PEMTec claim that with ECM they get surface reproducibility (I think?) down to 30 nanometers. This is an extremely promising figure for micro-engraving of information or machinery on metal surfaces using movable ECM electrodes. They claim to use a process gap of some microns, a salt-water electrolyte, “an exact current pulse”, “workpieces with an imaging accuracy in the lower micrometer range”, and oscillating die-sink tool electrodes, and get “a surface quality of up to 0.03 micrometers.”
(Magnetic impulse engraving is also potentially interesting: a large high-permittivity core brought down to a sharp needle point resting on a copper or aluminum surface, with a high-current pancake-stack coil wrapped around the core and connected through a step-down transformer to a high-voltage source with a fast switch such as a spark gap. This ought to produce, I think, enough force from the eddy currents around the sharp point to plastically indent the softer metal, but I haven’t done the math to check this. I guess I could do an experiment, but if it fails, that would only provide evidence that a particular configuration didn’t work, and I’d still have to do the math.)
ECM is also a potentially valuable technique for getting sharp metal conical points, flat surfaces, cylindrical surfaces, and spherical surfaces. By rotating the workpiece past an ECM “form tool” electrode you can do “ECM lathing”; if the form tool is a straight edge and also translates parallel to that edge, then small errors in the edge will be smoothed out, somewhat analogous to lapping — but permitting the formation of precise conical and hyperboloid shapes (depending on whether the edge intersects the axis) as well as cylindrical and flat. For a spherical surface, you want to use a concave circular form tool instead, and rotate it around its center of curvature.
A taut wire may be an adequate straight edge for many ECM purposes, and a taut wire being moved back and forth may be an adequate plane.
For high precision, this is superior to traditional lathing because the forces distorting and heating the workpiece and tool can be made arbitrarily low. Sometimes, though, it may be more desirable to maintain a positive fluid pressure in the gap in order to control the gap between the tool and workpiece more precisely than the position of either can be controlled independently. When this pressure can be spread over a large area, it should produce no local distortion, for example in the shape of the surface, only global distortion. However, in this case, only the cylindrical, flat, and spherical shapes achievable by lapping are achievable, not the wider range of shapes achievable on the traditional lathe.
(A different, widely-used electrochemical approach to sharpening is isotropic electrochemical etching; by isotropically eroding the metal by some distance d, any rounded features of radius less than d should in theory shrink to a point. This doesn’t produce precise shapes but it does produce sharp points.)
Plasmas, especially nonthermal plasmas, may be better working fluids for fluid-bearing purposes than traditional liquid electrolytes, particularly if they contain groups such as carbonyl which form low-boiling-point compounds with the workpiece metal. They would permit a much smaller process gap at a given pressure, and plasmas containing oxygen, hydrogen, or fluorine should be able to erode graphite, silicon, silicon carbide, and diamond, although in this case we are perhaps going a bit afield of ECM proper.
With a tool electrode shaped like an air-hockey puck with a needle stuck through it, the fluid-bearing technique will give extremely precise control of the process gap. To engrave precise three-dimensional shapes you still need precise positional control of the other two axes, though; while a kinematic mount consisting of six such fluid bearings able to swivel would achieve this, we wouldn’t be cutting inside those bearings, so traditional piezoelectric or galvanometer approaches are probably better.
(Probably EDM is a better fit than ECM for “lathing” and “lapping”, since its material removal rate is both higher and has a much sharper falloff with distance from the workpiece, but ECM will make it practical to do this with tungsten, copper, and tungsten-copper alloys, and with plasma, even semiconductors such as graphite.)
This technique should make it possible to produce, among other things, precise sharp-pointed electrodes for uses such as electrochemical engraving and scanning-probe microscopy.