Epoxidation of propylene with hydrogen peroxide (HP) is regarded as an attractive route to propylene oxide (PO) since it does not produce byproducts as conventional PO technologies do, but relatively high market prices, limited supply and toxic nature of the HP have been an obstacle to the route’s commercial acceptance. However, recent developing technologies involving onsite HP production followed by HP conversion to PO without refining are being looked over as promising technologies with good potentials for savings in PO capital and production costs. Dow and Degussa are working on HP based technologies in partnership with BASF and Udhe, respectively.
Lyondell has gone a step further by developing a direct oxidation technology, involving in-situ HP production with simultaneous conversion to PO in the same reactor. This technology, reportedly, is in the early phase of testing by Lyondell at its research center in Newtown Square, PA. Lyondell employs a composite catalyst consisting of palladium and titanium silicalite with certain additives such as platinum or (probably) gold for a hydrogen, oxygen and propylene reaction system. The catalyst additionally contains small amounts of substances used as modifier to augment PO selectivity by inhibiting byproducts generation. The exact chemistry and physical shape of the catalyst/s used in pilot plant is part of Lyondell’s proprietary information and hasn’t been disclosed in open literature. Results shown in some Lyondell’s patents indicate high conversion rates for hydrogen, oxygen and propylene, and up to 90% selectivity for PO based on propylene.
This Review presents a conceptual configuration and analysis of a direct-oxidation PO process founded upon the technical information given in US 6,710,194 and US 6,500,311 (assignee, Lyondell). The conceived process can briefly be described as below;
Hydrogen, oxygen and propylene (along with recycle unconverted reactants) are continuously fed to a slurry reactor containing the catalyst (palladium and titanium silicalite) particles suspended in methanol. Propane (nitrogen or any C1–C4 hydrocarbon can also be used) is used in the process system as an inert carrier. The reaction is carried out at 140°F (60°C) and 200 psia. Hydrogen conversion ratio is high (95%); oxygen and propylene conversions are moderate (77% and 53%, respectively). The reactor effluents are first flashed to remove bulk of the unreacted gases and then, after removing the solid catalyst particles via a proprietary system, passed through a series of distillation columns and a liquid-liquid extraction step to get the purified PO product. Solvent methanol is recycled to epoxidation reactor after reslurrying the solid catalyst particles back into reaction system.
Our engineering and economic estimates indicate that a grassroots 220-million lb/yr (100,000 metric ton/yr) PO plant based on the direct-oxidation technology of Lyondell may give a capital saving of up to 30% when compared to Lyondell’s conventional hydroperoxidation technology. However, productions cost of PO for such a plant offers only a marginal cost advantage of a few cents per lb of product. The technology is clean and does not produce byproducts. The new process cannot compete with chlorohydrin technologies, which have the lowest TFC costs, though production costs are close.