Calculating TS on psi4

1

Hello all. I’ve been having trouble converging my transition state structures on psi4 using mp2 level of theory.

My input code looks like this:

memory 3 gb

molecule TS {
0 2

geometry input here

}

set {
reference uhf
guess sad
opt_type ts
g_convergence gau_tight
scf_type df
basis cc-pvtz
df_basis_scf cc-pvtz-jkfit
df_basis_mp2 cc-pvtz-ri
freeze_core true
dynamic_level 3
}

optimize(‘mp2’)
frequencies(‘mp2’)

I’m dealing with a radical transition state structure that has been optimised using gaussian at ub3lyp/6-31g* level previously so I’m pretty sure it should not be a bad starting geometry.

Is there any way you guys can help me with this?

Thanks so much!

2

TS optimizations are notoriously tricky, without more detail of the system I’m not quite sure we can help. Perhaps @Rollin_King can say more.

3

Without more information (e.g., is a single-reference wavefunction reasonable here? are the forces small? have you tried computing the force constants prior to the optimization?) I can’t help.

The TS optimization code in Psi4 is not very sophisticated (PRFO steps) so you will need to begin the calculation in a region with the correct (or nearly correct) topology. Some amount of trial and error is likely.