Nanotechnology
I'm just getting started with this, so bear with me.  More to come in the future.
BELOW) What nano-technology IS NOT!
One of the things that drives me crazy is hearing someone say that we don't know how to separate the elements contained in lunar regolith and that we might have to use replicating nanobots to pick the molecules apart.  First of all, chemical processing of regolith was well worked out by O'Neill's group back in the seventies. See:<http://www.islandone.org/MMSG/aasm/AASM5E.htm>

Basically, on my lunar resources pages I have simply followed their work and replaced the Kroll process for titanium with the FFC process and the Alcoa process for aluminum with solar carbothermic reduction of alumina and added a few other tweaks.  Hydrogen reduction of ilmenite is already worked out and Carbotek even has a patent. Of the original suggestions on my part that I am aware of one is the use of fine grinding of fused particles of iron, titania and silicates and magnetic separation of iron from titania instead of acid leaching and another is the use of zone refining on a large scale taking advantage of the free vacuum and low gravity of the Moon or weightlessness of outer space.  Other tricks like boiling metals in solar furnaces, reduction with hydrogen, carbo-chlorination, electrostatic separation, use of devices similar to percolators, shifting CO and CO2 to methane, silicothermic reduction of magnesia and condensation of magnesium vapor, etc. are fairly standard. As for nanotechnology, we have know more about it.

From an article published in Chemical and Engineering News, Dec. 1, 2003, Vol. 81, No. 48, pp. 37-48 available on the web at: <http://pubs.acs.org/cen/coverstory/8148/8148counterpoint.html>  we read:

Dr. Richard E. Smalley does not think molecular assemblers are physically possible. He objects to the idea of molecular assemblers, because of the "fat fingers problem" and the "sticky fingers problem."
The fact is, little claws are going to be too "fat" to grab onto atoms and they will tend to react with atoms and become "sticky."  Dr. K. Eric Drexler does not disagree entirely with Dr. Smalley.  In reality, molecular assemblers will be much more sublime.

Drexler says,"like enzymes and ribosomes, proposed assemblers neither have nor need these "Smalley fingers." The task of positioning reactive molecules simply doesn't require them...My proposal is, and always has been  to guide the chemical synthesis of complex structures by mechanically positioning reactive molecules, not by manipulating individual atoms...The impossibility of Smalley fingers has raised no concern in the research community because these fingers solve no problems and thus appear in no proposals."

So we need to dispell notions of little crabs picking molecules apart and snipping chemical bonds. Not only would the little crab claws have to be impossibly small ("fat fingers"), the chemical bond is not a little stick shoved into two wooden balls like a tinker toy.  When atoms react to form molecules electrons whirling around the nucleus at high speed are either transferred or shared.  Electrostatic forces hold the atoms together. There is no little stick to snip and if there was the atoms would fly apart.  Also, when atoms come in contact they tend to react ("sticky fingers").  Drexler's devices are more like enzymes.  

Enzymes are complex protein molecules that have active receptor sites where other atoms and/or molecules called substrates attach, are brought into close proximity and react to form an end product. They have been described as templates with notches in them where other molecules fit in like puzzle pieces. In reality, the receptor sites have something to do with the complex alignment of electric charges in the atoms of the enzyme.  This causes the substrates to bind to the receptor sites and contributes to their reacting.  When the end product forms the alignments of the electric charges change and the formed molecule is not attracted to the enzyme and is released.  Well, that's what my father, a molecular biologist told me long ago when he was still alive.  One of the few gems he left me with.  If only scientists were all rich. He never was and neither am I.  Anyhow, reactions brought about by enzymes occur much faster than they would if atoms and molecules floating around in solution were just allowed to bump into each other by random chance, so enzymes are called cataysts.  Reactions can be speeded up by increasing chemical concentrations and raising temperatures to speed up thermal motion of atoms and molecules, but life can only exist with definite salt concentrations for instance and within a limited range of temperatures.  Thank God for enzymes.

Molecular assemblers will use nano-structures much like enzymes to work with reactive molecules.  Mechanical nano-bots might actually incorporate biological protein molecules to do some of their work.  The study of protein structures; enzymes and ribosomes, will guide us. Protein crystal growth in micro-gravity might be a part of this. Perhaps silicon based rather than carbon based complex "molecular templates" can be created. We know that some bacteria have little wheels that spin their flagella and that ribosomes can assemble protein molecules almost like "little sewing machines" as I was once told.  Nano-machines will attempt to imitate and improve upon Nature. Biological enzymes function in water within a specific range of temperatures, salt concentration and pH. While some microbes don't need oxygen and are even poisoned by it, the need for water is universal for life as we know it. Most enzymes acquire energy from the phosphate bonds of ATP molecules to do their stuff. Hybrid mechano-organic molecular assemblers may take the form of aquatic devices immersed in a solution containing various molecules, salts and ATP. We must ask,"Why bother creating hybrid mechano-organic molecular assemblers? Why not just use natural or genetically modified bacteria?"  Perhaps we will use bacteria. See: Bioleaching There may also be instances where an artificial creation does the job better.