Kinetic modeling of 1,2‐dichloropropane (PDC) free‐radical chlorination

Abstract

Recent government mandates have lowered the permissible global warming potential (GWP) for refrigerants in mobile air conditioning substantially below that of the hydrofluorocarbon products that are used currently. Potential replacements, hydrofluoro‐olefins (HFO), have a reduced impact on the ozone layer and lower GWP. Many desirable HFO compounds, such as HFO‐1234yf, can be produced utilizing chlorocarbons as feedstocks such as the preferred 1,1,2,3‐tetrachloropropene (TCPE). TCPE can be produced by several routes; however, producing TCPE from 1,2‐dichloropropane (PDC) is potentially more desirable environmentally and economically since PDC is a byproduct of propylene oxide and allyl chloride production. One process option is to convert PDC to pentachloropropane (PCP) intermediates by chlorination, followed by dehydrochlorination of the PCPs to produce TCPE. In this work, we show that PCPs can be produced through the chlorination of PDC in a free‐radical liquid phase reaction and have developed a kinetic model for PDC chlorination based on the relevant free radical elementary reactions. Thermodynamic properties including standard heats of formation, standard entropies of formation, and heat capacities for the radical and non‐radical species were estimated by using ab initio and COSMOtherm calculations and validated against available experimental data. The reaction equilibrium constants were determined from the Gibb's free energies of the reactants and products. Phase equilibria were calculated by means of a consistent set of thermodynamic properties of the species. In addition, physical properties such as the vapor pressure of pure components involved in the reaction network were also estimated. Ab initio transition state calculations were employed to estimate the rate parameters including pre‐exponential factors and activation energies for the relevant reactions. The activation energies of some key reactions were then adjusted to match experimental data. The resulting kinetic model provided a basis for process yield optimization and scale up. © 2016 American Institute of Chemical Engineers AIChE J, 62: 1174–1191, 2016