A Is Chern number exactly an integer of half-integer?

  • Thread starter Thread starter Mohammad-gl
  • Start date Start date
  • Tags Tags
    Berry phase
Click For Summary
The discussion centers on calculating the Chern number for a specific system, with the user reporting non-integer results like 0.9. It is noted that Chern numbers should generally be integers, but non-integer values can occur in metallic systems where the band structure is not gapped. The user is employing a tight-binding Hamiltonian and Kubo formula in MATLAB, and there are suggestions to adjust the numerical methods for better accuracy. The conversation also touches on the need for specific references and potential coding errors in the MATLAB implementation. Overall, the Chern number should ideally be an integer, but adjustments and workarounds can help clarify the results.
Mohammad-gl
Messages
29
Reaction score
4
TL;DR
I'm trying to calculate the Chern number for a specific system. When I calculate the Chern number, it's not an integer, for example, I have 0.9 for it. Is it the wrong result or can I consider it as 1?
I'm trying to calculate the Chern number for a specific system. When I calculate the Chern number, it's not an integer, for example, I have 0.9 for it. Is it the wrong result or can I consider it as 1?
 
Physics news on Phys.org
What is a "Chem number"? Google just gives me the phone number of a company called "Chem Care".
 
What is the system you are considering? In general, yes, it should be an integer.

You could have non-integer Chern numbers, for example, in a metallic system, because you do not have a gapped band structure at the Fermi level (Chern number non well defined in the conventional sense). But there are workarounds to still define an integer Chern number.

What methods are you using to calculate it? If it's just numerics you might adjust your code to get better results.
 
I'm usi
dRic2 said:
What is the system you are considering? In general, yes, it should be an integer.

You could have non-integer Chern numbers, for example, in a metallic system, because you do not have a gapped band structure at the Fermi level (Chern number non well defined in the conventional sense). But there are workarounds to still define an integer Chern number.

What methods are you using to calculate it? If it's just numerics you might adjust your code to get better results.
I'm using tight-binding Hamiltonian, Kubo formula and MATLAB programming language. Is it ok if I calculate Chern number 1.93 or 0.7?
 
Mohammad-gl said:
I'm usi

I'm using tight-binding Hamiltonian, Kubo formula and MATLAB programming language. Is it ok if I calculate Chern number 1.93 or 0.7?
Are you following a specific reference?
 
I'm using this paper.
DOI: 10.1103/PhysRevB.106.205416

Can you help me with Matlab codes? If you can I will send you the code maybe I made a mistake in coding.
 

Similar threads

Where does the Berry phase of $\pi$ come from in a topological insulat
  • · Replies 3 ·
Replies
3
Views
6K
A Berry phase and gauge dependence
  • · Replies 6 ·
Replies
6
Views
3K
A Why the Chern numbers (integral of Chern class) are integers?
  • · Replies 12 ·
Replies
12
Views
7K
A Berry phase in condensed matter
  • · Replies 10 ·
Replies
10
Views
2K
A Order parameter, symmetry breaking Landau style
  • · Replies 5 ·
Replies
5
Views
3K
A Proving BF action is a difference of Chern-Simons actions
  • · Replies 0 ·
Replies
0
Views
827
A On the relationship between Chern number and zeros of a section
  • · Replies 3 ·
Replies
3
Views
3K
A Hodge decomposition of a 1-form on a torus
  • · Replies 2 ·
Replies
2
Views
3K
Why there is mod 2pi in Berry phase?
  • · Replies 8 ·
Replies
8
Views
4K
A Question about Berry phase in 1D polyacetylene
  • · Replies 2 ·
Replies
2
Views
3K