Pseudopotential for M06-L

[August Melcher]

Hello all,

I'm interested in performing condensed phase AIMD simulations using the M06-L functional and I'm trying to decide which pseudopotential to use. Would anyone recommend one of the existing pseudopotentials over the others? Will libxc functionals work with the ATOM package for pseudopotential optimization? If so, is it likely to be worth optimizing new pseudopotentials?

Thanks,

August

[hut...@chem.uzh.ch]

Hi

usually another pseudo for a functional of similar type is a good
starting point. I would use PBE in this case.

You can use LIBXC functionals in the atomic code. However, there are
some additional restrictions, e.g. no Laplace functional.
A repotimization starting from PBE values should be easy, at least for
the lighter elements.

Juerg

--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----
To: cp...@googlegroups.com
From: August Melcher
Sent by: cp...@googlegroups.com
Date: 06/10/2014 11:47PM
Subject: [CP2K:5390] Pseudopotential for M06-L

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at http://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

[August Melcher]

Great, thanks! I've started doing some all-electron calculations with M06-L to compare with the pseudopotential calculations and I'm running into something weird. I ran a dozen calculations with the attached input file for a water dimer (with differing coordinates), some of them converge quickly (~30min), some of them don't converge after 2:30, but all of them print hundreds to thousands of lines like those below in between each SCF step. Is this necessary? Is there something wrong with my input file? Could this be related to the fact that some of them aren't converging?

I'd upload the output file as well but its over 1GB from all the extra lines.

  PROBLEM :::::            3          -3           3          -3           6           0

  PROBLEM :::::            3          -2           3          -2           6           0

  PROBLEM :::::            3          -1           3          -1           6           0

  PROBLEM :::::            3           1           3           1           6           0

  PROBLEM :::::            3           2           3           2           6           0

  PROBLEM :::::            3           3           3           3           6           0

  PROBLEM :::::            3          -3           3          -2           6           1

  PROBLEM :::::            3          -2           3          -3           6           1

  PROBLEM :::::            3          -2           3          -1           6           1

  PROBLEM :::::            3          -1           3          -2           6           1

  PROBLEM :::::            3           1           3           2           6           1

  PROBLEM :::::            3           2           3           1           6           1

  PROBLEM :::::            3           2           3           3           6           1

  PROBLEM :::::            3           3           3           2           6           1

Thanks,

August

[hut...@chem.uzh.ch]

Hi

these messages are related with a problem in the one-center
term for meta-functionals. As discussed here before, the current
code ignores a contribution to the kinetic energy density in GAPW
calculations. This causes a small error in energy and a small
inconsistency in the electronic forces.

You can try to set LMAXN1 in
CP2K_INPUT / FORCE_EVAL / DFT / QS
to a smaller value. I would use LMAXN1=4.

regards

Juerg

--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----
To: cp...@googlegroups.com
From: August Melcher
Sent by: cp...@googlegroups.com

Date: 06/16/2014 09:55PM
Subject: [CP2K:5411] Re: Pseudopotential for M06-L

Great, thanks! I've started doing some all-electron calculations with M06-L to compare with the pseudopotential calculations and I'm running into something weird. I ran a dozen calculations with the attached input file for a water dimer (with differing coordinates), some of them converge quickly (~30min), some of them don't converge after 2:30, but all of them print hundreds to thousands of lines like those below in between each SCF step. Is this necessary? Is there something wrong with my input file? Could this be related to the fact that some of them aren't converging?

  PROBLEM :::::            3          -3           3          -3           6           0
  PROBLEM :::::            3          -2           3          -2           6           0
  PROBLEM :::::            3          -1           3          -1           6           0
  PROBLEM :::::            3           1           3           1           6           0
  PROBLEM :::::            3           2           3           2           6           0
  PROBLEM :::::            3           3           3           3           6           0
  PROBLEM :::::            3          -3           3          -2           6           1
  PROBLEM :::::            3          -2           3          -3           6           1
  PROBLEM :::::            3          -2           3          -1           6           1
  PROBLEM :::::            3          -1           3          -2           6           1
  PROBLEM :::::            3           1           3           2           6           1
  PROBLEM :::::            3           2           3           1           6           1
  PROBLEM :::::            3           2           3           3           6           1
  PROBLEM :::::            3           3           3           2           6           1

Thanks,
August

[August Melcher]

Thanks! Is there any way to suppress the printing of the PROBLEM error message? The output files are extremely hard to look at when they're full of error messages. I tried PRINT_LEVEL SILENT but it didn't work.

The reason I'm doing GAPW calculations is for pseudopotential testing, comparing water dimer potential energy curves between all electron vs pseudopotential calculations. Is this still a valid method of testing given the shortcomings of meta-GGAs in GAPW? I was thinking I could instead compare with some other published M06-L data or is there a better way to test pseudopotentials?

[hut...@chem.uzh.ch]

Hi

The only way is to comment out the line in the source code.

qs_vxc_atom.F

regards

Juerg

--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----
To: cp...@googlegroups.com
From: August Melcher
Sent by: cp...@googlegroups.com

Date: 06/19/2014 09:24PM
Subject: [CP2K:5426] Re: Pseudopotential for M06-L

Thanks! Is there any way to suppress the printing of the PROBLEM error message? The output files are extremely hard to look at when they're full of error messages. I tried PRINT_LEVEL SILENT but it didn't work.

The reason I'm doing GAPW calculations is for pseudopotential testing, comparing water dimer potential energy curves between all electron vs pseudopotential calculations. Is this still a valid method of testing given the shortcomings of meta-GGAs in GAPW? I was thinking I could instead compare with some other published M06-L data or is there a better way to test pseudopotentials?

[August Melcher]

Hello again,

I'm running into a problem computing water dimer potential energy curves with GPW. I tried three different pseudopotentials (PBE, PBE0, and PBE "re-optimized" for M06-L) and two different basis sets (TZV2P and MOLOPT-TZV2PX) and all of them significantly underbind the dimer. 

Attached are the curves as well as a sample input file. The reference data was taken from "Further analysis and comparative study of intermolecular interactions using dimers from the S22 database" Laszlo Fusti Molnar, Xiao He, Bing Wang and Kenneth M. Merz Jr. J. Chem. Phys. 131, 065102 (2009); http://dx.doi.org/10.1063/1.3173809

I noticed the comment below in another thread. Could this be the cause of the underbinding I'm noticing? Or perhaps there is something wrong with my input file?

"I looked at some of the source code files related to the calling of the XC functionalities. In particular, the xc_libxc.f file states the following as a comment at the beginning of the file.

!>      WARNING: In the subroutine libxc_lsd_calc, it could be that the
!>      ordering for the 1st index of v2lapltau, v2rholapl, v2rhotau,
!>      v2sigmalapl and v2sigmatau is not correct. For the moment it does not
!>      matter since the calculation of the 2nd derivatives for meta-GGA
!>      functionals is not implemented in CP2K.

If this is the case, then it is possible that a meta-GGA functional (which generally is a function of the laplacian of the density) such as M06-2X may not work properly in CP2K. Sorry I could not be of more help."

Thanks,

August

[Christopher Knight]

Hi August,

The correlation part seems to be missing from that input, try the following: FUNCTIONAL XC_MGGA_X_M06_L XC_MGGA_C_M06_L

With the M06_L PP parameters below and a couple quick calculations (500Ry, TZV2P, 10 Angstrom^3), I get 

BE = 2*E_Monomer - E_Dimer = 5.48 kcal/mol.

chris

=============

#

H GTH-M06_L-q1

    1

    0.200000000000       2   -4.30444    0.799586

       0

#

O GTH-M06_L-q6

    2    4

     0.24084771    2   -17.00218284     2.49799489

    1

     0.22187778    1    18.01262971

On Jun 26, 2014, at 4:57 PM, August Melcher wrote:

Hello again,

I'm running into a problem computing water dimer potential energy curves with GPW. I tried three different pseudopotentials (PBE, PBE0, and PBE "re-optimized" for M06-L) and two different basis sets (TZV2P and MOLOPT-TZV2PX) and all of them significantly underbind the dimer. 

Attached are the curves as well as a sample input file. The reference data was taken from "Further analysis and comparative study of intermolecular interactions using dimers from the S22 database" Laszlo Fusti Molnar, Xiao He, Bing Wang and Kenneth M. Merz Jr. J. Chem. Phys. 131, 065102 (2009); http://dx.doi.org/10.1063/1.3173809

I noticed the comment below in another thread. Could this be the cause of the underbinding I'm noticing?

"I looked at some of the source code files related to the calling of the XC functionalities. In particular, the xc_libxc.f file states the following as a comment at the beginning of the file.

!>      WARNING: In the subroutine libxc_lsd_calc, it could be that the
!>      ordering for the 1st index of v2lapltau, v2rholapl, v2rhotau,
!>      v2sigmalapl and v2sigmatau is not correct. For the moment it does not
!>      matter since the calculation of the 2nd derivatives for meta-GGA
!>      functionals is not implemented in CP2K.

If this is the case, then it is possible that a meta-GGA functional (which generally is a function of the laplacian of the density) such as M06-2X may not work properly in CP2K. Sorry I could not be of more help."

Thanks,

August

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at http://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

<m06-L.inp><pseudobench.xlsx>

[August Melcher]

Fantastic! Thanks a lot. So I take it that pseudopotential has already been optimized for M06-L?

[Christopher Knight]

Yes, those PP parameters were optimized with the M06-L functional.  For H and O, they're pretty similar to PP parameters for a handful of GGAs.

chris

On Jun 27, 2014, at 4:33 PM, August Melcher <august...@gmail.com> wrote:

Fantastic! Thanks a lot. So I take it that pseudopotential has already been optimized for M06-L?

[August Melcher]

Did you do it using the CP2K/ATOM/PSEUDOPOTENTIAL_OPTIMIZATION module? If so, would you mind showing me the input file you used? I'm having a hard time figuring out how to alter the regtest examples to give reasonable results.

[Christopher Knight]

Sure, a template is below.  I use this with another code to drive CP2K and sample several sets of initial conditions (the v_# arguments below) and resulting optimized solutions.

chris

=========

cjknight:M06_L>cat cp2k.template.inp 

&GLOBAL

  PROGRAM_NAME ATOM

  PRINT_LEVEL DEBUG

&END GLOBAL

&ATOM

  ATOMIC_NUMBER 1

  ELEMENT H

  RUN_TYPE PSEUDOPOTENTIAL_OPTIMIZATION

  CORE none

  ELECTRON_CONFIGURATION 1s1

  MAX_ANGULAR_MOMENTUM 2

  CALCULATE_STATES 4 3 3

  COULOMB_INTEGRALS ANALYTIC

  EXCHANGE_INTEGRALS ANALYTIC

  &METHOD

     METHOD_TYPE  KOHN-SHAM

     RELATIVISTIC DKH(2)

     &XC

       &XC_FUNCTIONAL

         &LIBXC

           FUNCTIONAL XC_MGGA_X_M06_L XC_MGGA_C_M06_L

         &END LIBXC

       &END XC_FUNCTIONAL

     &END XC

  &END METHOD

  &OPTIMIZATION

    EPS_SCF 1.e-10

  &END

  &AE_BASIS

     BASIS_TYPE GEOMETRICAL_GTO

  &END AE_BASIS

  &PP_BASIS

     BASIS_TYPE GEOMETRICAL_GTO

  &END PP_BASIS

  &POTENTIAL

    PSEUDO_TYPE GTH

    CONFINEMENT 0.0 4.0 2.0

    &GTH_POTENTIAL

      1

        v_S_RC        2    v_S_C1           v_S_C2

      0

    &END 

  &END POTENTIAL

  &POWELL

     ACCURACY   1.e-6

!     ACCURACY   1.e-9

     STEP_SIZE  0.005

     MAX_FUN 50

  &END

&END ATOM

On Jun 27, 2014, at 5:08 PM, August Melcher wrote:

Did you do it using the CP2K/ATOM/PSEUDOPOTENTIAL_OPTIMIZATION module? If so, would you mind showing me the input file you used? I'm having a hard time figuring out how to alter the regtest examples to give reasonable results.