(385b) Role of Co-ZSM-5 in the Low Temperature Aerobic Selective Oxidation of Neat Ethylbenzene

Selective partial oxidation of alkylaromatics with molecular O₂ is environmentally and economically highly desirable. Acetophenone (AP), a product of ethylbenzene (EB) oxidation, is an important precursor for many valuable chemicals such as resins, and cycrimine. Common methods to oxidize ethylbenzene either use stoichiometric amounts of expensive organic oxidants [1] or involve hash reaction conditions such as high temperature and/or high pressure [2]. In our laboratory, we have observed that the filtrate generated from cyclooctene (COE) epoxidation reaction was effective in the low temperature aerobic selective oxidation of ethylbenzene. Two types of COE filtrates could be used; one with solubilized Au nanoclusters [3] and one without. Surprisingly, the oxidation rate was significantly enhanced by the presence of Co-ZSM-5 in which cobalt ions were dispersed in ZSM-5. In fact, simultaneous presence of COE filtrate and Co-ZSM-5 was necessary for high activity.

In a typical reaction, 7 mL EB, 1 mL decane (internal standard), and 3 mL COE-filtrate were added to the reactor. At 100 ÌŠC, the reaction was started by bubbling in O₂ at 30 mL/min. The products, acetophenone (AP) and 1-phenylethanol (PEA) were identified using Agilent 6890 GC equipped with a mass spectrometer, and ¹H NMR was used to detect ethyl benzene hydroperoxide (EBHP) and quantify the conversion and product selectivity. Reaction proceeded slowly. Under the same reaction condition, if 32 mg Co-ZSM-5 (5% loading) was also present, the EB oxidation activity was three to six times faster. There was no enhancement in activity if HZSM-5 was used instead of Co-ZSM-5. On the other hand, although Co-ZSM-5 was present, there was no oxidation activity if COE filtrate was not used. Thus, only simultaneous presence of COE filtrate and Co-ZSM-5 resulted in high activity. The degree of enhancement depended on the cobalt loading on ZSM-5, being higher at higher cobalt loadings. The activity enhancement by Co-ZSM-5 was reversible, and removal of CO-ZSM-5 solid by hot-filtration in the course of the reaction returned the reaction rate to the level before its addition. The COE-mixture contained mostly cyclooctene oxide (~80%), with minor components that included cyclooctenone, cyclooctenol, cyclooctanediol and cyclooctene hydroperoxide. The amount of cyclooctene hydroperoxide present in the mixture was carefully quantified to be 0.3 mmol by titration using sodium thiosulfate, which was much less than the amounts of EB reacted (15% conversion is equivalent to 9 mmol EB consumption). Interestingly, if tert-butyl hydroperoxide (1.65 mmoles, TBHP/Co = 61) was used instead of COE filtrate, only trace amounts of reaction were detected. Possible mechanistic explanations and the role of Au and the organics in the COE filtrate will be presented.

1. A. V. Biradar and T. Asefa, Appl. Catal. A. 435-436, (2012) 19-26

2. H. Ma , J. Xu, C. Chen, Q. Zhang, N. Jianbo, H. Miao, and L. Zhou, X. Li. Catal. Lett. 113, (2007) 104-108.

3. L. Qian, Z. Wang, E. V. Beletskiy, J. Liu, H. J. dos Santos, T. Li, M. do C. Rangel, M. C. Kung, and H. H. Kung Nature Communications, March 28, 2017, DOI 10.1038/NCOMMS14881