Skip to content
Tutorials Center
My learning

Topics

Lesson 27 of 46 · Semiconductors and Metals

Intrinsic carriers and the mass-action law

Compute electron and hole densities in a pure semiconductor: effective density of states, the Fermi level at midgap, ni for silicon and the mass-action law.

At absolute zero a pure semiconductor is an insulator: the valence band is full and the conduction band is empty. As you warm it, a few electrons gain enough thermal energy to jump the gap. Each jump creates two carriers: an electron in the conduction band and a hole in the valence band. How many? The answer is an exponential in Eg/2kT, and it sets the scale for almost everything in semiconductor physics.

Thermal excitation across the gap makes electrons and holes in equal numbers
In an intrinsic crystal, the Fermi level sits near the middle of the gap and every electron in the conduction band leaves a hole in the valence band.

What you will learn

  • How to compute n and p from the effective density of states N_c, N_v and the Fermi level.
  • Why the intrinsic Fermi level sits near midgap and how effective masses shift it.
  • The mass-action law n p = n_i² = N_c N_v exp(−Eg/kT).
  • How n_i changes with temperature and gap, and how to extract Eg from data.

Key terms

Effective density of states N_c, N_v: the number of states per unit volume that effectively sit at the band edge, N_c = 2(2π m_e* kT/h²)^(3/2).

Intrinsic carrier density n_i: the density of electrons (equal to the density of holes) in a pure crystal at temperature T.

Boltzmann approximation: valid when E_c − E_F and E_F − E_v exceed a few kT, so the Fermi-Dirac function becomes a simple exponential.

Mass-action law: n p = n_i², independent of doping in equilibrium.

Counting carriers

Near the bottom of the conduction band the energy is E = E_c + ħ²k²/(2m_e*), so the density of states per volume is g_c(E) = (1/2π²)(2m_e*/ħ²)^(3/2) sqrt(E − E_c). The electron density is the integral of g_c(E) f(E) dE with the Fermi-Dirac function f. If E_c − E_F ≫ kT, f ≈ exp(−(E − E_F)/kT) and the integral gives

n = N_c exp(−(E_c − E_F)/kT), with N_c = 2 (2π m_e* kT / h²)^(3/2)

Similarly for holes: p = N_v exp(−(E_F − E_v)/kT). For m* = m₀ and T = 300 K, N = 2.51 × 10¹⁹ cm⁻³. Silicon has N_c ≈ 2.8 × 10¹⁹ and N_v ≈ 1.0 × 10¹⁹ cm⁻³ at 300 K, using density-of-states effective masses m_e* ≈ 1.08 m₀ and m_h* ≈ 0.56 m₀.

In a pure crystal, n = p. Setting the two expressions equal gives the Fermi level

E_F = (E_c + E_v)/2 + (kT/2) ln(N_v/N_c) = (E_c + E_v)/2 + (3kT/4) ln(m_h*/m_e*)

which is within a few percent of midgap, because the logarithm is small compared with Eg/kT ≈ 43 for silicon. Multiplying the two expressions eliminates E_F and gives the mass-action law:

n p = N_c N_v exp(−Eg/kT) = n_i²

so n_i = sqrt(N_c N_v) exp(−Eg/2kT). Notice that n_i depends on the gap through exp(−Eg/2kT): half the gap, because each excitation creates a pair.

Numbers for silicon

At 300 K, kT = 0.02585 eV. For silicon with Eg = 1.12 eV: Eg/2kT = 21.7, exp(−21.7) = 3.9 × 10⁻¹⁰, and sqrt(N_c N_v) ≈ 1.7 × 10¹⁹ cm⁻³, giving n_i ≈ 7 × 10⁹ cm⁻³ in this simple model. Measured values are close to 10¹⁰ cm⁻³; the difference comes from temperature dependence of Eg and from approximate effective masses. Silicon has about 5 × 10²² atoms per cm³, so only about one in 10¹³ atoms contributes an electron-hole pair. In germanium (Eg = 0.66 eV) n_i ≈ 2 × 10¹³ cm⁻³, a couple of thousand times larger. In GaAs (Eg = 1.42 eV) n_i ≈ 2 × 10⁶ cm⁻³.

Temperature dependence is dominated by the exponential. Between 300 K and 400 K, silicon's n_i increases by a factor of a few hundred (it roughly doubles for every 10 K near room temperature), which is why intrinsic behavior limits devices at high temperature and why wide-gap semiconductors are preferred for hot environments.

Reading Eg from data

Taking logarithms, ln(n_i / T^(3/2)) = const − Eg/(2kT). Plotting against 1/T gives a straight line with slope −Eg/2k. This is how gaps are measured from conductivity or Hall data: an Arrhenius plot of the intrinsic range. Optical absorption gives the gap directly and provides a cross-check.

Common misconceptions

"Intrinsic means undoped, so n_i is zero at room temperature." It is small but finite; it grows exponentially with temperature and is the reference density for every doped sample.

"The Fermi level is a real energy level that electrons occupy." In the gap there are no states. E_F is a statistical quantity: the energy at which the occupation probability would be one half.

"n p = n_i² holds only for pure crystals." It holds for any nondegenerate semiconductor in thermal equilibrium, doped or not. Doping changes n and p in opposite directions with their product fixed.

Try it

Python Effective density of states and n_i

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Python Arrhenius plot: recover the gap

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Check your understanding

Quick check · 5 questions

of 5 answered You scored / 5 · perfect!

  1. 1. In an intrinsic semiconductor, the electron and hole densities satisfy

    Each thermal excitation produces one electron and one hole.
  2. 2. The mass-action law states that in equilibrium

    n p = N_c N_v exp(−Eg/kT), independent of the doping level.
  3. 3. Which change raises n_i the most?

    n_i ∝ exp(−Eg/2kT), so halving Eg changes the exponent drastically.
  4. 4. Where does the Fermi level lie in an intrinsic semiconductor?

    Equal electron and hole densities put E_F at midgap, shifted slightly by the mass ratio.
  5. 5. The slope of ln(n_i / T^(3/2)) against 1/T equals

    n_i/T^(3/2) ∝ exp(−Eg/2kT), so the slope is −Eg/2k.

Summary

  • n = N_c exp(−(E_c − E_F)/kT) and p = N_v exp(−(E_F − E_v)/kT) in the Boltzmann regime.
  • In a pure crystal n = p = n_i = sqrt(N_c N_v) exp(−Eg/2kT), with E_F near midgap.
  • The mass-action law n p = n_i² holds for doped crystals too.
  • Silicon has n_i of about 10¹⁰ cm⁻³ at 300 K, germanium about 10¹³ and GaAs about 10⁶.

Next: Doping and the p-n junction