The ABINIT Users Mailing List ( CLOSED )

[Anglade Pierre-Matthieu <anglade@gmail.com>]

Hi,

The two approaches should yield (slightly) different results since the
pseudopotential is unchanged if you simply change ixc. This last
approach is not very consistent but may lead to better fortuituous
error compensation than the consistent one.
So, to answer your question, the effect of setting ixc is the same
whatever the pseudopotential original XC. However it will work on a
DIFFERENT system since two pseudopotentials made with different XC are
different. Is it any clearer ?

regards,

PMA


On Mon, Apr 20, 2009 at 9:10 AM, florine bonald
<florine.bonald@gmail.com> wrote:
> Hello,
>
> Thanks for the reply.
>
> As you know, if we insert ixc other than 1, we receive a warning that
> reminds the the difference of XC with the pseudo's. I want to know if this
> different XC (in this case P.&Z. parametrization &...) has its effect on the
> calculation the same as we regenerate the pseudopotential with the said
> parametrization?
>
> Best regards,
> Florin
>
> On Mon, Apr 20, 2009 at 11:53 AM, Anglade Pierre-Matthieu
> <anglade@gmail.com> wrote:
>>
>> Hi,
>>
>> It is not uncommon to use a pseudopotential with a different XC
>> functional than the one it was intended for. It is even not unusual to
>> use LDA pseudopotential with GGA and vice versa. So basically what you
>> have to do is simply to add "ixc 2" into your input file. Obviously if
>> you want to be a little more "self consistent" you would have to
>> create a new pseudopotential using this P.&Z. parametrization for XC
>> functional.
>>
>> regards
>>
>> PMA
>>
>> On Sun, Apr 19, 2009 at 7:20 PM, florine bonald
>> <florine.bonald@gmail.com> wrote:
>> > Hi,
>> >
>> > I have used TM pseudopotential for my calculation with ixc=1 but now I
>> > want
>> > to shift to ixc=2 which is parametrized by Perdew and Zunger. What
>> > changes I
>> > should consider on the pseudo to have this ixc in my input file?
>> >
>> > Regards,
>> > Florin
>> >
>>
>>
>>
>> --
>> Pierre-Matthieu Anglade
>
>



-- 
Pierre-Matthieu Anglade