The Running Fine-Structure Constant

Hello! Welcome back to our study of Quantum Field Theory.

In our last lesson, we established the on-shell renormalization scheme, a systematic procedure for taming the ultraviolet divergences that plague loop calculations. We saw that by defining counterterms to cancel these infinities, we are left with finite, but momentum-dependent, corrections to physical observables. The renormalized photon self-energy, $Π^(k^{2}) = Π_{loop} (k^{2}) - Π_{loop} (0)$ , was a key result.

Today, we will explore the profound physical consequences of this momentum dependence. This lesson's objective is to explain the physical significance of the running of the fine-structure constant $α$ . We will discover that the strength of the electromagnetic interaction is not a fundamental constant in the way classical physics might suggest, but rather an effective quantity that changes with the energy scale of an experiment. This phenomenon, known as "running," is a cornerstone of modern particle physics.

1. From Vacuum Polarization to a Running Coupling

Let's begin by connecting the renormalized vacuum polarization, $Π^(k^{2})$ , to the electromagnetic coupling.

In a tree-level scattering process, like electron-muon scattering, the interaction is mediated by a photon. The amplitude for this process is proportional to the vertex factors and the photon propagator, $- i g_{μν} / k^{2}$ . At one-loop, the vacuum polarization correction modifies this propagator. The sum of the tree-level propagator and the one-loop correction can be expressed as a geometric series, which sums to the "dressed" propagator:

\frac{- i g _{μν}}{k ^{2}} + \frac{- i g _{μν}}{k ^{2}} (i Π^(k^{2})) \frac{- i g _{μν}}{k ^{2}} + \dots = \frac{- i g _{μν}}{k ^{2} ( 1 - Π ^ ( k ^{2} ))}

Here, $k$ is the momentum carried by the photon. Notice the structure of this result. The amplitude for a process involving a photon with momentum $k$ is modified by the factor $1/ (1 - Π^(k^{2}))$ . We can absorb this correction into the definition of the coupling constant itself.

An interaction amplitude proportional to $\frac{e ^{2}}{k ^{2}}$ is replaced by $\frac{e ^{2}}{k ^{2} ( 1 - Π ^ ( k ^{2} ))}$ . This has the same form as the original tree-level amplitude if we define an effective, momentum-dependent fine-structure constant, $α (k^{2})$ :

α (k^{2}) = \frac{α ( 0 )}{1 - Π ^ ( k ^{2} )}

Here, $α (0) \equiv e^{2} /4 π \approx 1/137$ is the fine-structure constant we defined in our on-shell scheme—the value measured at zero momentum transfer (i.e., at long distances). The function $Π^(k^{2})$ describes how quantum fluctuations modify this value at different momentum scales.

2. The Physical Picture: Screening the Vacuum

What is the physical intuition behind this energy dependence? The term $Π^(k^{2})$ arises from vacuum polarization loops—virtual electron-positron pairs that are constantly popping in and out of existence. The vacuum is not empty; it's a dynamic, polarizable medium.

Qualitative Sketch of the Running Fine-Structure Constant — A qualitative sketch of the effective electromagnetic coupling constant $\alpha_{eff}$ as a function of distance $r$. At large distances (low energy), the coupling approaches the familiar value $1/137$. As we probe shorter distances (higher energy), we penetrate the screening cloud of virtual particles, and the effective charge appears stronger. This illustrates the "running" of the coupling.

Imagine placing a 'bare' electric charge in this quantum vacuum.

The positive charge attracts the virtual electrons and repels the virtual positrons.
This creates a cloud of virtual pairs around the bare charge, with a net positive charge density further out and a net negative charge density closer in.
This polarized cloud effectively screens the bare charge.

At large distances (corresponding to low momentum transfer, $k^{2} \to 0$ ), an observer sees the combined effect of the bare charge plus its screening cloud. This is the familiar, small, physical charge $e$ that defines $α \approx 1/137$ .

However, if we probe the charge with a high-energy photon (large $k^{2}$ ), we are essentially probing it at very short distances. We penetrate deeper into the screening cloud and see a larger fraction of the bare charge. Consequently, the effective coupling strength, $α (k^{2})$ , increases with energy.

The following video provides an excellent summary of this screening effect and presents the analytic form of the running coupling.

Particle Physics Lecture 27: Renormalization, QCD, Effective Theories and Strings

This clip from Alex Flournoy's lecture series explains the screening effect in QED, connecting it directly to the one-loop vacuum polarization diagram and deriving the functional form for the running of \alpha.

Please watch from 22:07 to 29:59. Pay attention to how the summation of bubble diagrams leads to the formula for \alpha(Q^2) and how the increasing nature of the coupling with momentum Q^2 corresponds to the physical picture of screening.

3. The Mathematical Form and Experimental Evidence

Your strong physics background will appreciate seeing the explicit mathematical form that underlies this physical picture. The one-loop calculation of the renormalized vacuum polarization for $k^{2} ≫ m_{e}^{2}$ yields:

Π^(k^{2}) \approx \frac{α}{3 π} (lo g \frac{- k ^{2}}{m _{e}^{2}} - \frac{5}{3})

Plugging this into our expression for $α (k^{2})$ and keeping the leading logarithmic term, we get:

α (k^{2}) \approx \frac{α ( 0 )}{1 - \frac{α ( 0 )}{3 π} lo g \frac{- k ^{2}}{m _{e}^{2}}}

This equation, often expressed as a solution to a Renormalization Group Equation (RGE), quantitatively describes the running.

Renormalization Group Techniques

For a more rigorous derivation, these lecture notes solve the one-loop Renormalization Group Equation for QED, providing the precise formula for the running \alpha(E).

Please read the sections 'Renormalization of QED' (pages 17-19) and the subsequent derivation for the running coupling (pages 19-20). Focus on eq. (100), which relates the QED beta-function to the field's anomalous dimension, and eq. (108), which is the final solution for 1/\alpha(E). This shows explicitly how the coupling increases with energy.

This running is not a mere theoretical curiosity; it is a precisely measured physical phenomenon.

At very low energies, experiments like the measurement of the electron's anomalous magnetic moment ( $g - 2$ ) determine $α (0)^{- 1} \approx 137.036$ .
At the energy scale of the Z boson mass, $E = M_{Z} \approx 91$ GeV, experiments at the Large Electron-Positron Collider (LEP) measured the effective coupling to be much stronger: $α (M_{Z}^{2})^{- 1} \approx 128$ .

This $\sim 6%$ increase is one of the most important radiative corrections in the Standard Model. Precision tests of the electroweak theory would fail without accounting for the running of $α$ .

Precision Tests of the Standard Model

This paper on precision tests of the Standard Model highlights the practical importance of the running of \alpha. It shows how the shift, denoted \Delta\alpha, is a numerically large and crucial correction.

Please read the short sections '2.1 Renormalization in QED' (focusing on eq. 26 and the definition of \Delta\alpha) and the bullet point about \Delta\alpha in section '4.1 Standard Model predictions' (eq. 110). This emphasizes how contributions from all fermions (leptons and quarks) are included and quantifies the numerical impact of the running.

4. Theoretical Consequences: The Landau Pole and a Glimpse Beyond

The formula for $α (k^{2})$ carries a startling implication. The denominator, $1 - \frac{α ( 0 )}{3 π} lo g \frac{- k ^{2}}{m _{e}^{2}}$ , will eventually become zero at some astronomical, but finite, energy scale. At this energy, the coupling $α (k^{2})$ would become infinite. This is known as a Landau pole.

The existence of a Landau pole suggests that QED, as an isolated theory, cannot be a complete description of physics up to arbitrarily high energies. The theory ceases to be predictive as the coupling blows up, and perturbation theory breaks down completely. This is a strong hint that new physics must emerge before the Landau pole is reached. Indeed, in the Standard Model, QED unifies with the weak force, altering this high-energy behavior.

Test your understanding!

The running of $α$ is caused by virtual fermion-antifermion loops. We've focused on electrons. If a new, heavy, charged lepton (like a "tau-prime" with mass $M ≫ m_{e}$ ) were discovered, how would it affect the running of $α$ at energies $E ≫ M$ ? Would the coupling run faster, slower, or the same?

Show answer

It would run faster. The total vacuum polarization $Π^(k^{2})$ is a sum over all charged fermions that can exist in the loops. At energies $E ≫ M$ , the new lepton would contribute its own vacuum polarization term, $Π^_{new} (k^{2})$ , roughly proportional to $\frac{α}{3 π} lo g (E^{2} / M^{2})$ . This adds to the electron's contribution, making the total $Π^(k^{2})$ larger and positive. Since $α (E^{2}) = α (0) / (1 - Π^(E^{2}))$ , a larger $Π^(E^{2})$ means the denominator decreases more rapidly with energy, causing $α (E^{2})$ to increase faster. Each new charged particle "threshold" we cross adds a new channel for vacuum screening.

To put QED's behavior in context, let's briefly contrast it with Quantum Chromodynamics (QCD), the theory of the strong force. This will be a key topic in a future module.

Particle Physics Lecture 27: Renormalization, QCD, Effective Theories and Strings

To fully appreciate the significance of QED's screening, it's illuminating to see the opposite behavior in QCD. This video explains how gluon self-interactions lead to anti-screening and 'asymptotic freedom'.

Watch from 29:59 to 38:08. Notice the extra diagrams involving gluon loops and how they contribute with an opposite sign to the fermion loops, leading to a completely different energy dependence for the strong coupling \alpha_s.

In QCD, the force carriers (gluons) also carry color charge and interact with each other. These gluon self-interaction loops contribute to vacuum polarization with an opposite sign to the quark loops. This effect, called anti-screening, dominates. As a result, the strong coupling, $α_{s}$ , decreases at high energies. This is called asymptotic freedom. It explains why quarks behave as almost free particles at high energies (e.g., inside a proton during a high-energy collision) but are strongly bound together (confined) at low energies.

Conclusion

The fact that coupling constants "run" with energy is a fundamental prediction of quantum field theory, with deep physical significance.

Key Takeaways:

The fine-structure constant, $α$ , is not a true constant but an effective parameter that depends on the energy scale of the interaction.
This running is caused by the screening of a bare charge by virtual electron-positron pairs polarizing the quantum vacuum.
Consequently, $α$ increases with energy. This is a real, experimentally verified effect, essential for precision tests of the Standard Model.
The behavior of QED suggests it is not a complete theory at all energy scales, as indicated by the Landau pole.
This behavior contrasts sharply with QCD, where gluon self-interactions lead to anti-screening and asymptotic freedom, causing the strong coupling $α_{s}$ to decrease with energy.

Preview of the Next Lesson:
In this and the previous lesson, we have relied on the fact that QED renormalization works out consistently. For example, the photon remained massless, and divergences cancelled cleanly. These are not accidents; they are consequences of the underlying gauge symmetry of the theory. In our next lesson, we will explore the Ward-Takahashi identity, the precise mathematical statement of this symmetry, and understand its crucial role in guaranteeing the renormalizability of QED.

Recommended resources

1. From Vacuum Polarization to a Running Coupling

2. The Physical Picture: Screening the Vacuum

3. The Mathematical Form and Experimental Evidence

4. Theoretical Consequences: The Landau Pole and a Glimpse Beyond

Conclusion