[noone special]

hi
when sb use Gaussian for optimizing the molecule just 4 yes answer in the last cycle means the structure is fully optimized? or it's not enough???

[Jorriss]

Uhh, the wording of your question is fairly confusing... but a job completed successfully if you write 'tail jobname.log' and at the end it says Job completed successfully or something. This doesn't mean it found the global minima though, just a local.

[vex]

The best way to check to see if a job is optimized is to make a plot of the potential energy at each step. If you see the line leveling off, you've found the local minimum; but, as the above poster said, that does not mean that your structure is in its lowest-energy conformation. Finding this "global energy minimum" can get pretty complicated.

[noone special]

thank you for your answers but is it possible to explain me a little easier  I am not a physical chemist  so I don't know how I can draw plot of the potential energy???
and if completing the job means only a local minima how I can obtain the global?
do I have to check the negative frequencies?

[vex]

1) There are several ways to get the potential energy out, but I haven't done a calculation in a while, and without knowing what operating system/software you have available I can't really give you any helpful advice. If there's a professor or graduate student around where you are that does these calculations, they should be able to take three minutes and help you out.

2) Obtaining the global minimum can be a challenging task that takes months of work, and it depends on the system that you're optimizing. (What is the system that you're optimizing?)

3) You should always check negative frequencies when you do vibrational analysis. A negative frequency is an obvious indication of an impossible solution to the vibrational modes. This usually comes about as a result of a structure that has not been totally optimized. However, if you change your basis set in-between the geometry optimization and the vibrational analysis (which sometimes is done on purpose) you will get imaginary frequencies. If you've used "opt freq" as keywords, as you've probably done, you should not get negative frequencies.

[noone special]

vex thank you very much for your help, my system is pentium4 and I use win xp with gaussian 98 A6 revision. I use ab initio basis set (hf/6-31g) for my calculations, with this command in route section:   #  HF/6-31G   OPT
I always thaught when I see four yes answer in my out it means that my molecule has been optimized but I have seen some notes in my friends computational note books that shows me I'm wrong.