CCL:G: IRC-TS-PES

 Sent to CCL by: Dan Maftei [dan.maftei:uaic.ro]
 Dear Sujit,
 First, I'd recommend not to double-click the send button when writing to
 CCL. Your message will appear twice.
 1) Second, why optimizing the IRC geometries? It's pointless, really.
 Starting from those 21 geometries found after a significant
 computational effort for locating the TS then for performing the IRC,
 you'll end up with 2 final geometries, one corresponding to the
 reactants and the other to the products. To verify that TS corresponds
 to THAT reaction pathway, a visual inspection of the IRC geometries is
 usually enough. Even in the TS, after a frequency run, you should be
 able to predict the reaction course by inspecting the displacement
 vector which is the Hessian eigenvector corresponding to the Hessian
 negative eigenvalue (imaginary frequency).
 To extract, however, the geometries for each IRC step, you may find
 useful the GAUSSTOOL package (a pack of C-shell scripts written by Cu G.
 Phung,
 http://www.ccl.net/cca/software/UNIX/gaussian-utilities/README.shtml)
 on
 Linux/UNIX or, if using Windows, GaussView.
 2) Third (I mean second), if the job for finding the TS yielded ONE (and
 ONLY one) imaginary frequency, the final geometry calculated IS a
 transition state, at the level or theory you are doing the computations.
 I suggest to check you output file carefully, as the noeigentest
 parameter for the OPT keyword will make Gaussian to keep
 searching/optimizing even if no imaginary frequency (oe negative Hessian
 eigenvalue) was found or the Hessian has two or more imaginary negative
 eigenvalues. However, if you've been able to perform IRC calculations,
 the TS may be the one of interest...
 Also it is not uncommon after a geometry optimization to obtain a
 different geometry with respect to the initial one. Otherwise, why
 running geometry optimization?
 What did you mean by delocalization inside a ring? Your TS geometry does
 not display "aromatic" bonds inside the ring? This doesn't mean the
 chemistry of your geometry has changed! That's why I hate visualization
 software which displays multiple bonds. Any visualization program will
 display single, aromatic, double or triple bonds based on some values
 and tolerances hardcoded. If you give a ethane molecule as input, with
 about 2.0 angstroms between the carbon atoms, it will display as two CH3
 groups. It doesn't mean it's chemistry has changed. You've only
 elongated a bond. It will be displayed broken as the bond length is
 sufficiently larger than the one the program knows or has been told.
 Finally, I don't know if that's the case in your TS geometry, but if it
 is, you may use the above considerations.
 3) If looking towards PES scans and use Gaussian for that, first
 identify your coordinates of interest (those beings scanned) and
 "freeze" them to some equidistant values while optimizing (relaxing)
 the
 rest of the structure. The Gaussian keyword(s) for that is
 opt(z-matrix,...) or opt(modredundant,...). Both are well documented in
 the user guide (online, http://www.gaussian.com/g_tech/g_ur/k_opt.htm).
 Regards,
 Dan
 On 10/25/2011 10:43 PM, Sujit Kumar Mondal sujitharvard|,|gmail.com wrote:
 > Hi
 > I am working on Intrinsic Reaction Co-ordinate for hetero cycle organic
 > molecules and transition metal system. I performed IRC calculation which
 > produced 21 step of a reaction profile path, I would like to optimized
 > all the 21 structures obtained in IRC path including the transition
 > state individually.
 > 1) Can any one tell me the keywords for optimizing all the structures
 > obtained in IRC? I do not know how to do that, please help me out.
 >
 > 2) secondly, I tried to optimize my transition state using keywords
 > opt=(Calcfc,TS,noeigentest,maxcycle=200) Scf=(conver=5,maxcycle=200)
 > but I found that my optimized transition state has changed all the
 > atomic property including delocalization inside the 5membered ring
 > compare to the transition state I build up (as input file). Can anyone
 > tell me what should be done here?
 > i need help......
 >
 > 3) I would like to perform a potential energy scan, could someone tell
 > me the keywords for a potential energy scan of a particular system?
 >
 > Looking forward for your reply
 > Sujit
 >
 >
 --
 Dan Maftei,
 Assistant Professor,
 Faculty of Chemistry/Department of Chemistry
 University Alexandru Ioan Cuza Iasi
 Bd. Carol 1, Nr. 11, 700506 Iasi, Romania
 Tel: +40 232 201307
 E-mail(s):
 dan.maftei:chem.uaic.ro
 dan.maftei:uaic.ro