Elsevier

Catalysis Today

Volume 369, 1 June 2021, Pages 129-143
Catalysis Today

Fischer-Tropsch synthesis: Direct cobalt nitrate reduction of promoted Co/Al2O3 catalysts

https://doi.org/10.1016/j.cattod.2020.03.033Get rights and content

Highlights

  • TPR/TPR-MS indicate Co reducibility slightly increases using direct nitrate reduction.

  • TPR-EXAFS results reveal that Co crystallites are smaller using direct nitrate reduction.

  • TPR-XANES: Co0 preceded by Co nitrate decomposition, and Co3O4 and CoO intermediates.

  • CSTR tests show improved activity for uncalcined unpromoted and Pt-promoted Co catalysts.

  • The best performing catalyst was 0.5 %Pt-25 %Co/Al2O3 from direct nitrate reduction.

Abstract

Direct reduction of cobalt nitrate versus conventional calcination/reduction treatment was conducted using alumina with identical methodology as previously applied to SiO2 and TiO2. Similar BET surface areas, pore volumes and pore size distributions were obtained for the activated calcined and uncalcined catalysts indicating no significant difference in morphological properties. However, the reducibility slightly increases and Co crystallite size is smaller for activated uncalcined samples. Reduction phenomena were analyzed by TPR-MS and TPR-EXAFS/XANES. Combining these techniques allows an explanation of the complex phenomena occurring during the direct reduction of cobalt nitrate, as both nitrate decomposition and cobalt oxide reduction are involved. Cobalt nitrate species are converted to CoOx intermediates. These species are oxidized by NOX (from nitrate decomposition) to Co3O4 spinel, which is converted to CoO prior to Co0 formation. Noble metals (Pt, Re, Ru and Ag) improve cobalt oxide reducibility, especially for the final reduction step (i.e., CoO to Co0). The effect of direct nitrate reduction on FT activity was investigated using a 1 L CSTR. Activated unpromoted and Pt-promoted uncalcined catalysts achieved higher initial and steady-state CO conversions in comparison to the corresponding calcined catalysts. The best performance was achieved with direct reduction of uncalcined 0.5 %Pt-25 %Co/Al2O3.

Introduction

Fischer-Tropsch synthesis (FTS) is a catalytic reaction which converts syngas, generated by natural gas, coal or biomass, to synthetic clean fuels [1]. The active metals for FTS are: iron, cobalt, ruthenium, and nickel. Ruthenium, though the most active, is impractical due to its high cost and low abundance [2], while nickel, though inexpensive, is basically a methanation catalyst [3]. Thus, cobalt and iron are the essential active industrial metals for FTS. Iron is commonly paired with Coal-to-Liquids (CtL) or Biomass-to-Liquids (BtL), where the syngas ratio (H2/CO ratio) is close to 0.5–1.2, because of its intrinsic catalytic activity for water-gas shift (WGS). In contrast, cobalt is preferred for the Gas-to-Liquids (GtL) process, as it possesses low water-gas shift activity and is well-suited for higher H2/CO ratios (close to 2) [4]. Cobalt is an expensive metal and is usually supported on an inert metal oxide carrier (i.e., Al2O3, SiO2, TiO2) in order to increase the reactive surface area. The most common technique to prepare supported cobalt-based catalysts utilizes the following steps: wet impregnation (IWI: incipient wetness impregnation; SI: slurry impregnation) of cobalt nitrate, drying and calcination in air [5]. For typical air calcined catalysts, the strength of the interaction between the cobalt oxide species and the support depends on the type of support. Strong cobalt-support interactions exist on Al2O3 where small cobalt clusters are stabilized, whereas moderate and weak interactions are observed with air calcined Co/TiO2 and Co/SiO2 catalysts, respectively [5,6].

Air calcination is the typical pretreatment method to convert cobalt nitrate to cobalt oxide and remove NOX gases, as well as H2O. However, the cobalt agglomerates into larger Co0 clusters (e.g., 50+ nm) when air calcination is used prior to H2-activation if the metal-support interactions are weak, as in the case of Co/SiO2 [5]. Previous studies revealed that direct reduction of Co(NO3)2·xH2O/SiO2, either unpromoted [7,8] or accompanied by a reduction promoter [9] such as platinum (Pt), ruthenium (Ru), rhenium (Re), or and silver (Ag), resulted into smaller cobalt metal nanoparticles than those produced from reduction of the respective air calcined catalysts. Furthermore, the reduction of the cobalt oxides generated from cobalt nitrate decomposition was improved by the addition of noble metals, leading to higher active site densities. C5+ selectivity and hydrocarbon productivity in FTS reactor testing were improved as well.

The effect of the direct reduction of cobalt nitrate was also investigated in the case of moderate metal-support interactions by characterizing and testing cobalt supported on TiO2 [10]. As in the case of SiO2, smaller cobalt clusters can be obtained by direct reduction of the uncalcined catalyst than by the calcination followed by reduction. However, the effect was not so dramatic as in the case of SiO2 (i.e., where the cluster size decreased from 52.2 to 10 nm) [9] as the cluster size decreases from 9.5 nm to 7.4 nm for Co/TiO2. This slight effect on the particle size is related to the moderate interaction between TiO2 and cobalt, which already stabilized a relatively small cluster even in the case of the calcined sample. Lower reducibility was observed for the uncalcined samples and the addition of reduction promoters (Pt and Ru) improved both cobalt dispersion and extent of reduction. The 12 %Co/TiO2 uncalcined catalyst had twice the activity compared to the corresponding unpromoted calcined catalyst, whereas the 0.5 %Pt-12 %Co/TiO2 uncalcined catalyst exhibited both higher steady-state CO conversion and a lower deactivation rate than the corresponding calcined catalyst.

In 2007, temperature programmed reduction coupled with EXAFS/XANES (TPR-EXAFS/XANES) was used to investigate the influence of the support and promoter [6]. The experiments revealed that the size of cobalt crystallites and the nature of reduction of cobalt oxide following activation depends on the strength of the interactions between the support and the cobalt oxide species prior to reduction, with silica offering a weak interaction and alumina providing a strong one. These experiments point out two reduction steps: (1) Co3O4 to CoO (2) and CoO to Co0. Moreover, this study showed that the addition of noble metal, such as platinum, enhances the reducibility of cobalt oxide species, in agreement with previous studies [5,6,[11], [12], [13]].

TPR-EXAFS/XANES provided information to reveal the changes in chemical identity (e.g., oxidation state changes) and local atomic structure occurring during the direct reduction of cobalt nitrate supported on SiO2 and TiO2 [9,10]. CoOx intermediate species are formed from the decomposition of cobalt nitrate. Subsequently, these CoOx species are re-oxidized to Co3O4 by NOx. Following the nitrate decomposition, two reduction steps occur where Co3O4 is converted to CoO, and finally, CoO is reduced to Co0. The addition of reduction promoters improves the extent of reduction of cobalt, with the ranking of promoter effectiveness being: Pt > Re > Ru > Ag > unpromoted.

Alumina is a common industrial support for cobalt FTS catalysts, and is characterized by strong interactions with the cobalt oxides. Many studies have been carried out in attempts to minimize or overcome interactions between cobalt oxides and the support. So, it is interesting to investigate if the direct reduction of cobalt nitrate can have any effect in controlling the metal-support interaction and the resulting sizes of Co0 crystallites and Co0 clusters.

The most important properties of the catalyst considered in the current work are: carbon monoxide conversion per gram of catalyst (conversion rate), stability and selectivity. Cobalt particle size should play a crucial role in tuning the CO conversion rate. Turnover frequencies (TOF) for catalysts having particle sizes between 10 and 210 nm are typically similar [14,15]. However, the results for particle sizes smaller than 10 nm are not consistent. For instance, Co particle size ranges of 3−5 nm, obtained by a deposition-precipitation method, displayed a linear trend between activity and cobalt surface area per gram of catalyst [16]. In contrast, in a different study, TOF and selectivity diminished on cobalt/carbon nanofiber (CNF) catalysts when the cobalt particle size was lower than 16 nm [15]. Thus, the effect of the direct reduction of the nitrate with Co/Al2O3 and the correlation between the cobalt particle size and the activity and C5+ selectivity needs to be investigated.

Regarding the catalyst stability, heavily loaded Co catalysts have been reported to have greater stability against catalyst oxidation (i.e., and cobalt-support compound formation) in Co/Al2O3 [17,18]. In fact, the deactivation rate for Pt-25 %Co/Al2O3 is lower than for Pt-15 %Co/Al2O3. In 2016, Hughes et al. [19] performed EXAFS/XANES experiments on Co/TiO2 catalysts and reported that while catalysts with small average nanoparticle size (2−3 nm) prepared by CVD of cobalt underwent oxidation under conditions mimicking 50 % conversion, catalysts having larger cobalt clusters (nanoparticles ≥ 10 nm) underwent net reduction at the same conditions. Thermodynamic studies point out that small cobalt nanoparticles lower than ≤ 4 nm may re-oxidize under reactions conditions, and that this may be exacerbated by more strongly interacting supports [20,21].

In this contribution we have investigated the effect of direct reduction of cobalt nitrate versus the standard method of calcination followed by reduction and incorporate reduction promoters to produce different extents of reduction and different cobalt crystallite sizes (i.e., as inferred from Co-Co coordination in EXAFS) for different catalysts. Measurements include BET surface area and porosity, TPR-EXAFS/XANES, TPR-MS, and catalytic activity data obtained using a 1 L CSTR.

Section snippets

Catalyst preparation

Co/Al2O3 catalysts were prepared following the same procedure reported in our previous work for Co/SiO2 and Co/TiO2 catalysts [9]. Briefly, a parent batch of 25 wt.% Co/Al2O3 catalysts was prepared using standard incipient wetness impregnation (IWI) of aqueous solutions of cobalt nitrate hexahydrate solution (Co(NO3)2*6H2O) and Catalox 150 Al2O3 (133.7 m2/g, calcined at 350 °C for 4 h) served as the catalyst support. Following impregnation, the catalysts were dried using a Heidolph rotary

Morphological and structural properties

The morphological properties (BET surface area, pore volume and average pore diameter) for the prepared catalysts are reported in Table 1 and Fig. 1. Considering γ-Al2O3 has a surface area of 133.7 m2/g, the addition of 25 % by weight metallic cobalt is equivalent to 34 wt.% of Co3O4 and 77 wt.% for Co(NO3)2. Thus, assuming no pore blocking, the surface area should decrease to 88 m2/g and 30 m2/g for the calcined and uncalcined sample, respectively. Co/Al2O3 and Pt-Co/Al2O3 calcined catalyst

Conclusions

The direct reduction of cobalt nitrate on alumina support has been compared with the more conventional procedure of calcination followed by reduction. The morphological properties are not modified by skipping the calcination step. Indeed, the catalyst after the reduction step has almost identical surface area, pore volume and pore size distribution. From using a combination of TPR, TPR-MS, and TPR-EXAFS/XANES experiments, the main events observed during the reduction of uncalcined samples are:

CRediT authorship contribution statement

Mohammad Mehrbod: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing - review & editing, Writing - original draft. Michela Martinelli: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Project administration, Resources, Supervision, Validation, Visualization, Writing - review & editing, Writing - original draft. Jonathan D. Castro: Data curation, Formal analysis, Investigation, Methodology,

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Research conducted at UTSA was supported by UTSA, the State of Texas, and the STARs program. The work carried out at the CAER was supported in part by funding from the Commonwealth of Kentucky. Argonne’s research was supported in part by the U.S. Department of Energy (DOE), Office of Fossil Energy, National Energy Technology Laboratory (NETL). Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No.

References (31)

Cited by (8)

  • EXAFS studies of inorganic catalytic materials

    2023, Comprehensive Inorganic Chemistry III, Third Edition
  • Fischer-Tropsch synthesis: Towards a highly-selective catalyst by lanthanide promotion under relevant CO<inf>2</inf> syngas mixtures

    2022, Applied Catalysis A: General
    Citation Excerpt :

    In general, two main peaks were observed in all samples. All promoted materials exhibited a reduction pattern similar to that of the reference catalyst, with a first peak in the range 260–330 ºC associated to the low temperature reduction of Co3O4 to CoO, and a second one, in the range 470–565 ºC, which corresponds to the high temperature reduction of CoO to Coº [23,37]. However, it is important to underline three facts.

  • Mn doping of Co-Al spinel as Fischer-Tropsch catalyst support

    2021, Applied Catalysis A: General
    Citation Excerpt :

    Stability for FTS reaction was tested on 23Co/CMA-20 catalyst, and the results are shown in Fig. 9. It shows that the decline of CO conversion is less than 5% during the 1200 h testing, indicating its good stability [49–51]. XRD, Raman, EPR and TEM results of lattice Mn-doped CMA-m spinels show that, compared with CA, Mn doping in the lattice sites of spinel destroyed the periodicity growth of crystal, and formed defects on crystal grains, which might cause more attachment points such as nanoscale concaves, steps and corners on the surface of supports.

View all citing articles on Scopus
View full text