Look at this doc it may be helpful for you topic. Good luck

Carbon will be deposited if you have any one of tool/workpiece or dielectric with carbon as an element in it. I think the question is how to remove the debris particles (including carbon). Well, you can use tubular electrode and find out if gas or air can pass thorugh the hole in tube and remove particles (just a suggestion!)

Using air or gas as dielectric you loose different advantages of the liquid dielectric:

- the cooling of particles is less efficient due to lower convection to air

- you have not any more the "pumping" effect since air or gas viscosity is very low

Particles being hot will stuck to the electrode since there is no transport mouvement strong enough to move them away.

Above suggested solution is a way but it requires a hollow electrode and would be efficient only if air/gas velocities in the gap are high enough.

Many techniques can be used as per opinion.

Hollow electrode can be used

Avoid the usage of hydrocarbon oil as the dielectric medium

Arcing effect can also cause for the carbon layer formation. By choosing proper air gap distance, the effect can be reduced.

Proper choosing of flushing pressure should be used.

Reducing the peak current and pulse duration may lead to reduction in deposition of carbon layer.

I thought about a detail which I do not know if will work :

could you change the polarity ?

@Nick Hamburger....my first work is to eliminate the use of dielectric and I know that may be loosing various advantages of using dielectric. But one question still persists in my mind that will using a gases like Oxygen, Nitrogen, and the various noble gases, have an effect same as that of dielectric??

   And as we are using Micro electrodes, will that be able to generate the spark properly like the electrodes used that were used in conventional EDM? And also how to control their motion, as they are quite light in weight, the problem of deflection of electrode, and many others are coming into picture? 

If I understand the question properly, then a carbon layer formed on the workpiece is to be removed or prevented.

The main causes of a carbon film on the electrodes are a poor flushing and completely incorrect process parameters. Since I do not know which electrode materials they use, the conversion to carbon-free dielectrics is not commentable. The primary problem of micro-EDM is, of course, the working gap regeneration (flushing). It would now be necessary to ask which micro-geometries they want to work on, especially small, particularly deep, or ... In many cases of the micro-EDM they are very limited with the flushing conditions, so that additional lifting movements are necessary. Of course, the productivity is drastically reduced, but it is necessary.

The possible process parameter changes to be performed are better, but not successful without sufficient flushing. Here, a lot must be tried and considered, i. Good knowledge of the ablation fundamentals.

To MS:

Sorry for the delayed answer but I was not at my place and did not have access to the net.

Of course a gas can be a dielectric but as I mentioned a gas has a very low convection coefficient in comparison with usual fluids which combined with the small gap does not allow a cooling of liquefied or only torn particles. Being hot they bond in physical way witj the wall which stops theit movement.

Now in this technology the lectrode is usually moved with different amplitudes and frequencies to generate a dielectric movement in case that you do not have a continus flow. The elctrode has not to touch the hole bottom so that the axial force is function of the "pumping" effect. If you have a gas this pumping which function in narrow gaps of the fluid viscosity will be small and the risk of buckling can be reduced via control of frequencies and strokes. You should estimate according to the electrode length which is the elastic buckling limit and control your "pumping" to generate such a pressure that the safety coefficnet gainst the limit will be 6 to 10.

But which ever way you go the risk of deposit is not avoided for the reasons listed above.

Air or gases may be used in very specific combination of  micro EDM machining parameters,In general it is recommended to have proper dielectric and flushing pressure/flushing method.