Cobalt Catalyzed Fischer-Tropsch Synthesis with O₂-Containing Syngas

Abstract

Provision of sustainable transportation fuels is required for the energetic transition. A new process is presented for the production of synthetic sulfur free maritime fuel. This fuel is produced by Co-catalyzed Fischer-Tropsch synthesis (FTS) using syngas based on a plasma technology that contains traces of O₂. Gravimetric experiments and steady state measurements with a Co/Pt/Al₂O₃ catalyst at low temperature FTS conditions (10–30 bar, 180–230 °C) show that, with H₂ present in the system, the catalyst remains active for FTS, and shows no influence on the distribution of C₂₊-hydrocarbons. O₂ is only converted to H₂O and CO₂ in varying proportions (H₂O: 70–80%, CO₂: 20–30%), whereby a higher CO concentration increases the CO₂ selectivity. This work may wield a new CO₂ source for carbon-neutral fuels.

1. Introduction

Many efforts have been made to mitigate CO₂ emissions contributing to climate change. Nevertheless, the CO₂ output is currently still rising, mainly driven by a huge fossil fuel demand [1,2,3,4]. Therefore, the need for alternative energy resources in all sectors of the world’s economy is imperative, above all in transportation, which today almost completely relies on oil-derived fuels such as gasoline, diesel/marine oil, and jet fuel.

The supply of the required electrical energy by renewables seems to be in principle possible on the long run, but their availability is fluctuating, and storage solutions are essential for a successful transformation of the current energy infrastructure mainly based on fossil fuels to a supply by renewables. Therefore, energy storage via chemicals is an important focus of today’s energy research. Especially the power-to-liquid (PtL) approach yielding synthetic liquid fuels attracts attention both in industry and academia [5,6,7].

Long-distance travel by ship or airplanes depends on fuels with a high gravimetric and volumetric energy density such as liquid hydrocarbons, and their substitution by battery electric or hydrogen fuel cell driven systems is not viable [8,9]. Hence, synthetic liquid fuels providing the required energy density will be needed to reduce the CO₂ emissions related to transportation by air and shipping [10,11]. In particular, the maritime sector is strongly looking for an alternative fuel. As the cargo-shipping business is highly competitive, transportation costs are important. As a result, the shipping companies today rely on heavy bunker oil which is low in price but also low in quality, i.e., the fuel contains high amounts of organic sulfur and nitrogen compounds, which leads to high emissions of SO_x, NO_x, and soot [12]. In order to solve this problem, the International Shipping Organization ruled new emission limits for maritime fuel. As the current fossil-based heavy oils by far do not meet these criteria, other kinds of fuels become attractive [11]. Considering that a whole infrastructure based on hydrocarbon fuels already exsists, the use of synthetic fuels as a substitute for heavy bunker oil is regarded as very promising [13].

Although the main problem with fossil fuels is the output of carbon dioxide, CO₂ can be beneficial for other processes. The direct capture of carbon dioxide from large industrial plants with high CO₂ pollution, like steel mills and cement production, or the use of CO₂ formed as by-product in the chemical industry (e.g., NH₃ synthesis) can prevent—to a certain extent—the increase in atmospheric CO₂ and may serve as a carbon source for energy storage via synthetic fuels [10]. When direct capture of CO₂ and H₂O electrolysis (to provide H₂) based on renewable electricity (wind, hydro, solar) are joined skillfully, a Fischer-Tropsch plant can be operated to produce various green synthetic fuels, above all jet and maritime fuel. The latter is a mixture of long-chain linear hydrocarbons, thus low-temperature Co-catalyzed Fischer-Tropsch synthesis (LT-FTS) is the best choice for production [14]. Here, a multi-tubular fixed-bed reactor cooled by boiling water is typically used.

In a joint project of MCT Transformatoren GmbH, Overspeed GmbH, Institute for Photovoltaics of the University of Stuttgart, and the Chair of Chemical Engineering of the University of Bayreuth, such a process has been investigated. A process scheme is shown in Figure 1.

In this process (Figure 1), CO₂ and H₂O are converted in one step to CO and H₂ by renewable electrical energy in a plasma-based CO₂ conversion combined with H₂O electrolysis, respectively, finally yielding the syngas needed for FTS; for details, see [15,16]. In this route, O₂ occurs as byproduct of the plasma process and can not be completely removed. Hence, FTS must be run with syngas containing traces of molecular O₂ (up to 2 vol.%) [15,16]. An alternative would be the catalytic conversion of oxygen with H₂ or CO in an upstream reactor, as also discussed below.

As only about 20% of the CO₂ is converted in the plasma process to CO per pass, CO₂ remains in the syngas as well. Previous works show that only for a high CO₂ concentration combined with a very low CO concentration carbon dioxide does affect the FTS—if cobalt is used as catalyst—otherwise, CO₂ acts as inert [17]. The latter is the case here. As seen in Figure 1, syngas is converted in the FTS unit to maritime fuel (and lighter hydrocarbons). Unconverted syngas and short chained hydrocarbons (such as C₁–C₄) are recycled into the plasma reactor to increase the overall CO₂ conversion to maritime fuel. The oxygen from the syngas production can be used in other industrial chemical processes such as the partial oxidation of methanol to formaldehyde [18].

Because no data exists for this process, the following questions had to be answered in this work with regard to the oxygen present in the syngas:

• Does the FT catalyst remain active for FTS in the presence of O₂?
• If the answer of 1. is yes, does the product composition of the FTS change?
• Where does the O₂ end up (selectivity to H₂O and CO₂)?
• What has to be considered in terms of process/FTS reactor safety?

In this work, the focus is on the general influence of oxygen and not on details of the consequences with regard to the detailed design and runaway aspects of an FT reactor.

The reaction scheme (Equations (1)–(4)) suggests that O₂ present in the feed gas only ends up in CO₂ and H₂O during FTS, but the possible oxidation of cobalt to CoO (not active for FTS [19]) is still also an issue. As CO₂ and H₂O do not affect the FTS reaction with Co as catalyst, at least an unwanted—but here unavoidable—loss of syngas (H₂, CO) occurs [14].

2 H₂ + CO → (-CH₂-) + H₂O_(g) Δ_RH₀²⁹⁸ = −152 kJ mol⁻¹

(1)

3 H₂ + CO ⇌ CH₄ + H₂O_(g) Δ_RH₀²⁹⁸ = −206 kJ mol⁻¹

(2)

2 H₂ + O₂ → 2 H₂O_(g) Δ_RH₀²⁹⁸ = −484 kJ mol⁻¹

(3)

2 CO + O₂ → 2 CO₂ Δ_RH₀²⁹⁸ = −566 kJ mol⁻¹

(4)

With regard to thermodynamics, O₂ is favourably converted with CO to CO₂, e.g., for a syngas consisting of 1% O₂, 66% H₂, and 33% CO and typical FT conditions (30 bar, 230 °C); hence, the selectivity to CO₂ is ca. 99% with regard to thermodynamics. However, this value may not be reached with regard to reaction kinetics, as under stoichiometric FT conditions, the H₂ concentration exceeds the CO concentration twofold.

To our knowledge, the use of O₂-containing syngas in Fischer-Tropsch synthesis has never been considered before. Since this would expand the use of FTS to a syngas from new sources, it is a very important finding that we present. The results presented herein show a new power-to-liquid approach with a highly interesting syngas source and the possibilty to use CO₂ for maritime fuel production.

2. Results and Discussion

2.1. Stability of the FT Co Catalyst against O₂ in O₂-Containing Gas Mixtures

At first, the stability of the activated (reduced) FT Co catalyst was studied in different O₂-containing gas mixtures. Since cobalt is only active for FTS in metallic state [20,21], the crucial question was whether Co remains metallic or is oxidized. Hence, the catalyst’s mass was monitored by a magnetic suspension balance (MSB). For example, a complete oxidation of cobalt to CoO would lead to an increase in the catalyst’s mass (with 10 wt.% Co) by 2.7%. Two feed gases were initially used, 99% N₂/1% O₂ and 39% N₂/60% H₂/1% O₂; a typical FT temperature of 230 °C was thereby applied. H₂ is expected to hinder the oxidation of Co as only a small amount of O₂ (1%) is present in the feed gas. During these experiments, no CO was added to avoid FTS to take place, as this leads to an increase in the catalyst’s mass by filling of pores with higher hydrocarbons, which would falsify the signal of the MSB. Thus, only the reaction of O₂ with H₂ to H₂O and/or the oxidation of Co was measured. For comparison, the catalyst mass was also observed in a 1% H₂O/99% N₂ mixture. After contact with one of the three gases, the catalyst batches were dried in pure N₂ followed by pure H₂; the initial temperature of 230 °C was thereby finally increased to 360 °C. Experimental details are given in Section 3.1. The results are shown in Figure 2.

In case of the O₂-containing but H₂-free feed gas (Figure 2a), a fast (relative) increase in the weight of the catalyst by 2.4% is observed at 230 °C. No O₂ conversion was detected via the gas analyzer. Switching to pure N₂ at 230 °C leaves the catalyst mass unaffected and only a small decrease is observed after rising the temperature to 360 °C. This implies that the mass increase is caused by the oxidation of Co. Previous investigations have shown that in a H₂ atmosphere, temperatures ≥ 360 °C are required to activate (reduce) the catalyst [22]. Therefore, cooling the reactor to 230 °C and changing from N₂ to H₂ does not decrease the catalyst’s mass whereas the initial weight, i.e., the previous achieved reduction degree, is finally reached after several hours at 360 °C. Thus, the catalyst is oxidized if O₂ is present without a reducing agent, which is expected since formation of cobalt oxides (CoO) is then favorable. The Co dispersion on the surface was determined to 10% [22]. If only cobalt atoms directly exposed to the feed gas react with O₂, oxidation of cobalt to CoO at the surface would lead to a mass increase of only about 0.3%. Hence, the experimentally found value of 2.4% indicates that practically all Co atoms and not only the surface atoms are oxidized in an O₂-containing feed, if H₂ is not present. XRD measurements support this conclusion, showing the formation of only CoO in a H₂-free O₂-containing feed at elevated temperature and pressure (see Figure S1 in the Supporting Information).

In Figure 2b, O₂ as well as H₂ are present in the feed gas. Again, the catalyst’s mass increases. However, the mass increase by 1.2% at 230 °C is only half the value compared to case a. Furthermore, the initial catalyst mass is again achieved in pure N₂, if the temperature is increased to 360 °C (no change at 230 °C). Hence, adsorption (of H₂O) instead of Co oxidation obviously causes the increase in the mass of the catalyst in the presence of H₂ and O₂, and cobalt remains in its metallic form. During this experiment, the conversion of O₂ is constant (here about 50%). This is also a clear indication that H₂O is formed by H₂ and O₂ and then adsorbs on the catalyst; in case of Co oxidation, the conversion of O₂ would decline with time on stream, if cobalt is more and more fully oxidized. In order to prove this hypothesis, a 1% H₂O/99% N₂ gas mixture was passed over the reduced catalyst. Hence, this content of steam is the value resulting from the measured oxygen conversion of about 50% for an initial O₂ content of 1%. The weight gain of the catalyst was then also 1.2% (Figure 2c), confirming that the formed H₂O adsorbs on a Co/Pt/Al₂O₃ catalyst in a H₂/O₂/N₂ gas mixture. Moreover, the initial catalyst weight is again obtained by drying (desorption) in pure N₂ at 360 °C as in Figure 2b. Similar measurements at a lower temperature (165 °C) show the same trend supporting this explanation (see

Table 1. Maximum weight change of FTS particle after exposure with different gas mixtures. Conditions: T = 165–360 °C; p = 10 bar; $\dot{V}$_total = 13 L/h (NTP); p_H2 = 0–10 bar, p_N2 = 0–10 bar; p_O2 = 0–0.1 bar; p_H2O = 0–0.1 bar; initial m_cat (Co/Pt/Al₂O₃) = 1 g.

	Case A: Rel. Weight Change in O2	Case B: Rel. Weight Change in H2/O2	Case C: Rel. Weight Change in H2O
Feed gas	99% N2, 1% O2	60% H2, 39% N2, 1% O2	98% N2, 2% H2O
165 °C	+2.3%	+2.4%	+2.1%
230 °C	+2.5%	+1.2%	+1.1%

).

Table 1 summarizes the results: The weight gain caused by catalyst oxidation (case a) remains at around 2.5% at 165 °C, revealing that almost all cobalt atoms are then already oxidized to CoO. In case of a H₂/O₂ gas mixture (case b), the weight gain depends on temperature and is by a factor of two lower at 165 °C than at 230 °C. This trend also confirms H₂O adsorption (in a H₂/O₂ gas), as the (exothermic) adsorption is in general favored by a lower temperature. In both cases, the mass increase can be undone by drying in pure N₂ at 360 °C, confirming that the catalyst weight increase in a H₂/O₂ gas feed is due to surface adsorption of H₂O.

2.2. Reactivity of Co Catalyst for Fischer-Tropsch Synthesis with O₂-Containing Syngas

After the stability of Co against O₂ in the presence of H₂ was proven, the FTS was carried out with the Co/Pt/Al₂O₃-catalyst in O₂-containing syngas. The experiments with the fixed bed reactor were conducted to check whether O₂ prevents FTS or alters the product distribution. Initially, the FTS was run with O₂-free syngas (72 h). Then, syngas with 2% O₂ was used for further 72 h. Finally, O₂-free syngas was again used for 72 h to check if deactivation has occurred. The N₂ content was 23 or 25% to ensure constant partial pressures of H₂ and CO (ratio of 2) for a syngas with or without 2% O₂. In addition, the dilution with N₂ helps to ensure isothermal conditions, although numerous exothermic reactions may take place, see Equations (1)–(4). The results with regard to CO conversion and methane selectivity are shown in Figure 3.

A constant CO conversion of 9% is reached for the first and third run (O₂-free syngas). Thus, the O₂ present in the second run does not lead to a change of the FT activity of the catalyst. In case of the O₂-containing syngas, the CO conversion is 10.5% and thus slightly higher compared to O₂-free syngas (9%), which can be attributed to the formation of CO₂ by reaction of CO with O₂. This is confirmed by the analysis of the produced hydrocarbons, which show no change of the product composition for an O₂-free or O₂-containing syngas. The O₂ conversion is 40%. Based on the amount of unconverted O₂ and formed CO₂ (determined based on each content detected via the gas analyzers downstream the reactor), it could be calculated that if all converted O₂ had reacted to CO_2, the CO conversion should have been 16% instead of the real value of 10.5% (for a CO feed content of 25%). In return, this leads to the conclusion that 21% of the converted O₂ is found in CO₂ (S_CO2 = 21%, S_H2O = 79%).

The CH₄ selectivity decreases if O₂ (2%) is present in the syngas and drops from 24 to 16%. It does not return to the initial value, if syngas without O₂ is then again used (Figure 3). A reasonable explanation is adsorbed H₂O on the catalyst. According to Bertole, adsorbed H₂O leads to a decreased CO dissociation barrier and thereby to a higher hydrocarbon chain propagation rate due to more activated carbon [23], i.e., in return, to less methane formation. As H₂O stays adsorbed on the catalyst at the chosen temperature (see Figure 2c), this effect may still be present even if O₂ is no longer present in the feed gas. This effect, although rather small, is beneficial as a yield of higher hydrocarbons (HCs) and a lower yield of CH₄ are desired. Hence, more maritime fuel, i.e., C₁₁₊-HCs, are obtained.

2.3. Selectivity of Reaction of Oxygen (to CO₂ or H₂O) If Present in the Syngas of FTS

Knowing that the FT catalyst is still active in O₂-containing syngas and that oxygen does not affect the FTS other than reducing to a small extent the CH₄ selectivity, the focus is subsequently on O₂ consumption and selectivity to CO₂ and H₂O, which turned out to be the only products. For a detailed understanding of the O₂ reaction, only the influence of the FTS had to be minimized: The FTS can have an influence because strongly exothermic reactions (Equations (1) and (2)) take place, and the conversion of CO by reaction of oxygen is hard to detect accurately, if conversion by FTS is dominating. Consequently, the CO conversion by FTS must be as low as possible, but the O₂ consumption rate also decreases with lower CO conversion, e.g., by a lower temperature or residence time. However, for example, for an O₂ conversion of about 20%, the CO conversion by FTS is then only around 2% (Figure 4). As a result, the selectivity of H₂O and CO₂ formation from O₂ could be determined more precisely as a function of temperature, pressure, and concentration of O₂, CO, and H₂.

At first, the temperature was varied. Figure 4a shows the reaction rate of CO (with and without 1% O₂), the rate of O₂ consumption, as well as the formation rate of CO₂; the corresponding selectivities to CO₂ and H₂O are also shown (Figure 4b). The adjusted O₂ conversion was about 20%.

Figure 4a depicts that the increase in the CO reaction rate by addition of O₂ is always equivalent to the formation rate of CO₂. This again indicates that the FTS only is not affected by O₂: For temperatures between 205 and 225 °C, the CO rate varies from 0.87 × 10⁻² to 2.65 × 10⁻² mol kg_cat⁻¹ s⁻¹ in case of 1% O₂ in the syngas, and from 0.7 × 10⁻² to 2.4 × 10⁻² mol kg_cat⁻¹ s⁻¹ for an O₂-free syngas. The difference of the CO rates matches the rate of CO₂ formation (0.19 × 10⁻² to 0.25 × 10⁻² mol kg_cat⁻¹ s⁻¹). The rate of O₂ consumption (0.36 × 10⁻² to 0.5 × 10⁻² mol kg_cat⁻¹ s⁻¹) is much higher compared to the case of CO₂ formation only (r_CO2 = 2 r_O2), indicating that in the given temperature range, around 75% of O₂ are converted to H₂O (Figure 4b).

In order to check whether Co and/or Pt catalyzes the O₂ consumption, experiments were conducted with catalysts not containing Co or Pt. In both cases, the individual metal content was the same as of the standard FT catalyst (10 wt.% Co, 0.03 wt.% Pt). The experiments were carried out with an O₂-containing syngas (60% H₂, 30% CO, 9% N₂ and 1% O₂) at 20 bar and 180 °C; the O₂ conversion was held at 20% by variation of the residence time. In addition, the Al₂O₃ support (without Co and Pt) was also tested. The results are shown in

Table 2. Selectivity of O₂ reaction for different catalysts and the support material. Conditions: p = 20 bar; T = 180 °C; $\dot{V}$_total = 7.5–75 L/h (NTP); p_CO = 6 bar; p_H2 = 12 bar, p_N2 = 1.8 bar; p_O2 = 0.2 bar; m_cat = 0.4–1.2 g; X_O2 = 0.2.

Catalyst	Pre-Reduction at 360 °C	Oxidation State of Co and Pt	rO2 in molO2 kgcat−1 s−1	SCO2 in %	SH2O in %
Al2O3	No	-	-	no O2 conversion no FTS-activity
	Yes	-	-
9.7 wt.% Co rest Al2O3	No	+2 (CoO)	0.3 × 10−2	80 ± 3	20 ± 3
	Yes	+2 (CoO)	0.3 × 10−2	80 ± 3	20 ± 3
0.03 wt.% Pt rest Al2O3	No	0 (Pt)	0.4 × 10−2	58 ± 3	42 ± 3
	Yes	0 (Pt)	0.4 × 10−2	58 ± 3	42 ± 3
10 wt.% Co, 0.03 wt.% Pt rest Al2O3	No	+2 (CoO), 0 (Pt)	0.7 × 10−2	63 ± 3	37 ± 3
	yes, fresh catalyst	0 (Co), 0 (Pt)	3.5 × 10−2	36 ± 3	64 ± 3
	yes, cat. In steady state (100 h TOS)	0 (Co), 0 (Pt)	0.2 × 10−2	27 ± 3	73 ± 3

. Each catalyst was tested with or without pre-treatment (reduction at 360 °C in pure H₂).

With the Al₂O₃ support, only no reaction (FTS or O₂-consumption) occurred (Table 2). If the catalyst containing only Co is used, FTS does not take place, and the CO₂ formation rate is four times faster than the H₂O formation, i.e., the selectivity to CO₂ is always 80%. Note that in this case, the pre-reduction does not lead to metallic Co, as Pt is needed as reduction promotor to convert CoO to Co at 360 °C, as confirmed by thermogravimetric analysis (see Figure S2 in Supporting Information) and also reported in [22]. This explains that FTS is suppressed as metallic Co is needed.

If only platinum on Al₂O₃ is used, the selectivity to CO₂ is always 58% (H₂O selectivity 42%), and the rate of O₂ consumption is then by a factor of 1.33 higher compared to the Co catalyst (without Pt). Again, the application of the pre-reduction does not influence the catalyst’s activity as Pt is always in the metallic state. If the Co/Pt/Al₂O₃ catalyst is used without pre-reduction (hence we have CoO), the resulting O₂ consumption rate is higher compared to Pt only, which reflects the contribution of CoO to the rate of O₂ consumption.

If the Co/Pt/Al₂O₃ catalyst is pre-reduced at 360 °C, this leads to a reduction in CoO due to Pt as reduction promotor [22]. Then, the oxygen consumption rate is very high, e.g., by a factor of 5 compared to the same catalyst without pre-reduction. Hence, metallic Co is quite active for conversion of oxygen. The selectivity to CO₂ with the reduced Co/Pt catalyst is about 36%.

For steady-state conditions, reached after 100 h TOS, the pores of the catalyst are completely filled with liquid HCs [15], and the rate of O₂ consumption is then strongly reduced to only about 6% of the value reached at identical conditions with the fresh catalyst. Hence, the liquid higher hydrocarbons present in the pores obviously limit the oxygen transport to the catalyst surface and act as a protective layer.

2.4. Selectivity of O₂ Reaction in Steady-State Operation of Fischer-Tropsch Synthesis

The influence of the total pressure and of the O₂ content in the feed gas on oxygen consumption at a temperature of 215 °C is depicted in Figure 5. These measurements were conducted with the pre-reduced Co/Pt/Al₂O₃ catalyst under steady state conditions of FTS. Figure 5 reveals that both the influence of the O₂ content and of the total pressure on the selectivity to CO₂ and H₂O is small (Figure 5a).

The O₂ reaction rate almost linearly increases with the O₂ content (at constant total pressure) and with pressure (at constant O₂ content) (Figure 5b). Hence, the rate is first order with regard to O₂. To investigate the influence of the H₂ and CO content on the O₂ consumption, the feed gas content of these two syngas components were varied individually while the concentration of the other component was kept constant by the respective content of N₂ in the syngas. Again, steady state FT conditions were applied (TOS > 100 h). The results (again at 215 °C) are presented in Figure 6 and Figure 7. Both figures (cases a) show that the influence of the content of H₂ and of CO on the selectivity to CO₂ and H₂O is marginal or rather low, respectively: The CO₂ selectivity decreases from 30 to 27% for a variation of the H₂ content between 7 to 60% (Figure 6a). For CO, the increase in the content from 3 to 30% leads to an increase in the CO₂ selectivity from 21 to 33% (Figure 7a). For both gases, each content has no measurable influence on the consumption rate of O₂ (Figure 6b and Figure 7b).

3. Materials and Methods

Two different sets of experiments were conducted to study the effect of oxygen on FTS: The change of mass of the catalyst was studied gravimetrical in a magnetic suspension balance. In addition, the activity and selectivity of the catalyst both for FTS and conversion of the oxygen present in the syngas (to either CO₂ or H₂O) were measured in a classical fixed bed reactor with and without O₂-containing syngas at temperatures typical for LT-FTS (180–230 °C).

3.1. Measurements in a Magnetic Suspension Balance

The setup of the gravimatrical measurements (Figure 8) consists of four individual gas supply lines for CO, H₂, N₂, and air, respectively. The flow of each gas is regulated by mass flow controllers (Bronkhorst F-201V, Bronkhorst High-Tech B.V., Ruurlo, The Netherlands). The measurements of the Co-based catalyst stability against O₂ were conducted in a magnetic suspension balance (MSB) (TA Instruments former RUBOTHERM, New Castle, DE, USA). The catalyst particles are put in a steel retainer inside the MSB; the MSB device is electrically heated, and the temperature is measured and controlled by a dual PT-100 thermocouple) (TA Instruments former RUBOTHERM, New Castle, DE, USA). Condensable products (i.e., higher hydrocarbons and H₂O) are collected in two serial cooling traps, one trap at room temperature and the other trap at 0 °C. The gas leaving the cooling traps passes through a gas analyzer (Emerson X-Stream Enhanced XEPG, Emerson Electric Company, St. Louis, MO, USA) and the gas flow is measured in a soap bubble flow meter at the outlet of the setup.

Prior to the FTS experiments, the cobalt catalyst is reduced and activated in pure H₂ for 16 h at 360 °C and then cooled to 100 °C. Thereafter, the partial pressures of the desired gases are adjusted, and the gas mixture passes through the MSB for 24 h at 165 °C or 230 °C. Afterwards, pure N₂ is introduced to remove adsorbed species from the catalyst surface. The desorption is performed in two steps. First, drying is conducted for at least 3 h under the chosen temperature, i.e., at either 165 °C or 230 °C, until the catalyst mass remains constant. Then, the reactor is heated up to the maximum temperature of 360 °C (5 K min⁻¹). The heating is stopped when the mass signal remains constant or the initial mass of the activated catalyst is restored. If the mass is still above the initial mass, the drying procedure is repeated, but instead of N₂ pure H₂ is used to reduce the catalyst until the initial mass is regained.

3.2. Fischer-Tropsch Synthesis

FTS experiments are conducted in a fixed-bed reactor. The mass flows of the reactant gases CO, H₂, N₂, and O₂ are supplied and controlled by mass flow controllers (Bronkhorst EL-Flow Prestige, Bronkhorst High-Tech B.V., Ruurlo, The Netherlands). The reactor is an electrical heated steel tube that is 600 mm in length and 15 mm in diameter; the packed catalyst bed has a height of 150 mm. A guiding tube in the center of the reactor (6 mm diameter) over the total length was used to measure the axial temperature profile by means of a movable thermocouple; for the given conditions, it turned out that the fixed bed could be regarded as isothermal. Figure 9 shows the setup for the FTS experiments.

The product gas stream passes through three cooling traps to collect the condensable products. These cooling traps are connected in series at 120 °C, room temperature, and 0 °C. The remaining gaseous products are analyzed by a GC (Perkin Elmer Clarus^® 690 GC, PerkinElmer Inc., Waltham, MA, USA). In order to protect the following analytical device, a dry-ice-cooling trap is installed to remove the remaining water from the gas stream. The dry gas stream is analyzed (CO, CH₄, and CO₂ content) by an IR gas analyzer (Emerson X-Stream Enhanced Emerson Electric Company, St. Louis, MO, USA) and a paramagnetic oxygen sensor. Finally, the gas flow is measured via a gas meter (Ritter Apparatebau drum–type gas meter; type TG-5, Dr.-Ing. RITTER Apparatebau GmbH & Co. KG, Bochum, Germany) or a soap bubble flow meter. The collected condensable products are analyzed in a second GC (Bruker Varian CP-3800, Bruker Corporation, Billerica, MA, USA) in periodic time intervals during the reaction.

The catalysts are inhouse-made catalysts containing 10 wt.% cobalt and/or 0.03 wt.% Pt supported on 5 × 5 mm γ-Al₂O₃ particles (Sasol Germany GmbH, Hamburg, Germany). The metals are deposited by wet-impregnation. Therefore, the support is stirred in an aqueous solution of Co(NO₃)₂⋅6 H₂O (>98%, Carl Roth GmbH + Co. KG, Karlsruhe, Germany) and/or Pt(NH₃)₄(NO₃)₂ (Alfa Aesar, Haverhill, MA, USA) in a round-bottomed flask, being part of a rotovap. After applying a reduced pressure of 30 Pa three times to improve the pore filling, the flask is rotated to guarantee homogeneous metal dispersion. The resulting particles are dried at room temperature for at least two days, then calcinated in an air stream heated up in 5 K min⁻¹ steps to 360 °C, and then held for 3 h. In

Table 3. Characteristics of the synthesized catalysts used in this work.

Parameter	Co/Pt/Al2O3-Catalyst	Co/Al2O3-Catalyst	Pt/Al2O3-Catalyst
Chemical composition	0.031 wt.% Pt, 10 wt.% Co, rest γ-Al2O3	9.7 wt.% Co, rest γ-Al2O3	0.03 wt.% Pt, rest γ-Al2O3
Size of cylindrical particles	5 × 5 mm	5 × 5 mm	5 × 5 mm
BET surface area	188 m2 gCat−1	196 m2 gCat−1	205 m2 gCat−1
Pore volume	0.36 cm3 gCat−1	0.4 cm3 gCat−1	0.43 cm3 gCat−1
Porosity εp	0.53	0.53	0.53

, the characteristics of the catalyst are summarized.

In order to achieve isothermal conditions inside the catalyst bed, the catalyst is diluted with quartz sand (d_p = 250 μm). The mixture is filled in the reactor and after a leak test, the catalyst is activated in 100% H₂ at 360 °C for 16 h. After reduction, the catalyst is cooled to 150 °C; reoxidation is thereby excluded by a gas phase still rich in hydrogen (H₂-to-N₂ ratio of 2). After initiating the FTS reaction by introduction of syngas (here a mixture of CO, H₂, N₂, and traces of O₂), the reactor is heated up to reaction temperature and the temperature is held for at least 100 h to ensure steady state operation. At unsteady state conditions the reaction rate would decrease over time. The cause of the disrobed behavior is the ongoing hydrocarbon filling that builds an increasing diffusion barrier [24]. Afterwards, the parameters such as CO or H₂ concentration, temperature, or pressure were changed individually and held for a further 72 h to observe the effect of each parameter.

3.3. Evaluation of Experimental Data

The CO conversion $X_{CO}$ and O₂ conversion $X_{O_2}$ are calculated according to Equations (5) and (6), respectively,

$$X_{CO}=\frac{{\dot{n}}_{CO,0}-{\dot{n}}_{CO}}{{\dot{n}}_{CO,0}}$$

(5)

$$X_{O_2}=\frac{{\dot{n}}_{O_2,0}-{\dot{n}}_{O_2}}{{\dot{n}}_{O_2,0}}$$

(6)

with ${\dot{n}}_i$ as the molar gas flow of component i. ${\dot{n}}_{i,0}$ and ${\dot{n}}_i$ is the in- and outgoing molar flow of component i (here O₂ or CO). Based on the CO conversion, the selectivity S for methane and higher hydrocarbons can be determined according to the Equations (7) and (8).

$$S_{C{H}_4}=\frac{{\dot{n}}_{C{H}_4}}{{\dot{n}}_{CO,0}-{\dot{n}}_{CO}}$$

(7)

$$S_{C_{2+}}=\frac{{\dot{n}}_{C_{2+}}}{{\dot{n}}_{CO,0}-{\dot{n}}_{CO}}$$

(8)

The selectivities of O₂ to H₂O and CO₂ ($S_{C{O}_2}, S_{H_{2}O}$) are determined by Equations (9) and (10). For the H₂O selectivity, the selectivity of CO₂ is used since the CO₂ content in the product gas is easily and accurately obtained by online gas analysis.

$$S_{C{O}_2}=\frac{{\dot{n}}_{C{O}_2}}{{\dot{n}}_{O_2,0}-{\dot{n}}_{O_2}}$$

(9)

$$S_{H_{2}O}=1-S_{C{O}_2}$$

(10)

The reaction rate related to the mass of catalyst is calculated via Equation (11)

$${\dot{r}}_i=\frac{{\dot{n}}_{i,0}-{\dot{n}}_i}{m_{cat}}$$

(11)

where $m_{cat}$ is the mass of the catalyst and ${\dot{n}}_i$ the molar gas flow of compound i.

4. Conclusions

This work shows that Fischer-Tropsch fixed bed synthesis can be applied with syngas containing traces of O₂ (e.g., about 1%, if syngas is produced in plasma based processes) as long as sufficient cooling of the reactor (tubes) is applied. During FTS with oxygen in the syngas, the catalyst does not oxidize and the O₂ is converted exclusively to H₂O and CO₂. The FTS (acticity of catalyst, product distribution) is not influenced by O₂, except that a small (unwanted, but for the given case unavoidable) consumption of CO and H₂ by reaction with O₂ takes place. The O₂ reaction, i.e., the formation of H₂O and CO₂, is of first order with regard to O₂, and the selectivities to CO₂ (about 30%) and H₂O (70%) are not altered by the O₂ content of the syngas, and also the influence of the CO and H₂ content on the selectivity is marginal.

For the proposed new synthesis route to produce CO₂-neutral maritime fuel via a plasma-based CO supply, the consumption of O₂ leads to an addtional heat release in the FTS, which may be critical with regard to a temperature runaway of a cooled multi-tubular FT reactor: For only 1% O₂ in the syngas, the adiabatic temperature rise is 170 K (compared to about 1500 K for FTS with syngas consisting of pure H₂ and CO in a ratio of 2). To avoid this additional heat release, O₂, if present in the syngas, could be also completely converted to H₂O and CO₂ by a catalytic conversion step and pre-reactor in front of the FT reactor. A similar case is, for example, also discussed and investigated for the removal of oxygen traces present in raw coke oven gas, if high purity H₂ should be produced by pressure swing adsorption. Respective investigations with regard to the case of FTS are currently conducted, and will be presented elsewhere.