CCL: NBO vs Loewdin Charges



Hi,

The difficulty with charge assignment schemes is largely due to historical artifacts. 
All of the earliest schemes for assigning net atomic charges were seriously flawed: Mulliken, Lowden, Hirshfeld, etc.
Even Bader's AIM theory, which has many beautiful theoretical properties, can give non-nuclear attractors in some materials.
The persistent use of methods like Mulliken and Lowden, which have been known for decades (longer than most of us have been alive)
to be highly unstable and have no well-defined basis set limit, has contributed to the confusion.

From a historical perspective, this is similar to people's early aversion to "flying machines".
For many decades, the concept of heavier than air flying machines (what we now call airplanes) was considered a crackpot idea (a "fiction").
This is because early attempts to build "flying machines" were dismal failures.
It was only after people were able to construct successful demonstrations that the societal resistance to airplanes began to reverse.
Nowadays, it is hard for us to imagine or justify the profound societal resistance to building airplanes that existed around 1900.

Some of the modern charge assignment methods are getting much better than the earliest ones.
There is mounting experimental and computational evidence that some of the most recent schemes
are orders of magnitude more reliable than those developed decades ago.
Hopefully, this recent progress will go a long way towards convincing computational chemistry researchers that 
the limitations we have become accustomed to can be reversed after all.
I realize all of this will take time, but as a computational chemistry community we need to make an effort for progress.

The continual use of explicitly basis set dependent methods like Mulliken and Lowden is a bit weary at times, and comes off as a 
serious shame in the face of better available recent methods. As a computational chemistry community what can we do to support progress?
Do you think it is appropriate at this time for computational chemistry journals to develop standards that Mulliken and Lowden
populations are not considered to be of publishable quality and that published populations should be from methods with a 
well-defined basis set limit? One could argue that it is important for published articles to show the fundamental limitations of
Mulliken and Lowden schemes, but this has been established in countless papers over the past 50 years. 
I recall that my master's thesis advisor, who is a purely experimental catalysis researcher, warned us many years ago
about the limitations of Mulliken charges. If experimentalists knew about this all the way back then, why as a leading
computational chemistry community are we still fighting the obvious decades later?

It may be that the computational chemistry community has been slow to develop better alternatives.
I believe the continued use of Mulliken and Lowden schemes as default methods in some
quantum chemistry programs is due to the inherent limitations of other available charge assignment schemes.
Until recently, the NPA/NBO scheme, which has many great properties, was non-convergent for many systems
near the complete basis set limit. My experience and tests have shown this improved within the last year.
The NBO website and manual says that recently they have improved the projection algorithm to make it
more reliable. So, at least until very recently, the convergence of the NPA/NBO scheme was not sufficiently
reliable to use it as a default charge assignment method. The issue of non-nuclear attractors has precluded
using Bader's scheme as a default method for assigning net atomic charges. The Hirshfeld method assigns
net atomic charges that are too small in magnitude. Methods such as CHELP, CHELPG, ESP, etc. do not work
for dense nonporous materials that lack a van der Waals surface. Computing the APT and Born effective charges 
requires a computationally expensive frequency (first-order perturbation theory) computation or a series of energy calculations with
various atomic displacements. Thus, it seems that until recently the computational chemistry community has been stuck with 
seriously flawed methods such as Mulliken and Lowden as defaults in popular quantum chemistry programs.

For the past couple of years, I and one of my Ph.D. students have been developing a better alternative.
The goal of our project is to develop a general-purpose atomic population analysis method that works
nearly flawlessly across a broad range of diverse material classes, including both molecular and solid-state
materials. We have made extensive comparisons to experimental properties across diverse materials.
Our objective is to develop a method that converges reliably and efficiently and would be ideally suited
for use as a default atomic population analysis method in popular quantum chemistry programs
irrespective of the basis set type. We currently have a manuscript on this new method under review.
We are hoping that it is not too late for the computational chemistry community to improve.
We earnestly solicit your support to enable the computational chemistry community to make progress in the important
area of atomic population analysis methods.

Sincerely,

Tom


On Wed, Sep 2, 2015 at 12:11 AM, Irena Efremenko irena.efremenko*weizmann.ac.il <owner-chemistry\a/ccl.net> wrote:
Thanks, Tom, agree. But my main idea is that atomic charges derived > from QC calculations is a kind of fiction. I would recommend not to use it at all, especially for transition metals, and even more for such heavy atoms as Pu.   
******
Dr. Irena Efremenko
Department of Organic Chemistry
Weizmann Institute of Science
76100 Rehovot, Israel
Phone: +972 8 9343680
FAX:   +972 8 9343029
Email: irena.efremenko[*]weizmann.ac.il
******

On Sep 1, 2015, at 22:37 , Thomas Manz thomasamanz__gmail.com <owner-chemistry[*]ccl.net> wrote:

Irena,

You said "the basis set dependence of atomic charges arises from the basis set dependence of the spatial distribution of the electron density". This is only true for charge assignment methods that are functionals of the electron density.

It is profoundly violated by Mulliken, Lowden, and other charge assignment methods that have explicit basis set dependence. This is why Mulliken and Lowden charges are not generally considered useful.

Tom

On Tue, Sep 1, 2015 at 8:47 AM, Irena Efremenko irena.efremenko---weizmann.ac.il <owner-chemistry,+,ccl.net> wrote:
The initial question is lost in the discussion of technical problems. 

In my opinion, the answer is that ATOMS IN MOLECULES is not a physical observable. (corollary: partial atomic charges are not physical observables.) The partitioning of the electron density according to the definition of atoms in each specific scheme is what determines atomic charges, and this can be somewhat arbitrary. For example, in the Mulliken partitioning scheme the electron density between two bonded atoms is divided half-and-half, while in QTAIM atoms are divided by surfaces of zero flux in the gradient vector field of the electron density. Thus, corresponding charges should be different. 

Again, the basis set dependence of atomic charges arises from the basis set dependence of the spatial distribution of the electron density, especially in the range of diffuse functions. Usually, it influences calculated charges to a lesser degree than the partitioning scheme.
  
******
Dr. Irena Efremenko
Department of Organic Chemistry
Weizmann Institute of Science
76100 Rehovot, Israel
Phone: +972 8 9343680
FAX:   +972 8 9343029
Email: irena.efremenko,weizmann.ac.il
******

On Sep 1, 2015, at 08:06 , Dr. Robert Molt Jr. r.molt.chemical.physics+/-gmail.com <owner-chemistry,ccl.net> wrote:

Atomic charge models are not basis independent; see Chris Cramer's book, chapter 9, p.314. Some models give charges which are consistent with increasing basis set size, some are not, but all depend on the basis.

If you have anions, you cannot describe the density correctly without diffuse functions; this is a basis set dependence.
Dr. Robert Molt Jr.
 r.molt.chemical.physics]-[gmail.com
On 08/31/2015 11:14 PM, John McKelvey jmmckel(a)gmail.com wrote:
Hello,

There is a paper from a group in Germany demonstrating essentially basis set independence of atomic charges....  Terrible to say that while I remember the paper, and that the method has been installed in ORCA for the next release, I will have to search for the paper...  just moved from Indiana to Missouri so things are in a bit of disarray.

Maybe someone has seen this paper?


John
 

On Mon, Aug 31, 2015 at 5:47 PM, Víctor Luaña Cabal victor_-_fluor.quimica.uniovi.es <owner-chemistry/a\ccl.net> wrote:

Sent to CCL by: =?iso-8859-1?Q?V=EDctor_Lua=F1a?= Cabal [victor ~~ fluor.quimica.uniovi.es]
On Mon, Aug 31, 2015 at 04:22:51PM -0400, Jim Kress jimkress35-x-gmail.com wrote:
>
> Sent to CCL by: "Jim Kress" [jimkress35]-[http://gmail.com/" rel="noreferrer" target="_blank">gmail.com]
> Did you use an ECP for Plutonium and if so, was it properly represented in
> the AIM analysis?

Jim,

QTAIM should refer to the all electron density. However there have
been explorations of using ECP valence densities and they can work if
you do it carefully. You can either add the core contributions to the
valence density (it works) or be advise that your electron density is
not complete, so there are properties of the real QTAIM analysis that
can not work.

For instance, the basins and the topological partition of the full
electron density (rho) is not equivalent if you are using only the
valence component (val-rho).

However, if the basins are well defined there is no problem in
integrating the valence density within the full electron rho. The same
can be said for the analysis of spin components, and so on.
Using the critic2 code a lot of strange numerical experiments of
this kind can be easily performed. The code also has techniques to
fulfill the valence density with core components.

Read on critic2 in this ref.

A. Otero-de-la-Roza et al, Comput phys Commun 182 (2013) 2232--2248

An updated version of the code and manual can be found on

<http://azufre.quimica.uniovi.es/software.html#critic2>
<http://schooner.chem.dal.ca/wiki/Critic2>

I may look for old work published on this subject, if you are truly
interested.

Best regards,
             Víctor Luaña
--
     .  .    "In science a person can be convinced by a good argument.
    / `' \   That is almost impossible in politics or religion"
   /(o)(o)\  (Adapted from Carl Sagan)
  /`. \/ .'\  "Lo mediocre es peor que lo bueno, pero también es peor
 /   '`'`   \ que lo malo, porque la mediocridad no es un grado, es una
 |  \'`'`/  | actitud" -- Jorge Wasenberg, 2015
 |  |'`'`|  | (Mediocre is worse than good, but it is also worse than
  \/`'`'`'\/  bad, because mediocrity is not a grade, it is an attitude)
===(((==)))==================================+=========================
! Dr.Víctor Luaña, in silico chemist & prof. ! I hate the bureaucracy
! Departamento de Química Física y Analítica ! imposed by companies to
! Universidad de Oviedo, 33006-Oviedo, Spain ! which I owe nothing:
! e-mail:   victor(a)http://fluor.quimica.uniovi.es/" rel="noreferrer" target="_blank">fluor.quimica.uniovi.es   ! amazon, ResearchGATE and
! phone: +34-985-103491  fax: +34-985-103125 ! the like.
+--------------------------------------------+
 GroupPage : http://azufre.quimica.uniovi.es/
             (being reworked)






--
John McKelvey
545 Legacy Pointe Dr