lattice-system

Legacy catalogue: Multi-mode fermion via Jordan–Wigner (P2 backbone) (part 4 of 4)

Interim authority. This lossless catalogue chunk remains authoritative for formalization status and capstone identification until Issue #5228. The version 1 JSON catalogue is still a non-authoritative prototype.

Interim catalogueFermions and Hubbard models

| Lean name | Statement | File | |—|—|—| | theorem_10_4_lieb_repulsive_half_filling | Theorem 10.4 (Tasaki §10.2.2, p. 350, AXIOM): at half-filling N = \|Λ\|, the ground subspace is nonzero, energy-minimal, consists entirely of total-spin S₀ = \|\|A\|−\|B\|\|/2 states (Casimir S₀(S₀+1)), and has dimension exactly \|A\|−\|B\|+1 (the unavoidable SU(2) multiplet degeneracy). Lieb’s reflection positivity via the Shiba transformation → faithful documented axiom. | Fermion/JordanWigner/Hubbard/LiebRepulsive.lean | | shibaConfig / shibaConfigEquiv / shibaConfig_up / shibaConfig_down / shibaConfig_involutive | Shiba configuration involution (Tasaki §9.3.3, eq. (9.3.48), p. 335): particle-hole flip on down-spin occupations (n̂↑, n̂↓) ↦ (n̂↑, 1−n̂↓) as the permutation part of the Shiba transformation for Theorem 10.4. shibaConfig is the down-species flip; shibaConfigEquiv packages it as an involution (Equiv); proofs establish it is self-inverse with fixed up-config and flipped down-config. PR #4950 (c2): axiom-free infrastructure. | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaTransform.lean | | shibaPermMatrix / shibaPermMatrix_apply / shibaPermMatrix_isHermitian / shibaPermMatrix_conj_diagonal | Shiba permutation matrix (Tasaki §9.3.3, eqs. (9.3.48)/(9.3.50), pp. 335–336): the permutation matrix realizing the Shiba involution on Fock configurations; Hermitian (Pᴴ=P); conjugating a diagonal by P reindexes via shibaConfig. PR #4950 (c3): unsigned permutation part of the full Shiba unitary; axiom-free infrastructure. | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaUnitary.lean | | symmetricRepulsiveInteractionDiag / symmetricRepulsiveHubbardInteraction_eq_diagonal / shibaConfig_apply_up / shibaConfig_apply_down / symmetricRepulsiveInteractionDiag_shibaConfig / shibaPermMatrix_conj_symmetricInteraction / shibaSignedUnitary / shibaSignedUnitary_conjTranspose / shibaSignedUnitary_conj_symmetricInteraction | Shiba interaction conjugation (Tasaki §9.3.3, eqs. (9.3.47)/(9.3.54), pp. 334–336): the symmetric repulsive interaction Ĥ_{int}' is diagonal; Shiba flip negates each term (n̂↓−½ changes sign); capstone shibaSignedUnitary_conj_symmetricInteraction: the signed-Shiba unitary satisfies Ûᴴ·Ĥ_{int}'·Û = −Ĥ_{int}' (eq. 9.3.54). PR #4950 (c5): discharges axiom-free interaction sign-flip; consumed by the c6 full-conjugation capstone (PR #4952). | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaInteraction.lean | | gaugeSign / shibaGauge / shibaCrossingSpecies / shibaJwFlipParity / shibaSignFn / gaugeSign_mul_of_bipartite / shibaGauge_star_mul_self / shibaSignFn_star_mul_self / shibaSignedUnitary_conj_symmetricKinetic | Shiba kinetic-term conjugation (Tasaki §9.3.3, eq. (9.3.52), p. 336): the Shiba transformation leaves the symmetric kinetic term invariant Ûᴴ·Ĥ_kin·Û = Ĥ_kin. Mechanism: on a bipartite bond x∈A, y∈B, the particle-hole flip converts ĉ†_x ĉ_y into −ĉ†_y ĉ_x (CAR sign −1) while the sublattice gauge contributes ε_x ε_y = −1; the two signs multiply to +1, restoring the symmetric term t_xy = t_yx. capstone shibaSignedUnitary_conj_symmetricKinetic: kinetic invariance; shibaSignFn_star_mul_self (|s|=1) feeds the c6 conjugation. PR #4951 (c4): axiom-free kinetic-term preservation. | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaKinetic.lean | | symmetricHubbardOnSite_expand / hubbardKinetic_add / hubbardKinetic_diagonal / neg_symmetricRepulsiveInteraction_eq / shibaSignedUnitary_conj_symmetricRepulsive_eq_attractive | Shiba full conjugation repulsive→attractive (Tasaki §10.2.2, eqs. (10.2.10)/(10.2.11), p. 352): combining c4 (kinetic invariance) and c5 (interaction sign-flip), capstone shibaSignedUnitary_conj_symmetricRepulsive_eq_attractive: Ûᴴ·(Ĥ_kin+Ĥ_int')·Û = attractiveHubbardHamiltonian (T + diag(U/2)) U − ¼Σ_x U_x (eq. 10.2.10). The chemical shift of −Ĥ_int' (from (n̂↑−½)(n̂↓−½)=n̂↑n̂↓−½(n̂↑+n̂↓)+¼, eq. 10.2.11) is absorbed into the hopping diagonal, leaving the negative scalar constant. PR #4952 (c6): axiom-free; consumes the c4/c5 milestones; feeds the c7 balanced-sector transport. | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaConjugation.lean | | sum_spinful_split / fermionTotalNumber_eq_diagonal / fermionTotalSpinZ_eq_diagonal / shibaSignedUnitary_conj_diagonal / shibaSignedUnitary_conj_totalNumber / shibaSignedUnitary_conj_totalSpinZ | Shiba number/spin-z charge exchange (Tasaki §§9.3.3/10.2.2, eqs. (9.3.48)/(9.3.50), pp. 334–352): the charges , Ŝ³ are diagonal in the Fock basis, and the Shiba flip interchanges them — Û·N̂·Ûᴴ = 2Ŝ³ + (N+1)·1 and Ûᴴ·Ŝ³·Û = ½(N̂ − (N+1)·1). PR #4953 (c7 foundation): axiom-free; the sector-correspondence identities feeding the balanced-sector transport. | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaSector.lean | | spinZSectorEuclidean / repulsiveSpinZSector_ground_unique | Theorem 10.4, general spin-z-sector portion (Tasaki §10.2.2, pp. 350–352, eqs. (10.2.10)/(10.2.11), PROVED axiom-free): for any N, an even electron number 0 < Ne < 2(N+1), connected real symmetric bipartite hopping and on-site repulsion U_x > 0, the symmetric (μ = U/2) repulsive Hubbard Hamiltonian has a unique ground state on the spin-z sector Ŝ³ = m with m = (Ne−(N+1))/2. Proved by transporting Theorem 10.2 (number sector N̂ = Ne) through the Shiba unitary (c6 conjugation + c7 charge exchange Û N̂ Ûᴴ = 2Ŝ³+(N+1)·1, so Ne ↦ m). The total-spin value is not claimed — Û maps Ŝ² to the η-pseudospin Casimir; identifying it needs the deferred degenerate perturbation axiom. Half-integer m (odd Ne) is out of scope (Theorem 10.2 requires Even Ne). PR #4955 (general-sector PR-1). | Fermion/JordanWigner/Hubbard/LiebRepulsiveBalancedGround.lean | | shibaCrossingSpecies_update_flip_product / shibaGauge_update_down_flip_product / shibaSignedUnitary_conj_siteSpinPlus / shibaSignedUnitary_conj_siteSpinMinus / shibaSignedUnitary_conj_spinPlusMinus | Shiba spin-operator conjugation (Tasaki §10.2.2, eq. (10.2.13), p. 353, PROVED axiom-free): the Shiba flip sends the transverse spin operators to the on-site pair (η-pairing) operators — Ûᴴ Ŝ⁺_x Û = ε_x·ĉ†_{x↑}ĉ†_{x↓}, Ûᴴ Ŝ⁻_x Û = ε_x·ĉ_{x↓}ĉ_{x↑} (ε_x = sublattice gauge), hence capstone shibaSignedUnitary_conj_spinPlusMinus: Ûᴴ(Ŝ⁺_xŜ⁻_y)Û = ε_xε_y·hubbardPairCorrelationOp N x y. Basis-vec computation mirroring the Shiba down-hop conjugation. | Fermion/JordanWigner/Hubbard/LiebRepulsiveShibaSpinOp.lean | | fermionSpinTransverse / vectorExpectation / SameSublattice / theorem_10_5_shen_qiu_tian_transverse_sign | Theorem 10.5 (Shen–Qiu–Tian; Tasaki §10.2.2, p. 351/353, eqs. (10.2.7)/(10.2.8)/(10.2.13), PROVED axiom-free on the general spin-z sector Ŝ³=m (PR #4956)): for any N, even electron number 0<Ne<2(N+1), the transverse spin correlation ⟨φ\| Ŝ¹_xŜ¹_y+Ŝ²_xŜ²_y \|φ⟩ (= ½(Ŝ⁺_xŜ⁻_y+Ŝ⁻_xŜ⁺_y)) in the unique repulsive-Hubbard ground state on the sector Ŝ³=m (m=(Ne−(N+1))/2) is real, strictly positive when x,y are in the same sublattice and strictly negative otherwise. Proof: transport φ=Ûφ_attr through the Shiba unitary via the general-sector uniqueness repulsiveSpinZSector_ground_unique, apply the spin↔pair conjugation (eq. 10.2.13) and Tian’s positivity (Theorem 10.3). Only the correlation sign is claimed (total-spin value not asserted); the balanced Ŝ³=0 sector is the Ne=N+1 special case. | Fermion/JordanWigner/Hubbard/LiebRepulsiveCorrelation.lean | | fermionStaggeredCasimirOp / theorem_10_6_lieb_ferrimagnetism | Theorem 10.6 (Shen–Qiu–Tian ferrimagnetism; Tasaki §10.2.3, p. 354, eqs. (10.2.16)/(10.2.17), AXIOM): every normalized repulsive-Hubbard ground state satisfies ⟨v\| (Ô_L)² \|v⟩ ≥ ((\|A\|−\|B\|)/2)², where (Ô_L)² = Σ_{x,y} ε_xε_y Ŝ_x·Ŝ_y (staggered sign ε_x=±1 per sublattice) — ferrimagnetic long-range order. Reuses IsLiebRepulsiveModel. Reflection positivity → faithful documented axiom. | Fermion/JordanWigner/Hubbard/LiebFerrimagnetism.lean | | bipartiteSignMatrix / proposition_10_7_charpoly_neg_eq / proposition_10_7_zero_mode_lower_bound | Proposition 10.7 (Tasaki §10.2.3, p. 356, PROVED axiom-free): for a bipartite real symmetric hopping T, (i) the single-electron spectrum is symmetric about zero — (-T).charpoly = T.charpoly via the gauge D T D = -T (D = diag(±1)); (ii) there are at least \|A\|−\|B\| zero modes — Module.finrank (ker T.mulVecLin) ≥ \|A\|−\|B\| via the A→B block map and rank–nullity. The one Chapter-10 item that is finite-dim linear algebra. | Fermion/JordanWigner/Hubbard/BipartiteSpectrum.lean | | totalPairAnnihilationOperator / totalPairCreationOperator / totalPairCorrelationOperator / symmetricAttractiveHubbardHamiltonian / liebShenQiuPairLowerBound / theorem_10_8_lieb_shen_qiu_superconductivity | Theorem 10.8 (Lieb–Shen–Qiu superconductivity; Tasaki §10.2.3, p. 359, eq. (10.2.22), AXIOM): for the symmetric attractive Hubbard model Ĥhop − Σ_x U_x(n̂_↑−½)(n̂_↓−½) (eq. (10.2.21)) on a bipartite lattice with even N, 2\|B\| ≤ N ≤ 2\|A\|, the unique ground state satisfies ⟨φ\| b̂† b̂ \|φ⟩ ≥ (\|A\|−N/2)(N/2−\|B\|) with b̂ = Σ_x ĉ_{x↓}ĉ_{x↑} — off-diagonal long-range order (fermion-pair condensation / superconductivity). Reflection positivity + Theorem 10.2 uniqueness → faithful documented axiom. | Fermion/JordanWigner/Hubbard/LiebShenQiu.lean |

Kubo–Kishi finite-temperature susceptibility bound (Tasaki §10.2.5, Theorem 10.11, AXIOM)

The finite-temperature version of Lieb’s theorem (Kubo–Kishi): at half-filling the repulsive Hubbard model’s charge and on-site pairing susceptibilities are bounded uniformly in temperature and wave number, so there is no CDW or superconducting long-range order. Tasaki states it without proof, citing Kubo–Kishi, Phys. Rev. B 41, 4866 (1990) → recorded as a documented axiom (same policy as mielke_theorem_11_13 / theorem_10_4).

Lean name Statement File
imagTimeEvolve / duhamelStaticSusceptibility (generic) the imaginary-time (Wick-rotated) Heisenberg evolution A(τ) = e^{τH}A e^{-τH} and the Duhamel (Kubo) static isothermal susceptibility χ_{AB}(β) = ∫₀^β (⟨A(τ)B⟩_β − ⟨A⟩_β⟨B⟩_β) dτ (Tasaki §10.2.5, the fluctuation–dissipation form of the second-derivative susceptibilities eqs. (10.2.53)/(10.2.54); prefactor 1, -valued intervalIntegral) Quantum/GibbsState/Duhamel.lean
grandCanonicalRepulsiveHubbard / fourierCharacter / chargeFourierMode / pairFieldFourierMode / chargeSusceptibility / pairSusceptibility the grand-canonical Hamiltonian Ĥ − μN̂ (core of eq. (10.2.52)); the 1D wave-number character w_x = exp(2πi k x/(N+1)) (eq. (10.2.55), |w_x|=1, genuine periodic lattice character); the charge Fourier mode ñ_q = |Λ|^{-1/2}Σ_x w_x n̂_x and the pairing Fourier mode p̂_q = |Λ|^{-1/2}Σ_x w_x(ĉ†_{x↑}ĉ†_{x↓}+ĉ_{x↓}ĉ_{x↑}); and the charge / on-site pairing susceptibilities χ^c_q, χ^p_q as Duhamel two-point functions of these modes at ±q (eqs. (10.2.53)/(10.2.54)) Fermion/JordanWigner/Hubbard/LiebKuboKishi.lean
theorem_10_11_kubo_kishi_susceptibility_bound Theorem 10.11 (Kubo–Kishi; Tasaki §10.2.5, pp. 368–369, eqs. (10.2.52)–(10.2.56), AXIOM, PR #4957): for the uniform repulsive Hubbard model (U > 0, eq. (10.2.5)) on a bipartite real symmetric connected hopping T (Theorem 10.4 conditions except electron number) at half-filling μ = U/2, for every β > 0 and every wave number k the susceptibilities are real and χ^c_k(β, U/2) ≤ 1/U, χ^p_k(β, U/2) ≤ 2/U (eq. (10.2.56)) — no CDW or superconducting order at finite temperature. Tasaki states it without proof, citing Kubo–Kishi, Phys. Rev. B 41, 4866 (1990) → faithful documented axiom. Fermion/JordanWigner/Hubbard/LiebKuboKishi.lean

Hubbard effective Hamiltonian on the hard-core sector (Tasaki §11.2)

Lean name Statement File
hubbardEffectiveHamiltonian N t U the hard-core compression Ĥ_eff = P̂_hc · H · P̂_hc of the full Hubbard Hamiltonian (Tasaki §11.2; 1st ed., pp. 381-388) Fermion/JordanWigner/Hubbard/EffectiveHamiltonian.lean
hubbardEffectiveHamiltonian_isHermitian Ĥ_eff is Hermitian for Hermitian t and real U Fermion/JordanWigner/Hubbard/EffectiveHamiltonian.lean
hubbardEffectiveHamiltonian_mulVec_eq_projected_kinetic_of_mem U → ∞ reduction: on a hard-core vector, Ĥ_eff ψ = P̂_hc (H_hop ψ) (interaction drops out) Fermion/JordanWigner/Hubbard/EffectiveHamiltonian.lean
hubbardEffectiveHamiltonian_mulVec_mem Ĥ_eff ψ always lies in hubbardHardcoreSubspace Fermion/JordanWigner/Hubbard/EffectiveHamiltonian.lean
hubbardHardcoreProjection_mulVec_effectiveHamiltonian P̂_hc fixes every vector in the range of Ĥ_eff Fermion/JordanWigner/Hubbard/EffectiveHamiltonian.lean

Tasaki ordered-creation basis (Tasaki §11.2, eq. (11.2.3))

Lean name Statement File
occupationOf js the configuration occupied exactly on the indices in a list js Fermion/JordanWigner/Hubbard/TasakiBasis.lean
prod_creation_mulVec_vacuum a strictly index-sorted product of creation operators applied to the vacuum yields basisVec (occupationOf js), every Jordan–Wigner string sign being 1 Fermion/JordanWigner/Hubbard/TasakiBasis.lean
tasakiIndexList N τ / tasakiIndexList_sorted / mem_tasakiIndexList_iff the increasing list of occupied JW indices of the all-filled spin-τ configuration; strictly sorted; membership iff the spin label selects that orbital Fermion/JordanWigner/Hubbard/TasakiBasis.lean
hubbardTasakiBasisState N x σ the (11.2.3) ordered-creation basis state \|Φ_{x,σ}⟩ = ĉ_{x,↑} (∏_{y} ĉ†_{y,σ̄_y}) \|vac⟩ Fermion/JordanWigner/Hubbard/TasakiBasis.lean
hubbardTasakiBasisState_eq_smul_basisVec \|Φ_{x,σ}⟩ = ε • basisVec (hubbardOneHoleConfig N x σ), where ε is the single annihilation string sign Fermion/JordanWigner/Hubbard/TasakiBasis.lean
hubbardTasakiBasisSign_mul_self the string sign squares to 1 Fermion/JordanWigner/Hubbard/TasakiBasis.lean
hubbardTasakiBasisState_inner / hubbardTasakiBasisState_self_inner orthonormality inherited from basisVec: the pairing is the sign product times the configuration-equality indicator; self-overlap is 1 Fermion/JordanWigner/Hubbard/TasakiBasis.lean

Uniform-sign hole-filling action (Tasaki §11.2, eq. (11.2.4))

Lean name Statement File
jwSign_eq_neg_one_pow jwSign N j c = (-1)^{∑_{k<j} (c k).val} (string sign as parity of occupied modes) Fermion/JordanWigner/Hubbard/TasakiHopActionCore.lean
sum_spinful_reindex reindex a sum over the 2N+2 JW modes into a double site/spin sum Fermion/JordanWigner/Hubbard/TasakiHopActionCore.lean
hubbardTasakiBasisSign_eq the Tasaki basis sign is ε = (-1)^x (independent of σ) Fermion/JordanWigner/Hubbard/TasakiHopActionCore.lean
hop_jwSign_source / hop_jwSign_target the two hop string signs as (-1) to the occupied-site count below the source / target orbital Fermion/JordanWigner/Hubbard/TasakiHopActionCore.lean
hubbardTasakiHop_mulVec eq. (11.2.4): ĉ†_{(x,s)} ĉ_{(z,s)} \|Φ_{x,σ}⟩ = -\|Φ_{z, σ_{z→x}}⟩ — the four fermion signs combine to the uniform -1 (parity 2(x+z)-1 is odd) Fermion/JordanWigner/Hubbard/TasakiHopAction.lean

Effective-Hamiltonian matrix element (Tasaki §11.2, eq. (11.2.5))

Lean name Statement File
hubbardEffective_tasaki_matrixElement eq. (11.2.5), off-diagonal (x ≠ y): ⟨Φ_{y,τ} \| Ĥ_eff \| Φ_{x,σ}⟩ = -t_{x,y} if τ = σ_{y→x} (configurations coincide), else 0. Only the hole-filling channel (x, y, σ_y) survives the hard-core projection Fermion/JordanWigner/Hubbard/EffectiveHamiltonianMatrix.lean
hubbardEffective_tasaki_matrixElement_diag eq. (11.2.5), diagonal (x = y) under no self-hopping (∀ i, t_{ii} = 0): ⟨Φ_{x,τ} \| Ĥ_eff \| Φ_{x,σ}⟩ = 0Ĥ_eff moves the hole off x, so its image is orthogonal to \|Φ_{x,τ}⟩ Fermion/JordanWigner/Hubbard/EffectiveHamiltonianMatrix.lean

Cauchy–Schwarz energy bound (Tasaki §11.2, eq. (11.2.9))

Lean name Statement File
HoleSpin N x / holeSpinMoveEquiv canonical hole-spin configurations (hole-site spin fixed ); the hole-move bijection HoleSpin N x ≃ HoleSpin N y Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
hubbardEffEnergy / hubbardEffEnergy_expand the real-bilinear effective-Hamiltonian energy; its quadratic-form expansion in the Tasaki basis Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
oneHoleConfig_move_eq_iff the hole-move bridge: the moved configuration matches C_{x,σ} iff τ is the hole-move image Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
hubbardEffEnergy_tasaki_quadratic eq. (11.2.9) line 2: ⟨Φ\|Ĥ_eff\|Φ⟩ = -Σ_{x≠y} t_{y,x} Σ_σ c_{x,σ} c_{y, σ_{x→y}} (no self-hopping) Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
tasakiQuadForm_ferro_le the Cauchy–Schwarz bound on the real quadratic form: Q(Φ_↑) ≤ Q(Φ) for t ≥ 0 Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
hubbardWeakNagaoka_energy_bound eq. (11.2.9): ⟨Φ_↑\|Ĥ_eff\|Φ_↑⟩ ≤ ⟨Φ\|Ĥ_eff\|Φ⟩ (t ≥ 0, t_{ii}=0) — the ferromagnetic state is also a ground state Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
tasakiState_orthonormal orthonormality of the Tasaki basis (indexed by hole site and canonical hole-spin) Fermion/JordanWigner/Hubbard/WeakNagaoka.lean
tasakiExpansion_normSq / ferroCoeff_normSq_eq / tasakiFerro_normSq_eq eqs. (11.2.7)–(11.2.8): ‖Σ ϕ_p Φ_p‖² = Σ ϕ_p², and the ferromagnetic state has the same norm ‖Φ_↑‖ = ‖Φ‖ Fermion/JordanWigner/Hubbard/WeakNagaoka.lean

SU(2) symmetry of the effective Hamiltonian (Tasaki §11.2)

Lean name Statement File
fermionTotalSpinPlus_commute_hubbardHardcoreProjection Ŝ^+_tot commutes with the hard-core projection P̂_hc (spin operators preserve the no-double-occupancy subspace) Fermion/JordanWigner/Hubbard/EffectiveHamiltonianSpinSymmetry.lean
fermionTotalSpinPlus_commute_hubbardEffectiveHamiltonian / fermionTotalSpinMinus_commute_hubbardEffectiveHamiltonian [Ĥ_eff, Ŝ^±_tot] = 0 — the effective Hamiltonian inherits SU(2) symmetry (the backbone of the (2S_max+1)-degeneracy in Theorem 11.5) Fermion/JordanWigner/Hubbard/EffectiveHamiltonianSpinSymmetry.lean
fermionTotalSpinZ_commute_hubbardEffectiveHamiltonian / fermionTotalSpinSquared_commute_hubbardEffectiveHamiltonian [Ĥ_eff, Ŝ^z_tot] = 0 and [Ĥ_eff, (Ŝ_tot)²] = 0Ĥ_eff conserves total spin, so its eigenspaces split into fixed-S_tot sectors (Theorem 11.5 / 11.7) Fermion/JordanWigner/Hubbard/EffectiveHamiltonianSpinSymmetry.lean

Weak Nagaoka spin multiplet (Tasaki §11.2.1, Theorem 11.5 core)

Lean name Statement File
fermionTotalSpinPlus_commutator_fermionTotalSpinMinus SU(2) ladder commutator [Ŝ^+_tot, Ŝ^-_tot] = 2 Ŝ^z_tot (Tasaki §9.3.3) Fermion/JordanWigner/Hubbard/WeakNagaokaTheoremCore.lean
fermionTotalSpinSquared_commute_fermionTotalSpinMinus [(Ŝ_tot)², Ŝ^-_tot] = 0 — the Casimir is constant along the spin-lowering tower Fermion/JordanWigner/Hubbard/WeakNagaokaTheoremCore.lean
fermionTotalSpinPlus_mul_fermionTotalSpinMinus Ŝ^+_tot Ŝ^-_tot = (Ŝ_tot)² − Ŝ^z_tot(Ŝ^z_tot − 1) — raising-after-lowering via the Casimir Fermion/JordanWigner/Hubbard/WeakNagaokaTheorem.lean
fermionTotalSpinZ_mulVec_spinMinusPow / fermionTotalSpinSquared_mulVec_spinMinusPow Ŝ^z / Casimir towers: (Ŝ^-_tot)^k v has Ŝ^z = N/2 − k and constant Casimir, for any highest-weight v Fermion/JordanWigner/Hubbard/WeakNagaokaTheoremCore.lean (fermionTotalSpinZ_mulVec_spinMinusPow) + Fermion/JordanWigner/Hubbard/WeakNagaokaTheorem.lean (fermionTotalSpinSquared_mulVec_spinMinusPow)
spinMinusPow_ne_zero / spinMinusPow_linearIndependent the N+1 lowered states (Ŝ^-_tot)^k v (k ≤ N) are nonzero (ladder eigenvalue (k+1)(N−k) ≠ 0) and linearly independent (distinct Ŝ^z) Fermion/JordanWigner/Hubbard/WeakNagaokaTheorem.lean
weakNagaoka_spinMultiplet Theorem 11.5 core: a nonzero highest-weight Ĥ_eff-eigenvector v (Ŝ^+_tot v = 0, Ŝ^z_tot v = (N/2) v) generates N+1 = 2 S_max + 1 linearly independent Ĥ_eff-eigenvectors (Ŝ^-_tot)^k v at the same energy, all with S_tot = S_max = N/2. Fermion/JordanWigner/Hubbard/WeakNagaokaTheorem.lean
tasakiEffMatrix / tasakiEffMatrixUp the Tasaki matrix M = Tᴴ Ĥ_eff T of Ĥ_eff in the one-hole basis (Hermitian, entries ⟨Φ_q\|Ĥ_eff\|Φ_p⟩) and its all-up principal block M_↑ (Hermitian); Ĥ_eff preserves the all-up sector Fermion/JordanWigner/Hubbard/WeakNagaokaGroundState.lean
hubbardEffectiveHamiltonian_mulVec_tasakiState / _tasakiExpansion operator lift: Ĥ_eff Φ_p = Σ_q ⟨Φ_q\|Ĥ_eff\|Φ_p⟩ Φ_q and Ĥ_eff (Σ c_q Φ_q) = Σ_q (M c)_q Φ_q (via particle-number conservation, sector completeness tasaki_completeness) Fermion/JordanWigner/Hubbard/WeakNagaokaGroundStateCore.lean + Fermion/JordanWigner/Hubbard/WeakNagaokaGroundState.lean
weakNagaoka_theorem_11_5 Tasaki Theorem 11.5 (weak Nagaoka, effective one-hole sector): there exist N+1 = 2 S_max+1 linearly independent Ĥ_eff-eigenvectors at the maximal-spin sector minimum energy, all with S_tot = S_max = N/2; the ground state is the all-up state from a minimum eigenvector of the all-up block M_↑ (no Perron–Frobenius). Fermion/JordanWigner/Hubbard/WeakNagaokaGroundState.lean
weakNagaoka_theorem_11_5_global Tasaki Theorem 11.5 (global form): for t ≥ 0 symmetric, the all-up minimum equals the global one-hole minimum (hermitianMinEigenvalue M_↑ = hermitianMinEigenvalue M, via the Schwarz bound (11.2.9) ferromagnetization on the real Tasaki matrix), so the N+1 degenerate eigenvectors are genuine ground states with S_tot = S_max Fermion/JordanWigner/Hubbard/WeakNagaokaGlobalMin.lean
mulVec_eq_smul_of_rayleighOnVec_eq_min Hermitian variational equality: a nonzero vector attaining the minimum-eigenvalue Rayleigh bound is a minimum eigenvector (companion to the variational lower bound) Quantum/SpinS/HermitianVariationalEquality.lean
Lean name Statement File
hubbardKineticOnGraph N G J spinful Hubbard kinetic operator from a SimpleGraph G and edge weight J Fermion/JordanWigner.lean
hubbardKineticOnGraph_commute_fermionTotalNumber / hubbardKineticOnGraph_isHermitian charge conservation always; Hermiticity for real J Fermion/JordanWigner/Hubbard/Graph.lean
hubbardHamiltonianOnGraph N G J U full Hubbard Hamiltonian from a graph + on-site coupling Fermion/JordanWigner.lean
hubbardHamiltonianOnGraph_commute_fermionTotalNumber / hubbardHamiltonianOnGraph_isHermitian charge conservation; Hermiticity for real J and real U Fermion/JordanWigner/Hubbard/Graph.lean
hubbardChainHamiltonian N J U the canonical 1D nearest-neighbour Hubbard chain −J Σ_{σ,⟨i,j⟩} c_{iσ}† c_{jσ} + U Σ_i n_{i↑} n_{i↓} (built from pathGraph (N+1)) Fermion/JordanWigner.lean
hubbardChainHamiltonian_isHermitian / hubbardChainHamiltonian_commute_fermionTotalNumber Hermiticity (real J, U) and charge conservation Fermion/JordanWigner/Hubbard/Graph.lean
hubbardHamiltonianOnGraph_mulVec_vacuum / hubbardChainHamiltonian_mulVec_vacuum both graph-built Hubbard Hamiltonians annihilate the JW vacuum Fermion/JordanWigner/Hubbard/Graph.lean
hubbardChainGibbsState N β J U Gibbs state of the 1D Hubbard chain Fermion/JordanWigner.lean
hubbardChainGibbsState_isHermitian / hubbardChainGibbsState_commute_hamiltonian Hermiticity (real J, U) and commute with the Hamiltonian Fermion/JordanWigner/Hubbard/Graph.lean
hubbardCycleGibbsState_commute_hamiltonian the periodic Hubbard Gibbs state commutes with the periodic Hubbard Hamiltonian (companion of the open-chain version, free corollary of gibbsState_commute_hamiltonian) Fermion/JordanWigner/Hubbard/Graph.lean
hubbardChainGibbsExpectation_zero / _im_of_isHermitian / _commutator_hamiltonian / _hamiltonian_im / _hamiltonian_pow_im / hubbardChain_partitionFn_im / _ofReal_re_eq / hubbardChainGibbsState_pow_trace open-chain Hubbard expectation companions (β = 0 closed form, Hermitian-observable real, conservation, energy / energy-power expectations real, partition function real, real-cast, Rényi-n trace) Fermion/JordanWigner/Hubbard/Graph.lean
hubbardCycleGibbsExpectation_zero / _im_of_isHermitian / _commutator_hamiltonian / _hamiltonian_im / _hamiltonian_pow_im / hubbardCycle_partitionFn_im / _ofReal_re_eq / hubbardCycleGibbsState_pow_trace periodic Hubbard chain expectation companions (same family as the open chain) Fermion/JordanWigner/Hubbard/Graph.lean

Nagaoka’s theorem on a magnetization sector (Tasaki §11.2.2, Theorem 11.7 / Lemma 11.9)

Lean name Statement File
tasakiEffReMatrixOnSector_ground_finrank_le_one Theorem 11.7 core (Tasaki, 1st ed. (2020), §11.2.2, Theorem 11.7, p. 385; PR #5149): on a non-empty magnetization sector whose Perron–Frobenius companion nagaokaPFMatrixOnSector is irreducible (the per-sector instance of the Definition 11.6 connectivity condition nagaokaConnectivity), the sector effective matrix tasakiEffReMatrixOnSector has a strictly positive eigenvector at the eigenvalue −μ (μ = the Perron eigenvalue of the negated matrix) and that eigenspace has finrank ≤ 1. Since −μ = min spec M\|_sector, this is the non-degeneracy of the sector ground state that Nagaoka’s theorem asserts Fermion/JordanWigner/Hubbard/NagaokaMagnetizationSector.lean
reachSwapOff_of_exchangeReachable Lemma 11.9: spin-swap generation along exchange bonds (Tasaki §11.2.2, proof of Lemma 11.9 and footnote 13, p. 387; PR #5149): if x ≠ y are joined in the exchange-bond graph of nagaokaBondGraph, then the transposition {x, y} is reachable off a finite site set — from every hole position avoiding the support of the connecting exchange-bond walk, (p, σ) reaches (p, swap_{x,y} σ). Under ConnectedByExchangeBonds the hypothesis holds for every pair, which is exactly the step “we can generate any permutation of spin configurations by successive exchanges on the exchange bonds” that Tasaki’s footnote 13 defers to the property (iii) of p. 41 Fermion/JordanWigner/Hubbard/NagaokaStateQuiverReach.lean

General flat-band ground states: the annihilation peel behind eq. (11.3.46) (Tasaki §11.3.4)

Lean name Statement File
spinfulAnnihilation_rangeT_mulVec_eq_zero_of_mem_groundSubmodule Range-T annihilation of a ground-submodule vector (Tasaki, 1st ed. (2020), §11.3.4, premise of eq. (11.3.46), p. 412; PR #5149): for PSD hopping T and U > 0, every Φ in generalFlatBandGroundSubmodule T U is killed by every smeared annihilator built from the rows of a factorisation T = Cᴴ · C, i.e. Ĉ_σ(∑_k a_k C_k) Φ = 0 for all coefficient vectors a. This is the ground-submodule packaging of the Ĥ_hop Φ_GS = 0 step from which Tasaki concludes that a ground state is written only in terms of the ↠operators Fermion/JordanWigner/Hubbard/GeneralFlatBandGroundAnnihilation.lean
site_annihilation_mulVec_generalFlatBandSlaterState Site annihilator kills a Slater state vanishing at that site (Tasaki §11.3.4, peel step toward eq. (11.3.46), p. 412; PR #5149): if μ_q(z) = 0 for every mode q in the list, then ĉ_{z,σ} anticommutes through every ↠factor of generalFlatBandSlaterState μ qs (site-dual CAR with vanishing pairing) and annihilates the vacuum. On the index set I this is the biorthogonality statement for z ∈ I not among the listed modes Fermion/JordanWigner/Hubbard/GeneralFlatBandOperators.lean

← Multi-mode fermion via Jordan–Wigner (P2 backbone) · Catalogue · End →