High voltage phenomena

[Allcanadian]

Abstract
When dielectric materials are brought into contact and then separated, they develop static electricity. For centuries, it has been assumed that such contact charging derives from the spatially homogeneous material properties (along the material's surface) and that within a given pair of materials, one charges uniformly positively and the other negatively. We demonstrate that this picture of contact charging is incorrect. Whereas each contact-electrified piece develops a net charge of either positive or negative polarity, each surface supports a random "mosaic" of oppositely charged regions of nanoscopic dimensions. These mosaics of surface charge have the same topological characteristics for different types of electrified dielectrics and accommodate significantly more charge per unit area than previously thought.
The mosaics accommodate significantly more charge per unit area than previously estimated for contact electrification, but the overall/"net" charge on an electrified surface remains relatively small due to the "compensation" between the (+) and the (–) regions.

http://www.sciencemag.org/content/333/6040/308.full

[EMdevices]

sounds much like an analog of feromagnetic materials, that are not magnetized.  they have small domains inside that are magnetized, but they are distributed at random and they cancel each other out.  When an external magnetic field is applied,  the domains alighned in that direction begin to grow at the expense of the other, and the "domain walls" shift appropriately.