Evaluation of Iron-based Oxygen Carrier Supported on Alumina/Titania for Charcoal Combustion through Chemical Looping Process

Received: August 18, 2018
Revised: April 28, 2019
Accepted: June 19, 2019

Download Citation: RIS|BibTeX|https://doi.org/10.4209/aaqr.2018.06.0222  

Cite this article:

Wu, H.C. and Ku, Y. (2019). Evaluation of Iron-based Oxygen Carrier Supported on Alumina/Titania for Charcoal Combustion through Chemical Looping Process. Aerosol Air Qual. Res. 19: 1920-1936. https://doi.org/10.4209/aaqr.2018.06.0222

HIGHLIGHTS

• Fe₂O₃/Al₂O₃/TiO₂ and Fe₂O₃/Al₂O₃ particles provided with high reactivity for CO/H₂.
• Higher mass transfer coefficient was observed for Fe₂O₃/Al₂O₃/TiO₂ reduction.
• Over 47% of maximum carbon conversions were reached for charcoal combustion in MBR.
• Less oxygen carriers are required for charcoal combustion with Fe₂O₃/Al₂O₃/TiO₂.
• Roughly 255W was released for CLC of charcoal with Fe₂O₃/Al₂O₃/TiO₂.

Keywords: Chemical looping process; Charcoal combustion; Hematite; Reduction kinetic; Moving bed reactor.

INTRODUCTION

Coal is the second most important energy source and distributed in many countries (WEC, 2016). Based on the Monthly Energy Review reported by the U.S. Energy Information Administration (EIA), coal contributes about 30% to electricity net generation (EIA, 2017). CO₂ is a major category of greenhouse gases (GHGs) generated from traditional thermal power plants and needs reducing emission to mitigate global warming. In the past few decades, the chemical looping process (CLP) has been recognized as an innovative substitute technology for fuel conversion to implement the efficient generation of energy and the inherent separation of CO₂. For chemical looping combustion (CLC) operation, the metal oxides are wrought as the oxygen source to react with fuel, as described by Eq. (1). Subsequently, the reduced particles are oxidized by air for cyclic applications, as described by Eq. (2). 

The Ellingham diagram illustrates the variation of oxidation Gibbs free energy (ΔG) with temperature for various compounds. Metal oxides in the Ellingham diagram, such as CuO, NiO, Fe₂O₃, Mn₂O₃, CoO, and CaSO₄, have strong oxidizing properties and can work as oxygen carriers for both full oxidation and partial oxidation (Fan et al., 2015; Luo et al., 2015). Zhou et al. (2014) studied the CLC of methane in a fixed bed reactor and a fluidized bed reactor using nickel oxide as the oxygen carrier and reported that the CO₂ selectivity and fuel conversion of methane combustion reached above 95%. Cao et al. (2014) demonstrated high reactivity for CuO/Al₂O₃ oxygen carrier reduction with H₂, and also reported that the oxygen releasing performance and thermal stability of the copper-based oxygen carrier were significantly impacted by the supporting material (Al₂O₃). The reaction mechanisms and kinetic parameters were determined by Wang et al. (2016) using Mn-based oxygen carriers for oxygen release in a fixed bed reactor. Compared to other materials, high melting point, high mechanical strength, and low cost are major advantages have been reported for an iron-based oxygen carrier (Chiu and Ku, 2012). Hence, hematite (Fe₂O₃) is considered to be a reliable and appropriate oxygen carrier for CLC implementation (Luo et al., 2014a; Mattisson et al., 2004). In order to retard agglomeration and attrition of the oxygen carrier, coupling of hematite with support materials, such as MgAl₂O₄, SiO₂, ZrO₂, MgO, Al₂O₃ and TiO₂, can improve the thermal and mechanical stability of oxygen carriers for CLC operation (Abad et al., 2007; Leion et al., 2008; Adánez et al., 2010; Chiu et al., 2014a; Ku et al., 2014). Ismail et al. (2014) and Sun et al. (2017) found that the addition of CaO and CeO₂ can promote the reactivity of iron-based oxygen carrier reaction with gaseous fuel. Zafar et al. (2006) found that the Fe₂O₃/MgAl₂O₄ oxygen carrier showed high reactivity with methane, whereas the SiO₂-supported Fe₂O₃ may not be feasible oxygen carrier due to the unreactive iron silicate was formed at high temperature. Dharanipragada et al. (2017) indicated that the activation energies of Fe₂O₃/MgAl₂O₄ reaction with H₂ for reaction and diffusion were 104 ± 2.3 and 38 ± 5 kJ mol^–1, respectively. Qin et al. (2013) reported that the energy barriers of Fe₂O₃, Fe₂O₃/Al₂O₃, Fe₂O₃/MgO and Fe₂O₃/ZrO₂ reduction with CO in the fuel reactor (corresponding to Fe₂O₃ to FeO) was calculated as 2.59, 0.99, 0.98 and 0.83, respectively, while the energy barriers of reduced Fe₂O₃, Fe₂O₃/Al₂O₃, Fe₂O₃/MgO and Fe₂O₃/ZrO₂ oxidization with O₂ in the air reactor was determined to be 0.08, 0.79, 0.2 and 1.14, respectively. The addition of support materials decreased the energy barriers in the reduction period, whereas the energy barriers in the oxidization period are increased. Moghtaderi and Song (2010) and Monazam et al. (2013) evaluated the activation energies of 25 kJ mol^–1 and 50.2–64.8 kJ mol^–1 for Fe₂O₃/Al₂O₃ and Fe₂O₃/MgO reduction with methane. Li et al. (2011) indicated the O^2– diffusivity of iron-based oxygen carrier was enhanced by using TiO₂ as support because of the lower energy barrier for O^2– migration within the dense solid phase. In comparison with the above support materials, Al₂O₃ and TiO₂ are a promising support material due to its high mechanical resistance, proper chemical and physical stability, and high melting point, and frequently applied to improve the thermal and mechanical stability of oxygen carriers for CLC operation. Chiu et al. (2014a) characterized the reduction mechanism of Fe₂O₃/Al₂O₃ oxygen carrier and observed high CO₂ yields for syngas and methane combustion in a fixed bed reactor. Liu et al. (2016) indicated that Fe₂O₃/TiO₂ oxygen carriers containing 20–30 wt.% TiO₂ sintered at 1100°C delivered high CO₂ concentration and H₂ conversion. However, the disintegration of Fe₂O₃/TiO₂ oxygen carrier structure was caused by the interaction between metallic Fe and reduced TiO_2-x or titania species occurred during the redox cycle operation. Ku et al. (2017) proposed the iron cations’ diffusion mechanism of Fe₂O₃/Al₂O₃ and Fe₂O₃/TiO₂ pellets during reduction and oxidation operations and also indicated that the Fe₂O₃/TiO₂ pellets exhibited approximately 2 times more attrition loss than Fe₂O₃/Al₂O₃ pellets. Kuo et al. (2015) found that the NiFeAlO₄ oxygen carrier provided high CO conversion and H₂ generation. Recently, more complex materials are developed and studied for chemical looping process, including perovskite-type oxides, spinel-type oxides, and mixed-metal oxides (Ksepko et al., 2016). Nevertheless, most of the studies were limited to mono-support oxygen carrier. Co-support oxygen carrier has not been extensively applied for the chemical looping process. Therefore, in this study, Fe₂O₃ supported on Al₂O₃-TiO₂ co-support was prepared and used in evaluating its feasibility to be an oxygen carrier for CLC application.

The classification of coal can be divided into four categories, including anthracite, bituminous coal, sub-bituminous coal and lignite (brown coal) (IEA, 2009). Several researches have been performed in the continuous CLC facilities for coal as feedstock (Berguerand and Lyngfelt, 2008; Sozinho et al., 2012; Gu et al., 2014; Mendiara et al., 2014; Ströhle et al., 2015). Song et al. (2012) utilized the Shenhua coal and Xuzhou coal as solid fuels to evaluate the effect of char gasification efficiency on the CLC with hematite ore in a 1 kW_th fluidized bed reactor, and also reported that CO₂ concentration at the fuel reactor outlet was increased with increasing char gasification efficiency, further validating that char gasification products react fast with the hematite ore. Markström et al. (2013) presented the design and operation of CLC with Colombian coal in a 100 kW_th fluidized bed reactor with ilmenite and indicated that the highest gas conversion and CO₂ capture efficiency of coal combustion reached about 84.1% and 96.4–99.5% for experiments conducted at 940–980°C in the fuel reactor, respectively. Furthermore, Linderholm and Schmitz (2016) reported that the highest gas conversion of nearly 87% was achieved for both Colombian coal and Calenturitas coal combustion with hematite ore, while the CO₂ capture efficiencies were obtained to be 94–98%.

Thon et al. (2014) applied a 25 kW_th two-stage fluidized bed reactor to investigate the CLC of lignite dust with Australian ilmenite and greater than 90 vol.% of CO₂ concentration could be achieved at the fuel reactor outlet. However, the reason is needed to be further investigated for the unconverted combustible gases appeared in the fuel reactor exhaust gas. Bayham et al. (2013) indicated that the carbon conversions were confirmed more than 90% for the combustion of both sub-bituminous coal and lignite coal using a 25 kW_th moving bed reactor operated with iron-based oxygen carriers, while CO₂ concentration in the fuel reactor outlet stream was obtained to be 99.5 vol.%. Kim et al. (2013) reported that carbon conversion for sub-bituminous coal and metallurgical coke combustion were found to be 97% and 81%. The carbon conversion for metallurgical coke combustion was lower than that for sub-bituminous coal combustion due to the reactivity of metallurgical coke was less than sub-bituminous coal. Hence, longer residence time was required for complete conversion of metallurgical coke. In addition, high CO₂ purity (> 99%) was obtained for both solid fuel combustion, further indicating that the counter-current moving-bed configuration for the fuel reactor is more effective in fully converting fuels to CO₂ and H₂O as compared to the fluidized-bed system. Biomass is considered to be a promising alternative resource for partial replacement of coal to generate sustainable energy in the future. In recent years, several biomasses have been developed and applied for chemical looping technology (Shen et al., 2009; Gu et al., 2011; Mendiara et al., 2013). Charcoal is a promising solid fuel due to its lower moisture content, higher calorific value, lower tendency to biodegradation, and easy to pulverize like coal (Gucho et al., 2015; Strandberg et al., 2015). Wu and Ku (2016) indicated that the maximum fuel conversion and oxygen carrier conversion for chemical looping gasification (CLG) of charcoal with Fe₂O₃/Al₂O₃ oxygen carrier were found to be 11.5% and 38.3%, respectively, and also reported that more syngas (CO/H₂) were generated for charcoal gasification conducted in the moving bed reactor operated with higher oxygen carrier-to-fuel ratios.

In this study, alumina/titania-supported Fe₂O₃ oxygen carrier was fabricated to investigate its reactivity with CO by comparing to the Al₂O₃ supported Fe₂O₃ oxygen carrier. In order to understand the kinetics of the reduction with CO, the influences of various experimental parameters on the reduction rate of Fe₂O₃/Al₂O₃ or Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers in a fixed bed reactor were discussed and expressed using a shrinking core model (SCM). Furthermore, the present study was also aimed at applying the chemical looping process using the Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers to combust charcoal. Carbon conversion and oxygen carrier conversion were determined from the gaseous composition at the moving bed reactor exhaust. The crystalline phases of the reduced oxygen carrier were identified for the samples drawn out from the reactor.

EXPERIMENTAL

Oxygen Carriers

Hematite (99.9% Fe₂O₃; China Steel Corp.) particles were dispersed in deionized water as Fe₂O₃ slurry. Alumina (99% Al₂O₃; Chin Jung Trading Co.), utilized as a support material, was added into the Fe₂O₃ slurry solution to enhance the mechanical strength of Fe₂O₃/Al₂O₃ particles. The mixed solution was stirred vigorously for 10 min and dried at 130°C. The cakes were then pulverized and screened their particle sizes from 1.2 to 1.4 mm. Finally, the particles were calcined in air at 1300°C for 2 h. For the preparation of Fe₂O₃/Al₂O₃/TiO₂ particles, a predetermined amount of titania (95.5% TiO₂; Unique Enterprise Co.) was added into the slurry solution containing Fe₂O₃ and Al₂O₃. The Fe₂O₃/Al₂O₃/TiO₂ particle was thereafter fabricated by mechanical mixing described in the above section. The prepared Fe₂O₃/Al₂O₃/TiO₂ particles were then put in a muffle furnace and calcined isothermally at 1200°C for 2 h. The weight percentage for Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ particles were determined to be 60/40 and 70/20/10. Crush strength and attrition of prepared particles were measured by a texture machine (TA.XTPlus) and an attrition analyzer following ASTM methods D4058-96. The true density, particle density and bulk density of prepared oxygen carriers were determined by Eqs. (3) and (4), respectively, similar defines and results were reported by previous researchers (Markström and Lyngfelt, 2012):

where ρ_t is the true density of the oxygen carrier without the pores which presented in the oxygen carrier; ρ_p is the particle density of the oxygen carrier including the pores, which was measured by the Archimedes method; ρ_b is the bulk density of the oxygen carrier; φ is 49.13% and 59.60% as the porosity of prepared Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂, respectively; ε is the fraction of void contained in the packed Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers, which were determined to be 41.61% and 40.17%, respectively. The main physicochemical properties of both oxygen carriers are shown in Table 1. Firck et al. (2016) and Ku et al. (2017) illustrated that the crushing strength of oxygen carriers is enhanced with increasing calcination temperature, hence, less crushing strength of Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers are obtained may be ascribed to the Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers calcined at the lower calcination temperature. Breault et al. (2016) indicated that the specific surface area becomes smaller, while the grain size and porosity become larger because the growth of larger grain promoted by high calcination temperature. The physical properties of the iron-based oxygen carriers reported by previous studies were similar to the results of this study that the specific surface areas were obtained to be in the range of 9.06 cm² s^–1 to 18 m² s^–1 (Gao et al., 2016; Huang et al., 2016; Ksepko et al., 2017). Huang et al. (2016) indicated that the decline of iron-based oxygen carrier reactivity is generally major caused by the specific surface area of oxygen carrier is decreased. However, the satisfactory reactivity of iron ore is observed for experiments carried out with low specific surface areas of particles. The reason is due to the release of lattice oxygen in the oxygen carrier was suggested to be a nucleation and nuclei growth process, thus the reactivity of oxygen carrier is not individually influenced by the surface structure of the particles (Huang et al., 2016). Additionally, Ksepko et al. (2017) illustrated that the reactivity and oxygen transfer capacity of the calcined Sinai ore cannot be ascribed only to the small surface area of particles, while the bulk diffusion is considered to be contributed to the reaction kinetics. As mentioned above, the surface area is a key governing factor for evaluating the basic properties of the oxygen carrier, which also may influence the reaction kinetics of the oxygen carrier with fuel gas.

Solid Fuel

Charcoal was applied as the solid fuel for studying the reaction behavior of solid fuels with oxygen carrier in a moving bed reactor. The commercial grade charcoal used in this study was purchased from Puguang Co., Ltd. and all experimental charcoal particles were pulverized and sieved for size between 1.2 mm and 1.4 mm. Thermogravimetric analyzer (TGA; STA 449 F3 Jupiter; Netzsch) was utilized to investigate the proximate analysis of charcoal up to 700°C with a heating rate of 20°C min^–1. The elemental analyzer (vario EL III; Elementar) was used to identify the elemental contents of C, H, O, N, and S. The fuel characteristics of the charcoal obtained based on the experimental results are listed in Table 2.

Fixed Bed Reactor System

Experiments were carried out in a cylindrical stainless steel 310S column of 200 mm height and 25.4 mm inner diameter, which was kept isothermal at the desired temperature, ranging from 875 to 925°C. A porous plate was installed inside the reactor to support the predetermined amount of prepared Fe₂O₃/Al₂O₃ or Fe₂O₃/Al₂O₃/TiO₂ particles. For reduction experiments, carbon monoxide was introducing into the reactor at a flowrate of 1 L min^–1 at different concentrations, ranging from 5 to 20%. After reduction, the nitrogen gas flowed to the reactor until the concentration of reducing gas in the reactor was removed. The reduced particles were subsequently oxidized for 30 min at an air flow rate of 1 L min^–1. The discharged streams were sampled at predetermined time intervals and measuring the concentration of gaseous product by a non-dispersive infrared sensor (NDIR; Model 6000i Series; Molecular Analysis). 

Moving Bed Reactor System

Fig. 1 shows the reaction system employed in this research contained an annular dual-tube moving bed reactor (ADMBR) made entirely of stainless steel 310S (outlet diameter: 76.20 mm) as depicted in our previous research (Wu et al., 2015). The operating temperature of the reactor was maintained at 850–900°C by a PID-controlled electric heating element. The flow rate of charcoal throughout the experiment was kept at 1.05 g min^–1. Steam molar flow rates were maintained constant at the desired rate with the N₂ flow rate of 2.2 L min^–1. Before starting the CLC experiments, the ADMBR was filled with the Fe₂O₃/Al₂O₃ particles (3.0 kg) or Fe₂O₃/Al₂O₃/TiO₂ particles (2.5 kg) of a known amount. Subsequently, the Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ feeding rate were fixed at 21.79 g min^–1 and 7.55 g min^–1, respectively. For experiments conducted under different conditions, samples of gaseous products were collected at predetermined time intervals from the ADMBR. The screw conveyor was used to draw out the solid samples and residues from the bottom of the ADMBR. All of the solid products were then oxidized at 600°C for 12 h in a muffle furnace. Here, the oxygen carrier particles were separated manually from the ash. A CLC system usually consists of a fuel reactor, an air reactor, and a cyclone. In the whole CLC system, smaller ash is relatively easier to be separated from the reactor through the cyclone. An NDIR sensor (Model 6000i Series; Molecular Analysis) was used for analyzing the gaseous product concentrations in the discharged streams. The concentrations of hydrogen in the effluent streams were determined with a China Chromatography Model 2000 gas chromatograph (GC) equipped with a thermal conductivity detector.

Fig. 1. Moving bed reactor system experimental setup.

RESULTS AND DISCUSSION

Kinetic Parameter Determination

The mechanisms of gaseous and solid fuel combustion with oxygen carrier have been proposed by previous studies (Adánez et al., 2012; Chiu and Ku, 2012) and are illustrated in Fig. 2. Gaseous fuel can directly react with the oxygen carrier to form CO₂ and H₂O, while the chemical reaction of solid fuel with an oxygen carrier takes place in two steps. The combustible gases generated during the solid fuel gasification, and then oxygen carrier was reduced by the gasified syngas (CO/H₂). Hence, the kinetic study of solid fuel gasification and oxygen carrier reduction should be simultaneously considered for CLC of solid fuel in order to preliminarily assess the performance of the prepared oxygen carrier for solid fuel combustion. In this study, carbon monoxide (CO) used as reducing gas to determine the reduction kinetic parameters of Fe-based oxygen carriers.

Fig. 2. Reaction mechanisms for gaseous and solid fuel combustion with oxygen carrier by CLC.

The iron-based oxygen carriers were fully reduced for carbon monoxide combustion as described by Eq. (5) (Fan, 2010):

Based on the mass balances of oxygen and carbon, the reduction conversion of iron-based oxygen carriers for carbon monoxide combustion is determined as:

where F_CO,in denotes the initial flow rate (mmol min^–1) of CO; F_CO and F_CO2 are the flowrate (mmol min^–1) of CO and CO₂ in the effluent stream after time t; M_Fe2O3 denotes the molar mass of Fe₂O₃; m_OC denotes the packing weight (g) of iron-based particles; x_Fe2O3 denotes the Fe₂O₃ content of the iron-based particles. Fig. 3 shows the effect of operating temperature on the reduction of iron-based oxygen carriers with CO. The reduction conversions of Fe₂O₃/Al₂O₃ oxygen carriers for CO combustion were increased with operation time and were slightly influenced with increasing operating temperatures. However, X_red for CLC of CO with Fe₂O₃/Al₂O₃/TiO₂ particles was enhanced for experiments conducted with longer operation time, and gradually maintained to about 33%, while reduction of Fe₂O₃/Al₂O₃/TiO₂ particles was similar for experiments carried out at operating temperature ranged from 875 to 925°C. Fig. 4 exhibits the effect of inlet carbon monoxide concentrations on the CLC of CO by iron-based oxygen carriers. The conversions of Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ reduction were increased for experiments conducted with greater inlet CO concentrations, possibly because of the existence of more carbon monoxide molecules. More carbon monoxide molecules are more ready to react with Fe₂O₃ contained in the iron-based oxygen carriers to create carbon dioxide.

Fig. 3. Effect of operating temperature on the reduction of Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ with CO in the fixed bed.

Fig. 4. Effect of inlet CO concentrations on the reduction of Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ with CO in the fixed bed.

In this study, the temporal reduction behavior of iron-based oxygen carrier was described by a shrinking core model (SCM). The similar applications were exercised in the previous researches (Abad et al., 2007; Cabello et al., 2014; Abad et al., 2015).

where R and r are the radius (m) of the fresh particle and the unreacted core; τ_gf and τ_pl are the time (s) required for complete conversion of the oxygen carrier when the reaction is controlled by the gas-film diffusion and the product layer diffusion; C_Ag is the inlet molar concentration (mol m^–3) of the gaseous fuel; b denotes the stoichiometric coefficient, which is determined to be 1/3 by Eq. (5); k_g denotes the mass transfer coefficient (m s^–1) between gaseous fuel and oxygen carriers; D_e denotes the effective diffusion coefficient (m² s^–1) of gaseous fuel in the product layer; D_pl,0 (m² s^–1) and Ea_pl (kJ mol^–1) are the pre-exponential factor and activation energy for the product layer diffusion reaction; k_d, k_d,0 and Ea_d (kJ mol^–1) are the decay constant, pre-exponential factor and activation energy for the product layer diffusivity; R_g denotes the ideal gas constant; T (K) denotes the operation temperature.

The reduction kinetics of Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ with carbon monoxide were regressed and calculated by the shrinking core model (SCM), as described in Eqs. (7)–(12). After fitting the conversion curves, the kinetic parameters were summarized in Table 3. The mass transfer coefficient (k_g) of around 0.16 mm s^–1 was calculated for the reduction of Fe₂O₃/Al₂O₃ with carbon monoxide, while the k_g of Fe₂O₃/Al₂O₃/TiO₂ for carbon monoxide combustion reached 0.22 mm s^–1, indicating that the reaction rate of carbon monoxide for Fe₂O₃/Al₂O₃/TiO₂ reduction was faster than that for Fe₂O₃/Al₂O₃ reduction. Less D_pl,0 was obtained for experiments with Fe₂O₃/Al₂O₃ oxygen carriers than those with Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers, explicitly demonstrating that the carbon monoxide more easily diffused through the product layer of Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers. Similar results were also observed for Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ reduction conducted with same operating temperature, as shown in Fig. 3, because the activation energies for the product layer diffusion reaction (Ea_pl) were respectively determined to be 255.35 kJ mol^–1 and 265.91 kJ mol^–1 for Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ reduction with CO. For experiments conducted at operating temperature ranged from 875 to 925°C, the decay constants (k_d) of Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ were determined to be in the range of 7.8 to 10.1 and 12.7 to 14.2, respectively. Higher k_d was obtained for experiments with Fe₂O₃/Al₂O₃/TiO₂ particles than those with Fe₂O₃/Al₂O₃ particles, demonstrating that the D_e of Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers was more easily depended on the oxygen carrier conversion. Hence, for experiments conducted with the conversion of Fe₂O₃/Al₂O₃/TiO₂ oxygen carrier higher than 33.33%, the slope of the reduction conversion vs. time curves was found to be significantly decreased. Furthermore, the slope of the reduction conversion vs. time curves was observed to be slightly altered for experiments conducted at various operating temperature, which might be also attributed to the decay constant (k_d) was dependent on the operating temperatures, consistent with the result found in literature (de Diego et al., 2014; Abad et al., 2015). For experiments conducted with an operating temperature of 900°C, the slope of the Fe₂O₃/Al₂O₃ reduction conversion vs. time curves were changed with inlet CO concentration ranged from 5 to 20%, as shown in Fig. 4, possibly due to high k_d,0 was obtained for Fe₂O₃/Al₂O₃ reduction with CO (Cabello et al., 2014).

Charcoal Gasification with Steam in an Empty Bed Reactor

Steam was employed to be a gasification agent, and the main mechanism of the charcoal gasification in the steam-containing gaseous stream could be described as the following (Gayán et al., 2010; Cuadrat et al., 2011; Wu et al., 2015):

The concentrations of the gaseous products for the gasification of charcoal by steam at various operating temperature are shown in Fig. 5. Elevated operating temperature increased the CO concentration, whereas diminished the CO₂ and CH₄ concentrations. It is ascribed that the endothermic reactions (steam reforming reaction and char gasification) are favorably occurred at higher operating temperature, instead the exothermic reaction (water gas shift reaction) is not favorable (Chiu et al., 2014b). In addition, the H₂ concentration was slightly declined with the operating temperature between 850°C and 900°C, due to the less volatile matter contained in charcoal. Fig. 6 shows the gaseous product concentrations during charcoal gasification at various steam flow rate.

Fig. 5. Effect of operating temperature on the gasification of charcoal in the empty bed.

Fig. 6. Effect of steam flow rate on the gasification of charcoal in the empty bed.

Increased steam flow rate diminished the CO and CH₄ concentrations, whereas enhanced the CO₂ and H₂ concentrations. Experiments conducted with high steam flow rate might be beneficial for steam reforming reaction and water gas shift reaction. 

CLC of Charcoal by Fe₂O₃/Al₂O₃ Particles

A series of experiments were carried out to study the CLC of charcoal by Fe₂O₃/Al₂O₃ particles. Results for experiments conducted at different steam flow rates are presented in Fig. 7. The maximum concentration of CO₂ was obtained at inlet steam flow rate of 157.7 mmol min^–1. The result implied that the combustible gaseous products (H₂, CO, and CH₄) generated from charcoal gasification were completely combusted by the oxygen carrier. However, the CO₂ concentration was significantly dwindled by increasing the steam flow rate. The enhancement of H₂ and CO concentrations could be attributed to the incomplete combustion of fuel gas generated from charcoal gasification. The carbon conversion (X_C) and the oxygen carrier conversion (X_OC) for charcoal combustion were determined using the following relationship:

where F_i is the effluent molar flowrate (mmol min^–1) of species i; i is denoted as CO₂, CO, H₂O, H₂ and CH₄; F_C,in, F_H,in and F_O,in are the inlet carbon, hydrogen and oxygen mole flow rate (mmol min^–1) of charcoal, respectively; F_O,out and F_H,out are the outlet oxygen and hydrogen mole flow rate (mmol min^–1) of unburned charcoal, respectively; F_H2O,in is the inlet mole flow rate (mmol min^–1) of H₂O; ṁ_Charcoal,in and ṁ_Charcoal,out are the inlet and outlet mass flow rate (g min^–1) of charcoal; ṁ_OC is the mass flow rate (g min^–1) of the oxygen carriers in the ADMBR. Fig. 8 shows the effect of inlet steam flow rate on the combustion of charcoal by the CLC process with Fe₂O₃/Al₂O₃ oxygen carriers in the ADMBR operated at 900°C. For steam feeding rate more than 157.7 mmol min^–1, the carbon conversion was retained to about 60%, indicating that charcoal gasification was not significantly enhanced with the application of more steam, due to charcoal gasification is the limiting step, consistent with the data found in the literature (Luo et al., 2014b). Hence, the longer residence time is required for the gasification of charcoal occurred in the fuel reactor. Experimental results illustrated in Fig. 8 also exhibited that the almost complete reaction between Fe₂O₃/Al₂O₃ oxygen carriers and fuel gases generated from the charcoal gasification could be accomplished for the experiment conducted with inlet steam flow rate of 157.7 mmol min^–1. However, as for experiments conducted with carbon conversion of about 60%, the effluent concentrations of H₂ and CO were enhanced with steam feeding rate ranged from 221.4 to 387.4 mmol min^–1, because of the insufficient residence time was provided for the CLC of fuel gases with Fe₂O₃/Al₂O₃ particles. As also indicated in Fig. 8, the maximum conversion of Fe₂O₃/Al₂O₃ particles (nearly 38%) was obtained for the combustion of charcoal with inlet steam flow rate of 157.7 mmol min^–1, demonstrating the component of reduced oxygen carriers was majorly FeO, consistent with the result found in the literature (Luo et al., 2014b). Nevertheless, the oxygen carrier conversions of Fe₂O₃/Al₂O₃ particles were found to decrease with increasing inlet steam flowrate, because less Fe₂O₃/Al₂O₃ oxygen carriers were considered to be reduced by fuel gases, such as CH₄, CO, and H₂, which is generated by char gasification and combustion.

Fig. 7. Effect of steam flow rates on the combustion of charcoal with Fe₂O₃/Al₂O₃ in the moving bed.

Fig. 8. Effect of steam flow rates on fuel and oxygen carrier conversions for CLC of charcoal with Fe₂O₃/Al₂O₃ in the moving bed.

CLC of Charcoal by Fe₂O₃/Al₂O₃/TiO₂ Particles

CLC experiments were conducted in this study to examine the combustion of charcoal under various inlet steam flow rate. As the experimental results illustrated in Fig. 9, outlet gas stream containing about 99% CO₂ and 1% CH₄ were achieved for experiments with inlet steam flowrate operated between 157.7 mmol min^–1 and 221.4 mmol min^–1, but the effluent concentration of CO₂ was dropped for experiments carried out with inlet steam flow rate of 276.7 mmol min^–1. These experimental observations may be attributed due to more combustible gas (CH₄, H₂, and CO) generated by Eqs. (13)–(16) were higher than that consumed by Fe₂O₃/Al₂O₃/TiO₂ reduction reaction. Further increases of inlet steam flow rate (> 276.7 mmol min^–1), however, could enhance the CO₂ and CH₄ effluent concentrations, and decrease the H₂ and CO effluent concentrations. The only slightly enhanced methane concentration may be ascribed to the low reactivity of the iron-based oxygen carriers with methane. The injections of excessive steam inhibited charcoal gasification due to insufficient residence time was provided. Thus, less charcoal was gasified to generate methane, hydrogen and carbon monoxide, and then hydrogen and carbon monoxide were further combusted with Fe₂O₃/Al₂O₃/TiO₂ oxygen carrier to generate carbon dioxide and steam. Besides, fewer H₂ and CO outlet concentration were observed for charcoal combustion with Fe₂O₃/Al₂O₃/TiO₂ particle than those with Fe₂O₃/Al₂O₃ particle, because of the hydrogen and carbon monoxide were considered to be consumed more rapidly by Fe₂O₃/Al₂O₃/TiO₂ reduction reaction than by Fe₂O₃/Al₂O₃ reduction reaction. Fig. 10 shows that carbon conversions were obviously increased for charcoal combustion conducted with 7.5 g min^–1 Fe₂O₃/Al₂O₃/TiO₂ particles, and inlet steam flow rate ranged from 157.7 to 221.4 mmol min^–1. Luo et al. (2014b) indicated that the kinetic of charcoal gasification was enhanced by increasing the flow rate of the gasification agent. More H₂O and CO₂ were produced from fuel gases combustion and was subsequently reacted with charcoal to perform gasification reaction, consistent with the results found in the literatures(Fan, 2010; Wu and Ku, 2016). However, the carbon conversions were found to be diminished by the addition of excessive steam (> 221.4 mmol min^–1). The carbon conversions were decreased for CLC of charcoal is possibly caused by the fuel gases not immediately combusted with Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers, and more H₂ and CO were presented in the fuel reactor effluent. As shown in Figs. 8 and 10, maximum carbon conversion of around 61.8% was achieved for charcoal combustion with Fe₂O₃/Al₂O₃, while the maximum carbon conversion of charcoal combustion with Fe₂O₃/Al₂O₃/TiO₂ reached about 47.2% for experiments conducted with inlet steam flow rate of 221.4 mmol min^–1, indicating insufficient oxygen was provided to fully combust charcoal. Fewer carbon conversions were observed for the combustion of charcoal with Fe₂O₃/Al₂O₃/TiO₂ than that for charcoal combustion with Fe₂O₃/Al₂O₃, further demonstrating charcoal gasification was impeded because the amount of Fe₂O₃/Al₂O₃/TiO₂ mass flow for charcoal combustion was lower than that of Fe₂O₃/Al₂O₃ mass flow.

Fig. 9. Effect of steam flow rates on the combustion of charcoal with Fe₂O₃/Al₂O₃/TiO₂ in the moving bed.

Fig. 10. Effect of steam flow rates on fuel and oxygen carrier conversions for CLC of charcoal with Fe₂O₃/Al₂O₃/TiO₂ in the moving bed.

Fig. 10 reveals the X_OC of Fe₂O₃/Al₂O₃/TiO₂ particles were enhanced for experiments conducted with inlet steam flow rate ranged from 157.7 to 221.4 mmol min^–1, demonstrating increased steam flow rate to promote the charcoal gasification, and further enhance the Fe₂O₃/Al₂O₃/TiO₂ reduction. Subsequently, the oxygen carrier conversion was markedly dropped with inlet steam flow rate ranged from 221.4 to 387.4 mmol min^–1, indicating not enough residence time was provided to reduce Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers with the fuel gases generated by char gasification and combustion. Thus, experiments conducted with excessive inlet steam flow rate might not be favorable to perform iron-based oxygen carrier reduction. Less oxygen carrier was demanded for the combustion of charcoal with Fe₂O₃/Al₂O₃/TiO₂ than that for charcoal combustion with Fe₂O₃/Al₂O₃, probably because lattice oxygen present in the Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers are more ready to react with combustible gas, especially hydrogen and carbon monoxide. Moreover, fewer oxygen carriers are required may also be ascribed to the higher Fe₂O₃ contained in the Fe₂O₃/Al₂O₃/TiO₂ oxygen carrier. Abad et al. (2007) analyzed the design criteria for a CLC system and reported that the solid circulation rate was decreased with increasing the metal oxide content for the combustion of fuel gases (H₂, CO, and CH₄) with different metal oxides (CuO, NiO, and Fe₂O₃).

Crystalline Phase Identification and Heat Analysis

The XRD diffraction peaks of prepared Fe₂O₃/Al₂O₃ particles for the crystal-phase transformation are presented in Fig. 11, indicating that only Fe₂O₃ and Al₂O₃ are found for fresh oxygen carriers. However, the crystal phases of the reduced Fe₂O₃/Al₂O₃ particles identified from the XRD spectra can be classified into three components of Al₂FeO₄, Fe₃O₄, and Al₂O₃, indicating that the amount of oxygen utilization from F₂O₃ to slightly Fe₃O₄ and FeO. A similar observation was obtained by Ishida et al. (2005) and Ku et al. (2014). Zhu et al. (2016) analyzed the atomic structural evolution during the reduction of α-Fe₂O₃ nanowires and also reported that more oxygen vacancies were formed as the reduction continues. Thus, the rhombohedral structure of α-Fe₂O₃ become unstable and transformed into the cubic structure of Fe₃O₄. In this study, Fe₃O₄ was further reduced to form the cubic structure of Al₂FeO₄. Few Fe₃O₄ was observed in the reduced Fe₂O₃/Al₂O₃ particles possibly due to the presence of insufficient residence time. Subsequently, the reduced Fe₂O₃/Al₂O₃ oxygen carriers were completely oxidized by air to generate Fe₂O₃ and Al₂O₃, which were observed in the XRD pattern.

Fig. 11. X-ray diffraction patterns of fresh, reduced and regenerated Fe₂O₃/Al₂O₃ oxygen carriers for charcoal combustion.

As shown in Fig. 12, Fe₂O₃, Fe₂TiO₅ and Al₂O₃ were the major crystalline phases of fresh Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers. For an experiment conducted with inlet steam flow rate of 387.4 mmol min^–1 and Fe₂O₃/Al₂O₃/TiO₂ flow rate of 7.55 g min^–1, the reduced Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers were sampled from moving bed operation. The absence of Fe₂O₃ and Fe₂TiO₅ in the XRD pattern indicated that Fe₂O₃ and Fe₂TiO₅ were completely reduced. Hence, the rhombohedral structure of α-Fe₂O₃ and the orthorhombic structure of Fe₂TiO₅ were transformed into the cubic structure of Fe₃O₄ and Al₂FeO₄, as illustrated in Fig. 12. Similar results were obtained in the investigation of Abad et al. (2011) and Zhu et al. (2016). For Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers sampled after regeneration, Fe₂O₃, Fe₂TiO₅, and Al₂O₃ were observed in the XRD pattern, demonstrating that Fe₃O₄, Al₂FeO₄, Al₂O₃ and slightly TiO₂ contained in the reduced Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers were completely oxidized to Fe₂O₃, Fe₂TiO₅, and Al₂O₃.

Fig. 12. X-ray diffraction patterns of fresh, reduced and regenerated Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers for charcoal combustion.

The mechanism for reduction of Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers with CO or H₂ are proposed as Eqs. (23)–(26) based on the experimental results of this study. Unfortunately, the oxygen carrier conversion contributed by each reduction reactions were not possible to be separated because these reduction reactions may occur simultaneously. Therefore, the overall reduction reaction of iron-based oxygen carrier for methane combustion is summarized and simplified by Integrated Rate of Reduction (IRoR) model, which has been widely used for evaluating the total reduction rate of iron ores (Abad et al., 2011).

For Fe₂O₃/Al₂O₃ reduction:

For Fe₂O₃/Al₂O₃/TiO₂ reduction:

The definition of input processing capacity (Q_in) and output processing capacity for charcoal combustion (Q_C,out) and unburned (Q_unC,out) are expressed, respectively, as follows:

where ṁ is the charcoal feeding rate (g s^–1); ΔH_Fuel denotes the higher heating value (MJ kg^–1) of charcoal; F_j,out,g and F_j,out,C are the molar flowrate of combustible species j in the effluent stream such as CO, CH₄, H₂, and C; the combustible species generated from charcoal gasification and combustion are denoted as g and C; ΔH_rxn,j denotes the heat of combustion of species j. The output processing capacity for charcoal gasification (Q_g,out) was calculated by the heat balance, as described by Eq. (30):

As revealed in Fig. 13, Q_g,out for the combustion of charcoal with Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers were determined to be roughly 49 W and 30 W, respectively, further demonstrating that roughly 9% and 5% of Q_in released for full charcoal decomposition in the ADMBR. A maximum value of output processing capacity for charcoal combustion with Fe₂O₃/Al₂O₃ oxygen carriers was estimated to be 285 W as CO₂ concentration reached nearly 100%. Afterward, Q_C,out was gradually decreased with the increase of steam feeding rate. Moreover, Q_C,out of charcoal combustion by Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers increased as steam feeding rate increased from 157.7 to 221.4 mmol min^–1, and reached around 217 W as CO₂ concentration reached 99%, as shown in Fig. 13. However, Q_C,out decreased obviously with increasing inlet steam flow rate and reduced to around 49 W, indicating that the excessive amount of steam might be negative for burning the charcoal with iron-based oxygen carriers. The heat release of charcoal combustion using the Fe₂O₃/Al₂O₃ particles was more than that of using the Fe₂O₃/Al₂O₃/TiO₂ particles because the high feeding rate of Fe₂O₃/Al₂O₃ was provided. It should be pointed out that although the oxygen carrier flow rate of Fe₂O₃/Al₂O₃/TiO₂ only operated at 7.55 g min^–1, practically 255 W (about 46% of Q_in) was released for the combustion of charcoal by using Fe₂O₃/Al₂O₃/TiO₂, probably because combustible gases were more ready to react with the Fe₂O₃ contained in the Fe₂O₃/Al₂O₃/TiO₂ oxygen carriers.

Fig. 13. Effect of steam flow rate on processing capacity for charcoal combustion in the moving bed.

CONCLUSIONS

Fe₂O₃ supported by Al₂O₃ and Al₂O₃/TiO₂ was investigated as an oxygen carrier for the CLC of charcoal. The kinetics of reducing Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂ with hydrogen were well described by a shrinking core model. The results indicated that the Fe₂O₃/Al₂O₃/TiO₂ was reduced more quickly than the Fe₂O₃/Al₂O₃ by carbon monoxide, which diffused more easily through the product layer of the former. When the inlet steam flow rate was set to 221.4 mmol min^–1, approximately 61.8% and 47.2% of the charcoal was combusted in the moving bed reactor with Fe₂O₃/Al₂O₃ and Fe₂O₃/Al₂O₃/TiO₂, respectively; additionally, these oxygen carriers achieved conversion rates of ~29.2% and ~72.5%, respectively. Based on the XRD characterization, a phase transformation from Fe₂O₃ (rhombohedral) and/or Fe₂TiO₅ (orthorhombic) to Fe₃O₄ (cubic) and Al₂FeO₄ (cubic) occurred during the CLC of charcoal. Because a high flow rate was used for Fe₂O₃/Al₂O₃, more heat was released when the charcoal was combusted with this oxygen carrier than with Fe₂O₃/Al₂O₃/TiO₂. Furthermore, the Fe₂O₃/Al₂O₃TiO₂ particles reacted strongly to syngas (CO–H₂) due to their abundant Fe₂O₃ content; hence, less Fe₂O₃/Al₂O₃/TiO₂ was needed for combusting the charcoal.

ACKNOWLEDGMENTS

The authors would like to thank the Ministry of Science and Technology of the Republic of China (Taiwan) for financially supporting this research under Contract No. MOST 106-3113-E-007-002-. The contribution on the supply of hematite powders by China Steel Corp. is gratefully acknowledged.