Transmission powder X-ray diffraction technique for precise identification and quantification of drug polymorphs – a case study using metformin embonate

In this study, we present the results from the powder X-ray diffraction (PXRD) analyses conducted on two polymorphic samples of the antidiabetic drug metformin embonate (ME forms I and II), employing three distinct geometries: capillary transmission, foil transmission, and Bragg-Brentano reflection. Our objective is to determine which geometry is the most straightforward and effective for accurate phase identification and quantification of active pharmaceutical ingredients using metformin embonate as a model example. We assess the quality of the diffraction patterns of the two polymorphs across the three different geometries by comparing them to the reference patterns. The diffraction peaks in the foil and capillary transmission data are symmetric and well resolved, whereas peaks in the reflection data are broader and merged, with inherent peak asymmetry towards lower angles. Rietveld refinement was performed to evaluate the quality of data generated by these three geometries. The capillary transmission geometry yielded the best profile fit with a lower residual factor (<i>R</i>_wp) and goodness of fit (GOF), followed by the foil transmission and reflection geometries. Consequently, the foil transmission geometry can be viewed as a bridge-configuration that combines the advantages of both capillary transmission and reflection techniques, while addressing several challenges related to sample preparation, data quality, sample homogeneity, preferred orientation effects, <i>etc.</i> To further emphasize the significance of the foil transmission method and its utility for phase quantification studies compared to the commonly used reflection method, we have conducted PXRD analysis on polymorph mixtures (ranging from 5 to 95% of form I in II) and analyzed the results using both the single peak method and Rietveld refinement. The foil transmission method demonstrated superior profile fitting compared to the reflection data, exhibiting excellent linearity between the predicted and experimental compositions. In particular, when dealing with heterogeneous samples that exhibit preferred orientation, the foil transmission geometry yields better results compared to reflection geometry. When the polymorphic samples are homogeneous and lack preferred orientation, both reflection and transmission geometries are however effective and yield satisfactory results. To the best of our knowledge, this is the first systematic study to compare reflection and transmission geometries, while also demonstrating the effectiveness of the foil transmission method for the polymorphic identification and quantification of an active pharmaceutical ingredient. The study also offers suggestions for polymorph quantification and holds relevance for the pharmaceutical industry.