From c7f70d5a2c605a14ee27b7e5ff9fda29b51f943f Mon Sep 17 00:00:00 2001 From: Nikolai Kashcheev Date: Sun, 15 Nov 2026 05:33:19 -0500 Subject: [PATCH] feat(StandardModel): characterize local Higgs gauge orbits Replace TODO 6V2NX with a local orbit criterion for Higgs doublets: away from zeros, equality of pointwise squared norms is equivalent to a weak SU(2) transformation with the same regularity as the fields. Document why the criterion cannot be extended across arbitrary zeros. --- .../StandardModel/HiggsBoson/API-map.md | 34 ++-- .../StandardModel/HiggsBoson/Basic.lean | 157 ++++++++++++++++-- 2 files changed, 160 insertions(+), 31 deletions(-) diff --git a/Physlib/Particles/StandardModel/HiggsBoson/API-map.md b/Physlib/Particles/StandardModel/HiggsBoson/API-map.md index 3761cb2c24..5352e89df2 100644 --- a/Physlib/Particles/StandardModel/HiggsBoson/API-map.md +++ b/Physlib/Particles/StandardModel/HiggsBoson/API-map.md @@ -2,9 +2,9 @@ ## Abstract (Alpha) -A Higgs field is a smooth section of the Higgs bundle over spacetime, on which the global Standard -Model gauge group acts; for now the bundle is the trivial one, whose fibre is the two-dimensional -complex vector space of pointwise Higgs values. The pointwise values carry an orthonormal basis, +A Higgs field in this module is a real analytic section of the Higgs bundle over spacetime. The +bundle is currently trivial, with fibre the two-dimensional complex vector space of pointwise Higgs +values. The pointwise values carry an orthonormal basis, real scalar coordinates, and a gauge representation whose orbits are classified by the norm, while fields carry constant fields, smooth components, a pointwise inner product and norm squared, the Standard Model Higgs potential with its minimization theory, and invariant effective potentials on @@ -33,18 +33,23 @@ The Higgs bundle over spacetime {3} is a smooth vector bundle, currently the tri fibre the pointwise Higgs values. [8] It is the bundle associated to the gauge group, rather than the trivial one. [TODO] -Higgs fields are smooth sections of the Higgs bundle; each has an underlying map to pointwise -Higgs values, and every Higgs vector gives a constant field. [9] The components of a Higgs field -are smooth, real and imaginary parts alike. [10] The pointwise inner product of two Higgs fields -has an explicit expansion, is conjugate symmetric and additive, and is smooth. [11] The pointwise +Each real analytic section has an underlying map to pointwise Higgs values, and every Higgs vector +gives a constant field. [9] Its real and imaginary components are real analytic. [10] The pointwise +inner product of two Higgs fields has an explicit expansion, is conjugate symmetric and additive, +and is smooth. [11] The pointwise norm squared of a Higgs field is the real part of its inner product with itself, is nonnegative, and is smooth. [12] -Gauge transformations act on Higgs fields pointwise; two Higgs fields lie in the same orbit -exactly when their pointwise norms squared agree, and every smooth nonnegative function on -spacetime is the norm squared of some Higgs field. [13][TODO] Since the gauge group preserves -inner products and norms of Higgs vectors, the pointwise inner product and norm squared of Higgs -fields are gauge invariant. [5] +In a trivialization, an explicit weak SU(2) rotation sends each nonzero doublet to the neutral +representative whose second component is its norm. This is the scalar part of unitary gauge and also +works when the norm varies across spacetime. On a region where one of two doublets is nonzero, their +pointwise squared norms agree exactly when one weak SU(2) transformation carries one to the other. +The matrix coefficients have the same regularity as the input fields: finite differentiability, +smoothness, or real analyticity. At common zeros, equal norms alone do not guarantee a continuous +transformation. [13] + +The general gauge action on sections and the characterization of possible squared norm functions +are separate formalization tasks. [13][TODO] ### The Higgs potential @@ -100,8 +105,9 @@ gauge connection over spacetime. [TODO] 12. Physlib/Particles/StandardModel/HiggsBoson/Basic.lean — `normSq`, `inner_self_eq_normSq`, `normSq_eq_inner_self_re`, `normSq_expand`, `normSq_nonneg`, `normSq_zero`, `normSq_smooth`, `const_normSq` -13. Physlib/Particles/StandardModel/HiggsBoson/Basic.lean — `gaugeAction`, `guage_orbit`, - `gauge_orbit_surject` +13. Physlib/Particles/StandardModel/HiggsBoson/Basic.lean — `toRealGroupElem_toSU2_apply_self`, + `contDiffOn_toRealGroupElem_toSU2`, `exists_contDiffOn_su2_rep_eq_iff_norm_sq_eq`, + `exists_analytic_su2_rep_eq_iff_normSq_eq`, `gaugeAction`, `gauge_orbit_surject` 14. Physlib/Particles/StandardModel/HiggsBoson/Potential.lean — `Potential`, `toFun`, `toFun_smooth`, `neg`, `toFun_neg`, `μ2_neg`, `𝓵_neg`, `toFun_zero`, `toFun_eq_of_repGaugeGroupI` diff --git a/Physlib/Particles/StandardModel/HiggsBoson/Basic.lean b/Physlib/Particles/StandardModel/HiggsBoson/Basic.lean index 0ea827f5d4..c297c76322 100644 --- a/Physlib/Particles/StandardModel/HiggsBoson/Basic.lean +++ b/Physlib/Particles/StandardModel/HiggsBoson/Basic.lean @@ -1,20 +1,23 @@ /- Copyright (c) 2024 Joseph Tooby-Smith. All rights reserved. Released under Apache 2.0 license as described in the file LICENSE. -Authors: Joseph Tooby-Smith +Authors: Nikolai Kashcheev, Joseph Tooby-Smith -/ module public import Physlib.Particles.StandardModel.GaugeGroup.Unbroken public import Mathlib.RepresentationTheory.Stabilizer public import Mathlib.Geometry.Manifold.VectorBundle.ContMDiffSection +import Mathlib.Analysis.InnerProductSpace.Calculus /-! # The Higgs field +The Higgs vector space, its gauge representation, and the space of analytic Higgs fields. + ## i. Overview -The Higgs field describes is the underlying field of the Higgs boson. +The Higgs field is the underlying field of the Higgs boson. It is a map from SpaceTime to a 2-dimensional complex vector space. In this module we define the Higgs field and prove some basic properties. @@ -23,9 +26,12 @@ In this module we define the Higgs field and prove some basic properties. - `HiggsVec`: The 2-dimensional complex vector space which is the target space of the Higgs field. This vector space is equipped with an action of the global gauge group of the Standard Model. - `HiggsBundle`: The trivial vector bundle over `SpaceTime` with fiber `HiggsVec`. -- `HiggsField`: The type of smooth sections of the `HiggsBundle`, i.e., the type of Higgs fields. +- `HiggsField`: The type of real analytic sections of the `HiggsBundle`. - `stabilizer_ofReal`: The stability group of the Higgs vacuum `(0, √a)`, i.e. the gauge group left unbroken by it, is `SU(3) × U(1)_EM`. +- `HiggsVec.toRealGroupElem_toSU2_apply_self`: The weak rotation to the neutral representative. +- `HiggsVec.exists_contDiffOn_su2_rep_eq_iff_norm_sq_eq`: The squared norm determines the local + weak gauge orbit of a nonzero Higgs doublet, at any order of differentiability. ## iii. Table of contents @@ -40,6 +46,7 @@ In this module we define the Higgs field and prove some basic properties. - A.5. The Gauge orbit of a Higgs vector - A.5.1. The rotation matrix to ofReal - A.5.2. Members of orbits + - A.5.3. Local gauge orbits of Higgs doublets - A.6. The stability group of a Higgs vector - A.7. The electromagnetic U(1) and the phase of the first component - A.8. To real scalars @@ -80,6 +87,7 @@ open Matrix open Complex open ComplexConjugate open SpaceTime +open scoped ContDiff /-! @@ -294,6 +302,54 @@ lemma toRealGroupElem_apply_self (φ : HiggsVec) : rw [h0'] ring +private lemma toRealGroupElem_toSU2_of_ne_zero (v : HiggsVec) (hv : v ≠ 0) : + (toRealGroupElem v).toSU2.val = + !![v 1 / ‖v‖, -v 0 / ‖v‖; conj (v 0) / ‖v‖, conj (v 1) / ‖v‖] := by + simp [toRealGroupElem, hv, GaugeGroupI.toSU2] + +private lemma toRealGroupElem_eq_toSU2 (v : HiggsVec) : + toRealGroupElem v = (1, (toRealGroupElem v).toSU2, 1) := by + by_cases hv : v = 0 + · simp [toRealGroupElem, hv, GaugeGroupI.toSU2] + rfl + · simp [toRealGroupElem, hv, GaugeGroupI.toSU2] + +/-- The weak factor of `toRealGroupElem` alone rotates the Higgs doublet to +the neutral representative `(0, ‖v‖)`, as used in unitary gauge. -/ +lemma toRealGroupElem_toSU2_apply_self (v : HiggsVec) : + repGaugeGroupI (1, (toRealGroupElem v).toSU2, 1) v = ofReal (‖v‖ ^ 2) := by + rw [← toRealGroupElem_eq_toSU2 v, toRealGroupElem_apply_self] + +/-- The matrix entries of the explicit weak rotation to the neutral representative +are `C^n` away from the zero vector, including for `n = ∞` (smooth) and `n = ω` (analytic). -/ +lemma contDiffOn_toRealGroupElem_toSU2 {n : ℕ∞ω} : + ContDiffOn ℝ n (fun (v : HiggsVec) (i j : Fin 2) => (toRealGroupElem v).toSU2.val i j) + {v | v ≠ 0} := by + refine contDiffOn_pi.mpr fun i => contDiffOn_pi.mpr fun j => ?_ + let s : Set HiggsVec := {v | v ≠ 0} + have hn : ContDiffOn ℝ n (fun v : HiggsVec => ‖v‖) s := + (contDiff_id : ContDiff ℝ n (id : HiggsVec → HiggsVec)).contDiffOn.norm ℂ fun _ hv => hv + have hnC : ContDiffOn ℝ n (fun v : HiggsVec => (‖v‖ : ℂ)) s := + Complex.ofRealCLM.contDiff.comp_contDiffOn hn + have hn0 (v : HiggsVec) (hv : v ∈ s) : (‖v‖ : ℂ) ≠ 0 := by + exact_mod_cast norm_ne_zero_iff.mpr hv + have hInv : ContDiffOn ℝ n (fun v : HiggsVec => (‖v‖ : ℂ)⁻¹) s := hnC.inv hn0 + have hdiv {f : HiggsVec → ℂ} (hf : ContDiffOn ℝ n f s) : + ContDiffOn ℝ n (fun v => f v / (‖v‖ : ℂ)) s := by + simpa only [div_eq_mul_inv] using hf.mul hInv + have hi (k : Fin 2) : ContDiffOn ℝ n (fun v : HiggsVec => v k) s := + (contDiff_piLp 2 |>.mp contDiff_id k).contDiffOn + have hc (k : Fin 2) : ContDiffOn ℝ n (fun v : HiggsVec => conj (v k) / (‖v‖ : ℂ)) s := by + have hconj : ContDiffOn ℝ n (fun v : HiggsVec => conj (v k)) s := by + refine (Complex.conjCLE.contDiff.comp_contDiffOn (hi k)).congr fun v _ => ?_ + exact (Complex.conjCLE_apply (v k)).symm + exact hdiv hconj + fin_cases i <;> fin_cases j + · exact (hdiv (hi 1)).congr (fun v hv => by rw [toRealGroupElem_toSU2_of_ne_zero v hv]; rfl) + · exact (hdiv (hi 0).neg).congr (fun v hv => by rw [toRealGroupElem_toSU2_of_ne_zero v hv]; rfl) + · exact (hc 0).congr (fun v hv => by rw [toRealGroupElem_toSU2_of_ne_zero v hv]; rfl) + · exact (hc 1).congr (fun v hv => by rw [toRealGroupElem_toSU2_of_ne_zero v hv]; rfl) + /-! #### A.5.2. Members of orbits @@ -318,6 +374,69 @@ lemma exists_repGaugeGroupI_eq_iff_norm_eq (φ : HiggsVec) (ψ : HiggsVec) : /-! +#### A.5.3. Local gauge orbits of Higgs doublets + +In a trivialization, a doublet field is a map into `HiggsVec`. The weak rotation +to the neutral representative preserves the regularity of the field wherever +it is nonzero. The following results concern the action on the doublet itself. + +-/ + +private lemma contDiffOn_weakAlignment {E : Type*} [NormedAddCommGroup E] [NormedSpace ℝ E] + {n : ℕ∞ω} {s : Set E} {φ ψ : E → HiggsVec} (hφ : ContDiffOn ℝ n φ s) (hψ : ContDiffOn ℝ n ψ s) + (hnφ : ∀ x ∈ s, φ x ≠ 0) (hnψ : ∀ x ∈ s, ψ x ≠ 0) : + ContDiffOn ℝ n (fun (x : E) (i j : Fin 2) => + ((toRealGroupElem (ψ x)).toSU2⁻¹ * (toRealGroupElem (φ x)).toSU2).val i j) s := by + have h (χ : E → HiggsVec) (hχ : ContDiffOn ℝ n χ s) (hnχ : ∀ x ∈ s, χ x ≠ 0) + (k l : Fin 2) : ContDiffOn ℝ n (fun x => (toRealGroupElem (χ x)).toSU2.val k l) s := + contDiffOn_pi.mp (contDiffOn_pi.mp (contDiffOn_toRealGroupElem_toSU2.comp hχ hnχ) k) l + refine contDiffOn_pi.mpr fun i => contDiffOn_pi.mpr fun j => ?_ + have hs (k : Fin 2) : + ContDiffOn ℝ n (fun x => conj ((toRealGroupElem (ψ x)).toSU2.val k i)) s := by + refine (Complex.conjCLE.contDiff.comp_contDiffOn (h ψ hψ hnψ k i)).congr fun x _ => ?_ + exact (Complex.conjCLE_apply _).symm + change ContDiffOn ℝ n (fun x : E => + (star (toRealGroupElem (ψ x)).toSU2.val * (toRealGroupElem (φ x)).toSU2.val) i j) s + simp only [Matrix.mul_apply, Matrix.star_apply] + exact ContDiffOn.sum fun k _ => (hs k).mul (h φ hφ hnφ k j) + +/-- Where the first doublet is nonzero, two `C^n` doublets have the same pointwise squared norm +exactly when a single weak `SU(2)` transformation with `C^n` matrix entries takes one to the other. +The transformation first rotates the source to the neutral representative using +`toRealGroupElem`, then applies the inverse rotation for the target. The regularity and +action are asserted only on the given set. + +Equal norms alone are not sufficient at common zeros: in two real coordinates, the +doublets `(x + iy, 0)` and `(x - iy, 0)` require the transforming matrix to act with opposite +signs on the first basis vector along the two axes, so the matrix cannot be continuous at +their intersection. -/ +lemma exists_contDiffOn_su2_rep_eq_iff_norm_sq_eq {E : Type*} + [NormedAddCommGroup E] [NormedSpace ℝ E] + {n : ℕ∞ω} {s : Set E} {φ ψ : E → HiggsVec} (hφ : ContDiffOn ℝ n φ s) (hψ : ContDiffOn ℝ n ψ s) + (hnφ : ∀ x ∈ s, φ x ≠ 0) : + (∃ g : E → specialUnitaryGroup (Fin 2) ℂ, + ContDiffOn ℝ n (fun x (i j : Fin 2) => (g x).val i j) s ∧ + ∀ x ∈ s, repGaugeGroupI (1, g x, 1) (φ x) = ψ x) ↔ + ∀ x ∈ s, ‖φ x‖ ^ 2 = ‖ψ x‖ ^ 2 := by + constructor + · rintro ⟨g, _, hg⟩ x hx + simpa only [repGaugeGroupI_norm] using congrArg (fun v : HiggsVec => ‖v‖ ^ 2) (hg x hx) + · intro heq + have hnψ (x : E) (hx : x ∈ s) : ψ x ≠ 0 := by + intro h + have hz : ‖φ x‖ ^ 2 = 0 := by simpa [h] using heq x hx + exact hnφ x hx (norm_eq_zero.mp (sq_eq_zero_iff.mp hz)) + let g (x : E) := (toRealGroupElem (ψ x)).toSU2⁻¹ * (toRealGroupElem (φ x)).toSU2 + refine ⟨g, contDiffOn_weakAlignment hφ hψ hnφ hnψ, fun x hx => ?_⟩ + have hg : ((1, g x, 1) : GaugeGroupI) = + ((1, (toRealGroupElem (ψ x)).toSU2, 1) : GaugeGroupI)⁻¹ * + (1, (toRealGroupElem (φ x)).toSU2, 1) := by simp [g] + rw [hg, repGaugeGroupI_mul_apply, repGaugeGroupI_inv_apply_eq_iff, + toRealGroupElem_toSU2_apply_self, toRealGroupElem_toSU2_apply_self] + exact congrArg ofReal (heq x hx) + +/-! + ### A.6. The stability group of a Higgs vector The stability group of a Higgs vector `φ` is the subgroup `repGaugeGroupI.stabilizer φ` of @@ -543,16 +662,14 @@ instance : ContMDiffVectorBundle ⊤ HiggsVec HiggsBundle (Lorentz.Vector.asSmoo ## C. The Higgs fields -Higgs fields are smooth sections of the Higgs bundle. -This corresponds to smooth maps from `SpaceTime` to `HiggsVec`. +`HiggsField` consists of real analytic sections of the trivial Higgs bundle. +These correspond to real analytic maps from `SpaceTime` to `HiggsVec`. We here define the type of Higgs fields and create an API around them. -/ -/-- The type `HiggsField` is defined such that elements are smooth sections of the trivial - vector bundle `HiggsBundle`. Such elements are Higgs fields. Since `HiggsField` is - trivial as a vector bundle, a Higgs field is equivalent to a smooth map - from `SpaceTime` to `HiggsVec`. -/ +/-- Real analytic sections of the trivial Higgs bundle, equivalently real analytic maps from +`SpaceTime` to `HiggsVec`. The regularity `⊤` here means `ω`, not ordinary `C^∞`. -/ abbrev HiggsField : Type := ContMDiffSection (Lorentz.Vector.asSmoothManifold 3) HiggsVec ⊤ HiggsBundle @@ -811,22 +928,28 @@ lemma const_normSq (φ : HiggsVec) (x : SpaceTime) : ### C.5. The action of the gauge group on Higgs fields -The results in this section are currently informal. +For `HiggsField`, the local result in A.5.3 gives a weak gauge transformation with real analytic +matrix entries. This is the scalar part of unitary gauge in the trivial Higgs bundle. +The general action of the smooth gauge group on sections is a separate task. -/ +/-- The `HiggsField` specialization of `HiggsVec.exists_contDiffOn_su2_rep_eq_iff_norm_sq_eq`. +The action and real analyticity of the matrix entries hold on the given region. -/ +lemma exists_analytic_su2_rep_eq_iff_normSq_eq (φ ψ : HiggsField) (s : Set SpaceTime) + (hφ : ∀ x ∈ s, φ x ≠ 0) : + (∃ g : SpaceTime → specialUnitaryGroup (Fin 2) ℂ, + ContDiffOn ℝ ω (fun x (i j : Fin 2) => (g x).val i j) s ∧ + ∀ x ∈ s, HiggsVec.repGaugeGroupI (1, g x, 1) (φ x) = ψ x) ↔ + ∀ x ∈ s, φ.normSq x = ψ.normSq x := + HiggsVec.exists_contDiffOn_su2_rep_eq_iff_norm_sq_eq + φ.contDiff.contDiffOn ψ.contDiff.contDiffOn hφ + /-- The action of `gaugeTransformI` on `HiggsField` acting pointwise through `HiggsVec.repGaugeGroupI`. -/ informal_definition gaugeAction where deps := [``gaugeTransformI] tag := "6V2NP" -/-- There exists a `g` in `gaugeTransformI` such that `gaugeAction g φ = φ'` iff -`φ(x)^† φ(x) = φ'(x)^† φ'(x)`. --/ -informal_lemma guage_orbit where - deps := [``gaugeAction] - tag := "6V2NX" - /-- For every smooth map `f` from `SpaceTime` to `ℝ` such that `f` is positive semidefinite, there exists a Higgs field `φ` such that `f = φ^† φ`. -/