545568 Co-Ni Supported on Reducible Oxides As Highly Active and Selective Fischer-Tropsch Catalysts

Tuesday, June 4, 2019: 10:39 AM
Texas Ballroom EF (Grand Hyatt San Antonio)
Carlos Hernández Mejía, Jessi E.S. van der Hoeven, Petra E. de Jongh and Krijn de Jong, Inorganic Chemistry and Catalysis, Utrecht University, Utrecht, Netherlands

The increasing worldwide demand for transportation fuels has stimulated the research towards efficient fuel production, particularly for liquid fuels since these possess a high energy density.1 Fuels from processes such as natural gas to liquids (GTL) or coal to liquids (CTL) have emerged as an attractive alternative to the traditional fuels derived from crude oil. The GTL and CTL processes have several advantages like allowing for a feedstock diversification and generating ultra-clean fuels. A crucial step in GTL and CTL is the catalytic transformation of synthesis gas (a mixture of carbon monoxide and hydrogen) to hydrocarbons, also known as the Fischer-Tropsch (FT) synthesis. Cobalt-based catalysts are commonly employed in FT due to their high selectivity towards long-chain hydrocarbons (C5+), low water-gas shift activity and stable catalytic performance.2 A drawback however is that cobalt is a relatively expensive metal which in the last year has tripled in price,3 harming the competitiveness of FT.

In this research, we investigated the effect of gradual cobalt substitution by nickel, a more abundant and affordable metal, on supported FT catalysts and the influence of the support’s nature on the catalytic behaviour of the metals. For this, different cobalt to nickel molar ratios (x = Ni/(Ni+Co): 0, 0.25, 0.5, 0.75 or 1 mol/mol) supported on reducible (TiO2 and Nb2O5) or non-reducible (a-Al2O3) oxides were studied. Several characterization techniques were employed to understand the effect of combining cobalt and nickel on the different supports. Ex-situ analysis of the reduced samples by transmission electron microscopy showed an increase in metal dispersion upon nickel addition independently of the support. EDX-mapping additionally showed an overlap of cobalt and nickel signals for the mixed compositions, indicating the formation of a cobalt-nickel alloy after the reduction treatment. H2-chemisorption measurements of the reduced samples confirmed the increased metal dispersion upon nickel addition for the Al2O3-supported samples. In the case of the TiO2- and Nb2O5-supported samples, the hydrogen uptake was lower than expected for their metal particle size and relatively constant for all metal compositions. The inhibition in hydrogen uptake is characteristic of reducible supports, known as Strong Metal-Support Interaction (SMSI).4 During reductive conditions suboxides from the support might cover the metal nanoparticles, hindering hydrogen uptake. However, the metal-suboxide interface may give rise to promotional effects for FT.

The FT performance of the samples was evaluated at various reaction pressures and temperatures with an H2 to CO ratio of 2 v/v. Figure 1 shows the metal-normalized activities, expressed as Metal-Time-Yield (MTY, 10-5molCO·gmetal-1·s-1), for TiO2- and Nb2O5-supported catalysts as a function of the total pressure at 220 °C. At 1 bar, the Co-Ni samples displayed increased activities and C5+ selectivities, doubling the catalytic activity compared to the monometallic catalysts and reaching C5+ selectivities up to 77 % for the samples with 50 and 75 % Ni (x = 0.5 and 0.75). Contrastingly, Al2O3-supported catalysts had similar metal-normalized activities (~ 1·10-5molCO·gmetal-1·s-1) and high selectivity towards methane for all Co-Ni compositions. Moreover, these catalysts deactivated severely throughout the experiment contrary to the more stable TiO2- and Nb2O5-supported catalysts. Electron microscopy analysis of the spent catalysts revealed extensive metal particle growth for the Al2O3-supported catalysts containing nickel, explaining their fast deactivation, whereas only a slight growth was observed for the reducible supports. At these reaction conditions nickel readily forms volatile nickel carbonyl, leading to metal transport and particle growth. The use of reducible oxides appeared to have modified the reactivity of nickel in the alloy.

Increasing total pressure benefited the activity and selectivity of the samples with high cobalt content (x = 0.0, 0.25 and 0.5, Figure 1), particularly for the TiO2-supported catalysts. Likewise, higher pressures improved their C5+ selectivity whereas a lesser increase was observed upon increasing nickel content. In contrast, the activity of the samples with a high nickel content (x = 0.75) diminished by increasing the total pressure.

Figure 1. Metal-normalized catalytic activities against the different cobalt to nickel ratios supported on titania (left) or niobia (right). Reaction conditions: 1, 5 or 20 bar, 220 °C, H2/CO = 2 v/v and CO conversions 5 – 35 %.

The catalytic behaviour at 20 bar, 220 °C, H2/CO = 2 v/v and TOS = 100 h displayed similar trends for the TiO2- and Nb2O5-supported catalysts. The samples with 25 % Ni had an activation period at the first hours of the experiment resulting eventually in the highest metal-normalized activities and turn-over-frequencies of the set of samples, moreover their C5+ selectivity was the same as the samples containing only cobalt (83 % for TiO2 and 86 % for Nb2O5). Replacement of 50 % cobalt by nickel (x = 0.5) showed similar activities as the cobalt-based catalysts (x = 0.0) however with a slightly lower C5+ selectivity. STEM-EDX was employed to characterize the catalysts after reaction, minor particle growth was observed for all samples. Additionally, the initial alloyed Co-Ni nanoparticles observed after reduction showed changes in their metal arrangement after catalysis (Figure 2) with some of the particles having a cobalt-enriched surface.

Figure 2. STEM-EDX mapping of Co-Ni (x = 0.75) supported on niobia after catalysis (20 bar, 220 °C, H2/CO = 2 v/v and TOS = 100 h). Cobalt is shown in red and nickel in green.

The catalytic performance was additionally evaluated at different temperatures (220 – 260 °C) and 20 bar for the different Co-Ni compositions on both reducible supports. Apparent activation energies were derived from these results (Figure 3) and showed that at high nickel content (x = 0.75) the activation energy increases compared to the 100 % cobalt catalysts (x = 0.0) on both supports. A substitution of 50 % of the cobalt by nickel had the same activation energy as for x = 0.0. Finally, substitution of cobalt by 25 % nickel decreased the activation energy, indicating that up to a certain nickel concentration, it functioned as a promotor, which was also reflected on increased turnover frequencies. Interestingly, the selectivity was less affected at high temperatures when the nickel content was higher.

Figure 3. Apparent activation energies for the catalysts with different cobalt to nickel ratios supported on titania (red circles) or niobia (green squares). Reaction conditions: 20 bar, 220 - 260 °C, H2/CO = 2 v/v.

In summary, we showed that the combination of cobalt and nickel supported on reducible oxides allow for 25 to 50 % cobalt substitution for nickel with increased FT activity and without sacrificing much C5+ selectivity. Furthermore, nickel-rich catalysts on reducible supports allowed for higher reaction temperatures while maintaining a good selectivity to C5+ products. Reducible oxides used as support material strongly modified the reactivity of Co-Ni alloys, opening new possibilities for more efficient and affordable FT catalysts.

1. De Jong, K. P. et al. The frontiers of energy. Nat. Energy 1, (2016).

2. Dry, M. E. The Fischer–Tropsch process: 1950–2000. Catal. Today 71, 227–241 (2002).

3. The Economist. Goblin metals. Econ. 426, (2018).

4. Tauster, S. J., Fung, S. C. & Garten, R. L. Strong metal-support interactions. Group 8 noble metals supported on titanium dioxide. J. Am. Chem. Soc. 100, 170–175 (1978).

 


Extended Abstract: File Not Uploaded
See more of this Session: FTS on Co Catalysts II
See more of this Group/Topical: General Submissions