Vanished band gap after core levels manipulations

Forum » Forum / Support Questions » Vanished band gap after core levels manipulations

Started by:

Ahmad Abdolmaleki
Date: 21 Feb 2016 01:02
Number of posts: 6

RSS: New posts

Unfold All Fold All More Options

Fold

Vanished band gap after core levels manipulations

Ahmad Abdolmaleki 21 Feb 2016 01:02

Hi exciting developers and users

I want to calculate XANES spectra of phosphorene nanosheet. Based on the website tutorial, at first I should run a simple scf and find the energy of interested core level, and change species file to add local orbital.
I run a scf and the band gap is about 0.03 hartree that is reasonable, but when I remove 1S from cores and add local orbital, the band gap vanishes!
Used parameters are as below:

   <groundstate 
      stype="Fermi Dirac" 
      swidth="0.001d0" 
      dipolecorrection="true" 
      rgkmax="8"  
      ngridk="10 10 1"   
      gmaxvr="14.0"  
      xctype="GGA_PBE"  
      nempty="15"/>

and the changed species file of phosphorus:

 <?xml version="1.0" encoding="UTF-8"?>  
<spdb xsi:noNamespaceSchemaLocation="../../xml/species.xsd" xmlns:xsi="Removed/2001/XMLSchema-instance">  
  <sp chemicalSymbol="P" name="phosphorus" z="-15.0000" mass="56461.71408">  
    <muffinTin rmin="0.100000E-04" radius="2.0000" rinf="22.3534" radialmeshPoints="300"/> 
    <atomicState n="1" l="0" kappa="1" occ="2.00000" core="false"/>  
    <atomicState n="2" l="0" kappa="1" occ="2.00000" core="true"/>  
    <atomicState n="2" l="1" kappa="1" occ="2.00000" core="true"/>  
    <atomicState n="2" l="1" kappa="2" occ="4.00000" core="true"/>  
    <atomicState n="3" l="0" kappa="1" occ="2.00000" core="false"/>  
    <atomicState n="3" l="1" kappa="1" occ="1.00000" core="false"/>  
    <atomicState n="3" l="1" kappa="2" occ="2.00000" core="false"/>  
    <basis>  
      <default type="lapw" trialEnergy="0.1500" searchE="false"/>  
      <custom l="0" type="apw+lo" trialEnergy="0.1500" searchE="true"/>  
      <custom l="1" type="apw+lo" trialEnergy="0.1500" searchE="true"/>  
      <!-- Here below a local orbital is added for the state 1s -->  
      <lo l="0">  
         <wf matchingOrder="0" trialEnergy="0.1500" searchE="false"/> 
         <wf matchingOrder="1" trialEnergy="0.1500" searchE="false"/> 
         <wf matchingOrder="0" trialEnergy="-76.50" searchE="true"/>  
      </lo>  
    </basis> 
    <!-- Exciting code version: Lithium -->  
    <!-- Generation Scheme: simple -->  
  </sp>  
</spdb>

Can anyone help me to solve the gap problem? I changed K points, gmaxvar,epsband, deband, rgkmax, stype and swidth; but there was no improvement.
(While this problem does not exist with using ELK code.)

Best wishes

Reply Options

Unfold Vanished band gap after core levels manipulations by

Ahmad Abdolmaleki, 21 Feb 2016 01:02

Fold

Re: Vanished band gap after core levels manipulations

Caterina Cocchi 22 Feb 2016 09:27

Dear Ahmad,

could you please post the entire input file that you are using?

Best regards,
Caterina Cocchi
(exciting team)

Reply Options

Unfold Re: Vanished band gap after core levels manipulations by

Caterina Cocchi, 22 Feb 2016 09:27

Fold

Re: Vanished band gap after core levels manipulations

Ahmad Abdolmaleki 22 Feb 2016 10:39

Dear Caterina
Thank you for your quick response.

This is my input file:

<input>
   <title>phosphorene</title>
   <structure autormt="false" speciespath="./">
      <crystal scale="1">
         <basevect>6.277084      0.000000     0.000000</basevect>
         <basevect>0.000000      8.606720     0.000000</basevect>
         <basevect>0.000000      0.000000     32.125344</basevect>
      </crystal>
      <species speciesfile="P_xanes.xml" >
         <atom coord="0.99912200   0.41325505   0.26718847" />
         <atom coord="0.99913000   0.58674519   0.14238404" />
         <atom coord="0.49911300   0.08674939   0.26719053" />
         <atom coord="0.49912100   0.91325037   0.14238198" />
      </species>
   </structure>

   <groundstate
      stype="Fermi Dirac"
      swidth="0.001d0"
      epsband="1.0d-10"
      deband="0.003d0"
      rgkmax="8"
      ngridk="21 21 1"
      gmaxvr="14.0"
      xctype="GGA_PBE"
      nempty="15"/>
</input>

and P_xanes.xml :

 <?xml version="1.0" encoding="UTF-8"?>  
<spdb xsi:noNamespaceSchemaLocation="../../xml/species.xsd" xmlns:xsi="Removed/2001/XMLSchema-instance">  
  <sp chemicalSymbol="P" name="phosphorus" z="-15.0000" mass="56461.71408">  
    <muffinTin rmin="0.100000E-04" radius="2.0000" rinf="22.3534" radialmeshPoints="300"/> 
    <atomicState n="1" l="0" kappa="1" occ="2.00000" core="false"/>  
    <atomicState n="2" l="0" kappa="1" occ="2.00000" core="true"/>  
    <atomicState n="2" l="1" kappa="1" occ="2.00000" core="true"/>  
    <atomicState n="2" l="1" kappa="2" occ="4.00000" core="true"/>  
    <atomicState n="3" l="0" kappa="1" occ="2.00000" core="false"/>  
    <atomicState n="3" l="1" kappa="1" occ="1.00000" core="false"/>  
    <atomicState n="3" l="1" kappa="2" occ="2.00000" core="false"/>  
    <basis>  
      <default type="lapw" trialEnergy="0.1500" searchE="false"/>  
      <custom l="0" type="apw+lo" trialEnergy="0.1500" searchE="true"/>  
      <custom l="1" type="apw+lo" trialEnergy="0.1500" searchE="true"/>  
      <!-- Here below a local orbital is added for the state 1s -->  
      <lo l="0">  
         <wf matchingOrder="0" trialEnergy="0.1500" searchE="false"/> 
         <wf matchingOrder="1" trialEnergy="0.1500" searchE="false"/> 
         <wf matchingOrder="0" trialEnergy="-76.50" searchE="true"/>  
      </lo>  
    </basis> 
    <!-- Exciting code version: Lithium -->  
    <!-- Generation Scheme: simple -->  
  </sp>  
</spdb>

Also I ran a ground state calculation with below input file but, again, no gap was obtained:

<input>
   <title>phosphorene</title>
   <structure autormt="true" speciespath="./">
      <crystal scale="1">
         <basevect>6.277084 0.0 0.0</basevect>
         <basevect>0.0 8.606720 0.0</basevect>
         <basevect>0.0 0.0 32.125344</basevect>
      </crystal>
      <species speciesfile="P_xanes.xml" >
         <atom coord="0.49911300   0.08674939   0.26719053" />
      </species>
      <species speciesfile="P.xml" >
         <atom coord="0.99912200   0.41325505   0.26718847" />
         <atom coord="0.99913000   0.58674519   0.14238404" />
         <atom coord="0.49912100   0.91325037   0.14238198" />
      </species>
   </structure>

   <groundstate 
      ngridk="4 4 1"
      gmaxvr="14.0" />
</input>

Best wishes

Reply Options

Unfold Re: Vanished band gap after core levels manipulations by

Ahmad Abdolmaleki, 22 Feb 2016 10:39

Fold

Re: Vanished band gap after core levels manipulations

Caterina Cocchi 23 Feb 2016 17:20

Dear Ahmad,

in semiconductor systems, it is recommended to set the swidth parameter small enough to prevent unphysical electronic features, such as the disappearing of the fundamental gap. This is generally valid, independently of the manipulation of the species files.

In your specific case, where a P 1s electron is described by a local orbital, you have to be very careful with the electronic structure of the system, also far from the gap region. In fact, the representation of such deep and extremely localized core states with local orbitals might not be straightforward, and problems can appear in the eigenvalue spectrum.

Best regards,
Caterina Cocchi
(exciting team)

Reply Options

Unfold Re: Vanished band gap after core levels manipulations by

Caterina Cocchi, 23 Feb 2016 17:20

Fold

Re: Vanished band gap after core levels manipulations

Ahmad Abdolmaleki 29 Feb 2016 10:17

Dear Caterina
Thank you for your answer.

I ran a new calculation via following parameters:

   <groundstate
      stype="Fermi Dirac"
      swidth="0.00001d0"
      dipolecorrection="true"
      epsband="1.0d-14"
      deband="0.003d0"
      rgkmax="8"
      ngridk="31 31 1"
      gmaxvr="14.0"
      xctype="GGA_PBE"
      nempty="15"/>

But the problem still remains. So as you noted, I guess it is because of the definition of the local orbital.
Would you please direct me somehow to make a better local orbital?

Best regards,
Ahmad Abdolmaleki

Reply Options

Unfold Re: Vanished band gap after core levels manipulations by

Ahmad Abdolmaleki, 29 Feb 2016 10:17

Fold

Re: Vanished band gap after core levels manipulations

Caterina Cocchi 01 Mar 2016 15:34

Dear Ahmad,

such deep core states cannot be easily described with local orbitals, due to their very pronounced localization. We are aware of the problem and we are working to overcome it by introducing an improved treatment of core states, especially in view of excitations. We expect to include these features in the next release of exciting, which will be presented in the "How exciting!" workshop and hands-on tutorial, that we are organizing for next summer. For more information, please visit the related webpage: http://how-exciting-2016.physik.hu-berlin.de/.

Best regards,
Caterina Cocchi
(exciting team)

Reply Options

Unfold Re: Vanished band gap after core levels manipulations by

Caterina Cocchi, 01 Mar 2016 15:34

New Post

exciting

exciting, all electron DFT

Home

Downloads

How to reach us

Development