2023 EDIT:
It basically comes to possibility of creating an electromagnet with a piece of plastic (a dielectric) and a rotating electric field around the plastic. This rotating electric field should force the distorted/polarized electron clouds of each atom to rotate around the nucleus, creating sort of a current loop around each atom. Although this is not a full current loop like in a wire where there's continuous charge distribution along wire length. It is rather a localized charge "clump" rotating around a nucleus.
The question asks if it's possible to create a magnetic field with such setup and possibly, the order of magnitude of the generated field.
As for rotating electric field, it might be created by placing three electrodes around a dielectric material cylinder/rod and driving them with 3-phase electrostatic potential. This should create a rotating field around the rod. The higher the frequency of 3-phase signal, the higher the angular velocity of the distorted electron cloud around the nucleus.
I'm not sure but this may be related to Maxwell's displacement current.
Original 2016 question:
Consider a solid dielectric disk (lets say it is made of PTFE/Teflon) placed in uniform electric field so all its dipoles are oriented in parallel with the field. Let's assume that only electronic polarization occurs, that is electron clouds get distorted in response to applied electric field.
Now consider that electric field is continuously rotating around the disk's center. That is, the axis of rotation is placed in the disk center and is perpendicular to the top surface (it is pointing towards the screen). This will also rotate each atom's dipole in the disk.
Is it true that during rotation only electron clouds are distorted and respective nuclei do not move or move very little?
If yes then could these displaced electrons form a tiny current loops and produce magnetic fields similar to regular current loops? I'm aware that this is movement of a localized charge so that instantaneous magnetic field would be different from that originating from many moving charges in a regular current loop but when averaged over entire rotation cycle around the nucleus it should produce similar magnetic field.
Naturally all of these magnetic moments would add up to form a total magnetic field that is parallel to the axis of rotation of the electric field.
How to calculate or approximate the resulting magnetic field magnitude? Or even the order of magnitude if the former would be too complex to calculate.
It seems that magnetic field strength will be proportional to applied voltage and rotation frequency and also will depend on relative permittivity (dielectric constant) of the material.
