Electroplating at the nanoscale.

Electroplating is most easily performed if the substrate is conducting. Could you first make a non-conducting substrate conducting by passing a high voltage through it to exceed its breakdown voltage? With gases exceeding the breakdown voltage produces a highly ionized plasma. To electroplate onto a substrate you want it to retain its solid form. Could this be achieved while exceeding its breakdown voltage or would it also become a plasma gas? Alternatively, a non-conductor can be electrostatically charged by rubbing with a wool or silk cloth. If it is negatively charged then there are an excess of electrons attached to the atoms near the surface. Could these electrons be made to flow as a current if a voltage source was applied?
Bob Clark

[snip]
Chip fab photolithography, then deposition. Forget aluminum. Copper or tungsten.
No.

There are established techniques for doing this job, including evaporation of metals, semiconductor-manufacturing steps, and printing with conducting ink. I doubt that they can be done well without the right equipment.
As for peeling the fine wires off, remember that they will be delicate.

You might try electroless plating. Much better results when dealing with non-conductors. :-)
What's wrong with silver? Electroless deposition of silver is widely used for mirrors. It can even be applied as a spray, for example to plate glass windows in storefronts.
How about hollow tubes?
100 nm? That is indeed quite small. It may be easier to find a conducting molecule of that size. Perhaps the graphite fibrils from Hyperion Catalysis ? Electroless plating is easy -- nothing but chemicals and beakers. Watch out for the explosion hazard if you use silver. Copper is also a well-developed technology, used for plated-thru holes in circuit boards.
You might also consider depositing a conductive seed layer prior to conventional electroplating. This is commonly done with palladium or carbon, however carbon might be too lumpy to meet your thickness requirement. A palladium seed layer by itself might be thick enough. Here's a review of electroless metallization:

It is possible to evaporate or sputter thin metalic film on an insulating substrate. For example, I routinely evaporate 10nm thick Chromium on glass, it is used as an adhesion layer for a 1um thick (copper or gold) microstrip transmission line. This issue you will have is with the width of the lines. The cost goes up as the smallest feature goes down. You will need a fab. If you don't want to use industry or a university then you will need to build a fab in your garage.
Dwayne

The electroplating is not a real problem. Either Ni or Cu can easily be done by an amateur, maybe even gold could be done. The normal way to do this would be to vapor deposit gold onto an Si wafer (or onto float glass) about 10 nm thick, spin coat with PMMA resist, photolithographic patterning, etch away the pattern of the wires then electroplate the wires. The gold would be the release layer. For an amateur, hmmmm............ You might try getting an old inkjet cartridge, filling it with Aquadag carbon based conducting paint and then printing the finest lines you can, still, 100 nm is .1 micron......then electroplating. Another method, buy some ultra-thin Al foil from Goodfellow (kinda expensive), sandwich it between two pieces of glass, seal all but one edge, lap this edge with ultra-fine abrasive just prior to electroplating so that you have exposed a fresh layer of Al, electroplate for 6 min at 10 amp/sq ft (nickel sulfamate bath) to get your .1 micron. Take out your Al foil and put it in sodium hydroxide to dissolve the Al but leaving the Ni wire along the edge. Most of the time when people want ultra-thin foils, they vapor deposit them onto substrates coated with a soap film. Carbon "stripper foils" for accelerator targets are made this way to a thickness of 10 ANGSTROMS believe it or not about 5mm diameter free standing on an Al support with a hole in the middle. I have done this and it isnt too hard with a vapor deposition system. If you can think of a way to pattern your wires, all electron microscope labs (universities etc.) have either carbon coaters or gold sputter coaters for sample prep and these cost nothing to run so they might do this for you. I think a project like this is doable even by an amateur but am not exactly sure how yet.

[hanson] Bob, you received a whole set of good suggestion form a bunch of posters. Here is different, additional approach, that may solve the "peel off" problem and provide you with a great abundance of specimens to experiment with, namely fishing "metal whiskers" out of an aqueous solution. Google for it and see whether you can get a recipe for it. Post also into sci.chem.electrochem. Those dudes do run across such "metal whiskers" as a nuisance particularly with Sn & Cu. Good luck. hanson

Somebody suggested electroless deposition. Either electrolytic or electroless may work, however, I find the chemicals for electroplating to be much "nicer" than for electroless. However, electroless has the ability to plate deep down into pores and into slots whereas electroplating may not be able to get down into deep structures. I do not think this is a problem for you. Another possibility is to electroplate (or electroless) onto thin carbon fibers. I've done this but have no idea of the fiber size as it was an accident when I did it. Another possibility is "silvering" using the method that was long ago used to silver mirrors which deposits a thin layer of silver onto glass. This can easily be done by amateurs and is discussed in old copies of Amateur Telescope Making or in How to Make a Telescope by Jean Texereau. You would coat silver onto thin whiskers of some non-conducting material or onto a patterned substrate. Some hints:
Electroplated nickel does not stick to stainless steel so you might use stainless as a substrate so your wires will release easily. Gold will not stick well to glass or Si wafers so you can deposit a little of it and use it as a release layer. Copper will not stick well to glass so you might use it as a release layer and wire material. If you can get an electron microscope lab to deposit a thin layer of carbon with their specimen coating system, it will coat the entire thing all around and then you can electroplate onto it. Carbon also makes a good release layer. Since the carbon coats all around, you can actually coat individual salt crystals and then dissolve the salt leaving little hollow cubes, its like magic.
OK, HERE IS HOW TO DO IT, just thought of this.
Coat pieces of glass with the previously mentioned silvering process. Stack them on top of each other with edges exposed. Electroplate for desired time (I say 6 minutes ) and then seperate the stack. Dissolve the silver leaving the Ni wires. With many pieces of glass stacked, say microscope slides, you will get many samples. All of this can be done easily by an amateur.

If I'm understanding your suggestion correctly wouldn't this leave nickel plate with a width that is the thickness of the glass slides? Or perhaps you mean only coat a single side of each glass slide with silver while pressing an edge of the glass slide firmly against a hard, flat surface so this edge gets no silver coating. The sides covered by the silver are the ones pressed together. Apply nickel electroplate to an edge of the stuck-together glass slides. Then the nickel will only adhere to the conducting silver between the slides. There are high quality optical glass plates available that are flat at the nanoscale. Then the separation between the glass slides will only be nanometers, so this will form a nanometers wide nickel plating between the slides. So now you dissolve the silver and separate the slides. What do you suggest for dissolving the silver?
Bob Clark

Yes, you have the idea. Dissolving silver but not Ni? Hmm....look in the CRC manual.

If you only need a few, how about using wire that have already been created on a silicon chip? Larger chips will have long straight runs insulated from the layers below and I think that more recent ones are on the scale you want. If you need to separate the wires from the chip you might have problems, but maybe you could electroplate something like gold onto them and then dissolve them away to release the gold?
Scrim

You must've read about copper- plating plastic toys ... spray with aluminum paint, then ....
What about just using aluminum paint (or gold, bronze, ...), the paint vehicle muck things up?
Atty :-)