Step 6 - Set up Potential Refinement
Add the standard EPSR-style refinement layer to your simulation:
Layer ⇨ Create ⇨ Refinement ⇨ Standard EPSR
Select the
EPSR module if it isn’t already. We’ll need to set a relatively high starting value for EReq as we are basically dealing with point charges in a box, and so whatever empirical potential is necessary to drive the structure of the system needs to grow large enough to have some effect against the strong Coulombic forces present.
Go to the
Refine (EPSR) tab
If we run the simulation a bit longer:In go to the Control settings group and set EReq to
50.0
Simulation ⇨ Run
On the R-Factor tab of the
EPSR module you will see that the magnitude of the potential (top graph) increases rather quickly to 50 (within 100 steps in the plot below), but the quality of fit as measure by the r-factor (bottom graph) takes rather longer to reach a steady state.
Evolution of empirical potential magnitude (max = 50.0)
Evolution of r-factor (EReq max = 50.0)
The structure factor of the system at this point looks something like this:
Stable structure factor at EReq = 50.0
The question is, can it be better? The answer is, of course, yes - the intensity of the first strong peak in the F(Q) at 1.47 Å is a little low. The empirical potential will eventually sort this out (perhaps with an increase of EReq to 75 or so) but it takes time to get it exactly right.
What does the peak at Q = 1.47 Å correspond to, and why might it be difficult to reproduce?
For the purposes of the example, let’s move on and calculate something from the (semi-) refined structure.
First, stop the simulation:
Esc