Synthesis, Characterization, and Attrition Resistance of Kaolin and Boehmite Alumina-Reinforced La_0.7Sr_0.3FeO₃ Perovskite Catalysts for Chemical Looping Partial Oxidation of Methane

Abstract

This study investigates the impact of kaolin and boehmite alumina binders on the synthesis, catalytic properties, and attrition resistance of a La_0.7Sr_0.3FeO₃ (LSF) perovskite catalyst designed for the chemical looping partial oxidation (CLPO) of methane to produce synthesis gas sustainably. The as-synthesized and used catalysts with varying kaolin and boehmite alumina contents (KB_(x,y)/LSF) were scrutinized by a variety of characterization methods, including XRD, FE-SEM/EDS, BET, TPD-NH₃, and TPD-O₂ techniques. The catalytic activity of the synthesized samples was tested at 800 to 900 °C in a fixed-bed reactor producing syngas through the CLPO process over the consecutive redox cycles. Additionally, the attrition resistance of the fresh and used catalyst samples was examined in a jet cup apparatus to assess their durability against the stresses induced by thermal shocks or changes in the crystal lattice caused by chemical reactions. The characterization results showed the pure perovskite crystal structure of KB_(x,y)/LSF catalysts demonstrating adequate oxygen adsorption capacity, effective coke mitigation capability, robust thermal stability, and resilience to agglomeration during repetitive redox cycles. Among the tested catalysts, KB_(25,15)/LSF was identified as the superior sample, as it could consistently produce syngas with a suitable H₂:CO molar ratio varying from 2 to 3 within ten redox cycles at 900 °C, with CH₄ conversion and CO selectivity values up to 64% and 87%, respectively. The synthesized catalysts demonstrated a logarithmic attrition pattern in the jet cup tests at room temperature, featuring high attrition resistance after the erosion of particle shape irregularities or weakly bound particles. Moreover, the KB_(25,15)/LSF catalyst used at 900 °C showed great resistance in the attrition test, warranting its endurance in the face of extraordinarily harsh conditions in fluidized bed reactors employed for the CLPO process.

1. Introduction

The pursuit of sustainable and efficient energy sources has become a global imperative in the face of depleting fossil fuels and escalating environmental concerns. Methane, a common component in natural gas, shale gas, clathrates, and even landfills and biogas, represents one of the most prevalent hydrocarbons on our planet. It is used as fuel for electricity and heat generation, as well as a raw input material for producing substances such as hydrogen for clean energy, fertilizer, and widely used petrochemical products including cosmetics, medicine, synthetic fibers, and plastics.

Among the key pathways for methane utilization is its conversion into syngas, a precursor for a wide range of chemicals. This process encompasses several well-established techniques, including the steam reforming of methane (SRM) as the most mature conversion approach, followed by dry reforming (DRM), and partial oxidation (POM) [1]. POM has garnered attention due to its relatively low energy consumption and its ideal H₂/CO ratio that suits subsequent industrial applications [2]. However, POM necessitates a supply of pure oxygen, making it complex and costly [3]. To address this challenge, the concept of the chemical looping partial oxidation of methane (CLPO) has emerged as a promising approach, sidestepping the need for air separation and offering an economically viable route for carbon dioxide sequestration. Furthermore, it enhances operational safety by isolating methane and pure oxygen [4]. Nevertheless, CLPO is a complex and intricate redox process involving various steps, like the adsorption and dissociation of methane, the formation and detachment of oxygenated intermediates, and the evolution and migration of lattice oxygen in oxygen carriers (OCs). As a result, the performance of oxygen carriers is a crucial factor for the success of the CLPO process [5].

In the CLPO processes, various OCs composed of transition metal oxides, such as Ni, Cu, Fe, Co, and Mn oxides, have been studied. Oxygen vacancies within these transition metal oxides promote methane partial oxidation by reducing the activation barrier of the C—H bond in CH_x radicals, thus enhancing the reaction efficiency. Additionally, inert supports like ZrO₂, TiO₂, SiO₂, Al₂O₃, NiAl₂O₄, and MgAl₂O₄ have been employed to enhance the mechanical strength, reactivity, anti-coking properties, and regenerability of OCs [6,7].

Perovskite-type oxides that are generally represented as ABO₃ (with A typically being alkaline earth or rare earth metal cations and B being transition metal cations) [8], possess enormous potential as oxygen carriers due to their outstanding thermal stability, oxygen mobility, redox recyclability, and catalytic activity [9]. LaFeO₃, a representative perovskite-type oxygen carrier, can be further improved by substituting Fe with certain transition metal cations. This substitution leads to the creation of complex perovskite-type oxides, such as La_(1−x)Sr_xMO₃ (M = Mn, Ni, Fe) [10], LaFe_(1−x)Ni_xO₃, and LaFe_(1−x)Co_xO₃ [11], which have been synthesized and investigated for chemical looping oxidation of methane. For instance, it has been observed that the addition of Sr²⁺ to LaFeO₃ improves the oxygen storage capacity of perovskite materials due to the facilitated reduction of Fe⁴⁺ to Fe³⁺ in the resulting doped La_0.7Sr_0.3FeO₃ (LSF) perovskites, leading to an improved redox performance [12,13].

The LSF catalysts, known for their remarkable oxygen ion conductivity and robust thermal stability, have shown great potential for advancing sustainable syngas production through alternative cycles of methane partial oxidation and CO₂ splitting [14]. These catalysts exhibit a unique blend of properties that make them highly suitable for applications requiring rapid oxygen transfer and high-temperature operations. Despite their advantageous attributes, LSF-based catalysts are not without their challenges such as the issues related to material stability, coke formation, carbon deposition, and catalyst deactivation [15]. In the CLPO reactors, which are running under harsh operating conditions over extended redox cycles, the preservation of the mechanical stability and durability of the LSF catalysts is a major issue that should be addressed. This is particularly the case in fluidized bed reactors, which are popular reactor configurations for the CLPO process due to their enhanced conversion efficiency, reduced unburnt slip, and optimized reaction control [16]. However, in the fluidized bed reactors, the catalytic particles are subject to attrition due to constant collision with each other and the reactor walls [17,18,19]. Adjusting the composition of perovskite materials, including LSF, to improve their reactivity can impact their stability, posing challenges in optimizing these catalysts for efficient and durable performances [20]. For instance, the LSF stability issue becomes more critical when it is doped with elements like nickel that are used to enhance the reaction rates at intermediate temperatures [21].

Research on enhancing the durability of LSF-based catalysts in the CLPO process focuses on innovations to withstand operational demands, including the use of dopants [22], composite materials [23], and protective surface treatments [23], to mitigate the effects of high temperatures, mechanical wear, and chemical contaminants. These improvements aim to extend the service life and performance of LSF-based catalysts, which is critical for sustainable and efficient CLPO-based industrial applications. While the research on perovskites in catalysis often focuses on enhancing the thermal stability, selectivity, and activity of the catalysts [24], the studies carried out on enhancing the attrition resistance of these catalysts are scarce.

With the aim of developing an LSF catalyst with adequate catalytic features and enhanced attrition resistance, the present study explores the synthesis, characterization, and optimization of perovskite LSF catalysts modified with kaolin and boehmite alumina as binders. Kaolin and boehmite are both alumina-based compounds known for their ability to support catalytic activities. Although these materials have been exploited to improve the properties of some catalysts used in processes such as fluid catalytic cracking (FCC) [25] or low-temperature CO₂ methanation [26], to the best of our knowledge, there is no evidence that they have been directly used to enhance perovskite catalysts. However, given the adaptability of perovskite structures, it is plausible that kaolin and boehmite could contribute to improving the properties of perovskite catalysts.

Accordingly, this research introduces a novel approach to enhancing the performance and durability of LSF-based catalysts in the CLPO process, by incorporating a blend of kaolin and boehmite alumina, referred to as KB_(x,y)/LSF, where x and y represent the weight percentages of kaolin and boehmite alumina, respectively. The impact of these additives on the performance of the catalysts was investigated by a variety of analytical techniques and reactivity tests. Moreover, the structural stability of the synthesized catalysts with various KB_(x,y)/LSF formulations was assessed through the attrition tests. The findings of these tests provide insights into optimizing the catalyst formulation for enhanced catalytic and structural performance in the CLPO process.

2. Results and Discussion

2.1. Characterization of the Materials

2.1.1. XRD Analysis

Figure 1a demonstrates the XRD patterns of the synthesized KB_(x,y)/LSF. The diffraction pattern of the LSF and its main peaks (occurred at 2θ = 0.86, 5.32, 25.77, 29.52, 52.46, 56.72, 70.57, 71.87, 73.67, 85.22, and 99.39) align with an orthorhombic perovskite phase (powder diffraction file (PDF): 01-089-1269), with no discernible peaks pertaining to La₂O₃, SrO, or Fe₂O₃. The absence of such peaks suggests the formation of a pure perovskite crystal structure within the freshly created KB_(x,y)/LSF.

Regardless of the fraction of kaolin and boehmite alumina, all the samples shown in Figure 1a exhibit an orthorhombic crystal structure, with cell dimensions of a = 502.5 Å, b = 544.5 Å, and c = 846.7 Å, as obtained from the Scherrer equation [27]. In all the samples, the primary phase formed was perovskite, and the phases Al₂O₃ and kaolin were not detectable in the XRD patterns. These observations, which are in agreement with the findings of He et al. [28], corroborate the successful synthesis of the catalyst.

Figure 1b presents a comparison of the XRD patterns between the fresh KB_(25,15)/LSF catalyst and the KB_(25,15)/LSF catalyst used in the fixed-bed reactor at 900 °C over 10 consecutive redox cycles. The consistent peak positions observed at various 2θ values in both the fresh and used samples indicate that the crystal structure of the catalyst remained intact after undergoing 10 redox cycles. This analysis sets the stage for the forthcoming discussion on the KB_(25,15)/LSF catalyst, recognized as the optimal catalyst, which will be further detailed in the following sections.

2.1.2. FE-SEM/EDS Analyses

FE-SEM/EDS analyses were carried out to investigate the microstructure of the KB_(25,15)/LSF catalyst. Figure 2 illustrates the porous structure of the fresh catalysts, and the catalysts used after 10 redox cycles at 900 °C.

The FE-SEM image of the used catalyst (Figure 2c) closely resembles that of the fresh catalyst (Figure 2a) with no noticeable particle aggregation as a result of its exposure to high temperature and redox conditions. The stability of the catalyst during the examined redox cycles was also understood by assessing the comparability between the chemical composition of the fresh and used catalysts obtained from the EDS analyses shown in Figure 2b,d. These observations, which are in line with the results of the XRD analyses (Figure 1), collectively suggest that the microstructure of the catalyst remains physically and chemically consistent throughout 10 redox cycles. The catalyst stability would guarantee the stability of the chemical looping process for continuous syngas production. Upon the examination of the FE-SEM/EDS images of the fresh and used samples, it became apparent that the relatively uniform distribution of the additives across the surface was preserved after using the catalyst under the tested operating conditions.

2.1.3. BET/BJH Analysis

The analysis of the textural properties of the examined catalysts was conducted using the Brunauer–Emmett–Teller (BET) method. According to the BET results shown in

Table 1. Comparison of BET results.

Catalyst	Specific Surface Area (m2/g)	Total Pore Volume (cm3/g)	Average Pore Diameter (nm)
KB(30,10)/LSF	14.04	0.094	26.9
KB(25,15)/LSF	16.95	0.100	24.9
Used KB(25,15)/LSF	16.62	0.098	23.8
KB(10,30)/LSF	25.82	0.124	19.1

, all the tested catalysts demonstrated a well-developed porous structure. However, the BET surface area (a_s) and the total pore volume at the relative pressure (p/p₀) of 0.989 increased when increasing the fraction of boehmite alumina in the catalyst, while the average pore diameter of the catalyst decreased, as shown in Table 1. The correlation between the specific surface area of the catalyst and its boehmite alumina content is attributed to the transformation of the boehmite alumina into one of the alumina phases during the calcination process [29], resulting in an expansion of the surface area.

Comparing the BET analyses of the used KB_(25,15)/LSF catalyst with the corresponding fresh catalyst in Table 1 reveals that the specific surface area and the total pore volume of the used catalyst experienced a slight reduction of approximately 2% relative to the fresh one, indicating that despite exposure to high temperature (900 °C) within the reactor over 10 redox cycles, the specific surface area and pore volume of the catalyst remained almost stable and intact during the reactor tests.

Considering the classification by the International Union of Pure and Applied Chemistry (IUPAC), the adsorption and desorption isotherms of the fresh catalysts KB_(10,30)/LSF and KB_(30,10)/LSF, shown in Figure 3, are identified as Type III isotherms (amongst the synthesized samples, KB_(10,30)/LSF and KB_(30,10)/LSF contain the highest contents of boehmite and kaolin, respectively), A type III isotherm, characterized by an increase in the amount of gas adsorbed without limit as its relative saturation approaches unity, indicates a microporous structure with weak adsorbent interactions. This type of isotherm is indicative of an indefinite multi-layer formation after the monolayer is complete. Pore size distribution plays a crucial role in such isotherms, as it affects the adsorption capacity and the rate at which gas molecules are adsorbed [30].

The BJH analysis shown in Figure 4 compares the pore size distributions of fresh and used KB_(25,15)/LSF catalyst samples. The fresh catalyst features a sharp peak at a 5 nm pore radius, highlighting a uniform micro/mesoporous structure that likely enhances catalytic activity due to efficient molecular transport. The curve then tapers off, indicating a minor presence of larger pores. Conversely, the used catalyst displays a broader peak with a wider range of pore sizes, suggesting structural changes during use. The steeper initial slope for the fresh sample implies a greater total pore volume compared to the used sample as seen in Table 1.

The highly similar XRD patterns of the fresh and used catalyst samples, along with the negligible reduction in the specific surface area and the total pore volume of the used catalyst inferred from the BET/BJH analysis, suggest that the structure and morphology of the particles have not been markedly affected by internal or external diffusional factors during the reactivity tests.

2.1.4. TPD-O₂ and TPD-NH₃ Tests

Figure 5 shows the oxygen adsorption behavior of KB_(25,15)/LSF as obtained from the TPD-O₂ test conducted with a temperature ramp of 10 °C per minute from 100 °C to 900 °C. As inferred from Figure 5, oxygen adsorption begins at approximately 175 °C, which marks the temperature at which the catalyst starts to adsorb oxygen. As the temperature increases, so does the rate of oxygen adsorption, as evidenced by the ascending slope of the curve. This trend continues until a peak at about 657 °C, representing the point of maximum oxygen adsorption. Beyond this temperature, the curve shows a decline, indicating the onset of oxygen desorption, where the catalyst releases the adsorbed oxygen.

The observed peak in oxygen adsorption at approximately 657 °C has critical implications for the catalytic process, specifically for the partial oxidation of methane. To achieve an optimal performance, the reaction should be conducted at temperatures above 657 °C where the catalyst’s oxygen adsorption is at its highest. This ensures an adequate supply of oxygen for the reaction. Moreover, at higher temperatures, the catalyst not only adsorbs, but also releases oxygen more efficiently, a factor that can significantly enhance the reaction by ensuring the continuous availability of oxygen for the conversion of methane.

The TPD-NH₃ test for the KB_(25,15)/LSF catalyst was conducted from 100 °C to 900 °C with a heating rate of 10 °C/min. As shown in Figure 6, the test results exhibit three peaks at temperatures of 273 °C, 529 °C, and 652 °C. These peaks correspond to the density of acidic sites. The first and second peaks represent weak and moderate acidic sites, whereas the third peak indicates strong acidic sites. The occurrence of the third peak at relatively high temperatures suggests that the catalyst is slow to form coke.

The presence of a higher total density of weak and semi-strong acidic sites compared to strong acidic sites suggests that the formed coke is likely to be soft and easily combustible. This type of coke is generally considered less problematic for the catalyst, as it can be readily eliminated during the oxidation process. The soft nature of the coke implies that it is less tightly bound to the catalyst surface, making it more susceptible to oxidation and removal. Consequently, the coke formation observed in this case is expected to have a minimal impact on the catalyst’s performance and can be effectively mitigated through regeneration cycles involving oxidation steps [31].

It should be remarked that the decomposition of desorbed ammonia at elevated temperatures is anticipated. If ammonia decomposition occurs concurrently with the desorption of ammonia from the oxygen carrier, it may result in a broader or overlapping desorption peak. This overlap could complicate the accurate determination of the desorption temperature of ammonia from the carrier. However, it is important to note that the broadening of the peak observed at temperatures above 600 °C does not alter the above conclusion regarding the higher total density of weak and semi-strong acidic sites compared to strong acidic sites. In the absence of ammonia decomposition, one would expect to see a sharper peak with a smaller area at temperatures above 600 °C, which would provide stronger evidence for our conclusion.

2.2. Catalyst Testing of KB_(25,15)/LSF

The plots in Figure 7 depict the cycle-averaged concentrations of H₂ (Figure 7a), CO (Figure 7b), and CO₂ (Figure 7c) in the gaseous products of the reduction steps of the fixed-bed tests carried out at 800, 850, and 900 °C, using KB_(25,15)/LSF as the oxygen carrier. To obtain the value of each data point shown in Figure 7, the average molar concentration of H₂ or CO at corresponding sampling times was calculated. Each error bar in the figures represents the standard error (SE) of the corresponding data acquired for all the cycles at a given sampling time.

As understood from the null or extremely low concentrations of H₂ and CO, but the high concentrations of CO₂ at the first 5–10 min of the reduction step of each test, the predominant reaction at the beginning of the reduction step was the full oxidation of methane (Equation (1)) since the oxygen provided by KB_(25,15)/LSF was present in sufficient (stoichiometric) quantities during this period. The duration of this step was shortened by increasing the operating temperature.

$${\mathrm{CH}}_4+2{\mathrm{O}}_2\to {\mathrm{CO}}_2+2{\mathrm{H}}_2\mathrm{O}$$

(1)

The complete combustion was followed by partial oxidation of methane characterized by the formation of CO and H₂ (Equation (2)). As the highly reactive surface oxygen of KB_(25,15)/LSF is depleted during the complete oxidation of methane to CO₂ and H₂O, the bulk lattice oxygen primarily contributes to partially oxidize methane to CO and H₂ [32]. In fact, the diffusion rate of bulk oxygen through the lattice dictates the selectivity of the process between partial and complete oxidation products [33].

$${\mathrm{CH}}_4+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2\to \mathrm{CO}+2{\mathrm{H}}_2$$

(2)

Regardless of the tested operating temperature, the molar concentration of H₂ constantly increased in the reduction step up to about 30% as the tests progressed. However, the CO concentration followed an increasing–decreasing pattern during the test. The drop in the CO concentration is attributed to coke formation, as observed in the experiments. This occurs due to a decrease in the oxygen content of KB_(25,15)/LSF, resulting in methane decomposition, producing carbon (coke) and H₂ (Equation (3)). The instant of the occurrence of the maximum CO concentration in the gaseous products corresponds to the moment of transition from the prevalence of the partial oxidation to methane decomposition. It is noteworthy that the slope of H₂ molar concentration vs. time changed at this moment. Similar to the complete oxidation phase, the duration of the POM phase was shortened by increasing the temperature, resulting in the earlier initiation of this phase at higher temperatures relative to lower temperatures.

$${\mathrm{CH}}_4\to \mathrm{C}+2{\mathrm{H}}_2$$

(3)

As shown in Figure 7b, the cycle-averaged CO mole concentration during the initial 15 min of the reduction steps in tests conducted at 900 °C exhibits substantial error bars. This is attributable to the varying (decreasing) trend exhibited by the CO mole concentration at the corresponding time intervals across consecutive cycles.

It is well-established that elevated temperatures typically enhance conversion rates; however, they can also promote coke formation, which adversely affects CO production. Thus, at 900 °C, the rigorous formation and accumulation of coke on the surface of the KB_(25,15)/LSF catalyst, which may not completely decompose during the oxidation step, can lead to increased coke buildup over successive redox cycles. As the reactivity tests progress, this coke layer on the oxygen carrier surface introduces additional mass transfer resistance, impeding the access of methane to the catalyst surface during the reduction steps. Consequently, the diffusion of methane through the coke layer is delayed, and the partial oxidation reaction is less efficient, particularly in the initial moments of the subsequent cycles. As a result, CO conversion gradually declines during the initial intervals of the reduction step. However, this issue diminishes as the reduction step progresses in each cycle, allowing methane adequate time to penetrate the coke layer and react with the oxygen present in the KB_(25,15)/LSF.

Figure 8a illustrates the cycle-averaged evolution of CH₄ conversion over time at 800, 850, and 900 °C. As seen, methane conversion was fairly high at all the tested temperatures; however, it increased with increasing temperature. The decreasing–increasing trend of CH₄ conversion denotes the competition between the complete oxidation, partial oxidation, and decomposition of methane, depending on the availability of oxygen provided by KB_(25,15)/LSF. Given the discussion above, the partial oxidation of methane is the prevailing reaction during the period when the lowest CH₄ conversion is observed. By increasing the reactor temperature from 800 to 900 °C, the minimum CH₄ conversion rises from approximately 39% to 64%, which represents an 64% increase.

Figure 8b shows the plots of the cycle-averaged H₂:CO molar ratio vs. the time for the tested temperatures. Considering the stoichiometric H₂:CO ratio in the POM reaction (Equation (2)), once this ratio is within two and three (marked with yellow dashed lines in Figure 8b), it can be concluded that the POM prevails over other reactions. This range of H₂:CO molar ratio is desirable for the syngas used as the feedstock of chemical synthesis processes. As seen in Figure 8b, increasing the temperature from 800 to 900 °C resulted in shortening the waiting period to reach this range from about 25 min to 10 min. The time window at which the desirable range of the H₂:CO molar ratio was achieved for each temperature correlates with the moment at which the lowest CH₄ conversion was seen in Figure 8a.

Figure 8c compares the evolution of CO selectivity in the reduction stages carried out at 800, 850, and 900 °C. For all tested temperatures, the selectivity towards CO exhibited a sharp increase during the initial stages of the process, followed by a plateau as time progressed. At a given time, the CO selectivity was greater for a higher temperature, as it increased from 79% at 800 °C to 87% at 900 °C.

Figure 9 presents the molar concentrations of H₂, CO, and CO₂, as well as the CH₄ conversion and CO selectivity for the time window (10–15 min) of the reduction phases of each redox cycle at 900 °C, where the POM reaction was dominant. The concentration of H₂ and CO₂ remained nearly constant over the ten cycles, though the CO concentration slightly dropped after the fourth cycle, leading to a small decrease in the CO selectivity. In general, KB_(25,15)/LSF showed favorable cyclic stability, as inferred from the consistency of CH₄ conversion and CO selectivity within ten cycles.

As outlined in the Supplementary Information (SI), the results of the activity tests are unlikely to be influenced by either internal or external diffusional contributions.

In fluidized bed reactors, perovskite catalysts often exhibit enhanced mass transfer and better contact between the catalyst and reactants, leading to higher methane conversion rates. The dynamic nature of fluidized beds allows for more uniform temperature distribution and reduced catalyst deactivation due to coking, which is a common issue in fixed-bed systems [34,35]. Fixed-bed reactors may show lower conversion rates due to potential hot spots and channeling effects, which can lead to uneven flow distribution and reduced effective catalyst surface area over time [36].

Fluidized bed systems can also provide better control over reaction conditions, which may enhance CO selectivity. The continuous movement of catalyst particles helps maintain optimal conditions for the desired reactions, potentially increasing the yield of CO from CH₄ reforming processes [35,37]. In contrast, fixed-bed reactors may lead to lower CO selectivity, as the reaction environment can become less optimal over time, especially if the catalyst experiences deactivation or if there is insufficient contact with the reactants [36].

The H₂:CO ratio is often influenced by the reaction conditions, including temperature and pressure, which can be more easily controlled in a fluidized bed reactor. This flexibility allows for tuning the ratio to favor hydrogen production, which is crucial for applications like fuel cells [37,38]. Fixed-bed reactors may struggle to maintain an optimal H₂:CO ratio due to the aforementioned issues of temperature gradients and catalyst deactivation, leading to less efficient reforming processes [34,36].

The above evidence indicates that the tested catalyst is expected to perform considerably better in a fluidized bed reactor than in a fixed-bed reactor. Consequently, the catalytic activity reported for the KB_(25,15)/LSF catalyst—including CH₄ conversion, CO selectivity, and the H₂:CO ratio—should be regarded as a lower bound of the corresponding values that could be achieved in a fluidized bed reactor under comparable operating conditions.

2.3. Analysis of the Attrition Tests

The results of the attrition tests for the fresh (as synthesized) perovskite as well as the used KB_(25,15)/LSF are presented in Figure 10. All tested materials followed a logarithmic attrition pattern, i.e., a relatively considerable loss of catalyst materials was observed during the first 30 min, followed by a modest level of attrition within the second 30 min. Such a logarithmic attrition behavior can be explained in light of the high rate of attrition due to the non-spherical shape and the sharp edges of the catalyst particles, as well as the attachment of small satellite particles to the main particle body. Once such irregularities or weakly bound particles have eroded, the remaining material often has high attrition resistance [39].

It is noteworthy that the used KB_(25,15)/LSF catalyst exhibited slightly higher attrition susceptibility compared to the fresh KB(25,15)/LSF. This finding suggests that the catalyst can withstand harsh conditions and maintain its integrity during an operation. Furthermore, the attrition evolution pattern of the used KB_(25,15)/LSF sample closely resembled that of the fresh LSF samples. This similarity implies that the conclusions drawn from the attrition tests of the fresh LSF samples studied in this work can be generally applicable to the used catalysts as well. The close correlation between the attrition behavior of fresh and used catalysts provides valuable insights into the long-term performance and durability of the KB_(25,15)/LSF catalyst under real-world operating conditions. This information can be used to optimize catalysts’ design, predict catalysts’ lifetime, and ensure the reliable performance of the catalytic system.

Figure 11 compares A_tot and A_i values calculated for the perovskite samples and shows that the A_tot of the tested samples greatly depends on their kaolin content, particularly for the samples with a relatively low kaolin content. However, such dependence is considerably weaker for A_i. This suggests that the formation of satellite particles and/or shape irregularities is highly likely for the adopted synthesis method when a relatively low amount of kaolin is used.

It is evident from Figure 11 that samples with a higher content of kaolin have shown higher resistance against attrition. Thus, KB_(25,15)/LSF and KB_(30,10)/LSF are suggested as the most appropriate LSF samples to be used in the redox cycles operated in fluidized bed reactors. It is important to note that exceeding the kaolin content of the catalysts beyond a certain level may lead to a reduction in the catalytic activity due to the decreased surface area, as discussed in Section 2.1.3. However, the negative effect of kaolin can be counterbalanced by adding boehmite to the catalyst. Boehmite can increase the surface area of the catalyst and thus improve its performance [29]. Nonetheless, it should be remarked that an excessively large BET surface area would not be beneficial for the partial oxidation of methane, as it favors complete combustion over partial oxidation [40].

It should be remarked that A_tot and A_i describe the attrition resistance of tested particles under very harsh conditions in the jet cup, resulting in excessively accelerated mechanical attrition. Such circumstances do not normally exist or prevail in an actual fluidized bed reactor. Thus, the reported values of A_tot and A_i underestimate the expected lifetime of oxygen carrier particles in a real-world chemical looping process. Despite the inherent limitations of the jet cup method, such as ignoring the effect of elevated temperature or chemical reactions on the attrition of particles, it is common practice in the industry to measure the attrition resistance of the fluidizable particles, including the oxygen carriers used at extreme operating conditions in cold rigs, such as a jet cup, since it has been observed that the materials performing well during the continuous operations at elevated temperatures typically exhibit high attrition resistance during jet cup tests at room temperature [39].

The attrition resistance of the KB_(25,15)/LSF sample used in the CLPO reactivity tests conducted in the fixed-bed reactor at 900 °C for 10 consecutive cycles was also examined in the jet cup apparatus. As seen in Figure 11, while the A_tot of the used catalyst is larger than the A_tot of the fresh catalyst, a reverse relationship was observed for the corresponding A_i values (A_i of used catalyst < A_i of fresh catalyst). This suggests that using KB_(25,15)/LSF particles in the redox cycles and their exposure to the feed flow in a fixed bed may result in the irregular shape or low sphericity of particles. Nonetheless, the long-term attrition resistivity of fresh KB_(25,15)/LSF has been preserved and even improved slightly after being used, indicating its satisfactory resistance against the stresses induced by thermal shocks or the changes in the crystal lattice caused by chemical reactions.

3. Materials and Methods

3.1. Materials

All the essential components required for the synthesis of LSF perovskite catalysts, and the preparation of nickel catalyst supported on alumina, were sourced from Merck. These materials include lanthanum nitrate hexahydrate (La(NO₃)₃.6H₂O), strontium nitrate (Sr(NO₃)₂), iron(III) nitrate nonahydrate (Fe(NO₃)₃.9H₂O), ethylenediaminetetraacetic acid (EDTA), citric acid, aqueous ammonia, glycine, kaolin, boehmite alumina, nitric acid (HNO₃), γ-alumina (Al₂O₃-γ), and nickel (II) nitrate (Ni(NO₃)₂).

3.2. Preparation of Oxygen Carrier Materials

3.2.1. Synthesis of LSF

The synthesis of the LSF perovskite catalyst involved an auto-combustion or auto-ignition method. In this method, various salts are initially dissolved in distilled water, resulting in a homogeneous solution. By allowing the solution to stay at a temperature of 80 °C, a gel is obtained, which is then placed in an oven to undergo ignition. The ignition of this gel produces a porous powder, which is then calcined in a furnace to obtain the final catalyst. The advantage of this low-cost and facile synthesis method is that the resulting material has high porosity, making it suitable for use as a catalyst [41].

To synthesize the LSF perovskite catalyst, 66.4 g of La(NO₃)₃.6H₂O was first dissolved in 200 mL of distilled water heated up to 80 °C by a homogenizer heater. Then, 88.18 g of Fe(NO₃)₃.9H₂O and 39.61 g of Sr(NO₃)₂ were added to the solution and allowed to dissolve completely, resulting in a homogeneous solution. Once a homogeneous solution of nitrate salts was obtained, glycine was added to the solution, serving as the fuel in the substance. By adding glycine, a dark coffee-colored complex was obtained after the binding between the metal ions and the amide part of glycine. This solution was left on the homogenizer–heater to form a gel. The gel was then placed in an oven in which ignition took place once the temperature reached 120 °C. During the ignition process, NO_X gases were released, followed by the formation of a porous powder. For the calcination of the synthesized catalyst, the obtained powder was initially held at a temperature of 800 °C for 2 h and then maintained at 900 °C for 6 h in a furnace. Afterward, the calcined powder was pressed into pellets to increase its strength. Finally, the pellets were ground and classified into particles ranging from 180 μm to 250 μm.

3.2.2. Synthesis of KB_(x,y)/LSF

For the modification with kaolin and boehmite alumina, (KB), boehmite alumina was dissolved in nitric acid, forming a viscous solution that served as a binder for LSF and kaolin particles. LSF and kaolin were added to this solution, creating a coffee-like paste, which was subsequently dried at 80 °C for 24 h. After drying, the sample was pelletized, ground, and subjected to calcination at 800 °C for two hours, followed by 900 °C for six hours, resulting in the formation of KB/LSF catalysts. The proportions of KB and LSF were controlled at (10, 30) wt. %, (30, 10) wt. %, (15, 25) wt. %, and (25, 15) wt. % of kaolin and boehmite alumina, respectively, for different designations of KB_(x,y)/LSF. In all the synthesized catalysts, the weight percent of LSF was maintained at 60%, as previous studies have shown this composition results in the highest catalytic activity [42]. However, increasing the LSF content beyond 60% would likely improve the catalyst’s mechanical resistance and durability.

3.3. Characterizations

The prepared catalyst samples were characterized using powder X-ray diffraction (PXRD), field emission scanning electron microscopy/energy dispersive X-ray spectroscopy (FE-SEM/EDS), and a Brunauer–Emmett–Teller (BET) surface analysis. To determine phase compositions and structural changes in the catalysts, PXRD patterns were collected on a PHILIPS PW1730 diffractometer with Cu Kα (λ = 1.5406 Å) radiation, with a step size of 0.026°. FE-SEM/EDS (model VEGA3, MIRA II and III, TESCAN Co., Brno, Czech Republic) was used to characterize the morphologies and the compositional characteristics of the catalyst samples. For surface area analysis, samples were degassed for 5 h at 150 °C, then nitrogen physisorption at −196.15 °C was carried out by BELSORP-MINI II and BEL PREP VAC II. The pore size distributions were derived from the nitrogen adsorption isotherms at 87 K using the BJH (Barrett–Joyner–Halenda) method.

The catalytic features of the catalyst samples were characterized by the temperature-programmed desorption of ammonia (TPD-NH₃) and temperature-programmed desorption of oxygen (TPD-O₂). A TPD-NH₃ analysis evaluated the strength and quantity of the acidic sites on catalyst surfaces, utilizing Thermo Scientific™ (Waltham, MA, USA) TPDRO 1100.

The TPD-NH₃ analysis provides insights into the impact of acidity on the catalytic performance, including activity, selectivity, and coking. TPD-O₂, conducted with a Micromeritics Auto Chem (Repentigny, QC, Canada) II 2920, measured oxygen storage capacity by tracking oxygen adsorption and desorption, delivering key data for optimizing catalysts in oxygen-intensive reactions like CLPO.

To identify and quantify the reaction products, the products of CLPO tests were analyzed by gas chromatography coupled with mass spectrometry (GC–MS: Agilent (Santa Clara, CA, USA) GC7890-MS5975) using a 50 m × 250 μm × 0.25 μm HP-5MS alumina PLOT (porous layer open tubular) column capillary column. The used carrier gas was helium with a flow rate of 5 mL/min and the split ratio was 100:1. The temperature program used was the initial temperature of 40 °C for 5 min, followed by a heating rate of 10 °C/min to 200 °C hold for 10 min.

3.4. Reactivity Tests

After screening the catalyst particles into the range of 180 μm to 250 μm, they were loaded into a tubular microreactor, with an inner diameter of 2.5 mm, schematized in Figure 12. To this end, carbon carbide particles were first put at the bottom of the reactor, covered by a layer of quartz particles classified in the range of 180 μm to 250 μm. Then, around 2 g of the catalyst was poured into the reactor to fill about 6.5 cm of the reactor height. Finally, a layer of carbon carbide was added on top of the catalyst. A thermocouple was inserted into the reactor, positioning its tip in close proximity to the sample for accurate temperature monitoring.

The feed flow rate for the oxidation stage consisted of 20% helium and 80% air. The total feed flow rate for the oxidation stage was 22.48 mL/min (GHSV = 7000 NmL·g_cat⁻¹·h⁻¹). Prior to commencing the methane conversion process, the oxygen carrier underwent a heating phase, gradually increasing at a rate of 10 °C/min in a continuous flow of 100 mL/min of helium (He) until reaching the target temperature of each test. Subsequently, a 50 mL/min methane mixture (GHSV = 15,000 NmL·g_cat⁻¹·h⁻¹, comprising 30 vol % CH₄ and 70 vol % helium) was introduced into the reactor. Helium was used in oxidation and reduction streams for mass balance calculations, both at the beginning and end of the process. The flow rate of each gas stream was controlled by a mass flow controller (MFC).

Each test began with the oxidation stage. After the oxidation stage, nitrogen gas was injected for 30 min to remove any remaining air from the reactor. During the oxidation stage, gas samples were collected at 5-min intervals and analyzed using a gas chromatograph. Once the nitrogen-to-oxygen ratio in the exhaust gas reached 87.3, indicating the saturation of the catalyst with oxygen, the reduction stage began. In this stage, the gas samples continued to be collected every 5 min from the reactor and analyzed using the GC-MS.

The warm tests were conducted under the atmospheric pressure at three temperatures: 800 °C, 850 °C, and 900 °C. The tests were performed for 4 cycles at 800 °C and 850 °C and for 10 cycles at 900 °C. Throughout the entire reaction, the pressure drop in the reactor was less than 0.5 bar. A tube furnace heated the reactor to the reaction temperature.

The detected products of the reaction were hydrogen, carbon monoxide, carbon dioxide, and water vapor. The water vapor content of the gas output of the reactor, leaving it at temperatures ranging from 800 to 900 °C, was condensed while passing through the outlet pipe at room temperature. If there was any residue left, it was also condensed at the laboratory’s room temperature. The gaseous byproducts were collected in the Tedlar™ (Wilmington, DE, USA) gas sample bags at five-minute intervals during the catalytic tests. Following each collection, the gas products in the sample bag were promptly analyzed using GC-MS to determine their molar concentrations.

The CH₄ conversion (CH_{4 conv}) and the selectivity of the reaction towards CO were calculated by Equations (4) and (5)

$${\mathrm{CH}}_4 \mathrm{conv} ={\displaystyle \frac{(F_{\mathrm{CH}4,\mathrm{in}}-F_{\mathrm{CH}4,\mathrm{out}})}{F_{\mathrm{CH}4,\mathrm{in}}}}$$

(4)

$$\mathrm{Selectivity} ={\displaystyle \frac{F_{\mathrm{CO}}}{F_{\mathrm{CO}2}+F_{\mathrm{CO}}}}$$

(5)

where F_CH4,in and F_CH4,out signify the molar flow rate of CH₄ entering and exiting the reactor, respectively. F_CO2 and F_CO denote the molar flow rate of produced CO₂ and CO in the outlet of the reactor.

The total outlet molar flow rate (F) was determined from the outlet volumetric flow rate (Q), measured by a bubble flow meter and the ideal gas law:

$$F={\displaystyle \frac{PQ}{RT}}$$

(6)

where P and T are the absolute pressure (Pa) and absolute temperature (K) of the outlet flow, respectively and R is the universal gas constant (8.314 J/(mol·K)).

Once the total molar flow rate is known, the molar flow rate of each constituent (F_i) can be calculated using the mole fractions obtained from the GC-MS analysis:

$$F_i=F{X}_i$$

(7)

3.5. Attrition Tests

The attrition tests were conducted using a jet cup device, which consists of an entry nozzle, with a diameter of 1.5 mm, located under the jet cup. Air, controlled by a MFC, entered the jet cup for one hour at a velocity of approximately 100 m/s, creating a vortex of particles swirling upwards through the cup. As a result of the combination of mechanisms similar to those occurring in the grid zone of fluidized beds (due to high-velocity jets) and in cyclones (due to the intensive swirling flow), the particles were subject to accelerated attrition in the jet cup [39].

The apparatus was run under ambient conditions (room temperature and nearly atmospheric pressure). To avoid the adherence of particles to the inner wall of the jet cup due to the electrostatic effects, the inlet air was humidified using a humidifier. The particles subjected to attrition were reduced to a fine size and entrained through the top outlet of the apparatus. These fine particles were then collected in a thimble filter with a 99.99% removal efficiency for 0.8 μm particles positioned at the top outlet of the device.

The mass ratio of particles collected in the filter to the initial mass of particles in the device (5.0 g) serves as an indicator of the susceptibility of each particulate sample to attrition. The mass of entrained particles collected in the thimble filter was weighed at 10-min intervals for one hour after opening the airflow.

Following the approach introduced by Rydén et al. [39], two numerical indices, (Equations (8) and (9)), were calculated to express the attrition rate of each material to illustrate the logarithmic attrition pattern of the explored particles. A_tot and A_i in Equations (8) and (9) denote attrition rates expressed as wt % fines caught in the filter during the entire test period (0–60 min) and the second half of the test period (30–60 min), respectively.

$$A_{\mathrm{tot}}=100\left({\displaystyle \frac{m_{\mathrm{f},60}}{m_{\mathrm{s}}}}\right)$$

(8)

$$A_{\mathrm{i}}=100\left({\displaystyle \frac{60}{30}}\right)\left({\displaystyle \frac{m_{\mathrm{f},60}-m_{\mathrm{f},30}}{m_{\mathrm{s}}}}\right)$$

(9)

In Equations (8) and (9), m_s, m_f,60, and m_f,30 are the initial weight of the particle sample in the apparatus and the accumulated weight of fines collected in the filter after 60 and 30 min, respectively. If A_tot = A_i, particles would be subject to perfectly linear attrition, i.e., the attrition rate remains unchanged over the entire test period. However, A_tot > A_i denotes the logarithmic attrition pattern of particles. For materials exhibiting a logarithmic attrition pattern, as observed in this study, A_i is preferred over A_tot, as A_i more accurately reflects the long-term stability of particles. This is because the initial 30 min of the attrition test typically involves the erosion of shape irregularities and weakly bound particles, making A_tot a less reliable indicator of long-term stability.

4. Conclusions

This research introduces a novel approach to enhance the performance and durability of La_0.7Sr_0.3FeO₃-based catalysts in the chemical looping partial oxidation of methane, by incorporating a blend of kaolin and boehmite alumina. The impact of kaolin and boehmite alumina binders on the synthesis, catalytic properties, and attrition resistance of the LSF perovskite catalyst was explored using a combination of characterization techniques, reactivity tests, and attrition tests, and the following key findings emerged:

• The characterization results confirmed the pure perovskite crystal structure of KB_(x,y)/LSF catalysts. This structure demonstrated an adequate oxygen adsorption capacity, effective coke mitigation capability, robust thermal stability, and resilience to agglomeration during repetitive redox cycles;
• Among the tested catalysts, KB_(25,15)/LSF emerged as the superior sample. Consistently, this catalyst produced syngas with a suitable H₂:CO molar ratio (ranging from 2 to 3) within ten redox cycles, at 900 °C. The CH₄ conversion and CO selectivity values reached up to 64% and 87%, respectively, showcasing its superb performance;
• The synthesized catalysts exhibited a logarithmic attrition pattern in the jet cup tests conducted at room temperature. These tests revealed high attrition resistance after the erosion of particle shape irregularities or weakly bound particles, indicating their durability;
• The KB_(25,15)/LSF catalyst, when used at 900 °C, demonstrated great resistance in the attrition test. This finding suggests its potential endurance in the harsh conditions of fluidized bed reactors employed for the CLPO process.

Fluidized bed reactors generally offer enhanced catalytic performances for perovskite catalysts compared to fixed-bed reactors. This improvement can be attributed to superior mass transfer, uniform temperature distribution, and better reaction control, leading to increased CH₄ conversion, CO selectivity, and optimized H₂:CO ratios. Therefore, considering the promising results obtained in this study, it is anticipated that the tested LSF catalysts will demonstrate exceptional catalytic activity and attrition resistance when utilized in a fluidized bed reactor.