Data from: Phase transition of topological index driven by dephasing
Data files
Nov 28, 2025 version files 6.50 MB
-
fig1a_data.csv
3.20 MB
-
fig1b_data.csv
3.19 MB
-
fig1c_data.csv
66.62 KB
-
fig1d_data.csv
5.78 KB
-
fig1e_data.csv
5.74 KB
-
fig2_data.csv
11.76 KB
-
fig3a_data.csv
538 B
-
fig3b_data.csv
13.28 KB
-
fig3c_data.csv
6.87 KB
-
README.md
3.05 KB
Abstract
We study topological insulators under dephasing noise. With examples of both a 2d Chern insulator and a 3d topological insulator protected by time-reversal symmetry, we demonstrate that there is a phase transition at finite dephasing strength between phases with nontrivial and trivial topological indices. Here, the topological index is defined through the correlation matrix. The transition can be diagnosed through the spectrum of the whole correlation matrix or of a local subsystem. Interestingly, even if the topological insulator is very close to the topological-trivial critical point in its Hamiltonian, it still takes a finite strength of dephasing to change the topological index, suggesting the robustness of topological insulators under dephasing. In the case of Chern insulator, this robustness of the phase with nontrivial Chern number persists near the critical point between the topological and Anderson insulator, which is tuned by the strength of disorder in the Hamiltonian.
Dataset DOI: 10.5061/dryad.44j0zpct9
Description of the data and file structure
Files and variables
File: fig1c_data.csv
Description: Data for Fig. 1c of the manuscript. Lists values of (\gamma, g_c) denoting the analytic phase boundary (in red) of the dephased Chern insulator, where \gamma is the model tuning parameter (Eq. 3), and g_c is the critical dephasing strength for that value of \gamma.
Variables
- \gamma versus g_c
File: fig1a_data.csv
Description: Data for Fig. 1a of the manuscript. Lists all eigenvalues of the flattened Hamiltonian H_C at \gamma=1.5 for each value of dephasing, g.
Variables
- g versus eigenvalue number
File: fig1b_data.csv
Description: Data for Fig. 1b of the manuscript. Lists all eigenvalues of the flattened Hamiltonian H_C at \gamma=2.5 for each value of dephasing, g.
Variables
- g versus eigenvalue number
File: fig1e_data.csv
Description: Data for Fig. 1e of the manuscript. Lists values of d\Delta_{sub}/dg, the derivative of the sub-matrix spectral gap with respect to dephasing strength, g, for the various system sizes at each value of g.
Variables
- d\Delta_{\sub}/dg versus g
File: fig3a_data.csv
Description: Data for Fig. 3a of the manuscript. Lists disorder-averaged values of the Chern number, C, for each value of disorder strength, \delta, at various system sizes.
Variables
- C versus \delta
File: fig2_data.csv
Description: Data for Fig. 2 of the manuscript. Lists values of (\delta t, g_c) denoting the analytic phase boundary (in red) of the 3D topological insulator, where \delta t is the model tuning parameter (Eq. 5), and g_c is the critical dephasing strength.
Variables
- g_c versus \delta t
File: fig1d_data.csv
Description: Data for Fig. 1d of the manuscript. Lists values of the sub-matrix spectral gap, \Delta_{sub}, for the various system sizes at each value of dephasing strength, g.
Variables
- \Delta_{sub} versus g
File: fig3c_data.csv
Description: This table contains the raw Chern number values (C) used to build the histogram in Fig. 3c.
Each row shows one realization of the calculation, and the two columns correspond to different lattice sizes: a 6×6 system (left) and an 8×8 system (right).
The numbers reflect how the computed Chern number fluctuates across realizations for each system size, with some values close to 0 and others near 2, showing the spread that the histogram summarizes.
File: fig3b_data.csv
Description: This table lists the individual Chern number values used to construct Fig. 3b.
Each row corresponds to one numerical realization, and the two columns show results for two lattice sizes: 6×6 (left) and 8×8 (right).
The values illustrate how the computed Chern number varies between realizations and between system sizes, forming the distribution shown in the figure.