Selective Oxidation of Ethylene to Ethylene Oxide on Silver Catalysts at Industrial Conditions: Reactor Profiles, Kinetics, and Chlorine Inhibition

Journal
Industrial & Engineering Chemistry Research  
Volume
63
Issue
9
Start Page
3891
End Page
3909
Citation
Industrial and Engineering Chemistry Research 63 (9): 3891-3909 (2024-03-06)
Publisher DOI
10.1021/acs.iecr.3c04345
Scopus ID
2-s2.0-85186183924
Publisher
American Chemical Society
ISSN
08885885
Selectivity is the key parameter in industrial ethylene oxide (EO) production by oxidation of ethylene with oxygen on Ag/α-Al2O3 catalysts. Accurate temperature control in wall-cooled multitubular fixed-bed reactors and chlorination of the silver surface by feeding small chlorinated hydrocarbons such as 1,2-dichloroethane (DCE) are required to fine-tune electrophilicity and surface oxygen coverage for maximum EO selectivity at economic ethylene conversion. Temperature and molar flow rate profiles of C2H4, O2, EO, CO2, H2O, DCE, and chlorine-containing reaction products vinyl chloride (VC) and ethyl chloride (EC) were measured in a compact profile reactor (CPR) and in a pilot-scale profile reactor (PSPR) to explore the spatial interplay between DCE concentration, temperature, inlet flow rate, and O2 conversion. Chlorine and oxygen compete for the same active silver sites despite more than 4 orders of magnitude different concentrations (ppm vs vol %). Chlorine coverage increases from inlet to outlet due to the decreasing partial pressure of O2 along the bed, leading to shutdown of all reactions if all active Ag sites are blocked by chlorine. A kinetic model is derived from a dual-site mechanism taken from the literature. Kinetic parameters are determined from differential initial rate measurements, Arrhenius plots, and by fitting the rate expressions implemented in a plug flow model to the species and temperature profiles in the CPR. A very good agreement is reached. PSPR profiles are modeled by implementing the derived kinetic model into a 2D pseudohomogeneous reactor model. At conversions <10%, the experimental profiles are well captured, but the model fails to accurately reproduce the point of thermal runaway in the catalyst bed of the PSPR caused by a too low reactor temperature and resulting insufficient chlorine coverage of the silver surface.