468857 Iron-Manganese Catalysts for the CO Hydrogenation to Alkenes: Effect of the Catalyst Synthesis on the Selectivity and Activity
Iron-manganese catalysts for the CO hydrogenation to alkenes: effect of the catalyst synthesis on the selectivity and activity.
H. Karroum1,2, V.Iablokov1, Y. Xiang1, V. Dubois2, N. Kruse1
1 Voiland School of Chemical Engineering, Washington State University, Pullman, WA 99164
2 Institut Meurice, Brussels, Belgium
Fischer-Tropsch Synthesis (FTS) consists of the catalytic conversion of syngas into a large range of hydrocarbons and/or alcohols. Research into FTS has spurred in recent times due to the possibility of producing chemical raw materials such as oxygenates and olefins. These materials are being considered as important building blocks for industrial organic and polymer chemistry and provide the feedstock for lubricants, detergents, etc. The main challenge is to tune the selectivity of the Fischer-Tropsch reaction to favor one product class over others. Here we are interested in favoring short chain olefins and investigate the prospects of innovative catalyst preparation to boost olefins versus paraffins production.
It has been previously demonstrated that the size and morphology as well as the composition of metal particles have a major influence on catalyst performance (activity, selectivity and stability). The use of colloidal recipes seems most promising in controlling the size and morphology of metallic nanoparticles. Iron-manganese catalysts have been prepared in our study using this colloidal approach. Furthermore, and quite differently, the same catalysts have been prepared via oxalate co-precipitation as developed in our laboratory in recent years. The catalytic properties of the various catalysts prepared by either method will be compared in the high-pressure FTS synthesis.
The oxalate route consists of the co-precipitation of Fe2+ and Mn2+ with oxalic acid. The precipitate is known to form an extended polymeric oxalate structure incorporating both metal cations in the same structure. Subsequent hydrogen-assisted thermal decomposition (at 390°C) of the mixed-metal oxalate leads to an intimately mixed iron-manganese catalyst (‘cat 1’). On the other hand, the colloidal route consists of the thermal decomposition of Fe(acac)3 and Mn(acac)2 using oleylamine simultaneously as solvent and reductant. Typically, the suspension is first heated under vacuum at 100°C in order to dissolve the salts, followed by increasing the temperature up to 290°C under nitrogen atmosphere. Finally, iron-manganese nanoparticles are precipitated with methanol and after centrifugation the precipitate is redisolved in chloroform. The particles have been impregnated into silica (Davisil S = 260m2/g) support (‘cat 2’).
Prior to catalytic tests, both catalysts have been activated in situ under hydrogen using temperature programmed decomposition. XRD analyses have been performed on both activated catalysts. According to the XRD patterns, the bulk of ‘cat 1’ is mainly composed of spinel-type Jacobstite (MnFe2O4) while the bulk of ‘cat 2’ indicates broad diffraction lines tentatively assigned to Fe-oxide structures (Fe2O3 and Fe3O4). The absence of diffraction patterns associated with Mn-related structures is indicative of high Mn dispersion providing XRD-amorphous states in “cat 2’. TEM micrographs show that the particle sizes of ‘cat 1’ are ranged between 7-15 nm while ‘cat 2’ has a narrow size distribution ranged between 6-6.5nm. For both catalyst preparations metal particles appear in spherical morphology.
Catalytic tests in the CO hydrogenation have been performed in a fixed-bed plug-flow reactor after activating both catalysts under hydrogen. The catalytic performances of both catalysts have been tested at 300°C and 20 bar. ‘Cat 1’ shows an alkane selectivity of 27% and an alkene selectivity of 69% at 11 % CO conversion while ‘cat 2’ shows an alkane selectivity of 61% and an alkene selectivity of 39% at 28% CO conversion. Obviously these data demonstrate a tradeoff between conversion and olefins selectivity: while high CO conversion is associated with relatively low olefins production (39% is, however, not low!), low conversion increases the selectivities dramatically. The higher activity of ‘cat 2’ becomes obvious when comparing the reaction rates per gram of iron. In fact, the reaction rate of ‘cat 1’ is 14.7 umol gFe-1 s-1 while the reaction rate of ‘cat 2’ is 56.6 umol gFe-1 s-1. The TOFs (turnover frequencies) of ‘cat 1’ and ‘cat 2’ can then be calculated to obtain 0.009 s-1 and 0.02 s-1, respectively. Note that the reaction rate in terms of TOF values takes into account a surface iron density of 0.0613 nm-2 assuming Fe sites provide the active sites.
We have demonstrated here that the catalyst preparation procedures have a dramatic influence on the selectivity and activity toward the formation of the various product classes in FTS. Superior activity of silica-supported colloidal particles has been observed which is promising with regard to a further optimization of the process. Future efforts will therefore address the preparation of Fe-Mn bimetallic nanosized catalysts in different ratios to tune the production of short chain olefins.
 V. Iablokov, Y.Xiang, A. Meffre, P-F Fazzini, B. Chaudret, and N. Kruse, ’Size-dependent Activity and selectivity of Fe(0) Nanoparticles in the catalytic Hydrogenation of Carbon Monoxide’, ACS Catal., 2016, 6 (4), pp 2496–2500