From owner-chemistry@ccl.net Thu Aug 25 07:58:00 2016 From: "Brian Skinn bskinn++alum.mit.edu" To: CCL Subject: CCL:G: Constrained optimization and frequency calculation Message-Id: <-52360-160825073538-25247-Fb6F9LPm7gfPrelA5BzhAA+/-server.ccl.net> X-Original-From: Brian Skinn Content-Type: multipart/alternative; boundary=001a113cd7aea5ba9f053ae3cbdb Date: Thu, 25 Aug 2016 07:35:11 -0400 MIME-Version: 1.0 Sent to CCL by: Brian Skinn [bskinn*_*alum.mit.edu] --001a113cd7aea5ba9f053ae3cbdb Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Dr. Jensen, Apologies for the pedantry, but is "one-dimensional quantity" the proper term? Wouldn't, say, "order-one tensor quantity" be more accurate? That is to say, the gradient and each of the normal modes are individually 3N-dimensional, order-one tensor quantities, are they not? Best regards, Brian On Wed, Aug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk < owner-chemistry+/-ccl.net> wrote: > Gaussian by default assumes that the frequency analysis is done at a > stationary point, and projects out the T+R to get 3N-6 frequencies. > > If you are at a non-stationary point, use Freq=3DProjected to also projec= t > out the gradient, and thus get 3N-7 frequencies. > > Note that this provides 3N-7 frequencies, regardless of the number of > geometry constraints imposed, since the non-zero gradient is still only a > one-dimensional quantity. > > > > Frank > > > > Frank Jensen > > Assoc. Prof., Vice-Chair > > Dept. of Chemistry > > Aarhus University > > http://old.chem.au.dk/~frj > > > > *From:* owner-chemistry+frj=3D=3Dchem.au.dk+/-ccl.net [mailto: > owner-chemistry+frj=3D=3Dchem.au.dk+/-ccl.net] *On Behalf Of *Ankur Gupta > ankkgupt**umail.iu.edu > *Sent:* 24. august 2016 20:00 > *To:* Frank Jensen > *Subject:* CCL:G: Constrained optimization and frequency calculation > > > > Hello, > > Thank you Prof. Dr. M. Swart for answering my question. I found Baker's > paper really helpful. It discusses constrained optimization thoroughly bu= t > it does not focus much on normal mode analysis. I am more concerned about > the frequencies that we get from the Hessian after constrained > optimization. The algorithm for constrained optimization has been > implemented in most of the computational chemistry software. But I am not > able to understand the frequencies that it shows after the constrained > optimization. > > Thank you > > Ankur > > > > On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel.swart/./icrea.cat < > owner-chemistry[-]ccl.net> wrote: > > Dear Ankur, > > > > I would suggest to have a look at PQS (Baker, Pulay and co-workers) or > QUILD (Swart and co-workers). > > Both use Baker=E2=80=99s elegant solution to constrained optimizations. > > > > Baker, "Constrained optimization in delocalized internal coordinates=E2= =80=9D > > Journal of Computational Chemistry 18, 1079 (1997) > > http://dx.doi.org/10.1002/(SICI)1096-987X(199706)18:8% > 3C1079::AID-JCC12%3E3.0.CO;2-8 > > > > PQS: > > http://www.pqs-chem.com/capabilities.php > > > > QUILD: > > http://www.marcelswart.eu/quild > > https://www.scm.com/documentation/Quild/index/index > > > > Marcel > > > > On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt*indiana.edu < > owner-chemistry*ccl.net> wrote: > > > > > Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu] > Hello, > > I have been reading about constrained optimization. I have read several > papers related to the topic including the classic Reaction path Hamiltoni= an > for polyatomic molecules by Miller et al. This and other research article= s > describe what is known as 'projection operator' method to do optimization > keeping one or more internal coordinates constant. Theoretically, we shou= ld > get 3N-6 non-zero eigenvalues from the force constant matrix (for a > molecule having N nuclei) but if we apply m number of constraints in the > molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Als= o, > in cases where the constraint corresponds to a non-equilibrium geometry, > there will be coupling between rotational and vibrational motion due to > which the number of non-zero eigenvalues might change. But for the sake o= f > simplicity, we can talk about equilibrium geometries only. I use Gaussian > 09 and I observed that the number of non-zero eigenvalues did not change > after constrained optimization. ! > I know there are many computational chemistry softwares out there and I > would like to know if there is a software which can do constrained > optimization correctly and give me the right number and magnitude of > eigenvalues (frequencies) after the optimization. > > Thank you > Ankur > > > > -=3D This is automatically added to each message by the mailing script = =3D-> the strange characters on the top line to the * sign. You can also> > E-mail to subscribers: CHEMISTRY*ccl.net or use:> > E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use>
> > > > _____________________________________ > Prof. Dr. Marcel Swart, FRSC > > ICREA Research Professor at > Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC) > Univ. Girona (Spain) > > COST Action CM1305 (ECOSTBio) chair > Girona Seminar 2016 organizer > > IQCC director > > RSC Advances associate editor > > Young Academy of Europe member > > > > web > http://www.marcelswart.eu > vCard > addressbook://www.marcelswart.eu/MSwart.vcf > > > > > > > > --001a113cd7aea5ba9f053ae3cbdb Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Dr. Jensen,

Apologies for the pedantry,= but is "one-dimensional quantity" the proper term?=C2=A0 Wouldn&= #39;t, say, "order-one tensor quantity" be more accurate?

That is to say, the gradient and each of the normal modes= are individually 3N-dimensional, order-one tensor quantities, are they not= ?


Best regards,
Brian

On Wed, A= ug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk <owner-chemistry+/-ccl.net> wrote:

Gaussian by default assum= es that the frequency analysis is done at a stationary point, and projects = out the T+R to get 3N-6 frequencies.

If you are at a non-stati= onary point, use Freq=3DProjected to also project out the gradient, and thu= s get 3N-7 frequencies.

Note that this provides 3= N-7 frequencies, regardless of the number of geometry constraints imposed, = since the non-zero gradient is still only a one-dimensional quantity.

=C2=A0

Frank

=C2=A0

Frank Jensen

Assoc. Prof., Vice-Chair<= u>

Dept. of Chemistry=

Aarhus University<= u>

http://old.chem.au.dk/~frj

=C2=A0

From: owner-ch= emistry+frj=3D=3Dch= em.au.dk+/-ccl.net [mailto:owner-chemistry+frj=3D=3Dchem.au.dk+/-ccl.net] On Behalf Of Ankur Gupta ankkgupt**umail.iu.edu
Sent: 24. august 2016 20:00
To: Frank Jensen
Subject: CCL:G: Constrained optimization and frequency calculation

=C2=A0

Hello,<= /p>

Thank you Prof. Dr. M= . Swart for answering my question. I found Baker's paper really helpful= . It discusses constrained optimization thoroughly but it does not focus mu= ch on normal mode analysis. I am more concerned about the frequencies that we get from the Hessian after constrained optim= ization. The algorithm for constrained optimization has been implemented in= most of the computational chemistry software. But I am not able to underst= and the frequencies that it shows after the constrained optimization.

Thank you

Ankur

=C2=A0

On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel= .swart/./icrea.cat <<= a href=3D"mailto:owner-chemistry[-]ccl.net" target=3D"_blank">owner-chemist= ry[-]ccl.net> wrote:

Dear Ankur,

=C2=A0

I would suggest to have a look at PQS (Baker, Pulay = and co-workers) or QUILD (Swart and co-workers).

Both use Baker=E2=80=99s elegant solution to constra= ined optimizations.

=C2=A0

Baker, "Constrained optimization in delocalized= internal coordinates=E2=80=9D

Journal of Computational Chemistry 18, 1079 (1997)

=C2=A0

PQS:

=C2=A0

QUILD:

=C2=A0

Marcel

=C2=A0

On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt= *indiana.edu <owner-chemistry*ccl= .net> wrote:

=C2=A0


Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu]
Hello,

I have been reading about constrained optimization. I have read several pap= ers related to the topic including the classic Reaction path Hamiltonian fo= r polyatomic molecules by Miller et al. This and other research articles de= scribe what is known as 'projection operator' method to do optimization keeping one or more internal coord= inates constant. Theoretically, we should get 3N-6 non-zero eigenvalues fro= m the force constant matrix (for a molecule having N nuclei) but if we appl= y m number of constraints in the molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Also, in cases= where the constraint corresponds to a non-equilibrium geometry, there will= be coupling between rotational and vibrational motion due to which the num= ber of non-zero eigenvalues might change. But for the sake of simplicity, we can talk about equilibrium geom= etries only. I use Gaussian 09 and I observed that the number of non-zero e= igenvalues did not change after constrained optimization. !
I know there are many computational chemistry softwares out there and I wou= ld like to know if there is a software which can do constrained optimizatio= n correctly and give me the right number and magnitude of eigenvalues (freq= uencies) after the optimization.

Thank you
Ankur



-=3D This is automatically added to each message by the mailing script =3D-= the strange characters on the top line to the * sign. You can also
E-mail to subscribers: CHEMISTRY*ccl.net or use:
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/chemistry/sub_unsub.s= html

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.n= et/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/
Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml
<br
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/

=C2=A0


_____________________________________
Prof. Dr. Marcel Swart, FRSC

ICREA Research Professor at
Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC)
Univ. Girona (Spain)

COST Action CM1305 (ECOSTBio) chair
Girona Seminar 2016 organizer

IQCC director<= /u>

RSC Advances associate e= ditor

Young Academy of Europe = member

=C2=A0

=C2=A0

=C2=A0


--001a113cd7aea5ba9f053ae3cbdb-- From owner-chemistry@ccl.net Thu Aug 25 11:07:01 2016 From: "Ankur Gupta ankkgupt(!)umail.iu.edu" To: CCL Subject: CCL:G: Constrained optimization and frequency calculation Message-Id: <-52361-160825110324-17500-7jb4iJb3bU5U1Mj7msU9vw!A!server.ccl.net> X-Original-From: Ankur Gupta Content-Type: multipart/alternative; boundary=94eb2c12542ad03c2e053ae6b18b Date: Thu, 25 Aug 2016 11:03:02 -0400 MIME-Version: 1.0 Sent to CCL by: Ankur Gupta [ankkgupt*o*umail.iu.edu] --94eb2c12542ad03c2e053ae6b18b Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Thank you Professor Jensen and Dr. Skinn. The two widely used methods to do constrained optimization are Lagrange multiplier method by J. Baker ( http://onlinelibrary.wiley.com/doi/10.1002/(SICI)1096-987X(199706)18:8%3C10= 79::AID-JCC12%3E3.0.CO;2-8/abstract) and Projection operator method by D. G. Truhlar ( http://onlinelibrary.wiley.com/doi/10.1002/jcc.540120311/abstract). Of the two, Lagrange multiplier method has been implemented in most of the computational chemistry packages to do the constrained optimization. But, the Projection operator method seems to give the correct number of non-zero eigenvalues from the projected Hessian corresponding to the number of geometry constraints present in the molecule. It was my understanding that the Lagrangian Hessian (obtained in the Lagrange multiplier method) should give the correct number and values of the frequencies for a molecule at a non-stationary point, but that is probably not the case. I sincerely thank Professor Jensen for writing Introduction to Computational Chemistry. Ankur On Thu, Aug 25, 2016 at 7:35 AM, Brian Skinn bskinn++alum.mit.edu < owner-chemistry(~)ccl.net> wrote: > Dr. Jensen, > > Apologies for the pedantry, but is "one-dimensional quantity" the proper > term? Wouldn't, say, "order-one tensor quantity" be more accurate? > > That is to say, the gradient and each of the normal modes are individuall= y > 3N-dimensional, order-one tensor quantities, are they not? > > > Best regards, > Brian > > On Wed, Aug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk < > owner-chemistry]_[ccl.net> wrote: > >> Gaussian by default assumes that the frequency analysis is done at a >> stationary point, and projects out the T+R to get 3N-6 frequencies. >> >> If you are at a non-stationary point, use Freq=3DProjected to also proje= ct >> out the gradient, and thus get 3N-7 frequencies. >> >> Note that this provides 3N-7 frequencies, regardless of the number of >> geometry constraints imposed, since the non-zero gradient is still only = a >> one-dimensional quantity. >> >> >> >> Frank >> >> >> >> Frank Jensen >> >> Assoc. Prof., Vice-Chair >> >> Dept. of Chemistry >> >> Aarhus University >> >> http://old.chem.au.dk/~frj >> >> >> >> *From:* owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net [mailto: >> owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net] *On Behalf Of *Ankur Gupt= a >> ankkgupt**umail.iu.edu >> *Sent:* 24. august 2016 20:00 >> *To:* Frank Jensen >> *Subject:* CCL:G: Constrained optimization and frequency calculation >> >> >> >> Hello, >> >> Thank you Prof. Dr. M. Swart for answering my question. I found Baker's >> paper really helpful. It discusses constrained optimization thoroughly b= ut >> it does not focus much on normal mode analysis. I am more concerned abou= t >> the frequencies that we get from the Hessian after constrained >> optimization. The algorithm for constrained optimization has been >> implemented in most of the computational chemistry software. But I am no= t >> able to understand the frequencies that it shows after the constrained >> optimization. >> >> Thank you >> >> Ankur >> >> >> >> On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel.swart/./icrea.cat < >> owner-chemistry[-]ccl.net> wrote: >> >> Dear Ankur, >> >> >> >> I would suggest to have a look at PQS (Baker, Pulay and co-workers) or >> QUILD (Swart and co-workers). >> >> Both use Baker=E2=80=99s elegant solution to constrained optimizations. >> >> >> >> Baker, "Constrained optimization in delocalized internal coordinates=E2= =80=9D >> >> Journal of Computational Chemistry 18, 1079 (1997) >> >> http://dx.doi.org/10.1002/(SICI)1096-987X(199706)18:8%3C1079 >> ::AID-JCC12%3E3.0.CO;2-8 >> >> >> >> PQS: >> >> http://www.pqs-chem.com/capabilities.php >> >> >> >> QUILD: >> >> http://www.marcelswart.eu/quild >> >> https://www.scm.com/documentation/Quild/index/index >> >> >> >> Marcel >> >> >> >> On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt*indiana.edu < >> owner-chemistry*ccl.net> wrote: >> >> >> >> >> Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu] >> Hello, >> >> I have been reading about constrained optimization. I have read several >> papers related to the topic including the classic Reaction path Hamilton= ian >> for polyatomic molecules by Miller et al. This and other research articl= es >> describe what is known as 'projection operator' method to do optimizatio= n >> keeping one or more internal coordinates constant. Theoretically, we sho= uld >> get 3N-6 non-zero eigenvalues from the force constant matrix (for a >> molecule having N nuclei) but if we apply m number of constraints in the >> molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Al= so, >> in cases where the constraint corresponds to a non-equilibrium geometry, >> there will be coupling between rotational and vibrational motion due to >> which the number of non-zero eigenvalues might change. But for the sake = of >> simplicity, we can talk about equilibrium geometries only. I use Gaussia= n >> 09 and I observed that the number of non-zero eigenvalues did not change >> after constrained optimization. ! >> I know there are many computational chemistry softwares out there and I >> would like to know if there is a software which can do constrained >> optimization correctly and give me the right number and magnitude of >> eigenvalues (frequencies) after the optimization. >> >> Thank you >> Ankur >> >> >> >> -=3D This is automatically added to each message by the mailing script = =3D- >> the strange characters on the top line to the * sign. You can also >> E-mail to subscribers: CHEMISTRY*ccl.net or use:>> >> E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use>>
> >> >> >> >> _____________________________________ >> Prof. Dr. Marcel Swart, FRSC >> >> ICREA Research Professor at >> Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC) >> Univ. Girona (Spain) >> >> COST Action CM1305 (ECOSTBio) chair >> Girona Seminar 2016 organizer >> >> IQCC director >> >> RSC Advances associate editor >> >> Young Academy of Europe member >> >> >> >> web >> http://www.marcelswart.eu >> vCard >> addressbook://www.marcelswart.eu/MSwart.vcf >> >> >> >> >> >> >> >> > > On Thu, Aug 25, 2016 at 7:35 AM, Brian Skinn bskinn++alum.mit.edu < owner-chemistry(~)ccl.net> wrote: > Dr. Jensen, > > Apologies for the pedantry, but is "one-dimensional quantity" the proper > term? Wouldn't, say, "order-one tensor quantity" be more accurate? > > That is to say, the gradient and each of the normal modes are individuall= y > 3N-dimensional, order-one tensor quantities, are they not? > > > Best regards, > Brian > > On Wed, Aug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk < > owner-chemistry]_[ccl.net> wrote: > >> Gaussian by default assumes that the frequency analysis is done at a >> stationary point, and projects out the T+R to get 3N-6 frequencies. >> >> If you are at a non-stationary point, use Freq=3DProjected to also proje= ct >> out the gradient, and thus get 3N-7 frequencies. >> >> Note that this provides 3N-7 frequencies, regardless of the number of >> geometry constraints imposed, since the non-zero gradient is still only = a >> one-dimensional quantity. >> >> >> >> Frank >> >> >> >> Frank Jensen >> >> Assoc. Prof., Vice-Chair >> >> Dept. of Chemistry >> >> Aarhus University >> >> http://old.chem.au.dk/~frj >> >> >> >> *From:* owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net [mailto: >> owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net] *On Behalf Of *Ankur Gupt= a >> ankkgupt**umail.iu.edu >> *Sent:* 24. august 2016 20:00 >> *To:* Frank Jensen >> *Subject:* CCL:G: Constrained optimization and frequency calculation >> >> >> >> Hello, >> >> Thank you Prof. Dr. M. Swart for answering my question. I found Baker's >> paper really helpful. It discusses constrained optimization thoroughly b= ut >> it does not focus much on normal mode analysis. I am more concerned abou= t >> the frequencies that we get from the Hessian after constrained >> optimization. The algorithm for constrained optimization has been >> implemented in most of the computational chemistry software. But I am no= t >> able to understand the frequencies that it shows after the constrained >> optimization. >> >> Thank you >> >> Ankur >> >> >> >> On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel.swart/./icrea.cat < >> owner-chemistry[-]ccl.net> wrote: >> >> Dear Ankur, >> >> >> >> I would suggest to have a look at PQS (Baker, Pulay and co-workers) or >> QUILD (Swart and co-workers). >> >> Both use Baker=E2=80=99s elegant solution to constrained optimizations. >> >> >> >> Baker, "Constrained optimization in delocalized internal coordinates=E2= =80=9D >> >> Journal of Computational Chemistry 18, 1079 (1997) >> >> http://dx.doi.org/10.1002/(SICI)1096-987X(199706)18:8%3C1079 >> ::AID-JCC12%3E3.0.CO;2-8 >> >> >> >> PQS: >> >> http://www.pqs-chem.com/capabilities.php >> >> >> >> QUILD: >> >> http://www.marcelswart.eu/quild >> >> https://www.scm.com/documentation/Quild/index/index >> >> >> >> Marcel >> >> >> >> On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt*indiana.edu < >> owner-chemistry*ccl.net> wrote: >> >> >> >> >> Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu] >> Hello, >> >> I have been reading about constrained optimization. I have read several >> papers related to the topic including the classic Reaction path Hamilton= ian >> for polyatomic molecules by Miller et al. This and other research articl= es >> describe what is known as 'projection operator' method to do optimizatio= n >> keeping one or more internal coordinates constant. Theoretically, we sho= uld >> get 3N-6 non-zero eigenvalues from the force constant matrix (for a >> molecule having N nuclei) but if we apply m number of constraints in the >> molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Al= so, >> in cases where the constraint corresponds to a non-equilibrium geometry, >> there will be coupling between rotational and vibrational motion due to >> which the number of non-zero eigenvalues might change. But for the sake = of >> simplicity, we can talk about equilibrium geometries only. I use Gaussia= n >> 09 and I observed that the number of non-zero eigenvalues did not change >> after constrained optimization. ! >> I know there are many computational chemistry softwares out there and I >> would like to know if there is a software which can do constrained >> optimization correctly and give me the right number and magnitude of >> eigenvalues (frequencies) after the optimization. >> >> Thank you >> Ankur >> >> >> >> -=3D This is automatically added to each message by the mailing script = =3D- >> the strange characters on the top line to the * sign. You can also >> E-mail to subscribers: CHEMISTRY*ccl.net or use:>> >> E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use>>
> >> >> >> >> _____________________________________ >> Prof. Dr. Marcel Swart, FRSC >> >> ICREA Research Professor at >> Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC) >> Univ. Girona (Spain) >> >> COST Action CM1305 (ECOSTBio) chair >> Girona Seminar 2016 organizer >> >> IQCC director >> >> RSC Advances associate editor >> >> Young Academy of Europe member >> >> >> >> web >> http://www.marcelswart.eu >> vCard >> addressbook://www.marcelswart.eu/MSwart.vcf >> >> >> >> >> >> >> >> > > On Thu, Aug 25, 2016 at 7:35 AM, Brian Skinn bskinn++alum.mit.edu < owner-chemistry(~)ccl.net> wrote: > Dr. Jensen, > > Apologies for the pedantry, but is "one-dimensional quantity" the proper > term? Wouldn't, say, "order-one tensor quantity" be more accurate? > > That is to say, the gradient and each of the normal modes are individuall= y > 3N-dimensional, order-one tensor quantities, are they not? > > > Best regards, > Brian > > On Wed, Aug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk < > owner-chemistry]_[ccl.net> wrote: > >> Gaussian by default assumes that the frequency analysis is done at a >> stationary point, and projects out the T+R to get 3N-6 frequencies. >> >> If you are at a non-stationary point, use Freq=3DProjected to also proje= ct >> out the gradient, and thus get 3N-7 frequencies. >> >> Note that this provides 3N-7 frequencies, regardless of the number of >> geometry constraints imposed, since the non-zero gradient is still only = a >> one-dimensional quantity. >> >> >> >> Frank >> >> >> >> Frank Jensen >> >> Assoc. Prof., Vice-Chair >> >> Dept. of Chemistry >> >> Aarhus University >> >> http://old.chem.au.dk/~frj >> >> >> >> *From:* owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net [mailto: >> owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net] *On Behalf Of *Ankur Gupt= a >> ankkgupt**umail.iu.edu >> *Sent:* 24. august 2016 20:00 >> *To:* Frank Jensen >> *Subject:* CCL:G: Constrained optimization and frequency calculation >> >> >> >> Hello, >> >> Thank you Prof. Dr. M. Swart for answering my question. I found Baker's >> paper really helpful. It discusses constrained optimization thoroughly b= ut >> it does not focus much on normal mode analysis. I am more concerned abou= t >> the frequencies that we get from the Hessian after constrained >> optimization. The algorithm for constrained optimization has been >> implemented in most of the computational chemistry software. But I am no= t >> able to understand the frequencies that it shows after the constrained >> optimization. >> >> Thank you >> >> Ankur >> >> >> >> On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel.swart/./icrea.cat < >> owner-chemistry[-]ccl.net> wrote: >> >> Dear Ankur, >> >> >> >> I would suggest to have a look at PQS (Baker, Pulay and co-workers) or >> QUILD (Swart and co-workers). >> >> Both use Baker=E2=80=99s elegant solution to constrained optimizations. >> >> >> >> Baker, "Constrained optimization in delocalized internal coordinates=E2= =80=9D >> >> Journal of Computational Chemistry 18, 1079 (1997) >> >> http://dx.doi.org/10.1002/(SICI)1096-987X(199706)18:8%3C1079 >> ::AID-JCC12%3E3.0.CO;2-8 >> >> >> >> PQS: >> >> http://www.pqs-chem.com/capabilities.php >> >> >> >> QUILD: >> >> http://www.marcelswart.eu/quild >> >> https://www.scm.com/documentation/Quild/index/index >> >> >> >> Marcel >> >> >> >> On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt*indiana.edu < >> owner-chemistry*ccl.net> wrote: >> >> >> >> >> Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu] >> Hello, >> >> I have been reading about constrained optimization. I have read several >> papers related to the topic including the classic Reaction path Hamilton= ian >> for polyatomic molecules by Miller et al. This and other research articl= es >> describe what is known as 'projection operator' method to do optimizatio= n >> keeping one or more internal coordinates constant. Theoretically, we sho= uld >> get 3N-6 non-zero eigenvalues from the force constant matrix (for a >> molecule having N nuclei) but if we apply m number of constraints in the >> molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Al= so, >> in cases where the constraint corresponds to a non-equilibrium geometry, >> there will be coupling between rotational and vibrational motion due to >> which the number of non-zero eigenvalues might change. But for the sake = of >> simplicity, we can talk about equilibrium geometries only. I use Gaussia= n >> 09 and I observed that the number of non-zero eigenvalues did not change >> after constrained optimization. ! >> I know there are many computational chemistry softwares out there and I >> would like to know if there is a software which can do constrained >> optimization correctly and give me the right number and magnitude of >> eigenvalues (frequencies) after the optimization. >> >> Thank you >> Ankur >> >> >> >> -=3D This is automatically added to each message by the mailing script = =3D- >> the strange characters on the top line to the * sign. You can also >> E-mail to subscribers: CHEMISTRY*ccl.net or use:>> >> E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use>>
> >> >> >> >> _____________________________________ >> Prof. Dr. Marcel Swart, FRSC >> >> ICREA Research Professor at >> Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC) >> Univ. Girona (Spain) >> >> COST Action CM1305 (ECOSTBio) chair >> Girona Seminar 2016 organizer >> >> IQCC director >> >> RSC Advances associate editor >> >> Young Academy of Europe member >> >> >> >> web >> http://www.marcelswart.eu >> vCard >> addressbook://www.marcelswart.eu/MSwart.vcf >> >> >> >> >> >> >> >> > > --94eb2c12542ad03c2e053ae6b18b Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Thank you Professor Jensen and Dr. Skinn. <= br>
The two widely used methods to do constrained optimization are Lag= range multiplier method by J. Baker (http://onlinelibrary.wiley.com/doi/10.1002/(SICI)1096-987X(199706= )18:8%3C1079::AID-JCC12%3E3.0.CO;2-8/abstract) and Projection operator = method by D. G. Truhlar (http://onlinelibrary.wiley.com/doi/10.1002/jcc.= 540120311/abstract). Of the two, Lagrange multiplier method has been im= plemented in most of the computational chemistry packages to do the constra= ined optimization. But, the Projection operator method seems to give the co= rrect number of non-zero eigenvalues from the projected Hessian correspondi= ng to the number of geometry constraints present in the molecule. It was my= understanding that the Lagrangian Hessian (obtained in the Lagrange multip= lier method) should give the correct number and values of the frequencies f= or a molecule at a non-stationary point, but that is probably not the case.=

I sincerely thank Professor Jensen for writing Introduction = to Computational Chemistry.

Ankur

On Thu, Aug 25, = 2016 at 7:35 AM, Brian Skinn bskinn++alum.mit.edu <owner-chemistry(~)ccl.net>= wrote:
Dr. Jensen,

Apologies for the pedantry, but is "= one-dimensional quantity" the proper term?=C2=A0 Wouldn't, say, &q= uot;order-one tensor quantity" be more accurate?

<= div>That is to say, the gradient and each of the normal modes are individua= lly 3N-dimensional, order-one tensor quantities, are they not?

Best regards,
Brian

On Wed, Aug 24, 2016 at= 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk <owner-chemistry]_[ccl.net> w= rote:

Gaussian by default = assumes that the frequency analysis is done at a stationary point, and proj= ects out the T+R to get 3N-6 frequencies.

If you are at a non-= stationary point, use Freq=3DProjected to also project out the gradient, an= d thus get 3N-7 frequencies.

Note that this provi= des 3N-7 frequencies, regardless of the number of geometry constraints impo= sed, since the non-zero gradient is still only a one-dimensional quantity.

=C2=A0=

Frank<= /span>

=C2=A0=

Frank Jensen<= u>

Assoc. Prof., Vice-C= hair

Dept. of Chemistry

Aarhus University=

http://old.chem.au.dk/~frj<= u>

=C2=A0=

From: owner-chemis= try+frj=3D=3Dchem= .au.dk]_[ccl.net [mailto:owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net] On Behalf Of Ankur Gupta ankkgupt**umail.iu.edu
Sent: 24. august 2016 20:00
To: Frank Jensen
Subject: CCL:G: Constrained optimization and frequency calculation

=C2=A0

Hello,

Thank you Prof. Dr. M. = Swart for answering my question. I found Baker's paper really helpful. = It discusses constrained optimization thoroughly but it does not focus much= on normal mode analysis. I am more concerned about the frequencies that we get from the Hessian after constrained optim= ization. The algorithm for constrained optimization has been implemented in= most of the computational chemistry software. But I am not able to underst= and the frequencies that it shows after the constrained optimization.

Thank you

Ankur

=C2=A0

On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel= .swart/./icrea.cat <<= a href=3D"mailto:owner-chemistry[-]ccl.net" target=3D"_blank">owner-chemist= ry[-]ccl.net> wrote:

Dear Ankur,

=C2=A0

I would suggest to have a look at PQS (Baker, Pulay = and co-workers) or QUILD (Swart and co-workers).

Both use Baker=E2=80=99s elegant solution to constra= ined optimizations.

=C2=A0

Baker, "Constrained optimization in delocalized= internal coordinates=E2=80=9D

Journal of Computational Chemistry 18, 1079 (1997)

=C2=A0

PQS:

=C2=A0

QUILD:

=C2=A0

Marcel

=C2=A0

On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt= *indiana.edu <owner-chemistry*ccl= .net> wrote:

=C2=A0


Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu]
Hello,

I have been reading about constrained optimization. I have read several pap= ers related to the topic including the classic Reaction path Hamiltonian fo= r polyatomic molecules by Miller et al. This and other research articles de= scribe what is known as 'projection operator' method to do optimization keeping one or more internal coord= inates constant. Theoretically, we should get 3N-6 non-zero eigenvalues fro= m the force constant matrix (for a molecule having N nuclei) but if we appl= y m number of constraints in the molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Also, in cases= where the constraint corresponds to a non-equilibrium geometry, there will= be coupling between rotational and vibrational motion due to which the num= ber of non-zero eigenvalues might change. But for the sake of simplicity, we can talk about equilibrium geom= etries only. I use Gaussian 09 and I observed that the number of non-zero e= igenvalues did not change after constrained optimization. !
I know there are many computational chemistry softwares out there and I wou= ld like to know if there is a software which can do constrained optimizatio= n correctly and give me the right number and magnitude of eigenvalues (freq= uencies) after the optimization.

Thank you
Ankur



-=3D This is automatically added to each message by the mailing script =3D-= the strange characters on the top line to the * sign. You can also
E-mail to subscribers: CHEMISTRY*ccl.net or use:
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/chemistry/sub_unsub.s= html

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.n= et/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml
<br
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/

=C2=A0


_____________________________________
Prof. Dr. Marcel Swart, FRSC

ICREA Research Professor at
Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC)
Univ. Girona (Spain)

COST Action CM1305 (ECOSTBio) chair
Girona Seminar 2016 organizer

IQCC director<= /u>

RSC Advances associate e= ditor

Young Academy of Europe = member

=C2=A0

=C2=A0

=C2=A0




On Thu, Aug 25, 2016 at 7:35 AM, Brian Skinn bskinn++alum.mit.edu = <owner-chem= istry(~)ccl.net> wrote:
Dr. Jensen,

Apologies for= the pedantry, but is "one-dimensional quantity" the proper term?= =C2=A0 Wouldn't, say, "order-one tensor quantity" be more acc= urate?

That is to say, the gradient and each of th= e normal modes are individually 3N-dimensional, order-one tensor quantities= , are they not?


Best regards,
=
Brian

On Wed, Aug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk <owner-chemistr= y]_[ccl.net> wrote:

Gaussian by default = assumes that the frequency analysis is done at a stationary point, and proj= ects out the T+R to get 3N-6 frequencies.

If you are at a non-= stationary point, use Freq=3DProjected to also project out the gradient, an= d thus get 3N-7 frequencies.

Note that this provi= des 3N-7 frequencies, regardless of the number of geometry constraints impo= sed, since the non-zero gradient is still only a one-dimensional quantity.

=C2=A0=

Frank<= /span>

=C2=A0=

Frank Jensen<= u>

Assoc. Prof., Vice-C= hair

Dept. of Chemistry

Aarhus University=

http://old.chem.au.dk/~frj<= u>

=C2=A0=

From: owner-chemis= try+frj=3D=3Dchem= .au.dk]_[ccl.net [mailto:owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net] On Behalf Of Ankur Gupta ankkgupt**umail.iu.edu
Sent: 24. august 2016 20:00
To: Frank Jensen
Subject: CCL:G: Constrained optimization and frequency calculation

=C2=A0

Hello,

Thank you Prof. Dr. M. = Swart for answering my question. I found Baker's paper really helpful. = It discusses constrained optimization thoroughly but it does not focus much= on normal mode analysis. I am more concerned about the frequencies that we get from the Hessian after constrained optim= ization. The algorithm for constrained optimization has been implemented in= most of the computational chemistry software. But I am not able to underst= and the frequencies that it shows after the constrained optimization.

Thank you

Ankur

=C2=A0

On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel= .swart/./icrea.cat <<= a href=3D"mailto:owner-chemistry[-]ccl.net" target=3D"_blank">owner-chemist= ry[-]ccl.net> wrote:

Dear Ankur,

=C2=A0

I would suggest to have a look at PQS (Baker, Pulay = and co-workers) or QUILD (Swart and co-workers).

Both use Baker=E2=80=99s elegant solution to constra= ined optimizations.

=C2=A0

Baker, "Constrained optimization in delocalized= internal coordinates=E2=80=9D

Journal of Computational Chemistry 18, 1079 (1997)

=C2=A0

PQS:

=C2=A0

QUILD:

=C2=A0

Marcel

=C2=A0

On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt= *indiana.edu <owner-chemistry*ccl= .net> wrote:

=C2=A0


Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu]
Hello,

I have been reading about constrained optimization. I have read several pap= ers related to the topic including the classic Reaction path Hamiltonian fo= r polyatomic molecules by Miller et al. This and other research articles de= scribe what is known as 'projection operator' method to do optimization keeping one or more internal coord= inates constant. Theoretically, we should get 3N-6 non-zero eigenvalues fro= m the force constant matrix (for a molecule having N nuclei) but if we appl= y m number of constraints in the molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Also, in cases= where the constraint corresponds to a non-equilibrium geometry, there will= be coupling between rotational and vibrational motion due to which the num= ber of non-zero eigenvalues might change. But for the sake of simplicity, we can talk about equilibrium geom= etries only. I use Gaussian 09 and I observed that the number of non-zero e= igenvalues did not change after constrained optimization. !
I know there are many computational chemistry softwares out there and I wou= ld like to know if there is a software which can do constrained optimizatio= n correctly and give me the right number and magnitude of eigenvalues (freq= uencies) after the optimization.

Thank you
Ankur



-=3D This is automatically added to each message by the mailing script =3D-= the strange characters on the top line to the * sign. You can also
E-mail to subscribers: CHEMISTRY*ccl.net or use:
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/chemistry/sub_unsub.s= html

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.n= et/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml
<br
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/

=C2=A0


_____________________________________
Prof. Dr. Marcel Swart, FRSC

ICREA Research Professor at
Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC)
Univ. Girona (Spain)

COST Action CM1305 (ECOSTBio) chair
Girona Seminar 2016 organizer

IQCC director<= /u>

RSC Advances associate e= ditor

Young Academy of Europe = member

=C2=A0

=C2=A0

=C2=A0




On Thu, Aug 25, 2016 at 7:3= 5 AM, Brian Skinn bskinn++alum.mit.edu = <owner-chemistry(~)ccl.net> wrote:
Dr. Jensen,

Apologies for= the pedantry, but is "one-dimensional quantity" the proper term?= =C2=A0 Wouldn't, say, "order-one tensor quantity" be more acc= urate?

That is to say, the gradient and each of th= e normal modes are individually 3N-dimensional, order-one tensor quantities= , are they not?


Best regards,
=
Brian

On Wed, Aug 24, 2016 at 3:27 PM, Frank Jensen frj=3D=3D=3Dchem.au.dk <owner-chemistr= y]_[ccl.net> wrote:

Gaussian by default assum= es that the frequency analysis is done at a stationary point, and projects = out the T+R to get 3N-6 frequencies.

If you are at a non-stati= onary point, use Freq=3DProjected to also project out the gradient, and thu= s get 3N-7 frequencies.

Note that this provides 3= N-7 frequencies, regardless of the number of geometry constraints imposed, = since the non-zero gradient is still only a one-dimensional quantity.

=C2=A0

Frank

=C2=A0

Frank Jensen

Assoc. Prof., Vice-Chair<= u>

Dept. of Chemistry=

Aarhus University<= u>

http://old.chem.au.dk/~frj

=C2=A0

From: owner-ch= emistry+frj=3D=3D= chem.au.dk]_[ccl.net [mailto:owner-chemistry+frj=3D=3Dchem.au.dk]_[ccl.net] On Behalf Of Ankur Gupta ankkgupt**umail.iu.edu
Sent: 24. august 2016 20:00
To: Frank Jensen
Subject: CCL:G: Constrained optimization and frequency calculation

=C2=A0

Hello,<= /p>

Thank you Prof. Dr. M= . Swart for answering my question. I found Baker's paper really helpful= . It discusses constrained optimization thoroughly but it does not focus mu= ch on normal mode analysis. I am more concerned about the frequencies that we get from the Hessian after constrained optim= ization. The algorithm for constrained optimization has been implemented in= most of the computational chemistry software. But I am not able to underst= and the frequencies that it shows after the constrained optimization.

Thank you

Ankur

=C2=A0

On Sat, Aug 20, 2016 at 5:03 AM, Marcel Swart marcel= .swart/./icrea.cat <<= a href=3D"mailto:owner-chemistry[-]ccl.net" target=3D"_blank">owner-chemist= ry[-]ccl.net> wrote:

Dear Ankur,

=C2=A0

I would suggest to have a look at PQS (Baker, Pulay = and co-workers) or QUILD (Swart and co-workers).

Both use Baker=E2=80=99s elegant solution to constra= ined optimizations.

=C2=A0

Baker, "Constrained optimization in delocalized= internal coordinates=E2=80=9D

Journal of Computational Chemistry 18, 1079 (1997)

=C2=A0

PQS:

=C2=A0

QUILD:

=C2=A0

Marcel

=C2=A0

On 19 Aug 2016, at 22:33, Ankur Kumar Gupta ankkgupt= *indiana.edu <owner-chemistry*ccl= .net> wrote:

=C2=A0


Sent to CCL by: "Ankur Kumar Gupta" [ankkgupt||indiana.edu]
Hello,

I have been reading about constrained optimization. I have read several pap= ers related to the topic including the classic Reaction path Hamiltonian fo= r polyatomic molecules by Miller et al. This and other research articles de= scribe what is known as 'projection operator' method to do optimization keeping one or more internal coord= inates constant. Theoretically, we should get 3N-6 non-zero eigenvalues fro= m the force constant matrix (for a molecule having N nuclei) but if we appl= y m number of constraints in the molecule, we should obtain 3N-6-m non-zero eigenvalues (frequencies). Also, in cases= where the constraint corresponds to a non-equilibrium geometry, there will= be coupling between rotational and vibrational motion due to which the num= ber of non-zero eigenvalues might change. But for the sake of simplicity, we can talk about equilibrium geom= etries only. I use Gaussian 09 and I observed that the number of non-zero e= igenvalues did not change after constrained optimization. !
I know there are many computational chemistry softwares out there and I wou= ld like to know if there is a software which can do constrained optimizatio= n correctly and give me the right number and magnitude of eigenvalues (freq= uencies) after the optimization.

Thank you
Ankur



-=3D This is automatically added to each message by the mailing script =3D-= the strange characters on the top line to the * sign. You can also
E-mail to subscribers: CHEMISTRY*ccl.net or use:
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

E-mail to administrators: CHEMISTRY-REQUEST*ccl.net or use
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/cgi-bin/ccl/send_c= cl_message

=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/chemistry/sub_unsub.s= html

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.n= et/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml
<br
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/

=C2=A0


_____________________________________
Prof. Dr. Marcel Swart, FRSC

ICREA Research Professor at
Institut de Qu=C3=ADmica Computacional i Cat=C3=A0lisi (IQCC)
Univ. Girona (Spain)

COST Action CM1305 (ECOSTBio) chair
Girona Seminar 2016 organizer

IQCC director<= /u>

RSC Advances associate e= ditor

Young Academy of Europe = member

=C2=A0

=C2=A0

=C2=A0



--94eb2c12542ad03c2e053ae6b18b--