# SUMMARY: Transition State Optimization

*From*: Roy Jensen <royj -AatT- uvic.ca>
*Subject*: SUMMARY: Transition State Optimization
*Date*: Wed, 07 Feb 2001 21:35:56 -0800

I wont summarize the individual responses, they were all the same:
1. Use the NOEIGENTEST keyword.
2. You have to be close to the TS for this to converge in 20
iterations. Look at the output file, manually adjust the geometry, and
try again.
3. Once the geometry is very close to the TS, removing the NOEIGENTEST
keyword _might_ work. If it does, the job will converge faster.
In summary, it is not a one-step job -- you have to work at it.
NOTE: I did try the NOEIGENTEST keyword before posting my question,
but the job did not optimize. As stated above, I was too far from the
transition state.
Thanks to all,
Roy Jensen
ORIGINAL POST
=============
> I am attempting a transition state optimization on G98W A.9, currently on
> HOCN. Using a multitude of methods, basis sets, and starting geometries,
> I fail to get a negative eigenvalue after the first step. The exact error:
> -- Wrong number of Negative eigenvalues: Desired= 1 Actual= 0
>
> Has anyone experienced this problem?
> An input file is copied below.
>
> Roy Jensen