Re: [Wien] spin configuration for charge state

2016-05-12 Thread Peter Blaha
Did you do a structure optimization (runsp -min ) ??? This is probably mandatory ! > Total energy is 516 meV greater than the previous case. You mean more negative (or the absolute value is "greater") ? We usually say the total energy of a more stable configuration is more negative The

Re: [Wien] spin configuration for charge state

2016-05-12 Thread Laurence Marks
The first BVS number is for the lattice you used, the second is an approximate estimate for an expanded lattice with the typical PBE expansion. You can/should put into a file ".latcalib" (note the "." at the front) the linear expansion/contraction for the DFT optimized volume. The numbers you

Re: [Wien] spin configuration for charge state

2016-05-12 Thread Komal Bapna
Dear Prof. Peter and Laurence, I am really thankful for your valuable suggestions. *Prof. Laurence::* These are the bond-valence sums, I am getting Atom 1 equiv 1 Sr Bond-Valence Sum 2.402.58 Atom 2 equiv 1 Co Bond-Valence Sum 3.033.20 Atom 3 equiv 1 O

Re: [Wien] spin configuration for charge state

2016-05-10 Thread Laurence Marks
To clarify one point, you have to be careful about the meaning of Co"4+". The "4+" is a formal valence, which is quite well reproduced by the bond-valence sums (BVS), e.g. do "x nn ; grep Bond *tnn". It is also what people use when they refer to different atom states as observed by XPS, EELS etc.

Re: [Wien] spin configuration for charge state

2016-05-10 Thread Peter Blaha
a) I don't think that you will ever get a 3d-occupation which is in agreement with the "formal" valence (therfore this is called "formal"). This starts even for a very ionic TiO2, where formally the 3d electrons of Ti4+ should all be empty, but are not in reality. b) Try to stabilize a

Re: [Wien] spin configuration for charge state

2016-05-10 Thread Komal Bapna
Dear Prof. Peter Thank you for your kind reply.. The Density matrix for Co up spin is.. 0.84607 0.0 0.0 0.0 -0.08661 0.0 0.92206 0.0 0.0 0.0 0.0 0.0 0.59673 0.0 0.0 0.0 0.0 0.0 0.92206

Re: [Wien] spin configuration for charge state

2016-05-09 Thread Peter Blaha
You can still try and change case.inst. The ionicity will come during scf. But as I said before, in GGA most likely only ONE state will be reached (independent of your start). You must do GGA+U and you will get a certain state. Save it. Then modify the density matrix (case.dmatup/dn) (a more

Re: [Wien] spin configuration for charge state

2016-05-09 Thread Komal Bapna
Sir, Its true that we can not generate "ionic" electron density with lstart/dstart. When I tried to modify .inst file according to the Co4+ ionic state, it showed error. My query is that how can we generate spin state configuration for such an ionic state, it spin states for Co and Co4+ are

Re: [Wien] spin configuration for charge state

2016-05-07 Thread tran
Hi, I think it is not possible to generate "ionic" electron density with lstart/dstart. Charge transfer is something that will occur during SCF iterations. F. Tran On Friday 2016-05-06 11:48, Komal Bapna wrote: Date: Fri, 6 May 2016 11:48:57 From: Komal Bapna