Combined Twinning and Disorder

2/5) Include twinning using SHELXL

In the last section we solved the structure, fixed the atom types, and applied a sensible numbering scheme. We then ran a few rounds of refinement, which resulted in the following (copied from the end of the previous section):

Twinning in uric acid dihydrate is by pseudo-merohedry, and occurs by virtue of the monoclinic β angle being very close to 90°. It is easy to deal with, just like the twinning tutorial. We simply add an appropriate TWIN instruction and include a guess for the fractional occupancy of the minor component as a BASF (that's BAtch Scale Factor). The BASF will refine to its optimum value, so your initial guess does not even need to be all that good. To prove this point, let's set it to something ridiculous (e.g., 0.8 or 80%) just to see what happens. Here's the *.res file (in a separate text editor, but you could just use the ShelXle text editor).

The TWIN and BASF commands are highlighted above. Go ahead and run this through SHELXL; the results look like this:

Note that the BASF parameter, which represents the fraction of the minor component, has refined down to about ~12%. Also note that the R-value has dropped from almost 14% to about 11%. That's still not great of course, but it's a real improvement. Here's what it looks like in ShelXle:

It doesn't look much different from the view prior to adding the twin operation. Even the difference map peaks are in pretty much the same positions as before. In the next section we'll use the SHELXL construct FRAG ... FEND to incorporate a whole-molecule minor component of disorder into the model.