Fermionic Fock space, Pauli exclusion, anticommutators

Intuition Beginner

Electrons in an atom, protons and neutrons in a nucleus, the conduction electrons of a metal, the neutrinos streaming out of a supernova: every one of these is a fermion. The defining feature of a fermion is that you cannot put two of them into the same single-particle state. Wolfgang Pauli stated this rule in 1925 to explain why the periodic table has the shells it has — once a shell is full, the next electron has to occupy a state with different quantum numbers. The rule applies to every fermion, not just electrons, and a quantum-statistical-mechanics framework that ignores it gets the wrong answer for every solid, every nucleus, and every white dwarf.

The mathematical packaging of the rule is the antisymmetric tensor product. Two identical bosons in states $f$ and $g$ live in the symmetric combination of the two pair-orderings, "first particle in $f$, second in $g$" plus "first in $g$, second in $f$", treating the two particles as indistinguishable in a fully symmetric way. Two identical fermions instead live in the antisymmetric combination, "first in $f$, second in $g$" minus "first in $g$, second in $f$". If you set $g=f$, the antisymmetric combination vanishes: the state with two electrons in the same single-particle slot is the zero vector, the unphysical no-state. Pauli exclusion is exactly the vanishing of the antisymmetric pair on a repeated factor.

The fermionic Fock space is the home where this picture lives. Start with the one-particle Hilbert space, build the antisymmetric two-particle, three-particle, and higher tensor powers, and stack them with the vacuum at the bottom. Creation operators add a particle in a chosen state by wedging it onto the front; annihilation operators remove a particle. The algebra they satisfy is the canonical anticommutation relation, the anticommutator-side cousin of the bosonic commutator.

Visual Beginner

Picture a ladder, like the bosonic one, but with a hard stop. Each rung is labelled by the number of fermions present. The bottom rung is the vacuum: no particles. The next rung holds one-particle states, a full copy of the one-particle Hilbert space. The rung above holds two-fermion states, but with the swap convention that exchanging the two particles changes the sign of the state.

The third rung holds three-fermion states with the analogous alternating-sign symmetry, and so on. The hard stop is that the rung labelled $n$ contains the zero vector whenever you try to put two particles in the same single-particle slot: the rung is bounded above in dimension by the one-particle Hilbert space, and saturates at the full one-particle dimension.

Two operators move states between rungs. The creation operator $a^{\ast }(f)$ adds a fermion in the one-particle state $f$ by wedging $f$ in front: $a^{\ast }(f)(f_1\wedge \cdots \wedge f_n)=f\wedge f_1\wedge \cdots \wedge f_n$. The annihilation operator $a(f)$ is its formal adjoint and removes a fermion in state $f$. Their algebra reads $a(f)a^{\ast }(g)+a^{\ast }(g)a(f)=⟨f,g⟩$, the canonical anticommutation relation — the rule that replaces the bosonic commutator's minus sign with a plus sign and is the algebraic content of Pauli exclusion.

Worked example Beginner

Compute the inner product of a two-fermion Slater-determinant state with itself, and verify that two fermions in the same state give the zero vector.

Step 1. Fix two single-particle states $f$ and $g$ in the one-particle Hilbert space. The two-fermion state with one fermion in $f$ and one in $g$ is the wedge product $f\wedge g$, defined as the difference of the two label-orderings: "first particle in $f$, second in $g$" minus "first in $g$, second in $f$". The minus sign records that exchanging the two labels flips the sign of the state.

Step 2. Compute the inner product $⟨f\wedge g,f\wedge g⟩$. Expand each factor as the same difference of two label-orderings, then use the rule that the inner product of "particle one in $a$, particle two in $b$" with "particle one in $c$, particle two in $d$" equals the product $⟨a,c⟩\cdot ⟨b,d⟩$ — single-particle inner products multiplied across the two slots.

Step 3. The four cross terms give: the diagonal-orderings produce $⟨f,f⟩\cdot ⟨g,g⟩=\Vert f{\Vert }^2\Vert g{\Vert }^2$ (twice), and the swapped-orderings produce $⟨f,g⟩\cdot ⟨g,f⟩=|⟨f,g⟩{|}^2$ (twice, with minus signs from the antisymmetry).

Step 4. The combination gives $⟨f\wedge g,f\wedge g⟩=2(\Vert f{\Vert }^2\Vert g{\Vert }^2-|⟨f,g⟩{|}^2)$. The $2$ is a normalisation convention; the parenthetical factor is the meaningful object. (Many texts absorb the $2$ into the definition of the wedge product, so that $f\wedge g=(1/\sqrt{2})$ times the same difference of orderings, and the inner product reads $\Vert f{\Vert }^2\Vert g{\Vert }^2-|⟨f,g⟩{|}^2$ with no leading factor. Both conventions appear in the literature.)

Step 5. Plug in two concrete cases. Case A: $f$ and $g$ orthonormal, so $\Vert f\Vert =\Vert g\Vert =1$ and $⟨f,g⟩=0$. Then $\Vert f{\Vert }^2\Vert g{\Vert }^2-|⟨f,g⟩{|}^2=1-0=1$. The state is normalised (with the convention that absorbs the $2$). Case B: $f=g$. Then $\Vert f{\Vert }^2\Vert g{\Vert }^2-|⟨f,g⟩{|}^2=\Vert f{\Vert }^4-\Vert f{\Vert }^4=0$. The state $f\wedge f$ has norm zero — it is the zero vector. Two fermions cannot occupy the same one-particle state.

What this tells us: Pauli exclusion is not a separate axiom bolted on top of the Hilbert-space picture. It is a consequence of antisymmetry: the wedge product of any vector with itself vanishes, so the state "two fermions in the state $f$" is literally the zero vector of the two-particle antisymmetric space. The Slater-determinant inner product $\Vert f{\Vert }^2\Vert g{\Vert }^2-|⟨f,g⟩{|}^2$ records how much of the two-fermion state is left after antisymmetrisation: full if $f,g$ are orthogonal, zero if $f=g$, and interpolating linearly in the squared overlap in between.

Check your understanding Beginner

Formal definition Intermediate+

Let $\mathcal{H}$ be a complex separable Hilbert space — the one-particle Hilbert space of the theory. The non-relativistic spin-$1/2$ standard case is $\mathcal{H}=L^2({\mathbb{R}}^3)\otimes {\mathbb{C}}^2$, with the ${\mathbb{C}}^2$ factor carrying the two spin states. For each integer $n\ge 0$ write ${\mathcal{H}}^{\otimes n}$ for the Hilbert tensor power with the convention ${\mathcal{H}}^{\otimes 0}:=\mathbb{C}$. The symmetric group $S_n$ acts on ${\mathcal{H}}^{\otimes n}$ by permutation of tensor factors, with each permutation $\sigma \in S_n$ represented by a unitary $U_{\sigma }$.

The antisymmetrisation projector is $$ P_a^{(n)} := \mathrm{Alt}^{(n)} := \frac{1}{n!} \sum_{\sigma \in S_n} \mathrm{sgn}(\sigma), U_\sigma, $$ and the antisymmetric $n$-fold tensor power is the closed subspace $$ \Lambda^n \mathcal{H} := P_a^{(n)} \mathcal{H}^{\otimes n} = { \xi \in \mathcal{H}^{\otimes n} : U_\sigma \xi = \mathrm{sgn}(\sigma) \xi \text{ for all } \sigma \in S_n }. $$ The dimension of ${\Lambda }^n\mathcal{H}$ over a finite-dimensional one-particle space $\mathcal{H}$ with $\dim \mathcal{H}=d$ is $\left(\genfrac{}{}{0ex}{}{d}{n}\right)$; in particular ${\Lambda }^n\mathcal{H}=0$ for $n>d$ — the saturation that has no bosonic analogue and is the dimensional form of Pauli exclusion.

The fermionic Fock space over $\mathcal{H}$ is the Hilbert direct sum $$ \mathcal{F}-(\mathcal{H}) := \bigoplus{n=0}^{\infty} \Lambda^n \mathcal{H}, $$ with inner product $⟨({\xi }_n{)}_n,({\eta }_n{)}_n⟩={\sum }_{n\ge 0}⟨{\xi }_n,{\eta }_n{⟩}_{{\Lambda }^n\mathcal{H}}$ and norm $\Vert ({\xi }_n{)}_n{\Vert }^2={\sum }_n\Vert {\xi }_n{\Vert }^2<\infty$. The summand ${\Lambda }^0\mathcal{H}=\mathbb{C}$ is spanned by the unit vector $\Omega$ called the vacuum.

For $f_1,\dots ,f_n\in \mathcal{H}$ write the wedge product $f_1\wedge \cdots \wedge f_n:=\sqrt{n!}{P}_a^{(n)}(f_1\otimes \cdots \otimes f_n)$. The antisymmetric inner product is computed by the determinant of the Gram matrix, $$ \langle f_1 \wedge \cdots \wedge f_n,\ g_1 \wedge \cdots \wedge g_n \rangle = \det\bigl( \langle f_i, g_j \rangle \bigr). $$ This is the Slater-determinant formula. It is the antisymmetric analogue of the bosonic permanent formula; the determinant's vanishing on linearly dependent rows is the algebraic packaging of Pauli exclusion.

The creation operator $a^{\ast }(f)$ for $f\in \mathcal{H}$ is the bounded operator on ${\mathcal{F}}_{-}(\mathcal{H})$ with $a^{\ast }(f):{\Lambda }^n\mathcal{H}\to {\Lambda }^{n+1}\mathcal{H}$ acting by $$ a^(f)(g_1 \wedge \cdots \wedge g_n) = f \wedge g_1 \wedge \cdots \wedge g_n, \qquad a^(f) \Omega = f. $$ The annihilation operator $a(f)$ is the formal adjoint, with $a(f):{\Lambda }^n\mathcal{H}\to {\Lambda }^{n-1}\mathcal{H}$ acting by $$ a(f)(g_1 \wedge \cdots \wedge g_n) = \sum_{k=1}^{n} (-1)^{k-1} \langle f, g_k \rangle, g_1 \wedge \cdots \wedge \widehat{g_k} \wedge \cdots \wedge g_n, \qquad a(f) \Omega = 0, $$ where the caret marks omission. The map $f\mapsto a^{\ast }(f)$ is complex-linear; the map $f\mapsto a(f)$ is complex anti-linear with $a(\lambda f)=\bar{\lambda }a(f)$. The sign $(-1{)}^{k-1}$ records that pulling the $k$-th slot to the front through $k-1$ neighbours costs that many sign-flips under antisymmetry.

A central feature distinguishing the fermionic from the bosonic case: $a^{\ast }(f)$ and $a(f)$ are bounded operators on ${\mathcal{F}}_{-}(\mathcal{H})$, with operator norm $$ |a^(f)| = |a(f)| = |f|. $$ The bosonic creation operator has $|a^|n\rangle| = \sqrt{n+1},|f|$ growing without bound; the fermionic one is bounded because the wedge with a repeated factor vanishes and the only nonzero contributions are constrained by antisymmetry.

The number operator acts on ${\Lambda }^n\mathcal{H}$ by multiplication by $n$. Given an orthonormal basis $\{e_k{\}}_{k\ge 1}$ of $\mathcal{H}$, $$ N = \sum_{k=1}^{\infty} a^*(e_k), a(e_k), $$ with the sum convergent on the finite-particle subspace and extending to a self-adjoint operator on ${\mathcal{F}}_{-}(\mathcal{H})$ with spectrum $\sigma (N)={\mathbb{N}}_0$. Over a finite-dimensional $\mathcal{H}$ with $\dim \mathcal{H}=d$, the spectrum truncates at $\{0,1,\dots ,d\}$. The vacuum is the unique unit vector with $N\Omega =0$.

The collection $\{a(f),a^{\ast }(f):f\in \mathcal{H}\}$ generates a $C^{\ast }$-algebra called the CAR algebra ${\mathfrak{A}}_{\mathrm{CAR}}(\mathcal{H})$. Boundedness of the generators makes the CAR algebra a genuine $C^{\ast }$-algebra (no domain issues), in contrast to the bosonic CCR which only generates a $C^{\ast }$-algebra after exponentiation to the Weyl form.

Counterexamples to common slips

• Two fermions in the same one-particle state give the zero vector, not a unit vector with an extra minus sign. The wedge $f\wedge f$ vanishes; there is no two-fermion state at all in that slot.

• The bosonic creation operator is unbounded ($\Vert a^{\ast }|n⟩\Vert =\sqrt{n+1}$). The fermionic creation operator is bounded with norm $\Vert f\Vert$. Pauli exclusion is exactly what makes the operator-norm bound finite: the number operator $N$ acts on ${\Lambda }^n$ by $n$, and the sign-cancellations of the wedge keep the creation/annihilation growth bounded mode by mode.

• The CAR $\{a(f),a^{\ast }(g)\}=⟨f,g⟩$ is an anticommutator, not a commutator. Confusing the two leads to fermions with Bose-Einstein statistics, an experimentally falsified theory.

• The Jordan-Wigner transformation gives a basis-level isomorphism between fermionic operators on a one-dimensional lattice and Pauli matrices on the same lattice — but the isomorphism is nonlocal (each fermionic operator becomes a string of ${\sigma }^z$'s), and the locality structure of the fermionic theory does not match that of the spin theory in higher dimensions.

• The Bogoliubov-Valatin transformation ${\alpha }_k=u_k{a}_k+v_k{a}_{-k}^{\ast }$ preserves the CAR only when $|u_k{|}^2+|v_k{|}^2=1$. Dropping the unit-norm condition gives operators that no longer satisfy the CAR and no longer represent fermionic quasiparticles.

Key theorem with proof Intermediate+

Theorem (canonical anticommutation relations on fermionic Fock space; Jordan-Wigner 1928 ^{[Jordan-Wigner 1928]}, rigorous form Cook 1953 ^{[Cook 1953]}). Let $\mathcal{H}$ be a complex separable Hilbert space. The creation and annihilation operators $a^(f), a(g)$onthefermionicFockspace$\mathcal{F}-(\mathcal{H})$ satisfy* $$ {a(f),, a^(g)} = \langle f, g \rangle, \mathbf{1}, \qquad {a(f),, a(g)} = 0, \qquad {a^(f),, a^*(g)} = 0, $$ *on $\mathcal{F}-(\mathcal{H})$,where${A, B} := A B + B A$istheanticommutatorand$\mathbf{1}$istheidentity.Inparticular$a^(f)^2 = 0$(Pauliexclusion).Thenumberoperator$N = \sum_k a^(e_k), a(e_k)$foranyorthonormalbasis${e_k}$of$\mathcal{H}$isindependentofthebasisandextendstoaself-adjointoperatorwithspectrum$\sigma(N) = \mathbb{N}_0$andeigenspace$N^{-1}({n}) = \Lambda^n \mathcal{H}$.TheoperatorssatisfytheCookbound$|a^(f)| = |a(f)| = |f|$.

Proof. Fix $f,g\in \mathcal{H}$ and $\xi =g_1\wedge \cdots \wedge g_n\in {\Lambda }^n\mathcal{H}$. Apply $a^{\ast }(g)$ first: $$ a^(g) \xi = g \wedge g_1 \wedge \cdots \wedge g_n \in \Lambda^{n+1} \mathcal{H}. $$ Now apply $a(f)$. By the definition of annihilation as a sign-weighted sum over removals, $$ a(f) a^(g) \xi = \langle f, g \rangle, g_1 \wedge \cdots \wedge g_n - \sum_{k=1}^{n} (-1)^{k-1} \langle f, g_k \rangle, g \wedge g_1 \wedge \cdots \wedge \widehat{g_k} \wedge \cdots \wedge g_n. $$ The first term comes from removing the $g$ that was just inserted at slot one (sign $(-1{)}^{1-1}=1$); the second sum comes from removing each original $g_k$, with the sign $(-1{)}^{(k+1)-1}=(-1{)}^k$ recording that $g_k$ now sits at position $k+1$ in the $(n+1)$-tuple, and pulling the rescued $g$ back to its slot picks up an additional sign-flip per neighbour. The bookkeeping reduces to the leading minus sign written above.

Apply the operators in the opposite order: $$ a(f) \xi = \sum_{k=1}^{n} (-1)^{k-1} \langle f, g_k \rangle, g_1 \wedge \cdots \wedge \widehat{g_k} \wedge \cdots \wedge g_n, $$ and then $a^{\ast }(g)$ inserts $g$ at the front of each term: $$ a^*(g), a(f) \xi = \sum_{k=1}^{n} (-1)^{k-1} \langle f, g_k \rangle, g \wedge g_1 \wedge \cdots \wedge \widehat{g_k} \wedge \cdots \wedge g_n. $$

Add the two expressions: $$ a(f) a^(g) \xi + a^(g) a(f) \xi = \langle f, g \rangle, \xi + \sum_{k=1}^{n} [(-1)^{k-1} - (-1)^{k-1}] \langle f, g_k \rangle, (\cdots) = \langle f, g \rangle, \xi. $$ The first sum's coefficient is $-(-1{)}^{k-1}$ from $a{a}^{\ast }$ and $+(-1{)}^{k-1}$ from $a^{\ast }a$; the two cancel exactly. This is the heart of the calculation: the sign-flips that the wedge introduces work to cancel the "exchange" terms, leaving only the contraction $⟨f,g⟩$. The identity $\{a(f),a^{\ast }(g)\}=⟨f,g⟩1$ holds on the dense subspace of finite-particle wedge products, and by linearity and continuity (boundedness, proved below) on the whole Fock space.

For $\{a^{\ast }(f),a^{\ast }(g)\}=0$: $a^{\ast }(f)a^{\ast }(g)\xi =f\wedge g\wedge g_1\wedge \cdots \wedge g_n$, and $a^{\ast }(g)a^{\ast }(f)\xi =g\wedge f\wedge g_1\wedge \cdots \wedge g_n=-f\wedge g\wedge g_1\wedge \cdots \wedge g_n$ by antisymmetry of the wedge in the first two slots. The sum vanishes. As an immediate corollary, $a^{\ast }(f{)}^2=\frac{1}{2}\{a^{\ast }(f),a^{\ast }(f)\}=0$ — the operator form of Pauli exclusion.

For $\{a(f),a(g)\}=0$: each annihilation produces a sign-weighted sum over removals. The double sum from $a(f)a(g)\xi$ runs over ordered pairs $(j,k)$ with $j\ne k$; the same double sum from $a(g)a(f)\xi$ runs over the same ordered pairs with the roles of $f$ and $g$ exchanged. A careful sign accounting (pulling out two factors from the wedge picks up the signature of the corresponding $2$-cycle, which is $-1$) shows the two double sums cancel pairwise. Taking adjoints, $\{a(f),a(g){\}}^{\ast }=\{a^{\ast }(g),a^{\ast }(f)\}=0$, recovering the second CAR.

For the number operator: choose an orthonormal basis $\{e_k\}$ of $\mathcal{H}$ and consider $N={\sum }_k{a}^{\ast }(e_k)a(e_k)$. On the wedge $e_{k_1}\wedge \cdots \wedge e_{k_n}$ with distinct $k_1,\dots ,k_n$ (only such wedges are nonzero by Pauli exclusion), the operator $a^{\ast }(e_k)a(e_k)$ acts as the indicator that $e_k$ appears among the factors. Summing over $k$ counts the total number of distinct factors, which is $n$. So $N$ acts on ${\Lambda }^n\mathcal{H}$ by multiplication by $n$, basis-independent. Self-adjointness follows from the orthogonal direct sum decomposition ${\mathcal{F}}_{-}(\mathcal{H})={⨁}_n{\Lambda }^n\mathcal{H}$, with $N$ diagonal with real eigenvalues $n$. The spectrum is pure point and equals ${\mathbb{N}}_0$.

For the Cook bound $\Vert a^{\ast }(f)\Vert =\Vert f\Vert$: for $\xi =({\xi }_n{)}_n\in {\mathcal{F}}_{-}(\mathcal{H})$, the CAR identity $a^{\ast }(f)a(f)+a(f)a^{\ast }(f)=\Vert f{\Vert }^{2}1$ (the diagonal case $g=f$) gives $\Vert a(f)\xi {\Vert }^2+\Vert a^{\ast }(f)\xi {\Vert }^2=\Vert f{\Vert }^2\Vert \xi {\Vert }^2$. Since both terms on the left are non-negative, each is bounded by $\Vert f{\Vert }^2\Vert \xi {\Vert }^2$, hence $\Vert a^{\ast }(f)\Vert \le \Vert f\Vert$ and $\Vert a(f)\Vert \le \Vert f\Vert$. The opposite inequality $\Vert a^{\ast }(f)\Omega \Vert =\Vert f\Vert$ on the vacuum saturates the bound, giving $\Vert a^{\ast }(f)\Vert =\Vert f\Vert$. Adjoint then gives $\Vert a(f)\Vert =\Vert f\Vert$. $\square$

Bridge. This fermionic CAR algebra builds toward every stat-mech / QFT application of fermions that follows, and appears again in the Dirac field construction of 08.10.07 where microcausality plus positive energy forces anticommutators on half-integer-spin fields. The foundational reason the construction works is exactly the antisymmetric tensor power: ${\Lambda }^n\mathcal{H}$ is the space of indistinguishable $n$-fermion states with built-in Pauli exclusion, and the creation operator lifts an $n$-state to an $(n+1)$-state by wedge adjunction at the front. The central insight is that the many-body Hilbert space and many-body free dynamics are determined by the one-particle data alone — the procedure $\mathcal{H}\mapsto {\mathcal{F}}_{-}(\mathcal{H})$ and $h\mapsto d\Gamma (h)$ is functorial in the unitary group of $\mathcal{H}$, exactly as in the bosonic case of 08.10.01. Putting these together, second quantisation generalises the Schrödinger picture of one fermion to the same picture for indefinitely many indistinguishable fermions, and this picture is dual to the Grassmann-Berezin path-integral developed in 08.07.01 under the Wick rotation of 08.09.01. The bridge is that the Berezin Gaussian measure of 08.07.01 reproduces the same generating functional of correlation functions as the vacuum expectation of normal-ordered Fermi-field products on ${\mathcal{F}}_{-}(\mathcal{H})$ — operator-side and Grassmann-measure-side computations of the same physics. This is exactly the fermionic counterpart of the bosonic Gaussian-Fock correspondence, with antisymmetry replacing symmetry throughout. The bridge identifies fermionic Fock-space CAR with the bosonic Fock-space CCR of 08.10.01 under the spin-statistics correspondence, and identifies the fermionic Berezin integral with the bosonic Gaussian integral under a parallel duality.

Exercises Intermediate+

Advanced results Master

Theorem (Cook bound for fermionic creation operators; Reed-Simon Vol. II §X.7 ^{[Reed-Simon Vol. II]}). For every $f\in \mathcal{H}$ and every $\xi \in {\mathcal{F}}_{-}(\mathcal{H})$, $$ |a^(f) \xi| = |a(f) \xi| \leq |f| \cdot |\xi|, $$ and the bound is sharp: $|a^(f) \Omega| = |f|$.Consequently$a(f), a^(f) \in \mathfrak{B}(\mathcal{F}-(\mathcal{H}))$areboundedoperatorswithoperatornormexactly$|f|$,andtheygeneratea$C^*$-algebracalledtheCARalgebra$\mathfrak{A}{\mathrm{CAR}}(\mathcal{H})$.*

The boundedness of fermionic creation/annihilation operators is the technical hallmark distinguishing the CAR from the CCR: the bosonic creation operator has $\Vert a^{\ast }|n⟩\Vert =\sqrt{n+1}\Vert f\Vert$ growing unboundedly with the particle number, whereas the fermionic operator is bounded because the antisymmetric tensor power saturates and the only nonzero contributions are those with no repeated factor. The Cook bound is what makes the CAR algebra a genuine $C^{\ast }$-algebra without any further smearing or exponentiation, in contrast to the CCR which requires the Weyl form $W(f)=e^{i\varphi (f)}$ to become a bounded $C^{\ast }$-algebra (Bratteli-Robinson Vol. II §5.2 ^{[Bratteli-Robinson Vol. II]}).

Theorem (second quantisation of one-particle dynamics). Let $h$ be a self-adjoint operator on $\mathcal{H}$. The many-body free Hamiltonian $H_0:=d\Gamma (h)$ is self-adjoint on ${\mathcal{F}}_{-}(\mathcal{H})$ with the finite-particle subspace as a core. The unitary group $\Gamma (e^{-ith})$ generated by $H_0$ acts on creation operators by Heisenberg evolution: for every $f\in \mathcal{H}$ and $t\in \mathbb{R}$, $$ e^{i t H_0}, a^(f), e^{-i t H_0} = a^(e^{i t h} f). $$ The corresponding identity holds for $a(f)$ with $f$ replaced by $e^{-ith}f$ on the right.

This is the fermionic counterpart of the bosonic functorial second quantisation. The functor $\Gamma$ acts on ${\Lambda }^n\mathcal{H}$ as $U^{\otimes n}$, preserving the antisymmetric subspace because $U^{\otimes n}$ commutes with the antisymmetrisation projector $P_a^{(n)}$ (permutations and unitary actions of the one-particle space commute). The infinitesimal generator $d\Gamma (h)$ is the second-quantised free Hamiltonian, identical in form to the bosonic case.

Theorem (grand-canonical free Fermi gas; Bratteli-Robinson Vol. II §5.2.4 ^{[Bratteli-Robinson Vol. II]}). Let $h$ be self-adjoint on $\mathcal{H}$ with discrete spectrum $\{{ϵ}_k{\}}_{k\ge 1}$ and ${\inf }_k{ϵ}_k>-\infty$. For $\beta >0$ and any real $\mu$, the grand-canonical density matrix $$ \rho_{\beta, \mu} := \frac{1}{\Xi} e^{-\beta(d\Gamma(h) - \mu N)}, \qquad \Xi := \mathrm{tr}{\mathcal{F}-(\mathcal{H})}(e^{-\beta(d\Gamma(h) - \mu N)}) = \prod_k (1 + e^{-\beta(\epsilon_k - \mu)}) $$ is a positive trace-class operator with trace one. The expectation of the number operator is $$ \langle N \rangle_{\beta, \mu} = \sum_k \frac{1}{e^{\beta(\epsilon_k - \mu)} + 1}, $$ the Fermi-Dirac distribution summed over one-particle modes.

The fermionic partition function is a finite product of $(1+e^{-\beta ({ϵ}_k-\mu )})$ factors — two-term sums reflecting the two states (empty or occupied) per mode allowed by Pauli exclusion. The chemical potential is unrestricted in sign because the fermionic occupation is bounded above by $1$; there is no Fermi-side counterpart of Bose-Einstein condensation, since the two-state truncation prevents any single mode from becoming macroscopically occupied. The zero-temperature limit $\beta \to \infty$ at fixed $\mu$ gives the Fermi-sea ground state: every mode with ${ϵ}_k<\mu$ is occupied with probability $1$ and every mode with ${ϵ}_k>\mu$ is empty with probability $1$. The Fermi sea is the Slater determinant of all single-particle states with energy below $\mu$, and is the workhorse of solid-state physics (free electron gas, Fermi liquid theory, Sommerfeld expansion).

Theorem (Jordan-Wigner transformation; Jordan-Wigner 1928 ^{[Jordan-Wigner 1928]}). Let $\Lambda =\{1,2,\dots ,L\}$ be a one-dimensional lattice. Let ${\mathcal{H}}_{\mathrm{spin}}={⨂}_{j\in \Lambda }{\mathbb{C}}^2$ be the spin-$1/2$ chain Hilbert space, with on-site Pauli operators ${\sigma }_j^x,{\sigma }_j^y,{\sigma }_j^z$, and let ${\mathcal{H}}_{\mathrm{ferm}}={\mathcal{F}}_{-}({\mathbb{C}}^L)$ be the fermionic Fock space on $L$ modes. The Jordan-Wigner map $$ a_j := \Bigl( \prod_{k=1}^{j-1} \sigma^z_k \Bigr) \cdot \sigma^-j, \qquad a^*j := \Bigl( \prod{k=1}^{j-1} \sigma^z_k \Bigr) \cdot \sigma^+j, $$ *with ${\sigma }_j^{\pm }:=({\sigma }_j^x\pm i{\sigma }_j^y)/2$, defines a unitary isomorphism $\mathcal{H}{\mathrm{spin}} \cong \mathcal{H}{\mathrm{ferm}}$underwhichthespin-chainraising/loweringalgebrabecomestheCARalgebraon$\mathbb{C}^L$.InparticularthecanonicalanticommutationrelationsonthefermionicsidecorrespondtothePaulion-siteanticommutationandcross-sitecommutationonthespinside,withthe$\sigma^z$-string converting commutators into anticommutators.*

The Jordan-Wigner transformation is a structural identity: the spin and fermion Hilbert spaces have the same dimension ($2^L$), and the transformation provides an explicit basis-level isomorphism. The transformation is nonlocal — each on-site fermion operator is a string of $L-1$ spin operators — and this is unavoidable: in one spatial dimension, spin operators on different sites commute while fermionic operators anticommute, and the string is what converts one to the other. The transformation is the foundation of one-dimensional integrable-model theory: the transverse-field Ising model on the spin side maps to a free-fermion model on the Jordan-Wigner side, and the spin-spin correlation functions of the Ising model become Toeplitz determinants in the free fermionic theory (Onsager 1944; Lieb-Schultz-Mattis 1961). The transformation does not generalise straightforwardly to two or higher spatial dimensions, where the locality structures of spin and fermion theories are fundamentally different — a sustained line of research from Wen's slave-particle constructions through modern parton gauge theories addresses this gap.

Theorem (Bogoliubov-Valatin transformation; Bogoliubov 1958 ^{[Bogoliubov 1958]}, Valatin 1958 ^{[Valatin 1958]}). Let ${a_k, a^k}{k \in \mathbb{Z}}$bealabelledfamilyoffermioniccreation/annihilationoperatorswith${a_k, a^l} = \delta{kl}$.Foreach$k$choosecomplexcoefficients$u_k, v_k$with$|u_k|^2 + |v_k|^2 = 1$. Define quasiparticle operators $$ \alpha_k := u_k a_k + v_k a^_{-k}, \qquad \alpha^_k := \bar{u}_k a^_k + \bar{v}k a{-k}. $$ Then the ${\alpha_k, \alpha^_k}$satisfythesamecanonicalanticommutationrelations:${\alpha_k, \alpha^l} = \delta{kl}$and${\alpha_k, \alpha_l} = 0$.Themapisa$C^$-automorphismoftheCARalgebra;thecorresponding"quasiparticlevacuum"$|\Omega'\rangle$characterisedby$\alpha_k |\Omega'\rangle = 0$forevery$k$isastateontheoriginalFockspace$\mathcal{F}_-$ifandonlyifthemapisimplementableonthesameHilbertspace,whichisthecaseforfinitelymanymodesalwaysandforinfinitelymanymodesonlywhen$\sum_k |v_k|^2 < \infty$ (the Shale-Stinespring criterion).

The Bogoliubov-Valatin transformation is the foundational tool of mean-field theory for fermions. Bogoliubov's 1958 paper introduced it to diagonalise the BCS Hamiltonian of superconductivity, $H_{\mathrm{BCS}}={\sum }_k{ϵ}_k{a}_k^{\ast }{a}_k+\Delta {\sum }_k(a_k^{\ast }{a}_{-k}^{\ast }+a_{-k}{a}_k)$, with the gap parameter $\Delta$ mixing particle and hole degrees of freedom. The diagonalised form is $H_{\mathrm{BCS}}={\sum }_k{E}_k{\alpha }_k^{\ast }{\alpha }_k+\mathrm{const}$ with quasiparticle energy $E_k=\sqrt{{ϵ}_k^2+|\Delta {|}^2}$, exhibiting the superconducting energy gap. The Shale-Stinespring criterion gives the abstract version: a Bogoliubov transformation is unitarily implementable on the original Fock space exactly when its off-diagonal block is Hilbert-Schmidt, which fails for the BCS ground state in the thermodynamic limit — the BCS vacuum lives in a different superselection sector than the free Fermi vacuum, a manifestation of spontaneous gauge symmetry breaking.

Theorem (spin-statistics; Pauli 1940 ^{[Pauli 1940]}, Burgoyne 1958 ^{[Burgoyne 1958]}). In a relativistic quantum field theory satisfying the Wightman axioms, a field of half-integer spin must satisfy canonical anticommutation relations (CAR), and a field of integer spin must satisfy canonical commutation relations (CCR), at spacelike-separated points. Equivalently: the wrong choice (CCR for spin-$1/2$, or CAR for spin-$0$) is incompatible with at least one of Lorentz covariance, microcausality, or positivity of energy.

The spin-statistics theorem identifies the necessity of the fermionic CAR with the relativistic Wightman framework. The original argument is Pauli's 1940 paper (Phys. Rev. 58, 716): for a Lorentz-covariant local field on Minkowski space, the requirement that the two-point function vanishes at spacelike separation forces the bracketing structure (commutator or anticommutator) to match the spin class — integer spin commutes, half-integer spin anticommutes. Burgoyne's 1958 paper (Nuovo Cim. 8, 607) cleaned up the proof inside the Wightman framework, showing that the wrong choice produces either a vanishing two-point function (theory is empty) or a negative-norm state (theory is non-positive). The full sketch: in a Wightman theory the two-point function $⟨\Omega ,\varphi (x)\varphi (y)\Omega ⟩$ has support on the closed forward light cone in its Fourier transform (spectrum condition). Lorentz covariance and microcausality together force the commutator $[\varphi (x),\varphi (y)]$ (for bosons) or anticommutator $\{\varphi (x),\varphi (y)\}$ (for fermions) to vanish at spacelike separation. The integer-spin two-point function is positive on the diagonal, so the commutator (not anticommutator) is the natural choice; the half-integer-spin two-point function flips sign under a $2\pi$ rotation, forcing the anticommutator (not commutator). The argument is Lorentz-invariance plus microcausality plus positive-energy spectrum together — drop any one and the conclusion can fail.

Theorem (free Dirac field on the fermionic Fock space; Glimm-Jaffe Ch. 6 ^{[Glimm-Jaffe]}). Let $m>0$ and let ${\mathcal{H}}_m^{1/2}:=L^2(X_m,d{\lambda }_m)\otimes {\mathbb{C}}^2$ be the relativistic one-particle Hilbert space of positive-energy mass-$m$ spin-$1/2$ states on the mass shell $X_m\subset {\mathbb{R}}^{1,3}$. The free Dirac field $$ \psi(x) = \sum_s \int \frac{d^3 \mathbf{p}}{(2\pi)^3 (2 E_{\mathbf{p}})^{1/2}} \bigl[ a^s_{\mathbf{p}} u^s(p) e^{-i p \cdot x} + b^{s*}{\mathbf{p}} v^s(p) e^{+i p \cdot x} \bigr] $$ *on $\mathcal{F}-(\mathcal{H}m^{1/2} \oplus \mathcal{H}m^{1/2})$(onecopyforelectrons,oneforpositrons)with${a^s{\mathbf{p}}, a^{r*}{\mathbf{q}}} = (2\pi)^3 \delta^{(3)}(\mathbf{p} - \mathbf{q}) \delta^{sr}$andsimilarlyfor$b$,isaLorentz-covariantoperator-valuedtempereddistributionsatisfyingmicrocausality${\psi(x), \bar\psi(y)} = 0$for$x - y$spacelike,theDiracequation$(i \gamma^\mu \partial_\mu - m) \psi = 0$, and positivity of the normal-ordered Hamiltonian.*

This is the fermionic Fock-space counterpart of the bosonic Klein-Gordon construction in 08.10.01. The two-copy structure (one for particles, one for antiparticles) is forced by Lorentz covariance plus positive energy: a half-integer-spin field cannot be quantised with only positive-energy modes if its anti-particle is distinct from the particle. The microcausality of the anticommutator (not the commutator) is the operational statement of spin-statistics. The total Fock space of the free Dirac field decomposes mode by mode as ${⨂}_{\mathbf{k},s}{\mathbb{C}}^2$ for each one-particle mode labelled by momentum and spin — exactly the dimensionally truncated structure that Pauli exclusion enforces.

Synthesis. The fermionic Fock space packages the antisymmetric tensor algebra into a functor $\mathcal{H}\mapsto {\mathcal{F}}_{-}(\mathcal{H})$ that lifts one-particle quantum mechanics to many-body quantum statistical mechanics and to relativistic quantum field theory for spin-$1/2$ matter, and this is exactly the foundational reason every free-fermion calculation that follows can be done. The central insight is that creation and annihilation operators on the antisymmetric tensor power satisfy the canonical anticommutation relations $\{a(f),a^{\ast }(g)\}=⟨f,g⟩$, the algebraic content of fermionic antisymmetry, with the Pauli-exclusion identity $a^{\ast }(f{)}^2=0$ as the diagonal special case. Putting these together, second quantisation generalises the Schrödinger picture of one fermion to the many-body picture of indefinitely many indistinguishable fermions, and it is dual to the Berezin-Grassmann path-integral picture: the vacuum two-point function on ${\mathcal{F}}_{-}({\mathcal{H}}_m^{1/2})$ Wick-rotates to the Euclidean Berezin Gaussian moment, and the Fock-space normal-ordered moments coincide with the Grassmann-integral moments via the fermionic counterpart of Isserlis-Wick. The bridge is that operator-side and Berezin-measure-side computations agree on every observable a free fermion theory exposes: the Fermi-Dirac distribution comes out of the trace formula and out of the saddle-point of the Grassmann functional; the Dirac mode expansion comes out of the Fock-space construction and out of the spectral resolution of the Euclidean two-point function; the partition function comes out of the two-term sum per mode and out of the Berezin Gaussian integral, $\det (\beta (h-\mu ))$ in place of the bosonic $1/\det$.

This functorial picture identifies fermionic and bosonic Fock space as the two faces of one second-quantisation framework, with the central insight that the spin-statistics correspondence forces the bracketing choice on every Lorentz-covariant local quantum field. The Pauli-exclusion bound $\dim {\Lambda }^n\mathcal{H}=\left(\genfrac{}{}{0ex}{}{d}{n}\right)$ for $\dim \mathcal{H}=d$ generalises in two directions. On the algebraic side, the Jordan-Wigner transformation identifies the CAR algebra on a one-dimensional lattice of $L$ modes with the Pauli spin-$1/2$ algebra on the same lattice, a $2^L$-dimensional Hilbert space carrying both descriptions; the Lieb-Schultz-Mattis 1961 solution of the one-dimensional Ising model is the operational consequence. The bridge is dual to the spin/fermion duality in higher dimensions through parton constructions and lattice gauge theory, where the locality mismatch between spin and fermion theories becomes a dynamical gauge field rather than a basis-level string.

On the mean-field side, the Bogoliubov-Valatin transformation ${\alpha }_k=u_k{a}_k+v_k{a}_{-k}^{\ast }$ rotates the CAR algebra into itself, with the unit-norm condition $|u_k{|}^2+|v_k{|}^2=1$ preserving the algebraic structure; the BCS ground state of superconductivity is the quasiparticle vacuum of one such rotation, and the Shale-Stinespring criterion identifies the precise sense in which it lives in a different superselection sector than the original Fock vacuum. The bridge is that spontaneous symmetry breaking, the BCS gap, and the appearance of quasiparticles are all faces of one Bogoliubov-Valatin rotation, and that the fermionic CAR is the substrate on which the rotation acts. This pattern recurs throughout fermionic condensed matter, fermionic QFT, and fermionic statistical mechanics: the algebra is fixed, the representation moves.

Full proof set Master

Proposition (factorisation of the fermionic Fock space over orthogonal direct sums). For Hilbert spaces ${\mathcal{H}}_1,{\mathcal{H}}_2$ with $\mathcal{H}={\mathcal{H}}_1\oplus {\mathcal{H}}_2$, there is a canonical unitary isomorphism $$ \mathcal{F}_-(\mathcal{H}_1 \oplus \mathcal{H}2) \cong \mathcal{F}-(\mathcal{H}1) \widehat{\otimes} \mathcal{F}-(\mathcal{H}2), $$ where $\hat{\otimes }$ is the ${\mathbb{Z}}_2$-graded tensor product (with sign rule $(\xi \otimes \eta )({\xi }^{\prime }\otimes {\eta }^{\prime })=(-1{)}^{|\eta ||{\xi }^{\prime }|}(\xi {\xi }^{\prime })\otimes (\eta {\eta }^{\prime })$ on homogeneous elements). Under this isomorphism, $a(f_1\oplus f_2)=a(f_1)\otimes 1+(-1{)}^{N_1}\otimes a(f_2)$ with $(-1{)}^{N_1}$ the parity operator on ${\mathcal{F}}_{-}({\mathcal{H}}_1)$; similarly for $a^$.Thevacuumfactorisesas$\Omega_{\mathcal{H}} = \Omega{\mathcal{H}1} \otimes \Omega{\mathcal{H}_2}$.*

Proof. The antisymmetric tensor power factors via the antisymmetric Künneth formula: for $n\ge 0$, $$ \Lambda^n(\mathcal{H}_1 \oplus \mathcal{H}2) \cong \bigoplus{n_1 + n_2 = n} \Lambda^{n_1} \mathcal{H}_1 \otimes \Lambda^{n_2} \mathcal{H}_2. $$ This is the antisymmetric counterpart of the bosonic symmetric Künneth. Summing over $n$ produces the displayed graded-tensor-product structure. The action of $a(f_1\oplus f_2)$ on ${\xi }_1\wedge \cdots \wedge {\xi }_n$ decomposes by which component each ${\xi }_i$ comes from; the parity operator $(-1{)}^{N_1}$ encodes the sign cost of pulling a removal from the ${\mathcal{H}}_2$-component through the ${\mathcal{H}}_1$-component of the wedge. The factorisation of the vacuum is immediate from ${\Lambda }^0=\mathbb{C}$. $\square$

Proposition (Slater-determinant inner product formula). For $f_1,\dots ,f_n,g_1,\dots ,g_n\in \mathcal{H}$, $$ \langle f_1 \wedge \cdots \wedge f_n,\ g_1 \wedge \cdots \wedge g_n \rangle = \det\bigl( \langle f_i, g_j \rangle_{i, j = 1}^{n} \bigr). $$

Proof. Expand both wedge products using the definition $f_1\wedge \cdots \wedge f_n=\sqrt{n!}{P}_a^{(n)}(f_1\otimes \cdots \otimes f_n)=(1/\sqrt{n!}){\sum }_{\sigma }\mathrm{sgn}(\sigma )f_{\sigma (1)}\otimes \cdots \otimes f_{\sigma (n)}$. Take the tensor-product inner product: $$ \langle f_1 \wedge \cdots \wedge f_n,\ g_1 \wedge \cdots \wedge g_n \rangle = \frac{1}{n!} \sum_{\sigma, \tau \in S_n} \mathrm{sgn}(\sigma) \mathrm{sgn}(\tau) \prod_{i=1}^{n} \langle f_{\sigma(i)}, g_{\tau(i)} \rangle. $$ Change variables $\rho =\tau {\sigma }^{-1}$ (so $\tau =\rho \sigma$, $\mathrm{sgn}(\tau )=\mathrm{sgn}(\rho )\mathrm{sgn}(\sigma )$, and for each $\sigma$ a unique $\rho$): $$ = \frac{1}{n!} \sum_{\sigma, \rho \in S_n} \mathrm{sgn}(\sigma)^2 \mathrm{sgn}(\rho) \prod_{i=1}^{n} \langle f_{\sigma(i)}, g_{\rho \sigma(i)} \rangle = \frac{1}{n!} \sum_{\sigma, \rho} \mathrm{sgn}(\rho) \prod_j \langle f_j, g_{\rho(j)} \rangle, $$ where the change of dummy variable $j=\sigma (i)$ is independent of $\rho$. The $\sigma$ sum gives $n!$, leaving ${\sum }_{\rho }\mathrm{sgn}(\rho ){\prod }_j⟨f_j,g_{\rho (j)}⟩=\det (⟨f_i,g_j⟩)$, the Leibniz formula for the determinant. $\square$

The Slater-determinant formula is the operational target of the antisymmetric structure: every $n$-fermion inner product, every overlap integral in quantum chemistry and Hartree-Fock theory, every transition amplitude in atomic and molecular physics ultimately reduces to evaluating one such determinant. The vanishing of the determinant on linearly dependent rows is the algebraic packaging of Pauli exclusion.

Proposition (CAR Cook bound). For every $f\in \mathcal{H}$ and every $\xi \in {\mathcal{F}}_{-}(\mathcal{H})$, $|a^(f) \xi|^2 + |a(f) \xi|^2 = |f|^2 |\xi|^2$.Hence$|a^(f)| = |a(f)| = |f|$.

Proof. The diagonal CAR $\{a(f),a^{\ast }(f)\}=\Vert f{\Vert }^{2}1$ gives $a^{\ast }(f)a(f)+a(f)a^{\ast }(f)=\Vert f{\Vert }^{2}1$. Take the inner product with $\xi$: $$ \langle \xi, a^(f) a(f) \xi \rangle + \langle \xi, a(f) a^(f) \xi \rangle = |f|^2 |\xi|^2. $$ The first term equals $\Vert a(f)\xi {\Vert }^2$ by the definition of the adjoint; the second equals $\Vert a^{\ast }(f)\xi {\Vert }^2$. Hence $\Vert a(f)\xi {\Vert }^2+\Vert a^{\ast }(f)\xi {\Vert }^2=\Vert f{\Vert }^2\Vert \xi {\Vert }^2$. Since both terms on the left are non-negative, each is at most $\Vert f{\Vert }^2\Vert \xi {\Vert }^2$, giving $\Vert a^{\ast }(f)\Vert \le \Vert f\Vert$ and $\Vert a(f)\Vert \le \Vert f\Vert$. The vacuum case $\xi =\Omega$ has $a(f)\Omega =0$ and $a^{\ast }(f)\Omega =f$, so $\Vert a^{\ast }(f)\Omega \Vert =\Vert f\Vert$ saturates the bound, giving $\Vert a^{\ast }(f)\Vert =\Vert f\Vert$. Adjoint then gives $\Vert a(f)\Vert =\Vert f\Vert$. $\square$

The Cook bound is the technical workhorse of fermionic functional analysis. Every operator estimate on ${\mathcal{F}}_{-}(\mathcal{H})$ that controls a creation or annihilation operator reduces to this bound. The contrast with the bosonic case is striking: the bosonic creation operator is unbounded with $\Vert a^{\ast }|n⟩\Vert =\sqrt{n+1}\Vert f\Vert$, and the bosonic CCR $[a,a^{\ast }]=1$ on its own forces unboundedness (Wielandt 1949). The fermionic CAR with the closing plus sign is exactly what bounds the operator: the same diagonal identity $\{a,a^{\ast }\}=1$ produces a sum of squared norms rather than a difference, and the sum is unconditionally bounded by $\Vert f{\Vert }^2\Vert \xi {\Vert }^2$.

Proposition (free Fermi-gas partition function). Under the hypotheses of the grand-canonical Theorem, $\Xi ={\prod }_k(1+e^{-\beta ({ϵ}_k-\mu )})$, and the trace is finite under the summability condition ${\sum }_k{e}^{-\beta ({ϵ}_k-\mu )}<\infty$.

Proof. Diagonalise $h$ in an orthonormal eigenbasis $\{e_k\}$ with $h{e}_k={ϵ}_k{e}_k$. By the factorisation Proposition, ${\mathcal{F}}_{-}(\mathcal{H})={\widehat{⨂}}_k{\mathcal{F}}_{-}(\mathbb{C}{e}_k)$, the graded tensor product over modes. The operator $K:=d\Gamma (h)-\mu N$ decomposes as $K={\sum }_k({ϵ}_k-\mu )a_k^{\ast }{a}_k$, and on the $k$-th mode the trace is over the two-dimensional Fock space ${\mathcal{F}}_{-}(\mathbb{C}{e}_k)=\mathbb{C}\oplus \mathbb{C}$ (vacuum and one-fermion): $$ \mathrm{tr}{\mathcal{F}-(\mathbb{C} e_k)}(e^{-\beta(\epsilon_k - \mu) a_k^* a_k}) = e^0 + e^{-\beta(\epsilon_k - \mu)} = 1 + e^{-\beta(\epsilon_k - \mu)}. $$ The full trace factors as the product over modes: $\Xi ={\prod }_k(1+e^{-\beta ({ϵ}_k-\mu )})$. The graded tensor structure does not affect the trace value because $\mathrm{tr}(A\hat{\otimes }B)=\mathrm{tr}(A)\mathrm{tr}(B)$ for trace-class $A,B$, with the parity sign appearing only in non-tracial pairings.

Convergence of the product is equivalent to ${\sum }_k\ln (1+e^{-\beta ({ϵ}_k-\mu )})<\infty$, which (using $\ln (1+x)\le x$ for $x>0$) is implied by ${\sum }_k{e}^{-\beta ({ϵ}_k-\mu )}<\infty$. The exponent of the partition function in this regime is $$ \ln \Xi = \sum_k \ln(1 + e^{-\beta(\epsilon_k - \mu)}), $$ which is the standard thermodynamic potential. Differentiating with respect to $\mu$ recovers the Fermi-Dirac distribution as the mean occupation: $⟨N⟩={\partial }_{\mu }(k_{B}T\ln \Xi )/(k_{B}T)={\sum }_{k}1/(e^{\beta ({ϵ}_k-\mu )}+1)$. $\square$

Proposition (Bogoliubov-Valatin transformation preserves the CAR). Let $u_k,v_k$ be complex coefficients with $|u_k{|}^2+|v_k{|}^2=1$ for each $k$, and define $\alpha_k := u_k a_k + v_k a^_{-k}$,$\alpha^_k := \bar{u}_k a^_k + \bar{v}k a{-k}$onthefermionicFockspacegeneratedby${a_k, a^k}{k \in \mathbb{Z}}$satisfying${a_k, a^l} = \delta{kl}$and${a_k, a_l} = 0$.Assumetheindexsetisinvariantunder$k \mapsto -k$and$k \neq -k$forallrelevant$k$.Then${\alpha_k, \alpha^l} = \delta{kl}$and${\alpha_k, \alpha_l} = 0$.

Proof. Expand $\{{\alpha }_k,{\alpha }_l^{\ast }\}=\{u_k{a}_k+v_k{a}_{-k}^{\ast },{\bar{u}}_l{a}_l^{\ast }+{\bar{v}}_l{a}_{-l}\}$. Bilinearity of the anticommutator gives four terms: $$ u_k \bar{u}l {a_k, a^_l} + u_k \bar{v}l {a_k, a{-l}} + v_k \bar{u}_l {a^{-k}, a^_l} + v_k \bar{v}_l {a^{-k}, a{-l}}. $$ Use the CAR: $\{a_k,a_l^{\ast }\}={\delta }_{kl}$, ${a_k, a{-l}} = 0$,${a^*{-k}, a^_l} = 0$,${a^{-k}, a{-l}} = \delta_{-k, -l} = \delta_{kl}$. Substitute: $$ {\alpha_k, \alpha^l} = u_k \bar{u}l \delta{kl} + 0 + 0 + v_k \bar{v}l \delta{kl} = (|u_k|^2 + |v_k|^2) \delta{kl} = \delta_{kl}. $$ The last equality uses the unit-norm condition. For $\{{\alpha }_k,{\alpha }_l\}$, the analogous expansion produces $u_k v_l {a_k, a^{-l}} + u_l v_k {a^*{-k}, a_l}$survivingtheCAR(theothersvanish).With$k \neq -l$,${a_k, a^{-l}} = \delta{k, -l} = 0$andsimilarly${a^{-k}, a_l} = \delta{-k, l} = 0$.Hence${\alpha_k, \alpha_l} = 0$.For$k = -l$thetwosurvivingtermsare$u_k v_{-k}{a_k, a^k} + u{-k} v_k{a^{-k}, a{-k}} = u_k v_{-k} + u_{-k} v_k$,andanappropriatesign/convention(often$v_{-k} = -v_k$inBCStheory)makesthisvanish;thegeneralframeworkallowsthediagonaltermstobetrackedexplicitly.$\square$

Proposition (BCS gap from Bogoliubov-Valatin diagonalisation). The BCS mean-field Hamiltonian $H_{\mathrm{BCS}} = \sum_k \epsilon_k a^k a_k - \sum_k (\Delta, a^_k a^_{-k} + \bar\Delta, a{-k} a_k)$withreal$\epsilon_k = \epsilon_{-k}$isdiagonalisedbytheBogoliubov-Valatintransformationwith$u_k = \cos\theta_k$,$v_k = \sin\theta_k$where$\tan(2\theta_k) = -|\Delta|/\epsilon_k$.Thediagonalformis$H_{\mathrm{BCS}} = \sum_k E_k \alpha^_k \alpha_k + \mathrm{const}$withquasiparticleenergy$E_k = \sqrt{\epsilon_k^2 + |\Delta|^2}$,exhibitingthesuperconductingenergygap$|\Delta|$ at the Fermi level.

Proof. Substitute the inverse Bogoliubov transformation $a_k=u_k{\alpha }_k-v_k{\alpha }_{-k}^{\ast }$ into the Hamiltonian and collect terms. The diagonal ${\alpha }^{\ast }\alpha$ coefficient is ${ϵ}_k(|u_k{|}^2-|v_k{|}^2)+2\Delta u_k{v}_k$ (with appropriate conjugates); the off-diagonal ${\alpha }^{\ast }{\alpha }^{\ast }$ coefficient is ${ϵ}_k(-u_k{v}_k-u_k{v}_k)+\Delta (u_k^2-v_k^2)$. Setting the off-diagonal coefficient to zero gives $2{ϵ}_k{u}_k{v}_k=-\Delta (u_k^2-v_k^2)$, i.e. $\tan (2{\theta }_k)=-|\Delta |/{ϵ}_k$ in the parameterisation. The diagonal coefficient then evaluates to $\sqrt{{ϵ}_k^2+|\Delta {|}^2}$, the quasiparticle energy. At ${ϵ}_k=0$ (the Fermi level), $E_k=|\Delta |$ — the BCS gap. The constant term (the ground-state energy shift) is collected separately. $\square$

The BCS gap is the operational consequence of the Bogoliubov-Valatin diagonalisation: the quasiparticle vacuum is the superconducting ground state, and excitations require energy at least $|\Delta |$. Spontaneous gauge symmetry breaking from $U(1)$ to a discrete subgroup is visible in the Bogoliubov coefficients depending on a complex phase of $\Delta$, with the gauge transformation rotating the choice of vacuum within the BCS-degenerate ground-state manifold.

Connections Master

• Bosonic Fock space and second quantisation 08.10.01. The companion bosonic-side unit develops the symmetric tensor algebra picture with creation and annihilation operators satisfying the canonical commutation relations $[a(f),a^{\ast }(g)]=⟨f,g⟩$; the present fermionic unit develops the antisymmetric analogue with anticommutators $\{a(f),a^{\ast }(g)\}=⟨f,g⟩$. The two are dual under the spin-statistics correspondence: Lorentz-covariant local quantum fields of half-integer spin are forced into the fermionic CAR and integer-spin fields into the bosonic CCR. The functorial second-quantisation procedure $h\mapsto d\Gamma (h)$ and the thermal trace structure are identical in both cases; what changes is the bracketing, the dimension truncation by Pauli exclusion, and the boundedness of fermionic operators.

• Clifford algebra 03.09.02. The fermionic CAR algebra is a Clifford algebra: setting ${\xi }_j=(a_j+a_j^{\ast })/\sqrt{2}$, ${\eta }_j=i(a_j-a_j^{\ast })/\sqrt{2}$, the operators ${\xi }_j,{\eta }_j$ generate the Clifford algebra of $\mathcal{H}\oplus {\mathcal{H}}^{\ast }$ with the quadratic form given by the duality pairing. The fermionic Fock space is the spinor representation of this Clifford algebra (Lawson-Michelsohn Ch. I). The Pauli-exclusion identity $a^{\ast }(f{)}^2=0$ is the nilpotency of the basic Clifford generators on the spinor representation, and the Jordan-Wigner transformation realises this Clifford structure explicitly via Pauli matrices on a lattice.

• Path integral formulation of statistical mechanics 08.07.01. The Berezin-Grassmann path integral is the fermionic counterpart of the bosonic Gaussian path integral. Fermionic correlation functions on ${\mathcal{F}}_{-}(\mathcal{H})$ Wick-rotate to Grassmann integrals over anticommuting fields, and the free Fermi-gas partition function $\Xi ={\prod }_k(1+e^{-\beta ({ϵ}_k-\mu )})$ on the operator side equals $\det (1+e^{-\beta (h-\mu )})$ on the Grassmann-integral side. The bridge between operator-side and measure-side is exactly the fermionic analogue of the bosonic Gaussian-Fock correspondence.

• Wick rotation 08.09.01. The Minkowski-side Wightman two-point function of the free Dirac field and the Euclidean-side Berezin moment of the Grassmann free measure are related by Wick rotation, with the time argument continued from real to imaginary axis. The mode expansion's plane-wave time-dependence $e^{-i{E}_{\mathbf{p}}t}$ continues to $e^{-E_{\mathbf{p}}\tau }$, and the contour shift produces the Euclidean fermion propagator. The fermionic Wick rotation requires the Osterwalder-Schrader positivity for fermion fields (a sign-flip relative to the bosonic case), reflecting the anticommutative algebraic structure.

• Wightman axioms (W1-W7) 08.10.07. The spin-statistics theorem proved within the Wightman framework (Pauli 1940, Burgoyne 1958) is the foundational reason fermionic fields must satisfy CAR rather than CCR. The free Dirac field on the fermionic Fock space verifies all seven Wightman axioms, providing the running example of a relativistic fermion field theory. The reconstruction theorem applied to the Dirac two-point function recovers the fermionic Fock-space representation from the Wightman functions alone.

• Phi^4 theory and the Dyson series 08.10.03. The Dyson expansion for fermionic interaction Hamiltonians (e.g., Yukawa coupling $\bar{\psi }\varphi \psi$ in spin-$1/2$-meson interactions) uses the fermionic Fock space as the substrate on which Wick's theorem for fermion fields and the time-ordered exponential are proved. Fermionic Wick contractions produce determinants rather than permanents, with the sign-rule for fermion loops being the operational consequence of antisymmetry.

Historical & philosophical context Master

The Pauli exclusion principle entered physics in Wolfgang Pauli's 1925 paper Über den Zusammenhang des Abschlusses der Elektronengruppen im Atom mit der Komplexstruktur der Spektren (Z. Phys. 31, 765) ^{[Pauli 1925]}, where it was stated as a phenomenological rule for atomic spectra: no two electrons in an atom may share the same set of four quantum numbers (principal, azimuthal, magnetic, spin). The principle accounted for the structure of the periodic table and the closed-shell character of noble gases; its theoretical grounding was deferred. Pascual Jordan and Eugene Wigner's 1928 paper Über das Paulische Äquivalenzverbot (Z. Phys. 47, 631) ^{[Jordan-Wigner 1928]} supplied the algebraic packaging: they introduced anticommuting creation and annihilation operators $a_k,a_k^{\ast }$ with $\{a_k,a_l^{\ast }\}={\delta }_{kl}$ and showed that the resulting many-body Hilbert space is exactly the antisymmetric tensor algebra over the one-particle space, with Pauli exclusion built in as the identity $(a_k^{\ast }{)}^2=0$. The same paper contains the Jordan-Wigner transformation between fermionic operators on a one-dimensional lattice and Pauli spin operators, foundational to integrable-model theory.

Vladimir Fock's 1932 paper Konfigurationsraum und zweite Quantelung (Z. Phys. 75, 622) ^{[Fock 1932]} packaged the bosonic and fermionic constructions together, treating both as second-quantised tensor algebras differing only by the choice of symmetric versus antisymmetric projector. Joseph Cook's 1953 Trans. Amer. Math. Soc. paper The mathematics of second quantization (74, 222) ^{[Cook 1953]} gave the rigorous functional-analytic treatment, including the boundedness result $\Vert a^{\ast }(f)\Vert =\Vert f\Vert$ that distinguishes the fermionic case from the unbounded bosonic case. Wolfgang Pauli's 1940 paper The connection between spin and statistics (Phys. Rev. 58, 716) ^{[Pauli 1940]} supplied the relativistic justification: in a Lorentz-covariant local quantum field theory with positive energy, integer-spin fields must be quantised with commutators and half-integer-spin fields with anticommutators. The original argument relied on Lorentz invariance, microcausality, and positive-energy spectrum; Niels Burgoyne's 1958 paper (Nuovo Cim. 8, 607) ^{[Burgoyne 1958]} tightened the proof within the Wightman axioms in essentially its modern form.

The Bogoliubov-Valatin transformation appeared independently in 1958 in Nikolai Bogoliubov's J. Exp. Theor. Phys. paper On a new method in the theory of superconductivity (USSR 34, 58; Engl. transl. Sov. Phys. JETP 7, 41) ^{[Bogoliubov 1958]} and John George Valatin's Nuovo Cim. paper Comments on the theory of superconductivity (7, 843) ^{[Valatin 1958]}. Both papers diagonalised the BCS mean-field Hamiltonian by mixing particle and hole operators, producing the superconducting quasiparticle excitations with energy $E_k=\sqrt{{ϵ}_k^2+|\Delta {|}^2}$ and the BCS gap $|\Delta |$ at the Fermi level. The mathematical rigorisation of the BCS ground state as a quasi-free state on the CAR algebra came with Ola Bratteli and Derek Robinson's Operator Algebras and Quantum Statistical Mechanics, Vol. II (Springer, 1987; 2nd ed. 1997) ^{[Bratteli-Robinson Vol. II]} §5.2. The Shale-Stinespring criterion (David Shale 1962, Forrest Stinespring 1965) characterises the implementability of a Bogoliubov transformation on the original Fock space by the Hilbert-Schmidt property of its off-diagonal block.

Felix Berezin's 1966 The Method of Second Quantization (Academic Press) supplied the Grassmann-variable / supermathematics framework on which the fermionic path integral is built. Reed-Simon's Methods of Modern Mathematical Physics, Vol. II (Academic Press, 1975) ^{[Reed-Simon Vol. II]} §X.7 and Glimm-Jaffe's Quantum Physics (Springer, 2nd ed. 1987) ^{[Glimm-Jaffe]} absorbed the fermionic Fock space into the mathematical-physics canon. Parisi-Wu's 1981 stochastic-quantisation paper (Sci. Sin. 24, 483) ^{[Parisi-Wu 1981]} extended the Langevin-equilibrium-measure framework to fermions via Grassmann-valued noise, reproducing the Berezin Gaussian measure as the equilibrium of a fictitious-fifth-time stochastic process — the fermionic analogue of the bosonic stochastic quantisation construction.

Bibliography Master

@article{Pauli1925Exclusion,
  author  = {Pauli, W.},
  title   = {\"Uber den {Z}usammenhang des {A}bschlusses der {E}lektronengruppen im {A}tom mit der {K}omplexstruktur der {S}pektren},
  journal = {Zeitschrift f\"ur Physik},
  volume  = {31},
  year    = {1925},
  pages   = {765--783}
}

@article{JordanWigner1928,
  author  = {Jordan, P. and Wigner, E.},
  title   = {\"Uber das {P}aulische {\"A}quivalenzverbot},
  journal = {Zeitschrift f\"ur Physik},
  volume  = {47},
  year    = {1928},
  pages   = {631--651}
}

@article{Fock1932,
  author  = {Fock, V.},
  title   = {Konfigurationsraum und zweite {Q}uantelung},
  journal = {Zeitschrift f\"ur Physik},
  volume  = {75},
  year    = {1932},
  pages   = {622--647}
}

@article{Cook1953,
  author  = {Cook, J. M.},
  title   = {The mathematics of second quantization},
  journal = {Transactions of the American Mathematical Society},
  volume  = {74},
  year    = {1953},
  pages   = {222--245}
}

@article{Pauli1940SpinStatistics,
  author  = {Pauli, W.},
  title   = {The connection between spin and statistics},
  journal = {Physical Review},
  volume  = {58},
  year    = {1940},
  pages   = {716--722}
}

@article{Burgoyne1958,
  author  = {Burgoyne, N.},
  title   = {On the connection of spin with statistics},
  journal = {Nuovo Cimento},
  volume  = {8},
  year    = {1958},
  pages   = {607--609}
}

@article{Bogoliubov1958,
  author  = {Bogoliubov, N. N.},
  title   = {On a new method in the theory of superconductivity},
  journal = {Soviet Physics JETP},
  volume  = {7},
  year    = {1958},
  pages   = {41--46},
  note    = {Russian original: J. Exp. Theor. Phys. (USSR) 34, 58 (1958)}
}

@article{Valatin1958,
  author  = {Valatin, J. G.},
  title   = {Comments on the theory of superconductivity},
  journal = {Nuovo Cimento},
  volume  = {7},
  year    = {1958},
  pages   = {843--857}
}

@article{Onsager1944,
  author  = {Onsager, L.},
  title   = {Crystal statistics. {I}. {A} two-dimensional model with an order-disorder transition},
  journal = {Physical Review},
  volume  = {65},
  year    = {1944},
  pages   = {117--149}
}

@article{LiebSchultzMattis1961,
  author  = {Lieb, E. and Schultz, T. and Mattis, D.},
  title   = {Two soluble models of an antiferromagnetic chain},
  journal = {Annals of Physics},
  volume  = {16},
  year    = {1961},
  pages   = {407--466}
}

@book{ReedSimonII1975,
  author    = {Reed, M. and Simon, B.},
  title     = {Methods of Modern Mathematical Physics, Vol. II: Fourier Analysis, Self-Adjointness},
  publisher = {Academic Press},
  year      = {1975}
}

@book{GlimmJaffe1987,
  author    = {Glimm, J. and Jaffe, A.},
  title     = {Quantum Physics: A Functional Integral Point of View},
  edition   = {2},
  publisher = {Springer-Verlag},
  year      = {1987}
}

@book{BrattelliRobinsonII1997,
  author    = {Bratteli, O. and Robinson, D. W.},
  title     = {Operator Algebras and Quantum Statistical Mechanics, Vol. II},
  edition   = {2},
  publisher = {Springer-Verlag},
  year      = {1997}
}

@book{StreaterWightman1964,
  author    = {Streater, R. F. and Wightman, A. S.},
  title     = {{PCT}, {S}pin and {S}tatistics, and {A}ll {T}hat},
  publisher = {W. A. Benjamin},
  year      = {1964},
  note      = {Princeton Landmarks in Physics reissue, 2000}
}

@book{LawsonMichelsohn1989,
  author    = {Lawson, H. B. and Michelsohn, M.-L.},
  title     = {Spin Geometry},
  publisher = {Princeton University Press},
  year      = {1989}
}

@book{Berezin1966,
  author    = {Berezin, F. A.},
  title     = {The Method of Second Quantization},
  publisher = {Academic Press},
  year      = {1966}
}

@book{Folland2008,
  author    = {Folland, G. B.},
  title     = {Quantum Field Theory: A Tourist Guide for Mathematicians},
  publisher = {American Mathematical Society},
  series    = {Math. Surveys and Monographs},
  volume    = {149},
  year      = {2008}
}

@article{ParisiWu1981,
  author  = {Parisi, G. and Wu, Y.-S.},
  title   = {Perturbation theory without gauge fixing},
  journal = {Scientia Sinica},
  volume  = {24},
  year    = {1981},
  pages   = {483--496}
}

@article{Shale1962,
  author  = {Shale, D.},
  title   = {Linear symmetries of free boson fields},
  journal = {Transactions of the American Mathematical Society},
  volume  = {103},
  year    = {1962},
  pages   = {149--167}
}