Dear Radhika,

As far= as I know, increasing MaxCycles is usually not that thing one should begin= with when dealing with convergence difficulties. What did I find helpful i= n such cases are

* try initial guess from calculation with sm= aller basis set (use checkpoint file in that step, then Geom=3DCheck Guess= =3DTCheck in final optimization with the basis set desired);

* try= to plot the energy and maybe also the convergence criteria as a function o= f calculation step (simple line plot); if energy goes down smoothly and the= n begins to oscillate, You should probably try to perform new optimization = starting from the last geometry before the calculation has gone fishing (wi= th Geom=3D(Check, Step=3DN) Guess=3DTCheck);

* or You can try Opt= =3DCalcFC, as You will anyway need force constants for Your IR calculation = (then You will be able to specify Freq=3DReadFC); if that helps not, try Op= t=3DCalcAll, but this is very heavy one;

* of course, increasing M= axCycle also can help occasionally.

With best regards,

Igors Mihailovs,

ISSP, University of Latvia

2014-10-28 9:40 GMT+02:00 Mariusz Radon mari= usz.radon/agmail.com &l= t;owner-chemis= try*_*ccl.net>:

Sent to CCL by: Mariusz Radon [mariusz.radon_+_gmail.com]
On 10/28/2014 04:39 AM, Radhika Narayanan
radhika.narayanan.n|*|gmail.com wrote:
> Sent to CCL by: "Radhika=C2=A0 Narayanan" [radhika.narayanan= .n- -gmail.com]
> Dear all,
>
> I am dealing with the IR spectra of organometallic compounds.While doi= ng
> the optimization using DFT functional, I encounter the following troub= le:
>
>=C2=A0 Maximum Force=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 0.000079= =C2=A0 =C2=A0 =C2=A00.000450=C2=A0 =C2=A0 =C2=A0YES
>=C2=A0 RMS=C2=A0 =C2=A0 =C2=A0Force=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 0.000018=C2=A0 =C2=A0 =C2=A00.000300=C2=A0 =C2=A0 =C2=A0YES
>=C2=A0 Maximum Displacement=C2=A0 =C2=A0 =C2=A00.116059=C2=A0 =C2=A0 = =C2=A00.001800=C2=A0 =C2=A0 =C2=A0NO
>=C2=A0 RMS=C2=A0 =C2=A0 =C2=A0Displacement=C2=A0 =C2=A0 =C2=A00.028645= =C2=A0 =C2=A0 =C2=A00.001200=C2=A0 =C2=A0 =C2=A0NO
>
> Error termination request processed by link 9999.
>
> However, optimization at HF level using the same basis set gave a
> converged geometry.
>
>=C2=A0 Maximum Force=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 0.000004= =C2=A0 =C2=A0 =C2=A00.000450=C2=A0 =C2=A0 =C2=A0YES
>=C2=A0 RMS=C2=A0 =C2=A0 =C2=A0Force=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 0.000001=C2=A0 =C2=A0 =C2=A00.000300=C2=A0 =C2=A0 =C2=A0YES
>=C2=A0 Maximum Displacement=C2=A0 =C2=A0 =C2=A00.001674=C2=A0 =C2=A0 = =C2=A00.001800=C2=A0 =C2=A0 =C2=A0YES
>=C2=A0 RMS=C2=A0 =C2=A0 =C2=A0Displacement=C2=A0 =C2=A0 =C2=A00.000428= =C2=A0 =C2=A0 =C2=A00.001200=C2=A0 =C2=A0 =C2=A0YES
>
> What might be the reason for this error?Could anyone suggest a solutio= n
> for this problem?
>

Dear Radhika,

Have you tried to increase the number of steps in geometry optimization
procedure? E.g., Opt(MaxCycles=3D300).

Take care,
Mariusz>
>

--
Dr Mariusz Radon, Ph.D.
Coordination Chemistry Group
Faculty of Chemistry
Jagiellonian University
ul. Ingardena 3, 30-060 Krakow, Poland
http://w= ww2.chemia.uj.edu.pl/~mradon

-=3D This is automatically added to each message by the mailing script =3D-=
E-mail to subscribers: CHEMISTRY*_*ccl.n= et or use:
=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/cgi-bin/ccl/send_ccl_message=

E-mail to administrators: CHEM= ISTRY-REQUEST*_*ccl.net or use
=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/cgi-bin/ccl/send_ccl_message=

Subscribe/Unsubscribe:
=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/chemistry/sub_unsub.shtml

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

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

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

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

--001a1141b5b025961e050681f430-- From owner-chemistry@ccl.net Tue Oct 28 17:37:01 2014 From: "Violeta Isabel Perez Nueno pereznueno:-:harmonicpharma.com" To: CCL Subject: CCL: "Quantitative and Systems Pharmacology" at IWBBIO 2015, Granada, SPAIN Message-Id: <-50653-141028122645-14830-UREJTdP+emxyfK6IDhvQ+w[#]server.ccl.net> X-Original-From: "Violeta Isabel Perez Nueno" Date: Tue, 28 Oct 2014 12:26:43 -0400 Sent to CCL by: "Violeta Isabel Perez Nueno" [pereznueno : harmonicpharma.com] "Quantitative and Systems Pharmacology: Thinking in a wider "systems-level" context accelerates drug discovery and enlightens our understanding of drug action" at the 3rd International Work-Conference on Bioinformatics and Biomedical Engineering (IWBBIO 2015), 15-17 April, 2015. Granada (SPAIN). http://iwbbio.ugr.es/ Quantitative and Systems Pharmacology (QSP) is an emerging discipline focused on identifying and validating drug targets, understanding existing therapeutics and discovering new ones. The goal of QSP is to understand, in a precise, predictive manner, how drugs modulate cellular networks in space and time and how they impact human pathophysiology. (QSP White Paper - October, 2011) Over the past three decades, the predominant paradigm in drug discovery was designing selective ligands for a specific target to avoid unwanted side effects. However, in the current post-genomic era, the aim is to design drugs that perturb biological networks rather than individual targets. The challenge is to be able to consider the complexity of physiological responses to treatments at very early stages of the drug development. In this way, current effort has been put into combining chemogenomics with network biology to implement new network-pharmacology approaches to drug discovery; i.e. polypharmacology approaches combined with systems biology information, which advance further in both improving efficacy and predicting unwanted off-target effects. Furthermore, the use of network biology to understand drug action outputs treasured information, i.e for pharmaceutical companies, such as alternative therapeutic indications for approved drugs, associations between proteins and drug side effects, drugdrug interactions, or pathways and gene associations which provide leads for new drug targets that may drive drug development. Following the line of QSP Workshops I and II (2008, 2010), the QSP White Paper (2011), or QSP Pittsburgh Workshop (2013), the goal of this symposium is to bring together interdisciplinary experts to help advance the understanding of how drugs act, with regard to their beneficial and toxic effects, by sharing new integrative, systems-based computational or experimental approaches/ tools / ideas which allow to increase the probability that the newly discovered drugs will prove therapeutically beneficial, together with a reduction in the risk of serious adverse events. A list of contributions with very high quality will be selected to be extended and submitted in different special issues (Biomedical Engineering Online (BMC), Current Bioinformatics, Genomics and Computational Biology, Current Topics in Medicinal Chemistry). More information can be found at: http://iwbbio.ugr.es/issues.php DEADLINE SUBMISSION OF PAPERS: November 30th, 2015 Submission must be done through the submission platform: https://easychair.org/conferences/? conf=iwbbio2015. There are two different categories for submissions: Full paper contributions (published under Lecture Notes in Bioinformatics (LNBI)) and Abstract contributions (published in the conference proceedings under both ISBN and ISSN references). More information can be found at: http://iwbbio.ugr.es/submissions.php IMPORTANT DATES: http://iwbbio.ugr.es/index.php You can learn more about IWBBIO by visiting the website http://iwbbio.ugr.es/index.php I am happy to answer any of your questions. Thank you. Best regards, Violeta Isabel Perez Nueno -- Violeta I. Perez-Nueno, Ph.D. Senior Scientist Harmonic Pharma 615 rue du Jardin Botanique 54600 Villers ls Nancy - France Cell +33(0) 638 878 547 ; Tel +33(0) 354 958 604 Fax +33(0) 383 275 652 pereznueno^harmonicpharma.com W:http://www.loria.fr/~pereznue/ www.harmonicpharma.com From owner-chemistry@ccl.net Tue Oct 28 18:12:01 2014 From: "James Robinson jameschums%%yahoo.com" To: CCL Subject: CCL: Problem with the convergence of Maximium and RMS displacement Message-Id: <-50654-141028175319-5885-qqGb/l9X8qiULCtcxl61fw||server.ccl.net> X-Original-From: "James Robinson" Date: Tue, 28 Oct 2014 17:53:18 -0400 Sent to CCL by: "James Robinson" [jameschums]|[yahoo.com] Dear Radhika If my terrible memory is still working, sometimes changing the step size using IOp(1/8=n) might help, I think the default is 30 (?) ..perhaps reduce this.. try this too.. http://www.ccl.net/chemistry/resources/messages/2002/07/09.006-dir/ Regards, James J J Robinson - personal email - opinions are personal only. From owner-chemistry@ccl.net Tue Oct 28 21:17:01 2014 From: "Brian Skinn bskinn,alum.mit.edu" To: CCL Subject: CCL:G: Problem with the convergence of Maximium and RMS displacement Message-Id: <-50655-141028211545-21414-62q+mcOdTVhGie3Iq3WBzg!=!server.ccl.net> X-Original-From: Brian Skinn Content-Type: multipart/alternative; boundary=047d7b3a833c7b76bd050685808e Date: Tue, 28 Oct 2014 21:15:19 -0400 MIME-Version: 1.0 Sent to CCL by: Brian Skinn [bskinn-*-alum.mit.edu] --047d7b3a833c7b76bd050685808e Content-Type: text/plain; charset=UTF-8 Radhika, I have had similar problems optimizing species with flat potential energy surfaces (e.g., nitromethane) with DFT in ORCA, where the forces are relatively small but the BFGS update algorithm executes disproportionately large steps. The approach I have used to address this is to tighten the quadrature grids (e.g., !GRID4 and !GRIDX9) and/or to constrain the trust region of the geometry optimization (e.g., %geom Trust -0.002 end). I unfortunately don't know the corresponding commands in Gaussian, though James's advice above sounds like it might pertain to the latter approach. I would guess the dramatic difference you are seeing in the behavior of HF vs DFT is due to the numerical quadrature required for DFT, which is absent in HF. Thus, tightening the quadrature grid might be the best place to start. Good luck, Brian On Tue, Oct 28, 2014 at 5:53 PM, James Robinson jameschums%%yahoo.com < owner-chemistry||ccl.net> wrote: > > Sent to CCL by: "James Robinson" [jameschums]|[yahoo.com] > Dear Radhika > > If my terrible memory is still working, sometimes changing the step size > using IOp(1/8=n) might help, I think the default is 30 (?) ..perhaps reduce > this.. try this too.. > > http://www.ccl.net/chemistry/resources/messages/2002/07/09.006-dir/ > > Regards, James > J J Robinson - personal email - opinions are personal only.> > > --047d7b3a833c7b76bd050685808e Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable

Radhika,=C2=A0

I have had similar probl= ems optimizing species with flat potential energy surfaces (e.g., nitrometh= ane) with DFT in ORCA, where the forces are relatively small but the BFGS u= pdate algorithm executes disproportionately large steps.=C2=A0 The approach= I have used to address this is to tighten the quadrature grids (e.g., !GRI= D4 and !GRIDX9) and/or to constrain the trust region of the geometry optimi= zation (e.g., %geom Trust -0.002 end).=C2=A0 I unfortunately don't know= the corresponding commands in Gaussian, though James's advice above so= unds like it might pertain to the latter approach.

I wou= ld guess the dramatic difference you are seeing in the behavior of HF vs DF= T is due to the numerical quadrature required for DFT, which is absent in H= F.=C2=A0 Thus, tightening the quadrature grid might be the best place to st= art.

Good luck,

Brian

On = Tue, Oct 28, 2014 at 5:53 PM, James Robinson jameschums%%yahoo.com <owner-chemistry||ccl.net> wro= te:

Sent to CCL by: "James=C2=A0 Robinson" [jameschums]|[yahoo.com]
Dear Radhika

If my terrible memory is still working, sometimes changing the step size us= ing IOp(1/8=3Dn) might help, I think the default is 30 (?) ..perhaps reduce= this.. try this too..

http://www.ccl.net/chemistry/resources/messages/2002= /07/09.006-dir/

Regards, James
J J Robinson - personal email - opinions are personal only.

-=3D This is automatically added to each message by the mailing script =3D-=
To recover the email address of the author of the message, please ch= ange
E-mail to subscribers: CHEMISTRY||ccl.n= et or use:
=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/cgi-bin/ccl/send_ccl_message=

E-mail to administrators: CHEM= ISTRY-REQUEST||ccl.net or use
=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/cgi-bin/ccl/sen= d_ccl_message

Subscribe/Unsubscribe:
=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/chemistry/sub_unsub.shtml

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

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

Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml

=C2=A0 =C2=A0 =C2=A0 http://www.ccl.net/spammers= .txt

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

--047d7b3a833c7b76bd050685808e--