Thank you Xavier and Guillaume. One more small question. If I have ferromagnetic system (experimental data), calculated total energy in spinpolarized ferromagnetic system (let's call it E(SP)) then in order to get E(FM) from your formula, I have to calculate total energy of a non-spinpolarized system (E(NSP)), then I can calculate E(FM) like E(FM) = E(SP)-E(NSP). Am I right?
Thank you! Date: Wed, 18 May 2011 14:11:18 +0200 From: xavier.rocquefe...@cnrs-imn.fr To: wien at zeus.theochem.tuwien.ac.at Subject: Re: [Wien] exchange coupling constants calculations Dear Vincent, The rule is quite simple. Let's take the example of a monoatomic linear chain. To obtain the J parameter coupling two adjacent magnetic centers, you simply have to do 2 calculations: - one with the ferromagnetic order: FM - one with the antiferromagnetic order: AFM You will then obtain two total energies, i.e. E(FM) and E(AFM). Considering an Heisenberg Hamiltonian, you have: E(FM) = N?/4 * (-J) and E(AFM) = N?/4 * (+J) Then E(AFM) - E(FM) = N?/4 * (2*J) = N?/2 * J where N is the number of unpaired electrons on the magnetic center. For example, for N = 1: J = 2 * (E(AFM) - E(FM)) The present example is very simple and you could imagine that for more complex structures it will be more difficult to realize a mapping analysis. Then the idea is: - to use more magnetic orders than unknown J parameters - to do a least-square procedure to find the J values ... Here is and example of calculations I did with Peter and Karlheinz on CuO: http://iopscience.iop.org/0953-8984/22/4/045502 If you do not have access to the paper, I could send it to you. Regards Xavier Le 18/05/2011 12:31, Linux PC a ?crit : Dear Xavier, Thank you very much for the answer! I actually was looking through the manual before, unfortunately, did not succeed in finding any information on exchange interactions. Would you know, or may be other users, if any practical manual exists for calculating of J. Thank you! Vincent > Date: Wed, 18 May 2011 07:24:52 +0200 > From: Xavier.Rocquefelte at cnrs-imn.fr > To: wien at zeus.theochem.tuwien.ac.at > Subject: Re: [Wien] exchange coupling constants calculations > > Dear Vincent, > > Yes for sure. To estimate magnetic exchange coupling parameters you > need to do a mapping based on total energies of different magnetic > orders. WIEN2k allows to have total energies with high accuracy and > using different functionals (depending of the system you are > considering). So yes, using WIEN2k to estimate J values is an > excellent choice. > > Regards > > Xavier > > > > "Linux PC" <linuxfarm3m at hotmail.fr> a ?crit : > > > > > > > Dear Dr. Blaha and Wien2k users, > > > > Is it possible to use the Wien2k to calculate j magnetic exchange > > coupling constants? > > > > Thank you very much, > > Vincent > > > > > _______________________________________________ > Wien mailing list > Wien at zeus.theochem.tuwien.ac.at > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien _______________________________________________ Wien mailing list Wien at zeus.theochem.tuwien.ac.at http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien _______________________________________________ Wien mailing list Wien at zeus.theochem.tuwien.ac.at http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20110518/ad93d0a2/attachment.htm>