[Wien] Reproducing the data of PRB, 46, 1321

[Peter Blaha]

I just repeated a few tio2 calculations:

a,a,c: 8.681750  8.681750  5.591610      ATOM  -2: X=0.30497680
Titanium   NPT=  381  R0=.000022390 RMT=   2.00000   Z:  22.00000
O xygen    NPT=  381  R0=.000017910 RMT=   1.60000   Z:   8.00000

LDA, RKmax=8, kmesh just 500k, (6,6,10; 30k in IBZ, not tested, should 
be tested nowadays)

"default setup": Ti =    -2.09   *10**21  V / m**2 eta=0.41
  +Ti-d LO (-0.3) O  =     1.81   *10**21  V / m**2 eta=0.70

"O-p as "LAPW"": Ti =    -2.09   *10**21  V / m**2 eta=0.41
                  O  =     1.70   *10**21  V / m**2 eta=0.64

"O-sp as "APW"": Ti =    -2.09   *10**21  V / m**2 eta=0.41
+ O-p LO at -2Ry O  =     2.11   *10**21  V / m**2 eta=0.83

"O-sp as "LAPW": Ti =    -2.09   *10**21  V / m**2 eta=0.41
+ O-p LO at -2Ry O  =     2.08   *10**21  V / m**2 eta=0.82

PRB-paper from 1992: Ti:  -2.09                    eta=0.43
                      O     2.13                    eta=0.84

As you can see, the most important factor for the O-EFG is the 
additional O-p-local orbital at low energy. At this energy there are 
Ti-3p semicore states, which reach into the O-atomic sphere and must be 
"reexpanded" by an O-wavefunction. The O-p-LO gives the necessary 
variational freedom for these tails.


Singh, David J. schrieb:
> I found that in that case it is important to use
> local orbitals plus the lapw basis for O (not
> APW+lo).
> 
> 
> -----Original Message-----
> From: wien-bounces at zeus.theochem.tuwien.ac.at on behalf of Peter Blaha
> Sent: Fri 10/6/2006 11:37 AM
> To: A Mailing list for WIEN2k users
> Subject: Re: [Wien] Reproducing the data of PRB, 46, 1321
>  
> I don't know what "near" means, but anyway.
> 
> It is essential that you use the same structural parameters as in the 
> paper (u=.305), a and c probably from Ref.8
> Furthermore I'd assume this was a LDA calculation (not GGA).
> 
> Nowadays one can easily use more k-points or a better RKMAX, but in any 
> case you should come "close" to the published values.
> 
> 
> jadhikari at clarku.edu schrieb:
>> Hi,
>> I am trying to exactly reproduce the data according to the article PRB,
>> 46, 1321. The EFG parameters for Ti and O never come out to be near the
>> reported values. Despite trying with different inputs, the output is
>> always way off.
>> May be I am missing some basic factor or that I do not know all the minute
>> details- as I am a novice (1 month old) in this field.
>>
>> I would be very grateful for any suggestions.
>> Thanking you.
>>
>> Sincerely,
>> Subin Adhikari
>> Clark University
>> Worcester, MA 01610
>> USA
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> 

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: 
http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------