This page is a shooting range : every kind of question the Bose-Einstein parent note can throw at you, worked from zero. Before the examples, we lay out a matrix of every case class so you can see there are no gaps.
Everything below uses one formula, the Bose-Einstein mean occupancy:
n ˉ ( ε ) = e ( ε − μ ) / k B T − 1 1 .
Let us first name every symbol so nothing is used before it is earned.
Definition The symbols, in plain words
ε — the energy of one quantum state (think: the price of a seat ). Units: joules (J) or, conveniently, in units of k B T .
μ — the chemical potential (the discount ): how much energy the reservoir gives up to hand you one particle. See Chemical potential .
T — absolute temperature in kelvin (K); k B — Boltzmann's constant 1.381 × 1 0 − 23 J/K.
k B T — the thermal energy scale , the natural yardstick against which we measure ε − μ .
We abbreviate the single dimensionless number that runs the whole show:
y ≡ k B T ε − μ .
So n ˉ = e y − 1 1 . Everything below is just: pick y , compute n ˉ .
y controls everything
The formula never asks for ε , μ , T separately — only their combination y = ( ε − μ ) / k B T . It is the seat price measured in thermal units, after the discount . Small y (cheap net seat) → huge crowd; large y (expensive) → almost empty. The single curve below is the entire physics.
Look at the curve: it dives to infinity as y → 0 + (left, amber), and decays like e − y for large y (right, where it kisses the classical dashed line). Every example is a labelled dot on this curve.
Every question about n ˉ falls into exactly one of these case classes . The examples that follow each cover one (or more) cells.
Cell
Case class
Defining condition
Covered by
A
Photon gas (no μ )
μ = 0 , so y = ε / k B T
Ex 1, Ex 6
B
Intermediate y (quantum bunching)
y ∼ 0.1 –1
Ex 2
C
Classical tail (dilute limit)
y ≫ 1
Ex 3
D
Degenerate: y → 0 + (near-condensation)
μ → ε m i n −
Ex 4
E
Forbidden input (y < 0 )
μ > ε → negative n ˉ
Ex 5
F
Real-world word problem
back out T or ε from data
Ex 6
G
Exam twist : ratio / BE-vs-FD
compare two statistics at same y
Ex 7
H
Temperature sweep (limiting behaviour)
fixed ε − μ , vary T
Ex 8
Recall The whole matrix in one sentence
Question ::: Every BE problem reduces to "what is y = ( ε − μ ) / k B T ?" — then read n ˉ = 1/ ( e y − 1 ) , checking y > 0 (else forbidden) and watching y → 0 (near-BEC) and y → ∞ (classical).
Worked example Ex 1 · A light mode with energy
ε = k B T
Statement. A cavity holds photons at temperature T . Consider the mode whose energy equals the thermal energy, ε = k B T . How many photons, on average, sit in this mode?
Forecast: guess before computing — more than 1, or less than 1? (Photons are social, but this seat is not cheap…)
Step 1 — Set μ = 0 .
Why this step? Photons can be created and destroyed by the cavity walls, so their number is not conserved ; the thermodynamic cost of adding one is zero, giving μ = 0 (see the Planck radiation law derivation). Then y = k B T ε − 0 = k B T ε .
Step 2 — Plug in ε = k B T .
Why this step? We're told the numerator equals the denominator, so y = 1 exactly — no constants needed.
n ˉ = e 1 − 1 1 = 2.71828 − 1 1 = 1.71828 1 .
Step 3 — Evaluate.
n ˉ ≈ 0.582.
Verify: n ˉ < 1 , matching the forecast — a seat priced at exactly one thermal unit is more empty than full. Sanity: multiply back, 0.582 × ( e − 1 ) = 0.582 × 1.718 = 1.000 . ✓ This under-population above k B T is why blackbody intensity falls at high frequency.
Worked example Ex 2 · Bosons pile up near
μ
Statement. A boson level has y = ( ε − μ ) / k B T = 0.1 . Find n ˉ .
Forecast: the net seat is very cheap (y small). Do you expect a handful, or many?
Step 1 — Recognise the small-y cell.
Why this step? y = 0.1 ≪ 1 means we are far left on the s01 curve, in the steep region where n ˉ blows up.
Step 2 — Compute the exponential then subtract one.
Why this step? The whole quantum effect lives in the "− 1 ". Never approximate it away here.
e 0.1 = 1.10517 , e 0.1 − 1 = 0.10517.
Step 3 — Invert.
n ˉ = 0.10517 1 ≈ 9.51.
Verify: For small y , e y − 1 ≈ y , so n ˉ ≈ 1/ y = 1/0.1 = 10 — close to 9.51 , confirming the small-y divergence law n ˉ ≈ k B T / ( ε − μ ) . ✓ The classical guess e − 0.1 = 0.905 is 10× smaller : this factor-of-10 gap is boson bunching , the seed of lasers and Bose-Einstein condensation .
Worked example Ex 3 · High-energy state recovers Maxwell–Boltzmann
Statement. A boson level has y = 5 . Compare n ˉ BE with the classical n ˉ MB = e − y .
Forecast: far to the right on s01 — do BE and classical agree here, or clash?
Step 1 — BE value.
Why this step? Direct substitution; we expect the "− 1 " to barely matter since e 5 is huge.
e 5 = 148.413 , n ˉ BE = 148.413 − 1 1 = 147.413 1 = 0.006784.
Step 2 — Classical value.
Why this step? This is the Maxwell-Boltzmann statistics estimate — the limit BE should approach.
n ˉ MB = e − 5 = 0.006738.
Step 3 — Percent difference.
0.006738 0.006784 − 0.006738 × 100% = 0.68%.
Verify: They agree to under 1% , exactly as the "dilute limit" promises: when ε − μ ≫ k B T the − 1 is negligible and quantum statistics wash out. ✓ Units: n ˉ is a pure count (dimensionless) in both. ✓
Worked example Ex 4 · Ground state as
μ → ε m i n −
Statement. The ground state sits at ε m i n . Track n ˉ 0 as μ climbs toward it: take y = 0.01 , then y = 0.001 .
Forecast: the discount μ almost equals the seat price. Does the crowd stay finite?
Step 1 — y = 0.01 .
Why this step? We probe the approach to the forbidden point y = 0 from above.
n ˉ 0 = e 0.01 − 1 1 = 0.0100502 1 ≈ 99.5.
Step 2 — y = 0.001 .
Why this step? Halving-and-more of y should roughly multiply the crowd, showing the 1/ y divergence.
n ˉ 0 = e 0.001 − 1 1 = 0.00100050 1 ≈ 999.5.
Step 3 — Read the trend.
Why this step? From Step 1 to Step 2, y shrank × 10 and n ˉ 0 grew × ∼ 10 . So n ˉ 0 → ∞ as y → 0 + .
Verify: The law n ˉ 0 ≈ 1/ y gives 100 and 1000 — matching 99.5 and 999.5 . ✓ This divergence is the mathematical face of Bose-Einstein condensation : at μ = ε m i n the ground-state partition sum 1/ ( 1 − x ) blows up, so you must split the ground state off and treat it alone. μ may approach but never reach ε m i n .
Worked example Ex 5 · What if
μ > ε ? (the trap)
Statement. A student sets μ = 2 k B T for a state at ε = k B T , giving y = − 1 . Compute n ˉ and interpret.
Forecast: what sign will n ˉ come out — and is that physical?
Step 1 — Form y .
y = k B T ε − μ = k B T k B T − 2 k B T = − 1.
Step 2 — Substitute (blindly, to expose the trap).
Why this step? We plug in to see why negative y is banned, not because it's allowed.
e − 1 = 0.3679 , e − 1 − 1 = − 0.6321 , n ˉ = − 0.6321 1 = − 1.582.
Step 3 — Reject on physical grounds.
Why this step? A count of particles can never be negative. The formula returned n ˉ < 0 , which is nonsense — the flag that μ > ε is forbidden for that level. The condition μ < ε m i n (from the Grand canonical ensemble geometric-sum convergence x < 1 ) must hold for every level.
Verify: − 1.582 × ( e − 1 − 1 ) = − 1.582 × ( − 0.6321 ) = 1.000 — the algebra is self-consistent, so the error is physical , not arithmetic. ✓ Rule locked in: always check y > 0 before trusting n ˉ .
Common mistake Trusting a negative occupancy
Why it feels right: the formula "gave an answer," so it must mean something. Fix: BE is only valid for y > 0 . A negative n ˉ is the theory screaming that you violated μ < ε m i n .
Worked example Ex 6 · Cosmic Microwave Background photon mode
Statement. The CMB fills space at T = 2.725 K. Consider a photon mode of frequency ν = 160.2 GHz (near the CMB peak), energy ε = h ν with h = 6.626 × 1 0 − 34 J·s. Find n ˉ .
Forecast: at the spectral peak , is n ˉ order-1, or tiny like the k B T mode?
Step 1 — Photon ⇒ μ = 0 .
Why this step? Same reason as Ex 1: photon number isn't conserved, μ = 0 , so y = h ν / k B T .
Step 2 — Compute ε = h ν .
ε = 6.626 × 1 0 − 34 × 1.602 × 1 0 11 = 1.0615 × 1 0 − 22 J .
Step 3 — Compute k B T .
k B T = 1.381 × 1 0 − 23 × 2.725 = 3.763 × 1 0 − 23 J .
Step 4 — Form y and evaluate.
Why this step? y must be dimensionless — joules cancel joules, a units check built in.
y = 3.763 × 1 0 − 23 1.0615 × 1 0 − 22 = 2.821 , n ˉ = e 2.821 − 1 1 = 16.79 − 1 1 = 15.79 1 ≈ 0.0633.
Verify: y ≈ 2.82 is close to the Wien-peak value ≈ 2.82 for the radiance peak of a blackbody — a lovely consistency with the Planck radiation law . ✓ Units cancelled to a pure number. ✓ n ˉ ≈ 0.063 : even at its peak each CMB mode is sparsely filled.
Worked example Ex 7 · Same
y , three statistics
Statement. At y = 0.5 , compute n ˉ for Bose-Einstein (− 1 ), Fermi-Dirac (+ 1 ), and classical (e − y ). Rank them.
Forecast: which is biggest — the social boson, the antisocial fermion, or the classical middle?
Step 1 — Common exponential.
Why this step? All three share e 0.5 = 1.64872 ; compute once, reuse.
Step 2 — BE.
n ˉ BE = 1.64872 − 1 1 = 0.64872 1 = 1.5415.
Step 3 — FD (see Fermi-Dirac statistics — fermions ).
Why this step? Only the sign flips — the "+ 1 " caps occupancy below 1.
n ˉ FD = 1.64872 + 1 1 = 2.64872 1 = 0.3775.
Step 4 — Classical.
n ˉ MB = e − 0.5 = 0.60653.
Step 5 — Rank. n ˉ BE ( 1.54 ) > n ˉ MB ( 0.61 ) > n ˉ FD ( 0.38 ) .
Verify: Bosons bunch above classical, fermions anti-bunch below it — the classical value always sits between the two, exactly as theory demands. ✓ Note n ˉ FD < 1 always (Pauli cap), while n ˉ BE has no ceiling.
The bar chart makes the ordering visual: cyan boson tower highest, amber fermion lowest, white classical between.
Worked example Ex 8 · Fix
ε − μ , vary T
Statement. A level has fixed ε − μ = Δ . Find n ˉ at (a) k B T = Δ (so y = 1 ), (b) k B T = 10Δ (hot, y = 0.1 ), (c) k B T = Δ/5 (cold, y = 5 ). What happens as T → 0 and T → ∞ ?
Forecast: heating a boson gas — does each level's crowd grow or shrink?
Step 1 — (a) y = 1 .
Why this step? Baseline, same as Ex 1's arithmetic.
n ˉ = e 1 − 1 1 = 0.5820.
Step 2 — (b) hot, y = 0.1 .
Why this step? Raising T shrinks y (thermal yardstick grows), pushing us left on s01 → crowd grows.
n ˉ = e 0.1 − 1 1 = 9.508.
Step 3 — (c) cold, y = 5 .
Why this step? Lowering T grows y → right side of s01 → crowd empties out.
n ˉ = e 5 − 1 1 = 0.006784.
Step 4 — Read the limits.
Why this step? Extrapolate the trend to the endpoints.
T → ∞ : y → 0 + , n ˉ → ∞ (levels flood).
T → 0 : y → ∞ , n ˉ → 0 (excited levels empty; particles fall to ground state → Bose-Einstein condensation ).
Verify: From (c)→(a)→(b), heating (T up) monotonically raised n ˉ from 0.0068 → 0.582 → 9.51 . ✓ The two limits bracket the physics: hot = classical flood, cold = quantum ground-state collapse. Units: y = Δ/ k B T is dimensionless throughout. ✓
Mnemonic One-number workflow
See y , solve y . Compute y = ( ε − μ ) / k B T → check y > 0 → read n ˉ = 1/ ( e y − 1 ) → small y means big crowd (near BEC), big y means classical. B oson = B elow-by-one = the − 1 .