UT theoretical chemistry code forum

Hi,

I wonder how the dimer separation is defined (or maybe rather how it is used). Is it right that to get the coordinates of the two images in the dimer are

IMAGE1 = POSCAR +.5*DdR*MODECAR
IMAGE2 = POSCAR - .5*DdR*MODECAR

where DdR is the dimer separation? MODECAR is suppose to be a unit vector, right? Does this mean that for a system where a relatively large number of atoms are involved in the transition you need a larger dimer separation than for a system with a relatively small number of atoms taking part in the transition, in order to have similar separations between the respective atoms in the two images?

Thankful for answers.

/Jonas

In vasp, the definition is:

IMAGE1 = POSCAR + DdR*MODECAR
IMAGE2 = POSCAR - DdR*MODECAR

The MODECAR is a unit vector. As long as your processes are local, there is no need for a larger dimer separation between the dimer images. If all atoms are moving, perhaps a larger DdR would be required. But using a small value of ediff=1e-8 should let you use even very small displacements (per atom) of on the order of 0.001 Ang.