[Wien] optimization for 225 (cubic) space group

[tran at theochem.tuwien.ac.at]

Hi,
If the SCF calculations were done with a value for a parameter (e.g.,
RKmax, number of k-points or energy convergence criteria) that was
not good enough, then this may introduce some noise such that the
Murnaghan EOS can not fit perfectly the data and, therefore,
leads to an optimised lattice parameter that depends on the chosen
volumes to do the fit.

In your case the fit does not seem perfect, but still reasonable
since the variations in the lattice parameter are only at the
third digit (at most tiny changes in the electronic properties).

FT

On Thursday 2017-05-04 20:34, fatima DFT wrote:

>Date: Thu, 4 May 2017 20:34:34
>From: fatima DFT <fatimadft5 at gmail.com>
>Reply-To: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
>To: A Mailing list for WIEN2k users <Wien at zeus.theochem.tuwien.ac.at>
>Subject: [Wien] optimization for 225 (cubic) space group
>
>Dear All,I finished an optimisation for 225 space group with -15% to 25% volume change.
>
>If I plot the optimization cure for different vol parameters, I see the shape of ENE vs. Vol curve changes and the optimised lattice parametes also changes (at
>third digit).
>
>If I plot the curve for -15 to 25% then I get the below curve
>https://sites.google.com/site/dftfatima5/letter/mailing-list-queries/1.ps
>
>
>and the optimised lattice parameters are
>
>21.1924 bohr =    11.2145 Ang
>
>For -10 to 25%, 
>https://sites.google.com/site/dftfatima5/letter/mailing-list-queries/2.ps
>
>21.1984 bohr =    11.2177 Ang
>
>For -5 to 25%
>https://sites.google.com/site/dftfatima5/letter/mailing-list-queries/3.ps
>
>21.1915 bohr =    11.2140 Ang
>
>0.0 to 25 %
>https://sites.google.com/site/dftfatima5/letter/mailing-list-queries/4.ps
>
>21.1917 bohr =    11.2141 Ang
> for 5.0  to 25 %
>https://sites.google.com/site/dftfatima5/letter/mailing-list-queries/5.ps
>21.2026 bohr =    11.2199 Ang
>
>
>My query is:
>
>1. Why the optimised lattice parameters are changing if we select different set of data points? The ground state should be same in all cases. Is it?
>2.  We see from the lattice parameter the change is at third order, can we take any lattice parameters (a0)? How much the third digit changes the electronic
>properties?
>
>
>Warm regards
>Fatima
>
>
>