SUPRAMOLECULAR STRUCTURES FROM HIGH RESOLUTION POWDER DIFFRACTION. R. E. Dinnebier, Lehrstuhl fuer Kristllographie, University of Bayreuth, 95440 Bayreuth, Germany

Over the past few years, the feasibility of determining crystal structures ab initio from powder diffraction data has been steadily improved. Although a number of complicated inorganic crystal structures have been solved by this method, very little has been done in the field of supramolecular structure determination, namely for organic- and organometallic compounds. Assuming the material itself is well crystallized, the use of Synchrotron radiation is necessary to get a resolution as high as possible over the entire angular range of the powder pattern. Besides the higher resolution, advances in the computational aspects of the problem are also crucial for structure determination. Especially the development of more sophisticated grid search techniques of molecular fragments, considering geometrical and physical aspects of the crystal, is an important step forward in finding the right solution. The structure solutions presented here stand for some of the most complicated organic and organometallic structures which have ever been solved ab inito from high resolution powder data. They include the high and the low temperature phase of the Ru-sawhorse dimer [Ru₂(O₂PMe₂)₂CO)₄]n (1), the industrial important (Kolbe-Schmitt-synthesis) phenolates C₆H₅OA (A= K, Rb, Cs) (2), C₆H₅OK 2(C₆H₅OH) (2), C₆H₅OK 3(C₆H₅OH) (2), Na-Para-Hydroxy-Benzoate NaC₇H₅O₃ (2), base free alkaline-cyclo-pentadienide C₅H₅A (A=Li,Na,K,Rb) and the triclinic low temperature form of C₆₀Br₂₄(Br₂)₂ (3). In addition to well known techniques such as direct methods, difference Fourier synthesis, Patterson maps and Rietveld analysis, the newly proposed pseudo-atom method proved to be a very efficient tool to solve all structures containing well defined molecular fragments. In the case of C₆₀Br₂₄(Br₂)₂ the orientation of the well defined C₆₀Br₂₄ molecules in a triclinic distorted fcc lattice could be found unambiguously by maximizing of nearest neighbor distances. The structural motive of the high temperature form of the Ru-sawhorse dimer was found by conventional direct methods. The similarity criterion for the low temperature phase resulted in a restricted 4-dimensional grid search.

It can be shown that the precision is comparable to that achievable with single crystal techniques, and, therefore, allows for the interpretation of binding mechanism and reactions. Nevertheless, much more work is required in developing these structure solution methods further.

(1) Dinnebier, R.E. et al., NSLS annual report (1994), Beamline X3B1, Brookhaven National Laboratory

(2) Dinnebier, R.E. et al., NSLS annual report (1995), Beamline X3B1, Brookhaven National Laboratory

(3) Dinnebier, R.E. et al., J. Appl. Cryst. (1995), 28, 327-334