The Zak phase in topologically insulating chains: invariants and quaternionic constraints

This paper investigates the topological content of the Zak phase in one-dimensional topological insulators across all Altland-Zirnbauer-Cartan symmetry classes, demonstrating how to construct a $\mathbb{Z}_2$-valued invariant from the Zak phase while revealing that quaternionic structures imposed by specific anti-unitary symmetries force this invariant to vanish.

The Gist

Imagine you are walking through a vast, endless city made of identical blocks. This city represents a crystal (like a piece of metal or a semiconductor). In this city, electrons are the people walking around.

Usually, we think of electrons just as tiny balls bouncing around. But in Topological Insulators, these electrons behave more like dancers in a choreographed routine. They can't just stop or change their dance moves easily; they are locked into a specific pattern by the "rules of the city" (symmetries).

This paper is a mathematical detective story about how to count and classify these dance routines, specifically in one-dimensional chains (like a single long line of atoms).

Here is the breakdown of the paper's story, using simple analogies:

1. The Map and the Compass (The Zak Phase)

Imagine you are walking around a giant circular track (the "Brillouin Zone"). As you walk, you carry a compass. In normal materials, if you walk a full circle, your compass points in the same direction when you get back to the start.

But in these special topological materials, your compass might spin around once, twice, or three times before pointing the same way again. This "spin" is called the Zak Phase.

• The Paper's Job: The authors wanted to know: Can we use this spinning compass to tell us exactly what kind of topological "dance" the electrons are doing?

2. The Ten-Fold Rulebook (Symmetry Classes)

In physics, there is a famous "Periodic Table" (called the Ten-Fold Way) that sorts all possible topological materials into 10 different categories. These categories depend on three "rules" the electrons must follow:

• Time Reversal: If you play the movie of the electrons backward, does it look the same?
• Particle-Hole: If you swap an electron for a "hole" (a missing electron), does the physics stay the same?
• Chiral: Is there a specific left-right handedness to the dance?

The authors checked all 10 categories to see if the "spinning compass" (Zak Phase) could act as a unique ID card for each one.

3. The Magic Trick: The "Quaternionic" Trap

Here is the big discovery. The authors found that for most categories, the compass spin tells you a simple "Yes" or "No" (or a number like 0 or 1). This is a Z2 invariant—a simple binary switch.

However, they found a special "trap" in some categories.
Imagine you have a special pair of glasses (a Quaternionic Structure) that forces the compass to behave in a very specific way. If the material has this special "glasses" property (caused by a specific type of symmetry where the math squares to -1), the compass is forced to spin an even number of times.

• The Result: Because it must spin an even number of times, the "Yes/No" switch (the Z2 invariant) always reads Zero.
• The Lesson: The compass isn't broken; it's just that the "glasses" force the answer to be zero. The compass loses its ability to tell you if the material is "topologically interesting" or "boring" because the geometry of the space itself forces it to be zero.

4. The Kitaev Chain (The Real-World Example)

To prove this, the authors looked at a famous model called the Kitaev Chain (a theoretical chain of atoms that can host "Majorana particles," which are their own antiparticles).

• They took a standard Kitaev chain and added more complex connections (hopping further away, not just to the neighbor).
• They showed that even with these complex connections, the "spinning compass" still worked perfectly to tell them the parity (whether the number is odd or even) of the material's topological complexity.
• It's like checking if a number is odd or even by looking at the last digit. The compass can't tell you the exact number (like 5 or 7), but it can tell you if it's odd or even.

Summary: What did they actually do?

1. Built a Toolkit: They created a rigorous mathematical method to define a "symmetric compass" (Bloch basis) that respects the rules of the material.
2. Tested the Compass: They checked if this compass works for all 10 types of topological materials.
3. Found the Limitation: They discovered that in materials with a specific "quaternionic" symmetry, the compass is forced to give a zero answer, hiding the true topological nature of the material.
4. Verified with Examples: They showed that for the most common interesting case (Class BDI), the compass successfully identifies whether the material is in an "odd" or "even" topological state.

Why does this matter?

In the real world, we want to build quantum computers using these materials. These computers rely on "Majorana particles" at the edges of the chain. To build them, we need to know exactly what state our material is in.

This paper tells us: "Hey, if you use the Zak Phase (the compass) to check your material, be careful! If your material has a specific symmetry (the quaternionic structure), the compass will always say 'Zero,' even if the material is actually very complex. You need to know your material's 'glasses' before you trust the compass."

It's a guide for physicists to avoid false negatives when trying to identify these exotic quantum states.

Technical Summary

1. Problem Statement

The paper addresses the classification of topological phases in one-dimensional (1D) translation-invariant topological insulators. While the "ten-fold way" (Altland–Zirnbauer–Cartan or AZC classification) provides a rigorous K-theoretic framework for categorizing these phases based on discrete symmetries (Time-Reversal $T$, Particle-Hole $C$, and Chiral $S$), there is a need to connect these abstract classifications to concrete, computable physical markers.

Specifically, the authors investigate the Zak phase (the geometric Berry phase accumulated by a Bloch wavefunction across the Brillouin zone) as a topological marker. The central questions are:

1. To what extent does the Zak phase capture the full topological content of all 10 AZC symmetry classes in 1D?
2. Can a gauge-invariant topological invariant be constructed from the Zak phase for all symmetry classes?
3. How do specific geometric structures, particularly quaternionic structures (induced by anti-unitary symmetries squaring to $-1$), constrain or trivialize these invariants?

2. Methodology

The authors employ a rigorous mathematical framework combining fiber bundle theory, spectral analysis, and functional analysis.

• Model Setup: They model non-interacting topological insulators using tight-binding lattice Hamiltonians on a 1D lattice ($\mathbb{Z}$) with $N$ internal degrees of freedom. The Hamiltonian is translation-invariant and has finite-range hopping.
• Bloch-Floquet Decomposition: Using the Bloch-Floquet transform, the Hamiltonian is decomposed into a family of fiber Hamiltonians $H_k$ parameterized by the quasi-momentum $k$ on the Brillouin torus ($S^1$).
• Spectral Projections: Assuming a spectral gap around zero energy, they define spectral eigenprojections $P_k^{(-)}$ onto the negative energy bands using the Riesz formula. These projections define the Bloch bundle.
• Symmetric Bloch Bases: A critical step is the construction of symmetric Bloch bases $\{v_j(k)\}$ that respect the discrete symmetries of the system ($T, C, S$).
  • They utilize parallel transport operators ($T_k$) to construct a global frame for the bundle.
  • They explicitly define how these bases transform under symmetry operators $O(\epsilon_k, \epsilon_E)$, distinguishing between cases where symmetries square to $+1$ or $-1$.
• Zak Phase Definition: The Zak phase is defined as the integral of the Berry connection over the Brillouin zone. The authors analyze how this phase changes under gauge transformations (changes of basis) compatible with the symmetries.
• Quaternionic Analysis: They investigate the specific case where an anti-unitary symmetry operator squares to $-1$ (inducing a quaternionic structure on the Hilbert space). They analyze the constraints this imposes on the determinant of the transition matrices and the resulting winding numbers.

3. Key Contributions

A. Construction of a $\mathbb{Z}_2$-Valued Invariant

The authors generalize previous results (which focused on chiral or particle-hole symmetric classes) to all AZC symmetry classes. They define a topological invariant $I^{(AZC-class)}(H)$ derived from the Zak phase:
$$I^{(AZC-class)}(H) = \left[ \frac{1}{\pi i} \int_{S^1} \sum_{j=1}^m \langle v_j(k), \partial_k v_j(k) \rangle dk \right] \mod 2$$
where $\{v_j(k)\}$ is a symmetric Bloch basis for the occupied bands.

• Gauge Invariance: They prove that for symmetry classes possessing chiral ($S$) or particle-hole ($C$) symmetries (specifically types $O(+,-)$ and $O(-,-)$), the change in the Zak phase under a symmetric gauge transformation is always an even integer. Consequently, the Zak phase modulo 2 is a well-defined, gauge-invariant topological quantity.
• Triviality in Other Classes: For classes with only Time-Reversal symmetry ($T$) or no symmetry, the invariant is trivial (always 0 or undefined in this specific $\mathbb{Z}_2$ formulation).

B. The Quaternionic Constraint (Vanishing Theorem)

A major theoretical contribution is the demonstration that the presence of a quaternionic structure forces the invariant to vanish.

• If the system possesses an anti-unitary symmetry $O$ such that $O^2 = -1$ (e.g., Time-Reversal in class AII or DIII, or Particle-Hole in class CII), the Hilbert space admits a quaternionic structure.
• Theorem 5.1: In such classes, the winding number associated with the Zak phase is necessarily an even integer.
• Result: The $\mathbb{Z}_2$ invariant $I^{(AZC-class)}(H)$ becomes identically zero ($0 \in \mathbb{Z}_2$).
• Implication: The vanishing of the invariant in these classes does not necessarily imply a topologically trivial phase (as predicted by K-theory, some classes like AII have trivial K-theory, but DIII has $\mathbb{Z}_2$). Instead, the vanishing is an artifact of the geometric quaternionic structure preventing the Zak phase from detecting the non-trivial topology. The Zak phase is "blind" to the topology in these specific symmetry classes.

C. Application to Generalized Kitaev Chains (Class BDI)

To illustrate the utility of the invariant, the authors analyze Class BDI (which has $T^2=1, C^2=1, S^2=1$).

• K-Theory Prediction: Class BDI in 1D is predicted to have a $\mathbb{Z}$-valued topological invariant (the winding number).
• Zak Phase Result: The constructed invariant $I^{(BDI)}(H)$ yields the parity of the K-theoretic winding number.
• Generalization: They extend this to generalized Kitaev chains with arbitrary finite-range hopping and multiple chiral channels (multi-band). They show that the invariant correctly computes the parity of the winding number of the determinant of the off-diagonal block of the Hamiltonian.

4. Results Summary

Symmetry Class	K-Theory Invariant (1D)	Zak Phase Invariant $I^{(AZC)}$	Reason
A, AI, AII	0	0	No chiral/particle-hole symmetry to enforce even winding.
AIII	$\mathbb{Z}$	$\mathbb{Z}_2$	Captures parity of winding number.
BDI	$\mathbb{Z}$	$\mathbb{Z}_2$	Captures parity of winding number.
D	$\mathbb{Z}_2$	$\mathbb{Z}_2$	Matches K-theory.
DIII, CII, CI	$\mathbb{Z}_2$ or $\mathbb{Z}$	0	Quaternionic structure ($O^2=-1$) forces even winding, trivializing the $\mathbb{Z}_2$ marker.
C	0	0	Trivial.

• Key Finding: The Zak phase is a sensitive probe of the geometry of the Bloch bundle. In classes with quaternionic structures, the geometry forces the topological marker to vanish, even if the K-theoretic classification suggests a non-trivial phase (e.g., Class DIII).

5. Significance and Outlook

• Bridging Theory and Computation: The paper provides a concrete computational recipe (via symmetric Bloch bases) to extract topological information from 1D models, bridging the gap between abstract K-theory and physical observables.
• Limitations of the Zak Phase: It highlights a fundamental limitation: the Zak phase (and its $\mathbb{Z}_2$ reduction) is not a complete invariant for all symmetry classes. Specifically, it fails to distinguish topological phases in classes with quaternionic structures (like DIII and CII) because the geometric constraints of the symmetry force the invariant to zero.
• Geometric vs. Topological: The work distinguishes between topological invariants (K-theory) and geometric constraints (quaternionic structure). The vanishing of the Zak phase invariant in certain classes is a signature of the underlying quaternionic geometry rather than topological triviality.
• Future Directions: The authors suggest extending this analysis to higher dimensions (2D and 3D) to investigate "weak invariants" and the robustness of these markers in the presence of disorder or interactions.

In conclusion, the paper rigorously establishes that while the Zak phase is a powerful tool for characterizing 1D topological insulators, its efficacy is strictly bounded by the specific discrete symmetries of the system, particularly the presence of anti-unitary symmetries squaring to $-1$.