twisted transistors: part one

As promised, I'm giving a quick overview of field-effect transistors. First, I have to ask if there are any physicists or engineers reading CBC. If you are either, you get to point out where I'm wrong. I refuse to assume any authority here, since I've never actually made a field-effect transistor. For this reason, my understanding is incomplete, but here are the basics.

A transistor can basically work as a switch. There are three electrodes: gate, source, and drain. The gate electrode is separated from the source and drain electrodes by the gate dielectric, and there is a semiconductor between the source and drain electrodes. You can see this from the picture.

The gate electrode is the key here. If there is no gate bias, the semiconductor is not...conductive, and current cannot flow from source to drain.[1] What happens if the gate is biased? This is where the gate dielectric is important. It must be polarizable in order for acceptable device function. See the picture? When a gate bias is applied, the gate dielectric is polarized so that there is a partial negative charge of sorts at the interface between the dielectric and the semiconductor. This attracts charge carriers--holes, in the case of a p-type semiconductor--to the "channel."[2] Now, since charge carriers are present, current can flow from source to drain.

[1] I'm SO not getting into traps or threshold voltages or interface morphology or any numbers-followed-by-units here. You have to wait. (This is really just my way of forcing you to keep reading CBC.)

[2] Show me a material with acceptably high electron mobility, and I will start talking about n-type semiconductors. It cannot contain any icky metals.