Chemical Looping Reforming with Perovskite-Based Catalysts for Thermochemical Energy Storage

Abstract

The performance of a perovskite-based oxygen carrier for the partial oxidation of methane in thermochemical energy storage applications has been investigated. A synthetic perovskite with formula La_0.6Sr_0.4FeO₃ has been scrutinized for Chemical Looping Reforming (CLR) of CH₄ under fixed-bed and fluidized-bed conditions. Temperature-programmed reduction and oxidation steps were carried out under fixed-bed conditions, together with isothermal reduction/oxidation cycles, to evaluate long-term perovskite performance. Under fluidized-bed conditions, isothermal reduction/oxidation cycles were carried out as well. Results obtained under fixed-bed and fluidized-bed conditions were compared in terms of oxygen carrier reactivity and stability. The oxygen carrier showed good reactivity and stability in the range 800–1000 °C. An overall yield of 0.6 Nm³ of syngas per kg of perovskite can be reached per cycle. The decomposition of CH₄ catalyzed by the reduced oxide can also occur during the reduction step. However, deposited carbon is easily re-gasified through the Boudouard reaction, without affecting the reactivity of the material. Fluidized-bed tests showed higher conversion rates compared to fixed-bed conditions and allowed better control of CH₄ decomposition, with a H₂:CO ratio of around 2 and CO selectivity of around 0.8. However, particle attrition was observed and might be responsible for a loss of the inventory of up to 9%_w.

1. Introduction

To fulfil the Paris Agreement, governments all around the world will have to sustain the transition toward sustainable technologies and enforce private companies to lower their carbon emissions. In this frame, renewable energies will steadily increase their share of the energy sector and thermal energy storage (TES) technologies can be very useful to integrating high shares of renewable energy in power generation, industry, and buildings. Clean energy sources such as solar, wind, and biomass have the additional advantages of being unlimited and distributed in every continent, which are relevant aspects also for energy security. The market of solar technologies has considerably grown in the past decade [1]. Concentrated Solar Thermal (CST) technologies use solar radiation as a heat source and today are exploited to supply steam power plants. A CST plant usually requires a field of heliostats to direct solar rays on a receiver. Despite large capital costs and land use, CST displays some promising features:

• exploitation of the whole solar spectrum, rather than a selected spectral range;
• simple integration with Thermal Energy Storage (TES) to allow continuous operation [1];
• smaller requirements for critical raw materials compared to competing technologies [2].

TES is recognized as a key technology to overcome solar inherent intermittency, to increase the dispatchability of solar energy, and to expand the applications of CST technologies. Several TES technologies have been proposed to be integrated with concentrated solar plants for power generation. They can be distinguished as: sensible, latent, and thermochemical energy storage. Sensible TES uses solid or liquid media to store thermal energy without changing their phase [3]. Currently, molten salts are generally used as the heat transfer and energy storage medium in sensible TES systems, but they are limited by a maximum operating temperature of 565 °C. Latent TES is based on the heat absorption or release when a storage material undergoes a phase change from solid to liquid or liquid to gas or vice versa. The main advantages of these systems rely on higher storage capacities at similar temperature range between charging and discharging steps. Several materials have been investigated and proposed to enhance heat transfer, such as inorganic salt-based phase-change materials, metallic alloys, and composites. Thermochemical Energy Storage (TCES) consists of exploiting available heat to drive an endothermic reaction. Solar energy is thus stored in the chemical products, rather than as sensible energy. The products can be stored for an unlimited time and can even be transported over large distances before being employed.

Several TCES processes are being developed. Most of them consist of chemical looping cycles, by which a solid material undergoes two or more reaction steps cyclically [4]. TCES may allow either the chemical storage/release of thermal energy for the generation of electricity or the production of sustainable fuels, chemicals, and materials. In the former case, the decomposition/recombination of metal hydrides, carbonates, hydroxides, and oxides is frequently considered [5]. In the latter case, the production of hydrogen (H₂), carbon monoxide (CO), or a mixture of them (synthesis gas) is pursued [6]. These gases can then be used to power gas turbines or fuel cells, or rather to produce advanced fuels and chemicals. Thermochemical splitting of water (H₂O) and carbon dioxide (CO₂) is an example: the endothermic step is the reduction of a metal oxide with the release of gaseous oxygen (O₂):

MO_x → MO_x−δ + δ/2 O₂,

(1)

The reduced oxide then reacts exothermically with H₂O and\or CO₂, producing H₂ and\or CO:

MO_x−δ + δ H₂O → MO_x + δ H₂,

(2)

MO_x−δ + δ CO₂ → MO_x + δ CO,

(3)

The reduced oxide allows splitting of H₂O or CO₂ at temperatures lower than through the direct homogeneous reaction. In addition, it works as an oxygen carrier: O₂ is released separately during the reduction step without mixing with the synthesis gas, and no explosive mixtures are formed. The reduction step typically requires temperatures of 1200–1500 °C and extremely low O₂ partial pressures, typically obtained by vacuum pumping or inert-gas sweeping. The introduction of a reducing agent allows more moderate thermodynamic conditions. If methane (CH₄) is used, the reaction temperature can be reduced [7]:

MO_x + δ CH₄ → MO_x−δ + 2δ H₂ + δ CO,

(4)

The resulting overall reaction in this case is steam/dry reforming and the process is known as Chemical Looping Reforming (CLR). In principle, 4 moles of syngas can be produced from 1 mole of CH₄ per cycle, with a H₂:CO ratio that varies between 1 and 3 depending on whether CO₂ or H₂O is split during the oxidation step. The total energy content of the produced syngas is higher than the CH₄ feed by a factor U (upgrade factor) between 1.25 and 1.31, calculated as:

U = (n_H2 LHV_H2 + n_CO LHV_CO)/(n_CH4 LHV_CH4),

(5)

The selection of suitable metal oxides for thermochemical splitting is a wide research field [8,9]. The desired properties are:

▪

low reduction temperatures;

▪

high re-oxidation rates;

▪

high oxygen capacities (δ);

▪

good stability over cycles.

Volatile oxides release metal vapors when reduced. Zinc oxide (ZnO) falls into this category and has long been studied [10]. The efficient separation of the metal vapor from the product gas, to avoid recombination with O₂, is a fundamental aspect to consider [11]. The use of CH₄ [12] or carbon [13] as a reducing agent has been proposed for the combined production of Zn and synthesis gas. This shows how TCES can be integrated with other industrial sectors, opening to different markets. On the other side, non-volatile oxides have the advantage of remaining in a condensed state when reduced. Magnetite (Fe₃O₄) was the first oxide proposed for solar-driven thermochemical splitting and belongs to this category [14]. Mixed oxides have also been investigated as ferrites [15,16,17], aluminates [18], and perovskites [19], proving a higher stability during the high temperature reduction step. Despite limiting the oxygen capacity (δ), partial reduction to non-stoichiometric oxides shows beneficial aspects. Non-stoichiometric oxides preserve their crystal phase by letting oxygen diffuse through their lattice through the formation of vacancies. Among the advantages are moderate reduction temperatures, fast kinetics of re-oxidation, and good stability. Ceria (CeO₂) is a reference material of this kind [20]. The use of dopants enhances the stability of lattice defects and yields higher reduction degrees [21,22]. Perovskites are also non-stoichiometric oxygen carriers [19]. They are characterized by the presence of two cations and represented with the general formula ABO₃. Partial substitution of the A and B cations can be used to tailor the redox properties [23]. A Mars–van Krevelen-type reaction mechanism has been suggested for CLR [7,24,25]: adsorbed CH₄ reacts with lattice oxygen on the catalyst surface. The diffusion rate of bulk oxygen through the lattice dictates the selectivity of the process between partial and total oxidation products. Lanthanum (La)–iron (Fe) and La–manganese (Mn) perovskites have been found well-suited to CLR thanks to their high selectivity toward syngas (reaction (4)) [26,27]. La-substituted perovskites have been also investigated to enhance the mobility of oxygen and decrease C deposition. La_1−xSr_xFeO₃ and La_1−xSr_xMnO₃ perovskites were investigated in fixed-bed and membrane reactors [28,29], while a specific perovskite, La_0.85Sr_0.15FeO₃, was investigated in a fluidized-bed reactor [30]. Results were promising in terms of perovskite reactivity, partial oxidation of CH₄, and H₂O/CO₂ splitting, especially under fluidization conditions. However, only few studies are available in the literature on the use of La_1−xSr_xFeO₃ perovskites under different reactor configurations, stimulating further research on the topic.

As TCES usually involves heterogeneous gas–solid reactions, the design of multiphase reactor technologies is fundamental for the development of these processes, and several examples are found in the literature [31]. Packed bed or porous solid structured reactors have been studied in the literature, with interesting results [32,33,34,35]. They benefit from low costs and easy operability, as they do not require the handling of the solids. However, poor mass transfer and temperature gradients due to nonuniform heating may limit the scale-up of these systems. Conventional and unconventional fluidized beds have been proposed for solar applications and have been recently reviewed [36]. Spouted [37] and internally circulating fluidized beds [38] have been studied under direct concentrated radiation to achieve uniform heating of the particles. Many authors have proposed concentration of solar radiation on the tube walls of fluidized beds [39], exploiting their large heat transfer coefficients. A pair of twin indirectly irradiated fluidized beds have been recently tested for thermochemical splitting with hercynite particles [39]. Dense circulating systems have been developed, both with up- [40] and down-flowing particle suspensions [41], to have a continuous operation. The recovery of the sensible energy of the outlet solid stream to preheat the inlet solid in a countercurrent heat exchanger has been studied both experimentally [42,43] and through modeling [44]. Finally, a multistage solar reactor, composed of multiple adjacent fluidized beds connected in series, has been recently proposed to narrow the residence time distribution of the solid phase [45]. Fluidized beds have higher operating costs due to the pumping of the fluidizing gas and solid attrition and erosion, resulting in losses of inventory and wear of the equipment. However, a breakeven point can probably be reached, considering the undiscussed advantages of optimal temperature control, enhanced mass transfer, and better scalability.

This study investigates the redox properties of a laboratory-prepared perovskite for CLR of CH₄ in terms of reaction temperatures, oxygen capacity, and selectivity. A La-Sr-Fe-O system, specifically La_0.6Sr_0.4FeO₃, is tested as an oxygen carrier for the first time both in fixed-bed and fluidized-bed conditions. For this purpose, the time-resolved concentration profiles of the outlet gas were analyzed during reduction and oxidation steps. Oxidation was carried out with either H₂O or CO₂. The effect of temperature on reaction times has been qualitatively discussed with reference to a Mars–van Krevelen reaction mechanism. The material stability has been assessed by determining conversion degrees over iterated cycles. A critical comparison between fixed- and fluidized-bed performance in terms of reaction times and control of selectivity has been performed, to scrutinize the effect of gas–solid contact patterns.

2. Materials and Methods

2.1. Oxygen Carrier

The perovskite material was synthetized following the procedure developed in a previous study [23], to obtain a perovskite powder with chemical formula La_0.6Sr_0.4FeO₃. Briefly, stoichiometric amounts of strontium nitrate, lanthanum nitrate hydrate, and iron (III) nitrate nonahydrate (Sigma-Aldrich, St. Louis, MO, USA) were dissolved in 75 mL bi-distilled water and stirred for 3 h. The obtained solution was heated for about 10 min in a microwave oven (CEM SAM-155) up to the formation of a homogeneous gel, which was further calcined at 1100 °C for 4 h (heating rate: 10 °C/min). The powder was pressed into small cylindrical pads (2–3 mm diameter and height) with a bench-top press at 10 ton. Eventually, the pads were crushed into a mortar, and the fragments sieved in the size range 200–400 μm. Larger particles were crushed again, while finer particles were pressed again into pads, until granulation of the whole powder sample within the desired size range.

2.2. Experiments under Fixed-Bed Conditions

Figure 1 shows the set-up for the experiments under fixed-bed conditions. A perovskite sample (0.5 g) was placed on a porous plate inside a tubular quartz reactor with an internal diameter of 10 mm and heated by a cylindrical electric furnace (Lenton). A thermocouple, sheathed with a 6 mm tube, was vertically plunged into the sample along the center line of the reactor. A PID-type controller was used to set a heating rate or a fixed temperature. The reactants were diluted in a nitrogen (N₂) flow and fed at the top of the reactor. The gaseous species were fed from pure-gas cylinders through mass flow controllers (Brooks 5850S). Water was fed with a pump (Dionex P680 HPLC Pump) and vaporized by heating at 120 °C. A pressure transducer monitored the pressure upstream of the reactor inlet for safety reasons. The outlet gaseous stream was analyzed with two gas analyzers in series (Gas 3100 Syngas Analyzer and NGA 2000 from Rosemount) to detect the main chemical species (CH₄, CO, CO₂, H₂, O₂), after drying in a CaCl₂ trap. The reactor can be by-passed using a four-way valve to allow analysis of the feed.

The following experiments were run:

• Temperature-Programmed Reduction (TPR) from 25 to 1000 °C at 10 °C/min with CH₄
• Temperature-Programmed Oxidation (TPO) from 550 to 1000 °C at 10 °C/min with CO₂
• Isothermal cycles with CO₂ as oxidizer at 860, 920, and 1000 °C
• Isothermal cycles with H₂O as oxidizer at 920 °C

For every test, the gas was fed from the top with a total flowrate of 10 NL/h, corresponding to a gas velocity between 0.14 and 0.17 m/s for the operating temperature range (860–1000 °C). The gas composition was 5% CH₄ for the reduction steps and 5% CO₂ for the oxidation steps. The concentration of H₂O could not be measured, but the pump was set to supply the oxidizer at the same concentration. A pure N₂ flow was used to sweep the reactor between consecutive reaction steps. After the TPO with CO₂, a TPO was performed with O₂ to ensure the full re-oxidation of the catalyst and to eventually burn unconverted coke deposited on the perovskite surface.

2.3. Experiments under Fluidized-Bed Conditions

The set-up for the experiments under fluidized-bed conditions was very similar (Figure 2). The perovskite sample (3.2 g), sieved in the range 250–355 μm, was placed in a tubular quartz reactor with an internal diameter of 10 mm inside a cylindrical electric furnace (Watlow Italy, Corsico (MI), Italy). The temperature of the catalyst was controlled with a PID controller. The fluidizing and reacting gases were fed from gas cylinders through mass flow controllers (Bronkhorst High-Tech B.V., Ruurlo, The Netherlands). The outlet gas passed through a bubble flowmeter, a water condenser, and finally through a gas analyzer (ABB, Advance Optima AO2020).

Similarly, for fixed-bed experiments, different tests were performed within the fluidized-bed reactor by varying the temperature and gas composition, to perform either the reduction or the oxidation step of the process. Isothermal cycles were conducted at 900 °C using CO₂ as oxidizer. A single long reduction step at 970 °C was conducted to study the decomposition of CH₄, followed by an oxidation step at 855 °C. The gas was fed from the bottom with a total flowrate of 70 NL/h, in order to have about the same contact time of the fixed-bed experiments. The corresponding fluidization velocity was about 1.1 m/s at 900 °C. The gas composition was 5% CH₄ for all the reduction steps, 15% CO₂ for the oxidation steps at 900 °C, and 8% CO₂ for the oxidation step at 855 °C.

3. Results and Discussion

3.1. Fixed-Bed Experiments

Figure 3 shows the temperature and concentration profiles from the TPR experiments. At about 800 °C, the reaction starts to take place with a high selectivity toward the products of partial oxidation and with a limited conversion of CH₄, as inferred from the H₂:CO concentration ratio of about 2. These observations suggest that the reaction is kinetically controlled by oxygen diffusion through the lattice of the perovskite, and as a consequence the concentration of oxygen on the surface of the catalyst is very low. A very little amount of CO₂ appears to be produced according to the reaction:

MO_x + δ/4 CH₄→ MO_x−δ + δ/2 H₂O + δ/4 CO₂,

(6)

When the temperature exceeds 900 °C, the concentration of CO decreases but both the conversion of CH₄ and the production of H₂ continue. It is deduced that, as the catalyst is progressively reduced, CH₄ decomposition takes over, according to the reaction:

CH₄ → C + 2 H₂,

(7)

as confirmed by the ratio between the concentration of H₂ and that of consumed CH₄. The amount produced or reacted from each gaseous species can be calculated as follows:

n_i = N_tot ∫ |y_iⁱⁿ − y_i^out| dt,

(8)

where N_tot is the total molar flowrate, and y_iⁱⁿ and y_i^out are the inlet and outlet concentration of species i. From the atomic balances on hydrogen, carbon, and oxygen, it is possible to calculate the amount produced of H₂O and C and the amount of oxygen released by the perovskite:

4 n_CH4 = 2 n_H2 +2 n_H2O,

(9)

n_CH4 = n_C + n_CO + n_CO2,

(10)

n_O = n_CO + 2 n_CO2 + n_H2O,

(11)

By dividing n_O by the moles of the sample (n_perovskite), it is observed that 1.5 moles of oxygen were released per mole of perovskite (δ = 1.5), corresponding to the total reduction of iron from Fe⁺³ to Fe⁰ and thus to the complete conversion of the oxygen carrier. The formation of Fe⁰ may be responsible for the catalytic decomposition of CH₄ during the reduction step.

Figure 4 shows the temperature and concentration profiles during the TPO experiments. The reaction starts a little above the initial temperature of 550 °C, and above 800 °C the conversion of CO₂ is almost complete and then falls down as the catalyst is totally regenerated. The molar ratio between produced CO and converted CO₂ reaches values greater than unity, showing that gasification of deposited carbon according to the Boudouard reaction occurs simultaneously to perovskite oxidation, according to the reaction:

CO₂ + C → 2 CO,

(12)

The amounts involved in the reaction of the gaseous species can be calculated again from Equation (7). The atomic balances on carbon and oxygen give the amount of C re-gasified according to reaction (12) and the amount of oxygen uptaken by the perovskite:

n_C + n_CO2 = n_CO,

(13)

2 n_CO2 = n_O + n_CO,

(14)

The calculations prove that the deposited carbon (C) is totally re-gasified according to reaction (12) and that the perovskite is completely re-oxidized. According to these results, a yield of 0.6 Nm³ of syngas per kg of catalyst per cycle can be obtained from reactions (2–4), corresponding to the maximum value of δ = 1.5. The value of δ and of syngas yield are in good agreement with the literature data on a similar perovskite [30].

Isothermal reaction profiles at different temperatures during the reduction steps are shown in Figure 5. During the course of the reaction, the selectivity toward CO decreases and the H₂:CO ratio increases, indicating that CH₄ thermocatalytic decomposition also occurs. The reduction steps were interrupted in order to limit the decomposition of CH₄ and the deposition of C on the catalyst. To this end, the concentration of CH₄, H₂, and CO were monitored and the feed of CH₄ was interrupted as soon as the selectivity toward CO appeared to drop below 80%. Except for an initial peak of CO₂, the favored products are H₂ and CO in every case and a quasi-steady condition is reached at 860 °C. These observations fit very well the interpretation of a fast reaction regime kinetically controlled by the diffusion of oxygen through the lattice sites: at the beginning of the reaction, CH₄ reacts with the oxygen-rich surface of the catalyst through complete oxidation; due to the low diffusion rate, the oxygen concentration at the surface rapidly drops down, approaching the minimum value allowed by the thermodynamic equilibrium. This is evident at 860 °C, as the gas composition probably mirrors an equilibrium condition. At higher temperatures, the reaction speeds up thanks to higher diffusivities and more favorable thermodynamic conditions, as witnessed by the concentration profiles. However, the reaction also slows down more rapidly, due to the faster reduction of the catalyst.

In Figure 6, the isothermal oxidation steps are shown. The increase in temperature has opposite effects on the reaction rate. The initial peak of CO concentration indicates an increase in the kinetic rates of surface reactions. However, the reaction times do not increase monotonically. This can again be explained considering oxygen diffusion as the limiting step. At the beginning, the surface of the reaction is rapidly oxidized, producing the peak of CO, but it quickly saturates, reaching the highest oxygen concentration allowed by thermodynamics. Then, the reaction proceeds slowly, controlled by oxygen diffusion from the surface toward the bulk of the catalyst. As temperature is increased, the diffusivity also increases, and this can explain the reduction in the reaction time from 860 °C to 920 °C. However, at higher temperatures, unfavorable thermodynamic equilibrium limits the surface concentration of oxygen and thus slows the diffusion process. This might explain why the reaction time increases when raising the temperature from 920 °C to 1000 °C.

Equations (8)–(11), (13) and (14) were solved for each of the isothermal cycles, and the conversion degrees of the sample were calculated as:

X = n_O/(1.5 n_perovskite),

(15)

Figure 7 shows the results. A reduction degree around 50% and complete re-oxidation are observed for every process temperature. The cycles show good repeatability, and no deactivation trend is observed.

The results of the isothermal cycles at 920 °C using H₂O as oxidizer are shown in Figure 8. Six consecutive cycles were performed, using a shorter reaction time for the reduction step (6 min) to further limit CH₄ cracking. A reduction degree of about 40% was obtained. As for the CO₂ oxidation tests, the perovskite material showed complete re-oxidation and good cyclability.

3.2. Fluidized-Bed Experiments

Four isothermal cycles were conducted at 900 °C. During the first cycle the CH₄ feed was fed for 19 min to assess the time needed for the total reduction of the perovskite. For the following two cycles, the CH₄ feed was interrupted after 7 min to avoid CH₄ cracking and C deposition. Finally, the fourth cycle was carried on for 16 min to reach again full conversion of the perovskite. Figure 9 shows the reaction profiles of the second and fourth cycles. The reduction profiles (Figure 9a) show good repeatability. Total conversion of the perovskite is achieved in about 12 min and CH₄ decomposition initiates immediately after. The oxidation profiles (Figure 9b) differ for the Boudouard reaction, occurring in the fourth cycle.

Figure 10 shows the perovskite conversion degrees. The reduction degree for 7 min at 900 °C was about 72%, whereas in fixed-bed experiments at 920 °C the reduction degree for about the same time was 50%, suggesting that, in fluidized-bed conditions, higher conversion rates can be obtained.

Finally, results of the long reduction test at 970 °C are shown in Figure 11. This test allowed investigation of the occurrence of deactivation phenomenon due to C deposition on the catalyst. Complete reduction of the perovskite is achieved in about 9 min. It can be pointed out that CH₄ is almost fully converted in this phase, showing again that the conversion rate is higher with respect to the case of the fixed bed (see Figure 5c). Then, CH₄ decomposition started and continued for more than 35 min. The decomposition rate showed a gradual but constant decrease, indicating an inhibiting effect of the deposited carbon. About 3 mmol of C was produced per mmol of catalyst, indicating that the reaction proceeds beyond the total coverage of catalytic Fe⁰ sites.

Figure 12 shows the profiles of the following regeneration step, performed at 855 °C with 8% CO₂. Again, complete regeneration of the catalyst could be achieved.

After the fluidized-bed tests, the perovskite sample was recovered, weighted, and sieved in order to assess eventual mass loss and particle size reduction. A 0.3 g amount of perovskite was found missing, and 0.1 g of the recovered perovskite had reduced its size below 250 μm. This indicates that the particles reduced their size by attrition and up to 9%_w of the bed inventory might have been lost by elutriation.

3.3. Comparison between Fixed- and Fluidized-Bed Experiments

Finally, Figure 13 shows a comparison between the time profiles of the H₂:CO ratio, and the selectivity toward CO, for the reduction step at 1000 °C in fixed-bed conditions and the reduction step at 970 °C in fluidized-bed conditions. Under fluidized-bed conditions, CH₄ thermocatalytic decomposition initiates only at the end of the perovskite reduction, as inferred from the obtainment of a nearly constant H₂:CO ratio of 2 and selectivity of 0.8 throughout most of the experimental run. Differently, under fixed-bed conditions, the H₂:CO ratio soon exceeds the value of 2, with a concomitant decrease in the CO selectivity, indicating that CH₄ thermocatalytic decomposition starts much sooner and gradually takes over. This might be explained by the different contacting pattern between the gas reactant and the perovskite bed of particles. The fixed-bed conditions combined with the plug flow pattern of the gas phase determine that the layers of the catalyst at the inlet are reduced faster than those at the outlet. This leads to an early start of the decomposition reaction with respect to the complete reduction of catalyst particles. In fluidized-bed tests, instead, the solids perfect mixing conditions allow a uniform reduction degree of the particles, delaying CH₄ decomposition at the end of the bed solids reduction step.

4. Conclusions

A laboratory-produced perovskite catalyst has been tested in experiments of CLR of CH₄ under fixed- and fluidized-bed conditions. Temperature-programmed experiments indicate that the reduction can be performed at temperatures above 800 °C and the oxidation above 600 °C. Non-stoichiometric reduction yields a maximum value of δ = 1.5, so 0.6 Nm³ of syngas per kg of catalyst per cycle can be obtained from reactions (2–4). Except for a brief initial stage of the reduction, when the catalyst surface is rich of oxygen, the reaction evolves with a selectivity toward synthesis gas around 90%. Thermocatalytic decomposition of CH₄ catalyzed by the reduced oxide can occur during the reduction step. This drawback can be limited by interrupting the reduction step. C deposition does not deactivate the oxygen carrier, which is fully regenerated during the oxidation step thanks to gasification reactions. The overall H₂:CO ratio can be adjusted between 1 and 3, depending on whether CO₂ or H₂O is split during the oxidation step. CH₄ decomposition provides an additional degree of freedom to tune the H₂:CO ratio during the reduction step. No sign of a loss of reactivity was evidenced by determining the perovskite conversion degrees for several cycles at different temperatures. The reduction times decrease with temperature, whereas the oxidation times have a non-monotonical trend. Fluidized-bed tests showed higher conversion rates compared to fixed-bed conditions and allowed better control of CH₄ thermocatalytic decomposition. However, particle attrition was observed and might be responsible for a loss of the inventory of up to 9%_w. An improvement of the mechanical properties of the catalyst is an important issue to be addressed for scaled-up applications in fluidized beds. On the other hand, these results are very promising and provide a basis for the selection of operative conditions and reactor design, by considering the effect of temperature, reaction times, and gas–solid contact patterns.