🔬 optics

Efficient local classification of parity-based material topology

This paper introduces a numerically efficient real-space framework utilizing the spectral localizer and a scalable sparse factorization algorithm to classify parity-based $\mathbb{Z}_2$ topology in aperiodic materials, overcoming the limitations of conventional momentum-space approaches by enabling local, energy-resolved topological diagnosis in systems lacking translational symmetry.

Original authors: Stephan Wong, Ichitaro Yamazaki, Chris Siefert, Iain Duff, Terry A. Loring, Alexander Cerjan

Published 2026-10-14

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Stephan Wong, Ichitaro Yamazaki, Chris Siefert, Iain Duff, Terry A. Loring, Alexander Cerjan

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Materials science has long relied on the idea that the world is made of repeating patterns. Just as a brick wall is built from identical bricks stacked in a regular grid, the atoms in most solids form a crystal lattice that repeats over and over. This repetition allows physicists to use powerful mathematical tools to predict how electrons move through the material, revealing whether the substance is a simple conductor, an insulator, or something far more exotic. Among the most fascinating of these exotic states are topological insulators. These are materials that act as insulators in their interior but conduct electricity perfectly along their edges, a property that is protected by the fundamental geometry of the electron waves. This "topological" protection means the conducting edge states are incredibly robust; they cannot be easily destroyed by impurities or defects, making them promising candidates for future electronics and quantum computers.

However, this entire framework of understanding depends on that perfect repetition. When a material lacks a repeating pattern—such as in a quasicrystal, where atoms are arranged in an ordered but non-repeating way, or in a disordered alloy—the standard mathematical tools break down. Without the repeating grid, the usual methods for classifying these materials fail, leaving scientists unable to tell if a strange, aperiodic material is a topological insulator or just a messy conductor. This gap in knowledge has been a significant hurdle, especially as new materials like twisted layers of graphene create complex, non-repeating structures that could host these valuable states.

A team of researchers has now developed a new way to look at these materials that does not require them to be perfectly ordered. Instead of trying to force the material into a repeating pattern to analyze it, they created a method that examines the material's local properties directly, point by point, in real space. Their approach allows them to identify topological phases in materials that are aperiodic, disordered, or even lack a clear energy gap between conducting and insulating states. By focusing on the local geometry of the electron interactions rather than the global symmetry of the crystal, they have unlocked a way to diagnose these materials that works where previous methods could not.

The core of this new method involves a mathematical tool called a spectral localizer. Imagine trying to understand the shape of a complex object by touching it with your hand at a single point. You can feel the curvature and the texture right under your fingers without needing to see the whole object. The spectral localizer does something similar for the quantum world. It takes the energy and position of the electrons in a specific small region of the material and combines them into a single mathematical structure. This structure reveals a hidden property called a topological invariant, which acts like a label telling the researchers whether that specific spot is part of a topological phase. Crucially, this label comes with a built-in measure of how "protected" the state is, telling scientists how much disturbance the material can withstand before losing its special properties.

To make this work for real-world materials, which can be incredibly large and complex, the researchers had to solve a difficult computational problem. The mathematical structure they use is a massive, sparse matrix—a giant grid of numbers where most of the entries are zero. To read the topological label, they needed to determine the sign of a specific value derived from this grid, known as the Pfaffian. Calculating this for such huge grids is usually impossible with standard computer methods because the calculations become too slow and memory-intensive. The team developed a new, highly efficient algorithm that can break down these massive, sparse grids into smaller, manageable pieces without losing the essential information. This allowed them to run the calculations on standard workstations, even for systems containing millions of atoms, a feat that was previously out of reach.

The researchers tested their new framework on two very different types of materials to prove it worked. First, they looked at a theoretical model of a quasicrystalline heterostructure, a material made of two layers of non-repeating patterns stacked together. In this system, they successfully identified the quantum spin Hall effect, a specific type of topological behavior where electrons flow without resistance along the edges. They were able to map out exactly where these edge states appeared and how they changed as they adjusted the coupling between the layers. Importantly, their method worked even when the material did not have a clear energy gap, a situation where older methods would have failed completely. They found that the topological state remained robust and could be precisely located, even in the presence of disorder.

Next, they applied the method to a photonic quasicrystal, a structure designed to control light rather than electricity. This material was built from rods and squares arranged in an Ammann-Beenker pattern, a type of non-repeating tiling. The goal was to find a fragile topological phase, a delicate state that can be easily destroyed by certain perturbations but is still topologically distinct. Using their efficient algorithm, they analyzed a system with over three million unknown variables, a scale that would have been computationally impossible with previous dense-matrix approaches. They successfully identified the topological phase and pinpointed the exact frequencies of light where the material's behavior changed. The results showed that the topological phase transition occurred exactly where the local protection measure dropped to zero, confirming that their method could reliably diagnose these complex states.

The significance of this work lies in its ability to unify the study of topological materials. Previously, scientists had to use different, often incompatible tools for crystals, quasicrystals, and disordered systems. This new framework provides a single, robust tool that works across all these different types of matter. It removes the need for the material to have a perfect repeating structure or a clean energy gap, opening the door to exploring topological phases in a vast array of natural and synthetic materials that were previously too difficult to analyze. By combining a deep mathematical insight with a practical, scalable computer algorithm, the researchers have provided a way to see the hidden order in the apparent disorder of complex materials, paving the way for the discovery of new states of matter in the next generation of electronic and photonic devices.

Technical Summary: Efficient Local Classification of Parity-Based Material Topology

Problem Statement
The classification of topological phases in crystalline materials is well-established through momentum-space approaches relying on translational symmetry, such as Chern numbers and the $Z_2$ index. However, these conventional methods fail for aperiodic systems (e.g., quasicrystals, amorphous materials, and disordered systems) where translational symmetry is absent. Existing real-space alternatives for $Z_2$ topological insulators (Class AII) often rely on projected position operators (e.g., Bott index, spin Bott index) or scattering matrices. These methods face significant limitations: they typically assume the existence of a spectral gap in the Hamiltonian and sometimes a gap in spin projections, rendering them inapplicable to gapless systems common in realistic disordered models. Furthermore, projector-based approaches necessitate the use of dense matrices, restricting analysis to low-energy approximations and requiring the prior identification of such approximations. There is a critical need for a numerically efficient, real-space framework that can classify parity-based topology in aperiodic systems without relying on spectral gaps, spin resolution, or low-energy approximations.

Methodology
The authors present a framework based on the spectral localizer combined with a novel, scalable sparse numerical algorithm.

Spectral Localizer Framework:
The approach utilizes the spectral localizer operator $L_{(x,E)}$ , constructed from the system's Hamiltonian $H$ and position operators $X, Y$ . For 2D Class AII systems with fermionic time-reversal symmetry ( $T^2 = -I$ ), the framework exploits the system's symmetries to transform $L_{(x,E)}$ into a purely imaginary, skew-symmetric matrix in a specific basis. The local topological invariant is determined by the sign of the Pfaffian of this skew-symmetric matrix:
$\xi^L_{(x,E)} = \text{sign}[\text{Pf}(Q^\dagger L_{(x,E)} Q)] \in Z_2$
Crucially, this method does not require the system to possess a bulk spectral gap or a gapped spin sector. It provides a local, energy-resolved invariant accompanied by a local gap measure $\mu_{(x,E)}$ , which quantifies topological protection and indicates the presence of localized boundary states when $\mu \to 0$ .
Sparse Pfaffian Sign Algorithm:
To make this framework computationally feasible for large physical systems, the authors developed a scalable algorithm to compute the sign of the Pfaffian for large sparse skew-symmetric matrices without calculating the full determinant or Pfaffian.
- Factorization: The method relies on the factorization $S = L T L^\top$ , where $S$ is the sparse skew-symmetric matrix, $L$ is a unit-diagonal lower-triangular matrix, and $T$ is a skew-symmetric block-diagonal matrix.
- Sign Computation: The sign of the Pfaffian is derived from the signs of the diagonal entries of $L$ and the off-diagonal blocks of $T$ : $\text{sign}[\text{Pf}(S)] = \text{sign}[\det(L)] \cdot \text{sign}[\text{Pf}(T)]$ .
- Implementation: The algorithm utilizes the Tacho software package (part of the Trilinos framework) and employs a two-stage permutation strategy. First, a maximum cardinality or weighted matching algorithm ( $P_s$ ) moves large numerical entries to the subdiagonal to ensure stable $2\times2$ pivots. Second, a nested dissection algorithm ( $P_n$ , via METIS) reorders the matrix to minimize fill-in during factorization. This approach allows the method to handle matrices with millions of degrees of freedom on standard workstations, significantly outperforming dense algorithms in runtime scaling.

Key Results
The framework was applied to two distinct aperiodic systems:

Quasicrystalline Quantum Spin Hall (QSH) Systems:
The authors modeled a heterostructure of two coupled quasicrystalline layers (based on Penrose tiling) with opposite Chern numbers.
- Gapped Systems: The method successfully identified the non-trivial $Z_2$ topology and associated edge states in the gapped regime.
- Gapless Systems: The framework was applied to a gapless heterostructure where one side was gapped and the other gapless. Despite the absence of a global spectral gap, the local index $\xi^L$ and local gap $\mu$ correctly diagnosed the topological phase and predicted "leaky" boundary-localized states where the local gap closed. This demonstrates the method's ability to function where projector-based methods fail.
Photonic Quasicrystals with Fragile Topology:
The method was applied to a 2D photonic quasicrystal (Ammann-Beenker tiling) with $C_2T$ symmetry (bosonic time-reversal).
- The system was discretized using finite differences, resulting in a sparse matrix of approximately $3.2 \times 10^6$ dimensions.
- The algorithm successfully computed the fragile topological invariant across a range of frequencies in 1–2 hours on a standard workstation.
- The identified topological phase transitions coincided with the vanishing of the local gap, confirming the accuracy of the sparse factorization strategy.

Significance and Claims
The paper claims that the combination of the spectral localizer framework and efficient sparse numerical methods provides a unified and robust tool for diagnosing parity-based topological phases in aperiodic electronic, photonic, and acoustic materials.

Overcoming Limitations: Unlike previous approaches, this method does not rely on translational symmetry, bulk spectral gaps, or gapped auxiliary operators (like spin projections). It operates directly on the full Hamiltonian, eliminating the need for low-energy approximations.
Scalability: The development of the sparse Pfaffian sign algorithm enables the analysis of realistic material models with millions of degrees of freedom, a scale previously inaccessible to dense linear algebra methods.
Physical Insight: The framework provides not just a binary classification but an intrinsic measure of topological protection (the local gap) and a direct link to bulk-boundary correspondence via the localization of states where the gap closes.
Broad Applicability: The authors anticipate these methods will be valuable for the systematic discovery of topological phases in complex platforms such as twisted van der Waals systems and large-scale metamaterials where conventional momentum-space diagnostics are inapplicable.

The work establishes a direct mathematical connection between real-space topological classification and the homotopy classes of large sparse operators, offering a practical pathway for analyzing topological matter in the absence of periodicity.

Technical Summary: Efficient Local Classification of Parity-Based Material Topology

More like this