Reduction of Iron Oxides for CO₂ Capture Materials

Abstract

The iron industry is the largest energy-consuming manufacturing sector in the world, emitting 4–5% of the total carbon dioxide (CO₂). The development of iron-based systems for CO₂ capture and storage could effectively contribute to reducing CO₂ emissions. A wide set of different iron oxides, such as hematite (Fe₂O₃), magnetite (Fe₃O₄), and wüstite (Fe_(1−y)O) could in fact be employed for CO₂ capture at room temperature and pressure upon an investigation of their capturing properties. In order to achieve the most functional iron oxide form for CO₂ capture, starting from Fe₂O₃, a reducing agent such as hydrogen (H₂) or carbon monoxide (CO) can be employed. In this review, we present the state-of-the-art and recent advances on the different iron oxide materials employed, as well as on their reduction reactions with H₂ and CO.

1. Introduction

In the last few decades, CO₂ emissions have significantly increased, attracting the attention of the scientific community because these emissions are the main cause of the greenhouse effect [1,2]. The rise of the CO₂ concentration in the atmosphere is due to deforestation [3] and highly energetic industries, particularly in the sectors of chemicals [4], power, and steel, largely sustained by fossil fuels [5,6]. The steel industry alone is responsible for ~7% of the total CO₂ emission because of the wide use of fossil fuels [7,8].

Despite the increasing share of renewables for power generation and the acceleration towards the utilization of H₂ as a combustible in the steel industry [9,10], the International Energy Agency (IEA) reports that currently more than 80% of the world’s energy is still based on fossil fuel combustion [11,12], and in a mid-term scenario all large scale/industrial plants should employ some type of equipment in order to reduce CO₂ emissions [13,14]. For this reason, a wide set of energy companies are developing and testing strategies of carbon capture and sequestration (CCS) [15,16]

CCS [17,18] includes the separation, liquefaction, and finally storage of CO₂. The CO₂ can be stored (i) in deep ocean masses, (ii) in deep geological formations, or (iii) in the form of carbonates [19,20]. The most promising sequestration sites are the geological formations due to the possibility of storing ~2000 Gigatons of CO₂ [21,22]. In particular, the injection of CO₂ in oil reservoirs, mixed with different types of surfactants [23,24,25,26,27], is a validated technology to enhance oil recovery (EOR) [28,29]. Indeed, the combination based on the capture of CO₂ and its employment in EOR technology effectively reduces the CCS costs. Unfortunately, nowadays there are no power plants present that can operate with complete CCS methods because of the high cost and energy penalty of the capture stage, especially for hot and dilute flows derived from power plants [30,31]. Recent advances in research are primarily aimed at reducing costs and increasing the efficiency and selectivity of CO₂ separation and capture [32,33].

CO₂ capture technologies are based on either chemical [34] or physical methods [35]. In physical methods, CO₂ is adsorbed by a solid and is further released by decompression and/or heating under different conditions of temperature and pressure [36,37]. Contrariwise, the chemical methods are based on chemical reactions between CO₂ and liquids, such as amine solutions, or reactive solid sorbents, such as CaO [38,39]. The current benchmark for CO₂ capture in coal, oil, or natural gas power plants and other heavy industrial processes is absorption with amine solutions [40,41,42,43,44,45], while calcium looping capture is one step back in the Technology Readiness Level (TRL) [46]. Unfortunately, capture with amine solutions is very energy-intensive and expensive [47,48]; indeed, amine systems are capable of capturing ~85% of the CO₂ from the exhaust gas of a fossil-fuel power plant but with a penalty of ~25% of the plant electricity production [49,50].

CO₂ capture with membranes is a more recent solution but does not yet represent a mature alternative to amine capture. Indeed, their convenience is hampered by the large amount of exhaust gas that has to be treated [51,52]. The scientific community is, in fact, focusing on the enhancement of CO₂-nitrogen (N₂) membrane selectivity by using the facilitated mechanism membranes [53,54].

Another approach which has been recently gaining a prominent role is the use of metal oxides, such as iron oxides, magnesium oxides (MgO), and calcium oxides (CaO), for CO₂ capture, via the formation of carbonates [55,56]. In the second step, carbonates can be heated up to release back pure CO₂ gas and regenerate metal oxides [57,58]. Overall a cyclic process can be set, which includes the exothermic carbonation reaction and the endothermic carbonate decomposition [59,60].

In this review, we discuss the prospects of iron oxides in the context of energy transition, with a particular focus on their application for CCS. The features of the different types of iron oxides and their suitability for CO₂ capture will be surveyed. The thermodynamic and kinetic aspects of iron oxides reduction with H₂ or CO are also addressed.

2. Iron Oxides in the Context of Energy Transition and CCS

2.1. Motivation for the Use of Iron Oxides in CCS

The use of iron in the context of CCS is particularly appealing thanks to its large availability across the globe [61,62,63] and relative economy.

Iron is present in the planet mainly as hematite (Fe₂O₃), magnetite (Fe₃O₄), and wüstite (FeO). Other iron-based compounds include sulfides, hydroxides, and carbonates [64,65,66].

Iron oxides have been proposed both for pre-capture processes, such as chemical-looping combustion (CLC) [67,68,69], and for post combustion CO₂ capture. CLC processes have been developed in order to increase the concentration of CO₂ in the exhaust gases, thus reducing the costs of post combustion CO₂ capture processes. The high concentration of CO₂ is attained by performing carbon oxidation in a Fuel reactor, where the airflow is replaced by recycled flue gas. Here the oxygen necessary for carbon oxidation is released from the metal oxygen carriers, in particular iron oxides, which reduce to iron. The latter is, then, re-oxidized in a reactor, named the Air reactor, and recycled back to the Fuel reactor.

As far as the post combustion capture applications are concerned, iron has been proposed for the preparation of enhanced sorbents and used in combination with carbonaceous supports. However, the simplest and probably most direct and economic way to exploit iron oxides, in particular Fe₂O₃ and Fe₃O₄, is for capturing CO₂ from exhaust gases through the formation of carbonates (siderite FeCO₃). The carbonation stage can be followed by a high temperature decarbonation step to regenerate iron and release gaseous CO₂. Such a solution appears particularly appealing for capturing CO₂ from steel plants, where iron powders (even waste) are available on site and can be profitably exploited. Iron carbonates can be regenerated through the decarbonation reaction either in the same facility [70], or at locations far from the CO₂ emission/capture sites.

In this context, another important advantage can be envisaged: siderites can be safely transported via trucks, allowing CO₂ transport without the need of pipelines. However, in the techno-economic analysis, it must be considered that for the transport of 44 ton of CO₂, approximately 116 ton of material (in the form of siderite) must be handled (for a ratio of 44/156 = 0.38 g/g, which can be compared with the values of 0.1–0.3 g/g reported for CO₂ capture over metal organic frameworks (MOFs) [71]. Another issue that has to be considered for the economy of the process is that, in order to capture CO_2, the divalent or trivalent oxides of iron must be previously reduced in the metallic state [72,73] by a reduction agent, such as H₂ or CO. This step is crucial and might end up being the bottleneck of the process. This step will, indeed, be addressed in detail in the next paragraphs. Finally, the possibility of the utilization of iron materials over repeated cycles needs to be further investigated.

2.2. Hematite

Fe₂O₃ is one of the most important ores for industrial applications and represents 60%, of the iron ore reserve worldwide [74]. Oxygen (O₂) carriers using Fe₂O₃ as an active phase [75] or natural Fe₂O₃ have been widely investigated in the context of CLC. Notably, Al₂O₃ and SiO₂ in combination with Fe₂O₃, can further enhance the properties of O₂ carriers in CLC, their durability and reactivity [76].

Some studies investigated the use of Fe₂O₃ in natural ores and in waste materials as oxygen carriers in the CLC of fossil fuels, including coal and biomass with gas. Indeed, Zhang et al. studied the use of natural Fe₂O₃ in the CLC of coal under pressurized conditions and their results found no sintering on the Fe₂O₃ surface and good reaction performance [77]. Gu et al. used natural Fe₂O₃ [78] in the CLC of sawdust at 800 °C. Song et al. investigated the natural Fe₂O₃ CLC combustion of ShenHua bituminous coal and HuaiBei anthracite [79].

Furthermore, Fe₃O₄ and Fe₂O₃ can be employed as sorbents for CO₂ capture. Indeed, Mendoza et al. have investigated CO₂ capture and release employing Fe₂O₃ and Fe₃O₄. In particular, the chemical reactions among CO₂ and iron oxides were explored versus both the CO₂ pressure and ball mill process parameters. The experimental results indicated that the carbonation of iron oxides can be achieved at 10–30 bar and at room temperature. Furthermore, the complete calcination of FeCO₃, was investigated under Ar and vacuum atmospheres and it was found that at 367 °C FeCO₃ decomposed, yielding carbon and or ion and Fe₃O₄. In particular, it was highlighted that this mixture can reversibly capture CO₂ in several carbonation–calcination cycles [36].

2.3. Magnetite

Fe₃O₄ is an iron oxide widely employed as a component of industrial heterogeneous catalysts for its availability, stability, and low cost [80]. It has a spinel structure, and the mixed valence state, +2/+3, of the Fe can catalyze both acid–base and oxidation–reduction reactions [81]. For instance, in the Haber–Bosch process, Fe₃O₄ is part of the catalysts employed for ammonia production, in which Fe₃O₄ is reduced by H₂ to the active form α-Fe, allowing the adsorption and dissociation of the N₂ molecules [82].

In the last decade, in the CO₂ capture scenario, Fe₃O₄ has also been used to produce amine-functionalized nanofluids with higher CO₂ absorption rates than pure solvents [83]. Fe₃O₄ nanoparticles attracted scientific attention thanks to their low toxicity and high stability [84]. Park et al. have developed novel nano absorbents composed of Fe₃O₄ nanoparticles functionalized with different organic and inorganic materials such as (3-Aminopropyl) triethoxysilane (ATS), tetraethyl orthosilicate (TeOS), and L (+)-ascorbic acid (A). In order to synthesize an Fe₃O₄ double-functionalization, a primary coating with A followed by a secondary SiO₂ shell and subsequently, the amine functionalization by means of ATS was applied. The CO₂ absorption capacity of newly functionalized material, with A, was 11% higher than that of the un-functionalized ones, Fe₃O₄-SiO₂-NH₂. However, in order to determine the absorption improvement of functionalized material at 25 °C, concentrations ranging from 0.1 to 0.4 wt% were compared with those in which water was used as a reference. The highest CO₂ absorption improvement of 59.2% was achieved at a concentration of 0.3 wt% which is higher than blank Fe₃O₄, Fe₃O₄-A-SiO₂ at 0.3 wt%. Furthermore, a demonstration of cyclic performances was established in order to test its potential employment for industrial applications [85].

The observed enhancement of the CO₂ absorption of iron-amine-based nanofluids has been attributed to improved mass transfer compared to conventional amines solutions, thanks to a combination of effects [86]: the hydrodynamic effect [87,88], the shuttle effect [86,89], and the bubble-breaking effect [90,91]. However, these nanofluids still have limitations for industrial applications because they are susceptible to acid and oxidation conditions.

Furthermore, recently, a mechanochemical process aiming at capturing CO₂ by using Fe₂O₃ and Fe₃O₄ has been employed in the CO₂ capture field; this process led to the formation of FeCO₃ via carbonation reactions (1) and (2) [36].

$$F{e}_3{O}_4(s)+Fe(s)+4C{O}_2\to 4FeC{O}_3(s)$$

(1)

$$F{e}_2{O}_3(s)+Fe(s)+3C{O}_2\to 3FeC{O}_3(s)$$

(2)

In particular, Mendoza et al. [36], have investigated the ball milling condition effects (revolution speed, pressure, and reaction time) for both reactions (1) and (2) and found that the capturing capacity of the Fe₂O₃ and Fe system was higher than that of Fe₃O₄ and Fe at the same conditions of pressure, temperature, and reaction. In particular, the results demonstrated a capture capacity of 7.98 mmol CO₂ g⁻¹_sorbent in 3 h at 200 °C and 10 bar CO₂ pressure. Furthermore, the oxides could be completely regenerated by applying an oxidation step at 350 °C under airflow [15].

2.4. Wüstite

Wüstite is not commonly found as a natural material but is an intermediate derived from the reduction of iron ores, Fe₂O₃ and Fe₃O₄. The chemical formula of wüstite is Fe_(1−y)O, where y is the relative number of the ferrous iron ion vacancies. The O₂ content can change at different temperatures from a 23.16 wt% to 25.60 wt% range in variation. Fe_(1−y)O is thermodynamically unstable and prone to re-oxidize at ambient conditions; however, it can be also further reduced to metallic iron using agents such as carbon (C), CO, and H₂.

Recently O₂ carriers composed of Fe_(1−y)O mixed with other oxides, such as those of manganese (Mn) have been used in the chemical looping combustion of coal. It has been found that a fraction of O₂ was transferred via an uncoupling mechanism [92]. Perèz-Vega et al. prepared particles starting from Mn₃O₄ and Fe₂O₃ powders with an atomic ratio fixed at Mn:Fe, 77:23 by a ball mill. Subsequently, the solids mixture was calcined for 4 h at 1050 °C, and the calcined solid was milled for 300 min. Furthermore, the material behavior was evaluated during both the decomposition–regeneration of the [(Mn_0.77Fe_0.23)₂O₃] phase, bixbyite, in chemical looping with O₂ uncoupling, and the reduction–oxidation of [(Mn_0.77Fe_0.23)₃O₄], spinel phase, with CO, H₂, and CH₄. In particular, the chemical looping combustion redox cycle, using gaseous fuels, highlighted high reactivity with CO and H₂ and high O₂ transport capacity due to the reduction of [(Mn_0.77Fe_0.23)O], mangano-wüstite phase. Therefore, this material is considered a promising candidate to be employed in CLC with both coal and gas [93].

Furthermore, even if the wüstite is not stable, it has been used in combination with MgO in order to increase the CO₂ capture capability of MgO sorbents. Chen et al. investigated the ferric salt effect on the adsorption of strong solid bases in order to determine the optimal FeO amount to increase the CO₂ capture in flue-gas at T > 150 °C. They synthesized the FeO–MgO using magnesium acetate and ferrous acetate. In particular, after grinding and carbonization treatments of the two salts at 550 °C, a composite FeO-MgO is obtained in which the ferric species are surrounded by magnesia particles with a surface area higher than 160 m² g⁻¹. Furthermore, with a 3% mass ratio of FeO the basic sorbent can capture 35 mg g⁻¹ CO₂ at 150°. The CO₂ capture capacity and the ratio of strong basic sites results are higher than those of other MgO sorbents, resulting in an optimal situation for warm CO₂ capture [94].

3. Thermodynamics of Carbon Capture with Iron Oxides

In order to obtain low-cost iron products with high performances from natural iron ores, the reactions with reducing agents such as H₂ and CO need to be investigated in detail. The kinetics and thermodynamics of these reduction reactions will be discussed in the following chapter.

The reduction of Fe₂O₃ to the metallic iron (Fe) occurs stepwise. At temperatures > 570 °C, reduction is composed of three steps (Fe₂O₃ → Fe₃O₄ → Fe_(1−y)O → Fe) according to the Equation (3a–c). At temperatures < 570 °C, Fe₂O₃ reduces firstly to Fe₃O₄ and then to Fe (Fe₂O₃ → Fe₃O₄ → Fe), according to the two-step reaction reported in Equation (4a,b), while the Fe_(1−y)O intermediate is unstable.

$$3F{e}_2{O}_3+H_2\to 2F{e}_3{O}_4+H_{2}O\hspace{1em}\mathbf{T}>570 °\mathrm{C}$$

(3a)

$$(1-y)F{e}_3{O}_4+(1-4y)H_2\to 3F{e}_{(1-y)}O+(1-4y)H_{2}O$$

(3b)

$$F{e}_{(1-y)}O+H_2\to (1-y)Fe+H_{2}O$$

(3c)

$$3F{e}_2{O}_3+H_2\to 2F{e}_3{O}_4+H_{2}O,\hspace{1em}\mathbf{T}<570 °\mathrm{C}$$

(4a)

$$F{e}_3{O}_4+4H_2\to 3Fe+4H_{2}O,$$

(4b)

The stability of different iron oxide forms can be analyzed by the diagram composed of temperature versus O (wt%/100) reported in Figure 1.

The oxygen-to-iron content in the Fe-based systems scales in the following order: Fe₂O₃ > Fe₃O₄ > FeO. As previously mentioned, FeO is stable at T > 570 °C while at T < 570 °C Fe₂O₃ decomposes to Fe₃O₄ and Fe. However, the FeO stability area expands by increasing the temperature, because not all places in the lattice are occupied by iron ions. For this reason, the FeO formula has to be considered as Fe_(1−y)O in which 1−y represents vacancies in the iron lattice [96].

The Baur–Glässner diagram based on the thermodynamics of iron ore reduction is shown in Figure 2.

In particular, the diagram describes the stability areas of the different iron oxide phases during the reaction with H₂ (continuous three-phase line) and with CO (dashed three-phase line) as a function of gas oxidation degree (GOD) and temperature [95,97]. The GOD value is determined as the ratio of oxidized gas components over the sum of oxidized and oxidizable gas components (e.g., for a CO-CO₂ mixture, GOD = CO/(CO + CO₂)) and is an indicator of the reduction force of a gas mixture (the lower the GOD, the higher the reduction force of the gas mixture). It is possible to note from the diagram that when reduction is carried out with H₂, the stability region of Fe increases with temperature and GOD. Contrariwise, when reduction is carried out with CO, Fe is never stable for GOD > 0.5 and the iron stability region decreases with temperature for GOD < 0.5.

Therefore, in order to achieve Fe with H₂ the temperature of reduction must be lowered as the GOD decreases. Instead, in order to achieve Fe with CO, GOD must be kept above 0.5 and the temperature of reduction must be increased as the GOD decreases. For the best possible gas utilization, it is therefore recommended to operate at high temperatures with H₂ and at low temperatures with CO.

For the Fe-C-O system, the Boudouard reaction, see Equation (5), must also be considered. For this reason, the corresponding equilibrium line (calculated assuming a carbon activity of 1 and a pressure of 1 bar) is also reported in the diagram.

$$2CO\rightleftarrows C{O}_2+C$$

(5)

The Boudouard equilibrium line delimitates the Baur–Glässner diagram into two sections. For a CO-CO₂ gas mixture with composition and temperature below the equilibrium line, carbon precipitation can occur and hinder the reduction.

Furthermore, it is necessary to highlight that the reduction with CO is exothermic, while the reduction with H₂ is endothermic; therefore, in this case energy must be added to the system in order to maintain a constant reaction temperature. From an industrial point of view, an adequate process design, that involves the pre-heating of the materials, could aid in overcoming issues due to the supply of energy to the system [95].

4. Morphology and Microstructural Properties of Iron Oxides

Iron oxides involved in industrial processes and research activities have different features such as particle size, porosity, and purity degree. The true density of iron oxides decreases in the order Fe > Fe_(1−y)O > Fe₂O₃ > Fe₃O₄ (7.86 > 5.46 > 5.20 > 5.16 kg/m³ [98]. However, bulk or apparent density depends on porosity. In several industrial applications cm-size pellets or briquettes are used, compacting finer grains, from tens of micrometers to a few millimeters in size. The interstices between the grains generate large macroporosity from micro to even millimetric scale, while pores inside the fine grains contribute with smaller (micro and mesopore) range porosity.

Hakim et al. investigated the microstructure of fine powders of different iron oxides by SEM, shown in Figure 3. In particular, a honeycomb structure can be observed for FeO, a rhombohedral lattice for Fe₂O₃, and a clear cubical structure for Fe₃O₄. In terms of porosity, the following average pore diameters (d_pore) and pore volumes (V_pore) are reported: d_pore= 62.5 nm and V_pore = 0.0003 cm³/g for FeO; d_pore = 17.5 nm and V_pore = 0.029 cm³/g for Fe₂O₃; and d_pore = 15.5 nm and V_pore = 0.018 cm³/g for Fe₃O₄ [99].

4.1. Porosity Changes in Reduction of Iron Ores

The porosity and density of the iron ore materials are key parameters controlling the reduction kinetics of the iron ore materials [100]. An analysis of the cross sections of large and hard particles after partial reduction reveals that reduction follows a topochemical pattern and concentric layers of different oxides are established, composed of Fe₂O₃, Fe₃O₄, Fe_(1−y)O and Fe, respectively. Concentric layers are not observed for porous particles, yet a topochemical type of reduction could still be possible at a smaller scale, within the micro-grains which constitute larger pellets, Figure 4.

In several papers, the mutual relationship between porosity and reduction rate or progress was addressed [96,101,102].

Ghadi et al. compared several pellets of iron ores and highlighted the porosity, grain size, and particle size effect and found that the rate of reduction increased linearly with the initial porosity degree of the samples, Figure 5 [103,104].

Sarkar et al. [100] reported, instead, for twenty different samples of Fe₂O₃ pellets (diameter 12 mm), a progressive decrease in porosity and increase in density upon reduction at 900 °C with CO and H₂.

El-Geassy et al. studied the effects of temperature on the structure of Fe₂O₃ pellets upon reduction with H₂ and reported different trends in the evolution of porosity, according to the compaction degree and grains particles size. Indeed, for the less porous/more compact pellets and larger grains size, high temperature and prolonged reduction treatments induce sintering, which reduces the reactivity of the samples [101]. On the other hand, for highly porous pellets and small grain sizes [105], sintering effects are less important, while the densification of the unreduced oxide grains results in an overall increase in porosity and reduction reactivity.

4.2. Effect of Gangue on Reduction Kinetics and Porosity

Iron ores also contain additional gangue, such as Al₂O₃, SiO₂, CaO, and MgO, in different amounts, depending on the ore and the enrichment process. In general, gangue can have a prominent effect on the reduction rate of iron oxides and on sintering [101].

Teplov et al. investigated the rate of reduction with H₂ of natural Fe₃O₄ concentrates in the range 300–570 °C and observed that stronger oxides, such as Al₂O₃, Ti₂O₃, V₂O₃, Mn₃O₄, and MgO, that form solid solutions with Fe₃O₄, decrease reduction rate, because they decrease the Gibbs thermodynamic potential and the equilibrium oxygen pressure over Fe₃O₄. Among these oxides, Al₂O₃ strongly decreases the rate of hydrogen reduction of Fe₃O₄, while MgO showed a weak effect [106].

Kapelyushin et al. investigated the effect of Al₂O₃ on the reactivity of iron ore reduction. In particular, they compared the reduction rate of undoped and doped Fe₃O₄ with CO-CO₂ gas mixtures at different temperatures. As shown in Figure 5 of [107], an amount of alumina, 3%, enhances the reduction reactivity of iron ore, but for higher amounts of alumina the effect is reversed and reduction rate decreases [107,108].

Paananen et al. also investigated the effect of alumina on Fe₃O₄ during reduction in a CO/CO₂ gas mixture at 850 °C and found that an Al-doped Fe₃O₄ sample reduces more rapidly than the undoped Fe₃O₄ sample [109]. They attributed this to swelling and crack formation, suggesting that, upon doping, Al cations diffuse into the Fe₃O₄ structure forming hercynite. Upon reduction, high tension forces in the Fe₃O₄-hercynite phase close to the interface with the FeO would cause the disruption of the structure and generation of creaks.

As far as Ca is concerned, different papers addressed the effects of lime as well as of calcium carbonate on the reducibility of iron ores, and showed that their effect is beneficial, resulting in an enhancement of porosity upon reduction [101,110]. On the contrary, MgO would have a detrimental effect on iron reducibility [111]. As far as SiO₂ is concerned, controversial results have been reported in the literature; however, it seems that the effect of silica alone on the reduction reactivity of iron ores [112,113] can be considered of lower importance than that of other gangues, however SiO₂ can be involved in the formation of several phases when it is combined with Fe and Ca, such as in fayalite and calcio-ferrite.

4.3. Microstructural Changes in CO₂ Capture by Iron Oxides

The relation between microstructure and CO₂ captures by Fe₂O₃, Fe₃O₄, and FeO has been widely investigated in the literature [96,99,114].

Hakim et al. [99] investigated the microstructure and morphology of different iron oxides upon reaction with CO₂. In particular, they compared materials before and after CO₂ exposure for 4, 24, and 48 h and observed differences in carbonate growth on the surface, as shown in Figure 6.

In particular, the formation of bicarbonate, monodentate carbonate, and bidentate carbonates, as well as the morphological changes of the iron particles were observed. For FeO, SEM micrographs highlighted the honeycomb structure that became smooth and groove-like upon exposure with CO₂, suggesting the chemical interaction of the iron structure with several carbonate species. For Fe₃O₄ some sharp nanoparticles, indicative of carbonate formation, were clearly identified after exposure to CO₂; however, the cubic lattice of the iron oxide was preserved. A similar, but more remarkable effect was observed for Fe₃O₄, where nanoparticles form larger aggregates. Among all iron oxides, the uptake of CO₂ was consistently highest with Fe₂O₃ (at 3.01 mgCO₂/g).

5. Kinetics of Iron Ore Reduction

Several papers reported kinetic studies on iron ores reduction in different atmospheres. The majority, in fact, addressed the kinetics of reduction with H₂ and used thermogravimetric methods, either isothermal or non-isothermal. The results obtained are very scattered, as can be observed from the values of activation energies reported in

Table 1. Reduction of iron oxides by hydrogen with their amounts of apparent activation energy [95].

Operative Conditions	Reduction Reaction	Ea (kJ/mol)	Reference
Non-Isothermal with H2	Fe2O3 → Fe3O4	246	[115]
	Fe3O4 → Fe	93.2
	Fe2O3 → Fe3O4	162.1
	Fe3O4 → Fe	103.6
	Fe2O3 → Fe3O4	139.2	[116]
	Fe3O4 → FeO	77.3
	FeO → Fe	85.7
	Fe2O3 → Fe3O4	89.1	[117]
	Fe3O4 → Fe	70.4
	FeO → Fe	104.0	[118]
Isothermal with H2	Fe2O3 → Fe	57.1	[119]
	Fe2O3 → Fe	72.7
	Fe2O3 → Fe	89.9
	Fe2O3 → Fe3O4	30.1	[120]
	Fe2O3 → FeO	47.0	[121]
	FeO → Fe	30.0
	Fe2O3 → Fe	47.2	[122]
	Fe2O3 → Fe	51.5
	Fe2O3 → FeO	42.0	[123]
	FeO → Fe	55.0
	Fe2O3 → FeO	33.0	[124]
	FeO → Fe	11.0

, which indeed span between 57.1 (isothermal test) and 246 (non-isothermal test) kJ/mol. Kinetic models also spanned from simple first-order reactions to more complex models.

Notably, the reliability of kinetic expressions obtained by conventional (thermogravimetric) methods for any heterogenous (gas–solid) reaction relies on the fulfilment of some conditions:

1. Reactions result in a single stage of mass loss or in well-resolved and defined sequential stages.

2. The structure of the solid reactant is relatively constant throughout the experiment.

3. Inter- and intra-particle mass and temperature diffusion resistances are negligible.

It is evident that, due to the complexity of its reaction scheme, the reduction of iron by hydrogen does not fulfil the first condition.

The second condition is not fulfilled either, because the reduction of iron oxides leads to different metallic iron formations (Fe₂O₃, Fe₃O₄, and FeO).

As for the third condition, the mass transfer limitation (of the gaseous reactants) can be avoided by performing experiments at low temperature and with fine particles; however, in the reduction of iron ores, even for small particles size, an additional mechanism must be considered, that is, the transfer of electrons and iron ions generated by changes in iron oxides.

The complex interactions between physical and chemical phenomena and their impacts on the rate of the reaction of iron oxides with H₂/CO will be better explained in the following paragraph.

6. Effect of Physical and Chemical Phenomena on the Reaction Rate

The rate of reduction of iron ore particles depends on a complex series of chemical and transport phenomena, as reported in Scheme 1.

The first step is the mass transfer of the reducing gas from the bulk of the gas phase to the particles’ outer surface. This is followed by diffusion within pores and chemical reaction (including adsorption on the iron oxide internal surface and phase boundary reaction). Chemical reactions inside the particles produce gaseous products (CO₂ or H₂O), which have to diffuse towards the particle’s outer surface, through particle pores. Electrons and iron ions generated by changes in iron oxides, instead, diffuse towards the particle’s inner core by means of a solid-state diffusion mechanism. Reaction fronts and even concentrical layers of solid reduction products can therefore be established inside the particle as already shown in Figure 4 [125].

Notably, solid-state diffusion, which involves interstitial sites and defects, is much slower than gas diffusion and strongly dependent on the form of iron oxide. Gas transport in porous particles, instead, occurs through molecular diffusion mechanisms in pores with diameters larger than the mean free path, and according to a Knudsen diffusion mechanism in pores with diameters smaller than the mean free path [126].

As reported in previous paragraphs, reduction and heat treatment can progressively alter the porosity and microstructure of the material and modify the controlling reaction pattern. Two opposite phenomena are possible: on one side, the formation of creaks upon reduction can generate new porosity, which facilitates gas diffusion, and on the other side, sintering and the formation of dense iron layers reduces porosity and slows down gas diffusion.

Generally, in the literature, at least for fine particles [127,128,129,130], the first stage of reduction, Fe₂O₃ → Fe₃O₄, is relatively fast and limited by gas diffusion, while the second stage of the reaction, whereby Fe₃O₄ is reduced to Fe_(1−y)O or Fe, is slower and parameters such as temperature, concentration, particle size, and porosity all contribute to determine the rate-limiting step [123,124]. In particular, Table 2 reports a selection of experimental works that highlighted the rate-limiting process in the reduction of Fe₃O₄ to Fe_(1−y)O or Fe for different particles typology (grains or pellets), size, morphology, and porosity, and for different process parameters such as temperature, pressure, and gas composition. Results are also summarized in Figure 7a, for the reaction with H₂ and H₂/CO mixtures, and in Figure 7b, for CO and CO/H₂ mixtures.

At high temperatures and for large particles size or large porosity, the rate-limiting step is generally external mass transfer. On the contrary, for low temperatures and fine particles size, intrinsic kinetics/phase boundary reaction controls the overall rate [140,141,142]. In intermediate cases, the controlling rate mechanism could be either gas or solid-state diffusion and even mixed regimes are possible where, for example, transport and phase boundary reaction have comparable relevance [106,132].

Table 2. The effect of different parameters on iron oxide reduction.

Reducing Agent	Particle Size	Porosity (ε)	T (°C)	P (bar)	Ea (kJ/mol)	Limitation Steps (2nd Reaction Step)	Reference
H2	4 mm × 4 mm × 8 mm	-	900–1100	-	99.2	Solid-state diffusion	[138]
H2, CO	125 to 500 μm	0.15	350–600	10	91	Phase boundary reaction	[125]
CO	100–150 μm	-	700–850	-	80	Nucleation of FeO and then internal diffusion	[136]
H2	<74 µm	0.54	500–1100	68	51 for T < 650 °C; 84 for 650–900 °C; 176 for T > 900 °C	T < 650 °C chemical reaction, 650–900 °C chemical reaction + solid state diffusion, T > 900 °C solid state diffusion	[105]
CO-CO2 H2	50.8 µm	0.0115	240–417	-	71	Phase boundary reaction	[134]
H2/CO/mixtures	1.07–1.24 cm	0.06	850	-	-	Mixed (chemical reaction + pore diffusion) at the beginning, pore diffusion at the end	[143]
H2/CO	-	-	800–1000	-	48.64 with CO 63.19 with H2	First phase boundary reaction, then pore diffusion and mixed regime (at 800 °C and 900 °C).	[144]
H2	75–180 µm	0.27	700–1000	0.25–1	33	Pore diffusion	[123]
H2/CO	-	-	150–900	-	-	Solid-state diffusion	[135]
H2/CO	1.5–4.4 cm	0.31	700–1200	1–2	-	Gas diffusion At 700 °C in the late stages solid state diffusion	[133]
H2 CO/CO2	14 mm	-	900	-	-	Pore diffusion	[101]
H2/CO	12 mm	-	800–1000	-	-	Chemical reaction at the beginning, pore diffusion at the end	[137]
H2	50–160 µm	0.5	400	1	-	Chemical reaction	[106]
H2	100–160 µm	0.5	500	1	-	External gas transport + chemical reaction	[106]
H2	pellets	0.5	300–500	1	-	Transport in the external H2O layer	[106]
CO-CO2	~150 µm	0.69	590–1000	1	-	T < 700 °C, mixed control in the pores T > 700 °C external gas transport	[130]
CO-CO2	150–500 µm	0.42	590–1000	1	-	Chemical reaction	[130]
CO-CO2	<75 µm	0.23	590–1000	1	-	Chemical reaction	[130]
H2	0.249 µm	-	400–570		131.5 Fe3O4 → FeO 76.0 FeO → Fe	Chemical reaction	[131]
H2 CO	<100 µm	-	800–1100	-	41.15 with H2 54.19 with CO	Phase boundary reaction	[145]

Table 2. The effect of different parameters on iron oxide reduction.

Reducing Agent	Particle Size	Porosity (ε)	T (°C)	P (bar)	Ea (kJ/mol)	Limitation Steps (2nd Reaction Step)	Reference
H2	4 mm × 4 mm × 8 mm	-	900–1100	-	99.2	Solid-state diffusion	[138]
H2, CO	125 to 500 μm	0.15	350–600	10	91	Phase boundary reaction	[125]
CO	100–150 μm	-	700–850	-	80	Nucleation of FeO and then internal diffusion	[136]
H2	<74 µm	0.54	500–1100	68	51 for T < 650 °C; 84 for 650–900 °C; 176 for T > 900 °C	T < 650 °C chemical reaction, 650–900 °C chemical reaction + solid state diffusion, T > 900 °C solid state diffusion	[105]
CO-CO2 H2	50.8 µm	0.0115	240–417	-	71	Phase boundary reaction	[134]
H2/CO/mixtures	1.07–1.24 cm	0.06	850	-	-	Mixed (chemical reaction + pore diffusion) at the beginning, pore diffusion at the end	[143]
H2/CO	-	-	800–1000	-	48.64 with CO 63.19 with H2	First phase boundary reaction, then pore diffusion and mixed regime (at 800 °C and 900 °C).	[144]
H2	75–180 µm	0.27	700–1000	0.25–1	33	Pore diffusion	[123]
H2/CO	-	-	150–900	-	-	Solid-state diffusion	[135]
H2/CO	1.5–4.4 cm	0.31	700–1200	1–2	-	Gas diffusion At 700 °C in the late stages solid state diffusion	[133]
H2 CO/CO2	14 mm	-	900	-	-	Pore diffusion	[101]
H2/CO	12 mm	-	800–1000	-	-	Chemical reaction at the beginning, pore diffusion at the end	[137]
H2	50–160 µm	0.5	400	1	-	Chemical reaction	[106]
H2	100–160 µm	0.5	500	1	-	External gas transport + chemical reaction	[106]
H2	pellets	0.5	300–500	1	-	Transport in the external H2O layer	[106]
CO-CO2	~150 µm	0.69	590–1000	1	-	T < 700 °C, mixed control in the pores T > 700 °C external gas transport	[130]
CO-CO2	150–500 µm	0.42	590–1000	1	-	Chemical reaction	[130]
CO-CO2	<75 µm	0.23	590–1000	1	-	Chemical reaction	[130]
H2	0.249 µm	-	400–570		131.5 Fe3O4 → FeO 76.0 FeO → Fe	Chemical reaction	[131]
H2 CO	<100 µm	-	800–1100	-	41.15 with H2 54.19 with CO	Phase boundary reaction	[145]

Notably it has also been reported that the rate-controlling step can change throughout reduction because of changes in the microstructure of the particles [133,136,143,144]. Kawasaki et al. performed the reduction of porous Fe₂O₃ spherical pellets of a few centimeters in size in streams of pure H₂ or CO at temperatures between 700° and 1200 °C. At the higher temperature, the reduction rate was entirely diffusion controlled and a shell of reduction was clearly visible in sections of partially reduced specimens, which moved concentrically into the core of the samples. However at the lower reduction temperatures, Fe_(1−y)O was trapped inside dense shells of iron impervious to the reducing gas and the reaction proceeded by solid-state diffusion [133].

Variations in the controlling regime were also observed for finer particles. Chen et al. studied the reduction kinetics of 100–150 μm Fe₂O₃ by CO between 700–850 °C and concluded that the initial stage the reduction process was controlled by the gas–solid reaction occurring at the Fe/FeO interface, but subsequently it was controlled by the nucleation of FeO, and finally by diffusion. Moreover, the reactions of Fe₂O₃ → FeO and FeO → Fe occurred simultaneously but with different time dependences [136].

The changes in reaction regimes throughout the reduction of iron ores can be explained considering that the true density of the different oxides decreases in the order Fe > Fe_(1−y)O > Fe₂O₃ > Fe₃O₄ and that denser iron oxide forms imply slower diffusion [146,147,148,149].

Thermodynamic considerations, summarized in the Baur–Glässner diagram, shown in Figure 2, indicate that transition from Fe₃O₄ to the denser Fe and Fe_(1−y)O occurs above a critical temperature, whose value, however, varies according to the reduction potential of the gaseous atmosphere (which is the inverse of the GOD reported in Figure 2). In CO/CO₂ mixtures, a transition occurs at a temperature of ~570 °C for GOD above 50%, while it occurs at 1000 °C above 30% of GOD transition. In H₂/H₂O mixtures for GOD above 20%, the transition occurs above ~570 °C. As far as the differences between H₂ and CO as reducing agents are concerned, it can be observed that H₂ and CO have the same reduction potential at ~810 °C, while at higher temperatures the reduction potential of H₂ is higher than that of CO, while the CO’s reduction potential is higher than H₂′s at low temperatures.

Several authors investigated different gas composition effects on the reduction rate of iron ores [104,150]. Turkdogan et al. highlighted that the reduction rate of iron ore pellets in H₂/CO mixtures increased with H₂ concentration, likely due to its better gas diffusivity [101]. Also, El-Geassy et al. found that for the reduction of FeO micropellets between 900 and 1100 °C, the rate increased with the percentage of H₂ in the gas mixture, and they observed a neat formation of Fe nuclei. On the contrary, when reduction was carried out with pure CO, they detected a delay in reduction resulting in an incubation time [151]. These results confirm that the thermodynamically driven changes of iron forms end up also having an impact on the reaction rate, because of the different densities of the iron forms that determine different diffusional resistances which might ultimately control the overall reaction rate.

In addition, the diffusivity of the reducing agent molecule should be considered.

Indeed, Zuo et al. investigated the reduction rate in a H₂/CO atmosphere at 800 °C, where the thermodynamic transition between more or less dense iron forms is comparable for H₂ and CO, see Figure 3 of [137]. Even under these conditions, they observed that the reduction rate is higher with H₂, thanks to the better diffusivity of the H₂ molecule. They also conclude that increasing the CO amount, even by a small amount, in a gas mixture hinders the diffusion path of H₂ leading to an inhibition of the chemical reaction [137]. The decrease in the rate when CO is present in the reducing gas was also observed by El Geassy et al. who attributed it to a poisoning effect of CO molecules on the Fe_(1−y)O surface [152].

7. Conclusions, Challenges and Future Perspectives

The use of iron-based materials for CO₂ capture is still under exploration; however, research executed in the last five years has shown that iron oxide materials could facilitate CO₂ capture thanks to their utilization directly in industrial plants. An interesting route for CO₂ capture, especially in the context of steel industry, is the reaction of Fe₂O₃ or Fe₃O₄ with reduced iron and with CO₂ to form FeCO₃. The CO₂, captured in the form of FeCO₃, can be easily transported by truck, and on a decarbonation stage, can free the CO₂ and regenerate iron.

While the main advantages of this route are clearly the large availability and relative cheapness of iron across the globe, a successful application of natural iron ores requires a careful analysis and comprehension of the relationships between their structural properties (density, porosity, etc.) and process conditions (temperature, gas composition). The pre-reduction step, necessary to reduce iron ores to metallic iron, may, in this respect, play a crucial role.

Luckily, a great deal of information on the microstructural properties of the different iron oxide forms and on the kinetics and thermodynamics of their reduction can be obtained from the large amount of literature which has been produced over the last decades in the context of steel making. The following general indications can be pointed out:

• H₂, for thermodynamic and kinetic reasons, is a better reduction agent than CO thanks to better diffusion behavior. Indeed, the viscosity and the molecule size of H₂ are lower than CO, affecting the diffusion behavior.
• The first step of reduction, leading from Fe₂O₃ to Fe₃O₄, below 570 °C, is relatively fast compared to further reduction from Fe₃O₄ to Fe_(1−y)O or Fe; however, it is important because it opens up porosity.

The rate of reduction of Fe₃O₄ to Fe_(1−y)O or Fe can be limited by structural features of the starting material and by microstructural changes occurring upon reduction and heat treatment. The rate-controlling step at high temperatures and for large particle size or large porosity is generally external gas transport. On the contrary, for low temperatures and finer particles size, solid-state diffusion or phase boundary reaction can hamper the overall rate.

It is clear from the current state-of-the-art that the best conditions for the production of iron-based materials, which are not only efficient in capturing CO₂, but that can also be re-used over several cycles, cannot be superficially selected and require a careful analysis of the material properties, like grain size, porosity, mineralogy, and the presence of gangue. Further research activity in this field is warmly encouraged to provide guidelines for the utilization of natural iron ores in CO₂ capture, as well as to unlock the potential of advances/engineered iron-based materials for CCS.