# Why are ion thruster grids separated by vacuum?

Or to rephrase: Why aren't the grids two sides of a single insulator?

I'd see many upsides (mechanical and electrical properties separated, holes can't be misaligned, tuning of grid-capacitance, ...) and few downsides (extra sputtered material, current leakage). Since there isn't a single thruster using an insulator to separate grids, I must be missing something. However, I wasn't able to find a paper or explanation why two grids are better than one.

1. Current leakage due to surface discharge or through surface contamination build-up or roughening is End of Mission. Nobody is out there to clean or replace the insulators if they get dirty.

2. Insulator is going to be a dielectric with a significant dielectric constant. Increased capacitance means more stored energy, so if there is a spark it might have a significant higher energy. Also the spark may leave a conductive path then leading back to problem #1.

3. Dielectric insulator changes the shape of the field lines between the two grids, may cause some focusing or defocusing or decrease opacity or ions and/or electrons impacting the insulator. In addition to lower efficiencies that can cause damage to the insulator, producing a better path for breakdown or leakage once again leading back to problem #1.

• that wraps that up! :) Nov 1, 2018 at 3:23
• Breakthrough voltage is definitely a point I missed and that seems very likely to be the culprit. I disagree with the third point, though. Dielectrics behave predictably, so shaping the grid-holes appropriately would take care of changing ion paths. At the same time, other effects, such as gradual potential change, could be used to assert better ion control. Nov 1, 2018 at 8:58
• @GammaSQ predictably does not at all mean you can get the field you require. That's just wishful thinking. While I used conditional language on purpose, the solution to the Laplace equation is going to generate the field that really may not provide the best ion optics. Now, if you've done the calculation or seen it done and you can show the required field shape is achievable, that would be great. However, electric field lines love dielectrics and love terminating on surface charges. You'll see your ions hitting them. It will be hard to stop. It's just the nature of the beast.
– uhoh
Nov 1, 2018 at 11:06
• The grids could not be separated by vacuum only, some insulators are needed to fix them in place.
– Uwe
Aug 6, 2019 at 18:22
• @Uwe but there is certainly no requirement to put the supporting insulators directly between the grids. You want to keep paths for surface discharge as long as possible.
– uhoh
Aug 6, 2019 at 21:23