transport takes place by the carbonate ion flow due to the fuel cell reaction in the anode and cathode. The amount of carbonate ion transported from the cathode side to the anode side is proportional to the current density (I/A) and the electrolyte velocity can be expressed as:

Therefore, flux of Ni2+ or H2 becomes :

The boundary conditions at both ends of the matrix are:

CNi2+ at the cathode-electrolyte interface can be calculated assuming that NiO dissolution reaction:

NiO + C02 Ni2+ + CO,2- (16) -n is always in equilibrium. Then Cne+ is proportional to the partial pressure of C02 at the cathode that is the average partial pressure of C02 determined by the material balance in the cathode for given current density. Cm" can be obtained from the Henry's law using average partial pressure of H2 at the anode side. The differential equation (2) and (6) together with initial and boundary conditions (3), (14) and (15) were solved by finite difference method. The amount of Ni deposited in the matrix was obtained by integrating the pore radius along the pore axis. The kinetic parameters were optimized so that the evolution of the amount of nickel in the matrix during the cell operation was fitted well to the 3cm2-cell data from the literature [1,2].

### Distribution of Ni in the matrix

The distribution of Ni in the matrix depends on the operating conditions and matrix properties. We concentrate in this abstract on the effect of gas composition on the Ni distribution. Fig. 2 and Fig. 3 show the distribution of Ni in the matrix at 2,000h as a function of cathode C02 partial pressure (Pco2) and anode hydrogen partial pressure (Pm). respectively. The location of Ni deposit moves toward the anode side with increasing Pco2- In addition, the amount of Ni deposited in the matrix increases with increasing Pco2- Similar results were obtained when Pm at the anode side were varied. As P»2 decreases, the film location moves to the anode side and Ni distribution becomes broad.

It was found that the amount of Ni deposited in the matrix of a 3cm2 cell was higher than the amount found in the bench-scale cell or stack even if the same feed compositions were used [1]. Since 3cm2 cell is usually operated at lower utilization, average Pco2 and Pm in the 3cm2 cell should be higher than those in the bench-scale cell or stack. Since the amount of Ni increases with increasing Pco2 and PH2 as shown in Fig. 2 and Fig. 3, it is expected that the amount of Ni deposited in the 3cm2 cell is higher than that in the bench-scale cell or stack. Furthermore, the simulation results indicate that the thickness of Ni deposit is lower for the 3cm2 cell.

Previously, Kasai and Suzuki found from the co-flow type 100cm2-cell experiment that the Ni distribution in the matrix was not uniform along the gas flowing direction [7]. For the matrix placed near the gas inlet, Ni was deposited in a narrow region close to the cathode side. On the other hand, Ni deposit was found in a broad region close to the anode side of the matrix near the gas outlet.. These results can be clearly explained in view of the results obtained from the simulation. Since fuel utilization is usually higher than C02 utilization in the bench-scale cell experiment, gradient of PH2 will be higher along the gas flowing direction. Near the gas inlet, PH2 is relatively high and Ni deposition would occur far from the anode side of the matrix. As Ph2 decreases along the gas flowing direction, the location of Ni deposition moves to the anode side as shown in Fig. 3. The Ni distribution is relatively broad in the matrix near the gas oudet since both Ph2 and PCo2 are low at this location.

### References

1. D.A. Shores, J.R. Selman, S. Irani and E.T. Ong, Proceedings of the second Symposium on the MCFC Technology, Electrochem. Soc., Pennington, NJ, p290 (1990).

2. J.B.J. Veldhuis, S.B. van der Molen, R.C. Makkus and G.H.J. Broers, Ber. Bunsenges. Phys. Chem., 94,947 (1990).

3. H.R. Kunz and J.W. Pandolof, J. Electorchem. Soc., 139,1549 (1992).

4. Y. Mugikura, T. Abe, S. Yoshioka and H. Urushibata, J. Electorchem. Soc., 142,2971 (1995).

5. G.R. Gavalas, C.E. Megiris and S.W. Nam, Chem. Eng. Sci., 44,1829 (1989).

6. Y.S. Lin and A.J. Burggraaf, J. Membrane Sci., 46,3067 (1991).

7. H. Kasai and A. Suzuki, Proceedings of the second Symposium on the MCFC Technology, Electrochem. Soc., Pennington, NJ, p240 (1990).

0 0