What Is a Semiconductor?

Every phone, laptop, car, and satellite you’ve ever used depends on a thin wafer of purified sand doing something remarkable: acting as an electrical material that is neither a good conductor nor a good insulator, but something in between — and something we can control. That material is called a semiconductor, and understanding it is the key to understanding how modern electronics actually work.

Conductors, insulators, and the material in between

Every material can be sorted, roughly, by how easily electricity flows through it.

Conductors (copper, silver, aluminum) let electric current flow almost freely. Their outer electrons are loosely held and wander easily from atom to atom.
Insulators (glass, rubber, plastic) barely let current flow at all. Their electrons are locked tightly in place.
Semiconductors (silicon, germanium) sit in between. In their pure form they conduct only a little, but their conductivity can be changed — deliberately and precisely — by adding tiny amounts of other elements, applying voltage, or shining light on them.

That controllability is the entire point. A material that always conducts (like copper) or never conducts (like glass) is useless for building a switch you can turn on and off electrically. A material whose conductivity you can dial up or down on command is exactly what you need to build the switches that make up every digital circuit in existence.

🧩 Think of it like… a water valve instead of a plain pipe. An open pipe (conductor) always lets water through. A solid wall (insulator) never does. A valve (semiconductor) can do either, depending on how you turn the handle. Electronics needs valves, not just pipes and walls — and semiconductors are how we build electrical valves out of solid material with no moving parts.

Why silicon?

Silicon is the second most abundant element in Earth’s crust (after oxygen) — it’s the main ingredient in ordinary sand. That alone makes it cheap. But silicon earns its place in every chip factory on the planet for a few more specific reasons:

It has 4 valence electrons, which lets it form a stable crystal lattice and also makes it easy to “dope” with neighboring elements (more on this in Doping: N-Type and P-Type).
Its band gap (~1.12 eV) is just right — large enough that pure silicon doesn’t conduct too much on its own at room temperature, but small enough that doping and voltage can control it precisely. See Energy Bands for what a band gap actually is.
It forms an excellent natural insulating oxide (silicon dioxide, essentially glass) when exposed to oxygen, which turned out to be crucial for building the tiny insulated gates used in transistors.
It’s mechanically strong and stable, which matters when you’re etching features smaller than a virus onto a wafer.

Germanium was used in the very first transistors in the 1940s and 50s, but silicon’s superior oxide and thermal stability made it the industry standard by the 1960s, and it has stayed dominant ever since.

How a semiconductor fits into a chip

A modern processor might contain tens of billions of transistors, and every single one starts as a region of silicon whose conductivity has been carefully engineered. Zoom out far enough, and the entire chip industry is really an industry devoted to controlling one property — how easily electrons move through a piece of silicon — with atomic precision.

    flowchart LR
    A["Raw sand<br/>(silicon dioxide)"] --> B["Purified silicon crystal"]
    B --> C["Doped regions<br/>(N-type / P-type)"]
    C --> D["Transistors<br/>(diodes, MOSFETs...)"]
    D --> E["Logic gates"]
    E --> F["A working chip"]

“Semiconductor” is used two ways in everyday language: it can mean the material (silicon), or — especially in news and industry contexts — the chip itself (“the semiconductor shortage”). Both usages trace back to the same underlying material property described on this page.

What makes conductivity controllable

Pure silicon’s conductivity depends on how many free charge carriers (electrons that can move) exist inside it. At room temperature, only a small number of electrons have enough thermal energy to break free and conduct. Engineers exploit two big levers to change this:

Doping — deliberately adding a tiny amount of another element (about 1 dopant atom per million silicon atoms) to add extra free electrons or extra “holes” (missing electrons that behave like positive charges).
Electric fields — applying a voltage near a semiconductor region can push charge carriers away or pull them in, turning conduction on or off without any moving parts.

Both of these ideas get their own deep dive later: doping in Doping: N-Type and P-Type, and field control in The MOSFET.

Key takeaways

Materials range from conductors (easy current flow) to insulators (no current flow), with semiconductors in between.
What makes semiconductors special isn’t being “medium conductive” — it’s that their conductivity can be controlled on demand.
Silicon dominates the chip industry because it’s abundant, has a well-suited band gap, and forms a excellent natural insulating oxide.
Controlling conductivity precisely, at a scale of billions of tiny regions, is what allows an entire processor to be built from silicon.

Atoms and Electrons