The Higgs potential
Note Authors: Joseph Tooby-Smith
These notes are created using an interactive theorem
prover called Lean.
Lean formally checks definitions, theorems and proofs for correctness.
These notes are part of a much larger project called
PhysLean, which aims to digitalize
physics into Lean. Please consider contributing to this project.
Please provide feedback or suggestions for improvements by creating a GitHub issue
here.
Table of content
1. Introduction
2. The Higgs field
3. The Higgs potential
3.1. Properties of the Higgs potential
3.2. Boundedness of the Higgs potential
3.3. Maximum and minimum of the Higgs potential
1. Introduction
The Higgs potential is a key part of the Standard Model of particle physics. It is a scalar potential which is used to give mass to the elementary particles. The Higgs potential is a polynomial of degree four in the Higgs field.
2. The Higgs field
The vector space HiggsVec
is defined to be the complex Euclidean space of dimension 2.
For a given spacetime point a Higgs field gives a value in HiggsVec
.
Show Lean code:
abbrev HiggsVec := EuclideanSpace ℂ (Fin 2)
The HiggsBundle
is defined as the trivial vector bundle with base SpaceTime
and
fiber HiggsVec
. Thus as a manifold it corresponds to ℝ⁴ × ℂ²
.
Show Lean code:
abbrev HiggsBundle := Bundle.Trivial SpaceTime HiggsVec
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
.
Show Lean code:
abbrev HiggsField : Type := ContMDiffSection SpaceTime.asSmoothManifold HiggsVec ⊤ HiggsBundle
3. The Higgs potential
The structure Potential
is defined with two fields, μ2
corresponding
to the mass-squared of the Higgs boson, and l
corresponding to the coefficent
of the quartic term in the Higgs potential. Note that l
is usually denoted λ
.
Show Lean code:
structure Potential where
/-- The mass-squared of the Higgs boson. -/
μ2 : ℝ
/-- The quartic coupling of the Higgs boson. Usually denoted λ. -/
𝓵 : ℝ
Given an element φ
of HiggsField
, normSq φ
is defined as the
the function SpaceTime → ℝ
obtained by taking the square norm of the
pointwise Higgs vector. In other words, normSq φ x = ‖φ x‖ ^ 2
.
The notation ‖φ‖_H^2
is used for the normSq φ
.
Show Lean code:
def normSq (φ : HiggsField) : SpaceTime → ℝ := fun x => ‖φ x‖ ^ 2
Given a element P
of Potential
, P.toFun
is Higgs potential.
It is defined for a Higgs field φ
and a spacetime point x
as
-μ² ‖φ‖_H^2 x + l * ‖φ‖_H^2 x * ‖φ‖_H^2 x
.
Show Lean code:
def toFun (φ : HiggsField) (x : SpaceTime) : ℝ :=
- P.μ2 * ‖φ‖_H^2 x + P.𝓵 * ‖φ‖_H^2 x * ‖φ‖_H^2 x
3.1. Properties of the Higgs potential
For an element P
of Potential
, if l < 0
then the following upper bound for the potential
exists
P.toFun φ x ≤ - μ2 ^ 2 / (4 * 𝓵)
.
Show Lean code:
lemma neg_𝓵_quadDiscrim_zero_bound (h : P.𝓵 < 0) (φ : HiggsField) (x : SpaceTime) :
P.toFun φ x ≤ - P.μ2 ^ 2 / (4 * P.𝓵) := by
have h1 := P.quadDiscrim_nonneg (ne_of_lt h) φ x
simp only [quadDiscrim, discrim, even_two, Even.neg_pow] at h1
ring_nf at h1
rw [← neg_le_iff_add_nonneg',
show P.𝓵 * P.toFun φ x * 4 = (- 4 * P.𝓵) * (- P.toFun φ x) by ring] at h1
have h2 := le_neg_of_le_neg <| (div_le_iff₀' (by linarith : 0 < - 4 * P.𝓵)).mpr h1
ring_nf at h2 ⊢
exact h2
For an element P
of Potential
, if 0 < l
then the following lower bound for the potential
exists
- μ2 ^ 2 / (4 * 𝓵) ≤ P.toFun φ x
.
Show Lean code:
lemma pos_𝓵_quadDiscrim_zero_bound (h : 0 < P.𝓵) (φ : HiggsField) (x : SpaceTime) :
- P.μ2 ^ 2 / (4 * P.𝓵) ≤ P.toFun φ x := by
have h1 := P.neg.neg_𝓵_quadDiscrim_zero_bound (by simpa [neg] using h) φ x
simp only [toFun_neg, μ2_neg, even_two, Even.neg_pow, 𝓵_neg, mul_neg, neg_div_neg_eq] at h1
rw [neg_le, neg_div'] at h1
exact h1
For an element P
of Potential
with l < 0
and a real c : ℝ
, there exists
a Higgs field φ
and a spacetime point x
such that P.toFun φ x = c
iff one of the
following two conditions hold:
0 < μ2
andc ≤ 0
. That is, ifl
is negative andμ2
positive, then the potential takes every non-positive value.- or
μ2 ≤ 0
andc ≤ - μ2 ^ 2 / (4 * 𝓵)
. That is, ifl
is negative andμ2
non-positive, then the potential takes every value less then or equal to its bound.
Show Lean code:
lemma neg_𝓵_sol_exists_iff (h𝓵 : P.𝓵 < 0) (c : ℝ) : (∃ φ x, P.toFun φ x = c) ↔ (0 < P.μ2 ∧ c ≤ 0) ∨
(P.μ2 ≤ 0 ∧ c ≤ - P.μ2 ^ 2 / (4 * P.𝓵)) := by
refine Iff.intro (fun ⟨φ, x, hV⟩ => ?_) (fun h => ?_)
· rcases P.neg_𝓵_toFun_neg h𝓵 φ x with hr | hr
· rw [← hV]
exact Or.inl hr
· rw [← hV]
exact Or.inr (And.intro hr (P.neg_𝓵_quadDiscrim_zero_bound h𝓵 φ x))
· simp only [toFun, neg_mul]
simp only [← sub_eq_zero, sub_zero]
ring_nf
let a := (P.μ2 - Real.sqrt (discrim P.𝓵 (- P.μ2) (- c))) / (2 * P.𝓵)
have ha : 0 ≤ a := by
simp only [discrim, even_two, Even.neg_pow, mul_neg, sub_neg_eq_add, a]
rw [div_nonneg_iff]
refine Or.inr (And.intro ?_ ?_)
· rw [sub_nonpos]
by_cases hμ : P.μ2 < 0
· have h1 : 0 ≤ √(P.μ2 ^ 2 + 4 * P.𝓵 * c) := Real.sqrt_nonneg (P.μ2 ^ 2 + 4 * P.𝓵 * c)
linarith
· refine Real.le_sqrt_of_sq_le ?_
rw [le_add_iff_nonneg_right]
refine mul_nonneg_of_nonpos_of_nonpos ?_ ?_
· refine mul_nonpos_of_nonneg_of_nonpos ?_ ?_
· linarith
· linarith
· rcases h with h | h
· linarith
· have h1 : P.μ2 = 0 := by linarith
rw [h1] at h
simpa using h.2
· linarith
use (ofReal a)
use 0
rw [ofReal_normSq ha]
trans P.𝓵 * a * a + (- P.μ2) * a + (- c)
· ring
have hd : 0 ≤ (discrim P.𝓵 (- P.μ2) (-c)) := by
simp only [discrim, even_two, Even.neg_pow, mul_neg, sub_neg_eq_add]
rcases h with h | h
· refine Left.add_nonneg (sq_nonneg P.μ2) ?_
refine mul_nonneg_of_nonpos_of_nonpos ?_ h.2
linarith
· rw [← @neg_le_iff_add_nonneg']
rw [← le_div_iff_of_neg']
· exact h.2
· linarith
have hdd : discrim P.𝓵 (- P.μ2) (-c) = Real.sqrt (discrim P.𝓵 (- P.μ2) (-c))
* Real.sqrt (discrim P.𝓵 (- P.μ2) (-c)) := by
exact (Real.mul_self_sqrt hd).symm
rw [mul_assoc]
refine (quadratic_eq_zero_iff (ne_of_gt h𝓵).symm hdd _).mpr ?_
simp only [neg_neg, or_true, a]
For an element P
of Potential
with 0 < l
and a real c : ℝ
, there exists
a Higgs field φ
and a spacetime point x
such that P.toFun φ x = c
iff one of the
following two conditions hold:
μ2 < 0
and0 ≤ c
. That is, ifl
is positive andμ2
negative, then the potential takes every non-negative value.- or
0 ≤ μ2
and- μ2 ^ 2 / (4 * 𝓵) ≤ c
. That is, ifl
is positive andμ2
non-negative, then the potential takes every value greater then or equal to its bound.
Show Lean code:
lemma pos_𝓵_sol_exists_iff (h𝓵 : 0 < P.𝓵) (c : ℝ) : (∃ φ x, P.toFun φ x = c) ↔ (P.μ2 < 0 ∧ 0 ≤ c) ∨
(0 ≤ P.μ2 ∧ - P.μ2 ^ 2 / (4 * P.𝓵) ≤ c) := by
have h1 := P.neg.neg_𝓵_sol_exists_iff (by simpa using h𝓵) (- c)
simp only [toFun_neg, neg_inj, μ2_neg, Left.neg_pos_iff, Left.neg_nonpos_iff, even_two,
Even.neg_pow, 𝓵_neg, mul_neg, neg_div_neg_eq] at h1
rw [neg_le, neg_div'] at h1
exact h1
3.2. Boundedness of the Higgs potential
Given a element P
of Potential
, the proposition IsBounded P
is true if and only if
there exists a real c
such that for all Higgs fields φ
and spacetime points x
,
the Higgs potential corresponding to φ
at x
is greater then or equal toc
. I.e.
∀ Φ x, c ≤ P.toFun Φ x
.
Show Lean code:
def IsBounded : Prop :=
∃ c, ∀ Φ x, c ≤ P.toFun Φ x
Given a element P
of Potential
which is bounded,
the quartic coefficent 𝓵
of P
is non-negative.
Show Lean code:
lemma isBounded_𝓵_nonneg (h : P.IsBounded) : 0 ≤ P.𝓵 := by
by_contra hl
rw [not_le] at hl
obtain ⟨c, hc⟩ := h
by_cases hμ : P.μ2 ≤ 0
· by_cases hcz : c ≤ - P.μ2 ^ 2 / (4 * P.𝓵)
· have hcm1 : ∃ φ x, P.toFun φ x = c - 1 := by
rw [P.neg_𝓵_sol_exists_iff hl (c - 1)]
apply Or.inr
simp_all
linarith
obtain ⟨φ, x, hφ⟩ := hcm1
have hc2 := hc φ x
rw [hφ] at hc2
linarith
· rw [not_le] at hcz
have hcm1 : ∃ φ x, P.toFun φ x = - P.μ2 ^ 2 / (4 * P.𝓵) - 1 := by
rw [P.neg_𝓵_sol_exists_iff hl _]
apply Or.inr
simp_all
obtain ⟨φ, x, hφ⟩ := hcm1
have hc2 := hc φ x
rw [hφ] at hc2
linarith
· rw [not_le] at hμ
by_cases hcz : c ≤ 0
· have hcm1 : ∃ φ x, P.toFun φ x = c - 1 := by
rw [P.neg_𝓵_sol_exists_iff hl (c - 1)]
apply Or.inl
simp_all
linarith
obtain ⟨φ, x, hφ⟩ := hcm1
have hc2 := hc φ x
rw [hφ] at hc2
linarith
· have hcm1 : ∃ φ x, P.toFun φ x = 0 := by
rw [P.neg_𝓵_sol_exists_iff hl 0]
apply Or.inl
simp_all
obtain ⟨φ, x, hφ⟩ := hcm1
have hc2 := hc φ x
rw [hφ] at hc2
linarith
Given a element P
of Potential
with 0 < 𝓵
, then the potential is bounded.
Show Lean code:
lemma isBounded_of_𝓵_pos (h : 0 < P.𝓵) : P.IsBounded := by
simp only [IsBounded]
have h2 := P.pos_𝓵_quadDiscrim_zero_bound h
by_contra hn
simp only [not_exists, not_forall, not_le] at hn
obtain ⟨φ, x, hx⟩ := hn (-P.μ2 ^ 2 / (4 * P.𝓵))
have h2' := h2 φ x
linarith
3.3. Maximum and minimum of the Higgs potential
Given an element P
of Potential
with l < 0
, then the Higgs field φ
and
spacetime point x
maximizes the potential if and only if one of the following conditions
holds
μ2 ≤ 0
and‖φ‖_H^2 x = μ2 / (2 * 𝓵)
.- or
0 < μ2
andφ x = 0
.
Show Lean code:
lemma isMaxOn_iff_field_of_𝓵_neg (h𝓵 : P.𝓵 < 0) (φ : HiggsField) (x : SpaceTime) :
IsMaxOn (fun (φ, x) => P.toFun φ x) Set.univ (φ, x) ↔
(P.μ2 ≤ 0 ∧ ‖φ‖_H^2 x = P.μ2 /(2 * P.𝓵)) ∨ (0 < P.μ2 ∧ φ x = 0) := by
rw [P.isMaxOn_iff_isMinOn_neg,
P.neg.isMinOn_iff_field_of_𝓵_pos (by simpa using h𝓵)]
simp
Given an element P
of Potential
with 0 < l
, then the Higgs field φ
and
spacetime point x
minimize the potential if and only if one of the following conditions
holds
0 ≤ μ2
and‖φ‖_H^2 x = μ2 / (2 * 𝓵)
.- or
μ2 < 0
andφ x = 0
.
Show Lean code:
theorem isMinOn_iff_field_of_𝓵_pos (h𝓵 : 0 < P.𝓵) (φ : HiggsField) (x : SpaceTime) :
IsMinOn (fun (φ, x) => P.toFun φ x) Set.univ (φ, x) ↔
(0 ≤ P.μ2 ∧ ‖φ‖_H^2 x = P.μ2 /(2 * P.𝓵)) ∨ (P.μ2 < 0 ∧ φ x = 0) := by
by_cases hμ2 : 0 ≤ P.μ2
· simpa [not_lt.mpr hμ2, hμ2] using P.isMinOn_iff_field_of_μSq_nonneg_𝓵_pos h𝓵 hμ2 φ x
· simpa [hμ2, lt_of_not_ge hμ2] using P.isMinOn_iff_field_of_μSq_nonpos_𝓵_pos h𝓵 (by linarith) φ x