Physlib

Physlib.Electromagnetism.Kinematics.GaugeTransformation

Gauge Transformations of the Electromagnetic Potential

i. Overview

In this module we define gauge transformations of the electromagnetic potential `A^μ ↦ A^μ + ∂^μ χ` where `χ : SpaceTime d → ℝ` is a smooth gauge function, and prove that the field strength tensor is invariant under such transformations.

The raised-index gradient `∂^μ χ := η^{μν} ∂_ν χ` is necessary because the bare covariant gradient `∂_μ χ` does not make `F^{μν}` invariant. The formal witness is `fieldStrengthMatrix_bareGradient_inl_inr` (§B.5), which computes a specific nonzero component of the field strength of a bare-gradient potential. The invariance theorem `toFieldStrength_gaugeTransform` doubles as a correctness test of `ofGradient`.

ii. Key results

- `ofGradient` : The pure-gauge potential `A^μ = η^{μν} ∂_ν χ` built from a gauge function `χ`. - `gaugeTransform` : The gauge transformation `A^μ ↦ A^μ + ∂^μ χ`. - `toFieldStrength_ofGradient` : A pure-gauge potential has vanishing field strength. - `toFieldStrength_gaugeTransform` : The field strength tensor is invariant under gauge transformations. - `fieldStrengthMatrix_gaugeTransform` : The field strength matrix is invariant under gauge transformations. - `gaugeTransform_gaugeTransform` : Composing two gauge shifts equals shifting by the sum; upgrades one-step F-invariance to invariance along any finite chain. - `ofGradient_equivariant` : `ofGradient` intertwines the Lorentz action with function composition. - `gaugeTransform_equivariant` : Gauge transformations commute with Lorentz transformations. - `fieldStrengthMatrix_bareGradient_inl_inr` : The `(inl 0, inr i)` field-strength component of the bare-gradient potential `χ(x) = x⁰·xⁱ` equals `2`; in particular the bare gradient does not give a gauge-invariant field strength (necessity of the metric contraction in `ofGradient`).

iii. Table of contents

- A. The pure-gauge potential - A.1. Definition and basic lemmas - A.2. Differentiability of the pure-gauge potential - A.3. Vanishing field strength of the pure-gauge potential - A.4. Lorentz equivariance of the pure-gauge potential - B. Gauge transformations - B.1. Definition and basic lemmas - B.2. Invariance of the field strength - B.3. Group structure of gauge shifts - B.4. Equivariance under Lorentz transformations - B.5. Necessity: bare gradient does not give gauge invariance

iv. References

  • https://en.wikipedia.org/wiki/Mathematical_descriptions_of_the_electromagnetic_field#Gauge_freedom

A. The pure-gauge potential

A.1. Definition and basic lemmas

A.2. Differentiability of the pure-gauge potential

A.3. Vanishing field strength of the pure-gauge potential

A.4. Lorentz equivariance of the pure-gauge potential

`ofGradient` intertwines the Lorentz action on potentials with composition by Λ⁻¹ on the gauge function: `Λ • ofGradient χ = ofGradient (χ ∘ (Λ⁻¹ • ·))`. The proof reduces to the metric-commutativity identity `Λ * η = η * (Λ⁻¹)ᵀ`, which is the defining property of the Lorentz group (`LorentzGroup.comm_minkowskiMatrix`).

B. Gauge transformations

B.1. Definition and basic lemmas

B.2. Invariance of the field strength

The key ingredient — that a pure-gauge potential has vanishing field strength — is proved in §A.3 (`toFieldStrength_ofGradient`).

B.3. Group structure of gauge shifts

Composing two gauge shifts by χ₁ and χ₂ is the same as shifting by χ₁ + χ₂. Together with `gaugeTransform_zero` this shows that the map `χ ↦ (A ↦ gaugeTransform χ A)` is a group action of the additive group of smooth functions. (We do not build the formal `MulAction` here — the composition lemma is the agreeable core, and the rest is straightforward from it.)

The `ofGradient` lemmas used here — `ofGradient_zero` and `ofGradient_add` — are proved in §A.1.

B.4. Equivariance under Lorentz transformations

The gauge-transformation map commutes with the Lorentz group action: applying Λ to a potential and then gauge-transforming by χ is the same as gauge-transforming by `χ ∘ (Λ⁻¹ • ·)` and then applying Λ. The proof delegates to `ofGradient_equivariant` (§A.4).

B.5. Necessity: bare gradient does not give gauge invariance

We exhibit a concrete gauge function `χ(x) = x⁰·xⁱ` whose **bare** covariant gradient `B^μ := ∂_μ χ` has a nonzero field-strength component (equal to `2`), certifying that the metric contraction in `ofGradient` is required for gauge invariance.

18 declarations

definition

Pure-gauge potential Aμ=μχA^\mu = \partial^\mu \chi

Given a gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R}, this definition constructs the pure-gauge electromagnetic potential AμA^\mu whose components are given by the raised-index gradient Aμ(x)=κημκκχ(x)A^\mu(x) = \sum_{\kappa} \eta_{\mu\kappa} \partial_\kappa \chi(x). Here η\eta is the Minkowski matrix diag(1,1,,1)\text{diag}(1, -1, \dots, -1) and κχ\partial_\kappa \chi is the partial derivative of χ\chi with respect to the κ\kappa-th coordinate.

theorem

Aμ(x)=κημκκχ(x)A^\mu(x) = \sum_{\kappa} \eta_{\mu\kappa} \partial_\kappa \chi(x)

For any gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R}, point xSpaceTimedx \in \text{SpaceTime}_d, and coordinate index μ\mu, the μ\mu-th component of the pure-gauge electromagnetic potential A=ofGradient χA = \text{ofGradient } \chi at xx is given by the sum Aμ(x)=κημκκχ(x)A^\mu(x) = \sum_{\kappa} \eta_{\mu\kappa} \partial_\kappa \chi(x) where η\eta is the Minkowski matrix and κχ\partial_\kappa \chi is the partial derivative of χ\chi with respect to the κ\kappa-th coordinate.

theorem

Component Evaluation of Pure-Gauge Potential Aμ(x)=ημμμχ(x)A^\mu(x) = \eta_{\mu\mu} \partial_\mu \chi(x)

For a gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R}, a point xx in spacetime, and a coordinate index μ\mu, the μ\mu-th component of the pure-gauge potential Aμ(x)A^\mu(x) is given by Aμ(x)=ημμμχ(x) A^\mu(x) = \eta_{\mu\mu} \partial_\mu \chi(x) where η=diag(1,1,,1)\eta = \text{diag}(1, -1, \dots, -1) is the Minkowski metric and μχ\partial_\mu \chi is the partial derivative of χ\chi with respect to the μ\mu-th coordinate. This formula reflects that because the Minkowski metric is diagonal, the sum Aμ=κημκκχA^\mu = \sum_{\kappa} \eta_{\mu\kappa} \partial_\kappa \chi collapses to a single term.

theorem

The pure-gauge potential of the zero gauge function is zero

For any point xSpaceTimedx \in \text{SpaceTime}_d and any coordinate index μ\mu, the pure-gauge potential AμA^\mu derived from the zero gauge function χ(x)=0\chi(x) = 0 is zero. That is, (Agauge)μ(x)=0(A_{\text{gauge}})^\mu(x) = 0 whenever χ=0\chi = 0.

theorem

ofGradient(χ1+χ2)=ofGradient(χ1)+ofGradient(χ2)\text{ofGradient}(\chi_1 + \chi_2) = \text{ofGradient}(\chi_1) + \text{ofGradient}(\chi_2)

For any two differentiable gauge functions χ1,χ2:SpaceTimedR\chi_1, \chi_2 : \text{SpaceTime}_d \to \mathbb{R}, the pure-gauge potential associated with their sum χ1+χ2\chi_1 + \chi_2 is equal to the sum of the individual pure-gauge potentials associated with χ1\chi_1 and χ2\chi_2: ofGradient(χ1+χ2)=ofGradient(χ1)+ofGradient(χ2)\text{ofGradient}(\chi_1 + \chi_2) = \text{ofGradient}(\chi_1) + \text{ofGradient}(\chi_2) where ofGradient(χ)\text{ofGradient}(\chi) denotes the pure-gauge potential AμA^\mu whose components are defined by the raised-index gradient Aμ(x)=ημννχ(x)A^\mu(x) = \eta^{\mu\nu} \partial_\nu \chi(x).

theorem

The Pure-Gauge Potential Aμ=μχA^\mu = \partial^\mu \chi is Differentiable for C2C^2 Gauge Functions

For any gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R} that is twice continuously differentiable (of class C2C^2), the pure-gauge electromagnetic potential AμA^\mu, defined by Aμ(x)=ημννχ(x)A^\mu(x) = \eta^{\mu\nu} \partial_\nu \chi(x) where η\eta is the Minkowski metric, is a differentiable function.

theorem

χCn+1\chi \in C^{n+1} implies the pure-gauge potential AμCnA^\mu \in C^n

Let dd be the dimension of spacetime and nn be a natural number. If a gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R} is (n+1)(n+1)-times continuously differentiable (of class Cn+1C^{n+1}), then the corresponding pure-gauge potential Aμ=μχA^\mu = \partial^\mu \chi (defined as Aμ=ημννχA^\mu = \eta^{\mu\nu} \partial_\nu \chi) is nn-times continuously differentiable (of class CnC^n).

theorem

Pure-Gauge Potential Aμ=μχA^\mu = \partial^\mu \chi has Vanishing Field Strength F=0F = 0

For any spatial dimension dd and any gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R} that is twice continuously differentiable (of class C2C^2), the field strength tensor FF associated with the pure-gauge potential Aμ=μχA^\mu = \partial^\mu \chi vanishes at every point xx in spacetime: F(x)=0 F(x) = 0 where the potential is defined by Aμ=ημννχA^\mu = \eta^{\mu\nu} \partial_\nu \chi and the field strength tensor by Fμν=μAννAμF^{\mu\nu} = \partial^\mu A^\nu - \partial^\nu A^\mu.

theorem

Lorentz Equivariance of the Pure-Gauge Potential: Λμχ=μ(χΛ1)\Lambda \cdot \partial^\mu \chi = \partial^\mu (\chi \circ \Lambda^{-1})

Let χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R} be a differentiable gauge function and Λ\Lambda be a Lorentz transformation. The Lorentz action on the pure-gauge potential Aμ=μχA^\mu = \partial^\mu \chi (where μχ=ημννχ\partial^\mu \chi = \eta^{\mu\nu} \partial_\nu \chi) is equivariant with respect to the transformation of the gauge function, such that Λ(μχ)=μ(χΛ1)\Lambda \cdot (\partial^\mu \chi) = \partial^\mu (\chi \circ \Lambda^{-1}).

definition

Gauge transformation AμAμ+μχA^\mu \mapsto A^\mu + \partial^\mu \chi

For a given spatial dimension dd, a gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R}, and an electromagnetic potential AA, the gauge transformation of AA by χ\chi is the potential defined by the sum A+μχA + \partial^\mu \chi. This represents the pointwise addition of the potential AA and the pure-gauge potential μχ(x)=νημννχ(x)\partial^\mu \chi(x) = \sum_{\nu} \eta^{\mu\nu} \partial_\nu \chi(x) constructed from the gradient of χ\chi and the Minkowski metric η\eta.

theorem

Pointwise evaluation of the gauge transformation: (gaugeTransform χA)(x)=A(x)+μχ(x)(\text{gaugeTransform } \chi A)(x) = A(x) + \partial^\mu \chi(x)

For a given spatial dimension dd, a gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R}, an electromagnetic potential AA, and a point xSpaceTimedx \in \text{SpaceTime}_d, the gauge transformation of AA by χ\chi evaluated at xx is given by A(x)+μχ(x)A(x) + \partial^\mu \chi(x), where μχ\partial^\mu \chi is the pure-gauge potential (the raised-index gradient of χ\chi).

theorem

Invariance of the field strength tensor FμνF^{\mu\nu} under gauge transformations AA+χA \mapsto A + \partial \chi

Let AA be a differentiable electromagnetic potential on a (1+d)(1+d)-dimensional spacetime and χ:SpaceTimedR\chi: \text{SpaceTime}_d \to \mathbb{R} be a twice continuously differentiable (C2C^2) gauge function. The field strength tensor (Faraday tensor) Fμν=μAννAμF^{\mu\nu} = \partial^\mu A^\nu - \partial^\nu A^\mu is invariant under the gauge transformation AμAμ+μχA^\mu \mapsto A^\mu + \partial^\mu \chi, where μχ=ημννχ\partial^\mu \chi = \eta^{\mu\nu} \partial_\nu \chi is the raised-index gradient. That is, for any point xx in spacetime, the field strength of the transformed potential is equal to the field strength of the original potential: F[A+χ](x)=F[A](x) F[A + \partial \chi](x) = F[A](x)

theorem

Invariance of the Field Strength Matrix under Gauge Transformations

Let AA be a differentiable electromagnetic potential on a (1+d)(1+d)-dimensional spacetime and χ:SpaceTimedR\chi: \text{SpaceTime}_d \to \mathbb{R} be a twice continuously differentiable (C2C^2) gauge function. For any point xx in spacetime, the field strength matrix of the potential after a gauge transformation AμAμ+μχA^\mu \mapsto A^\mu + \partial^\mu \chi is identical to the field strength matrix of the original potential AA. That is, F[A+χ](x)=F[A](x) F[A + \partial \chi](x) = F[A](x) where FF denotes the matrix representation of the field strength tensor (Faraday tensor).

theorem

Gauge transformation by χ=0\chi = 0 is the identity map

For any electromagnetic potential AA in dd-dimensional spacetime, the gauge transformation of AA with respect to the constant zero function χ(x)=0\chi(x) = 0 leaves the potential unchanged, i.e., gaugeTransform(0,A)=A\text{gaugeTransform}(0, A) = A.

theorem

gaugeTransform χ1(gaugeTransform χ2A)=gaugeTransform (χ1+χ2)A\text{gaugeTransform } \chi_1 (\text{gaugeTransform } \chi_2 A) = \text{gaugeTransform } (\chi_1 + \chi_2) A

Let AA be an electromagnetic potential on a dd-dimensional spacetime. For any two differentiable gauge functions χ1,χ2:SpaceTimedR\chi_1, \chi_2 : \text{SpaceTime}_d \to \mathbb{R}, the composition of a gauge transformation by χ2\chi_2 followed by a gauge transformation by χ1\chi_1 is equivalent to a single gauge transformation by the sum of the functions χ1+χ2\chi_1 + \chi_2: gaugeTransform(χ1,gaugeTransform(χ2,A))=gaugeTransform(χ1+χ2,A)\text{gaugeTransform}(\chi_1, \text{gaugeTransform}(\chi_2, A)) = \text{gaugeTransform}(\chi_1 + \chi_2, A) where gaugeTransform(χ,A)\text{gaugeTransform}(\chi, A) denotes the transformation AμAμ+μχA^\mu \mapsto A^\mu + \partial^\mu \chi.

theorem

Lorentz Equivariance of Gauge Transformations: ΛgaugeTransform(χ,A)=gaugeTransform(χΛ1,ΛA)\Lambda \cdot \text{gaugeTransform}(\chi, A) = \text{gaugeTransform}(\chi \circ \Lambda^{-1}, \Lambda \cdot A)

For an electromagnetic potential AA in a dd-dimensional spacetime, a differentiable gauge function χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R}, and a Lorentz transformation ΛL\Lambda \in \mathcal{L}, the Lorentz transformation of the gauge-transformed potential is equal to the gauge transformation of the Lorentz-transformed potential by the pulled-back gauge function χΛ1\chi \circ \Lambda^{-1}: ΛgaugeTransform(χ,A)=gaugeTransform(χΛ1,ΛA)\Lambda \cdot \text{gaugeTransform}(\chi, A) = \text{gaugeTransform}(\chi \circ \Lambda^{-1}, \Lambda \cdot A) where (ΛA)(x)=Λ(A(Λ1x))(\Lambda \cdot A)(x) = \Lambda(A(\Lambda^{-1}x)) and gaugeTransform(χ,A)\text{gaugeTransform}(\chi, A) denotes the mapping AμAμ+μχA^\mu \mapsto A^\mu + \partial^\mu \chi.

theorem

The (0,i)(0, i)-component of the field strength matrix for the bare-gradient potential Bμ=μ(y0yi)B^\mu = \partial_\mu(y^0 y^i) equals 22

In a (1+d)(1+d)-dimensional spacetime, let χ:SpaceTimedR\chi : \text{SpaceTime}_d \to \mathbb{R} be a gauge function defined by χ(y)=y0yi\chi(y) = y^0 y^i, where y0y^0 is the temporal coordinate and yiy^i is the ii-th spatial coordinate for some index i{1,,d}i \in \{1, \dots, d\}. If BB is an electromagnetic potential defined by the bare covariant gradient Bμ=μχB^\mu = \partial_\mu \chi, then the (0,i)(0, i) component of the field strength matrix at any point xx is equal to 22: F0i(x)=2F^{0i}(x) = 2

theorem

The field strength of the bare-gradient potential Bμ=μ(y0yi)B_\mu = \partial_\mu(y^0 y^i) is non-zero

In a (1+d)(1+d)-dimensional spacetime, let χ:SpaceTimedR\chi: \text{SpaceTime}_d \to \mathbb{R} be the scalar function defined by χ(y)=y0yi\chi(y) = y^0 y^i, where y0y^0 is the temporal coordinate and yiy^i is the ii-th spatial coordinate for some index i{1,,d}i \in \{1, \dots, d\}. If the electromagnetic potential BB is defined by the bare gradient Bμ=μχB_\mu = \partial_\mu \chi, then its associated field strength tensor FF is non-zero at any point xx in spacetime, i.e., F(x)0F(x) \neq 0.