The ABINIT Users Mailing List ( CLOSED )

[Liam Damewood <ldamewood@gmail.com>]

Thank you, smaller tsmear got me closer to a 3.0 magnetization. This
makes sense. I also lowered toldfe to the value you recommended.

The issue must be with the pseudopotential I'm using. I also tried the
HGH pp using basically the same setup as my previous post and got a
smaller acell of 10.18. Are there more databases with GGA
pseudopotentials available for use in abinit?

Thanks.
-Liam
==========================================
acell     10.70 10.70 10.70
rprim     0.0  0.5  0.5
          0.5  0.0  0.5
          0.5  0.5  0.0
ntypat    2
znucl     24 33
natom     2
typat     1 2
xred      0.00 0.00 0.00
          0.25 0.25 0.25
nsppol    2
spinat    0.00 0.00 6.00
          0.00 0.00 -3.00
nband     16

optcell   1
ionmov    3
ntime     10
dilatmx   1.07
ecutsm    0.5

ecut      30.0

kptopt    1
ngkpt     27 27 27
nshiftk   4
shiftk    0.5 0.5 0.5
          0.5 0.0 0.0
          0.0 0.5 0.0
          0.0 0.0 0.5

occopt    4
nstep     20
toldfe    1.0d-10
tsmear    0.004
#prtdos    2



On Sat, Jan 17, 2009 at 1:47 PM, matthieu verstraete
<matthieu.jean.verstraete@gmail.com> wrote:
> Hello,
>
> for the magnetization in particular, your tsmear is quite high. This will
> occupy a few of the minority spin bands and lower the total M (a bit).
>
> The GGA overestimation of acell is normal, and possibly worsened by the
> tsmear... For these systems even more than others the pseudopotential is
> crucial. I am not surprised the GGA is worse - it can be more delicate to
> make good potentials. Try another one! Or just use LDA, it's miraculously
> good as a rule. Did you use the potentials from the full TM tables?
>
> Your toldfe is also a bit too high. 1e-10 is decent, depending on what
> properties you need.
>
> Matthieu
>
> On Sat, Jan 17, 2009 at 7:31 PM, <ldamewood@gmail.com> wrote:
>>
>> I've been calculating the gap and magnetization in ZB CrAs using abinit.
>> As a
>> reference, I've been comparing my calculations to Phys. Rev. B 67, 224420
>> (2003). When I preform an LDA calculation, I get good agreement with the
>> lattice parameter and magnetization. For the GGA calculation, I get a much
>> larger lattice parameter and a non-integer magnetization (an integer
>> magnetization would be expected for half metallic CrAs). Specifically,
>> acell
>> should be 10.69 whereas GGA gives me 11.2. magnetization should be exactly
>> 3
>> while GGA gives 2.95.
>>
>> I've included an input below. Can anyone tell me if I am doing something
>> incorrectly? I'm not doing anything vastly different in my GGA calculation
>> than
>> in my LDA calculation so my guess is that the pseudopotential is not well
>> suited for this calculation. I'm new at this so any help is appreciated.
>>
>> -Liam
>>
>> ----------------------------------------------------------
>> For LDA, I used the FHI pseudopotentials and for GGA, I used PBE from the
>> FHI
>> code. Here is my input file for a GGA calculation for the acell
>> convergence.
>>
>> [damewood@**** CrAs_1]$ cat CrAs_1.in
>> ndtset    15
>> acell:    10.0 10.0 10.0
>> acell+    00.1 00.1 00.1
>> rprim     0.0  0.5  0.5
>>          0.5  0.0  0.5
>>          0.5  0.5  0.0
>> ntypat    2
>> znucl     24 33
>> natom     2
>> typat     1 2
>> xred      0.00 0.00 0.00
>>          0.25 0.25 0.25
>> nsppol    2
>> spinat    0.00 0.00 6.00
>>          0.00 0.00 -3.00
>> nband     12
>>
>> ecut      30.0
>>
>> kptopt    1
>> ngkpt     27 27 27
>> nshiftk   4
>> shiftk    0.5 0.5 0.5
>>          0.5 0.0 0.0
>>          0.0 0.5 0.0
>>          0.0 0.0 0.5
>>
>> occopt    4
>> nstep     20
>> toldfe    1.0d-6
>> tsmear    0.01
>> prtdos    2
>
>
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Dr. Matthieu Verstraete
>
> European Theoretical Spectroscopy Facility (ETSF)
> Dpto. Fisica de Materiales,
> U. del Pais Vasco,
> Centro Joxe Mari Korta, Av. de Tolosa, 72,   Phone: +34-943018393
> E-20018 Donostia-San Sebastian, Spain        Fax  : +34-943018390
>
> Mail : matthieu.jean.verstraete@gmail.com
> http://www-users.york.ac.uk/~mjv500
>