Molecular Catalyst Synthesis Strategies to Prepare Atomically Dispersed Fe-N-C Heterogeneous Catalysts

Abstract

We report a strategy to integrate atomically dispersed iron within a heterogeneous nitrogen-doped carbon (N-C) support, inspired by routes for metalation of molecular macrocyclic iron complexes. The N-C support, derived from pyrolysis of a ZIF-8 metal-organic framework, is metalated via solution-phase reaction with FeCl2 and tributyl amine, as a Brønsted base, at 150 °C. Fe active sites are characterized by 57Fe Mössbauer spectroscopy and aberration-corrected scanning transmission electron microscopy. The site density can be increased by selective removal of Zn2+ ions from the N-C support prior to metalation, resembling the transmetalation strategy commonly employed for the preparation of molecular Fe-macrocycles. The utility of this approach is validated by the higher catalytic rates (per total Fe) of these materials relative to established Fe-N-C catalysts, benchmarked using an aerobic oxidation reaction.

Development and Validation of a Deep Learning Model for Detection of Allergic Reactions Using Safety Event Reports Across Hospitals

“…4 Patients in health care settings are at particularly high risk for developing an allergic reaction given their many new exposures. [5][6][7] Given that allergic reactions can cause patient harm 8 and result in malpractice litigation, [9][10][11] timely allergic event detection, monitoring, and characterization are critical for improving health care quality and patient safety. 6,12 Hospital safety event reporting systems, which collect voluntarily reported safety event data from frontline personnel, are integral to the detection of patient safety signals in health care.…”

“…12 Among the 9107 reports retrieved by keywords, only 431 reports (4.7%) were confirmed as true allergic reactions, and it was unclear how many cases were missed. 12 Machine learning studies have detected adverse drug reactions (ADRs) from electronic health records, 5,20,21 safety reports, 22 and social media data, 23 but few studies have focused on allergic reaction identification.…”