[DFTB-Plus-User] eigvev.out & oversqr.dat

[邓凤麟]

Dear DFTB developers and users,




I am a new user of DFTB+. Now, I cannot understand the relation between overlap matrix S in oversqr.dat and cofficients C in eigvev.out when I do a calculation of Silicon crystal.




In DFTB, the formula (C^T)_ij S_jk C_kl=delta_il shoud always be satisfied. I do a calculation of H2O molecule and obtain the right result delta_il, like the previous discussion (https://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/2017/002504.html).




However, for the Si crystal, I am puzzled when I calculate  (C^T)_ij S_jk C_kl at random k point. 

For the first k point and first eigen vector,

S=


     1.00855+3.40401e-20im  …     0.126487-0.0489968im
 6.09864e-20-0.0437937im          0.115195-0.0619647im
 1.73134e-18-0.0437937im          0.115195-0.0619647im
  1.7415e-18-0.0437937im         0.0258335+0.120866im
   -0.068126+0.357226im        4.68655e-17+0.0437937im
    0.126487+0.0489968im    …    0.0338919-5.55196e-17im
    0.126487+0.0489968im         0.0338919-5.55586e-17im
    0.126487+0.0489968im          0.751609-4.18909e-17im




C=





  0.585064 - 0.000778im
 -0.006024 + 0.041804im
 -0.006024 + 0.041804im
 -0.006024 + 0.041804im
 -0.111193 - 0.574401im
  0.039879 - 0.013912im
  0.039879 - 0.013912im
  0.039879 - 0.013912im




C^T=
0.585064+0.000778im  -0.006024-0.041804im  …  0.039879+0.013912im




I hope I use the data from output files in a right way. The input file can be found in attachments.




Best regards,


Fenglin Deng
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20210513/015ea7da/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: si.zip
Type: application/zip
Size: 76941 bytes
Desc: not available
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20210513/015ea7da/attachment.zip>