Hints on using Crystals to refine a higher symmetry framework (highly ordered host network) containing a lower symmetry guest molecule From David Watkin: Date: Thu, 8 Nov 2001 08:51:34 +0000 (GMT) From: David Watkin [david.watkin@chemistry.oxford.ac.uk] To: Lachlan Cranswick [l.m.d.cranswick@dl.ac.uk] Subject: RE: Crystals Request Hello Tis is not a problem. You enter the space group of lowest symmetry for the whole structure, then apply addtional constraints/restraints to the higher symmetry part. eg, for 2 atoms on opposite sides of a pseudo centre 1) Ensure that the starting values for the parameters are as you would expect (since non-linear LSQ only computes increments which are applied to the refernced paramters) 2) set up constraints. First ensure that the same shifts are applied to the x,y,z parameters pairwise RIDE c(1,x's) c(101,x's) Then make sure that the inverse shifts are applied to one of the atoms WEIGHT -1 c(101,x's) If you want to do whole groups LINK C(1,X'S) UNTIL C(6) AND C(101,X'S) UNTIL C(106) WEIGHT -1 C(101,X'S) UNTIL C(106) Alternatively, it can be done via restraints, eg RESTRAIN 0.0, 0.001 = C(1,X) + C(101,X) RESTRAIN 0.0, 0.001 = C)1,Y) + C(101,Y) CONSTRAINTS AND RESTRAINs can be mixed, though it is up to the user to ensure they make sense. When the CONSTRAINT list is processes, CRYSTALS checks to see if any of the atoms in the structure are subject to spacegroup symmetry restrictions. If they are, and they are also a subject to a user CONSTRAINT, the space group symmetry is applied as an additional RESTRAINT, otherwise an additional CONSTRAINT is applied Hope this helps Best wishes David