Tuning Chemical Looping Steam Reforming of Methane Performance via Ni-Fe-Al Interaction in Spinel Ferrites

Abstract

The chemical looping steam reforming of methane (CLSR) employing Fe-containing oxygen carriers can produce syngas and hydrogen simultaneously. However, Fe-based oxygen carriers exhibit low CH₄ activation ability and cyclic stability. In this work, oxygen carriers with fixed Fe content and different Fe/Ni ratios were synthesized by the sol–gel method to investigate the effects of Ni-Fe-Al interactions on CLSR performance. Ni-Fe-Al interactions promote the growth of the spinel structure and regulate both the catalytic sites and the available lattice oxygen, resulting in the CH₄ conversion and CO selectivity being maintained at 96–98% and above 98% for the most promising oxygen carrier, with an Fe₂O₃ content of 20 wt% and Fe/Ni molar ratio of 10. The surface, phase, and particle size were kept the same over 90 cycles, leading to high stability. During the CLSR cycles, conversion from Fe³⁺ to Fe²⁺/Fe⁰ occurs, along with transformation between Ni²⁺ in NiAl₂O₄ and Ni⁰. Overall, the results demonstrate the feasibility of using spinel containing earth-abundant elements in CLSR and the importance of cooperation between oxygen release and CH₄ activation.

1. Introduction

Global warming caused by the excessive combustion of fossil fuels has attracted increasing attention [1,2]. Hydrogen, as an alternative energy carrier, has a high calorific value (142 MJ/kg) and only produces H₂O during utilization. It is also an important industrial intermediate in crude oil refining and ammonia production. Therefore, the demand for H₂ is increasing [3,4]. However, most H₂ in industry is currently produced from the steam reforming of methane, leading to CO₂ emission problems [5,6]. Therefore, it is important to develop a novel process for producing H₂ with CO₂ capture from fuels.

The chemical looping steam reforming of methane (CLSR) is a process that decouples the steam reforming of methane into reactions occurring in a fuel reactor and a steam reactor with metal oxides as the oxygen carriers (OCs). As shown in Figure 1, methane is partially oxidized by OC in the fuel reactor to produce syngas, and H₂O is split in the steam reactor to produce H₂ and regenerate OC. Through the repeated reduction and oxidation of OC, methane is converted into syngas with an ideal molar ratio of 2 for downstream industries while producing pure hydrogen [7,8,9,10]. OC provides lattice oxygen and forms active sites for reactions during CLSR [11,12,13,14]. Although iron oxides are promising because of their favorable thermodynamics for water splitting, low price, wide availability, etc., they exhibit low reactivity towards CH₄ and suffer deactivation through sintering at high temperature [15,16,17,18,19]. Therefore, it is important to develop novel Fe-based OCs for CLSR.

To promote the reactivity and stability of a novel Fe-based OC, various supports [20,21,22] (Al₂O₃, CeO₂, LaFeO₃, etc.) and foreign promoters [23,24] (Ni, Co, Cu, etc.) are chosen to modify the bulk and surface. Among supports, Al₂O₃ is commonly used because of its low cost, high mechanical strength, and high melting point, which greatly improve stability at high temperature [25]. However, the formation of FeAl₂O₄ spinel hinders hydrogen production when applying the Fe₂O₃/Al₂O₃ OC for CLSR [26]. When foreign promoters are introduced, the catalytic sites of CH₄ and lattice oxygen transfer activity are influenced simultaneously [27]. According to the existing results, Ni clearly reduces the energy barrier for C-H bond cleavage [28], but the Fe-Ni bimetallic OC faces problems: severe carbon deposition due to the strong catalytic effect of the Fe-Ni alloy and deactivation because of phase separation [29,30]. Therefore, the NiFeAlO OC could be a feasible solution to the conflict between reactivity and stability. Cui et al. reports that a NiFeAlO OC showed a high oxygen transport capacity and redox stability in chemical looping combustion [31]. Kim et al. investigated an Al-incorporated NiFe₂O₄ particle for chemical looping hydrogen production and found that Al incorporation prevented the densification of the OC [32]. Based on characterization results, the NiFeAlO OC exhibits promising performance in chemical looping processes because of the formation of spinel structure and the exsolution of Ni during reduction. Recently, the coordination environment of the atom has been employed to regulate OC reactivity [33,34,35,36]. This strategy has also been proved feasible by our previous work through varying the Mg/Fe ratio of a MgFeAlO OC, but the CH₄ conversion was not satisfactory due to limited surface active sites. Similarly, Ni, Fe, and Al could exist in the tetrahedral and octahedral interstices of the spinel structure [37,38], potentially leading to different types of lattice oxygen transfer and surface active site formation in CLSR. Hence, the Ni-Fi-Al interactions and the reaction mechanism during CLSR are worth studying.

This work synthesized a series of NiFeAlO OCs with fixed Fe₂O₃ content and varying Fe/Ni ratios and explored the effects of Ni-Fe-Al interactions on CLSR performance in detail. Firstly, the phase was revealed using an X-ray diffractometer (XRD), and the bulk lattice oxygen activity was studied through H₂ temperature-programmed reduction (H₂-TPR) and density functional theory (DFT) calculation. Also, the surface reactivity toward CH₄ was investigated utilizing CH₄ temperature-programmed reduction (CH₄-TPR). Then, cyclic CLSR experiments were conducted in a lab-scale fluidized bed reactor and the changes in OCs were characterized. Finally, X-ray photoelectron spectroscopy (XPS) was applied to study the chemical states of the OCs, reflecting the reaction mechanism.

2. Materials and Methods

2.1. Oxygen Carrier Preparation

NiFeAlO OCs were prepared by the sol–gel method with Fe/Ni molar ratios of 20, 10, 7.5, and 5, and named FeNi20Al, FeNi10Al, FeNi7.5Al, and FeNi5Al. Fe(NO₃)₃·9H₂O (Shanghai Aladdin Biochemical Technology Co., Shanghai, China), Al(NO₃)₃·9H₂O (Shanghai Aladdin Biochemical Technology Co., Shanghai, China), and Ni(CH₃COO)₂·4H₂O (Shanghai Aladdin Biochemical Technology Co., Shanghai, China) acted as sources of Fe, Al, and Ni, with citric acid (Shanghai Aladdin Biochemical Technology Co., Shanghai, China) and ethanediol (Shanghai Aladdin Biochemical Technology Co., Shanghai, China) as complexing agents. The amount of starting materials was calculated according to three principles, namely, the predicted mass, the controlled content of Fe with Fe₂O₃ at 20 wt% [39], and the designed molar ratio of Ni and Al. Fe(NO₃)₃·9H₂O, Al(NO₃)₃·9H₂O, and Ni(CH₃COO)₂·4H₂O were dissolved in deionized water, where the molar ratio between the metal ions and H₂O was 1:10. Then, citric acid was introduced to the mixture, and the molar ratio between the citric acid and total metal ions was 1:1. The solution was then heated to 80 °C with continuous stirring, and ethanediol was added at 1.5 times the amount of citric acid in moles until the sol was formed. Then, the resulting sol was aged overnight and dried at 120 °C for 12 h. The dried gel was calcined at 350 °C for 2 h to remove carbonaceous residues. The obtained powder was tableted and calcined at 1000 °C for 5 h. Finally, the oxygen carriers were sieved to a size range of 0.15–0.30 mm.

2.2. Experimental Setup

A fluidized bed reactor with a quartz tube (inner diameter of 8 mm) was adapted to evaluate the CLSR performance. A total of 1 g of oxygen carriers was loaded in each test. The reaction temperature was monitored using a thermocouple near the middle part of the oxygen carrier bed, and the flow rates of the reaction gases were controlled using mass flow meters. One CLSR cycle comprised two stages, i.e., the reduction and oxidation stages. The flow rate of the inert carrier gas N₂ was 90 mL/min. After the OC was heated to 900 °C under an air atmosphere, 10 mL/min of CH₄ entered the quartz tube and reduced the OCs during the reduction stage. Thereafter, the purge stage was used to avoid mixing of the products in the reduction and oxidation stages. In the subsequent oxidation stage, steam generated from a 25 μL/min flow of water was applied for hydrogen production. The exhaust gas was dried and finally analyzed using a gas analyzer (MRU GmbH, Neckarsulm, Germany).

In the reduction stage, the oxygen carrier’s performance was evaluated before obvious carbon deposition occurred because the hydrogen purity can be affected by carbon entering the oxidation stage. The decline in CO flow rate was the starting signal for obvious carbon deposition. To ensure the reliability of the data, a carbon balance was considered. The error $E$ was calculated as

$$E=(n_{{\mathrm{CH}}_4,\mathrm{out}}+n_{{\mathrm{CO}}_2,\mathrm{out}}+n_{\mathrm{CO},\mathrm{out}}+0.5*(n_{{\mathrm{H}}_2,\mathrm{out}}-2*n_{\mathrm{CO},\mathrm{out}}))/n_{{\mathrm{CH}}_4,\mathrm{in}}*100\%$$

(1)

where $n_{{\mathrm{CH}}_4,\mathrm{in}}$ is the molar flow of CH₄ in the inlet feed. $n_{{\mathrm{CO}}_2,\mathrm{out}}$, $n_{\mathrm{CO},\mathrm{out}}$, $n_{{\mathrm{H}}_2,\mathrm{out}}$, and $n_{{\mathrm{CH}}_4,\mathrm{out}}$ are the molar flows of CO₂, CO, H₂, and CH₄, respectively, in the effluents. This error value was controlled between 90% and 110% throughout this work.

In addition, the CH₄ conversion $X_{{\mathrm{CH}}_4}$ (%), CO selectivity $S_{\mathrm{CO}}$ (%), oxygen capacity $n_{\mathrm{O}}$ (mmol/gOC), and oxygen release rate $\nu$ (μmol/(s·gOC)) can be obtained in the CH₄ reduction stage.

$$X_{{\mathrm{CH}}_4}=(n_{{\mathrm{CH}}_4,\mathrm{in}}-n_{{\mathrm{CH}}_4,\mathrm{out}})/n_{{\mathrm{CH}}_4,\mathrm{in}}*100\%$$

(2)

$$S_{\mathrm{CO}}=n_{\mathrm{CO},\mathrm{out}}/(n_{{\mathrm{CH}}_4,\mathrm{in}}-n_{{\mathrm{CH}}_4,\mathrm{out}})*100\%$$

(3)

$$n_{\mathrm{O}}=4*n_{{\mathrm{CO}}_2,\mathrm{out},1}+n_{\mathrm{CO},\mathrm{out}}$$

(4)

$$\nu =n_{\mathrm{O}}/t$$

(5)

2.3. Oxygen Carrier Characterization

The crystal phase was detected by XRD in a Smartlab diffractometer with Cu Kα radiation from 20° to 80° at a speed of 10°/min. H₂-TPR was performed to evaluate the activity of lattice oxygen in a fixed bed reactor and a TCD analyzer under a 10% H₂/Ar mixture (50 mL/min) over 100 mg of OC, using a heating rate of 10 °C/min from 200 °C to 1000 °C. A DFT calculation was carried out using the QUANTUM-ESPRESSO open-source software package, employing the PBE exchange functional and GBRV ultrasoft pseudopotentials. The kinetic energy cutoffs for wavefunctions and charge density were set at 40 Ry and 400 Ry. Meanwhile, a 5 × 5 × 5 Monkhorst−Pack k-point mesh and Gaussian smearing were applied. The oxygen vacancy formation energy was calculated from the energy of a stoichiometric crystal E_tot, the energy of a reduced crystal with one oxygen vacancy E_v, and the energy of an O₂ molecule E_O2 (E_v + 0.5*E_O2-E_tot). The model with a lower oxygen vacancy formation energy tends to release oxygen more easily. Meanwhile, CH₄-TPR experiments were performed to investigate the oxygen release of the oxygen carrier in a CH₄ atmosphere. In a typical experiment, 0.3 g of the OC sample was loaded into a fluidized bed reactor. The sample was then heated from 250 °C to 900 °C at a linear ramp rate of 10 °C/min under a continuous flow (100 mL/min) of a 10% CH₄/N₂ gas mixture. The composition of the reactor effluent was continuously monitored using an online gas analyzer (MRU VARIO Plus). The reaction mechanism was explored according to the valence state and surface element composition through XPS, which was conducted on a spectrometer equipped with a monochromatic Al Kα source. The C 1s at 284.8 eV was used for calibration.

3. Results and Discussion

3.1. Interaction Between Ni, Fe, and Al

The XRD patterns of oxygen carriers after calcination in Figure 2 were used to analyze the effects of Ni-Fe-Al interactions on the obtained phase. The diffractogram reveals that fresh NiFeAlO OCs predominantly consist of three phases: Fe₂O₃ (PDF #89-8104), Al₂O₃ (PDF #75-0782), and spinel (PDF #89-1696). Among the OCs, only the FeNi20Al exhibits the presence of the Fe₂O₃ phase, and no phases containing Ni are detected. The absence of Ni-related peaks is attributed to the low loading of nickel combined with its high dispersion on the support and peak overlap. Furthermore, as the Ni content increases, the full width at half maximum (FWHM) of the characteristic peak at 36.0° decreases from 0.29 in the FeNi20Al OC to 0.26 in the FeNi10Al OC to 0.22 in the FeNi7.5Al OC to 0.20 in the FeNi5Al OC. According to Scherrer’s formula, this suggests that the grain size of the spinel phase in the FeNi5Al OC is larger than that in the FeNi20Al OC. Therefore, the addition of Ni promotes the growth of spinel.

The effects of Ni-Fe-Al interactions on the oxygen release can be judged from the H₂ consumption curves for the H₂-TPR of the fresh NiFeAlO OCs in Figure 3. It is evident that oxygen release from the OC occurs after 800 °C, higher than the reduction temperature for nickel oxides and iron oxides in the existing reports. Combined with the XRD results, these results show that the oxygen release results from the reduction in spinel. Also, the Fe/Ni ratio has significant effects on the peak area but limited effects on the peak position. Therefore, the movement of Ni atoms into the spinel structure increases the amount of available lattice oxygen rather than the lattice oxygen activity, further proved by the following DFT results.

The oxygen vacancy formation energy calculated using density functional theory (DFT), shown in Figure 4, was also employed to explore the effects of the Ni-Fe-Al interactions on the lattice oxygen activity in spinel. Based on the FeAl₂O₄ lattice, the (NiFe)Al₂O₄ present in Figure 4b was obtained by substituting an Fe atom with a Ni atom, and Fe(FeAl)₂O₄ was constructed by replacing an Al atom with an Fe atom. Further, two kinds of (NiFe)(FeAl)₂O₄ were produced, as shown in Figure 4c,d, through replacing an Fe atom in the A site of Fe(FeAl)₂O₄ with a Ni atom. Upon comparing FeAl₂O₄ with Fe(FeAl)₂O₄, it can be observed that Fe atoms in the B site of the spinel structure (AB₂O₄) benefit the lattice activity. Doping the spinel structure with Ni atoms did not decrease the oxygen vacancy formation energy, meaning that there was no significant improvement in lattice oxygen activity.

The effects of Ni-Fe-Al interactions on the CH₄ activation can be observed from the CH₄-TPR results for the four OCs in Figure 5. The release of lattice oxygen is divided into two stages: low temperature (650–750 °C) and high temperature (750–950 °C). The low-temperature stage is attributed to the reaction of CH₄ with the highly active surface and near-surface lattice oxygen of the oxygen carrier. These oxygen species are more mobile or have weaker bonds with the metal cations, allowing them to be reduced by CH₄ at a lower temperature. The high-temperature stage corresponds to the reaction of CH₄ with the bulk lattice oxygen. The migration of bulk lattice oxygen to the surface to participate in the reaction requires a higher activation energy and thus higher temperature. The deviation of the ratio of H₂ to CO from 2 is caused by side reactions. The reaction at low temperature is steam methane reforming (CH₄ + H₂O → CO + 3H₂), while that at high temperature is methane cracking (CH₄ → CO + 2H₂). Also, the initial release temperature for lattice oxygen is 650 °C for all the OCs, significantly lower than the 800 °C during the H₂-TPR process. This suggests that a small quantity of Ni in the FeNi20Al OC provides sufficient catalytic sites. Meanwhile, the CH₄ conversion rate in the low-temperature region firstly rises and then declines as the Ni content increases, but a clear increasing trend in the oxygen release capacity of the OCs can be seen with increasing Ni content in the high-temperature region. Overall, the CH₄-TPR and H₂-TPR results demonstrate that the Ni-Fe-Al interactions enhance both the catalytic sites and the available lattice oxygen.

3.2. CLSR Performance of OCs

Figure 6 presents the outlet flows of CO₂, CO, H₂, and CH₄ during the 2nd to 10th CH₄ reduction stages. The variations in the CH₄ conversion, CO selectivity, oxygen capacity, and oxygen release rate for FeNi_xAl before carbon deposition are shown in Figure 7, from which the reactivity and stability of the oxygen carriers can be judged. The 2nd to 10th CH₄ reduction stages mainly consist of two stages, partial oxidation and rapid decomposition. Reducing the Ni content improves the CO selectivity, and the lowest CO selectivity of the FeNi5Al OC is more than 93%. The CO selectivity of FeNi20Al is close to 100% because it hardly detects CO₂ in the 2nd–10th CH₄ reduction stages. With an increase in CLSR cycles, the partial oxidation reaction of CH₄ tends to shift backward to varying degrees, resulting in a decrease in CH₄ conversion. The most obvious change in CH₄ conversion occurs with the FeNi20Al OC, from 92% to 85%. The FeNi10Al OC shows a slight decrease in CH₄ conversion, and the CH₄ conversions of the FeNi7.5Al and FeNi5Al OCs remain at 97%. However, the oxygen capacity of each OC before carbon deposition remains almost the same, and the oxygen capacities of FeNi20Al, FeNi10Al, FeNi7.5Al, and FeNi5Al are 0.7 mmol/gOC, 1.0 mmol/gOC, 0.9 mmol/gOC, and 1.0 mmol/gOC, respectively. The oxygen capacity does not monotonically increase, indicating that the oxygen release before carbon deposition is related not only to the amount of lattice oxygen available in the oxygen carrier but also to the surface catalytic sites such as Ni⁰ and Fe⁰. Too many catalytic sites cause rapid cracking of CH₄, and the generated carbon cannot be consumed immediately, leading to its transformation into inert carbon [40]. In addition, increasing the Ni content benefits the lattice oxygen release rate, but the FeNi10Al can reach 6.5 μmol/(s·gOC), which is close to the 7.0 μmol/(s·gOC) for the FeNi5Al.

The XRD patterns of the OCs after the 10th cycle are illustrated in Figure 8. Only two phases, Al₂O₃ and Fe-Al-O spinel, are observed in the used OCs. The spinel formed during the deep reduction of Fe₂O₃ cannot be regenerated by steam. No Ni-containing impurities are detected even after 10 cycles, suggesting the uniform dispersion of Ni atoms without significant precipitation. The consistency in the FWHM of the Al₂O₃ phase among various OCs indicates the high-temperature stability of Al₂O₃. The most significant change during cycling is observed for the spinel, with the characteristic peak shifting from 36.0° to 36.5°. This shift towards larger angles is ascribed to the incorporation of Ni atoms into the spinel, as the characteristic peak of NiAl₂O₄ is at a higher angle than FeAl₂O₄. Additionally, the reduction in the FWHM of the spinel phase around 36.5° for each cycled OC compared with fresh OC suggests a growth trend for the crystallite. Consequently, the increased Ni incorporation into the spinel and the growth of the spinel phase are identified as the main reasons for the decline in CH₄ conversion rate during cycling.

The H₂ consumption curves for the H₂-TPR of NiFeAlO OCs after the 10th CLSR cycle are illustrated in Figure 9. There is minimal fluctuation in the oxygen release temperature of the OC after the 10th cycle, compared with Figure 3. The high oxygen release temperature also indicates a reduction in spinel. However, the decrease in peak area indicates a reduction in the available lattice oxygen, and OCs with higher Ni content exhibit a more pronounced decline. The Fe-containing spinel phase grows with cycling, leading to a decrease in available lattice oxygen.

3.3. Cyclic Stability of FeNi10Al OC

According to the OC performance shown in Figure 7, FeNi10Al demonstrates the optimal CH₄ conversion and CO selectivity. A further increase in Ni content does not substantially enhance the oxygen capacity but leads to a higher oxygen carrier cost and lower CO selectivity. Consequently, the FeNi10Al OC was selected for 90 cycles of CLSR experiments to delve deeper into the cycling stability. The detailed gas flows of the FeNi10Al OC in the 10th, 50th, and 90th CH₄ reduction and H₂O oxidation stages are presented in Figure 10a,b. Additionally, the CH₄ conversion, CO selectivity, oxygen capacity, and oxygen release rate during the CH₄ reduction stage using FeNi10Al OC in long-term CLSR experiments are illustrated in Figure 10c. It can be inferred that the FeNi10Al OC exhibits robust stability during 90 CLSR cycles, with the CH₄ conversion rate remaining at 96–98%, the CO selectivity remaining above 98%, the oxygen capacity before carbon deposition being about 0.9 mmol/gOC, and the average oxygen release rate being around 6.5 μmol/(s·gOC). The outlet gas distributions in the 10th, 50th, and 90th cycles are largely consistent.

Regarding the phase changes detailed in Figure 11a, Al₂O₃ and spinel without other phases were identified through cycling. Analysis of the particle size distribution and N₂ adsorption and desorption curves (Figure 11b,c) of FeNi10Al after the 20th and 90th cycles indicate negligible changes in particle size and the specific surface area of the FeNi10Al OC across cycles. The sustained phase and physical strength could be the reason for the stability.

To summarize, the superior CLSR performance of the FeNi10Al oxygen carrier is attributed to suitable cooperation between the surface catalytic reaction and bulk oxygen supply, which is changed by the Fe/Ni ratio. As can be seen in Figure 6, a weaker catalytic reaction of FeNi20Al leads to low CH₄ conversion, while a stronger bulk oxygen supply of FeNi7.5Al and FeNi5Al causes CO₂ formation. Also, the sufficient amount of Al in FeNi10Al enhances the mechanical strength and high-temperature resistance. Therefore, Ni-Fe-Al interactions in FeNi10Al result in promising CO selectivity, oxygen release rates, and cycling stability.

3.4. Reaction Mechanism

The variation in surface elements in the FeNi10Al OC reflects the reaction mechanism of the oxygen carrier. XPS tests were conducted at five points, including complete H₂O oxidation in the 1st cycle (point 1), CH₄ reduction for 90 s in the 2nd cycle (point 2), CH₄ reduction for 150 s in the 2nd cycle (point 3), H₂O oxidation for 180 s in the 2nd cycle (point 4), and complete H₂O oxidation in the 10th cycle (point 5), as shown in Figure 12.

The Fe 2p spectra are divided into 2p3/2, a 2p3/2 satellite peak, 2p1/2, and a 2p1/2 satellite peak. The 2p3/2 peak can be separated into three subpeaks at 713.3 eV (Fe1), 711.1 eV (Fe2), and 709.4 eV (Fe3). According to the existing report, Fe1 and Fe2 represent Fe³⁺ ions, and Fe3 represents Fe²⁺ ions [41]. Comparing points 1–3, the Fe²⁺ ratio continuously increases during the second CH₄ reduction, also supported by the continuous attenuation of the satellite peaks. The Fe²⁺ content at point 4 is found to be higher than that at point 3, indicating that a deeper reduction can be achieved after the complete CH₄ reduction. The Fe²⁺ content at point 4 is significantly higher than that at point 1. Therefore, the proportion of lattice oxygen during the process from Fe²⁺ to Fe³⁺ is relatively low. It can also be judged from points 1 and 5 that the Fe²⁺/Fe³⁺ ratio after the 10th complete oxidation is lower.

The Ni 2p spectra are divided into 2p3/2, a 2p3/2 satellite peak, 2p1/2, and a 2p1/2 satellite peak. The peak fitting of 2p3/2 yields four subpeaks at 856.6 eV (Ni1), 855.5 eV (Ni2), 854.4 eV (Ni3), and 852.2 eV (Ni4). Ni1, Ni2, Ni3, and Ni4 represent Ni²⁺ in NiAl₂O₄, Ni²⁺ in NiFe₂O₄, Ni²⁺ in NiO, and Ni⁰, respectively [42,43]. According to the calculated ratio, NiAl₂O₄ is transformed into Ni⁰ and NiO, and is then converted to Ni⁰ and NiFe₂O₄ during the reduction process. The formation of Ni⁰ at the beginning of the reduction accounts for the high CH₄ conversion rate. During the H₂O oxidation stage, Ni exists initially in the form of Ni⁰ and NiFe₂O₄, and finally in the form of NiAl₂O₄.

The O 1s spectra are divided into O1 (>531 eV) and O2 (<531 eV), representing chemisorbed oxygen and surface lattice oxygen [44]. As the reduction reaction progresses, the surface lattice oxygen decreases, and the peak position of the surface lattice oxygen shifts towards lower binding energies. During the oxidation process with H₂O, the surface lattice oxygen is initially replenished. As the number of cycles increases, the content of chemisorbed oxygen increases after complete oxidation. Combined with the results of the cyclic experiments shown in Figure 7b, the increase in chemisorbed oxygen content is unfavorable for CH₄ conversion.

According to the XPS results, the reaction mechanism can be scheduled as shown in Figure 13. Fe mainly undergoes transformations between Fe³⁺ and Fe²⁺/Fe⁰. Ni primarily undergoes transformations between Ni²⁺ in NiAl₂O₄ and Ni⁰. The reaction process of oxygen mainly involves the release and replenishment of lattice oxygen.

4. Conclusions

This work prepared a series of NiFeAlO oxygen carriers with Fe/Ni molar ratios of 20, 10, 7.5, and 5 and conducted CLSR cycles in a small-scale fluidized bed reactor. The influence of Ni-Fe-Al interactions on the physicochemical properties of the oxygen carriers was characterized using XRD, H₂-TPR, CH₄-TPR, and XPS. The main conclusions are as follows:

(1) In terms of the Ni-Fe-Al interactions, Ni promotes the growth of a spinel structure. After 10 cycles, only Al₂O₃ and spinel phases are observed in NiFeAlO oxygen carriers. The incorporation of Ni atoms into the spinel structure enhances both the catalytic sites and the available lattice oxygen.

(2) In CLSR cycles, a decrease in the Ni content enhances CO selectivity. The oxygen capacity remains relatively constant before carbon deposition for all oxygen carriers. In terms of the average oxygen release rate, an increase in the Ni content improves the lattice oxygen release rate. FeNi10Al is a promising oxygen carrier and exhibits good stability over 90 cycles, with the CH₄ conversion rates maintained at 96–98% and CO selectivity above 98%. The oxygen capacity before carbon deposition is approximately 0.9 mmol/gOC, and the average oxygen release rate is around 6.5 μmol/(s·gOC).

(3) During CLSR cycles, Fe mainly undergoes transformations between Fe³⁺ and Fe²⁺/Fe⁰. Ni primarily undergoes transformations between Ni²⁺ in NiAl₂O₄ and Ni⁰. The oxygen reaction process mainly involves the release and replenishment of lattice oxygen.