ERROR: should be no new maxima in edge refinement

Bader charge density analysis

Moderator: moderators

Post Reply
Hanmei_Tang
Posts: 7
Joined: Wed Jan 27, 2016 6:26 pm

ERROR: should be no new maxima in edge refinement

Post by Hanmei_Tang »

Hi all,

Now I am doing Bader analysis on a 100-atom crystal unit. I think I ran into some kind of trouble.
1. run "bader CHGCAR -ref CHGCAR_sum" I got "ERROR: should be no new maxima in edge refinement"
2. from OUTCAR I know my unit cell should have 660 electrons.
Bader calculation gave me "NUMBER OF ELECTRONS: 634.73046"
If I continue to use a higher NGX/NGY/NGZ, the results are:
=========================================================
NGX = 100
NGY = 100
NGZ = 240
-----------------------------------------------------------------------------------------------------
NUMBER OF BADER MAXIMA FOUND: 90
SIGNIFICANT MAXIMA FOUND: 90
VACUUM CHARGE: -0.1817
NUMBER OF ELECTRONS: 635.76616
==========================================================
NGX =165
NGY =165
NGZ =395
-----------------------------------------------------------------------------------------------------
NUMBER OF BADER MAXIMA FOUND: 90
SIGNIFICANT MAXIMA FOUND: 90
VACUUM CHARGE: -0.1817
NUMBER OF ELECTRONS: 636.57041
==========================================================
graeme
Site Admin
Posts: 2256
Joined: Tue Apr 26, 2005 4:25 am
Contact:

Re: ERROR: should be no new maxima in edge refinement

Post by graeme »

It seems like something is not right. If you post the files I'll take a look at the analysis.
Hanmei_Tang
Posts: 7
Joined: Wed Jan 27, 2016 6:26 pm

Re: ERROR: should be no new maxima in edge refinement

Post by Hanmei_Tang »

Dear Dr. Graeme,

Just would like to let you know that I have solved my problem by running a previous version bader (Version 0.28a 07/12/12).

I can get output 'NUMBER OF ELECTRONS: 660.00000', which corresponds OUTCAR. Also ACF.dat is slightly different from previous result.
graeme
Site Admin
Posts: 2256
Joined: Tue Apr 26, 2005 4:25 am
Contact:

Re: ERROR: should be no new maxima in edge refinement

Post by graeme »

Thanks for the note; there was a bug in the new version which did not properly deal with charge density files containing negative values of the charge density. I'll update things right now.
Post Reply