| Diagrams 1 |
 |
 |
| This is the device to start with. It uses raw regolith (after roasting volatiles out of course) and electricity to make oxygen, Fe-Si alloy in metal ingots and ceramic bricks that can be used to build things including blast furnaces and as a source of other materials like CaO and even magnesium and aluminum, but we will probably use the bricks mostly for bricks. Oxymoronic?? Dr. Larry Haskin wrote that if the voltage was too high magnesium would come out of the magma. Mg b.p. 1120 C. In the furnace above it would react with oxygen and a fire would result Also, this could ruin the MgAl2O4 spinel of the ceramic. |
 |
| Molten silicate (magma) electrolysis. That's what we like. 5000 metric tons raw regolith, no benficiation (except roasting out volatiles) per year yeilds 1000 tons O2 from a 70 ton plant using 3 MWe. We could scale it down to a 7 ton or less plant and use 300 kWe or make it smaller even. The fluxed molten silicate electrolysis uses 3.5 MWe and amasses 80 tons but runs at lower temps. Pretty sure about that. The flux is LiF + CaF and if we don't need them why bother? Run higher temps. Just got to make the furnace right. |
 |
 |
 |
 |
 |
| Here, the use of zirconia cells to electrolyze CO2 back to CO and Oxygen is pictured. |
| Here, inert gas is used to blow oxidized sodium, potassium and possibly even magnesium vapors from magma furnaces. A ferrochrome catalytic reactor and the heat of off gas CO2 breaks the CO2 down into CO and oxygen to recycle them. |
 |
 |
 |
 |
 |
| Lost and found this article after long time: http://epsc.wustl.edu/admin/research/psmrg/nvm3_00/a_m00h02.pdf We know the power required (~13 MWh per tonne O2), the overvoltage required (~0.2V), the practical upper limit to cell voltage (lest Mg2+ be reduced and volatilized), effects of surface tension and how to relieve them (clinging of O2 bubbles at the anode, balling of metal at the cathode),compositional effects of magma composition, suitablecontainer composition and stability, suitable electrode composition and stability, etc |
| David A. Dietzler, 2007 |