Semiconductor Physics: Foundations of Modern Electronics

Semiconductor physics sits at the intersection of quantum mechanics and practical engineering — it is the science that explains why silicon runs the world. This page covers the foundational principles of semiconductor behavior, from band theory and carrier dynamics to doping mechanisms and device-level tradeoffs. The field is central to every transistor, solar cell, LED, and sensor built in the past seven decades.

• Definition and scope
• Core mechanics or structure
• Causal relationships or drivers
• Classification boundaries
• Tradeoffs and tensions
• Common misconceptions
• Checklist or steps
• Reference table or matrix

Definition and scope

A semiconductor is a material whose electrical conductivity falls between that of a conductor and an insulator — and, crucially, whose conductivity can be tuned with extraordinary precision. Silicon has a room-temperature resistivity of roughly 640 Ω·m in its pure form, compared to copper at about 1.7 × 10⁻⁸ Ω·m. That ten-order-of-magnitude gap is not a flaw in silicon's résumé; it is the entire point. The gap is what makes control possible.

The scope of semiconductor physics spans equilibrium carrier statistics, non-equilibrium transport, optical interactions, and surface and interface phenomena. It is grounded in quantum mechanics — specifically in Bloch's theorem and the band theory of solids — but it connects directly to the macroscopic current-voltage relationships that engineers measure in laboratories. The Physics Authority index situates semiconductor physics within the broader landscape of applied and condensed-matter physics.

The materials in scope include elemental semiconductors (silicon, germanium), compound semiconductors (gallium arsenide, indium phosphide, gallium nitride), and the growing class of two-dimensional semiconductors such as molybdenum disulfide. Each material family has a distinct band gap, carrier mobility profile, and suitability for different device types.

Core mechanics or structure

The operating principle of every semiconductor device traces back to one concept: the band gap. In a crystalline solid, quantum mechanical interactions between atoms split discrete atomic energy levels into continuous bands. The valence band holds electrons at equilibrium; the conduction band, separated by the band gap energy E_g, is where electrons go when they gain enough energy to conduct.

Silicon's band gap is 1.12 eV at 300 K ([Sze and Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007]). Gallium arsenide's is 1.42 eV — wider, and a direct gap, meaning electrons transition between bands without a change in crystal momentum. That directness is why GaAs emits light efficiently and silicon does not: the physics of optical emission depends on whether the transition is momentum-conserving.

At finite temperature, thermal energy (k_B T ≈ 0.026 eV at 300 K) promotes a small but non-zero concentration of electrons into the conduction band, leaving behind positively charged vacancies called holes. Both electrons and holes are charge carriers. Their concentrations in an intrinsic (undoped) material are governed by Fermi-Dirac statistics and the density-of-states functions in each band, producing the intrinsic carrier concentration n_i — approximately 1.5 × 10¹⁰ cm⁻³ for silicon at 300 K (Sze and Ng, 2007).

Transport of carriers is driven by two mechanisms: drift (response to an electric field) and diffusion (response to a concentration gradient). The Einstein relation links the two: D/μ = k_B T/q, where D is the diffusion coefficient and μ is the carrier mobility. Electron mobility in bulk silicon is approximately 1,400 cm²/V·s; hole mobility is approximately 450 cm²/V·s — a difference rooted in the curvature of their respective energy bands.

Causal relationships or drivers

Doping is the deliberate introduction of impurity atoms to shift carrier concentrations by orders of magnitude. Add phosphorus (a Group V element) to silicon, and each phosphorus atom donates a loosely bound electron to the conduction band, creating an n-type material with electron concentration far exceeding n_i. Add boron (Group III) and each atom accepts an electron, creating a hole and a p-type material.

A doping concentration of 10¹⁷ cm⁻³ — achievable by ion implantation with dose control — raises electron concentration from the intrinsic 1.5 × 10¹⁰ cm⁻³ to approximately 10¹⁷ cm⁻³, a seven-order-of-magnitude change in carrier density. This tunability is the mechanical heart of semiconductor technology.

The p-n junction arises when p-type and n-type regions meet. Electrons diffuse toward the p-side; holes diffuse toward the n-side. The resulting charge separation creates a built-in electric field that opposes further diffusion, reaching equilibrium at a built-in potential V_bi typically between 0.6 V and 0.9 V for silicon. Forward bias reduces this barrier; reverse bias increases it. The asymmetric current response — the Shockley diode equation — is not an approximation invented for convenience; it falls directly from the mathematics of minority carrier injection and diffusion across the junction (NIST, semiconductor device fundamentals literature).

Temperature drives carrier concentration exponentially. Every 10°C rise approximately doubles the intrinsic carrier concentration in silicon, which is why thermal management is a first-order design constraint rather than an afterthought in power semiconductor devices.

Classification boundaries

Semiconductors are classified along three primary axes:

Band gap type. Direct-gap materials (GaAs, InP, GaN) allow efficient radiative recombination. Indirect-gap materials (Si, Ge) require phonon assistance for electron-hole recombination, suppressing light emission but not electronic switching.

Elemental vs. compound. Elemental semiconductors (Si, Ge) have a single atomic species. Binary compounds (GaAs, SiC) combine two; ternary (AlGaAs) and quaternary (InGaAsP) compounds allow band gap engineering by composition tuning.

Doping polarity. N-type materials carry current primarily by electrons; p-type by holes. The minority carrier lifetime — a key parameter in bipolar devices — measures how long injected minority carriers survive before recombining. In high-purity silicon, minority carrier lifetime can exceed 1 millisecond; in heavily doped or defect-rich material, it may fall below 1 microsecond.

Understanding where a given material falls on these axes is foundational for anyone working through the conceptual frameworks outlined at How Science Works: Conceptual Overview.

Tradeoffs and tensions

The central tension in semiconductor device design is between speed and power dissipation. High carrier mobility enables fast switching but generates heat. Gallium nitride (GaN), with an electron mobility of approximately 2,000 cm²/V·s in bulk and a critical electric field of about 3.3 MV/cm (vs. silicon's 0.3 MV/cm), can switch faster and handle higher voltages — but GaN substrates cost significantly more than silicon wafers, and GaN processing is less mature.

Silicon carbide (SiC) presents a different tradeoff: its thermal conductivity of approximately 4.9 W/cm·K (vs. silicon's 1.5 W/cm·K) makes it superior for high-temperature power electronics, but its wide band gap (3.26 eV for the 4H polytype) requires higher activation energies for doping, complicating fabrication.

Scaling transistors to smaller geometries — the logic behind Moore's Law — reduces switching energy but introduces quantum tunneling leakage through gate oxides thinner than 2 nm, short-channel effects that erode the transistor's ability to switch cleanly between on and off states, and variability from statistical dopant fluctuations when a transistor contains fewer than 100 dopant atoms.

The field also grapples with the tension between model simplicity and physical accuracy. The drift-diffusion model describes most device behavior well but breaks down in sub-100 nm devices where carrier transport becomes quasi-ballistic and energy-dependent scattering requires full Boltzmann transport equation treatment.

Common misconceptions

Holes are not real particles. Holes are collective quantum mechanical constructs — the absence of an electron in the valence band behaves, for purposes of charge transport, like a positively charged particle. The math works because the valence band curvature is negative, giving holes a positive effective mass. They are a convenience that maps exactly onto measurable phenomena like the Hall effect and minority carrier diffusion.

Semiconductors do not conduct better when hotter because they "wake up." The increase in conductivity with temperature in intrinsic semiconductors occurs because n_i grows exponentially with temperature — more electron-hole pairs are thermally generated. In doped (extrinsic) semiconductors, increasing temperature above a certain range actually decreases mobility due to increased phonon scattering, which can reduce conductivity despite having more carriers.

A wider band gap does not automatically mean a better semiconductor. Diamond has a band gap of 5.5 eV and excellent thermal properties but suffers from extreme difficulty in p-type doping, making it impractical for most device applications despite decades of research interest.

Silicon is not used because it is the best semiconductor for every property. Silicon dominates the industry largely because silicon dioxide (SiO₂) forms a near-perfect native oxide that serves as an excellent gate dielectric, and because decades of manufacturing refinement have made silicon wafers defect-dense at fewer than 1 defect per cm² in leading-edge production.

Checklist or steps

The following sequence describes the physical analysis process for a p-n junction device — not a fabrication procedure, but the logical chain a physicist works through when characterizing junction behavior:

Reference table or matrix

Comparison of common semiconductor materials at 300 K

Mobility values are bulk, undoped approximations. Device-layer values differ due to confinement, interface scattering, and doping. Sources: Sze and Ng (2007); Madelung, Landolt-Börnstein Series.