LDA+U problem
Posted: Tue Jan 11, 2011 2:10 pm
Dear all,
I used the Bader code to analyze results obtained on vanadium oxides with VASP, with PAW pseudos and LSDA calculations.
Using the correct keyword, I was able to analyze the charge density file obtained from the addition of AECARR0 and AECCAR2 and the results looked fine.
Then, I did the same thing but using a LSDA+U calculation and I got a strange thing: the total number of electrons is larger than the formal one !
In my system, I have 22 oxygen, 8 vanadium and 4 copper atoms which makes a total of 476 electrons. With the LSDA calculation, this is the number given in the ACF.dat file. However with the +U calculation, I obtain a total of 478 electrons in the ACF.dat file !!!
Does anyone know what is the reason for this ?
Thank you.
Magali
I used the Bader code to analyze results obtained on vanadium oxides with VASP, with PAW pseudos and LSDA calculations.
Using the correct keyword, I was able to analyze the charge density file obtained from the addition of AECARR0 and AECCAR2 and the results looked fine.
Then, I did the same thing but using a LSDA+U calculation and I got a strange thing: the total number of electrons is larger than the formal one !
In my system, I have 22 oxygen, 8 vanadium and 4 copper atoms which makes a total of 476 electrons. With the LSDA calculation, this is the number given in the ACF.dat file. However with the +U calculation, I obtain a total of 478 electrons in the ACF.dat file !!!
Does anyone know what is the reason for this ?
Thank you.
Magali