In electro-etching, an electrolyte selectively removes metal from a metal surface by anodic dissolution; typically a vinyl mask is applied to the surface to shield some areas, although of course a conventional photolithographic resist like SU-8 epoxy or PMMA can also be used, and painted-on coatings such as Sharpie are also frequently used. This can produce deep etching fairly quickly with high current.
If the surface is first uniformly electroplated (or otherwise coated) with a differently reactive metal, even very shallow electro-etching ought to be able to produce dramatic visual effects by selectively removing the plating, followed by a subsequent treatment such as acid etching, etching with alum, bluing, toning with sulfur, or possibly even autocatalytic “electroless” plating. This ought to enable strongly nonlinear threshold effects as well: where the plating is completely removed, the reactive surface is the substrate, and where it is not completely removed, the reactive surface is the plating.
In cases where the underlying substrate is more reactive than the plating, it ought to be possible to use further uniform electro-etching at a carefully controlled voltage in between the (modified, not standard) electrode potentials of the two materials to selectively remove the substrate material where it is exposed, thus deepening the initially patterned etch.
This is all prequel to suggesting that electro-etching or electroplating with a foam of soap bubbles, as from dishwashing liquid, should make a freaking awesome pattern. The air in the bubbles would play the role of the vinyl resist. Thanks to sbp for the idea.
A variant of this commonly happens in a variety of electrolytic processes (anodization, electro-etching, electroplating, electroforming, batteries, and so on) where the bubbles form from electrolysis of the liquid; generally this is considered a nuisance, since the bubbles spawn at unpredictable places, and in batteries “depolarizers” like manganese dioxide are used to counter it. But it might also provide an interesting artistic texture.
In addition to soap bubbles, there are several other surface-patterning approaches that come to mind.
Stamping patterns onto the surface of metal with a conductive rubber stamp (graphite-filled or copper-filled, say) and electrolyte “ink” is another possible form of electrolytic rapid patterning of metal surfaces.
Earlier I’d suggested selective electro-etching or electrodeposition with one or many moving electrodes very close to a metal workpiece as a way to produce precise surface contours, or similarly electrolytic anodization as a way to precisely produce colors. The above-suggested methods of “developing” an extremely thin initial etch or plating with nonlinear effects should enable this process to pattern a surface orders of magnitude faster, either by selectively etching away part of a surface coating, by selectively depositing plating, or both. (See also the note on ECM engraving.
Another possible way to selectively electroplate a surface is with localized laser heating; for example, in a standard acid blue vitriol electroplating solution, even a 5-watt blue laser has been reported to produce this effect by locally heating the solution and thus slightly shifting the electrode potentials.
The more common way to modify a metal surface with a laser is of course to heat it up in the air, which, depending on the degree of heating, can oxidize it, explode tiny holes in it that expose fresh metal, or both. The oxide layer may also be usable as a selective resist. If the laser heating is carried out in a reducing atmosphere such as hydrogen, carbon monoxide, acetylene, or vitriolic air, it could simply remove the oxide, exposing raw metal, rather than depositing it.
By using selective corona or other glow discharge, for example from carbon fibers, platinum electrodes, or sharp aluminum wires, rather than a laser, we could gain a number of other advantages. We could easily pattern the surface at scales well below the wavelength of light, limited only by the diffusion of the plasma, which in turn is largely limited by the precision with which we can control the distance from the tooltip to the substrate. If we are reducing a surface oxide coating, we can use much smaller amounts of reducing gases (or dielectric liquids), and using above-atmospheric or below-atmospheric pressures may be more practical than they would be with a laser. By giving the workpiece a negative charge, we can encourage anions from the plasma to smash into it, reducing lateral plasma diffusion, and the anions can be more reactive than non-ionized molecules would be. (Butane gas, for example, is fairly inert, but a butane plasma will contain all kinds of hydrocarbon free radicals.) This will also tend to vaporize the tooltip electrode faster than a glow discharge would; the electrode can contribute other helpful materials to the mix, including in particular metals for vapor deposition.
These processes, too, can sharpen the boundaries between surface regions using the same kind of differential deposition-then-removal process described earlier for electrolytic processing; for example, first reduce the surface oxide coating everywhere, then selectively deposit it in some places, then selectively remove it in others to steepen its boundaries, and then apply some other reaction, specific to either the oxide or the underlying metal, to use the pattern thus deposited. As another example, you could selectively deposit aluminum in an argon atmosphere by plasma-vaporizing it, then use an oxidizing atmosphere to selectively oxidize areas where you don’t want the aluminum.
Using a cold plasma pencil instead of just a glow discharge may permit more flexibility, for example by allowing a higher degree of ionization than a glow discharge can achieve, or allowing short-lived ionized species to decay. But it probably can’t achieve as fine precision.
Another way to pattern a surface by local heating is by resistance heating, like a spot welder does. At short distances you can invoke field electron emission (20–40 V/μm, lower with a low-work-function coating) or thermionic emission to liberate electrodes from your “write head” with which to bombard the surface. (At short distances at atmospheric pressures there isn’t enough gas to sustain an avalanche discharge.) This is actually the same process described above for generating plasma, but with a different purpose, of heating the surface rather than generating ions, so the current is in the opposite direction. By pulsing the discharge, greater peak temperatures can be achieved at a given average power, changing the attainable reaction products. This heating can provoke many of the same kinds of reactions as described above. Also, despite what I said above, this current direction is probably better for sputtering atoms off the tooltip electrode.
For thus sputtering metal onto a non-conductive substrate you might want to use two separate electrodes. I suspect such sputtering at atmospheric pressure should be feasible at very small scales.
Local heating and reaction is most precisely attainable with focused electron beams or focused ion beams, but these of course require hard vacuum and thus cannot be used to provoke reactions with gases or volatile liquids, nor reactions that produce much of them. Many semiconductor photoresists are routinely patterned in this way.
Semiconductor etching processes offer further possibilities for amplifying surface patterning, including not only the acid etching mentioned above but also mass anisotropic etching with reactive ion plasmas which react selectively with the exposed substrate.
If you have patterned a metal surface in such a way, you could etch away the substrate metal underneath it — for example, etching steel with alum, or aluminum with lye — to get a very thin foil of the deposited pattern. I understand that Drexler prototyped a solar-sail material in a way similar to this, but you could also use the resulting perforated metal foil as a photolithography mask. A three-layer technique may be the best solution here: first a massive, rigid, etchable substrate; then a uniform thin foil of microns up to hundreds of microns, which is also etchable, but resists at least one etchant that attacks the substrate; then a “resist” mask, perhaps of metal or metal oxide, deposited on top and patterned with submicron thickness. Once the “resist” is patterned, you etch the foil away where it is exposed by the resist; once the foil has been etched all the way through, you switch etchants and etch away the substrate while leaving the foil unharmed.
If you start by depositing a thin film of a resist on the surface, you can selectively remove it more easily than the thick films discussed above. Langmuir–Blodgett films of poly(N-alkylmethacrylamides) are already used as UV photoresist for photolithography, but the other patterning techniques described above can also be used with Langmuir–Blodgett films. That includes not only LB films of that photoresist, but also of a variety of inert surface coating materials, and also the opposite — surface coating materials that functionalize an otherwise inert substrate to react with materials it will be exposed to later, or that eventually react with a surface coating that is already present, for example of oxide.