The MOSFET (metal-oxide-semiconductor field-effect transistor) is the most important microelectronic device. I will discuss a portion of the MOSFET called the MOS capacitor. A good way to visualize the MOSFET is as a capacitor (a device that stores electric charge) with a power-supply voltage (something that moves the stored charge, creating current) across the bottom plate of the capacitor*. See below for diagrams.
* Unfortunately this isn't necessarily an accurate portrayal of a MOSFET. As MOSFETs get smaller and smaller, this picture gets less and less accurate. Regardless, everyone who researches MOSFETs thinks about them in this manner and treats the inaccuracies as second-order effects. Dealing with the inaccuracies is the focus of modern MOSFET research.
The MOSFET
Vgate
|
|
-------
|Metal|
GND-----| ------- |-----VDS
| |Oxide| |
=====-------=====
| S | | D |
| | | |
===== =====
Semiconductor
body
-----------------
|
|
Vbody
The MOS capacitor
Vgate
|
|
-------
|Metal|
-------
|Oxide|
-------
|Body |
-------
|
|
GND
Notice that the MOS capacitor is just a portion of the MOSFET. I need to explain some things. In modern MOSFETS, highly doped polysilicon is used as the gate material. Technically polysilicon is not a metal, but it behaves much like one. When polysilicon replaced aluminum as the gate material, there was no reason to change the term MOSFET. The "oxide" under the gate is almost always silicon dioxide. However, it could be any dielectric material--its purpose is to provide electrical insulation between the gate and body. The MOSFET body is almost always silicon. The body is doped p-type to make NMOS transistors or n-type to make PMOS transistors. CMOS, a mix of NMOS and PMOS, requires n-wells and p-wells to provide both body doping types.
The MOS capacitor is different from a normal parallel-plate capacitor because the bottom "plate" is a semiconductor. Unlike a metal, a semiconductor doesn't have a huge amount of free charge. When a voltage is applied between the gate and body, charge piles up on the oxide/gate interface. How the charge in the body is dispersed depends on the regimes of operation (a range of Vgate. I will consider the operation regions of an NMOS capacitor--the PMOS capacitor behaves in an opposite manner.
Regimes of operation of an NMOS capacitor
In the diagrams below I'll use + to indicate positive charge and : to indicate negative charge.
Accumulation
In the accumulation regime, the voltage applied to gate draws more holes to the oxide/body interface. Thus for an NMOS capacitor (with a p-type body), a negative voltage (and correspondingly an electric field that points from the body to the gate) puts the MOS capacitor in accumulation.
Accumulation Regime
Vgate < 0V
|
|
-------
|metal|
|:::::|
|-----|
|oxide|
-------
|+++++|<-----accumulation of holes
|p |
|type |
|body |
-------
|
|
GND
Depletion
In the depletion regime, the applied gate voltage pushes holes away from the oxide/body interface, depleting the interface of mobile charges. The holes that are pushed away leave behind negatively charged acceptor ions.
Depletion Regime
Vgate > 0V but < Vt
|
|
-------
|metal|
|+++++|
-------
|oxide|
-------
|:::::|<------Acceptor ions
|:::::|
|p |
|type |
|body |
-------
|
|
GND
Inversion
In the inversion regime, the applied voltage is enough to draw a large amount of mobile electrons to the oxide/semiconductor interface. Since the previously p-type semiconductor becomes n-type at the interface, it is said to be inverted. The voltage at which a significant number (this can be arbitarily defined) of mobile electrons exists at the interface is called the threshold voltage of the MOS capacitor and is denoted by Vt.
Inversion Regime
Vgate > Vt
|
|
-------
|metal|
|+++++|
-------
|oxide|
-------
|:::::|<------Mobile electrons
|:::::|
|:::::|<------Depletion region
|:::::| still exists.
|p |
|type |
|body |
-------
|
|
GND
This analysis has been qualitative. Quantitative analysis requires a background of semiconductor physics. Several books on semiconductor devices provide quantitative analysis. The semiconductor textbooks by Pierret and Muller/Kamins come to mind.