Topological Defects & Homotopy Groups

Introduction

In our last lesson, we discovered the deep connection between the topology of a system's order parameter space, $M$ , and the existence of stable topological defects. We established the fundamental classification principle: a defect of dimension $m$ in a $d$ -dimensional space is stable if the order parameter field, when evaluated on a surrounding sphere $S^{d - m - 1}$ , defines a non-trivial map into $M$ . This non-triviality is precisely what the homotopy groups $π_{d - m - 1} (M)$ measure.

Today, we transition from this principle to practice. Knowing that homotopy groups classify defects is one thing; knowing how to compute them for a given physical system is another. This lesson will equip you with the primary computational machinery used in physics to do just that. We will see that order parameter spaces often have the elegant mathematical structure of a coset space, $M = G / H$ , which allows us to use a powerful tool from algebraic topology—the long exact sequence of a fibration—to systematically determine the catalogue of possible defects.

Order Parameter Spaces as Coset Spaces

The structure $M = G / H$ is not an accident; it is a direct consequence of spontaneous symmetry breaking. Recall that a system described by a Hamiltonian with a symmetry group $G$ may settle into a ground state (or vacuum) that is only invariant under a smaller subgroup $H \subset G$ . This subgroup $H$ is called the isotropy group or stabilizer of that particular ground state.

The order parameter space $M$ is the set of all possible degenerate ground states. Since any ground state can be obtained by acting on a chosen reference state $ϕ_{0}$ with an element of $G$ , we can identify the points in $M$ with the elements of $G$ . However, this is not a one-to-one correspondence. Any two group elements $g_{1}$ and $g_{2}$ that differ only by an element of the isotropy group ( $g_{1} = g_{2} h$ for some $h \in H$ ) will map the reference state to the same new state: $g_{1} ϕ_{0} = (g_{2} h) ϕ_{0} = g_{2} (h ϕ_{0}) = g_{2} ϕ_{0}$ .

This means that all elements within a given left coset $g H$ are equivalent. The space of distinct ground states is therefore the space of these cosets, written as $M = G / H$ .

Let's ground this with a few key examples you've encountered:

3D Heisenberg Ferromagnet: A system of 3D spins has full rotational symmetry in its disordered phase, so $G = S O (3)$ . When it cools, the spins align along a specific, arbitrary direction, say $z^$ . The system is still symmetric under rotations around the $z^$ -axis. This remaining symmetry group is $H = S O (2)$ . The space of all possible magnetization directions (the surface of a sphere) is therefore $M = S O (3) / S O (2) ≃ S^{2}$ .
Nematic Liquid Crystals: These systems consist of rod-like molecules. The symmetry group is again $G = S O (3)$ . In the ordered phase, the molecules align along a common axis, but crucially, they have no head or tail (they are described by a headless vector or "director"). This means the ground state is invariant not only under any rotation around the director axis, but also under a $π$ rotation about any axis perpendicular to it. This isotropy group is the infinite dihedral group, $H = D_{\infty}$ . The order parameter space is $M = S O (3) / D_{\infty}$ , which is topologically equivalent to the real projective plane, $R P^{2}$ . This space is what you would get if you took a sphere and identified every point with its antipode.

The following thesis provides an excellent, mathematically sound overview of this formulation.

Applications of homotopy theory to topological solitons

This text formalizes the idea that order parameter spaces are homogeneous spaces, which can be identified with coset spaces.

Please read Section IV.2.1, "Coset spaces". The main goal is to understand the argument that for a compact Lie group G acting transitively on a space X, we have X \simeq G/H, where H is the stabilizer (isotropy group) of a point in X.

The Main Computational Tool: The Long Exact Sequence

Recognizing that $M = G / H$ is the crucial first step. The second is to exploit this structure. The projection map $p : G \to G / H$ that sends each group element $g$ to its coset $g H$ has a special structure known as a fibration. For our purposes, you can think of a fibration as a well-behaved projection where the "fibers" (the pre-images of points, which in our case are the cosets themselves) are all topologically identical to the subgroup $H$ .

The power of a fibration $H \to G \to G / H$ is that it induces a "long exact sequence" that connects the homotopy groups of all three spaces.

Applications of homotopy theory to topological solitons

This section introduces the concept of a fibration and its associated long exact sequence, which is the central theorem for our computations.

Read Section IV.1, "Fibrations". Focus on understanding the statement of Theorem 3, which presents the long exact sequence. Don't worry about the proof involving the Puppe sequence; focus on the structure of the sequence itself: \dots \to \pi_n(H) \to \pi_n(G) \to \pi_n(G/H) \to \pi_{n-1}(H) \to \dots \to \pi_1(G/H) \to \pi_0(H)

An exact sequence is a sequence of groups and homomorphisms where the image of each map is precisely the kernel of the next. The practical utility of this is immense: if we know some of the groups in the sequence are trivial (i.e., contain only the identity element, ${e}$ ), we can deduce relationships, often isomorphisms, between the non-trivial ones.

The Recipe for Computing Homotopy Groups

The long exact sequence provides a direct computational recipe. A particularly useful result arises when the larger group $G$ is topologically simple.

Theorem: For a fibration $H \to G \to G / H$ , if $G$ is path-connected and $π_{n} (G) = π_{n - 1} (G) = {e}$ , then the long exact sequence implies an isomorphism:
$π_{n} (G / H) ≃ π_{n - 1} (H)$

This powerful relation allows us to find the $n$ -th homotopy group of the order parameter space by calculating the ( $n - 1$ )-th homotopy group of the much simpler isotropy group.

There is one subtlety. Many important Lie groups, like $S O (3)$ , are not simply connected ( $π_{1} (S O (3)) = Z_{2} \neq = {e}$ ). To handle this, we can replace $G$ with its universal covering group $G ~$ , which is always simply connected ( $π_{1} (G ~) = {e}$ ). For any $n \geq 2$ , the homotopy groups are unchanged: $π_{n} (G) ≃ π_{n} (G ~)$ . The coset space can then be rewritten as $M = G / H ≃ G ~ / H ~$ , where $H ~$ is the corresponding subgroup in $G ~$ .

For $G = S O (3)$ , its universal covering group is $G ~ = S U (2)$ . Since $S U (2)$ is topologically a 3-sphere ( $S^{3}$ ), we know its homotopy groups are $π_{1} (S U (2)) = {e}$ , $π_{2} (S U (2)) = {e}$ , and $π_{3} (S U (2)) = Z$ . This makes $S U (2)$ the perfect starting point for our calculations.

Applications: Classifying Defects in Practice

Let's now apply this machinery to predict the defects in a few key physical systems. The following text provides detailed examples.

[PDF] Topological Defects in Physics

This document applies the group-theoretic approach to classify defects in magnetic systems and liquid crystals.

First, read Section 3.1, "Spins". Pay close attention to the calculation for the 3D spin system (S^2), where the author uses the SU(2)/U(1) representation to find \pi_1(S^2) and \pi_2(S^2). Next, read Section 3.2, "Nematics". This section tackles the more complex case of nematic liquid crystals where \mathcal{M} = \mathbb{R}P^2, and shows that \pi_1(\mathbb{R}P^2) = \mathbb{Z}_2.

Let's summarize the results and the logic:

Example 1: 3D Ferromagnet ( $M = S^{2}$ )

Symmetry Breaking: $G = S O (3) \to H = S O (2)$ , so $M = S O (3) / S O (2) ≃ S^{2}$ .
Universal Cover: We use $G ~ = S U (2)$ . The isotropy subgroup $S O (2)$ lifts to $H ~ = U (1)$ in $S U (2)$ . So we analyze the fibration $U (1) \to S U (2) \to S^{2}$ .
Compute $π_{1} (S^{2})$ (Vortices):
The relevant part of the long exact sequence is:
$\dots \to π_{1} (S U (2)) \to π_{1} (S^{2}) \to π_{0} (U (1)) \to \dots$
We know $π_{1} (S U (2)) = {e}$ . The group $U (1)$ is a circle, which is path-connected, so $π_{0} (U (1)) = {e}$ . The sequence becomes ${e} \to π_{1} (S^{2}) \to {e}$ , which forces $π_{1} (S^{2}) = {e}$ .
- Physical Conclusion: No stable line defects (vortices) in a 3D ferromagnet.
Compute $π_{2} (S^{2})$ (Monopoles):
The relevant part of the sequence is:
$\dots \to π_{2} (S U (2)) \to π_{2} (S^{2}) \to π_{1} (U (1)) \to π_{1} (S U (2)) \to \dots$
We know $π_{2} (S U (2)) = {e}$ and $π_{1} (S U (2)) = {e}$ . The sequence becomes ${e} \to π_{2} (S^{2}) \to π_{1} (U (1)) \to {e}$ . This implies an isomorphism: $π_{2} (S^{2}) ≃ π_{1} (U (1))$ . Since $U (1) ≃ S^{1}$ , we have $π_{1} (U (1)) = Z$ .
- Physical Conclusion: $π_{2} (S^{2}) = Z$ . The system supports a countably infinite number of topologically distinct point defects (monopoles and anti-monopoles), classified by an integer "charge".

Example 2: Nematic Liquid Crystals ( $M = R P^{2}$ )

Symmetry Breaking: $G = S O (3) \to H = D_{\infty}$ , so $M = R P^{2}$ .
Compute $π_{1} (R P^{2})$ (Vortices/Disclinations):
As shown in the resource, the calculation using the universal cover yields $π_{1} (R P^{2}) = Z_{2} = {0, 1}$ .
- Physical Conclusion: Nematic liquid crystals support one type of non-trivial line defect (called a disclination). The group structure $Z_{2}$ is fascinating: it means that two identical disclinations can combine and annihilate each other ( $1 + 1 = 0 (mod 2)$ ). A single disclination, however, is stable.

Left: A $\pi$ disclination in a nematic liquid crystal, where directors rotate by 180 degrees around the defect core. Center: a $2\pi$ disclination. Right: The order parameter space, a sphere with antipodal points identified ($\mathbb{R}P^2$). A loop from the south pole (A) to the north pole (A) is a non-trivial loop in $\mathbb{R}P^2$, corresponding to a $\pi_1(\mathbb{R}P^2) = \mathbb{Z}_2$ defect.

Compute $π_{2} (R P^{2})$ (Monopoles):
For $n \geq 2$ , the homotopy groups of a space and its universal covering space are identical. The universal cover of $R P^{2}$ is $S^{2}$ . Therefore, $π_{2} (R P^{2}) ≃ π_{2} (S^{2}) = Z$ .
- Physical Conclusion: Nematic liquid crystals also support stable point defects, classified by an integer charge, just like ferromagnets.

These examples show the power and generality of the method. By identifying the symmetry breaking pattern, you can unlock a systematic procedure to predict the entire taxonomy of topological defects a system can host.

Conclusion

In this lesson, we have moved from the abstract principle of defect classification to a concrete computational methodology rooted in the symmetries of the physical system.

Key Takeaways:

Order parameter spaces arising from spontaneous symmetry breaking $G \to H$ are naturally described as coset spaces $M = G / H$ .
The map $p : G \to G / H$ is a fibration, which induces a long exact sequence of homotopy groups.
This sequence is a powerful computational tool. For a simply connected group $G$ , it often yields the isomorphism $π_{n} (G / H) ≃ π_{n - 1} (H)$ .
When $G$ is not simply connected (like $S O (3)$ ), we use its universal covering group ( $S U (2)$ ) to simplify calculations.
By applying this recipe, we can compute the homotopy groups of $M$ and thereby predict the complete classification of stable topological defects (domain walls, vortices, monopoles, etc.).

Preview of the Next Lesson:
We have now concluded our exploration of topological defects that arise from the static, "ground state" properties of the order parameter manifold. Our focus will now shift to dynamics. In the next lesson, we will introduce the quantum adiabatic theorem, which governs how a system behaves when its parameters are changed slowly over time. This will lead us to discover another profound topological concept in quantum physics: the Berry phase, a geometric phase acquired by a quantum state during such an adiabatic evolution.

Recommended resources