Advancing the Compositional Analysis of Olefin Polymerization Catalysts with High-Throughput Fluorescence Microscopy

Abstract

To optimize the performance of supported olefin polymerization catalysts, novel methodologies are required to evaluate the composition, structure, and morphology of both pristine and prepolymerized samples in a resource-efficient, high-throughput manner. Here, we report on a unique combination of laboratory-based confocal fluorescence microscopy and advanced image processing that allowed us to quantitatively assess support fragmentation in a large number of autofluorescent metallocene-based catalyst particles. Using this approach, significant inter- and intraparticle heterogeneities were detected and quantified in a representative number of prepolymerized catalyst particles (2D: ≥135, 3D: 40). The heterogeneity that was observed over several stages of slurry-phase ethylene polymerization (10 bar) is primarily attributed to the catalyst particles' diverse support structures and to the inhomogeneities in the metallocene distribution. From a mechanistic point of view, the 2D and 3D analyses revealed extensive contributions from a layer-by-layer fragmentation mechanism in synergy with a less pronounced sectioning mechanism. A significant number of catalyst particles were also found to display limited support fragmentation at the onset of the reaction (i.e., at lower polymer yields). This delay in activity or "dormancy" is believed to contribute to a broadening of the particle size distribution during the early stages of polymerization. 2D and 3D catalyst screening via confocal fluorescence microscopy represents an accessible and fast approach to characterize the structure of heterogeneous catalysts and assess the distribution of their fluorescent components and reaction products. The automation of both image segmentation and postprocessing with machine learning can yield a powerful diagnostic tool for future research as well as quality control on industrial catalysts.