What Is a Semiconductor? How Conductivity Can Be Controlled

14 min read · Updated: October 23, 2026

A semiconductor is a material that conducts electricity better than glass or rubber (insulators) but worse than metals (conductors). What matters most is not this middle position. It is that tiny amounts of impurities, or voltage, light and temperature, can change how well it conducts by a huge amount. This property lets us build switches that turn current on and off. That is why semiconductors are the foundation of phones, computers and the electronics in cars.

What you'll learn

  • What the name "semiconductor" means, and how it differs from a conductor and an insulator
  • Four ways to change how easily a semiconductor conducts
  • Why silicon is the material of choice
  • What n-type and p-type mean (how impurities create carriers of electric charge)
  • Whether "semiconductor" in the news means a material, a component or a chip

Prerequisites: basic school science (current, voltage, atoms and electrons). No equations.

Where you meet semiconductors

Your phone contains parts that compute, parts that store photos and contacts, and parts that send and receive radio signals. At the heart of each of these parts is a semiconductor. The camera part that turns light into an electrical signal is a semiconductor too. It is called an image sensor.

At home, the glowing part of an LED bulb and the solar panels on a roof are semiconductors. An LED turns electricity into light, and a solar cell turns light into electricity. Air conditioners and electric cars have parts that switch large currents on and off very quickly. These parts control the power sent to the motor. This kind of part is called a power semiconductor.

When the news talks about a "semiconductor shortage" or a "semiconductor plant," it usually does not mean the material itself. It means parts made from semiconductors, especially "chips": small slabs packed with circuits. We sort out these meanings later, in "Three meanings of the word 'semiconductor'."

What the name "semiconductor" means

Materials can be roughly sorted into three groups by how easily they conduct electricity. Metals such as copper and aluminum conduct well, and we call them "conductors." Glass, rubber and plastic barely conduct at all, and we call them "insulators." Semiconductors fall in between. The name says so directly: "semi" (half) plus "conductor."

To compare materials fairly, we measure how strongly a piece of a fixed shape and size resists current. This value is called "resistivity," and its unit is the ohm-centimeter (Ω·cm). The smaller the value, the more easily the material conducts. Copper and silicon dioxide (SiO₂), the main ingredient of glass, differ in resistivity by about 10²⁰ times. That is a 1 followed by 20 zeros.

Resistivity of conductors, semiconductors and insulators, and how far impurities lower the resistivity of silicon ← conducts more conducts less → Conductor Semiconductor Insulator 10⁻⁶ 1 10⁶ 10¹² Copper Pure silicon SiO₂ A little impurity makes it conduct orders of magnitude better Resistivity (Ω·cm), ×10⁶ per tick
The horizontal axis is resistivity (further right means harder to conduct). What this figure leaves out: the borders between the three groups are rough guides, not sharp lines. How far impurities can shift silicon is also approximate.

Still, if you think of semiconductors as materials that "conduct only halfway," you miss their most important point. As the figure shows, pure silicon is actually quite a poor conductor. The real feature of a semiconductor, as we will see next, is that its conductivity can be changed a lot from the outside.

The key point: conductivity can be changed

Electric current flows because tiny particles carrying electric charge move. Metals naturally have many electrons that can move freely. In a semiconductor, the number of these "charge carriers" can be raised or lowered after the fact.

There are four ways to change how well a semiconductor conducts. None of them changes a metal's conductivity much.

  • Add impurities: Mix in a very small amount of another element. This sets a part's properties, so it is the central method of making semiconductors. The amount is tiny. Even one atom per 5 million silicon atoms makes silicon conduct hundreds of thousands of times better.
  • Apply a voltage: Place a metal plate (an electrode) on a thin insulating film and apply a voltage. Charge carriers are drawn toward it, and only the region just under the film starts to conduct. The switch called a transistor works this way.
  • Shine light on it: The energy of light creates charge carriers, so current flows more easily. Camera image sensors and solar cells use this property.
  • Change the temperature: A pure semiconductor conducts better as it gets hotter. Metals behave the opposite way. This property is also used in parts that measure temperature.

For electronics, the first two matter most. Impurities set the type and amount of charge carriers. A voltage then switches the flow on and off where it is needed. This combination is the basis of semiconductor parts.

Why silicon?

Many substances can act as semiconductors. Even so, most chips that compute or store data in computers and phones are made of silicon. There are three main reasons.

1. There is plenty of it

Earth's crust is the layer of rock that covers the planet. Silicon makes up more than a quarter of its weight. It is the second most abundant element there, after oxygen. Sand and quartz are also mainly silicon bonded to oxygen. Sand cannot be used as it is, though. It has to be purified in many steps.

2. It easily forms a good insulating film

Heat silicon to a high temperature in oxygen, and a clean film of silicon dioxide (SiO₂) grows on its surface. This method is called thermal oxidation. SiO₂ barely conducts, and its boundary with the silicon has few defects. So it works well as a film that separates conducting parts, such as the thin film under a transistor's electrode.

In fact, the first transistor (1947, Bell Labs in the US) was made of a different semiconductor, germanium. Silicon later took over, largely because it can form this good film. Silicon has another advantage, too: it leaks less current as the temperature rises.

3. Very pure, large crystals can be made

Semiconductor parts are built on crystals, in which atoms line up in a regular pattern. A piece that is one continuous crystal throughout is called a "single crystal." For silicon, the technology to grow single crystals 30 cm across is well established. These are sliced into thin round disks called wafers, and parts are built on them.

Raw silicon for chips reaches a purity of 99.999999999%. There are eleven 9s, so this level is called "eleven nines." It means no more than one impurity atom per 100 billion atoms. Purity is pushed this far because the properties are set by adding a tiny, chosen amount of impurity. If unwanted atoms are mixed in from the start, the material will not turn out as planned.

n-type and p-type: making charge carriers with impurities

A silicon atom has four electrons in its outermost shell. In a crystal, each atom shares electrons with its four neighbors, as if holding hands with them. All of these electrons are used in bonds, so pure silicon has almost no electrons that can move freely. That is why pure silicon conducts poorly.

Now mix in a little phosphorus (P), which has five outer electrons. A phosphorus atom takes the place of a silicon atom and uses four electrons for bonds. The fifth electron is only loosely held. At room temperature, almost all of these electrons break free and move around. An impurity that gives away electrons like this is called a donor.

Silicon with donors added carries current with electrons, which have negative charge. It is called an "n-type semiconductor," where n stands for negative. The phosphorus keeps an equal amount of positive charge behind. So the material as a whole does not become negatively charged.

If you mix in boron (B) instead, which has three outer electrons, one bond is short of an electron. An electron from a neighboring bond can move into this empty spot. The empty spot then moves the opposite way. It behaves like a particle with positive charge, so it is called a "hole."

An impurity that takes electrons, like boron, is called an acceptor. Silicon with acceptors added, where holes carry the current, is a "p-type semiconductor." The p stands for positive.

An n-type silicon crystal with phosphorus added and a p-type silicon crystal with boron added n-type: phosphorus p-type: boron Si Si Si Si P Si Si Si Si Si Si Si Si B Si Si Si Si Extra electron Empty spot (hole)
Lines are bonds between atoms. On the left, the extra electron from phosphorus (P) carries current. On the right, the empty spot next to boron (B) does. What this figure leaves out: a real crystal is three-dimensional, and each bond line holds two electrons. Real impurities are much sparser, for example about one per several million atoms.

Particles that carry electric charge, such as electrons and holes, are together called "carriers." In n-type silicon, electrons are the more numerous carrier. In p-type silicon, holes are. The more numerous kind is called the majority carrier. In both types, the added impurity is tiny compared with the number of silicon atoms.

One step further: the band gap

So far, we have pictured bonds between atoms. Physics textbooks explain the same thing with "levels" of electron energy. This view lets us compare conductors, semiconductors and insulators in a single picture.

Electrons in a crystal can only have energies within certain ranges, called bands. In semiconductors and insulators, the lower band is full of electrons, so as it is, they cannot move. Electrons that rise into the upper band can carry current. So can the empty spots (holes) they leave behind in the lower band. The gap between the two bands, where electrons cannot be, is called the band gap.

The size of the gap is given in a unit called the electronvolt (eV). One eV is the energy one electron gains when pushed by a voltage of 1 volt. Silicon's gap is about 1.1 eV, and that of the insulator silicon dioxide is about 9 eV. In a metal, there is free room just above the band holding the electrons, so there is no gap.

Size of the band gap in a metal, silicon and silicon dioxide ↑ higher = more electron energy Empty Filled Empty Full Empty Full Metal Silicon SiO₂ No gap about 1.1 eV about 9 eV
Shaded bands are filled with electrons; outlined bands are empty. The vertical lines are the gaps (band gaps), drawn roughly to scale. What this figure leaves out: the width and shape of the bands, and the levels that donors and acceptors create.

Silicon's gap is small enough that room-temperature heat lifts a very small number of electrons across it. More electrons cross as the temperature rises. So a pure semiconductor conducts better when it is hotter. Light helps current flow for the same reason: its energy lifts electrons across the gap. Donor electrons, by contrast, need far less energy than the gap to move into the upper band.

That is why we can build switches

Join p-type and n-type silicon, and current passes through the boundary much more easily in one direction. Make the p-type side positive, and current flows well. Reverse the voltage, and almost no current flows. This one-way property is called "rectification."

A part with this property is a diode. One use is in circuits that turn the alternating current (AC) from a wall outlet into direct current (DC) for electronics. AC keeps switching direction, while DC always flows the same way. An LED is also a kind of diode, one that turns electricity into light. You can see what happens at the p–n boundary, and change the conditions yourself, in the p–n junction figure.

Combine n-type and p-type regions and add electrodes, and you can build a part with a special property. A path for current appears or disappears depending on the voltage on one electrode. This part is the transistor. It is a switch run by electricity: a small voltage signal turns a current on and off. The next article, "What Is a Transistor?", explains how it works.

A switch that is on or off can stand for two states, 1 and 0. Combine many transistors, and you get circuits that calculate and remember with 0s and 1s. The computing part of a phone, such as its CPU, is in the end a collection of these tiny switches (see "What Is a CPU?").

Three meanings of the word "semiconductor"

The word "semiconductor" refers to three different things, depending on the context. When you read the news, it helps to notice which one is meant.

What it refers toExamplesWhere it often comes up
MaterialSilicon, silicon carbide (SiC), gallium nitride (GaN)Physics and materials
Component (one part, one job)Diodes, power semiconductorsPower supplies and motors
Chip (integrated circuit)CPUs, memory"Semiconductor shortage," "semiconductor plant"

The third, a chip, is a small piece of silicon with many transistors and wires built onto it. Its formal name is integrated circuit, or IC for short. Most uses of "semiconductor" in the news mean ICs.

Parts like those in the second row do one job each, as a single part. To set them apart from ICs, they are called discrete semiconductors. Semiconductors are essential to everything from home appliances to cars. In Japan, they have long been called "the rice of industry," meaning a staple that every industry needs.

Common misconceptions

Myth: A semiconductor "conducts electricity halfway"

In fact, its key feature is that its conductivity can be changed a lot. Pure silicon is actually a poor conductor. Depending on the amount of impurity, it can range from quite poor to nearly as good as a metal. That is a change of many orders of magnitude. The myth comes from the name "half conductor," spread together with the idea that its resistivity is "in the middle."

Myth: "Semiconductor" means a chip (an IC)

In fact, the word can mean a material, a component or a chip. The glowing part of an LED bulb and solar cells are semiconductors too. But they are not ICs packed with many transistors. The myth arises because the news often shortens "semiconductor chip" to just "semiconductor."

Myth: All semiconductors are made of silicon

In fact, other materials are used for some jobs. Blue and white LEDs are made from gallium nitride (GaN) and related materials, because silicon is poor at giving off light. Power semiconductors, such as those in electric cars, also widely use silicon carbide (SiC), which withstands high voltages. The myth arises because most chips for computing and memory are made of silicon.

By the numbers

NumberWhat it describes
About 10²⁰ timesDifference in resistivity between copper and silicon dioxide
Hundreds of thousands of Ω·cm → about 0.5 Ω·cmResistivity of pure silicon, and of silicon with one phosphorus atom per 5 million atoms (room temperature)
99.999999999%Purity level of raw silicon for semiconductors ("eleven nines")
About 1.1 eVBand gap of silicon (room temperature). About 9 eV for SiO₂
More than a quarterShare of the weight of Earth's crust that is silicon
300 mmDiameter of the largest wafers used in mass production
$791.7 billionWorldwide semiconductor sales in 2025 (Semiconductor Industry Association, SIA, announced February 2026)

Put two of these numbers side by side. Raw silicon is purified until impurities are no more than one in 100 billion atoms. Then, for example, one phosphorus atom per 5 million is added on purpose. Making semiconductors is, in part, the craft of controlling which impurity goes where, and how much.

Learn more

Read next

Articles in the series “Semiconductors for beginners”

  1. How to Start Learning Semiconductors: Beginner Reading Order
  2. What Is a Semiconductor? How Conductivity Can Be Controlled
  3. What Is a Transistor? How a Transistor Works as a Switch
  4. What Is an Integrated Circuit? IC vs Chip vs Die vs SoC
  5. What Does 2nm or 3nm Mean? Chip Process Nodes Explained
  6. What Is a Diode? How It Lets Current Flow Only One Way
  7. What Is CMOS? How nMOS and pMOS Pairs Save Power
  8. What Is a GAA Transistor? Gate-All-Around Nanosheets

Questions on this topic: Semiconductor physics

Sources

All explainers