From owner-chemistry@ccl.net Mon Oct 10 04:33:00 2005 From: "Noel O Boyle noel.oboyle2-#-mail.dcu.ie" To: CCL Subject: CCL: Summary: What keeps molecule adsorbed? Message-Id: <-29526-051010042616-27519-blmwIXlZWwMdPzxhHfSJ2w[#]server.ccl.net> X-Original-From: "Noel O'Boyle" Content-Transfer-Encoding: 7bit Content-Type: text/plain Date: Mon, 10 Oct 2005 09:26:01 +0100 Mime-Version: 1.0 Sent to CCL by: "Noel O'Boyle" [noel.oboyle2]^[mail.dcu.ie] Thanks to all those who answered my rather naive question. Here is the original question plus the answers received (not all were posted on the mailing list) Question: --------- I have been trying to explain the results of a DFT study (using G03) where a molecule adsorbs onto a gold cluster through its nitrogen atom. I created an isosurface of the change in the electron density associated with the adsorption (by calculating the electron densities of the molecule and the gold cluster separately at the geometry of the adsorbed system). This shows that electron density *decreases* in the area between the adsorbate and the surface. This is the opposite of what I expected. A paper by Bilic, Reimers, Hush and Hafner (JCP, 2002, 116, 8981) shows the same effect for ammonia on an infinite surface of gold(111). It shows the same result, and says that this is evidence that covalent bonding effects are not so important (they also have other evidence). It seems that I do not understand bonding very well - what keeps the molecule stuck to the surface, if it isn't a shared electron or two? Regards, Noel Answers: -------- Sent to CCL by: Mariusz Sterzel [msterzel^_^buffalo.edu] Dear Noel, Following Ruedenberg, the simple answer is electrons kinetic energy. For more details look at K. Ruedenberg, Reviews of Modern Physics, 34(2):326-376, (1962) and M.S. Gordon, J.H.Jensen, Theoretical Chemistry Accounts, 103:248-251, (2000) Mariusz -- Mariusz Sterzel Sent to CCL by: Joseph Han [jhh3851/a\yahoo.com] If the adsorption energy is small (< 5 kcal/mol), it might also be one of the following: (1) Basis set superposition error. You might need to do a counterpoise correction to address this. (2) Dipole-dipole interaction (e.g. van der Waals) where you induce a dipole in the gold cluster leading to weak electrostaic attraction. Joseph Sent to CCL by: Per-Ola Norrby [pon;;kemi.dtu.dk] To continue on this thread, there are also some important adsorption forces that are NOT captured by standard DFT. The simplest would be the London dispersion (induced dipole-induced dipole). There are also some recent, interesting results by Schwarz and coworkers showing that binding energies change drastically if the gold is treated relativistically. I don't have the reference handy, but I believe I saw numbers like standard relativistic ECP: no binding; fully relativistic calculation: 20 kcal/mol binding (something like that). Of course fully relativistic calculations are out of the question for more than a few atoms. And if your question really was "what constitutes binding in my calculation" as opposed to binding in reality, then both of the above effects are irrelevant, since they wouldn't appear in your calculations. Regards, Per-Ola Remember, there are other forms of bonding besides covalent which is probably the only form you will see via density differencing. You can have ionic, Van der Waals/ dispersion, etc. Have you looked at the charge transfer between the gold and the nitrogen? What about polarization effects? Try using the AOMix program to analyze your results. It may give you some insights. Also the NEDA portion of the NBO program may help. If you can, do a MP2 or MP4 calculation or use a DFT functional (like x3lyp or PW6B95, MPW1B95, B98, B97-1, TPSS1KCIS, PWB6K, MPWB1K, BB1K, and MPW1K) that do a much better job - especially for dispersion forces. Any other DFT functional (e.g. b3lyp) will probably not be a very good choice. Jim I don't know anything about the details of your system, but the fact that it is adsorption on a metal surface, would seem to suggest the possibility of an "image-charge" role. Could the reduction of "bond-charge" correspond to a change in the effective net charge "above the surface"? I would think characteristic vibration frequencies might provide key evidence as to whether the bond becomes more akin to the image-charge type of bond compared to a genuine covalent bond. Again, I don't know whether that is the case or not, but it's just a thought. Joe Harrison Dept. of Physics U. of Alabama at Birmingham If the adsorption energy is small (< 5 kcal/mol), it might also be one of the following: (1) Basis set superposition error. You might need to do a counterpoise correction to address this. (2) Dipole-dipole interaction (e.g. van der Waals) where you induce a dipole in the gold cluster leading to weak electrostaic attraction. Joseph Dear Dr. O'Boyle! It might be that the density difference that you calculated does not correspond exclusively to the "interaction" between the cluster and the system. The ligating lone pairs of the system may displace the (supposedly easily polarizable) electrons of the cluster. There would be several aspects to such a displacement: (i) from the atoms by which ligation takes place; (ii) to some other location, perhaps other gold atoms or (in case a charge transfer takes place upon coordination) to the system being coordinated. In effect, the process of adsorption may roughly, and, of course, arbitrarily, be divided into two parts, (i) the promotion (deformation) of the cluster, and (ii) the interaction between the promoted cluster and the adsorbate. As an example, take the Fe2+ d6 ion coordinating six strong ligands. I imagine the ground state of the ion is not the singlet with (t2g)6 configuration, but some higher spin, with perhaps 2 electrons in the two (eg) orbitals. But if the ligands are indeed strong, the 6 electrons would pair up, and the resulting octahedral species would have (qualitatively, ignoring back-donation) 0 electrons in the (eg) orbitals. So if you calculate the density difference using the ground state of Fe2+, the metal-ligands sigma system would appear to have lost 2 electrons upon the ligation. Yet this would not mean that there is no chemical bonding in the sigma system. Most chemists would probably try to use the "promoted" singlet (t2g)6 configuration of Fe2+ right away, being aware of the promotion side of the complexation. In your case, once again, only hypothetically, an electon may be removed > from the 6s orbital of the Au atom(s) that coordinate the adsorbate. The high density of electronic states in the metal cluster may mean that no spin pairing is necessary to promote the cluster into the configuration that is used for adsorption. The "promoted" state would probably be impossible to relate to any true, adiabatic eigenstate of the bare cluster. Some diabatic/valence bond states/wave functions may be necessary. Looking at the density *increase* regions of the density difference may reveal some aspects of the promotion. We put these ideas, at a very primitive level of relation to previous work, into a paper: Khoroshun, D. V.; Musaev, D. G.; Morokuma, K. Mol. Phys. 2002, 100, 523. Once again, the connection to your situation is only hypothetical to me, because there might be other important reasons for the depletion of the density in the binding region. Sincerely, Dmitry Khoroshun dima++mpi-muelheim.mpg.de From owner-chemistry@ccl.net Mon Oct 10 10:03:00 2005 From: "James Kubicki kubicki^^geosc.psu.edu" To: CCL Subject: CCL: Summary: What keeps molecule adsorbed? Message-Id: <-29527-051010100147-10601-/jgHoIzzWk9UoKMjo01e8g:_:server.ccl.net> X-Original-From: James Kubicki Content-Type: multipart/alternative; boundary="=====================_2906328==_.ALT" Date: Mon, 10 Oct 2005 09:10:22 -0400 Mime-Version: 1.0 Sent to CCL by: James Kubicki [kubicki(0)geosc.psu.edu] --=====================_2906328==_.ALT Content-Type: text/plain; charset="us-ascii"; format=flowed I find it interesting that no one mentioned the overall Gibbs free energy change of the reaction. If you have Gold surface/solvent + sorbate in solvent --> Gold surface/sorbate + solvent/solvent (where the "/" signifies interaction between the two components), then even if the gold surface/sorbate interaction delta-G is positive (i.e., not favorable), you can have adsorption if this positive delta-G is overcompensated by a more negative delta-G for releasing the solvent molecules from the surface and allowing them better interaction with other solvent molecules. The solvophilicity of the sorbate also plays a role. This is typically a delta-S term for water where releasing H2O molecules from their relatively rigid positions on a surface can create a significant amount of positive entropy for the system. Modelers tend to only focus on the energy of the sorbate/sorbent interaction which is only half the story. What needs to be calculated is delta-G for all components. I think the neglect of the second half of the reaction comes > from the past history of modeling adsorption in vacuum. At 04:26 AM 10/10/2005, you wrote: >Sent to CCL by: "Noel O'Boyle" [noel.oboyle2]^[mail.dcu.ie] >Thanks to all those who answered my rather naive question. Here is the >original question plus the answers received (not all were posted on the >mailing list) > >Question: >--------- >I have been trying to explain the results of a DFT study (using G03) >where a molecule adsorbs onto a gold cluster through its nitrogen atom. > >I created an isosurface of the change in the electron density associated >with the adsorption (by calculating the electron densities of the >molecule and the gold cluster separately at the geometry of the adsorbed >system). This shows that electron density *decreases* in the area >between the adsorbate and the surface. This is the opposite of what I >expected. > >A paper by Bilic, Reimers, Hush and Hafner (JCP, 2002, 116, 8981) shows >the same effect for ammonia on an infinite surface of gold(111). It >shows the same result, and says that this is evidence that covalent >bonding effects are not so important (they also have other evidence). > >It seems that I do not understand bonding very well - what keeps the >molecule stuck to the surface, if it isn't a shared electron or two? > >Regards, >Noel > >Answers: >-------- >Sent to CCL by: Mariusz Sterzel [msterzel^_^buffalo.edu] >Dear Noel, > >Following Ruedenberg, the simple answer is electrons >kinetic energy. For more details look at K. Ruedenberg, Reviews of >Modern >Physics, 34(2):326-376, (1962) and M.S. Gordon, J.H.Jensen, Theoretical >Chemistry Accounts, 103:248-251, (2000) > > >Mariusz > >-- >Mariusz Sterzel > >Sent to CCL by: Joseph Han [jhh3851/a\yahoo.com] >If the adsorption energy is small (< 5 kcal/mol), it might also be one >of the >following: > >(1) Basis set superposition error. You might need to do a counterpoise >correction to address this. > >(2) Dipole-dipole interaction (e.g. van der Waals) where you induce a >dipole >in the gold cluster leading to weak electrostaic attraction. > >Joseph > >Sent to CCL by: Per-Ola Norrby [pon;;kemi.dtu.dk] > To continue on this thread, there are also some important >adsorption forces that are NOT captured by standard DFT. The >simplest would be the London dispersion (induced dipole-induced >dipole). There are also some recent, interesting results by Schwarz >and coworkers showing that binding energies change drastically if the >gold is treated relativistically. I don't have the reference handy, >but I believe I saw numbers like standard relativistic ECP: no >binding; fully relativistic calculation: 20 kcal/mol binding >(something like that). Of course fully relativistic calculations are >out of the question for more than a few atoms. And if your question >really was "what constitutes binding in my calculation" as opposed to >binding in reality, then both of the above effects are irrelevant, >since they wouldn't appear in your calculations. > > Regards, > > Per-Ola > >Remember, there are other forms of bonding besides covalent which is >probably the only form you will see via density differencing. You can >have >ionic, Van der Waals/ dispersion, etc. > >Have you looked at the charge transfer between the gold and the >nitrogen? >What about polarization effects? Try using the AOMix program to analyze >your results. It may give you some insights. Also the NEDA portion of >the >NBO program may help. > >If you can, do a MP2 or MP4 calculation or use a DFT functional (like >x3lyp >or PW6B95, MPW1B95, >B98, B97-1, TPSS1KCIS, PWB6K, MPWB1K, BB1K, and MPW1K) that do a much >better >job - especially for dispersion forces. Any other DFT functional (e.g. >b3lyp) will probably not be a very good choice. > >Jim > >I don't know anything about the details of your system, but the fact >that >it is adsorption on a metal surface, would seem to suggest the >possibility >of an "image-charge" role. Could the reduction of "bond-charge" >correspond to a change in the effective net charge "above the surface"? >I >would think characteristic vibration frequencies might provide key >evidence >as to whether the bond becomes more akin to the image-charge type of >bond >compared to a genuine covalent bond. Again, I don't know whether that >is >the case or not, but it's just a thought. > >Joe Harrison >Dept. of Physics >U. of Alabama at Birmingham > >If the adsorption energy is small (< 5 kcal/mol), it might also be one >of the >following: > >(1) Basis set superposition error. You might need to do a counterpoise >correction to address this. > >(2) Dipole-dipole interaction (e.g. van der Waals) where you induce a >dipole >in the gold cluster leading to weak electrostaic attraction. > >Joseph > >Dear Dr. O'Boyle! > >It might be that the density difference that you calculated does not >correspond exclusively to the "interaction" between the cluster and the >system. > >The ligating lone pairs of the system may displace the (supposedly >easily >polarizable) electrons of the cluster. There would be several aspects to >such a displacement: (i) from the atoms by which ligation takes place; >(ii) to some other location, perhaps other gold atoms or (in case a >charge >transfer takes place upon coordination) to the system being coordinated. > >In effect, the process of adsorption may roughly, and, of course, >arbitrarily, be divided into two parts, (i) the promotion (deformation) >of >the cluster, and (ii) the interaction between the promoted cluster and >the >adsorbate. > >As an example, take the Fe2+ d6 ion coordinating six strong ligands. I >imagine the ground state of the ion is not the singlet with (t2g)6 >configuration, but some higher spin, with perhaps 2 electrons in the two >(eg) orbitals. But if the ligands are indeed strong, the 6 electrons >would >pair up, and the resulting octahedral species would have (qualitatively, >ignoring back-donation) 0 electrons in the (eg) orbitals. So if you >calculate the density difference using the ground state of Fe2+, the >metal-ligands sigma system would appear to have lost 2 electrons upon >the >ligation. Yet this would not mean that there is no chemical bonding in >the >sigma system. Most chemists would probably try to use the "promoted" >singlet (t2g)6 configuration of Fe2+ right away, being aware of the >promotion side of the complexation. > >In your case, once again, only hypothetically, an electon may be removed > > from the 6s orbital of the Au atom(s) that coordinate the adsorbate. The >high density of electronic states in the metal cluster may mean that no >spin pairing is necessary to promote the cluster into the configuration >that is used for adsorption. > >The "promoted" state would probably be impossible to relate to any true, >adiabatic eigenstate of the bare cluster. Some diabatic/valence bond >states/wave functions may be necessary. > >Looking at the density *increase* regions of the density difference may >reveal some aspects of the promotion. > >We put these ideas, at a very primitive level of relation to previous >work, into a paper: > >Khoroshun, D. V.; Musaev, D. G.; Morokuma, K. Mol. Phys. 2002, 100, 523. > >Once again, the connection to your situation is only hypothetical to me, >because there might be other important reasons for the depletion of the >density in the binding region. > >Sincerely, >Dmitry Khoroshun >dima(~)mpi-muelheim.mpg.deJames Kubicki 308 Deike Bldg. Dept. of Geosciences The Pennsylvania State University University Park, PA 16802 814-865-3951 814-574-7379 (Cell) 814-863-7823 (Fax) --=====================_2906328==_.ALT Content-Type: text/html; charset="us-ascii" I find it interesting that no one mentioned the overall Gibbs free energy
change of the reaction.
If you have
Gold surface/solvent + sorbate in solvent --> Gold surface/sorbate + solvent/solvent
(where the "/" signifies interaction between the two components), then even if
the gold surface/sorbate interaction delta-G is positive (i.e., not favorable), you can have
adsorption if this positive delta-G is overcompensated by a more negative delta-G
for releasing the solvent molecules from the surface and allowing them better
interaction with other solvent molecules. The solvophilicity of the sorbate also plays
a role.  This is typically a delta-S term for water
where releasing H2O molecules from their relatively rigid positions on a surface
can create a significant amount of positive entropy for the system.
Modelers tend to only focus on the energy of the sorbate/sorbent interaction
which is only half the story.  What needs to be calculated is delta-G for
all components.  I think the neglect of the second half of the reaction comes
> from the past history of modeling adsorption in vacuum.

At 04:26 AM 10/10/2005, you wrote:

Sent to CCL by: "Noel O'Boyle" [noel.oboyle2]^[mail.dcu.ie]
Thanks to all those who answered my rather naive question. Here is the
original question plus the answers received (not all were posted on the
mailing list)

Question:
---------
I have been trying to explain the results of a DFT study (using G03)
where a molecule adsorbs onto a gold cluster through its nitrogen atom.

I created an isosurface of the change in the electron density associated
with the adsorption (by calculating the electron densities of the
molecule and the gold cluster separately at the geometry of the adsorbed
system). This shows that electron density *decreases* in the area
between the adsorbate and the surface. This is the opposite of what I
expected.

A paper by Bilic, Reimers, Hush and Hafner (JCP, 2002, 116, 8981) shows
the same effect for ammonia on an infinite surface of gold(111). It
shows the same result, and says that this is evidence that covalent
bonding effects are not so important (they also have other evidence).

It seems that I do not understand bonding very well - what keeps the
molecule stuck to the surface, if it isn't a shared electron or two?

Regards,
Noel

Answers:
--------
Sent to CCL by: Mariusz Sterzel [msterzel^_^buffalo.edu]
Dear Noel,

Following Ruedenberg, the simple answer is electrons
kinetic energy. For more details look at K. Ruedenberg, Reviews of
Modern
Physics, 34(2):326-376, (1962) and M.S. Gordon, J.H.Jensen, Theoretical
Chemistry Accounts, 103:248-251, (2000)


Mariusz

--
Mariusz Sterzel

Sent to CCL by: Joseph Han [jhh3851/a\yahoo.com]
If the adsorption energy is small (< 5 kcal/mol), it might also be one
of the
following:

(1)  Basis set superposition error.  You might need to do a counterpoise
correction to address this.

(2)  Dipole-dipole interaction (e.g. van der Waals) where you induce a
dipole
in the gold cluster leading to weak electrostaic attraction.

Joseph

Sent to CCL by: Per-Ola Norrby [pon;;kemi.dtu.dk]
        To continue on this thread, there are also some important
adsorption forces that are NOT captured by standard DFT.  The
simplest would be the London dispersion (induced dipole-induced
dipole).  There are also some recent, interesting results by Schwarz
and coworkers showing that binding energies change drastically if the
gold is treated relativistically.  I don't have the reference handy,
but I believe I saw numbers like standard relativistic ECP: no
binding; fully relativistic calculation: 20 kcal/mol binding
(something like that).  Of course fully relativistic calculations are
out of the question for more than a few atoms.  And if your question
really was "what constitutes binding in my calculation" as opposed to
binding in reality, then both of the above effects are irrelevant,
since they wouldn't appear in your calculations.

        Regards,

        Per-Ola

Remember, there are other forms of bonding besides covalent which is
probably the only form you will see via density differencing.  You can
have
ionic, Van der Waals/ dispersion, etc.

Have you looked at the charge transfer between the gold and the
nitrogen?
What about polarization effects?  Try using the AOMix program to analyze
your results.  It may give you some insights.  Also the NEDA portion of
the
NBO program may help.

If you can, do a MP2 or MP4 calculation or use a DFT functional (like
x3lyp
or PW6B95, MPW1B95,
B98, B97-1, TPSS1KCIS, PWB6K, MPWB1K, BB1K, and MPW1K) that do a much
better
job - especially for dispersion forces.  Any other DFT functional (e.g.
b3lyp) will probably not be a very good choice.

Jim

I don't know anything about the details of your system, but the fact
that
it is adsorption on a metal surface, would seem to suggest the
possibility 
of an "image-charge" role.  Could the reduction of "bond-charge"
correspond to a change in the effective net charge "above the surface"?
I
would think characteristic vibration frequencies might provide key
evidence
as to whether the bond becomes more akin to the image-charge type of
bond
compared to a genuine covalent bond.  Again, I don't know whether that
is
the case or not, but it's just a thought.

Joe Harrison
Dept. of Physics
U. of Alabama at Birmingham

If the adsorption energy is small (< 5 kcal/mol), it might also be one
of the
following:

(1)  Basis set superposition error.  You might need to do a counterpoise
correction to address this.

(2)  Dipole-dipole interaction (e.g. van der Waals) where you induce a
dipole
in the gold cluster leading to weak electrostaic attraction.

Joseph

Dear Dr. O'Boyle!

It might be that the density difference that you calculated does not
correspond exclusively to the "interaction" between the cluster and the
system.

The ligating lone pairs of the system may displace the (supposedly
easily
polarizable) electrons of the cluster. There would be several aspects to
such a displacement: (i) from the atoms by which ligation takes place;
(ii) to some other location, perhaps other gold atoms or (in case a
charge
transfer takes place upon coordination) to the system being coordinated.

In effect, the process of adsorption may roughly, and, of course,
arbitrarily, be divided into two parts, (i) the promotion (deformation)
of
the cluster, and (ii) the interaction between the promoted cluster and
the
adsorbate.

As an example, take the Fe2+ d6 ion coordinating six strong ligands. I
imagine the ground state of the ion is not the singlet with (t2g)6
configuration, but some higher spin, with perhaps 2 electrons in the two
(eg) orbitals. But if the ligands are indeed strong, the 6 electrons
would
pair up, and the resulting octahedral species would have (qualitatively,
ignoring back-donation) 0 electrons in the (eg) orbitals. So if you
calculate the density difference using the ground state of Fe2+, the
metal-ligands sigma system would appear to have lost 2 electrons upon
the
ligation. Yet this would not mean that there is no chemical bonding in
the
sigma system. Most chemists would probably try to use the "promoted"
singlet (t2g)6 configuration of Fe2+ right away, being aware of the
promotion side of the complexation.

In your case, once again, only hypothetically, an electon may be removed
> from the 6s orbital of the Au atom(s) that coordinate the adsorbate. The
high density of electronic states in the metal cluster may mean that no
spin pairing is necessary to promote the cluster into the configuration
that is used for adsorption.

The "promoted" state would probably be impossible to relate to any true,
adiabatic eigenstate of the bare cluster. Some diabatic/valence bond
states/wave functions may be necessary.

Looking at the density *increase* regions of the density difference may
reveal some aspects of the promotion.

We put these ideas, at a very primitive level of relation to previous
work, into a paper:

Khoroshun, D. V.; Musaev, D. G.; Morokuma, K. Mol. Phys. 2002, 100, 523.

Once again, the connection to your situation is only hypothetical to me,
because there might be other important reasons for the depletion of the
density in the binding region.

Sincerely,
Dmitry Khoroshun
dima(~)mpi-muelheim.mpg.de


      http://www.ccl.net/cgi-bin/ccl/send_ccl_message
      http://www.ccl.net/cgi-bin/ccl/send_ccl_message
      http://www.ccl.net/chemistry/sub_unsub.shtml

Before posting check wait time for next message at: http://www.ccl.net

Job advertisements: http://www.ccl.net/jobs
      http://www.ccl.net/spammers.txt

James Kubicki
308 Deike Bldg.
Dept. of Geosciences
The Pennsylvania State University
University Park, PA 16802
814-865-3951
814-574-7379 (Cell)
814-863-7823 (Fax)
 --=====================_2906328==_.ALT-- From owner-chemistry@ccl.net Mon Oct 10 10:38:00 2005 From: "b wafaa wafaab2*_*yahoo.fr" To: CCL Subject: CCL: Condensed Fukui Functions Message-Id: <-29528-051010095142-9436-1R4IqcLSQ7t8nU1vHTpwLA++server.ccl.net> X-Original-From: b wafaa Content-Transfer-Encoding: 8bit Content-Type: multipart/alternative; boundary="0-1144409699-1128948697=:17669" Date: Mon, 10 Oct 2005 14:51:37 +0200 (CEST) MIME-Version: 1.0 Sent to CCL by: b wafaa [wafaab2=yahoo.fr] --0-1144409699-1128948697=:17669 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Dear CCLers, We have some problems in the prediction of reactivity when we calculate condensed Fukui functions with the finite difference approximation using N, N-1 and N+1 systems. In your opinion, what is the best method for the calculation of these indexes ? Which population analysis do you suggest ? Sincerely --------------------------------- Appel audio GRATUIT partout dans le monde avec le nouveau Yahoo! Messenger Téléchargez le ici ! --0-1144409699-1128948697=:17669 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit

Dear CCLers,
We have some problems in the prediction of reactivity when we calculate condensed Fukui functions with the finite difference approximation using N, N-1 and N+1 systems.
In your opinion, what is the best method for the calculation of these indexes ?
Which population analysis do you suggest ?
Sincerely

Appel audio GRATUIT partout dans le monde avec le nouveau Yahoo! Messenger
Téléchargez le ici ! --0-1144409699-1128948697=:17669-- From owner-chemistry@ccl.net Mon Oct 10 11:25:00 2005 From: "Noel O Boyle noel.oboyle2+/-mail.dcu.ie" To: CCL Subject: CCL: Condensed Fukui Functions Message-Id: <-29529-051010112025-12170-RICuBMjlUAVOJVotSkrg+w%x%server.ccl.net> X-Original-From: "Noel O'Boyle" Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 Date: Mon, 10 Oct 2005 16:20:11 +0100 Mime-Version: 1.0 Sent to CCL by: "Noel O'Boyle" [noel.oboyle2:+:mail.dcu.ie] I think that Hirshfeld population analysis has been shown to be one of the best methods for calculating condensed Fukui functions (at least compared to mulliken population analysis and its close relatives). Noel On Mon, 2005-10-10 at 14:51 +0200, b wafaa wafaab2*_*yahoo.fr wrote: > Dear CCLers, > We have some problems in the prediction of reactivity when we > calculate condensed Fukui functions with the finite difference > approximation using N, N-1 and N+1 systems. > In your opinion, what is the best method for the calculation of these > indexes ? > Which population analysis do you suggest ? > Sincerely > > > ______________________________________________________________________ > Appel audio GRATUIT partout dans le monde avec le nouveau Yahoo! > Messenger > Téléchargez le ici ! From owner-chemistry@ccl.net Mon Oct 10 12:39:01 2005 From: "James Kirkpatrick james.kirkpatrick .. imperial.ac.uk" To: CCL Subject: CCL: constrained geometry optmisation of molecular complexes Message-Id: <-29530-051010122928-16578-eA8LMfjt548O/LFSy9iH9Q=-=server.ccl.net> X-Original-From: James Kirkpatrick Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Mon, 10 Oct 2005 17:29:31 +0100 MIME-Version: 1.0 Sent to CCL by: James Kirkpatrick [james.kirkpatrick__imperial.ac.uk] Hi CClers, I wish to perform geometry calculations on two of molecules, each molecule contains rougly 20 heavy atoms. So far as I understand, HF methods, semi-empirical methods and most DFT functionals will not describe van-der-Waals forces (responsible for a good part of the non-binding interaction?) correctly. In order to obtain a half sensible geometry therefore I presume that it may be a good idea to constrain certain geometrical quantities so as to force the two molecules to, e..g, be at a certain distance or to be coplanar. Once I run my geometry optimisation, I wish to check that I am at a minimum by calculating the Hessian of the system. I seem to obtain imaginary frequencies, suggesting that the geometry is not at an actual minimum. This could be because of the geometric parameters that I have constrained to particular value simply do not bring the molecule at a minimum. Does this mean I should ditch those parameters and try different ones? Does anyone have any suggestions on how to perform optimizations on complexes of molecules? Is using constrained parameters the best possible strategy? Or is using HF, DFT simply not sufficient? I would be most grateful for being pointed in the correct direction! Thanks for your attention and advice James -- James Kirkpatrick ------------------------------------------- Centre for Electronic Materials and devices Imperial College ------------------------------------------- 020 759 47519 From owner-chemistry@ccl.net Mon Oct 10 14:32:00 2005 From: "Cory Pye cpye- -crux.smu.ca" To: CCL Subject: CCL: constrained geometry optmisation of molecular complexes Message-Id: <-29531-051010142825-19454-v/JzkyeJAYaPf2dx2PS0lA---server.ccl.net> X-Original-From: Cory Pye Content-Type: TEXT/PLAIN; charset=US-ASCII Date: Mon, 10 Oct 2005 15:28:16 -0300 (ADT) MIME-Version: 1.0 Sent to CCL by: Cory Pye [cpye:+:crux.smu.ca] On Mon, 10 Oct 2005, James Kirkpatrick james.kirkpatrick .. imperial.ac.uk wrote: > Sent to CCL by: James Kirkpatrick [james.kirkpatrick__imperial.ac.uk] > Hi CClers, > > I wish to perform geometry calculations on two of molecules, each > molecule contains rougly 20 heavy atoms. So far as I understand, HF > methods, semi-empirical methods and most DFT functionals will not > describe van-der-Waals forces (responsible for a good part of the > non-binding interaction?) correctly. These techniques miss out on the dispersion interaction. MP2 might be a better bet for these systems, but unfortunately your molecules may be too big for such a calculation. > > In order to obtain a half sensible geometry therefore I presume that it > may be a good idea to constrain certain geometrical quantities so as to > force the two molecules to, e..g, be at a certain distance or to be > coplanar. Once I run my geometry optimisation, I wish to check that I am > at a minimum by calculating the Hessian of the system. I seem to obtain > imaginary frequencies, suggesting that the geometry is not at an actual > minimum. This could be because of the geometric parameters that I have > constrained to particular value simply do not bring the molecule at a > minimum. Does this mean I should ditch those parameters and try > different ones? As soon as you introduce a constraint that is not related to the symmetry of the system, the technique of using the Hessian is no longer valid and you may indeed get imaginary frequencies. If the fundamental interaction is not described by your Hamiltonian, then the Hessian (second derivative wrt nuclear displacements) will likewise not be described. > > Does anyone have any suggestions on how to perform optimizations on > complexes of molecules? Is using constrained parameters the best > possible strategy? Or is using HF, DFT simply not sufficient? I would be > most grateful for being pointed in the correct direction! > If you have certain key parts of your molecule that are "doing" the interaction, you might wish to try the ONIOM approach, say putting the key parts at MP2/6-31+G* or better and describing the rest with your favorite level say HF or B3LYP/6-31G*. The ONIOM technique can do optimizations and frequency calculations, and you wouldn't need to use constraints. > Thanks for your attention and advice > James > > ************* ! Dr. Cory C. Pye ***************** ! Associate Professor *** ** ** ** ! Theoretical and Computational Chemistry ** * **** ! Department of Chemistry, Saint Mary's University ** * * ! 923 Robie Street, Halifax, NS B3H 3C3 ** * * ! cpye/./crux.stmarys.ca http://apwww.stmarys.ca/~cpye *** * * ** ! Ph: (902)-420-5654 FAX:(902)-496-8104 ***************** ! ************* ! Les Hartree-Focks (Apologies to Montreal Canadien Fans) From owner-chemistry@ccl.net Mon Oct 10 15:11:00 2005 From: "Jim Kress ccl_nospam[-]kressworks.com" To: CCL Subject: CCL: constrained geometry optmisation of molecular complexes Message-Id: <-29532-051010144225-27964-zR8JNX55YSzZ/Ct6xuCJMg~!~server.ccl.net> X-Original-From: "Jim Kress" Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii" Date: Mon, 10 Oct 2005 14:42:11 -0400 MIME-Version: 1.0 Sent to CCL by: "Jim Kress" [ccl_nospam~~kressworks.com] X3lyp does a reasonable job of including dispersion forces, though some on the list will disagree. A good overview is presented in: Design of Density Functionals That Are Broadly Accurate for Thermochemistry, Thermochemical Kinetics, and Nonbonded Interactions, Yan Zhao and Donald G. Truhlar*, J. Phys. Chem. A 2005, 109, 5656-5667 Jim > -----Original Message----- > From: James Kirkpatrick james.kirkpatrick .. imperial.ac.uk > [mailto:owner-chemistry**ccl.net] > Sent: Monday, October 10, 2005 12:30 PM > To: Kress, Jim > Subject: CCL: constrained geometry optmisation of molecular complexes > > > Sent to CCL by: James Kirkpatrick [james.kirkpatrick__imperial.ac.uk] > Hi CClers, > > I wish to perform geometry calculations on two of molecules, > each molecule contains rougly 20 heavy atoms. So far as I > understand, HF methods, semi-empirical methods and most DFT > functionals will not describe van-der-Waals forces > (responsible for a good part of the non-binding interaction?) > correctly. > > In order to obtain a half sensible geometry therefore I > presume that it may be a good idea to constrain certain > geometrical quantities so as to force the two molecules to, > e..g, be at a certain distance or to be coplanar. Once I run > my geometry optimisation, I wish to check that I am at a > minimum by calculating the Hessian of the system. I seem to > obtain imaginary frequencies, suggesting that the geometry is > not at an actual minimum. This could be because of the > geometric parameters that I have constrained to particular > value simply do not bring the molecule at a minimum. Does > this mean I should ditch those parameters and try different ones? > > Does anyone have any suggestions on how to perform > optimizations on complexes of molecules? Is using constrained > parameters the best possible strategy? Or is using HF, DFT > simply not sufficient? I would be most grateful for being > pointed in the correct direction! > > Thanks for your attention and advice > James > > > -- > James Kirkpatrick > > ------------------------------------------- > Centre for Electronic Materials and devices Imperial College > ------------------------------------------- > > 020 759 47519 > > > > -= This is automatically added to each message by the mailing > script =- To recover the email address of the author of the > message, please change the strange characters on the top line > to the ** sign. You can also look up the X-Original-From: line > in the mail header.> > -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ > -+-+-+-+-+ > > > > > > From owner-chemistry@ccl.net Mon Oct 10 15:46:00 2005 From: "Andreas Serr serr::theorie.physik.uni-muenchen.de" To: CCL Subject: CCL: constrained geometry optmisation of molecular complexes Message-Id: <-29533-051010144512-29675-PJt8ydCk2WxgzvfEA1Aeog:-:server.ccl.net> X-Original-From: Andreas Serr Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=ISO-8859-1 Date: Mon, 10 Oct 2005 19:54:09 +0200 (CEST) MIME-Version: 1.0 Sent to CCL by: Andreas Serr [serr/a\theorie.physik.uni-muenchen.de] Dear James, sorry I cannot answer your questions, but I would like to make a short comment on > So far as I understand, HF > methods, semi-empirical methods and most DFT functionals will not > describe van-der-Waals forces (responsible for a good part of the > non-binding interaction?) correctly. A standard HF or DFT calculation *does* incorporate Keesom and Debye interactions (dipole-dipole, dipole-induced dipole) -- it will simply be reflected in charge re-distributions of the molecules. You are right that they do *not* account for London (dispersion) interaction (induced dipole-induced dipole). Depending on your system (mainly: magnitude of static dipole) either Keesom, Debye or London interact. may dominate van-der-Waals forces. Sorry again, if it does not help for your question. Regards, Andreas Andreas Serr Biological Soft-Matter Theory Sektion Physik - Ludwig-Maximilians-Universitaet Theresienstr. 37, D-80333 Munich, Germany tel: +49-(0)89-2180-4603 / fax: +49-(0)89-2180-4517 room no: 304 / e-mail: serr(-)theorie.physik.uni-muenchen.de http://www.theorie.physik.uni-muenchen.de/biophysics/ From owner-chemistry@ccl.net Mon Oct 10 23:26:00 2005 From: "Hua Ma g0403132;;nus.edu.sg" To: CCL Subject: CCL: W:VASP--helping Message-Id: <-29534-051010212631-12308-+D1IAzVPPgWw3z0cvnJZMg]^[server.ccl.net> X-Original-From: "Hua Ma" Sent to CCL by: "Hua Ma" [g0403132^-^nus.edu.sg] Dear all, I can not find any errors in the program, but it does not run correctly. The following is the message involved. The calculation is about the surface DOS. INCAR: SYSTEM = MoS2(3-4.4-Au) clean surface EDIFFG = -0.01 ENMAX = 280.00 eV ISTART = 0; ICHARG = 2 ISMEAR = 2; SIGMA = .25 ALGO = V NSW = 100 IBRION = 1 POTIM = 0.2 POSCAR: hcp: 3.16 4.000 0.000 0.000 2.000 3.464 0.000 0.000 0.000 5.000 16 28 1 slective dynamics cartesian 0.50000000 0.2887 0.5019 F F F 1.50000000 0.2887 0.5019 F F F 2.50000000 0.2887 0.5019 F F F 3.50000000 0.2887 0.5019 F F F 1.00000000 1.1547 0.5019 F F F 2.00000000 1.1547 0.5019 F F F 3.00000000 1.1547 0.5019 F F F 4.00000000 1.1547 0.5019 F F F 1.50000000 2.0207 0.5019 F F F 2.50000000 2.0207 0.5019 F F F 3.50000000 2.0207 0.5019 F F F 4.50000000 2.0207 0.5019 F F F 2.00000000 2.8867 0.5019 F F F 3.00000000 2.8867 0.5019 F F F 4.00000000 2.8867 0.5019 F F F 5.00000000 2.8867 0.5019 F F F 0.00000000 0.0000 0.0000 F F F 1.00000000 0.0000 0.0000 F F F 2.00000000 0.0000 0.0000 F F F 3.00000000 0.0000 0.0000 F F F 0.50000000 0.8660 0.0000 F F F 1.50000000 0.8660 0.0000 F F F 2.50000000 0.8660 0.0000 F F F 3.50000000 0.8660 0.0000 F F F 1.00000000 1.7320 0.0000 F F F 2.00000000 1.7320 0.0000 F F F 3.00000000 1.7320 0.0000 F F F 4.00000000 1.7320 0.0000 F F F 1.50000000 2.5980 0.0000 F F F 2.50000000 2.5980 0.0000 F F F 3.50000000 2.5980 0.0000 F F F 4.50000000 2.5980 0.0000 F F F 0.00000000 0.0000 1.0038 F F F 1.00000000 0.0000 1.0038 F F T 2.00000000 0.0000 1.0038 F F T 3.00000000 0.0000 1.0038 F F F 1.00000000 1.7320 1.0038 F F T 2.00000000 1.7320 1.0038 F F T 3.00000000 1.7320 1.0038 F F F 4.00000000 1.7320 1.0038 F F F 1.50000000 2.5980 1.0038 F F F 2.50000000 2.5980 1.0038 F F F 3.50000000 2.5980 1.0038 F F F 4.50000000 2.5980 1.0038 F F F 1.50000000 0.8660 1.0300 F F T KPOINTS: Monkhosrt Pack 0 Monkhost Pack 9 9 1 0 0 0 POTCAR: Mo S Au running on 2 nodes distr: one band on 1 nodes, 2 groups vasp.4.6.21 23Feb03 complex POSCAR found : 3 types and 45 ions forrtl: No such file or directory forrtl: No such file or directory forrtl: info: Fortran error message number is 29. forrtl: warning: Could not open message catalog: ifcore_msg.cat. forrtl: info: Check environment variable NLSPATH and protection of /usr/lib/ifcore_msg.cat. forrtl: info: Fortran error message number is 29. forrtl: warning: Could not open message catalog: ifcore_msg.cat. forrtl: info: Check environment variable NLSPATH and protection of /usr/lib/ifcore_msg.cat.