Lunar soil, regolith, or Moon dust is mostly oxygen and silicon along with iron, calcium, aluminum, magnesium, titanium and traces of chromium, manganese, sodium, potassium, phosporus,sulphur and miniscule amounts of many other elements. It also contains traces of hydrogen, helium, methane, CO, CO2 and nitrogen implanted by the solar wind that can be extracted by mining millions of tons of regolith and heating it to about 700 C. Numerous processes for extracting and separating these elements have been described. Ideally, it should be possible to heat regolith until it decomposes and separate the elements with something similar to a mass spectrometer. Dr. Peter Schubert has designed a device pictured above that heats and decomposes regolith into oxygen and other elements like iron, aluminum and silicon. A slag of calcium and magnesium oxides forms too. These are all separated in what he calls an "All Isotope Separator" (not pictured here) that resembles a mass spectrometer or calutron for enriching uranium. This device holds a lot of promise. If it fails to meet up to expectations economically, Moon miners will have to use a different process. Proposals involving the use of HF acid, fluorine gas and/or many other chemical reagents not available on the Moon have been made. Below is an attempt to deliniate an alternative process for extracting resources from lunar regolith that relies on heat and electricity moreso than chemicals.
Producing Lunar Materials with Heat and Electricity
Numerous processes for extracting oxygen and metals from lunar regolith have been proposed. Most of them use chemical reagents like HF, fluorine, chlorine and other substances not common on the Moon. A process that doesn't require water and/or lots of corrosive and imported chemicals is desired. Solar wind implanted volatiles like hydrogen, helium, carbon and nitrogen can be extracted by roasting huge quantities of regolith at 700 C. Mare regolith can be excavated, pressed into forms and sintered with heat from solar or electrical furnaces, or melted and cast to make numerous basalt items. Highland regolith can be sintered or cast to make glassy products with a melting point of about 1500 C. Meteoric iron fines can be magnetically extracted from large amounts of regolith just about anywhere on the Moon. After some grinding, screening and a second magnetic separation, 99% pure iron-nickel particles can be obtained . This can be refined further with carbon monoxide to extract nickel. Another possibility is the use of a device similar to a mass spectrometer to separate iron, nickel, cobalt, gallium, germanium and platinum group metals from this meteoric material. It will also be possible to melt and cast the iron fines in sand molds and it should be possible to run these particles in 3D printers after sizing them to make pure iron items for low stress applications. It will also be possible to melt and cast the iron into rods, pack them in carbon powder, and get them red hot for a few days in solar or electrical furnaces to convert the iron into strong steel. So far, so good. Mining the Moon for these materials doesn't require any water, acids, halogens or other substances rare or practically non-existent on the Moon.
Heating regolith to 700 C. will release hydrogen and helium. Carbon will react with oxygen in the regolith to form CO and CO2 gases. Hydrogen will react with carbon to form methane and it will react with oxides to form water. At the University of Wisconsin in Madison, scientists have designed a machine that can dig up and heat regolith from a square kilometer of lunar surface to a depth of three meters in a years' time. They predict harvesting of substantial quantities of volatiles .
Materials Harvested per Year in Tonnes
Polar ices may also serve as a source of water and thus hydrogen and oxygen. The ice might also contain ammonia (NH3), methane (CH4), carbon monoxide and carbon dioxide.
Sulfur, potassium and sodium can be roasted out of large quantities of regolith by heating the Moon dust up to 1200 C. and possibly higher . At 1200 C. in the vacuum FeO will volatilize. At 1500 C. and hotter SiO2 and MgO will boil out. If anorthite is heated to 2000 C. in the vacuum SiO2 will boil out and calcium aluminate, CaAl2O4 will remain . Calcium aluminate can be used to make hydraulic cement. Plain sand mixed with molten sulfur also makes a kind of cement. Regolith should work as well as terrestrial sand.
What about oxygen? Regolith will give up oxygen simply by heating it to 2000 C. and hotter. Ilmenite obtained by electrostatic refining of high-Ti mare regolith can be reduced with hot hydrogen gas. The resultant water can be electrolyzed to gain oxygen and recycle hydrogen. Magma electrolysis at 1600 C. can release oxygen and produce ferrosilicon and a spinel rich ceramic. This last process requires platinum anodes and equipment that can handle high temperatures. An actively cooled magma cell and oxide ceramic electrodes instead of platinum might be used. More research is called for.
At this point, Moon miners will have hydrogen, helium, carbon monoxide, carbon dioxide, methane and nitrogen, basalt, a glassy ceramic, iron, nickel, sodium, sulfur, potassium, oxides of iron, silicon and magnesium as well as cement and oxygen. Silicon dioxide is glass, a useful material. Add some Na, K and CaAl2O4 and we can make a lower melting point glass that is more workable. Given a way to separate SiO2 and MgO from roasted regolith, perhaps electrostatic separation, we can take MgO and reduce it with ferrosilicon at 1200 C. under low pressure to obtain magnesium. It should also be possible to reduce FeO with hot hydrogen to get iron and oxygen. What can we do with basalt, ceramics, glass, cement, iron and magnesium on the Moon? Since these can be obtained with purely electrical and thermal processes without imported chemicals it seems these will be comparatively inexpensive.
Steel could be made on the Moon if iron is available. Carbon from mining for solar wind implanted volatiles and/or polar ice mining can be combined with iron by the simple "crucible" or "blister steel" process. Since steel is 0.05% to 1.5% carbon, a little carbon can make a lot of steel. When pure iron won't do, we could make some steel.
What about titanium? We can get titanium without water but we need imported chlorine to make CaCl2 electrolyte for FFC process cells. We will need carbon for electrodes and a system to recycle the carbon as it burns up...unless we can use imported non-consumable electrodes made of tin oxide or calcium ruthenate. If the CaCl2 electrolyte breaks down we will need to recover and reuse the chlorine. The chlorine would be shipped to the Moon in salt form and this will cost thousands of dollars per pound; therefore titanium might be rather expensive.
Aluminum can be obtained on the Moon. It might be possible to reduce CaAl2O4 with hot potassium and sodium. If this is true, then aluminum could be produced without water and without imported chemicals and the price of aluminum should be reasonable.
Highland regolith consists mostly of anorthite (CaAl2Si2O8). There is no bauxite on the Moon because this is a sedimentary rock and there has never been any water on the Moon. Anorthite could be concentrated using electrostatic separation processes. I've read about aluminum production using a complex process that involves leaching anorthite with H2SO4, carbochlorinating the aluminum sulfate and electrolysis in a flux of lithium and sodium chloride. Sulfuric acid and carbon could be obtained on the Moon and recycled. Chlorine and lithium would be imported and also recycled. This process is complicated and puts precious water at risk. It is also possible to electrolyze CaAl2O4 in an electrolyte of LiF and CaF with electrodes made of some exotic metals that was developed by EMEC Consultants . Imports would be required but water would not, so I prefer the EMEC process to the acid leach-carbochlorination-electrolysis process. Even so, it would be preferrable to use a process that doesn't require any imports of chemical reagents. Something resembling a mass spectrometer or simply heating anorthite to drive off the silicon to get calcium aluminate (CaAl2O4) and reducing this with lunar sodium and potassium seems more desirable.
It seems then that we can get all the major lunar materials with purely thermal and electrical processes, but we need some imports for titanium and perhaps for aluminum. The chemistry of these processes is not to complicated, but the solar and electrical furnaces, electrolysis cells, retorts and other pieces of equipment are going to be a challenge for chemical engineers to figure out how to construct on the Moon with lunar materials. One mechanical engineer that I know designed a silicon carbide furnace that could reach 2000 C. and concluded that there would be a lot of thermal management problems. Heat radiating from the furnace could actually melt nearby regolith! We will need to look at more "bone headed" solutions like building a furnace from a massive pile of bricks that has so much "thermal inertia" that it never gets red hot and instead of using conduction to heat the whole furnace as well as the material in the furnace we could just hit "ore" with an electron beam in the free vacuum or use radio waves (induction heating). Ceramic furnaces and magma electrolysis cells could have cooling passages drilled in them and CO2 could be used as coolant. The waste heat could be disposed of with metal space radiators coated by vapor deposition with a very thin layer of dark basalt to increase emissivity. Regolith is such a good insulator that waste heat cannot be rejected into the ground.
Since basalt will be so easily obtained it seems that there should be a focus on using basalt whenever possible. See: Unique Lunar Products
1) Dr. William Agosto. "Lunar Beneficiation." http://www.nss.org/settlement/nasa/spaceresvol3/lunarben1b.htm.
2) Matthew E. Gajda et al. "A Lunar Volatiles Miner." nasa-academy.org/soffen/travelgrant/gadja.pdf
3) Handbook of Lunar Materials, NASA Reference Publication 1057, 1980, pg. 115, http://www.nss.org/settlement/moon/library/1980-HandbookOfLunarMaterials.pdf
4) Rudolf Keller and D.B.Stofesky. "Selective Evaporation of Lunar Oxide Components." Space Manufacturing 10, pp. 130-135, Abstract: http://ssi.org/ssi-conference-abstracts/space-manufacturing-10/
5) Christian W. Knudsen and Michael A. Gibson, Processing Lunar Soils for Oxygen and Other Materials, http://www.nss.org/settlement/nasa/spaceresvol3/plsoom1.htm