Role of Step Sites and Surface Vacancies in the Adsorption and Activation of CO on -Fe 5 C 2 Surfaces Melissa A. Petersen,* Jan-Albert van den Berg, and Werner Janse van Rensburg Sasol Technology (Pty) Ltd., R&D DiVision, 1 Klasie HaVenga Road, Sasolburg 1947, South Africa ReceiVed: December 11, 2009; ReVised Manuscript ReceiVed: March 9, 2010 Density functional theory calculations have been used to investigate CO adsorption on three surface terminations of -Fe 5 C 2 in the presence of carbon vacancy sites. CO did not show a strong energetic preference for a particular adsorption site on each surface, since similar adsorption energies were obtained for structurally distinct adsorption configurations. In addition, it was found that the adsorption of CO in a vacancy site is not necessarily more favorable energetically, compared with adsorption in alternative Fe sites. The presence of a subsurface carbon atom directly below a 4-fold site was found to inhibit or significantly destabilize adsorption of CO in that site. The role of step sites in activating CO has been investigated by comparing the calculated adsorption energies, structural properties, and vibrational stretching frequencies of CO adsorbed in equivalent sites in the presence or absence of steps. Coordination of CO to the surface through both ends of the molecule was associated with a lengthening of the C-O bond and a red-shift of the C-O stretching frequency, and such geometries were readily obtained for adsorption at the bottom of a step. Activation energies were calculated for the dissociation of CO initially adsorbed in a vacancy site in the presence and absence of steps. The step sites were found to lower the activation energy by at least 0.3 to 0.6 eV, without destabilizing the initial state. 1. Introduction The Fischer-Tropsch synthesis (FTS) reaction converts synthesis gas (a mixture of H 2 and CO) derived from coal, natural gas, or biomass into chemicals and hydrocarbon fuels. 1,2 Commercially, catalysts based mainly on Fe or Co are used, the choice of which is dependent on factors such as the source from which the synthesis gas is derived, the cost of the catalyst material, and the desired product distribution. In particular, iron- based catalysts are suited to coal-to-liquid (CTL) technologies due to their ability to catalyze the water-gas shift reaction, which can increase the low H 2 /CO ratio typical of synthesis gas derived from coal gasification. 3 This, coupled with the abundance and low cost of iron, has led to the use of iron- based catalysts in commercial CTL operations, such as at the Sasol facilities based in Secunda, South Africa. 1 Iron-based Fischer-Tropsch (FT) catalysts undergo complex phase transformations both during catalyst pretreatment and during FTS. 4–6 Typically, the catalyst precursor phase (such as R-Fe 2 O 3 ) is treated in H 2 , CO, or synthesis gas, followed by FTS, during which a mixture of carbide, oxide, and metallic iron phases may coexist. 5 In particular, the formation of iron carbide phases has been associated with FT activity, 7–9 and in some instances a correlation between the extent of bulk carburization of the catalyst and the FT activity has been reported. 7,8 At least five different carbide phases have been identified as forming under FTS conditions, including the ε′- Fe 2.2 C, ε-Fe 2 C, -Fe 5 C 2 , θ-Fe 3 C, and Fe 7 C 3 phases. 10–13 The extent of carburization of the catalyst and identity of the carbide phases present both prior to and during FTS have been observed to be sensitively dependent on the preparation, pretreatment, and process conditions. 9,11–14 Although the formation of iron carbide phases has been cited to be a necessary prerequisite for FTS activity, 9,15 a correlation between the bulk carbide phase composition and catalyst activity has not been consistently observed, 3,16,17 and the role of bulk carbide phases in FTS remains under debate. 4 Nevertheless, FT activity has in a number of cases been related to the amount of -Fe 5 C 2 phase present in the catalyst, 5,14,18,19 suggesting that the Ha ¨gg iron carbide phase (-Fe 5 C 2 ) may play an important role in FT activity. Although it should be acknowledged that the working catalyst is likely to be a mixture of phases 4,20 that may exhibit amorphous regions, 5 and in which the surface composition may differ from the bulk, 17,21 an understanding of the intrinsic reactivity of an iron carbide phase that has been associated with FT activity, such as -Fe 5 C 2 , is likely to yield fundamental insight into the underlying catalytic role of carbide phases during iron-based FTS. One way to obtain such insight is through the use of first principles calculations based on density functional theory (DFT), which have been successfully applied in the elucidation of fundamental aspects of catalytic processes of industrial relevance. 22,23 With respect to iron carbides, a number of studies based on DFT calculations have emerged in recent years, focusing on the bulk and surface properties of Fe 3 C, 24,25 Fe 4 C, 26 and Fe 5 C 2 , 24,27 or on the chemisorption of CO and/or H 2 on Fe 3 C, 28 Fe 4 C, 29 and Fe 5 C 2 surfaces. 30–33 Two further studies on ketene hydrogenation 34 and carbon species 35 adsorbed on Fe 5 C 2 (001) have also been reported, and very recently the role of vacancy sites on iron carbides, formed through surface carbon hydrogenation, has been highlighted. 36,37 A common feature of all of these studies is that they have focused on the low Miller index iron carbide surfaces, without placing any particular emphasis on the relationship between similar adsorption sites in different surface environments. FTS necessarily starts with the adsorption and activation of CO on the catalyst surface. Recent DFT studies that have focused on the structure sensitivity of elementary catalytic reactions such as CO dissociation, have revealed that steps sites can significantly lower the activation energy for CO dissociation on a range of transition metals, including Fe, 38 Co, 39–41 Ni, 42 Ru, 43 Rh, 44,45 and Pd. 44 Specifically * E-mail: melissa.petersen@sasol.com. J. Phys. Chem. C 2010, 114, 7863–7879 7863 10.1021/jp911725u  2010 American Chemical Society Published on Web 04/01/2010