Skip to main content

Welcome to a Grasp lesson

Create a free account to keep studying and create your own study path.

Lesson 65

Gauge Fields and Covariant Derivatives from Local U(1) Invariance

Recommended resources

Welcome back. In our previous lesson, we established a crucial distinction between global and local symmetries and saw how the demand for local, or gauge, invariance presents a fundamental challenge. A Lagrangian for a free field, like the Dirac field, is invariant under a global phase rotation but fails to be invariant when that rotation becomes a function of spacetime coordinates. The culprit, we found, is the ordinary derivative , which compares fields at infinitesimally separated points without accounting for the point-to-point change in the "phase reference frame."

This lesson will address that challenge head-on. Guided by the principle of local gauge invariance, our goal is to systematically modify the theory to restore the broken symmetry. We will construct a new type of derivative, the covariant derivative , which is designed to transform "covariantly" under local gauge transformations. In building this object, we will find that we are forced to introduce a new vector field—the gauge field —and derive its precise transformation properties. This process is the heart of gauge theory and provides the blueprint for deriving fundamental interactions from symmetry principles.

The Geometric Analogy: Derivatives and Parallel Transport

Before diving into the algebra, let's build an intuition based on a concept you are familiar with from your study of General Relativity: differentiating vectors on a curved manifold.

On a flat plane, subtracting two vectors at different points is straightforward because the basis vectors are the same everywhere. On a curved surface, like a sphere, the tangent space at each point is different. You cannot simply subtract a vector at point from a vector at a nearby point because they live in different vector spaces. To make a meaningful comparison, you must first parallel transport the vector from to along the connecting path. The covariant derivative is then the difference between the actual vector at and the parallel-transported version of the vector from .

This diagram illustrates the concept of the covariant derivative \(\nabla_v(w)\) for a vector field \(w\) on a curved manifold. To find the change in \(w\) along the direction of a vector \(v\) at point \(p\), we compare the vector \(w|_{p+\epsilon v}\) with the vector \(w|_p\) after it has been parallel transported along the curve \(C\) to the point \(p+\epsilon v\). The object that defines this parallel transport is the connection (in GR, the Christoffel symbols).

The situation with local gauge symmetry is perfectly analogous. A local U(1) transformation attaches a different "internal" phase reference frame at each spacetime point . Just as we need a connection to compare vectors in different tangent spaces in GR, we need a "gauge connection" to compare the phase of a field at with its phase at . The gauge field, , will play the role of this connection.


The Comparator and the Covariant Derivative

Let's formalize this idea of parallel transport for our U(1) phase symmetry. We need a way to compare and that respects the local gauge freedom. This requires an object, a "comparator" or "parallel transporter" , that connects the internal space at point to the internal space at point .

The following reading develops this idea from first principles, showing how the definition of a well-behaved derivative in the presence of local symmetry naturally leads to the introduction of a gauge field.

Quantum Theory I, Recitation 5 Notes

These notes from MIT OCW's Quantum Theory I course provide a wonderfully intuitive construction of the covariant derivative using the concept of a "comparator."

Please read the first two pages, focusing on these key steps: The problem: Review how promoting a global symmetry to a local one breaks the Schrödinger equation and makes the ordinary finite-difference derivative ill-defined. The solution: Understand the definition of the covariant derivative D_\mu in terms of the comparator U(y,x) and how this construction ensures the derivative transforms covariantly by design. The construction: Follow the derivation of the gauge field A_\mu from the infinitesimal form of the comparator U(x+n\epsilon, x) and see how this leads to the final expression for the covariant derivative.

The key insight from this reading is that by demanding a derivative that transforms simply under local phase rotations, we are forced to introduce a new entity, the vector gauge field , which acts as an infinitesimal parallel transporter. The final form, , shows that the covariant derivative is the ordinary derivative plus a correction term involving this new field. The transformation rule for the gauge field, , is precisely what is needed to make the whole structure work.


An Algebraic Derivation

The "comparator" approach provides excellent physical intuition. Now, let's re-derive the same result using a more direct algebraic method common in QFT literature. This approach is less about building from infinitesimal transport and more about enforcing a desired algebraic transformation property on the derivative operator itself.

We start by postulating the form of the covariant derivative, , where we treat as an unknown gauge field. (Note the different sign convention; this is a matter of definition). We then demand that the operator as a whole transforms in a way that preserves the covariance of .

Geometry of Mathematical Physics III (2021-22)

This excerpt from Stefano Cremonesi's notes on gauge theory presents a concise and rigorous derivation of the U(1) gauge transformation laws.

Please read section 5.2, "U(1) gauge symmetry." Pay close attention to the argument starting from Remark 5. The core idea is to require that the covariant derivative operator transforms as D'_\mu = e^{i\alpha} D_\mu e^{-i\alpha}. Follow the steps of the proof to see how this requirement uniquely determines the transformation law for the gauge field A_\mu.

This operator-based derivation beautifully complements the comparator approach. It shows that if we want to transform like (i.e., picking up a factor of ), the operator itself must transform "in the adjoint representation" of the gauge group. Working this out leads to the same transformation for the gauge field we found earlier:

and confirms the form of the covariant derivative:

The lecture in the following video walks through this same derivation, providing a dynamic explanation that ties the pieces together.

Particle Physics Lecture 12: Interactions via Local Gauge Symmetry (The Abelian Case)

This segment from Alex Flournoy's lecture on local gauge symmetry provides a clear, step-by-step derivation of the covariant derivative and the gauge field transformation.

Please watch from 18:14 to 32:06. The lecturer introduces the covariant derivative to fix the non-invariance of the Lagrangian and derives the necessary transformation law for the new gauge field A_\mu. Notice that he uses the convention D_\mu = \partial_\mu + iqA_\mu, which leads to A'_\mu = A_\mu - \frac{1}{q}\partial_\mu \alpha. This highlights how different conventions for the coupling constant (q, e, or g) and the sign in D_\mu are related.

At the end of this segment, the lecturer correctly identifies the transformation of as the familiar gauge transformation from electromagnetism. By demanding local phase invariance for a charged field, we have been forced to invent a vector field that transforms exactly like the electromagnetic four-potential.


The Deeper Picture: Gauge Fields as Connections on Fiber Bundles

The analogy between the covariant derivative in GR and the gauge covariant derivative is not merely a pedagogical tool; it reflects a deep, underlying mathematical unity. Both theories are described by the geometry of fiber bundles. In this framework, the gauge field is the local representation of a connection one-form on a principal bundle.

This is a highly abstract and mathematically advanced topic, but given your background, it is valuable to see how the concepts we've developed physically are realized in their full geometric glory. The following optional video gives a glimpse into this formalism.

Covariant derivatives - Lec 25 - Frederic Schuller

This excerpt from Frederic Schuller's lectures on the "Geometry of Mathematical Physics" provides a rigorous derivation of the covariant derivative from the perspective of fiber bundles. This is an advanced viewpoint but offers profound insight.

As an optional exploration, watch the segment from 59:43 to 1:13:01. The lecturer starts with abstract objects on the "principal bundle" (the space of all possible internal reference frames) and "pulls them back" to our familiar spacetime (the "base manifold"). You will see the local expression for the covariant derivative, \nabla_T S = dS(T) + A(T) \cdot S emerge naturally, where A is the pullback of the connection one-form—what physicists call the gauge field (or Yang-Mills field). Don't worry about catching every detail; the goal is to appreciate how the physical objects we've constructed fit into a larger, elegant mathematical structure.


Conclusion

In this lesson, we have taken the crucial step from identifying a problem (the failure of local invariance) to constructing the solution. We have derived the fundamental tools of gauge theory from the ground up by demanding that the laws of physics respect a local symmetry.

Key Takeaways:

  • The problem of differentiating fields under local gauge transformations is analogous to differentiating vectors on a curved manifold. Both require a connection to enable parallel transport.
  • The covariant derivative, , is a modified derivative that transforms covariantly under gauge transformations, meaning transforms just like .
  • To achieve this, we must introduce a gauge field (or gauge connection) . For a U(1) symmetry where a field transforms as , the covariant derivative and gauge field are:

    • (Note: conventions for the sign and placement of the charge can vary, but the underlying structure is the same).
  • The gauge field we were forced to introduce has exactly the transformation properties of the electromagnetic four-potential.

We now have the necessary machinery. We have a derivative, , that behaves properly under local gauge transformations. In our next lesson, we will replace the problematic ordinary derivative with our new covariant derivative in the Dirac Lagrangian. We will see that this simple replacement not only restores gauge invariance but also automatically generates a new term in the Lagrangian—the interaction term between the electron and the photon, which is the foundation of Quantum Electrodynamics (QED).