RE: Re: CPCM TS location



Please take a look at Wang, Nakamura, Balbuena, JACS, 2002, Vol. 124, 4408.
We had used PCM-B3PW91 in the paper as well as CPCM (to be published) to
optimize reductive decomposition TS (page 4410, structure 7) of vinylene carbonate.
 
Yixuan Wang
 
Department of Chem. Engineering
Univ. South Carolina 
 
-----Original Message-----
From: zjwu [mailto:zjwu %! at !% ns.ciac.jl.cn]
Sent: Thursday, May 01, 2003 8:27 PM
To: help %! at !% gaussian.com
Cc: chemistry %! at !% ccl.net
Subject: CCL:Re: CPCM TS location

Dear Dr. Hess,

 

Do you know any publications containing cpcm ts structure calculations? I do find some papers on local minima optimizations using cpcm, but did not see any paper on transition state calculation using cpcm model?

 

Any papers on this aspect are appreaciated.

 

Best wishes,

 

Zhijian Wu


Dear Dr. Zhijian Wu,
    You have performed the steps I'd recommend, i.e. using the gas-phase  
Hessian or the Hessian from an Onsager calculation. Since these options  
did not work, it might be necessary to calculate the CPCM Hessian. You  
cannot do this with "opt=(ts,calcfc)" in G98, but you can perform a  
two-step job in which the first step uses "freq" and the second job  
step uses "opt=(ts,rcfc)". You might also consider performing an  
"opt=qst3" calculation in which you use your Onsager model optimized TS  
as the initial guess structure for the CPCM transition state.
 
Regards,
Jim Hess
 
On Thursday, April 24, 2003, at 01:13 AM, zjwu wrote:
 
>>  Dear Dr. Hess,
>>
>>  ?/DIV>
>>
>>  I mean the job failing because the structure is not the TS structure  
>>  anymore. The calculation may converge but goes to different geometry  
>>  which I did not want. I have tried to read force constants at cpcm  
>>  calculation from either gas phase frequency calculation or Onsager  
>>  frequency calculation (these frequencies are obtained based on the  
>>  optimized geometry) (calculation using Onsager model works well for  
>>  the TS calculation).?n addition, I also tried opt=(ts,calcfc), but as  
>>  you know,?alcfc does not exist for?pcm.?t seems g03 is possible to  
>>  calculate fc at cpcm level.
>>
>>  ?/DIV>
>>
>>  Hope I have made myself clear.
>>
>>  ?/DIV>
>>
>>  Thank you very much for your answer.
>>
>>  ?/DIV>
>>
>>  Best regards,
>>
>>  ?/DIV>
>>
>>  Zhijian Wu
>>
>>  ?/DIV>
>>
>>
>>
<< image.tiff>
>>
>>  Dear?r.?hijian?u,
>>  ??Is?his?ob?ailing?ecause?t?uns?ut?f?ptimization?ycles??'d 
>>  ?/DIV>
>>  need?o?now?ore?etails?bout?hat?s?oing?rong?efore??ake?/DIV>
>>  detailed?uggestions,?ut?ne?ossible?ay?o?mprove?his?alculation?/DIV>
>>  is?o?alculate?he?orce?onstants?n?he?olution?hase,?hen?se?/DIV>
>>  these?orce?onstants?o?tart?he?ransition?tate?earch.>.?sually?he 
>>  ?/DIV>
>>  gas-phase?essian?s??ood?nough?pproximation,?ut?n?ome?ases?/DIV>
>>  where?he?as-phase?nd?olution-phase?otential?nergy?urfaces?re?/DIV>
>>  different?nough,?xplicitly?alculating?he?econd?erivatives?n?he?/DIV>
>>  solution?hase?an?e?elpful.?ncidentally,?f?ou?re?sing?98,?he?/DIV>
>>  CPCM?econd?erivatives?ill?e?erformed?y?umerical?ifferentiation?/DIV>
>>  of?he?nalytic?radients?nd?f?ou?re?sing?03,?he?econd?/DIV>
>>  derivatives?ill?e?alculated?nalytically.
>>  ?/DIV>
>>  Regards,
>>  Jim?ess
>>  ?/DIV>
>>  ++++++++++++++++++++++++++
>>  James?ess,?h.D.
>>  Customer?upport?cientist
>>  Gaussian,?nc.
>>  e-mail:?elp %! at !% gaussian.com
>>  ++++++++++++++++++++++++++
>>  ?/DIV>
>>  ?/DIV>
>>  On?uesday,?pril?2,?003,?t?8:29?M,?jwu?rote:
>>  ?/DIV>
>>  > >  ?ear?embers,
>>  > >
>>  > >  ?/DIV>
>>  > >
>>  > >  ??m?rying?o?ocate?he?ransition?tate?sing?pcm?odel.?he?/DIV>
>>  > >  ?nitial?tructure?s?rom?as?hase?ptimized?ransition?tructure?/DIV>
>>  > >   
>>  ?with?orrect?ibrational?ode).?he?ptimization?t?pcm?evel?ails?/DIV>
>>  > >  ?y?he?ollowing?eywords
>>  > >
>>  > >  ?/DIV>
>>  > >
>>  > >   
>>  ?b3lyp/6-31g(d,p)?pt=(ts,?eadfc,?oeigen)?crf=(cpcm,solvent=water)
>>  > >
>>  > >  ?/DIV>
>>  > >
>>  > >  ?o?nybody?as?xperience?n?pcm?ransition?tructure?alculation?
>>  > >
>>  > >  ?/DIV>
>>  > >
>>  > >  ?hanks??ot.
>>  > >
>>  > >  ?/DIV>
>>  > >
>>  > >  ?est?egards,
>>  > >
>>  > >  ?/DIV>
>>  > >
>>  > >  ?hijian?u
>>  > >
Dear Dr. Zhijian Wu,
 
   You have performed the steps I'd recommend, i.e. using the
gas-phase Hessian or the Hessian from an Onsager calculation. Since
these options did not work, it might be necessary to calculate the
CPCM Hessian. You cannot do this with "opt=(ts,calcfc)" in G98, but
you can perform a two-step job in which the first step uses "freq" and
the second job step uses "opt=(ts,rcfc)". You might also consider
performing an "opt=qst3" calculation in which you use your Onsager
model optimized TS as the initial guess structure for the CPCM
transition state. 
 
 
Regards,
 
Jim Hess
 
 
On Thursday, April 24, 2003, at 01:13 AM, zjwu wrote:
 
 
<< excerpt> Dear Dr. Hess,
 
 
?/DIV>
 
 
I mean the job failing because the structure is not the TS structure
anymore. The calculation may converge but goes to different geometry
which I did not want. I have tried to read force constants at cpcm
calculation from either gas phase frequency calculation or Onsager
frequency calculation (these frequencies are obtained based on the
optimized geometry) (calculation using Onsager model works well for
the TS calculation).?n addition, I also tried opt=(ts,calcfc), but as
you know,?alcfc does not exist for?pcm.?t seems g03 is possible to
calculate fc at cpcm level.
 
 
?/DIV>
 
 
Hope I have made myself clear.
 
 
?/DIV>
 
 
Thank you very much for your answer.
 
 
?/DIV>
 
 
Best regards,
 
 
?/DIV>
 
 
Zhijian Wu
 
 
?/DIV>
 
 
 
 
<< /excerpt> < < image.tiff>
 
<< excerpt>
 
Dear?r.?hijian?u,
 
??Is?his?ob?ailing?ecause?t?uns?ut?f?ptimization?ycles??'d?/DIV>
 
need?o?now?ore?etails?bout?hat?s?oing?rong?efore??ake?/DIV>
 
detailed?uggestions,?ut?ne?ossible?ay?o?mprove?his?alculation?/DIV>
 
is?o?alculate?he?orce?onstants?n?he?olution?hase,?hen?se?/DIV>
 
these?orce?onstants?o?tart?he?ransition?tate?earch..?sually?he?/DIV>
 
gas-phase?essian?s??ood?nough?pproximation,?ut?n?ome?ases?/DIV>
 
where?he?as-phase?nd?olution-phase?otential?nergy?urfaces?re?/DIV>
 
different?nough,?xplicitly?alculating?he?econd?erivatives?n?he?/DIV>
 
solution?hase?an?e?elpful.?ncidentally,?f?ou?re?sing?98,?he?/DIV>
 
CPCM?econd?erivatives?ill?e?erformed?y?umerical?ifferentiation?/DIV>
 
of?he?nalytic?radients?nd?f?ou?re?sing?03,?he?econd?/DIV>
 
derivatives?ill?e?alculated?nalytically.
 
?/DIV>
 
Regards,
 
Jim?ess
 
?/DIV>
 
++++++++++++++++++++++++++
 
James?ess,?h.D.
 
Customer?upport?cientist
 
Gaussian,?nc.
 
e-mail:?elp %! at !% gaussian.com
 
++++++++++++++++++++++++++
 
?/DIV>
 
?/DIV>
 
On?uesday,?pril?2,?003,?t?8:29?M,?jwu?rote:
 
?/DIV>
 
>> >  ?ear?embers,
 
>> >
 
>> >  ?/DIV>
 
>> >
 
>> >  ??m?rying?o?ocate?he?ransition?tate?sing?pcm?odel.?he?/DIV>
 
>> >  ?nitial?tructure?s?rom?as?hase?ptimized?ransition?tructure?/DIV>
 
>> >
?with?orrect?ibrational?ode).?he?ptimization?t?pcm?evel?ails?/DIV>
 
>> >  ?y?he?ollowing?eywords
 
>> >
 
>> >  ?/DIV>
 
>> >
 
>> >
?b3lyp/6-31g(d,p)?pt=(ts,?eadfc,?oeigen)?crf=(cpcm,solvent=water)
 
>> >
 
>> >  ?/DIV>
 
>> >
 
>> >  ?o?nybody?as?xperience?n?pcm?ransition?tructure?alculation?
 
>> >
 
>> >  ?/DIV>
 
>> >
 
>> >  ?hanks??ot.
 
>> >
 
>> >  ?/DIV>
 
>> >
 
>> >  ?est?egards,
 
>> >
 
>> >  ?/DIV>
 
>> >
 
>> >  ?hijian?u
 
>> >
 
<< /excerpt>