HYPOTHETICAL, ENTIRELY
       
David A. Dietzler, 2007
On other pages in the Lunar Resources section I have described a sulfuric acid leaching process followed by roasting of Al2(SO4)3 to Al2O3 , carbochlorination to AlCl3 and electrolysis or solar carbothermal reduction.  I have suggested using formed CaSO4 to make CaO flux by hi temp calcining.  This whole process is very complicated and requires elablorate devices as well as various chemical reagents. If only there was a simpler way that used fewer chemicals.  Hi temp magma electrolysis might be the way to get Al. Ca, etc. or fluxed electrolysis at lower temp. but these involve materials problems that must be solved.

The following is wild speculation:

1) Locate and mine 95% pure anorthite.  CaAl2Si2O8

2) Heat to 2000 C. to 2200 C. with focused solar rays as proposed by Dr. T.D. Lin to make a cement worthy formula approx.  40% CaO  50% Al2O3  10% SiO2 (1)

3) Sinter the powder into lumps or briquettes.

4) Load into Direct Reduction Furnace.

@ 2000 C  Al2O3 + 3C ===>  2Al + 3CO

@ 1600 - 1800 C.  SiO2 + 2C ===> Si + 2CO

@ 2200 C.  CaO + C ===> Ca + CO   Don't operate furnace at temps this high.   CaO is more valuable than calcium metal as an iron and steel making flux. 

Just go up to 2000 C.   The furnace will use CO gas and this will react CO + CO ===> CO2 + C to form the carbon that reduces the alumina and silica. CO2 and CO off gas will go into a ceramic heat exchanger to preheat incoming CO gas.  The CO2 and CO will then go to a Reverse Water Gas Shift Reactor, bascially just a chamber with an iron-chrome catalyst where it is mixed with hydrogen gas to form H2O and reform CO gas for reuse.

The liquid Al / Si alloy will pour out.  Further purifcation will yeild almost pure Al, although the Al/Si alloy may have many uses too.

There is an alloy called Alusil made in blast furnaces.  see: http://www.indiainfoline.com/sect/alum/ch04.html 

to quote the article: "
Direct Reduction Process adopts the blast furnace route to reduce clay/bauxite with coke and produce 70 % aluminum and 30 % silicon alloy-Alusil. The main advantage of the process is the substitution of expensive electricity by cheaper thermal energy.
Carbo-Thermic Process for the production of aluminum may either use bauxite or alumina.

In the case of bauxite the material is reduced in an electrical arc furnace whereby fused alumina is produced and impurities are removed. The fused alumina is then reduced in another furnace in the presence of pure carbon where aluminum is produced. The process consumes lower energy as compared to the Hall and Herault Process."

I can find no more information about the Alusil blast furnace production process on the web or in a Missouri wide university library search with Mobius.  Of course, by not being in school presently I am locked out of most journal databases.  My question is, does direct contact with hot coke reduce the alumina and silica rich clay or does the CO generated by the burning coke do the redux as it does with iron ore in a blast furnace?   I envison something like a reflux furnace to capture Al vapors because this is a problem because the redux takes place at almost the b.p. of Al. Perhaps inert gas pressure could prevent Al volatilization.  Much research remains to be done, but this could be an excellent way to produce Al-Si alloy for structural purposes on the Moon. 

Solid CaO will be removed thru a tap hole.

Al mp ~ 700 C  bp 2300 C.
Si mp 1700 C  bp  3300 C.

eutectic alloy should have lower mp and bp

CaO mp 2570 C.

While this process may seem complex in some ways, it does not demand mining for hydrogen and sulfur and making H2SO4 that must be recycled with elaborate devices and plumbing.

No chlorine imported from Earth.

No carbochlorination step  to make AlCl3 and recycling of carbon by shifting CO gas to methane and water and pyrolyzing CH4 and electrolysis of H2O.

No AlCL3 electrolysis and related power equipment.

No roasting of CaSO4 to CaO or electrostatic separation of CaSO4 and SiO2.

Though the CO gas recycling system for the process described here seems complex, it is a constant flow system and needs no reagents other than hydrogen. 

See the system at: 
Lunar vs. Mars steel.

THE ONLY QUESTION IS, WILL THIS SIMPLETON SYSTEM WORK?????
Al/Si at 2000 C., even super alloy pipes might need active cooling  with inert gas and space radiators.
1) 1)  T.D. Lin "Concrete for Lunar Base Construction"  Lunar Bases and Space Activities of the 21st Century  ed. W.W.Mendell  Lunar and Planetary Institute, Houston pg. 381
<http://ads.harvard.edu/books/lbsa/toc.html>
We might wind up with a Ca-Al-Si alloy. Although Ca is only slightly soluble in Al, some contamination might result. Since Ca is a better conductor than copper, and silicon improves Al ductility but reduces tensile strength,  this alloy might still be useful for electrical wires and high power cables.  it should still be possible to separate the alloy with intense solar heat and fractional distillation, something not recommended in the oxygen rich air of Earth, but on the solar energy rich Moon with free vacuum, no problem.