The useful middle ground
A light switch controls electricity by physically opening or closing a path. A transistor also controls current, but it has no moving lever. Its behavior comes from the material inside it: a semiconductor. Unlike a good metal conductor or a strong electrical insulator, a semiconductor can be engineered so that electrical charge moves through it readily in some circumstances and with difficulty in others.
Silicon is a common example. Each silicon atom bonds with its neighbors in a repeating crystal structure. For an electron to move freely through that structure, it must have enough energy to leave a bond and enter a range of allowed energies called the conduction band. The energy separation between the bonded state and that band is known as the band gap. This gap helps explain why pure silicon is neither as conductive as copper nor as resistant to current as glass.
Changing silicon with tiny additions
Pure silicon alone is not enough to make most practical transistors. Manufacturers deliberately add small amounts of other elements in a process called doping. Some added atoms contribute electrons that can move through the crystal more easily, producing n-type silicon. Others create places where an electron is missing from a bond. These vacancies, called holes, can move through the material as nearby electrons shift to fill them, producing p-type silicon.
The names describe the main mobile charge carriers: negatively charged electrons in n-type material and positively behaving holes in p-type material. They do not mean that every piece of doped silicon has a large overall electrical charge. Doping changes how readily current can flow and allows regions with different electrical properties to be built next to one another on the same chip.
What happens where regions meet
Put p-type and n-type semiconductor together, and some electrons and holes near the boundary combine. This leaves a narrow region with relatively few mobile charge carriers. An electric field develops there, making it harder for more carriers to cross. The resulting p–n junction conducts differently depending on which way an external voltage is applied.
Junctions are important building blocks, but a transistor does more than provide a one-way preference for current. It uses one electrical signal to influence another flow of charge. Different transistor designs accomplish this in different ways. To see how semiconductor properties become a controllable switch, it helps to look at a widely used design: the metal–oxide–semiconductor field-effect transistor, or MOSFET.
An electric field becomes a switch
A MOSFET has two terminals, called the source and drain, between which current may flow. A third terminal, the gate, sits above the region between them, separated from the semiconductor by a thin insulating layer. Applying a voltage to the gate creates an electric field in the material below it. Because the gate is insulated, it can change the conditions for current flow without needing a continuous direct current through that insulating layer.
In a common n-channel MOSFET, a suitable gate voltage attracts electrons toward the surface of a p-type region. When enough electrons gather there, they form a conductive channel between the source and drain. With a lower gate voltage, that channel is absent or much less conductive. The gate therefore controls whether current can pass. The exact voltage needed depends on the transistor’s design; a real transistor also does not become a perfect conductor or a perfect insulator.
From one device to useful circuits
A transistor can represent a digital state by making a circuit path relatively conductive or resistant to current. Arranging transistors together lets a circuit perform logical operations, store a bit of information or direct signals elsewhere. Transistors can also work between those extremes, controlling a varying current in an amplifier. The same basic ability to influence charge flow supports many kinds of electronics.
The key is not that silicon naturally behaves like a miniature mechanical switch. It is that its band gap, its response to carefully chosen dopants and its sensitivity to nearby electric fields give designers ways to shape current flow. By building regions with different properties and connecting them to electrical terminals, engineers turn a piece of semiconductor into a device that one signal can control.
