Anomalous X-ray diffraction (AXD) and neutron diffraction can be used to crystallographically distinguish between metals of similar electron density. Despite the use of AXD for structural characterization in mixed metal clusters, there are no benchmark studies evaluating the accuracy of AXD toward assessing elemental occupancy in molecules with comparisons with what is determined via neutron diffraction. We collected resonant diffraction data on several homo and heterometallic clusters and refined their anomalous scattering components to determine metal site occupancies. Theoretical resonant scattering terms for Fe0, Co0, and Zn0 were compared against experimental values, revealing theoretical values are ill-suited to serve as references for occupancy determination. The cluster featuring distinct cation and anion metal compositions [CoCp2*][(tbsL)Fe3(μ3-NAr)] was used to assess the accuracy of different f' references for occupancy determination (f'theoretical ± 15-17%; f'experimental ± 10%). This methodology was applied toward calculating the occupancy of three different clusters: (tbsL)Fe2Zn(py) (6), (tbsL)Fe2Zn(μ3-NAr)(py) (7), and [CoCp*2][(tbsL)Fe2Zn(μ3-NAr)] (8). The first two clusters maintain 100% Fe/Zn site isolation, whereas 8 showed metal mixing within the sites. The large crystal size of 8 enabled collection of neutron diffraction data which was compared against the results found with AXD. The ability of AXD to replicate the metal occupancies as determined by neutron diffraction supports the AXD occupancy methodology developed herein. Furthermore, the advantages innate to AXD (e.g., smaller crystal sizes, shorter collection times, and greater availability of synchrotron resources) versus neutron diffraction further support the need for its development as a standard technique.
Read full abstract