RESUMO
Data were collected from a crystal of CoZnPO-CZP {sodium cobalt-zinc phosphate hydrate, Na(6)[Co(0.2)Zn(0.8)PO(4)](6).6H(2)O} using synchrotron radiation at ELETTRA at the inflection point and 'white line' for both the cobalt and zinc K edges, and at 1.45 Å, a wavelength remote from the K edges of both metals. The data were processed using the programs DENZO and SCALEPACK. The CCP4 program suite was used for the scaling of data sets and the subsequent calculation of dispersive difference Fourier maps. Optimal scaling was achieved by using a subset of reflections with little or no contribution from the metal atoms (i.e. which were essentially wavelength independent in their intensities) and using weights based on the sigma's to obtain an overall scale factor in each case. Phases were calculated with SHELXL97 based on the refined structure using a much higher resolution and complete Cu Kalpha data set. An occupancy of 100% by zinc at the two metal-atom sites was assumed. The dispersive difference Fourier map calculated for zinc gave two peaks above the background of similar heights at the expected metal-atom sites. The peak height at the Zn1 site was a little higher than at the Zn2 site. The dispersive difference Fourier map calculated for cobalt gave just one peak above the background, at the Zn1 site, and only a small peak at the Zn2 site, thus indicating that incorporation of cobalt takes place mainly at one site. Refinement of the zinc occupancies using MLPHARE reinforces this conclusion. The chemical environment of each site is discussed.