A transistor is a type of semiconductor. [2] A transistor can work like a switch but doesn't need physical moving parts or human controls. It's tiny (~22 nm wide) and works much faster than switches. E.g., A phone has hundreth of transistors. Even more in a computer and countless other modern devices.
[2] A transistor uses semiconductors, by using silicon, which conducts current better than insulators but not as well as metals. That's due to a silicon atom has 4 electrons in its outer shell/valence shell, allowing it to form bonds to other silicons, forming a tetrahedral crystal. But as the crystal are stuck in bonds, few get enough energy to escape their bonds and travel through the lattuce. So having small number of mobile charges is what make silicons semiconductors.
[2] Doping is the process to inject foreign substances for better performance, like in semiconductors. Dopings are either p (positive) or n-types (negative).
[2] To make n-type semiconductors, we inject small element amounts with 5 valence electrons, like phosphorus as it's akin enough to silicon to fit into the lattuce but cann bring an extra electron. Thus the semiconductor has more mobile charges, conducting current better.
To make p-type semiconductors, we inject small element amount of 3 valence electrons added to the lattuce, like boron, creating an electron hole (acting as positive charges) where there should be an electron (which is negatively charged), increasing silicon's conductivity as electrons can move into the holes.
But this doesn't mean p-types are positively charged and n-types are negatively charged. Rather, both are neutral as both have the same proton and electron amount. 'p' and 'n' only refer to their charges. A transistor has both 'n' and 'p'-type semiconductors (NPN and PNPs).
[2] A transistor's junctions has depletion layers, where holes and electrons diffuse between each other, leaving behind immobile ionized dopant atoms (positive ions in the n-side and negative ions in the p-side) in the region, creating a built-in electric field that stops further significant diffusion.
E.g., n-type electrons move to p-type electrons to fill the holes. There moving charges are being depleted.
diffuse:
deplete:
Since the n-type is out of free electrons as they've filled the p-type's holes, making the p-type negative due to added electrons. As a result, the p-type now repels electrons coming from the n-type. So the depletion layer is a barrier stopping current flow across the transistor. Currently the transistor is off, like an open switch (0 state).
[2] To turn it on, we apply small positive voltage to the gate to attract the electrons over and overcomes the repulsion from the depletion layer, shrinking the layer for electrons to move through and form a conducting channel. So the transistor is now is on (1 state).
[2] (NPN) transistors have 1 p-type in the middle of 2 n-types (like a sandwitch). Like switches, a transistor has electrical contacts at its n-types called the drain and source. But IO of a mechanical switch, the p-type in the middle has a 3rd contact, the gate, insulated from the semiconductor by an oxide layer.
A transistor is often 22 nm wide today. But they must be smaller to keep up to Moore's law, which states the transistor number per chip must double per year, with a limit.
As the terminals approach, quantum effects become significant and electrons can tunnel from a side to another, so we may not make a barrier high enough to stop current flow. This becomes an issue of future transistors
Darlington transistor is a
[1] Wikipedia