This study is led by Prof. Landong Li (College of Chemistry, Nankai University) and Prof. Ding Ma (College of Chemistry and Molecular Engineering, Peking University). The direct epoxidation of propylene with molecular oxygen has 100% theoretical atomic economy and is the most ideal route for the production of propylene oxide. However, this aerobic epoxidation reaction suffers from the apparent trade-off between propylene conversion and PO selectivity, and remains a key challenge in catalysis. The research team develops titanium-containing zeolite catalyst (Ti-Beta) for the reaction, achieving very high selectivity to propylene oxide at reasonable propylene conversions.
Ti-Beta zeolites with high titanium content are prepared via a two-step post-synthetic modification strategy. The presence of isolated pentacoordinated titanium sites in as-prepared Ti-Beta is explicitly demonstrated by various characterization techniques including UV-Vis, Cs-corrected ADF-STEM and TOF-SIMS.
In catalytic tests, stable propylene conversion of 17% and PO selectivity of 85%–90% can be achieved over Ti-Beta at the same time, matching the levels of industrial ethylene aerobic epoxidation process. H-terminated pentacoordinated Ti species within Beta zeolite frameworks are identified as the preferred active sites for propylene aerobic epoxidation and the reaction is initiated by the participation of lattice oxygen in Ti–OH.
The reaction mechanism is further interpreted with DFT calculations. In the PO pathway at the H-terminated pentacoordinated Ti site, H transfer from OH* to an adjacent O atom has the highest energy barrier of 0.92 eV, making it the rate-determining step (RDS). For the acrolein pathway, the RDS involves a higher energy barrier (1.18 eV) for CH2=CHCH2* attacking O* at the Ti site, leading to acrolein byproduct. Hence, PO formation is more favorable than acrolein formation due to the lower energy barrier, corresponding to the high PO selectivity observed in propylene aerobic oxidation.
Transient kinetic analysis reveals a maximal PO yield of 0.08 molPO/molTi upon feeding propylene to Ti-Beta due to the reaction between propylene molecules and the lattice oxygen in Ti–OH. The left vacant Ti sites are ideal for dioxygen chemisorption to form Ti–OO* motif. Subsequent feeding propylene to regenerated Ti-Beta containing Ti–OO* species gives PO yield of 0.17 molPO/molTi, nearly double the initial yield. These observations undoubtedly confirm that the catalytic process follows the Mars-van-Krevelen mechanism, in good consistency with that predicted by DFT calculations.
The breakthrough presented in this study is expected to spark new technology for the industrial production of PO toward more sustainable chemistry and chemical engineering.