A SPRAY REACTOR CONCEPT FOR CATALYTIC OXIDATION OF P-XYLENE TO PRODUCE HIGH-PURITY TEREPHTHALIC ACID

Abstract

Terephthalic acid (TPA), with current annual world capacity of exceeding 50 million metric tons, is a commercially important chemical used primarily in the manufacture of polyesters. A spray reactor in which the liquid phase, containing dissolved p-xylene (pX) and the catalyst (Co/Mn/Br), is dispersed as fine droplets by a nozzle into a continuous vapor phase containing the oxidant (O2) is shown to produce high-purity TPA with less than 25 ppm 4-carboxybenzaldehyde (4-CBA) in the solid TPA product. In sharp contrast, the solid TPA product obtained from a conventional stirred reactor similar to the configuration used in the conventional Mid-Century (MC) process contains nearly 1000 ppm 4-CBA even though the reactor is operated at similar pressure and temperature (15 bar and 200 °C) but with the gas phase dispersed into the liquid phase. The dramatic improvement in TPA product quality during spray reactor operation is attributed to two main factors: the alleviation of interphase gas-liquid mass transfer limitations that facilitates more complete oxidation of the pX and the intermediate oxidation products to TPA, and reduced backmixing that enhances the oxidation rates. Kinetic studies of pX oxidation to TPA performed in a well-stirred 50 mL reactor confirm that the intermediate oxidation steps are subject to mass transfer limitations even at the highest rpm used. Theoretical calculations show that the time constants for O2 diffusion in typical spray droplets (assumed to be 50 μm diameter) are one to two orders of magnitude lower than the kinetic rate constant confirming complete O2 penetration and saturation of the droplets. Gas phase concentration measurements show that in the spray reactor gas phase CO formation is roughly one-fourth of that in the MC process, indicative of solvent burning. This decrease is attributed to the shorter residence times in the spray reactor. Further, the usage of CO2 as an inert gas and the dominance of acetic acid (50 mol%) in the vapor phase under reaction conditions create a gas phase environment that falls outside of the flammability envelope. Mathematical modeling of the stirred reactor using MC process conditions accurately predicts the steady state temperatures observed in industrial reactors (195 °C). The model also clearly divulges that the cooling provided by partial evaporation of the acetic acid solvent, upon absorbing the heat of reaction at the set reactor pressure, is vital to maintain stable steady state operation. Experimental results clearly attest to the significance of reliable pressure control to prevent undesired temperature rises. Comparative economic analyses and gate-to-gate and cradle-to-gate life cycle assessments show that the spray process significantly reduces capital and operating costs by 55% and 16% respectively, and also imposes less adverse environmental impacts than the MC process. These benefits of the CEBC spray process are mainly derived from the non-requirement of the hydrogenation step required in the conventional process for purifying the crude TPA. Thus, the spray reactor concept has the potential to be a greener and more sustainable process for making polymer-grade dicarboxylic acids in one step. The results from this dissertation provide valuable guidance for the rational design and development of a continuous spray reactor.