Solvent-free, improved synthesis of pure bixbyite phase of iron and manganese mixed oxides as low-cost, potential oxygen carrier for chemical looping with oxygen uncoupling

Introduction

The considerable use of fossil fuels for energy generation by thermal power plants will continue in the coming years for both developed and developing countries, due to the huge energy demand [1]. Therefore, carbon dioxide (CO₂) as the primary greenhouse gas (GHG), which accounts for one third of the total CO₂ emissions solely from thermal power plants, would emit continuously to the environment due to the combustion of fossil fuels [2], [3], [4]. CO₂ capture and sequestration (CCS) is one of the viable among other options to generate cleaner energy from fossil fuels, where CO₂ can be captured and stored permanently or it can be used to produce the value added chemicals [5], [6]. The high capital and running cost as well as energy are needed in the existing CCS techniques due to the separation of diluted CO₂ from the exhaust as it contains largely N₂, unreacted O₂, H₂O and other gasses [7]. Chemical looping combustion (CLC) and chemical looping with uncoupling (CLOU) possess the inherent promise of CO₂-separation from the off-gasses than that of existing CCS techniques as they produce virtually CO₂ as the result of enhanced fuel combustion due to the inherent supply of pure gaseous O₂ from the oxygen carrier (e.g. metal oxides, mixed metal oxides, etc.) upon heating [8], [9], [10].

The pure phase of mixed metal oxides of Cu and Mn (CuMn₂O₄), Fe and Mn (FeMnO₃), etc. as low-cost, environmentally benign oxygen carriers are developed recently that can be used as alternative to the currently available oxygen carriers for CLC [11], [12], [13], [14], [15]. These mixed oxides form MnO as one of the products after chemical looping combustion reaction using methane, syngas, etc. and it provides the structural stability to their reduced co-products, which facilitates the regeneration of these oxygen carriers upon heating in air (self-supported oxygen carrier) [11], [12], [13]. The developed oxygen carrier showed enhanced CLC activity for methane and syngas combustion in terms of multi-cycle performance, stability, CO₂-selectivity, etc. under harsh CLC operational conditions than that of its counterparts (Cu-, Fe- and Mn-oxides) [11], [12], [13]. In addition, these oxygen carriers possess CLOU property, which is the tendency of releasing gaseous oxygen from their lattice upon heating [16], [17], [18]. In addition to the combustion of gaseous fuels, the CLOU property of oxygen carrier can be exploited for direct combustion of the solid fuels (e.g. coal, biomass, etc.) without prior gasification for thermal power plants applications [17], [19]. The production of pure gaseous oxygen would also find several promising applications in fuel cell-related reactions (e.g. cathodic oxygen reduction reaction), separation of O₂ from air, etc. The cost of these potential oxygen carriers, however, is still high, which restricts the implementation of chemical looping-based technology at a large scale for cleaner energy generation from fossil fuels due to their synthesis methods (e.g. template synthesis, co-precipitation, high energy ball mill, mechanical alloying, etc.), as they require huge amounts of energy, high purity metallic salts, various precipitants, pH control, washing solvents, etc [20], [21], [22], [23], [24], [25], [26]. Li et al. reported a facile route for the synthesis of FeMnO₃ hollow sphere/reduced graphene oxide composite, where FeMnO₃ was synthesized using carbon spheres as sacrificial template and demonstrated its superior electrochemical storage performance for supercapacitor applications [20]. The ferromagnetic FeMnO₃ was synthesized through mechano-chemical method using high energy ball milling and its structure was investigated using powder neutron diffraction method [21]. CeO₂-supported FeMnO₃ was synthesized through incipient wetness method as oxygen carrier for chemical looping combustion of H₂ and CH₄ [22]_. The pure phase of FeMnO₃ was synthesized through the oxidative thermal decomposition of Mn[Fe(CN)₅NO]·2H₂O at 1095°C [23]. Iron-manganese mixed oxides at nano-order were synthesized through sonochemical method, however, the synthesized bixbyite FeMnO₃ nanoparticles showed a small impurity of Mn₂O₃ [24]. The CO oxidation reaction on FeMnO₃ and Pd/FeMnO₃ catalysts was performed, where these catalysts were prepared through co-precipitation method [25]. The mixed oxides of Fe-Mn were synthesized through co-precipitation methods which were used for low temperature selective catalytic reduction of NO with NH₃ [26]. The authors of this article also demonstrated the stable and multi-cycle CLC performance of the pure phase of FeMnO₃ synthesized by co-precipitation method using the nitrate precursors of Fe and Mn and Na₂CO₃ as precipitating agent [12]. Although several methods have been reported for the synthesis of pure bixbyite FeMnO₃ either in the form of supported- or unsupported, however, these methods require high synthesis cost due to the use of organic templates, various solvents including washing purpose, high purity Fe and Mn precursors, high energy ball milling, high energy ultrasonicator, precipitating agents etc., which restrict their implementation in chemical looping-based technology at a larger scale for cleaner energy generation from fossil fuels. The expenditure for the synthesis of these oxygen carriers can be reduced significantly for the wide exploitation of chemical looping-based technology by the selection of low-cost synthesis method, abundantly available minerals, ores, etc.

Here, we successfully developed a solvent-free, improved synthesis method to prepare the pure bixbyite FeMnO₃ phase as potential oxygen carrier without using any solvent, precipitants and any sophisticated technique or process. FeMnO₃ was successfully synthesized by sequentially exposing of the stoichiometric precursors (3:1) of MnO and Fe₃O₄ under N₂ or reducing conditions (H₂ or CH₄) followed by air at 900°C, which showed a stable and enhanced multi-cycle, CLOU performance (10 cycles) (Scheme 1). These oxides were selected as the precursors for the synthesis due to the formation of Fe₃O₄ and MnO as thermodynamically stable products after CLC reaction of FeMnO₃ for CH₄ combustion [12].

Scheme 1:

Schematic illustration of solvent-free, improved synthesis of bixbyite, FeMnO₃ phase and its multi-cycle chemical looping with oxygen uncoupling performance. Ball and stick models show the crystal structures of (Fe, Mn)₂O₃ and (Fe, Mn)₃O₄. Brown, purple and red balls correspond to Fe, Mn, and O, respectively.

Experimental

Results and discussion

Figure 1a shows the pXRD profile of the products synthesized through the solvent-free, improved synthesis method. The products showed the characteristic diffraction peaks at 23.1°, 32.9°, 38.3°, 45.2°, 49.3°, 55.2° and 65.8° for the reflection planes of (211), (222), (400), (332), (431), (440) and (622), respectively, correspond to the pure bixbyite FeMnO₃ phase (Ia3̅, a=b=c=0.94 nm) (hereafter denoted as FMO). The pXRD profile for the physical mixture shows the characteristic diffraction peaks at 29.5°, 34.8°, 42.3°, 52.4°, 55.8°, 61.4° and 61.5° for the reflection planes of (112), (103), (220), (312), (321), (224) and (400) respectively, correspond to Fe₃O₄ (Imma; a=b=0.59 nm, c=0.83 nm) along with the characteristic diffraction peaks at 34.8°, 40.4°, 58.7°, 69.7° and 87.7° for the reflection planes of (111), (200), (220), (311) and (400) respectively, correspond to MnO (Fm3̅m; a=b=c=0.44 nm) (Fig. 1b). It should be noted that the pure bixbyite phase of FMO without any impurities was obtained when the precursors were subjected also under H₂ or CH₄ followed under air at 900°C (see Supporting Information, Fig. S1). These results suggest that the pure bixbyite phase can be synthesized by treating the abundantly available precursors under reduction followed by oxidation conditions.

Fig. 1:

pXRD profiles of FeMnO₃ prepared through solvent-free, improved synthesis method (a) and the physical mixture of MnO and Fe₃O₄ (b). The simulated XRD lines for FeMnO₃ are shown for reference purpose.

The precursors treated under N₂, H₂ and CH₄ (i.e. reductions conditions) at 900°C were analysed by pXRD to examine the influence of reduction conditions to the precursors. Figure 2 shows the pXRD profile of the physical mixture of MnO and Fe₃O₄ after the reduction treatment under N₂, H₂ and CH₄ atmosphere. No significant difference, however, the characteristic diffraction peaks of Fe₃O₄ and MnO are slightly shifted to the lower diffraction angle when the precursors were treated under N₂ atmosphere at 900°C (Fig. 2b). The precursors treated under CH₄ atmosphere showed the characteristic diffraction peaks at 29.6°, 34.9°, 42.4°, 56° and 61.5° for the reflection planes of (220), (311), (400), (333) and (440), respectively, correspond to Mn_1.03Fe_1.97O₄ (Fd3̅ m; a=b=c=0.85 nm) along with the characteristic diffraction peaks of MnO (Fig. 2c). Whereas, the precursors treated under H₂ atmosphere showed the characteristic diffraction peaks at 35.4°, 41.1°, 59.6°, 71.3°, 75° and 89.3° for the reflection planes of (111), (200), (220), (311), (222) and (400), respectively, correspond to (FeO)_0.497(MnO)_0.503( Fm3̅m; a=b=c=0.43 nm) (Fig. 2d). These results imply that the precursors are uniformly mixed through solid-state reaction under reduction conditions at elevated temperature. This would most likely facilitate the formation of pure mixed oxide form upon heating in air at elevated temperature through exothermic reaction which was further rationalized by performing the FE-SEM and elemental mapping of the precursors of Fe₃O₄ and MnO after the reduction treatment at 900°C (see Figs. 3 and 4 for the details).

Fig. 2:

pXRD profiles of the physical mixture of MnO and Fe₃O₄ (a) after subjecting under N₂ (b) CH₄ (c) and H₂ (d). MnO (■), Fe₃O₄(★), Mn_1.03Fe_1.97O₄(▲), (FeO)_0.497(MnO)_0.503(●).

Fig. 3:

FE-SEM image of the physical mixture of Fe₃O₄ and MnO (a) and its elemental mapping of Fe, Mn, and O (b–d).

Fig. 4:

FE-SEM image of the physical mixture of Fe₃O₄ and MnO after reduction treatment at 900°C (a) and its elemental mapping of Fe, Mn, and O (b–d).

Figure 3a shows the FE-SEM image of the physical mixture of Fe₃O₄ and MnO. The precursors were thoroughly mixed for 2 h and performed the microscopic observation of the resultant products. It shows the large-sized, agglomerated particles with irregular shapes and sizes. It also shows the non-uniform Fe and Mn distributions throughout the products as shown in Fig. 3b–c.

Figure 4a shows the FE-SEM image of the physical mixture of Fe₃O₄ and MnO after reduction treatment at 900°C. The large-sized, agglomerated particles with random in shapes and sizes were again observed after the reduction treatment. Whereas, the elemental mapping images (Fig. 4b–d) show the uniform distribution of Fe, Mn and O throughout the surface which most likely facilitates the formation of pure bixbyite phase on aerial oxidation at 900°C. The physical mixture of Fe and Mn-precursors undergoes solid-state reaction when they expose under reduction conditions that induce the formation of mixed oxides. The composition of mixed oxides under different reduction conditions is strongly dependent on the reduction potential of reducing agents. Despite the significant difference in the composition of mixed oxides after reduction treatment using various reducing agents, the pure bixbyite FeMnO₃ phase was observed when they treat under air at 900°C. These results suggest that the formation of mixed oxide phase of the precursors is the key to realize the pure mixed oxide of bixbyite FeMnO₃ phase. It should be noted that the direct calcination of the physical mixture of Fe and Mn-precursors in air at 900°C was not resulted the pure bixbyite FeMnO₃ phase and it forms always with some impurities. Therefore, the pure bixbyite FeMnO₃ phase can be prepared easily through the solvent-free, improved synthesis method without using any solvents, precipitants, sophisticated techniques, etc, under reduction followed by oxidation conditions.

The large-sized, agglomerated bixbyite FMO clusters with random in shapes and sizes were observed as shown in Fig. 5a. The FMO particles were agglomerated to form the large-sized clusters due to the harsh synthesis conditions (Fig. 5b). It should be noted that the morphology of oxygen carriers is not influenced significantly for both CLC and/or CLOU performance in the case of mixed oxides of Fe-Mn and Cu-Mn. In fact, we reported previously that the significant difference in morphology was observed both in the fresh-and CLC-treated oxygen carrier, however, their oxygen carrying capacity was not at all compromised with the increase of CLC cycles [11], [12]. The observed stable, enhanced multi-cycle CLC performance of the oxygen carrier is rationalized to the uniform distribution of their counter oxides of MnO and Fe₃O₄ as the final reduced products [11], [12]. Therefore, elemental mapping of FMO was performed on the randomly selected agglomerate to examine the elemental distribution of Fe, Mn, and O as shown in the FE-SEM image (Fig. 5c). As expected, Fe, Mn and O of FMO are uniformly distributed throughout the agglomerates as shown in Fig. 5d–f, which likely facilitate the regeneration in air upon heating for multi-cycle performance. However, the elemental mapping of Fe, Mn, and O of physical mixture of Fe₃O₄ and MnO showed that they are not uniformly distributed in the agglomerates (Fig. 3) which likely affect the CLC and CLOU performance (see Fig. 6b for the details).

Fig. 5:

FE-SEM image of bixbyite FeMnO₃ (a) and its magnified image (b). Elemental mapping of Fe, Mn and O (d–f) of the randomly selected agglomerate of bixbyite FeMnO₃ (c).

Fig. 6:

Temperature programmed oxygen desorption profile of bixbyite FeMnO₃ prepared through solvent-free, improved synthesis method (a), the physical mixture of Fe₃O₄ and MnO (b), Thermogravimetric analysis of bixbyite FeMnO₃ (c).

The CLOU performance (i.e. O₂-releasing behavior upon heating) of FMO and physical mixture of MnO and Fe₃O₄ for the reference purpose was examined in the presence of helium by the TPD method. The oxygen carrier that releases gaseous oxygen at higher temperature (750–1000°C) is usually preferred owing to the advantages of heat transfer for power production in the thermal power plants [27], [28]. FMO showed a strong, single O₂ desorption peak at 920°C without showing any O₂ desorption at lower temperatures as shown in Fig. 6a. The physical mixture of MnO and Fe₃O₄ (Fig. 6b) also showed the O₂ desorption peaks, however, at relatively weaker O₂ desorption and at a lower temperature range (Fig. 6b) compared to that of FMO, which is usually not favorable for oxygen carrier for thermal power plants applications. The amount of desorbed O₂ upon heating was estimated for both FMO and the physical mixture of MnO and Fe₃O₄. The estimated amount of the desorbed O₂ from FMO is to be 0.99 mmoles g⁻¹, which is significantly higher than that of its stoichiometric physical mixture of Fe₃O₄ and MnO (0.19 mmoles g⁻¹). These results clearly infer that the physical mixture of the individual oxides as oxygen carrier possesses inferior performance and may not be suitable for long-term CLC or CLOU applications. The estimated desorbed O₂ amount of FMO is 3.2 wt.% compared to the mass of FMO, which is almost same to its theoretical O₂ desorption amount of FMO under CLOU conditions (3(Fe, Mn)₂O₃→2(Fe, Mn)₃O₄+1/2O₂) [15]. TGA of FMO was performed further to examine the weight loss due to the desorption of oxygen through CLOU. FMO showed the sharp weight loss of 0.15 mg between 150 and 1010°C, which is again equal to 3.3 wt.% to the mass of FMO (Fig. 6c). Both TPD and TGA results are in excellent agreement with the theoretical O₂ desorbed amount of FMO under CLOU conditions. The reduced oxygen carrier after CLOU performance by TPD and TGA methods was analysed further to identify the product after desorption of gaseous O₂, which shows the spinel phase of FeMn₂O₄ (Fd3̅m, a=b=c=0.85 nm) (see Supporting Information, Fig. S2). A minor phase of Fe₂O₃ was obtained due to the decomposition of FeMn₂O₄ at higher temperature above 1000°C in the case of the reduced oxygen carrier by the TGA analysis as shown in the inset of Fig. 6c (see Supporting Information, Fig. S3).

Temperature programmed pXRD analysis of FMO was performed to examine the structural changes in particular lattice due to the release of oxygen upon heating from room temperature to 950°C (Fig. 7a). It should be noted that the oxide materials induce lattice strains at higher temperatures [29], [30]. FMO was heated gradually at set temperatures in N₂-environment and cooled down to room temperature in air. FMO showed the characteristic diffraction peak at 32.9° for the reflection plane of (222), which showed no significant changes up to 100°C, however, it showed the linear shift towards lower diffraction angle upon heating up to 950°C. Interestingly, the characteristic diffraction peak was virtually retained back to its original peak position upon cooling down to room temperature in air. The lattice constant of FMO (a=0.940 nm at 50°C) showed the linear increment after 100°C with the increase of temperature and reached to 0.9474 nm at 950°C. It was, however, retained virtually back to the original lattice value upon cooling in air to room temperature (Fig. 7b). These results infer that the lattice of FMO was altered significantly upon heating and cooling in N₂-environment and air, respectively. FMO is not released the oxygen when it is heated up to 100°C, however, it starts releasing the oxygen from their lattice which creates vacant sites in its lattice that induces the relaxation in the lattice structure of FMO. Hence, the characteristic diffraction peaks of FMO are shifted to lower diffraction angles upon heating above 100°C. When the reduced FMO is cooled down in air from 950°C to room temperature, the oxygen vacant sites are re-filled again by the oxygen derived from air to its original sites and thus the characteristic diffraction peaks of FMO are re-appeared again. The lattice constant also retained back upon cooling down in air from 950°C to room temperature.

Fig. 7:

Temperature programmed pXRD profiles of FeMnO₃ at set temperatures in N₂ and air (a). The temperature dependence of lattice constant of FeMnO₃ (b). FeMnO₃ was heated in N₂ (closed circle) up to 950°C and cooled down to 25°C in air (open square).

The stable and multi-cycle performance of oxygen carrier is essentially required to realize them as potential oxygen carrier. The oxygen carrier was re-oxidized after the CLOU reaction (i.e. complete exhaustion of removable O₂ from oxygen carrier) after each successive cycle to examine its multi-cycle CLOU performance. Spinel FeMn₂O₄ as the reduced product of oxygen carrier was re-oxidized in air at 900°C for 5 h and the product was characterized by pXRD. As expected, the bixbyite phase of FMO was regenerated upon re-oxidation (see Supporting Information, Fig. S4). The regenerated oxygen carrier was examined for its multi-cycle CLOU performance. Figure 8 shows the TPD profiles of FMO for 10 cycles, which showed no significant difference in the TPD profiles. FMO showed the strong O₂ desorption peak at 920°C and showed no change at all in O₂ desorption temperature. It is noteworthy to mention that the bixbyite FeMnO₃ phase was retained without forming any impurities even after 10 CLOU cycles. The amount of desorbed O₂ in the case of FMO was quantified by integration of the desorption peak and presented in the inset of Fig. 8. The amount of desorbed O₂ for the first cycle was estimated to be 0.99 mmoles g⁻¹ which was not changed significantly with the increase of cycles. These results imply that the solvent-free, improved synthesis method developed in this work can be employed for the large scale synthesis of bixbyite FeMnO₃ which possesses the potential as oxygen carrier for the CLOU application compared to the individual oxides.

Fig. 8:

Cyclic performance of temperature programmed oxygen desorption (TPD) profiles of bixbyite FeMnO₃ prepared through solvent-free, improved synthesis method. Inset shows the dependence of oxygen concentration as a function of cycles.

Conclusion

The simple solvent-free, improved synthesis method was developed to prepare the pure bixbyite FeMnO₃ phase as low-cost, potential oxygen carrier without using any solvents, precipitants and/or any other sophisticated techniques. The abundantly available precursors of Fe₃O₄ and MnO with 1:3 stoichiometry were mixed uniformly and calcined in the presence of N₂ or reduction conditions (H₂ or CH₄) followed by air at 900°C to yield the pure bixbyite FeMnO₃ phase. The large-sized, agglomerated clusters of FeMnO₃ showed the uniform elemental distribution of Fe, Mn and O which facilitate its excellent regeneration upon heating in air for multicycle performance. FeMnO₃ released the lattice O₂ upon heating through CLOU mechanism which was confirmed by O₂-TPD and TGA. The desorbed oxygen amount (3.2 wt.% compared to the mass of FeMnO₃) is almost equal to the theoretical O₂ desorption amount of FeMnO₃ under CLOU conditions. The lattice constant of FeMnO₃ increased linearly with the increase of temperature, however, without decomposition of bixbyite structure at 900°C and it is virtually retained back to its original structure upon cooling in air at 900°C. FeMnO₃ showed enhanced, multi-cycle CLOU performance than that of the physical mixture of its precursors MnO and Fe₃O₄. The spinel phase, FeMn₂O₄ was formed as the reduced product after CLOU reaction and can be readily regenerated to bixbyite FeMnO₃ upon heating in air at 900°C. The developed solvent-free, improved method for the synthesis of mixed oxides in pure form would find not only in CCS- but also in catalysis-related applications.