UT theoretical chemistry code forum

Dear Graeme,

I am trying a rather standard TS search of CH4 on Ni(111), I know you've tried it and published. I must admit that I am not in a good grasp of how to tweak parameters (or input coordinates, for that matter) to obtain the converged TS search, such as this one, most of the times it's hit and miss (most of the times it's a miss). Probably because I am an eigenvector TS guy using other sofwtare packages so NEB is somewhat new to me.

So, back to my calculations, they run out of 500 cycles. I was wondering if you could provide some suggestions on how to prepare TS like that so I could achieve succesfull TS optimization.

See unsuccessful run attached

thanks

Jonas

The calculation attached looks like it reached the specified force criterion (of 0.02 eV/Ang) in 399 iterations.

You can probably play with optimizer parameters and save a little time, but I expect what could save you a lot of time is to relax your band at a lower level of accuracy (e.g. a 2x2x1 point mesh, ENCUT=270). From a converged band, it will not be much work to reconverge when you increase the k-point mesh and cutoff.

So specifically, I relax convergence criteria, locate TS, then I take what for the restart? the resulting 01 through 08 CONCTAR files and restart with them? thanks
Jonas

yes