Pseudo-tetragonal twinning

6/6: Complete four-fold P2/n twin refinement

All that is left is to refine the structure against the 'HKLF 5' format data. For this, a few changes are needed in the res file from ADDSYM. In particular, the HKLF command must be 'HKLF 5', without a transformation matrix because the data are now indexed for P2/n. Other changes in the file below include 'MERG 0', 'BASF 0.25 0.25 0.25', etc.

Subsequent refinement is completely stable, no restraints needed. Nevertheless, even at the end, there's some ambiguity over hydroxyl and water hydrogen positions. In a structure with such complex twinning, that's not too surprising. With a sensible numbering scheme and after a few rounds of refinement etc., the structure looks like this:

Here's a view of the packing down the pseudo-tetragonal twin axis:

Once the twinning and proper space-group symmetry are locked down, completing the structure is straightforward.