[DFTB-Plus-User] charges in MD runs

[Hu Shuanglin]

Dear all,

When I did spin-polarized plusU MD runs with DFTB+, I found the charges
printed in detailed.out and the number of electrons in output .xyz file did
not agree to each other.

For example, I have a TiO2 surface with one oxygen vacancy, and I chose the
first two Ti as Ti3+ ions. The two Ti3+ ions distinguished well among other
Ti4+ ions well when I did geometry relaxation and static energy
calculations, which can be seen from the correspoding detailed.out and .xyz
file:
$ grep -A 10 'Net atomic charges' detailed.out
 Net atomic charges (e)
  Atom       Net charge
     1       0.68142312
     2       0.68089722
     3       1.07348552
     4       1.06186570
     5       1.04718640
...

$ head geom.out.xyz
  196
Geometry Step: 1432
   Ti      4.54958356     13.22965494     14.57128888      3.31857688
   Ti      4.53579697      6.61120482     14.58567434      3.31910278
   Ti     12.12797386      0.00488244     17.73313427      2.92651448
   Ti     12.12729010      6.62579500     17.69843422      2.93813430
   Ti      3.03171520     13.22590270     17.75939140      2.95281360
...

But in the MD runs, it is:
$ grep -A 10 'Net atomic charges' detailed.out
 Net atomic charges (e)
  Atom       Net charge
     1       0.67984703
     2       0.66752988
     3       1.07816246
     4       1.06265469
     5       1.02547626

$ grep -A 5 'MD iter: 1048' geom.md1.xyz
MD iter: 1048
   Ti      4.56682261     13.10980045     14.62200850      2.27988180
4.64045954      3.35909318     -0.39461966
   Ti      4.49531732      6.58520316     14.60760767      4.36272611
-1.98856435      0.10046890      1.90791931
   Ti     12.13250734     -0.03765839     17.76186059      2.92051271
-3.39619380     -1.00271714     -0.09378173
   Ti     12.14473925      6.59728454     17.74358229      2.93774829
-1.93502820     -1.83963283     -1.66329078
   Ti      3.01352703     13.21447665     17.74017744      2.96803645
1.46811080     -1.73095809      2.41670206

while the Atomic populations in detailed.out seems correct
 Atom populations (up)                          Atom populations (down)
  Atom       Population                             Atom       Population
     1       1.13877329                                  1       2.18056149
     2       2.18051732                                  2       1.15105036
     3       1.46156280                                  3       1.46203088
     4       1.46927778                                  4       1.46858315
     5       1.48308108                                  5       1.48980877
...                                                      ...

Is this due to any bugs for the number of elections printed  in the .xyz
file when we do MD runs? Or is there anything else happened in the MD?

Thank you!

Best Regards,
Shuanglin Hu
--
Department of Chemistry - the Angstrom Laboratory,
Uppsala University, Uppsala, Sweden
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20130122/d4522abd/attachment.htm>