CCL:G: Not able to optimize structure

I am sorry, I sent the -email while editing.
Hello everybody,

Thanks a lot for the replies.

Dear Dr. Layec, I just realized that the only try that passed had the following input:

# opt=(calcfc, loose, cartesian) pm6 nosymm scf=qc

Dear Dr. Molt, I am not sure if I understood your question correctly, but I believe tungsten is a singlet in these cases, the spin question, I am not sure but, should it be zero? I am not sure how to answer this question right about now (need to read my organometallics textbook! :))
By the way I forgot to mention on the original post that I am trying DFT methods as well but since they are taking so long to finish I started the semi-empirical calculations thinking they would be straight forward (oh was I wrong). For the DFT I am using a combination of basis sets combined with pseudo-core potentials for the tungsten

C H O P S 0

6-311G**

****

W 0

MWB

****

W 0

ECP60MWB 5 60

Ultimately, it did not pass on time so I will restart from the last geometry soon.

There were a couple of PM6 that did not finish on time on the interactive session so I do not know if I should restart them now that you said it was ambitious.

I also thought about making the calculation with smaller atoms such as chromium and maybe changing the phenyl ligands for methyl groups, but being a "dimeric" structure, I do not want to obtain something completely different using smaller ligands.

Dear Dr. Shobe,
With the Amber it crashed immediately, but with the UFF it did go through some steps. When I search for Energy in the output I find:

" Convergence criteria 0.00011250 0.00007500 0.00045000 0.00030000

Step NS ND Rises OKQ Scale Max. Force RMS Force Max. Disp. RMS Disp. Energy Flag

.....bunch of steps

2093 0 0 F T 1.61D+00 0.00009841 0.00001833 0.00009354 0.00002815 0.3068418 ====

Energy= 0.306841771"
Then,
" Item Value Threshold Converged?

Maximum Force 0.000097 0.000450 YES

RMS Force 0.000018 0.000300 YES

Maximum Displacement 0.110212 0.001800 NO

RMS Displacement 0.014639 0.001200 NO

Predicted change in Energy=-1.662896D-05"
Then,
"Convergence limit is 0.750E-04 MaxStp= 10000 StMxLn= 3.00D-04 StpMin= 1.00D-06.

Convergence criteria 0.00011250 0.00007500 0.00045000 0.00030000

Step NS ND Rises OKQ Scale Max. Force RMS Force Max. Disp. RMS Disp. Energy Flag

1 0 0 F T 1.00D+00 0.00017298 0.00001595 0.00017298 0.00001595 0.3068252 -+++

2 0 0 F T 1.47D+00 0.00007243 0.00000955 0.00013885 0.00001595 0.3068252 ====

Energy= 0.306825172 "

Finally,

" Item Value Threshold Converged?

Maximum Force 0.000070 0.000450 YES

RMS Force 0.000010 0.000300 YES

Maximum Displacement NaN 0.001800 YES

RMS Displacement NaN 0.001200 YES

Predicted change in Energy= NaN"

For the other UFF trial, there were more results for the Energy word, so I guess it went slightly better, I will just paste shorter versions starting from the top of the output:

"Step NS ND Rises OKQ Scale Max. Force RMS Force Max. Disp. RMS Disp. Energy Flag

1 0 0 F T 1.00D+00 0.00000000 0.00000000 0.00000000 0.00000000 0.8772618 ====

Energy= 0.877261757 NIter= 0."
" Energy= 0.849329056"
" Energy= 0.813651610"
" Energy= 0.771900194"

"Energy= 0.724817194 "
.....
" Energy= 0.623211212"
" Energy= 0.612600532"

Item Value Threshold Converged?

Maximum Force 0.037731 0.000450 NO

RMS Force 0.005556 0.000300 NO

Maximum Displacement 0.051140 0.001800 NO

RMS Displacement 0.007356 0.001200 NO

Predicted change in Energy=-1.136353D-02

Optimization stopped.

-- Number of steps exceeded, NStep= 10

I do not have an atom type or anything for tungsten, the input of the UFF looks like:

# opt=calcfc uff geom=connectivity

I used the ones available in Gaussian 09 , UFF, as I said before.

It has 60 atoms, I realized the typo, and I was waiting for someone to notice thanks.

Yes, for the later jobs I am, but those did not finish on time during the interactive session.

Unfortunately I am way past my allocated hours per month so I get very low priorities in my jobs now.

Guess=Read does seem to be working with the later jobs, and I do believe the tungsten is in the correct electronic state.

Just a friendly reminder that before Gaussian, I preoptimized the structures first in chem3d using MM2 then in Avogadro using UFF. So I was pretty surprised to see that Gaussian failed.

Thank you everybody for the replies. I hope my DFT on queue will pass this time.

Sincerely,

Daniel Morales Salazar

Bachelor of Science in Chemistry, Georgia Institute of Technology

PhD student, Department of Chemistry - Ångström Laboratory

Uppsala University

Box 523

SE-751 20 Uppsala

Sweden

On Mon, May 16, 2016 at 10:15 PM, Daniel Morales Salazar <danielmoralessalazar91|,|gmail.com> wrote:

Hello everybody,

Thanks a lot for the replies.

Dear Dr. Layec, I just realized that the only try that passed had the following input:

# opt=(calcfc, loose, cartesian) pm6 nosymm scf=qc

Dear Dr. Molt, I am not sure if I understood your question correctly, but I believe tungsten is a singlet in these cases, the spin question, I am not sure but, should it be zero? I am not sure how to answer this question right about now (need to read my organometallics textbook! :))
By the way I forgot to mention on the original post that I am trying DFT methods as well but since they are taking so long to finish I started the semi-empirical calculations thinking they would be straight forward (oh was I wrong). For the DFT I am using a combination of basis sets combined with pseudo-core potentials for the tungsten
C H O P S 0
6-311G**
****
W 0
MWB
****

W 0
ECP60MWB 5 60

Ultimately, it did not pass on time so I will restart from the last geometry soon.

There were a couple of PM6

I also thought about making the calculation with smaller atoms such as chromium and maybe changing the phenyl ligands for methyl groups, but being a "dimeric" structure, I do not want to obtain something completely different using smaller ligands.

Dear Dr. Shobe,
With the Amber it crashed immediately, I am assuming tungsten was not parametr

Daniel--
A few questions, that may be useful to others even if not to me :-)
First regarding the MM methods. The error message given is pretty generic. Does the calculation produce any intermediate geometries, or does it crash on the first geometric-optimization step? If there are intermediate geometries,does the energy decrease from step to step, or is there a sudden increase in energy? Do you even have an atom type and parameters for tungsten? Which MM force field are you using?
[I would have advised opt=calcall as a last resort, but given the size of your molecule (which is surely even bigger than the formula you actually wrote implies: how many carbon atoms again? :-) I might have second thoughts on that. Hopefully someone has a better idea.]
About the semiemprical method (PM6). It appears from the error message that you are not getting a decent wavefunction. Are you using guess=read to get the wavefunction from the checkpoint file in which you were able to optimize with opt=loose? (Or does guess=read even work with semiempirical methods?) Are you sure everything (especially the tungsten moieties) are in the correct electronic state?
Hope this helps,--David Shobe

SCF Energy R -1.229246327147848E+00
Total Energy R -1.229246327147848E+00
RMS Force R 2.869324126432187E-04

On Sat, May 14, 2016 at 10:42 PM, David Shobe avidshobe-$-yahoo.com <owner-chemistry|,|ccl.net> wrote:
Daniel--
A few questions, that may be useful to others even if not to me :-)
First regarding the MM methods. The error message given is pretty generic. Does the calculation produce any intermediate geometries, or does it crash on the first geometric-optimization step? If there are intermediate geometries,does the energy decrease from step to step, or is there a sudden increase in energy? Do you even have an atom type and parameters for tungsten? Which MM force field are you using?
[I would have advised opt=calcall as a last resort, but given the size of your molecule (which is surely even bigger than the formula you actually wrote implies: how many carbon atoms again? :-) I might have second thoughts on that. Hopefully someone has a better idea.]
About the semiemprical method (PM6). It appears from the error message that you are not getting a decent wavefunction. Are you using guess=read to get the wavefunction from the checkpoint file in which you were able to optimize with opt=loose? (Or does guess=read even work with semiempirical methods?) Are you sure everything (especially the tungsten moieties) are in the correct electronic state?
Hope this helps,--David Shobe

On Saturday, May 14, 2016 9:10 AM, Daniel Morales Salazar danielmoralessalazar91]_[gmail.com <owner-chemistry++ccl.net> wrote:

Sent to CCL by: "Daniel Morales Salazar" [danielmoralessalazar91*|*gmail.com]
Hello,

I am trying to optimize a molecule with the chemical formula CH52O18P4S4W2, which contains thiophenes-phospholes, phenyl groups, and
tungsten pentacarbonyl substituents.

It is a large and especially bulky molecule. I have tried using MM and
Semi-empirical methods (PM6) unsuccessfully. The only converged trial so
far was one with the "loose" keyword using PM6, which crashed in the
following calculation with stricter parameters when I did not use the
keyword. I have tried the keywords calcfc, qc, cartesian, etc.

Any advice for these types of cases?

Semi-empirical jobs crash with the message:

"Symmetry turned off by external request.
DSYEV returned Info= 2 IAlg= 2 N= 3 NDim= 3 NE2=
2.
Diagonalization in DiagDN failed."

MM methods crash with the message:

" Error termination request processed by link 9999."

I have the feeling that the initial geometries are not so bad even! Thank
you very much for your help.

Sincerely,

Daniel Morales Salazar

-= This is automatically added to each message by the mailing script =- Subscribe/Unsubscribe:
Job: http://www.ccl.net/jobs ------_Part_2073303_152685841.1463258551881
Content-Type: text/html; charsetUTF-8
Content-Transfer-Encoding: quoted-printable

<html><head></head><body><div style="color:#000; background-color:#fff; font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif;font-size:13px"><div id="yiv3733130765"><div id="yui_3_16_0_ym19_1_1463251158989_27956"><div style="color:#000;background-color:#fff;font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif;font-size:13px;" id="yui_3_16_0_ym19_1_1463251158989_27955"><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609">Daniel--</div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609"><br></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609">A few questions, that may be useful to others even if not to me :-)</div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609"><br></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609">First regarding the MM methods.  The error message given is pretty generic.  Does the calculation produce any intermediate geometries, or does it crash on the first geometric-optimization step?  If there are intermediate geometries,does the energy decrease from step to step, or is there a sudden increase in energy?  Do you even have an atom type and parameters for tungsten?  Which MM force field are you using?</div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609"><br></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609" dir="ltr">[I would have advised opt=calcall as a last resort, but given the size of your molecule (which is surely even bigger than the formula you actually wrote implies: how many carbon atoms again? :-) I might have second thoughts on that.  Hopefully someone has a better idea.]</div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609"><br clear="none"></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609" dir="ltr">About the semiemprical method (PM6).  It appears from the error message that you are not getting a decent wavefunction.  Are you using guess=read to get the wavefunction from the checkpoint file in which you were able to optimize with opt=loose?  (Or does guess=read even work with semiempirical methods?)  Are you sure everything (especially the tungsten moieties) are in the correct electronic state? </div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609" dir="ltr"><br></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609" dir="ltr">Hope this helps,</div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609" dir="ltr">--David Shobe</div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609"><br clear="none"></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22609"><br clear="none"></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_22613"><br clear="none"></div><div id="yiv3733130765yui_3_16_0_ym19_1_1463251158989_23060"><br clear="none"></div><div class="yiv3733130765qtdSeparateBR" id="yui_3_16_0_ym19_1_1463251158989_27957"><br clear="none"><br clear="none"></div></div></div></div><div class=".yiv3733130765yahoo_quoted"> <div style="font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif;font-size:13px;"> <div style="font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif;font-size:16px;"> <div class="qtdSeparateBR"><br><br></div><div class="yiv3733130765yqt3542491216" id="yiv3733130765yqtfd16282"><div dir="ltr"><font size="2" face="Arial"> On Saturday, May 14, 2016 9:10 AM, Daniel Morales Salazar danielmoralessalazar91]_[gmail.com <owner-chemistry++ccl.net> wrote:<br clear="none"></font></div> <br clear="none"><br clear="none"> <div class="yiv3733130765y_msg_container"><br clear="none">Sent to CCL by: "Daniel  Morales Salazar" [danielmoralessalazar91*|*gmail.com]<br clear="none">Hello, <br clear="none"><br clear="none">I am trying to optimize a molecule with the chemical formula CH52O18P4S4W2, which contains thiophenes-phospholes, phenyl groups, and <br clear="none">tungsten pentacarbonyl substituents.<br clear="none"><br clear="none">It is a large and especially bulky molecule. I have tried using MM and <br clear="none">Semi-empirical methods (PM6) unsuccessfully. The only converged trial so <br clear="none">far was one with the "loose" keyword using PM6, which crashed in the <br clear="none">following calculation with stricter parameters when I did not use the <br clear="none">keyword. I have tried the keywords calcfc, qc, cartesian, etc. <br clear="none"><br clear="none">Any advice for these types of cases?<br clear="none"><br clear="none">Semi-empirical jobs crash with the message:<br clear="none"><br clear="none">"Symmetry turned off by external request.<br clear="none"> DSYEV returned Info=          2 IAlg= 2 N=    3 NDim=    3 NE2=          <br clear="none">2.<br clear="none"> Diagonalization in DiagDN failed."<br clear="none"><br clear="none">MM methods crash with the message:<br clear="none"><br clear="none">" Error termination request processed by link 9999."<br clear="none"><br clear="none">I have the feeling that the initial geometries are not so bad even! Thank <br clear="none">you very much for your help.<br clear="none"><br clear="none">Sincerely,<br clear="none"><br clear="none">Daniel Morales Salazar<br clear="none"><br clear="none"><br clear="none"><br clear="none">-= This is automatically added to each message by the mailing script =-<br clear="none">To recover the email address of the author of the message, please change<br clear="none">the strange characters on the top line to the ++ sign. You can also<br clear="none"><br clear="none"><br clear="none">E-mail to subscribers: <a rel="nofollow" shape="rect" ymailto="mailto:CHEMISTRY++ccl.net" target="_blank" href="" href="mailto:CHEMISTRY%2B%2Bccl.net" target="_blank">CHEMISTRY++ccl.net">CHEMISTRY++ccl.net</a> or use:<br clear="none">      <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/cgi-bin/ccl/send_ccl_message" rel="noreferrer" target="_blank">http://www.ccl.net/cgi-bin/ccl/send_ccl_message">http://www.ccl.net/cgi-bin/ccl/send_ccl_message" rel="noreferrer" target="_blank">http://www.ccl.net/cgi-bin/ccl/send_ccl_message</a><br clear="none"><br clear="none">E-mail to administrators: <a rel="nofollow" shape="rect" ymailto="mailto:CHEMISTRY-REQUEST++ccl.net" target="_blank" href="" href="mailto:CHEMISTRY-REQUEST%2B%2Bccl.net" target="_blank">CHEMISTRY-REQUEST++ccl.net">CHEMISTRY-REQUEST++ccl.net</a> or use<br clear="none">      <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/cgi-bin/ccl/send_ccl_message" rel="noreferrer" target="_blank">http://www.ccl.net/cgi-bin/ccl/send_ccl_message">http://www.ccl.net/cgi-bin/ccl/send_ccl_message</a><br clear="none"><br clear="none"<br clear="none">      <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/chemistry/sub_unsub.shtml" rel="noreferrer" target="_blank">http://www.ccl.net/chemistry/sub_unsub.shtml">http://www.ccl.net/chemistry/sub_unsub.shtml" rel="noreferrer" target="_blank">http://www.ccl.net/chemistry/sub_unsub.shtml</a><br clear="none"><br clear="none">Before posting, check wait time at: <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/" rel="noreferrer" target="_blank">http://www.ccl.net/">http://www.ccl.net</a><br clear="none"><br clear="none">Job: <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/jobs" rel="noreferrer" target="_blank">http://www.ccl.net/jobs">http://www.ccl.net/jobs </a><br clear="none">Conferences: <a rel="nofollow" shape="rect" target="_blank" href="" href="http://server.ccl.net/chemistry/announcements/conferences/" rel="noreferrer" target="_blank">http://server.ccl.net/chemistry/announcements/conferences/">http://server.ccl.net/chemistry/announcements/conferences/</a><br clear="none"><br clear="none">Search Messages: <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/chemistry/searchccl/index.shtml" rel="noreferrer" target="_blank">http://www.ccl.net/chemistry/searchccl/index.shtml">http://www.ccl.net/chemistry/searchccl/index.shtml</a><br clear="none"><br clear="none"<br clear="none">      <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/spammers.txt" rel="noreferrer" target="_blank">http://www.ccl.net/spammers.txt">http://www.ccl.net/spammers.txt</a><br clear="none"><br clear="none">RTFI: <a rel="nofollow" shape="rect" target="_blank" href="" href="http://www.ccl.net/chemistry/aboutccl/instructions/" rel="noreferrer" target="_blank">http://www.ccl.net/chemistry/aboutccl/instructions/">http://www.ccl.net/chemistry/aboutccl/instructions/</a><br clear="none"><br clear="none"><br clear="none"><br clear="none"><br clear="none"></div> </div></div><div class="yiv3733130765yqt3542491216" id="yiv3733130765yqtfd26208"> </div></div><div class="yiv3733130765yqt3542491216" id="yiv3733130765yqtfd37150"> </div></div></div></body></html>

--
Daniel Morales Salazar
Bachelor of Science in Chemistry, Georgia Institute of Technology
PhD student, Department of Chemistry - Ångström Laboratory
Uppsala University
Box 523
SE-751 20 Uppsala
Sweden