Elsevier

Catalysis Today

Volume 275, 15 October 2016, Pages 49-58
Catalysis Today

Fischer–Tropsch synthesis and water gas shift kinetics for a precipitated iron catalyst

https://doi.org/10.1016/j.cattod.2016.01.006Get rights and content

Highlights

  • The kinetics of FTS and WGS reactions for a precipitated Fe–K catalyst were studied simultaneously.

  • The kinetics at low (<70%) and high conversion (>70%) ranges were modelled separately.

  • A new FTS model with CO2 inhibition described the data more accurately.

  • A new WGS model with a CO2 term was effective in the high conversion region.

  • H assisted CO dissociation and formate mechanisms were suggested for the Fe catalyst.

Abstract

A large number of kinetic data points (83 sets) was obtained over a wide range of CO conversion (7–90%), pressure (1.3–2.5 MPa) and H2/CO ratio (0.67–1.5) with an iron catalyst (100 Fe/5.1 Si/1.25 K). The kinetics of the catalyst in the low (XCO < 70%) and high conversion (XCO > 70%) regions were studied separately. Twenty six Fischer–Tropsch synthesis (FTS) and water gas shift (WGS) kinetic models were tested and discriminated. Water and CO2 inhibition was evaluated. While all thirteen FTS models gave a satisfactory fit, the new FTS models that included CO2 inhibition surpassed the others. Water inhibition of the FTS rate was insignificant over both low and high conversion ranges. For the WGS kinetics of the iron catalyst, a newly constructed empirical model and one from the literature provided the best fits of the WGS rates, while nine mechanistic models and one power law WGS model were unable to satisfactorily fit the WGS kinetic data. Water did not significantly limit the WGS rate and CO2 only inhibited the rate at high CO conversions. The equations obtained for the low and high CO conversion ranges varied greatly. The errors for the models for 85% of the FTS and WGS data points were less than 10%, and the errors of the remaining points fell in the range of 10–15%.

Introduction

Because of the importance of defining the kinetics of Fischer–Tropsch synthesis (FTS) in the development of XTL (X = coal, natural gas and biomass) processes, numerous kinetic investigations of FTS over various catalysts have been conducted over the past several decades. In a large number of kinetic studies published [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], many mechanistic or semi-mechanistic kinetic models for Fe-based FTS have been developed on the basis of classic carbide and enol mechanisms (Table 1). However, little consensus in the kinetic equations was obtained in these earlier studies, which may be attributed to the complexity of the FTS reaction mechanism itself [23], [24], and/or due to complexities associated with differences in catalyst preparation, catalyst composition, pretreatment, and process conditions [24]. The simple first-order kinetics in H2 for the iron based FTS catalyst is an important model, and it was proven to be useful to predict FTS rates at low CO conversions, i.e., <70%, of fused or promoted Fe catalysts by Dry [1], [15]. Meanwhile, Dry [8], Anderson et al [1] and Huff and Satterfield [3] derived the same form of a mechanistic kinetic model for fused Fe or precipitated Fe catalysts according to enol and/or carbide mechanisms (Model 10 in Table 1). van Steen and Schulz [4] developed a model (Model 13 in Table 1) by assuming the formation of CH monomer being the rate determination step, and it successfully fit the kinetic behavior of several unpromoted and promoted iron and cobalt catalysts. All these models include a water effect term and suggest a negative role of water on the FTS reaction. Ledakowicz et al. [2], [12] developed a mechanistic FTS model (Model 9) in terms of an enol mechanism that includes both the effect of water and CO2 on the rate. More prevalent FTS mechanistic kinetic models developed by FTS researchers (e.g., Atwood and Bennett [9], Decker et al. [10], Zimmerman and Bukur [14], van der Lann and Beenackers [7], Bote and Breman [5] and Zhou et al. [6]) and a power law model without CO2 and water inhibition (PCOaPH2b) are also summarized in Table 1. In general, these mechanistic kinetic models can be lumped as functions of partial pressures of CO and/or H2, kPCOPH2a/(1+KiPi)b (a = 0.5, or 1, b = 1 or 2; i = CO, H2O, CO2), and terms for contributions such as H2O, CO2, and/or vacant sites are reflected in the denominator.

Another issue in the kinetic study of Fe catalysts may be linked to the conversion level; it has not been widely discussed to date. The reason for this is likely due to the difficulty of obtaining reliable kinetic data in the high conversion region i.e. (e.g., >70%), since deactivation of Fe catalysts occurs to a greater degree and at a more rapid rate in this region. This is exacerbated at certain process conditions (e.g., high PH2O in the reactor). It has been reported [25], [26] that the FTS rate varies greatly with conversion over Fe based catalysts, and that this is likely due to changes in the extent of the water gas shift reaction (WGS) with CO conversion. The H2 produced by the WGS reaction, in turn, influences the FTS rate and product selectivity. Fig. 1 shows changes in CO, H2 and FTS rates with CO conversion/contact time over an iron based catalyst at 270 °C, indicating that productivity varies significantly in moving from low to high CO conversion levels. In the range of low CO conversion (below 65%, short contact time region), the productivity is high but decreases rapidly as contact time is increased in a nonlinear manner. On the other hand, the changes in the FTS rate, CO rate and H2 rate become smooth and show nearly linear relationships with contact time at high CO conversion levels (greater than 65%), suggesting that the WGS reaction limits the FTS rate, since H2 provided by WGS is needed to increase the extent of the FTS reaction. In Fig. 2, the partial pressures of CO, H2, H2O and CO2 are plotted against contact time. Over the entire range of CO conversion, the partial pressures of CO and H2 decreased almost linearly with increases in contact time or CO conversion; this is a different trend from those of the CO, H2 and FT rates (Fig. 1). Thus, the rate and pressure curves suggest that FTS kinetic behavior at low and high conversion levels may be different. However, most earlier kinetic studies were conducted at low to medium conversion levels or with only a few high conversion data points. In short, the kinetics of Fe catalysts at high conversion has not been systematically studied.

The partial pressure of CO2, as shown in Fig. 2, increases remarkably, i.e., 0.1 to 0.7 MPa, with increases in CO conversion from 25 to 87%. This is probably a reason why many previous FTS kinetic models contain a CO2 inhibition term. Fig. 2 also shows that the partial pressure of water inside the reactor was low (<0.08 MPa), about ten times lower than those of CO, H2, or CO2, which implies that the majority of the water that formed during FTS was consumed by WGS on the Fe catalyst. Moreover, the water curve passes through a maximum at about 60% CO conversion. The slight increase in water partial pressure with increasing CO conversion in the low CO conversion range probably results from a greater increase in the FTS rate relative to WGS, while the decrease in the partial pressure of water with increases in CO conversion above 60% CO conversion likely result from a relatively higher WGS rate. A number of previous kinetic studies [2], [3], [4], [7], [8], [9], [10], [11], [12], [14] have reported that water inhibits the FTS rate of Fe catalysts. Recently, Bote and Breman [5] and Zhou et al. [6] developed a kinetic model, rFT = kPCOPH20.5/(1 + mPCO)2 (Model 6 in Table 1) based on the assumptions of H-assisted CO dissociation and the formation of the CHx monomer from the intermediate species, HCO-s, being the rate determining step. The model was reported to give a good fit to a group of experimental data of iron catalysts obtained at 250 °C, but the role of H2O on the FTS rate was unnoticeable. Therefore, discrepancies in the literature regarding the role of H2O still exists, and additional studies are needed to investigate the inhibition of water on the FT rate of an iron catalyst.

As with H2O, a consensus of the CO2 effect on the FTS kinetics of Fe based catalyst has not been reached. As noted above, few studies of FTS kinetics on Fe catalysts considered the CO2 effect when constructing Fe FTS kinetic models [1], [3], [4], [5], [6], [7], [8], [9], [10], [11], [14]. Yates and Satterfield [13] conducted a CO2 effect study on the kinetics using a fused Fe catalyst by co-feeding 20–50% CO2 to a slurry phase reactor. The authors reported that CO2 was inert and had little effect on the fused Fe catalyst. They suggested that the inhibition attributed to CO2 was as a result of water formed from the reverse WGS reaction. However, Ledakowicz et al. [2] conducted a kinetic study on a K-promoted precipitated iron catalyst that exhibited high WGS activity and they obtained a pronounced CO2 effect on FTS kinetics. Therefore, despite the fact that CO2 has a low competitive adsorption and low solubility in slurry relative to H2O, CO2 inhibition is possible when a large amount of CO2 is produced on some Fe catalysts.

According to the above review, one important task of this study is to determine whether the same kinetic equation is applicable for both the low- (i.e., <70%) and the high-conversion (i.e., >70%) regions. Moreover, because of strong water adsorption, and CO2 being one of the most abundant product species on the Fe catalyst surface, the possibility of inhibition of the FT rate by H2O and CO2 for the Fe catalyst must be considered. Furthermore, how well the various mechanistic and empirical kinetic models in the literature fit the current results needs to be assessed.

The WGS reaction proceeds simultaneously with the FTS reaction on iron based catalysts for which CO rich syngas is usually used. Depending on the catalyst components and process conditions, about half of the CO is converted through the WGS reaction. Therefore, a study of the kinetics of the WGS reaction under FTS conditions is of the same significance as that of FTS reaction since it can give further insights into the relationship between the two reactions as well as provide fundamental knowledge for the design and scale up of XTL processes. So far, a few studies [7], [14], [27], [28], [29], [30] have reported the kinetics of WGS and FTS simultaneously over Fe catalysts. Various WGS kinetic models have been proposed based on certain regenerative redox and formate (-COOH) mechanisms [14], [27], [28], [29], [30], [31], [32], [33] or empirical equations [33], [34] and they are summarized in Table 2. However, a lack of a good fit for these specified mechanisms in the current context does not rule out the operation of a different formate or redox mechanism, as these are families of mechanisms. In most cases, CO2 and H2 inhibition were included in the rate equations as a consideration of how far away they were from the WGS equilibrium (PCO2PH2/Kp). The WGS model of Zimmerman and Bukur [14] included CO2 and water terms in the denominator for a 100 Fe/0.3Cu/0.2 K catalyst (rWGS=k(PCOPH2O/PCO2PH2/Kp)/(PCO+cPH2O+dPCO2)), but CO2 adsorption was reported to be insignificant. The simple power law model, rWGS=k(PCOa) [8], [21] is also summarized in Table 2. Note that different active sites (FexC and Fe3O4) [7], [28], [29], [31] or the same active sites (such as FexC) [3], [14], [27] for the WGS and FTS reactions were proposed when constructing these WGS models. Additional studies are needed to validate the effectiveness of these models using data collected from different catalysts at different conditions.

The WGS and FTS reactions may influence each other on the Fe catalyst. The WGS reaction could increase or decrease the FTS reaction on Fe catalysts, depending on the conditions. For example, the WGS rate changes significantly with K loading and CO conversion level; high K loading increases the WGS reaction rate constant, causing the WGS to approach equilibrium. The approach to equilibrium also increases greatly with CO conversion [26]. Therefore, it is important to systematically evaluate the kinetics of WGS at different conditions (i.e., low and high conversion levels).

Consequently, three important tasks were undertaken in this study. First, we examined many literature and newly constructed FTS and WGS models by assessing eighty three sets of reliable kinetic data obtained in this study and then chose the most suitable models to examine water and CO2 or H2 and CO2 inhibition on the FTS and WGS reaction rates. Kinetic data at both low and high conversion levels were compared to elucidate whether different equations apply in the different ranges. A typical K promoted Fe–Si–Cu catalyst prepared by co-precipitation was used. Conditions relevant to industrial application were employed in the kinetic experiments: a constant temperature of 270 °C; pressure and H2/CO ratio ranges of 1.3–2.5 MPa and 0.67–1.5, respectively; and CO conversions in the range of 7–90%. All kinetic data were collected with a low extent of catalyst deactivation in order to ensure accurate kinetic results. Furthermore, to obtain models that can best describe Fe catalyst kinetic performance, generalized FTS and WGS models were built according to the existing Fe catalyst kinetic models and knowledge of reactant and product adsorption on Fe based catalysts.

Section snippets

Catalyst preparation

A base Fe catalyst 100Fe/5.1Si/1.25 K was prepared using a precipitation method and details of the catalyst preparation can be found elsewhere [35]. The 100 Fe/5.1 Si/1.25 K/2.0Cu catalyst used in this study was prepared by impregnation of an appropriate amount of Cu on the base catalyst using a Cu(NO3)2∙2.5H2O precursor solution. The catalyst was dried under vacuum in a rotary evaporator at 80 °C and the temperature was slowly increased to 95 °C. The catalyst was calcined under air flow at 350 °C

Generalized FTS and WGS models

As shown in Table 1, many popular mechanistic kinetic models are the best candidates for various iron based catalysts. This could mean that the iron FTS kinetic model is a function of catalyst component; or CO and H2 adsorption and dissociation on the Fe catalyst surface were too complicated to use a single mechanism or rate determination step to mimic their behavior. Therefore, the best model describing the FTS behavior of Fe catalysts could be a lumped form if all major adsorption species CO,

Fischer–Tropsch synthesis kinetics

The development of a kinetic equation to cover a wide CO conversion range (e.g., 7–90%) is a complex task. For the lower CO conversion range, the FTS rate appears to be more rapid than the WGS rate, whereas for the higher conversion region the FTS reaction rate is close to that of the WGS reaction (Fig. 1, Fig. 3). If this is the case, the kinetic model would appear to require a different form of the rate equation for the low- and high-conversion regions for CO and some form of an equation to

Conclusions

The kinetics of FTS and WGS reactions over 100 Fe/5.1 Si/1.25 K/2.0Cu catalyst were studied in two CO conversion regions, i.e., a low level (7–63%) and a high level (70–87%), using a 1-L CSTR reactor at 270 °C and wide ranges of pressures (1.3–2.5 MPa) and H2/CO ratios (0.67–1.5). Thirteen FTS models and thirteen WGS models either from the literature or constructed in this study were tested and discriminated using 83 sets of kinetic data. The new FTS and WGS models were constructed by considering

Acknowledgments

The authors would like to thank NASA for financial support (NASA research contract NNX07AB93A), as well as the Commonwealth of Kentucky.

References (37)

  • E. van Steen et al.

    Appl. Catal. A: Gen.

    (1999)
  • G.P. van der Laan et al.

    Appl. Catal. A: Gen.

    (2000)
  • M.E. Dry et al.

    J. Catal.

    (1972)
  • G. Bub et al.

    Chem. Eng. Sci.

    (1980)
  • R.A. Dictor et al.

    J. Catal.

    (1986)
  • H. Schulz

    Appl. Catal.

    (1999)
  • A.P. Raje et al.

    Catal Today

    (1997)
  • Y. Wang et al.

    Fuel

    (2003)
  • F.G. Botes

    Appl. Catal.

    (2007)
  • A.N. Pour et al.

    J. Nat. Gas Chem.

    (2010)
  • Y. Li et al.

    Appl. Catal. B: Env.

    (2000)
  • C.S. Kellner et al.

    J. Catal.

    (1981)
  • A.P. Raje et al.

    J. Catal.

    (1998)
  • M.E. Dry et al.

    J. Catal.

    (1969)
  • R.B. Anderson

    Kinetics and reaction mechanism of the Fischer–Tropsch synthesis

  • S. Ledakowicz et al.

    Ind. Eng. Chem. Process Des. Dev.

    (1985)
  • G.A. Huff et al.

    Ind. Eng. Chem. Process Des. Dev.

    (1984)
  • F.G. Botes et al.

    Ind. Eng. Chem. Res.

    (2006)
  • Cited by (27)

    • Renewable and sustainable energy from CO<inf>2</inf> following the green process

      2023, Green Chemistry Approaches to Environmental Sustainability: Status, Challenges and Prospective
    • A hybrid kinetics integrating feed-consumption rate and product selectivity models for Fischer-Tropsch synthesis over an industrial cobalt-based catalyst

      2023, Chemical Engineering Journal
      Citation Excerpt :

      Furthermore, kinetic models for raw materials consumption or products formation have practical importance for engineering design of FTS reactor and the industrial process optimization [16]. Over the past decades, numerous kinetic studies of the FTS have been performed over Cobalt [17–28], Iron [16,28–36], or Ruthenium-based [37,38] catalyst. Moreover, for the prediction of FTS reaction behavior and product distribution, different types of kinetic model have been published, such as the hybrid model (a combination of raw material consumption rate model and product distribution model), or the detailed mechanic kinetic model [31,39].

    • Active phases for high temperature Fischer-Tropsch synthesis in the silica supported iron catalysts promoted with antimony and tin

      2021, Applied Catalysis B: Environmental
      Citation Excerpt :

      Let us discuss the effect of the promotion with antimony and tin on the characteristics of iron catalysts such as dispersion of active phase and extent of carbidization. There is a general consensus in the literature that the activity of iron catalysis in FT synthesis can be principally attributed to iron carbides, though iron oxides can contribute in a lesser extent by affecting to some extent the intrinsic activity of iron carbide species, enhancing water gas shift and secondary reactions [44–46]. Iron dispersion and extent of iron carbidization are therefore, important parameters, which should be considered in the interpretation of the catalytic data.

    View all citing articles on Scopus
    View full text