reached required accuracy - stopping

Vasp transition state theory tools

Moderator: moderators

Post Reply
Posts: 24
Joined: Mon Jul 03, 2017 9:40 am

reached required accuracy - stopping

Post by mmg016 »

Dear all,

I am relatively new to NEB calculations and I am trying to find the transition sate of a simple reaction such as CH2 --> CH + H on a Ni(100) slab.

After very well optimizing the initial and final states, I have made 10 images between initial and final states. As a first step of NEB, I am using a fairly loose criteria for convergence of this calculation with the intention of using more stringent convergence criteria for 5 additional images around the left and the right of the saddle point resulting from this step. My exact setup in the incar file is:

EDIFF = 1e-6
POTIM = 0.01
NSW = 500
EDIFFG = -0.1

As can be seen from the above, I have set a criteria of max force of 0.1 ev/A for convergence.

My above calculation ends with a seemingly positive result in OUTCARS and with the line "reached required accuracy - stopping structural energy minimisation" at the end of the stdout of vasp job.

However, when I look at the forces and energies of the individual images with the script '', I see that the forces on many images have not still reached the convergence criteria of 0.1 ev/AA. I am attaching the vaspout plots for each image. I am also attaching the vaspjob stdout 'vasp.out' file.

Has anybody else experienced this kind of 'false-positive' result in the past?

I would greatly appreciate it if you can share your thoughts on why this is occuring and how to overcome this.

Best regards,
(320.4 KiB) Downloaded 516 times
(420.05 KiB) Downloaded 492 times
Site Admin
Posts: 1999
Joined: Tue Apr 26, 2005 4:25 am

Re: reached required accuracy - stopping

Post by graeme »

The true forces on the NEB images will not, in general, drop below EDIFFG. Remember, the images along a minimum energy path climb up over a barrier. It is the NEB forces that matter for convergence. You can find these in the OUTCAR. The NEB forces are calculated as the true forces with the component along the path projected out. There is also a spring force which keeps them equidistance, but this is typically not the dominant component.
Post Reply