Buy article online - an online subscription or single-article purchase is required to access this article.
research papers
New methods for the determination of site occupancy factors are described. The methods are based on the analysis of differences between intensities of Friedel reflections in noncentrosymmetric crystals. In the first method (Anomalous-Expert) the site occupancy factor is determined by the condition that it is identical for two data sets: (1) initial data without averaging of Friedel intensities and (2) data that are averaged on Friedel pairs after the reduction of the anomalous scattering contribution. In the second method (anomalous anisotropic intermeasurement minimization method, Anomalous-AniMMM) the site occupancy factor is refined to satisfy the condition that the differences between the intensities of Friedel reflections that are reduced on the anomalous scattering contribution must be minimal. The methods were checked for three samples of RbTi1-xZrxOPO4 crystals (A, B and C) with KTiOPO4 structure, at 295 and 105 K (five experimental data sets). Microprobe measurements yield compositions xA,B = 0.034 (5) and xC = 0.022 (4). The corresponding site occupancy factors are QA,B = 0.932 (10) and QC = 0.956 (8). Using Anomalous-AniMMM and three independent refinements for the first and second samples, the initial occupancy factor of QA,B = 0.963 (15) was improved to QA,B = 0.938 (7). Of the three room-temperature data sets, one was improved to QA,B = 0.934 (2). For the third sample and one data set, the initial occupancy factor of QC = 1.000 was improved to QC = 0.956 (1). The methods improve the Hirshfeld rigid-bond test. It is discussed how the description of chemical bonding influences the site occupancy factor.