Semiconductor

Definition: A semiconductor is a material whose electrical conductivity falls between that of a conductor and an insulator, and can be precisely controlled, most commonly silicon.

How It Works

Pure (“intrinsic”) silicon has four valence electrons that bond with neighboring atoms, leaving few free carriers at room temperature, so it conducts poorly on its own
Conductivity is tuned by “doping,” deliberately introducing trace impurities into the crystal lattice
Adding a pentavalent element like phosphorus contributes extra free electrons, creating n-type material with negative charge carriers
Adding a trivalent element like boron leaves “holes,” missing electrons that behave like mobile positive charges, creating p-type material
Combining n-type and p-type regions creates a p-n junction, the fundamental building block of diodes, transistors, and integrated circuits
At the junction, electrons and holes diffuse across and recombine, forming a depletion region with a built-in electric field that resists further current flow
Applying an external voltage can widen or narrow that depletion region, giving the junction its ability to control current direction and flow
Conductivity in semiconductors also rises with temperature, opposite to metals, because more electrons gain enough energy to jump into the conduction band
The energy gap between the valence band (bound electrons) and the conduction band (free electrons) is called the bandgap, and it determines how easily a material conducts
Intrinsic (undoped) semiconductors have equal numbers of electrons and holes; extrinsic (doped) semiconductors have one carrier type dominating, called the majority carrier
Light can also excite electrons across the bandgap, which is the basis for photodiodes, solar cells, and image sensors built from semiconductor material
Crystal purity matters enormously: fabrication starts from ultra-pure single-crystal ingots, since stray impurities or lattice defects disrupt the carefully engineered doping profile

Illustration

Under the Hood

Conductivity relationship (simplified) showing dependence on carrier concentration:

σ = q × (n × µn + p × µp)

q: elementary charge
n, p: electron and hole concentrations
µn, µp: electron and hole mobility

Diode equation describing current through a p-n junction:

I = Is × (e^(V/(n×VT)) - 1)

Is: reverse saturation current
VT: thermal voltage, about 0.026V at room temperature
n: ideality factor, close to 1 for a typical diode

Bandgap energy determines whether a material is a conductor, semiconductor, or insulator:

Silicon Eg ≈ 1.12 eV
Germanium Eg ≈ 0.66 eV

Worked Problem 1: Estimating forward voltage drop Given: A silicon p-n junction has Is = 1×10⁻¹² A, room-temperature VT = 0.026V, n = 1, and carries 10 mA (0.01A) forward current. Step 1: Rearranged diode equation: V = VT × ln(I/Is + 1) Step 2: V = 0.026 × ln(0.01 / 1e-12 + 1) ≈ 0.026 × ln(1×10¹⁰) ≈ 0.026 × 23.03 Answer: V ≈ 0.60V, consistent with the typical ~0.6-0.7V forward drop of a silicon diode.

Worked Problem 2: Comparing bandgaps Given: Silicon has Eg ≈ 1.12 eV, germanium has Eg ≈ 0.66 eV. Step 1: A smaller bandgap means less energy is needed to free a charge carrier into conduction Step 2: Germanium devices conduct more easily at lower voltage but leak more current at high temperature Answer: Silicon’s larger bandgap gives it better high-temperature stability, which is why it displaced germanium in most modern devices.

Worked Problem 3: Doping direction Given: Pure silicon (group IV) is doped with arsenic (group V, 5 valence electrons). Step 1: Arsenic contributes one extra electron beyond what’s needed for bonding Step 2: That extra electron is loosely bound and becomes a free (negative) charge carrier Answer: The result is n-type silicon, since the majority carriers are electrons.

Worked Problem 4: Doping the opposite direction Given: Pure silicon (group IV) is doped with boron (group III, 3 valence electrons). Step 1: Boron has one fewer valence electron than needed to complete the bonding structure Step 2: This leaves a “hole,” a missing bond that a neighboring electron can move into, effectively moving the hole Answer: The result is p-type silicon, since the majority carriers behave as mobile positive holes.

Why It Matters

Semiconductors are the physical basis for diodes, transistors, and integrated circuits, essentially all modern electronics
Doping lets engineers custom-tailor a material’s electrical behavior at the atomic level, something impossible with fixed conductors or insulators
The same base material, silicon, can be processed into memory, logic, sensors, and power devices depending on doping pattern and structure
Semiconductor manufacturing (fabrication) scaling has driven the exponential growth in computing power described by Moore’s Law for decades
Wide-bandgap semiconductors like GaN and SiC are enabling smaller, more efficient power supplies and faster electric vehicle chargers than silicon alone could achieve
Semiconductor supply chains are now considered strategically critical infrastructure by many governments, given how concentrated advanced chip fabrication capacity is globally

Common Pitfalls

Assuming “semiconductor” means “half a conductor” in terms of resistance value, when it actually refers to a material class whose conductivity can be engineered across a huge range
Forgetting that semiconductor behavior is strongly temperature-dependent, unlike a plain resistor, so devices can fail or behave unpredictably outside their rated temperature range
Confusing n-type/p-type material with actual net electric charge, doped silicon is still electrically neutral overall, only the type of majority mobile carrier differs
Handling static-sensitive semiconductor devices without ESD precautions, a static discharge can punch through microscopic junction structures instantly
Assuming all semiconductors are silicon, when compound semiconductors like gallium arsenide (GaAs), gallium nitride (GaN), and silicon carbide (SiC) are used for high-frequency, high-power, or optical applications
Overheating a semiconductor junction beyond its rated maximum, since excess heat can trigger thermal runaway where rising temperature increases current, which raises temperature further
Ignoring that “n-type” and “p-type” refer to majority carrier type only, both materials remain electrically neutral overall and are not themselves charged

Comparison

Material Class	Conductivity	Example	Typical Use
Conductor	High, ~10^6-10^8 S/m	Copper, silver	Wires, connectors
Semiconductor	Tunable, ~10^-6-10^4 S/m	Silicon, germanium	Diodes, transistors, chips
Insulator	Very low, <10^-10 S/m	Glass, rubber, ceramic	Wire coating, isolation
Superconductor	Effectively infinite below critical temp	Niobium-titanium	MRI magnets, research
Wide-bandgap semiconductor	Tunable, higher voltage/temp tolerance	Gallium nitride, silicon carbide	High-power, high-frequency devices

Example

Silicon wafers doped in precise, microscopic patterns are etched, layered, and interconnected into the billions of transistors inside a modern CPU, each transistor acting as a tiny switch built from carefully arranged p-n junctions.

History

The semiconducting properties of materials like selenium and galena were observed in the 19th century, long before the physics was understood.
Point-contact diodes (“cat’s whisker” detectors) made from galena crystals were used in early crystal radios in the early 20th century.
The first practical transistor was demonstrated at Bell Labs in 1947 by John Bardeen, Walter Brattain, and William Shockley, who shared the 1956 Nobel Prize in Physics for the work.
Jack Kilby and Robert Noyce independently developed the integrated circuit around 1958-1959, combining multiple semiconductor devices on a single chip.
Silicon overtook germanium as the dominant semiconductor material by the 1960s due to its higher melting point, better thermal stability, and abundant natural silicon dioxide for insulation.
The term “Silicon Valley” itself traces back to the concentration of semiconductor companies that grew around Stanford University and Shockley Semiconductor Laboratory starting in the late 1950s.

FAQ

Why is silicon used more than other semiconductors? Silicon is abundant, forms a stable native oxide (silicon dioxide) useful as an insulating layer during fabrication, and has a bandgap that balances good conductivity control with reasonable high-temperature stability.

What does “doping” actually change physically? It changes the number and type of mobile charge carriers available in the crystal, without significantly altering the material’s other bulk properties, allowing precise control of conductivity and junction behavior.

Is a semiconductor always solid-state? In modern electronics, essentially yes, semiconductors refer to crystalline solid materials. The term “solid-state electronics” itself originated to distinguish semiconductor-based devices from earlier vacuum tube technology.

Why do semiconductors conduct better as they heat up, unlike metals? In metals, heating increases atomic vibration, which scatters electrons and increases resistance. In semiconductors, heating instead frees more electrons across the bandgap into the conduction band, and this carrier increase outweighs the scattering effect.

What’s the difference between intrinsic and extrinsic semiconductors? Intrinsic semiconductors are undoped, pure material with carrier concentration set entirely by temperature. Extrinsic semiconductors have been deliberately doped, giving engineers direct control over carrier type and concentration independent of temperature.

Can a semiconductor become an insulator or conductor under different conditions? Yes. At very low temperatures, a semiconductor can behave almost like an insulator since few carriers have enough energy to conduct. At very high doping levels, it can approach metal-like conductivity, which is exploited in the heavily doped regions of real transistors.

Why are semiconductor chips manufactured in cleanrooms? Doping patterns and transistor features are measured in nanometers, so even microscopic dust particles can ruin a device by disrupting a junction or short-circuiting adjacent structures, making extreme contamination control essential during fabrication.

What does “doping concentration” actually control? It sets how many majority carriers are available to conduct, which in turn shapes a device’s resistance, breakdown voltage, and switching speed, engineers tune doping profiles precisely to hit a target electrical behavior for each region of a chip.

Referenced by

Dig deeper

Spot an error in this note?