From owner-chemistry@ccl.net Sun Aug 14 03:39:01 2011
From: "Patrick Bultinck patrick.bultinck,ugent.be" <owner-chemistry ~~ server.ccl.net>
To: CCL
Subject: CCL:G: fukuifunction
Message-Id: <-45257-110814033753-4124-JXALAelqc2I1ZAonizTumw ~~ server.ccl.net>
X-Original-From: Patrick Bultinck <patrick.bultinck+/-ugent.be>
Content-Type: multipart/alternative; boundary=Apple-Mail-1--7186543
Date: Sun, 14 Aug 2011 09:37:42 +0200
Mime-Version: 1.0 (Apple Message framework v1084)


Sent to CCL by: Patrick Bultinck [patrick.bultinck(~)ugent.be]

--Apple-Mail-1--7186543
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii

The Fukui function, be it atom condensed or not, is a far trickier thing =
than many people anticipate but your question is also not quite =
accurately formulated (pop does not really give you the density, it =
gives you atom condensed densities). Depending on what you want, pop =
will do or you will need to dig into the formatted checkpoint file.

Here are some thoughts:
- With DFT and Hartree-Fock the Fukui function can ALWAYS be negative in =
parts of space. This can even be proven mathematically so there is =
little argument to opt for methods that would get rid of such =
possibility. It's in the physics/maths/chemistry that it has to be =
there. There is no proof that it *must* turn negative at some point but =
it *can*. At correlated levels of theory you cannot give such a closed =
form proof but my experiments show that the Fukui function can turn =
negative easily at some points in space while retaining normalization. =
So it is something to live with.
- Any approximation as HOMO or LUMO density is actually a dirty trick =
inspired on frontier MO theory but is and remains a trick. I guess you =
can say that you get 98% of the Fukui function right with those =
densities but is chemistry not all about getting it really right ?
- Beware that is like with a density matrix: a negative eigenvalue tells =
you that density can turn negative, not that it will. Still, you cannot =
ignore the possibility. There is recent literature on a Fukui matrix =
that uses only basic linear algebra and shows when/where things can turn =
negative.

- Atom condensed Fukui functions can, as a consequence, also turn =
negative. There is also nothing mathematically wrong with that although =
it is a bit awkward. Situations where a Fukui function *has* to turn =
negative have been identified from the hardness kernel/matrix in =
electronegativity equalization.
- Beware that arbitrary decisions have sometimes been made by different =
authors on the way to compute atom condensed Fukui functions. There is =
much more than meets the eye here. A difference in charge (as you seem =
to use) is a tricky thing for all that is not Mulliken. This is the =
essence of a little understood (not the fault of the authors) sentence =
in the original Yang/Mortier condensed Fukui function paper; namely that =
differentiation with respect to electron number (N) and atomic =
projection/condensation should commute. They do not in NPA, only in MPA =
where you mix Hilbert and position space. In HPA they commute because of =
a completely arbitrary choice of weight function (that has been =
corrected since a few years).
- Even if you deal with the above issue, there is another issue on the =
extent to which you want your atomic "projectors" to be N-dependent. It =
can turn the Fukui function upside down, depending on your projectors =
(so except for MPA).

I think the best is to simply do a literature search on the negative =
character of Fukui functions to understand these issues and then to =
write a simple algebraic routine to get what you need from Gaussian. I =
will not send a list of my papers to CCL as this should not be a =
publicity channel for own work (...) but for those interested, they can =
mail me for a list of references in which we address issues like the =
above. That, in turn, will cite a lot of previous work from where you =
can start your adventures.

Patrick Bultinck
Ghent University
Belgium

On 13 Aug 2011, at 06:56, Mina Haghdadi mhaghdadi2-,-yahoo.co.uk wrote:

>=20
> Hi,
> I am working  on a molecule to calculate local fukui function but with =
some methods (MPA,NPA,HPA) the values be negative and I can't compare =
the reactivity of the nucleophilicity or electrophilicity of atomes in =
molecule. I have seen some articles that calculated the frontier =
molecular density that HOMO and LUMU is like f(+) and f(-) with positve =
values. Could you tell me what keyword to use the calculation molecular =
density in G03 I tried with keyeord :pop but  the gaussian does't work? =
and How can find them in the output results.
> I express my thanks for your time and generous help.
> Sincerely yours
> M.Haghdadi


--Apple-Mail-1--7186543
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=us-ascii

<html><head></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: after-white-space; =
"><div>The Fukui function, be it atom condensed or not, is a far =
trickier thing than many people anticipate but your question is also not =
quite accurately formulated (pop does not really give you the density, =
it gives you atom condensed densities). Depending on what you want, pop =
will do or you will need to dig into the formatted checkpoint =
file.</div><div><br></div><div>Here are some thoughts:</div><div>- With =
DFT and Hartree-Fock the Fukui function can ALWAYS be negative in parts =
of space. This can even be proven mathematically so there is little =
argument to opt for methods that would get rid of such possibility. It's =
in the physics/maths/chemistry that it has to be there. There is no =
proof that it *must* turn negative at some point but it *can*. At =
correlated levels of theory you cannot give such a closed form proof but =
my experiments show that the Fukui function can turn negative easily at =
some points in space while retaining normalization. So it is something =
to live with.</div><div>- Any approximation as HOMO or LUMO density is =
actually a dirty trick inspired on frontier MO theory but is and remains =
a trick. I guess you can say that you get 98% of the Fukui function =
right with those densities but is chemistry not all about getting it =
really right ?</div><div>- Beware that is like with a density matrix: a =
negative eigenvalue tells you that density can turn negative, not that =
it will. Still, you cannot ignore the possibility. There is recent =
literature on a Fukui matrix that uses only basic linear algebra and =
shows when/where things can turn negative.</div><div><br></div><div>- =
Atom condensed Fukui functions can, as a consequence, also turn =
negative. There is also nothing mathematically wrong with that although =
it is a bit awkward. Situations where a Fukui function *has* to turn =
negative have been identified from the hardness kernel/matrix in =
electronegativity equalization.</div><div>- Beware that arbitrary =
decisions have sometimes been made by different authors on the way to =
compute atom condensed Fukui functions. There is much more than meets =
the eye here. A difference in charge (as you seem to use) is a tricky =
thing for all that is not Mulliken. This is the essence of a little =
understood (not the fault of the authors) sentence in the original =
Yang/Mortier condensed Fukui function paper; namely that differentiation =
with respect to electron number (N) and atomic projection/condensation =
should commute. They do not in NPA, only in MPA where you mix Hilbert =
and position space. In HPA they commute because of a completely =
arbitrary choice of weight function (that has been corrected since a few =
years).</div><div>- Even if you deal with the above issue, there is =
another issue on the extent to which you want your atomic "projectors" =
to be N-dependent. It can turn the Fukui function upside down, depending =
on your projectors (so except for MPA).</div><div><br></div><div>I think =
the best is to simply do a literature search on the negative character =
of Fukui functions to understand these issues and then to write a simple =
algebraic routine to get what you need from Gaussian. I will not send a =
list of my papers to CCL as this should not be a publicity channel for =
own work (...) but for those interested, they can mail me for a list of =
references in which we address issues like the above. That, in turn, =
will cite a lot of previous work from where you can start your =
adventures.</div><div><br></div><div>Patrick Bultinck</div><div>Ghent =
University</div><div>Belgium</div><br><div><div>On 13 Aug 2011, at =
06:56, Mina Haghdadi mhaghdadi2-,<a =
href=3D"http://-yahoo.co.uk">-yahoo.co.uk</a> wrote:</div><br =
class=3D"Apple-interchange-newline"><blockquote type=3D"cite"><table =
cellspacing=3D"0" cellpadding=3D"0" border=3D"0"><tbody><tr><td =
valign=3D"top" style=3D"font: inherit;"><table border=3D"0" =
cellspacing=3D"0" cellpadding=3D"0">
<tbody>
<tr>
<td valign=3D"top">
<div>Hi,</div>
<div>I am working&nbsp; on a molecule&nbsp;to calculate local&nbsp;fukui =
function but with some methods (MPA,NPA,HPA)&nbsp;the values be negative =
and I can't compare the reactivity of the&nbsp;nucleophilicity or =
electrophilicity of atomes in molecule. I have seen some articles that =
calculated the frontier molecular density that&nbsp;HOMO and LUMU is =
like f(+) and f(-)&nbsp;with positve values.&nbsp;Could you tell me what =
keyword to use the calculation molecular density in G03 I tried with =
keyeord :pop but &nbsp;the gaussian does't work? and How can find them =
in the output results.</div>
<div>I express my thanks for your time and generous help.</div>
<div>Sincerely yours</div>
=
<div>M.Haghdadi</div></td></tr></tbody></table></td></tr></tbody></table><=
/blockquote></div><br></body></html>=

--Apple-Mail-1--7186543--


From owner-chemistry@ccl.net Sun Aug 14 11:14:01 2011
From: "Dario Fernando Coral ferchocoralg16!A!gmail.com" <owner-chemistry()server.ccl.net>
To: CCL
Subject: CCL: single point failure
Message-Id: <-45258-110814111126-23667-UGt1podhaaXxBSBoNzkyUQ()server.ccl.net>
X-Original-From: "Dario Fernando Coral" <ferchocoralg16/./gmail.com>
Date: Sun, 14 Aug 2011 11:11:23 -0400


Sent to CCL by: "Dario Fernando Coral" [ferchocoralg16 _ gmail.com]
I optimized a complex dissociation with the level of theory ub3lyp/6-31g(d). I have obtained the frequencies with the same method and indicate that indeed is a minimum. The problem is the calculation of single point. When I try to do with the level of theory ub3lyp/6-311+G(3df,2p), this calculation dies. The output file indicates that the error occurs during execution of the link 502, ie, Iteratively solves the SCF equations (UHF & ROHF conven., all direct Methods, SCRF).

Someone knows why this happens?

I must say that I have run this calculation on three different machines.

Dario Fernando Coral Obando
Chemistry student
Universidad de Nario - Colombia


From owner-chemistry@ccl.net Sun Aug 14 12:03:01 2011
From: "eurisco1[-]pochta.ru" <owner-chemistry%server.ccl.net>
To: CCL
Subject: CCL: single point failure
Message-Id: <-45259-110814115839-28039-PPjiwcUKLGxX5BD/pmgNrQ%server.ccl.net>
X-Original-From: <eurisco1]=[pochta.ru>
Content-Transfer-Encoding: 8bit
Content-Type: text/plain;
	format=flowed;
	charset="koi8-r";
	reply-type=original
Date: Sun, 14 Aug 2011 19:58:27 +0400
MIME-Version: 1.0


Sent to CCL by: [eurisco1[#]pochta.ru]
Dear Dario Fernando Coral,

Maybe SCF(XQC,vshift=150,MaxCycle=512) will help.

Sincerely,
Ol Ga


-----Исходное сообщение----- 
> From: Dario Fernando Coral ferchocoralg16!A!gmail.com
Sent: Sunday, August 14, 2011 7:11 PM
To: Ga, Ol 
Subject: CCL: single point failure


Sent to CCL by: "Dario Fernando Coral" [ferchocoralg16 _ gmail.com]
I optimized a complex dissociation with the level of theory ub3lyp/6-31g(d). 
I have obtained the frequencies with the same method and indicate that 
indeed is a minimum. The problem is the calculation of single point. When I 
try to do with the level of theory ub3lyp/6-311+G(3df,2p), this calculation 
dies. The output file indicates that the error occurs during execution of 
the link 502, ie, Iteratively solves the SCF equations (UHF & ROHF conven., 
all direct Methods, SCRF).

Someone knows why this happens?

I must say that I have run this calculation on three different machines.

Dario Fernando Coral Obando
Chemistry student
Universidad de Nario - Colombiahttp://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txt