[CP2K:9170] wB97X-D functional with libxc-3.0.0

[hut... at chem.uzh.ch]

Hi

- the damping functions (2 variations) are hard coded for the
  empirical D3
- there is a known bug for meta-functionals and GAPW, you will not be
  able to get sub milli-Hartree accuracy
- You can test different functionals to see what accuracy can be achieved
  e.g. PBE, PBE0, TPSS, ...

regards

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

-----cp... at googlegroups.com wrote: -----To: cp2k <cp... at googlegroups.com>
From: John Herr 
Sent by: cp... at googlegroups.com
Date: 06/28/2017 05:30PM
Subject: [CP2K:9170] wB97X-D functional with libxc-3.0.0

I'm trying to get the wB97X-D functional working in cp2k-4.1. I have already compiled cp2k with libxc-3.0.0, and I am trying to compare a calculation of a single water molecule with q-chem to make sure it's correct (will compare with more later, just using this as a starting point). For reference I am using the 6-311g** basis. In q-chem I can get an energy component breakdown like this


 ---------------------------------------
  Cycle       Energy         DIIS Error
 ---------------------------------------
    1     -76.3663998805      1.11E-01
    2     -76.3620453973      2.72E-02
    3     -76.3749719700      2.59E-02
    4     -76.4212961721      1.07E-03
    5     -76.4213693768      2.28E-04
    6     -76.4213747966      2.02E-05
    7     -76.4213748557      1.79E-06 Convergence criterion met
 ---------------------------------------
 One-Electron    Energy =      -123.5178738137
 Total Coulomb   Energy =        47.1112629588
 Alpha Exchange  Energy =       -1.4239467336
 Beta  Exchange  Energy =       -1.4239467336
 DFT   Exchange  Energy =      -6.0846516212
 DFT Correlation Energy =        -0.4328050558
 Nuclear Repu.   Energy =         9.3505861434
 Nuclear Attr.   Energy =       -199.8742482744
 Kinetic         Energy =            76.3563744607


which I would like to compare with cp2k to make sure the components of the energy are right, but I can't seem to figure out how to get these to print, or what to compare them to out of what does print out. Here is the energy breakdown after the last iteration from cp2k

    11 DIIS/Diag.  0.73E-05    1.6     0.00000731       -76.4226387492 -1.09E-09

  *** SCF run converged in    11 steps ***


  Electronic density on regular grids:               -7.9264937891        2.0735062109
  Core density on regular grids:                       10.0000000000       -0.0000000000
  Hard and soft densities (Lebedev):               -16.3078231164      -14.2343170120
  Total Rho_soft + Rho1_hard - Rho1_soft       -9.9999998935
  Total charge density (r-space):                      0.0000001065
  Total Rho_soft + Rho0_soft (g-space):           0.0000001024

  Overlap energy of the core charge distribution:                0.00000020808927
  Self energy of the core charge distribution:                   -75.73106814560732
  Core Hamiltonian energy:                                             -5.18060334470231
  Hartree energy:                                                           21.19955249945549
  Exchange-correlation energy:                                       -2.68058148507007
  Hartree-Fock Exchange energy:                                   -2.84801596711676

  GAPW| Exc from hard and soft atomic rho1:                 -3.83867168544622
  GAPW| local Eh = 1 center integrals:                           -7.34325082876699

  Total energy:                                                            -76.42263874916490

I have attached my cp2k input file. I think that I have the HF interaction potential set correctly, but not 100% sure on that. As far as the vdW potential, I am not sure what to set, as it seems that cp2k only has the zero damping and Becke-Johnson damping function, but the damping function in wB97X-D is different from either of these. Any help with this would be much appreciated.








-- 

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... at googlegroups.com.

To post to this group, send email to cp... at googlegroups.com.

Visit this group at https://groups.google.com/group/cp2k.

For more options, visit https://groups.google.com/d/optout.



[attachment "h2o-wb97xd-cp2k.inp" removed by Jürg Hutter/at/UZH]