[Pw_forum] problem with Xspectra

[matteo calandra]

Dear Min Wu,

    usually the mistake you did (using the 1s wavefunction with the core-hole
instead of that without the core-hole) leads to small differences in the
case of K-edge, as the two wavefunction are probably very similar.
This is not true for L1 edges (that you can calculate with the
current program) or generally speaking, less deep edges (L,M and so on).
However, even in the case of K-edges it can be important.
So I suggest you repeat the last step of your calculation
(the XSPECTRA one) using the core wfc without a core-hole
as this is the correct thing to do.

All the best, M.

>
> Message: 1
> Date: Wed, 06 Oct 2010 02:53:07 +0800
> From: "wumindt2" <wumindt2 at zju.edu.cn>
> Subject: Re: [Pw_forum] problem with Xspectra
> To: pw_forum at pwscf.org
> Message-ID: <LTRHDZGOVULYBIDYMMCUNFWGFNIV.wumindt2 at zju.edu.cn>
> Content-Type: text/plain;	charset="gb2312"
>
> Dear matteo,
>
> Thank you very much!
> I've got the picture now.
>
> BTW, in my previous calculation, i mistakenly extract the core wavefunction
> from the pseudopotential WITH a core hole. But the result seems good  
> when compared
> with the experiment. Does it happen often?
>
> Cheers,
>
> Min Wu
> 2010-10-5
>
>
>
>> From: matteo calandra <matteo.calandra at impmc.jussieu.fr>
>> Reply-To: PWSCF Forum <pw_forum at pwscf.org>
>> To: pw_forum at pwscf.org
>> Subject: Re:[Pw_forum] problem with Xspectra
>> Date: Tue,  5 Oct 2010 10:49:48 +0200
>>
>> Dear Min Wu,
>>
>>  the pseudopotentials to be used for the absorbing atom
>> in Xspectra are those having
>> the label _gipaw, two examples are:
>>
>> Fe.pbe-sp-mt_gipaw.UPF
>> Fe.star1s-pbe-sp-mt_gipaw.UPF
>>
>> The first one has no core-hole and has full gipaw informations
>> while the second one has core-hole and full gipaw informations.
>> Gipaw informations mean many things. What you need to know is that
>> GIPAW informations include the presence of the
>> all-electron wavefunction for the 1s core state.
>>
>> Now when you run a simulation you first need to calculate the charge
>> density in the presence of a core-hole and thus use pseudo
>> Fe.star1s-pbe-sp-mt_gipaw.UPF . This is done by a standard pw.x
>> scf calculation using this pseudo.
>> In a second step you need to calcuate the Xanes cross
>> section and this is done by the Xspectra code.
>>
>> In the calculation of the  matrix element appearing in the
>> Xanes cross-section (Fermi golden rule) the initial 1s state
>> IN THE ABSENCE of a core hole is needed. This information is contanined
>> in the pseudo potential WITHOUT the core hole (Fe.pbe-sp-mt_gipaw.UPF).
>> Thus you simply extract this wavefunction using the script given with the
>> Xspectra disribution and you put the name of the file produced by the script
>> in the input file of the Xspectra program.
>>
>> You cannot use the 1s core function with the core hole simply  
>> because this is
>> wrong as the matrix element inthe cross section involves the product
>> between the initial and final states and the initial state has no core hole.
>>
>> M.
>>> Message: 1
>>> Date: Tue, 05 Oct 2010 00:42:50 +0800
>>> From: "wumindt2"
>
>>> Subject: [Pw_forum] problem with Xspectra
>>> To: pw_forum at pwscf.org
>>> Message-ID:
>
>>> Content-Type: text/plain; charset="gb2312"
>>>
>>> Hi,
>>>
>>> I'm a new user of Xspectra.
>>>
>>> In the Xspectra example given in the pwscf code, it calculates the
>>> XAS using the pseudopotential
>>> without core hole level. In this case, we need to extract the core
>>> wavefunction from the GIPAW
>>> pseudopotential when we run the xspectra.x.
>>>
>>> Now there are pseudopotentials with core hole, normally named as
>>> *.star1s*.UPF. So here is
>>> my question, after the SCF calculation, do we still need to extract
>>> the core wavefunction from the
>>> peusopotential of the absorbing atom?
>>>
>>> If the answer is yes, then why we choose the core wavefunction from
>>> the pseudopotential as the ground
>>> state, rather than the core wavefunction after the SCF calculation?
>>>
>>> Thanks!
>>>
>>> Min Wu
>>> 2010-10-4
>>
>>
>> ----------------------------------------------------------------
>> This message was sent using IMP, the Internet Messaging Program.
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
>


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.