ABSTRACT
One major method for producing terephthalic acid (1,4-dicarboxybenzene) is via the reaction of p-xylene and dioxygen in acetic acid. The catalysts used for this reaction are soluble forms of cobalt(II), manganese(II), and bromide salts. The number of elementary reactions that occur between bromide, Co(II), Co(III), Mn(II), Mn(III), acetic acid, and p-xylene have been characterized. These are used to rationalize why the Amoco MC (Mid-Century) process is more active and selective than all of the current alternative processes to make terephthalic acid.
Kinetic and thermodynamic arguments can be used to explain why p-xylene does not react with dioxygen. This problem is overcome by forming a highly unstable radical of p-xylene. The formation of this radical (the initiation step) can occur in acetic acid via the reaction of cobalt(II) acetate with dioxygen. This initiation sequence becomes greatly amplified via a propagation step with the p-xylene radicals. Using only cobalt as the catalyst produces very poor yields of terephthalic acid because of the electron-withdrawing nature of the carboxy group. The yield cannot be improved by use 322of high temperatures due to the decarboxylation of acetic acid by cobalt(III). This has led to alternative methods of terephthalic acid manufacture such as partial oxidation and esterification (Witten process), changing the cobalt catalyst and cooxidation (Eastman process). A large increase in activity and selectivity (at a given temperature) and an elimination of a thermal barrier occurs when bromide is added to cobalt. A series of reactions and their relative rates are proposed to explain why this occurs. The addition of manganese to cobalt and bromide results in increased activity and selectivity because of a rapid electron transfer reaction between cobalt(III) and manganese(II) and because manganese(III) decarboxylates the solvent at a rate much less than that of cobalt(II).
