Anodes of elemental carbon may be discharged in a galvanic cell using a molten carbonate electrolyte, a nickel-foam anode-current collector, and a porous nickel air cathode to achieve power densities of 40-100 mW/cm 2 . We report cell and anode polarization, surface area, primary particle size and a crystallization index for nine particulate carbon samples derived from fuel oil, methane, coal, charred biological material and petroleum coke. At 800 °C, current densities of 50-125 mA/cm 2 were measured at a representative cell voltage of 0.8 V. Power densities for cells with two carbon-anode materials were found to be nearly the same on scales of 2.8-and 60 cm 2 active area. Constant current operation of a small cell was accompanied by constant voltage during multiple tests of 10-30 hour duration. Cell voltage fell off after the carbon inventory was consumed. Three different cathode structures are compared, indicating that an LLNL fabricated porous nickel electrode with <10 µm pores provides improved rates compared with nickel foam with 100-300 µm pores.Petroleum coke containing substantial sulfur and ash discharges at a slightly lower rate than purified petroleum coke. The sulfur leads to degradation of the anode current collector over time. A conceptual model for electrochemical reactivity of carbon is presented which indicates the importance of (1) bulk lattice disorder, which continually provides surface reactive sites during anodic dissolution and (2) electrical conductivity, which lowers the ohmic component of anode polarization.

An investigation was conducted into alternative techniques for zinc and electrolyte regeneration and reuse in the refuelable zinc/air battery that was developed by LLNL and previously tested on a moving electric bus using cut wire. Mossy zinc was electrodeposited onto a bipolar array of inclined Ni plates with an energy consumption of 1.8 kWh/kg. Using a H 2depolarized anode, zinc was deposited at 0.6 V (0.8 kA/m 2 ); the open circuit voltage was 0.45 V. Three types of fuel pellets were tested and compared with results for 0.75 mm cut wire: spheres produced in a spouted bed (UCB); coarse powder produced by gas-atomization (Noranda); and irregular pellets produced by chopping 1-mm plates of compacted zinc fines (Eagle-Picher, Inc.). All three types transported within the cell. The coarse powder fed continuously from hopper to cell, as did the compacted pellets (< 0.83 mm). Large particles (> 0.83 mm; Eagle-Picher and UCB) failed to feed from hopper into cell, being held up in the 2.5 mm wide channel connecting hopper to cell. Increasing channel width to ~3.5 mm should allow all three types to be used.Energy losses were determined for shorting of cells during refueling. The shorting currents between adjacent hoppers through zinc particle bridges were determined using both coarse powder and chopped compressed zinc plates. A physical model was developed allowing scaling our results for electrode polarization and bed resistance Shorting was found to consume < 0.02% of the capacity of the cell and to dissipate ~0.2 W/cell of heat. Corrosion rates were determined for cut wire in contact with current collector materials and battery-produced ZnO-saturated electrolyte. The rates were 1.7% of cell capacity per month at ambient temperatures; and 0.08% of capacity for 12 hours at 57 °C.The total energy conversion efficiency for zinc recovery using the hydrogen was estimated at 34% (natural gas to battery terminals)--comparable to fuel cells. Producing zinc shot was quoted at 1.5-3 ¢/lb above base price (52¢/lb, ASM) for super purity ingot.Both the spouted-bed and the Eagle-Picher processes might conceivably be configured for fleet operation in user-owned and operated equipment located a the fleet's home base. This would eliminate the need for green-field industrial plants and fuels distribution systems. Scaleup of the spouted bed process and detailed examination of the Eagle-Picher process are recommended.

A model is formulated by combining carbonate dissociation with pre-existing anode mechanisms involving heterogeneous reaction kinetics. The proposed model accounts for both the observed preponderance of CO 2 evolution and dependence of rate on carbon anode microstructure. Implications of the model for the design of carbon batteries and fuel cells are discussed, and the laboratory cells used in earlier research are described. High coulombic efficiencies for the net reaction C + O 2 = CO 2 require severely limiting the thickness of paste anodes in powder-fed fuel cells while the unreacting surfaces of solid prismatic anodes must be isolated from the CO 2 product atmosphere to prevent Boudouard corrosion, according to C + CO 2 = 2CO.

No abstract