[gpaw-users] 2D Metal Sheets

[Jens Jørgen Mortensen]

Den 26-02-2018 kl. 19:59 skrev Tomlinson, Warren (CDR) via gpaw-users:
> All-
> I have run into a curious result with my 2D magnetic sheet 
> calculations and I was thinking you might be able to help. I have done 
> calculations on the following 2D sheets:

Are your numbers converged with respect to number of k-points and 
Fermi-Dirac width?  Maybe try with more k-points and smaller width?

Jens Jørgen

> NiCl2
> NiBr2
> CoCl2
> FeCl2
> MnCl2
> MnBr2
>
> Step one was to find the correct geometry and cell size. I did this 
> with the following code:
>
> for a in [6.87, 6.88, 6.89, 6.90, 6.91]:
>
>    sheet = mx2(formula='FeCl2', kind='1T', a=a/2, thickness=2.5, 
> size=(2,2,1), vacuum=17)
>    sheet.set_pbc([1,1,1])
>    name = '%.2f' % a
>
>    calc = GPAW(mode=PW(600),
>            kpts=(10,10,1),
>            xc='PBE',
>            parallel={'kpt': 30, 'band': 1},
>            basis = 'dzp',
>            convergence={'energy': 0.0001},
>            occupations=FermiDirac(0.1),
>            setups={'Fe': ':d,3.0'},
>            txt=name + '.txt',
>            maxiter=1000
>            )
>
>    sheet.set_calculator(calc)
>    opt = BFGS(sheet)
>    opt.run(fmax=0.01)
>    E = sheet.get_potential_energy()
>    parprint('**********')
>    parprint(name + ':\t %f' % E)
>    parprint('**********')
>    write(name + '.traj', sheet)
>
> —————
>
> Next I did spin polarized calculations to determined what (if any) 
> magnetic state was most preferred. Here is the code for that:
>
> a = read('FeCl2.traj')
> mm = [[1.5, 0.1, 0.1, 1.5, 0.1, 0.1, 1.5, 0.1, 0.1, 1.5, 0.1, 0.1],
>      [-1.5, -0.1, -0.1, 1.5, 0.1, 0.1, 1.5, 0.1, 0.1, -1.5, -0.1, -0.1],
>      [0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.]]
> type = ['FM', 'AFM', 'None']
> name = 'FeCl2'
> E = 0
>
> for t in type:
>    if t == 'FM': a.set_initial_magnetic_moments(mm[0])
>    if t == 'AFM': a.set_initial_magnetic_moments(mm[1])
>    if t == 'None': a.set_initial_magnetic_moments(mm[2])
>
>    calc = GPAW(mode=PW(600),
>            kpts=(10,10,1),
>            xc='PBE',
>            spinpol=True,
>            parallel={'kpt': 30, 'band': 1},
>            basis = 'dzp',
>            convergence={'energy': 0.0001},
>            occupations=FermiDirac(0.1),
>            setups={'Fe': ':d,3.0'},
>            txt=t + '_' + t + '.out',
>            maxiter=1000
>            )
>
>    a.calc = calc
>    energy = a.get_potential_energy()
>    parprint('$$$$$$')
>    parprint('E for ' + t + ' mode: %f' % energy)
>    parprint('$$$$$$')
>
> My results show that FM is preferred for all but FeCl2, MnCl2 and 
> MnBr2. This matches the experimental results from another group I am 
> working with, except for FeCl2. While they show the Mn structures 
> preferring an AFM state, they show FeCl2 preferring an FM state. The 
> results I got for FM vs AFM for FeCl2 are not even close. AFM is about 
> 1.6 eV lower in energy than FM. This is a considerably larger 
> difference then what I found for the other structures. The other 
> structures had FM-AFM energy differences of around 50-100 meV, though 
> the CoCl2 difference was relatively high at ~700 meV. Can you think of 
> any reason why my results for FeCl2 are so far off? Should I be 
> treating Fe differently for some reason? Am I missing an important 
> element of my spinpol calculations that I’m just not seeing?
> Thanks,
> Warren
>
>
> CDR Warren Tomlinson, Ph.D.
> Department of Physics
> Naval Postgraduate School
> Sp. 118
> (831) 656-2877
> wwtomlin at nps.edu <mailto:wwtomlin at nps.edu>
>
>
>
> CDR Warren Tomlinson, Ph.D.
> Department of Physics
> Naval Postgraduate School
> Sp. 118
> (831) 656-2877
> wwtomlin at nps.edu <mailto:wwtomlin at nps.edu>
>
>
>
>
>
>
> _______________________________________________
> gpaw-users mailing list
> gpaw-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/gpaw-users/attachments/20180226/1f5f33ca/attachment-0001.html>