Robert-Jan Slager,Adrien Bouhon,F. Nur Ünal
Abstract:We comment on the recent paper ``Floquet non-Abelian topological insulator and multifold bulk-edge correspondence" by Tianyu Li and Haiping Hu, Nat. Comm. {\bf 14}, 6418 (2023). Apart from the fact that the authors unjustly imply to study multi-gap topology in Floquet systems for the first time, only known homotopic relations are presented. While such insights are used to present interesting Floquet phenomena and phases, which is an attractive result in itself, they cannot be used to deduce the total bulk characterization in the dynamical context without further proof. In fact, the authors essentially rephrase a Zak phase description. These results should in particular be contrasted to earlier results, <a class="link-https" data-arxiv-id="2208.12824" href="https://arxiv.org/abs/2208.12824">arXiv:2208.12824</a>, in which static-compatible Zak phases {\it and} dynamical Dirac strings were shown to be able to {\it distinguish} rather similar non-Abelian Floquet phases in $2+1$ dimensional systems. As a result, the claim of a sharp multifold bulk-edge correspondence cannot be concluded from the given arguments.
Mesoscale and Nanoscale Physics,Materials Science,Quantum Gases,Mathematical Physics,Quantum Physics
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to comment on and question the conclusions in the paper "Floquet non - Abelian topological insulators and multi - fold bulk - edge correspondence" published by Tianyu Li and Haiping Hu in Nature Communications. Specifically, the authors (Robert - Jan Slager, Adrien Bouhon, and F. Nur Ünal) believe that Li and Hu's research has the following problems:
1. **Inaccurate claim of the first study**:
- Li and Hu claimed that they were the first to study the characteristics of multi - band - gap topology in Floquet systems, but in fact, there have been similar research results before (such as arXiv:2208.12824), and these studies have already explored similar non - Abelian topological phenomena.
2. **Lack of sufficient proof**:
- Li and Hu used known homotopy relations to describe interesting Floquet phenomena and phases, but failed to provide enough evidence to prove that these results can be used to derive the overall characteristics of dynamic systems.
- They essentially reformulated the Zak phase description without further proof or new insights.
3. **Controversy over multi - fold bulk - edge correspondence**:
- Li and Hu proposed a "multi - fold bulk - edge correspondence", that is, different phases can be marked with the same invariant. However, this correspondence cannot explain different edge - state spectra, so it cannot comprehensively quantify the bulk invariants.
- Specifically, the authors pointed out that even if two phases have the same total charge \(q\), they may also show different edge - state spectra, which means that \(q\) is not sufficient to fully classify all dynamic Floquet phases.
4. **Comparison between static and dynamic systems**:
- Li and Hu's research mainly focused on 1 + 1 - dimensional systems and did not involve the braiding process in 2 + 1 - dimensional systems, which limits the identification of the Euler class.
- Some details in the static context (such as the definition of 1D charge) require more precise mathematical treatment to ensure the correct definition of parallel transport.
5. **Citation and presentation of previous work**:
- Li and Hu's citation of arXiv:2208.12824 in their paper was insufficient, and when answering reviewers' questions, their emphasis on their work as "the first" seemed inaccurate. In fact, arXiv:2208.12824 has already explored multi - band - gap - dependent non - Abelian charges and their braiding under periodic driving.
In summary, this review paper aims to point out some key problems in Li and Hu's research and emphasizes that more theoretical and experimental evidence is needed to support their conclusions. In particular, regarding the concept of "multi - fold bulk - edge correspondence", the authors believe that the current argument is not sufficient to support this statement.