MULLIKEN CHARGE

Josip Lovrić

unread,

Dec 19, 2016, 10:56:55 AM12/19/16

to cp2k

Dear CP2K community,

Recently I was working on NO2-dimer and now I need help from you. Thing is that I am running gas phase calculation of NO2 cis-dimer. After program successfully finish GEOMETRIC OPTIMIZATION output configuration is consistent with already published papers. Problem is in the MULLIKEN charges. Charges are too small and not consistent with published values(I am using same basis set).

!-----------------------------------------------------------------------------!

Mulliken Population Analysis

# Atom Element Kind Atomic population Net charge

1 N 1 4.968657 0.031343

2 O 2 5.852787 0.147213

3 O 2 6.191534 -0.191534

4 N 1 4.928901 0.071099

5 O 2 6.056358 -0.056358

6 O 2 6.001764 -0.001764

# Total charge 34.000000 0.000000

!-----------------------------------------------------------------------------!

I am using BLYP functional with dispersion corrections and TZV2P basis set with pseudo-potentials. Charges are becoming more physical when performing all electron calculation, they increase for factor ~10.

For DZVP basis set charges increase little, but still to low:

!-----------------------------------------------------------------------------!

Mulliken Population Analysis

# Atom Element Kind Atomic population Net charge

1 N 1 4.951639 0.048361

2 O 2 5.822911 0.177089

3 O 2 6.202144 -0.202144

4 N 1 4.858609 0.141391

5 O 2 6.103330 -0.103330

6 O 2 6.061367 -0.061367

# Total charge 34.000000 -0.000000

!-----------------------------------------------------------------------------!

Am I doing something wrong? Do you have any explanation for this results?

I would really appreciate.

I am putting in attachment my input and output files.

Best wishes

Josip

cis_dimer_all_electrons.inp

cis_dimer_all_electrons.out

mulliken.dat--1.mulliken

Matt W

unread,

Dec 19, 2016, 11:03:50 AM12/19/16

to cp2k

Hi Josip,

Mulliken charges are not real observables - they are strongly dependent on the basis set, and they do not converge as you make the basis set complete. Check a quantum chemistry textbook for a good discussion ('Computational Chemistry', Errol Lewars for instance).

Something more quantitative can come from something like Bader charge analysis (which needs external processing of a total_charge_density cube file).

Matt

Josip Lovrić

unread,

Dec 19, 2016, 11:09:18 AM12/19/16

to cp2k

Dear Matt,

I totally agree with you and I am aware of fact that atomic charge are just model.

Maybe I was not clear enough but my problem is that charges are not consistent with published values.

P.S. input files are named "*_all_electrons*". This is mistake, calculation is not all electron, it is a relict from previous calculation.

Josip

hut...@chem.uzh.ch

unread,

Dec 19, 2016, 11:16:49 AM12/19/16

to cp...@googlegroups.com

Hi

can you show us an example?

Literature: Basis Pseudopotential/All electron Mulliken charges

CP2K: Basis Pseudopotential/All electron Mulliken charges

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: cp2k <cp...@googlegroups.com>
From: Josip Lovrić
Sent by: cp...@googlegroups.com
Date: 12/19/2016 05:09PM
Subject: [CP2K:8491] Re: MULLIKEN CHARGE

Dear Matt,
I totally agree with you and I am aware of fact that atomic charge are just model. Maybe I was not clear enough but my problem is that charges are not consistent with published values.
P.S. input files are named "*_all_electrons*". This is mistake, calculation is not all electron, it is a relict from previous calculation.
Josip

On Monday, December 19, 2016 at 5:03:50 PM UTC+1, Matt W wrote:Hi Josip,
Mulliken charges are not real observables - they are strongly dependent on the basis set, and they do not converge as you make the basis set complete. Check a quantum chemistry textbook for a good discussion ('Computational Chemistry', Errol Lewars for instance).
Something more quantitative can come from something like Bader charge analysis (which needs external processing of a total_charge_density cube file).
Matt
On Monday, December 19, 2016 at 3:56:55 PM UTC, Josip Lovrić wrote:Dear CP2K community,

Recently I was working on NO2-dimer and now I need help from you. Thing is that I am running gas phase calculation of NO2 cis-dimer. After program successfully finish GEOMETRIC OPTIMIZATION output configuration is consistent with already published papers. Problem is in the MULLIKEN charges. Charges are too small and not consistent with published values(I am using same basis set).!-----------------------------------------------------------------------------! Mulliken Population Analysis

# Atom Element Kind Atomic population Net charge 1 N 1 4.968657 0.031343 2 O 2 5.852787 0.147213 3 O 2 6.191534 -0.191534 4 N 1 4.928901 0.071099 5 O 2 6.056358 -0.056358 6 O 2 6.001764 -0.001764 # Total charge 34.000000 0.000000

!-----------------------------------------------------------------------------!I am using BLYP functional with dispersion corrections and TZV2P basis set with pseudo-potentials. Charges are becoming more physical when performing all electron calculation, they increase for factor ~10. For DZVP basis set charges increase little, but still to low: !-----------------------------------------------------------------------------! Mulliken Population Analysis

# Atom Element Kind Atomic population Net charge 1 N 1 4.951639 0.048361 2 O 2 5.822911 0.177089 3 O 2 6.202144 -0.202144 4 N 1 4.858609 0.141391 5 O 2 6.103330 -0.103330 6 O 2 6.061367 -0.061367 # Total charge 34.000000 -0.000000

!-----------------------------------------------------------------------------! Am I doing something wrong? Do you have any explanation for this results?I would really appreciate.

I am putting in attachment my input and output files.
Best wishesJosip

--

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 https://groups.google.com/group/cp2k.

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

Matt W

unread,

Dec 19, 2016, 11:19:53 AM12/19/16

to cp2k

Hi,

apologies for patronising. Which published values are you comparing to?

At a glance your values don't look unreasonable to me.

Your input looks OK, except the cutoff is rather small - to get away with that small a value you need quite heavy smoothing of the XC - do the values change if you increase the cutoff?

Matt

Matt W

unread,

Dec 19, 2016, 11:29:54 AM12/19/16

to cp2k

Also, your box is too small, and maybe you don't centre your molecule

&CELL

ABC 10.0 10.0 10.0

PERIODIC NONE

&END CELL

&COORD

N 2.233559 -3.621843 1.337023

O 2.067990 -2.527433 1.053663

O 3.021496 -3.912949 2.863941

N 3.415616 -2.732165 3.611083

O 3.805883 -1.768317 2.949635

O 3.319372 -2.851822 4.823878

&END

&TOPOLOGY

&CENTER_COORDINATES

&END

&END TOPOLOGY

Try something like above and see if it helps.

Matt

Josip Lovrić

unread,

Dec 19, 2016, 1:05:55 PM12/19/16

to cp2k

Sorry for not responding immediately, I was of computer for some time.

Published article on which I am referring is this http://pubs.acs.org/doi/suppl/10.1021/jz3014985 and they give results for BASIS PSEUDOPOTENTIAL. I have deleted my all electron calculations. I will repeat them if necessary.

Now when I wanted to copy paste their results here I see that they actually performed gas phase calculation with NO2 dimer + water and HCl molecule.

So my comparison make no sense, sorry for wasting your time. Nevertheless this stays strange for me because I have also performed this calculation (cis dimer isolate) in Gaussian(basis downloaded from https://bse.pnl.gov/bse/portal) and charges are closed to one in paper.