Optimization of continuous flow synthesis of fluoropolyimide monomer via advanced real-time process analytics

This study demonstrates the integration of two complementary process analytical technology tools (inline UV/Vis and FTIR spectroscopy) within a continuous flow reactor for the multistep synthesis of the fluoropolyimide monomer 2,2′-bis(3-amino-4-hydroxyphenyl)hexafluoropropane (6FAP). Advanced data analysis models, including partial least squares regression (PLS) and artificial neural networks (ANN), were developed to enable real-time quantitative analysis of both target and intermediate species throughout the process. The synergy of process analytics and machine learning algorithms allowed for rapid optimization of reaction conditions and elucidation of reaction pathways, particularly for multifunctional compounds in homogeneous nitration and heterogeneous hydrogenation steps. This approach achieved yields of 95 % for the nitro intermediate 6FNP and up to 94 % for the final amine product 6FAP. These results underscore the transformative potential of integrating advanced process analytics and data-driven models into the continuous flow synthesis of high-value monomers.