Author Topic: Pseudopotential transfer between ATK and SIESTA  (Read 15236 times)

0 Members and 1 Guest are viewing this topic.

Offline vihardabest

  • Regular QuantumATK user
  • **
  • Posts: 22
  • Country: gb
  • Reputation: 0
  • Vihar Georgiev
    • View Profile
Pseudopotential transfer between ATK and SIESTA
« on: April 22, 2009, 18:41 »
Dear all,
I would like to know if it is possible  to transfer psoudopotentials between ATK and SIESTA.

I made N-pseudopotential in SIESTA, the beginning of the file is :


SIESTA---------- NITROGEN.psf


  N ca nrl nc
 ATM 3.2.2 22-APR-09 Kerker                                   
 2s 2.00  r= 0.92/2p 3.00  r= 0.80/3d 0.00  r= 0.79/4f 0.00  r= 0.79/ 
   4  0 1018  0.354107453809E-03  0.125000000000E-01   5.00000000000   
 Radial grid follows
  0.445412344792E-05  0.896427287179E-05  0.135311529916E-04  0.181554773917E-04
  0.228379686322E-04  0.275793583619E-04  0.323803874325E-04  0.372418060146E-04
  0.421643737147E-04  0.471488596942E-04  0.521960427889E-04  0.573067116317E-04
  0.624816647748E-04  0.677217108153E-04  0.730276685210E-04  0.784003669585E-04
  0.838406456230E-04  0.893493545690E-04  0.949273545435E-04  0.100575517120E-03
  0.106294724837E-03  0.112085871330E-03  0.117949861479E-03  0.123887611544E-03
  0.129900049310E-03  0.135988114233E-03  0.142152757586E-03  0.148394942607E-03
  0.154715644649E-03  0.161115851336E-03  0.167596562712E-03  0.174158791402E-03
-----------------------------------------------------------------------------------------------------------------------

The ATK pseudopotential file with name - n_pz.upf  contain this information in the begining:

<PP_INFO>
 ATM3      no_date   Troullier-Martins
</PP_INFO>
<PP_HEADER>
    0                  Version Number
    N                  Element
   NC                  Norm - Conserving pseudopotential
    T                  Nonlinear Core Correction
 SLA  PZ   NOGX NOGC   PZ   Exchange-Correlation functional
                5      Z valence
                0      Total energy
  0.0000000  0.0000000 Suggested cutoff for wfc and rho
    3                  Max angular momentum component
 1018                  Number of points in mesh
    4    4             Number of Wavefunctions, Number of Projectors
 Wavefunctions         nl  l   occ
                       2S  0  2
                       2P  1  3
                       3D  2  0
                       4F  3  0
</PP_HEADER>
<PP_MESH>
<PP_R>
     4.45412344792e-06     8.96427287179e-06     1.35311529916e-05     1.81554773917e-05
     2.28379686322e-05     2.75793583619e-05     3.23803874325e-05     3.72418060146e-05
     4.21643737147e-05     4.71488596942e-05     5.21960427889e-05     5.73067116317e-05
     6.24816647748e-05     6.77217108153e-05      7.3027668521e-05     7.84003669585e-05
      8.3840645623e-05      8.9349354569e-05     9.49273545435e-05      0.00010057551712
     0.000106294724837      0.00011208587133     0.000117949861479     0.000123887611544
      0.00012990004931     0.000135988114233     0.000142152757586     0.000148394942607
     0.000154715644649     0.000161115851336     0.000167596562712     0.000174158791402
     0.000180803562768     0.000187531915069     0.000194344899623     0.000201243580973
     0.000208229037052     0.000215302359352     0.000222464653093       0.0002297170374
     0.000237060645471      0.00024449662476     0.000252026137154     0.000259650359155
------------------------------------------------------------------------------------------------------------------------------
The beginning of the radial grid is the same. I checked the file further and the numbers are almost match. Therefore my question is:
Can I use the redial grid from ATK program in my SIESTA calculation? ???
Obviously their is some relation.
One more question:
what is the difference between the files:
'chemical-element' _pz.upf and 'hemical-element' _pbe.upf
Thank you very much for your help. ;) ::)

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5565
  • Country: dk
  • Reputation: 93
    • View Profile
    • QuantumATK at Synopsys
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #1 on: April 22, 2009, 20:51 »
Theoretically this should be quite possible, if you can match the individual grids, etc. Just try :)

The difference in the two files lies in the exchange-correlation:
PZ = LDA Perdew-Zunger
PBE = GGA Perdew-Burke-Ernzerhof

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5565
  • Country: dk
  • Reputation: 93
    • View Profile
    • QuantumATK at Synopsys
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #2 on: April 22, 2009, 20:54 »
There are some software tools out there which can convert different pseudopotential formats between each other. Google something like "upf psf convert pseudopotential".

Offline vihardabest

  • Regular QuantumATK user
  • **
  • Posts: 22
  • Country: gb
  • Reputation: 0
  • Vihar Georgiev
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #3 on: April 22, 2009, 23:22 »
Thank you Anders :)
I will try and let you know.
Have a nice work tomorrow :)

Offline xhsh

  • Heavy QuantumATK user
  • ***
  • Posts: 32
  • Reputation: 0
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #4 on: October 26, 2009, 16:52 »
Have you got any tools to transfer pseudopotentials between upf and psf format? Or have anybody else got some tools to do that?

I googled with great efforts, but I got nothing.

Offline ipsecog

  • Heavy QuantumATK user
  • ***
  • Posts: 35
  • Country: se
  • Reputation: 0
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #5 on: October 26, 2009, 17:18 »

Offline xhsh

  • Heavy QuantumATK user
  • ***
  • Posts: 32
  • Reputation: 0
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #6 on: October 27, 2009, 02:49 »
I don't have personal experience of this, but I know of some resources that might be worth checking out:



Hello, ipsecog. Thank you very much!!

Maybe we have to DIY on this matter.

Offline zhangguangping

  • QuantumATK Guru
  • ****
  • Posts: 193
  • Country: cn
  • Reputation: 2
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #7 on: September 28, 2010, 10:50 »
There are some software tools out there which can convert different pseudopotential formats between each other. Google something like "upf psf convert pseudopotential".
Dear Anders Blom,
Can you give some hints on the  software tools which can convert different pseudopotential formats between each other.I need one software to convert the psf format which siesta can use to upf format which ATK uses.

I have found a tool called "pspconvert",but I can not download it.http://www.tddft.org/trac/APE/browser/pspconvert?desc=1

Can you suggest another one!


Thanks in advance.

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #8 on: September 29, 2010, 02:57 »
see  here to find out how to download it:
Quote
http://www.tddft.org/programs/APE/

Offline zhangguangping

  • QuantumATK Guru
  • ****
  • Posts: 193
  • Country: cn
  • Reputation: 2
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #9 on: September 29, 2010, 05:57 »
see  here to find out how to download it:
Quote
http://www.tddft.org/programs/APE/
Dear zh,
Thanks for your help.APE can generate all the kinds of PP for a input files?

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #10 on: September 29, 2010, 13:29 »
Quote
APE can generate all the kinds of PP for a input files?

Sorry, I can not understand your questions. Please state your question more clear.  Anyway, about the detailed usage of APE, please refer to its manual.

Offline zhangguangping

  • QuantumATK Guru
  • ****
  • Posts: 193
  • Country: cn
  • Reputation: 2
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #11 on: October 1, 2010, 11:02 »
Quote
APE can generate all the kinds of PP for a input files?

Sorry, I can not understand your questions. Please state your question more clear.  Anyway, about the detailed usage of APE, please refer to its manual.

I followed your instruction,and download the APE,and learn how to use,but when give the generated psf format pp to sieta,siesta always complains it.Oh....and the upf can not be used by atk.The attachment is the pps generate by APE.
« Last Edit: October 1, 2010, 11:20 by zhangguangping »

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #12 on: October 1, 2010, 13:10 »
You had better check the difference in the UPF format used in APE and ATK, and adjust the format of upf file used in APE and make it be compatible with ATK. The UPF format proposed initially by the PWSCF code has been updated. Maybe, the version of UPF format used in APE and ATK are different.
« Last Edit: October 1, 2010, 13:25 by zh »

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5565
  • Country: dk
  • Reputation: 93
    • View Profile
    • QuantumATK at Synopsys
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #13 on: October 1, 2010, 18:04 »
I just tried your pp with ATK 10.8, it works fine.

Offline zhangguangping

  • QuantumATK Guru
  • ****
  • Posts: 193
  • Country: cn
  • Reputation: 2
    • View Profile
Re: Pseudopotential transfer between ATK and SIESTA
« Reply #14 on: October 2, 2010, 09:41 »
I just tried your pp with ATK 10.8, it works fine.
I use ATK-2008.10,and it does not work.Maybe there is difference as zh suggested between the upf I generated and AKT use.Is there different in format for the pps used in 2008.10 and 2010.8
But when I tried to make the format the same ,2008 yet does not work
Thank you and zh again.
« Last Edit: October 2, 2010, 10:44 by zhangguangping »