Skip to content
Tutorials Center
My learning

Topics

Lesson 27 of 39 · Quantum Statistics: Fermi-Dirac and Bose-Einstein

Bose-Einstein Condensation

Derive the critical temperature of an ideal Bose gas, the condensate fraction and the role of the zeta function, and compute numbers for liquid helium and trapped ultracold atoms.

Fermions are forced apart by the Pauli principle. Bosons do the opposite: below a critical temperature, a macroscopic fraction of an ideal Bose gas falls into the single lowest-energy state. This is Bose-Einstein condensation, predicted by Einstein in 1925 (building on Bose's work on photon statistics) and first observed in dilute atomic vapors in 1995. Here you derive the transition temperature and the condensate fraction.

Why the ordinary calculation fails

For a gas of non-interacting massive bosons in a box of volume V, the number of particles is the sum of Bose-Einstein occupations:

Text
N = Σ_i 1 / ( e^((ε_i - μ)/kT) - 1 )

Because occupations must be non-negative, μ ≤ ε₀ (the ground-state energy, which we set to 0), so μ ≤ 0. Replace the sum by an integral using the density of states g(ε) = (V/4π²)(2m/ħ²)^(3/2) √ε (a single spin state, no degeneracy factor):

Text
N = ∫₀^∞ g(ε) dε / ( e^((ε - μ)/kT) - 1 )

The largest this integral can get, as μ → 0⁻, is

Text
N_max = V ζ(3/2) / λ³,     ζ(3/2) = 2.612375...

with λ = h/√(2π m k T) the thermal wavelength. At fixed N and V, as T decreases λ grows, and N_max shrinks. When N_max < N, the integral cannot hold all the particles. The missing ones are in the ground state, which the integral missed because g(0) = 0. They form the condensate.

Critical temperature and condensate fraction

Setting N = N_max defines the critical temperature, equivalently the condition n λ³ = ζ(3/2):

Text
T_c = (2π ħ² / (m k)) ( n / ζ(3/2) )^(2/3)

Below T_c, the number of particles in excited states is N_ex = N (T/T_c)^(3/2), so the condensate fraction is

Text
N₀ / N = 1 - (T/T_c)^(3/2)       (T < T_c)

At T = 0 all particles are in the ground state; at T_c the condensate vanishes continuously. This is a genuine phase transition driven purely by statistics, with no interaction.

Condensation does not mean the particles are close together in space. It means they occupy the same quantum state, so the gas acts like one giant matter wave. The ideal-gas model overlooks interactions, which matter in real systems.

Numbers: liquid helium and a laboratory gas

Helium-4 atoms are bosons. Liquid helium has a mass density of about 0.145 g/cm³ near its transition. Treated as an ideal Bose gas, the formula gives T_c of about 3.1 K. The real superfluid transition (the lambda point) occurs at 2.17 K: the same order of magnitude, and a long-standing hint that superfluidity is related to condensation, although interactions in liquid helium are strong.

In dilute alkali-atom gases, the density is about a million times lower than in liquids, so T_c falls to hundreds of nanokelvin, which is why laser cooling and evaporative cooling are needed. The code evaluates the formula for three cases (the rubidium density is an illustrative value in the typical range, not a particular experiment).

Python Bose-Einstein critical temperatures

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.

Condensate fraction and the zeta integral

Two numerical checks make the theory concrete. First, the integral that defines N_max equals Γ(3/2) ζ(3/2) = (√π/2) ζ(3/2). Second, the plot shows how the ground state fills up as the gas is cooled below T_c.

Python The zeta integral and the condensate fraction

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.

For particles in a harmonic trap the density of states differs, so the transition temperature changes to T_c = (ħω̄/k)(N/ζ(3))^(1/3) and the condensate fraction is 1 - (T/T_c)³. This is the version relevant to the 1995 experiments with trapped atoms.

Why bosons and not fermions

Condensation requires an unlimited number of particles in one state. A Fermi gas has no such transition: occupations cannot exceed 1, so the Fermi sea fills smoothly. Photons are bosons too, but their number is not conserved (μ = 0 always), so they simply disappear at low temperature instead of condensing.

Check your understanding

Quick check · 4 questions

of 4 answered You scored / 4 · perfect!

  1. 1. Why must the chemical potential of an ideal Bose gas satisfy μ ≤ 0 (taking ε₀ = 0)?

    If μ exceeded the lowest energy, the Bose-Einstein occupation would be negative or infinite.
  2. 2. At T_c, which condition holds?

    That is when the excited states can no longer hold all N particles.
  3. 3. Below T_c in a uniform ideal Bose gas, the fraction of particles in the ground state is

    The excited states hold a fraction (T/T_c)^(3/2); the rest sit in the ground state.
  4. 4. Why does ideal-gas T_c for liquid He-4 (about 3.1 K) not match the observed 2.17 K exactly?

    Interactions are strong in the liquid, but the same order of magnitude is a sign that the condensation idea is relevant.

Summary

  • The excited states of an ideal 3D Bose gas can hold only N_max = V ζ(3/2)/λ³ particles; the rest condense into the ground state.
  • T_c = (2πħ²/mk)(n/ζ(3/2))^(2/3), and N₀/N = 1 - (T/T_c)^(3/2) below it.
  • Liquid helium gives about 3.1 K in this model (observed lambda point 2.17 K); dilute atomic gases condense at hundreds of nanokelvin.
  • Fermions and photons do not condense in this way.

Next: The Ising Model in One Dimension