electrolyte. The anhydrous cell does not have the graphite feeding problem so chlorine is more easily collected. Current efficiency at the outset can be of the order of 90%, but decreases as the semiwall deteriorates to a level of 75% to 85%. There is a trade-off between current efficiency and cell potential. As the cathodeanode spacing is decreased to minimize the voltage, the chlorine and magnesium rising in the cell are brought closer together and recombination is increased, lowering current efficiency. Thus, ignoring semiwall deterioration for the moment, there is an optimum electrode spacing. Halliday and McIntosh (5) have been able to demonstrate that the hydrodynamics of the electrolyte and thickness of the gas layer in a 2000-amp pilot magnesium cell could be simulated by flowing nitrogen through a porous electrode in an aqueous modeling system. This hydrodynamic technique may be extremely useful in simulating cell parameters under conditions of reduced gravity. While their fused salt cell used an LiCl-KCl electrolyte that is lighter than magnesium, the cell and aqueous system should be applicable with some modification to studying heavier electrolytes as well. Optimizing current efficiency with a possible decrease in cell voltage could add up to an effect equivalent to improvement of 10% in current efficiency. It has been shown that clean steel is preferentially wet by magnesium rather than fused salt. This leads to the concept of a porous cathode through which the magnesium deposited on the surface permeates and is collected on the back side. Such a system will allow minimal electrode spacing and could lead to cell potentials as low as 4.0 V. Obviously the gravitational force needed for collecting magnesium through a porous, wettable cathode is much less critical than at a solid cathode from which magnesium must float to the surface. Halliday and McIntosh (5) chose to use a low density electrolyte to simplify separation of chlorine and magnesium, since the latter dropped to the bottom of the cell. Electrodes were slightly sloped to enhance the separation as demonstrated in the simulated aqueous system and the fused salt cell. Current efficiencies of over 90% were obtained at current densities of 1.5 amp/cm2 (1400 amp/ft2), some three times the operating level of the conventional cells. Even at this current density, cell potentials of the order of only 5 V were obtained. Thus, use of these light electrolytes offers a way of improving cell efficiency with energy savings similar to those cited for a 4 V cell. Use of light electrolytes, based on a composition containing lithium chloride, has been patented by Dow. However, there appears to be no major effort devoted to use of a lighter-than-magnesium electrolyte at the present time. Aluminum cells operate with lighter-than-metal electrolyte, which is an advantage over conventional magnesium cells. It is probable that use of an electrolyte with lower density than the metal will facilitate separation of chlorine and magnesium under conditions of reduced gravity since the gas will rise and the metal sink. SPECIAL CONSIDERATIONS FOR MAGNESIUM PRODUCTION IN SPACE Feed Materials Prior discussion has indicated the suitability of MgO as a feed material. It may either be converted to the chloride, with recycling of chlorine, or used directly for reduction by silicon or ferrosilicon.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==