Back to Table of Contents

45642-AC4
Improving Terephthalic Acid Synthesis in High-Temperature Water

Phillip E. Savage, University of Michigan

Introduction Terephthalic acid is a commodity chemical, and its main use is in the production of polyethylene terephthalate, a polymer used to make synthetic fibers and beverage containers. Terephthalic acid is produced commercially in acetic acid, and then purified by hydrogenation in high-temperature water. If water were used as the medium for the synthesis step, a simpler process may be possible because there would be no need to change solvents for the two steps. Moreover, water is a byproduct from the oxidation so the current process includes a tall distillation column for separating acetic acid, so it can be recycled, and water. If water were the reaction medium for terephthalic acid synthesis, the need for this column and its operating costs would be eliminated. The cost of this tower is significant both because of its size and its internals requiring titanium to withstand the corrosive conditions. Finally, the bromide catalysts used react with acetic acid to form small amounts of methyl bromide. This pollutant must be removed from the gaseous effluent stream. If water were the solvent, no methyl bromide would form so the process would be less polluting. There are both environmental and economic payoffs if terephthalic acid synthesis were to be practiced in water commercially. Background Previous work in our lab has demonstrated that high-temperature liquid water (HTW) at 300 °C is a very promising medium. High terephthalic acid yields (80 – 90%) and very high selectivities (essentially 100%) are available. Though the technical feasibility of a HTW process for terephthalic acid synthesis has been demonstrated, the reaction must be improved and further refined if it is going to displace the current acetic acid based process, which has benefited from years of continuous refinement and optimization. One barrier that accompanies the HTW process at present is the low p-xylene concentrations (about 0.02 mol/L) that have been used experimentally. Low p-xylene concentrations mean low production of terephthalic acid per unit reactor volume. Thus, there is a need to explore synthesis at higher concentrations of p-xylene to get greater productivity. Work to date on this project has focused on this objective of increasing productivity of terephthalic acid synthesis in HTW by running at high p-xylene concentrations. To work at higher reactant concentrations, though, involves running the reaction as a two-phase aqueous suspension since higher concentrations exceed the solubility of p-xylene in HTW. It is not known whether this two-phase processing scheme will be successful. Results We have conducted several sets of stirred, batch oxidation experiments at 300 °C with p-xylene concentrations higher than previously explored. Two experiments were done with p-xylene loadings ten times higher than those used in previous research. Figure 1 shows the composition of samples withdrawn periodically from the batch reactor during one of these oxidation reactions. Over the course of the two-hour run, the composition transitions from being largely the reactant (p-xyelene), as in sample number 1, to being largely the desired product (terephthalic acid (TPA)), as in sample number 11. The terephthalic acid selectivity in the last two samples (10 and 11) is about 93%, but products of incomplete oxidation (e.g., 4-CBA, toluic acid, tolualdehyde) are still present, so even higher selectivities appear to be feasible. At the end of the experiment we recovered and analyzed the solid material remaining in the reactor. This material was 98% terephthalic acid, and the balance was benzoic acid (a product from terephthalic acid decomposition). Thus, the terephthalic acid selectivity increased between the last sample withdrawn from the reactor prior to terminating the run and the final material remaining in the reactor after reactor cool-down, depressurization, and shut-down had occurred. Adding together the material recovered at the end of the run and the material removed during sampling, we can account for 92% of the moles of aromatic rings initially loaded into the reactor. Therefore, the terephthalic acid yield we demonstrated with this experiment is 90% (92% recovery times 98% selectivity). The results from this run indicate that terephthalic acid can be synthesized in high selectivity and high yield even at 10-fold higher p-xylene concentrations than used previously. The current commercial technology produces terephthalic acid in about 95% yield, however, so we plan to do additional experiments at this p-xylene loading in an attempt to demonstrate even higher yields, which would make HTW synthesis competitive with current technology. This future work will involve higher catalyst loadings.

Back to top