The ABINIT Users Mailing List ( CLOSED )

["张�s" <zhangting1980323@gmail.com>]

Hi

I think abinit can not perform the calculations that you described. In my opinion, all bands must be calculated at the same time. If you want to calculate states near fermi level in a much detailed scale, you can try the keywords 'ndivk' and ' kptbounds', which can locate the energy band calculaiton in the k-spcae range that you are interested in. That is , you can just calculate the k-points that near fermi k-point, and about just several eVs around fermi level.  Even in this kind of tasks, for each k-point, all bands are taken into account. Then you can select the states you are interested in.

Regards

                                         Zhang Ting
                                         Peking Univ.
                                         Oct. 20th 2007

2007/10/19, shinekin@hotmail.com <shinekin@hotmail.com>:

hi Lan,
thanks for your reply, but i just want to know how to calculate the bands within 1 or 2 eV around the fermi level, i mean that other band could be neglect.

certainly 'nband' can specify the numbers of bands i want to calculate,but i found that abinit generates eigenvalues in a order from the low to high enery according to the corresponding valence states. for instance, if the valence states that abinit considers is 2d3s3p and the 'naband' was set to be 1,then just 1 band of 2d state will be calculated because of the 2d state locates at the lowest energy, in my question, how to set parameters so that abinit can just calculates the band which locates at the higher energy levels in a single band structure calculation?

       Regards
Shinekin