Mechanism of discharge in DBD? (dieletric barrier discharge)

[Moaaz]

Hallo all,

I'm investigating in using a dieletric barrier discharge for gas treatment, if I have two electrodes isolated by dieletric and the discharge gap between both of them has AIR..
how the ionization of air will take place, from where the source of electron which is going to start the Avalanche since both of the eletrodes are isolated? Is the charge also leaving the source of the dieletric ?!

I'm not a physicist xD
Thanks in advance

 

[Henryk]

Good question.
To start a discharge in air you need two things: the electric field strong enough (above 30 kV/cm) and some 'seed' electrons. It is commonly accepted that the seed electrons are generated by cosmic radiation. In any material, there is also a small number of radioactive elements that can generate seed electrons as well.

 

[Moaaz]

Nice, I really appreciate your answer.. but what about isolating one electrode or both electrodes?
I've seen some modules for DBD some people isolate (cover) one or both with dielectric
as i know... Two electrodes = less Enet = less ionization ( discharge) ... But the advantage is limiting the current and spark formation
Now, One electrode covered is also achieving the same advantage but with higher E net so = higher discharge

So which one is better and why?

 

[Henryk]

You are correct, the purpose of a dielectric layer is to prevent runaway ionization, that is arcing. The dielectric surface quickly charges up reducing the field in the gap to below the value needed to sustain avalanche.
In fact, in air, the DBD has a filamentary character: that is multiple discharges, limited in extent to a fraction of a mm and each lasting something like 10 nanoseconds. If you look at the current, you find out that there is a quite a few of these filamentary discharges at every change of the applied voltage polarity. (You are aware that the DBD has to be powered by an AC high voltage of frequency in the kilohertz range).
Your question, which is better: one or two dielectric layer? I'm not quite sure. It is enough to have just one layer to prevent the discharge. Also, there are other considerations. One thing that discharge does is sputter the material. Sometimes, the application of a DBD device prohibits any sputtered metal and in that case, it is better to have a layer of a dielectric on both sides of the discharge gap.

 

[Moaaz]

You want to say if i applied a high voltage and i have only one electrode covered .. the second one might has some kind of ( sputtering) happens or in another words can i say ( erosion )? so it's better to have the two electrodes covered ... But that leads me to another questions, why then in industry we use Corona discharge which is almost the same technique but without covering ? ( which has the risk of spark formation and sputtering ) ?

 

[Henryk]

Covering metal surface does not eliminate sputtering only now, the insulator surface gets sputtered. It will be at a different rate, different molecules are emitted.
The second point, why corona discharge does not need dielectric.
There are basically two geometries for corona discharge: point corona and wire corona. Point corona is better known in science literature. Wire corona is much more widely used in practice; all the laser printers and photocopiers use wire corona. Point corona is essentially a sharp needle with high voltage applied to it. Wire corona uses a small diameter wire. In either case, there is a large curvature (small radius) near the surface that produces very high electric field sufficient to cause the avalanche. However, the strength of the field drops very quickly with the distance from the electrode and within a millimeter or so, becomes insufficient to maintain the avalanche. That's how arcing of sparking is prevented. The ionization is limited to a very small volume near the electrode only.
And yes, there is sputtering of the electrode. That's why the corona electrode material is usually hard metal, like tungsten. It has relatively low sputtering rate.

 

[Moaaz]

That is interesting, because I was always thinking about formation of Arc as the limitation factor, but I didn't think about sputtering. Actually I'm doing an experiment to test the effect of both corona discharge and DBD on a mixture of gases... because of that I'm thinking about which will give a higher ionization..so it means will give a higher yield of chemical reaction with catalyst existence.

In corona discharge there are different types of discharge before reaching to Arc .. but as i understand in DBD it's a filament (streamer) then it turns into a micro discharge channel of plasma with increasing the voltage ( ions and electrons) ...Is that the only mechanism?

I really appreciate your help.

 

[Henryk]

Sputtering is actually very slow. I don't think it has any noticeable effect on the discharge itself, it just limits the lifetime of the device operation continuously.
As far as the type of the discharge, it depends very much on the gas. I was doing a lot of work with DBD discharge in air and always I observed filamentary discharge. How did I know? just monitoring the DBD current: on top of a sine wave, there were many spikes per cycle.
At one point, I changed the discharge gas to argon and immediately I noticed a big change: instead of many short current spikes per cycle I saw one current pulse lasting about 1 microsecond every time polarity changed. This was a clear case of a diffuse discharge.
Later, I came around a paper that studied nitrogen/oxygen controlled mixture and found out that in pure nitrogen, the DBD discharge is diffuse but adding oxygen changes it to filamentary.
Anyway, discharge phenomena in gases are actually not quite understood. One of the paper reviewing research emphasised a frequent use of words and expressions like "appears to be", "likely", "probably", "it is considered". In other words, not many thing are actually confirmed true (or false).

 

[Moaaz]

That is true. I've noticed that it's not clearly investigated and it makes it difficult specially for the people who came from a different background.. I'm investigating in the of treating mixture of gases by plasma with catalyst.. so mainly I've to focus on Corona discharge and DBD. but as the literature used to mention that DBD is more promising and easy to control..etc
But I still can't find a clear information about the frequency effect and the dielectric thickness on the discharge ( electric field )
As far as I understood, Increasing the thickness of dielectric for the same discharge gap means lowering the net electric field but the frequency is a bit tricky

 

[Henryk]

Hi,
I don't think that dielectric thickness plays a role other than it has to be thick enough to prevent breakdown.
As for frequency, I did see effects of DBD frequency on the chemical species. DBD was connected to a mass spectrometer and at 100 kHz the mass spectra were different than at 30 kHz. I never found an explanation for the effect neither found any mention of it in the literature.
Good luck with your research. Lots of unknowns, lots of fun.

 

[Moaaz]

Hopefully Thanks for your help. I would be glad to keep in touch with you
But for the same discharge gap with increasing the thickness of dielectric it means lowering the electric field due two increasing the distance between the electrodes.
The frequency If increased will lower the dielectric constant which consequently has an impact on the electric field in the discharge gap so all parameters are connected somehow