Metals conduct electricity easily; rubber and glass essentially don't conduct at all. Semiconductors, materials like silicon, sit deliberately in between — poor conductors in their pure form, but conductors that can be precisely, controllably switched on by adding tiny, specific impurities. That controllability, not raw conductivity, is what makes them the foundation of essentially all modern electronics.
Why "in between" is a feature, not a limitation
Pure silicon has a specific number of electrons tightly bound in place, leaving very few free to carry current, which is why it barely conducts on its own. Introducing tiny, deliberate impurities — a process called doping — changes that dramatically. Adding an element with one extra electron, like phosphorus, creates a surplus of free electrons, called an n-type semiconductor. Adding an element with one fewer electron, like boron, creates "holes" — missing electrons that behave like mobile positive charges — called a p-type semiconductor. Neither type alone is especially useful. Placing them next to each other is where things get interesting.
A junction that only lets current through one way
Where a p-type and n-type semiconductor meet, called a p-n junction, something asymmetric happens: current flows easily in one direction across the junction but is blocked in the other, creating a diode, the simplest and most fundamental semiconductor device. That one-way behaviour is the basic building block behind rectifying alternating current into direct current, and layering multiple p-n junctions together in more complex patterns is how transistors — the on-off switches inside every computer chip — are built.
What we're still unsure about
Semiconductor manufacturing has followed a remarkably consistent pattern of miniaturisation for decades — transistors have kept shrinking, roughly doubling in density on a chip every couple of years, a trend popularly known as Moore's Law — but transistors are now approaching physical limits measured in just a handful of atoms across, where quantum effects start interfering with reliable on-off switching. Whether continued miniaturisation, entirely new device physics, or a shift toward different computing architectures altogether will carry the industry's performance gains forward from here is a genuinely open, actively contested question in semiconductor engineering, not a settled roadmap.
This sits inside Semiconductor Physics & Diodes, one of eight topics in Electrical Engineering, one of four domains in Engineering, one of seventeen subjects the app can quiz you on.