The definition of exciton spin in ypp

[xueshanxihua]

Dear developers,

I'm calculating spin-flip excitons with yambo by including spinor wavefunctions in DFT step (lspinorb=.true. in qe). The material has a Mott-insulator-like band structure at the band edge, so I simply used one spin-up valance band and one spin-down conduction band to construct the BSE Hamiltonian in the calculation, and I thought it should produce merely spin-triplet excitons. Then I try the function "-e spin" in ypp. It gives two files "exc_E+spin" and "exc_I+spin" which contain exciton eigenvalues and are quite similar to those created by "-e s", but with two additional lines in each file dubbed as "S_z" and "S_square". However, I found that all the values in the two lines are "0.000", although there are only spin-triplet S=1 excitons herein.

So I think the "spin" here doesn't refer to the total moment of exciton, but something like an averaged spin value among the associated electrons and holes. Is that the case? Would you please tell me the exact definition of "exciton spin" given in the file?


Thanks a lot,

[Davide Sangalli]

Dear Zeyu Jiang,
unfortunately the implementation of the spin on the exciton state is broken.
We are currently working to fix it. This is why you find just zeros in the output.

I'll let you know as soon as there is a working implementation you can use.

Best,
D.

[pangrt]

Dear developers,
I analyzed the absorption spectrum with ypp -e sp, and obtained as:
# E [ev] Strength Index S_z S_square
#
1.3376 0.0055 1.0000 0.9399 0.0000
1.3376 0.0049 2.0000 -0.9399 0.0000
1.34 0.897E-09 3.00 0.459E-05 0.00
1.339 0.2301E-7 4.000 -.7379E-5 0.000
1.345 0.2845E-8 5.000 -.4364E-4 0.000
1.346 0.6218E-5 6.000 -.9264 0.000
1.346 0.9481E-5 7.000 0.9272 0.000
The S_z is the projection of total spin along z direction? And why it is not equal in some pairs like index 3 and 4?
Thanks!
Rongtian