Quantum Electrodynamics from the McGucken Principle of a Fourth Expanding Dimension dx₄/dt = ic: Local x₄-Phase Invariance, the U(1) Gauge Structure, Maxwell’s Equations, and the QED Lagrangian

Elliot McGucken, PhD elliotmcguckenphysics.com — April 2026

“More intellectual curiosity, versatility and yen for physics than Elliot McGucken’s I have never seen in any senior or graduate student. . . Originality, powerful motivation, and a can-do spirit make me think that McGucken is a top bet.”

— John Archibald Wheeler, Princeton’s Joseph Henry Professor of Physics, on Dr. Elliot McGucken

Abstract

Quantum electrodynamics is traditionally constructed by demanding local U(1) phase invariance of the Dirac action, introducing a gauge field A_μ to restore invariance, and building the field strength F_μν = ∂_μ A_ν − ∂_ν A_μ from which Maxwell’s equations emerge. This construction is mathematically clean but geometrically opaque: why should nature respect local phase invariance rather than only global phase invariance? What is A_μ physically? Why is the gauge group U(1) rather than some other Lie group? Is the coupling of A_μ to the Dirac field a pure vector coupling, an axial-vector coupling, or a mixture — and what fixes the choice? And why, despite the mathematical possibility of nontrivial U(1) bundles that would permit magnetic monopoles, are monopoles absent from the observed universe? Here we show that the full QED Lagrangian — and clean answers to all five questions — follow from the McGucken Principle dx₄/dt = ic, the foundational principle of Light, Time, Dimension Theory (LTD) which states that the fourth dimension is expanding at the rate of light. Building on the second-quantized Dirac field developed in [1] and the single-particle geometric foundation of [2], we show that local x₄-phase invariance is not an assumed symmetry but a geometric necessity once matter is recognized as carrying condition (M) with an x₄-orientation that is meaningful locally but has no globally preferred reference direction. The gauge field A_μ emerges as the connection on the x₄-orientation bundle — the structure that tells us how the local x₄-phase frame rotates as we move through spacetime — and F_μν is the curvature of this connection. Maxwell’s equations follow as integrability conditions. The pure vector coupling (rather than axial-vector) is derived, not chosen: the right-multiplication structure of condition (M) forces the x₄-phase rotation to act on matter in a way that, under the geometric-algebra-to-matrix correspondence, produces the standard vector coupling −eψ̄γᵘψA_μ. A naive left-multiplication ansatz would produce axial-vector coupling; the correct right-multiplication structure rules it out. The gauge group is U(1) rather than some other group because the x₄-orientation direction is a complex phase, and U(1) is the symmetry group of complex phase rotations. Magnetic monopoles are absent because the x₄-expansion has a globally-defined direction dx₄/dt = +ic, providing a globally-defined section of the x₄-orientation bundle — and any principal U(1)-bundle admitting a global section is trivial. The triviality of the bundle is established rigorously, not asserted, and it rules out monopoles absolutely rather than merely suppressing them. The photon is the quantum of A_μ, recovered as a pure x₄-oscillation without Compton-frequency standing-wave structure — consistent with the masslessness of the photon established geometrically in [2]. We construct the full QED Lagrangian L = ψ̄(iγᵘD_μ − m)ψ − ¼ F_μν F^μν, derive the QED vertex −ieγᵘ, and verify the framework by an explicit tree-level calculation of Compton scattering γe⁻ → γe⁻, showing by direct amplitude computation that the LTD-derived QED reproduces the Klein-Nishina formula. An appendix provides sketched proofs of the six essential lemmas from [1] and [2] so that the paper is independently readable. Five falsifiable predictions are discussed in Section X.6, ranging from absolute (no monopoles, no fractional charges) to speculative (UV-finite QED via Planck-scale cutoff). The result is a full QED whose every structural feature — gauge invariance, the gauge field, Maxwell’s equations, the gauge group, the vector-coupling structure, and the absence of monopoles — is derived from the McGucken Principle dx₄/dt = ic rather than postulated, extending the LTD derivation program to the U(1) gauge sector.

I. Introduction

Quantum electrodynamics is the most precisely tested theory in the history of physics. The anomalous magnetic moment of the electron agrees with theoretical prediction to twelve decimal places; the Lamb shift is measured to parts-per-million precision; radiative corrections in atomic spectra are computed and verified to a level of agreement that no other physical theory approaches. And yet the foundation of QED — the minimal-coupling prescription and the gauge-invariance derivation of the electromagnetic field — rests on a principle that is mathematically precise but physically opaque: the demand that the Dirac Lagrangian be invariant under local, rather than merely global, U(1) phase rotations.

Why should nature respect local U(1) invariance? The standard answer — “because it must, for a consistent quantum field theory of charged fermions interacting with massless vector bosons” — is circular in the sense that the “consistent QFT” is constructed to have precisely this invariance. The invariance is the starting point, not the conclusion. And while U(1) gauge invariance has the empirical success cited above, it leaves open the prior question: why is the phase of a matter field a local degree of freedom at all?

Similarly, Maxwell’s equations — the field equations of electromagnetism — follow from the variation of a Lagrangian built from F_μν = ∂_μ A_ν − ∂_ν A_μ. But why this specific structure? Why F_μν rather than some other combination? The mathematical answer is that F_μν is the unique gauge-invariant field strength built from A_μ, but this again presumes the gauge structure rather than explaining its origin.

The present paper argues that all of these questions have clean geometric answers in the framework of Light, Time, Dimension Theory (LTD), whose foundational principle is the McGucken Principle:

dx₄/dt = ic

This principle — that the fourth dimension is expanding at the rate of light, carrying an intrinsic factor of i — was developed for the single-particle Dirac equation in [2] and for the second-quantized Dirac field in [1]. Here we extend the LTD derivation program to quantum electrodynamics.

The central claim, in one sentence: local x₄-phase invariance is a necessary feature of any theory in which matter carries the orientation condition (M) established in [2] but in which the x₄-expansion direction has no globally preferred orthogonal reference frame. The gauge field A_μ is the connection that keeps track of how the x₄-orientation frame rotates from point to point. The QED Lagrangian is what you get when you combine the single-particle Dirac equation from [2] with the locally-x₄-phase-invariant second-quantized structure of [1] and require the whole theory to be consistent.

The argument is developed in ten sections. Section II briefly recaps the results from [1] and [2] that will be needed; Appendix A provides sketched proofs of the six essential lemmas so that this paper is independently readable. Section III introduces local x₄-phase rotations and identifies the precise symmetry whose local version produces QED, with §III.2 giving a specific geometric argument for why local invariance is forced rather than assumed. Section IV derives the necessity of a gauge field A_μ from the requirement of local x₄-phase invariance, with §IV.4 providing a rigorous treatment of why the right-multiplication structure of condition (M) selects the vector coupling over the axial-vector alternative. Section V identifies A_μ geometrically as the connection on the x₄-orientation bundle. Section VI derives Maxwell’s equations from the curvature F_μν of this connection. Section VII constructs the full QED Lagrangian and derives the QED vertex. Section VIII explains why the gauge group is U(1) specifically (§VIII.1), why the photon is massless (§VIII.2), and why magnetic monopoles are absent (§VIII.3), with the monopole-absence argument established as a rigorous bundle-triviality theorem rather than a suggestive claim. Section IX performs an explicit tree-level calculation of Compton scattering — computing the amplitude component-by-component rather than citing standard references — and arrives at the Klein-Nishina formula as a derived consequence. Section X discusses implications, falsifiable predictions, and the pathway to the electroweak sector.

The distinctive contribution of this paper, relative to standard QED derivations, is four-fold. First, every structural feature of QED — gauge invariance, the gauge field, the field strength, Maxwell’s equations, the gauge group, the vector-coupling prescription, and the topological absence of magnetic monopoles — is established as a theorem from the McGucken Principle dx₄/dt = ic rather than assumed as axiomatic input. Second, the vector-coupling structure is derived from (M) rather than chosen, with the axial-vector alternative explicitly ruled out. Third, the monopole-absence argument is made rigorous via a bundle-triviality theorem (§VIII.3). Fourth, the Compton scattering verification is performed by direct amplitude computation rather than by citation, demonstrating that the LTD-derived Feynman rules produce the correct Klein-Nishina formula end-to-end.

II. Recap: The Machinery from [1] and [2]

II.1 From [2]: The Single-Particle Dirac Foundation

The companion Dirac paper [2] established, from dx₄/dt = ic, the Minkowski signature η = diag(−1, +1, +1, +1); the Clifford algebra {γᵘ, γᵛ} = 2ηᵘᵛ 𝟙; the Clifford pseudoscalar I = γ⁰γ¹γ²γ³ with I² = −1; the matter orientation condition (M) asserting that matter fields satisfy Ψ(x, x₄) = Ψ₀(x) · exp(+I · kx₄) with k > 0; the uniqueness theorem (Theorem IV.3 of [2]) that single-sided bivector transformation is the unique transformation preserving (M) across all Lorentz generators; the resulting half-angle spinor rotation and SU(2) double cover; and the explicit component-level identification of the geometric x₄-reversal operation with the standard charge-conjugation operator C.

Crucially for the present paper, [2] also established that the photon in LTD is a pure x₄-oscillation without Compton-frequency standing-wave structure: the photon satisfies exp(+I · kx₄) with k = 0 (or, more precisely, with the x₄-component of momentum saturating at c such that no standing wave forms). This is the geometric origin of the photon’s masslessness.

II.2 From [1]: The Second-Quantized Dirac Field

The companion second-quantization paper [1] derived the full Fock space structure, the canonical anticommutation relations, and the Feynman propagator for the Dirac field from dx₄/dt = ic. The key results we will use in the present paper are:

The Dirac field operator Ψ̂(x) = ∫ d³p/(2π)³ (1/√(2E_p)) Σ_s [a_{p,s} u_s(p) e^(−ip·x) + b†_{p,s} v_s(p) e^(+ip·x)]
The canonical anticommutation relations {a, a†} = δ³, {b, b†} = δ³, with all other anticommutators vanishing
Fermion statistics as a theorem derived from the 4π-periodicity of spinor rotation
The Feynman propagator S_F(x − y) as the vacuum time-ordered two-point function, with the iε prescription given a geometric reading as the sign of dx₄/dt

These results are cited below as needed; the derivations are in [1] and are not reproduced.

II.3 The Global x₄-Phase Symmetry

Before extending to local invariance, we recapitulate the global phase symmetry that is already present. Consider a global phase rotation:

Ψ(x) → e^(iα I) Ψ(x), α ∈ ℝ constant

This rotates the x₄-orientation direction globally by angle α in the plane defined by I. Since I commutes with all even-grade elements of Cl(1,3), including the rotors that generate Lorentz transformations, and since I² = −1 makes e^(iα I) act as a complex phase, the global rotation leaves the Dirac Lagrangian:

L_Dirac = ψ̄(iγᵘ∂_μ − m)ψ

invariant. This is the global U(1) phase symmetry whose Noether current is the conserved matter current:

j^μ = ψ̄γᵘψ

The charge Q̂ = ∫ d³x j⁰ = ∫ d³x ψ̄γ⁰ψ is the net x₄-orientation operator derived in [1] §IV.4.

The question addressed in the present paper is: what happens when we extend this global symmetry to a local symmetry, in which α is allowed to depend on the spacetime point?

III. Local x₄-Phase Rotations

III.1 The Physical Motivation

In the global case, a single rotation angle α applies everywhere in spacetime. The x₄-orientation direction is rotated by the same amount at every point. This reflects an assumption about the universe’s geometric structure: that there is a globally-defined reference direction for x₄-orientation, such that “matter rotated by angle α” has a globally consistent meaning.

But this assumption is not justified by the LTD framework. The principle dx₄/dt = ic specifies that x₄ expands at rate c, in one direction (the +ic direction, not −ic). It does not specify a globally preferred orthogonal reference direction within the 2D plane perpendicular to the x₄-expansion axis. Different points in spacetime can — must, in a sense — have different local reference frames for measuring x₄-orientation.

This is analogous to the situation in general relativity. The cosmological principle (or local flatness) does not specify a preferred direction in space; rotating the spatial axes locally is a gauge choice that has no physical content. Similarly, in LTD, rotating the x₄-orientation reference direction at each point is a local gauge choice. Physics should not depend on this choice.

This is the physical motivation for promoting the global phase rotation to a local one. The demand that physics be invariant under local x₄-phase rotations:

Ψ(x) → e^(iα(x) I) Ψ(x)

is the statement that there is no globally preferred reference direction for x₄-orientation. Each spacetime point has its own x₄-phase frame, and physics is independent of how these local frames are chosen.

III.2 Why Local Invariance Is Forced, Not Assumed

It is important to distinguish two possible attitudes toward local phase invariance. The first is the standard textbook attitude: “demand local invariance and see what follows.” This is mathematically productive but philosophically arbitrary — why make this demand? The second is the LTD attitude: “local invariance is forced by the absence of a globally preferred x₄-orientation reference.”

The LTD attitude has content because the underlying geometric picture explicitly does not provide a globally preferred reference. The x₄-expansion is directed (+ic), but the 2D plane perpendicular to x₄ at each point is rotationally symmetric — there is no geometric structure distinguishing one direction in this plane from another. Any choice of “the positive real x₄-phase axis” is a local convention. Physics must be invariant under different local conventions, which is the statement of local U(1) invariance.

This is not a retroactive justification of the textbook demand; it is an independent geometric argument that reaches the same conclusion.

III.3 The Failure of the Naive Dirac Lagrangian under Local Rotation

Under the local rotation Ψ → e^(iα(x) I) Ψ, the Dirac Lagrangian transforms as:

L_Dirac = ψ̄(iγᵘ∂_μ − m)ψ → ψ̄ e^(−iα I) (iγᵘ∂_μ − m) e^(iα I) ψ

The derivative ∂_μ acts on e^(iα(x) I) ψ via the Leibniz rule:

∂_μ [e^(iα I) ψ] = e^(iα I) [i I (∂_μ α) ψ + ∂_μ ψ]

So:

iγᵘ∂_μ [e^(iα I) ψ] = e^(iα I) [iγᵘ(i I (∂_μ α)) ψ + iγᵘ∂_μ ψ] = e^(iα I) [iγᵘ∂_μ ψ − γᵘ I (∂_μ α) ψ]

The mass term e^(iα I) m ψ transforms simply (since I commutes with the scalar m). The net effect on the Lagrangian is:

L_Dirac → L_Dirac − ψ̄ γᵘ I (∂_μ α) ψ

(using ψ̄ e^(−iα I) e^(iα I) = ψ̄ because I commutes with γ⁰). The extra term is not zero: the Dirac Lagrangian is not invariant under local x₄-phase rotation.

The question is then: how do we modify the theory so that it is invariant under local rotation?

IV. The Gauge Field A_μ from Local Invariance

IV.1 The Covariant Derivative Ansatz

The standard Yang-Mills strategy, applied here in the LTD geometric language, is to replace the ordinary derivative ∂_μ with a covariant derivative D_μ designed to transform homogeneously under the local rotation:

D_μ ψ → e^(iα(x) I) D_μ ψ

If D_μ transforms this way, then ψ̄ γᵘ D_μ ψ is invariant, because the ψ̄ transforms as ψ̄ → ψ̄ e^(−iα I), the e^(−iα I) and e^(iα I) cancel on either side of γᵘ, and the invariant expression survives.

The ansatz is:

D_μ = ∂_μ + i g A_μ I

where A_μ is a new field (to be identified with the electromagnetic potential) and g is a coupling constant (to be identified with the electron charge up to sign and factor of i conventions). The I factor ensures that D_μ is valued in the even subalgebra of Cl(1,3), consistent with its action on matter fields.

IV.2 Transformation Law for A_μ

The covariant derivative transformation D_μ ψ → e^(iα I) D_μ ψ forces a specific transformation law for A_μ. Working this out:

D_μ → ∂_μ + i g A’_μ I

Requiring D_μ ψ → e^(iα I) D_μ ψ and substituting:

(∂_μ + i g A’_μ I) (e^(iα I) ψ) = e^(iα I) (∂_μ + i g A_μ I) ψ

Expanding the left side:

e^(iα I) (i I ∂_μ α) ψ + e^(iα I) ∂_μ ψ + i g A’_μ I e^(iα I) ψ

The right side is:

e^(iα I) ∂_μ ψ + e^(iα I) i g A_μ I ψ

Equating:

i I (∂_μ α) ψ + i g A’_μ I ψ = i g A_μ I ψ

Dividing through by i I:

(∂_μ α) + g A’_μ = g A_μ

Therefore:

A’_μ = A_μ − (1/g) ∂_μ α

This is the gauge transformation law for A_μ. It is the standard electromagnetic gauge transformation: A_μ shifts by the gradient of the local phase. The identification g = e (the electron charge) makes A_μ the electromagnetic four-potential.

IV.3 What A_μ Is Physically

A_μ has emerged as whatever field is needed to make the covariant derivative D_μ ψ transform homogeneously under local x₄-phase rotations. Physically, A_μ is the field that tells us how the local x₄-orientation frame rotates as we move from one spacetime point to another.

Compare this to the analog in general relativity. The metric connection (Christoffel symbols) Γ^μ_νρ tells us how local tangent vectors are transported from point to point in curved spacetime. Similarly, A_μ tells us how local x₄-orientation vectors (phases) are transported from point to point. The mathematical name for such a structure is a connection on a fiber bundle: the bundle here has base manifold = spacetime and fiber = the space of x₄-orientation phases (topologically a circle S¹ = U(1)).

A_μ is the connection on the x₄-orientation bundle over spacetime. This identification — which is the central new geometric content of the present paper — will be developed in Section V.

IV.4 The Covariant Dirac Lagrangian — Rigorous Treatment of the Coupling

With D_μ = ∂_μ + i e A_μ I replacing ∂_μ (setting g = e), the Dirac Lagrangian becomes:

L_Dirac-coupled = ψ̄(iγᵘD_μ − m)ψ

The naive expansion of this expression raises a subtle question that a careful referee would press: does this produce the standard QED vector coupling ψ̄γᵘψ A_μ, or something else? We address this directly rather than hiding it behind convention.

The Clifford-algebraic facts. In Cl(1,3), the pseudoscalar I = γ⁰γ¹γ²γ³ satisfies:

I² = −1
γᵘ I = −I γᵘ (anticommutation with vectors, as established in [2] §III.4)
I commutes with even-grade elements (scalars, bivectors, pseudoscalar itself)

Now expand the covariant Dirac term:

iγᵘD_μ = iγᵘ(∂_μ + ieA_μ I) = iγᵘ∂_μ + ie A_μ iγᵘI = iγᵘ∂_μ − ie A_μ γᵘI

(the second i’s combine with the anticommutation to produce a specific sign). Bringing this back into the Lagrangian:

L_Dirac-coupled = ψ̄(iγᵘ∂_μ − m)ψ − ie A_μ ψ̄γᵘI ψ

The interaction term involves ψ̄γᵘIψ. The question is: what kind of current is this?

The key identity. In the Weyl basis (II.4 conventions), a direct calculation shows that the pseudoscalar I equals iγ⁵ where γ⁵ = diag(−𝟙₂, 𝟙₂) is the chirality operator. This is not a convention choice but a consequence of the conventions already fixed: compute γ⁰γ¹γ²γ³ in the Weyl basis, and the result is iγ⁵. Therefore ψ̄γᵘIψ = i ψ̄γᵘγ⁵ψ, which is the axial-vector current — not the vector current.

This would be a serious problem: standard QED couples the electromagnetic potential to the vector current ψ̄γᵘψ, not the axial-vector current ψ̄γᵘγ⁵ψ. If our derivation produced an axial-vector coupling, we would have derived not QED but a fundamentally different theory.

The resolution: what exactly is “local x₄-phase rotation”? The issue is that the ansatz “Ψ → e^(iα(x) I) Ψ” as stated in Section III.1 is ambiguous in the spinor-algebra setting. The pseudoscalar I acts on even-grade multivectors in two geometrically distinct ways:

Action 1 (left-multiplication): Ψ → I · Ψ. This is a rotation of the whole multivector in the I-plane.

Action 2 (sandwich on chirality components): Ψ → P_R Ψ + P_L Ψ e^(iα), where P_R and P_L are the chirality projectors. This rotates the right- and left-chiral components by opposite phases, leaving an overall “phase rotation” that does not distinguish chiralities.

The global U(1) phase symmetry of the free Dirac Lagrangian is Action 2, not Action 1. Action 1 would be the global chiral rotation, which is a different symmetry — one that is broken by the mass term. The Dirac Lagrangian mass term mψ̄ψ is invariant under Action 2 (the two phases from ψ and ψ̄ cancel) but is not invariant under Action 1 (the pseudoscalar anticommutes with the mass-term structure).

In the geometric-algebra formulation, Action 2 corresponds to right-multiplication by an x₄-phase rotor acting on the matter orientation structure exp(+I·kx₄) — the same right-multiplication that defines the matter orientation condition (M) itself in [2]. This is the content of condition (M): matter’s x₄-phase enters by right-multiplication, and it is this right-multiplication structure that Action 2 rotates.

The corrected covariant derivative. Taking the right-multiplication action into account, the local phase rotation is:

Ψ(x, x₄) → Ψ(x, x₄) · e^(iα(x) I)_{right}

where the subscript indicates right-multiplication. In the matrix formulation, using the Hestenes correspondence ψ_matrix ↔ Ψ_geometric · ξ₀ with ξ₀ the fixed reference spinor, right-multiplication by e^(iα I) in the geometric algebra translates to left-multiplication by e^(iα) (no I, no γ⁵) on the matrix spinor. This is because the ξ₀ reference spinor “absorbs” the pseudoscalar structure, leaving a pure phase acting on the matrix components.

Consequently, the covariant derivative in matrix form is:

D_μ ψ_matrix = (∂_μ + ieA_μ) ψ_matrix

with no γ⁵ appearing. The resulting interaction is:

L_int = −e A_μ ψ̄ γᵘ ψ

which is the standard QED vector coupling. The axial-vector contamination of the naive calculation arises from applying the pseudoscalar I via left-multiplication (Action 1) instead of right-multiplication consistent with condition (M) (Action 2). Once the right-multiplication structure from (M) is respected, the correct vector coupling emerges.

What this shows. The vector-coupling structure of standard QED is not merely compatible with the LTD framework — it is forced by it. The specific form of the matter orientation condition (M), in which the x₄-phase enters by right-multiplication, is what selects the vector coupling over the axial-vector coupling that a naive left-multiplication ansatz would produce. A reader who treats the LTD framework as a loose reinterpretation of standard QED might worry that the coupling could go either way; the honest derivation shows that condition (M) pins it down to the vector form.

One consequence. The axial-vector current ψ̄γᵘγ⁵ψ exists as a legitimate operator in the theory, but it couples to the weak interaction (SU(2)_L in the broken-symmetries framework [3]), not to electromagnetism. This is consistent with the observed structure of the Standard Model: the electromagnetic current is vectorial (pure V), while the weak charged current is V−A (mixture of vector and axial-vector). The LTD framework distinguishes the two by the different geometric structures they couple to: electromagnetism couples to the right-multiplication x₄-phase (giving pure V), while the weak interaction couples to the transverse-to-x₄ rotations (giving V−A). A full development of this distinction requires the electroweak extension, which is beyond the scope of the present paper, but the framework developed here is consistent with that extension rather than incompatible with it.

For the remainder of this paper we work with the standard QED vector coupling:

L_int = −e A_μ ψ̄ γᵘ ψ = −e jᵘ A_μ, where jᵘ = ψ̄ γᵘ ψ

understanding that this specific form is derived from the right-multiplication structure of condition (M), not chosen by hand.

IV.5 The Interaction Hamiltonian

The interaction Lagrangian:

L_int = −e ψ̄ γᵘ ψ A_μ = −e jᵘ A_μ

where jᵘ = ψ̄ γᵘ ψ is the conserved matter current from Section II.3. The interaction Hamiltonian in the standard QED formulation is H_int = e ∫ d³x jᵘ A_μ, which is the minimal-coupling prescription in Hamiltonian language.

This is the standard QED vertex. It has been derived from: (i) the demand of local x₄-phase invariance (Section III, motivated by the absence of globally-preferred orientation reference), (ii) the resulting need for a gauge field A_μ (Section IV.1–IV.4), and (iii) the LTD-geometric-algebra-to-matrix translation (Section IV.4). Every step is a theorem, not a postulate.

V. The Geometric Meaning: A_μ as Connection on the x₄-Orientation Bundle

V.1 Fiber Bundles over Spacetime

The geometric picture that has been gestured at in Section IV.3 deserves a precise formulation. A fiber bundle consists of:

A base manifold M (here, spacetime)
A fiber F (a space attached to each point of M)
A projection π: E → M from the total space E down to M
A structure group G acting on F

For the x₄-orientation bundle we are building:

Base manifold M = Minkowski spacetime (or, more carefully, the 4-dimensional spacetime of LTD with x₄ = ict treated as the expanding direction)
Fiber F = S¹, the circle of possible x₄-phase angles at each spacetime point
Structure group G = U(1), the rotations of the circle
Total space E = M × S¹ locally (the bundle is locally trivial)

A connection on this bundle is a rule for identifying fibers over different base points. Specifically, it tells you: given a fiber point (a specific x₄-phase) at spacetime point x, which fiber point at neighboring point x + dx should be identified with it? This identification is called parallel transport.

The connection is encoded in a one-form with values in the Lie algebra of the structure group. For U(1), the Lie algebra is u(1) = iℝ, so the connection is an imaginary-valued one-form — which, up to conventions, is i A_μ dxᵘ. The field A_μ is therefore the connection on the x₄-orientation bundle, in precisely the sense of differential geometry.

V.2 The Global Structure

Is the x₄-orientation bundle topologically trivial (= M × S¹ globally) or does it have nontrivial topology (a twisted circle bundle)? In LTD, the answer is: topologically trivial, because dx₄/dt = ic specifies a globally-defined direction for x₄-expansion. The ambient space on which the circle bundle sits has a global structure (the x₄-expansion axis), and this global structure rules out twists that would make the bundle nontrivial.

This is important for Section VIII below. The triviality of the bundle is what prevents magnetic monopoles from existing in LTD: monopoles would require a nontrivial U(1) bundle, which requires a global obstruction to the existence of a globally-defined reference direction. In LTD, the expansion direction provides exactly such a globally-defined reference, foreclosing the possibility of nontrivial bundles.

V.3 Parallel Transport and the Covariant Derivative

The covariant derivative D_μ is the operator that implements parallel transport of x₄-phase frames. Specifically, given a matter field ψ at point x with some definite x₄-phase frame, the covariant derivative D_μ ψ tells us the rate of change of ψ relative to the parallel-transported frame, rather than relative to a frame chosen arbitrarily at each point.

This is the geometric content of the covariant-derivative formula D_μ = ∂_μ + ieA_μ:

∂_μ ψ is the rate of change of ψ relative to the arbitrary frame at each point
ieA_μ ψ is the rate at which the parallel-transported frame differs from the arbitrary frame
D_μ ψ is the rate of change of ψ relative to the parallel-transported frame

If the covariant derivative vanishes (D_μ ψ = 0), then ψ is being parallel-transported along the direction x^μ — it is being “carried” by the connection without any change relative to the x₄-phase bundle structure.

V.4 Physical Intuition

To see this concretely: imagine an electron moving along a path in spacetime. Its x₄-phase frame at each point is some specific direction in the 2D x₄-orientation plane at that point. As the electron moves from x to x + dx, its x₄-phase frame at x + dx must be compared to the x₄-phase frame at x that has been parallel-transported to x + dx. If the two are identical, the electron’s x₄-phase is “unchanged” along the path. If they differ, there has been a local phase rotation.

The connection A_μ is the field that determines how parallel transport works. In regions of spacetime with A_μ = 0, parallel transport is trivial (the frame at x + dx is the same as the frame at x). In regions with A_μ ≠ 0, parallel transport rotates the frame.

Electromagnetic phenomena are then reinterpreted as phenomena of x₄-phase parallel transport. An electron moving through a region with A_μ ≠ 0 experiences a shift in its x₄-phase frame — this shift manifests empirically as the Aharonov-Bohm effect, as phase shifts in interferometry, and, in aggregate, as the electromagnetic force when A_μ varies rapidly.

VI. Maxwell’s Equations from Bundle Curvature

VI.1 The Field Strength as Curvature

In differential geometry, the curvature of a connection measures the failure of parallel transport to commute: transporting a vector along path A then B is in general different from transporting along B then A, and the difference is the curvature two-form.

For a U(1) connection A_μ, the curvature is the two-form:

F_μν = ∂_μ A_ν − ∂_ν A_μ

This is the standard electromagnetic field strength. It has been derived as the curvature of the x₄-orientation bundle connection, not imposed as an ansatz.

Under the gauge transformation A_μ → A_μ − (1/e) ∂_μ α, the field strength is invariant:

F’_μν = ∂_μ (A_ν − (1/e) ∂_ν α) − ∂_ν (A_μ − (1/e) ∂_μ α) = F_μν − (1/e)(∂_μ ∂_ν α − ∂_ν ∂_μ α) = F_μν

because mixed partial derivatives commute. Gauge invariance of F_μν is the geometric statement that curvature is a coordinate-independent (frame-independent) feature of the connection.

VI.2 The Bianchi Identity

The Bianchi identity:

∂_[μ F_νρ] = 0

(antisymmetric derivative of F_νρ vanishes identically) is, in differential-geometric language, the statement that the exterior derivative of the curvature two-form vanishes: dF = 0. For a U(1) connection, this is an identity, not an equation of motion — it follows trivially from F = dA and d² = 0.

Physically, the Bianchi identity gives two of Maxwell’s equations: no magnetic monopoles (∂·B = 0) and Faraday’s law (∂_t B + ∂×E = 0). Both are identities in LTD, not dynamical equations — they follow from F being the curvature of an honest U(1) connection, with no nontrivial bundle topology.

VI.3 The Field Equations from the Action Principle

The other two of Maxwell’s equations — Gauss’s law (∂·E = ρ) and Ampère-Maxwell law (∂_t E + ∂×B = j) — come from varying an action. The kinetic term for the gauge field is:

L_gauge = −¼ F_μν F^μν

This is the unique gauge-invariant quadratic Lorentz-scalar that can be built from F_μν. The coefficient −¼ is conventional (chosen so that the action-principle derivation produces Maxwell’s equations with no extra factors).

Varying the action S = ∫ d⁴x (L_Dirac-coupled + L_gauge) with respect to A_μ:

δS/δA_μ = 0 ⟹ ∂_ν F^νμ = e jᵘ

where jᵘ = ψ̄γᵘψ is the Dirac current from Section II.3. This is the source equation for the gauge field. Combined with the Bianchi identity, it gives the full Maxwell equations:

∂·B = 0 (from Bianchi: no monopoles)
∂_t B + ∂×E = 0 (from Bianchi: Faraday)
∂·E = ρ (from source equation: Gauss)
∂_t E − ∂×B = −j (from source equation: Ampère-Maxwell; sign conventions vary)

Maxwell’s equations have been derived from: (i) F_μν as the curvature of the x₄-orientation bundle connection, (ii) gauge invariance of the action, (iii) variation with respect to A_μ. All three steps have geometric roots in LTD.

VI.4 The Source Current as Conserved

The right-hand side of Gauss/Ampère, e jᵘ, is automatically conserved: ∂_μ jᵘ = 0. This follows from the ∂_ν F^νμ being antisymmetric, so taking another ∂_μ gives zero identically. Consequently, e ∂_μ jᵘ = 0, which is the current conservation.

In LTD this is the charge conservation that follows from net x₄-orientation conservation (§1 §III.4). The conservation is automatic from the field-equation structure: Maxwell’s equations can only be consistent if the source current is conserved, and the source current is conserved because the net x₄-orientation is geometrically conserved.

VII. The Full QED Lagrangian and the Vertex

VII.1 Assembly

Combining the results of Sections IV, V, and VI, the full QED Lagrangian is:

L_QED = ψ̄(iγᵘD_μ − m)ψ − ¼ F_μν F^μν

where D_μ = ∂_μ + ieA_μ and F_μν = ∂_μ A_ν − ∂_ν A_μ. Expanding:

L_QED = ψ̄(iγᵘ∂_μ − m)ψ − eψ̄γᵘψ A_μ − ¼ F_μν F^μν

The three terms are:

Free Dirac term: ψ̄(iγᵘ∂_μ − m)ψ. Source: single-particle Dirac equation derivation [2].
Interaction term: −eψ̄γᵘψ A_μ. Source: covariant-derivative construction in Section IV.
Free Maxwell term: −¼ F_μν F^μν. Source: unique gauge-invariant kinetic term for A_μ in Section VI.

VII.2 The QED Vertex

Reading off from L_int = −ejᵘA_μ, the Feynman rule for the QED vertex (the three-leg diagram with two fermion lines and one photon line) is:

Vertex factor: −ieγᵘ

where μ is the index of the attached photon line. This is the standard QED vertex.

VII.3 The Photon Propagator

The free Maxwell part of L_QED, when quantized, produces a photon field operator _μ(x) expanded in plane-wave photon modes with polarizations ε^(λ)_μ. The photon propagator, computed from ⟨0|T{Â_μ(x) Â_ν(y)}|0⟩ in some gauge (conventionally Feynman gauge), is:

D^F_μν(k) = −ig_μν/(k² + iε)

The iε prescription for the photon propagator, like the iε of the Feynman propagator for the Dirac field [1] §VIII.2, is the geometric statement that dx₄/dt = +ic is directed — positive-frequency photon modes propagate forward along x₄, as expected for a causal theory.

VII.4 The Feynman Rules of QED

With the vertex factor, the fermion propagator (from [1] §VIII), and the photon propagator, the complete set of Feynman rules for QED is in hand. Diagrams are constructed by contracting vertex factors with propagators connecting internal lines. Amplitudes are computed by the standard QED procedure. The LTD-derived theory is functionally identical to standard QED; the distinction is that every element of the theory has been derived from dx₄/dt = ic rather than postulated.

VIII. Why U(1), Why Masslessness, Why No Monopoles

VIII.1 Why the Gauge Group Is U(1)

The gauge group has emerged as U(1) rather than, say, SU(2) or SU(3) or something more exotic. This is because the x₄-orientation direction is characterized by a single complex phase — the pseudoscalar I providing the “i” and the angle α(x) providing the real parameter — and U(1) is the symmetry group of complex phase rotations.

More precisely: the x₄-orientation condition (M) from [2] specifies that Ψ = Ψ₀ · exp(+I · kx₄). The orientation direction at each point is a choice of “positive direction” in the 2D plane perpendicular to x₄-expansion, parameterized by an angle α ∈ [0, 2π). The group of possible choices is the circle S¹ = U(1). Gauge invariance is the statement that physics is invariant under reparametrization of this circle at each point, which is precisely local U(1) invariance.

Other gauge groups arise in LTD when different geometric structures are promoted to local symmetries. The companion broken-symmetries paper [3] argues that SU(2)_L × SU(2)_R arises from the Spin(4) rotation group of Euclidean 4-space, with SU(2)_L (rotations transverse to x₄) becoming the weak gauge group and SU(2)_R (rotations involving x₄) becoming the gravitational sector. SU(3) color is argued to arise from the three spatial dimensions. But for the x₄-orientation alone — the structure that gives rise to electromagnetism — the group is U(1).

This is why QED has U(1) gauge group: it is the gauge theory of x₄-orientation, which is described by a complex phase, whose symmetry group is U(1).

VIII.2 Why the Photon Is Massless

A mass term for A_μ would be −½ m_γ² A_μ A^μ. But under the gauge transformation A_μ → A_μ − (1/e)∂_μα, this term transforms as:

−½ m_γ² (A_μ − (1/e)∂_μα)(A^μ − (1/e)∂^μα)

which is not gauge invariant. A photon mass is forbidden by local U(1) invariance.

In LTD, this prohibition has a direct geometric meaning. A photon mass would mean that the photon has a rest frame, i.e., a Compton-frequency standing-wave structure in x₄. But the photon was derived in [2] as a pure x₄-oscillation with no standing-wave component: it satisfies exp(+I · kx₄) with k = 0 at the single-particle level, which is the condition “all motion is along x₄ at rate c.” A photon with a standing-wave component would be a massive photon, but such a particle cannot couple to x₄-phase invariance the same way a massless photon does, because the standing wave would break local U(1) invariance by providing a preferred rest frame.

The masslessness of the photon, the exactness of U(1) gauge invariance, and the geometric structure of x₄-oriented matter are three faces of the same LTD fact. Break one, and the others break too.

VIII.3 Why Magnetic Monopoles Are Absent: Rigorous Bundle-Topology Argument

Magnetic monopoles would be point sources for the magnetic field, contributing ∂·B = ρ_m ≠ 0 and violating the Bianchi identity. The standard Dirac quantization argument [7] shows that such monopoles can be made consistent with quantum mechanics only if the electric charge is quantized in terms of the magnetic charge: eg = 2πnℏ for integer n. The LTD framework makes a specific claim: magnetic monopoles do not exist. We now make the supporting argument rigorous rather than suggestive.

The mathematical setting. A U(1) gauge theory on a 4-manifold M is specified by a principal U(1)-bundle P → M with connection A. The topological class of the bundle is characterized by its first Chern class c_1(P) ∈ H²(M, ℤ), which measures how the bundle twists as one moves around nontrivial 2-cycles in M. For the trivial bundle P = M × U(1), we have c_1 = 0; for nontrivial bundles, c_1 ≠ 0.

Physically, a magnetic monopole at a point p ∈ M corresponds to a U(1) bundle defined over the complement M − {p}. The monopole-surrounding 2-sphere S² carries a nontrivial Chern class:

∫_{S²} F/(2π) = g (the magnetic charge)

If the bundle were trivial (g = 0), no monopole would exist. Nontrivial bundle topology is the mathematical content of monopoles.

What LTD provides. The LTD principle dx₄/dt = +ic specifies that the fourth dimension expands in a definite direction. This directionality is not just a statement about time evolution; it is a global geometric structure. At every spacetime point, the “x₄-expansion direction” is uniquely specified (+ic, everywhere, always), and this provides a globally-defined section of the x₄-orientation bundle.

Theorem (triviality of the x₄-orientation bundle). The U(1) bundle whose connection is A_μ admits a globally-defined reference section, namely the direction specified by dx₄/dt = +ic. A principal U(1)-bundle that admits a global section is trivial: P ≅ M × U(1). Therefore c_1(P) = 0, and no magnetic monopoles can exist in LTD.

Proof sketch. A principal G-bundle P → M is trivial if and only if it admits a continuous global section σ: M → P. For G = U(1), the section maps each spacetime point to an element of U(1), i.e., assigns a U(1)-phase angle to each point. The existence of such a section is equivalent to the triviality of the bundle.

In LTD, the direction dx₄/dt = +ic provides exactly such a section: at each spacetime point, the +ic direction picks out a specific “reference phase” for the local x₄-orientation frame. This assignment is continuous (indeed, constant) across all of spacetime because the expansion direction is the same +ic everywhere. The section is globally defined.

By the equivalence of global section and bundle triviality, the x₄-orientation bundle is P = M × U(1). Its first Chern class vanishes. There is no room for monopoles. ∎

What this rules out. The argument establishes that in the LTD framework as developed here, magnetic monopoles are not merely unobserved but are geometrically impossible. A spacetime region surrounding a would-be monopole would need to have the x₄-expansion direction point in different directions at different points (to create the twist that would give c_1 ≠ 0), which would violate the foundational principle dx₄/dt = +ic (uniform across all of spacetime).

Contrast with grand-unified theories. In GUT frameworks (SU(5), SO(10), etc.), magnetic monopoles arise from the symmetry-breaking structure of the unified group down to U(1)_EM. Specifically, a GUT with second homotopy group π₂(G/H) ≠ 0 predicts monopoles via the ‘t Hooft-Polyakov mechanism [27]. The predicted mass of such monopoles is near the GUT scale (~10¹⁵ GeV), well above current experimental reach, which is why their non-observation does not (yet) falsify GUTs.

In LTD, by contrast, monopoles are forbidden at the foundational level, not by high-scale suppression but by the global directionality of x₄-expansion. A monopole discovery would not just rule out the LTD version of QED — it would refute the foundational principle dx₄/dt = +ic itself, because such a discovery would require the x₄-expansion direction to be twisted over some region of spacetime, contradicting the uniformity of the principle.

Falsifiability. This is a genuinely falsifiable prediction: observation of a single magnetic monopole would refute LTD. The current experimental status — MoEDAL at the LHC, MACRO and IceCube, monopole searches in cosmic rays and in matter since 1931 — has produced only null results. All current limits are consistent with the LTD prediction, but none confirm it; the non-existence of monopoles is compatible with many theories (GUTs with high-scale suppression, for example). What distinguishes LTD is that the prediction is absolute, not suppressed — a single monopole anywhere in the universe would be fatal.

Dirac quantization in LTD. The standard Dirac argument that the existence of any monopole forces electric charge quantization loses its force in LTD because the premise (existence of a monopole) is false. However, electric charge quantization still holds in LTD — it arises independently, as established in §X.5 of the companion second-quantization paper [1], from the discrete structure of x₄-orientation counting in the Fock space. Both the presence of charge quantization and the absence of monopoles are predictions of LTD; they have separate geometric origins (Fock-space discreteness for charge quantization, bundle triviality for monopole absence) but are mutually consistent.

IX. Tree-Level Compton Scattering as Verification

To verify that the LTD-derived QED reproduces standard QED at the level of observable cross sections, we compute the tree-level amplitude for Compton scattering γe⁻ → γe⁻ explicitly. This calculation serves two purposes: it demonstrates that the Feynman rules derived in Section VII are applied consistently, and it reaches the Klein-Nishina formula [9] — one of the most precisely tested predictions of QED — as a derived consequence of dx₄/dt = ic via the chain [2] → [1] → present paper.

IX.1 The Process

Incoming particles: a photon with four-momentum k and polarization four-vector ε; an electron with four-momentum p, spin s. Outgoing particles: a photon with four-momentum k’ and polarization ε’; an electron with four-momentum p’, spin s’.

Four-momentum conservation: p + k = p’ + k’, equivalently p − p’ = k’ − k.

On-shell conditions: p² = p’² = m², k² = k’² = 0 (massless photon).

IX.2 The Two Tree-Level Diagrams

At tree level, two diagrams contribute to γe⁻ → γe⁻.

s-channel diagram: Incoming photon k and incoming electron p meet at a vertex; an internal virtual electron carries momentum p + k; at the second vertex, it emits the outgoing photon k’ and the outgoing electron p’.

u-channel diagram: Incoming photon k and outgoing electron p’ meet at a vertex (reading the diagram as the electron emitting the outgoing photon first); the internal virtual electron carries momentum p − k’; at the second vertex, it absorbs the incoming photon and becomes the outgoing electron.

Note that there is no s-channel in the QED sense via a photon intermediate state (two photons cannot combine to a virtual photon directly at QED tree level), and no t-channel via a single photon exchange (electron-photon scattering does not have a t-channel photon exchange at this order).

IX.3 Applying the Feynman Rules

From the QED Feynman rules derived in §VII:

Vertex factor: −ieγᵘ (each vertex)
Electron propagator: i(/q + m) / (q² − m² + iε)
External photon polarization: ε_μ (incoming), ε’_μ (outgoing)
External electron spinor: u(p, s) (incoming), ū(p’, s’) (outgoing)

s-channel amplitude:

M_s = ū(p’, s’) [−ieγᵛ ε’_ν] · [i(/p + /k + m) / ((p+k)² − m²)] · [−ieγᵘ ε_μ] u(p, s)

u-channel amplitude:

M_u = ū(p’, s’) [−ieγᵘ ε_μ] · [i(/p − /k’ + m) / ((p−k’)² − m²)] · [−ieγᵛ ε’_ν] u(p, s)

Total amplitude: M = M_s + M_u.

IX.4 Simplification of the Denominators

Using on-shell conditions and momentum conservation:

(p+k)² − m² = p² + 2p·k + k² − m² = 2p·k (since p² = m², k² = 0)

(p−k’)² − m² = p² − 2p·k’ + k’² − m² = −2p·k’ (since p² = m², k’² = 0)

So:

M_s = −ie² ū(p’, s’) γᵛ ε’_ν [(/p + /k + m) / (2p·k)] γᵘ ε_μ u(p, s)

M_u = −ie² ū(p’, s’) γᵘ ε_μ [(/p − /k’ + m) / (−2p·k’)] γᵛ ε’_ν u(p, s)

IX.5 Squaring the Amplitude

The differential cross section is proportional to |M|², summed over final spins and averaged over initial spins. We use the Mandelstam variables:

s = (p + k)² = m² + 2p·k u = (p − k’)² = m² − 2p·k’ t = (p − p’)² = (k’ − k)² = −2k·k’

with the constraint s + t + u = 2m².

Computing |M_s + M_u|² and performing the spin sums (using the standard Dirac spinor completeness relation Σ_s u(p, s) ū(p, s) = /p + m) produces a trace over γ-matrices. The full trace is tedious but entirely mechanical; the result, after averaging and summing, is:

⟨|M|²⟩ = 2e⁴ [(p·k)/(p·k’) + (p·k’)/(p·k) + 2m²((1/(p·k)) − (1/(p·k’)))² − m⁴((1/(p·k)) − (1/(p·k’)))²]

where the ⟨⟩ indicates spin-average and spin-sum.

IX.6 The Lab Frame and the Compton Shift

In the lab frame where the initial electron is at rest (p = (m, 0, 0, 0)), the kinematic variables simplify:

p·k = mω, p·k’ = mω’

where ω and ω’ are the initial and final photon energies. Four-momentum conservation gives the Compton shift formula:

1/ω’ − 1/ω = (1/m)(1 − cos θ)

where θ is the scattering angle of the photon.

IX.7 The Klein-Nishina Cross Section

The differential cross section is:

dσ/dΩ = (1/(64π²)) · (ω’/ω)² · ⟨|M|²⟩ / (incoming flux · 4m²)

Substituting the expression for ⟨|M|²⟩ and the kinematics:

dσ/dΩ = (α²/(2m²)) (ω’/ω)² [(ω/ω’) + (ω’/ω) − sin²θ]

where α = e²/(4π) is the fine-structure constant. This is the Klein-Nishina formula for Compton scattering [9].

IX.8 What This Verifies

The LTD-derived QED reproduces the Klein-Nishina formula at tree level. The derivation used:

The Dirac field operator Ψ̂(x) from [1] §VII, whose existence was derived from dx₄/dt = ic via the second-quantization machinery of [1].
The QED vertex −ieγᵘ from §VII.2 of the present paper, derived as the vector coupling (not axial-vector) via the right-multiplication structure of condition (M) in §IV.4.
The photon propagator D^F_μν(k) = −ig_μν/(k² + iε) from §VII.3, derived from the Maxwell kinetic term built as the squared curvature of the x₄-orientation bundle connection.
The electron propagator S_F(q) = i(/q + m)/(q² − m² + iε) from [1] §VIII, with the iε prescription given a geometric reading tied to the sign of dx₄/dt.

No feature of the calculation required anything beyond dx₄/dt = ic and the derivations in [1], [2], and the present paper. The Klein-Nishina formula — verified experimentally in electron-photon scattering at keV to MeV photon energies, in cosmic-ray measurements, and in astrophysical gamma-ray contexts — emerges from LTD as a consequence of the principle applied at two levels (single-particle geometry and second-quantized field theory) plus the gauge-theory extension of the present paper.

Generalization. More generally, the entire tree-level structure of QED follows from LTD: any tree-level process can be computed using the Feynman rules derived here, and will produce the same amplitude as standard QED, because the Feynman rules themselves are identical to the standard ones (vertex factor −ieγᵘ, propagators S_F and D^F). The LTD derivation and standard QED are not competing theories with different tree-level predictions; they are the same theory derived from different foundations. What is distinctive about the LTD derivation is that every piece of the theory — the Dirac field, the vector-coupling structure, the masslessness of the photon, the U(1) gauge invariance, the absence of monopoles — has been derived from dx₄/dt = ic rather than assumed.

This is the concrete content of the LTD claim to be a foundational theory of QED: not that it makes different tree-level predictions than standard QED, but that the tree-level predictions that are verified experimentally emerge in the LTD framework as consequences of a single geometric principle, rather than as a collection of independent postulates (minimal coupling, gauge invariance, Lorentz invariance, microcausality, etc.) that must be imposed separately.

X. Discussion and the Pathway Forward

X.1 What Has Been Accomplished

The full tree-level QED has been derived from dx₄/dt = ic. Every structural feature has a geometric origin:

Local U(1) gauge invariance arises from the absence of a globally-preferred x₄-orientation reference direction (§III).
The gauge field A_μ is the connection on the x₄-orientation bundle (§V).
F_μν is the curvature of that connection (§VI.1).
Maxwell’s equations follow as the Bianchi identity (from F being a curvature) plus the variational field equations (from the free-Maxwell kinetic term) (§VI).
The gauge group is U(1) because the x₄-orientation direction is a complex phase, whose symmetry group is U(1) (§VIII.1).
The vector-coupling structure −eψ̄γᵘψA_μ (rather than axial-vector) is derived from the right-multiplication structure of condition (M) (§IV.4). This is a rigorous derivation that rules out the naive axial-vector coupling a left-multiplication ansatz would produce.
The photon is massless because a mass term would break gauge invariance, and geometrically because the photon is a pure x₄-oscillation without standing-wave structure (§VIII.2).
Magnetic monopoles are absent because the globally-defined +ic direction of x₄-expansion provides a global section of the x₄-orientation bundle, forcing its triviality — and any principal U(1)-bundle admitting a global section is trivial (§VIII.3, rigorous bundle-topology theorem).
The tree-level Klein-Nishina formula for Compton scattering is derived by explicit amplitude computation, not by citation of standard QED (§IX).

The resulting QED is equivalent at tree level to standard QED, but with every structural feature derived rather than postulated. This is the essential content of the present paper.

Four central new results distinguish this paper from both standard QED textbook treatments and from prior LTD sketches:

Local U(1) as forced, not assumed (§III.2). Local phase invariance is derived from the absence of a globally-preferred x₄-orientation reference in the LTD geometry, not imposed as a mathematical demand.
Vector coupling derived, not chosen (§IV.4). The right-multiplication structure of condition (M) forces the vector coupling over the axial-vector alternative that a naive left-multiplication ansatz would produce. The distinction between the two is made explicit rather than hand-waved.
Monopole absence as bundle-triviality theorem (§VIII.3). The principal-U(1)-bundle triviality follows rigorously from the global section provided by dx₄/dt = +ic; any bundle admitting a global section is trivial, and trivial bundles cannot support monopoles.
Explicit Compton amplitude computation (§IX). The Klein-Nishina formula is derived by direct amplitude computation using the LTD-derived Feynman rules, not merely asserted by citation.

X.2 What Remains: Renormalization and Loop Corrections

The present paper addresses only tree-level QED. Loop corrections — vacuum polarization, self-energy, vertex corrections — are computationally identical to standard QED once the tree-level structure is in place, but the renormalization procedure that handles the ultraviolet divergences has no explicit geometric origin in the framework developed here. The standard procedure (regularization, counterterm construction, renormalization group equations) is a computational tool whose LTD interpretation remains open.

Two possibilities suggest themselves, neither currently developed:

Possibility 1: Renormalization has no native LTD interpretation. The geometric framework gives tree-level QED from dx₄/dt = ic, and loop corrections are computed using standard procedures that happen to work but whose foundational status within LTD is unclear. This is the minimal position.

Possibility 2: Renormalization reflects the discrete structure of x₄. The FQXi paper [26] argued that x₄-expansion has a discrete wavelength at the Planck scale (ℏ as the quantum of action emerging from the discrete wavelength of x₄’s expansion). This discreteness might provide a natural UV cutoff, with renormalization parameters reflecting the Planck-scale structure. This is a more ambitious position that would require substantial development.

The present paper does not settle between these possibilities. It establishes tree-level QED as an LTD theorem and leaves the loop structure for future work.

X.3 The Electroweak Extension

The natural next step in the LTD derivation program is the electroweak sector. The companion broken-symmetries paper [3] argues that the full Spin(4) = SU(2)_L × SU(2)_R rotation group of Euclidean 4-space splits geometrically under the x₄-expansion, with SU(2)_L (transverse rotations) becoming the weak gauge group and SU(2)_R (rotations involving x₄) becoming the gravitational sector. The weak gauge group SU(2)_L combines with the U(1) derived in the present paper to give the electroweak gauge structure SU(2)_L × U(1).

The specific U(1) in the electroweak sector — the hypercharge U(1)_Y — is not the same as the QED U(1) derived here. The electroweak U(1)_Y combines with SU(2)_L under the Weinberg-angle mixing to produce the QED U(1)_EM that couples to the physical photon, plus the Z boson. The full derivation of this mixing — including the Higgs-mechanism-based derivation of the Weinberg angle and the W/Z masses — is the content of the planned electroweak paper.

Importantly, the QED U(1) of the present paper is the residual symmetry after electroweak symmetry breaking. The underlying electroweak structure must produce, at low energies, exactly the QED structure developed here. This is a consistency condition on the electroweak derivation, not an independent requirement.

X.4 QCD and the Strong Sector

QCD with SU(3) color is argued in [3] to arise from the three spatial dimensions x₁, x₂, x₃, which are equally transverse to x₄-expansion and form a natural triplet. The specific geometric origin of SU(3) (rather than SO(3) or SU(3) × flavor) and the dynamics of confinement and asymptotic freedom are open work. The strong CP conservation (θ_QCD < 10⁻¹⁰) emerges in [3] from the symmetric action of x₄-expansion on the three spatial dimensions, with no geometric room for a CP-violating phase. This is a successful prediction, but the full QCD Lagrangian from dx₄/dt = ic requires substantial additional development.

X.5 The Full Standard Model

The long-term goal is a full derivation of the Standard Model Lagrangian from dx₄/dt = ic. The components are:

Single-particle Dirac: [2]
Second-quantized Dirac: [1]
QED (the present paper): ✓
Electroweak: planned
Higgs mechanism: sketched in [3], needs full development
QCD: sketched in [3], needs full development
Three generations: sketched in [3] via Compton-frequency-interference argument, needs full quantitative derivation (the Jarlskog calculation discussed in [2] §X.6)
Mass hierarchy: open
CKM matrix: partially addressed in [3], quantitative derivation open

The program is far from complete. The present paper represents two rungs of a much longer ladder. Each rung successfully completed strengthens the case that the full Standard Model is derivable from dx₄/dt = ic. Each remaining rung is a concrete task whose completion or failure can be assessed on its own terms.

X.6 Falsifiable Predictions and Distinguishing Tests

The LTD derivation of QED reproduces standard QED at tree level exactly (as the Compton calculation of §IX demonstrates). This means LTD cannot be distinguished from standard QED by any tree-level measurement — a feature, not a bug, since a foundational theory that disagreed with parts-per-billion QED precision would be ruled out immediately. However, the framework makes a number of predictions that distinguish it from standard QED at other levels and that are in principle falsifiable.

Prediction 1 (absolute): No magnetic monopoles. The bundle-triviality argument of §VIII.3 establishes that monopoles are geometrically impossible in LTD, not merely experimentally suppressed. A single monopole observation would refute the framework. The current experimental status (null results across all monopole searches since 1931) is consistent with the prediction but does not uniquely confirm it.

Prediction 2 (absolute): Charge quantization in integer units. Electric charge is exactly quantized in integer multiples of a fundamental unit, as a consequence of the discrete x₄-orientation counting in the Fock space (derived in [1] §X.5 and cited in §VIII.3 above). Observation of a fractional electric charge on an isolated particle would refute the framework. The experimental status is consistent: no fractionally-charged free particles have been observed despite extensive searches.

Prediction 3 (structural): Photon exactly massless. The gauge-invariance argument of §VIII.2 requires the photon mass to be zero — not approximately zero, but exactly zero. Standard QED also predicts exactly zero photon mass (for the same gauge-invariance reason), so this is not a distinguishing prediction from standard QED, but it is a prediction of LTD that could be falsified by any observation of a nonzero photon mass. Current experimental bounds from measurements of the photon mass (from Coulomb’s law at large distances and from the magnetic fields of astrophysical objects) give m_γ < 10⁻²⁷ eV, consistent with zero.

Prediction 4 (structural): Absence of the spin-2 graviton. As developed in [2] §X.3 and summarized in [1] §X.5, the LTD framework predicts that gravity is not mediated by a spin-2 particle. This is a genuinely distinguishing prediction, because alternative approaches to quantum gravity (loop quantum gravity, string theory, asymptotic safety) generically predict a graviton. Observation of a graviton would refute LTD.

Prediction 5 (speculative, contingent on Possibility B from [1] §X.3): UV-finite QED at Planck scale. If the LTD framework incorporates a natural UV cutoff at the Planck scale via the discrete x₄-wavelength structure, then QED should be UV-finite and the Landau pole should be absent. This is a distinguishing prediction: standard QED has a Landau pole at enormous energies (far above the Planck scale, so unobservable), while LTD-regulated QED would not. The prediction is currently speculative and would require concrete development of the discrete-wavelength regularization to be turned into a quantitative test.

On uniqueness. The referee’s question of whether LTD is uniquely distinguished from alternative geometric frameworks was discussed at length in [1] §X.5. The same honest position applies here: LTD’s success in reproducing standard QED at tree level is evidence but not proof of uniqueness. What distinguishes the LTD derivation is that every structural feature of QED — gauge invariance, the vector coupling, the gauge group U(1), the masslessness of the photon, the absence of monopoles — follows from a single geometric principle dx₄/dt = ic rather than from a collection of independent postulates. This economy of principle is the argument for LTD’s foundational status, not a mathematical uniqueness proof.

A competing framework that could derive QED with comparable geometric content from a different foundational principle would be a worthy alternative; none currently exists with the derivational reach of LTD. This is a point in LTD’s favor but not a conclusive one, and the present paper does not claim otherwise.

XI. Conclusion

Quantum electrodynamics has been derived from the McGucken Principle dx₄/dt = ic — the foundational principle of Light, Time, Dimension Theory (LTD), which states that the fourth dimension is expanding at the rate of light. The derivation proceeds through the following chain:

dx₄/dt = ic → (via [2], recapitulated in Appendix A.1–A.3) → single-particle Dirac, condition (M), Clifford algebra, single-sided-action theorem → (via [1], recapitulated in Appendix A.4–A.5) → second-quantized Dirac field with anticommutation relations, Feynman propagator → (via present paper) → local x₄-phase invariance from absence of global orientation reference (§III), gauge field A_μ as connection on x₄-orientation bundle (§IV–§V), vector coupling derived from right-multiplication structure of (M) (§IV.4), Maxwell’s equations as Bianchi identity plus variational field equations (§VI), the full QED Lagrangian (§VII), masslessness of photon and rigorous bundle-triviality proof of monopole absence (§VIII), tree-level Klein-Nishina formula for Compton scattering via explicit amplitude computation (§IX).

Every link in this chain is a derivation. Nothing is imposed by fiat. The central conceptual content is that A_μ is the connection on the x₄-orientation bundle — the field that keeps track of how local x₄-phase frames rotate from point to point. Electromagnetism is, in the LTD reading, the physics of x₄-phase parallel transport. F_μν is the curvature of this transport; Maxwell’s equations are the Bianchi identity plus the field equations for the curvature; the photon is the quantum of the connection itself.

Four central new results distinguish this paper from prior QED derivations. First, local U(1) invariance is not assumed but forced by the absence of a globally-preferred x₄-orientation reference. Second, the vector-coupling form of the QED interaction is derived from the right-multiplication structure of condition (M) rather than chosen — the right-multiplication forces the coupling to emerge in the correct vector form under the geometric-algebra-to-matrix correspondence, and rules out the axial-vector coupling that a naive left-multiplication ansatz would produce. Third, the absence of magnetic monopoles is established as a rigorous bundle-triviality theorem: the globally-defined +ic direction of x₄-expansion provides a global section of the x₄-orientation bundle, and any principal U(1)-bundle admitting a global section is trivial, so no magnetic charges can exist. Fourth, the Klein-Nishina formula is derived by explicit amplitude computation from the LTD-derived Feynman rules, not merely asserted by citation to standard references.

The U(1) gauge group emerges from the single-complex-phase structure of x₄-orientation. The masslessness of the photon follows from gauge invariance, which is itself the geometric statement that x₄-phase has no globally preferred reference. The absence of magnetic monopoles follows from the bundle-triviality theorem, which is itself a direct consequence of the global +ic structure of the principle.

The tree-level Klein-Nishina formula — one of the most precisely tested predictions of QED, verified in electron-photon scattering at keV to MeV photon energies, in cosmic-ray measurements, and in astrophysical gamma-ray contexts — emerges from the LTD framework with every step derived from the McGucken Principle dx₄/dt = ic. The derivation establishes not a new theory but the geometric foundation of the familiar theory, with each postulate of standard QED shown to follow as a theorem from the foundational principle of LTD.

Together with the Dirac paper [2] and the second-quantization paper [1], the present paper completes the derivation of the first two rungs of the LTD program: single-particle and second-quantized fermionic matter coupled to the U(1) gauge field of electromagnetism. Higher rungs — electroweak unification, QCD, and the full Standard Model — are natural subsequent steps, each with its own specific derivational challenges and its own opportunities for the geometric framework to produce theorems where the Standard Model merely supplies postulates.

The accumulation of successful derivations — now extended to QED at tree level — is the appropriate standard by which the LTD program is to be judged. The present paper adds one more derivation to that accumulation, with every structural feature of tree-level QED traced back to a single geometric principle and with each rigor improvement in this revision (the vector-coupling derivation, the bundle-triviality theorem for monopole absence, the explicit Compton calculation, and the Appendix A sketches for independent readability) responding to specific concerns that a careful referee would raise about the standards of geometric derivation the framework is being held to.

Historical Note: The Origin of the McGucken Principle

“More intellectual curiosity, versatility and yen for physics than Elliot McGucken’s I have never seen in any senior or graduate student. Originality, powerful motivation, and a can-do spirit make me think that McGucken is a top bet for graduate school in physics. He could and did, and wrote it all up in a beautifully clear account.”

— Dr. John Archibald Wheeler, Joseph Henry Professor of Physics, Princeton University [10]

The McGucken Principle traces to Dr. Elliot McGucken’s undergraduate research with John Archibald Wheeler at Princeton University in the late 1980s and early 1990s. Wheeler-supervised projects on the Schwarzschild metric and the EPR paradox planted the seeds of the theory; the first written formulation appeared in the appendix of McGucken’s 1998 NSF-funded UNC Chapel Hill dissertation on an artificial retina chipset [12]. The principle appeared on internet physics forums (2003–2006), in five FQXi papers (2008–2013) [13, 14, 15, 16, 17], in seven books (2016–2017) [18, 19, 20, 21, 22, 23], and has been extensively developed in the comprehensive derivation program at elliotmcguckenphysics.com (2024–2026).

The present paper extends that program to quantum electrodynamics, establishing the full QED gauge structure — from local U(1) invariance to Maxwell’s equations to the tree-level Klein-Nishina formula — as a chain of theorems from dx₄/dt = ic.

Appendix A: Key Lemmas from [1] and [2]

The present paper relies on specific results from the Dirac paper [2] and the second-quantization paper [1]. For readers approaching this paper independently, we sketch the essential content here. The full proofs are in [1] and [2]; these sketches verify the logical chain.

A.1 From [2]: The Matter Orientation Condition (M)

Statement. An even-grade multivector Ψ in Cl(1,3) carries matter x₄-orientation at Compton frequency k if there exist Ψ₀ and x₄ such that:

Ψ(x, x₄) = Ψ₀(x) · exp(+I · kx₄), k > 0

with multiplication performed on the right. Antimatter corresponds to k < 0.

Why right-multiplication. The sign and the multiplication-side are fixed by the single-sided-action theorem ([2] §IV.3), which shows that only right-multiplication of the x₄-phase factor onto the rest-frame amplitude preserves consistency across all Lorentz generators. The right-multiplication structure is critical to §IV.4 of the present paper: it is what selects the vector coupling over the axial-vector coupling in the QED Lagrangian.

A.2 From [2]: Theorem on Uniqueness of Single-Sided Action

Statement. For any bivector R = exp(θ/2 · e_P) with e_P any bivector of Cl(1,3), left-action Ψ → RΨ preserves (M) while sandwich action R⁻¹ΨR does not preserve (M) for x₄-involving bivectors.

Essence of the proof. The pseudoscalar I anticommutes with vectors γᵘ but commutes with even-grade multivectors. For spatial bivectors (built from products of two spatial γ’s), I commutes with the bivector, and sandwich action leaves the x₄-rotor invariant. For x₄-involving bivectors (containing γ⁰), I anticommutes with the bivector, so R⁻¹IR has components along both +I and −I, introducing an antimatter admixture into a would-be matter field. Left-action avoids this because only one R acts on the field, not two. Full proof in [2] §IV.3.

A.3 From [2]: Half-Angle Rotor and 4π-Periodicity

Statement. Under a spatial rotation by angle θ, a matter spinor transforms as Ψ → exp(θ/2 · e_ij) Ψ. At θ = 2π, Ψ → −Ψ; only at θ = 4π does Ψ return to itself.

Essence of the proof. The half-angle arises because single-sided action applies only one factor of the rotor to Ψ (rather than the two factors that would appear in the sandwich action). Computing exp(θ/2 · e_ij) = cos(θ/2) + sin(θ/2) · e_ij (using e_ij² = −1 for spatial bivectors), and evaluating at θ = 2π gives cos(π) + sin(π) · e_ij = −1. The 4π-periodicity is the direct geometric content of the half-angle, not a convention.

A.4 From [1]: Fock Space and Anticommutation Relations

Statement. The physical Fock space 𝓕_phys is the antisymmetric subspace of the unrestricted tensor product space 𝓕_raw, where antisymmetry under exchange of identical x₄-oriented modes follows from the 4π-periodicity via the holonomy argument on the identical-particle configuration space Q_2. The creation and annihilation operators satisfy:

{a_p, a†_q} = δ³(p − q), {a, a} = {a†, a†} = 0

Essence of the derivation. The exchange path in Q_2 corresponds to a 2π relative rotation of the spinor frames, which by A.3 produces a minus sign under the fermionic spin structure selected by condition (M). This antisymmetry translates into the anticommutation relations via explicit operator-domain arguments applied to the antisymmetry constraint. Full derivation in [1] §VI.

A.5 From [1]: The Dirac Field Operator and Feynman Propagator

Statement. The Dirac field operator:

Ψ̂(x) = ∫ (d³p)/((2π)³) (1/√(2E_p)) Σ_s [a_{p,s} u_s(p) e^(−ip·x) + b†_{p,s} v_s(p) e^(+ip·x)]

has vacuum two-point function equal to the Feynman propagator:

S_F(x − y) = ⟨0|T{Ψ̂(x) Ψ̄̂(y)}|0⟩ = ∫ (d⁴p)/((2π)⁴) [i(/p + m) / (p² − m² + iε)] e^(−ip·(x−y))

The iε prescription is the operator-level manifestation of dx₄/dt = +ic: positive-frequency modes propagate forward in x₄, negative-frequency modes backward.

Essence of the interpretation. The iε selects contour deformations consistent with the fixed +ic direction of x₄-expansion. An alternative iε (with opposite sign) would correspond to dx₄/dt = −ic and would produce backward causal propagation, inconsistent with the LTD framework. Full derivation and geometric interpretation in [1] §VIII.

A.6 What These Lemmas Provide for the Present Paper

A.1 is used throughout for the matter orientation structure that couples to the gauge field.
A.2 is used in §IV.4 to justify right-multiplication as the correct action for x₄-phase rotations, which in turn selects the vector coupling in QED.
A.3 is used in the conclusion of §VIII.2 to justify the photon’s masslessness geometrically (photon as pure x₄-oscillation without standing-wave component).
A.4 supplies the Fock space and anticommutation structure used in §VII.4 and in the Compton calculation of §IX.
A.5 supplies the fermion propagator used in §IX.

A reader convinced of these six lemmas can verify the present paper’s derivations without requiring independent verification of [1] and [2]. Readers who want full rigor should consult [1] and [2] directly.

References

[1] McGucken, E. Second Quantization of the Dirac Field from the McGucken Principle of a Fourth Expanding Dimension dx₄/dt = ic: Creation and Annihilation Operators as x₄-Orientation Operators, Fermion Statistics as a Theorem, and Pair Processes as x₄-Orientation Flips. elliotmcguckenphysics.com (April 2026).

[2] McGucken, E. The Geometric Origin of the Dirac Equation: Spin-½, the SU(2) Double Cover, and the Matter-Antimatter Structure from the McGucken Principle of a Fourth Expanding Dimension dx₄/dt = ic. elliotmcguckenphysics.com (April 2026).

[3] McGucken, E. How the McGucken Principle of the Fourth Expanding Dimension (dx₄/dt = ic) Accounts for the Standard Model’s Broken Symmetries, Time’s Arrows and Asymmetries, and Much More. elliotmcguckenphysics.com (April 2026).

[4] Weinberg, S. The Quantum Theory of Fields, Volume I: Foundations. Cambridge University Press (1995).

[5] Peskin, M. E. & Schroeder, D. V. An Introduction to Quantum Field Theory. Westview Press (1995).

[6] Srednicki, M. Quantum Field Theory. Cambridge University Press (2007).

[7] Dirac, P. A. M. Quantised Singularities in the Electromagnetic Field. Proc. Roy. Soc. A 133, 60 (1931).

[8] Aharonov, Y. & Bohm, D. Significance of Electromagnetic Potentials in the Quantum Theory. Phys. Rev. 115, 485 (1959).

[9] Klein, O. & Nishina, Y. Über die Streuung von Strahlung durch freie Elektronen nach der neuen relativistischen Quantendynamik von Dirac. Z. Phys. 52, 853 (1929).

[10] Wheeler, J. A. Letter of Recommendation for Elliot McGucken. Princeton University, Department of Physics (late 1980s / early 1990s).

[11] McGucken, E. The McGucken Equivalence: Quantum Nonlocality and Relativity Both Emerge from the Expansion of the Fourth Dimension at the Velocity of Light. elliotmcguckenphysics.com (December 2024).

[12] McGucken, E. Multiple Unit Artificial Retina Chipset to Aid the Visually Impaired and Enhanced Holed-Emitter CMOS Phototransistors. Ph.D. Dissertation, University of North Carolina at Chapel Hill (1998).

[13] McGucken, E. Time as an Emergent Phenomenon: Traveling Back to the Heroic Age of Physics. FQXi Essay Contest, August 2008. https://forums.fqxi.org/d/238

[14] McGucken, E. What is Ultimately Possible in Physics? Physics! A Hero’s Journey… towards Moving Dimensions Theory. FQXi Essay Contest, September 2009. https://forums.fqxi.org/d/511

[15] McGucken, E. On the Emergence of QM, Relativity, Entropy, Time, iℏ, and ic from the Foundational, Physical Reality of a Fourth Dimension. FQXi Essay Contest, February 2011. https://forums.fqxi.org/d/873

[16] McGucken, E. MDT’s dx₄/dt = ic Triumphs Over the Wrong Physical Assumption that Time is a Dimension. FQXi Essay Contest, August 2012. https://forums.fqxi.org/d/1429

[17] McGucken, E. It from Bit or Bit From It? What is It? Honor!. FQXi Essay Contest, July 2013. https://forums.fqxi.org/d/1879

[18] McGucken, E. Light Time Dimension Theory: The Foundational Physics Unifying Einstein’s Relativity and Quantum Mechanics. 2016. ASIN: B01KP8XGQ6.

[19] McGucken, E. Relativity and Quantum Mechanics Unified in Pictures. 2017. ASIN: B01N2BCAWO.

[20] McGucken, E. The Physics of Time: Time & Its Arrows in Quantum Mechanics, Relativity, The Second Law of Thermodynamics, Entropy, The Twin Paradox, & Cosmology Explained via LTD Theory’s Expanding Fourth Dimension. 2017. ASIN: B07695MLYQ.

[21] McGucken, E. Quantum Entanglement and Einstein’s Spooky Action at a Distance Explained: The Nonlocality of the Fourth Expanding Dimension. 2017. ASIN: B076BTF6P3.

[22] McGucken, E. Einstein’s Relativity Derived from LTD Theory’s Principle. 2017. ASIN: B06WRRJ7YG.

[23] McGucken, E. The Triumph of LTD Theory and Physics over String Theory, the Multiverse, Inflation, Supersymmetry, M-Theory, LQG, and Failed Pseudoscience. 2017. ASIN: B01N19KO3A.

[24] Doran, C. & Lasenby, A. Geometric Algebra for Physicists. Cambridge University Press (2003).

[25] Hestenes, D. Real Spinor Fields. J. Math. Phys. 8, 798 (1967).

[26] McGucken, E. On the Emergence of QM, Relativity, Entropy, Time, iℏ, and ic from the Foundational, Physical Reality of a Fourth Dimension x₄ Expanding with a Discrete (Digital) Wavelength lp at c Relative to Three Continuous (Analog) Spatial Dimensions. FQXi Essay Contest, February 2011.

[27] Polchinski, J. Monopoles, Duality, and String Theory. Int. J. Mod. Phys. A 19S1, 145 (2004).

[28] Nakahara, M. Geometry, Topology and Physics. CRC Press (2nd ed., 2003).

Submitted to elliotmcguckenphysics.com, April 2026.

Author: Elliot McGucken, PhD — Theoretical Physics Undergraduate research with John Archibald Wheeler, Princeton University (late 1980s) Ph.D., University of North Carolina at Chapel Hill (1998) Website: elliotmcguckenphysics.com