Experimental and Theoretical Study of the Interactions between Fe₂O₃/Al₂O₃ and CO

Abstract

The behavior of Fe₂O₃/Al₂O₃ particles as oxygen carriers (OCs) for CO chemical looping combustion (CLC) under different reaction temperatures (700 °C, 800 °C, 900 °C, and 1000 °C) were tested in a lab-scale fluidized bed and a thermogravimetric analysis (TGA) unit. The results show that the oxygen carrier presents the highest reactivity at 800 °C, even after 30 cycles of redox reaction in a fluidized bed, while more obvious carbon deposition occurred for the case at 700 °C, and agglomeration for the case at 1000 °C. Moreover, the detailed behavior of the prepared Fe₂O₃/Al₂O₃ particle was detected in the TGA apparatus at different reaction temperatures. Furthermore, temperature-programming TGA experiments were performed to investigate the influence of different CO concentrations and CO/CO₂ concentrations on the reaction between CO and OC during the chemical looping combustion processes. Based on these experimental behaviors of the prepared Fe₂O₃/Al₂O₃ during the CLC of CO, the detailed models and electronic properties of the pure and reduced Fe₂O₃/Al₂O₃ supported the slabs, CO adsorption, and oxidation, and the decomposition reactions on these surfaces were revealed using density functional theory (DFT) calculations which went deep into the nature of the synergetic effect of the support of Al₂O₃ on the activity of Fe₂O₃ for the CLC of CO.

1. Introduction

Recently, chemical looping combustion (CLC) has been suggested as an effective technology to capture CO₂ without extra energy consumption and with nearly zero emission of pollutants [1,2]. The system contains two interconnected reactors: the fuel reactor (FR) and the air reactor (AR), as depicted in Figure 1. In the FR, fossil fuel is oxidized into CO₂ and H₂O by an oxygen carrier (OC), while the OC is reduced to lower valence states (see Reaction 1). After that, the reduced OC is transferred to the AR, and is oxidized into its original state by air (Reaction 2). The outlet gases from the FR are mainly CO₂ and H₂O, which can be easily separated by condensing and drying to obtain a high concentration of CO₂ [3,4,5]. Although Reaction 1 is often endothermic, Reaction 2 is exothermic and therefore the oxidized OC can also act as a heat carrier to transfer the energy needed to maintain the reduction reaction that has happened in the FR. Furthermore, the total amount of heat obtained from the CLC system is even more than that of conventional combustion because of the lower irreversibility of the two reaction courses [6,7].

$$\left(2n+m\right){\mathrm{M}}_x{\mathrm{O}}_y+{\mathrm{C}}_n{\mathrm{H}}_{2m}\to \left(2n+m\right){\mathrm{M}}_x{\mathrm{O}}_{y-1}+m{\mathrm{H}}_2\mathrm{O}+n{\mathrm{CO}}_2$$

(1)

$$2{\mathrm{M}}_x{\mathrm{O}}_{y-1}+{\mathrm{O}}_2\left(\mathrm{Air}\right)\to {\mathrm{M}}_x{\mathrm{O}}_y$$

(2)

However, an applicable OC usually acts as one of the key factors in the CLC system because of its vital role in transferring thermal energy and oxygen. It should possess excellent properties, including enough reactivity, high oxygen transfer capacity, low production cost, environmental friendliness, and a low tendency of fragmentation and attrition [8,9,10]. The main materials used as OCs are transition metal oxides, such as Mn₃O₄, Fe₂O₃, NiO, CuO, and CoO [11,12,13,14,15,16]. Among these metal oxides, Fe-based OC has been recognized as one of the most promising candidates, because it possesses the advantages of being low cost, having a good thermal stability, and being environmentally friendly [17,18,19,20]. Though Fe-based OC has high theoretical oxygen capacity (30%) from Fe₂O₃ to Fe, it has lower reactivity and oxygen transfer rate compared to Ni-based OC and Cu-based OC during CLC process. Reduction of Fe₂O₃ experiences the process as below: Fe₂O₃ → Fe₃O₄ → Fe_1−xO → Fe, which shows relatively high reactivity during the first stage from Fe₂O₃ to Fe₃O₄, while further reduction into FeO or Fe shows lower reaction rate [21,22]. Tang et al. [23] found that the transformation from Fe₂O₃ to Fe₃O₄ was suitable for the total oxidation of CH₄, while partial oxidation of CH₄ is more favored during the reduced stage from FeO to Fe, which implies that the reactivity of Fe-based OC depends greatly on its oxidation state.

In addition, while being used as OC, the pure metal oxides will be gradually reduced, which tend to agglomerate under high temperature in the FR, and the reduced OC is difficult to re-oxidize back to its original structure in the AR [24]. It has been proved that loading metal oxides onto inert oxide supports—especially onto Al₂O₃—can act as a very efficient method to solve these problems [25,26]. Cabello et al. [27] claimed that Al₂O₃ present in the Fe-based OC particles promotes full combustion of fuel into CO₂ and H₂O. Al₂O₃ not only influences the structure of Fe-based OC, but also alters its chemical properties for CLC of fuel. However, the synergetic effect of Al₂O₃ on the reactivity of the prefect and reduced Fe₂O₃ during the CLC process has not been adequately revealed. In particular, the mechanisms of CO oxidations and catalyzed CO decomposition on the surface of the gradually reduced Fe₂O₃ supported on Al₂O₃ remain unknown.

To address these points, a comprehensive study based on experiments was performed to detect the influences of temperature on carbon deposition and the reactivity of the oxygen carrier (Fe₂O₃/Al₂O₃) in a lab-scale fluidized bed, thermogravimetric analysis (TGA) unit. The synergetic effects of Al₂O₃ on the reactivity of the pure and reduced Fe₂O₃ for CO oxidations and decomposition during the CLC process were detected by performing density functional theory (DFT) calculations. The results may provide a deep insight into the gradually changing activity of the reduced Fe₂O₃/Al₂O₃ particle for CO adsorption, oxidation, and decomposition reactions.

2. Experimental and Theoretical Method

2.1. Preparation of the Oxygen Carrier

The properties of the oxygen carrier are largely dependent on its active components and the preparation methods [28]. A modified incipient impregnation method was used to prepare an oxygen carrier of Fe₂O₃/Al₂O₃ (the mass fraction of Fe₂O₃ is 15 wt %). Stoichiometric amounts of iron nitrate precursor are dissolved in water under continuous stirring at 40 °C in order to increase the solubility of the nitrate. After the complete dissolution of nitrate, the alumina particles were slowly added to the solution. Then the mixtures were further treated by ultrasonic for 12 h, and were dried at 110 °C overnight. Finally, the dried samples were calcined at 550 °C for 5 h under oxidizing conditions, then crushed and sieved to a particle size of 0.1–0.3 mm for the CLC experiments.

2.2. Combustion Experiment

CLC experiments were conducted in a fluidized-bed reactor, where a quartz tube (with an inner diameter of 16 mm and a total length of 1500 mm) was placed in an oven with a porous distributor plate located 550 mm from the bottom, and the oxygen carrier particles were placed on the porous quartz plate, as illustrated in Figure 2. Four different operating temperatures (700 °C, 800 °C, 900 °C, and 1000 °C) were used, and the real-time temperature was measured at a point of 10 mm above the porous quartz plate with an accuracy of ±2 °C. A 6 g sample of Fe₂O₃/Al₂O₃ particles with a size of 0.1–0.3 mm was placed on the porous plate and then initially heated to the reaction temperatures in an inert atmosphere (N₂). Then, the sample was alternately exposed to CO, and the air atmosphere was introduced into the reactor at the bottom of the tube, respectively. In order to prevent the mixing of fuel gas and air, N₂ was introduced to the reactor for 180 s between each reducing period and oxidizing period. The gas flow of N₂, air, and CO was 500 mL/min. The outlet gases were tested using a gas analyzer (Intelligent Analytical Instruments XLZ-1090, Sielins, Beijing, China) to record the concentration of CO₂. Furthermore, the fresh and the used oxygen carriers were tested by X-ray Diffraction (XRD, D/MAX-RB, Rigaku, Tokyo, Japan), Scanning Electron Microscope (SEM, LEO-1450, Carl Zeiss AG, Oberkochen, Germany), and Brunauer-Emmett-Teller (BET, Autosorb-iQ-MP, Quantachrome, Boynton Beach, FL, USA).

2.3. Thermal Gravimetric Analysis (TGA)

The reactivity of Fe₂O₃/Al₂O₃ with CO was carried out in a thermo gravimetric analyzer (PerkinElmer, STA6000, Boston, MA, USA). Between 40 and 50 mg of Fe₂O₃/Al₂O₃ was loaded in a platinum basket and heated up to 120 °C/min at 30 °C/min and held for 10 min to remove the moisture in the N₂ atmosphere. After weight stabilization, the sample was further heated to the desired temperature with the heating rate of 50 °C/min. Once the set temperature was reached, the experiment was started by exposing the sample to alternating CO and air atmosphere for the reduction and oxidation steps. To avoid the mixing of CO and air, nitrogen was introduced for 3 min after each reducing and oxidizing period. In the reducing period, Fe₂O₃ is mostly reduced to other lower oxidation states of iron oxide, such as magnetite (Fe₃O₄) and wustite (FeO). The degree of the oxygen carrier conversion (X_oc) in the reduction reaction is defined as:

$$X_{\mathrm{oc}}=\left(m_{\mathrm{ox}}-m\right)/\left(m_{\mathrm{ox}}-m_{\mathrm{red}}\right)$$

(3)

where m is the instantaneous weight of the sample, m_ox is the mass of the sample at its fully oxidized state, and m_red is the weight of the sample at its fully reduced state. The difference between m_ox and m_red in Equation (3) is the theoretical maximum amount of oxygen that can react with the CO.

2.4. Thermodynamic Analysis

Thermodynamic analysis was performed by FACTSAGE 7.1 (GTT-Technologies, Herzogenrath, Germany), and was used to predict the phases and compounds that can be formed, and their fractions and thermodynamic properties at given temperatures and pressures similar to the experimental conditions employed [29]. The Gibbs free energy change (ΔG) of the reduction reaction of Fe₂O₃ was calculated to discuss its oxidizing capability with CO. ΔG is a measure of the thermodynamic driving force that motivates the reaction process. Figure 3 presents the standard Gibbs free energy changes for all possible reactions between CO and iron oxide of different stable phases under various temperatures. The results indicated that all of reactions are more favored at higher temperature due to more negative values of Gibbs free energy, which is similar with the tendency of reactions between iron-based oxygen carrier and CH₄ in the literature in 2005 by Xin et al. [30]. The more negative the value of ΔG, the more likely that the reaction towards positive will proceed. Moreover, it also confirmed that the reactions between Fe-based oxygen carrier and CO became more difficult with decreasing oxidation sates. Several conditions for the generation of CO₂ via the reaction of iron based oxygen carrier with CO are as follows:

$$3{\mathrm{Fe}}_2{\mathrm{O}}_3+\mathrm{CO}\to 2{\mathrm{Fe}}_3{\mathrm{O}}_4+{\mathrm{CO}}_2$$

(4)

$${\mathrm{Fe}}_2{\mathrm{O}}_3+\mathrm{CO}\to 2\mathrm{FeO}+{\mathrm{CO}}_2$$

(5)

$${\mathrm{Fe}}_2{\mathrm{O}}_3+3\mathrm{CO}\to 2\mathrm{Fe}+3{\mathrm{CO}}_2$$

(6)

$${\mathrm{Fe}}_3{\mathrm{O}}_4+\mathrm{CO}\to 3\mathrm{FeO}+{\mathrm{CO}}_2$$

(7)

$${\mathrm{Fe}}_3{\mathrm{O}}_4+4\mathrm{CO}\to 3\mathrm{Fe}+4{\mathrm{CO}}_2$$

(8)

$$\mathrm{FeO}+\mathrm{CO}\to \mathrm{Fe}+{\mathrm{CO}}_2$$

(9)

2.5. Theoretical Method

Fe₂O₃(001) is one of the dominant growth faces of the natural α-Fe₂O₃; five atomic layers of Fe₂O₃(001) were modeled and supported on four matching atomic layers of Al₂O₃(001) to simulate the Fe₂O₃/Al₂O₃ system. Then, the reduced Fe₂O₃/Al₂O₃ was modeled by gradually removing the O atom on the top layer. The stable configurations of the pure and reduced Fe₂O₃/Al₂O₃ supported slabs are shown in Figure 4. Based on these optimized surfaces, CO adsorption, oxidation, and decomposition reactions were investigated.

All calculations were performed using DFT with the generalized gradient approximation (GGA) [31] of Perdew, Burke, and Ernzerhof (PBE) [32] to realize the exchange correlation potential [33]. The electronic wave functions were expanded in a plane-wave basis set with a cut-off energy of 350 eV in all calculations, and the convergence criteria was set at the medium level. All reactants, products, intermediates, and transition states were completely optimized, and were confirmed to be the local minimum energy by frequency analysis. Optimization of the energy, displacement, and force convergence was used as a criterion. The convergence valves were 2.0 × 10⁻⁵ Ha/atom, 5.0 × 10⁻³ Å, and 4.0 × 10⁻³ Ha/Å, respectively. The binding energies (E_B) for the studied interaction systems were calculated by using the following equation:

$$E_B=E\left(\mathrm{CO}-\mathrm{OC}\right)-E\left(\mathrm{OC}\right)-E\left(\mathrm{CO}\right)$$

(10)

where E(OC) is the energy of the isolated oxygen carrier, E(CO) is the energy of an isolated CO molecule, and E(CO–OC) is the total energy of the CO interacting with the Fe₂O₃ system under consideration. The transition state of the CO oxidation was searched using the complete linear/quadratic synchronous transit (LST/QST) [34,35].

3. Results and Discussion

3.1. Experimental Results in the Fluidized-Bed Reactor

Firstly, we focused on the performance of Fe₂O₃/Al₂O₃ for the CLC of CO in a fluidized bed reactor for five redox cycles. The reduction efficiency of Fe₂O₃/Al₂O₃ in the AR was characterized by analyzing the concentration of the outlet CO₂. Figure 5a displays the CO₂ concentration of the fifth reduction period at different reaction temperatures. Because CO was introduced into the FR, the reaction between Fe₂O₃/Al₂O₃ and CO made the concentration of CO₂ rapidly increase and reach 100% at 700 °C, 800 °C, and 900 °C. Then the concentration of CO₂ declined quickly after the Fe₂O₃/Al₂O₃ was reduced to a certain oxidation state, as the oxygen in the Fe₂O₃/Al₂O₃ particle was gradually depleted with the CLC process. The results indicated that the oxygen carrier has a high reactivity and can fully oxidize CO into CO₂ at a certain reduction period, and that pure CO₂ could be captured by controlling the reaction time of the oxygen carrier in the FR. However, the concentration of CO₂ only reached 80% at 1000 °C, which may be attributed to the agglomeration of the oxygen carrier and the occurrence of less oxygen transfer during the reduction period, because less carbon deposition happened in this case in comparison with the other cases.

Except for the reduction efficiency, carbon deposition on the reduced Fe₂O₃/Al₂O₃ is another important factor to characterize the behavior of the prepared Fe₂O₃/Al₂O₃. If carbon deposits on the reduced Fe₂O₃/Al₂O₃, carbon will be brought into the AR, leading to the generation of CO₂ in the oxidation period [36]. Figure 5b shows the CO₂ concentration for the fifth oxidation period in the AR under different temperatures. As can be seen in Figure 5b, the most obvious carbon deposition happens for the case at 700 °C, corresponding to the report that low temperature favors carbon deposition due to the Boudouard reaction of CO during the CO CLC process [37].

We further calculated the conversion of Fe₂O₃/Al₂O₃ through the yield of CO₂, which can be measured by the concentration of CO₂ and the rate of CO flow inlet into the reactor. The results indicate that the conversions of oxygen carrier are 0.503, 0.571, 0.514, and 0.322 at 700 °C, 800 °C, 900 °C, and 1000 °C within the period of 0 to 44 s (as displayed in Figure 5a), which means that a large number of Fe₂O₃ was reduced to FeO and Fe.

In order to analyze the carbon conversion and the amount of carbon deposition, the relative fraction (f_i) of CO and CO₂ in the outlet gases were calculated using the following equation [38]:

$$f_i=\frac{{{\displaystyle \int }}_0^t{n}_{out}{x}_i\mathrm{dt}}{{{\displaystyle \int }}_0^t{n}_{out}\left(x_{\mathrm{CO}}+x_{{\mathrm{CO}}_2}\right)\mathrm{dt}}$$

(11)

where x_i is the molar fraction of i indicated by the subscript in the outlet gases during the reduction period, n_out is molar flow rate of the outlet gases, and t is the reaction time. The carbon dioxide yield (γ_CO2) was used to quantify the conversion from CO into CO₂, which was calculated using the following equation:

$${\gamma }_{{\mathrm{CO}}_2}=\frac{f_{{\mathrm{CO}}_2}}{f_{\mathrm{CO}}+f_{{\mathrm{CO}}_2}}$$

(12)

where γ_CO2 = 1 denotes that all CO was completely converted into CO₂.

Figure 6a shows the yield of CO₂ as a function of temperature, where γ_CO2 first increased and then declined with the increase of temperature, and the maximum value was obtained at 800 °C. This result indicates that the oxygen carrier has the highest reactivity to oxidize CO into CO₂ at a temperature of around 800 °C. This trend means that there is an equilibrium selection between the favorable carbon deposition under relatively lower temperatures (<800 °C) and the agglomeration of the oxygen carrier at relatively higher temperatures (>800 °C). Figure 6b shows a decrease in the average amounts of carbon deposition with a temperature increase from 700 °C to 1000 °C, corresponding to the fact that low temperature favors reactions for reducing CO into C [39,40]. Although carbon formation occurred in the reduction period, the total amount of carbon deposited on the oxygen carrier was not more than 3% of the mole ratio of CO that is used in each experiment.

Further, long-term experiments were carried out to investigate the thermal stability of Fe₂O₃/Al₂O₃ under the temperature of 800 °C. After 30 reducing/oxidizing reaction cycles, the CO₂ concentration could still reach 100% and keep for seconds, which confirmed that the Fe₂O₃/Al₂O₃ has a good reactivity and stability for CO oxidation, as shown in Figure 7a. Figure 7b shows that carbon deposition increased slightly with reaction cycles, which implies that part of the active components had been deactivated. Generally, carbon conversion at 800 °C in long-term experiments still keeps a relatively high value, suggesting that Fe₂O₃/Al₂O₃ is suitable as an oxygen carrier for the CLC system.

3.2. TGA Experiments

Moreover, the detailed behavior of the prepared Fe₂O₃/Al₂O₃ particle was detected in the TGA apparatus at different reaction temperatures (700 °C, 800 °C, 900 °C, and 1000 °C). During the CLC process, in an ideal case, the active component of Fe₂O₃ could be reduced by CO to different final oxidation states, such as Fe₃O₄, FeO, and Fe, with the corresponding conversion χ of 0.11, 0.33, and 1, respectively. Figure 8 compares the TGA results under four different temperatures, where the Fe₂O₃/Al₂O₃ particle shows the highest conversation at 800 °C and the lowest at 1000 °C due to the agglomeration under high temperature reported by Bao et al. [41]. The TGA result corresponds to that of the fluidized-bed experiments. Moreover, all the conversation curves suggest that the reduction of Fe₂O₃/Al₂O₃ under various temperatures experienced three distinct reaction stages. The first stage is for the reduction of Fe₂O₃ into the oxidation state between Fe₃O₄ and FeO with the highest reaction rate and shortest time occupied. The second stage is for further reduction into another oxidation state lower than FeO with a lower reaction rate. Finally, the third stage is for deeper reduction of the OC, but where the OC was not completely reduced into Fe. The same results were likewise obtained from our previous work [42]. In short, deeper reduction degrees show lower reactivity. In addition, the reaction between Fe₂O₃/Al₂O₃ and CO shows the highest efficiency under 800 °C.

Further, temperature-programming TGA experiments were performed to investigate the influence of different CO concentrations and CO/CO₂ concentrations on the equilibrium of the reaction between CO and the OC during CLC processes. Figure 9a illustrates the results of a temperature-programming TGA investigation into the reaction between Fe₂O₃/Al₂O₃ and CO at different concentrations (and different CO/CO₂ concentrations). With temperature increases from ambient temperature to 900 °C, the CLC reaction between the OC and the CO under the conditions of different CO concentrations and different CO/CO₂ concentrations experiences two stages. The first stage corresponds to weight loss, since Fe₂O₃ was reduced by CO gradually with the formation of CO₂. After Fe₂O₃ was reduced to a certain reduction state, more Fe atoms appeared on the surface of the reduced Fe₂O₃, which could act as active sites to catalyze the decomposition of CO, resulting in carbon deposition on the surface showing a weight increase as the second stage. As could be seen in Figure 9, relatively lower CO concentration favors the reduction of Fe₂O₃, while higher CO concentration leads to more obvious weight increase at the second stage, where more carbon deposition happens on the surface of the reduced Fe₂O₃, leading to obvious weight increase. These results can be attributed to a higher concentration of CO resulting in quicker consumption of O atoms on the surface of Fe₂O₃/Al₂O₃, then more Fe atoms exposed on the surface promoting the catalytic decomposition of CO, resulting in carbon deposition on the surface. Figure 9b shows that a reasonable CO/CO₂ ratio favors the conversion of Fe₂O₃/Al₂O₃ with less carbon decomposition.

3.3. Characterization of the Oxygen Carrier

The BET test of the freshly prepared Fe₂O₃/Al₂O₃ particle shows a high specific surface area (192.75 m²/g) compared to other oxygen carriers in previous studies [43,44]. However, the specific surface area decreases to 113.76 m²/g, 96.68 m²/g, 76.35 m²/g, and 40.71 m²/g after use in a fluidized bed for five cycles at 700 °C, 800 °C, 900 °C, and 1000 °C, respectively. These results correspond very well to the analysis above that suggests that changes in the oxygen carrier reactivity and the CO₂ yield occur when the temperature increases.

Figure 10 presents the XRD patterns for the fresh Fe₂O₃/Al₂O₃ and the used Fe₂O₃/Al₂O₃ after five redox reaction cycles in the fluidized bed reactor. As can be seen, the reflections of Fe₂O₃ and Al₂O₃ in the fresh oxygen carrier are weak and broad. Both the characteristic peaks of the oxygen carrier became intensified and narrow after reacting with CO in the fluidized bed reactor. This illustrates that the lattice size of the crystals in the oxygen carrier was increased. Further, the characteristic peaks of Fe₃O₄ and FeO were detected in the used oxygen carriers, which indicated that most of the Fe₂O₃ was reduced to FeO and Fe₃O₄. In addition, the FeAl₂O₄ was also detected in XRD patterns. Hafizi et al. [45] also found that FeAl₂O₄ was generated in their oxygen carrier of Fe/Ca/Al₂O₃ prepared by the sequential impregnation method. Therefore, this irreversible reaction between active components and inert carriers is one of the problems that needs to be addressed during the CLC process [44].

The morphologies of the fresh Fe₂O₃/Al₂O₃ and the used Fe₂O₃/Al₂O₃ after 5 and 30 oxidation-reduction cycles under 800 °C in a fluidized-bed reactor were analyzed using SEM, as shown in Figure 11. In Figure 11a, the image of the fresh Fe₂O₃/Al₂O₃ has a dense surface and the grains are regular. However, the images of the used Fe₂O₃/Al₂O₃ for the 5th and 30th cycles had many morphological changes, which became coarser and more porous, as shown in Figure 11b,c. The SEM images indicate that the oxygen carrier had not agglomerated in any obvious way after 30 cycles.

3.4. The Mechanism of CO CLC on Fe₂O₃/Al₂O₃

Further, the internal association between the reactivity and the gradually changed structure (and property) of the reduced Fe₂O₃/Al₂O₃, as well as the synergetic effect of Al₂O₃ on the active component (the pure and the reduced Fe₂O₃) were discussed by investigating the electronic property of the Fe₂O₃/Al₂O₃ layer structure, and the mechanisms of adsorption, oxidation, and decomposition of CO on the pure Fe₂O₃/Al₂O₃ and the reduced Fe₂O₃/Al₂O₃ particle.

Figure 12 depicts the electron density difference for the stable Fe₂O₃/Al₂O₃ slab. According to Figure 4 and Figure 12, it can be observed that a hybrid occurs at the interface between Fe₂O₃ and Al₂O₃, leading to a longer Fe-O bond (an average value of 0.211 nm) at the interface than that in the bulk phase (an average value of 0.202 nm), and the newly formed Al–O bond at the interface is of 0.194 nm, slightly shorter than the bulk Al–O bond (0.191 nm). The newly formed Al–O shows a symmetric electron density different from that of the Al–O of the Al₂O₃ slab, showing an obvious electronic interaction at the interface. Figure 12b depicts the total density of state (DOS) for the O atom on the bottom layer of Fe₂O₃, as well as that for the O and Al atoms on the top layer of Al₂O₃, which shows an obvious energy overlap between the O atom on the bottom layer of Fe₂O₃ and the Al atom on the top layer of Al₂O₃, leading to a strong hybrid. The electron population around the O atom on the bottom layer of Fe₂O₃ is similar to that of the O atom on the top layer of Al₂O₃. This result confirms the stability of such supported Fe₂O₃/Al₂O₃ systems and the formation of FeAl₂O₄ species analyzed above by XRD.

Based on the stable Fe₂O₃/Al₂O₃, we want to go deeper into the mechanisms of adsorption, oxidation, and decomposition of CO during CLC. Because CO prefers to interact with the Fe atom of Fe₂O₃ through the C atom [37], the interaction between CO and the α-Fe₂O₃(001) surface was done by approaching the CO molecule to the Fe atom to do geometric optimization. Then, the adsorption energy was calculated according to Equation (10). Figure 11 compares the adsorption energy of CO on the pure and reduced Fe₂O₃/Al₂O₃ (seen in Figure 4), in comparison with the adsorption energy of CO on the pure and reduced Fe₂O₃ in our previous work [46]. The support that Al₂O₃ promotes in the interaction between CO and the surfaces can be observed in Figure 13, with the adsorption energy of –1.51 eV, –1.61 eV, –1.97 eV, –2.13 eV, –2.16 eV, and –2.25 eV for the CO-Fe₂O₃/Al₂O₃, CO-Fe₂O_2.86/Al₂O₃, CO-Fe₂O_2.57/Al₂O₃, CO-Fe₂O_2.29/Al₂O₃, CO-Fe₂O₂/Al₂O₃, and CO-Fe₂O_1.7/Al₂O₃ systems, respectively. Generally, deeper reduction shows stronger interaction, which corresponds to the case between CO and the surface without the support [46].

After the adsorption of CO on the OC surface, CO reacts to lattice oxygen on the surface, breaking O–Fe bonds and forming oxygen vacancy, and thus leading to the gradually reduced process of Fe₂O₃/Al₂O₃. Therefore, the corresponding oxygen vacancy formation energy (E_vac) could be used to characterize the reduction in the reactivity of the surface. The E_vac was calculated using the following equation:

$$E_{vac}=\frac{E_{{\mathrm{O}}_2}}{2}+E_{slab,r}^{k-1}-E_{slab,r}^k$$

(13)

where k refers to the number of O atoms, E_slab,r is the total energy of the optimized slab, and E_O2 is the total energy of the O₂ molecule. E_vac is the obtained energy for the O vacancy formation.

Figure 14 illustrates the energy profiles between E_vac and the theoretical valence of the Fe atom of the pure and reduced Fe₂O₃/Al₂O₃, in comparison with those for the pure and reduced Fe₂O₃(001) with the supported Al₂O₃ [47] range from 3.25 eV to 3.61 eV. However, E_vac from the pure and reduced Fe₂O₃/Al₂O₃ into the Fe₂O₂/Al₂O₃ range is between 2.86 eV to 3.30 eV, which is lower than those on the un-supported surfaces, respectively. The results suggest that Al₂O₃ has a synergetic effect in improving the oxygen transfer capacity from the bulk of the iron-based oxygen to the surface during the CLC of CO. The improvement in the reduction property of iron-based oxygen is due to the hybrid between O and Al at the interface.

As the experimental analysis results and our previous works [48,49] show, after the reduction of Fe₂O₃ into the lower oxidation state around FeO, a more metallic Fe atomic site appears on the surface, which can act as an active site for catalyzed CO decomposition. Herein, we compared the catalytic CO decomposition by Fe₂O₂/Al₂O₃ and Fe₂O_1.7/Al₂O₃, which is compared to the catalytic CO decomposition by Fe₂O₂ and Fe₂O_1.7. The reactions initiated from the stable adsorption of CO on the surface of the OCs, and the stable decomposition structure of CO on the surface of the OC was set as the final state. The results indicate that the decompositions of CO catalyzed by Fe₂O₂/Al₂O₃ and Fe₂O_1.7/Al₂O₃ pass through transition states with an energy barrier (E_a) of 3.89 eV and 3.16 eV, respectively. Analogously, CO decomposition on Fe₂O₂ and Fe₂O_1.7 had to pass through transition states with an E_a of 4.27 eV and 3.38 eV, respectively. These results imply that the carbon deposition rate increases with the reduction of the OC, but no evidence could verify that Al₂O₃ favors the prevention of carbon deposition on the reduced Fe₂O₃/Al₂O₃.

4. Conclusions

The reactivity, thermal stability, redox properties, and structural evolution of Fe₂O₃/Al₂O₃ were investigated in detail by multicycle tests in a fluidized bed and thermogravimetric analyzer at different reaction temperatures ranging from 700 to 1000 °C. The carbon deposition occurred during the reduction period and showed a declined tendency with increasing temperature. However, the oxygen carrier showed a high reactivity and did not show an obvious decrease even after 30 cycles at 800 °C. Temperature-programming TGA experiments suggested that a reasonable CO/CO₂ ratio favors the conversion of Fe₂O₃/Al₂O₃, thus avoiding high carbon decomposition. The SEM images and XRD spectrum indicated that the structure of oxygen carriers used at different temperatures had no obvious agglomeration except at 1000 °C. Further, density functional theory calculations showed that a hybrid occurs at the interface between Fe₂O₃ and Al₂O₃, with obvious electronic interaction at the interface. The electronic synergy of Al₂O₃ promotes the interaction between CO and the surfaces, hence favoring oxygen transfer from the bulk of iron-based oxygen to the surface during the CLC of CO. However, no evidence could verify that Al₂O₃ favors the prevention of carbon deposition on the reduced Fe₂O₃/Al₂O₃. These results may provide a fundamental understanding of the relationship between the behavior of Fe₂O₃/Al₂O₃ and CO.