Epoxidation of propylene in the gas phase Blazˇej Horva´ th a, *, Milan Hronec a , and Robert Glaum b a Department of Organic Technology, Slovak University of Technology, Radlinske ´ho 9, Bratislava, 81237 Slovak Republic b Institut fu ¨r Anorganische Chemie, Rheinische Friedrich-Wilhelms-Universita ¨t, Gerhard-Domagk-Str. 1, Bonn, D-53121 Germany The gas-phase epoxidation of propylene using N 2 O, air and air-ammonia mixture as an oxidants was studied. Propylene can be epoxidized by nitrous oxide with a yield as high as 13.3% over silica supported iron oxide catalysts modified by amines. The iron oxide dispersion, the acidity of the support and the nitrogen-containing modifiers are the key factors determining the catalytic performance. We suggest a reaction pathway involving two concurrent mechanisms: the radical oxidation of propylene to acroleine, hexanediene, etc., and a non-radical oxidation leading to epoxide. Propylene is epoxidized with air over silica-supported iron oxide catalysts at a conversion of about 0.2%. Using air as an oxidizing agent, the presence of gaseous ammonia improves the propylene conversion by 10-fold preserving the considerable selectivity (up to 60%). This observation suggests a reaction mechanism involving the oxidation of ammonia to nitrous oxide in the first step, which subsequently produces active oxygen species, which selectively oxidize propylene to propylene oxide (PO). KEY WORDS: epoxidation; propylene; iron oxide; nitrous oxide; oxygen. 1. Introduction The direct gas-phase epoxidation of propylene and other compounds containing reactive allylic hydrogen atoms continues to be a challenge. Lately, new catalytic systems have been designed for the direct epoxidation of propylene using hydrogen peroxide, nitrous oxide, pure molecular oxygen (often with halogenide or NO addi- tion) or hydrogen-oxygen mixture [1–8]. The above mentioned oxidants are more costly than molecular oxygen from air. (table 1) Propylene oxide (PO) manufacture is one of the largest consumers of propylene. The increased reactivity of allylic hydrogen, thus the difficulty of oxygen addi- tion to the double bond makes PO a high-value-added product. The simple analogy with the ethylene epoxi- dation over silver catalysts (even modified) is still not practically viable. Over the Ag/a-Al 2 O 3 catalyst using molecular oxygen Geenen [7] obtained the following results: (scheme 1.) The fact that the selectivity towards the epoxide is independent on the oxygen conversion indicates that k 1 >> k 3 . This points out that assuming molecular oxygen as oxidant (thus the radical reaction mechanism) the primary route to CO 2 formation is the oxidation of the substrate and not the oxidation of PO once formed. This is in accordance with the findings obtained by molecular modeling. Conversely, using a different catalytic system, Fe 2 O 3 / SiO 2 , exploiting N 2 O as an oxygen source, Ananieva [4] suggests a reaction pathway involving the formation of PO, its consecutive isomerization on acidic sites and the overoxidation of originally formed PO and isomeriza- tion products. Authors [4] stated that in scheme 2, using N 2 O, under working conditions k 2 /k 1 ‡ 4. The pore shape and diameter of the support and iron oxide particle size have been shown to be crucial parameters in epoxidation of propylene by N 2 O [5]. 2. Experimental Several different parent silicas were used as supports: Silica aerogel Aerosil 300 (Degussa) with iron impurity level <30 ppm Fe 2 O 3 with rather uniform micelle size of 8 nm, silica xerogel from Merck; as-synthesized xerogel prepared by fast precipitation of sodium silicate by acetic acid; and SBA-15 (a mesoporous silica), syn- thesized according to [9] using PluronicÒ template. Iron oxide was deposited on the supports by impreg- nation in a 0.1 wt.% solution of Fe(III)acetylacetonate in toluene, as well as by hydrolysis of 0.1% iron(III) nitrate or iron(III) chloride in 1 wt.% aqueous urea solution at 363 K. In some cases ethylenediamine was added to the Fe(III)acetylacetonate solution during the Scheme 1. Scheme 2. * To whom correspondence should be addressed. E-mail: blazej.horvath@stuba.sk Topics in Catalysis Vol. 46, Nos. 1–2, September 2007 (Ó 2007) 129 DOI: 10.1007/s11244-007-0323-7 1022-5528/07/0900-0129/0 Ó 2007 Springer Science+Business Media, LLC