How would this capacitor configuration work?

1. Feb 6, 2010

azure kitsune

Assume you have two parallel plates with charges +q and -q.

Now, you fill the space in between with two different dielectrics so that if you were to move straight from one plate to the other, you would be in the same dielectric. (In other words, the two dielectrics do not form "layers.")

My question is: What happens to the charges on the plates?

Do they stay uniformly distributed? This would mean the potential difference across the capacitor is different at different locations. Wouldn't this make the electric field inside the capacitor non-conservative?

Do the charges move reposition themselves so that the potential difference remains constant? If this were true, wouldn't the charges have a tendency to go back to uniform distribution because the closer-packed ones would repel each other more?

If the capacitor is connected to an EMF source with constant voltage, then I believe the charges would reposition themselves to maintain the same potential drop. But what if there is no current?

2. Feb 6, 2010

kcdodd

The conductors will always be equipotential, due to E = 0 inside the conductor. The dielectrics are polarized in the direction of E, which means there is an effective charge density next to each conductor. This allows the charge inside the conductor to be re-arranged.

3. Feb 7, 2010

azure kitsune

Thank you for explaining! I think I understand. Is the following statement correct?:

For a charged conducting plate by itself, the charge would be uniformly distributed on the two surfaces because there is no other way for E to be 0 inside. For a capacitor with two dielectrics, the charge is not uniformly distributed, so the electric field inside the plate due to the charges on the surface alone is not zero. However, the net electric field inside the conductor (which includes E from the polarized dielectrics) is zero.

4. Feb 7, 2010

kcdodd

It's not really safe to assume charge is distributed uniformly, even for isolated conductor. Except perhaps spheres or infinite plates. You usually start with the assumption of a constant voltage on the conductor, and then solve for the charge density.