If you’ve ever asked how does a pn junction work, here is the short answer. Joining p-type and n-type silicon creates a depletion region, and that region lets current pass in only one direction.
This page is part of the semiconductor basics section, which covers the building blocks behind diodes and transistors.
| Item | Value |
|---|---|
| Cut-in voltage, silicon diode | About 0.8 V |
| Cut-in voltage, GaAs diode | About 1.2 V |
| Junction invention usually attributed to Russell Ohl, Bell Laboratories | 1939 |
| Modern junction theory elucidated by William Shockley | 1950 |
These figures come from the University of Liverpool PHYS389 lecture notes and the Wikipedia p-n junction article. Real devices vary and sources get revised, so check the source before you rely on them.
What is a pn junction made of?
A pn junction is a combination of p-type and n-type semiconductor material in a single crystal, as the Wikipedia p-n junction article describes it. The n side holds freely moving electrons, while the p side holds freely moving electron holes.
Each side gets its character from added impurities. The detail behind those impurities is covered in what is doping in semiconductors.
- p-type: trivalent atoms such as boron are added to silicon, leaving a deficiency of electrons that creates holes.
- n-type: pentavalent atoms such as phosphorus or arsenic are added to silicon, which increases the number of free electrons.
- One crystal: the two regions sit side by side in the same piece of silicon, with a continuous boundary between them.
That last point matters. Newcastle University’s lecture notes explain that a practical junction can only be created by inserting different impurities into different parts of a single crystal.
How the depletion region forms
The interesting part happens the moment the two regions meet. Nothing external is needed, because the junction sets itself up through a short sequence of events.
- Free electrons from the n side wander into the p side by random thermal diffusion, and holes wander the other way.
- Near the boundary the electrons and holes recombine and cancel each other out.
- What remains are fixed, charged dopant atoms that cannot move: positive on the n side, negative on the p side.
- Those fixed charges create an electric field that opposes further diffusion, so the process settles into equilibrium.
The charge-stripped zone around the boundary is the depletion region. Across it sits a built-in potential, which ROHM’s TechWeb article puts at typically 0.6-0.7 V in silicon at room temperature.
Think of the built-in potential as a hill. Without outside help, carriers cannot climb it, so no large current flows.
Forward bias: why current flows
Bias simply means applying a voltage across the junction. Attaching the positive terminal to the p side and the negative terminal to the n side is called forward bias, according to ROHM’s TechWeb article.
That voltage works against the built-in potential. The depletion layer narrows, the barrier drops, and electrons and holes can cross the junction far more easily.
Seen from the carriers’ side, three things change at once when you forward-bias a junction.
- The energy barrier that electrons and holes face is reduced.
- The depletion region gets narrower, so the carriers have a shorter gap to cross.
- Charge carriers can then flow freely across the boundary, as the Wikipedia p-n junction article explains.
Current does not switch on at a single sharp point, but there is a useful rule of thumb. The Liverpool notes call the voltage at which forward current becomes significant the cut-in voltage, and they give it as about 0.8 V for silicon diodes and about 1.2 V for GaAs diodes (Liverpool lecture notes).
Once the applied voltage passes the built-in potential, the junction stops resisting. As ROHM’s TechWeb article puts it, if the forward voltage surpasses the built-in potential, a large current begins to flow.
The rise is steep rather than gradual. Under forward bias the current increases exponentially with the voltage, so a small extra push gives a much larger current. That steepness is also a hazard, because at high current densities the device may heat and suffer burnout.
Reverse bias: why current stops
Flip the connections and the picture reverses. Applying the negative terminal to the p side and the positive terminal to the n side is reverse bias, and it raises the effective barrier and blocks carrier flow.
The depletion region widens, because carriers are pulled further from the boundary. What is left is a very small leakage current, which ROHM’s TechWeb article calls practically negligible in normal operation.
This one-way behaviour is the whole point. It is what makes functions like AC-to-DC conversion and switching possible in a diode.
Reverse bias has a limit. Under very high electric fields, electrons gain enough energy to knock other electrons into the conduction band, a process called impact ionisation, and the junction starts to break down.
The diode equation in plain terms
Engineers describe all of this with the Shockley diode equation. You do not need to solve it to use the idea, because it says three things.
- Under forward bias, the current increases exponentially with the voltage.
- Under reverse bias, the current levels off at a small saturation value.
- The result is not Ohm’s-law behaviour but strongly rectifying behaviour.
The Liverpool notes sum it up neatly: this strong asymmetry in the diode current is what makes the p-n diode attractive for many applications. Push forward current too high and the device may heat up and suffer burnout, so real circuits limit it.
Where pn junctions show up
A pn junction on its own, placed in a circuit, is a diode. More complex devices come from combining p-type and n-type regions in further layers.
- Transistors such as the BJT, MOSFET, JFET and IGBT all use pn junctions for operation.
- Solar cells and LEDs are essentially pn junctions, designed to absorb or emit light.
- Integrated circuits combine many such devices on a single chip.
If the transistor case is what you came for, the next step is how does a transistor work.
The history is short. Russell Ohl of Bell Laboratories is usually credited with the invention in 1939, and William Shockley laid out the modern theory in his 1950 book Electrons and Holes in Semiconductors, per the Wikipedia p-n junction article.
Frequently asked questions
Is a pn junction the same as a diode?
Show answer
Almost. A pn junction by itself, when placed in a circuit, is a diode, and more complex parts build on the same structure.
Why does the depletion region stop growing?
Show answer
Diffusion keeps creating fixed charge, but that charge builds an electric field that opposes further diffusion. When the two effects balance, the junction is in equilibrium.
What happens if the reverse voltage is too high?
Show answer
Very high fields can trigger impact ionisation, where energetic electrons free more electrons, and current rises sharply. The Liverpool lecture notes describe this breakdown behaviour.
Can you make a pn junction by pressing two crystals together?
Show answer
Not usefully. A practical junction is made by putting different impurities into different parts of one crystal.
