Questions about the paper not the code

20 views
Skip to first unread message

Jaime Garrido

unread,
Jul 8, 2024, 9:24:06 AMJul 8
to TB2J
Hi everyone!

I have been revisiting the original TB2J paper in order to try to try to fully understand the derivation of the Hamiltonian parameters. I thought about contacting the authors through email, but then I realized that this place might be a better option since the answer can be read later by the rest of the users.

I have two questions basically. Sorry if the answer is obvious :/ 
  1.  The first one is about the paragraph between equations (7) and (8) where we can read: "In condensed form, the spin and orbital matrix for site i is now Pi = pi^0*I + \vec{p}i \cdot \vec{sigma}  ". I am getting confused with the dimensions of the matrices. Pi is 2Norb X 2Norb. and the Pauli matrices are just 2x2. Then, how is the equality possible? I guess there is something implicit that I am not following.
  2. Before arriving to equation (13): "The change of H due to the rotation of the spin is \delta \vec{\phi} x \vec{p}...". Why is it exactly that quantity? I do not have any clue about where that relation is coming from. 

My ultimate goal is being able to write equation (13) from equation (12).

Thanks in advance!

Jaime

Xu He

unread,
Jul 8, 2024, 3:09:14 PMJul 8
to TB2J
Hello Jaime, 
Thanks. Indeed these questions are valuable to the community. 
1. This operation is for each orbital of one atom.   Pi is thus a 2x2 matrix, pi is a scalar, and \vec{p}_i is a 3-vector. 
2. \delta \vec{\phi} \cross \vec{p} is the viariation of \vec{p} when it is rotated along the axis \vec{\phi}.  
(I take this figure from google.) When the vector is rotated from A by \delta\theta to B, you can see easily B-A has the amplitude of |A  \delta \theta| , and the direction is perperdicular to both A and \delta \theta (the rotation axis can be seen as a vector). 


A vector $\\vec {A}$ is rotated by a small angle $\\Delta \\theta$ radians  ($\\Delta \\theta 1$) to get a new vector $\\vec {B}$. In that case $|\\vec{ A}-\\vec {B}|$ isA. $|\\vec{A}| (1-\\dfrac {\\Delta {\\
To go from Eq. 12 to Eq.13 is indeed not trivial. We did not put the details as it has been done in several pervious work and it would take a few pages to do so. 
I recommend this recent review paper which gives more detailed. (section V). 

Best regards, 
HeXu

Jaime Garrido

unread,
Jul 9, 2024, 4:45:29 AMJul 9
to TB2J
Thanks a lot for your answer.

Regarding the recent review you mentioned, I do not see any link or hyperlink on the message. Maybe you forgot to add it? I would love to take a look to the section you say.

Thanks again!

Jaime

Xu He

unread,
Jul 9, 2024, 5:02:19 AMJul 9
to TB2J
Sorry. Here is the link:
https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.95.035004

Best regards, 
HeXu

Reply all
Reply to author
Forward
0 new messages