Quantum statistics — distinguishable vs indistinguishable particles: Question bank

The reveals use Prompt ::: Reasoning. The answer side always explains, never bare yes/no.

Symbols on this page (read first)

Before any trap, here is every symbol used below, in plain words, so the page stands alone.

Definition The symbol dictionary

$ε$ (epsilon) — the energy of one single-particle level (one "box" an electron/photon/atom can sit in). Bigger $ε$ = higher rung on the energy ladder.
$T$ — absolute temperature of the reservoir the particles talk to (in kelvin).
$k_{B}$ — Boltzmann's constant, the conversion factor between temperature and energy.
$β$ (beta) — shorthand for $β = \frac{1}{k _{B} T}$ , an inverse temperature. Hot $\Rightarrow$ $β$ small; cold $\Rightarrow$ $β$ large. It appears because energies always enter the Boltzmann weight $e^{- β ε}$ .
$μ$ (mu) — the chemical potential, the energy "cost/reward" of adding one more particle to the system. It is the knob that fixes the average number of particles.
$N$ — the total number of particles.
$G$ — the number of boxes (single-particle states available) in a counting problem.
$n$ — the occupation number of one level: how many particles sit in that one box. $⟨ n ⟩$ is its average.
$ψ$ (psi) — the wavefunction; $∣ ψ ∣^{2}$ is the probability density (the only directly measurable thing about position).
$ψ_{-}$ — the antisymmetric combination (flips sign under swap → fermions); $ψ_{+}$ the symmetric one (unchanged under swap → bosons).
$\hat{P}$ — the swap operator that exchanges particle 1 and particle 2.
$1$ — the identity operation (do nothing).
$Z = \sum_{n} e^{- β (ε - μ) n}$ — the grand partition function of one level: the sum, over every allowed occupation $n$ , of the weight $e^{- β (ε - μ) n}$ . Fermions stop at $n = 1$ ; bosons run to infinity.

These are exactly the objects from the grand canonical ensemble — the framework where each level trades particles with a big reservoir at fixed $T$ and $μ$ .

The three occupation formulas (stated in full)

So the traps below have something concrete to point at, here are all three mean-occupation results explicitly, with the mechanistic derivation of why they follow from $Z$ .

Figure — Quantum statistics — distinguishable vs indistinguishable particles

The occupancy curves you will keep referring to are the two quantum boxed distributions above. The figure plots both against energy so you can see the $\leq 1$ ceiling of fermions and the blow-up of bosons near $ε = μ$ .

The symmetry figure above shows the swap $\hat{P}$ : the symmetric state ( $+$ ) is unmoved, the antisymmetric state ( $-$ ) flips sign and dies when the two states coincide (Pauli).

The stars-and-bars figure above is the intuition for the boson count $(2 4) = 6$ used in the first section — read the caption on it before the "True or false" items.

True or false — justify

Indistinguishability means two particles are so alike we merely fail to tell them apart in practice.

False. It is a fundamental absence of any label even in principle — swapping them is a non-operation, so "which is which" has no answer, not just an unknown answer.

Pauli exclusion is a repulsive force acting between fermions.

False. It is the vanishing of the antisymmetric wavefunction

ψ_{-}

when two fermions share a state; no potential energy term is involved, it is a counting constraint. See Pauli Exclusion Principle.

For two particles in three boxes, bosons have fewer microstates than distinguishable particles.

True. Distinguishable gives

3^{2} = 9

(each particle independently picks one of 3 boxes), bosons give

(2 4) = 6

by stars-and-bars; ignoring labels removes the double-counted "swapped" arrangements.

Fermions always have the fewest microstates of the three statistics.

True (when doubling-up is possible). Forbidding same-box configurations removes states the other two keep, so here

3 < 6 < 9

(Fermi

<

Bose

<

distinguishable).

The Bose–Einstein and Fermi–Dirac distributions become identical at very high temperature.

True. When

e^{β (ε - μ)} ≫ 1

the

\pm 1

is tiny next to it, so

\frac{1}{e ^{β (ε - μ)} \pm 1} \approx e^{- β (ε - μ)}

— both collapse to the Maxwell–Boltzmann form.

The chemical potential

μ

for a Bose gas can take any value.

False. Occupancy

⟨ n ⟩_{B E} \geq 0

requires

e^{β (ε - μ)} > 1

at every level, so

μ

must stay below the lowest energy

ε_{min}

; hitting that bound triggers Bose–Einstein Condensation.

Symmetrizing a boson wavefunction produces a genuine attractive potential.

False. There is no potential; symmetrization raises the statistical weight of same-state configurations, an "exchange" effect that only mimics attraction.

The

1/ N!

factor in the classical partition function is an arbitrary fudge.

False. It corrects the exact over-counting from pretending identical particles carry labels — each true microstate was counted

N!

times, once per relabelling. See Gibbs Paradox & Entropy of Mixing.

Half-integer spin implies antisymmetric wavefunctions purely by convention.

False. It is forced by the Spin–Statistics Theorem; half-integer spin must be antisymmetric (fermionic) in any relativistic quantum field theory, not by choice.

The result "swap twice ⇒ phase is

\pm 1

" holds for particles in any number of dimensions.

False. It assumes 3D, where swapping twice is topologically a trivial loop. In strictly 2D, exchange paths can wind, allowing any phase

e^{i ϕ}

— these are anyons, which the

\pm 1

argument silently excludes.

Spot the error

⟨ n ⟩_{F D} = \frac{1}{e ^{β (ε - μ)} - 1}

, so it can exceed 1."

Wrong sign in the denominator. Fermions use

+ 1

, giving

⟨ n ⟩_{F D} \leq 1

; the

- 1

form is Bose–Einstein and is what allows unbounded occupation.

"Two fermions in the same state give

ψ_{-} = \frac{1}{2} [ψ_{a} (1) ψ_{a} (2) + ψ_{a} (2) ψ_{a} (1)]

The antisymmetric combination uses a minus sign, so setting

a = b

gives

ψ_{-} = 0

, not a doubled state — that vanishing is exactly Pauli exclusion.

"Bosons cluster, so their wavefunction must be antisymmetric."

Reversed. Clustering (bunching) comes from the symmetric (

+

) combination; antisymmetric wavefunctions belong to fermions, which avoid the same state.

"For bosons the geometric series

\sum e^{- β (ε - μ) n}

always converges."

Only if

e^{- β (ε - μ)} < 1

, i.e.

ε > μ

. If

μ \geq ε

the series diverges — physically the signal that the level's occupation blows up.

"Distinguishable counting treats 'A in box 1, B in box 2' and 'B in box 1, A in box 2' as the same microstate."

No — distinguishable counting treats them as different, because labels matter. Merging them is precisely the indistinguishable (Bose) count.

"The classical limit keeps the full

\frac{1}{e ^{x} \pm 1}

with no approximation."

No — the classical form drops the

\pm 1

. Expanding

\frac{1}{e ^{x} \pm 1} = e^{- x} (1 \mp e^{- x} + \dots)

, we neglect the higher-order terms

e^{- 2 x}, \dots

because

e^{- x} ≪ 1

when

e^{x} ≫ 1

, leaving only

⟨ n ⟩_{M B} = e^{- x}

"The

+ 1

in Fermi–Dirac comes from spin being

1/2

It comes from the occupation sum stopping at

n = 1

(the two-term series

Z = 1 + e^{- β (ε - μ)}

), which is exclusion; spin enters only via the spin–statistics link that makes the particle a fermion.

"At the classical limit only Maxwell–Boltzmann is valid; Bose and Fermi break down."

All three agree there. The classical result is the shared limit of the quantum ones when levels are rarely doubly occupied, not a rival that survives while others fail.

Why questions

Why does swapping two identical particles have to leave

∣ ψ ∣^{2}

unchanged?

Because

∣ ψ ∣^{2}

is the only measurable prediction about their positions; identical particles must give identical measurable predictions, so the swap cannot alter the density.

Why does the exchange phase have to be exactly

\pm 1

(in 3D)?

\hat{P} ψ = e^{i ϕ} ψ

, applying

\hat{P}

twice gives

e^{2 i ϕ} ψ

; but swapping twice is

\hat{P}^{2} = 1

(do nothing), so

e^{2 i ϕ} = 1 \Rightarrow e^{i ϕ} = \pm 1

. This uses that the double-swap loop is topologically trivial, true in 3D but not 2D (anyons).

Why does the classical

\pm 1

become negligible at high

T

— show the steps?

Write

x = β (ε - μ)

. Hot

T

(small

β

) with

ε > μ

makes

x

large, so

e^{x} ≫ 1

; then

\frac{1}{e ^{x} \pm 1} = \frac{1}{e ^{x} ( 1 \pm e ^{- x} )} = e^{- x} (1 \mp e^{- x} + \dots)

, and we drop the higher-order

e^{- 2 x}

and smaller terms to get

\approx e^{- x}

Why can we derive

⟨ n ⟩

by differentiating

ln Z

instead of summing by hand?

Because

- \partial_{x} ln Z = \frac{1}{Z} \sum_{n} n e^{- x n}

is exactly the weighted-mean occupation; the derivative auto-generates the factor

n

and the division by

Z

, sparing us the explicit sum.

Why does the grand canonical ensemble make deriving these distributions clean?

Levels exchange particles independently with a reservoir, so we can treat one level at a time and let

μ

handle particle number, turning the sum over occupations

Z = \sum_{n} e^{- β (ε - μ) n}

into a short geometric or two-term series. See Grand Canonical Ensemble.

Why does hitting

μ = ε_{min}

signal condensation rather than nonsense?

ε \to μ

the exponent

β (ε - μ) \to 0

, so

⟨ n ⟩_{B E} = \frac{1}{e ^{β (ε - μ)} - 1} \to \infty

, meaning macroscopic numbers pile into the ground state — the defining feature of a condensate.

Why do electrons in an atom fill successive shells instead of all sitting lowest?

Antisymmetry forbids two electrons sharing the same complete state, so once a level (with its spin slots) is full the next electron must occupy a higher one — building shell structure without any repulsive force. Relevant to Fermi Gas & Fermi Energy.

Why does the

1/ N!

patch restore extensive entropy — what does the

N!

count?

N!

is the number of ways to relabel

N

identical particles; the labelled classical count is

N!

times too big, and dividing it out removes a

ln N!

term whose absence otherwise makes entropy non-extensive, curing the fake mixing entropy. See Gibbs Paradox & Entropy of Mixing.

Edge cases

What is

⟨ n ⟩_{F D}

exactly at

ε = μ

\frac{1}{e ^{0} + 1} = \frac{1}{2}

, since

β (ε - μ) = 0

. The chemical potential (Fermi level at

T = 0

) is where a fermion state is half-occupied on average.

What happens to the Fermi–Dirac distribution as

T \to 0

(i.e.

β \to \infty

It becomes a step:

⟨ n ⟩ = 1

for

ε < μ

and

0

for

ε > μ

, because

β (ε - μ)

goes to

- \infty

+ \infty

. This defines the Fermi energy.

What does

⟨ n ⟩_{B E}

do as

ε \to μ^{+}

The denominator

e^{β (ε - μ)} - 1 \to 0^{+}

, so

⟨ n ⟩ \to + \infty

— the level can hold unlimited particles, the seed of condensation.

For photons, what is

μ

and why?

μ = 0

, because photon number is not conserved (they are freely created and absorbed), so no constraint fixes

N

; this reduces Bose–Einstein to the Planck form. See Blackbody Radiation (Photon Gas).

If two particles are placed in the same box, which statistics still allow it?

Distinguishable and Bose–Einstein both allow double occupancy; only Fermi–Dirac forbids it, because

ψ_{-}

vanishes when the states coincide.

In the dilute high-temperature limit, why does distinguishable-vs-not stop mattering?

Occupancy per level is tiny, so configurations with two particles in one state are negligibly rare — the very cases where the statistics differ almost never occur, so all three coincide with Maxwell–Boltzmann.

For a single particle (

N = 1

), do the three statistics differ?

No. With one particle there is no pair to swap and no shared-state question, so labelling, symmetrization, and exclusion all give identical counting.