Permanently charging a sphere by induction using a high voltage I want to charge a sphere using a Van de Graff generator. It would sound easy, but using a 1 MV or a 900 KV is really hard, it is enough to break like 1 meter of air so whenever i try to ground the sphere there is spark. So I wanted to know, if it will help if I covered the sphere with a layer of plastic or an insulator? Or if there was a vacuum between the the generator and the sphere. And another thing, can a graphite rod be charged like a metal one?
 A: Anything that increases the strength of the dielectric between the sphere and the ground will improve your odds of success.  Achieving a nearly perfect vacuum will help, but in general, a better, but imperfect vacuum will actually hurt (the lower air density reduces the collision rate faster than it reduces the number of carriers).  You are better off attempting to insulate with known insulators (rubber, glass, etc.)
A: In high voltage labs with Van de Graaff generators, they typically will encase the whole device in high-pressure, high mass gas, like Sulfur Hexafluoride, pumped to ~5 atmospheres of pressure.  This decreases the mean free path of ions and electrons that cause the breakdown, so that they don't get up enough speed to ionize anything before they run into a gas molecule and lose speed.
In perfect vacuum, there is nothing for those free electrons and ions to hit, so at ultra high vacuum, you get really good insulation, but as the prior answer said, it is really hard to get ultra high vacuum, and when you do, you always have a container around the device (the vacuum chamber walls) which can get hit by those free particles.  If they get hit hard enough to liberate atoms from the chamber walls and ionize them, the ions fly back and forth between walls and electrode repeating the process over and over, and discharge it anyway.
Coating it with an insulator works for a while, but over time the charge on the ball will leak into the insulator - it is a constant DC field, so no matter the insulator's resistance, eventually it leaks into the insulator.  This can lead to space charge within the thickness of the insulator; the charges don't just all sit on the surface like with a conductor, they slowly diffuse through the material.  Given that what you want is a permanently charged sphere, space charge may actually be ideal for you; it is hard to discharge all the accumulated space charge from insulating materials all at once, so it probably increases the retention of charge the way you are looking for (though, I'm speculating about that last part - I have not tried it).
