[molpro-user] Optimisation to a minimum

[Jayashree]

Thanks for the replies. I apparently did have my dihedral angles fixed,
which I have been keeping throughout, and never thought of changing. Here is
my input for locating a Cs minimum from a C3v geometry by doing an
optimisation with casscf. Since I would like to locate a Cs minimum, I am
keeping two of the bond lengths and two bond angles fixed. My guess for
casscf here is C3v geometry wavefunction. The only thing is, though the
following run gives me all positive frequencies, it gives a warning -
"Gradient norm not zero!!" Should I be concerned?

 basis=6-311++G
 geomtype={zmat}
 geometry={Angstrom,nosym;
          c;
          o,c,oc;
          h1,c,h1c,o,h1co;
          h2,c,h2c,o,h2co,h1,dih1;
          h3,c,h2c,o,h2co,h1,dih2}
 oc=1.475858059,
 h1c=1.078404519,
 h2c=1.078404519,
 h1co=108.48495694,
 h2co=108.48495694,
 dih1=120.0
 dih2=240.0
{casscf;
   occ,12;
   closed,6;
   wf,17,1,1;state,1
   CPMCSCF,GRAD,1.1}
{optg,root=1,startcmd=casscf,space=zmat}
{frequencies}


On Fri, Oct 24, 2008 at 9:45 AM, Kirk Peterson <kipeters at wsu.edu> wrote:

> Jayashree,
>
> do you have any symmetry restrictions in your z-matrix input (2 or bond
> lengths set equal to each other, dihedral angles fixed to 180, etc.)?  If
> you optimize under symmetry constraints you will not locate the global
> minimum if it has lower symmetry.
>
> just a guess,
>
> -Kirk
>
> On Oct 23, 2008, at 3:42 PM, Jayashree wrote:
>
>  I am trying to optimise to a minimum geometry, which means that I should
>> get all positive frequencies for the normal modes. By default, I know that
>> MOLPRO does a search for a minimum. But in my output, even though it has
>> completed the optimisation in 5 steps, it still gives me one imaginary
>> frequency in the search for a minimum. I am not sure what is wrong, since
>> the optimisation seems to be complete, yet its unable to locate a minimum.
>> Any input/suggestion would be helpful..
>>
>> --
>> Jayashree
>> _______________________________________________
>> Molpro-user mailing list
>> Molpro-user at molpro.net
>> http://www.molpro.net/mailman/listinfo/molpro-user
>>
>
>


-- 
Jayashree
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20081024/e943c036/attachment.html>